Els Delaey

A comparative study of the photosensitizing characteristics of some cyanine dyes

The present work has been carried out to explore the potential application of cyanines in photodynamic therapy. After photosensitization, the in vitro cytotoxic and antiproliferative activity on HeLa cells of a total of 35 cyanines belonging to several chemical subgroups is explored. Most of these cyanines have never been used before in similar experimental work. From a first set of experiments, it is found that none of the krypto-, oxa- and imidacyanines is photobiologically active on HeLa cells. Conversely, five thiacyanines (Thiacl-5), one rhodacyanine (Rhodac) and four indocyanines (Indoc2, Indoc4, Indoc5, Indoc7) show photodependent cytotoxicity or antiproliferative effects. A more detailed study shows that out of the ten selected compounds, eight cyanines feature significant photodependent cytotoxic and antiproliferative effects. All possess maximum absorption ranges between 545 and 824 nm. In particular, Rhodac, a tetramethinemeromonomethine rhodacyanine dye with an absorption maximum of 655 nm (ethanol) and a molar absorption coefficient epsilon = 108000 shows very promising photo-dependent biological activity. In general, the measured singlet oxygen quantum yield of the selected cyanines is low (< 0.08) and does not correlate with the degree of photosensitization. Furthermore, the present study shows that cyanines with a partition coefficient close to 1.5 accumulate to the highest extent in HeLa cells, while the more hydrophobic compounds (e.g., indocyanines) concentrate less intracellularly.

Cytotoxicity and antiproliferative effect of hypericin and derivatives after photosensitization.

The toxicity on three human tumor cell lines (A431, HeLa and MCF7) of five phenanthroperylenequinones (hypericin and derivatives) and two perylenequinones (cercosporin and calphostin C) was investigated after photosensitization (4 J/cm2). Furthermore, the antiproliferative effect on HeLa cells was studied for the phenanthroperylenequinones. Hypericin, 2,5‐dibromohypericin, 2,5,9,12‐tetrabromohypericin and perylenequinones displayed a potent cytotoxic and antiproliferative effect in the nanomolar range. Hypericin dicarboxylic acid exhibited no photoactivity. In general, the antiproliferative activity correlated well with the photocytotoxicity. However, the nonphotocytotoxic compound hexamethylhypericin showed potent antiproliferative activity in the nanomolar range, probably exerting its action by protein kinase C inhibition. Without light irradiation, no cytotoxic and antiproliferative effect was observed for any photocytotoxic phenanthroperylenequinone compound. Furthermore, confocal laser microscopy revealed that the subcellular localization in A431 cells was similar for the photoactive compounds; the photosensitizers were mainly concentrated in the perinuclear region, probably corresponding with the Golgi apparatus and the endoplasmic reticulum. In addition, the accumulation of the photosensitizers in HeLa cells was investigated. All compounds except hypericin dicarboxylic acid were found to concentrate to a large extent in the cells. The compound 2,5,9,12‐tetrabromohypericin seemed intrinsically more effective than hypericin since the intracellular concentration of the bromoderivative was a magnitude of order lower than that of hypericin although both compounds showed similar photobiological activity.

Efficacy of antitumoral photodynamic therapy with hypericin: Relationship between biodistribution and photodynamic effects in the RIF-1 mouse tumor model

We investigated the hypericin‐mediated PDT effects on the tumor and normal skin and in correlation with its biodistribution. These studies were carried out on C3H mice bearing RIF‐1 tumors. The hypericin distribution and PDT effects were recorded at different intervals (0.5–24 hr) after intravenous injection of a 5‐mg/kg dose of hypericin. After administration, rapid biphasic exponential decay was observed in the plasma drug concentration. It was found that hypericin was preferentially bound to the plasma lipoproteins. The tumor drug levels increased rapidly over the first few hours and reached a maximum around 6 hr after injection. In contrast, PDT efficacy was maximal when irradiation was performed at 0.5 hr after hypericin administration, which led to 100% cure. The PDT efficacy decreased rapidly as the administration‐irradiation interval was prolonged. No tumor cure was obtained at the 6‐hr interval, even though it was at this time that the tumor drug level peaked. Fluorescence microscopic studies showed that hypericin was mainly confined within the tumor vasculature at 0.5 hr after injection, whereupon it rapidly diffused to the surrounding tumor tissue. At 6 hr, a strong hypericin fluorescence was observed in the tumor tissue with only faint fluorescence within the vasculature, whereas at 24 hr the fluorescence in the tumor also decreased and became more diffused, and no fluorescence could be seen in the tumor vasculature. Like the tumor response, skin reactions were also found to be much more dramatic at short administration‐irradiation intervals. Hypericin distribution and PDT response studies revealed a close correlation between the plasma drug level and the PDT effects, which suggests that vascular damage is the primary effect of hypericin‐mediated PDT in this tumor model.

Photocytotoxic effect of pseudohypericin versus hypericin

Pseudohypericin and hypericin, the major photosensitizing constituents of Hypericum perforatum, are believed to cause hypericism. Since hypericin has been proposed as a photosensitizer for photodynamic cancer therapy, the photocytotoxicity of its congener pseudohypericin has been investigated. The presence of foetal calf serum (FCS) or albumin extensively inhibits the photocytotoxic effect of pseudohypericin against A431 tumour cells, and is associated with a large decrease in cellular uptake of the compound. These results suggest that pseudohypericin, in contrast to hypericin, interacts strongly with constituents of FCS, lowering its interaction with cells. Since pseudohypericin is two to three times more abundant in Hypericum than hypericin and the bioavailabilities of pseudohypericin and hypericin after oral administration are similar, these results suggest that hypericin, and not pseudohypericin, is likely to be the constituent responsible for hypericism. Moreover, the dramatic decrease of photosensitizing activity of pseudohypericin in the presence of serum may restrict its applicability in clinical situations.

In Vitro Study of the Photocytotoxicity of Some Hypericin Analogs on Different Cell Lines

In the present study, hypericin analogs with an increased hydrophilic character were synthesized. As chemical modifications alter the lipophilicity/hydrophilicity balance together with the photophysical/chemical background of the molecule the influence of these structural changes on the cellular uptake, retention and subcellular localization in HeLa cells was investigated. Besides, their photocytotoxic effects using three cell lines (HeLa, MCF‐7, A431), as well as their plasma protein binding were also assessed. To assess the relative hydrophilic/lipophilic character of hypericin and analogs their retention times were determined on a reversed phase high performance liquid chromatography (C‐18) column. The retention time of all the hypericin analogs was <46 min, except for dibenzyltetramethylhypericin (118 min), while the retention time of hypericin was >200 min (solvent system: methanol/citrate buffer 30 mM pH 7; 70/30). Hypericin, hexa‐, penta‐ and dibenzyltetramethylhypericin displayed a potent antiproliferative effect at the nanomolar range after photosensitization (3.6 J/cm2). On the contrary, photoactivated tetrasulfonhypericin and fringelite D had no antiproliferative effect on the three cell lines, whereas hypericin polyethylene glycol showed only an intermediate cytotoxic effect on A431 cells. In dark conditions no antiproliferative effect was observed for any photosensitizer. The antiproliferative photoeffect correlated well with the intracellular accumulation as measured using HeLa cells. In general, the photocytotoxic hypericin analogs concentrated to a large extent, while the noncytotoxic compounds were not taken up by the HeLa cells. Furthermore, confocal laser microscopy revealed that all photosensitizers mainly concentrated in the perinuclear region, probably corresponding with Golgi apparatus and the endoplasmic reticulum, except for tetrasulfonhypericin which located at the plasma membrane. In addition, the plasma protein binding studies illustrated that hypericin bind extensively to the low‐density lipoproteins, while the other hypericin analogs were mainly bound to heavy proteins (mostly albumin) and to a small extent to low‐density lipoproteins.

Photocytotoxicity of hypericin in normoxic and hypoxic conditions

The normoxic and hypoxic photocytotoxicity of hypericin has been examined on A431 cells as assessed by the Neutral Red method, using cell-culture flasks made of polystyrene and glass, different hypericin concentrations and light fluences. Using polystyrene flasks, lower hypoxic photoactivities of hypericin than those in normoxic conditions are seen under low fluence. In these conditions the hypoxic photocytotoxic effect can be (partially) rescued by increasing the fluence. However, a completely different outcome is observed when using glass flasks, since most of the hypoxic photocytotoxicity is lost under these conditions. The differences can be explained in terms of efficiency of deoxygenation of the medium present in polystyrene or glass flasks. Polystyrene holds large amounts of oxygen that effuses very slowly. Glass, on the other hand, does not cause this inconvenience. Therefore the type of material of the container used to investigate the oxygen dependency of the photobiological activity of photosensitizers dramatically influences the outcome of the hypoxic experiments. Our results unequivocally prove that the cytotoxic effect induced by photoactivated hypericin is completely oxygen dependent. Hence hypericin does not differ from other phototherapeutics used in photodynamic therapy of cancer, since haematoporphyrin derivative and the second-generation photosensitizers used all seem to depend on the presence of oxygen for their antitumour activity.

Bromohypericins Are Potent Photoactive Antiviral Agents

Several hypericin derivatives, previously shown to have interesting light‐mediated biological activities, were evaluated for antiviral activities against herpes simplex virus and influenza virus. Three brominated hypericins, the di‐bromo‐ and tetrabromo‐derivatives and the natural compound gymnochrome B were all very active against both viruses, particularly herpes simplex virus, although light was required in all cases for maximum activity. The di‐bromohypericin was the most potent, under standard assay conditions, gymnochrome B was approximately as active as hypericin itself and tetrabromohypericin significantly less so. Surprisingly, hexamethylhypericin, which is known to have potent anti‐protein kinase (PK) C activity, as well as anticell proliferation properties, showed no antiviral activity at all. The compounds were also evaluated in different serum concentrations. All the active compounds were inhibited by increasing concentrations of serum, but to different degrees, such that their relative antiviral potencies changed to some extent Thus, in summary, there was no correlation between antiviral and anti‐PK or anticellular activities, and consequently it is not possible at present to define those structural features of hypericin‐type molecules that are required for their various biological activities.

In Vitro Photobiological Evaluation of Rhodac, A New Rhodacyanine Photosensitizer¶

We have previously shown that the rhodacyanine dye, Rhodac, exhibits a potent photocytotoxic activity in HeLa cells. In this study several aspects of the photobiological activity of Rhodac were further examined. Rhodac displayed no selective cytotoxicity toward several malignant cell lines after photosensitization (3.6 J/cm2), although HeLa cells were found to be the most sensitive. Interestingly, MCF‐7/Adr cells, a multidrug‐resistant subline, were less sensitive to the antiproliferative effect of photoactivated Rhodac. The subcellular localization, as revealed by confocal laser microscopy, demonstrated that the dye was mainly concentrated in the cytosolic membranes of the perinuclear region. The Rhodac‐induced inhibition of HeLa cell proliferation after light exposure was found to be strictly oxygen dependent. In addition, photoactivated Rhodac induced poly(adenosine 5′ diphosphate‐ribose)polymerase cleavage, caspase‐3 activation and apoptosis in HeLa cells. In the current work it was further demonstrated that Rhodac binds specifically to high‐density lipoproteins and low‐density lipoproteins, while no binding was observed to very low‐density and heavy proteins. To sum up, our results show that Rhodac is an interesting and potent photosensitizer. Further in vivo experiments are required to elucidate whether the lipoprotein binding leads to a selective uptake of Rhodac in tumor cells and to address its efficacy in photodynamic therapy.