Synthesis, crystal structure, and in vitro biological evaluation of C-6 pyrimidine derivatives: New lead structures for monitoring gene expression in vivo

Article abstract of DOI:10.1080/15257770.2011.581258

Novel C-6 substituted pyrimidine derivatives are good substrates of herpes simplex virus type 1 thymidine kinase (HSV1-TK). Enzyme kinetic experiments showed that our lead compound, N-methyl DHBT (N-methyl-6-(1,3-dihydroxyisobutyl) thymine; N-Me DHBT), is

Full text of DOI:10.1080/15257770.2011.581258

This article was downloaded by: [Dipartmento di Studi E Reicerche]
On: 23 May 2013, At: 18:02
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
Synthesis, Crystal Structure, and in Vitro
Biological Evaluation of C-6 Pyrimidine
Derivatives: New Lead Structures for
Monitoring Gene Expression in Vivo
Miljen Martić a , Lucile Pernot b , Yvonne Westermaier b , Remo
Perozzo b , Tatjana Gazivoda Kraljević c , Svjetlana Krištafor c
,
Silvana Raić-Malić c , Leonardo Scapozza b & Simon Ametamey a
a Center for Pharmaceutical Science of ETH, PSI and USZ, ETH
Zürich, Zürich, Switzerland
b Pharmaceutical Biochemistry group, School of Pharmaceutical
Sciences, University of Geneva, University of Lausanne, Geneva,
Switzerland
c Department of Organic Chemistry, Faculty of Chemical Engineering
and Technology, University of Zagreb, Zagreb, Croatia
Published online: 27 May 2011.
To cite this article: Miljen Martić , Lucile Pernot , Yvonne Westermaier , Remo Perozzo , Tatjana
Gazivoda Kraljević , Svjetlana Krištafor , Silvana Raić-Malić , Leonardo Scapozza & Simon Ametamey
(2011): Synthesis, Crystal Structure, and in Vitro Biological Evaluation of C-6 Pyrimidine Derivatives:
New Lead Structures for Monitoring Gene Expression in Vivo, Nucleosides, Nucleotides and Nucleic
Acids, 30:4, 293-315
To link to this article: http://dx.doi.org/10.1080/15257770.2011.581258
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 and Nucleic Acids, 30:293–315, 2011
Copyright ꢀC Taylor and Francis Group, LLC
ISSN: 1525-7770 print / 1532-2335 online
DOI: 10.1080/15257770.2011.581258
SYNTHESIS, CRYSTAL STRUCTURE, AND IN VITRO BIOLOGICAL
EVALUATION OF C-6 PYRIMIDINE DERIVATIVES: NEW LEAD
STRUCTURES FOR MONITORING GENE EXPRESSION IN VIVO
Miljen Martic´,1 Lucile Pernot,2 Yvonne Westermaier,2 Remo Perozzo,2
Tatjana Gazivoda Kraljevic´,3 Svjetlana Krisˇtafor,3 Silvana Raic´-Malic´,3
Leonardo Scapozza,2 and Simon Ametamey1
1Center for Pharmaceutical Science of ETH, PSI and USZ, ETH Zu¨rich, Zu¨rich, Switzerland
2Pharmaceutical Biochemistry group, School of Pharmaceutical Sciences, University of
Geneva, University of Lausanne, Geneva, Switzerland
3Department of Organic Chemistry, Faculty of Chemical Engineering and Technology,
University of Zagreb, Zagreb, Croatia
Novel C-6 substituted pyrimidine derivatives are good substrates of herpes simplex virus type
2
1 thymidine kinase (HSV1-TK). Enzyme kinetic experiments showed that our lead compound, N-
methyl DHBT (N-methyl-6-(1,3-dihydroxyisobutyl) thymine; N-Me DHBT), is phosphorylated at a
similar rate compared to “gold standard” 9-[4-fluoro-3-(hydroxymethyl)butyl]guanine, FHBG, (Km
= 10 0.3 µM; kcat = 0.036 0.015 sec−1). Additionally, it does not show cytotoxic properties
on B16F1 cells up to a concentration of 10 mM. The x-ray analysis of the crystal structures of HSV1-
TK with N-Me DHBT and of HSV1-TK with the fluorinated derivative N-Me FHBT confirmed the
binding mode predicted by docking studies and their substrate characteristics. Moreover, the crystal
structure of HSV1-TK with N-Me DHBT revealed an additional water-mediated H-bond interesting
for the design of further analogues.
Keywords C-6 pyrimidine derivatives; N-Me DHBT; HSV1-TK; positron emission
tomography
INTRODUCTION
A noninvasive imaging modality, such as positron emission tomogra-
phy (PET), offering the possibility of monitoring the location, magnitude,
Received 21 February 2011; accepted 11 April 2011.
M. M. and L. P. contributed equally to this article.
We thank the x06sa and the x06da beamline scientists at the SLS for their help in collecting x-ray
data. L. P. was first a Biomedizin-Naturwissenschaft-Forschung fellow and subsequently a recipient of a
grant from the Novartis Consumer Health Foundation. This work was supported by the grant SNF No.
3100A0-113830/1.
Address correspondence to Leonardo Scapozza, Pharmaceutical Biochemistry Group, School
of Pharmaceutical Science, University of Geneva, University of Lausanne, Quai Ernest Ansermet 30
CH-1211 Geneva 4, Switzerland. E-mail: leonardo.scapozza@unige.ch
293

294
M. Martic et al.
and duration of gene expression over time is critically important for the
application of gene therapy to human beings. To be considered as suit-
able agents for monitoring herpes simplex virus type 1 thymidine kinase
(HSV1-TK) expression in vivo by PET, compounds have to fulfill several
requirements[1]: good catalytic efficacy for HSV1-TK, high specificity to-
ward the HSV1-TK enzyme, low specificity toward other human enzymes,
low cytotoxicity in normal and TK-transfected cells, a suitable structure for
radiolabeling and favorable pharmacokinetics. Non-fluorinated acyclic nu-
cleoside analogues are known to be potent anti-herpes virus agents.[2,3] The
most efficient ones, aciclovir (ACV) and ganciclovir (GCV) possess low host
toxicity and are active against herpes simplex virus type 1 and type 2.[3,4]
The selectivity of these acyclic nucleoside analogues is due, in part, to
the fact that they are only phosphorylated in virus-infected cells, where a
virus specific thymidine kinase of low substrate specificity converts the nu-
cleoside analogues to their monophosphate derivatives. Our investigations
were prompted by the need to develop PET imaging agents which lack the
disadvantages of already existing reporter probes of HSV1-TK, such as 9-
[4-[18F]fluoro-3-(hydroxymethyl)-butyl]guanine ([18F]FHBG), which shows
cytotoxicity and unfavorable pharmacokinetics (Figure 1). Previous re-
sults with a C-6 substituted pyrimidine derivative, 6-(2,3-dihydroxypropyl)-5-
[18F]fluoromethyl-(1H,3H)-pyrimidine-2,4-dione ([18F]fluoromethyl-HHT)
in a murine melanoma cell line B16 F1 tumor model, showed that
[18F]fluoromethyl-HHT is indeed far superior to [18F]FHBG, due to
FIGURE 1 Structure of FHBG and C-6 alkylated pyrimidine derivatives. The compound N-Me DHBT
bears a methyl substitution in the N-1 and the acyclic sugar extends from the carbon C-6.

C-6 Pyrimidine Derivatives
295
a greater sensitivity and contrast, as well as lower levels of abdomi-
nal background radioactivity.[5] However, it has the disadvantage of in
vivo defluorination. In a previous work,[6,7] we described the lead com-
pound DHBT (6-(1,3-dihydroxyisobutyl)thymine) and discussed its ad-
vantages over existing compounds. Those promising results prompted
us to further explore the chemical space of novel C-6 thymine deriva-
tives and the substrate acceptability of HSV1-TK. Unlike existing PET re-
porter probes that bear aliphatic side chains or sugar moieties in the
N-1 position of purine or pyrimidine, we intend to evaluate new fluorinated
compounds that are C-6 alkylated pyrimidine derivatives. As we aimed to
improve the in vivo stability of the new compound by first placing the 18F
radiolabel in the acyclic side chain extending from the C-6 position of the
thymine ring, we introduced a methyl group in the N-1 position in order to
circumvent the intramolecular cyclization of the compound (Figure 1).
Herein, we describe a docking study on a N-1 methylated deriva-
tive, a synthesis of the lead compound N-Me DHBT (N-methyl-6-(1,3-
dihydroxyisobutyl)thymine) and the x-ray structures of HSV1-TK in com-
plex respectively, with N-Me DHBT and with the fluorinated derivative N-Me
FHBT, respectively (Figure 1). Additionally, we provide the binding affinity
of N-Me DHBT to HSV1-TK, its kinetic turnover constant and an evaluation
of the effect on the proliferation of HSV-1 TK transduced and nontrans-
duced B16F1 cell lines.
EXPERIMENTAL
Docking Studies Using GOLD
The human thymidine kinase (PDB entry code 1W4R) and HSV1-TK
(PDB entry code 1VTK) x-ray structures were retrieved from the Protein
Data Bank in pdb format. For each structure, monomer A was kept and
hydrogens and Gasteiger charges added using Sybyl 7.2 (Tripos Inc., St.
Louis, MO, USA). Then, both structures were stored in mol2 format. For the
putative ligands of hTK1 and HSV1-TK, hydrogens and Gasteiger-Huckel
partial charges were assigned. Subsequently, the energy of each ligand was
minimized using the Powell method with an initial simplex optimization.
The termination criterion was set to a gradient of 0.05 kcal/mol × A and
the number of maximal iterations to 1000.
Standard GOLD parameters were applied, except for the definition of
the active site, where an origin was defined as the center of a sphere with
˚
a 6.5 A radius. For hTK1, its coordinates were 37.67, 93.58, and −19.48 for
x, y, and z, respectively. This corresponds to the centroid (center of mass)
of dTTP (thymidine triphosphate) in subunit A of 1W4R. For HSV1-TK,

296
M. Martic et al.
these centroid coordinates were set to 39.50, 16.73, and 31.94 (reflecting
the centroid of dTMP (thymidine monophosphate) in subunit A of 1VTK),
respectively. For each compound, 10 GOLD runs were done. The docking
protocol was validated by redocking several ligands (thymidine, penciclovir
(PCV), ganciclovir (GCV), and DHBT) into the corresponding x-ray crystal
structures.
Synthesis of the Lead Compound N-Me DHBT
The synthesis of N-Me DHBT is shown in Figure 2.
1
The H NMR and 13C NMR spectra were recorded on a 300 MHz Var-
ian Gemini 2000 (Varian Inc., USA) or a Bruker 400 MHz spectrometer
(Bruker, Germany) with tetramethylsilane as an internal standard. All data
were recorded in DMSO-d6 at 298 K. Chemical shifts were referenced to
the residual solvent signal of DMSO at δ 2.50 ppm for 1H and δ 39.50 ppm
for 13C. Individual resonances were assigned on the basis of their chemical
shifts, signal intensities, multiplicity of resonances, and H-H coupling con-
stants. Chemical shifts, δ, are reported in parts per million referred to TMS,
and signals are expressed as s (singlet), d (doublet), t (triplet), q (quartet),
m (multiplet), or br (broad). Coupling constants (J) are expressed in Hz.
Mass spectroscopy analysis was performed with a Micromass Quattro micro
TM API LC-ESI or a LCT Premier ESI-TOF from Waters (France). The TLC
was done on Merck silica gel 60 F254 precoated plates. Compound was visu-
alized under ultraviolet (UV) light (254 nm). The silica gel used for column
chromatography was silica gel 60, particle size 0.040–0.063 mm (230–400
mesh ASTM), Fluka, Germany. Evaporations were performed using a Bu¨chi
Rotavapor R-215 system at 40◦C (Bu¨chi, Switzerland).
2,4-Dichloro-5,6-dimethylpyrimidine (2)
A mixture of 2,4-dihydroxy-5,6-dimethylpyrimidine 1 (6.33 g, 45 mmol)
and POCl3 (30 ml, 328 mmol) was heated under reflux for 4 hours. Excess of
POCl3 was then removed under reduced pressure and the residue was added
to ice, washed with ether, and dried over sodium sulphate. The crude product
was recrystallized from ethanol to give 2 (7.54 g, 94.3%, m.p. 70–71◦C);
(C6H6N2Cl2), Rf = 0.65 (n-hexane/EtOAc = 5/1).
1H NMR (DMSO-d6): 2.26 (s, 3H, CH3), 2.47 (s, 3H, CH3).
13C NMR (DMSO-d6): δ 155.24 (C-2), 171.48 (C-4), 127.74 (C-5), 160.54
(C-6), 22.83 (CH3), 14.40 (CH3).

C-6 Pyrimidine Derivatives
297
FIGURE 2 Synthetic pathway to the novel lead compound N-Me DHBT. Reagents and conditions:
(i) POCl3, reflux; (ii) NaOCH3, MeOH; (iii) LDA, THF; (iv) N-chlorosuccinimide, Et3N, toluene;
(v) −55◦C, THF; (vi) C3H3ClO3, DMAP, CH3CN; (vii) Bu3SnH, AIBN, Toluene; (viii) CH3I; (ix) −78◦C,
BCl3, CH2Cl2; (x) AcCl, H2O; NH3, MeOH. Synthetic yields for each reaction ranged from 45 to 100%
and are reported. (Color figure available online.)

298
M. Martic et al.
2,4-Dimethoxy-5,6-dimethylpyrimidine (3)
2 (7.54 g, 42.6 mmol) was added to a solution of sodium (2 g, 87.0
mol) in methanol (50 ml). The reaction mixture was refluxed for 6 hours.
The solvent was evaporated and water was added to dissolve NaCl. The oily
layer was extracted with dichloromethane, dried over sodium sulphate and
concentrated under reduced pressure. The residue was kept in refrigerator
and gave colorless crystals of 3 (7.10 g, 99.1%, m.p. 39–40◦C).
1H NMR (DMSO-d6): 1.98 (s, 3H, CH3), 2.29 (s, 3H, CH3); 3.82 (s, 3H,
OCH3), 3.87 (s, 3H, CH3).
13C NMR (DMSO-d6): δ 162.10 (C-2), 168.67 (C-4), 107.25 (C-5), 165.67
(C-6), 53.96 (OCH3), 53.74 (OCH3), 21.46 (CH3), 9.67 (CH3).
1,3-Bis(benzyloxy)propan-2-one (6)
N-Chlorosuccinimide (18.0 g, 134.8 mmol) was suspended in dry toluene
(210 ml), and the mixture was cooled in a dry ice bath. Dimethyl sulfide (15
ml, 204.2 mmol) was added, and the mixture was cooled to −25◦C in a
dry ice-CH2Cl2 bath. 1,3-bis(benzyloxy)-2-propanol 5 (25.0 g, 91.8 mmol)
in toluene (21 ml) was added to the mixture, and the mixture was kept
under argon for 4 hours. Triethylamine (100 ml, 717.5 mol) was added
and the reaction was allowed to warm until reaching room temperature.
After 30 minutes of stirring at room temperature, the solution was passed
through a filter paper. The residue was washed with a diethyl ether (250
ml). The filtrate and the washings were combined, neutralized with 5%
aqueous HCl to pH 7, washed with a saturated NaCl solution (3 × 50 ml)
and then with water (3 × 50 ml), and finally dried over sodium sulphate.
The solvent was removed in vacuum and the resulting oil was purified by
column chromatography (n-hexane/EtOAc = 4/1) to give 13.5 g (54.4%
yield) of compound 6 (C17H18O3). m.p. 40◦C.
MS m/z: 270.1 (M+).
1H NMR (DMSO-d6): 7.23 (10H, Ph), 4.46 (s, 4H, CH2Ph), 4.13 (s, 4H,
CH2O).
6-(3-Benzyloxy)-2-benzyloxymethyl-2-hydroxypropyl)-2,4-
dimethoxy-5-methylpyrimidine (7)
LDA (1.5 M, 45 ml, 113 mmol) was added dropwise to a solution of 2,4
dimethoxy-5,6-dimethylpyrimidine 3 (7.50 g, 44.6 mmol) in THF (90 ml)
at −70◦C. The temperature was raised to −55◦C, and the solution was stirred
for 30 minutes. 6 (12.00 g, 44.4 mmol) in THF (65 ml) was added dropwise

C-6 Pyrimidine Derivatives
299
to this solution and stirring was continued for 2.5 hours while the tem-
perature was maintained at −55◦C. The solution was neutralized by the
addition of AcOH to pH 7 and then the temperature was raised to 25◦C
and the solvent removed. The residue was partitioned between AcOEt and
H2O. The organic layer was separated, dried over sodium sulphate and
removed by evaporation. The residue was isolated by chromatography on
silica gel column, eluting with n-hexane/EtOAc = 3/1. The desired frac-
tions were concentrated in vacuum to afford compound 7 in 59% yield
(1.25 g).
MS m/z: 439.3 (M+).
1H NMR (DMSO-d6): 7.28 (10H, Ph), 5.09 (s, 1H, OH), 4.48 (s, 4H,
CH2Ph), 3.87 (s, 2H, CH2O), 3.76 (s, 3H, OCH3), 3.44 (s, 3H, OCH3), 2.83
(s, 2H, CH2), 1.99 (s, 3H, CH3).
6-[3-Benzyloxy-2-(benzyloxymethyl)propyl]-2,4-dimethoxy-5-
methylpyrimidine (9)
Methyl oxalyl chloride (5.9 ml, 63.7 mmol) was added to a mix-
ture of 6-3-(Benzyloxy-2-benzyloxymethyl-2-hydroxypropyl-2,4-dimethoxy-5-
methylpyrimidine 7 (6.00 g, 13.7 mmol) and 4-(dimethylamino) pyridine
(DMAP, 3.7 g, 29.6 mmol) in dry CH3CN (50 ml) at 0◦C under argon. The
mixture was stirred for 14 hours under argon at room temperature and then
diluted with EtOAc (400 ml). The mixture was washed successively with a
saturated aqueous NaHCO3 solution (120 ml) and H2O (120 ml). The sep-
arated organic phase was dried over sodium sulphate and the solvent was
removed under reduced pressure. The residue was co-evaporated twice with
dried toluene (40 ml) to afford the methyl oxalyl ester 8. MS m/z: 481 (M+).
A mixture of Bu3SnH (9.0 ml, 36 mmol) and 2,2ꢁ-azobis (isobutyronitrile)
(AIBN; 180 mg, 1.1 mmol) in dry toluene (50 ml) was added to a solution of 8
in dry toluene (130 ml) under an argon atmosphere. The mixture was heated
at 110◦C for 3 hours and the solvent was removed by evaporation under
reduced pressure. The residue was purified by column chromatography
(mixture of n-hexane/EtOAc = 4/1 as mobile phase) to afford 2.77 g of
6-[3-benzyloxy-2-[(benzyloxymethyl)propyl]thymine 9 in 48% yield.
MS m/z: 423.3 (M+).
1H NMR (DMSO-d6): 7.29 (10H, Ph), 4.42 (s, 4H, CH2Ph), 3.87 (s, 3H,
OCH3), 3.79 (s, 3H, OCH3), 3.43 (d, 4H, OCH2), 2.68 (d, 2H, CH2), 1.97
(s, 3H, CH3).

300
M. Martic et al.
6-[3-(Benzyloxy-2-(benzyloxymethyl)propyl]-4-methoxy-1,5-
dimethylpyrimidin-2(1H)-one (10)
A mixture of 9 (3.6 g, 7.57 mol) and CH3I (50 ml, 803 mol) was refluxed
at 50◦C for 4 days and evaporated to dryness. The raw product was purified
on silica column (n-pentane/EtOAc = 1/1) to afford 3.5 g of pure 10 in
97% yield.
MS: 423.3 (M+)
1H NMR (CDCl3): 7.24 (m, 10H, Ph), 4.40 (q, 4H, CH2), 3.88 (s, 3H,
CH3), 3.45 (s, 3H, CH3), 3.42 (d, 4H, CH2), 2.78 (d, 2H, CH2), 2.10 (m, 1H,
CH), 2.95 (s, 3H, CH3).
6-[3-Hydroxy-2-(hydroxymethyl)propyl]-4-methoxy-1,5-
dimethylpyrimidin-2(1H)-one (11)
A mixture of 10 (3.2 g, 7.57 mmol) in dry CH2Cl2 (40 ml) was cooled
to −78◦C. BCl3 (1 M in CH2Cl2, 35 ml, 35 mol) was added via a syringe and
under argon. The mixture was stirred at −78◦C for 2 hours and then the
temperature was raised to −40◦C. A mixture of CH2Cl2/MeOH (1/1; 100 ml)
was added and the cooling bath was removed. The solution was neutralized
with a saturated NaHCO3 solution to pH 7 and stirred for additional 30
minutes. The water layer was separated and solvent was removed under
reduced pressure to give 1.65 g of pure 11 in 90% yield.
MS m/z: 243.3 (M+).
1H NMR (CDCl3): 1.96 (s, 3H, CH3), 2.01 (br, 1H, CH), 2.83 (m, 2H,
CH2), 3.47 (s, 3H, CH3), 3.58 (m, 4H, CH2 + CH2), 3.87 (s, 3H, OCH3).
6-[3-Hydroxy-2-(hydroxymethyl)propyl]-1,5-
dimethylpyrimidin-2,4(1H,3H)-dione (N-Me DHBT) (12)
A solution of 11 (3.0 g, 12.4 mmol) dissolved in acetyl chloride (10 ml)
and containing a few drops of water was stirred at reflux for 48 hours. The
solvent was removed under reduced pressure and the residue was dissolved
in saturated methanolic ammonia (50 ml) stirred at room temperature for
additional 4 hours, and then evaporated to dryness. The product was purified
on a silica column (CHCl3/i-PrOH = 4/1) to give pure N-Me DHBT (1.3 g)
in 46% yield.
MS m/z: 229.17 (M+).
1H NMR (CDCl3): 1.86 (s, 3H, CH3), 1.96 (m, 1H, CH), 2.73 (d, 2H,
CH2), 3.35 (s, 3H, CH3), 3.58 (m, 4H, CH2 + CH2).
13
ꢁ
ꢁ
C NMR (CDCl3): 10.96 (CH3), 28.08 (C1 ); 31.80 (N-CH3), 41.65 (C2 ),
ꢁ
ꢁ
61.15 (C3 ,C4 ), 109.44 (C5), 153.85 (C2, C4).

C-6 Pyrimidine Derivatives
301
HPLC Assay for Monitoring the Phosphorylation of N-Me DHBT
A previously published protocol was used.[8,9] Briefly, reactions were car-
ried out in a final volume of 75 µl containing H2O (58.2 µl), 1 M Tris
buffer, pH 7.4 (3.75 µl), 100 mM ATP (3.75 µl), 100 mM MgCl2 (3.75 µl),
1 mM N-Me DHBT (3.75 µl), and 2–10 µg of HSV1-TK. The reaction were
carried at 37◦C and were stopped by addition of 2.5 mM EDTA (675 µl)
at time-points 10, 20, 30, 45, 60, and 90 minutes. The obtained mixture
was then injected into an HPLC column for the analysis (column, C-18; sol-
vent, 0.2 M NaH2PO4, 25 mM tetrabutylammonium hydrogen sulphate, 3%
methanol; flow 1.1 ml/min; detection: diode array detector). The formation
of the monophosphorylated compound as well as ADP, was monitored quali-
tatively. Two different blank reactions (no enzyme or no substrate) were run
concurrently to account for the occurring minimal reaction, independent
of ATP hydrolysis. All reactions were performed in triplicates.
Determination of Km and kcat by Spectrophotometric Assays
The Km and kcat values were determined as described in the litera-
ture.[9,10] Reaction mixtures with a final volume of 75 µl containing 50 mM
Tris buffer pH 7.2, 1 mM 1,4-dithio-DL-threitol, 0.21 mM phosphoenolpyru-
vate, 2.5 mM MgCl2, 5 mM ATP, 0.18 mM NaDH, 0.8 µg pyruvate kinase,
0.5 µg L-lactate dehydrogenase, and different concentrations of the substrate
(0.01 mM, 0.05 mM, 0.1 mM, 0.25 mM, 0.5 mM, 1 mM) were incubated at
37◦C. Two minutes later, 5 µg of HSV1-TK were added in order to initiate
the reaction. The changes in absorbance at 340 nm corresponding to ADP
formation of the TK-dependent reaction at 37◦C were monitored during
20 minutes. The data presented are the results of the measurements per-
formed in triplicates. Control experiments were performed in order to take
into account the spontaneous hydrolysis of ATP under the experimental
conditions.
Cell Proliferation Assay
Confluent B16F1 cells stably transduced with HSV1-TK (B16F1 HSV1-
TK+) and wild-type cells (B16F1 wt) were trypsinisated and resuspended in
10 ml of fresh RPMI 1640 medium supplemented with 10% fetal calf serum
and 1% antibiotic PSF solution. The cell concentration for a 1:1 dilution
with trypan blue was determined using a Neubauer’s counting chamber.
The suspension was diluted with the medium to a final concentration of
20,000 cells per milliliter.
The protocol used was similar to one previously published.[9] Of the
suspension, 100 µl (2000 cells) were seeded in each well of a 96 well Nun-
clon surface plate from Nunc (ThermoFisher Scientific, Denmark). After
1.5 hours of incubation at 37◦C, the cells were attached to the bottom of the

302
M. Martic et al.
wells. 100 µl of sterile solutions of ganciclovir (GCV) and N-Me DHBT in
RPMI 1640 were added to the cells leading to final substrate concentrations
of 0.05, 0.1, 1, 5, 10, 50, 100, 500, and 1000 µM. All samples were prepared
in triplicates for each cell line. The wells without thymidine analogue allow
calculating the standard value of 100% of cell growth. The plates were incu-
bated for 3.5 days at 37◦C in the presence of 5% CO2. Viability of the cells was
measured by a colorimetric assay using the XTT labeling reagent (sodium 3ꢁ-
[1-(phenylaminocarbonyl)- 3,4-tetrazolium]-bis(4-methoxy-6-nitro)benzene
sulfonic acid hydrate). 50 µl of the XTT reagent were added to each
well.[11,12] After 9 hours of incubation at 37◦C in the presence of 5% CO2,
the number of viable cells in each well was determined by a colorimetric
UV absorption measurement of the converted XTT with a plate reader at a
wavelength of 450 nm. To account for the absorbance due to the presence of
both varying numbers of cells and nonmetabolized XTT, the optical density
(OD) at a reference wavelength of 750 nm was subtracted.[11,12]
Protein Expression, Purification, and Crystallization
The HSV1-TK was expressed as a GST-TK fusion protein and further
purified as HSV1-TK by using glutathione affinity chromatography following
protocols that have been reported earlier.[13,14]
Best crystals of HSV1-TK were obtained at 23◦C (296 K) with the vapor
diffusion technique using the geometry of the sitting-drop. Crystals grew in
0.9–1.2 M Li2SO4, 1 mM DTT and 0.1 M HEPES buffer at pH 7.5–8.0.[14] Crys-
tals of HSV1-TK were soaked for 30 minutes in the high salt crystallization so-
lution to which 5 mM of either N-methyl-6-(1,3-dihydroxyisobutyl)thymine,
N-Me DHBT, or N-Me FHBT were added, respectively.
Prior to data collection, the crystals were immerged for a short time in the
crystallization buffer containing 30% (v/v) glycerol and were subsequently
frozen in a gaseous nitrogen stream at 100 K. The crystals belong to the
˚
orthorhombic space group C2221 with unit cells parameters a = 113.0 A,
◦
˚
˚
b = 117.4 A, c = 107.8 A, α = β = γ = 90 for the HSV1-TK:N-Me DHBT
◦
˚
˚
˚
crystal and a = 114.2 A, b = 118.3 A, c = 108.9 A, α = β = γ = 90 for the
HSV1-TK:N-Me FHBT crystal.
Data Collection, Structure Solution, and Refinement
The x-ray data set of HSV1-TK:N-Me DHBT crystal was collected at the
beam-line x06sa, equipped with the pixel-detector PILATUS and tuned
˚
at a wavelength of 1 A, at the synchrotron Swiss Light Source (Villigen,
Switzerland). An angular increment between the diffraction images of 1◦
and a crystal-to-detector distance of 300 mm were chosen. The x-ray data set
of HSV1-TK:N-Me FHBT crystal was collected at the synchrotron Swiss Light

C-6 Pyrimidine Derivatives
303
Source on the beam-line x06da, equipped with the MAR CCD detector and
˚
tuned at a wavelength of 1 A. The crystal-to-detector distance was fixed at 255
mm and 0.5◦ was chosen as an angular increment between the diffraction
images. The datasets were obtained from a single crystal, respectively. Raw
diffraction images were indexed and integrated with MOSFLM7.0.1.[15] Data
scaling, merging, and reduction was carried out with programs of the CCP4
suite.[16] Relevant statistics are given in Table 2.
The structure of HSV1-TK containing the substrate analogue N-Me
DHBT was determined by molecular replacement with the program
PHASER[17] using the model of HSV1-TK complexed with the substrate ana-
logue 6-(1,3-dihydroxyisobutyl)thymine, DHBT (1E2P) as a search probe.
Before starting the molecular replacement procedure, the substrate ana-
logue DHBT and solvent molecules were removed from the initial model.
The rotation and the translation functions gave two clear single solutions,
which corresponded to two molecules in the asymmetric unit.
These two molecules located in the asymmetric unit yielded a Matthews
3
˚
coefficient (V M) of 2.41 A /Da and a corresponding solvent content of
49.1%.[18] Both solutions were employed in the refinement procedure per-
formed with CNS 1.1.[19]
From the total of independent reflections, 5% were used for calculating
the Rfree and thus monitor the progress of the refinement. After the first
cycle of refinement, composed of a rigid-body and an energy minimization
procedure, residues Gly61 to Ser74 for subunit A and residues Lys62 to Ile78
for subunit B of the HSV1-TK homodimer were reconstructed. Subsequently,
several cycles (simulated annealing, energy minimization, and grouped tem-
perature factor refinement) were performed and were interchanged with
manual rebuilding sessions using the molecular graphics program O.[20]
Within the binding pocket of both subunits A and B of the HSV1-TK
homodimer, a residual density was present in the (F obs − F calc) electron
density map contoured at 2σ and was clearly identified as the substrate
analogue N-Me DHBT. The sulfate ions were introduced also directly in the
ATP-binding pocket. Water molecules were positioned progressively if they
were observed in an environment susceptible to provide hydrogen bonds. In
the final stages of the refinement, individual restrained B-factor refinement
was performed.
Subsequently, the model of HSV1-TK complexed with the N-Me DHBT
was used as a search probe to solve by molecular replacement with the pro-
gram PHASER[17] the structure of HSV1-TK containing the fluorinated ana-
logue N-Me FHBT. Iterative model refinement carried out with PHENIX[21]
and manual building done with the program COOT[22] completed the struc-
ture of HSV1-TK:N-Me FHBT. Likewise, water molecules and sulfate ions
were introduced at the end of the refinement procedure. A residual density

304
M. Martic et al.
was present in the (F obs − F calc) electron density map contoured at 2.5σ and
located in the active site of both subunits A and B. The LigandFit subroutine
of PHENIX localized the N-Me FDHT derivative within this residual electron
density map.[23,24] The eLBOW subroutine was used to generate the geom-
etry restraint information of the N-Me FHBT derivative for performing the
final refinement step with the introduced derivative.[25]
The stereochemical quality of the two final models was assessed with
PROCHECK.[26] The coordinates and structure factors are deposited in the
Protein Data Bank under the accession codes 3F0T and sf3f0t for HSV1-TK
complexed with N-Me DHBT and under 3RDP and sf3rdp for HSV1-TK
complexed with N-Me FHBT.
Figures 4–7 were produced with the program PyMOL (http://www.
pymol.sourceforge.net; http://www.pymol.org).
RESULTS AND DISCUSSION
Docking Study Results
The choice of the docking program among the ones available (Autodock,
Dock, GOLD, FlexX) was made based on a study of Kellenberger et al.[27]
stating that both the capacity of reproducing the x-ray pose (accuracy) as
well as the discrimination of known thymidine kinase inhibitors were best
achieved with the docking program GOLD.
GOLD 2.2 (Genetic Optimisation for Ligand Docking) is especially
suited for complexes displaying predominantly H-bonds.[28] Apart from vi-
sual inspection, the main parameter for estimating docking success is the
fitness function, which has been optimized for the prediction of ligand bind-
ing poses rather than affinity. Docking studies were performed in order to
investigate whether N-Me DHBT would, indeed, be a substrate for HSV1-TK.
The docking model was tested by examining binding modes of thymidine,
PCV, GCV, and DHBT. It was successfully validated since all binding modes
corresponded to the binding modes seen in their respective crystal structures
with HSV1-TK. Root mean square deviation (RMSD) values were always close
˚
to 1 A.
The docking results predicted that N-Me DHBT would indeed be a HSV1-
TK substrate since N-Me DHBT shows the same interactions known to thymi-
dine and DHBT. Specifically, 3ꢁ-OH groups of the N-Me DHBT and N-Me
FHBT point toward the catalytic center of the enzyme, formed by Glu83 and
Arg163. Due to these interactions, it was predicted that N-Me DHBT and
N-Me FHBT would undergo phosphorylation, since their hydroxyl function-
alities mimic the 5ꢁ-hydroxyl group of thymidine as later confirmed by the
x-ray structures reported here. The second OH group of N-Me DHBT (4ꢁ-
OH) and the F atom of N-Me FHBT point towards the LID domain (residues

C-6 Pyrimidine Derivatives
305
219–226) of HSV1-TK. Based on these results, both compounds fulfill the
prerequisite for being HSV1-TK substrates.
Synthesis of N-Me DHBT
The synthesis of the key intermediate N-Me DHBT (Figure 2) was
accomplished in analogy to published procedures with slight modifica-
tions in the work up.[29,30] The starting compound, 2,4-dihydroxy-5,6-
dimethylpyrimidine (1), was converted to the corresponding chloride 2,
which was used without further purification. The treatment of 2 with sodium
methanolate gave the dimethoxy compound 3 in 95% yield after purification
by column chromatography. In parallel, 1,3-dibenzyloxy-2-propanol (5) was
oxidized using Corey oxydation to yield ketone 6. The product of the reac-
tion between the lithiated compound 4 and ketone 6 was the C-6 substituted
pyrimidine 7. The removal of the tertiary hydroxyl group was accomplised
using radical deoxygenation via oxalyl ester 8. Introduction of the methyl
group in the N-1 position of the pyrimidine ring was achieved by the reaction
of 9 with methyl iodide. The cleavage of two benzyl protecting groups, fol-
lowed by the removal of the methoxy group gave N-Me DHBT in 13% overall
yield. The insertion of the methyl group on the N-1 compared to DHBT al-
lows avoiding intramolecular cyclisation while performing fluorination on
one of the OH groups for producing the 18F labeled N-Me FHBT.
In Vitro Validation of N-Me DHBT as Substrate of HSV1-TK
In analogy to the previously published protocol,[6,9] we initially investi-
gated whether N-Me DHBT would be phosphorylated by HSV1-TK. Prior to
test the phosphorylation of N-Me DHBT, we always assessed HSV1-TK activ-
ity using thymidine and DHBT as positive controls. The formation of ADP
and the monophosphorylated substrate, as well as the disappearance of the
substrate over time, were clearly visible. The retention times of N-Me DHBT
and ADP differed only by 0.2 minutes, making a clear distinction between
the two components difficult. However, recording the UV spectra produced
by the DAD (Diode Array Detector) allowed to detect the UV-maxima at 267
nm and 258 nm that corresponded to ADP and N-Me DHBT, respectively.
The incubation of N-Me DHBT with HSV1-TK, also revealed the presence
TABLE 1 Enzyme kinetics data obtained by the spectrophotometric assay
Compound
N-Me DHBT
dT∗
PCV∗
DHBT∗
Km (µM)
10
0.036
0.2
0.35
1.5
0.05
35.3
0.035
kcat (s−1
)
∗Data from the literature.[6,10]

306
M. Martic et al.
FIGURE 3 HPLC-based phosphorylation pattern assay. The formation of new peaks (ADP and
monophosphorylated product) has been monitored by HPLC coupled to a diode array detector. To
assess the functionality of the HSV1-TK enzyme, thymidine (dT) and DHBT were used as positive con-
trols. These chromatograms clearly show the formation of ADP and the monophosphorylated products
dTMP and DHBT-P, respectively. For N-Me DHBT, the formation of N-Me DHBT monophosphate (N-
Me DHBT-P) is clearly visible, while the peaks of N-Me DHBT and ADP overlap. (Color figure available
online.)
of a new peak at 7.05 minutes, which corresponded to the monophospho-
rylated N-Me DHBT (Figure 3). An increase over time of this newly formed
peak provided another evidence that N-Me DHBT is indeed a substrate of
HSV1-TK.
Determination of the Binding and Catalytic Constant of N-Me
DHBT as Substrate of HSV1-TK
Quantative enzyme kinetics measurements to determine kcat and Km
values were performed by a spectrophotometric assay using a previously
published protocol.[9,10] The experiments, performed in triplicates, showed
that N-Me DHBT was phosphorylated at a similar rate as FHBG (Km = 10
0.3 µM; kcat = 0.036 0.015 sec−1) (see Table 1). Compared to DHBT (Km
= 35.3 1.3 µM; kcat = 0.035 0.01 sec−1), N-Me DHBT exhibits a rather
similar Km value, suggesting that the introduction of a methyl group at the
N-1 position of the pyrimidine ring of DHBT has no dramatic influence on
the binding affinity of N-Me DHBT.
Cell Proliferation Assay
Cell proliferation measurements were performed on B16F1 cells using
the XTT proliferation assay protocol.[9] No inhibition of cell proliferation
was observed for the HSV-1 TK transduced B16F1 cells (B16F1 HSV1-TK+)
and B16F1 wildtype cells up to 500 µM. Cell proliferation was partially in-
hibited at 1 mM. The later concentration is significantly higher than the one
where GCV shows high cytotoxicity.[9]

C-6 Pyrimidine Derivatives
307
TABLE 2 Data collection and refinement statistics of the complexes HSV1-TK:N-Me DHBT and
HSV1-TK:N-Me FHBTa
Data collection
HSV1-TK : N-Me DHBT
HSV1-TK : N-Me FHBT
Beamline
SLS – X06SA
31 – 2.0
487690
48655
10 (9.9)
100 (100)
17.6 (5.4)
11.1 (38.5)
SLS – X06DA
35.3–2.8
60375
˚
Resolution limits (A)
No. of measured reflections
No. of unique reflections
Redundancy
Completeness (%)
I/σ(I)
16788
3.6 (3.6)
91.3 (93)
5.9 (2.3)
18.6 (56.6)
Rsym (%)
Refinement and final model
No. of reflections
No. of omitted reflections
No. of proteins residues
No. of water molecules
No. of ions
48627
2438
630
303
6
2
15969
819
634
81
3
2
No. of bound ligand
R (%)/Rfree (%)
Mean on B-factors (A )
19.5/22.8
20/24.4
2
˚
Protein/solvent
Compound
30.6/41
34
36.3/30
40.2
RMSD. from ideal geometry
˚
Bond length (A)
0.006
1.3
0.002
0.55
Bond angles (◦)
a
˚
Figures in parentheses correspond to the value for the highest shell of resolution (2.1–2.0 A) for
˚
HSV1-TK:N-Me DHBT and (2.95–2.8 A) for HSV1-TK:N-Me FHBT.
Overall Structure of HSV1-TK:N-Me DHBT
With the aim to improve the design of a PET-tracer harbouring better
pharmacokinetics, the structure of HSV1-TK complexed with the substrate
˚
analogue N-Me DHBT was solved by molecular replacement at 2.0 A resolu-
tion in order to gain insight into the binding mode of this new C-6-alkylated
pyrimidine derivative within the nucleoside binding pocket. The structure
of HSV1-TK:N-Me DHBT is a homodimer composed of the two subunits A
and B. For several loops, due to their high mobility, the electron density
map was discontinuous and, therefore, they could not be modeled. The
final model of HSV1-TK:N-Me DHBT contains residues 45–147, 153–264,
280–373 in subunit A and residues 46–147, 153–220, 223–264, 274–374 in
subunit B, where the missing residues 221 and 222 belong to the loop cor-
responding to the LID domain. In addition, the side chains of Met A130,
Arg A293, Met B130, Met B179, Met B182, and Cys B362 were modeled with
a double conformation. The relevant statistics of both data collection and
model refinement are given in Table 2. The structure was refined to a Rfactor
of 19.5% and a Rfree of 22.8%. The geometry analysis of the final model

308
M. Martic et al.
revealed that most of the residues of the two subunits are located in the most
favored regions of the Ramachandran plot, that is, 94.2% for the subunit
A and 93% for the subunit B. Only one residue, Arg163, was found in the
nonallowed region, although its main-chain geometry is in good agreement
with the electron density map. This observation correlates with other HSV1-
TK structures reported. The unfavorable geometry of Arg163 is due to its
particular role: Arg163 participates in the phosphate transfer during the
enzymatic reaction.
The overall α/β-fold of HSV1-TK:N-Me DHBT, which is composed of
12 α-helices and 7 β-sheets, is conserved compared to the previously solved
HSV1-TK structures. The only noticeable difference appears in helix α1 at
the residues Lys62-Ser74, following just the P-loop (residues Gly56–Gly61).
Compared to most of the HSV1-TK structure that have been reported, either
this helix α1 does not superimpose very well or is not fully constructed. For
instance, the RMSD between the two subunits A and B of the HSV1-TK:N-Me
˚
DHBT is 0.96 A.
Binding of the Substrate N-Me DHBT in the Nucleoside
Binding Pocket
The position and orientation of the substrate N-Me DHBT in the thymi-
dine binding pocket was clearly interpretable in the (F obs − F calc) electron
density contoured at 2σ and both substrates for each subunit were intro-
duced into the active site with no ambiguity.
The interactions of N-Me DHBT with the binding site are depicted in
Figure 4. The structure reveals that the newly introduced N-methyl group
FIGURE 4 Stereo view of the interactions of N-Me DHBT with the active site of HSV1-TK. The (3F obs
–
˚
2F calc) electron density map at a resolution of 2.0 A obtained at the end of the refinement is contoured at
1σ and colored in purple. The final HSV1-TK:N-Me DHBT model is superposed. The water molecules are
depicted as red spheres and the hydrogen bonds are displayed as red dashed lines. The representation
was done with the subunit A of the homodimer. (Color figure available online.)

C-6 Pyrimidine Derivatives
309
FIGURE 5 Superposition of the active sites of subunit A (colored in green) and subunit B (colored in
red) of HSV1-TK:N-Me DHBT. The water molecules are displayed as blue spheres. The hydrogen bonds
displayed as red dashed lines are shown for subunit A only. (Color figure available online.)
stabilizes the aliphatic moiety of the molecule due to its increased bulkiness.
The interactions of the thymine moiety of N-Me DHBT with residues of the
binding site are conserved (Figure 4).
First, the pyrimidine ring of the thymidine moiety is fixed between
Met128 and Tyr172, forming a sandwich-like complex. The thymidine moi-
ety is further held in place via a hydrogen bond network, which involves
the carboxamide of Gln125, and the side chain of Arg176 and two water
molecules.
The dihydroxy-isobutyl side chain of N-Me DHBT is maintained within
the substrate active site by hydrogen bonds. The 3ꢁ-OH is connected via a
hydrogen bond with the side chain of the catalytically important residue
Arg163 (via the atom NE) and the sulfate ion via a water molecule. One
new feature appears: A water molecule mediates interactions between the
dihydroxy-isobutyl chain of N-Me DHBT (atom 4ꢁ-OH) and the side chains
of residues Glu225 and Tyr101. Glu225 belongs to the LID domain, a lysine-
and arginine-rich segment that is able to form a flap enclosing the active
site.
As shown in Figure 5, this water molecule is also present in the subunit
B connecting the 4ꢁ-OH of N-Me DHBT with the side chains of residues
Glu225 and Tyr101 via hydrogen bonds. All the interactions between the N-
Me DHBT substrate present in the active site of the subunit A are conserved
in the subunit B of the homodimer. The two thymidine rings of the substrate

310
M. Martic et al.
analogue are almost positioned at the same place. However, the dihydroxy-
isobutyl side chain appears to be more flexible in the binding site of HSV1-
TK.
Following the binding of N-Me DHBT to the active site, the LID domain
(residues Lys219–Arg226) is in a closed conformation in both subunits.
Nevertheless, in subunit B, the end of the loop (Gln221–Arg222) seems
flexible and could not be modeled because the electron density map was not
continuous.
Interestingly, in subunit A, residue Lys62 located by the helix α1
(Lys62–Ser74) is pointing outside of the ATP-binding pocket, whereas in
subunit B, it is pointing in direction of the sulfate ion located in the ATP-
binding pocket.
Differences in the Binding Mode Between HSV1-TK:N-Me DHBT
and HSV1-TK:DHBT
The superposition between the active site of HSV1-TK:N-Me DHBT and
HSV1-TK:DHBT (PDB entry code 1E2P) reveals that there are no relevant
differences at the level of the backbone and of the side chains (Figure 6A).
The only noticeable difference is located at the level of the helix α1 that
does not superpose very well for each subunit. In this case also, residue
Lys62 in subunit A of HSV1-TK:N-Me DHBT points out of the ATP-binding
site whereas in subunit A of HSV1-TK:DHBT, it points towards the sulfate
ion. Most of the interactions established in the active site by the substrate
N-Me DHBT are conserved in the binding of DHBT. The thymidine ring
is sandwiched between Met128 and Tyr172 and maintained in place by hy-
drogen bonds involving the side chains of Gln125 and Arg176. Likewise, the
3ꢁ-OH of the dihydroxy-isobutyl side chain of DHBT establishes hydrogen
bonds with the side chain of Arg163 (via the atom NH2). However, there is
no water molecule present between the 4ꢁ-OH of the dihydroxy-isobutyl side
chain and the side chains of residues Glu225 and Tyr101 (Figure 6A). The
dihydroxy-isobutyl side chain of DHBT appears to be also flexible.
Differences in the Binding Pattern Between HSV1-TK:N-Me
DHBT and HSV1-TK:dT or HSV1-TK:dTMP:ADP
Figure 6B shows the differences in the binding pattern between HSV1-
TK:N-Me DHBT and HSV1-TK:dTMP:ADP (PDB entry code 1VTK;[31] 2.75A
˚
resolution). There are no relevant differences at the level of the backbone
and the side chains of HSV1-TK:dTMP:ADP. The only difference is located
at the level of the helices α1 that do not superpose well. The thymidine
ring of dTMP is maintained in the nucleoside binding pocket by the same
interactions that are established by the thymidine ring of N-Me DHBT. How-
ever, the dihydroxy-isobutyl side chains of N-Me DHBT and the side chain
of dTMP do not occupy the same place. In the case of HSV1-TK:dTMP:ADP,

C-6 Pyrimidine Derivatives
311
FIGURE 6 A) Superposition of the active sites of HSV1-TK:N-Me DHBT (colored in green) and HSV1-
TK:DHBT (HSV1-TK is colored in dark pink and DHBT is colored in orange). B) Superposition of the
active sites of HSV1-TK:N-Me DHBT (colored in green) and HSV1-TK:dTMP:ADP (HSV1-TK is colored
in cyan and dTMP and ADP are dark blue). C) Superposition of the active sites of HSV1-TK:N-Me DHBT
(colored in green) and HSV1-TK:ganciclovir (HSV1-TK is colored in purple and ganciclovir in yellow).
The pictures are done with both subunits A of each homodimer. The hydrogen bonds, depicted as red
dashed lines, are shown only for subunit A of HSV-1 TK complexed with N-Me DHBT. (Color figure
available online.)
the 3ꢁ-OH of the ribose moiety of dTMP makes hydrogen bond interac-
tions with the side chains of residues Glu225 and Tyr101. The position
occupied by the 3ꢁ-OH of dTMP corresponds to the one occupied by the
water molecule in the structure of HSV1-TK:N-Me DHBT, establishing hy-
drogen bonds between the 4ꢁ-OH atom and the residues Glu225 and Tyr101
(Figure 6B).

312
M. Martic et al.
Differences in the Binding Mode Between HSV1-TK:N-Me DHBT
and HSV1-TK in Complex with a Purine Substrate Analogue
Figure 6C depicts the differences in the binding pattern between HSV1-
TK:N-Me DHBT and HSV1-TK:ganciclovir (PDB entry code 1KI2[32]).
The guanine moiety of ganciclovir and the thymidine moiety of N-Me
DHBT occupy a different location with the nucleoside binding pocket of
HSV1-TK. However, the binding mode observed for ganciclovir displays some
similarities with the one of HSV1-TK:N-Me DHBT. The guanine moiety of
ganciclovir is stacked between Met128 and Tyr172. The carboxamide of
Gln125 forms hydrogen bonds with the atoms N1, N2 and O6 of the guanine
moiety. The side chain of ganciclovir is hydrogen bonded to the side chains
of Tyr101 and Glu225 via the 4ꢁ-OH atom. The position occupied by the
4ꢁ-OH atom of ganciclovir matches the one occupied by the water molecule
in the structure of HSV1-TK:N-Me DHBT.
Binding of the Substrate N-Me FHBT in the Nucleoside
Binding Pocket
˚
The final model of HSV1-TK:N-Me FHBT solved at 2.8 A resolution con-
tains residues 46–264, 280–374 in subunit A and residues 46–264, 274–374
in subunit B. The relevant statistics of both data collection and model re-
finement are given in Table 2. The structure was refined to a Rfactor of 20%
and a Rfree of 24.4%. The geometry analysis of the final model revealed that
most of the residues of the two subunits are located in the most favored re-
gions of the Ramachandran plot, 90.8% for the subunit A and 89.1% for the
subunit B.
The N-Me FHBT derivative was clearly located in the thymidine binding
pocket of the two subunits A and B of the HSV1-TK homodimer. The binding
interactions of N-Me FHBT within the active site are depicted in Figure
7A. N-Me FHBT establishes mostly the same interactions as N-Me DHBT
(Figure 7B). Likewise, the pyrimidine ring is inserted in a sandwich-like
complex between Met128 and Tyr172. The thymidine moiety is kept in place
by a hydrogen bond network, established with the side chains of Gln125 and
of Arg176 and one water molecule. The acyclic side chain of N-Me FHBT is
kept in the active site through hydrogen bonds: The 3ꢁ-OH atom is connected
via hydrogen bonds with the side chain of the Arg163 (atom NE), the side
chain of Glu83 (atom OE1) and a water molecule. The fluorine atom points
in the direction of Glu225, Tyr101, and His58 and establishes no interactions.
The interactions established by N-Me FHBT in the active site of the subunit
A are also present in the subunit B of the homodimer. The additional water
molecule found in the structure of HSV1-TK:N-Me DHBT is not present in
the structure of HSV1-TK:N-Me FHBT. Upon the binding of N-Me FHBT

C-6 Pyrimidine Derivatives
313
FIGURE 7 View of the binding interactions of N-Me FHBT within the active site of HSV1-TK. A) The
˚
(2F obs – F calc) electron density map at a resolution of 2.8 A obtained at the end of the refinement is
contoured at 1σ and colored in blue. The final HSV1-TK:N-Me FHBT model is superposed. N-Me FHBT
is in sticks and colored in violet. The water molecules are depicted as red spheres and the hydrogen bonds
are displayed as red dashed lines. The representation was done with the subunit A of the homodimer.
B) Superposition of the active sites of HSV1-TK:N-Me DHBT (colored in green) and HSV1-TK:N-Me
FHBT (colored in violet). Only the hydrogen bonds (red dashed lines) established by N-Me DHBT are
displayed and the water molecules in the structure of HSV1-TK:N-Me DHBT are depicted as red spheres.
The representation was done with the subunits A of the homodimer. (Color figure available online.)
to the active site, the LID domain (residues Lys219–Arg226) is in a closed
conformation in both subunits.
The structure of HSV1-TK:N-Me FHBT revealed that the fluorine atom
is not lost and that the internal cyclization of the acyclic side chain did
not occur. It also confirmed that the fluorinated derivative N-Me FHBT is a
putative substrate of HSV1-TK.
CONCLUSION
Molecular modeling was successfully employed to select a novel C-6 sub-
stituted pyrimidine derivative for the development as a PET ligand for gene
expression monitoring. The introduction of a methyl group at the N-1 posi-
tion of the pyrimidine ring did not have any dramatic effect on the binding
properties of the methylated derivative. The novel lead compound, N-Me
DHBT was shown to be a substrate for HSV1-TK besides being non-cytotoxic.
The x-ray structure of HSV1-TK in complex with N-Me DHBT confirmed the
molecular modeling prediction and gives new insights for the design and syn-
thesis of further C-6 substituted pyrimidine derivative with improved phar-
macokinetics. Likewise, the structure of HSV1-TK:N-Me FHBT confirmed
that the fluorinated derivative is a possible substrate for HSV1-TK. While in
vitro validation including cell uptake studies, as well as in vivo PET studies

314
M. Martic et al.
are necessary to fully assess the biological properties of [18F]N-Me FHBT
for monitoring HSV1-TK expression in vivo, we expect that the radiosyn-
thesis of [18F]N-Me FHBT will be facilitated by the presence of the methyl
group on N-1, preventing intramolecular cyclization observed with N-1-free
C-6 pyrimidine derivatives such as DHBT and undesired side reactions.[6]
According to the observation made for other C-6 pyrimidine derivatives,[6]
the fluorinated derivative is expected to have similar biochemical properties
as N-Me DHBT.
REFERENCES
1. Serganova, I.; Ponomarev, V.; Blasberg, R. Human reporter genes: potential use in clinical studies.
Nucl. Med. Biol. 2007, 34, 791–807.
2. Cai, H.; Yin, D.; Zhang, L.; Yang, X.; Xu, X.; Liu, W.; Zheng, X.; Zhang, H.; Wang, J.; Xu, Y.;
Cheng, D.; Zheng, M.; Han, Y.; Wu, M.; Wang, Y. Preparation and biological evaluation of 2-amino-
6-[18F]fluoro-9-(4-hydroxy-3-hydroxy-methylbutyl) purine (6-[18F]FPCV) as a novel PET probe for
imaging HSV1-tk reporter gene expression. Nucl. Med. Biol. 2007, 34, 717–725.
3. Tjuvajev, J. G.; Doubrovin, M.; Akhurst, T.; Cai, S.; Balatoni, J.; Alauddin, M. M.; Finn, R.; Bornmann,
W.; Thaler, H.; Conti, P. S.; Blasberg, R. G. Comparison of radiolabeled nucleoside probes (FIAU,
FHBG, and FHPG) for PET imaging of HSV1-tk gene expression. J. Nucl. Med. 2002, 43, 1072–1083.
4. Iyer, M.; Barrio, J. R.; Namavari, M.; Bauer, E.; Satyamurthy, N.; Nguyen, K.; Toyokuni, T.; Phelps,
M. E.; Herschman, H. R.; Gambhir, S. S. 8-[18F]Fluoropenciclovir: an improved reporter probe for
imaging HSV1-tk reporter gene expression in vivo using PET. J. Nucl. Med. 2001, 42, 96–105.
5. Johayem, A. Development of novel nucleoside analogues for the PET imaging of herpes simplex virus type
1 thymidine kinase gene expression. Ph. D. Thesis, Diss. ETH No.:15450, Eidgeno¨ssische Technische
Hochschule Zu¨rich, Switzerland, 2004.
6. Johayem, A.; Raic-Malic, S.; Lazzati, K.; Schubiger, P. A.; Scapozza, L.; Ametamey, S. M. Synthesis
and characterization of a C6 nucleoside analogue for the in vivo imaging of the gene expression of
herpes simplex virus type-1 thymidine kinase (HSV1 TK). Chem Biodivers 2006, 3, 274–283.
7. Raic-Malic, S.; Johayem, A.; Ametamey, S. M.; Batinac, S.; De Clercq, E.; Folkers, G.; Scapozza,
L. Synthesis, 18F-radiolabelling and biological evaluations of C-6 alkylated pyrimidine nucleoside
analogues. Nucleosides Nucleotides Nucleic Acids 2004, 23, 1707–1721.
8. Pilger, B. D.; Perozzo, R.; Alber, F.; Wurth, C.; Folkers, G.; Scapozza, L. Substrate diversity of herpes
simplex virus thymidine kinase. Impact Of the kinematics of the enzyme. J. Biol. Chem. 1999, 274,
31967–31973.
9. Schelling, P.; Claus, M. T.; Johner, R.; Marquez, V. E.; Schulz, G. E.; Scapozza, L. Biochemical and
structural characterization of South methanocarba-thymidine that specifically inhibits growth of
herpes simplex virus type 1 thymidine kinase transduced osteosarcoma cells. J. Biol. Chem. 2004, 279,
32832–32838.
10. Schelling, P.; Folkers, G.; Scapozza, L. A spectrophotometric assay for quantitative determination
of kcat of herpes simplex virus type 1 thymidine kinase substrates. Anal. Biochem. 2001, 295, 82–
87.
11. Jost, L. M.; Kirkwood, J. M.; Whiteside, T. L. Improved short- and long-term XTT-based colorimet-
ric cellular cytotoxicity assay for melanoma and other tumor cells. J. Immunol. Methods 1992, 147,
153–165.
12. Roehm, N. W.; Rodgers, G. H.; Hatfield, S. M.; Glasebrook, A. L. An improved colorimetric assay
for cell proliferation and viability utilizing the tetrazolium salt XTT. J. Immunol. Methods 1991, 142,
257–265.
13. Fetzer, J.; Michael, M.; Bohner, T.; Hofbauer, R.; Folkers, G. A fast method for obtaining highly pure
recombinant herpes simplex virus type 1 thymidine kinase. Protein Expr. Purif. 1994, 5, 432–441.
14. Vogt, J.; Perozzo, R.; Pautsch, A.; Prota, A.; Schelling, P.; Pilger, B.; Folkers, G.; Scapozza, L.; Schulz,
G. E. Nucleoside binding site of herpes simplex type 1 thymidine kinase analyzed by x-ray crystallog-
raphy. Proteins 2000, 41, 545–553.

C-6 Pyrimidine Derivatives
315
15. Leslie, A. G. The integration of macromolecular diffraction data. Acta Crystallogr. D Biol. Crystallogr.
2006, 62, 48–57.
16. Evans, P. Scaling and assessment of data quality. Acta. Crystallogr. D Biol. Crystallogr. 2006, 62, 72–82.
17. McCoy, A. J.; Grosse-Kunstleve, R. W.; Storoni, L. C.; Read, R. J. Likelihood-enhanced fast translation
functions. Acta. Crystallogr. D Biol. Crystallogr. 2005, 61, 458–464.
18. Matthews, B. W. Solvent content of protein crystals. J. Mol. Biol. 1968, 33, 491–497.
19. Brunger, A. T.; Adams, P. D.; Clore, G. M.; DeLano, W. L.; Gros, P.; Grosse-Kunstleve, R. W.; Jiang,
J. S.; Kuszewski, J.; Nilges, M.; Pannu, N. S.; Read, R. J.; Rice, L. M.; Simonson, T.; Warren, G. L.
Crystallography & NMR system: A new software suite for macromolecular structure determination.
Acta Crystallogr. D Biol. Crystallogr. 1998, 54, 905;921.
20. Jones, T. A.; Zou, J. Y.; Cowan, S. W.; Kjeldgaard, M. Improved methods for building protein models
in electron density maps and the location of errors in these models. Acta Crystallogr. A. 1991, 47,
110–119.
21. Adams, P. D.; Grosse-Kunstleve, R. W.; Hung, L. W.; Ioerger, T. R.; McCoy, A. J.; Moriarty, N. W.; Read,
R. J.; Sacchettini, J. C.; Sauter, N. K.; Terwilliger, T. C. PHENIX: building new software for automated
crystallographic structure determination. Acta Crystallogr. D Biol. Crystallogr. 2002, 58, 1948–1954.
22. Emsley, P.; Cowtan, K. Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol.
Crystallogr. 2004, 60, 2126–2132.
23. Terwilliger, T. C.; Adams, P. D.; Moriarty, N. W.; Cohn, J. D. Ligand identification using electron-
density map correlations. Acta Crystallogr. D Biol. Crystallogr. 2007, 63, 101–107.
24. Terwilliger, T. C.; Klei, H.; Adams, P. D.; Moriarty, N. W.; Cohn, J. D. Automated ligand fitting by
core-fragment fitting and extension into density. Acta Crystallogr. D Biol. Crystallogr. 2006, 62, 915–
922.
25. Moriarty, N. W.; Grosse-Kunstleve, R. W.; Adams, P. D. electronic Ligand Builder and Optimization
Workbench (eLBOW): a tool for ligand coordinate and restraint generation. Acta Crystallogr. D Biol.
Crystallogr. 2009, 65, 1074–1080.
26. Laskowski, R. A.; Moss, D. S.; Thornton, J. M. Main-chain bond lengths and bond angles in protein
structures. J. Mol. Biol. 1993, 231, 1049–1067.
27. Kellenberger, E.; Rodrigo, J.; Muller, P.; Rognan, D. Comparative evaluation of eight docking tools
for docking and virtual screening accuracy. Proteins 2004, 57, 225–242.
28. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. Development and validation of a genetic
algorithm for flexible docking. J. Mol. Biol. 1997, 267, 727–748.
29. Hsu, L.-Y.; Wise, D. S.; Kucera, L. S.; Drach, J. C.; Townsend, L. B. Synthesis of Anti-Restricted
pyrimidine acyclic Nucleosides. J. Org. Chem. 1992, 57, 3354–3358.
30. Hsu, L.-Y.; Wise, D. S.; Shannon, W. M.; Drach, J. C.; Townsend, L. B. Synthesis of C-6 pyrimidine
acyclic nucleoside analogs as potential antiviral agents. Nucleosides Nucleotides 1994, 13, 563–584.
31. Wild, K.; Bohner, T.; Folkers, G.; Schulz, G. E. The structures of thymidine kinase from herpes simplex
virus type 1 in complex with substrates and a substrate analogue. Protein Sci. 1997, 6, 2097–2106.
32. Champness, J. N.; Bennett, M. S.; Wien, F.; Visse, R.; Summers, W. C.; Herdewijn, P.; de Clerq, E.;
Ostrowski, T.; Jarvest, R. L.; Sanderson, M. R. Exploring the active site of herpes simplex virus type-1
thymidine kinase by x-ray crystallography of complexes with aciclovir and other ligands. Proteins 1998,
32, 350–361.