A first line of defence against ROS is, of course, the protection against their formation. However, the interception of the damaging species once formed, to prevent it from further deleterious reactions, is also a deactivation strategy of defence. In general, free radical scavengers neutralize the radical species by donating one of their own electrons. The quenching agents themselves are not particularly toxic before and after the electron donation [105].

Three different types of quenching are possible, which include the transfer of the radical character with the formation of a reactive scavenger-derived radical; trapping of free radicals with the formation of a stable or inert free radical trap; and molecules which mimic quenching enzyme activities. In general, scavenger molecules either prevent free radicals from being formed or remove them before they can damage vital molecular components [105].

Several free radical scavengers have been used to evaluate the specific involvement of type I mechanism during mammalian viruses and bacteriophages PDI with different PS (Table 2).

In general, the action of chemical singlet oxygen quenchers involves the reaction of singlet oxygen with the quenching agent, producing an oxidized product. Another possibility is the deactivation of singlet oxygen to ground state (³O₂) by physical quenching, achieved by either energy or charge transfer, without consumption of oxygen or product formation [101,109]. Residues of histidine, tryptophan and tyrosine in proteins are considered to be major natural quenchers of singlet oxygen [110].

Several singlet oxygen quenchers have been used to evaluate the specific involvement of type II mechanism during viral PDI with different PS (Table 3).

From the four DNA bases, guanine is the most susceptible component to suffer a type I photosensitization reaction, due to the fact that it exhibits the lowest oxidation potential among DNA bases and it is the only base that can be oxidized by singlet oxygen (type II process) [116].

The treatment of viruses with MB and other heterocyclic dyes resulted in the damage of viral DNA [53,65,75,76] either by base modification or base loss, single strand breaks (SSB), or cross-links of DNA with proteins [34,75,81,88,117]. It is known that cationic porphyrins can bind to nucleic acids via intercalation into base pairs or self-stacking, inducing lesions upon photoinactivation due to the easy oxidation of guanine residues [118,119,120].

The binding of cationic porphyrins to DNA is presumably due to the electrostatic interaction between the positively-charged substituents in the porphyrin macrocycle and the negatively charged phosphate oxygen atoms of DNA [120]. However, porphyrin binding to DNA is not a prerequisite for an efficient photosensitization, since free porphyrins are more effective in virus inactivation than the DNA-bound species [88]. This observation, which is in conflict with the generally accepted idea that the porphyrin molecule must be in close vicinity with the site of photosensitized damage, may be explained by the lower quantum yield of singlet oxygen by the bound porphyrin when compared with the free one [88].

RNA has been suggested to be a key factor in viral PDI with many PS, but direct evidence of a correlation between RNA damage and infectivity loss has not been reported yet, as is the case of VSV when treated with phthalocyanine derivatives [81]. In RNA, as for DNA [71], guanine is suggested as the major target for oxidation by photosensitizing agents and light.

It has been shown that enveloped viruses can be inactivated due to protein damage [30,82,83,131]. However, while the same treatment is reported to be ineffective against some non-enveloped viruses [83,131], the results from Wong et al. [85] showed that even a non-enveloped virus can be efficiently inactivated due to the damage induced by PDI to its viral proteins (Table 4).

The proteins in the viral envelope of HSV-1 were considered to be major targets of merocyanine 540 photosensitization [45]. Some phthalocyanine derivatives have been shown to induce cross-links in HSV protein that might be responsible for the observed loss of infectivity [132]. Protein analysis by SDS-PAGE, after treatment with phthalocyanine derivatives, revealed irreversible changes in the HSV-1 envelope proteins, which were reflected by the loss of many proteins, the appearance of cross‑linked material on the top of the gel and by alterations in the molecular mass and molecular charge of the proteins. These alterations contribute, in all likelihood to HSV-1 inactivation [83].

In VSV treated with 3.75–30 μL mL⁻¹ of chlorophyll derivatives and light, the M protein band was not detected, which was accompanied by a decrease in the intensity of the G protein band [77]. Large complexes of proteins were also detected on the top of the gel, indicating that viral PDI cross-linked the proteins [77]. Using a fusion assay and protein analysis, it was shown that MB and AlPcS₄ caused a decrease in the intensity of the G-protein (which is known to play a crucial role in binding VSV to the host cell) band and a slight decrease in the intensity of M protein (matrix protein) band and protein cross-links. However, the observed damage in viral proteins could not account for VSV PDI [82]. VSV was inactivated by MB and phthalocyanine derivatives, which inhibited the fusion of the virus envelope to Vero cells. However, the degree of this inhibition was small compared to the extent of virus inactivation (43% inhibition vs. 4.7 log or 99.998% inactivation, for MB) [81]. Abe and Wagner [81] also found few changes in the relative abundance of VSV G protein after MB and AlPcS4 phototreatment, and they also observed additional protein bands on SDS-PAGE analysis [81]. It was found, by Western blot analysis, that HIV-1 p24 and gp120 proteins were altered in size, possibly due to cross-linking, after MB phototreatment [34]. However, using the same PS, AlpcS₄ and MB, no changes in protein patterns after SDS-PAGE of the viral proteins were observed, under conditions that caused complete VSV inactivation [135].

The results from Vzorov et al. [36] indicated that the porphyrins inhibited the cell fusion activity of HIV Env proteins (a biological function that is important for viral entry as well as induction of viral cytopathic effects) when expressed from recombinant vectors. These results showed that the viral Env protein is an important target of these compounds [36].

PDI of influenza virus by rose bengal altered the HA fusion protein and led to protein cross-links [108].

Photoinactivation of vaccinia virus with rose bengal significantly altered the concentration and oxidized histidine in vaccinia virus protein, suggesting that inactivation was attributed to alterations in viral proteins, as opposed to nucleic acids [86].

Treatment of of influenza and Sindbis viruses by hypericin [30], lead to an extensive cross-linking of the envelope proteins, which may have impaired the capacity of the viruses to adhere to and penetrate the host cells.

The protein profile of the non-enveloped enterovirus 71 was considerably altered after a low dose PDI and a MB concentration ≥0.5 μM, as revealed by a smearing and the disappearance of several protein bands [85]. However, enterovirus 71 PDI was also due to damages in the viral genome [85].

In spite of the limited available data for enveloped bacteriophages, substantially higher photoinactivation rates compared with other non-enveloped phages were described [94]. The photoinactivation by merocyanine 540 of four bacteriophages, two non-enveloped phages without lipids (phi X174 and T7), a non-enveloped phage with lipids (PRD1), and an enveloped phage with an external lipoprotein envelope (phi 6) was studied by Lytle et al. [94]. The survival curves of the different viruses clearly demonstrated different levels of sensitivity to photoinactivation by this PS, with phi 6 being the most sensitive, followed by T7 (21-fold less sensitive). While both PRD1 and phi 6 have lipid components, only phi 6 was photoinactivated by the PS. Thus, the internal lipid components of PRD1 were not sufficient to allow photoinactivation by merocyanine 540. A higher inactivation rate with a fullerene derivative was also observed by Hotze et al. [57] for a phage without lipids (T7 phage) than for PRD1 phage. The dissimilarities in phage composition resulted from differential resistance to singlet oxygen by the outer structures, since PRD1 has a double capsid with an internal lipid membrane, whereas T7 has a single proteinaceous capsid lacking lipids, and both phages contain double stranded DNA with similar GC content (48% for T7 and 51% for PRD1) [57]. Phage proteins were significantly affected by photosensitization (30–92%) when compared to the relatively smaller effect on nucleic acids in both PRD1 and T7, and lipids in PRD1 phage (≤13%), as assessed by FTIR spectra analysis [57]. The higher T7 phage inactivation is consistent with greater damage to its proteinaceous capsid. Besides this, SDS-PAGE analysis further evidenced that oxidative cross-linking of capsid proteins induced by exogenous singlet oxygen is the likely cause of phage inactivation [57]. The high propensity for MS2 phage inactivation by this PS (compared to PRD1 and T7 phages) possibly arises from damage to its A protein, which is necessary for infecting its host Escherichia coli since it contains highly reactive amino acids such as methionine, cysteine, histidine, and tyrosine and not to damages to the nucleic acid [57]. Glycosylated substituted porphyrins led to structural changes at the protein capsid and/or loosening of the protein-DNA interaction, which can be responsible for T7 phage inactivation [64]. Besides of the alteration of the DNA structure, the phototreatment pointed to significant alterations in the protein structure and/or in the DNA-protein interaction, which may be the cause of photodynamic inactivation [87,88]. The alterations in the DNA secondary structure might also be the result of photochemical damage in phage capsid proteins and consequent disruption of the phage particle. Photomodification of core proteins can also lead to phage inactivation, even if the primary structure of the DNA part is preserved, since these proteins play an important role in the early events of infection and DNA penetration [87]. The damage of T7 nucleoprotein is a complex process and clearly both phage DNA and protein capsid are affected by photoreactions [88]. Irradiation of Qβ bacteriophage in the presence of increasing concentrations of MB resulted in exponentially increasing amounts of viral RNA-protein cross-linkage products, and this is probably the most important event in viral inactivation [92]. The RNA genome of Qβ bacteriophage contained sufficiently lethal lesions following MB plus light exposure to account for the resulting phage inactivation. Nevertheless, the data also indicate that the protein component of the phage somehow contributes to the inactivation of the phage [122]. The protein component of Qβ phage is involved in the process of photoinactivation because the formation of protein carbonyls and RNA-protein cross-links were efficiently formed by MB plus light exposure [89]. The close correlation of cross-link formation with phage inactivation and the expectation that even one such cross-link in a phage genome would be lethal makes the RNA-protein cross-link lesion a strong candidate for the primary inactivating lesion of Qβ phage exposed to MB and light [122].

Little alteration of M13 phage proteins on SDS-PAGE after MB and AlPcS₄ photoinactivation was observed by Abe and Wagner [81]. The results of Zupán et al. [136], suggested that the tetracationic porphyrin meso-tetrakis(1-methylpyridinium-4-yl)porphyrin did not interact with capsid proteins and did not disturb protein-DNA interaction, even if it has a strong stabilization effect on the intraphage DNA.

Complete inactivation of VSV (4.2 log) can be obtained by treating it with 1.0 μM of the anionic phthalocyanine derivative AlPcS₄ and 5 min illumination with red light. For the neutral phthalocyanine derivative (Pc₄), complete inactivation (4 log) was achieved using a much lower amount of PS (4.5 nM) in combination with 10 min illumination [82]. The inactivation of VSV in PBS showed a linear relationship with illumination time [82]. Inactivation of the fusion activity of VSV, influenza and Sendai viruses was reached with nanomolar concentrations of hypericin and rose bengal and was absolutely dependent upon light and increased with increasing time of illumination [30]. HAV in PBS or plasma was completely inactivated within 10 min (>3.7 log) by the cationic symmetric porphyrin meso-tetrakis(1-methylpyridinium-4-yl)porphyrin. In contrast, inactivation of HAV to 3.6 log with the anionic symmetric porphyrin meso-tetrakis(4-sulfonatephenyl)porphyrin required 90 min [44]. The rate and extent of inactivation appeared to vary with the nature of the meso substituent groups [44]. HIV and VSV lost infectivity upon illumination with hypericin and rose bengal in a concentration-dependent manner [30].

MS2 phage inactivation has been observed with neutral porphyrin derivatives. However, this required higher irradiation periods (30 min) than for the cationic ones (1 min) [44]. Neutral glycosylated substituted porphyrins can also significantly photoinactivate the T7 phage [64,87]. The T4-like phage PDI was achieved by exposing the phage in the presence of six cationic porphyrins at different concentrations (0.5, 1.0 and 5.0 μM) to white light for 270 min. The results showed that phage photoinactivation varied according with the PS concentration, with higher concentrations being the most efficient ones [66]. The T4-like phage PDI also varied with the number of porphyrin charges, with tri- and tetracationic porphyrin derivatives being more effective in viral inactivation that the dicationic ones, which inactivated the phage below the limit of detection. Tetra- and tricationic porphyrin derivatives (meso-tetrakis(1-methylpyridinium-4-yl)porphyrin and 5-(pentafluorophenyl)-10,15,20-tris(1-methylpyridinium-4-yl)porphyrin, respectively) lead to complete T4-like phage inactivation (~7 log) after 270 min of irradiation with 40 W m⁻² [66]. This tetracationic porphyrin showed similar results in another study (7 log of reduction) for lambda phage inactivation, when irradiated with light of 658 nm [58]. Increasing porphyrin concentration at a fixed light dose leads to increased viral inactivation [58]. A concentration-dependent effect was also detected with a porphyrin derivative [87], but over 2.0 μM of PS the process was saturated. A further increase in porphyrin concentration did not lead to a higher inactivation rate of T7 phage. Aggregation and/or photobleaching of PS are likely explanations [87]. Cationic meso-tetrakis(1-alkylpyridinium-4-yl)porphyrin derivatives with different alkyl substituent groups were tested for MS2 phage inactivation but, with the exception of 5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrin, showed toxicity even in the absence of light [44].

In a study conducted by Gábor et al. [64], the porphyrin derivative with symmetrical glycosylated groups was found to be twice as effective as the asymmetrical one on the inactivation process of T7 phage. According to Costa and colleagues [66], the rate of T4-like phage inactivation was also dependent on the lipophilic character of the meso-substituent groups. The presence of a lipophilic aryl group in one of the meso positions of the porphyrin core appears to have an important role in phage inactivation, affecting the rate and efficiency of T4-like phage [66]. Casteel et al. [44] have also observed differences in the photoinactivation rate of MS2 phage when they used PS with different alkyl substituent groups and concluded that the rate and extent of inactivation appeared to vary with the nature of the meso substituent groups.

The effects of dengue virus inactivation were increased with the increase of MB concentration, the enhancement of power density of the light source and the extension of illumination time, as well as the decrease of illumination distance. This enabled the narrow bandwidth light system to kill or inactivate the enveloped virus at much greater distance in much shorter time [74]. VSV in the presence of MB was rapidly inactivated by red (provided by LED incident light at 272 W cm⁻²) or green-yellow light (provided by low-pressure sodium lamps at a fluence rate of 165 W cm⁻²) but slower by white light (provided by a bank of fluorescent tubes at a fluence rate of 42 W cm⁻²) [46], showing that higher power densities produce a high rate of viral inactivation than low fluence rates. Wagner et al. [164] also showed that red light of 9 W m⁻², given at a total dose of 1.8 × 10⁴ and 3.2 × 10⁴ J m^-2, inactivated MB-treated VSV by 6 and ≥7 log, respectively. VSV inactivation was linearly dependent on the fluence rate of red light illumination [165].

In terms of what is known about phage PDI, only one study focusing on the effect of different light sources and power densities [67] exists. In this study, cationic porphyrin derivatives (meso-tetrakis(1-methylpyridinium-4-yl)porphyrin and 5-(pentafluorophenyl)-10,15,20-tris(1-methylpyridinium-4-yl)porphyrin), when irradiated with different sources of light (fluorescent PAR lamps, sun light and halogen lamp) with fluence rates ranging from 40 W m⁻² to 1690 W m⁻², efficiently photoinactivated non-enveloped phages. All light sources tested lead to reductions of about 7 log for the somatic T4-like phage. However, the rate and the extent of inactivation were dependent on the light source, namely when low fluence rates were used (40 W m⁻²) and on the energy dose, being considerably more effective when light was delivered at a lower fluence rate. However, depending on the light source used, different irradiation periods were required to inactivate T4-like phage to the limits of detection. The results also showed that the efficacy of T4-like phage inactivation, using the same fluence rate, was dependent on the light source used, in particular when the light is delivered at a low fluence rate. M13 phage was phototreated with 5.0 μM MB and was inactivated in an irradiation dose-dependent manner [52]. Kastury and Platz [58] showed that increasing the concentration of a PS at a fixed light dose leads to increased viral inactivation as does an increase in the total light exposure at a fixed PS concentration. The inactivation rate of T1 bacteriophage increased with increasing fluence rate, indicating that the distance of the sample from the light source is a variable which must be controlled [73]. At higher PS concentrations, the inactivation rate reaches a maximum and then decreases, because the filtering effect of the dye decreases the effective fluence rate [73]. In a simple model purposed by Lee et al. [56], the phage survival ratio can also be considered as a decreasing exponential fraction of the light fluence (assuming that the fluence is uniform throughout the system).