Side-chain conformational restriction in template-competitive inhibitors of E. coli DNA polymerase I Klenow fragment: Synthesis, structural characterization and inhibition activity

Article abstract of DOI:10.1081/NCN-200034042

Nucleotide triphosphate α-(4-azidophenyl)-1,N⁶-etheno-dATP 3 and its monophosphate 3m were synthesized by condensation of 2-halo-2-(4-azidophenyl)acetaldehyes with dATP and dAMP, respectively. Structure analysis shows that the azidophenyl side

Full text of DOI:10.1081/NCN-200034042

This article was downloaded by: [Monash University Library]
On: 10 May 2013, At: 02:43
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Nucleosides, Nucleotides and Nucleic Acids
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/lncn20
Side‐Chain Conformational Restriction in
Template‐Competitive Inhibitors of E. coliDNA
Polymerase I Klenow Fragment: Synthesis, Structural
Characterization and Inhibition Activity
Michael B. Doughty a , Karam Aboudehen a , Garland Anderson a , Ke Li a , Bob Moore II a b
Tina Poolson a
&
a Department of Chemistry and Physics, Southeastern Louisiana University, SLU Box 10878,
Hammond, Louisiana, 70402, USA
b Department of Pharmaceutical Sciences, University of Tennessee Health Science Center,
Memphis, Tennessee, 38163, USA
Published online: 01 Jun 2007.
To cite this article: Michael B. Doughty , Karam Aboudehen , Garland Anderson , Ke Li , Bob Moore II & Tina Poolson (2004):
Side‐Chain Conformational Restriction in Template‐Competitive Inhibitors of E. coliDNA Polymerase I Klenow Fragment:
Synthesis, Structural Characterization and Inhibition Activity , Nucleosides, Nucleotides and Nucleic Acids, 23:11, 1751-1765
To link to this article: http://dx.doi.org/10.1081/NCN-200034042
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.

NUCLEOSIDES, NUCLEOTIDES & NUCLEIC ACIDS
Vol. 23, No. 11, pp. 1751–1765, 2004
Side-Chain Conformational Restriction in
Template-Competitive Inhibitors of E. coli DNA
Polymerase I Klenow Fragment: Synthesis,
Structural Characterization and Inhibition Activity#
1
1
1
Michael B. Doughty,1, Karam Aboudehen, Garland Anderson, Ke Li,
*
Bob Moore II,1,2 and Tina Poolson1
1Department of Chemistry and Physics, Southeastern Louisiana University,
Hammond, Louisiana, USA
2Department of Pharmaceutical Sciences, University of Tennessee Health Science
Center, Memphis, Tennessee, USA
ABSTRACT
Nucleotide triphosphate a-(4-azidophenyl)-1,N6-etheno-dATP 3 and its monophos-
phate 3m were synthesized by condensation of 2-halo-2-(4-azidophenyl)acetaldehyes
with dATP and dAMP, respectively. Structure analysis shows that the azidophenyl
side chain is attached to the a-position of the etheno ring (i.e., the carbon attached to
N1 of the purine), and conformation calculations show minima in the etheno-phenyl
bond rotation at 50 and 130° where the bulk of the phenyl ring projects out from the
plane of the etheno group. Like DNA Pol inhibitor 2-(4-azidophenacyl)thio-2’-
deoxyadenosine 5’-triphosphate 1, nucleotide 3 is a template-competitive DNA
polymerase inhibitor (TCPI), with a competitive Ki for Pol I KF of 3.41 mM, but has
only weak activity as an HIV RT inhibitor relative to the template-competitive reverse
#This work was funded in part by the Louisiana Board of Regents 044UG-02 and the National
Institutes of Health GM067686.
*
Correspondence: Michael B. Doughty, Department of Chemistry and Physics, Southeastern
Louisiana University, SLU Box 10878, Hammond, LA 70402, USA; Fax: (985) 549-5126;
E-mail: mdoughty@selu.edu.
1751
DOI: 10.1081/NCN-200034042
1525-7770 (Print); 1532-2335 (Online)
www.dekker.com
Copyright D 2004 by Marcel Dekker, Inc.

1752
Doughty et al.
transcriptase inhibitor 2-(4-azidophenacyl)thio-1,N6-etheno-2’-deoxyadenosine 5’-
triphosphate 2. Additionally, 3 photoinactivates KF in a time-dependent manner,
confirming the kinetic data that 3 binds to the free form of KF. The TCPI activity of 3
provides evidence for an extended side chain conformational preference in the
combined substrate polymerase inhibitors.
Key Words: DNA polymerase inhibitors; Reverse transcriptase inhibitors; Klenow
fragment.
INTRODUCTION
We previously reported inhibitors of DNA polymerase I Klenow fragment (KF)
and HIV-1 reverse transcriptase (RT) based on 2-substituted dATP derivatives. These
nucleotide inhibitors bind to the free enzyme in competition with template-primer.
These so-called template-competitive inhibitors of Pol I and reverse transcriptase differ
primarily based on the absence or presence of a 1,N6-etheno group. Thus the dATP
analog 1 is a KF inhibitor with a competitive Ki versus template/primer of 2.1 mM,[1]
and its exact etheno analog 2 is a template-competitive reverse transcriptase inhibitor
(TCRTI) versus HIV RT, with a competitive Ki versus template-primer of 10 mM.[2]
The initial template-competitive polymerase inhibitor (TCPI) 1 was designed as a
photoprobe based on a thermodynamic model for substrate binding to DNA
polymerase developed by Doronin and Kolocheva.[3,4] In our model, the dATP moiety
of 1 binds to the dNTP site and the 2-phenacylthio side chain binds to a lipophilic
pocket in the template binding site predicted by the thermodynamic model and
observed by photolabeling (Fig. 1A).[5] The phenacyl side chain of 1 increases binding
to the free form of KF by about 20 fold relative to shorter side chain or natural
dNTPs.[1] However, this model was unable to account for the activity of the etheno
derivative 2 in the TCRTI series, where the combination of the 2-etheno and 2-
phenacylthio side chain increased potency as TCRTIs by >300-fold over natural dNTPs
and 1 but reduced polymerase inhibition activity.[2,6] Conformational analysis led us to
propose that the main effect of the etheno group in 2 is to shift the 2-side chain
conformational equilibrium to a 180° C1 (N1-C2-S-CH2) torsional angle (Fig. 1B),
whereas C1 in TCPI 1 is 90°, allowing the side chain to project out from the adenine
base.[7]
As a test of this conformational hypothesis, we designed TCPI 3, where the side
chain is a phenyl ring appended to the a position of the etheno group, and additionally
contains an azido group as a photoprobe for independent analysis of binding
interactions. This conformationally restricted analog has the 1,N6-etheno group of
TCRTI 2 but the side chain, appended to the etheno group, cannot adopt the same
conformation proposed for the side chain of 2; in fact, the phenyl ring in 3 is rigidly
held even relative to the 2-side chain in 1. One negative aspect of this design is the
˚
decreased length of the side chain in 3, about 2 A, relative to the proposed side chain

DNA Polymerase I Klenow Fragment, Inhibitors of
1753
Figure 1. A (Top): Binding model proposed for the interaction of template-competitive DNA
polymerase inhibitor 1 with the Klenow Fragment; B (Bottom): Proposed active structures of TCPI
1 (top) and TCRTI 2 (bottom).
extension in 1. In the present communication we report the synthesis, structure, and
enzyme inhibitory activities of the side-chain restricted TCPI 3.
RESULTS AND DISCUSSION
TCPI 3 Synthesis and Photochemistry
In general, a-substituted-1,N6-etheno-2’-deoxyadenosine 5’-triphosphate analogues
are prepared by condensation of a-halogenated phenylacetaldehydes with dAMP or
dATP as the key reaction. The synthesis of a-halogenated substituted phenyl-
acetaldehydes is depicted in Scheme 1. The 4-azidophenethyl alcohol (5), prepared
from 4-amino-phenethyl alcohol by the general method of Lewis et. al.,[8] was
Scheme 1. Synthesis of side chain precursors.

1754
Doughty et al.
converted to the phenylacetaldehyde derivative by two methods, the Swern oxidation
with DMSO and oxalyl chloride[9] and oxidation using pyridinium dichromate
(PDC),[10] followed by silica gel chromatography purification. In the presence of the
para-azido group, the oxidation reaction needs to be conducted in very mild conditions
to avoid decomposition of the product. Both PDC and Swern oxidation selectively
oxidized 4 to 4-azidophenylacetaldehyde (5) without affecting the azido group.
However, higher yields were obtained with PDC because the base used in the Swern
oxidation, e.g., triethylamine, caused decomposition of the product, while PDC
oxidation requires neither strongly basic nor acidic conditions.
Phenylacetaldehydes are halogenated using N-bromo succinimide (NBS), bromine
or t-butyl hypochlorite (Scheme 1). Hanze et al. reported a one step oxidation/a-
halogenation of 3-hydroxysteroids using t-butyl hypochlorite.[11] However, when
phenethyl alcohol was treated with t-butyl hypochloric acid, no reaction was observed,
although reaction of benzyl alcohol with t-butyl-hypochlorite generated the
corresponding aldehyde, and chlorination of phenylacetaldehyde by t-butyl hypochlorite
in acetic acid generated a-chloro-phenylacetaldehyde. Alternatively, bromination of
phenylacetaldehyde with bromine or NBS both generated a-bromo-phenylacetaldehyde
in satisfactory yield.
The preparation of 2-halo-2-(4-azidophenyl)acetaldehydes was problematic owing
to the extremely poor stability of the product. Chlorination of 4-azido-phenyl-
acetaldehyde (5) by t-butyl hypochlorite suffered from low yield and poor
reproducibility due to significant decomposition under the reaction conditions.
1
Formation of the corresponding carboxylic acid (about 30% based on H-NMR) was
observed in the chlorination of 6, although the byproduct did not interfere with the
coupling reaction with dATP or dAMP. Additionally, bromination of 5 was
accomplished by reaction with NBS in THF only in the presence of a catalytic amount
of p-toluenesulfonic acid. The bromination usually took 0.5 to 1 hr at room temperature
in the presence of p-toluenesulfonic acid, while no reaction occurred in the absence of
the catalyst. Compounds 6 and 7 were extremely unstable, and extraction with sodium
bicarbonate solution or purification on a silica column caused decomposition. Therefore,
both 6 and 7 were prepared fresh and carried to the next step without purification.
Treatment of dAMP and dATP with an excess of 6 or 7 in aqueous acetonitrile
(50%) generated nucleotides 3m and 3, respectively, in about 20% yield (Scheme 2).
Scheme 2. Synthesis of a-substituted ethenoadensoine analogs.

DNA Polymerase I Klenow Fragment, Inhibitors of
1755
˚
Figure 2. Photodecomposition of 3 at 3000 A. Aliquots of the photolysis reaction were taken at
times (from top to bottom at 298 nm) 0, 0.25, 0.5, 1.0, 2.0, and 4.0 min, and spectra of the aliquots
were taken in 0.1 M HCl.
The low yield of this reaction was probably related to the extremely poor stability of 2-
halo-2-(4-azidophenyl)acetaldehyde. The yield was very low (<10%) when the reaction
was carried out in H2O, and no product was formed in aqueous 50% DMF. Al-
ternatively, 3 can be prepared from 3m by activation with phenyl phosphoryl chloride
and cleavage with pyrophosphate.[2]
The photo-activity of 3 was examined by monitoring UV absorbance changes upon
˚
irradiation at 3000 A. As shown in Fig. 2, compound 3m has an absorption maxima at
298 nm, corresponding to the maximum absorption of the electron-deficient
phenacylazides in both 1 and 2.[1,6] Overlay of the UV spectra taken at different
times of irradiation showed isobestic points at 250 and 330 nm, indicating a single
reaction pathway for the photodecomposition of the aryl azide group. Photodecompo-
sition followed first-order kinetics, with a half-life of 0.9 min, which correlates well
with the t1/2 of 0.6 min for the electron-deficient phenacylazide in 2.[6] Our results show
that the photodecomposition is significantly faster in quartz cuvettes than in pyrex and
clear plastic tubes (data not shown). Therefore, all photolysis experiments were
conducted in water-jacketed quartz cuvettes silanized in order to prevent nonspecific
absorption of protein to the glass surface.
Structure and Conformational Properties of TCPI 3
The reaction of adenine nucleotides with a-halo-acetaldehyde is suggested to occur
by initial alkylation at N-1 of the adenosine ring followed by ring closure and
dehydration based on alkylation studies in the 2-aminopyridine series.[12] The
regiochemistry of the substitution groups in the products were assigned based on this
mechanism. For example, when cAMP is reacted with a-bromoacetophenone, the
structure of the product (13% yield) is assigned to have the phenyl group at the a
position,[13,14] while when reacted with a-bromo-phenylacetaldehyde, the product is
assigned as a phenyl-etheno-cAMP.[14]
The regiochemistry of the phenyl group in the 1,N6-etheno-dATP derivatives
was examined by difference NOE spectra of a-phenyl-1,N6-etheno-dAMP synthesized
from condensation of dAMP with a-bromo-2-phenylacetaldehyde (data not shown).
Observed NOE enhancement of the ortho phenyl protons on selective irradiation of the

1756
Doughty et al.
C2 proton of the purine ring proves that phenyl ring is on the a-carbon (connected to
˚
N1), placing the ortho protons within at least 5 A of C2. Thus the phenyl ring side
chain of 3 projects into the minor groove region of the purine ring as suggested by the
structures shown.
The orientation of the phenyl ring relative to the 1,N6-etheno-adenine ring was
examined by calculation of the potential energy using a grid search method pushing the
psi angle only from 0 to 180° in 5° increments (a 360° rotation is not required because
of the planar symmetry of the two ring systems). As shown in Fig. 3A, two energy
minima at 50 and 130° were observed, and two energy maxima were observed, one at
0/180° where the planes of the two rings are parallel, and a second at 90° where the
planes of the two rings are perpendicular, with the parallel orientation being the highest
energy conformation. This conformational preference is a combined effect of the
destabilizing steric interaction between the ortho-phenyl and ring protons and
stabilizing p-electron overlap between the phenyl and etheno-adenosine rings, both in
Figure 3. Conformational analysis of the structure of 3. A: Rotation of Psi (the a-etheno-phenyl
bond) from 0 to 180°; B: Pseudorotation cycle for sugar conformation; C: Rotation of the glycosidic
bond Chi (C1’ –N9) from –180 to 180° at two conformations of gamma (C4’–C5’).

DNA Polymerase I Klenow Fragment, Inhibitors of
1757
the parallel orientation. Because the phenyl ring adopts a pseudo antiparallel
orientation, the bulk of the phenyl ring extends out from the plane of the etheno
ring. In addition, rotation of this single bond does not change the relative position of
the para-substituted azido group relative to the base.
The conformational state of the furanose ring of nucleotide 3 was examined by
computation of relative energy as a function of pseudorotation phase angle P.[15] Grid
searches were performed in which each endocyclic torsional angle of the furanose ring
was driven from À40° to 40° by an increment of 3° with torsional angles w and g
starting at anti and +sc, respectively. Pseudorotational phase angle P values for each
conformation were then calculated to give the energy profile for the entire
pseudorotational pathway shown in Fig. 3B. As with TCRTI 2 reported earlier,[7]
two energy minima at C3’-endo (P ꢀ 10) and C3’-exo/C2’-endo twist (P ꢀ 180) were
observed for 3, and the minimum at the twist conformation was broad and included C2’-
endo and C3’-exo. A western potential energy barrier (P = 270) of about 3 kcal/mole
and an eastern energy barrier (P = 90) of about 0.5 kcal/mol was also observed. These
results are consistent with reported furanose conformational preferences, particularly
with that found for TCPI 1,[5] indicating that the structural modifications on the adenine
ring in 3 do not exert a major effect on the conformational states of the furanose ring.
Nucleotide geometry is affected by glycosidic torsional angle chi (w), which changes
the orientation of the base relative to the sugar and phosphate moieties. Therefore, energy
profiles for 3 relative to w were examined using grid searches in which w was driven from
À180° to 180° by an increment of 5°. The conformation of the furanose ring was first set
at C3’-endo and the torsional angle g at either +sc or ap but not Àsc since the later is
rarely observed.[15] Shown in Fig. 3C are the energy profiles of 3 as a function of w with
g at ap or +sc. The +sc conformation is favored over the ap conformation and the energy-
minimum fell in the range corresponding to the high anti conformation. This minimum
energy conformation is also consistent with the preferred conformations for the DNA
polymerase inhibitor 1[5] and the reverse transcriptase inhibitor 2.[7]
Reversible and Irreversible Enzyme Inhibition
The Lineaweaver-Burk plot illustrated in Fig. 4 demonstrates that 3 is a com-
petitive inhibitor of the Klenow Fragment versus template/primer, and the average Ki
of 3.41 mM is about 10 times the activity of natural dNTPs such dATP as measured
using steady state kinetics with non-complementary template/primer. A summary of the
kinetic data is given in Table 1. Nucleotide 3 is nearly equipotent with 1 as a KF
inhibitor, and the steady state kinetics for 3 are much cleaner. Although 1 showed
mixed competitive inhibition versus template/primer,[1] the inhibition data with 3 for all
three experiments fit best to the competitive model even when challenged in a mixed
inhibition model. These results indicate that TCPI 3 binds to the free enzyme form of
KF in competition with template/primer, and does not bind to the binary enzyme-
template/primer complex as a dNTP substrate.
The inhibitory activities of 3 and dATP as a control versus HIV-1 RT are compiled
in Table 1; the data are presented as IC50’s because of the weak activity. Control dATP
had no inhibitory effect up to 300 mM, indicating an IC50 > 3 mM, whereas 3 had weak
activity with IC50 = 121 mM. The phenyletheno 3 follows the activity profile observed
for the polymerase inhibitors upon which it was designed, with a Pol selectivity ratio of

1758
Doughty et al.
Figure 4. Kinetic pattern for inhibition of Klenow fragment by 3 versus variable template/primer
poly(dA) Á (dT10) at inhibitor concentrations of 19 (circle with solid line), 13 (triangles with dot-
dash line), 6 (circles with dotted line), and 0 (diamonds with dash line) mM.
33 (Table 1). This TCPI selectivity indicates that the side chain conformation, with an
extended conformation being preferred in the polymerase inhibitors, and not the
absence or presence of the etheno ring, is determinant of enzyme selectivity and
activity in this series.
In design, an anchoring interaction of both the TCPIs and the TCRTIs is binding of
the triphosphate to the triphosphate coordination site in the dNTP binding site. Thus
photoinactivation of and photoincorporation of 1 into KF requires magnesium ion and
is inhibited by dATP as well as template/primer.[1,5] The monophosphate analogs of 1
and 2 as well as their derivatives are inactive as inhibitors, certainly since they lack the
ability to complex magnesium either in solution or at the triphosphate binding site,[2,5]
and in crystal structures of ternary complexes of both DNA polymerase b and HIV-1
reverse transcriptase, the triphosphates are complexed to two magnesium ions.[16,17]
Likewise, 3m, the monophosphate of TCPI 3, is inactive at its solubility limit as an
Table 1. Summary of inhibition data.
Inhibition constant (mM)
Inhibitor
HIV-1 RT
DNA Pol I Klenow
Pol I inhibition selectivity
1
>3000a
10c
2.11b
1430
0.109
35
NAe
86
2
92c
3
121 10.6d
NAf
3.41 0.12e
NAf
34.8 2.52e
3m
dATP
>3000f
aNo activity observed in IC50 assays up to 300 mM (3 mM magnesium), so activity must be greater
than 10 times that maximum.
bCompetitive Ki’s taken from Ref. [1].
cCompetitive Ki’s taken from Ref. [2].
dIC50 data (average of triplicate determinations).
eCompetitive Ki’s from steady-state kinetic studies (average of triplicate determinations).
fNo activity at highest concentration available for testing.

DNA Polymerase I Klenow Fragment, Inhibitors of
1759
Figure 5. Photoinactivation of KF by TCPI 3 and protection by substrate. In each experiment,
Klenow fragment, with and without 100 mM inhibitor 3, and in the presence of saturating
˚
poly(dA)–(dT)10, were photolyzed at 3000 A for the indicated times. Activity assays were
performed as describe in experimental.
inhibitor of both KF and RT (Table 1). Thus by this criterion 3 also anchors at the
triphosphate site as originally proposed in the binding model for 1 (Fig. 1A).
˚
Irradiation of KF at 3000 A in the presence of 3 results in a time-dependent loss of
polymerase activity with only 24% activity remaining at 4 min (Fig. 5). Saturating
template/primer poly(dA) Á (dT)10 protected 75% of KF from photoinactivation by 3.
This behavior is very similar to the activity of inhibitor 1 versus KF[5] and of 2 versus
RT.[6] This photoinactivation data is further supporting evidence that TCPI 3
specifically binds to the free form of KF at a binding site that simultaneously includes
part(s) of the dNTP and template/primer binding sites.
CONCLUSIONS
The original TCPI 1 was designed as a KF affinity probe that would bind with
good affinity to the dNTP site. Because dNTPs as DNA polymerase affinity probes
suffer from weak affinity and poor covalent labeling, the 2-phenacylthio side chain was
appended to the minor groove edge of the purine to stabilize binding by interaction at
the template site. The 20-fold increase in binding activity to the free enzyme, 100%
photolabeling with protection by both template/primer and dATP, and finally affinity
labeling in the template site, provided strong evidence for the model.[1,5] But the 40-
fold poorer activity of the etheno analog 2 as a TCPI forced us to reexamine the model.
TCPI 3 is truly a template-competitive inhibitor of KF in that it both competes
with template/primer in kinetic inhibition assays and efficiently photoinactivates the
free form of KF. The conformational properties of 3 include classic nucleotide
parameters with C3’-endo sugar conformation, a high-anti glycosidic bond, and a
preferred +sc phosphate conformation. Most importantly, the psi bond rotation minima
at 50 and 130° indicate that the bulk of the phenyl ring projects out from the plane of
the etheno ring. The activity of 3 is 10-fold greater than dNTPs as a TCPI in binding to
the free enzyme form of KF, indicating that the etheno itself is not detrimental to KF
inhibitory activity; indeed, the near equipotent activity of the restricted analog 3 and
the phenacyl analog 1 suggests a side chain conformation that is fully extended in the
TCPI series. The slightly weaker activity of 3 vs 1 as a TCPI is likely due to the

1760
Doughty et al.
decreased length of the phenyl ring side chain in 3, and we are currently synthesizing
analogs of 3 with increasing size at the 4 position of the phenyl ring to increase the
TCPI activity in this side chain restricted polymerase inhibitor series.
EXPERIMENTAL METHODS
Materials and Methods
Chemicals and solvents of reagent grade were purchased from commercial
suppliers and were used without purification except as follows: NBS was crystallized
from water; tetrahydrofuran (THF) was distilled from sodium/benzophenone prior to
use; and t-butyl hypochlorite was prepared by the method of Teeter.[18] HIV-1 RT was
purchased from Worthington (Lakewood, NJ), and the Klenow fragment of DNA
polymerase I was purchased from United States Biochemical Corporation (Cleveland,
Ohio). Polynucleotides, oligonucleotides and nonradioactive nucleotides were pur-
chased from Pharmacia (Piscataway, NJ). Tritiated thymidine ([3H]TTP) was purchased
from DuPont New England Nuclear (Boston, MA). Bovine serum albumin (BSA) was
purchased from Sigma (St. Louis, MO).
Enzyme concentration was measured using Pierce BCA protein assay reagent
(Pierce Chemical Co., Rockford, IL) relative to BSA as standard. Measurement of UV
absorbance on micro-titer plates was performed in a Bio-Tek Powerwave XS
Microplate Reader (Menlo Park, CA). Proton (1H) nuclear magnetic resonance spectra
were recorded on General Electric QE-300 and Varian VXR-500S 500 MHz
instruments at the University of Kansas NMR Laboratory. Chemical shifts are reported
as parts per million (ppm) relative to the CHCl3 peak ( = 7.24) in CDCl3 and HOD
peak (d = 4.80) in D2O. Mass spectra (MS) data for electron ionization and chemical
ionization were obtained on a Nermag R10-10 quadruple GC/MS system at the
University of Kansas Mass Spec Laboratory. Positive and negative fast atom
bombardment (FAB) MS were obtained on a VG Analytical ZAB spectrophotometer.
Infrared spectra (IR) were determined on a Perkin–Elmer 16PC FTIR. The ultraviolet
spectra were recorded on a Hewlett–Packard 8450A diode array spectrophotometer.
Photolysis was conducted in a Rayonet Photochemical Mini-Reactor chamber Model
RMR-600 purchased from the Southern New England Ultraviolet Company (Branford,
CT) in 0.7 mL water jacketed quartz cells. Liquid scintillation counting was conducted
on a Packard 1900 TR Liquid Scintillation Counter (Meridian, CT).
Caution
Sodium azide and aryl azides are explosive, and hydrazoic acid is explosive and
poisonous. Reactions using or generating these reagents should be performed in a well-
ventilated hood behind an explosion-proof shield. We have worked with these reagents
and procedures on mg to gram scales, and to date have observed no explosions.
Nevertheless, use extreme care when repeating or scaling up these procedures. All
manipulations of phenylazides were conducted in the dark or under a red light.
2-(4-Azidophenyl)acetaldehyde (5). A solution of 4-azidophenethyl alcohol (4,
0.5 g, 3.1 mmol) in dry methylene chloride (2 mL) was added to PDC (2 g, 1.5

DNA Polymerase I Klenow Fragment, Inhibitors of
1761
equivalent) suspended in dry methylene chloride (2 mL). After stirring at room tem-
perature for 5 hr under nitrogen, the reaction mixture was diluted with ethyl ether
(50 mL) The black solid was removed by suction filtration and washed with ethyl
ether (50 mL). The combined ether layer was extracted with sodium bicarbonate (10%,
50 mL). The organic layer was evaporated to remove solvent. The product was then
purified by silica gel chromatography, eluting with hexane/ethyl acetate (9:1). The
appropriate fractions were pooled and roto-evaporated, yielding 0.4 g (80%) of crude
1
product as a yellowish oil. H-NMR (CDCl3) d 3.67 (d, J = 2.1, 2H, C2H), 7.07 (d,
J = 6.6, 2H, PhH), 7.18 (d, J = 6.6, 2H, PhH), 9.72 (t, J = 2.1, 1H, C1H).
2-Chloro-2-(4-azidophenyl)acetaldehyde (6): Freshly prepared 4-azidophenylac-
etaldehyde (4, 0.2 g, 1.24 mmol) was dissolved in acetic acid (2 mL), and t-butyl
hypochlorite (170 mL, 1.1 equivalent) in acetic acid (1 mL) was added dropwise. After
stirring 1 hr at room temperature, the reaction mixture was diluted with water (5 mL)
and extracted with three volumes of ethyl ether. The combined ether extracts were
washed with water until the aqueous phase was neutral, dried, and filtered (do not
extract with base as the product decomposes). The final crude product contained 70%
1
aldeyhyde product and 30% acid by NMR: H-NMR (CDCl3) d 5.19 (d, J = 2.7 Hz,
1H, C2H), 6.95–7.21 (m, 4H, PhH), 9.49 (d, J = 2.7 Hz, 1H, C1H).
2-Bromo-2-(4-azidophenyl)acetaldehyde (7). Freshly-prepared 2-(4-azidophenyl)-
acetaldehyde (4, 0.1g, 0.62 mmol) and p-toluenesulfonic acid (about 0.1 equivalent)
were dissolved in dry THF (2 mL), and NBS (0.11 g, 1 equivalent) was added in small
portions. The reaction was stirred at room temperature until the brown color
disappeared (about 30 min). The reaction mixture was diluted with methylene chloride
(20 mL) and extracted with three volumes of water. Evaporation of the organic layer
1
yielded the crude product as a yellowish oil. H-NMR (CDCl3) d 5.19 (d, J = 2.7 Hz,
1H, C2H), 6.95–7.21 (m, 4H, PhH), 9.49 (d, J = 2.7 Hz, 1H, C1H).
a-(4-Azidophenyl)-1,N6-etheno-2’-deoxyadenosine 5’-monophosphate (3m). A
solution of dAMP (50 mg, 0.14 mmol) and sodium acetate (0.24 g, 1.76 mmol ) in
water (1 mL) was added to a 10-fold molar excess of freshly prepared a-halo-2-(4-
azidophenyl)acetaldehyde (6 or 7) in acetonitrile (1 mL), and the heterogeneous
solution was stirred at 40°C for 24 hr. The reaction was then diluted with water
(10 mL), neutralized with 0.1N NaOH and extracted with three volumes of ethyl ether.
The aqueous phase was lyophilized, and the product was purified by ion-exchange
chromatography on a DEAE-sephadex column with an average purification yield of
20%: IR (KBr) 2120 (N3); UV lMax (e  10À4): 0.05N NaOH, 275 (1.46), 298 (1.96);
0.05N HCl, 275 (1.72), 298 (1.93); 0.01M ammonium acetate, 276 (1.38), 296 (1.83);
1H-NMR d 2.58 (m, 1H, C2’H1), 2.88 (m, 1H, C2’H2), 3.97 (t, J = 3.9 Hz, 2H, C5’H),
4.21 (m, 1H, C4’H), 4.68 (br s, 1H, C3’H), 6.55 (t, J = 6.9 Hz, 1H, C1’H), 7.01 (d,
J = 7.8 Hz, 2H, PhH), 7.46 (d, J = 7.8 Hz, 2H, PhH), 7.50 (s, 1H, C10H), 8.52 (s, 1H,
C2H), 8.92 (s, 1H, C8H); high resolution FAB-MS m/e 471.0953; calcd. for
C18H16N8O6P1 (M–H+), 471.0930.
a-(4-Azidophenyl)-1,N6-etheno-2’-deoxyadenosine 5’-triphosphate (3) was syn-
thesized from dATP and a 10-fold molar excess of 6 (or 7) as described in the synthesis
of 3 with an average yield of 20%, or by the pyrophosphorylation of 4m as described

1762
Doughty et al.
in the synthesis of 2[2] with yield of about 50%. The product was purified by either
preparative HPLC eluting from a C4 column with 0–50% acetonitrile in 10 mM
triethylammonium bicarbonate, pH 7.5, or by FPLC ion-exchange chromatography
eluting from Mono-Q column with a 0.1–1.6 M ammonium acetate, pH 8.0 gradient.
The final product was >95% pure based on elution from analytical reverse phase HPLC
in the solvent given above: MS (-FAB, glycerol) m/e 631 (M–H+).
Photochemistry
Nucleotide monophosphate (3m, ꢁ 0.5 mM) was dissolved in distilled water
˚
(800 mL) and irradiated at 3000 A at room temperature in a quartz cuvette or in a clear
polypropylene tube. At indicated time points, aliquots of 50 mL were removed and
diluted in 0.1 N HCl (2.95 mL). The UV spectra of the diluted samples were measured;
the absorbance at 293 nm was plotted as a function of time and the half-life for
photodecomposition was calculated by fitting the data to a first order kinetic equation.
Structure and Conformation of TCPI 3
Nucleotide a-phenyl-1,N6-etheno-dAMP was synthesized from dAMP and a-
bromo-phenylacetaldeyde by the method used to synthesize the 4-azido analog 3m. The
regiochemistry of the phenyl ring was deduced using the NOE difference spectra
observed after selective irradiation of the C2 proton.
The solution structure of the polymerase inhibitor 3m was studied by molecular
mechanics calculations as described earlier for nucleotides 1[5] and 2.[7] Briefly, Sybyl
6.0 (Tripos Associates, Inc., St. Louis, MO) computations were run on IBM’s UNIX
variant AIX (version 4.25). TCPI 3 was constructed by modification of the structure of
dAMP taken from BIOPOLYMER. Semi-empirical calculations were performed using
the MOPAC package and MNDO method to optimize the charge and geometry of
nucleotide. Potential energy calculations of purine analogues were performed using the
Tripos force field, and those of nucleotide analogues were performed using the
Kollman all atom force field[19,20] with addition of the C4’-O4’-C1’-N* anomeric
torsional parameters for cyclic compounds to correct for the anomeric effect.[21]
A
dielectric constant of 1 was used in these calculations. Systematic sampling of torsional
space was accomplished using the GRID SEARCH option in Sybyl 6.0. The criterion
for termination of minimization was an energy change less than 0.0001 kcal/mol. The
conformations generated by grid search were analyzed using the MOLECULAR
SPREADSHEET function, which allows the measurement and calculation of
parameters such as torsional angels, interproton distances, and the pseudorotational
phase angle P.[15]
Polymerase Assays
For assay of the polymerase activity of the Klenow fragment, poly(dA) Á (dT)10
and thymidine 5’ triphosphate (TTP) were used as template/primer and nucleoside
triphosphate substrate, respectively, as previously described.[1] Briefly, reaction
mixtures containing 10 concentrations of template/primer (from 5–50 nM in 0.5 Km
increments), a fixed concentration of 3H-TTP (25 mM), and in the absence and
presence of various concentrations of inhibitor 3 (0–19 mM) or dATP (0–200 mM)

DNA Polymerase I Klenow Fragment, Inhibitors of
1763
were intiated by addition of 2.6 E-4 units of KF in a total volume of 25 mL and
incubated in 50 mM tris-HCl buffer, pH 7.4, containing 3 mM magnesium chloride and
1 mg/mL BSA at 37°C for 30 min. The reactions were terminated by addition of 25 mL
of 30 mM EDTA solution, pH 8, and quenched aliquots (10 mL) were spotted on
DEAE cellulose filter paper (squares of 2 Â 2 cm2 or 2.5 cm diameter circles). The
filter papers were washed three times in ammonium formate solution (0.3 M, pH 8,
>5 mL/square) with gentle stirring, dehydrated in ethanol and ethyl ether and air-dried.
The radioactivity on each filter paper was then measured in triplicate by scintillation
counting in 5 mL of Scintiverse II. The initial rate was determined as the picomoles of
[3H]TTP incorporated into DNA per min. The initial velocity versus substrate
concentration data were fit to a MM equation using non-linear regression analysis in
KINETICS to determine Km and Vmax. Inhibition data (40 datapoints per assay) was
fit by non-linear regression analysis to competitive, non-competitive, uncompetitive,
and mixed inhibition models, and the best fit was chosen based on statistical
parameters (r2, cod, and model selection criterion). The Ki’s reported are the average
of three determinations.
The polymerase activity of HIV-1 reverse transcriptase was assayed using
poly(rA) Á (dT)10 as template/primer and TTP as the nucleoside triphosphate substrate
as previously described.[2] For IC50 assays, RT (2.4 E-4 units) in 50 mM tris-HCl buffer
(pH 7.4) containing 3 MgCl2 and 1 mg/mL of BSA was incubated with template/primer
3
(20 nM) and H-TTP (25 mM) for 30 min, and radioactivity incorporated into DNA was
determined by the DEAE absorption method as described above for the DNA pol assay.
IC50’s were determined at 10 concentrations of inhibitor over 2.5 log units from 0 to
100% inactivation or up to solubility limits (for monophosphates) or up to 300 mM (for
triphosphates). Non-linear regression analysis using the method of Chou[22] was used to
determine the IC50’s, and the average and standard error of at least triplicate
determinations is reported.
For irreversible photo-inactivation studies, a standard irradiation mixture contained
2 mM KF, b-mercaptoethanol (BME), 3 mM magnesium chloride and the photoprobe
(100 mM) in 100 mM Tris-HCl buffer (pH7.4) in a total volume of 40 mL; for substrate
protection experiments, saturating (2000 nM) template-primer was added to the
photolysis mixture. The concentration of BME used in the photolysis was 20 times the
photoprobe or 2 mM in the absence of the inhibitor. The photolysis mixture was
˚
incubated at 37°C for 10 min in the dark and irradiated at 3000 A at 25°C. Aliquots of
5 mL were removed at indicated times and diluted 10-fold in the assay buffer. An
aliquot (5 mL) of the diluted sample was assayed for polymerase activity in a total
volume of 50 mL using 20 nM of template/primer and 25 mM [3H]-TTP. The %
photoinactivation at each time point was calculated as the percentage loss of enzyme
activity relative to the enzyme activity of the control in which the enzyme was
photolyzed in the absence of the photoprobe.
REFERENCES
1. Moore, B.M.; Li, K.; Doughty, M.B. Deoxyadenosine-based DNA polymerase
photoprobes: design, synthesis and characterization as inhibitors of DNA
polymerase I Klenow fragment. Biochemistry 1996, 35, 11634–11641.
2. Li, K.; Lin, W.; Chong, K.H.; Moore, B.M.; Doughty, M.B. Template-competitive

1764
Doughty et al.
inhibitors of HIV-1 reverse transcriptase: design, synthesis and inhibitory activity.
Bioorganic Med. Chem. 2002, 10, 507–515.
3. Doronin, S.V.; Lavrik, O.I.; Nevinsky, G.A.; Podust, V.N. The efficiency of dNTP
complex formation with human placenta DNA polymerase a as demonstrated by
affinity modification. FEBS Lett. 1987, 216, 221–224.
4. Kolocheva, T.I.; Nevinsky, G.A.; Levina, A.S.; Khomov, V.V.; Lavrik, O.I. The
mechanism of recognition of templates by DNA polymerases from pro-and eu-
karyotes as revealed by affinity modification data. J. Biomol. Struc. Dyn. 1991, 9,
169–186.
5. Moore, B.M.; Jalluri, R.; Doughty, M.B. DNA Polymerase photoprobe 2-(4-
azidophenacyl)thio-dATP identifies the template binding region of the DNA
polymerase I Klenow fragment. Biochemistry 1996, 35, 11642–11651.
6. Lin, W.; Li, K.; Doughty, M. Characterization of a binding site for template
competitive inhibitors of HIV-1 reverse transcriptase using photolabeling
derivatives. Bioorganic Med. Chem. 2002, 10, 4131–4141.
7. Lin, W.; Li, K.; Moore, B.M.; Doughty, M.B. Conformational properties of
template competitive HIV-1 reverse transcriptase inhibitors: analysis of enzyme
binding modes. Nucleosides Nucleotides Nucleic Acids 2003, 22, 283–297.
8. Lewis, R.V.; Roberts, M.F.; Dennis, E.A.; Allison, W.S. Photoactivated hetero-
bifunctional cross-linking reagents which demonstrate the aggregation state of
phospholipase A. Biochemistry 1979, 16, 5650–5654.
9. Mancuso, A.J.; Swern, D. Activated dimethyl sulfoxide: useful reagents for syn-
thesis. Synthesis 1981, 165–185.
10. Corey, E.J.; Schmidt, G. Useful procedures for the oxidation of alcohols involving
pyridinium dichromate in aprotic media. Tetrahedron Lett. 1979, 20, 399–402.
11. Hanze, A.R.; Fonken, G.S.; McIntosh, J.A.V.; Searcy, A.M.; Levin, R.H. Chemical
studies with 11-oxygenated steroids. V. A one-step oxidation-halogenation of 3-
hydroxysteroids. J. Am. Chem. Soc. 1954, 76, 3179–3181.
12. Schram, K.H.; Townsend, L.B. Fluorescent nucleoside derivatives of imidazo[1,2-
C]pyrrolo[2,3-d]pyrimidine: a new and novel heterocyclic ring system (1).
Tetrahedron Lett. 1974, 15, 1345–1348.
13. Meyer, R.B., Jr.; Shuman, D.A.; Robins, R.K.; Miller, J.P.; Simon, L.N. Synthesis
and enzymic studies of 5-aminoimidazole and N-1-and N-6-substituted adenine
ribonucleoside cyclic 3’,5’-phosphates prepared from adenosine cyclic 3’,5’-
phosphate. J. Med. Chem. 1973, 16, 1319–1323.
14. Jones, G.H.; Murthy, D.V.K.; Tegg, D.; Golling, R.; MOffatt, J.G. Analogs of
adenosine 3’,5’-cyclic phosphate, II. Synthesis and enzymatic activity of derivatives
of 1,N6-ethenoadenosine 3’,5’-cyclic phosphate. Biochem. Biophys. Res. Commun.
1973, 53, 1338–1343.
15. de Leeuw, H.P.M.; Haasnoot, C.A.G.; Altona, C. Empirical correlations between
conformational parameters in b-D-furanoside fragments derived from a statistical
survey of crystal structures of nucleic acid constituents. Full description of nucleo-
side molecular geometries in terms of four parameters. Isr. J. Chem. 1980, 20, 108–
126.
16. Pelletier, H.; Sawaya, M.R.; Kumar, A.; Wilson, S.H.; Kraut, W.J. Structures of
ternary complexes of rat DNA polymerase b, a DNA template-primer, and ddCTP.
Science 1994, 264, 1891–1903.

DNA Polymerase I Klenow Fragment, Inhibitors of
1765
17. Huang, H.; Chopra, R.; Verdine, G.L.; Harrison, S.C. Structure of a covalently
trapped catalytic complex of HIV-1 reverse transcriptase: implications for drug
resistance. Science 1998, 282, 1669–1675.
18. Teeter, H.M.; Bell, E.W. tert-Butyl hypochlorite. Org. Synt. 1952, 32, 20–22.
19. Weiner, S.J.; Kollman, P.A.; Case, D.A.; Singh, U.C.; Ghio, C.; Alagona, G.;
Profeta, S.; Weiner, P. A new force field for molecular mechanical simulation
of nucleic acids and proteins. J. Am. Chem. Soc. 1984, 106, 765–784.
20. Weiner, S.J.; Kollman, P.A.; Nguyen, D.T.; Case, D.A. An all atom force field for
simulations of proteins and nucleic acids. J. Comput. Chem. 1985, 7, 230–252.
21. Jalluri, R.K.; Yuh, Y.H.; Taylor, E.W. O-C-N anomeric effect in nucleosides: a
major factor underlying the experimentally observed eastern barrier to pseudo-
rotation. In The Anomeric Effect and Associated Steroelectronic Effects; Thatcher,
G.R.J., Ed.; American Chemical Society: Washington, 1993; 277–293.
22. Chou, T.-C. Relationships between inhibition constants and fractional inhibition in
enzyme-catalyzed reactions with different numbers of reactants, different reaction
mechanisms, and different types and mechanism of inhibition. Mol. Pharmacol.
1974, 10, 235–247.