Review
Photophysical properties of porphyrinoid sensitizers non-covalently bound to host molecules; models for photodynamic therapy

https://doi.org/10.1016/j.ccr.2004.02.004Get rights and content

Abstract

The binding of photosensitizers to host molecules is discussed from the perspective of how the confinement in a molecular assembly influences the sensitizer’s photophysical properties. In connection with photodynamic therapy (PDT) of cancer during which the administered sensitizer necessarily interacts with the biological material the problem becomes of utmost importance. This review surveys changes of photophysical behaviour of porphyrins, metalloporphyrins and other porphyrinoid sensitizers induced by their interaction with biopolymers (proteins, nucleic acids), liposomes or synthetic sensitizer carriers (cyclodextrins, calixarenes). The structure, charge, and physicochemical properties of the sensitizer predetermine the type of interaction with the surrounding microenvironment and are manifested by changes in absorption, fluorescence, kinetics of deactivation of the excited states, and generation of singlet oxygen. As follows from the collected data, binding of the sensitizer does not restrict formation of the excited states but influences the kinetics. Various consequences of binding on the form and photophysical parameters of the sensitizers are discussed and general features of the mutual interaction are outlined.

Introduction

Photosensitized reactions of molecular oxygen have recently found far-reaching applications in biology and medicine. Complete understanding of these reactions in the complex biological environment became a topical interdisciplinary problem that spans from photophysics over photochemistry and photobiology to photomedicine [1], [2], [3], [4], [5], [6], [7], [8], [9], [10]. The so-called photodynamic effect rests in the oxidative damage of biological material by reactive forms of oxygen generated by sensitized reactions. The photodynamically active species is singlet oxygen 1O2(1Δg) generated in situ by energy transfer from an excited sensitizer to oxygen molecule. Superoxide O2, the product of electron transfer [1], [2] is also involved, to a lesser extent. The photodynamic effect has been utilized, e.g., in photodynamic therapy (PDT) of cancer or atherosclerosis, in inactivation of some bacteria and viruses and in insecticides. The most intensively studied area is the photodynamic therapy of tumors. The treatment is based on administering the sensitizer usually by intravenous injections and, after a period necessary for retention of the sensitizer in the tumor, irradiation of the tumor by visible light. This way singlet oxygen and other reactive particles are formed directly in a tumor and destroy it from inside. Though numerous sensitizers produce singlet oxygen, almost all sensitizers studied in this context belong to the group of porphyrinoid ligands or their metallocomplexes. The collective term porphyrinoids denotes porphyrins and structurally similar macrocyclic compounds with four or more pyrrole rings.

The observed poor correlation between photophysical parameters of a sole sensitizer and its photodynamic efficacy turned attention to the influence of the biological environment. To be photodynamically active, the sensitizer needs to be closely associated with the target. Consequently, the favourable pattern of localization depends on the nature of the sensitizer (structural and physicochemical), nature of the sensitizer carrier, and on complex environmental conditions, which makes the final effect rather unpredictable. In chemical terms the influence of the environment can be attributed to non-covalent interactions of the sensitizer with surrounding molecules. Several papers have been dedicated to changes of photophysical parameters of sensitizers in the presence of proteins, nucleic acids, and some other molecules that may act as host molecules. It is worth noting the impact non-covalent interaction exerts on such intrinsic properties of the sensitizer molecule, as the photophysical quantities [3], [4], [5], [6]. Vice versa, the change of photophysical properties can be a useful tool for getting information on the topology of binding sites and on the nature of interactions by which they were evoked. Our review of this developing area of chemistry is not intended to be exhaustive. Instead, it is intended to serve as a report highlighting photophysical characteristics of non-covalently bound sensitizers and the ways how knowledge of these characteristics can help in the prediction of photodynamic action caused by novel photosensitizers.

The interaction often affects the molecular form of the sensitizer, in most cases the equilibrium between monomeric and aggregated species. According to the charge and steric arrangement of either of the interacting components, the interaction can promote monomerization or support self-assembling into organized, often chiral, aggregates. Binding-induced aggregation is not a desirable process since photodynamic efficiency decreases as a result of the poor or absent sensitizing ability of dimers and higher aggregates [6], [11]. In many respects, photodynamic effect bears upon self-assembling, formation of supramolecular structures and molecular recognition and it may be inspiring to treat the problem from this point of view. Furthermore, time-resolved methods may fill in for the lack of information about the time course of processes.

The organization of the present review is as follows. The theoretical background of sensitization, singlet oxygen formation and specific requirements on suitable sensitizers are summarized in Sections 1.1. and 1.2. In Section 2 we review the photophysical properties of singlet oxygen producing porphyrinoid sensitizers1 bound to well-defined molecules of biological importance. The photophysical characteristics including fluorescence quantum yields (Φf), quantum yields of the triplet states (ΦT), quantum yields of the singlet oxygen formation (ΦΔ), fluorescence (τf) and triplet (τT) lifetimes and rate constants of the triplet states quenching by oxygen (kq) are summarized in Tables throughout the text. Figures documenting the described effects are taken from the authors’ experiments. To the best of our knowledge, no attempt has been made at a direct comparison of the effect of binding on photophysical properties of porphyrinoid sensitizers. We believe that it can contribute to understanding and prediction of sensitizer behavior in the presence of biologically active compounds. Careful assessment of the binding effects and comparison of biopolymers and potential host molecules, such as cyclodextrins and calixarenes on the molecular level is a rational way to understand photosensitized processes in complicated biological environments. The future development of the topic is outlined in Section 3.

The oxygen molecule exhibits a series of absorption bands in the ultraviolet, visible and infrared. Although the direct photoexcitation of molecular oxygen [12] or oxygen–organic molecule charge transfer complexes [13] to produce singlet oxygen is possible, this method is not of much interest for preparative applications due to small yields of 1O2. An indirect path of excitation, the photosensitized reaction (Scheme 1A), is therefore the core of photo-initiated reactions involving oxygen. Photosensitized oxygen reactions are classified as Type I and Type II according to the nature of a quencher [14] (Scheme 1B). Quenching of the excited sensitizer by molecular oxygen (Type II reactions) proceeds as energy transfer yielding singlet oxygen 1O2 or as electron transfer yielding superoxide anion radical O2. Energy transfer from the excited triplet state of the sensitizer to the ground state (triplet) oxygen is a spin allowed process, coupled with spin inversion of oxygen to two forms of singlet oxygen 1O2(1Δg) and 1O2(1Σg) (Scheme 2)3Sens+3O21Sens+1O2Occupation of the highest antibonding orbitals 2 in the ground state and excited states is demonstrated in Scheme 2.

Electron transfer to oxygen generates doublet particles, a sensitizer cation radical and superoxide O23Sens+3O23(2Sens+2O2)↔2Sens++2O2Quenching by substrate or solvent molecules yields the corresponding radicals (Type I reactions). The reactions producing O2 in the primary step (Eq. (2)) are of Type II—oxygen reactions [14], [15], though they were originally described as Type I—radical reactions [9], [16].

The excited singlet states S1 of most sensitizers are too short-lived to be effectively quenched by oxygen and yield 1O2. Nonetheless, singlet oxygen 1O2(1Δg) can be generated provided that the energy difference ΔE (S1T1) of the sensitizer exceeds 94.1 kJ mol−1. Singlet oxygen is generally accepted as a decisive species in the photodynamic action. The less frequently generated O2 is the parent species of OHradical dot radicals formed via dismutation of O2/HO2radical dot and the catalyzed Haber–Weiss reaction. As recent studies indicate, radical reactions of Type I and thermal processes connected with radiationless transitions of the excited sensitizer may also contribute to the final photodynamic effect [11], [17], [18].

In the absence of oxygen or any chemical reaction, the lifetime τ of an excited sensitizer is related to the rate constant of the monomolecular deactivation processes byτT=1kdecayT=1kphosphorescence+kiscTfor the triplet states, andτS=1kdecayS=1kfluorescence+kic+kiscSfor the singlet states (Scheme 1). In the presence of oxygen, the observed quenching rate of the triplet sensitizer is given bykobs=kdecayT+kq[O2]where kq is the rate constant characterizing bimolecular quenching by oxygen and is expressed as the sum of oxygen dependent rate constants namely of energy transfer, electron transfer and enhanced intersystem crossing. Since diffusion controlled kq is of the order of 109–1010 M−1 s−1, the determination of the correct value of kdecayT is extremely sensitive to traces of oxygen.

The quantum yield of singlet oxygen formation ΦΔ depends on the quantum yield of the triplet states ΦT according toΦΔTSΔSqwhere SΔ is the fraction of triplet molecules quenched by oxygen and yielding 1O2 and is given bySΔ=ketkqwhere ket is the rate constant of energy transfer leading to the formation of 1O2(1Δg) and/or 1O2(1Σg) (Scheme 2, Table 1), and Sq is the fraction of oxygen dependent triplet deactivationsSq=kq[O2]kq[O2]+kdecayTThe denominator represents all pathways of triplet deactivations. If kdecayTkq[O2], then Sq≅1 and for ΦΔ holds the simplified equation [11], [19]ΦΔTSΔ

In the context of photodynamic processes, the more stable 1O2(1Δg) form has been considered and monitored so far, and is referred to simply as “singlet oxygen”. The growing interest in the formation and deactivation of 1O2(1Σg) in a solution is motivated by an effort to acquire important information about the mechanism of photosensitized oxygen reactions and about solvent effects on forbidden radiative transitions [20], [21]. Studies on 1O2(1Σg) have been fostered by the development of sophisticated time-resolved spectroscopic methods [21], [22], [23], [24].

Quenching of the sensitizers in the triplet states by oxygen in some organic solvents was shown to directly produce 1O2(1Σg) and 1O2(1Δg) in the primary step of energy transfer [25].2 The obvious stipulation is that the T1 energies of the sensitizer must exceed the energy difference 157 kJ mol−1 between 1O2(1Σg) and ground state oxygen. The branching ratios 1Σ/1Δ depend on the nature of the sensitizer and the solvent and for the reported systems varied between 1.7 and 0.4. Apparently, a considerable fraction of 1O2(1Σg) is generated in solution.

The energy level diagram and transitions between electronic states of oxygen are shown in Scheme 2. The competing processes of radiative and non-radiative decay of 1O2(1Σg) and 1O2(1Δg) are solvent dependent because solvent perturbations enhance the probability of spin forbidden transitions 1Σg3Σg and 1Δg3Σg. The rate constants of radiative decay of 1O2(1Σg) and 1O2(1Δg) in a solution are significantly lower than are those of non-radiative processes. Hence, it is the non-radiative decay that prevails in solutions. With respect to deactivation, the spin allowed path 1O2(1Σg) → 1O2(1Δg) is dominant with almost unit efficiency. The viability of the path is supported by the fact that a lesser amount of energy ΔE=62.8 kJ mol−1 needs to be dissipated. It follows that 1O2(1Δg) originates from oxygen quenching of triplet sensitizer through two reaction paths—one direct, the other indirect via 1O2(1Σg) [21], [25]. Speculation that 1O2(1Σg) can participate in photosensitized oxidation have not been confirmed [25]. This justifies the general belief that 1O2(1Δg) is the decisive oxidation agent in photodynamic processes. The relevance of 1O2(1Σg) for photodynamic processes rests in the fact that it can be a precursor of 1O2(1Δg) since practically all 1O2(1Σg) decays to 1O2(1Δg) [25]. Some of the characteristics of both forms of 1O2 are summarized in Table 1.

Singlet oxygen can be consumed in two competing ways: (i) physical quenching of 1O2 by a quencher that becomes electronically excited (bimolecular) or deactivation proceeding by vibrational excitation of solvent molecules (monomolecular). (ii) Oxidation of a molecule by 1O2 (chemical reaction). Quenching and oxidation of substrates in the ground singlet state are spin allowed reactions. The mechanism of singlet oxygen reactions—solvent deactivation and reactions with quenchers and substrates—are visualized in Scheme 3.

The fraction of 1O2 molecules that react with the substrate S to oxidized substrate Sox isfr=kr[S]kr[S]+kd+kp[Q]where kr is the rate constant of oxidation reaction, kp the rate constant of physical quenching by a quencher Q, and kd the rate constant of deactivation in the absence of S and Q. If no quencher is present, the quantum yield of the oxidized substrate Φr is given byΦrΔfrΔkr[S]kr[S]+kdClearly, when kr[S]⪢kd, the above equation simplifies to Φr=ΦΔ; in this case entire 1O2 produced is consumed in a reaction with the substrate (saturation effect) [29].

Due to the high reactivity of 1O2 with substrates, the chemistry of 1O2 is very rich [19], [29], [30], [31]. Still, the reactions have certain selective features. Typical reactions involving CC bonds, isolated or conjugated, are oxidations of olefins (ene-type reactions, [2+2] cycloadditions), 1,3-dienes ([4+2] cycloadditions), aromatic compounds and heterocycles. The intermediates are peroxo species, such as perepoxides, dioxethanes and endoperoxides. Thiocompounds are oxidized to sulfoxides, and phosphines to phosphine oxides [29]. In the context of this review singlet oxygen reactions with constituents of proteins, lipids and DNA are relevant.
Amino acids (amino acid residues in proteins). Among essential amino acids the most prone to oxidation with 1O2 are cysteine, methionine, histidine, and tryptophan [26], [29], [32]. Due to their reactivity, these amino acids are the primary target of an oxidative attack on proteins. The reaction mechanisms are rather complex and as a rule lead to a number of final products. Cysteine and methionine are oxidized mainly to sulfoxides, histidine yields a thermally unstable endoperoxide, tryptophan reacts by a complicated mechanism to give N-formylkynurenine.
Lipids. Unsaturated lipids typically undergo ene-type reactions [33].
DNA. Of the four nucleobases guanine is the most susceptible to oxidation by 1O2. The reaction mechanism has been extensively studied in connection with oxidative cleavage of DNA [34], [35], [36]. The first step is a [4+2] cycloaddition to the C-4 and C-8 carbons of the purine ring leading to an unstable endoperoxide. The subsequent complicated sequence of reactions and the final products depend on whether the guanine moiety is bound in an oligonucleotide or a double-stranded DNA [37].

Generally, the reactions of 1O2 are sensitive to steric factors. A novel approach pays special attention to the effects of the environment namely to the ability of supramolecular structures to control conformation of the built-in substrates and/or sensitizers. This approach is expected to yield relevant information on reactions proceeding in biological systems on the cellular level. The supramolecular structures can be composed of micelles, cyclodextrins, zeolites, calixarenes, etc. Recently, the effects of micelles, cyclodextrins or synthetic membranes as host structures on singlet oxygen reactions were reviewed [31].

Numerous sensitizers produce singlet oxygen and can be considered suitable for PDT, nevertheless, the majority of sensitizers investigated in this context were compounds with a porphyrinoid structure [7], [8], [9], [10], [33]. The reason for this preference is the extensive knowledge of their chemistry together with the inherent similarity to natural porphyrins frequently occurring in living matter. Of non-porphyrinoid sensitizers eosine, acridine, rose bengal, methylene blue, perylenequinones, triarylmethane dyes, etc. have been taken into consideration. The effort to correlate the structure of the sensitizer to intermolecular non-covalent interactions produced a categorical requirement to use well-defined individual sensitizers. Mixtures of derivatives, species differing in the number of substituents or even different regioisomers, interacting differently with biomolecules in the tissue, often yield misleading results on structure–activity relationships [7], [33].

Sensitizers suitable for PDT have to meet a number of specific requirements:

  • Maximum absorption in the region 600–800 nm [7], [33], [38]. The incident intensity of light—irradiance—is reduced by absorption by chromophores in tissue or by scattering. The efficiency of scattering increases as the wavelength is decreased. On the other side, absorption by water molecules increases at wavelengths above 800 nm. Consequently, the window for optimum penetration lies between 600 and 800 nm, i.e. in the region of red light. The characteristic quantity is the penetration depth δ defined as the depth in which the irradiance is reduced to 1/e of the initial value. Typical values of δ vary between 1 and 3 mm, but apparently the photodynamic effect reaches beyond this limit. The upper limit of the wavelength to produce 1O2 is given by the energy necessary for the 1O2 formation (λ<1269 nm, ΔE>94.1 kJ mol−1 for a one-photon process).

  • Minimum absorption in the region 400–600 nm [11], [39], [40]. Sensitizers absorbing in this region, i.e. in the maximum of the spectroscopic distribution of daylight, enhance photosensitivity of the skin, which is a complicating side effect of photodynamic treatment. Photosensitization of skin has been the major drawback of first generation sensitizers based on hematoporphyrin derivatives.

  • High quantum yields of 1O2. The quantum yields ΦΔ vary in the range 0.3–0.8 for most sensitizers. The significance of ΦΔ should not be overestimated, since the amount of 1O2 produced depends strongly on other factors, namely on interaction of the sensitizer with surrounding biopolymers, aggregation of the sensitizer, oxygen depletion and side reactions. The demand of high ΦΔ includes the prerequisite of adequately high ΦT, the triplet state energy sufficient for the 1O2 formation and relatively long lifetime of the triplet states τT.

  • Photostability. A sensitizer should be stable against photodegradation and against oxidation by 1O2 or other reactive oxygen species generated in situ. Photobleaching of sensitizers in biological systems is a complex process, not necessarily oxygen dependent. Photobleaching plays an important role in decreasing skin sensitization and the specificity of phototreatment [7], [8], [40], [41].

  • Non-toxicity and phototoxicity. Low “dark” toxicity (e.g. nephrotoxicity, neurotoxicity) is desirable so as to avoid unnecessary strain on the organism prior to irradiation. The overall destructive photodynamic effect of the sensitizer on biological material in vitro or in vivo is called phototoxicity.

  • Specific retention in the malignant tissue. The specific retention of a sensitizer in the malignant tissue is a consequence of different kinetics of sensitizer removal from the malignant and healthy tissue. The removal from the healthy tissue is faster. The concentration difference is adjusted within several hours after sensitizer administration and depends on the nature of the sensitizer. Effectual removal of the sensitizer from the healthy tissue precludes its photodynamic damage [4].

  • Single substance. The use of a single, well-defined substance is necessary for reliable evaluation of the sensitizer–biopolymer interaction [7], [33].

  • Fluorescence. Fluorescence of the sensitizer enables detection of the sensitizer distribution in vivo. To retain both functions of the sensitizer, namely fluorescence and 1O2 production, the ratio Φf/Φisc should be optimised [11].

  • Solubility. Sufficient solubility of the sensitizer in aqueous media is important for direct intravenous application and transport to the intended target location. Hydrophobic, insoluble sensitizers can be transported by water-soluble carriers [17], [31], [42]. The role of possible carriers will be discussed in Section 2.

It is not the intention of this review to provide an exhaustive description of the numerous porphyrinoid sensitizers applicable in PDT. Several excellent reviews exist [7], [8], [9], [10], [33], [43], [44], [45], [46]. Basic skeletons are presented in Scheme 4. Structures 14 were originally natural products, whose synthetic analogues are used as a rule. Synthetic porphyrins 1 are often substituted in the meso-positions. Chlorin 2 and bacteriochlorin 3 are partly hydrogenated porphyrins; hydrogenation shifts the absorption bands to the more advantageous region above 600 nm. Purpurins 4 are degradation products of chlorophylls. Compounds 512 are purely synthetic and have no counterpart in nature. Porphyrazine or tetraazaporphyrin 5 is the basic skeleton of the extensively used phthalocyanines 6 and naphthalocyanines 7 [47]. Among novel types of porphyrinoid compounds 8–12 are new sensitizers, some of which are very promising for PDT. Porphycenes 8 are isomeric porphyrins with differing length of the pyrrole–pyrrole bridges [48], [49]. Expanded porphyrins 9 are characterized by an increased number of atoms separating the pyrrole rings. The most important expanded porphyrins are sapphyrins 10 and texaphyrins 11 [9], [43], [50]. A trend opposite to expansion of the macrocycle follow subphthalocyanines 12 with three diiminoisoindole rings bound to a central boron atom [51]. Generally, porphyrinoid sensitizers can be free ligands or metallocomplexes with Al, Zn, Mg, Ga, Si, Ge, Sn, or lanthanides [9], [33], [43], [48], [49] central ions. Meso-substituted porphyrins, chlorins and some expanded porphyrins are usually applied as free ligands (non-metallated). On the other hand, phthalocyanines and naphthalocyanines are always metallated since the free ligand is less chemically stable [7]. Complexes with transition metals are poor sensitizers because of their short triplet lifetimes ranging from picoseconds to nanoseconds.

The following paragraphs review the principal types of sensitizer arranged according to the charge of the peripheral groups. The charge, its sign and distribution, and hydrophilicity or hydrophobicity of the sensitizer predestine the mode of interaction with biomolecules and/or carriers and, consequently, photophysical properties, the fate, and effectiveness of the sensitizer in a biological system. The tetrapyrrolic sensitizers, anionic or cationic, with three or four charged substituents are hydrophilic (polar). Monosubstituted and disubstituted sensitizers behave as amphiphilic molecules (vide infra). Symmetric disubstitued sensitizers have more hydrophilic character than the asymmetric [8], [33].

Of the anionic sensitizers sulfonated or carboxylated metallophthalocyanines and meso-tetraphenylporphyrins have been most frequently investigated. In vitro and in vivo studies have been carried out in order to map a correlation between sulfonation degree and photodynamic activity [4], [7], [8], [11], [33], [52]. Anionic sensitizers interact with proteins possessing positively charged side chains—protonated amino nitrogen atoms. Highly basic proteins can compensate repulsion between the negatively charged sensitizer molecules and promote aggregation of the sensitizer [53]. The observation that the cationic sensitizer TMPyP intercalates into DNA at the G–C base pairs [54] and the capacity to penetrate into the nucleus [55] gave impetus to extensive studies of novel cationic sensitizers. The positive charge is usually localized on the protonated pyrrole nitrogen atoms (e.g. sapphyrins) and the quaternized pyridinium or ammonium nitrogen. The cationic sensitizers are less numerous than the anionic. Besides the most widely used TMPyP [55] cationic tetratolylporphyrins [56], [57], pyridinoporphyrazines [47], [58], and pyridinium or trimethylammonium phthalocyanines [59], [60] have also been investigated.

The class of hydrophobic sensitizers involves a number of uncharged species among which phthalocyanines and naphthalocyanines prevail over porphyrins and porphycenes. The advantage of hydrophobic sensitizers in PDT lies in their affinity to lipid membranes. Axial ligands, such as cholesterol coordinated to central ions of phthalocyanines and naphthalocyanines or apolar peripheral substituents (carotenoid chain, butoxy groups) can be used to enhance the hydrophobicity/lipophilicity of the sensitizer and hence its uptake by the cells. Due to the insolubility of the hydrophobic sensitizers, a suitable carrier (lipid emulsion, liposomes, micelles, cyclodextrins) is to be used for their administration [6], [7].

The amphiphilic sensitizers possess separated hydrophilic and hydrophobic regions that can independently interact with other adjacent molecules [7]. The amphiphilic sensitizers are photodynamically more active than symmetric hydrophilic or hydrophobic sensitizers [8]. The activity is not necessarily correlated with photophysical properties of the isolated molecule in solution. Systematic studies of variously sulfonated Al phthalocyanines AlPcSn and tetraphenylporphyrins (1≤n≤4, n is the number of the sulfonato groups) as model sensitizers have unambiguously shown maximal activity of unsymmetrical disulfonated compounds. Important amphiphilic sensitizers, confirming good prospects for this class in PDT, are 5,10,15,20-tetrakis(m-hydroxyphenyl)chlorin (THPC) and benzoporphyrin derivatives (verteporphin) [7], [9], [33].

The approximately planar porphyrinoid sensitizers tend to form stacked dimers and higher aggregates in polar solvents, held together by π–π interactions of the aromatic rings and by hydrophobic interactions. As it is known from experimental observation, aggregated sensitizers produce very little 1O2 and have low, if any, photodynamic activity [6], [11]. Dis-aggregation as a consequence of interaction between the sensitizer and biopolymers (proteins, DNA) will be discussed in the following sections.

Section snippets

Non-covalent interactions

In this section, we present brief introduction to the types of non-covalent interactions that are responsible for molecular assembling. By non-covalent interactions are understood weak binding forces, by whose action assemblies of molecules arise, but not new molecules, as with a chemical reaction. Interactions govern the structure and stability of assemblies, and play a decisive role in molecular recognition.

The common features of non-covalent interactions are the distinctly lower bond

Conclusions and outlook

In this review, we have concentrated on porphyrinoid sensitizers the changes of whose photophysical properties have been a consequence of non-covalent interaction with other molecules-biopolymers or abiotic container molecules. From the data summarized throughout the text it can be concluded that the non-covalent interaction of the sensitizers does not restrict the formation of the excited singlet states, triplet states and hence the formation of 1O2 which plays the dominant role in

Acknowledgements

This research was supported by the Grant Agency of the Czech Republic (grants no. 203/01/0634, 203/02/0420 and 203/02/1483.) We thank the many colleagues, who participated in various parts of the projects described in our review. In particular, we are indebted to Pavel Kubát, Pavel Anzenbacher, Pavel Lhoták, Jan Sejbal and Vladimı́r Král.

References (193)

  • B.W. Henderson et al.

    Photochem. Photobiol.

    (1992)
  • R. Schmidt et al.

    Chem. Phys. Lett.

    (1993)
  • L.A. Martinez et al.

    J. Photochem. Photobiol. B: Biol.

    (2000)
  • E.L. Clennan

    Tetrahedron

    (2000)
  • S. Stolik et al.

    J. Photochem. Photobiol. B: Biol.

    (2000)
  • G. Jori

    J. Photochem. Photobiol. A: Chem.

    (1992)
  • M.C. DeRosa et al.

    Coord. Chem. Rev.

    (2002)
  • K. Lang et al.

    J. Photochem. Photobiol. A: Chem.

    (1998)
  • A. Villanueva

    J. Photochem. Photobiol. B: Biol.

    (1993)
  • A. Minnock et al.

    J. Photochem. Photobiol. B: Biol.

    (1996)
  • D.A. Fernández et al.

    J. Photochem. Photobiol. B: Biol.

    (1997)
  • H.-J. Schneider et al.

    Tetrahedron

    (2002)
  • P. Lugo-Ponce et al.

    Coord. Chem. Rev.

    (2000)
  • K. Lang et al.

    J. Photochem. Photobiol. B: Biol.

    (2000)
  • V.S. Chirvony et al.

    J. Photochem. Photobiol. B: Biol.

    (1997)
  • N.N. Kruk et al.

    J. Photochem. Photobiol. B: Biol.

    (1998)
  • N.N. Kruk et al.

    J. Photochem. Photobiol. B: Biol.

    (1998)
  • D. Kessel

    Cancer Lett.

    (1986)
  • R. Nilsson et al.

    Photochem. Photobiol.

    (1972)
  • C. Schweitzer et al.

    Chem. Rev.

    (2003)
  • B.M. Aveline et al.

    Photochem. Photobiol.

    (1999)
  • A.E. Lissi et al.

    Chem. Rev.

    (1993)
  • F. Riccheli et al.

    Photobiol. B: Biol.

    (1995)
  • R.B. Boyle et al.

    Photochem. Photobiol.

    (1996)
  • I.J. MacDonald et al.

    J. Porphyrins Phthalocyanines

    (2001)
  • H. Ali et al.

    Chem. Rev.

    (1999)
  • C.M. Allen et al.

    J. Porphyrins Phthalocyanines

    (2001)
  • D. Phillips, in: T.J. Kemp, R.J. Donovan, M.A.J. Rodgers (Eds.), Progress in Reaction Kinetics, vol. 22, 1997, p....
  • I.B.C. Matheson et al.

    J. Am. Chem. Soc.

    (1972)
  • P.R. Ogilby et al.

    Macromolecules

    (1990)
  • C.S. Foote

    Photochem. Photobiol.

    (1991)
  • D.M. Wagnerová

    Z. Phys. Chem.

    (2001)
  • K. Gollnick

    Adv. Photochem.

    (1968)
  • G. Jori, in: W.M. Horspool, Pill-Soon Song (Eds.), Organic Photochemistry and Photobiology, CRC Press, Boca Raton,...
  • A. Harriman, in: W.M. Horspool, Pill-Soon Song (Eds.), Organic Photochemistry and Photobiology, CRC Press, Boca Raton,...
  • R.V. Bensasson, E.J. Land, T.G. Truscott, Excited States and Free Radicals in Biology and Medicine, Oxford University...
  • P.R. Ogilby

    Acc. Chem. Res.

    (1999)
  • D. Weldon et al.

    Photochem. Photobiol.

    (1999)
  • T. Keszthelyi et al.

    Photochem. Photobiol.

    (1999)
  • T. Keszthelyi et al.

    J. Phys. Chem. A

    (2000)
  • C. Schweitzer et al.

    J. Phys. Chem.

    (2003)
  • F. Wilkins et al.

    J. Phys. Chem. Ref. Data

    (1981)
  • R.Y.N. Ho, J.F. Liebman, J.S. Valentine, in: C.S. Foote, J.S. Valentine, A. Greenberg, J.F. Liebman (Eds.), Active...
  • C.S. Foote, E.L. Clennan, J.S. Valentine, in: C.S. Foote, J.S. Valentine, A. Greenberg, J.F. Liebman (Eds.), Active...
  • B.M. Monroe, in: A.A. Frimer (Ed.), Singlet O2, vol. I, CRC Press, Boca Raton, Florida, 1985, p....
  • P. Kubát et al.

    Z. Phys. Chem.

    (1999)
  • R. Bonnet, Chem. Soc. Rev. (1995)...
  • J. Cadet, P. Vigny, in: H. Morrison (Ed.), Bioorganic Photochemistry; Photochemistry and the Nucleic Acids, vol. 1,...
  • I.E. Kochevar, D.A. Dunn, in: H. Morrison (Ed.), Bioorganic Photochemistry; Photochemistry and the Nucleic Acids, vol....
  • B. Armitage

    Chem. Rev.

    (1998)
  • Cited by (0)

    View full text