Inhibition of protein kinase C by synthetic xanthone derivatives

Article abstract of DOI:10.1016/S0968-0896(02)00641-7

The modulatory activity of two xanthones (3,4-dihydroxyxanthone and 1-formyl-4-hydroxy-3-methoxyxanthone) on isoforms α, βI, δ, η and ζ of protein kinase C (PKC) was evaluated using an in vivo yeast phenotypic assay. Both xanthones caused an effect compatible with PKC inhibition, similar to that elicited by known PKC inhibitors (chelerythrine and NPC 15437). PKC inhibition caused by xanthones was confirmed using an in vitro kinase assay. The yeast phenotypic assay revealed that xanthones present differences on their potency towards the distinct PKC isoforms tested. It is concluded that 3,4-dihydroxyxanthone and 1-formyl-4-hydroxy-3-methoxyxanthone may become useful PKC inhibitors and xanthone derivatives can be explored to develop new isoform-selective PKC inhibitors.

Full text of DOI:10.1016/S0968-0896(02)00641-7

Bioorganic & Medicinal Chemistry 11 (2003) 1215–1225
Inhibition of Protein Kinase C by Synthetic Xanthone Derivatives
Lucılia Saraiva,a,b Paula Fresco,a Eugenia Pinto,b Emılia Sousa,c
Madalena Pintoc and Jorge Goncalvesa,*
a
´
´
´
Servic¸odeFarmacologia, CEQOFFUP, FaculdadedeFarmacia, UniversidadedoPorto, ruaAnıbalCunha, 164, 4050-047Porto, Portugal
b
c
´
Servic¸odeMicrobiologia,CEQOFFUP,FaculdadedeFarmacia,UniversidadedoPorto,ruaAnıbalCunha,164,4050-047Porto,Portugal
´
´
Servic¸o de Quımica Orgaˆnica, CEQOFFUP, Faculdade de Farmacia, Universidade do Porto, rua Anıbal Cunha, 164, 4050-047 Porto,
´
Portugal
Received 7 June 2002; accepted 4 December 2002
Abstract—The modulatory activity of two xanthones (3,4-dihydroxyxanthone and 1-formyl-4-hydroxy-3-methoxyxanthone) on
isoforms a, bI, d, Z and z of protein kinase C (PKC) was evaluated using an in vivo yeast phenotypic assay. Both xanthones caused
an effect compatible with PKC inhibition, similar to that elicited by known PKC inhibitors (chelerythrine and NPC 15437). PKC
inhibition caused by xanthones was confirmed using an in vitro kinase assay. The yeast phenotypic assay revealed that xanthones
present differences on their potency towards the distinct PKC isoforms tested. It is concluded that 3,4-dihydroxyxanthone and
1-formyl-4-hydroxy-3-methoxyxanthone may become useful PKC inhibitors and xanthone derivatives can be explored to develop
new isoform-selective PKC inhibitors.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
of several isoforms in the cells used for in vivo assays,
and the difficulties to reproduce in vitro the interactions
occurring in vivo between PKC and other cellular con-
stituents. These limitations can be circumvented using
the yeast phenotypic assay. This is an in vivo assay,
using yeast expressing an individual mammalian PKC
isoform, based on the growth inhibition (reflecting an
increase in the cell doubling time) which is proportional
to the degree of PKC activation.4,5 In the yeast pheno-
typic assay, PKC activators cause growth inhibition of
transformed yeast4,5 while PKC inhibitors block the
growth inhibition caused by a PKC activator.6,7
Protein kinase C (PKC) is a family of serine-threonine
kinases with important roles in cellular functions such
as growth, differentiation, tumor promotion and apop-
tosis.1 PKC isoforms are grouped into at least three
groups: the classical PKCs (cPKCs), which include the
isoforms a, bI, bII and g; the novel PKCs (nPKCs),
which include the isoforms d, e, y, and Z; and the atyp-
ical PKCs (aPKCs), which include the isoforms z and l/
i.2,3 More recently, a fourth group of structurally dis-
tinct PKCs has been identified, the so-called PKC-related
kinases (PRKs).2,3 A new member of the PKC family,
PKCm/PKD has also been reported, but its inclusion
remains controversial.2
Natural and synthetic xanthones have been reported to
mediate several important biological activities namely
anti-tumor,8ꢀ12 anti-inflammatory,13 anti-thrombotic14
and neuropharmacological effects.15ꢀ17 Recently, it has
been suggested that these compounds may act, at least in
part, by interacting with PKC, and cause effects compat-
ible not only with PKC activation18 but also with PKC
inhibition. Prenylated xanthones were shown to be
potent inhibitors of eukaryote kinases in vitro, including
PKC19 and norathyriol (1,3,6,7-tetrahydroxyxanthone)
reduced the PMA-induced respiratory burst and aggre-
gation, an effect ascribed to inhibition of PKC.20
Knowledge of specific roles attributable to individual
PKC isoforms has been hampered by the lack of iso-
form-selective drugs and the search for selective drugs has
been difficult because of several methodological limit-
ations. Examples of these limitations are the coexistence
*Corresponding author. Tel.: +351-22-2078932; fax: +351-22-
2078969; e-mail: jorge.goncalves@ff.up.pt
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0968-0896(02)00641-7

1216
L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
The modulatory activity of a series of twenty-two simple
xanthones (9H-xanthen-9-ones) on PKC isoforms a and
bI (cPKCs), d and Z (nPKCs) and z (aPKCs) were
characterized in our laboratory using the yeast pheno-
typic assay. The majority of these compounds caused an
effect compatible with PKC activation.21 In the present
study, we report that two of the xanthones tested
(3,4-dihydroxyxanthone and 1-formyl-4-hydroxy-3-
methoxyxanthone) caused effects compatible with PKC
inhibition. The effects of these xanthones were com-
pared with two established potent and selective PKC
inhibitors (chelerythrine22 and NPC 1543723,24). The
two compounds presented differences on their potency
towards the individual PKC isoforms tested. Further-
more, for some isoforms, they showed potencies even
higher than those of the PKC inhibitors used. There-
fore, 3,4-dihydroxyxanthone and 1-formyl-4-hydroxy-3-
methoxyxanthone can be used as PKC inhibitors for
characterization of PKC-mediated effects and for the
development of new PKC isoform-selective inhibitors.
Figure 1. Immunodetection of PKC-a, -bI, -d, -Z and -z isoforms
expressed in transformed Saccharomyces cerevisiae (CG379). Indivi-
dual immunoblots are presented in a horizontal arrangement and were
obtained from proteins extracts (ꢁ40 mg protein/lane) from cultures
grown in selective medium in the presence of 2% galactose (lanes A
and B; duplicate samples) or in the absence of 2% galactose (lane D).
Positive controls (lane C; 4 mg) were obtained using recombinant pro-
teins PKC-a (MW 76,799 Da), PKC-bI (MW 76,790 Da), PKC-d (MW
77,517 Da), PKC-Z (MW 77,600 Da) and PKC-z (MW 67,740 Da).
Results
Expression of mammalian PKC-a, -bI, -d, -Z and -z was
confirmed by immunoblotting, of protein extracts of
yeast cells, transformed with the gene of a single mam-
malian PKC isoform, grown in the presence of the
transcription inducer (2% galactose). Expression of
each of these PKC isoforms resulted in a single anti-
genic band, which co-migrated with the respective
recombinant protein. Protein extracts of transformed
yeast cells grown in the absence of galactose did not
present antigenic bands (Fig. 1).
Figure 2. Chemical structure of 3,4-dihydroxyxanthone (1) and 1-formyl-
4-hydroxy-3-methoxyxanthone (2).
Yeast phenotypic assay
PMA (considered a standard activator for the classical
and novel PKC isoforms) was tested in yeast expressing
PKC-a, -bI, -d or -Z, in concentrations up to 10ꢀ5 M
(higher concentrations were not possible to test due to
its low solubility in the culture medium). Since atypical
PKC isoforms are not activated by phorbol esters,25
arachidonic acid (up to 10ꢀ5 M) was used as the stan-
dard activator for PKC-z.26 Effects of drugs were
expressed as percentage of the maximal growth inhibi-
tion caused by the standard PKC activator (PMA for
PKC-a, -bI, -d, and -Z or arachidonic acid for PKC-z;
100% growth inhibition was assumed to be that caused
by 10ꢀ5 M of the respective PKC activator).
the concentration of 10ꢀ5 M, did not influence yeast
growth. NPC 15437, in concentrations higher than
10ꢀ6 M, inhibited yeast growth. Therefore, to avoid
these effects non-mediated by the expressed mammalian
PKC isoform, 10ꢀ6 M was the maximal concentration
of NPC 15437 tested when expression of mammalian
PKC isoforms were induced (galactose present in the
medium).
In the presence of galactose, PMA caused a concen-
tration-dependent inhibition on growth of yeast expres-
Table 1. Yeast growth inhibition caused by 10ꢀ5 M PMA on the
PKC isoforms studied
In the absence of galactose, neither PMA nor arachi-
donic acid altered yeast growth. The solvent used
(DMSO; final concentration 0.1%) did not change yeast
growth either in the presence or in the absence of
galactose (not shown).
PKC isoforms
Growth inhibition caused
by 10ꢀ5 M PMA
a
bI
d
Z
z
40.6ꢂ1.9 (n=36)
36.1ꢂ1.1 (n=56)
26.6ꢂ0.6 (n=52)
21.3ꢂ0.7 (n=52)
0.2ꢂ1.4 (n=36)a
The influence of established PKC inhibitors (chelery-
thrine and NPC 15437) and of 3,4-dihydroxyxanthone
(1) and 1-formyl-4-hydroxy-3-methoxyxanthone (2)
(Fig. 2) on the growth of yeast expressing individual
PKC isoforms was also studied. In the absence of
galactose, chelerythrine, xanthones 1 and 2, all tested in
Growth in the presence of solvent was considered to be 0% growth
inhibition (100% growth; see Experimental for details). Each value
represents the mean ꢂ SEM of the indicated n determinations.
aGrowth inhibition caused by 10ꢀ5 M arachidonic acid was 23.9ꢂ0.7
(n=56).

L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
1217
sing cPKCs or nPKCs isoforms. EC50 values were (nM)
99.4ꢂ10.1 (PKC-a), 214.7ꢂ48.0 (PKC-bI), 453.1ꢂ32.8
(PKC-d) and 7.6ꢂ0.6 (PKC-Z) (n=64), for the isoform
indicated. In yeast expressing PKC-z, arachidonic acid,
In the yeast phenotypic assay, a PKC inhibitor may be
characterized by its capacity to block the growth inhibi-
tion caused by a PKC activator6,7 and may cause, per se, a
stimulation of yeast growth.6 Therefore, chelerythrine,
NPC 15437 and xanthones 1 and 2 were tested alone and
combined with the appropriate PKC activator (PMA for
PKC-a, -bI, -d and -Z; arachidonic acid for PKC-z).
but not PMA, caused
a concentration-dependent
growth inhibition of yeast expressing this isoform, with
an EC50 of 208.2ꢂ30.3 nM (n=64). Maximal values of
growth inhibition caused by 10ꢀ5 M PMA (or arachi-
donic acid for PKC-z), on the PKC isoforms tested are
presented on Table 1.
On a first approach, chelerythrine, NPC 15437 and
xanthones were tested alone and only on the maximal
Figure 3. Effects of xanthones 1 and 2 and of the PKC inhibitors (chelerythrine and NPC 15437) on the growth of yeast expressing individual PKC
isoforms (a, bI, d, Z or z). Yeast cells expressing the indicated mammalian PKC isoform were incubated with each compound at the concentration
indicated or solvent (DMSO, 0.1% final concentration). Results are expressed as % of growth. Shown are means ꢂ SEM of 16–24 determinations.
Significantly different from growth in the presence of solvent: *P <0.05, **P <0.001 (paired Student’s t test).

1218
L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
Figure 4. Interaction experiments on cPKCs (a and bI): concentration-response curves for PMA alone (open circles) and in the presence of 10ꢀ5 M
xanthone 1, xanthone 2, chelerythrine and 10ꢀ6 M NPC 15437 (filled circles) or 10ꢀ8 M (filled squares). Results are expressed as % of the maximal
effect caused by 10ꢀ5 M PMA. Shown are means ꢂ SEM of 16–20 determinations. Significantly different from growth inhibition caused by PMA
alone: *P <0.05, **P <0.001 (unpaired Student’s t test).

L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
1219
Figure 5. Interaction experiments on nPKCs (d and Z): concentration-response curves for PMA alone (open circles) and in the presence of 10ꢀ5 M
xanthone 1, xanthone 2, chelerythrine and 10ꢀ6 M NPC 15437 (filled circles) or 10ꢀ8 M (filled squares). Results are expressed as % of the maximal
effect caused by 10ꢀ5 M PMA. Shown are means ꢂ SEM of 16–20 determinations. Significantly different from growth inhibition caused by PMA
alone: *P <0.05, **P <0.001 (unpaired Student’s t test).

1220
L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
concentration feasible under these experimental condi-
tions: 10ꢀ5 M for chelerythrine and xanthones 1 and 2;
10ꢀ6 M for NPC 15437. In general, both PKC inhibitors
chelerythrine and NPC 15437, and xanthones 1 and 2
stimulated growth of yeast expressing the individual
mammalian PKC isoforms (Fig. 3), although this
growth stimulation was less marked with chelerythrine
than with NPC 15437 or xanthones 1 or 2. A decrease
on concentration reduced the growth stimulation
induced by these drugs. At 10ꢀ8 M, chelerythrine, NPC
15437, xanthones 1 and 2 failed to cause a significant
stimulation of yeast growth (Fig. 3).
On a subsequent series of experiments, the ability of
chelerythrine, NPC 15437, xanthones 1 and 2 to block
the PKC activator-induced growth inhibition was
investigated. These compounds were tested both at the
maximal concentration they showed to be selective for
PKC (10ꢀ5 M for chelerythrine and xanthones 1 and 2;
10ꢀ6 M for NPC 15437) and at 10ꢀ8 M (concentration
at which all failed to significantly change growth of
transformed yeast). Effects of chelerythrine, NPC 15437
and xanthones 1 and 2 on the growth inhibition caused
by the PKC activator PMA on yeast expressing PKC-a
or -bI and on yeast expressing PKC-d or -Z are shown
in Figs. 4 and 5, respectively. Figure 6 shows the effects
of chelerythrine, NPC 15437 and xanthones 1 and 2 on
the growth inhibition caused by arachidonic acid on
yeast expressing PKC-z.
In general, both chelerythrine and NPC 15437 reduced
the growth inhibition caused by the PKC activator,
causing a rightward shift on the concentration-response
curve of the PKC activator (compare open and filled
symbols in lower panels of Figs. 4–6). The rightward
shift was, generally, concentration-dependent being the
displacement caused by 10ꢀ8 M less pronounced than
that caused by 10ꢀ5 M chelerythrine or by 10ꢀ6 M NPC
15437 (compare filled squares and circles in lower panels
of Figs 4–6). Xanthones 1 and 2 also caused, in general,
a rightward shift of the concentration-response curve of
the PKC activator (see upper panels of Figs 4–6). As
observed with chelerythrine and NPC 15437, the right-
ward shift was, in general, concentration-dependent
being the displacement caused by 10ꢀ8 M less pro-
nounced than that caused by 10ꢀ5 M xanthones 1 or 2
(compare filled squares and circles in upper panels of
Figs 4–6). However, in some cases (xanthone 1 on PKC-
z, xanthone 2 on PKC-a and-d, NPC 15437 on PKC-bI
and-d), no concentration-dependence was found, being
the rightward shift identical for the highest and the
lowest concentration of PKC inhibitor tested (P >0.05;
unpaired Student’s t test).
In order to compare the potency of the PKC inhibitors
to that of xanthones 1 and 2 in blocking the effect of
PKC activators, EC50 ratios of the PKC activator,
obtained in the presence and in the absence of PKC
inhibitors or xanthones 1 or 2, were calculated (EC50 is
the concentration of the PKC activator that caused half
of the growth inhibition caused by 10ꢀ5 M PMA; ara-
chidonic acid for PKC-z). Because the highest concen-
tration of NPC 15437 and xanthones 1 and 2 caused
Figure 6. Interaction experiments on aPKC (z): concentration–
response curves obtained for arachidonic acid alone (open circles) and
in the presence of 10ꢀ5 M xanthone 1, xanthone 2, chelerythrine and
10ꢀ6 M NPC 15437 (filled circles), or 10ꢀ8 M (filled squares). Results
are expressed as % of the maximal effect caused by 10ꢀ5 M of arachi-
donic acid. Shown are means ꢂ SEM of 16–20 determinations. Sig-
nificantly different from growth inhibition caused by arachidonic acid
alone: *P <0.05, **P <0.001 (unpaired Student’s t test).

L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
1221
Table 2. EC50 ratios for xanthones 1 and 2, NPC 15437 and chelerythrine on the individual PKC isoforms tested
Compound
EC50 ratioa
PKC-a
PKC-bI
PKC-d
PKC-Z
PKC-z
Xanthone 1
Xanthone 2
NPC 15437
Chelerythrine
1.8ꢂ0.2*x
79.2ꢂ6.9
5.1ꢂ0.5*x
1.9ꢂ0.3*x
0.8ꢂ0.1*x
28.7ꢂ2.4y
3.7ꢂ0.3*x
2.9ꢂ0.3*x
4.5ꢂ0.8*x
24.0ꢂ1.6y
2.4ꢂ0.1*x
1.3ꢂ0.1*x
600.0ꢂ25.4
20.0ꢂ2.8{y
556.0ꢂ44.7
63.2ꢂ1.7{
40.1ꢂ1.8x
10.4ꢂ1.1{y
4.5ꢂ0.1{x
1.4ꢂ0.2{x
The EC50 values were considered the concentration of PKC activator that caused half of the growth inhibition caused by 10ꢀ5 M of PMA (arachi-
donic acid for PKC-z). Shown are means ꢂ SEM of 16–20 determinations. Significant differences: from xanthone 2, *P <0.05; from xanthone 1, {P
<0.05; from PKC-a, yP <0.05; from PKC-Z, xP <0.05 (one way ANOVA, followed by Tukey’s post-hoc test).
aEC50 ratio=EC50 (PKC activator + 10ꢀ8 M compound)/EC50 (PKC activator).
growth stimulation, only ratios obtained in the presence
of 10ꢀ8 M (which did not cause a significant stimulation
of growth) were used for comparisons (Table 2).
According to the EC50 ratios obtained, chelerythrine,
NPC 15437 and xanthones 1 and 2 presented differences
on their potencies towards the PKC isoforms tested and
the following order of potencies for the compounds
studied was proposed: for chelerythrine, PKC-Z>
-bI>-a >-z=-d; for NPC 15437, PKC-Z >-a=-bI=-z
>-d; for xanthone 1, PKC-Z >-z >-d >-a=-bI and
for xanthone 2, PKC-a >-bI=-d=-Z >-z (for defini-
tion of the rank order of potency, the signal >was
applied only when the EC50 ratio for the isoform placed
at the left was significantly higher than that of the iso-
form placed at the right of the signal; otherwise=was
applied; P <0.05; unpaired Student’s t test). Like NPC
15437 and chelerythrine, xanthone 1 presented an
higher EC50 ratio on PKC-Z. On the other hand, the
EC50 ratios obtained with xanthone 2 on the different
isoforms tested did not differ so markedly, although it
was shown to be higher on PKC-a than on other isoforms.
Xanthone 1 was the compound which presented the high-
est EC50 ratio on PKC-Z and -z, whereas xanthone 2
showed the highest EC50 ratio on PKC-a, -bI and -d.
In vitro kinase assay
To directly test if xanthones 1 and 2 were interacting
with PKC, their effects were studied in vitro, on purified
rat brain PKC. The test used is based on the ability of
PKC inhibitors to revert the decrease on the relative
fluorescence units (RFU) caused by the recommended
endogenous PKC activator phosphatidyl-l-serine
(which is assumed to reflect inhibition of the PKC cat-
alytic activity). Xanthones 1 and 2, chelerythrine and
NPC 15437 were used on the lowest concentration tes-
ted (10ꢀ8 M), the same used to obtain the EC50 ratios
presented on Table 2. Results are shown in Figure 7.
Xanthones 1 and 2 were able to revert the effect of the
endogenous PKC activator being the effects similar to
those caused by chelerythrine and NPC 15437.
Discussion
The yeast phenotypic assay, which uses transformed
yeast expressing individual mammalian PKC isoforms,
has been proposed as an alternative rapid and simple in
vivo assay for the screening of potential PKC modulators.
In the present study, the mammalian PKC-a, -bI, -d, -Z
or -z isoforms were expressed in the same yeast strain.
According to previous studies,21 exposure of trans-
formed yeast expressing a mammalian PKC isoform to
a PKC activator, caused a concentration-dependent
inhibition of yeast growth, reflecting an increase in the
cell doubling time. The inhibition of yeast growth
caused by the PKC activator was assumed to be due to
its interaction with the PKC isoform expressed, because
it did not occur when transformed yeast cells were
incubated in medium lacking the expression inducer,
galactose. Confirmation that mammalian PKC isoforms
were not expressed when galactose was not added to the
medium was obtained by immunobloting. PMA failed
to influence growth of yeast expressing PKC-z, con-
firming the well known inactivity of phorbol esters on
isoforms of the atypical PKC family25 and provided
additional evidence that the growth inhibition caused by
Figure 7. In vitro kinase assay using a purified rat brain PKC enzyme:
effects of chelerythrine, NPC 15437, xanthone 1 and xanthone 2 (all
tested at the concentration of 10ꢀ8 M). Relative fluorescence units
(RFU) were measured in samples without enzyme (100%; control) or
with PKC, which caused a decrease on RFU (see Experimental for
details). Phosphatidyl-l-serine, the recommended PKC activator, was
present in all samples; drugs or solvent, were tested in the absence or
in the presence of PKC. Effects of drugs was expressed as % of RFU
obtained in the absence of PKC. Shown are means ꢂ SEM of four
assays (8 assays for phosphatidyl-l-serine alone). Significantly differ-
ent from solvent (phosphatidyl-l-serine alone): *P <0.05 (one way
ANOVA, followed by Tukey’s post-hoc test).

1222
L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
PMA, in this assay, is due to PKC activation. There-
fore, in the present experimental conditions, growth
inhibition of yeast expressing an individual mammalian
PKC isoform caused by the known PKC activators was
assumed to reflect PKC activation.
Comparison of the potency was based on the comparison
of EC50 ratios caused by concentrations of chelerythrine,
NPC 15437, xanthone 1 and xanthone 2 that, per se, did
not stimulate yeast growth. Under these conditions xan-
thone 1 caused a pattern of effects similar to that of NPC
15437: both xanthone 1 and NPC 15437 were very potent
upon PKC-Z and presented a low potency upon PKC-a,
-bI and -d. Chelerythrine presented also a marked potency
on PKC-Z, although it was less potent than xanthone 1
and NPC 15437. The high potency of xanthone 1 on PKC-
z must also be emphasised. On PKC-z, xanthone 1 seems
to be much more potent than NPC 15437 and cheler-
ythrine. Xanthone 2 did not share the pattern of effects
presented by NPC 15437 and chelerythrine, but presented
the highest potency on PKC-a, -bI and -d.
Absolute values of growth inhibition caused by the PKC
activators used (PMA or arachidonic acid) differ among
the expressed PKC isoforms. These differences may be
ascribed to variations on the degree of expression of the
mammalian PKC isoform or to differences on the cata-
lytic potential of the expressed isoform. The later possi-
bility is supported by results obtained in vitro that
showed that identical amounts of PKC isoforms, in the
presence of the same concentration of a PKC activator,
led to different degrees of phosphorylation of a given
substract.27 In order to avoid the influence of different
absolute values on data interpretation, growth inhibition
caused by PKC activators was standardized and, for each
isoform, growth inhibition caused by a maximal concen-
tration of PKC activator assumed to be the 100% of
growth inhibition reachable under the present experi-
mental conditions and EC50 values estimated afterwards.
Like other PKC inhibitors, differences on the potency of
xanthones 1 and 2 to block the effect of PKC activators
may be influenced by their ability to reach their intra-
cellular receptor (PKC) in mammalian cells. The yeast
phenotypic assay is an in vivo assay, and factors that
influence the access of compounds to intracellular
receptors in yeast cells are similar to those present in
more complex eukariotic cells. Therefore, the activity
showed by xanthones 1 and 2 should reflect their
potency in vivo.
PKC inhibitors, like chelerythrine and NPC 15437,
caused, per se, a stimulation of yeast growth. A similar
effect was reported to occur with bryostatin 1 on yeast
expressing PKC-g,6 an effect probably due to blockade
of the PKC activation caused by endogenous activators.
Xanthones 1 and 2, like NPC 15437, also stimulated
growth of yeast expressing mammalian PKC isoforms.
Nevertheless, chelerythrine, described as a potent PKC
inhibitor, only marginally changed growth, what ques-
tions the use of growth stimulation as an index of
potency for PKC inhibitors. Reasons for the lack of
correlation between potency of PKC inhibitors and
their ability to increase yeast growth are, at present, not
fully understood, but may rely on the different site of
interaction between these inhibitors and PKC: NPC
15437 interacts with the C1 region of the regulatory
domain23,24 whereas chelerythrine interacts with the
ATP region of the catalytic domain.22 Xanthones 1 and
2, when tested alone, also stimulated yeast growth, with
a pattern of effects similar to those of NPC 15437. This
resemblance may indicate that these xanthones interact
with PKC at the same domain of NPC 15437.
Confirmation that, xanthones 1 and 2 blocked the effect
of PKC activators on yeast growth by interacting with
PKC was obtained using an in vitro kinase assay. At the
lowest concentration used on the yeast phenotypic assay,
xanthones 1 and 2, like NPC 15437 and chelerythrine,
inhibited the PKC catalytic activity, an effect similar to
that caused by the established PKC inhibitors.
Previous studies showed that xanthones may act as
PKC activators.21 The present study extends this obser-
vation and shows that xanthone derivatives may also
act as potent PKC inhibitors. Given the implication of
PKC activation in carcinogenesis and other human dis-
eases, PKC inhibition have become a major research
target28 and xanthone derivatives may become an
important family to look for potent and selective PKC
inhibitors.
Conclusions
Blockade of the PKC activators-induced growth inhibi-
tion by established PKC inhibitors has been previously
reported to reflect inhibition of the mammalian PKC
isoform expressed.6,7 Under these conditions, PKC
inhibitors caused a rightward shift of the concentration-
response curve for the PKC activator. The rightward
shift was estimated by the ratio between the EC50 of the
PKC activator obtained in the presence and in the
absence of the PKC inhibitor. The obtained EC50 ratios
were used as an index of the potency of the PKC inhi-
bitor on the mammalian PKC isoform expressed.
This study shows, by using in vitro and in vivo assays, that
simple xanthones can act as inhibitors of mammalian a,
bI, d, Z and z PKC isoforms. The potencies presented
by 3,4-dihydroxyxanthone and 1-formyl-4-hydroxy-3-
methoxyxanthone on different PKC isoforms suggest
that xanthones may be an important group to look for
potent and isoform-selective PKC inhibitors.
Experimental
Chemistry
Chelerythrine and NPC 15437 blocked the growth inhi-
bition caused by PMA (PKC-a, -bI, -d and -Z) or by ara-
chidonic acid (PKC-z). Xanthones 1 and 2 also blocked
PKC activation caused by PMA or by arachidonic acid.
Purifications of compounds were performed by column
chromatography; using Merck silica gel 60 (0.50–0.20 mm)

L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
1223
and preparative thin layer chromatography (tlc), using
Merck silica gel 60 (GF254). Melting points were
obtained in a Kofler microscope and are uncorrected.
IR spectra were recorded on a Perkin Elmer 257 in KBr.
1H and 13C NMR spectra were taken in CDCl3 or
DMSO-d6 at room temperature, on Bruker AC 200
instrument. Chemical shifts are expressed in d (ppm)
values relative to tetramethylsilane (TMS) as an internal
reference. MS spectra were recorded as EI (Electronic
Impact) mode on a Hitachi Perkin–Elmer.
Wayne University, Detroit, USA) and YEplac181,
encoding the cDNA for the rat PKC-d, mouse PKC-Z
or PKC-z (kindly offered by Dr. Nigel Goode, Royal
Veterinary College, London, UK) were amplified in
Escherichia coli DH5a and confirmed by restriction
analysis. The plasmids used contain galactose-inducible
transcriptional elements and the leu2gene for selection.
Saccharomyces cerevisiae (S. cerevisiae; strain CG379; a
ade5 his7-2 leu2-112 trp1-289a ura3-52 [Kil-O]; Yeast
Genetic Stock Center, University of California, Berkeley,
USA) was transformed using the lithium acetate
method.33 To ensure the selection of transformed yeast,
cells were grown in leucine-free medium, in 1.5% agar
plates, at 30 ꢃC.
The synthesis of the following compounds have been
carried out as follow.
3,4-Dihydroxyxanthone (1). This compound was
obtained by demethylation of 3,4-dimethoxyxanthone
according to the procedure described.29 Mp >330 ꢃC,
chloroform (240–241 ꢃC, ethanol/aqueous;30 238–
240 ꢃC, methanol31); umax (cmꢀ1) KBr: 3527, 3391, 3088,
1593, 1459, 1340, 1229, 1056, 756; 1H NMR (DMSO-d6,
200.13 MHz) d: 8.14 (1H, dd, J=8.1 and 1.6 Hz, H-8),
7.81 (1H, ddd, J=8.6, 6.9 and 1.7 Hz, H-6), 7.62(1H,
dd, J=8.6 and 0.9 Hz, H-5), 7.56 (1H, d, J=8.6 Hz,
H-1), 7.42(1H, ddd, J=8.1, 6.9 and 0.9 Hz, H-7), 6.93
(1H, d, J=8.6 Hz, H-2); 13C NMR (DMSO-d6,
50.03 MHz) d: 175.3 (C-9), 155.5 (C-4b), 151.6 (C-3),
146.4 (C-4a), 134.8 (C-6), 132.7 (C-4), 125.9 (C-8), 124.0
(C-7), 120.8 (C-8a), 118.0 (C-5), 116.6 (C-1), 114.7
(C-9a), 113.2(C-2); MS m/z (rel int): 230 (3, [M+2]+.),
229 (22, [M+1]+.), 228 (100, [M+.]), 200 (14), 171 (9),
126 (10), 115 (12), 100 (8), 77 (5), 63 (7).
For the yeast phenotypic assay, transformed cells were
incubated in leucine free-medium, with slow shaking, at
30 ꢃC. The leucine free-medium contained 0.7% yeast
nitrogen base, 2% glucose (w/v) or the indicated carbon
source, amino acids, purines and pyrimidines, according
to the transformed yeast requirements. Galactose (2%;
w/v), instead of glucose, was included to induce tran-
scription of the mammalian PKC gene.
Cell lysis and immunoblotting. Cell lysis was performed
basically as described.21 The protein concentration was
determined using a kit for protein quantification (Co-
omassie1 Protein Assay Reagent Kit, Pierce, Biocontec,
Lisbon, Portugal). Similar amounts of protein (ꢁ40 mg)
from protein extracts were then separated on 10% SDS-
polyacrylamide gels (Mini-Protean II, BioRad, Her-
cules, CA, USA). Positive controls (4 mg) were obtained
using recombinant proteins PKC-a (MW 76,799 Da),
PKC-bI (MW 76,790 Da), PKC-d (MW 77,517 Da),
PKC-Z (MW 77,600 Da) and PKC-z (MW 67,740 Da).
Proteins were electrophoretically transferred to nitro-
cellulose membranes and probed on immunoblots with
specific rabbit antibodies to the individual mammalian
PKC isoforms and revealed with a secondary alkaline
phosphatase-conjugated anti-rabbit IgG (AP-10,
Oxford Biomedical Research, LabClinics, Barcelona,
Spain).
1-Formyl-4-hydroxy-3-methoxyxanthone (2). This com-
pound was obtained according to the procedure descri-
bed.32
General methods for the in vivo yeast phenotypic assay
Yeast nitrogen base was from DIFCO (Merck Portugal,
Lisboa, Portugal). The kit for protein quantification
was from Pierce (Biocontec, Lisboa, Portugal). The
secondary alkaline phosphatase-conjugated anti-rabbit
IgG detection kit (AP-10), recombinant proteins PKC-a
(PK11), PKC-bI (PK16), PKC-d (PK31), PKC-Z
(PK46) and PKC-z (PK41) were from Oxford Biomedi-
cal Research (LabClinics, Barcelona, Spain). Nitro-
cellulose membranes and all the reagents for sodium
dodecyl sulfate (SDS)-polyacrylamide gel electrophor-
esis and immunoblots were from BioRad (PACI, Lis-
boa, Portugal). Acid-washed glass beads, antibodies to
PKC-a, PKC-bI, PKC-d, PKC-Z and PKC-z, aprotinin,
arachidonic acid sodium salt, chelerythrine cloride,
R-2,6-diamino-N-[[1-(1-oxotridecyl)-2-piperidinyl]methyl]-
hexanamide dihydrochloride (NPC 15437 dihydro-
chloride), leupeptin, pepstatin A, phenylmethylsulfonyl
fluoride, phorbol 12-myristate 13-acetate (PMA) were
from Sigma Aldrich (Sintra, Portugal). All other che-
micals used were of analytical grade.
Yeast phenotypic assay. Transformed yeast cultures
were incubated in leucine-free medium. Optical density
values, measured at 620 nm (OD620; Cary 1E Varian
spectrophotometer, Palo Alto, CA, USA), were used as
an indicator of growth. Transformed yeast were grown
to an OD620 of approximately 1, collected by centri-
fugation and diluted to an OD620 of 0.05, in medium
containing 2% (w/v) galactose (gene transcription-
inducer) and 3% (v/v) glycerol (alternative carbon
source). Diluted cultures (200 mL) were transferred to
96-wells microtitre plates and incubated for up to 100 h,
with slow shaking at 30 ꢃC, either in the presence of
drugs or solvent (DMSO 0.1%; final concentration).
Growth was monitored by determining the OD620 using
a plate reader (BioRad Benchmark Microplate Reader;
Hercules, CA, USA). In preliminary experiments,
growth curves for individual isoforms were determined
and the duration of the logarithmic and stationary
phases identified. Estimation of drug effects was based
on OD620 measurements at fixed time points (at 65 h for
Yeast transformation and cell cultures. Constructed
yeast expression plasmids YEp52and YEp51, encoding
the cDNA for bovine PKC-a and for rat PKC-bI,
respectively (kindly offered by Dr. Heimo Riedel,

1224
L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
cPKC isoforms or at 48 h incubation for nPKC and
aPKC), times occurring during the respective logarith-
mic phase and where a ‘steady-state growth inhibition’
(period of time during which maximal inhibition of
growth was reached and remained constant or changed
only slightly) was reached. In individual experiments,
OD620 was routinely monitored for up to 100 h to con-
firm whether the ‘fixed time points’ chosen were appro-
priate for all the compounds studied (PMA or
arachidonic acid). The difference between the maximal
OD620 reached and that measured at the beginning of
incubation was used as an index of yeast growth. Drugs
or solvent were added at the beginning and kept
throughout the incubation. Yeast growth in the pre-
sence of drugs was expressed as percentage of growth
observed in parallel experiments in the presence of sol-
vent; it was further transformed into growth inhibition
by subtracting that value from 100. Because growth
inhibition caused by a maximal concentration of the
standard PKC activator varied between isoforms, 100%
growth inhibition was assumed to be that caused by
10ꢀ5 M PMA (or arachidonic acid for PKC-z), in order
to standardize the maximal inhibition reachable; 0%
growth inhibition would occur when growth in the pre-
sence of drugs was identical to that in the presence of
solvent. Effects of PKC activators were expressed as
percentage of that effect.
phosphorylation is reduced causing an increase on
RFU. RFU in samples containing no PKC were con-
sidered as 100%. Effects of drugs were expressed as
percentage of that value.
Statistical analysis. Results are given as arithmetic
means ꢂ SEM and n represents the number of deter-
minations. Differences between means were tested for
significance using either paired Student’s t test, unpaired
Student’s t test or one way ANOVA, followed by
Tukey’s post-hoc test. A P value less than 0.05 was
taken to be statistically significant.
Acknowledgements
The authors thank Fundacao para a Ciencia e Tecno-
logia (FCT; I&D n. 226/94), POCTI (QCA III) and
FEDER for financial support. E. Sousa is a recipient of
a PhD grant from FCT (PRAXIS XXI/BD/15663/98).
References and Notes
1. Musashi, M.; Ota, S.; Shiroshita, N. Int. J. Hematol. 2000,
72, 12 .
2. Webb, B. L.; Hirst, S. J.; Giembycz, M. A. Br. J. Pharma-
col. 2000, 130, 1433.
For interaction experiments, a single concentration of
PKC inhibitors or xanthones 1 or 2 were added to PMA
(10ꢀ8–10ꢀ5 M; arachidonic acid for PKC-z). Concen-
tration-response curves for the PKC activator alone in
the absence or in the presence of PKC inhibitors or
xanthones 1 or 2, were obtained and the concentration
of PKC activator that caused 50% growth inhibition
(EC50) determined. For each compound the EC50
ratios [EC50 (PKC activator+10ꢀ8 M compound)/
EC50 (PKC activator)] were calculated on each PKC
isoform tested.
3. Ron, D.; Kazanietz, M. G. FASEB J. 1999, 13, 1658.
4. Riedel, H.; Hansen, H.; Parissenti, A. M.; Su, L.; Shieh,
H. L.; Zhu, J. J. Cell. Biochem. 1993, 52, 320.
5. Shieh, H. L.; Hansen, H.; Zhu, J.; Riedel, H. Mol. Carci-
nog. 1995, 12, 166.
6. Keenan, C.; Goode, N.; Pears, C. FEBS Lett. 1997, 415,
101.
7. Keenan, C.; Goode, N.; Pears, C. Biochem. Soc. Trans.
1997, 25, S591.
8. Abou-Shoer, M.; Boettner, F. E.; Chang, C. J.; Cassady, J.
Phytochemistry 1988, 27, 2795.
9. Liou, S. S.; Shieh, W. L.; Cheng, T. H.; Won, S. J.; Lin,
C-N. J. Pharm. Pharmacol. 1993, 45, 791.
10. Lin, C. N.; Liou, S. J.; Lee, T. H.; Chuang, Y. C.; Won,
S-J. J. Pharm. Pharmacol. 1996, 48, 539.
11. Liu, H. S.; Lin, C. N.; Won, S. J. Anticancer Res. 1997, 17,
1107.
12. Yoshimi, N.; Matsunaga, K.; Katayama, M.; Yamada,
Y.; Kuno, T.; Qiao, Z.; Hara, A.; Yamahara, J.; Mori, H.
Cancer Lett. 2001, 163, 163.
13. Lin, C. N.; Chung, M. I.; Liou, S. J.; Lee, T. H.; Wang,
J-P. J. Pharm. Pharmacol. 1996, 48, 532.
14. Lin, C. N.; Hsieh, H. K.; Liou, S. J.; Ko, H. H.; Lin, H. C.;
Chung, M. I.; Ko, F. N.; Liu, H. W.; Teng, C. M. J. Pharm.
Pharmacol. 1996, 48, 887.
15. Marona, H. Pharmazie 1998, 53, 672.
16. Marona, H.; Gorka, Z.; Szneler, E. Pharmazie 1998, 53,
219.
17. Mak, N. K.; Li, W. K.; Zhang, M.; Wong, R. N. S.; Tai,
L-S.; Yung, K. K. L.; Leung, H. W. Life Sci. 2000, 66, 347.
18. Mak, N. K.; Lung, H. L.; Wong, R. N. S.; Leung, H. W.;
Tsang, H. Y.; Leung, K. N. Planta Med. 2001, 67, 400.
19. Lu, Z. X.; Hasmeda, M.; Mahabusarakam, W.; Ternai,
B.; Ternai, P. C.; Polya, G. M. Chem. Biol. Interact. 1998, 114,
121.
In vitro kinase assay
The direct interaction between the compounds tested
(PKC inhibitors and xanthones 1 and 2) with PKC was
TM
studied in vitro using the IQ PKC Assay Kit—Pseudo-
substrate Peptide Substrate from Pierce Biotechnology
(PACI, Lisboa, Portugal). The assay was performed in
agreement with the kit procedure for 96-well plate for-
mat. Xanthones 1 and 2, and the standard PKC inhibi-
tors (chelerythrine and NPC 15437) were tested at the
concentration of 10ꢀ8 M. Purified PKC enzyme stan-
dard from rat brain (containing a mixture of a, b and g
isoforms) was from Pierce Biotechnology (PACI, Lis-
boa, Portugal); it was used at the final concentration of
0.01 Units. Reaction mixtures containing the endogen-
ous PKC activator phosphatidyl-l-serine, were incu-
bated for 90 min and fluorescence intensity determined
using
a
microplate spectrofluorometer SPEC-
TRAmax1GEMINI XS (Molecular Devices Corpora-
tion, Sunnyvale, CA, USA). In the presence of PKC,
there was a decrease on relative fluorescence units
(RFU), reflecting a phosphorylation of the fluorescent
substrate. In the presence of PKC inhibitors, substrate
20. Wang, J. P.; Raung, S. L.; Tsao, L. T.; Lin, C. N. Eur. J.
Pharmacol. 1997, 336, 81.

L. Saraiva et al. / Bioorg. Med. Chem. 11 (2003) 1215–1225
1225
21. Saraiva, L.; Fresco, P.; Pinto, E.; Sousa, E.; Pinto, M.;
Goncalves, J. Bioorg. Med. Chem. 2002, 10, 3219.
22. Herbert, J. M.; Augereau, J. M.; Gleye, J.; Maffrand, J. P.
Biochem. Biophys. Res. Commun. 1990, 172, 993.
23. Sullivan, J. P.; Connor, J. R.; Shearer, B. G.; Burch, R. M.
Agents Actions 1991, 34, 142.
28. Da Rocha, A. B.; Mans, D. R. A.; Regner, A.; Schwarts-
mann, G. Oncologist 2002, 7, 17.
29. Gottlieb, O. R.; Mesquita, A. A. L.; Oliveira, G. G.;
Melo, M. T. Phytochemistry 1970, 9, 2537.
30. Grover, P. K.; Shah, G. D.; Shah, R. C. J. Chem. Soc.
1955, 3982.
24. Sullivan, J. P.; Connor, J. R.; Shearer, B. G.; Burch, R. M.
Mol. Pharmacol. 1992, 41, 38.
31. Lin, C. N.; Liou, S. S.; Ko, F. N.; Teng, C. M. J. Pharm.
Sci. 1993, 82, 11.
25. Ways, D. K.; Cook, P. P.; Webster, C.; Parker, P. J.
J. Biol. Chem. 1992, 267, 4799.
26. Nakanishi, H.; Exton, J. H. J. Biol. Chem. 1992, 267, 16347.
27. Ryves, W. J.; Evans, A. T.; Olivier, A. R.; Parker, P. J.;
Evans, F. T. FEBS Lett. 1991, 288, 5.
32. Fernandes, E. G. R.; Silva, A. M. S.; Cavaleiro, J. A. S.;
Silva, F. M.; Borges, F. M.; Pinto, M. M. M. Magn. Reson.
Chem. 1998, 36, 305.
33. Ito, H.; Fukuda, Y.; Murata, K.; Kimura, A. J. Bacteriol.
1983, 153, 163.