Download In EVALUATION OF PHOTOTOXIC PROPERTIES OF FOUR STRUCTURALLY RELATED

Photochemistry and Photobiology Val. 52,No. 3, pp. 491-500, 1990
Printed in Great Britain. All rights reserved
0031-8655/90$03.00+0.00
Copyright 0 1990 Pergamon Press plc
In vitro EVALUATION OF PHOTOTOXIC PROPERTIES
OF FOUR STRUCTURALLY RELATED
BENZOPORPHYRIN DERIVATIVES
ANNAM. RICHTER,~
ELIZABETH
WATERFIELD,3 ASHOKK. JAIN,' ETHAND. STERNBERG,*
DAVIDDOLPHIN^ and JULIAG . LEVY'*
Departments of 'Microbiology and 'Chemistry, University of British Columbia, Vancouver, BC,
Canada V6T 1W5 and 'Quadra Logic Technologies, Vancouver, BC, Canada
(Received 20 October 1989; accepted 15 February 1990)
Abstract-Four structural analogs of benzoporphyrin derivative (BPD) have been studied and compared for photosensitizing activity in v i m . All analogs have an identical reduced tetrapyrrol porphyrin
ring, and differ by the position of a cyclohexadiene ring (fused at either ring A or ring B of the
porphyrin) and the presence of either two acid groups or one acid and one ester group at rings C and
D of the porphyrin. Photosensitizer activity was tested with the M1 tumor cell line using an assay (the
MTT assay) which detects mitochondria1 hydrogenases as a measure of cell viability. This assay was
shown to be equivalent to the standard clonogenicity or ['Hlthymidine uptake assay. Comparative
studies with the BPD analogs showed that the monoacid derivatives had equivalent cytotoxicity and
were about five-fold more active than the diacid forms. This was the case whether the assays were
performed in the presence or absence of fetal calf serum.
and plants, and phthalocyanines which are synthetic
molecules. A few other structurally different comPhotodynamic therapy (PDT)?, a very promising
pounds have been studied (for a comprehensive
experimental therapy for cancer, employs photoreview see refs. Kreimer-Birnbaum, 1989; Sternberg
activatable drugs and light to destroy tumor cells.
and Dolphin, 1989). All the compounds currently
The efficiency of PDT depends to a large extent on
being investigated for prospective use in PDT have
the localization of the photosensitizer in tumor cells
the following characteristics in common: they are
or microvasculature and its ability to be efficiently
pure compounds whose structure is well known and
activated by tissue penetrating light. As wavelengths
absorb light in the range of 650-800 nm. In general,
between 700 and 1000 nm penetrate tissue best
it is difficult to compare the results obtained in
(Doiron, 1984a; Doiron et al., 1984b), a photovarious laboratories because in most instances difsensitizer for use in PDT should absorb well at
ferent experimental conditions and tumor models
these wavelengths. Extensive clinical studies with
are used. However, some comparative studies for
hematoporphyrin derivative (HpD) and Photofrins
different photosensitizers have been reported
(a mixture of dihematoporphyrin ethersiesters,
(Evensen and Moan, 1987; Moan et al., 1987).
(Dougherty, 1987; Kessel et al., 1987) have been
We are currently testing a chlorin-like porphyrin,
generally
encouraging
(Dougherty,
1986;
benzoporphyrin derivative (BPD), which is comMcCaughan, 1987; Gomer, 1989) and serve to mainposed of four structural analogs following synthesis.
tain interest in PDT and the search for new photoAll four analogs have an identical reduced tetrapyrsensitizers.
rol porphyrin ring. They differ only by the position
At present, the most widely studied compounds
of a cyclohexadiene ring which may be fused either
for potential use in PDT are from two main groups
at ring A or B of the porphyrin (ring A or B analogs)
of dyes. These are porphyrins and structurally
and the presence of either two acid groups (diacids)
related compounds, such as chlorins, chlorophylls
or one acid and one ester group (monoacids) at
and purpurins, which occur naturally in animals
rings C and D of the porphyrin p i g . 1). An four
analogs are hydrophobic, absorb red light at about
* To whom correspondence should be addressed.
700 nm wavelength, and produce singlet oxygen
t Abbreviations: BPD, benzoporphyrin derivative; BPDefficiently (Bensasson, personal communication).
MA, benzoporphyrin derivative, monoacid, ring A;
Despite the similarity among all four molecules,
BPD-DA, benzoporphyrin derivative, diacid ring A;
they differ in their light activated cytotoxicity in
BPD-DB, benzoporphyrin derivative, diacid ring B;
vitro and in vivo. In the present paper, we report
BPD-MB, benzoporphyrin derivative, monoacid, ring B;
DME, Dulbecco's Modified Eagle (medium); DMSO,
the results of comparative studies in vitro on photodimethyl sulfoxide; FCS, fetal calf serum; HpD, hematosensitizing activities of the analogs of BPD, in an
porphyrin derivative; LD,,, lethal dose killing 50% of
the population; M'IT, 3-(4,5-dimethylthiazol-2-y1)-2,5 attempt to define at a structural level those characteristics which contribute most to photosensitizer
diphenyl tetrazolium bromide; PBS, phosphate buffered
activity.
saline; PDT, photodynamic therapy.
INTRODUCTION
495
ANNAM. RICHTER
et al.
496
R
iR
Figure 1. Structure of benzoporphyrin derivative (BPD)
( 1 ) monoacid, ring A analog, (2) monoacid, ring B analog.
R = C0,Me. Diacid analogs differ from monoacid
analogs only in that they have the ester group replaced
with the acid group, therefore, they have two acid groups
at C and D rings of the porphyrin.
MATERIALS AND METHODS
Synthesis of BPD analogs. Synthesis of BPD and separation of ring A and ring B monoacid and diacid analogs
have been described earlier (Richter et a l . , 1987, 1989).
The following analogs were obtained and tested: BPDmonoacid, ring A (BPD-MA), ring B (BPD-MB), BPD
diacid, ring A (BPD-DA) and ring B (BPD-DB).
All four BPD analogs were maintained in dimethyl
sulfoxide (DMSO) at a concentration of 400 pg/mf.
Immediately before use they were diluted in culture
medium to desired concentrations. The time between the
preparation of dilution and addition to the cells in culture
was kept to a minimum. The photosensitizers and cells
incubated with photosensitizers were protected from light
at all times except for the time of planned exposure.
Because final concentrations of the analogs rarely
exceeded 1 pg/me, the effective dilution of DMSO was at
least 1:400. At these concentrations it has no effect on
cell viability.
Tumor cell lines. M1 tumor cells were obtained directly
from the M1 tumor (a methylcholanthrene-induced rhabdomyosarcoma of DBAR mice) grown subcutaneously in
mature, male DBARJ mice as described earlier (Richter
et al., 1987). The cells, obtained by teasing apart the
excised tumors, were washed and cultured in 96-well plates
(lo4 cells/well) in Dulbecco’s Modified Eagle (DME)
medium (Gibco, Grand Island, NY) supplemented with
10% fetal calf serum (FCS, Sigma Chemical Co., St.
Louis, MO). Medium was changed every 24 h and the
cells were used for the cytotoxicity assays before reaching
full confluency, when they were still in the logarithmic
growth, usually at 72 h in culture.
P815 cells (mastocytoma from DBA/2 mice), KGl
(human myeloid leukemia) and K562 (human myelomonocytic leukemia), all maintained in our laboratory for
many years, were cultured in the same medium as M1
cells in a 10% C 0 2 humidified incubator. All cells were
used for assays when they were in logarithmic growth.
Cyfofoxicifyassay. The protocols for the assay differed
slightly depending on the adherent or non-adherent type
of cells used. M1 cells (adherent) were tested as follows:
cells grown in 96-well plates, and still in logarithmic
growth, were washed with DME in order to remove serum
and dead, non-adherent cells. Viability of cells at this
stage (as assessed by the trypan blue exclusion method)
was 90% or greater. Various concentrations of photosensitizers (as well as controls) were added to wells in
quadruplicate and incubated for 1 h at 37°C in the dark.
Dark controls were tested on separate plates and kept
under the light but covered with aluminum foil. In all tests
carried out, it was found that dark controls (with or without photosensitizer) gave essentially the same cell number/
viability values as control cells (no photosensitizer) which
were exposed to light. Controls of this type were run
routinely with each experiment and indicated that the light
had no measurable toxic effect and that the BPD analogs,
at the concentrations used, had no dark toxicity. We have
observed some dark toxicity with these compounds at
concentrations above 5.0 pm/me. In assays testing the
effect of serum, the incubation of BPD analogs with
cells was carried out in the presence of 5 , 10 or 20% FCS.
Immediately after the incubation, the cells were washed
with DME and exposed to light in 100 pe DME/well for
1 h (5.4 J/cm2) and then incubated further in DME-5%
FCS at 37°C in the dark in a 10% C 0 2 humidified incubator for 18-24 h. At that time, the viability of cells
was tested using MTT [3-(4,5-dimethylthiazol-2-yl)-2,5diphenyl tetrazolium bromide] (Sigma Chemical Co., St
Louis, MO), as described by Mosmann (1983).
The cytotoxicity assay on suspension grown cells (PS15,
KG1 and K562) has been described in detail previously
(Richter et a l . , 1987). These cells were used mainly to
determine the relationship between the M l T assay and
[3H]thymidine incorporation and clonogenicity assays. The
protocol was as follows: cells in logarithmic growth (at
1O6/me) were incubated in 12 X 75 mm tubes for 1 h with
various concentrations of photosensitizers in the absence
of serum. Appropriate controls (no drug and dark) were
included. Following the incubation the cells were washed
and exposed to fluorescent light for 1 h (5.4 J/cmZ) in
0.9 me DME in tubes inclined under the light, as described
earlier (Richter et al., 1987). Immediately after the
exposure to light the cells were divided between the MTT
assay and either the [3H]thymidine incorporation assay or
clonogenic assay. At least 3 or more separate experiments
were carried out for each set of conditions.
MTT assay. This assay was carried out as described by
Mosmann (1983). Briefly, at 18-20 h post light irradiation
10 pf of MTT at 5 mg/me PBS was added per well in 96well plates containing cells in 100 pe DME-5% FCS and
the plates were placed back in the incubator for 1-3 h, at
which time the reaction was stopped with 150 pe of isopropyl alcohol containing 0.04 N hydrochloric acid (Fisher
Scientific). The cells were lysed and the blue formazan
crystals, produced by mitochondria1 hydrogenases in living
cells from the MTT substrate, were dissolved by vigorous
mixing using an octapette. Full intensity of color
developed within 10-15 min and the plates were read in
a Bio-Rad Model 2550 EIA reader using a 600 nm filter.
The percentage of cells killed was calculated in relation
to control cells incubated without photosensitizer.
[3H]thymidine incorporation assay. This protocol has
been described previously (Richter et al., 1987). Briefly,
following treatment, cells were plated in quadruplicate in
96-well plates at 1.0-1.5 x 1O5/well in DME-10% FCS
and cultured in the presence of 2 pCi [3H]thymidine/well
for 24 h. The amount of radioactivity incorporated by the
cells served as a measure of cell viability and the ability
to proliferate. The percentage of cells killed was calculated
in relation to control cells incubated with medium alone.
Clonogenicity assay. The protocol was as follows: following treatment, cells were cultured in duplicate in
35 mm plastic culture dishes (LUX 5221-R, Miles Scientific, Naperville, IL) at 5 x lo4 KG1 cells/dish in 1 mf of
DME medium containing 10% FCS, 1% methylcellulose
(Fluka AG, CH-9470 Buchs, Switzerland) and 50 p M of
2-mercaptoethanol (Eastman Kodak Co., Rochester,
NY). The dishes were incubated in a CO, incubator and
colonies were counted by inverted microscope 14 days
later. The percentage of proliferating cells killed was calculated by relating the number of colonies formed after
treatment to the number of colonies formed by control
Photosensitizing potency of benzoporphyrin derivatives
cell preparations. The large inoculum reflects the low
cloning efficiency of KG1 due to the presence in the
cultures of large numbers of terminally differentiated cells
which were no longer capable of cell division. Inoculum
size was based empirically on numbers which yielded
between 100 and 200 colonies per plate. The minimum
number of cells defined as a colony was 12-15 cells.
Light source. For light irradiation, a bank of 4 fluorescent lamps (General Electric F20T12, Delux Cool
White) was used. The spectrum of light was between 300
and 800 nm with the highest radiant power at 600 nm
(G.E. Bulletin, unpublished data). The incident light density at 11 cm distance from the lamp (routine distance for
the light irradiation of plates and tubes) was 1.5 mWI
cmz as measured by YSI Kettering Model 65 radiometer
(Yellow Springs, OH). This instrument is a wide spectrum
radiometer with greatest sensitivity at about 700 nm. The
temperature at this distance did not exceed 25°C at any
time.
RESULTS
Characteristics of BPD analogs
The structures of BPD analogs are shown in Fig.
1. The molecular weights for diacids and monoacids
are 704 and 718, respectively. They are lipophi!ic
molecules with limited solubility in aqueous solutions. All four analogs have very similar absorption
spectra (Fig. 2) and absorb light efficiently in the
red part of the spectrum. The extinction coefficient
in dichloromethane is 34 000 M-lcm-' for BPDMA and within 25% of this value for other analogs.
In aqueous solutions the 688 nm absorption peak
shifts to 692 nm. Further shift to the right is indicative of aggregation and increases with the concentration of BPD and with time in an aqueous solution. For this reason, comparative absorption
spectra were consistently measured in 50% methanol. All four analogs emit red fluorescence
i
--+--.+--+-d----d
400
500
600
700
800
497
(690 nm) when excited at 420 nm. Efficient excitation can also be achieved within the UV range
(355 nm; Jamieson et al., 1989).
Assays for determination of cell survival after
photodynamic treatment
In order to compare the 4 BPD analogs' ability
to kill test cell lines, we have developed an assay
system which measures viability colorimetrically by
measuring the activity of mitochondria1 hydrogenases. The assay has been described previously
(Mosmann, 1983) and the concentration of blue
formazan, produced by cleavage of the yellow tetrazolium salt (MTT) by hydrogenases in mitochondria
of living cells, has been shown to correlate both
with the number of viable cells and their metabolic
activity, including the ability to proliferate. This test
has been adapted in our laboratory for its speed
and convenience which enabled us to screen many
photosensitizers under diverse conditions. Incubation of cells after treatment, for about 24 h prior
to the MTT assay, was selected in order to standardize the assay and has been determined experimentally. Preliminary studies, in which the M'TT assay
was compared to trypan blue exclusion as a measure
of cell viability following photosensitization of cells,
indicated good correlation. We have also confirmed
the correlation, reported by Mosmann (1983),
between the MTT assay and ["Hlthymidine incorporation using various cell lines (Fig. 3). In the
present study, we also compared the MITT assay to
a clonogenic assay using the KG1 cell line and
various doses of the BPD analog, BPD-MA. In
these experiments cells were exposed to BPD-MA
in the absence of serum. Following light exposure,
cells were split and incubated for 24 h in DME-5%
FCS (MTT assay) or cultured for 14 days in the
100 1
0
2 0
4 0
60
8 0
100
3H-THYMIDINE ASSAY
WAVELENGTH (nrn)
Figure 2. Absorption spectrum of BPD-MA (15 pg/mt)
in 50% methanol-PBS solvent. The peaks, indicated by
arrows, are at 354, 418, 574, 626 and 688 nm wavelength.
All 4 BPD analogs have almost identical spectra (not
shown for clarity) and very close extinction coefficients
(within 25% range).
Figure 3. Correlation between the results of MTT assay
and [3H]thymidine incorporation assay. The percentage of
cells (P815, KG1, K562 cell lines) killed after the exposure
to the photosensitizer (BPD-DA or -DB) and light (5.4JI
cmz) was determined parallely by MTT assay (20-23 h
post exposure) and [3H]thymidine incorporation (0-24 h
post exposure).
ANNAM. RICHTER
et a1
498
presence of 10% FCS under the conditions
described in the Materials and Methods section.
The results (Fig. 4) show that the MTT assay correlates well with the clonogenic assay in determining
cell survival. Similar results were obtained when
MTT and clonogenicity were compared with the
P815 cell line (data not shown).
Cytotoxicity of BPD analogs
In the cytotoxicity assay developed, we routinely
use serum-free conditions, because serum might
complicate the in vitro system, without actually
duplicating the in vivo situation. The presence of
serum elements could modify drug-cell interaction
in vitro without mimicking the in vivo situation.
However, we have compared the 4 analogs of BPD
both in the absence and in the presence of 5-20%
FCS. The photosensitizing activity of the 4 BPD
analogs tested using M1 cells in serum-free medium
differed, especially between monoacid and diacid
analogs, monoacids being at least 5 times more
potent than diacids (Fig. 5). Clearly, the presence
of one or two acid groups on rings C or D of the
porphyrin macrocycle has a major effect on the
photosensitizing activity of this family of molecules.
The concentrations of BPD-MA, -MA, -DA, and
-DB required to kill 50% of cells (LD50) in the
absence of serum were 21, 19, 75, and 105 ng/me
culture medium, respectively. The presence of FCS
during the incubation with BPD analogs did not
change the order of their photosensitizing activity
(BPD-MA > = -MB
-DA > -DB), merely
increasing the concentration of the analogs required
to achieve the same photodynamic killing (Fig. 6).
The LDSo
of BPD analogs in the presence of serum
~.
were proportional to the concentration of serum,
roughly doubling with doubling serum concentrations.
*
DISCUSSION
The physical properties of the BPD analogs studied here do not differ significantly, nor does their
ability to produce singlet oxygen (Dr. Bensasson,
personal communication). The present study was
undertaken in an attempt to determine whether any
of the features which distinguish the analogs from
each other would contribute to greater phototoxicity
of malignant cells in vitro.
During the past two years, we have utilized a
rapid assay system which appears to correlate well
with clonogenic assays. The MTT test system
involves the colorimetric detection of active mitochondrial hydrogenases as a measure of cell
viability. The assay is rapid and easy to perform
and permits a reliable procedure for screening
photosensitizer activity. The MTT assay has been
reported by others to be useful in determining the
effects of chemotherapeutic drugs on human leukemia (Campling et al., 1988; Pieters et al., 1988) and
lung cells (Cole, 1986). We also found that the
most reliable procedure for establishing values for
phototoxic killing involved the incubation of treated
cells overnight before determining rates of killing.
In the study reported here, we have shown that this
procedure, using the M'IT assay, correlates closely
with more standard procedures involving [)H]thymidine incorporation (Fig. 3) and estimates of clonogenicity (Fig. 4), both of which also determine
viability of cells following the time required for
phototoxic killing. McHale and McHale (1988) have
also shown the correlation between the M7T and
the standard clonogenic assay while determining
human breast and melanoma cell survival following
1001
100 -
80
-
)
.
>
20
.
a
v
fn 6 0 -
a
.
t
I-
s
40-
:
0
2 0
4 0
60
80
100
CLONOGENIC ASSAY
Figure 4. Correlation between M'IT and clonogenic
assays. The percentage of KGl cells killed by exposure to
various concentrations of BPD-MA and light (5.4 Jim2)
was determined parallely by MTT and clonogenic assays.
Error bars indicate standard error.
-
-
BPD-DB
Photosensitizing potency of benzoporphyrin derivatives
100-
(A)
8060
-
40
-
100
0 % FCS
5 % FCS
10 % FCS
20 % FCS
10'
102
1
o3
Log. Conc. (ng/ml)
Log. Conc. (nglml)
Figure 6. Photosensitizing potency of BPD-MA (A) and
BPD-DA (B) toward M1 tumor cells in culture. The cells
were incubated with either of the photosensitizers in the
absence of serum or in the presence of 5, 10, or 20% of
serum for 1 h and then washed and exposed to light (5.4 J/
cm'). The percentage of cells killed was determined by
MTT assay at 18-24 h post exposure. Each point represents the average of 4-6 experiments. Error bars indicate
the standard error.
photoradiation therapy with HpD. Good correlation
between the MTT assay and ["Hluridine incorporation in human tumor cell lines after treatment with
doxorubicin or vindesine has also been reported
(Ford et al., 1988).
We have carried out many experiments in the
absence of serum. Our reasoning for performing
experiments in this way was based on our wish to
evaluate these compounds directly, in terms of their
action on cells, in the absence of any possible modifying effects introduced by serum proteins. Not
surprisingly, serum components were found to compete with the cells for photosensitizer; therefore,
more photosensitizer molecules were required for
the same photodynamic damage, while the mechanism of photosensitization (as judged from the paralleled slopes, Fig. 6) is the same. In the presence
of serum, the hierarchy of efficacy between BPD
analogs established in the absence of serum
remained unchanged.
We have shown here using the 4 BPD analogs that
relatively minor structural differences may affect the
photosensitizing activity of molecules. It was evident
that the position of the cyclohexadiene ring at the
A or B ring of porphyrin does not affect the photosensitizing efficiency of the isomers as much as the
presence of two acids or one acid and one ester
group at rings C and D of the porphyrin macrocycle.
However, the presence of the cyclohexadiene ring
499
fused to the porphyrin macrocycle may be responsible for relatively high photosensitizing potency of
BPD analogs. Morgan et al., (1987) have found that
photosensitizing potency was related to the presence
of "bulky substitution" in the form of a five or six
member ring at the reduced porphyrin macrocycle.
The major difference between mono and diacid
forms of BPD is their relative hydrophobicity and
lipophilicity. Although all BPD analogs have very
limited solubility in aqueous solutions, monoacids
are more soluble than diacids. Limited water solubility has been reported to make disulfonated zinc
phthalocyanine more efficient as a photosensitizer
than other forms, either insoluble or more soluble
(Brasseur et al., 1988). Also, BPD monoacids are
more lipophilic than the diacid forms. It is possible
that this feature could enhance their association and
uptake by cell membranes. It was reported earlier
by Kessel (1989) that BPD-MA was more readily
taken up by L1210 cells than was BPD-DA. BPD
diacids are more negatively charged than monoacid
forms in neutral aqueous solutions (or in serum). It
is not possible to speculate at this time as to whether
or not the net negative charge on cell surfaces influences relative uptake of the various BPD analogs.
Cellular uptake and localization of the BPD analogs
is currently under investigation.
Acknowledgements-The authors express their gratitude
to Ms. Kathy Lamb and Mr. Stephen Yip for their skillful
technical assistance.
Research supported in part by a grant from the Natural
Sciences and Engineering Research Council of Canada,
Grant No. 5-80268.
REFERENCES
Brasseur, N., H. Ali, R. Langlois and J. E. van Lier (1988)
Biological activities of phthalocyanines-IX. Photosensitization of V-79 Chinese hamster cells and EMT-6
mouse mammary tumor by selectively sulfonated zinc
phthalocyanines Photochem. Photobiol. 47, 705-71 1.
Campling, B. G . , J. Pym, P. R . Galbraith and S. P. C.
Cole (1988) Use of the MTT assay for rapid determination of chemosensitivity of human leukemic blast
cells. Leuk. Res. 12, 823-831.
Cole, S. P. C. (1986) Rapid chemosensitivity testing of
human lung tumor cells using the MTI' assay. Cancer
Chemother. Pharmacol. 17, 259-263.
Doiron D. R. (1984a) Photophysics and instrumentation
for porphyrin detection and activation. In Porphyrin
Localization und Treatment of Tumors (Edited by D .
R. Doiron and C. J. Gomer), p.41. Liss, New York.
Doiron, D. R., C. J. Gomer, S. W. Fountain and N. J .
Razum (1984b) Photophysics and dosimetry of photoradiation therapy. In Porphyrins in Tumor Phototherapy
(Edited by A. Andreoni and R. Cubeddu), pp. 281-291.
Plenum Press, New York.
Dougherty, T. J. (1986) Photosensitization of malignant
tumors. Semin. Surg. Oncol. 2, 24-37.
Dougherty, T. J. (1987) Studies on the structure of porphyrins contained in Photofrin 11. Photochem. Photobiol.
46, 569-573.
Evensen, J . F. and J . Moan (1987) A test of different
photosensitizers for photodynamic treatment of cancer
in a murine tumor model. Photochem. Photobiol. 46,
859-865.