Temperature dependence of hemolysin production during psoralen photooxidation

Vladislava O. Melnikova^a,b, Mikhail V. Malakhov^a, Elena S. Andina^a, Eugene P. Lysenko^a, Irina V. Belichenko^a,b, Lina N. Bezdetnaya^b, Francois Guillemin^b and Alexander Ya. Potapenko^a*.

^aDepartment of Medical and Biological Physics, Russian State Medical University, Moscow, Russia

^bUnite de Recherche en Therapie Photodynamique, Centre Alexis Vautrin, Vandoeuvre-les-Nancy, France

Abstract

There are indications that toxic and therapeutic effects induced by PUVA-therapy (psoralen+UVA) are at least partly mediated by products of psoralen photooxidation (POP). In the present study psoralen photooxidation in aqueous solution was investigated in temperature range from 4 to 45° C. Chemical identification of biologically active POP products is highly difficult because of their instability. That is why POP products were revealed with the use of very sensitive biologic test system - by their hemolytic effects in suspensions of human erythrocytes. The production of POP hemolysins (POP_hem) increased with temperature in the range from 4 to 25° C. At temperature higher than 25° C, the production of POP_hem dropped down. Analogous temperature dependence was observed when POP products were estimated by chemiluminescence induced by addition of ferrous ions to aqueous POP solution (POP_ChL). However, despite the resemblance in temperature dependencies of POP production revealed by hemolysis and chemiluminescence, these two methods gave information about different products of psoralen photooxidation because POP_ChL were more labile in storage than POP_hem. Therefore, estimation of products of psoralen photooxidation by their hemolytic effect is a unique and specific approach for monitoring the production and properties of biologically active products of photooxidation.

Keywords: Psoralen; photooxidized psoralen; hemolysis; chemiluminescence; temperature effect

Abbreviations: POP is photooxidized psoralen; ChL_Fe is Fe(II)-dependent chemiluminescence; PBS is phosphate-buffered saline; RBCs is human red blood cells; UVA is 320-400 nm radiation; PUVA is psoralen+UVA phototherapy

*Corresponding author: Department of Medical and Biological Physics, Russian State Medical University, Ostrovityanova Street 1, Moscow 117869, Russia. Fax: +007-095-2451220; e-mail: potap@hotmail.com

1. Introduction

Psoralen under exposure to UVA yields large amount of photoproducts. These products are being formed either with or without participation of molecular oxygen. Dark addition of products of psoralen photooxidation (POP) to biological substrates in vitro induces a variety of biological effects: POP oxidizes unsaturated lipids and proteins [1,2], induces hemolysis [2], modifies the "respiratory burst" of phagocytes [3]. POP products induce a systemic effect in vivo – they are capable of modulating immunological response when injected to mice [4]. Biological activity of POP is of our particular interest because POP is not only able to induce toxic effects but acts also as a therapeutic agent. For instance, promising results on POP-induced inhibition of grafted EL-4 lymphoma in mice were obtained [4].

In some cases, the chemical nature of POP products responsible for certain types of biological effects was identified. For example, Mizuno et al. [5] have found that 2,3-dihydro-2,9-dimethoxy-3-hydroxy-7-oxo-7H-furo[3,2-g][1]benzopyran inhibits chemotaxis of neutrophils. Hydroperoxides and dioxetanes of furocoumarins induce mutagenic effects [6,7]. As for the other types of biological activity of POP, chemical isolation of the acting photoproduct is not a trivial task to perform for the following reasons. Biologically active POP compounds are thermolabile [4], readily destroyed by electron donors [8] and even by simple dilution by the same solvent in which they were produced and stored [9].

Biological testing of the identified POP products showed very high sensitivity of such tests. For example, peroxides of alloimperatorin and imperatorin in nanomolar concentrations completely suppressed the delayed type hypersensitivity reaction in mice [10]. Biological testing, therefore, is very sensitive method for detection of labile POP products. POP-induced hemolysis is also a very sensitive test system. If, for example, formation of a single cation permeability canal in membrane leads to a single hemolytic act, then one molecule of an active agent can potentially cause lysis of one cell. In the present paper we investigated the effect of temperature on the formation of POP hemolysins (POP_hem) in psoralen aqueous solutions, as well as thermolability of POP_hem. It is known that POP_hem can be destroyed by Fe(II)-ions [8]. Interaction of POP with ferrous ions can be registered by chemiluminescence flash (ChL_Fe) [11]. Kinetics of the formation of POP hemolysins and destructible by Fe(II)-ions POP compounds (POP_ChL) has common features. The yield of both POP products increases with increase in psoralen concentration during irradiation [9,11].

In the present paper an effect of temperature on photooxidation of psoralen in solution was monitored by two methods: POP-induced hemolysis and registration of Fe(II)-dependent chemiluminescence of POP. Thermolability of photooxidative products of psoralen was studied.

2. Materials and methods

2.1. General materials

Psoralen (98% purity) was purchased from the Institute of Chemistry of Plant Substances (Tashkent, Uzbekistan). Ferrous sulfate (FeSO_4´ 7H₂O) was from Reachim (Russia). Ethanol was distilled before use.

2.2. Photooxidation of psoralen solution

Five milliliter of 0.1 mM psoralen (1% ethanol in PBS) was irradiated by the focused light of high-pressure quartz mercury lamp DRSH-250-3 (Russia) in glass cuvette under continuous stirring at different temperature maintained by water thermostat. "Hg-Mon 365" filter (Carl Zeiss, Jena, Germany) and water heat filter were used to isolate the 366 nm line. The fluence rate in the UVA region, as measured with a type 585100 Waldman UV meter (Germany), was 640 W/m².

2.3. Registration of hemolysis induced by photooxidized psoralen

Human red blood cells (RBCs) were isolated from fresh heparinized donor blood. Cells were washed twice in PBS by centrifugation (10 min, 400 g). Concentration of RBCs was further adjusted to 2 ´ 10⁷ cells/ml. Solution of photooxidized psoralen was mixed with RBCs suspension in ratio 1:1 V/V. The mixture was incubated at 37° C. In control experiments RBCs were incubated with nonirradiated psoralen solution. Hemolysis was monitored by the turbity measurement (670 nm) by means of a photoelectric colorimeter KFK-2MP (Russia). The transmittance of RBC suspension (10⁷ cells/ml) was equal to 70% (Fig. 1, inset, T_min) and increased up to 96-97% if all the cells in suspension were lysed (complete hemolysis, Fig.1, inset, T_max). A linear dependence of the transmittance on the cell concentration was observed in this transmittance range. Depending on the irradiation fluence, POP solution could induce an incomplete hemolysis when only part of the cells lysed with transmittance curves reaching the plateau below T_max (Fig. 1, inset, T_i ). The portion of the cells lysed was called an amplitude of hemolysis and calculated as: amplitude of hemolysis = [(T_i - T_min)/(T_max - T_min)] ´ 100%. Dependence of an amplitude of hemolysis on psoralen irradiation fluence allowed to estimate a fluence at which lysis of 50% suspended cells could be achieved (Fig. 1, F₅₀).

Fig. 1. Dependence of the amplitude of hemolysis on the fluence of psoralen irradiation. Psoralen was irradiated at different fluences and mixed with RBCs suspension with following registration of hemolysis at 37° C. F₅₀ - fluence of psoralen preirradiation which was used for calculation of hemolytic efficiency. Inset: T_min - starting value of transmittance before hemolysis; T_max - transmittance after complete hemolysis; T_i - transmittance after incomplete hemolysis. Each point is the mean of three experiments ± SEM.

2.4. Registration of Fe(II)-dependent chemiluminescence of POP solution

To register a ChL_Fe flash, to the mixture of (5 ml PBS + 4 ml POP) was added 1 ml of FeSO_4´ 7H₂O (1 mM in distilled water) under vigorous stirring. ChL_Fe kinetics was monitored at the same temperature as during irradiation of psoralen or storage of POP by chemiluminometer (photomultiplier FEU-39, Russia, assembled in our laboratory) connected to MacLab/2e - Macintosh LC475 controll system. An intensity of chemiluminescence was determined as an amplitude of ChL_Fe flash in arbitrary units.

3. Results

3.1. Effect of temperature on psoralen photooxidation in solution and formation of POP_ChL and POP_hem

The psoralen solution was irradiated at different temperatures. Addition of POP solution to the suspension of RBCs induced cell lysis. For quantitative evaluation of POP hemolytic activity, we assumed that (1) the same hemolytic effect (for example 50% amplitude of hemolysis) is induced in different experiments by the same amount of POP hemolysin and (2) the concentration of POP_hem is in direct proportion to the fluence of irradiation. Then the relative quantity of POP_hem produced per one unit of fluence can be determined as 1/F₅₀, where F₅₀ is the fluence of preirradiation of psoralen solution at which the amount of POP hemolysin that is formed is necessary for the destruction of 50% of the suspended cells. The value 1/F₅₀ is defined as "hemolytic efficiency," and we propose that it is directly proportional to the yield of POP hemolysin formation.

In Fig. 2, the dependence of hemolytic efficiency of POP on the temperature of psoralen solution during irradiation is presented. It is seen that POP hemolytic efficiency augmented with the temperature increase from 4 to 25° C. Further temperature increase up to 45° C leaded to the decrease in POP hemolytic efficiency. Thus, maximal yield of POP hemolysin was observed at 25° C.

ChL flash is generated upon the addition of ferrous ions to POP. Intensity of ChL_Fe is signaling about the amount of destructible by Fe(II)-ions POP compounds. Psoralen solution was irradiated at various temperature with following registration of ChL_Fe (Fig. 2). Dependence of intensity of ChL_Fe on the temperature had its maximum at 25° C like that of POP hemolysin production.

Fig. 2. Effect of temperature during psoralen irradiation on production of POP hemolysins and POP_ChL. The relative yield of hemolysins was estimated by hemolytic efficiency (curve 1), the relative yield of POP_ChL - by intensity of ChL accompaning an addition of Fe(II) to POP. Each point is the mean of three experiments ± SEM.

3.2 Effect of temperature on postirradiation storage of POP_ChL and POP_hem

Photooxidized at different temperatures psoralen was then stored at 4 or 25° C with following addition to erythrocytes or registration of ChL_Fe. Fig. 3 depicts the dependence of hemolytic efficiency of POP produced at 4 (POP₄) or 25° C (POP₂₅) on the time of its storage at either 4 or 25° C. One can see that POP₄ hemolytic efficiency was being preserved during storage at 4° C (Fig.3, curve 4/4) and decreased in storage at 25° C (Fig.3, curve 4/25). Hemolytic efficiency of POP₂₅ decreased in storage at both 4 (Fig.3, curve 25/4) and 25° C (Fig.3, curve 25/25). As can be seen, the decrease in hemolytic efficiency of POP₂₅ stored at 4° C reached the plateau after one hour and stayed equal to the hemolytic efficiency of POP₄ stored at 4° C.

Fig. 3. Destruction of POP hemolysin in postirradiational storage. Numbers at the curves indicate: the first one - temperature during irradiation, the second - temperature during storage. Points are mean of three experiments ± SEM.

Fig. 4. Destruction of POP_ChL in postirradiational storage. Numbers at the curves indicate: the first one - temperature during irradiation, the second - temperature during storage. Points are mean of three experiments ± SEM.

The kinetics of POP₄ and POP₂₅ degradation in storage at 4 or 25° C as registered by ChL_Fe method are presented in Fig. 4. The intensity of ChL_Fe of POP decreased in storage of both POP₄ and POP₂₅ at either temperature used: 4 (Fig.4, curves 4/4, 25/4) or 25° C (Fig.4, curves 4/25, 25/25). In contrast to POP_hem produced at 4° C, POP_ChL produced at 4° C was unstable at any temperature of storage and its degradation accelerated with increase in storage temperature.

4. Discussion

In the present study we tryed to clarify whether hemolysis and Fe(II)-dependent chemiluminescence were mediated by the same products of psoralen photooxidation. We have found that temperature of psoralen solution during irradiation strongly affected hemolysis and chemiluminescence. The dependencies bared an analogous extreme character. Increase in irradiation temperature from 4 to 25° C resulted in augmentation of hemolytic efficiency and chemiluminescence of POP, whereas further increase in temperature up to 45° C resulted in decrease in both effects. Such correlation argued for the identity of POP_ChL and POP_hem.

The increase in POP_hem at 25° C in comparison to 4° C could be explained either by increase in quantity of POP of the same type that was produced at 4° C or by generation of additional new POP product. To elucidate this we have studied the stability of POP produced at 4 and 25° C in storage at different temperatures. We have found that hemolytically active POP produced at 4° C consists only of the product(s) stable at 4° C (Fig.3, curve 4/4). Whereas degradation of POP_hem produced at 25° C and stored at 4° C is described by at least two exponents and therefore POP_hem produced at 25 ° C consists of at least two different compounds - stable and degrading at 4° C (Fig.3, curve 25/4).

We propose the following kinetic scheme of formation of POP hemolysins. If irradiation is conducted at 4° C then POP_hem of only one type is being formed (H₁). This H₁ is stable at 4° C. Postirradiational storage of H₁ at 25° C showed that once temperature is higher than 4° C, H₁ degrades with formation of inactive product IP₁ (Fig.3, curve 4/25). Hence, if temperature of irradiation raised up to 25° C, at least two process occure: production of H₁ and its degradation to IP₁. To explain increase in hemolytic efficiency of POP produced at 25° C comparing to POP produced at 4° C, we further suggest that IP₁ preserves the features of the chromophore and can absorb the energy of photons hn ₂ yielding an active product H₂. H₂ is hemolytically more efficient and less stable than H₁. Thermal destruction of H₂ yields inactive product IP₂:

If POP_hem and POP_ChL consisted of chemically identical products, then their thermal degradation curves at 4 and 25° C should be similar. However, kinetics of degradation of POP_hem and POP_ChL was different. None of the POP_ChL degradation curves could be viewed as monoexponential, hence, several POP products are responsible for ChL_Fe. Unlike POP_hem, POP_ChL produced at 4° C was not stable at 4° C. Chemiluminescent POP products produced at any temperature degraded completely in storage at 25° C. Therefore, our comparative analysis showed that POP products responsible for hemolysis and ChL_Fe have different chemical nature.

5. Conclusion

Biological testing can be advantageous to chemical isolation because it allows to observe highly labile biologically active products of photooxidation in very low concentrations. Such unstable photoproducts would be unavoidably lost in a process of their chromatographical isolation. Our experiments demonstrated that photooxidation of psoralen strongly depends on temperature during irradiation. By varying the temperature we can affect not only the quantity, but also the set of biologically active photoproducts. Thus, temperature factor must be taken into consideration during irradiation as well as postirradiation stages in any analytical study of photooxidative products. Also by varying the temperature during the course of photopheresis, for example, we can perhaps influence the biological effects of the treatment.

References

[1] A.Ya. Potapenko, V.L. Sukhorukov, Photooxidative reactions of psoralens, Studia Biophys. 101 (1984) 89-98.

[2] A.Ya. Potapenko, Mechanism of photodynamic effects of furocoumarins, J. Photochem. Photobiol. B: Biol. 9 (1991) 1-33.

[3] A.A. Kyagova, L.G. Korkina, T.V. Snigereva, E.P. Lysenko, S.K. Tomashaeva, A.Ya. Potapenko, Psoralen-photosensitized damage of rat peritoneal exudate cells, Photochem. Photobiol. 53 (1991) 633-637.

[4] A.Ya. Potapenko, A.A. Kyagova, L.N. Bezdetnaya, E.P. Lysenko, I.Yu. Chernyakhovskaya, V.A. Bekhalo, E.V. Nagurskaya, V.G. Nesterenko, N.G. Korotky, S.N. Akhtyamov and T.M. Lanshchikova, Products of psoralen photooxidation possess immunomodulative and antileukemic effects, Photochem. Photobiol. 60 (1994) 171-174.

[5] N. Mizuno, K. Esaki, J. Sakakibara, N. Murakami and S. Nagai, Structural elucidation of the 8-methoxypsoralen oxidized product that inhibits the chemotactic activity of polymorphonuclear neutrophils toward anaphylotoxin C5a, Photochem. Photobiol. 54 (1991) 689-701.

[6] W. Adam, M. Berger, J. Cadet., F. Dall’Aqua, B. Epe, S. Frank, D. Ramaiah, S. Raoul, C.R. Saha-Moeller and D. Vedaldi, Photochemistry and photobiology of furocoumarin hydroperoxides derived from imperatorin: novel intercalating photo-Fenton reagents for oxidative DNA modification by hydroxyl radicals, Photochem. Photobiol. 63 (1996) 768-778.

[7] W. Adam, H. Hauer, T. Mosandl, C.R. Saha-Moeller, W. Wanger, D. Wild, Furocoumarine-, naphtofuran-, and furoquinoline-annulated 1,2-dioxetanes: synthesis, termolysis and mutagenicity, Liebigs Ann. Chem. (1990) 1227-1236.

[8] N.N. Zhuravel, I.V. Belichenko, A.A. Kyagova, E.P. Lysenko, E.M. Khalilov, A.Ya. Potapenko, Activation of hemolysis induced by photooxidized psoralen (POP) by Fe²⁺ ions. The role of Fe²⁺ reactions with POP and erythrocytes, Membr. and Cell Biol. 10 (1996) 381-387.

[9] A.A. Kyagova, N.N. Zhuravel, M.V. Malakhov, E.P. Lysenko, W. Adam, Ch.R. Saha-Möller and A.Ya. Potapenko, Suppression of delayed type hypersensitivity and hemolysis induced by previously photooxidized psoralen: effect of fluence rate and psoralen concentration, Photochem. Photobiol. 65 (1997) 694-700.

[10] E.S. Andina, A.A. Kyagova, A.Ya. Potapenko, M. Moeller, H. Stopper, W. Adam, C.R. Saha-Möller, Suppression of delayed type hypersensitivity reaction by furocoumarin photooxidation products, J. Invest. Dermatol. 108 (1997) 665.

[11] I.N. Rodenko, A.N. Osipov, E.P. Lysenko, A.Ya. Potapenko, Degradation of psoralen photo-oxidation products induced by ferrous ions, J. Photochem. Photobiol. B: Biol. 19 (1993) 39-48.