Blue light photoinhibition of Streptococcus mutans: potential chromophores and mechanisms

Endogeneous porphyrins have been found in a range of anaerobic and facultative anaerobic bacterial species [30‐34]. Porphyrins share common basic structures that include four pyrrole rings linked together by a methine bridge at their alpha carbon bonds. Due to their common structure, they exhibit blue light absorbance and fluorescence in the red spectrum (λ = 600–700 nm). Similar to PDT, different classes of ROS are generated after light absorption. In the type I reaction, hydroxyl and superoxide anion radicals are produced. In type II reactions, the energy released from the photosensitiser is taken up by a triplet oxygen, transforming it into a much more excited singlet oxygen which is highly reactive and understood to have the major effect in the photodisinfection process [35‐41].

ALA is not a known photosensitiser [42] but is a heme/cytochrome intrinsic precursor [43]. ALA stimulates oxidative phosphorylation by upregulating cytochrome c oxidase activity and inhibiting glycolysis in eukaryotic cells [44, 45]. ALA is also known to similarly increase cytochrome levels in Escherichia coli (E. coli) [46]. The supplementation of bacteria capable of synthesising heme with ALA enhances their porphyrin production; examples include Pseudomonas aeruginosa (P. aeruginosa), Staphlycoccus aureus (S. aureus), and E. coli. Increased levels of porphyrins have subsequently been shown to increase the bacteria’s susceptibility to light exposure in both the blue (λ = 407–420 nm) and the red (λ = 635 nm) portions of the visible spectrum [47‐50]. Interestingly, ALA supplementation has also been shown to be capable of enhancing the inhibitory effect of both blue (λ = 440 nm) and red (λ = 635 nm) light in S. mutans [51, 52] even though this bacterium is incapable of synthesising heme. Similarly, Enterococcus faecalis (E. faecalis) — another lactic acid-producing bacterium — did not show any porphyrin synthesis following ALA treatment [47], yet also exhibited photo-inhibitory effects following ALA supplementation and red light (λ = 633 nm) irradiation [53]. It has also been shown that depending on the bacterial species, certain concentrations of ALA can induce inhibition without light irradiation. For example, 40 mM ALA decreased the viability of E. coli while 5–10 mM induced similar effects in P. aeruginosa [48, 50]. Unexpectedly, 30 mM ALA can also induce antimicrobial effects in S. mutans unaccompanied by light exposure [51, 52].

It should be noted that heme/hemin (ferric iron (Fe³⁺)-containing porphyrin) acquisition and utilisation by bacteria remains a subject that has not entirely been elucidated and there are still unanswered questions concerning how bacteria utilise exogenous heme. It is unclear as to whether exogenous or endogenously synthesised forms are preferentially taken up by the bacterial cell, and once absorbed do they exist in a free state or as a chaperone inside the bacterial cell. However, heme synthesis is not limiting to heme-protein/cytochrome production, meaning that heme synthesis can take place at much higher rates than heme-protein production. It has also been suggested that all Gram-positive bacteria use non-canonical pathways for protoheme synthesis compared with Gram-negative bacteria and eukaryotic cells. Gram negative bacteria and eukaryotes rely on protoporphyrin as a final intermediate, while Gram-positive bacteria use coproporphyrin [54‐56]. Intriguingly, heme inhibits the growth of species with confirmed endogenous porphyrins, while enhancing the growth and aero-tolerance of lactic acid-producing species which reportedly do not rely on heme for growth and where the presence of porphyrins has not yet been identified [57‐62]. Nitzan et al. reported that hemin affects the growth and viability of S. aureus rapidly. This was evidenced by cessation of both glucose uptake and production of carbon dioxide. The antibacterial effects were reversed in the presence of albumin — which has a binding capacity to hemin — confirming the binding of hemin to bacterial cells. The toxic effect of hemin was not attributed to singlet oxygen or hydroxyl radical formation. This was confirmed by adding scavengers/quenchers such as mannitol and histidine, which failed to prevent the inhibitory effect of hemin on bacterial growth. Furthermore, reduced — but not oxidised — cysteine and glutathione inhibited the antibacterial effect of hemin. These results supported the hypothesis that the toxic effects of hemin are mediated through oxidation of intracellular components [57, 58, 63]. Conversely, hemin showed opposite effects on lactic acid-generating bacteria, such as Lactococcus Lactis (L. lactis), resulting in enhanced growth and glucose utilisation. Hydrogen peroxide (H₂O₂) levels were also supressed after hemin addition, despite a lack of catalase activity [59]. Exogenous heme and hemin have also promoted cytochrome formation in L. lactis, as well as catalase and cytochrome bd activity in E. faecalis. Notably, heme was consumed only to synthesise cytochromes and not to alter the mode of respiration [64]. All these outcomes support the proposition that due to their unique potential to adapt to various environmental changes, the ancestors of all lactic acid-producing bacteria had the cytochrome gene — cyd — but multiple gene loss has occurred over time [61]. The cytochrome bd genes (cydABCD) within L. lactis improve cellular survival and oxidative stress resistance by inducing a proton motive force. Genes involved in heme biosynthesis have been identified in species not requiring heme for growth. E. faecalis contains the gene HemH, while L. Lactis has the HemZ gene. These genes encode ferrochelatase, the only function of which is to insert Fe²⁺ into protoporphyrin IX — to form heme [60, 65]. However, these genes have not been identified in other lactic acid-producing bacteria, including S. mutans [60‐62].

Since the photochemical mechanism of blue light is mediated through generating toxic levels of ROS, it is essential to understand how S. mutans survives different oxidative challenges. S. mutans is a Gram-positive, facultative anaerobic bacterium that grows optimally in a microaerophilic atmosphere; however, it can also survive under aerobic conditions. It grows rapidly at a temperature of 37 °C, ferments a range of sugars (i.e. lactose, sorbitol, mannitol, raffinose, mannose, salicin, and trehalose) to produce lactic acid. Compared to other streptococci, it is the most acidogenic and cariogenic species. It can synthesise extracellular polysaccharides (e.g. glucan–fructan) from sucrose, which enhance bacterial colonisation and growth on the tooth surface [66‐68]. Even though S. mutans does not perform oxidative phosphorylation, it utilises an incomplete tricarboxylic acid (TCA) cycle and depends mainly on glycolysis for adenosine triphosphate (ATP) generation [69‐72].

Superoxide dismutase (SOD), peroxidase, and catalase are the main enzymes that bacteria use to manage oxidative stress [73]. SOD converts superoxide anions to less reactive H₂O₂ and molecular oxygen, while catalase is the enzyme responsible for degrading H₂O₂ to water and oxygen [60]. Since S. mutans lacks heme-bound catalase and cannot synthesise cytochromes (heme-containing redox proteins) for full aerobic respiration and energy metabolism, it depends mainly on a set of proteins/enzymes that are either metal- or FAD-bound. Regarding metal-bound enzymes, S. mutans has SOD that requires either iron or manganese cofactors [74, 75]. S. mutans also has an ability to resist oxidative stress by means of a DNA-binding protein from starved cells (Dps)-like peroxide resistance protein (Dpr), albeit, not through a direct interaction with ROS. Having ferritin-like properties, Dpr sequestrates free iron, which prevents further hydroxyl radical formation through the Fenton reaction [73, 76].

Flavins, which are essential for growth and energy metabolism, have also been reported to mediate several redox functions in the process of oxygen metabolism [77]. S. mutans adapts to aerobic environments using FAD-bound enzymes such as reduced nicotinamide adenine dinucleotide (NADH) oxidases (Nox), alkyl hydroperoxide reductases (Ahp), and thioredoxin reductase (TRX). Nox-1 (AhpF homologue) and Nox-2 enable lactate dehydrogenase to oxidise NADH by reducing molecular oxygen. However, Nox-1 generates H₂O₂, while Nox-2 generates water and plays a greater role in the regeneration of NAD⁺ for the steady operation of glycolysis. H₂O₂ generated by both SOD and Nox-1 is eliminated by AhpC (non-FAD). TRX maintains proteins in their reduced state by eliminating reversable disulphide bridges formed following oxidative stress (see Fig. 1). Notably, S. mutans can also reduce oxidised proteins using a non-FAD glutaredoxin system [29, 72, 78, 79]. To combat oxidative stress, it is understandable that there is a rise in the levels of the antioxidant enzymes once blue light is absorbed in its designated chromophore(s). Interestingly however, if an entire group of these enzymes is bound to a blue light absorbing molecule (FAD) [80], this means that blue light might be ironically inhibiting S. mutans through its own defence system.

Even though there are many studies in the literature confirming the inhibitory effects of blue light against S. mutans, no studies have investigated the mechanism underpinning these antimicrobial effects. Earlier reports in the literature stated that streptococci have an endogenous chromophore with two absorbance peaks (λ = 455 and 504 nm) that have variable intensities at different pH, while having an isosbestic absorbance at λ = 400 nm [81]. Recently, flavins have been reported as blue light chromophores — besides porphyrins — in both S. aureus and E. coli [34]. In this section, we discuss the evidence supporting the hypothesis that porphyrins and flavins are the potential chromophores in S. mutans.

There is evidence supporting the hypothesis that blue light might be absorbed by porphyrins in S. mutans even though this species does not require heme for growth. First, S. mutans has a hemin binding capacity, which is mediated through membrane proteins and not siderophores. Notably, the iron ion bound to hemin is Fe³⁺ (oxidised), unlike that bound to heme which is Fe²⁺ (reduced). In all cases, both forms of iron have been identified in S. mutans. It binds Fe³⁺ prior to transporting it intracellularly in its Fe²⁺ form; however, it is unclear what the fate of the iron-free porphyrin is after hemin binding and Fe³⁺ acquisition, and why iron is only transported in the Fe²⁺ form. When grown aerobically, the Dpr protein binds both forms of iron, limiting the free iron pool. The iron-active SOD is also upregulated when S. mutans are grown aerobically (see Fig. 1). Interestingly however, Fe³⁺ uptake and intracellular Fe²⁺ iron concentrations are both higher under anaerobic conditions [72, 73, 75, 76, 82, 83]. In E. coli, porphyrins/heme synthesis enzymes are slightly upregulated when bacteria are grown anaerobically [84]. This means that S. mutans utilises Fe²⁺ for other functions when growing anaerobically. S. mutans cultures showed red fluorescence (λ = 630–670 nm) — following λ = 405 nm blue light excitation — when grown anaerobically, and the fluorescence peaks were matching those of specific porphyrin classes, such as protoporphyrin IX [85, 86]. Secondly, it has been reported that knocking out the genes involved in the porphyrin enzymatic cascade in species capable of synthesising heme (E. coli) made the bacteria more sensitive to blue light compared to their wild-type counterparts. The phenomenon was associated with the accumulation of certain classes of porphyrins through an incomplete cascade [87, 88]. Porphyrin accumulation might explain why bacterial species with incomplete or blocked heme synthesis pathways are still susceptible to blue light irradiation. Thirdly, although there are few studies investigating the effects of ALA and blue light combined on S. mutans [51], it is conceivable that ALA might potentiate the effects of blue light due to porphyrin induction. Further investigations to confirm porphyrin accumulation were not carried out in this study. However, there are other possible explanations as to why ALA synergistically potentiated the effects of blue light or even induced inhibition in the absence of light. This evidence is related to the similarity between prokaryotes and eukaryotes in some steps of the heme synthesis pathway [55, 56]. Cyclic diadenosine monophosphate (C-di-AMP) helps S. mutans grow into biofilms and combat oxidative stress through upregulating the glucan-producing enzyme, glucosyltransferases. Moreover, S. mutans relies on F-type adenosine triphosphatase (F-ATPase) to fight acid stress by pumping out hydrogen ions to maintain intracellular pH [29, 89, 90]. ALA inhibits that action of both ATPase and cAMP in eukaryotic cells [91, 92]; therefore, it is possible that ALA also inhibits both C-di-AMP and F-Atpase in S. mutans through a similar biochemical process. Fourthly, S. mutans possess an incomplete TCA cycle [70, 71], and bacteria having incomplete TCA cycle can synthesise and utilise ALA through alternative pathways [93]. However, concentrations and distributions of intermediate compounds in the pathway differ according to growth conditions and between species. In all cases, the pathway is directly or indirectly linked to both the TCA cycle and cytochrome synthesis. Lactic acid produced can also reverse the function of the TCA cycle from catabolic to anabolic pathways promoting cytochrome synthesis [94]. Glutamyl tRNA (ALA precursor) is an alternative pathway for heme synthesis depending on glutamine [95, 96], and S. mutans depends on glutamine for biofilm growth and acid tolerance [97, 98] (see Fig. 1). Interestingly, stimulating intrinsic cytochrome formation through supplementing cultures with exogeneous ALA is significantly more effective than exogenous heme supplementation [43]. ALA increases the expression of cytochromes irrespective of intracellular heme levels [54]. Although porphyrins maximally absorb blue light wavelengths and ALA is the precursor of porphyrin production [15, 35], some investigations rely on red light (λ = 633–635 nm) after incubating bacterial cells with ALA. Intriguingly, bacterial growth inhibition was successful with both E. faecalis and S. mutans [51‐53], which suggests that there might be other chromophores alongside porphyrins; cytochromes are possible candidates. E. faecalis cytochrome bd has absorption peaks at 561 and 626 nm [99]. H₂O₂ forming Nox is mainly found in species possessing cytochromes and heme-proteins [78]. Bacterial species that generated cytochromes after heme supplementation were also capable of producing flavins when grown in media devoid of exogeneous heme [64]. S. mutans contains an extracellular electron transport system, which likely contains both flavins and cytochromes for full operation — flavo-cytochromes or flavo-porphyrins [100‐102]. Accordingly, it is possible that S. mutans might possesses the potential of generating porphyrins, heme-proteins, and cytochromes. However, the underlying pathways are not elucidated yet.

With regard to FAD, not only does FAD absorb blue light (absorption maximum at λ = 450 nm) but also induces ROS production which is essential for the inhibitory effects of light to occur [80, 103]. Besides the fact that ROS can be generated from FAD on its own, there are several flavoenzymes in S. mutans that are also strong candidates for mediating blue light-induced toxicity. Regarding the Nox system, it could be hypothesised that Nox-1 acts as a blue light chromophore since it is triggered by increasing FAD concentrations and generates H₂O₂ [78]. However, the FAD in Nox-2 cannot be excluded since in the process of reducing diatomic oxygen to water, other ROS are generated such as hydroxyl radicals and superoxide anions [72]. On the other hand, and following a different mechanism, TRX also represents a potential candidate for blue light absorption in S. mutans. This enzyme has been deactivated by blue light (λ = 460 nm) in L. lactis. Deactivation effects declined when an oxygen scavenger (an oxidase/catalase combination) and potassium iodide (a quencher for photoexcited FAD) were used, confirming the role of both oxygen and excited FAD in the process. Intriguingly, the blue light deactivated TRX in L. Lactis shares similar crystal structures with its homologue AhpF in E. coli [104, 105]. Therefore, deactivating the Nox-1 (AhpF) in S. mutans — and hence blocking its ability to reduce molecular oxygen — is also a possibility. Consequently, there is also strong evidence supporting FAD-bound enzymes as blue light chromophores, but the process might be more complex compared to the photoinhibition mediated through porphyrins. It is difficult to determine whether the mechanism of blue light inhibition takes place due to ROS production from a photo-excited FAD molecule solely, Nox-1/Nox-2 excitation, TRX/ Nox-1 inhibition, or a combination of all of these mechanisms (see Fig. 2).

The variability identified in the inhibitory effects of blue light on different bacterial strains is due to the difference in the photo-absorber types/levels, the antioxidant potential, and the light irradiation parameters (light source, wavelength, irradiance, irradiation time, and energy density) [14, 106]. The chromophore levels in bacteria differ greatly in wild-type strains cultured in their natural environment [107], and the antioxidative potential of each strain fluctuates depending on culture conditions [108]. Following blue light excitation (λ = 405 nm), S. mutans exhibited red fluorescence when tryptone soy cultures were supplemented with either a spinach extract (λ = 674 nm) or a mixture of blood, hemin, and vitamin K (λ = 636 nm). Conversely however, there was no fluorescence signal in non-supplemented tryptone soy or when cultures were supplemented with blood only [85]. Non-supplemented casein peptone soybean cultures, as well as those supplemented with chlorophyll, showed fluorescence peaks (λ = 630–670 nm) attributed to hematin, hematoporphyrin, and protoporphyrin IX. However, peaks of much less intensity (only at λ = 630 nm) resulted when blood was added to the media, which likely took place due to limited porphyrin production via a negative feedback loop [86]. The Nox system also operates at the converging point of several pathways involved in oxidative stress responses and energy metabolism. It controls and is itself affected by these pathways to maintain S. mutans survival [109]. Nox-2-deficient S. mutans cannot utilise the lactate dehydrogenase pathway due to the inability of these bacteria to oxidise the NADH generated [78]. Additionally, different sugars affect Nox levels differently. Mannitol generates more NADH compared to glucose, and hence, more Nox is produced to balance the NADH/NAD⁺ levels [79]. Conversely, Nox is less active throughout biofilms and acidic/low pH environments [110]. Importantly, light absorption within a chromophore does not necessarily mean that bacterial killing will take place. The chromophore has to be situated within the cell at the correct biochemical orientation. Previously, blue light (λ = 450 nm) did not show any signs of inhibition of Streptococcus agalactiae growth although it contains granadaene — a blue light chromophore (peak absorption at λ = 438–485 nm). Furthermore, exogenous FAD only minimally enhanced the inhibitory effects of blue light. This also highlights the issue with PDT; it is not clear whether ROS is generated inside the cell or generated in the media post irradiation to cause bacterial killing [111, 112]. It is worth mentioning that S. mutans flavoenzymes are localised in the cytoplasm [78]; however, if it utilises flavo-cytochromes/flavo-porphyrins for its electron transport system at the membrane level [100‐102], then membrane depolarisation remains a possibility. Conversely, recent studies have confirmed that — sequentially — intracellular chromophores (porphyrins and FAD) are responsible for the production of ROS which in turn leads to a reduction in transmembrane potential of S. aureus [113, 114]. It is also important to consider other aspects such as the half-life and the diffusibility of the ROS generated. For example, singlet oxygen has a very short half-life in aqueous solutions (4 μs) and can only diffuse a distance of 100 nm. Accordingly, it is essential to produce cytotoxic levels of ROS relative to the amount of photons absorbed [115‐117].