ARTICLE IN PRESS

0944-7113/$ - se

doi:10.1016/j.ph

�CorrespondSciences, P. J. S

E-mail addr

Phytomedicine 14 (2007) 172–178

www.elsevier.de/phymed

Histomorphological changes in murine fibrosarcoma after hypericin-based

photodynamic therapy

Nikita Bobrova, Ivan Cavargab, Frantisek Longauera, Silvia Rybarovaa,Peter Fedorockoc, Peter Brezania, Pavol Miskovskyc,d,�, Ladislav Mirossaya,Jan Stubnaa

aFaculty of Medicine, P. J. Safarik University, Kosice, SlovakiabMedical School Hospital of L.Pasteur, Kosice, SlovakiacFaculty of Sciences, P. J. Safarik University, Kosice, SlovakiadInternational Laser Center, Bratislava, Slovakia

Received 23 December 2005; accepted 7 June 2006

Abstract

Histomorphological changes in murine fibrosarcoma after photodynamic therapy (PDT) based on the naturalphotosensitizer hypericin were evaluated. C3H/DiSn mice were inoculated with fibrosarcoma G5:1:13 cells. When thetumour reached a volume of 40–80mm3 the mice were intraperitoneally injected with hypericin, either in a single dose(5mg/kg; 1 or 6 h before laser irradiation) or two fractionated doses (2.5mg/kg; 6 and 1 h before irradiation with laserlight; 532 nm, 70mW/cm2, 168 J/cm2).

All groups of PDT-treated animals with single and fractionated hypericin dosing presented primary vascularreactions including vascular dilatation, congestion, thrombosis and oedema. Two hours after PDT there were necroticchanges with small, rather focal appearance. One day after therapy the necrotic areas were enhanced, often affecting acomplete superficial layer of tumour tissue. Necrotic areas were accompanied with inflammation and haemorrhages.r 2006 Elsevier GmbH. All rights reserved.

Keywords: Hypericin; Cancerotherapy; Photodynamic therapy; Tumour

Introduction

Photodynamic therapy (PDT) involves the activationof a photosensitizer with light of the appropriatewavelength to produce reactive oxygen species, includ-ing singlet oxygen, which produces focal damage totissue, e.g. tumour masses.

e front matter r 2006 Elsevier GmbH. All rights reserved.

ymed.2006.09.017

ing author. Department of Biophysics, Faculty ofˇ afarik University, Jesenna 5, 04154 Kosice, Slovakia.

ess: misko@kosice.upjs.sk (P. Miskovsky).

PDT has been approved in several countries fortreating tumours (Dougherty et al., 1998). Alongside thewell-established photosensitizers such as porphyrins,phtalocyanins and chlorins, hypericin is a modern anti-cancer drug which exhibits promising photodynamictherapeutical properties. Hypericin is a naturally occur-ring polycyclic aromatic naphtodiantrone, which isisolated from the plant Hypericum genus. Its photo-cytotoxic activity has been tested in several experimentalstudies in vivo and in vitro (Chung et al., 2000; Chenet al., 2001b; Cavarga et al., 2001; Blank et al., 2002;Solar et al., 2002; Hopfner et al., 2003).

ARTICLE IN PRESSN. Bobrov et al. / Phytomedicine 14 (2007) 172–178 173

Our recent study on murine fibrosarcoma modelshowed that fractionated administration of hypericinassociated with the PDT can produce better therapeuticresponse than single administration of the drug.Histomorphological aspects of this preliminary reportwill be analysed in this paper.

Materials and methods

Photosensitizer

Hypericin was purchased from Molecular Probes,Eugene, Oregon, USA. Hypericin was dissolved in amixture of ethanol and glycerol (1:1, vol/vol). Then asolution of 20% PEG 400 in PBS (vol/vol) was added tothe final concentration of hypericin 0.1% (wt/vol).Hypericin was prepared in subdued light conditions.The drug was used immediately after preparation.

Animals

Male C3H/DiSn inbred mice, 8–10 weeks old (weigh-ing E20 g), were obtained from VELAZ, Prague, CzechRepublic. The animals were quarantined for a period of2 weeks and were housed in rodent cages with 4–5animals per cage at about 23 1C, and they were givenVelaz/Altromin 1320 St lab chow and tap water acidifiedto pH 2.4 ad libitum. During therapy experiments, themice were housed in cages under subdued lightconditions. The research was conducted according tothe principles enunciated in the ‘‘Guide for the Care andUse of Laboratory Animals’’, prepared by the StateVeterinary Office of the Slovak Republic.

Cell line and tumour response studies

The N-methyl-N0-nitro-N-nitrosoguanidine-inducedG:5:1:13 fibrosarcoma cell line (H-2k) was kindlyprovided by Dr. Margaret L. Kripke (University ofTexas, M.D. Anderson Cancer Center, Houston, Texas,USA). G:5:1:13 cells were maintained in RPMI-1640culture medium at 37 1C with 5% CO2 atmosphere andwere used for experiments during the exponentialgrowth phase. Tumour cells were harvested by trypsi-nization, washed twice with serum-free medium andadjusted to the concentration of 8� 105 cells/ml (Hoferet al., 2003). Anaesthetized animals were injected with1� 105 viable tumour cells/mouse s.c. into the depilatedparamedian lumbar region. Incidence and approximatetumour size were recorded weekly throughout theexperimental period. Tumour location was deter-mined by palpation. Animals were euthanized forhistological observation 2 and 24 h after PDT bycervical dislocation.

Light delivery

PDT was performed using light provided by solid-state laser. The excitation wavelength used in theexperiments was 532 nm. The light dose of 168 J/cm2

was delivered at the fluence rate 70mW/cm2. Theirradiation spot centred on the tumour region was l cmin diameter.

Photodynamic therapy protocol

Mice bearing G:5:1:13 tumours were randomized intosix groups of 5–9 animals per group for the followingtreatments:

(1)

No treatment. (2) Laser light. (3) Hypericin alone i.p. (5mg/kg). (4) Single dose of hypericin i.p. 1 h before laser

irradiation (5mg/kg).

(5) Single dose of hypericin i.p. 6 h before laser

irradiation (5mg/kg).

(6) Fractionated dose of hypericin i.p. 2.5mg/kg 6 h and

2.5mg/kg 1 h before laser irradiation.

Mice were treated with hypericin when the tumourreached a volume of 40–80mm3 (�17 days afterinoculation). Hypericin was administered intraperitone-ally at a volume 200 ml of solution. With delays of 1 or6 h after application in treatment groups 4–6, tumourswere irradiated with laser light. Immediatelyprior to irradiation the mice were anaesthetized withNarkamon/Rometar. In order to avoid photosensitizeractivation by ambient light, animals were kept incovered cages in subdued light conditions after hypericinadministration.

Histomorphological investigation

Complex preparates of tumour and surroundingtissues of the animals’ dorsal area were precisely taken,fixed in 4% phosphate-buffered formol, subsequentlyprocessed using the routine histological techniqueand embedded in paraffin. Slices were stained withhaematoxylin and eosin and examined throughSoligor StereoZoom microscope and Leica labor-atory microscope. For hypericin distribution study byfluorescent microscopy (Olympus DP50; excitationfilter 390–440 nm, emission filter 610–685 nm), thetumour tissue was prepared by means of frozen tissuesection.

ARTICLE IN PRESS

Fig. 1. Macroappearance of intact MFS of control group (A). Tumour focal necroses at a depth of approximately 3mm under the

skin. The group with the application of a single dose of hypericin 6 h before laser irradiation (2 h interval after PDT) (B).

N. Bobrov et al. / Phytomedicine 14 (2007) 172–178174

Results

Control tumours (no treatment, laser light, hypericin

alone)

Because no significant differences in the histomor-phological picture were observed, data from all controlsare therefore referred to as control.

The macroscopic appearance of murine fibrosarcoma(MFS) tumours in the control groups of animals wascharacterized by subcutaneous dorsal apposition ofsolid structures tightly adhering to the underlying ribsand partially penetrating into the intercostal muscula-ture, and subcutaneous oedema is also evident. Majorintercostal blood vessels feeding the tumourous tissuewere visible on cross-sections of the posterior thoracicwall (Fig. 1A). Microscopically the penetrating growthof MFS towards the subcutaneous muscular tissue led topartial muscular destruction and local inflammatory andvascular reactions. Bizarre mitotic figures were observedin the fibrosarcoma tissue. No signs of tumour cellnecrosis were revealed.

The group with the application of a single dose of

hypericin 1 h before laser irradiation (2 h interval

after PDT)

Tumours show evident vascular reaction (dilatedarterioles and venules) particularly in the superficiallayers. Scattered focal necroses containing 10–20 kar-yopyknotic cells are localized mainly in superficial(‘‘skin cap’’) layers of the tumour and surrounding

blood venular vessels, and in some superficial areas thereare focal haemorrhages (photo not shown).

The group with the application of a single dose of

hypericin 6 h before laser irradiation (2 h interval

after PDT)

Tumour tissue focal necroses are arranged chain-likedown to a depth of 3mm under the skin surfaceaccording to the localization of vascular loops (Fig. 1B).Oedema of the dermis is evident.

The group with a fractionated dose of hypericin 6

and 1 h before laser irradiation (2 h interval after

PDT)

Tumour focal necroses are localized not only in thesuperficial layers, but in deep (costovertebral) layers aswell. There are foci of massive haemorrhagesattached to the borders of nectrotic areas. Fluorescentmicroscopy reveals evident hypericin fluorescence of theblood venular vessel walls in the tumour as well asdiffuse perivascular space fluorescence (photo notshown).

The group with the application of a single dose of

hypericin 1 h before laser irradiation (24 h interval

after PDT)

The overall appearance of the investigated skin andsubcutaneous area is characterized by moderate oedemaof interstitial tissue surrounding the subcutaneousmuscle, evident vascular reaction – dilating of capill-aries and venules, and stasis of blood in the tumoursuperficial layers (Fig. 2A). The range of necrotic

ARTICLE IN PRESS

Fig. 2. Evident vascular reaction in superficial layers of the tumour (arrows) (A). Thrombotic venules in the central part of the

necrotic area (B). The group with the application of a single dose of hypericin 1 h before laser irradiation (24 h interval after PDT).

Fig. 3. Massive blood vessel thromboses in superficial parts of the tumour (arrows) (A). Chain-like arranged necrotic areas in deep

layers of the tumour (B). The group with the application of a single dose of hypericin 6 h before laser irradiation (24 h interval after

PDT).

N. Bobrov et al. / Phytomedicine 14 (2007) 172–178 175

areas is wider in comparison with groups ofshorter survival, while the thrombotic venules arecharacteristically localized in central parts of thenecrotic areas (Fig. 2B).

The group with the application of a single dose of

hypericin 6 h before laser irradiation (24 h interval

after PDT)

Massive thromboses of arterioles and venules as partsof tumour tissue vascular loops are evident in superficiallayers of the tumour (Fig. 3A). The range of necrosis insuperficial and deep parts of the tumour is approxi-

mately equal. Deep necrotic foci are arranged chain-likein correspondence with the overall ‘‘pseudotrabecular’’pattern of tumour vasculature (Fig. 3B).

The group with a fractionated dose of hypericin 6

and 1 h before laser irradiation (24 h interval after

PDT)

The superficial layer of the tumour is almost totallynecrotic down to a depth of 3mm below the skinsurface, with evident oedema and reactive inflammatoryreaction of the dermis, subcutaneous muscle andsurrounding loose fibrous connecting tissue (Fig. 4A).

ARTICLE IN PRESS

Fig. 4. Total necrosis of the superficial part of the tumour (arrow) (A). Wide range of deep necroses with the haemorrhaging (B).

The group with the fractionated dose of hypericin 6 and 1 h before laser irradiation (24 h interval after PDT).

N. Bobrov et al. / Phytomedicine 14 (2007) 172–178176

The deep necroses are wide ranging, with haemorrhagefoci and signs of inflammation localized near the‘‘demarcation borders’’ of necrotic areas (Fig. 4B).

Discussion

For most sensitizers in clinical and preclinical use,three primary mechanisms of PDT-mediated tumourdestruction in vivo have been established: cellular,vascular and immunologic. The relative contributionof each depends, among other factors on the nature ofthe photosensitizer and its localization within thetumour tissue, the tumour type (vascularity and macro-phage content) and the time of light irradiation (which isone determinant of site of localization in the vascular vs.parenchymal compartment of the tumour).

Under some PDT protocols, vascular damage isconsidered to be the dominant mechanism of tumourdeath in vivo. Damage is believed to be initiated byrelease of factors such as eicosanoids, in particularthromboxane, histamines and tumour necrosis factor a(TNF-a). Macroscopically the PDT response is char-acterized by acute erythema, oedema, blanching andsometimes necrosis. Microscopically, the tumour tissueis characterized by endothelial cell damage, plateletaggregation, vasodilatation, vasoconstriction, cellularinflammatory response, necrosis and in certain condi-tions by apoptosis (Dougherty et al., 1998; Engelbrechtet al., 1999; Blank et al., 2002; Chen et al., 2002a, 2005;Dolmans et al., 2002).

Microscopic and macroscopic tumour response afterhypericin-based PDT was evaluated in our study too.We used the same treatment protocol with single andfractionated dosing schedules as published previously(Cavarga et al., 2005). We chose 2- and 24-h time

intervals after PDT for tumour tissue analysis. Anecrotic eschar creation was observed macroscopicallyin all PDT-treated groups of animals 24 h after laserlight exposure. At the shorter time interval after PDT(2 h) we expected only subclinical damage of tumourtissue, which could present an interesting histologicalcorrelate.

Tsoukas et al. (2000) described histological changeson rabbit skin after PDT with benzoporphyrins.Immediately after irradiation with 690 and 458 nm laserbeam the light microscopy showed stasis and inflamma-tory infiltrates in the papillary dermis. Later (days 1–2)PDT led to necrosis of skin appendages and epidermis.Necrosis was primarily due to vascular compromise.

Nakaseko et al. (2003) treated actinic keratoses with5-aminolaevulinic acid, a porphyrin precursor. Threehours after PDT they observed cells with eosinophiliccytoplasm and markedly stained nuclei, vacuolation ofsome tumour cells and low-level infiltration of lympho-cytes and neutrophils. One day after treatment all layersof the epidermis exhibited necrosis. TUNEL stainingrevealed apoptosis-positive cells within 1 day after PDT.

In our experiments, all groups of PDT-treatedanimals with single and fractionated hypericin dosingpresented primary vascular reactions. These includedvascular dilatation, congestion, thrombosis and oedema.Two hours after PDT the necrotic changes were smallwith rather focal appearance. One day after therapy thenecrotic areas were enhanced, often affecting thecomplete superficial layer of tumour tissue.

Vascular or tumour localization of the photosensitizercan determine the efficacy of PDT treatment. Shortlyafter administration (0.5 h) hypericin was located in thevascular compartment of the tumour, but 6 and 24 hlater it was located in the interstitial and cellularcompartments (Chen et al., 2001a). Some dilated and

ARTICLE IN PRESSN. Bobrov et al. / Phytomedicine 14 (2007) 172–178 177

occluded blood vessels were found immediately after the0.5 h-delayed PDT. Tumour cells remained intact.Evidence of vascular damage such as oedema anddilated and congested blood vessels was clear at 2 h after0.5 h delayed PDT. Extensive pycnotic tumour cells orruptured blood vessels were found at 24 h after 0.5 h-delayed PDT. Primary tumour vascular damage wasalso detected in tumours after 6 h-delayed PDT. Becauseof only partial vascular collapse, many viable tumourcells were detected 24 h after PDT (Chen et al., 2002a).

Vascular damage was observed in our experimentalgroups with 1 and 6 h delay of pharmacokinetics and 2 hafter PDT. We observed vascular dilatation, oedema ofthe dermis and some focal haemorrhages.

A combined PDT protocol with 5-aminolaevulinicacid (cellular targeted porphyrin prodrug) and photofrin(vessel targeted sensitizer) performed on the animaltumour model resulted in necrosis of neoplastic cells andsevere disruption of tumour microvasculature (Peng etal., 2001). This protocol is similar to our fractionateddosing schedule and we observed severe vascular andperivascular tissue damage, which progressed to almosttotal necrosis down to a depth of 3mm below the skinsurface (Fig. 4A).

Vascular accumulation of pyropheophorbide photo-sensitizer in murine mammary adenocarcinoma causesselective thrombosis in tumour vessels and 3 days afterPDT the tumour showed extensive necrosis (Dolmans etal., 2002). This is in correlation with our observationsfollowing hypericin-based PDT (Figs. 2A and 3B).

Intratumoral localization of hypericin was studied ona bladder transitional cell carcinoma implanted sub-cutaneously in rats. A short interval (0.5 h) after i.v.administration fluorescence microphotographs showedheterogenous distribution of hypericin in the tumour,with some isolated dense fluorescent areas. On the otherhand both the 6 and 24 h administration interval showedmore homogenous distribution of hypericin in thetumour stroma (Zupko et al., 2001). Our fluorescentmicroscopy observation reveals evident hypericin fluor-escence in blood vessel walls in the tumour as well asdiffuse perivascular space fluorescence.

Histological examination after hypericin-based PDTrevealed some dilated and congestive blood vesselsimmediately after treatment. Occluded or rupturedblood vessels together with extensive pycnotic tumourcells were observed at 15 h post treatment. Apparentlyviable tumour cells could be detected at the bottom ofthe tumour tissue at 24 h post treatment. Demarcationbetween pycnotic and viable tumour cells was relativelysharp (Chen et al., 2002b). A wide range of deepnecroses with haemorrhaging could be seen in ourexperimental observation (Fig. 4B). Necrotic areas wereaccompanied with inflammatory reactions (Fig. 4A, B).

Our histomorphological observation after hypericin-based PDT of G5:1:13 murine fibrosarcoma showed

that in all therapeutical groups of animals with singleand fractionated hypericin dosing, the consequence wasrepresented by primary vascular reactions which pro-gressed to continuous tumour tissue necrosis.

Acknowledgements

This work was supported by the Science andTechnology Assistance Agency under Contract no.APVT-20-022202, Grant no. 1/2329/05 of the ScientificGrant Agency of the Ministry of Education of theSlovak Republic and institutional Grant no. 8/2002/IG4from the Faculty of Medicine, P. J. Safarik University.Thanks also to Andrew J. Billingham for proofreadingthe manuscript.

References

Blank, M., Kostenich, G., Lavie, G., Kimel, S., Keisari, Y.,

Orenstein, A., 2002. Wavelength-dependent properties of

photodynamic therapy using hypericin in vitro and in an

animal model. Photochem. Photobiol. 76, 335–340.

Cavarga, I., Brezani, P., Cekanova-Figurova, M., Solar, P.,

Fedorocko, P., Miskovsky, P., 2001. Photodynamic ther-

apy of murine fibrosarcoma with topical and systemic

administration of hypericin. Phytomedicine 8, 325–330.

Cavarga, I., Brezani, P., Fedorocko, P., Miskovsky, P.,

Bobrov, N., Longauer, F., Rybarova, S., Mirossay, L.,

Stubna, J., 2005. Photoinduced antitumour effect of

hypericin can be enhanced by fractionated dosing. Phyto-

medicine 12, 680–683.

Chen, B., Xu, Y., Roskams, T., Delaey, E., Agostinis, P.,

Vandenheede, J.R., DeWitte, P., 2001a. Efficacy of anti-

tumoral photodynamic therapy with hypericin: relationship

between biodistribution and photodynamic effect in the

RIF–l mouse tumor model. Int. J. Cancer 93, 275–282.

Chen, B., Zupko, I., DeWitte, P., 2001b. Photodynamic

therapy with hypericin in a mouse P388 tumor model:

vascular effects determine the efficacy. Int. J. Oncol. 18,

737–742.

Chen, B., Roskams, T., DeWitte, P., 2002a. Antivascular

tumor eradication by hypericin mediated photodynamic

therapy. Photochem. Photobiol. 76, 509–513.

Chen, B., Roskams, T., DeWitte, P., 2002b. Enhancing the

antitumoral effect of hypericin-mediated photodynamic

therapy by hypertermia. Lasers Surg. Med. 31, 158–163.

Chen, B., Pogue, B.W., Hoopes, P.J., Hasan, Y., 2005.

Combining vascular and cellular targeting regimens en-

hances the efficacy of photodynamic therapy. Int. J. Radiat.

Oncol. Biol. Phys. 61, 1216–1226.

Chung, P.S., Rhee, C.K., Kim, K.H., Paek, W., Chung, J.,

Paiva, M.B., Eshraghi, A.A., Castro, D.J., Saxton, R.E.,

2000. Intratumoral hypericin and KPT laser therapy for

transplanted squamous cell carcinoma. Laryngoscope 110,

1312–1316.

Dolmans, D.E.J.G.J., Kadambi, A., Hill, J.S., Waters, C.A.,

Robinson, B.C., Walker, J.P., Fukumura, D., Jain, R.K.,

ARTICLE IN PRESSN. Bobrov et al. / Phytomedicine 14 (2007) 172–178178

2002. Vascular accumulation of novel photosensitizer

MV6401 cases selective thrombosis in tumor vessels

following photodynamic therapy. Cancer Res. 62,

2151–2156.

Dougherty, T.J., Gomer, C.J., Henderson, B.W., Jori, G.,

Kessel, D., Korbelik, M., Moan, J., Peng, Q., 1998.

Photodynamic therapy. J. Natl. Cancer Inst. 90, 889–905.

Engelbrecht, B.W., Menon, C.H., Kachur, A.V., Hahn, S.M.,

Fraker, D.L., 1999. Photofrin mediated photodynamic

therapy induces vascular occlusion and apoptosis in human

sarcoma xenograft model. Cancer Res. 59, 4334–4342.

Hofer, M., Hoferova, Z., Pospısil, M., Znojil, V., Chramostova,

K., 2003. Effects of adenosine on the growth of murine

G:5:113 fibrosarcoma cells in vitro. Folia Biol. 49, 207–210.

Hopfner, M., Maaser, K., Theiss, A., Lenz, M., Sutter, A.P.,

Kashtan, H., von Lampe, B., Riecken, E.O., Zeitz, M.,

Scherubl, H., 2003. Hypericin activated by incoherent light

source has photodynamic effects on esophageal cancer cells.

Int. J. Colorectal Dis. 18, 239–247.

Nakaseko, H., Kobayashi, M., Akita, Y., Tamada, Y.,

Matsumoto, Y., 2003. Histological changes and involve-

ment of apoptosis after photodynamic therapy for actinic

keratoses. Br. J. Dermatol. 148, 122–127.

Peng, Q., Warloe, T., Moan, J., Godal, A., Apricena,

F., Giercksky, K.E., Nesland, J.M., 2001. Antitumor

effect of 5-aminolevulinic acid-mediated photodynamic

therapy can be enhanced by the use of a low dose photofrin

in human tumor xenografts. Cancer Res. 61, 5824–5832.

Solar, P., Cavarga, I., Horfmanova, J., Cekanova-

Figurova, M., Miskovsky, P., Brezani, P., Hrckova, G.,

Kozubık, A., Fedorocko, P., 2002. Effect of acetazo-

lamide on hypericin photocytotoxicity. Planta Med. 68,

658–660.

Tsoukas, M.M., Gonzalez, S., Flotte, T.J., Anderson, R.R.,

Sherwood, M.E., Kollias, N., 2000. Wavelength and fluence

effect on vascular damage with photodynamic therapy on

skin. J. Invest. Dermatol. 114, 303–308.

Zupko, I., Kamuhabwa, A.R., D’Hallewin, M.A., Baert, L.,

DeWitte, P., 2001. In vivo photodynamic activity of

hypericin in transitional cell carcinoma bladder tumors.

Int. J. Oncol. 18, 1099–1105.