PAPER www.rsc.org/pps | Photochemical & Photobiological Sciences

Comparative characterization of the efficiency and cellular pharmacokineticsof Foscan R©- and Foslip R©-based photodynamic treatment in human biliarytract cancer cell lines

Tobias Kiesslich,a Juergen Berlanda,a Kristjan Plaetzer,a Barbara Krammera and Frieder Berr*b

Received 4th December 2006, Accepted 22nd January 2007First published as an Advance Article on the web 26th February 2007DOI: 10.1039/b617659c

Due to the poor prognosis and limited management options for perihilar cholangiocarcinoma (CC) thedevelopment of alternatives for treatment is an important topic. Photodynamic therapy (PDT) withporfimer as palliative or neoadjuvant endoscopic treatment of non-resectable perihilar CC hasimproved quality of life and survival time, but cannot eradicate the primary tumors because ofinadequate tumoricidal depth (4 mm only around the tumor stenoses). The use of meta-tetrahydroxy-phenyl chlorin (mTHPC) and photoactivation at higher wavelengths (650–660 nm) provides hightumoricidal depth (10 mm) for PDT of pancreatic cancer and should yield similar tumoricidal depth inCC. This study investigates the photodynamic characteristics of mTHPC in solvent-based formulation(Foscan R©) and in liposomal (water soluble) formulation (Foslip R©) in an in vitro model system consistingof two biliary cancer cell lines (GBC, gall bladder cancer and BDC, bile duct cancer cells). Dark toxicity,photodynamic efficiency, time-dependent uptake and retention and intracellular localization of Foscan R©

and Foslip R© were studied. The results prove mTHPC as a potent photosensitizing agent with highphototoxic potential in biliary cancer cells as a concentration of 600 ng ml−1 and irradiation with1.5 J cm−2 (660 ± 10 nm) is sufficient for about 90% cell killing. Addition of foetal bovine serum (FBS)to the incubation medium and analysis of the uptake and phototoxic properties reveals that bothphotosensitizer formulations bind to serum protein fractions, i.e. no difference between Foscan R© andFoslip R© can be found in the presence of FBS. Laser scanning fluorescence microscopy indicates a similarpattern of perinuclear localization of both sensitizers. This study demonstrates the potential of mTHPCfor treatment of bile duct malignancies and provides evidence that Foslip R© is an equivalent water-solubleformulation of mTHPC that should ease intravenous application and thus clinical use of mTHPC.

Introduction

The prognosis of extrahepatic and perihilar bile duct cancer(cholangiocarcinoma, CC) is very poor since patients do notbecome symptomatic until the disease has progressed to anadvanced stage.1 The overall survival rate is less than 5%1–4 and theonly curative treatment—resective surgery—is just appropriate inabout 20% of cases and associated with an overall 5-year survivalof not more than 10–30% after curative resection in selectedseries.2,3 Palliation can be achieved in most patients by placementof one or more biliary stents, however, the prognosis remainspoor with median survival times between 4 and 6 months.1,3 Non-surgical oncological approaches such as intraluminal brachyther-apy, even in combination with external-beam radiotherapy, orchemotherapy could not be proven to provide a significant survivalbenefit according to a systematic review of over 65 differentstudies.5

Photodynamic therapy (PDT) has been used in prospectiveclinical studies for palliative treatment of non-resectable CC

aDepartment of Molecular Biology, University of Salzburg, Hellbrunner-strasse 34, 5020, Salzburg, AustriabDepartment of Internal Medicine I, Paracelsus Medical University, Lan-deskrankenanstalten, Muellner Hauptstrasse 48, 5020, Salzburg, Austria.E-mail: f.berr@salk.at; Fax: +43 (0)662/4482-881; Tel: +43 (0)662/4482-2802

with moderate life extension and better quality of life.6,7 PDTis based on the light-activation of a photosensitizing compound(photosensitizer) which afore has been taken up by the target tissue(tumor cells), resulting in production of reactive oxygen species(ROS) and cell damage/cell death;8–10 the severity of damage, i.e.the initial PDT dose, is most important in determining the mode ofcellular response (dose-dependent induction of cellular survival,apoptosis and necrosis).10,11

PDT using the photosensitizer Photofrin R© (porfimer) and laserlight at 632 nm wavelength is an established procedure for thepalliation of non-resectable hilar bile duct cancer6,7 and hassuccessfully been applied in a neoadjuvant fashion for the down-staging of non-resectable bile duct cancer with subsequent R0-resection.12,13 However, the tumoricidal depth of the Photofrin R©-PDT is limited to 4 mm tissue penetration in the tumor stenosisaround the laser light applicator.12,13 This is a major disadvantageof the procedure, because bile duct carcinomas originating fromthe epithelial layer extend usually about 6–8 mm into the bile ductwall/surrounding tissue by the time of diagnosis.12,13

Meta-tetrahydroxyphenyl chlorin (mTHPC; trade name:Foscan R©) is a powerful second-generation photosensitizer withpotent photochemical properties; the absorption spectrum allowsirradiation at wavelengths of about 652 nm which apparentlyachieves photoactivation of mTHPC for up to more than 10 mmtissue penetration in irresectable pancreatic cancer.14 Similar

This journal is © The Royal Society of Chemistry and Owner Societies 2007 Photochem. Photobiol. Sci., 2007, 6, 619–627 | 619

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

CView Article Online / Journal Homepage / Table of Contents for this issue

tumoricidal depth could achieve complete eradication of theprimary tumor in non-resectable CCs. To overcome the problemsof precipitation and adsorption (e.g. to endothelium of the injectedvein and to adjacent subcutaneous fat tissue) observed in somepatients after intravenous administration of Foscan R© a novelformulation, Foslip R©, makes use of liposomes to render thephotosensitizer water soluble.

In this study the potential of Foscan R© and Foslip R© for light-mediated destruction of human biliary tract cancer cells is evalu-ated by analysis of dark toxicity, phototoxicity, cellular uptake andintracellular distribution for each sensitizer. For selected parame-ters the influence of serum in the incubation medium is studied.

Experimental

Cell culture

A highly differentiated human gall bladder cancer cell line (GBC)and an undifferentiated human bile duct cancer cell line (BDC)15,16

were maintained in Dulbecco’s modified Eagle’s medium contain-ing 4.5 g l−1 glucose supplemented with 10 mM HEPES, 4 mML-glutamine, 1 mM Na-pyruvate, 100 U ml−1 penicillin, 0.1 mg ml−1

streptomycin and 10% (v/v) FBS (foetal bovine serum; all fromPAA-laboratories, Linz, Austria), in a humidified atmosphereat 37◦ C and 7.5% CO2.16 Cells were regularly split by trypsintreatment at 80–95% confluency. All experiments were carriedout with cells from passage number 5 to 25. For measurementof dark toxicity or phototoxicity, cells were seeded in transparent96-well microplates or 96-well microplates with black walls andclear bottom (all from Greiner Bio-One, Kremsmuenster, Austria),respectively, at densities of 2.5E4 (GBC) or 2E4 (BDC) cells perwell in 100 ll medium. Photosensitizer uptake and release werestudied with cell samples cultured in 24-well microplates (GreinerBio-One) at densities of 1.5E5 or 1.25E5 cells per well for GBC orBDC cells, respectively (0.5 ml medium). For confocal fluorescencemicroscopy, cells were seeded on 32 mm coverslips at a density of6E5 per dish (1.5 ml medium).

Photodynamic treatment

Foscan R© and Foslip R© were obtained from Biolitec AG (Jena,Germany) supplied as an ethanol–propylene glycol solution(Foscan R©, 4 mg ml−1) or as a lyophilized powder that wasreconstituted with sterile water to 1.3 mg ml−1 in the case ofFoslip R©; for working stocks, these solutions were diluted withethanol (Foscan R©) or water (Foslip R©) to 400 lg ml−1 (concentrationof the photoactive substance). Twenty-four hours after seeding thecells were washed twice with FBS-free medium (supplemented asabove) and incubated with serum-free medium containing eitherFoscan R© or Foslip R©; the final concentration of the sensitizers was400 ng ml−1 for analysis of phototoxicity, uptake, release, FBS-dependent uptake/phototoxicity and intracellular localization.When indicated, a range of concentrations of Foscan R© and Foslip R©

was added to the cells or FBS was added at concentrations asindicated in the figures. Based on the analysis of the time-dependent uptake of the photosensitizers (see Fig. 3 later), forall other experiments an incubation period of 20 h was chosenwhere the photosensitizer uptake approaches a plateau phase.Subsequent to a washing step with serum-free medium, sensitized

cells were irradiated from below the culture receptacles using ared-light illumination diode array (see Pieslinger et al.17 for details;kmax = 660 ± 10 nm) with a constant light power of 2.33 mW cm−2.Different fluences were studied by setting the irradiation time inthe range of 0.5–16.5 min (amounting to fluences in the range of0.07–2.31 J cm−2). Samples with sensitizer were handled undersubdued light conditions before and after irradiation.

Photosensitizer uptake and release

Analysis of the uptake/release of Foscan R© and Foslip R© wascarried out with cells sensitized as described above by afluorescence-activated cell sorter (FACS). For this purpose, thesamples were washed twice with phosphate buffered saline (PBS),harvested by trypsin treatment and transferred to FACS tubesat the respective time points. Immediately thereafter, the redfluorescence (FL-3 channel, kEX = 485 nm) was analyzed with aBecton-Dickinson FACSCalibur flow cytometer (BD Bioscience,Schwechat, Austria).

Dark and light-dependent cytotoxicity, MTT assay

For assessment of the dark and light-dependent cytotoxicity, cellswere incubated with the photosensitizers for 20 h and either notirradiated or irradiated as described above, respectively. Twenty-four hours post irradiation (p.i.) cellular survival was assessed bymeans of reduction of 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) to the insoluble blue formazancatalyzed by mitochondrial and other cellular dehydrogenases18

as described in ref. 19. The resulting absorbance, indicative ofremaining cellular activity, was read at 565 nm on a Spectrafluormicroplate photometer (Tecan, Salzburg, Austria).

Confocal fluorescence microscopy

Microscopic images of cellular fluorescence were acquired us-ing a Zeiss laser scanning fluorescence microscope (model:LSM510METER, Carl-Zeiss, Jena, Germany). Samples wereincubated for 20 h with 400 ng ml−1 Foscan R© or Foslip R©

and co-stained with 200 nM MitoTracker R© Green FM or1 lM LysoTracker R© Yellow-HCK-123 (Invitrogen, Lofer, Austria)30 min before image acquisition. Images were taken using a Plan-Apochromat 63×/1.4 Oil DIC objective with a kEX = 364 nm UV-laser and a kEX = 458 nm Ar-laser for fluorescence excitation for thephotosensitizers and MitoTracker R©/LysoTracker R©, respectively.Images were processed using the Zeiss LSM Image Browsersoftware (v. 4.0.0.157).

Statistical analysis

Unless indicated otherwise, data points represent mean values ±SEM of at least three independent experiments. Statistical signifi-cance was evaluated using Student’s t test (p < 0.05 and p < 0.01for significant and highly significant differences, respectively).

Results and discussion

Results

Dark toxicity. The cytotoxic effect of incubation with Foscan R©

and Foslip R© is shown in Fig. 1. Up to a final concentration of

620 | Photochem. Photobiol. Sci., 2007, 6, 619–627 This journal is © The Royal Society of Chemistry and Owner Societies 2007

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

C

View Article Online

Fig. 1 Dark toxicity. Gall bladder cancer cells (GBC, (a), (c)) and bile duct cancer cells (BDC, (b), (d)) were incubated with varying concentrations ofFoscan R© and Foslip R© in 0% (v/v) FBS, (a) and (b), or 10% (v/v) FBS-containing medium, (c) and (d), for 20 h, washed and after further 24 h incubationthe survival was measured (MTT test). Data represent mean ± SEM; * and ** indicate Foslip R© results significantly (p < 0.05) or highly significantly(p < 0.01) different from the corresponding Foscan R© data points, respectively.

2–3 lg ml−1 of Foscan R© the MTT signal remains at about 100%of untreated controls (without sensitizer and light); above thisconcentration, the signal decreases to about 25% and 10% forGBC and BDC cells, respectively (diagrams (a) and (b), mediumwithout serum). As indicated in the diagrams, the MTT signalof Foslip R©-treated cells is significantly higher than that of theFoscan R© samples; a decrease of MTT activity is observable only atthe highest concentration (100 lg ml−1) for both cell lines. Additionof 10% foetal bovine serum (FBS) (v/v, final concentration)slightly increases the dark toxicity of both sensitizers with adecrease of the MTT signal >95% at a Foscan R© concentrationabove approximately 20 lg ml−1; the survival curves are similar forboth cell lines and over again, Foslip R© is significantly less toxic tothe cells.

Light-induced cytotoxicity. The survival of GBC cell and BDCcell samples incubated with 400 ng ml−1 in serum-free mediumand irradiated with varying fluences is shown in Fig. 2 (diagrams(a) and (b)). At a fluence of 0.1 J cm−2 the MTT signal isabout 90–95% of untreated controls (75% for Foscan R©-treatedBDC cells). With increasing light dose the survival continuouslydecreases ending up at 8–16% for GBC cells at 2.3 J cm−2 and<5% for BDC cells at fluences >1.2 J cm−2. Similar results are

obtained when the fluence was held constant for all samplesat 1.5 J cm−2 and the photosensitizer concentration was variedbetween 23 and 600 ng ml−1 (Fig. 2, diagrams (c) and (d)). For allexperiments, PDT-induced reduction of the MTT activity is morepronounced for BDC cells and the survival of Foslip R©-treated cellsis significantly greater than that of Foscan R©-treated cells.

Time-dependent uptake of Foscan R© and Foslip R©. Analysis ofthe time-dependent photosensitizer uptake (Fig. 3) indicates acontinuous increase in fluorescence up to 30 h post incubationwhereby the signal of Foslip R© is about 50% and 65% of theFoscan R©-induced cellular fluorescence at 36 h for GBC and BDCcells (diagrams (a) and (b)), respectively. Addition of 10% FBS inthe incubation medium results in similar uptake dynamics of bothFoscan R© and Foslip R© ending up at about 40–50% and 60–65% forGBC and BDC cells, respectively (relative to the FBS-free Foscan R©

signal at 36 h).

FBS-dependent uptake of Foscan R© and Foslip R©. A more de-tailed exploration of the effect of FBS on the uptake of Foscan R©

and Foslip R© is shown in Fig. 4; at low FBS concentrations theFoscan R© signal remains at the level of serum-free samples butdecreases continuously with higher FBS-concentrations (to about35% and 45% of the initial signal for Foscan R© in GBC (diagram

This journal is © The Royal Society of Chemistry and Owner Societies 2007 Photochem. Photobiol. Sci., 2007, 6, 619–627 | 621

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

C

View Article Online

Fig. 2 Light-induced toxicity. (a), (b): GBC and BDC cells were incubated with 400 ng ml−1 Foscan R© and Foslip R© in serum-free medium for 20 h,washed and irradiated with varying fluences (660 ± 10 nm). (c), (d): GBC and BDC cells were incubated with varying concentrations of Foscan R© andFoslip R© in serum-free medium for 20 h, washed and irradiated with 1.5 J cm−2 (660 ± 10 nm). Twenty-four h p.i. the survival was measured (MTT test).Data represent mean ± SEM; * and ** indicate Foslip R© results significantly (p < 0.05) or highly significantly (p < 0.01) different from the correspondingFoscan R© data points, respectively.

(a)) and BDC cells (diagram (b)), respectively). In the case ofFoslip R©, addition of 2.5% FBS causes an increase of the signalfrom 65% to about 120% (GBC) and 105% (BDC) of the referencesignal (Foscan R©, FBS-free). With higher concentrations of FBSthe signal decreases to about 38% (GBC) and 40% (BDC) of theinitial Foscan R© signal. Statistical significant differences betweenFoscan R© and Foslip R© are found only for FBS-free samples.

FBS-dependent phototoxicity of Foscan R© and Foslip R©. Similarto the survival curves in Fig. 2, samples incubated with 0%, 5%or 10% FBS show a decrease of the MTT activity with increasinglight dose (Fig. 5). The cytotoxic effect is generally more profoundfor BDC cells. Statistical comparison reveals significant differencesbetween Foscan R©- and Foslip R©-treated cells only under serum-freeconditions; in this case the signal of Foscan R©-treated samples islower than the signal of Foslip R©-treated samples (both cell lines).

Time-dependent release of Foscan R© and Foslip R©. For analysisof the retention characteristics, samples were incubated for 20 hwith 400 ng ml−1 (0% FBS), washed twice with serum free-medium

and analyzed by FACS after 0, 24 and 48 h. As shown in Fig. 6 thesignal decreases to about 70% and 55% for GBC cells incubatedwith Foscan R© at 24 and 48 h post medium change, respectively; theremaining fluorescence of Foslip R© samples is about 10% lower atboth time points. In BDC cells, the release of the photosensitizersis generally less obvious as the Foscan R© signal decreases to about85% only after 48 h. Again, the release of Foslip R© is significantlygreater; at both time points, the signal is about 20% lowercompared to the samples incubated with Foscan R©.

Intracellular localization of Foscan R© and Foslip R©. Following20 h of incubation with Foscan R© and Foslip R© (without FBS),the samples were analyzed by laser-scanning fluorescence mi-croscopy. As shown in Fig. 7, both sensitizers predominantlylocalize in the cell’s perinuclear region with diffuse fluorescencein the remaining cell. There is no obvious difference betweenFoscan R©- and Foslip R©-treated BDC or GBC cells (Fig. 7). Whencompared to the organelle-specific fluorescent dyes MitoTracker R©

(for mitochondria) and LysoTracker R© (for lysosomes) intracellularregions with overlapping fluorescent signatures can be found

622 | Photochem. Photobiol. Sci., 2007, 6, 619–627 This journal is © The Royal Society of Chemistry and Owner Societies 2007

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

C

View Article Online

Fig. 3 Photosensitizer uptake. GBC (a) and BDC (b) were incubated with400 ng ml−1 Foscan R© and Foslip R© in 0% or 10% (v/v) FBS-containingmedium. At 0, 6, 12, 24, 30 and 36 h post incubation, the cells were washedand analyzed for photosensitizer fluorescence by flow cytometry. Datarepresent mean ± Gaussian error related to the 36 h Foscan R© signal as astandard (=100%).

(indicated by the yellowish colour in the superimposed images).However, distinctly coloured mitochondria and lysosomes (green)do not have an obvious counterpart in the images showingphotosensitizer fluorescence.

Discussion

Photodynamic therapy has been proposed as an innovative ap-proach for neo-adjuvant or palliative treatment of non-resectablecholangiocarcinoma6–8,13 and recent results from clinical studiesprove an increased survival time6,7,20,21 and quality of life7,20–24

of patients with non-resectable CC treated with PDT. AlthoughPhotofrin R©-PDT is regarded as a selective method for efficientdestruction of CC within a superficial layer of 4 mm5,13 the useof other photosensitizing agents with deeper tissue penetrationwould be desirable since CC are usually characterized by athickness of 6–8 mm by the time of diagnosis.13 For non-resectablepancreatic cancer a tumoricidal tissue penetration of 10 mm hasbeen achieved14 using meta-tetrahydroxyphenyl chlorin (Foscan R©;photoactivation at about 652 nm). We therefore studied the

Fig. 4 FBS-dependent photosensitizer uptake. GBC (a) and BDC(b) were incubated with 400 ng ml−1 Foscan R© and Foslip R© in medium sup-plemented with different concentrations of FBS. Twenty h post incubation,the cells were washed and analyzed for photosensitizer fluorescence by flowcytometry. Data represent mean ± SEM related to the 0% FBS-Foscan R©

signal as a standard (=100%); * indicates Foslip R© results significantly(p < 0.05) different from the corresponding Foscan R© data points.

effectiveness and properties of this compound and a water-solubleformulation (Foslip R©) in an experimental biliary tract cancermodel consisting of a highly differentiated cancer cell line derivedfrom a gall bladder carcinoma (GBC) and a cell line derived froman undifferentiated bile duct carcinoma (BDC).15,16

One of the properties of an ideal photosensitizing agent is alow dark toxicity, i.e. a low cytotoxic effect without irradiation isadvantageous. After an incubation period of 20 h Foscan R© is toxicto BDC or GBC cells above a concentration of 2 lg ml−1 (Fig. 1).For Foslip R©, only the highest concentration used in this experi-ments (100 lg ml−1) shows dark toxicity. These results indicate ahigher dark toxicity as compared to a study by Oertel et al.16 whoused hematoporphyrin derivative and a bacteriochlorine (THP,tetrakis-pyridyl-tetrahydroporphyrin) as sensitizers and the samecell model. However, in the present study the survival was assessed24 h after the incubation period and this could partially explainthe greater dark toxicity since a portion of cells may die fromapoptotic cell death and this is manifested as a reduced MTT

This journal is © The Royal Society of Chemistry and Owner Societies 2007 Photochem. Photobiol. Sci., 2007, 6, 619–627 | 623

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

C

View Article Online

Fig. 5 FBS-dependent phototoxicity. GBC (a) and BDC (b) wereincubated with 400 ng ml−1 Foscan R© and Foslip R© in medium supplementedwith different concentrations of FBS. Twenty h post incubation, the cellswere washed and irradiated with 0.14, 0.42, 0.7 and 0.98 J cm−2 (660 ±10 nm). Twenty-four h p.i. the cells were analyzed for MTT activity. Datarepresent mean ± SEM related to untreated control samples; * and **indicate Foslip R© results significantly (p < 0.05) or highly significantly(p < 0.01) different from the corresponding Foscan R© data points,respectively.

signal only after a certain period of time;25 additionally, dilutionof the original stock solution (4 mg ml−1) to 100 lg ml−1 resultsin final concentration of the solvent (ethanol–propylene glycol

Fig. 6 Time-dependent photosensitizer retention. GBC and BDC wereincubated with 400 ng ml−1 Foscan R© and Foslip R© in FBS-free medium.Twenty h post incubation, the cells were washed twice and incubatedwith 0% FBS-containing medium without sensitizer. At 0, 24 and 48 hpost medium change the cellular fluorescence was analyzed by flowcytometry. Data represent mean ± SEM related to the correspondinginitial fluorescence signal after the 20 h incubation period; * and ** indicateFoslip R© results significantly (p < 0.05) or highly significantly (p < 0.01)different from the corresponding Foscan R© data points, respectively.

solution in the case of Foscan R©) of 2.5% (v/v) which might bealready toxic itself, without any specific effect of the compound.

For further experimentation, a concentration of 400 ng ml−1 forboth, Foscan R© and Foslip R© was chosen at which no dark toxicityis observed and at which a sufficient cellular uptake is achieved;preliminary experiments (data not shown) indicated that lowerdoses caused different cytotoxic effects at varying cell densitieswhich is not advantageous for in vitro testing. Investigation of thecytotoxic potential after irradiation with 660 ± 10 nm proves thereputation of Foscan R© as an effective photosensitizer;26 as shownin Fig. 2, a reduction of the cellular survival of 80–90% can beachieved at 24 h p.i. with a fluence >2.25 J cm−2. Similar survivalcurves are obtained when the concentration of Foscan R© or Foslip R©

is varied instead of the fluence (Fig. 2, (c), (d)). It is important tonote, that in these experiments cells were incubated with the drugsin serum-free medium and under these conditions, a generallyless profound effect is found for Foslip R©. Table 1 summarizes theefficiency of Foscan R© and Foslip R© as photosensitizers on GBC

Table 1 Photosensitizing efficiency of Foscan R© and Foslip R© in human biliary tract cancer cell lines

Dark toxicitya/ng ml−1 Phototoxicityb/ng ml−1 IC50

Photosensitizer Cell line LD90, DT LD50, DT LD90, PDT LD50, PDT (LD50, DT/LD50, PDT)

Foscan R© GBC N.a. 16000 222 45 ∼356BDC 105000 18000 110 44 ∼410

Foslip R© GBC N.a. N.a. 675 61 —BDC N.a. N.a. 195 56 —

a LD90, DT and LD50, DT indicate the concentration (ng ml−1) of Foscan R© or Foslip R© that is lethal for 90% and 50% of the cells subsequent to 20 h incubationin the dark, respectively (serum-free conditions). b LD90, PDT and LD50, PDT indicate the concentration (ng ml−1) of Foscan R© or Foslip R© that is lethal for 90%and 50% of the cells subsequent to 20 h incubation in the dark and irradiation with a fluence of 1.5 J cm−2 at a wavelength of 660 ± 10 nm, respectively(serum-free conditions). N.a. = not applicable.

624 | Photochem. Photobiol. Sci., 2007, 6, 619–627 This journal is © The Royal Society of Chemistry and Owner Societies 2007

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

C

View Article Online

Fig. 7 Intracellular photosensitizer localization. GBC (a) and BDC(b) were incubated with 400 ng ml−1 Foscan R© and Foslip R© in mediumwithout FBS. Twenty h post incubation, the cells were washed and mito-chondria or lysosomes were stained by addition of 200 nM MitoTracker R©

or 1 lM LysoTracker R© (final concentration), respectively. Confocal fluo-rescence microscopy was performed as described in ‘Experimental’; imagesin the three columns represent fluorescence patterns of Foscan R©/Foslip R©

(red), MitoTracker R©/LysoTracker R© (green) and superimposed images,respectively.

and BDC cells. The IC50 value (lethal dose (50% cell killing) in thedark divided by lethal dose (50% cell killing) following irradiation(1.5 J cm−2), LD50, DT/LD50, PDT) of about 356 and 410 for GBC and

BDC cells, respectively, indicates a high photosensitizing potentialof Foscan R© in this cell model. For Foslip R© the IC50 is even higher(in this case it could not be calculated since the LD50, DT was noteven reached with the highest Foslip R© concentration).

Analysis of the cellular uptake of the photosensitizers underserum-free conditions indicates a roughly linear temporal dynam-ics with a greater cellular fluorescence of the Foscan R© samplescompared to Foslip R©. Consistent with previous studies27–29 theuptake of Foscan R© is a slow process. After 36 h of incubation aplateau of the cellular fluorescence can be observed (Fig. 3). Inthese experiments an incubation medium supplemented with 10%FBS was also used and under these circumstances Foscan R©- andFoslip R©-treated GBC and BDC cells show comparable uptake atall time points studied. A more detailed analysis of the influenceof serum on the sensitizer uptake revealed that addition of 2.5%FBS causes an enhanced uptake of Foslip R© (factor: 1.5) whereasFoscan R©-induced fluorescence is not altered (Fig. 4) in bothcell lines. At higher FBS concentrations the signal decreases forboth sensitizers and no significant differences can be found. Thistrend is also evident in irradiated samples (Fig. 5) that wereincubated in varying FBS concentrations; only for serum-freesamples a significant difference between Foscan R© and Foslip R© canbe found. Based on these observations it can be concluded thatboth sensitizers attach to a serum constituent and this associationaffects the efficiency of photosensitizer uptake. Furthermore,the differences between Foscan R© and Foslip R© with respect tophotosensitizing efficiency and cellular uptake are only apparent inserum-free media where the different chemical composition andsolubility influences the internalization mechanism. Since bothformulations of mTHPC behave almost identical by the timeserum (regardless of the concentration) is added to the incubationmedium it can be further assumed that the portion of activesubstance that is bound by serum components and transferredinto the cells is irrespective of its initial chemical formulation.This conclusion is supported by several studies reporting anassociation between mTHPC and different plasma fractions suchas high density lipoprotein (HDL), low density lipoprotein (LDL)and albumin, although the relative importance of these fractionsdiffer between these studies.30–33 The association between LDL andphotosensitizer could contribute to its tumor-localizing propertiessince tumor cells are known to bear extensive amounts of LDLreceptors.34,35

The retention characteristics are also different for Foscan R© andFoslip R©, i.e. Foslip R© is released somewhat more rapidly from bothcell lines compared to Foscan R© (Fig. 6). This slightly more rapidefflux could be related to cellular efflux of liposomal mTHPC,if some of the mTHPC still had resided in form of liposomesin exocytotic vesicles. The cause for slightly more rapid effluxobserved with Foslip R© remains speculative, but the retention ofmTHPC was 48% in GBC cells and beyond 70% in BDC cells, highenough as cellular enrichment suitable for PDT in both cell lines.

Confocal fluorescent microscopy images of the cellular pho-tosensitizer fluorescence show no apparent differences betweenFoscan R© and Foslip R© and the cell lines used. It seems difficult toprovide evidence for an organelle-specific localization of Foscan R©

or Foslip R© based on comparison with mitochondrial or lysosomallocalizing dyes (Fig. 7). The localization characteristics may bedifferent at earlier time points after incubation; however, therecommended protocol for clinical use of Foscan R© suggests a

This journal is © The Royal Society of Chemistry and Owner Societies 2007 Photochem. Photobiol. Sci., 2007, 6, 619–627 | 625

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

C

View Article Online

drug-light interval of 96 h and the 20 h standard incubationperiod in this study was chosen also for localization studiesto allow comparison with the remaining results presented here.As mentioned in Leung et al.36 the intracellular localization ofmTHPC seems to be dependent on the cell model and incuba-tion parameters used. Consistent with this several publicationsreport localization of mTHPC in endoplasmatic reticulum (HT29(human colon adenocarcinoma) cells,37 MCF-7 (human breastcancer) cells38,39), Golgi apparatus (MCF-7 cells38,39), lysosomes(Colo 201 (human colon adenocarcinoma) cells36), mitochondria(HT29 (human colon adenocarcinoma) cells,37 NPC (humannasopharyngeal carcinoma) cells40 and murine myeloid leukaemiacells, p38831/M1 and JCS41) and even the nuclear envelope (mousemammary sarcoma (EMT6) cells42).

Conclusions

The present study compares the efficiency and cellular propertiesof Foscan R© and its liposomal formulation, Foslip R©, using twohuman biliary tract cancer cell lines as a model system; the resultscan be summarized as follows: (i) Foslip R© displays a markedlyreduced toxicity in the absence of light (dark toxicity) comparedto Foscan R©; (ii) under serum-free conditions less Foslip R© is takenup by both cell lines and this difference is also seen in the resulton the phototoxicity; and (iii) addition of serum, however, resultsin almost identical behaviour of both formulations with respectto uptake and phototoxicity. Taken together, both formulationsof mTHPC represent powerful and efficient photosensitizers forhuman biliary tract cancer cells. Given the similar behaviourin serum-supplemented media (which is more according to thephysiological situation when injected intravenously) of Foslip R©

and Foscan R© and the good water-solubility of Foslip R©, the use ofthe latter presents itself as an interesting alternative to previouslyemployed sensitizers due to higher penetration depths of theactivating light wavelengths. It can be considered—althoughthis has to be validated in further in vivo experiments—thatunder physiological conditions, Foslip R© displays tumor-localizingproperties similar to Foscan R© and comparable efficiency, retentionand intracellular localization.

Abbreviations

BDC, bile duct cancer cells; CC, cholangiocarcinoma; FACS,fluorescence-activated cell sorter; FBS, foetal bovine serum;GBC, gall bladder cancer cells; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid; HDL, high density lipoprotein;LD50, DT, lethal doses (50% cell killing), dark toxicity; LD50, PDT,lethal doses (50% cell killing), photodynamic toxicity; LDL, lowdensity lipoprotein; mTHPC, meta-tetrahydroxyphenyl chlorin;MTT, 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazoliumbromide; PBS, phosphate buffered saline; PDT, Photodynamictherapy/treatment; p.i., post irradiation; ROS, reactive oxygenspecies; THP, tetrakis-pyridyl-tetrahydroporphyrin.

Acknowledgements

This work was supported by the Medizinische Forschungsge-sellschaft Salzburg and the Forschungsburo of the ParacelsusMedical University Salzburg (project number 05/02/010). Juergen

Berlanda is supported by a research grant of the Dean ofthe Faculty of Natural Sciences (University of Salzburg). Theauthors are grateful to the Biolitec AG (Jena, Germany) forproviding the photosensitizing agents and to Dr Susanna Zierlerfor confocal microscopic imaging (Department of OrganismicBiology, University of Salzburg).

References

1 P. de Groen, G. Gores, N. LaRusso, L. Gunderson and D. Nagorney,Biliary Tract Cancers., N. Engl. J. Med., 1999, 341, 1368–1378.

2 L. Ayaru, S. G. Bown and S. P. Pereira, Photodynamic therapy forpancreatic and biliary tract carcinoma, Int. J. Gastrointest. Cancer,2005, 35, 1–13.

3 K. N. Lazaridis and G. J. Gores, Cholangiocarcinoma, Gastroenterol-ogy, 2005, 128, 1655–1667.

4 G. D. Leonard and E. M. O’Reilly, Biliary tract cancers: currentconcepts and controversies, Expert. Opin. Pharmacother., 2005, 6, 211–223.

5 M. Hejna, M. Pruckmayer and M. Raderer, The role of chemotherpyand radiation in the management of biliary cancer: a review of theliterature, Eur. J. Cancer, 1998, 34, 977–986.

6 F. Berr, Photodynamic therapy for cholangiocarcinoma, Semin. LiverDis., 2004, 24, 177–187.

7 M. E. Ortner, K. Caca, F. Berr, J. Liebetruth, U. Mansmann, D.Huster, W. Voderholzer, G. Schachschal, J. Mossner and H. Lochs, Suc-cessful photodynamic therapy for nonresectable cholangiocarcinoma:a randomized prospective study, Gastroenterology, 2003, 125, 1355–1363.

8 S. B. Brown, E. A. Brown and I. Walker, The present and future roleof photodynamic therapy in cancer treatment, Lancet Oncol., 2004, 5,497–508.

9 D. E. Dolmans, D. Fukumura and R. K. Jain, Photodynamic therapyfor cancer, Nat. Rev. Cancer, 2003, 3, 380–387.

10 K. Plaetzer, T. Kiesslich, T. Verwanger and B. Krammer, The Modesof Cell Death Induced by PDT: An Overview, Med. Laser Appl., 2003,18, 7–19.

11 K. Plaetzer, T. Kiesslich, C. B. Oberdanner and B. Krammer, Apoptosisfollowing photodynamic tumor therapy: induction, mechanisms anddetection, Curr. Pharm. Des., 2005, 11, 1151–1165.

12 F. Berr, A. Tannapfel, P. Lamesch, S. Pahernik, M. Wiedmann,U. Halm, A. E. Goetz, J. Mossner and J. Hauss, Neoadjuvantphotodynamic therapy before curative resection of proximal bile ductcarcinoma, J. Hepatol., 2000, 32, 352–357.

13 M. Wiedmann, K. Caca, F. Berr, I. Schiefke, A. Tannapfel, C.Wittekind, J. Mossner, J. Hauss and H. Witzigmann, Neoadjuvantphotodynamic therapy as a new approach to treating hilar cholangio-carcinoma: a phase II pilot study, Cancer, 2003, 97, 2783–2790.

14 S. G. Bown, A. Z. Rogowska, D. E. Whitelaw, W. R. Lees, L. B. Lovat, P.Ripley, L. Jones, P. Wyld, A. Gillams and A. W. Hatfield, Photodynamictherapy for cancer of the pancreas, Gut, 2002, 50, 549–557.

15 P. P. Purdum, 3rd, A. Ulissi, P. B. Hylemon, M. L. Shiffman and E. W.Moore, Cultured human gallbladder epithelia. Methods and partialcharacterization of a carcinoma-derived model, Lab. Invest., 1993, 68,345–353.

16 M. Oertel, S. I. Schastak, A. Tannapfel, R. Hermann, U. Sack,J. Mossner and F. Berr, Novel bacteriochlorine for high tissue-penetration: photodynamic properties in human biliary tract cancercells in vitro and in a mouse tumour model, J. Photochem. Photobiol.,B, 2003, 71, 1–10.

17 A. Pieslinger, K. Plaetzer, C. B. Oberdanner, J. Berlanda, H. Mair, B.Krammer and T. Kiesslich, Characterization of a simple and homo-geneous irradiation device based on light-emitting diodes: A possiblelow-cost supplement to conventionallight sources for photodynamictreatment, Med. Laser Appl., 2006, 21, 277–283.

18 R. J. Gonzalez and J. B. Tarloff, Evaluation of hepatic subcellularfractions for Alamar blue and MTT reductase activity, Toxicol. InVitro, 2001, 15, 257–259.

19 J. Berlanda, T. Kiesslich, C. B. Oberdanner, F. J. Obermair, B.Krammer and K. Plaetzer, Characterization of apoptosis induced byphotodynamic treatment with hypericin in A431 human epidermoidcarcinoma cells, J. Environ. Pathol. Toxicol. Oncol., 2006, 25, 173–188.

626 | Photochem. Photobiol. Sci., 2007, 6, 619–627 This journal is © The Royal Society of Chemistry and Owner Societies 2007

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

C

View Article Online

20 M. A. Ortner, J. Liebetruth, S. Schreiber, M. Hanft, U. Wruck, V. Fusco,J. M. Muller, H. Hortnagl and H. Lochs, Photodynamic therapy ofnonresectable cholangiocarcinoma, Gastroenterology, 1998, 114, 536–542.

21 T. Zoepf, R. Jakobs, J. C. Arnold, D. Apel, A. Rosenbaum andJ. F. Riemann, Photodynamic therapy for palliation of nonresectablebile duct cancer-preliminary results with a new diode laser system,Am. J. Gastroenterol., 2001, 96, 2093–2097.

22 F. Berr, M. Wiedmann, A. Tannapfel, U. Halm, K. R. Kohlhaw, F.Schmidt, C. Wittekind, J. Hauss and J. Mossner, Photodynamic therapyfor advanced bile duct cancer: evidence for improved palliation andextended survival, Hepatology, 2000, 31, 291–298.

23 F. L. Dumoulin, T. Gerhardt, S. Fuchs, C. Scheurlen, M. Neubrand,G. Layer and T. Sauerbruch, Phase II study of photodynamic therapyand metal stent as palliative treatment for nonresectable hilar cholan-giocarcinoma, Gastrointest. Endosc., 2003, 57, 860–867.

24 A. Rumalla, T. H. Baron, K. K. Wang, G. J. Gores, L. M. Stadheim andP. C. de Groen, Endoscopic application of photodynamic therapyfor cholangiocarcinoma, Gastrointest. Endosc., 2001, 53, 500–504.

25 K. Plaetzer, T. Kiesslich, B. Krammer and P. Hammerl, Characteri-zation of the cell death modes and the associated changes in cellularenergy supply in response to AlPcS4-PDT, Photochem. Photobiol. Sci.,2002, 1, 172–177.

26 R. R. Allison, G. H. Downie, R. Cuenca, X.-H. Hu, C. J. Childs andC. H. Sibata, Photosensitizers in clinical PDT, Photodiagn. Photodyn.Ther., 2004, 1, 27–42.

27 L. Morlet, V. Vonarx-Coinsmann, P. Lenz, M. T. Foultier,L. X. de Brito, C. Stewart and T. Patrice, Correlation betweenmeta(tetrahydroxyphenyl)chlorin (m-THPC) biodistribution and pho-todynamic effects in mice, J. Photochem. Photobiol., B, 1995, 28, 25–32.

28 Q. Peng, J. Moan, L. W. Ma and J. M. Nesland, Uptake, localization,and photodynamic effect of meso-tetra(hydroxyphenyl)porphine andits corresponding chlorin in normal and tumor tissues of mice bearingmammary carcinoma, Cancer Res., 1995, 55, 2620–2626.

29 F. Wierrani, D. Fiedler, G. Schnitzhofer, J. C. Stewart, K. Gharehbaghi,M. Henry, W. Grin, W. Grunberger and B. Krammer, A new approachto cancer therapy due to appropriate uptake and retention kinetics ofmeta-tetrahydroxy-phenylchlorin in a human fibroblast cell line, CancerBiochem. Biophys., 1996, 15, 171–176.

30 H. J. Hopkinson, D. I. Vernon and S. B. Brown, Identification andpartial characterization of an unusual distribution of the photosensi-tizer meta-tetrahydroxyphenyl chlorin (temoporfin) in human plasma,Photochem. Photobiol., 1999, 69, 482–488.

31 D. Kessel, Transport and localisation of m-THPC in vitro, Int. J. Clin.Pract., 1999, 53, 263–267.

32 A. T. Michael-Titus, R. Whelpton and Z. Yaqub, Binding of temoporfinto the lipoprotein fractions of human serum, Br. J. Clin. Pharmacol.,1995, 40, 594–597.

33 S. Sasnouski, V. Zorin, I. Khludeyev, M. A. D’Hallewin, F. Guilleminand L. Bezdetnaya, Investigation of Foscan interactions with plasmaproteins, Biochim. Biophys. Acta, 2005, 1725, 394–402.

34 T. J. Dougherty, Photosensitizers: therapy and detection of malignanttumors, Photochem. Photobiol., 1987, 45, 879–889.

35 R. A. Firestone, Low-density lipoprotein as a vehicle for targetingantitumor compounds to cancer cells, Bioconjugate Chem., 1994, 5,105–113.

36 W. N. Leung, X. Sun, N. K. Mak and C. M. Yow, Photodynamic effectsof mTHPC on human colon adenocarcinoma cells: photocytotoxicity,subcellular localization and apoptosis, Photochem. Photobiol., 2002,75, 406–411.

37 V. O. Melnikova, L. N. Bezdetnaya, C. Bour, E. Festor, M. P. Gramain,J. L. Merlin, A. Potapenko and F. Guillemin, Subcellular localizationof meta-tetra (hydroxyphenyl) chlorin in human tumor cells subjectedto photodynamic treatment, J. Photochem. Photobiol., B, 1999, 49, 96–103.

38 M. H. Teiten, L. Bezdetnaya, P. Morliere, R. Santus and F. Guillemin,Endoplasmic reticulum and Golgi apparatus are the preferential sitesof Foscan localisation in cultured tumour cells, Br. J. Cancer, 2003, 88,146–152.

39 M. H. Teiten, S. Marchal, M. A. D’Hallewin, F. Guillemin and L.Bezdetnaya, Primary photodamage sites and mitochondrial eventsafter Foscan photosensitization of MCF-7 human breast cancer cells,Photochem. Photobiol., 2003, 78, 9–14.

40 C. M. Yow, J. Y. Chen, N. K. Mak, N. H. Cheung and A. W. Leung,Cellular uptake, subcellular localization and photodamaging effect oftemoporfin (mTHPC) in nasopharyngeal carcinoma cells: comparisonwith hematoporphyrin derivative, Cancer Lett., 2000, 157, 123–131.

41 J. Y. Chen, N. K. Mak, C. M. Yow, M. C. Fung, L. C. Chiu,W. N. Leung and N. H. Cheung, The binding characteristics andintracellular localization of temoporfin (mTHPC) in myeloid leukemiacells: phototoxicity and mitochondrial damage, Photochem. Photobiol.,2000, 72, 541–547.

42 T. H. Foster, B. D. Pearson, S. Mitra and C. E. Bigelow, Fluorescenceanisotropy imaging reveals localization of meso-tetrahydroxyphenylchlorin in the nuclear envelope, Photochem. Photobiol., 2005, 81, 1544–1547.

This journal is © The Royal Society of Chemistry and Owner Societies 2007 Photochem. Photobiol. Sci., 2007, 6, 619–627 | 627

Dow

nloa

ded

by T

he U

nive

rsity

of

Mel

bour

ne L

ibra

ries

on

16/0

5/20

13 1

6:16

:35.

Pu

blis

hed

on 2

6 Fe

brua

ry 2

007

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/B61

7659

C

View Article Online