Antitumor effects of guanosine-analog phosphonates identified by molecular modelling

Article abstract of DOI:10.1016/j.ijpharm.2010.06.036

Aiming to address new drug targets, molecular modelling is gaining increasing importance although the prediction capability of the in silico method is still under debate. For an improved treatment of actinic keratosis and squamous cell carcinoma, inhibitors of human DNA polymerase α (pol α) are developed by docking nucleoside phosphonate diphosphates into the active site of pol α. The most promising prodrugs OxBu and OxHex were then prepared by total synthesis and tested in the squamous cancer cell line SCC25. OxBu and OxHex proved cytotoxic and antiproliferative in the nanomolar concentration range and thus exceeded activity of aphidicolin, the relevant model compound, and 5-fluorouracil, the current standard for the therapy of actinic keratosis. Interestingly, the cytotoxicity in normal human keratinocytes with OxHex was clearly less pronounced and even not detectable with OxBu. Moreover, cytotoxicity of OxBu in particular with the colorectal carcinoma cell line HT29 even surmounted cytotoxicity in SCC25, and other tumor cell lines were influenced, too, by both agents. Taken together, OxBu and OxHex may offer a new approach to cancer therapy, given the agents are sufficiently well tolerated in vivo which is to be suspected beside their chemical structure.

Full text of DOI:10.1016/j.ijpharm.2010.06.036

International Journal of Pharmaceutics 397 (2010) 9–18
Contents lists available at ScienceDirect
International Journal of Pharmaceutics
journal homepage: www.elsevier.com/locate/ijpharm
Antitumor effects of guanosine-analog phosphonates identified
by molecular modelling
Anja Schwankea, Caterina Murruzzub, Barbara Zdrazilc, Ralf Zuhsed, Maja Nateka, Monika Höltjec,
Hans Christian Kortinge, Hans-Ulrich Reissigb, Hans-Dieter Höltjec, Monika Schäfer-Kortinga,∗
a Institut für Pharmazie (Pharmakologie und Toxikologie) der Freien Universität Berlin, D-14195, Berlin, Germany
b Institut für Chemie und Biochemie (Organische Chemie) der Freien Universität Berlin, Germany
c Institut für Pharmazeutische und Medizinische Chemie, Heinrich Heine-Universität Düsseldorf, Germany
d Chiracon GmbH, Luckenwalde, Germany
e Klinik und Poliklinik für Dermatologie und Allergologie der Ludwig-Maximilians-Universität München, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Aiming to address new drug targets, molecular modelling is gaining increasing importance although the
prediction capability of the in silico method is still under debate. For an improved treatment of actinic
keratosis and squamous cell carcinoma, inhibitors of human DNA polymerase ␣ (pol ␣) are developed by
docking nucleoside phosphonate diphosphates into the active site of pol ␣. The most promising prodrugs
OxBu and OxHex were then prepared by total synthesis and tested in the squamous cancer cell line SCC25.
OxBu and OxHex proved cytotoxic and antiproliferative in the nanomolar concentration range and thus
exceeded activity of aphidicolin, the relevant model compound, and 5-fluorouracil, the current standard
for the therapy of actinic keratosis. Interestingly, the cytotoxicity in normal human keratinocytes with
OxHex was clearly less pronounced and even not detectable with OxBu. Moreover, cytotoxicity of OxBu
in particular with the colorectal carcinoma cell line HT29 even surmounted cytotoxicity in SCC25, and
other tumor cell lines were influenced, too, by both agents. Taken together, OxBu and OxHex may offer
a new approach to cancer therapy, given the agents are sufficiently well tolerated in vivo which is to be
suspected beside their chemical structure.
Received 17 March 2010
Received in revised form 18 June 2010
Accepted 19 June 2010
Available online 25 June 2010
Keywords:
Human DNA polymerase alpha
Polymerase inhibitors
Keratinocytes
Tumor cell lines
Nucleotide analogs
Skin cancer
© 2010 Elsevier B.V. All rights reserved.
1. Introduction
these tumors are basal cell carcinomas (80%) which are com-
paratively benign. A frequent precursor of the more dangerous
About one million new cases of basal and squamous cell carci-
noma are annually diagnosed in the USA and 100 000 in Germany.
Almost 90% result from excessive UV light exposure (Stockfleth,
2008), recently an association with human papilloma virus infec-
tion has been described, too (Weinstock, 2006). The majority of
squamous cell carcinoma (SCC) in photodamaged skin is actinic ker-
atosis (AK) – a proliferation of cytologically atypical keratinocytes
in the lower epidermal layers. Actinic keratosis lesions can either
regress or evolve further to invasive squamous cell carcinoma, then
all epidermal layers contain atypical keratinocytes and the tumor
cells may surmount the basement membrane to become invasive
(Anwar et al., 2004). The prevalence of actinic keratosis is about 20%
for people over 40 years (Jorizzo, 2004b), and the risk for actinic
keratosis to generate an invasive squamous cell carcinoma is about
5–10% (Pariser et al., 2008; Stockfleth, 2008; Stockfleth and Kerl,
2006).
Current first line therapy consists of the topical application of
5% 5-fluorouracil (5-FU; Fig. 1) ointment which interferes with
DNA synthesis by blocking thymidylate synthetase and is able to
cure about 43% of actinic keratosis lesions (Jorizzo, 2004a; Levy
et al., 2001). Unwanted effects include severe wound infections
(Jorizzo, 2004b). Loaded to microsponges, 5-FU efficacy increases
which allows a reduction of the in use-concentration to 0.5% with
an improved tolerability (Jorizzo et al., 2002). Since tumor cells
accumulate greater amounts of photosensitisers than normal ker-
atinocytes, 70–90% of actinic keratosis lesions can also be destroyed
Abbreviations: 5-FU, 5-fluorouracil; Ac, acetyl; ATCC, American Type Culture Col-
lection; DBU, 1,8-diazabicyclo[5.4.0]undec-7-ene; DCM, dichloro methane; DMAP,
4-dimetylaminopyridine; DMEM, Dulbecco’s Modified Eagle’s Medium; DMF,
dimethylformamide; DMSO, dimethylsulfoxide; EDTA, ethylenediamine tetraacetic
acid; FCS, fetal calf serum; HaCaT, spontaneously transformed human keratinocyte
cell line; HPLC, high pressure liquid chromatography; HT29, colorectal carcinoma
cell line; KGM, keratinocyte growth medium; MCF-7, human breast adenocarcinoma
cell line; MEME, minimum essential medium Earle; MTT, methylthiazolyldiphenyl-
tetrazolium bromide; NaOAc, sodium acetate; NHK, normal human keratinocytes;
PBS, phosphate buffered saline, pH 7.4; RPMI, Roosevelt Park Memorial Insti-
tute; SCC, squamous cell carcinoma; SISO, adenocervical carcinoma cell line; T24,
bladder carcinoma cell line; TBAF, tetrabutylammonium fluoride; TBDPS-Cl, tert.-
butyldiphenylsilyl chloride; THF, tetrahydrofuran; TMS-Cl, trimethylsilyl chloride;
TP, triphosphate; TsCl, p-toluenesulfonyl chloride.
∗
Corresponding author. Tel.: +49 30 83853283; fax: +49 30 83854399.
E-mail address: msk@zedat.fu-berlin.de (M. Schäfer-Korting).
0378-5173/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.ijpharm.2010.06.036

10
A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
has been published recently (Siller et al., 2009). Thus the agent
extracted from euphorbia may become another option; neutrophils
appear to be the key component of antitumor activity (Challacombe
et al., 2006) which is clearly different from the agents studied
here. Yet, as current therapy of actinic keratosis is either painful
as with 5-FU and photodynamic therapy and/or cure rates are not
totally satisfying for potentially invasive carcinoma as with 5-FU,
diclofenac/hyaluronic acid and imiquimod, alternative approaches
like interference with enhanced DNA synthesis of human ker-
atinocytes and/or virus assembly are worth looking for.
Catalysis of the very first steps in DNA replication makes
eukaryotic DNA polymerase pol ␣ an interesting target in hyper-
proliferative diseases. While aphidicolin (Fig. 1) being the gold
standard of pol ␣ inhibitors in experimental pharmacology failed
to be introduced into therapy (Sessa et al., 1991), some nucleoside
human pol ␣ inhibitors are already used systemically for leukemia
(cytarabine, fludarabine) and pancreatic carcinoma (gemcitabine)
and others preferentially inhibiting the viral pol ␣ (Fig. 1) are used
for hepatitis B or AIDS, respectively. Inhibition of DNA synthesis
correlates with the inhibition of pol ␣ activity by the activated
forms gemcitabine and cytarabine triphosphate in the breast can-
cer cell line MCF-7 (Gandhi et al., 1997; Jiang et al., 2000). Aiming
for skin tumor therapy, non-nucleoside inhibitors of pol ␣ such as
dehydroaltenusin, glycolipids from green tea and spinach as well
as acyl-catechins have been studied preclinically (Kuriyama et al.,
2005; Maeda et al., 2007a,b; Matsubara et al., 2006, 2007). These
inhibitors were identified without knowing the structure of the
activity domain of the human enzyme.
We have modelled the structure of the active site of pol ␣ and
were able to fit a series of known pol ␣ inhibitors into the active
site approving the validity of the model (Richartz et al., 2008;
Zdrazil et al., in press). By molecular modelling of the respective
triphosphates we also identified thymidine and guanosine analog
prodrugs which showed a mild to moderate cytotoxicity and
inhibition of keratinocyte proliferation which goes along with the
mRNA expression of relevant enzymes like thymidine kinase 1 as
well as the mitochondrial enzyme desoxyguanosine kinase in nor-
Fig. 1. Structures of related polymerase inhibitors and of the standard drug for
actinic keratosis 5-fluorouracil.
by photodynamic therapy (Pariser et al., 2008; Stockfleth and
Kerl, 2006). Another option for topical therapy is 5% imiquimod
cream (Stockfleth et al., 2002), 50% of the patients achieved a 83%-
reduction of their actinic keratosis lesions (Gupta et al., 2005;
Lebwohl et al., 2004). Interestingly, this topical also cures geni-
tal warts, a disease unquestionably caused by Human Papillloma
Virus (HPV) (Perry and Lamb, 1999). In fact, a viral impact on
AK formation has been described recently (Vasiljevic et al., 2008;
Weissenborn et al., 2009). Since exposure of human keratinocytes
to UVB irradiation increases prostaglandin E2 formation due to an
over-expression of cyclooxygenase 2 (COX-2), diclofenac is another
therapy option (Buckman et al., 1998) eliminating 60–80% of the
lesions when applied for 3 months (Rivers et al., 2002). Moreover,
a phase-II study based on 0.05% ingenol gel short-term therapy
Table 1
Structures, molecular weight and calculated log P of polymerase ␣ inhibitors designed by molecular modelling
.
General backbone structure for the congeners. R1 and R2 differentiate and are shown in the table below.
Compound
M (g/mol)
Log P
R1
R2
BuP-OH
414.47
2.63
OxBu
436.41
464.46
1.36
2.27
OxIsohex
OxHex
464.46
492.54
520.37
2.34
2.75
3.73
OxBu-DE
OxHex-DE

A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
11
mal human keratinocytes (NHK), the spontaneously transformed
keratinocyte cell line HaCaT and the squamous cell carcinoma
cell line SCC25 (Höltje et al., 2010; Schäfer-Korting et al., 2010).
In order to improve activity and selectivity for tumor cells we
have next modelled nucleoside phosphonates, the diphosphate
activation products appearing to fit particularly well to the binding
site. Kinases are known to introduce the first phosphate group
less efficiently as compared to the second and third one (Eriksson
et al., 2002). The most promising agents (OxBu, OxHex, OxIso-
hex) have been synthesized and tested for potential antitumor
effects in vitro. Since, we aim in particular for topical therapy of
skin cancer, we also examined the effect of the more lipophilic
diethylester-pre-prodrugs OxBu-DE and OxHex-DE.
2. Material and methods
2.1. Molecular modelling
Scheme 2. Synthesis of compound 11a (OxBu). Reagents and conditions: (a): pyri-
dine, toluene, 95 ◦C; (b): AlCl3/NaCl, 135 ◦C; (c): H2, Pd-charcoal, MeSO3H, AcOH,
ethyl acetate, ethanol; (d): KOH, methanol, H2O, 100 ◦C; (e): TBDPSCl, pyridine,
DMAP, 80 ◦C. Side chains for OxHex and OxIsohex are synthesized correspondingly.
Molecular modelling methods, as well as the design and the
properties of the pol ␣ active site model have been described
in detail previously (Richartz et al., 2008; Zdrazil et al., in press)
and will not be discussed here. Triphosphates (TP) of BuP-OH and
diphosphates of the nucleoside phosphonates OxBu, OxIsohex and
OxHex (for molecular structures see Table 1) were docked manually
into the active site of the target enzyme and the resulting interac-
tion energy of the complexes was minimized. As a confirmation of
this procedure, molecular dynamics simulations were performed
to investigate the stability of the active site–inhibitor complexes.
For the synthesis of 5-amino-2-hexyl-phenol (10b) 128.6 g KOH
was dissolved in 65 ml water and 100 ml methanol at 0 ◦C.
A solution of 29.28 g N-(4-hexyl-3-hydroxy-phenyl)-acetamide
(117 mmol) in 190 ml methanol was added, and it was stirred at
100 ◦C. After 3.5 h the cooled solution was added to 330 ml of water
and 490 ml of ethyl acetate and the phases were separated. The
aqueous phase was extracted three times with 150 ml ethyl acetate.
The combined organic extracts were dried with Na2SO4. Evapo-
ration yields 44.6 g brownish solid. After chromatography (ethyl
acetate/hexane 1:1) 14.86 g (77 mmol, 66%) product was isolated.
1H NMR (400 MHz, CDCl3) ı: 6.88 (d, 1H; J = 8 Hz), 6.23 (dd, 1H,
J = 2/8 Hz), 6.13 (d, 1H; J = 2 Hz), 2.47 (t, 2H, 7–8 Hz), 1.47–1.60 (m,
2H), 1.2–1.4 (m, 6H), 0.87 (t, 3H, J = 7 Hz).
2.2. Synthesis
In the following we report in detail the synthesis of OxBu and
OxHex (Schemes 1–3), OxIsohex synthesis is performed under
identical conditions. All chemicals needed for synthesis were com-
mercially available and were used without further purification. The
solvents were dried using standard procedures. The products were
purified by flash chromatography (FLC) on silica gel (230–400 mesh,
Merck, Darmstadt, Germany). TLC was performed on pre-coated sil-
ica gel Plates 60 F254 and visualized with a UV lamp (254 nm) or
using a solution of KMnO4/Na2CO3 in water followed by heating.
Preparative HPLC was carried out on a Nucleosil 50-5 column and
detected with a Shimadzu variable UV-detector (ꢀ = 255 nm, DAD
SPD – M10A and UV LC 2010C, Shimadzu, Berlin, Germany) and a
Shimadzu light scattering detector (ELSD LT II, Shimadzu, Berlin,
Germany). Unless stated otherwise, yields refer to analytically
pure samples. 1H and 13C NMR spectra were recorded on Varian
400 ultra shield (400 MHz; Varian, Palo Alto, CA) and Bruker AC
250 (250 MHz, NMR) instruments (Bruker, Karlsruhe, Germany). IR
spectra were measured with a FTIR spectrometer Nicolet 5 SXC and
a Nicolet 5 SX205 (Perkin-Elmer, Rodgau-Jügesheim, Germany).
5-amino-2-butyl-phenol (10a) was synthesized under the same
conditions in 67% yield.
1H NMR: (400 MHz, CD3OD) ı: 6.78 (d, J = 7.8 Hz, 1H), 6.22 (d,
J = 2.1 Hz, 1H), 6.19 (dd, J = 7.8, 2.1 Hz, 1H), 2.46 (t, J = 7.5 Hz, 2H),
1.47–1.53 (m, 2H), 1.29–1.37 (m, 2H), 0.91 (t, J = 7.3 Hz, 3H).
For the synthesis of 3-(tert-butyldiphenylsiloxy)-4-hexyl phenyl-
amine (11b) 27 g (140 mmol) 5-amino-2-hexyl-phenol (10b) was
dissolved in 250 ml dry pyridine. After addition of 2.06 g DMAP
and 54 ml TBDPS-Cl (27.8 mmol) the mixture was stirred at 125 ◦C.
3.5 ml TBDPS-Cl were added after 48 h and 3 ml after 78 h. After
144 h, 5.5 ml methanol were added and stirred for another 1.5 h.
The cooled solution was diluted with 700 ml DCM and washed
twice with 350 ml water. The aqueous phase was extracted three
times with 350 ml DCM. The combined organic phases were
extracted four times with 500 ml HCl (1 M), dried over Na2SO4 and
evaporated. The resulting product was further cleaned by three
coevaporations with toluene.
The crude product obtained from 290.4 mmol starting mate-
rial was combined and purified by chromatography (hexane/ethyl
acetate 10:1). 70.59 g of a black oil (164 mmol, 56%) was isolated.
1H NMR (400 MHz, CDCl3) ı: 7.71–7.76 (m, 4H), 7.34–7.46 (m,
6H), 6.91 (d, 1H, J = 8 Hz), 6.17 (dd, 1H, J = 8/2 Hz), 5.78 (d, 1H,
J = 2 Hz), 3.15 (br s, 2H), 2.62–2.70 (m, 2H), 1.52–1.72 (m, 2H),
1.25–1.45 (m, 6H), 1.08 (s, 9H), 0.89 (t, 3H, J = 7 Hz).
11a was produced in the same manner in a yield of 85%.
1H NMR: (400 MHz, CDCl3) ı: 7.76-7.70 (m, 4H), 7.35–7.43 (m,
6H), 6.92 (d, J = 8.0 Hz, 1H), 6.16 (dd, J = 8.0, 2.2 Hz, 1H), 5.78 (d,
J = 2.2 Hz, 1H), 3.14 (br s, 1H), 2.67 (t, J = 7.8 Hz, 2H), 1.61–1.66 (m,
2H), 1.38–1.46 (m, 2H), 1-07 (s, 9H), 0.95 (t, J = 7.4 Hz, 3H).
Scheme 1. Synthesis of compound 5. Reagents and conditions: (a): diethyl ether,
ZnCl2, AcCl; (b): (EtO)3P, 100 ◦C; (c): Dowex, 78 ◦C; (d): TsCl, DMAP, DCM.

12
A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
Scheme 3. Synthesis of compound 17a (OxBu). Reagents and conditions: (a): CH3OCH2CH2OH, H2O, 125 ◦C; (b): POCl3, N,N-dimethylaniline, 110 ◦C; (c): DBU, DMF, 80 ◦C;
(d): TBAF, THF; (e): NH3, methanol, 85 ◦C; (f): TMS-Br, MeCN. OxHex 17b and OxIsohex are synthesized accordingly.
For the synthesis of 2-[3-(tert-butyldiphenylsiloxy)-4-hexyl-
phenylamino]-1,4,5,9-tetrahydro-purin-6-one (12b) 69.13 g
(m, 4H), 7.43–7.49 (m, 6H), 7.12 (d, J = 8.3 Hz, 1H), 6.50 and 6.22 (br
s, 1H), 2.69 (t, J = 7.6 Hz, 2H), 1.59–1.64 (m, 2H), 1.37–1.43 (m, 2H),
1.06 (s, 9H), 0.92 (t, J = 7.3 Hz, 3H).
(160 mmol) 3-(tert-butyldiphenylsiloxy)-4-hexyl phenylamine
(11b) and 15.42 g (71.2 mmol) bromhypoxanthine were suspended
in 445 ml 2-methoxy ethanol and 125 ml water. After 16 h at 125 ◦C
the mixture was cooled down and 400 ml water were added for
complete precipitation of the product. The product was isolated by
filtration und washed twice with 60 ml ammonia (25%) and with
500 ml ethyl acetate until the filtrate was colourless. The residue
was dried under high vac. 34.37 g product (61 mmol; 85%) were
isolated as white powder.
For the synthesis of [3-(tert-butyldiphenylsiloxy)-4-hexyl-
phenyl]-(6-chloro-9H-purin-2-yl)-amine (13b) 37.37 g (61 mmol)
2-[3-(tert-butyldiphenylsiloxy)-4-hexyl-phenylamino]-1,4,5,9-
tetrahydro-purin-6-one (12b) was suspended in 1.1 l dry
acetonitrile. 33.83 g (122 mmol) tetrabutylammonium chloride,
8 ml N,N-dimethylaniline (63 mmol) and 34 ml phosphoroxy-
chloride (365 mmol) were added. After 3 h another 10 ml
phosphoroxychloride (107 mmol) were added. The mixture
was cooled down after 5.5 h and was added within 15 min to
a solution of 413 g NaOAc in 1.6 l water at 0 ◦C and remains
were dissolved in a few ml ethyl acetate. To the mixture 1 l
of ethyl acetate was added in order to dissolve the separating
product. It was stirred 15 h at room temperature. The mix-
ture was filtered and the phases were separated. The organic
phase was washed four times with 500 ml NaHCO3 and twice
with 500 ml brine. After drying with Na2SO4 53.52 g of brown-
ish oil were isolated and purified by chromatography (ethyl
acetate/hexane 1:1). 23.83 g product (41 mmol, 67%) were
isolated.
1H NMR (400 MHz, DMSO-d6) (2 tautomers) 12.98 (br s), 12.41
(br s, both together 1H), 10.2 (br s), 10.03 (br s, both together
1H), 8.2 (s, 0.7H), 8.07 (s, 0.4H), 7.98 (s, 0.4H), 7.65–7.80 (m, 4H),
7.4–7.55 (m, 6H), 7.11 (d, 1H), 6.51 (br s, br s) 6.23 (br s both together
1H), 2.68 (t, 2H, J = 7–8 Hz), 1.55–1.70 (m, 2H), 1.2–1.4 (m, 6H), 1.06
(s, 9H), 0.86 (t, 3H, J = 6–7 Hz).
Analogous to 12b 2-[3-(tert-butyldiphenylsiloxy)-4-butyl-
phenylamino]-1,4,5,9-tetrahydro-purin-6-one
synthe-sized in a yield of 85%.
(12a)
was
1H NMR: (400 MHz, DMSO-d6) (2 tautomers) ı: 13.03 and 12.47
(br s, 1H), 10.22 and 10.06 (br s, 1H), 8.10–8.23 (m, 2H), 7.70–7.74

A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
13
1H NMR (400 MHz, CDCl3) ı: 7.3–7.4 (m, 4H), 7.3–7.5 (m, 6H),
7.15 (s, 1H), 7.1 (d, 1H, J = 8 Hz), 6.91 (dd, 1H, J = 8/2 Hz), 6.83 (d, 1H,
J = 2 Hz), 6.91 (s, 1H), 2.72–2.80 (m, 2H), 1.5–1.8 (m, 2H), 1.2–1.5
(m, 6H), 1.08 (s, 9H), 0.89 (t, 3H).
For the synthesis of {2-[6-amino-2-(4-hexyl-3-hydroxy-
phenylamino)-purine-9-yl]-ethoxymethyl}-phosphonic acid diethyl
ester (16b) 8.18 g (16.1 mmol) {2-[6-chloro-2-(4-hexyl-3-hydroxy-
phenylamino)-purin-9-yl]-ethoxymethyl}-phosphonic
acid
diethyl ester (15b) were dissolved in 250 ml 7 M methanolic
ammonia, sealed in an autoclave and heated to 85 ◦C for 15 h. After
3 h of cooling the solvent was distilled off. 8.45 g of a brownish solid
were isolated. After chromatography (ethyl acetate/methanol 10:1)
3.32 g (6.4 mmol; 40%) product were isolated and 2.34 g (4.3 mmol;
27%) of the starting material.
Following the same procedure 62% of [3-(tert-butyldiphenyl-
siloxy)-4-butyl-phenyl]-(6-chloro-9H-purin-2-yl)-amine
(13a)
were isolated.
1H NMR: (400 MHz, CDCl3) ı: 10.94 (br s, 1H), 7.73–7.76 (m,
4H), 7.30–7.41 (m, 7H), 7.12 (d, J = 8.1 Hz, 1H), 7.08 (s, 1H), 6.94
(dd, J = 8.1, 2.1 Hz, 1H), 6.81 (s, 1H), 6.78 (d, J = 2.1 Hz, 1H), 2.78 (t,
J = 7.8 Hz, 2H), 1.67–1.70 (m, 2H), 1.43–1.50 (m, 2H), 1.09 (s, 9H),
0.98 (t, J = 7.3 Hz, 3H).
1H NMR: (400 MHz, CDCl3) ı: 8.71 (br s, 1H), 7.66 (d, 1H, J = 2 Hz),
7.59 (s, 1H), 7.19 (br s, 1H), 7.97 (d, 1H, J = 8 Hz), 5.99 (dd, 1H,
J = 2/8 Hz), 5.63 (br s, 2H), 4.24 (t, 2H, J = 5–6 Hz), 4.10 (t, 2H, J = 5 Hz),
4.10 (dquint, 4H, J = 7/7 Hz), 6.93 (d, 2H, 7 Hz), 7.57 (t, 2H, 7–8 Hz),
1.50–1.65 (m, 2H), 1.25–1.45 (m, 6H), 1.20 (t, 6H, J = 7 Hz), 0.87 (t,
3H, J = 7 Hz).
Synthesis of {2-[6-amino-2-(4-butyl-3-hydroxy-phenylamino)-
purin-9-yl]-ethoxymethyl}-phosphonic acid diethyl ester (16a) was
performed following the same procedure: 45% product and 43%
starting material were isolated.
Synthesis
of
(2-{2-[3-(tert-butyldiphenylsiloxy)-4-hexyl-
phenylamino]-6-chloro-purin-9-yl}-ethoxymethyl)-phosphonic acid
diethyl ester (14b) was performed dissolving 21.83 g (37.2 mmol)
[3-(tert-butyldiphenylsiloxy)-4-hexyl-phenyl]-(6-chloro-9H-
purin-2-yl)-amine (13b) in 175 ml dry DMF. 5.8 ml (38.6 mmol)
DBU were added and the mixture was stirred 30 min at room
temperature. 17.04 g (46.5 mmol) toluene-4-sulfonic acid 2-
(diethoxy-phosphorylmethoxy)-ethylester (5) was added. After
17 h at 70 ◦C the solvent was evaporated. Chromatography (ethyl
acetate/methanol 17:1) yielded 18.96 g (24.4 mmol; 66%) of
product as mixture with the desilylated derivative.
1H NMR: (400 MHz, CDCl3) ı: 7.69 (d, J = 2.2 Hz, 1H), 7.60 (s,
1H), 7.12 (s, 1H), 6.99 (d, J = 8.0 Hz, 1H), 6.57 (d, J = 8.0, 2.2 Hz, 1H),
4.27 (t, J = 5.7 Hz, 2H), 4.03–4.13 (m, 6H), 3.83 (d, J = 6.9 Hz, 2H),
2.59 (t, J = 7.6 Hz, 2H), 1.55–1.62 (m, 2H), 1.35–1.42 (m, 2H), 1.23 (t,
J = 7.1 Hz, 6H), 0.93 (t, J = 7.3 Hz, 3H).
1H NMR (CDCl3): mixture of two parts product and one part
desilylated product.
For the final synthesis of {2-[6-amino-2-(4-hexyl-3-
hydroxy-phenylamino)-purin-9-yl]-ethoxymethyl}-phosphonic
acid (17b; OxHex) 729 mg {2-[6-amino-2-(4-hexyl-3-hydroxy-
ı: s 7.87 (1H), 7.70–7.77 (m, 4H), 7.30–7.45 (m, 4H), 7.13 (d,
1H, J = 8 Hz), 6.80 (s, 1H), 6.57 (d, 1H, J = 2 Hz), 4.34 (t, 2H, J = 6 Hz),
4.0–4.2 (m, 6H), 3.7 (d, 2H, J = 8 Hz), 3.6–3.7 (m, 2H), 2.74 (t, 3H,
J = 8 Hz), 1.5–1.7 (m, 2H), 1.2–1.5 (m, 6H), 1.08 (s, 9H), 0.88 (t, 3H,
7 Hz).
phenylamino)-purine-9-yl]-ethoxymethyl}-phosphonic
acid
diethyl ester (16b) were suspended in 60 ml dry acetonitrile. 2 ml
trimethylsilylbromide (14.9 mmol) were added and the mixture
was stirred 16 h at room temperature. Solvent was distilled off and
the residue was dissolved in 60 ml methanol after drying at high
vac. After 2 h the solvent was again removed and the residue was
dried at high vac. The solid was suspended in 43 ml acetone and
11 ml water, stirred 45 min at room temperature and filtered. After
drying 527 mg OxHex (1.31 mmol; 88%) were isolated.
Desilylated product: see next step
In the same manner 97% of (2-{2-[3-(tert-butyldiphenylsiloxy)-
4-butyl-phenylamino]-6-chloro-purin-9-yl}-ethoxymethyl)-
phosphonic acid diethyl ester (14a) was synthesized.
1H NMR: (400 MHz, CDCl3) ı: 7.88 (s, 1H), 7.74–7.78 (m, 4H),
7.32–7.43 (m, 8H), 7.14 (d, J = 8.3 Hz, 1H), 6.84 (s, 1H), 6.59 (d,
J = 2.1 Hz, 1H), 4.06–4.11 (m, 6H), 3.98 (t, J = 4.9 Hz, 2H), 3.71 (d,
J = 8.3 Hz, 2H), 2.77 (t, J = 7.6 Hz, 2H), 1.65–1.75 (m, 2H), 1.41–1.49
(m, 2H), 1.28 (t, J = 7.1 Hz, 6H), 1.10 (s, 9H), 0.97 (t, J = 7.3 Hz,
3H).
1H NMR: (400 MHz, DMSO-d6) ı: 8.66 (s, 1H), 7.87 (s, 1H), 7.42
(d, 1H, J = 1–2 Hz), 7.07 (dd, 1H, J = 1–2/8 Hz), 6.86 (br s), 6.83 (br s
together 3H), 4.26 (t, 2H, J = 5 Hz), 3.91 (t, 2H, J = 5 Hz), 3.63 (d, 2H,
J = 8–9 Hz), 2.44 (t, 2H, J = 7–8 Hz), 1.41–1.53 (m, 2H), 1.20–1.35 (m,
6H), 0.86 (t, 3H, J = 6 Hz).
For the synthesis of {2-[6-chloro-2-(4-hexyl-3-hydroxy-
phenylamino)-purin-9-yl]-ethoxymethyl}-phosphonic acid diethyl
ester (15b) 17.62 g (22.6 mmol) [3-(tert-butyldiphenylsiloxy)-
13C NMR: (100 MHz, DMSO-d6) ı: 155.8, 155.2, 154.5, 150.7,
139.7, 138.6, 128.8, 120.5, 113.4, 109.4, 105.7, d 70.0 (J = 10 Hz), d
66.3 (J = 160 Hz), 42.1, 31.0, 29.4, 28.9, 28.4, 21.9, 13.6.
4-hexyl-phenyl]-(6-chloro-9H-purin-2-yl)-amine
(13b)
was
dissolved in 500 ml THF. After addition of 14.93 g TBAF (46 mmol)
the solution was stirred 2.5 h at room temperature. The solution
was diluted with 1 l ethyl acetate and washed twice with 500 ml
brine. A formed precipitate was dissolved by addition of water. The
phases were separated and the aqueous phase was extracted three
times with 300 ml ethyl acetate. The combined organic phases
were dried with Na2SO4. 29.24 g of a greenish-brown oil were
isolated. Purification by chromatography led to the isolation of
12.34 g product (100%).
Analogous
68%
{2-[6-amino-2-(4-butyl-3-hydroxy-
phenylamino)-purine-9-yl]-ethoxymethyl}-phosphonic acid (17a;
OxBu) were formed.
1H NMR: (400 MHz, DMSO-d6) ı: 8.72 (s, 1H), 7.88 (s, 1H), 7.41
(s, 1H), 7.08 (d, J = 8.2 Hz, 1H), 6.94 (br s, 2H) 6.86 (d, J = 8.2 Hz,
1H), 4.27 (t, J = 4.5 Hz, 2H), 3.92 (t, J = 4.5 Hz, 2H), 3.64 (d, J = 8.5 Hz),
2.45 (t, J = 7.5 Hz, 2H), 1.46–1.51 (m, 2H), 1.27–1.34 (m, 2H), 0.89 (t,
J = 7.3 Hz, 3H).
13C NMR: (100 MHz, DMSO-d6) ı: 155.82, 155.35, 154.80,
150.92, 139.91, 138.98, 129.11, 120.87, 113.58, 109.70, 106.02,
70.31, 70.21, 67.32, 65.73, 31.96, 28.86, 22.03, 13.94.
1H NMR (400 MHz, CDCl3) ı: 8.57 (s, 1H), 7.79 (s, 1H), 7.71 (d, 1H,
J = 2 Hz), 7.15 (s, 1H), 7.02 (d, 1H, J = 8 Hz), 6.48 (dd, 1H, J = 8/2 Hz),
4.34 (t, 2H, J = 6 Hz), 4.16 (t, 2H, J = 6 Hz), 4.07 (dq, 4H, J = 8/7 Hz), 3.83
(d, 2H, J = 7 Hz), 2.58 (t, 2H, J = 7–8 Hz), 1.5–1.7 (m, 2H), 1.25–1.45
(m, 6H), 1.21 (t, 6H, J = 7 Hz), 0.87 (br t, 3H, J = 7 Hz).
2.3. Lipophilicity
Using the same procedure 89% of {2-[6-chloro-2-(4-butyl-3-
hydroxy-phenylamino)-purine-9-yl]-ethoxymethyl}-phosphonic
acid diethyl ester (15a) were isolated.
Log P values (1-octanol/water) were calculated using the
ALOGPS 2.1 program which was developed with 12908 molecules
from the PHYSPROP database using 75 E-state indices. 64 neural
networks were trained using 50% of molecules selected by chance
from the whole set. The log P prediction accuracy is root mean
squared error rms = 0.35 and standard mean error s = 0.26 (Tetko
and Tanchuk, 2002; Tetko et al., 2001).
1H NMR: (400 MHz, CDCl3) ı: 8.60 (br s, 1H), 7.83 (s, 1H),
7.71 (d, J = 2.2 Hz, 1H), 7.22 (s, 1H), 7.04 (d, J = 8.0 Hz, 1H), 6.53 (d,
J = 8.0, 2.2 Hz, 1H), 4.36 (t, J = 7.8 Hz, 2H), 4.04–4.21 (m, 6H), 3.84 (d,
J = 6.8 Hz, 2H), 2.61 (t, J = 7.7 Hz, 2H), 1.57–1.63 (m, 2H), 1.26–1.41
(m, 2H), 1.22 (t, J = 7.1 Hz, 6H), 0.94 (t, J = 7.3 Hz, 3H).

14
A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
2.4. Materials for the biological experiments
2.7. DNA synthesis
Cell culture media were purchased from Sigma–Aldrich
(Schnelldorf, Germany), or – for primary keratinocytes – from
Cambrex (Landen, Belgium). Fetal calf serum (FCS) was from
Seromed Biochrom (Berlin, Germany). Chemicals for cell culture
experiments were purchased from Sigma–Aldrich or VWR (Berlin,
Germany). Stock solutions (each 2 × 10−2 M) were prepared by dis-
solving the nucleoside/nucleotide analogues and aphidicolin (Serva
Electrophoresis, Heidelberg, Germany) in DMSO, 5-FU was dis-
solved in phosphate buffered saline (PBS; pH 7.4). Stock solutions
were diluted with cell culture media to final concentrations of the
test agents of 10−4 to 10−14 M. Then the maximum concentration
of DMSO was 0.5%. In all experiments cells grown in the presence
of the respective pure solvents served as control.
Keratinocytes (4 × 104 cells per well) were grown in 24-well
plates at 37 ◦C and 5% CO2 for 24 h. Then the medium was replaced
with fresh growth medium and cells were incubated with indicated
agents for another 24 h. Pulsed with 1 mCi of [methyl-3H]thymidine
per well cells were incubated for another 3 h. The medium was
removed and cells were washed once with PBS and twice with
ice-cold trichloroacetic acid (5%). The precipitated material was
dissolved in 0.3 M NaOH solution and incorporated thymidine was
quantified by scintillation counting (MicroBeta Plus, Wallac Oy,
Turku, Finland) as described (Manggau et al., 2001).
2.8. Statistics
GraphPad Prism Software Version 4.02 (GraphPad Software Inc.,
San Diego, CA, USA) was used for curve fitting and calculation of IC50
2.5. Cell culture
values. Arithmetic mean
standard error were calculated from
Normal human keratinocytes (NHK) and fibroblasts were iso-
lated from foreskin as described before (Vogler et al., 2003).
For the experiments, cells from at least three donors were
pooled. NHK were cultured in keratinocyte growth medium
(KGM) which was prepared from keratinocyte basal medium
by the addition of 0.1 ng/ml epidermal growth factor, 5.0 ␮g/ml
insulin, 0.5 ␮g/ml hydrocortisone, 30 ␮g/ml bovine pituitary
extract, 50 ␮g/ml gentamicin sulfate, and 50 ng/ml amphotericin
B. For experiments NHK from the second to the fourth pas-
sage were used. Fibroblasts were cultured in Dulbecco’s modified
Eagle’s Medium (DMEM) supplemented with 7.5% FCS, 2 mM
l-glutamine, 100 U/ml penicillin and 100 ␮g/ml streptomycin;
cells from the third to the sixth passage were used for the
experiments. Medium for SCC25 cells (ATCC # CRL-1628, LGC
Promochem, Wesel, Germany) was prepared from Ham’s F12
: DMEM (1:1) by addition of 15% FCS, 0.4 ␮g/ml hydrocorti-
sone, 100 U/ml penicillin and 100 ␮g/ml streptomycin. HaCaT
cells were a gift from Prof. Fusenig (DKFZ, Heidelberg, Ger-
many), T24 cells were purchased from LG Promochem (ATCC
HTB-4). These cells were grown in RPMI 1640 supplemented with
10% FCS, 2 mM l-glutamine, 100 U/ml penicillin and 100 ␮g/ml
streptomycin. HT29 cells (ATCC HTB-38, Germany) were cul-
tured in DMEM containing 10% FCS, 100 U/ml penicillin and
100 ␮g/ml streptomycin. MCF7 cells (ATCC HTB-22) were grown
in Earle’s Minimum Essential Medium (MEME) with 2.0 g/l bicar-
bonate, 2 mM l-glutamine, 50 mg/l gentamicin (969 U/mg) and
110 mg/l sodium pyruvate. SISO cells (ACC327) purchased from
the DSMZ (Deutsche Sammlung von Mikroorganismen und Zel-
lkulturen, Braunschweig, Germany) were grown in RPMI 1640
with DMEM (1:1) supplemented with 10% FCS (heat inacti-
vated, 20 min, 56 ◦C), 2 mM l-glutamine, 100 U/ml penicillin and
100 ␮g/ml streptomycin. Cells were subcultured twice a week fol-
lowing trypsinization using 0.05% trypsin/0.02% ethylenediamine
tetraacetic acid (EDTA).
three independent experiments which were performed each in
triplicate. Cytotoxicity according to the inhibition of cell viability
(MTT test) and cell proliferation (thymidine incorporation assay)
were compared to their respective control by one-way ANOVA fol-
lowed by Dunnett’s test using built-in statistic analysis of GraphPad
Prism (GraphPad Software, Inc., La Jolla, CA), p ≤ 0.05 is considered
to indicate a difference.
3. Results
The 3D-structure of our pol ␣ model (Richartz et al., 2008)
enabled us to identify new promising inhibitors of the enzyme.
Proceeding from the first studies with thymidine and guanosine
analogs (Höltje et al., 2010), we investigated optimized compounds
with an open sugar analog structure and a phosphonate group
with a stable P–C bond which reflects the monophosphate of these
compounds (Table 1). The results of the molecular dynamic sim-
ulations with the triphosphate analog structure indicated OxBu,
OxIsohex and OxHex (Table 1) to be the most active compounds
(Zdrazil et al., in press) which were synthesized and subjected
to pharmacological testing in comparison to aphidicolin and 5-FU
(Fig. 1).
3.1. Synthesis
The desired structure 17 which well reflects its target contains
three well defined parts: the nucleic base core, an aromatic side
chain connected by an amine and a linear phosphonate side chain.
The linear phosphonate was synthesized according to literature
procedures (Bailey et al., 1995; Chen et al., 1996). The aromatic side
chain was synthesized by a strategy developed for BuP-OH (Höltje
et al., 2010) which was slightly modified.
To reach the desired target structures a six-step synthesis route
was successful. First the nucleic base core (bromohypoxanthine)
was coupled with the aromatic side chain (see scheme above).
The coupled product was chlorinated and the aromatic hydroxyl
group was protected with a TBPS-group. This protected compound
(13) can be condensed with the phosphoric side chain, prepared
according to Scheme 1. The substitution of the chlorine atom in 6
position of the purine part of 15 with an amino group gives the
ester 16. This ester can be saponificated by the use of trimethylsilyl
bromide in acetonitrile or with hydrochloric acid in water to give
the desired structure 17. This synthesis is performed for the two
derivatives (R = butyl (OxBu) and R = hexyl (OxHex)) in satisfying
yields.
2.6. MTT dye-reduction assay
Cells (4 × 104 cells/well, in 24-well plates) were grown in
growth medium overnight at 37 ◦C and 5% CO2. Medium was
replaced by fresh growth medium for keratinocytes or basal
medium for the other cell types, respectively. Then the indicated
agents were added to the medium and the cells were incubated
for 48 h. 3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium
bromide (MTT) solution (Sigma–Aldrich; final concentration:
0.5 mg/ml) was added for the last 4 h. After removing supernatants
and solubilizing formazan crystals in DMSO, optical density was
determined at 540 nm as described (Gysler et al., 1997).

A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
15
Table 2
Cytotoxicity −lg IC50 values [M] standard error (IC50 in ␮M) and maximum inhibition (%) derived from viability (MTT-test) assay in NHK, HaCaT cells, fibroblasts and
various tumor cell lines after stimulation with the indicated agents for 48 h. Three independent experiments were performed in triplicate. Cells exposed to the respective
solvent served as control. Reduction was calculated after nonlinear regression as 100 − E; E = activity at 10−4 M in percent; n.d.: −lg IC50 is not determinable with maximum
inhibition ≤15%.
Viability (48 h)
NHK
HaCaT
Fb
SCC25
HT29
MCF7
T24
SISO
Aphidicolin −lg IC50 [M]
(IC50 in ␮M)
8.26 0.41
(0.0055) 44%
6.34 0.21
(0.4571) 86%
8.07 0.25
(0.0085) 54%
6.29 0.29
(0.5129) 62%
7.66 0.23
(0.0219) 67%
8.61 0.17
(0.0025) 61%
8.67 0.21
(0.0021) 62%
8.57 0.19
(0.0027) 51%
inhibition
5-FU
−lg IC50 [M]
(IC50 in ␮M)
inhibition
5.08 0.48
(8318) 41%
6.22 0.16
(0.6026) 80%
4.94 0.25
(11.482) 88%
6.01 0.28
(0.9772) 42%
7.80 0.23
(0.0158) 66%
8.31 0.16
(0.0049) 70%
12.55 0.31
(0.0000003)
43%
8.39 0.15
(0.0041) 43%
BuP-OH
OxBu
−lg lC50 [M]
(IC50 in ␮M)
inhibition
−lg IC50 [M]
(IC50 in ␮M)
inhibition
−lg IC50 [M]
(IC50 in ␮M)
inhibition
−lg IC50 [M]
(IC50 in ␮M)
inhibition
4.58 0.24
(26.303) 66%
4.64 0.32
(22.909) 68%
7.61 0.25
(0.0245) 53%
5.05 0.37
(8.913) 65%
6.37 0.21
(0.4266) 67%
7.32 0.21
(0.0479) 75%
8.25 0.21
(0.0056) 66%
7.34 0.33
(0.0457) 38%
n.d.
n.d.
n.d.
n.d.
4.27 0.91
(53.703) 62%
9.43 0.22
(0.00037) 39% (4.8978) 58%
11.31 0.21
7.89 0.15
(0.0129) 60%
7.77 0.28
(0.0170) 53%
8.76 0.14
(0.0017) 46%
Oxlsohex
OxHex
-
4.93 0.26
(11.749) 20%
–
–
–
–
7.89 0.22
(0.0129) 43%
8.01 1.17
(0.0098) 17%
8.25 0.16
(0.0056) 65%
9.12 0.11
(0.00076) 74% (0.0040) 57%
8.40 0.23
9.25 0.23
(0.00056) 51% (0.0020) 65%
8.70 0.17
8.77 0.14
(0.0017) 47%
3.2. Cytotoxicity in skin cells
fibroblasts was seen in concentrations rather close to those toxic in
SCC25 cells with OxHex, yet OxBu was only toxic for fibroblasts at
very high concentrations.
In contrast to molecular modelling which has to be based on the
triphosphates of the nucleoside analogs, cytotoxicity was inves-
tigated with the nucleoside/nucleotide analogs themselves. The
phosphorylated substances would be too hydrophilic to penetrate
into the cells, activation of the substances by cellular kinases is to
be expected. To investigate the potency of our compounds as a new
therapy option in skin cancer we compared the test agents to 5-FU
and aphidicolin. To be suitable for clinical use, toxicity in SCC25
cells should exceed toxicity in NHK and possibly also in fibroblasts.
Moreover, introducing the phosphonate group we aimed for higher
activity exceeding those of BuP-OH (Höltje et al., 2010; Schäfer-
Korting et al., 2010).
Viability of normal and transformed skin cells was determined
by measuring formazan formation by the mitochondrial reduc-
tase system (MTT dye-reduction assay, Table 2). Only viable and
metabolically active cells will produce the blue formazan, thus the
amount of dye is set in direct proportion to the cell number. The
active concentrations of BuP-OH exceeding those of aphidicolin
by several magnitudes make this agent not well applicable in the
topical therapy of actinic keratosis. Moreover, BuP-OH affects nor-
mal and transformed keratinocytes as well as fibroblasts potently
(Table 2), thus there is no selective toxicity in skin tumor cells. In
fact, aphidicolin which failed to come into clinical use (Schäfer-
Korting et al., 2010) proved to be even more toxic for NHK as
compared to SCC25 cells.
Proliferation of skin cells was determined by quantification of
the thymidine incorporation into DNA (Table 3). Aphidicolin com-
pletely inhibited DNA synthesis in NHK whereas the tumor cell
line SCC-25 was less sensitive. The higher toxicity for normal ker-
atinocytes makes aphidicolin unsuitable for use in skin cancer.
BuP-OH inhibits NHK proliferation completely and SCC-25 cells are
strongly inhibited, too, yet only at rather high concentrations. The
overall influences of OxIsohex, OxBu and OxHex on skin cell prolif-
eration correspond to the results of viability testing (Table 2). OxBu
has no effect on NHK and HaCaT proliferation but attacks SCC25
proliferation even more than seen with the MTT test. The results
of OxHex once more underline the high potency of the substance.
OxIsohex had no effect on HaCaT cells but inhibited the prolifer-
ation of NHK and SCC25 cells by about 22% and 64%, respectively.
Due to the clearly superior activity of OxBu and OxHex, we excluded
OxIsohex from further testing.
The inhibitory effect of 5-FU on proliferation was not tested
with the thymidine uptake assay, since exogenous thymidine prob-
ably antagonizes the thymidine analog 5-FU related depletion of
Surprisingly the nucleoside phosphonate having an iso-hexyl
substituent (OxIsohex) did not affect NHK and OxIsohex reduced
viability of SCC25 cells only at high concentrations and by about
20%. By contrast, both phosphonates with a n-alkyl chain were
active at low concentrations and even appeared to differentiate
between NHK and SCC25 cells favourably (Fig. 2). Referring to the
IC50 value, OxBu and OxHex are 1000 fold more toxic in SCC25 than
5-FU. OxBu reached with 39% maximum inhibition of SCC25 viabil-
ity almost the potency of 5-FU but did not influence either NHK or
HaCaT cells. With 74% maximum inhibition OxHex is much more
active than 5-FU and even more toxic than aphidicolin. Although
OxHex affects NHK, too, active concentrations with NHK are at least
10 fold higher and the maximum inhibition is clearly (43%) less
pronounced than with SCC25 (74%). HaCaT showed a minor sus-
ceptibility to OxBu and a moderate one to OxHex. Cytotoxicity in
Fig. 2. Cytotoxicity (MTT dye-reduction assay) of guanosine phosphonate con-
geners. Dose–response curves. NHK (ꢀ: OxBu, ꢀ: OxHex) and SCC25 cells (ꢁ: OxBu,
♦: OxHex) were exposed for 48 h to OxBu (—) and to OxHex (- - - - -). Three inde-
pendent experiments were performed in triplicate. Cells exposed to the respective
solvent served as control.

16
A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
Table 3
Inhibition of proliferation. −lg IC50 values [M] standard error (IC50 in ␮M) and maximum inhibition (%) derived from thymidine incorporation (proliferation) assay in NHK,
HaCaT cells, fibroblasts and tumor cell lines after stimulation by the indicated agents for 24 h. Three independent experiments were performed in triplicate. Cells exposed to
the respective solvent served as control. Reduction was calculated after nonlinear regression as 100 − E; E = activity at 10−4 M in percent; n.d.: −lg IC50 is not determinable
with maximum inhibition ≤15%.
Proliferation (24 h)
NHK
HaCaT
Fb
SCC-25
HT29
–
MCF7
–
T24
–
SISO
–
Aphidicolin −lg IC50 [M]
(IC50 in ␮M)
7.08 0.05
(0.0832) 100% (0.0891) 99%
7.05 0.04
6.07 0.18
(0.8511) 58%
inhibition
BuP-OH
OxBu
−lg IC50 [M]
(IC50 in ␮M)
inhibition
−lg IC50 [M]
(IC50 in ␮M)
inhibition
5.59 0.27
(2.5704) 100% (9.7724) 67%
5.01 0.35
5.07 0.38
(8.5114) 87%
–
–
–
–
n.d.
n.d.
4.84 0.68
(14.454) 48%
9.14 0.16
(0.00072) 74% (0.0021) n.d.
8.68 1.44
8.96 0.33
(0.0011) 47%
5.98 0.23
(1.0471) 49%
7.81 0.16
(0.0155) 32%
Oxlsohex
OxHex
−lg IC50 [M]
(IC50 in ␮M)
inhibition
−lg IC50 [M]
(IC50 in ␮M)
inhibition
4.95 0.65
(11.220) 22%
7.21 1.27
(0.0617) n.d.
–
9.16 0.21
(0.00069) 64%
–
–
–
–
5.99 0.26
(1.0233) 41%
9.32 0.75
(0.00048) 17% (0.0490) 63%
7.31 0.20
7.39 0.15
(0.0410) 69%
7.16 0.23
(0.0692) 24%
7.01 0.24
(0.0977) 61%
7.62 0.34
(0.0240) 40%
8.19 0.09
(0.0065) 39%
Table 4
Cytotoxicity and inhibition of proliferation. −lg IC50 values [M] standard error (IC50 in ␮M) and maximum inhibition (%) derived from viability (MTT-test) and thymidine
incorporation (proliferation) assay in NHK and SCC25 cells after stimulation with the indicated agents. Three independent experiments were performed in triplicate. Cells
exposed to the respective solvent served as control. Reduction was calculated after nonlinear regression as 100 − E; E = activity at 10−4 M in percent.
Viability (48 h)
NHK
SCC25
OxBu-DE
OxHex-DE
−lg IC50 [M] (IC50 in ␮M) inhibition
−lg IC50 [M] (IC50 in ␮M) inhibition
8.88 0.20 (0.0013) 57%
7.99 0.15 (0.0102) 69%
7.81 0.19 (0.0155) 72%
8.93 0.14 (0.0012) 72%
Proliferation (24 h)
NHK
SCC25
OxBu-DE
OxHex-DE
−lg IC50 [M] (IC50 in ␮M) inhibition
−lg IC50 [M] (IC50 in ␮M) inhibition
8.11 0.25 (0.0078) 40%
7.01 0.24 (0.0978) 61%
7.36 0.20 (0.4365) 63%
7.38 0.15 (0.0417) 69%
endogenous dTTP, and thus an antiproliferative effect (Boucher et
al., 2006).
In fact cytotoxicity (MTT test) for NHK increased with the pre-
prodrugs (Table 4). By OxBu-DE the maximum inhibition of NHK
viability was 57%. Therewith the pre-prodrug had a clearly cyto-
toxic effect in NHK, while OxBu did not influence NHK at all. An
increased cytotoxicity was also seen with OxHex-DE, the maximum
inhibition in NHK cells increased from 43% (OxHex) to 69% while
the active concentrations remained almost the same. With respect
to the tumor cell line SCC25, OxBu-DE more efficiently inhibited
viability than OxBu admittedly at a more than 10 fold higher con-
centration. OxHex-DE effects were close to those of OxHex taking
the maximum inhibition and active concentrations into account.
Inhibition of thymidine incorporation into DNA well agreed with
MTT data. Once more, OxBu-DE did not show the selectivity for
tumor cells seen with OxBu. Taken together, the activity of the
phosphonate diethylesters is by no means superior to the activ-
ity of the non-esterified nucleoside phosphonates with partially
removed sugar moiety.
3.3. Cytotoxicity with other tumor cell lines
Due to the high potency and selectivity of OxBu and OxHex in
skin tumor cells, we next tested these compounds in other tumor
cell lines, too. These encompass colorectal carcinoma cells (HT29),
mammary carcinoma cells (MCF7), bladder carcinoma cells (T24),
and adenocervical carcinoma cells (SISO). In general, potency, max-
imum inhibitory effects and IC50 values of aphidicolin and OxHex
were rather close (Table 2), whereas OxBu appeared clearly more
active in HT29, while less active in MCF7 and T24 cells. Of note, only
very low concentrations of 5-FU were needed to affect T24 cells
and 5-FU’s potency with the other tumor cell lines, too, exceeded
cytotoxicity in SCC25 cells.
In addition, we examined the potency of OxBu and OxHex to
inhibit DNA synthesis (Table 3). −lg IC50 values were all close to
the range 7–8 which is well in accordance with the cytotoxicity, yet
maximum inhibition of thymidine incorporation was less marked
compared to the decline of viability.
4. Discussion
Since current therapy options for actinic keratosis and squa-
mous cell carcinoma are not satisfying (Jorizzo, 2004b; Stockfleth
and Kerl, 2006) we are searching for alternative approaches. DNA
pol ␣ is an interesting target since it catalyzes the very first steps
in DNA replication. Next to modelling of the active site of pol
␣ (Richartz et al., 2008) we have recently identified structurally
and pharmacologically selective nucleoside inhibitors by molecu-
lar docking simulations (Höltje et al., 2010; Zdrazil et al., in press).
As we aim for topical treatment, we have to assure sufficient bio-
availability in the tumor lesion when applied onto the skin which
implies a molecular mass under 500 g/mol and a well-balanced
hydrophilicity/lipophilicity (log P 1–3) (Korting and Schäfer-
Korting, 2009). Since drugs penetrate the skin at a rate of a few per-
cent at most (Korting and Schäfer-Korting, 2009) we aim for carrier-
3.4. Cytotoxic potency of OxBu and OxHex diethylesters
Since normal human keratinocytes produce esterases in great
amount (Gysler et al., 1997; Haberland et al., 2006; Ngawhirunpat
et al., 2003) we investigated whether the respective diethylesters
(Table 1) of OxBu and OxHex affect NHK and SCC25, too. The higher
lipophilicity may favour penetration of the target cell as well as
skin penetration when applied topically. As the diethylesters of
OxBu and OxHex first need to be cleaved to the phosphonate com-
pounds (prodrug formation) and then to be converted into the
three-phosphate analog, they can be designated as pre-prodrugs.
The results of cytotoxicity testing were disappointing.

A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
17
improved skin penetration, too, which becomes possible in particu-
lar by loading to lipidic (Jenning et al., 2000; Lombardi Borgia et al.,
2005; Santos Maia et al., 2002; Stecova et al., 2007) and non-lipidic
nanoparticles (Küchler et al., 2009). The entrapment efficiency by
lipophilic carriers increases with lipophilicity, too (Schäfer-Korting
et al., 2007). After penetrating into the skin tumor cells nucleoside
analog pol ␣ inhibitors have to be activated via stepwise phos-
phorylation. This can be done by cytosolic nucleoside kinases and
the mitochondrial deoxy guanosine kinase activating deoxyguano-
sine and analogues (Eriksson et al., 2002). We were able to detect
both, but also the anionic outward transporters MRP4 and MRP5 in
normal and transformed keratinocytes (Richartz et al., 2008).
To improve the activity of nucleoside analogues (Höltje et al.,
2010) we next aimed for an improved activation process as it is real-
ized with some antiviral agents (e. g. adefovir, tenofovir; Fig. 1). In
fact, it is the nucleoside monophosphate formation which is the rate
limited process (Spadari et al., 1998). While aiming to circumvent
this step as well as to avoid instability we developed phosphonate
derivatives, containing a metabolically stable P–C bond. These are
also less polar (Table 1) and therefore should penetrate the skin
to higher amounts than the corresponding phosphates. Moreover,
we aimed to overcome the lack of selectivity for tumor cells of
the nucleoside analogs. OxBu, OxIsohex, OxHex (Table 1) effects
were compared to 5-FU effects as we intend to surmount the lim-
ited efficacy of the current topical therapy of actinic keratosis. The
well-known inhibitor of pol ␣ and ␦ aphidicolin (Fig. 1) served
as reference, too. The agent interferes with the nucleotide incor-
poration into the elongated DNA during the S phase and induces
apoptosis and necrosis in normal and transformed keratinocytes
(Höltje et al., 2010). Interestingly, OxBu and OxHex were about
1000 fold more active than 5-FU and even seem to differenti-
ate between normal human keratinocytes and SCC25 cells (Fig. 2)
whereas 5-FU differentiates less efficiently between primary cells
and the cancer cell line. Aphidicolin which failed to be introduced
into therapy in fact is more toxic for NHK (Tables 2 and 3).
OxBu which is as effective as 5-FU shows no influence on NHK
at all and even none on HaCaT cells. Especially OxHex is clearly
more cytotoxic than aphidicolin or 5-FU, the maximum toxicity in
NHK, however, was close to the one seen with 5-FU. These findings
are well in accordance to the predictions made with our 3D-model
which suggested OxHex to be one of the best fitting substances
(Zdrazil et al., in press). On top of this, the most potent effects of
the phosphonate analogues on tumor cells might be related to the
abolition of the first, time consuming activation step which facili-
tates the formation of the activated tri-phosphate analogs within
the cell. According to the modelling data, however, OxIsohex should
have been more active as compared to OxBu which was not true
(Tables 2 and 3). This might be due to hindered access of the non-
linear side chain of OxIsohex to the active site of the enzyme.
Neverless our data demonstrate the suitability of molecular mod-
elling for the identification of new drug candidates. Moreover, the
close relation of predicted and observed activity and the structure
relation of BuP-OH, OxBu and OxHex to the well-known inhibitors
of viral pol ␣ inhibitors which have been used to build-up our
model (Richartz et al., 2008) indicates that pol ␣ inhibition is of
central importance with respect to cytotoxicity of the drug candi-
date. Moreover, enzyme inhibition needs to be studied, too, yet asks
for the synthesis of the activated OxBu and OxHex diphosphates.
BuP-OH which is discussed in detail elsewhere (Höltje et al.,
2010) is clearly less active than OxBu and OxHex and does not
differentiate between normal and tumor cells. OxHex decreases
the viability of normal human fibroblasts by 65% already at lower
concentrations (−lg IC50 8.25) whereas OxBu acts only at high con-
centrations (−lg IC50 4.27). Therefore, OxHex might be more toxic
as compared to OxBu, if the substance should penetrate into the
dermis at a relevant amount. Therefore, skin penetration studies as
well as controlled penetration into the epidermis (epidermal tar-
geting) by loading to nanoparticles (Jenning et al., 2000; Santos
Maia et al., 2002) will be of high relevance. Though loading should
be facilitated by the more lipophilic character of the ester pre-
prodrugs OxBu-DE and OxHex-DE, this is no option because of a
clear decline in selectivity for skin tumor cells, probably due to the
greater lipophilicity (Table 1) and thus a more efficient uptake by
NHK.
Moreover, a potential antiviral effect could be of interest. In fact
the HPV-DNA-load is higher in actinic keratosis lesions compared
to other types of skin cancer (Weissenborn et al., 2005). HPV might
be of particular interest in the context of relapse of actinic keratosis
which is a frequent phenomenon (Dianzani et al., 2008). Moreover,
patients suffering from squamous cell carcinoma of the skin show
higher seroreactivity against HPV as compared to patients suffering
from basal cell carcinoma (Andersson et al., 2008). In fact, we know
today that persistent HPV infection is linked to HPV-DNA amplifi-
cation by DNA polymerases of the host (Park et al., 1994). In the
context, it certainly also is of note that anti-virally active toll-like
receptor 7 and toll-like receptor 7/8 agonists like imiquimod do not
only treat genital warts which undoubtedly is an infectious disease
due to HPV, but also actinic keratoses (Miller et al., 2008).
Due to their great potential as antitumor agents for actinic ker-
atosis and squamous cell carcinoma we tested OxBu and OxHex in
other tumor cell lines like HT29, MCF7, T24 and SISO cells, too. The
outcome may open the horizon in the treatment of the respective
tumors like colorectal, mammary, bladder and adenocervical carci-
noma, respectively. Notably promising results with OxBu in HT29
cells were obtained, the effective concentrations being even below
the nanomolecular range.
5. Conclusion
In conclusion, since OxBu and OxHex preferentially inhibit
tumor cells in the nano molar range we identified innovative
nucleotide analogs which may outperform 5-FU. Thus these agents
may offer a new approach for the treatment of actinic keratosis and
other forms of skin cancer, if active by the topical application route
to control unwanted side effects.
Acknowledgment
Financial support of the German Ministry of Education and
Research (13N9062 and 13N9061) is gratefully acknowledged.
Declaration of interest: Two authors (Hans-Dieter Höltje, Monika
Schäfer-Korting) are inventors of the respective patent applied for
by RIEMSER Arzneimittel AG (Greifswald–Insel Riems), European
Appl. No.: 07090098.0. Moreover, Monika Schäfer-Korting acts as a
consultant for RIEMSER Arzneimittel AG (Greifswald–Insel Riems)
with respect to the development of dermatics.
References
Andersson, K., Waterboer, T., Kirnbauer, R., Slupetzky, K., Iftner, T., de Villiers, E.M.,
Forslund, O., Pawlita, M., Dillner, J., 2008. Seroreactivity to cutaneous human
papillomaviruses among patients with nonmelanoma skin cancer or benign skin
lesions. Cancer Epidemiol. Biomark. Prev. 17, 189–195.
Anwar, J., Wrone, D.A., Kimyai-Asadi, A., Alam, M., 2004. The development of actinic
keratosis into invasive squamous cell carcinoma: evidence and evolving classi-
fication schemes. Clin. Dermatol. 22, 189–196.
Bailey, W.F., Zarcone, L.M.J., Rivera, A.D., 1995. Selective protection of 1,2- and 1,3-
diols via acylative cleavage of cyclic formals. J. Org. Chem. 60, 2532–2536.
Boucher, P.D., Im, M.M., Freytag, S.O., Shewach, D.S., 2006. A novel mechanism of
synergistic cytotoxicity with 5-fluorocytosine and ganciclovir in double suicide
gene therapy. Cancer Res. 66, 3230–3237.
Buckman, S.Y., Gresham, A., Hale, P., Hruza, G., Anast, J., Masferrer, J., Pentland, A.P.,
1998. COX-2 expression is induced by UVB exposure in human skin: implications
for the development of skin cancer. Carcinogenesis 19, 723–729.

18
A. Schwanke et al. / International Journal of Pharmaceutics 397 (2010) 9–18
Challacombe, J.M., Suhrbier, A., Parsons, P.G., Jones, B., Hampson, P., Kavanagh, D.,
Rainger, G.E., Morris, M., Lord, J.M., Le, T.T., Hoang-Le, D., Ogbourne, S.M., 2006.
Neutrophils are a key component of the antitumor efficacy of topical chemother-
apy with ingenol-3-angelate. J. Immunol. 177, 8123–8132.
Chen, W., Flavin, M.T., Filler, R., Xu, Z.-Q., 1996. An improved synthesis of 9-
[2-(diethoxyphosphonomethoxy)ethyl]adenine and its analogues with other
purine bases utilizing the mitsunobu reaction. Nucleosides Nucleotides 15,
1771–1778.
Matsubara, K., Saito, A., Tanaka, A., Nakajima, N., Akagi, R., Mori, M., Mizushina, Y.,
2006. Catechin conjugated with fatty acid inhibits DNA polymerase and angio-
genesis. DNA Cell Biol. 25, 95–103.
Matsubara, K., Saito, A., Tanaka, A., Nakajima, N., Akagi, R., Mori, M., Mizushina,
Y., 2007. Epicatechin conjugated with fatty acid is a potent inhibitor of DNA
polymerase and angiogenesis. Life Sci. 80, 1578–1585.
Miller, R.L., Meng, T.C., Tomai, M.A., 2008. The antiviral activity of Toll-like receptor
7 and 7/8 agonists. Drug News Perspect. 21, 69–87.
Dianzani, C., Pierangeli, A., Chiricozzi, A., Avola, A., Degener, A.M., 2008. Cuta-
neous human papillomaviruses as recurrence factor in actinic keratoses. Int.
J. Immunopathol. Pharmacol. 21, 145–152.
Ngawhirunpat, T., Kawakami, N., Hatanaka, T., Kawakami, J., Adachi, I., 2003. Age
dependency of esterase activity in rat and human keratinocytes. Biol. Pharm.
Bull. 26, 1311–1314.
Eriksson, S., Munch-Petersen, B., Johansson, K., Eklund, H., 2002. Structure and
function of cellular deoxyribonucleoside kinases. Cell. Mol. Life Sci. 59,
1327–1346.
Gandhi, V., Huang, P., Chapman, A.J., Chen, F., Plunkett, W., 1997. Incorporation
of fludarabine and 1-beta-d-arabinofuranosylcytosine 5ꢂ-triphosphates by DNA
polymerase alpha: affinity, interaction, and consequences. Clin. Cancer Res. 3,
1347–1355.
Gupta, A.K., Davey, V., McPhail, H., 2005. Evaluation of the effectiveness of
imiquimod and 5-fluorouracil for the treatment of actinic keratosis: crit-
ical review and meta-analysis of efficacy studies. J. Cutan. Med. Surg. 9,
209–214.
Pariser, D., Loss, R., Jarratt, M., Abramovits, W., Spencer, J., Geronemus, R., Bailin, P.,
Bruce, S., 2008. Topical methyl-aminolevulinate photodynamic therapy using
red light-emitting diode light for treatment of multiple actinic keratoses: a
randomized, double-blind, placebo-controlled study. J. Am. Acad. Dermatol. 59,
569–576.
Park, P., Copeland, W., Yang, L., Wang, T., Botchan, M.R., Mohr, I.J., 1994. The cellular
DNA polymerase alpha-primase is required for papillomavirus DNA replica-
tion and associates with the viral E1 helicase. Proc. Natl. Acad. Sci. USA 91,
8700–8704.
Perry, C.M., Lamb, H.M., 1999. Topical imiquimod: a review of its use in genital warts.
Drugs 58, 375–390.
Gysler, A., Lange, K., Korting, H.C., Schäfer-Korting, M., 1997. Prednicarbate bio-
transformation in human foreskin keratinocytes and fibroblasts. Pharm. Res.
14, 793–797.
Haberland, A., Schreiber, S., Maia, C.S., Rübbelke, M.K., Schaller, M., Korting, H.C.,
Kleuser, B., Schimke, I., Schäfer-Korting, M., 2006. The impact of skin viability
on drug metabolism and permeation – BSA toxicity on primary keratinocytes.
Toxicol. In Vitro 20, 347–354.
Richartz, A., Höltje, M., Brandt, B., Schäfer-Korting, M., Höltje, H.D., 2008. Targeting
human DNA polymerase alpha for the inhibition of keratinocyte proliferation.
Part 1. Homology model, active site architecture and ligand binding. J. Enzyme
Inhib. Med. Chem. 23, 94–100.
Rivers, J.K., Arlette, J., Shear, N., Guenther, L., Carey, W., Poulin, Y., 2002. Topical
treatment of actinic keratoses with 3.0% diclofenac in 2.5% hyaluronan gel. Br. J.
Dermatol. 146, 94–100.
Höltje, M., Richartz, A., Zdrazil, B., Schwanke, A., Dugovic, B., Murruzzu, C., Reissig,
H.U., Korting, H.C., Kleuser, B., Höltje, H.D., Schäfer-Korting, M., 2010. Human
polymerase alpha inhibitors for skin tumors. Part 2. Modeling, synthesis and
influence on normal and transformed keratinocytes of new thymidine and
purine derivatives. J. Enzyme Inhib. Med. Chem. 25, 250–265.
Jenning, V., Schäfer-Korting, M., Gohla, S., 2000. Vitamin A-loaded solid lipid
nanoparticles for topical use: drug release properties. J. Control Release 66,
115–126.
Jiang, H.Y., Hickey, R.J., Abdel-Aziz, W., Malkas, L.H., 2000. Effects of gemcitabine
and araC on in vitro DNA synthesis mediated by the human breast cell DNA
synthesome. Cancer Chemother. Pharmacol. 45, 320–328.
Jorizzo, J., 2004a. Topical treatment of actinic keratosis with fluorouracil: is irritation
associated with efficacy? J. Drugs Dermatol. 3, 21–26.
Jorizzo, J., Stewart, D., Bucko, A., Davis, S.A., Espy, P., Hino, P., Rodriguez, D., Savin, R.,
Stough, D., Furst, K., Connolly, M., Levy, S., 2002. Randomized trial evaluating a
new 0.5% fluorouracil formulation demonstrates efficacy after 1-, 2-, or 4-week
treatment in patients with actinic keratosis. Cutis 70, 335–339.
Jorizzo, J.L., 2004b. Current and novel treatment options for actinic keratosis. J. Cutan.
Med. Surg. 8 (Suppl. 3), 13–21.
Santos Maia, C., Mehnert, W., Schaller, M., Korting, H.C., Gysler, A., Haberland, A.,
Schäfer-Korting, M., 2002. Drug targeting by solid lipid nanoparticles for dermal
use. J. Drug Target. 10, 489–495.
Schäfer-Korting, M., Höltje, M., Korting, H.C., Höltje, H.D., 2010. Innovative agents for
actinic keratosis and nanocarriers enhancing skin penetration. Skin Pharmacol.
Physiol. 23, 6–14.
Schäfer-Korting, M., Mehnert, W., Korting, H.C., 2007. Lipid nanoparticles for
improved topical application of drugs for skin diseases. Adv. Drug Deliv. Rev.
59, 427–443.
Sessa, C., Zucchetti, M., Davoli, E., Califano, R., Cavalli, F., Frustaci, S., Gumbrell, L.,
Sulkes, A., Winograd, B., D’Incalci, M., 1991. Phase I and clinical pharmacological
evaluation of aphidicolin glycinate. J. Natl. Cancer Inst. 83, 1160–1164.
Siller, G., Gebauer, K., Welburn, P., Katsamas, J., Ogbourne, S.M., 2009. PEP005
(ingenol mebutate) gel, a novel agent for the treatment of actinic keratosis:
results of a randomized, double-blind, vehicle-controlled, multicentre, phase
IIa study. Australas. J. Dermatol. 50, 16–22.
Spadari, S., Maga, G., Verri, A., Focher, F., 1998. Molecular basis for the antiviral
and anticancer activities of unnatural l-beta-nucleosides. Expert Opin. Investig.
Drugs 7, 1285–1300.
Korting, H.C., Schäfer-Korting, M., 2009. Carriers in the topical treatment of skin
disease. In: Schäfer-Korting, M. (Ed.), Handbook of Experimental Pharmacology,
Vol. 197 Drug Delivery, Springer, Heidelberg, pp. 435–468.
Küchler, S., Radowski, M.R., Blaschke, T., Dathe, M., Plendl, J., Haag, R., Schäfer-
Korting, M., Kramer, K.D., 2009. Nanoparticles for skin penetration enhancement
– a comparison of a dendritic core-multishell-nanotransporter and solid lipid
nanoparticles. Eur. J. Pharm. Biopharm. 71, 243–250.
Stecova, J., Mehnert, W., Blaschke, T., Kleuser, B., Sivaramakrishnan, R., Zouboulis,
C.C., Seltmann, H., Korting, H.C., Kramer, K.D., Schäfer-Korting, M., 2007. Cypro-
terone acetate loading to lipid nanoparticles for topical acne treatment: particle
characterisation and skin uptake. Pharm. Res. 24, 991–1000.
Stockfleth, E., 2008. Aktinische Keratose, AWMF-Leitlinien. AWMF.
Stockfleth, E., Kerl, H., 2006. Guidelines for the management of actinic keratoses.
Eur. J. Dermatol. 16, 599–606.
Kuriyama, I., Musumi, K., Yonezawa, Y., Takemura, M., Maeda, N., Iijima, H., Hada,
T., Yoshida, H., Mizushina, Y., 2005. Inhibitory effects of glycolipids fraction
from spinach on mammalian DNA polymerase activity and human cancer cell
proliferation. J. Nutr. Biochem. 16, 594–601.
Stockfleth, E., Meyer, T., Benninghoff, B., Salasche, S., Papadopoulos, L., Ulrich, C.,
Christophers, E., 2002. A randomized, double-blind, vehicle-controlled study to
assess 5% imiquimod cream for the treatment of multiple actinic keratoses. Arch.
Dermatol. 138, 1498–1502.
Lebwohl, M., Dinehart, S., Whiting, D., Lee, P.K., Tawfik, N., Jorizzo, J., Lee, J.H., Fox,
T.L., 2004. Imiquimod 5% cream for the treatment of actinic keratosis: results
from two phase III, randomized, double-blind, parallel group, vehicle-controlled
trials. J. Am. Acad. Dermatol. 50, 714–721.
Levy, S., Furst, K., Chern, W., 2001. A comparison of the skin permeation of three
topical 0.5% fluorouracil formulations with that of a 5% formulation. Clin. Ther.
23, 901–907.
Lombardi Borgia, S., Regehly, M., Sivaramakrishnan, R., Mehnert, W., Korting, H.C.,
Danker, K., Röder, B., Kramer, K.D., Schäfer-Korting, M., 2005. Lipid nanoparticles
for skin penetration enhancement–correlation to drug localization within the
particle matrix as determined by fluorescence and parelectric spectroscopy. J.
Control Release 110, 151–163.
Maeda, N., Kokai, Y., Ohtani, S., Sahara, H., Hada, T., Ishimaru, C., Kuriyama, I.,
Yonezawa, Y., Iijima, H., Yoshida, H., Sato, N., Mizushina, Y., 2007a. Anti-tumor
effects of the glycolipids fraction from spinach which inhibited DNA polymerase
activity. Nutr. Cancer 57, 216–223.
Maeda, N., Kokai, Y., Ohtani, S., Sahara, H., Kuriyama, I., Kamisuki, S., Takahashi, S.,
Sakaguchi, K., Sugawara, F., Yoshida, H., Sato, N., Mizushina, Y., 2007b. Anti-
tumor effects of dehydroaltenusin, a specific inhibitor of mammalian DNA
polymerase alpha. Biochem. Biophys. Res. Commun. 352, 390–396.
Manggau, M., Kim, D.S., Ruwisch, L., Vogler, R., Korting, H.C., Schäfer-Korting,
M., Kleuser, B., 2001. 1Alpha,25-dihydroxyvitamin D3 protects human ker-
atinocytes from apoptosis by the formation of sphingosine-1-phosphate. J.
Invest. Dermatol. 117, 1241–1249.
Tetko, I.V., Tanchuk, V.Y., 2002. Application of associative neural networks for pre-
diction of lipophilicity in ALOGPS 2.1 program. J. Chem. Inf. Comput. Sci. 42,
1136–1145.
Tetko, I.V., Tanchuk, V.Y., Villa, A.E., 2001. Prediction of n-octanol/water partition
coefficients from PHYSPROP database using artificial neural networks and E-
state indices. J. Chem. Inf. Comput. Sci. 41, 1407–1421.
Vasiljevic, N., Hazard, K., Dillner, J., Forslund, O., 2008. Four novel human betapapil-
lomaviruses of species 2 preferentially found in actinic keratosis. J. Gen. Virol.
89, 2467–2474.
Vogler, R., Sauer, B., Kim, D.S., Schäfer-Korting, M., Kleuser, B., 2003. Sphingosine-
1-phosphate and its potentially paradoxical effects on critical parameters of
cutaneous wound healing. J. Invest. Dermatol. 120, 693–700.
Weinstock, M.A., 2006. Controversies in the public health approach to keratinocyte
carcinomas. Br. J. Dermatol. 154 (Suppl. 1), 3–4.
Weissenborn, S.J., De Koning, M.N., Wieland, U., Quint, W.G., Pfister, H.J., 2009.
Intrafamilial transmission and family-specific spectra of cutaneous betapapillo-
maviruses. J. Virol. 83, 811–816.
Weissenborn, S.J., Nindl, I., Purdie, K., Harwood, C., Proby, C., Breuer, J., Majewski,
S., Pfister, H., Wieland, U., 2005. Human papillomavirus-DNA loads in actinic
keratoses exceed those in non-melanoma skin cancers. J. Invest. Dermatol. 125,
93–97.
Zdrazil, B., Schwanke, A., Schmitz, B., Schäfer-Korting, M., Höltje, H.D. Molecular
modeling studies of new potential human DNA polymerase alpha inhibitors. J.
Enzyme Inhib. Med. Chem., in press.