Inactivation of the human papillomavirus-16 E6 oncoprotein by organic disulfides

Article abstract of DOI:10.1016/S0968-0896(00)00193-0

We are investigating compounds that could be useful in the treatment of neoplastic lesions of the cervix by acting on the oncoprotein E6 of human papillomavirus-16. The E6 protein contains two potential zinc-binding domains that are required for most of its functions. We have published tests that measure (i) the release of zinc ions after chemical alteration of the cysteine groups of these zinc-binding domains (TSQ assay), (ii) the interaction of E6 with the cellular proteins E6AP and E6BP (BIACORE assay), and (iii) the viability of tumor cell lines that require the continuous expression of HPV oncoproteins (WST1 assay). Based on these tests, we identified 4,4'-dithiodimorpholine as a potential lead compound. In this study we examined whether the dithiobisamine moiety of 4,4'-dithiodimorpholine may be an important molecular prerequisite for further drug development in this system. We have evaluated 59 new substances including organic disulfides and those containing the dithiobisamine moiety, as well as structural analogues. The compounds with significant reactivity in all three assays were observed only for dithiobisamine derivatives with saturated cyclic amines and aryl substituted piperazines. The identity of these substances suggests that the N-S-S-N moiety is necessary but not sufficient for reactivity in our assays, and that dithiobisamine based substances are useful as lead compounds that target the cysteine groups of HPV-16 E6 zinc fingers. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(00)00193-0

Bioorganic & Medicinal Chemistry 8 (2000) 2549±2560
Inactivation of the Human Papillomavirus-16 E6 Oncoprotein
by Organic Disul®des
Walter Beerheide,*y Mui Mui Sim,{ Yee-Joo Tan, Hans-Ulrich Bernardx
and Anthony E. Ting*
Drug Screen Development Laboratory, Institute of Molecular and Cell Biology, 30 Medical Drive, Singapore 117609
Received 1 May 2000; accepted 27 June 2000
AbstractÐWe are investigating compounds that could be useful in the treatment of neoplastic lesions of the cervix by acting on the
oncoprotein E6 of human papillomavirus-16. The E6 protein contains two potential zinc-binding domains that are required for
most of its functions. We have published tests that measure (i) the release of zinc ions after chemical alteration of the cysteine
groups of these zinc-binding domains (TSQ assay), (ii) the interaction of E6 with the cellular proteins E6AP and E6BP (BIACORE
assay), and (iii) the viability of tumor cell lines that require the continuous expression of HPV oncoproteins (WST1 assay). Based
on these tests, we identi®ed 4,40-dithiodimorpholine as a potential lead compound. In this study we examined whether the dithio-
bisamine moiety of 4,40-dithiodimorpholine may be an important molecular prerequisite for further drug development in this sys-
tem. We have evaluated 59 new substances including organic disul®des and those containing the dithiobisamine moiety, as well as
structural analogues. The compounds with signi®cant reactivity in all three assays were observed only for dithiobisamine derivatives
with saturated cyclic amines and aryl substituted piperazines. The identity of these substances suggests that the N±S±S±N moiety is
necessary but not sucient for reactivity in our assays, and that dithiobisamine based substances are useful as lead compounds that
target the cysteine groups of HPV-16 E6 zinc ®ngers. # 2000 Elsevier Science Ltd. All rights reserved.
Introduction
E6 and E7 are expressed in benign lesions, which main-
tain a productive viral infection, as well as in malignant
lesions, which most often contain chromosomally inte-
grated viral genomes.4
Human papillomaviruses (HPVs) cause benign and malig-
nant neoplasia of mucosal and cutaneous epithelia.1,2
More than 80 HPV types are known, and they fall into
various distantly related groups. Within each group,
members are closely related to one another, infect simi-
lar target cells, and give rise to similar pathology.3 Most
research has been done on HPV-16 and HPV-18, the
most prevalent viruses found in cervical cancer and its
precursor lesions, and on HPV-6 and HPV-11, which
cause genital warts. HPV genomes contain eight genes,
among them the oncogenes E6 and E7. The oncopro-
teins E6 and E7 of di€erent HPV types are homologous
to one another, respectively, and have similar, but not
completely identical functions. In the case of HPV-16,
The function of these two oncoproteins is required for
focus formation in cell culture,5 establishment of benign
neoplasia,6 progression toward malignancy,7 and main-
tenance of the malignant phenotype.8,9 The E6 onco-
protein alone is sucient to induce tumors in transgenic
mice.10 E6 proteins exert their function by interacting
on the molecular level with a large number of cellular
proteins, although they have only a small size, approxi-
mately 150 amino acids. Presently, protein±protein
interactions have been reported between the E6 protein
and p53, E6AP, E6BP, paxillin, CBP, AP-1, hDLG,
IRF-3, Myc, hMCM7, Bak, and E6TP-1;6,11 22 for
reviews.23,24 A major part of the function of the E7
protein apparently stems from its ability to interact with
the cell cycle repressor Rb.25
*Corresponding authors. Fax: +65-779-1117; e-mail: mcbaet@
imcb.nus.edu.sg (A.E. Ting); tel.: +49-6221-42-3336; fax: +49-6221-42-
3342; e-mail: w.beerheide@DKFZ-Heidelberg.de (W. Beerheide).
yPresent address: C0700, Genetic Alterations in Carcinogenesis,
Deutsches Krebsforschungszentrum DKFZ, Im Neuenheimer Feld
280, D-69120 Heidelberg, Germany.
{Medicinal and Combinatorial Chemistry Laboratory, Institute of
Molecular and Cell Biology, 30 Medical Drive, Singapore 117609.
xLaboratory for Papillomavirus Biology, Institute of Molecular and
Cell Biology, 30 Medical Drive, Singapore 117609.
The comparison of the E6 oncoproteins of di€erent
HPV types has identi®ed numerous conserved amino
acid residues. The most conspicuous motifs are two
Cys-X2-Cys-X29-Cys-X2-Cys sequences (where
represents any amino acid and the number represents
X
0968-0896/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(00)00193-0

2550
W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
the number of residues), which are responsible for the
ability of E6 proteins to bind zinc,26 and are likely to
establish zinc ®ngers with structures speci®c for these
HPV proteins.27 The E7 oncoprotein has a singular
presumed zinc ®nger structure. Mutations of conserved
amino acids, in particular mutation of the cysteine resi-
dues of the Cys-X2-Cys-X29-Cys-X2-Cys motifs, have
most often pleiotropic functional consequences, sug-
gesting that mutations of these amino acid residues lead
to changes of the E6 structure that are incompatible with
many or all of the above mentioned protein±protein
interactions.6,23
Results
Design of organic disul®des with the potential to react with
zinc coordinating cysteine thiol groups of HPV-16 E6
In our previous screen of disul®des, azoics and nitroso
aromatics we identi®ed only one compound, the dithio-
bisamine dithiodimorpholine (C16), with activity in
three established assays and therefore as a potential lead
compound.28 We hypothesized that related dithiobisa-
mine derivatives might have a similar or better reactivity
with the E6 protein. To investigate this, we synthesized
dithiobisamines with the general formula: R1/R2±N±S±
S±N±R3/R4 and other disul®des derived from cyclic
amines, acyclic amines, aryl substituted piperazine, aro-
matic amines, and additional analogues.
These cysteine residues of E6 present themselves as tar-
gets for drugs, since their alteration by biochemical
reactions should negatively a€ect structure and function
in a manner similar to mutations. Papillomavirus pro-
teins are likely to be much better targets for antiviral
and anticancer drugs than the proteins of RNA viruses,
since the genomes of papillomaviruses do not evolve as
rapidly as RNA genomes, and remain completely
unchanged over time spans as long as decades.3 There-
fore it is likely that anti-HPV drugs will be more con-
sistently useful than drugs against RNA viruses.
The synthesis of bisthioamines was carried out accord-
ing to the condition reported by Katritzky et al.,30 with
some slight modi®cations. The secondary amine reacted
readily with disulfur dichloride in petroleum ether to
give the bisthioamines in 10±80% yield after puri®ca-
tion (Scheme 1). In our hands, we found that a low
temperature ( 78 ꢀC) and a very dilute reaction solution
suppress the formation of side products very eciently.
Sometimes, co-solvent such as THF was added to
increase the solubility of the amine in the reaction mix-
ture. Puri®cation and characterization by NMR and
MS of the bisthioamines were carried out.
We have recently described novel screens for anti-HPV-
E6 drugs.28 Our experiments were based on two
assumptions. Firstly, we assumed that compounds that
chemically alter any of the eight cysteine groups, either
by oxidation or by formation of adducts of the thiol
groups, would render E6 incapable of binding zinc and
of maintaining its tertiary structure. Secondly, we
established a hierarchy of tests that measured (1) the
release of zinc ions from E6, (2) the ability of E6 to bind
the cellular proteins E6AP and E6BP, and (3) the pro-
liferation of cell lines whose survival depends on con-
tinuous expression of E6 and E7 from endogenous HPV
genomes. In these tests we examined 36 compounds that
might react with thiol groups, and we identi®ed one
particular substance, 4,40-dithiodimorpholine (C16), as a
powerful and speci®c inhibitor of E6 function. Based on
these ®ndings, we examined 59 new compounds in this
study. Most of these substances were designed and syn-
thesized based on the assumption that the dithiobisa-
mine moiety of 4,40-dithiodimorpholine is an important
prerequisite for its reactivity, as it is generally known
that the bond-disassociation energy of disul®de in the
dithiobisamine is weaker than in the dithioether.29 As a
result, the formation of disul®de with the cysteine group
is probably the driving force for such an exchange to
take place. We further speculated that modi®cation of
the side chains around the N±S±S±N moiety might alter
its reactivity and ability to access the E6 protein. Yet
other substances were designed to test whether structu-
rally analogous substances without the dithiobisamine
moiety might fail to show reactivity with the E6 protein.
We have also designed and synthesized a series of
structural analogues of bisthioamines M50±M55 con-
taining the morpholine moiety of C16 but without the
disul®de or the diamine. They were prepared by reacting
the morpholine with phenyl isothiocyanate, 4-nitrophe-
nyl chloroformate, naphthalene-1-sulfonyl chloride,
benzoyl chloride, 1,10-thiocarbonyl diimidazole and sul-
furyl chloride respectively (Scheme 2). In the case of
M51, 0.5 equivalent of 4-nitrophenyl chloroformate was
used as our initial intention was to make a dimorpholine
urea. However, M51 was obtained because the carba-
mate was too stable to be displaced by another molecule
of morpholine. In the synthesis of M55, the more reac-
tive sulfuryl chloride was preferred over 1,1-sulfonyl
diimidazole as it gave better yield.
Table 1 summarizes all substances examined in this
study and introduces a shorthand nomenclature for this
paper. It shows disul®des derived from (A) cyclic
amines (C16 and M1±M11) and aryl and arylalkyl sub-
stituted piperazines (M12±M26), (B) acyclic amines
Scheme 2. (a) C6H5NCS; (b) 4-NO2C6H4OC(O)Cl; (c) 1-C10H7SO2Cl;
(d) C6H5COCl; (e) C(S)Im2; (f) SO2Cl2.
Scheme 1. (a) S2Cl2/ 78 ꢀC/ petroleum ether.

W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
2551
(M27, M29±M32) and formamide (M28), and (C) for-
mamidine (M33), imine and aromatic heterocycles
(M34±M40), benzyl amines (M41±M43), and anilines
(M44±M46). Table 1 also shows structurally analogous
substances (C) (M47±M59), including thiuram disul®des
(M48±M49) and compounds containing the morpholino
moiety of C16 but without the disul®de (M50±M55),
piperidinomethane (M56) as well as other disul®des
without the bisamine moiety (M57±M59).
activity, while M33, M38, M39 and M42 had moderate
activity with 30±50% zinc release. Among these three,
only M42 had the N±S±S±N moiety. Within the hetero-
geneous group of structural analogues (M47±M59) we
found that the dithiodiphthalimide M47 and the thiuram
disul®des (M48±M49) had the highest values in the TSQ
assay, whereas all the other structural analogues were
inactive (M50±M59).
E€ects of compounds on the protein interaction between
HPV-16 E6 and E6AP and E6BP
Zinc ejection after in vitro treatment of HPV-16 E6
protein with organic disul®des
BIACORE technology allows real time analysis of pro-
tein±protein interactions without the need for labeling the
reaction partners. It is based on the detection of small
changes in the refractive index on the surface of a thin
gold ®lm coated with a dextran matrix on which one of
the reactant proteins is covalently linked.31 Here, we
monitored the interaction of E6 with E6AP and with E6BP
by passing GST-E6 incubated with the chemical com-
pounds over immobilized E6AP or E6BP. The tests are
based on the assumption that zinc-releasing compounds
would lead to a change in the tertiary structure of E6 pro-
tein, which would result in the inability of E6 to form the
heterologous complexes with E6AP and E6BP. In contrast
to the TSQ assays, these reactions take place in the pre-
sence of glutathione and resemble the in vivo situation,
where glutathione may undergo disul®de exchange reac-
tions and compete as a substrate with the E6 protein.28
The TSQ assay measures the release of zinc after topolo-
gical or chemical change of the E6 protein. Our assay
conditions require the absence of glutathione, whose
reducing power can antagonize the reactivity of the com-
pound being tested.28 Zinc release is expressed as a per-
centage of the total zinc available in the E6 protein that
is determined by oxidizing all thiol groups with H2O2, a
strong oxidizing agent. Table 1 shows the outcome of
this experiment. Only two compounds, the dithiodiph-
thalimide M47 and the thiuram disul®de M49, were as
active as H2O2 in releasing zinc, and another nine sub-
stances in addition to the previously characterized 4,40-
dithiodimorpholine (C16) released between 50 and 100%
of the zinc compared with the H2O2 value and were
classi®ed as compounds of high TSQ-activity. Ten sub-
stances released between 30 and 50% of the zinc and were
classi®ed as compounds with moderate TSQ-activity, 21
substances released between 10 and 30% of the zinc and
were de®ned as compounds with little TSQ-activity, while
17 substances released less than 10% and can be con-
sidered inactive, as these values did not di€er much from
the TSQ ¯uorescence obtained with the solvent DMSO
alone (6.1%).
Of the 60 compounds examined for their ability to
inhibit E6 binding to E6BP or E6AP, 23 compounds
were observed to reduce the E6-E6BP and E6-E6AP
interaction by at least 50% (Figure 1, Table 1). The
majority of these compounds belonged to the group of
cyclic amines (11 out of 12 compounds, C16 and M1±
M10) that also scored positive (>30% zinc release) in
the TSQ assay, as well as from the aryl and arylalkyl
piperazines (6 out of 10, M12, M13, M15±M17 and
M19). Consistent with the outcome of the TSQ assay,
most of the compounds from groups B and C were not
active in the BIACORE assay except the acyclic amines
M27 and M30, formamidine M28, ethyl N-benzylgly-
cine ester disul®de M41, dithiodiphthalimide M47 and
thiuram disul®de M48. However, the other thiuram
disul®de M49 was not active in the BIACORE assays,
even though it had highest activity in the TSQ assay.
Most of the compounds tested were colorless but we
noted that some compounds with yellow color at 1 mM
stock solutions such as M18, M19 and M44±M46 had
TSQ values of 0%. When these compounds were exam-
ined for TSQ-activity in the presence of 1 mM ZnCl2,
reduced ¯uorescence was observed when compared to
DMSO only (data not shown), suggesting that these
compounds interfered with the TSQ-¯uorescence.
Disul®des derived from cyclic amines (C16 and M1±
M11) were highly (C16, M6 and M8±M10) or moder-
ately active (M1±M4 and M7) in the TSQ assay. Di-
sul®des derived from aryl and arylalkyl substituted
piperazines (M12±M26) were mostly inactive or had lit-
tle activity, except for three compounds, M12 and M13
with high activity and M14 with moderate activities.
Among them, the halogen-substituted aryl and arylalkyl
piperazine compounds (M23±M26) had particularly low
TSQ values (less than 12%).
E€ects of the compounds on cell viability
In addition to studying the e€ects of these 60 com-
pounds in in vitro assays, we also examined their e€ects
on cell viability (WST1 assay). Previously, we observed
that C16 was able to induce cell death (apoptosis) spe-
ci®cally in HPV-infected cell lines apparently due to the
increased stability of p53 after inactivation of the E6
protein.28 Similarly, here we compared the e€ects of our
new compounds on cell viability of HPV-positive cells
(SiHa, HeLa, and CaSki) and HPV-negative cells
(HaCat, MCF7 and HepG2). At a concentration of
50 mM, 17 of the compounds (M5, M6, M8, M9, M20,
M24, M27, M28, M34, M38, M41, M42, M44, M45 and
M47±M49) were found to be unspeci®cally cytotoxic
Within the group of acyclic amines and formamide
(M27±M32) only two disul®des (M27±M28) showed
high TSQ-activities whereas all the other compounds in
this group had little activity. Similarly, 10 out of 14
disul®des derived from formamidine, imine, aromatic
amines and heterocycles (M33±M46) had no or little

2552
W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
Table 1. Evaluation of disul®des and other compounds by in vitro and cell-based assays
Compoundsa
TSQb
BIAc
WSTd
WST-Spece
Groupf
A: Disul®des derived from cyclic amines
C16
M1
M2
M3
M4
M5
M6
M7
M8
M9
M10
63
40
42
45
37
28
71
31
55
51
54
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
1
1
1
1
1
4
3
3
3
3
1
+
+
+
+
M11
M12
M13
19
51
67
5
1
1
+
+
+
+
+
+
M14
41
+
+
2
M15
M16
24
24
+
+
+
+
+
+
2
2
M17
M18
M19
M20
M21
M22
12
0g
0g
24
19
22
+
+
+
+
+
+
2
4
4
5
5
5
+
(continued on next page)

W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
2553
Table 1 (continued)
Compoundsa
TSQb
BIAc
WSTd
WST-Spece
Groupf
M23
7
5
M24
M25
12
11
+
5
5
M26
9
5
B: Disul®des derived from acyclic amines and formamide
M27
57
55
+
+
+
+
3
3
M28
M29
M30
M31
12
11
28
5
4
5
+
M32
15
5
C: Disul®des derived from aromatic amines, heterocycles and other analogues
M33
M34
M35
M36
33
8
4
5
5
5
+
3
18
M37
M38
0
5
4
48
+
(continued on next page)

2554
W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
Table 1 (continued)
Compoundsa
TSQb
BIAc
WSTd
WST-Spece
Groupf
M39
M40
36
18
4
5
M41
22
+
+
+
4
M42
M43
38
11
4
5
M44
M45
0g
+
+
5
5
0g
M46
M47
0g
5
3
103
+
+
+
M48
M49
M50
81
106
10
++
+
3
4
5
M51
M52
M53
M54
M55
6
10
7
5
5
5
5
5
8
9
(continued on next page)

W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
2555
Table 1 (continued)
Compoundsa
TSQb
BIAc
WSTd
ND
WST-Spece
ND
Groupf
M56
M57
3
1
5
5
M58
M59
2
5
5
13
aAll the compounds are denoted by the letter M and a corresponding number, except dithiodimorpholine (C16).
bThe value represents the % of zinc released compared to release by 0.3% H2O2. The concentration of compound used was 13 mM. The values
represent 2±5 experiments done in duplicate with standard deviations of <5%.
cThis assay measures the compound's ability to interfere in the binding of E6 with the cellular coactivators E6BP and E6AP. A positive score indi-
cates those compounds having values less than 50% of the binding observed in the presence of DMSO. See Figure 1 for values obtained.
dA positive score indicates that the compound is cytotoxic at 50 mM for at least one of the following cell lines: HeLa, SiHa, CaSki, HaCat, MCF7
and HepG2.
eA positive score indicates that the compound at 50 mM is cytotoxic for HPV-infected cells, HeLa, SiHa and CaSki, compared to the HPV-negative
cell lines, HaCat, MCF7 and HepG2. See Figure 2 for details.
fThe compounds have been categorized into ®ve di€erent groups: Group 1 includes those compounds with TSQ values greater than 30% that were
positive in the BIACORE assay and speci®c in the WST1 assay. Group 2 consists of the compounds that were speci®c in the WST1 assay and either
positive in the BIACORE assay or with TSQ values greater than 30%. Group 3 contains compounds with TSQ values greater than 30% and positive
in the BIACORE assay, but were non-speci®cally cytotoxic. Group 4 is for those compounds that demonstrated activity in only one of the assays.
Compounds in Group 5 were inactive in all three assay systems.
gThese compounds were observed to interfere with TSQ activity in the presence of 1 mM ZnCl2.
against the cells examined (Table 1 and Figure 2). 12
compounds in addition to C16 speci®cally a€ected cell
viability in HPV-infected cells (Table 1, Figure 2). These
12 HPV-speci®c compounds consisted only of cyclic
amines (M1±M4 and M10), and aryl and arylalkyl sub-
stituted piperazines (M12±M18). These compounds
were also active in the TSQ and/or BIACORE assays,
except M18. No cytotoxicity was observed with the
remaining classes of compounds. Surprisingly, we noted
that many of the compounds that speci®cally inhibited
the growth of HPV containing cell lines had an opposite
stimulating e€ect on the growth of the control HepG2
cells (Fig. 2). The 30 compounds a€ecting speci®cally or
unspeci®cally the cell viability were further examined at
lower concentrations. Most compounds were found to
be inactive at 1±10 mM, except the thiuram disul®de
M48, which was still unspeci®cally toxic at 1 mM and
inactive at 0.1 mM (data not shown).
CORE assay (M15±M17) or with a TSQ value greater
than 30% (M14). Group 3 contains compounds (M6,
M7, M8, M9, M27, M28, M47 and M48) that were
positive in BIACORE assay with TSQ values greater
than 30%, but were not speci®c in the WST1 assay.
Group 4 is for those compounds that demonstrated
activity in only one of the assays. These include those
compounds with TSQ values greater than 30% (M33,
M38, M39, M42 and M49), positive in the BIACORE
assay (M5, M19, M30, M41) or speci®c in the WST1
assay (M18). Group 5 includes the compounds that
were inactive in all three assay systems.
Discussion
In an attempt to identify therapeutically useful new
compounds for the treatment of human papillomavirus
associated cervical cancer we previously established a
hierarchy of assay systems that target the HPV E6
oncoprotein.28 The initial screen of 36 substances included
a large group of disul®de based organic compounds.
These substances were chosen under the assumption
that disul®de exchange reactions would either lead to
adducts between the organic compound and the cysteine
residues in E6, or to oxidation of two or more cysteine
residues and disul®de bond formation within E6.32 In
both cases, the metal chelating property of the cysteine
residues would be disrupted, the zinc ion ejected, and
the integrity of the zinc ®nger disturbed. Among the
disul®des that we examined, 4,40-dithiodimorpholine
was the only substance that scored positive in all three
Grouping of compounds based on their reactivity in the
TSQ, BIACORE and WST1 assays
Based on the results from the three di€erent assays, we
have categorized the compounds tested into ®ve groups
(Table 1). Group 1 includes compounds that were spe-
ci®c in the WST1 assay (WST1-spec) and positive in the
BIACORE assay with TSQ values greater than 30%
(C16, M1±M4, M10, M12 and M13). These groups
represent the most promising compounds for further
evaluation.
Group 2 consists of the compounds that were speci®c in
the WST1 assay and were positive in either the BIA-

2556
W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
Figure 1. Interaction of HPV-16 E6 with E6AP measured with BIACORE technology. The interactions of E6 with E6AP were monitored using a
BIACORE 2000. GST-E6AP was immobilized on a CM-5 sensor chip. HPV16 GST-E6 was incubated with 400 mM compound for 2 h and subse-
quently passed over the immobilized ligand in sequential ¯ow mode. The amount of E6 bound to E6BP or E6AP was measured by the increase in
resonance units after E6 was passed over the ligand. The values represent the % binding compared to the value obtained in the presence of DMSO
only after subtracting the value obtained with GST alone (which represents non-speci®c binding). Similar results were obtained when E6BP was
examined (data not shown).
tests. Furthermore, dithiodimorpholine speci®cally
induced p53 protein levels, associated with the induction
of apoptosis, and thereby inhibited speci®cally the
growth of HPV-containing cell lines.28 All these data are
in agreement with established functions of E6.
were derivatives of monocyclic amines with the dithio-
bisamine moiety (C16, M1±M4 and M10) or were pyr-
idyl and pyrimidinyl piperazines with the dithiobisamine
moiety (M12 and M13). Dithiobisamines with acyclic
amines, aromatic amines or other structural analogues
were partially active in some of the assays, especially in
releasing zinc from E6 (TSQ), but lacked in general
activity in the BIACORE assay as well as the ability to
speci®cally inhibit the growth of HPV containing cell
lines. Many of these compounds were inactive in the
BIACORE as well as in cell culture, which might indi-
cate their inactivation under GSH conditions.28 This
possibility is excluded for compounds that were active
in the BIACORE assay and unspeci®cally toxic in cell
culture (M5, M6, M8, M9, M27, M28, M41, M47 and
M48) or only unspeci®cally toxic in cell culture (M20,
M24, M34, M38, M42, M44, M45 and M49). Probably
these unspeci®c toxic compounds a€ect cellular zinc
binding proteins that are crucial for cell survival. There
are three compounds (M7, M19 and M30) that are only
active in the BIACORE, but lack toxicity in cell culture.
Potentially the penetration of these compounds into
cells is inecient or another possibility could be the
intracellular stability of these compounds being a€ected
by additional factors than GSH.
Together with its ability (chemical reactivity) or access
(bulkiness) to react with the E6-zinc-binding domain in
vitro one additional crucial requirement in our initial
screen, that led to the identi®cation of dithiodimorpholine,
was the resistance of compounds in vitro (BIACORE
assay) and in cell culture (WST1 assay) towards GSH
mediated inactivation.28 Based on this observation, we
hypothesized that the dithiobisamine moiety of this
substance may render this disul®de bond speci®cally
reactive both in the TSQ and BIACORE assays and in
cultured cells and that related dithiobisamine derivatives
may have a similar or better reactivity with the E6 protein.
We have now examined 59 structurally related com-
pounds, most of them which were newly synthesized, in
order to de®ne the structural requirements necessary for
their inhibition of HPV E6 activity. Based on their
reactivity pro®le, we identi®ed 21 compounds with
activity >30% zinc release in the TSQ assay, of which
15 were also active in the BIACORE assay and 7 of
these compounds in addition to dithiodimorpholine
(C16) also speci®cally inhibited the growth of HPV
containing cell lines (Table 1, Group 1 compounds).
Those compounds that scored positive in all three assays
The mechanism of reaction between the disul®de and
the zinc-®nger is not fully understood. There are a few
possibilities. Firstly, it might be an electrophilic attack
on the sulfur atoms of the zinc-coordinated cysteines of

W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
2557
E6 by the disul®de.33 Secondly, the disul®de bond of
dithiobisamine might cleave to generate an R1R2NS
radical which reacts further with the sulfur atoms of the
zinc ®nger resulting in mixed disul®des. Thirdly, it
might also be a combination of both the electrophilic
and radical reactions. Our ®nding that compounds
M50±M56 were inactive was not at all surprising, since
these compounds are not capable of undergoing elec-
trophilic or radical reactions.
Based on our results, we have found that the dithiobis-
amine moiety N±S±S±N is crucial but not sucient for
activity and speci®city. While dithiodimorpholine (C16)
was active, structural analogues with the morpholino
ring such as thiourea (M50, M54), carbamate (M51)
sulfonamide (M52), amide (M53), and sulfone (M55)
derivatives were inactive. Similarly, while dithiodipiper-
idine (M10) was active, structural analogues that do not
contain the disul®de or amine such as piperidinomethane
(M56), cyclohexyldisul®de (M57), trans-1,2-dithiane-4,5-
diol (M58) and 3H-1,2-benzodithiol-3-one, 1,1-dioxide
(M59) were also found to be inactive. Finally, the most
active compounds (Group 1) are derived from a six-
membered saturated cyclic amine. While the Group 1
compounds di€ered in their extent of zinc release (37±
67%, Table 1) and inhibition of the complex formation
between E6 and E6AP or E6BP (Fig. 1), they were
similar in activity at 50 mM with respect to the speci®c
inhibition of HPV-infected cells (Fig. 2).
In addition to dithiobisamines we have also examined
thiuram disul®des (M48 and M49) with a cyclic amine
ring including the morpholino ring of C16 in M48.
These compounds showed highest reactivity in the TSQ
assay, in contrast to our previous study with tetra-
ethylthiuram disul®de (Disul®ram), which had only
moderate activity in the TSQ assay.28 Thiuram disul-
®des containing ring structures such as cyclopenta-
methylene rings were also reported to be highly reactive
with cysteine groups of the human immunode®ciency
virus 1 (HIV-1) p7 zinc ®nger.34 In this study, the com-
pounds were inactive in in vivo experiments determining
the replication of HIV-1. In our studies, M48 was found
to be positive in the BIACORE assay and cytotoxic to
all the cell lines tested even at 1 mM, suggesting that the
compound is highly resistant to intracellular inactiva-
tion (i.e., by GSH) compared to the compounds identi-
®ed in Group 1. Because thiuram disul®des and the
metabolite dithiocarbamates have been implicated in a
variety of therapeutic applications including chemo-
prevention,32,35 37 in¯ammation,38 anticancer enhan-
cers,39 inhibition of HIV-progression,40 anti-hepatotoxic
activity41 and inhibition of cisplatinum nephrotoxi-
city,42 derivatives of M48 should be of interest for fur-
ther evaluation.
Figure 2. Viability assays of HPV-positive (HeLa, SiHa) and HPV-
negative (HaCat, MCF7 and HepG2) cell lines incubated with dithio-
bisamine compounds. Indicated cell lines were incubated with 50 mM
of dithiobisamine compounds containing cyclic amine moieties (a)
C16, M1±M11 or (b) aryl substituted piperazine moieties M12±M18
for two times in three days as described.28 Cell viability was deter-
mined through absorption readings of WST1, a tetrazolium salt that
measured the activity of mitochondrial dehydogenase (WST1 assay).
Several of the compounds in (a) C16, M1±M4 and M10 and (b) M12±
M18 demonstrated substantial and speci®c inhibition of cell viability
in HPV-positive cervical epithelial tumor cell lines (HeLa and SiHa),
but had little or no e€ect on HPV-negative cell lines, immortalized
skin epithelial cells (HaCat) and mammary and liver epithelial cancer
cells (MCF7 and HepG2 respectively), except M13 which did not
a€ect MCF7 and HepG2 only. Other dithiobisamine compounds with
cyclic amine moieties such as M5 and M9 demonstrated unspeci®c
cytotoxicity towards HPV-positive and HPV-negative cell lines or did
not inhibit cell proliferation of any cell line examined as observed with
M11. CaSki, another HPV-positive cell line, demonstrated similar
pro®les of cell viability observed with the HeLa and SiHa cells after
treatment with compounds (data not shown). All values were normal-
ized to the values obtained in the presence of dimethyl sulfoxide
(DMSO) which was set as 100%. Values represent 2±8 independent
experiments with at least duplicated measurements, and standard
deviations of <10%.
Human papillomaviruses are associated with more than
95% of cervical cancers. Cervical cancer is the second-
most common cancer among women worldwide and the
leading cancer among women in Africa, Asia, and
South America with over 300,000 deaths per year. The
current treatment for advanced cervical cancer is sur-
gery, followed in some cases by radiation treatment.

2558
W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
Recently, it has been reported that the addition of the
drug cisplatin with radiation treatment reduced the risk
of death by 30 to 50% in the ®rst few years after diag-
nosis of cervical cancer.43 45 While these results are quite
dramatic, cisplatin is a non-speci®c drug with known side
e€ects42 and induces p53 in both HPV-positive and
HPV-negative cells.46,47 In contrast, dithiodimorpholine
(C16) strongly induced p53 only in HPV-positive cells,28
suggesting that dithiobisamine-based compounds may
provide greater speci®city than cisplatin in the treatment
of cervical cancer. Furthermore, now we de®ne a spe-
ci®c class of dithiobisamine compounds including cyclic
amines and aryl substituted piperazines that inactivate
the HPV-16 E6 protein and induce cell death speci®cally
in HPV-infected cells. Hopefully, further evaluation of
these compounds and their derivatives will allow for a
potent therapeutic agent against HPV related diseases.
8H), 3.10 (br t, J=4.9 Hz, 8H). 13C NMR (CDCl3) d: 28.8,
58.2. HRMS calcd for C8H16N2S4 (M+) 268.01962,
found 268.01968.
Di-4-methylpiperazinyl disul®de (M3). Yellow solid.
Yield: 22.4 mg (68%). 1H NMR (CDCl3) d: 2.25 (s, 6H),
2.45 (br s, 8H), 2.82±2.84 (m, 8H). 13C NMR (CDCl3) d:
45.78, 45.82, 55.57. HRMS calcd for C10H22N4S2 (M+)
262.12860, found 262.12728.
Di-4-(1-ethoxycarbonyl)piperazinyl disul®de (M4). White
1
solid. Yield: 15.4 mg (33%). H NMR (CDCl3) d: 1.27
(t, J=7.1 Hz, 6H), 2.78 (br s, 8H), 3.53 (br s, 8H), 4.14
(q, J=7.1Hz, 4H). 13C NMR (CDCl3) d: 14.6, 44.1, 55.5,
61.6, 155.3. HRMS calcd for C14H26N4O4S2 (M+)
378.13956, found 378.13895.
Dipiperidyl disul®de (M10). Yellow solid. Yield: 18.7 mg
(65%). The 1H NMR spectrum of M10 was slightly
di€ered from those reported by Diaz. 1H NMR (CDCl3)
d: 1.36±1.43 (m, 4H), 1.63±1.70 (m, 8H), 2.78 (br t,
J=5.3 Hz, 8H). 13C NMR (CDCl3) d: 22.8, 27.0, 57.4.
HRMS calcd for C10H20N2S2 (M+) 232.10680, found
232.10680.
Experimental
Chemicals and reagents
All reagents were from commercial sources unless
otherwise mentioned. The compounds M6, M7, M28
and M34 were obtained from the collection of the
National Cancer Institute, Bethesda, Maryland. C16
was purchased from Tee Hai Chemicals. Compounds
M33, M35±M38, M40, M47, M48 and M56±M59 were
obtained from Aldrich. M39 and M49 were from Sigma
and Acros Organics, respectively.
Di-4-(2-pyridyl)piperazinyl disul®de (M12). Yellow syrup.
1
Yield: 27.9 mg (58%). H NMR (CDCl3) d: 2.92±2.96
(m, 8H), 3.60±3.61 (m, 8H), 6.63±6.66 (m, 4H), 7.47±
7.51 (m, 2H), 8.18±8.19 (m, 2H). 13C NMR (CDCl3) d:
45.8, 55.4, 107.1, 113.6, 137.5, 147.9, 159.0. HRMS calcd
for C18H24N6S2 (M+) 388.15039, found 388.15038.
Di-4-(pyrimidin-2-yl)piperazinyl disul®de (M13). White
1
Dithiobisamines
solid. Yield: 25.1 mg (51%). H NMR (CDCl3) d: 2.87
The synthesis of all other dithiobisamines not men-
tioned above was according to Katritzky et al.,30 with
modi®cations as described. A solution of the secondary
amine (0.5 mmol) in petroleum ether (40 mL) was pre-
cooled to 78 ꢀC before disulfur dichloride (10 mL,
0.125 mmol) was added. The solution was vigorously
stirred for 15 min at 78 ꢀC and another 30 min at room
temperature. Water (20 mL) was added and the desired
compound was extracted into the organic phase using
diethyl ether (3Â10 mL). The combined organic layers
were dried over magnesium sulfate, ®ltered and con-
centrated in vacuo. The compound was puri®ed on pre-
parative TLC plate. High-resolution mass spectra were
obtained with a VG7035. 1H and 13C NMR spectra
were recorded with a Bruker DMX 400 and 100 MHz
respectively. Only data for Group 1 compounds are
reported below. The preparative TLC solvent system
used for M1 and M2 was 12.5% ethyl acetate in hexane,
for M4, M10 and M12 was 25% ethyl acetate in hexane
and for M3 was 9% methanol in chloroform.
(br t, J=4.5 Hz, 8H), 3.88 (br t, J=4.8 Hz, 8H), 6.50 (t,
J=4.7 Hz, 2H), 8.30 (d, J=4.7 Hz, 4H). 13C NMR
(CDCl3) d: 44.1, 55.5, 110.1, 157.7, 161.3. HRMS calcd
for C16H22N8S2 (M+) 390.14090, found 390.14205.
(Cyclohexylamino)morpholin-4-ylmethane-1-thione (M50).
To a solution of morpholine (87 mL, 1 mmol) in anhy-
drous CH2Cl2 was added cyclohexyl isothiocyanate
(142 mL, 1 mmol) and Et3N (70 mL, 0.5 mmol). The
reaction mixture was stirred for 3 h. Aminomethylated
resin (20 mg, 1.46 mmol/g Novabiochem) was added
and the mixture was stirred for another 2 h. The mixture
was ®ltered and then quenched with water. The com-
pound was extracted with CH2Cl2, dried over MgSO4
and ®ltered. The ®ltrate was concentrated and puri®ed
on preparative TLC plate (50% ethyl acetate in hexane)
to give M50 as yellow crystals (127 mg, 56%).
1H NMR (CDCl3) d: 1.13 (dd, J=3.5 Hz, 12.0 Hz, 1H),
1.19 (dd, J=3.5 Hz, 7.4 Hz, 1H), 1.43 (qt, J=3.2 Hz,
12.3 Hz, 2H), 1.65 (dt, J=3.5 Hz, 12.1 Hz, 2H), 1.71 (dt,
J=3.6 Hz, 13.5 Hz, 2H), 2.11 (br dd, J=3.2 Hz, 6.1 Hz,
2H), 3.72±3.75 (m, 4H), 3.77±3.79 (m, 4H), 4.34±4.36
(m, 1H), 5.37 (d, J=7.0 Hz, 1H). 13C NMR (CDCl3) d:
24.9, 25.5, 33.0, 47.2, 54.3, 66.1, 171.5. HRMS calcd for
C11H20N2OS (M+) 228.12964, found 228.12958.
Di-2,6-dimethylmorpholin-4-yl disul®de (M1). Yellow
solid. Yield: 13.5 mg (37%). M1 was a mixture of dia-
stereoisomers. 1H and 13C NMR were identical to those
reported.30 HRMS calcd for C12H24N2O2S2 (M+)
292.12793, found 292.12886.
Dithiomorholin-4-yl disul®de (M2). Yellow solid. Yield:
12.3 mg (37%). 1H NMR (CDCl3) d: 2.73 (br t, J=4.9Hz,
4-Nitrophenyl morpholine-4-carboxylate (M51). To an
ice-cooled solution of morpholine (87 mL, 1 mmol) in

W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
2559
anhydrous CH2Cl2 was added 4-nitrophenyl chloro-
formate (100 mg, 0.5 mmol) and Et3N (140 mL, 1 mmol).
The mixture was allowed to warm up to room tem-
perature and stirred for 3 h. The reaction was quenched
with water and extracted with CH2Cl2. The organic
layer was combined, dried over MgSO4 and ®ltered. The
®ltrate was concentrated and puri®ed on preparative
TLC plate (50% ethyl acetate in hexane) to yield M51
as white crystal (73 mg, 58%).
room temperature. The reaction mixture was stirred for
1 h at room temperature and puri®ed using the same
protocol for M51 to yield M55 as pale yellow powder
(26 mg, 22%).
1H NMR (CDCl3) d: 3.16 (br s, 4H), 3.87 (br t,
J=4.8 Hz, 4H). 13C NMR (CDCl3) d: 47.6, 65.2. HRMS
calcd for C8H16N2O4S (M+) 236.08308, found
236.08454.
1H NMR (CDCl3) d: 3.59 (br t, J=4.4Hz, 2H), 3.69 (br t,
J=4.0 Hz, 2H), 3.76 (t, J=4.6 Hz, 4H), 7.30 (d,
J=15.3 Hz, 2H), 8.26 (d, J=15.3 Hz, 2H). 13C NMR
(CDCl3) d: 44.2, 44.9, 66.3, 66.5, 122.2, 125.1, 145.0,
152.3, 156.0. HRMS calcd for C11H12N2O5 (M+)
252.07462, found 252.07570.
Expression of E6, E6AP and E6BP as GST-fusion pro-
teins. E6, E6AP (amino acids 391±408) and E6BP-Glu-
tathione S-transferase (GST) fusion proteins were
prepared as described.28 For direct use of GST-fusion
proteins bound on glutathione±Sepharose beads in the
zinc-release assay, the glutathione±Sepharose beads
were resuspended in PBS, Tris, pH 8.2. GST-fusion
proteins for BIACORE analysis were eluted with elu-
tion-bu€er (10 mM GSH, 50 mM Tris, PBS, pH 8.2).
4-(Naphthylsulfonyl)morpholine (M52). To an ice-cooled
solution of morpholine (87 mL, 1 mmol) in anhydrous
CH2Cl2 was added naphthalene-1-sulfonyl chloride
(226mg, 1 mmol) and Et3N (140 mL, 1 mmol). The reac-
tion was stirred for 3 h at room temperature. The same
work up protocol for M50 was used. M52 was isolated
as white powder (138 mg, 50%).
Zinc release assays. Release of zinc from HPV-16-E6
was monitored by the change in ¯uorescence of the zinc-
selective ¯uorophore TSQ ((N-6-methoxy-8-quinolyl)-p-
toluenesulfonamide, Molecular Probes) by published
procedures48,49 as modi®ed by us.28 In brief, 9 mg (1 mM)
recombinant GST-E6 protein, bound to glutathione±
Sepharose beads, was incubated with 13 mM compound
or 0.3% H2O2 in TSQ-assay bu€er (10 mM sodium
phosphate bu€er, pH 7.0, 10% glycerol) for 2 h at room
temperature (200 ml total volume in 96-well plates).
Immediately after addition of TSQ to a concentration of
100 mM, the increase in ¯uorescence was measured with
a Tecan platereader (355 nm excitation ®lter and 460 nm
emission ®lter).
1H NMR (CDCl3) d: 3.16 (br t, J=4.7 Hz, 4H), 3.69 (br
t, J=4.7 Hz, 4H), 7.55±7.68 (m, 3H), 7.94 (dd, J=1.1 Hz,
7.5 Hz, 1H), 8.10 (d, J=8.2 Hz, 1H), 8.21 (dd,
J=1.1 Hz, 7.4 Hz, 1H), 8.77 (d, J=8.7 Hz, 1H). 13C
NMR (CDCl3) d: 45.6, 66.2, 124.1, 125.1, 126.9, 128.2,
128.9, 129.0, 130.8, 131.8, 134.3, 134.7. HRMS calcd for
C14H15NO3S (M+) 277.07727, found 277.07636.
Morpholin-4-yl-phenyl ketone (M53). An ice-cooled
solution of morpholine (87 mL, 1 mmol), benzoyl chlo-
ride (116 mL, 1 mmol) and Et3N (140 mL, 1 mmol) in
anhydrous CH2Cl2 was stirred for 10 min before warm-
ing up to room temperature. Stirring was continued for
another 3 h at room temperature and puri®cation was
carried out following the same protocol for M50 to
yield M53 as translucent syrup (86 mg, 45%).
BIACORE assays. Binding of GST-E6 to GST-E6BP,
GST-E6AP and GST was monitored by surface plas-
mon resonance (SPR) on a BIACORE 2000 (BIACORE
AB) as described.28 Brie¯y, puri®ed ligand (GST, GST-
E6AP and GST-E6BP) was covalently amine coupled to
a CM-5 sensor chip. Typically 6000±10,000 RU of GST,
E6BP and E6AP were immobilized on three di€erent
¯owcells. Aliquots of puri®ed HPV-16 GST-E6 were
incubated with either a 400 mM solution of the examined
compound or 0.6% (170 mM) H2O2 for 2 h at room tem-
perature, and injected at 1 ml/min over the immobilized
ligands. Interactions between GST-E6 and the ligands
were monitored by the change of resonance signal in arbi-
trary units (RU). In between each sample, the surfaces
were regenerated with a 1-minute pulse of 50 mM NaOH.
1H NMR (CDCl3) d: 3.45 (br s, 2H), 3.54 (br s, 2H),
3.78 (br s, 4H), 7.40±7.45 (m, 5H). 13C NMR (CDCl3) d:
66.9, 127.1, 128.5, 129.9, 135.2, 170.2. HRMS calcd for
C11H13NO2 (M+) 191.09464, found 191.09745.
Di-morpholin-4-ylmethane-1-thione (M54). A mixture of
morpholine (87 mL, 1 mmol), 1,10-thiocarbonyl diimida-
zole (87 mg, 0.5 mmol) and Et3N (70 mL, 0.5 mmol) in
anhydrous CH2Cl2 was stirred at room temperature for
3 h. The same work up protocol as for M51 was used.
M54 was isolated as pale yellow powder (67 mg, 62%).
Determination of cell viability (WST1 assay). For this
study we used three cervical cancer cell lines: HeLa,
SiHa, and CaSki, that contain HPV genomes. Additional
cell lines studied include: HaCat, an immortalized skin
epithelial cell line, MCF7, a mammary epithelial tumor
cell line, and HepG2, a liver epithelial tumor cell line.
Cell viability was determined as previously described.28
1H NMR (CDCl3) d: 3.61 (br t, J=4.8 Hz, 8H), 3.74 (br t
J=4.8 Hz, 8H). 13C NMR (CDCl3) d: 52.0, 66.4, 174.8.
HRMS calcd for C9H16N2O2S (M+) 216.09325, found
216.09590.
Di-morpholin-4-yl sulfone (M55). An ice-cooled mixture
of morpholine (87 mL, 1 mmol), sulfuryl chloride (40 mL,
0.5 mmol) and Et3N (140 mL, 1 mmol) in anhydrous
CH2Cl2 was stirred for 10 min before warming up to
Acknowledgements
We thank P. M. Howley and E. J. Androphy for kindly
providing the E6AP and E6BP expression clones,

2560
W. Beerheide et al. / Bioorg. Med. Chem. 8 (2000) 2549±2560
respectively. We would like to acknowledge W. G. Rice
for providing the compounds from the NCI chemical
repository.
25. Dyson, N.; Howley, P. M.; Munger, K.; Harlow, E.
Science 1989, 243, 934.
26. Grossman, S. R.; Laimins, L. A. Oncogene 1989, 4, 1089.
27. Ullman, C. G.; Haris, P. I.; Galloway, D. A.; Emery, V.
C.; Perkins, S. J. Biochem J. 1996, 319, 229.
28. Beerheide, W.; Bernard, H. U.; Tan, Y. J.; Ganesan, A.;
Rice, W. G.; Ting, A. E. J. Natl. Cancer Inst. 1999, 91, 1211.
29. Diaz, C. V.; Arancibia, A. Phosphorus, Sulfur, and Silicon
1989, 44, 1.
30. Katritzky, A. R.; Zhao, X.; Hitchings, G. J. SYNLETT
1990, 8, 473.
31. Malmqvist, M. Curr Opin. Immunol. 1993, 5, 282.
32. Rothenberg, M. L.; Ostchega, Y.; Steinberg, S. M.;
Young, R. C.; Hummel, S.; Ozols, R. F. J. Natl. Cancer Inst.
1988, 80, 1488.
33. Hathout, Y.; Fabris, D.; Han, M. S.; Sowder, R. C.; Hen-
derson, L. E.; Fenselau, C. Drug Metab. Dispos. 1996, 24, 1395.
34. Rice, W. G.; Turpin, J. A.; Schae€er, C. A.; Graham, L.;
Clanton, D.; Buckheit, R. W. J.; Zaharevitz, D.; Summers, M.
F.; Wallqvist, A.; Covell, D. G. J. Med. Chem. 1996, 39, 3606.
35. Perchellet, J. P.; Abney, N. L.; Thomas, R. M.; Perchellet,
E. M.; Maatta, E. A. Cancer Res. 1987, 47, 6302.
36. Frank, N.; Christmann, A.; Frei, E. Toxicology 1995, 95,
113.
References
1. zur Hausen, H.; de Villiers, E. M. Annu. Rev. Microbiol.
1994, 48, 427.
2. International Agency for Research on Cancer (IARC).
Human papillomaviruses. Volume 64: IARC Monographs on
the evaluation of carcinogenic risks to humans, IARC: Lyon,
France, 1995.
3. Chan, S. Y.; Delius, H.; Halpern, A. L.; Bernard, H. U. J.
Virol. 1995, 69, 3074.
4. Stoler, M. H.; Rhodes, C. R.; Whitbeck, A.; Wolinsky, S.
M.; Chow, L. T.; Broker, T. R. Hum. Pathol. 1992, 23, 117.
5. Barbosa, M. S.; Schlegel, R. Oncogene 1989, 4, 1529.
6. Liu, Y.; Chen, J. J.; Gao, Q.; Dalal, S.; Hong, Y.; Mansur,
C. P.; Band, V.; Androphy, E. J. J. Virol. 1999, 73, 7297.
7. White, A. E.; Livanos, E. M.; Tlsty, T. D. Genes Dev. 1994,
8, 666.
8. von Knebel, D.; Rittmuller, C.; zur Hausen, H.; Durst, M.
Int. J Cancer 1992, 51, 831.
9. Tan, T. M.; Ting, R. C. Cancer Res. 1995, 55, 4599.
10. Song, S.; Pitot, H. C.; Lambert, P. F. J. Virol. 1999, 73,
5887.
11. Chen, J. J.; Reid, C. E.; Band, V.; Androphy, E. J. Science
1995, 269, 529.
12. Crook, T.; Tidy, J. A.; Vousden, K. H. Cell 1991, 67, 547.
13. Gao, Q.; Srinivasan, S.; Boyer, S. N.; Wazer, D. E.; Band,
V. Mol. Cell Biol. 1999, 19, 733.
37. Liu, G. Y.; Frank, N.; Bartsch, H.; Lin, J. K. Mol. Carci-
nog. 1998, 22, 235.
38. Kovacich, J. C.; Boyle, E. M. J.; Morgan, E. N.; Canty, T.
G. J.; Farr, A. L.; Caps, M. T.; Frank, N.; Pohlman, T. H.;
Verrier, E. D. J. Thorac. Cardiovasc. Surg. 1999, 118, 154.
39. Osswald, H.; Frank, N. J. Cancer Res. Clin. Oncol. 1990,
116, 448.
40. Reisinger, E. C.; Kern, P.; Ernst, M.; Bock, P.; Flad, H.
D.; Dietrich, M. Lancet 1990, 335, 679.
14. Gross-Mesilaty, S.; Reinstein, E.; Bercovich, B.; Tobias,
K. E.; Schwartz, A. L.; Kahana, C.; Ciechanover, A. Proc.
Natl. Acad. Sci. USA 1998, 95, 8058.
15. Huibregtse, J. M.; Sche€ner, M.; Howley, P. M. Mol. Cell
Biol. 1993, 13, 775.
16. Kukimoto, I.; Aihara, S.; Yoshiike, K.; Kanda, T. Bio-
chem. Biophys. Res. Commun. 1998, 249, 258.
17. Lamberti, C.; Morrissey, L. C.; Grossman, S. R.; Andro-
phy, E. J. EMBO J. 1990, 9, 1907.
18. Ronco, L. V.; Karpova, A. Y.; Vidal, M.; Howley, P. M.
Genes Dev. 1998, 12, 2061.
41. Siegers, C. P.; Strubelt, O.; Volpel, M. Arch. Toxicol.
1978, 41, 79.
42. Borch, R. F.; Pleasants, M. E. Proc. Natl. Acad. Sci. USA
1979, 76, 6611.
43. Keys, H. M.; Bundy, B. N.; Stehman, F. B.; Muderspach,
L. I.; Chafe, W. E.; Suggs, C. L.; Walker, J. L.; Gersell, D. N.
Engl. J. Med. 1999, 340, 1154.
44. Rose, P. G.; Bundy, B. N.; Watkins, E. B.; Thigpen, J. T.;
Deppe, G.; Maiman, M. A.; Clarke-Pearson, D. L.; Insalaco,
S. N. Engl. J. Med. 1999, 340, 1144.
45. Morris, M.; Eifel, P. J.; Lu, J.; Grigsby, P. W.; Levenback,
C.; Stevens, R. E.; Rotman, M.; Gershenson, D. M.; Mutch,
D. G. N. Engl. J. Med. 1999, 340, 1137.
19. Sche€ner, M.; Huibregtse, J. M.; Howley, P. M. Proc.
Natl. Acad. Sci. USA 1994, 91, 8797.
20. Thomas, M.; Banks, L. Oncogene 1998, 17, 2943.
21. Tong, X.; Howley, P. M. Proc. Natl. Acad. Sci. USA 1997,
94, 4412.
22. Zimmermann, H.; Degenkolbe, R.; Bernard, H. U.;
O'Connor, M. J. J. Virol. 1999, 73, 6209.
23. Myers, G., Androphy, E. J. The E6 Protein. In: Human
Papillomaviruses 1995; Myers, G.; Bernard, H. U.; Delius, H.;
Baker, C.; Icenogel, J.; Halpern, A. L.; Wheeler, C., Eds. Los
Alamos National Laboratory: Los Alamos, 1995; pp 47±57.
24. Rapp, L.; Chen, J. J. Biochim. Biophys. Acta 1998, 1378, 1.
46. Butz, K.; Shahabeddin, L.; Geisen, C.; Spitkovsky, D.;
Ullmann, A.; Hoppe-Seyler, F. Oncogene 1995, 10, 927.
47. Fritsche, M.; Haessler, C.; Brandner, G. Oncogene 1993,
8, 307.
48. Tummino, P. J.; Scholten, J. D.; Harvey, P. J.; Holler, T.
P.; Maloney, L.; Gogliotti, R.; Domagala, J.; Hupe, D. Proc.
Natl. Acad. Sci. USA 1996, 93, 969.
49. Rice, W. G.; Supko, J. G.; Malspeis, L.; Buckheit, R. W.
J.; Clanton, D.; Bu, M.; Graham, L.; Schae€er, C. A.; Turpin,
J. A.; Domagala, J. Science 1995, 270, 1194.