Interaction of photodynamic treatment and either hyperthermia or ionizing radiation and of ionizing radiation and hyperthermia with respect to cell killing of L929 fibroblasts, Chinese hamster ovary cells, and T24 human bladder carcinoma cells.

Both hyperthermia and photodynamic therapy of cancer are frequently used in combination with other treatment modalities in order to improve tumor control with minimal damage to normal tissues. The present results indicate that the most effective combination of treatment modalities is different in different cell types. For instance, ionizing irradiation and hyperthermia exhibited additivity when applied to L929 fibroblasts, in contrast to the synergistic interaction described before in many other cell lines. This aberrant behavior of L929 cells could be explained by the relative insensitivity of DNA repair in these cells to hyperthermia. Conversely, a synergistic interaction between photodynamic treatment and ionizing irradiation was observed with L929 fibroblasts, whereas these treatments were additive with Chinese hamster ovary and T24 cells. The synergistic interaction with L929 cells could be explained by the high sensitivity of DNA repair in these cells to photodynamic treatment. Photodynamic treatment and hyperthermia exhibited a synergistic interaction in L929, Chinese hamster ovary, and T24 cells. The critical target for cell killing by the combined treatment protocol in these cell lines has not yet been elucidated. In all three cell lines, however, analysis of the results according to the Arrhenius equation revealed a photodynamically induced change of both the frequency factor and the activation energy of subsequent thermal cell killing. It is considered that this may indicate a basic mechanism, in which a particular protein is a common, critical target of the two modalities of treatment.

The role of DNA damage and inhibition of poly(ADP-ribosyl)ation in loss of clonogenicity of murine L929 fibroblasts, caused by photodynamically induced oxidative stress.

Reactive oxygen species are used to eradicate malignant cells in photodynamic therapy as well as in other cancer therapies. Despite many efforts, the pathways leading to cellular damage and cell killing due to the action of these species are poorly understood. In previous studies with hematoporphyrin derivative-sensitized L929 murine fibroblasts, the only parameter for which a relation with photodynamically induced reproductive cell death could not be excluded was inhibition of DNA excision repair. The present results show that loss of clonogenicity of these cells in fact is related to a series of effects, including the development of slight, irreperable DNA damage, a virtually complete inhibition of poly(ADP-ribosyl)ation activation, a transient elevation of the intracellular calcium concentration and, after a lag time of about 8 h, DNA fragmentation caused by endonuclease activity. This conclusion is supported by the observation that photodynamic treatment inhibited the repair of X-ray-induced DNA strand breaks and suppressed X-ray- and methyl methanesulfonate-induced enhancement of poly(ADP-ribosyl)ation. Our experimental results further suggest that in this cell line the photodynamically induced inhibition of enhanced poly(ADP-ribosyl)ation could well be involved in inhibition of repair of DNA strand breaks and in activation of endonuclease activity.

Photodynamic effects of hematoporphyrin-derivative on transmembrane transport systems of murine L929 fibroblasts.

Photodynamic treatment of murine L929 fibroblasts with hematoporphyrin-derivative causes deterioration of various membrane functions. Most sensitive to photodynamic inactivation are the energy-coupled transport systems for aminoisobutyric acid and for Rb+. The facilitated diffusion system for 2-deoxy-D-glucose is slightly less sensitive. After longer illumination periods also the membrane barrier function is impaired, as reflected by K+ leakage and increased passive Rb+ uptake. After still longer illumination periods intermolecular protein crosslinking can be observed. This makes it unlikely that intermolecular protein crosslinking is causally involved in the deterioration of these membrane functions.

A survey of differences between membrane polypeptides of transformed and nontransformed chick embryo fibroblasts.

Plasma membrane-associated polypeptides of chick embryo fibroblasts and cells transformed by the Schmidt-Ruppin wild-type strain of Rous sarcoma virus and its temperature-sensitive tsNY68 mutant were compared by two-dimensional gel electrophoresis. Polypeptide and glycoprotein alterations were identified after incubation of cells with [35S]methionine and [3H]mannose and by staining of the gels with 125I-labeled concanavalin A and Coomassie brilliant blue. Polypeptides found to be consistently transformation-sensitive included a group of five polypeptides that were detected only by short-term labeling with methionine, fibronectin, a 180 kDa polypeptide with a pI of 5.6, a mannose-containing glycoprotein of 48 kDA and an unusually high pI of 8.4, and a 19 kDa polypeptide with a pI of approx. 4.5. Several of these polypeptides appear to be particularly interesting for further characterization.

Protein damage, induced by small amounts of photodynamically generated singlet oxygen or hydroxyl radicals.

The influence of limited oxidation of glyceraldehyde-3-phosphate dehydrogenase (D-glyceraldehyde-3-phosphate:NAD+ oxidoreductase (phosphorylating), EC 1.2.1.12), alcohol dehydrogenase (alcohol:NAD+ oxidoreductase, EC 1.1.1.1) and myoglobin by singlet oxygen and by hydroxyl radicals was investigated. The intrinsic fluorescence of glyceraldehyde-3-phosphate dehydrogenase and alcohol dehydrogenase decreased rapidly during oxidation, indicating a conformational change of the protein molecules. The free energy of isothermal unfolding in urea solutions was increased by singlet oxygen, but decreased by hydroxyl radical attack. The velocity of refolding of the denatured protein after dilution of the denaturant was increased by exposure to either singlet oxygen or hydroxyl radicals, with one exception: the velocity of refolding of myoglobin, oxidized by singlet oxygen, was strongly decreased. Hydroxyl radicals produced covalently crosslinked protein aggregates and some fragmentation, whereas singlet oxygen produced only crosslinked aggregates with glyceraldehyde-3-phosphate dehydrogenase and alcohol dehydrogenase. All oxidized proteins were more susceptible to proteolysis by elastase and proteinase K, as compared to the undamaged proteins. Singlet oxygen-induced crosslinked aggregates were degraded very rapidly by elastase. Hydroxyl radical-induced aggregates of glyceraldehyde-3-phosphate dehydrogenase were also degraded very rapidly by this enzyme, but hydroxyl radical-induced aggregates of alcohol dehydrogenase were resistent to enzymatic degradation. The results indicate that limited protein oxidation may have a pronounced influence on several properties of the protein. The effects vary, however, with varying proteins and with the oxidizing species.

Temperature dependence of photodynamic red cell membrane damage.

1. The protoporphyrin-sensitized photo-oxidation of free amino acids, amino acid residues in solubilized spectrin and amino acid residues in red blood cell membranes appeared to be virtually independent of temperature over the range 0-37 degrees C. The photodynamically produced increase in cation permeability in intact cells was also almost temperature independent. 2. The photodynamic cross-linking of the membrane proteins, on the other hand, was clearly temperature dependent, both when the proteins were present in the membrane structure and when isolated and purified. 3. With red cell membranes, illuminated in the presence of protoporphyrin at 0 degrees C, it could be shown that during subsequent incubation in the dark at 37 degrees C the protein cross-linking increased considerably. 4. The results indicate that cross-linking of membrane proteins is a secondary reaction in which rather stable photo-oxidation products of susceptible amino acid residues are involved. Furthermore, these experiments strongly suggest that the deterioration of membrane function, leading to increased caption permeability, is caused by photo-oxidation of amino acid residues rather than by cross-linking of membrane proteins.

Photodynamic protein cross-linking.

Exposure of spectrin to visible light in the presence of a photosensitizer results in photo-oxidation of sensitive amino acid residues and covalent cross-linking of the polypeptides. In a previous paper the cross-linking was ascribed to a secondary reaction between photo-oxidized histidine residues and amino groups. The following observations, described in this paper, are in accordance with this supposition. (1) During illumination of spectrin in the presence of a photosensitizer a pronounced photo-oxidation of histidine residues takes place. (2) Simultaneously a decrease of free amino groups is observed. (3) Semicarbazide protects against cross-linking and is bound to a histidine photo-oxidation product in spectrin. (4) The pH profile of histidine photo-oxidation and subsequent reaction with amino groups is similar to the pH profile of spectrin cross-linking. Amidination of NH2 groups in spectrin does not inhibit cross-linking, as visualized by gel electrophoresis. On the other hand aminidation of denatured myoglobin causes a 50% inhibition of cross-linking. These observations support the notion of NH2-involvement in cross-linking but also demonstrate, that other photodynamic cross-link mechanisms exist.

The influence of cupric ions on porphyrin-induced photodynamic membrane damage in human red blood cells.

Photooxidation of various susceptible substrates with hematoporphyrin derivative (photofrin) as sensitizer was strongly inhibited by simultaneous addition of cupric acetate to the reaction mixture. With sulfhydryl-containing compounds, however, an increased rate of photooxidation was observed under these experimental conditions. Preincubation of photofrin and cupric acetate at equimolar concentrations for 24 h at room temperature yielded a stable photofrin-Cu2+ complex. This complex did not act as photosensitizer with histidine, tryptophan, tyrosine, methionine or guanosine as substrates. With dithiothreitol, however, the photofrin-Cu2+ complex still acted as a photosensitizer, with an efficiency of about 30% as compared to free photofrin. Also in red blood cell membranes only sulfhydryl groups were photooxidized with the photofrin-Cu2+ complex as sensitizer. Illumination of intact erythrocytes in the presence of the photofrin-Cu2+ complex resulted in K+ leakage and, ultimately, photohemolysis. Pretreatment of the cells with N-ethylmaleimide and 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) inhibited this photodynamically induced K+ loss. Considering recent studies on the reactivity of distinct membrane SH-groups with various sulfhydryl reagents this suggests that a sulfhydryl group, located in the 17 kDa membrane-bound fragment of band 3, is involved in photodynamic K+ leakage with the photofrin-Cu2+ complex as sensitizer.

Peroxide-induced membrane damage in human erythrocytes.

Erythrocytes exposed to H2O2 or t-butyl hydroperoxide (tBHP) exhibited lipid peroxidation and increased passive cation permeability. In the case of tBHP a virtually complete inhibition of both processes was caused by butylated hydroxytoluene (BHT), whereas pretreatment of the cells with CO increased both lipid peroxidation and K+ leakage. In the experiments with H2O2, on the other hand, both BHT and CO strongly inhibited lipid peroxidation, without affecting the increased passive cation permeability. These observations indicate different mechanisms of oxidative damage, induced by H2O2 and tBHP, respectively. The SH-reagent diamide strongly inhibited H2O2-induced K+ leakage, indicating the involvement of SH oxidation in this process. With tBHP, on the contrary, K+ leakage was not significantly influenced by diamide. Thiourea inhibited tBHP-induced K+ leakage, without affecting lipid peroxidation. Together with other experimental evidence this contradicts a rigorous interdependence of tBHP-induced lipid peroxidation and K+ leakage.

Protoporphyrin-induced photodynamic effects on transport processes across the membrane of human erythrocytes.

Previous studies have shown that illumination of erythrocytes with visible light in the presence of protoporphyrin results in cross-linking of membrane proteins and deterioration of several membrane functions, e.g. active transport of K+ and Na+. In the present study it is shown that carrier-mediated transport of glucose, L-leucine, sulphate and glycerol is also inhibited by the photodynamic process, whereas non-specific permeability of glycerol and thiourea is increased. It is shown that these effects are not caused by lipid peroxidation, but by photooxidation of membrane proteins. The inhibition of carrier-mediated transport is caused either by photodynamic oxidation of susceptible essential amino acid residues of the carrier molecules, or by an aspectific perturbation of the membrane structure, leading to inhibition of carrier functions.

Photodynamic effects of protoporphyrin on human erythrocytes. Nature of the cross-linking of membrane proteins.

Protoporphyrin-sensitized photooxidation in human red blood cell membranes leads to severe deterioration of membrane structure and function. The membrane damage is caused by direct oxidation of amino acid residues, with subsequent cross-linking of membrane proteins. The chemical nature of these cross-links was studied in model systems, isolated spectrin and red cell ghosts. Cysteine and methionine are not involved in the cross-linking reaction. Further it could be shown that dityrosine formation, the crucial mechanism in oxidative cross-linking of proteins by peroxidase-H2O2 treatment, plays no role in photodynamic cross-linking. Experimental evidence indicated that a secondary reaction between free amino groups and a photooxidation product of histidine, tyrosine or tryptophan is involved in photodynamic cross-linking. This was deduced from the reaction observed between compounds containing a free amino group and photooxidation products of these amino acids, both in model systems, isolated spectrin and erythrocyte ghosts. In accordance, succinylation of free amino groups of membrane proteins or addition of compounds with free amino groups protected against cross-linking. Quantitative data and consideration of the reaction mechanisms of photodynamic oxidation of amino acids make it highly probable that an oxidation product of histidine rather than of tyrosine or tryptophan is involved in the cross-linking reaction, via a nucleophilic addition by free amino groups.

Protoporphyrin-induced photodynamic effects on band 3 protein of human erythrocyte membranes.

In previous studies it has been shown that protoporphyrin-induced photodynamic effects on red blood cells are caused by photooxidation of amino acid residues in membrane proteins and by the subsequent covalent cross-linking of these proteins. Band 3, the anion transport protein of the red blood cell membrane, has a relatively low sensitivity to photodynamic cross-linking. This cannot be attributed to sterical factors inherent in the specific localization of band 3 in the membrane structure. Solubilized band 3, for instance, showed a similar low sensitivity to cross-linking. By extracellular chymotrypsin cleavage of band 3 into fragments of 60,000 and 35,000 daltons it could be shown that both fragments were about equally sensitive to photodynamic cross-linking. The 17,000 dalton transmembrane segment, on the other hand, was completely insensitive. Inhibition of band 3-mediated sulfate transport proceeded much faster than band 3 interpeptide cross-linking, presumably indicating that the inhibition of transport is caused by photooxidation of essential amino acid residues or intrapeptide cross-linking. A close parallel was observed between photodynamic inhibition of anion transport and decreased binding of 4,4'-diisothiocyanodihydrostilbene-2,2'-disulfonate (H2DIDS), suggesting that a photooxidation in the immediate vicinity of the H2DIDS binding site may be responsible for transport inhibition.

Anomalous properties of water in macromolecular gels.

Low molecular weight solutes often exhibit elution characteristics on gel filtration columns which deviate from ideal behaviour. In many previous studies this anomalous behaviour was attributed to the existence of extremely narrow pores in the gel, inaccessible even to very small solute molecules, to explain Kd values lower than unity. Kd values of small solutes higher than unity were usually ascribed to adsorption of the solute to the gel matrix. In the present paper several observations are presented that contradict these suggestions. Experimental evidence indicates that with small solute molecules Kd values differing from unity can be fully explained by the anomalous properties of vicinal water layers at the gel matrix-water interface.

Isolation and characterization of polyphosphates from the yeast cell surface.

When cells of Saccharomyces fragilis are subjected to osmotic shock, they release a limited amount of inorganic polyphosphate into the medium, which represents about 10% of the total cellular content. The osmotic shock procedure causes no substantial membrane damage, as judged from the unimpaired cell viability, limited K+ leakage and low percentage of stained cells. It is therefore suggested that this polyphosphate fraction is localized outside the plasma membrane. The released polyphosphate fraction differs from the remaining cellular polyphosphates in two respects: the mean chain length of the shock-sensitive fraction is significantly higher than that of the total cellular polyphosphates and its metabolic turnover rate, subsequent to pulsing with [32P]orthophosphate is much lower compared to the rest of the cellular polyphosphate. Incubation of intact cells with the anion exchange resin Dowex AG 1-X4 results in the release of high molecular weight polyphosphates. These results suggest that the osmotic shock-sensitive polyphosphate fraction has specific characteristics in both its cellular localization and metabolism.

The influence of ozone on human red blood cells. Comparison with other mechanisms of oxidative stress.

Exposure of red blood cells to ozone resulted in K+ leakage, lipid peroxidation and inhibition of some membrane-associated enzymes. On the other hand, carrier-mediated transport of glucose, leucine, sulfate and glycerol and the nonspecific permeation of glycerol, L-glucose and erythritol were not affected by ozone. The cellular level of reduced glutathione declined, whereas the ATP content of the cells was quite insensitive to ozone exposure. It was shown that, most probably, lipid peroxidation and K+ leakage are not causally related. Further, K+ leakage did not reflect gradual, progressive loss of K+ from all cells simultaneously, but occurred in an all-or-none fashion. Finally, ozone-induced damage was compared to damage induced by H2O2, t-butyl hydroperoxide and photosensitizers plus light. It appeared that the pathways leading to membrane deterioration are quite dissimilar in these various forms of oxidative stress.

Inhibition of enzymes and oxidative damage of red blood cells induced by t-butylhydroperoxide-derived radicals.

The effects of t-butylhydroperoxide (tBHP), its alkoxyl radical (tBuO.) and its peroxyl radical (tBuOO.) in model systems and on red blood cells were studied. Glyceraldehyde-3-phosphate dehydrogenase was strongly inhibited by tBHP via a direct reaction of the hydroperoxide with an essential sulfhydryl group in the enzyme molecule. Several other enzymes were unaffected by tBHP. Alcohol dehydrogenase was strongly inhibited by tBuO. but was much less sensitive to tBuOO.. Lysozyme, lactate dehydrogenase and trypsin, on the other hand, were very sensitive to the peroxyl and not, or much less, to the alkoxyl radical, whereas acetylcholinesterase was very sensitive to both radicals. tBuOO. caused covalent binding of tryptophan, tyrosine, histidine and methionine to serum albumin. The corresponding alkoxyl radical was ineffective in this respect. Conversely, tBuO. caused peroxidation of linolenic acid, whereas tBuOO. did not. Incubation of human erythrocytes with tBHP caused lipid peroxidation and K+ leakage. Both effects were caused by tBHP-derived radicals generated in a reaction of the hydroperoxide with hemoglobin. With radical scavengers it was possible to dissociate tBHP-induced lipid peroxidation and K+ leakage, demonstrating that these two processes are not causally related. Experimental results indicate that tBuO. causes lipid peroxidation, whereas tBuOO. is responsible for K+ leakage.

Factors affecting the amount and the mode of merocyanine 540 binding to the membrane of human erythrocytes. A comparison with the binding to leukemia cells.

In the presence of albumin Merocyanine 540 (MC540) exhibits a very limited binding to the outer surface of the membrane of normal erythrocytes, whereas pronounced binding is observed to leukemia cells. To find out whether this difference is due to differences in the composition or structural organization of the cell membrane we analyzed effects of a number of covalent and non-covalent perturbations of the red cell membrane on the binding and fluorescence characteristics of membrane-bound MC540. It is shown that exposure of the cells to cationic chlorpromazine, neuraminidase or photodynamic treatment with AlPcS4 as sensitizer caused a limited increase (30-50%) of MC540 binding, together with a red shift of the fluorescence emission maximum and an increase of the relative fluorescence quantum yield of membrane-bound MC540. Other forms of perturbation of the membrane structure, like hyperthermia (48 degrees C) and treatments that produce a decrease of phospholipid asymmetry in addition to accelerated flip-flop, did not result in increased MC540 binding, but did cause a red shift of the fluorescence emission maximum and an increase of the relative fluorescence quantum yield. These changes in fluorescence properties indicate a penetration of the dye into more hydrophobic regions in the membrane. MC540, bound to Brown Norway myelocytic leukemia cells, exhibited a red shift of the fluorescence emission maximum and an increased relative fluorescence quantum yield as compared to MC540 bound to untreated erythrocytes. These changes were of the same order of magnitude as in photodynamically treated red blood cells. Dye binding per surface area, however, was about 3-times higher with these leukemia cells than with photodynamically treated red blood cells. This demonstrates that certain perturbations of the erythrocyte membrane evoked a MC540 binding that became qualitatively comparable to the dye binding to leukemia cells, although dye binding per surface area was still significantly lower.

Enhancement of transbilayer mobility of a membrane lipid probe accompanies formation of membrane leaks during photodynamic treatment of erythrocytes.

In order to further characterize membrane alterations in human erythrocytes subjected to photodynamic treatment the passive transbilayer mobility of a phospholipid analogue was studied in cells illuminated for various lengths of time in the presence of the photosensitizer, aluminum chlorotetrasulfophthalocyanine. These measurements were combined with the characterization of the membrane leaks for polar solutes occurring under the same conditions with respect to their apparent size, number and ion selectivity. The time-dependent photodynamic enhancement of leaks for K+ as well as choline or erythritol was paralleled by a marked increase of the transbilayer reorientation rate of the amphiphilic lipid probe, palmitoyllysophosphatidylcholine from 0.05% min-1 in native cells to 0.32% min-1 after 60 min illumination. The asymmetric orientation of native phospholipids was not affected by this treatment. The leak permeability proved to be due to the formation of pores with apparent radii of about 0.45 nm after 60 min illumination, and of 0.75 nm after 90 min. The number of pores per cell was calculated to be less than 1, the pores are slightly cation-selective (PK/PCl approximately 3:1). Since photodynamic treatment did not induce lipid peroxidation under the prevailing experimental conditions, protein modification must be the primary cause of both, leak permeability and flip enhancement. Since it is also likely that the leak permeability arises from oxidation of intrinsic membrane proteins, the results raise the interesting possibility that oxidative alteration of intrinsic membrane proteins may lead to enhanced transbilayer mobility of lipids.

Preferential uptake of cytotoxic porphyrins from hematoporphyrin derivative in murine L929 fibroblasts and Chinese hamster ovary K1 epithelial cells.

Photodynamically induced loss of clonogenicity of murine L929 fibroblasts and Chinese hamster ovary K1 epithelial cells was determined with two different assays. It appeared that the loss of clonogenicity was much higher when 20 cells/cm2 were incubated with hematoporphyrin derivative (HPD) and illuminated, than when confluent cell layers were incubated with the same amount of HPD and illuminated prior to plating out. This dependency of cell killing on the experimental protocol was also observed when protoporphyrin (90-95% pure) was used as photosensitizer, but not when the cells were photodynamically treated with rose bengal or exposed to mitomycin C. Further, when cell layers were incubated with the residual solution that remained after the previous incubation of a confluent cell layer with HPD, illumination of these layers appeared to be almost non-toxic, although the overall porphyrin concentration in the residual solution was only slightly lower than in HPD. These results indicate that the porphyrins, responsible for loss of clonogenicity, are present in relatively small amounts in HPD and unpurified protoporphyrin and are preferentially taken up by the cells. Although 2-aminoisobutyric acid transport and DNA synthesis are among the most photosensitive targets with HPD, photodynamic treatment of L929 cells with the residual solution did not result in inhibition of the transport system and DNA synthesis. In contrast, the K+ content of the cells still decreased considerably, when utilizing the porphyrins, remaining in the residual solution as sensitizer. This indicates that under the present experimental conditions the disturbance of the membrane barrier function does not contribute to loss of clonogenicity of these cells and, moreover, that the photodynamically induced K+ leakage is caused by a component of HPD other than inhibition of 2-aminoisobutyric acid transport and DNA synthesis.

Effects of the micro-environment on the photophysical properties of hematoporphyrin.

The photochemical degradation of histidine, cysteine and tyrosine with hematoporphyrin as sensitizer was potentiated by the presence of Sephadex, BioGel or Percoll particles. This effect could only partly be explained by binding of the sensitizer to the gel particles, leading a.o. to monomerization of aggregated sensitizer molecules in the aqueous environment. The hematoporphyrin triplet state life time increased from 250 microseconds in phosphate buffer to 1992 microseconds in the presence of 50% Percoll. Most likely the effect of the gel particles on the sensitizer triplet state is, at least partly, mediated by the solvent. A plausible explanation seems to be that the vicinal water structure at the particle interface stabilizes the sensitizer triplet.