**3. Discussion**

In accordance with our previous findings (Martins et al. [20]), the extracts from the "greenish" skin and the "yellowish" proboscis of *Eulalia viridis* contain multiple pigments and sustain very significant inter-organ variation. However, the extraction methods were distinct between the two works, albeit HPLC-DAD being used for fractioning in either case. Whereas we previously used HCl:acetonitrile to extract the pigments, in the present work we opted for extraction in PBS to obtain a physiologically compatible vehicle for the bioassays. The most notorious difference is the higher representativity of individual yellowish pigments in the skin found in the current work relatively to Martins et al. [20]. Nonetheless, the main pigments, in particular the yellow compound retrieved at retention time 4.40 min (Figures 3D and 5D) is seemingly present in the extracts from our preceding work. Moreover, this pigment shows the expected porphyrinoid signature. The methodological differences between the two works are likely responsible for disparities and render direct comparisons difficult, taking into account extraction efficiencies. However, it is clear from the current results that many porphyrinoid pigments in *E. viridis* are hydrophilic and compatible with PBS, which has important implications for biotechnological endeavors. It must be noted, though, that our previous paper already suggested the amphiphilic nature of porphyrinoids from *Eulalia*, similarly to what has already been described for other porphyrin-derived pigments [22]. In addition, the advantages of amphiphilic substances in therapeutics, as they can be transported through the blood stream and have facilitated crossing through the phospholipid by-layer, leads to substantial efforts to produce these compounds synthetically for the purpose of PDT (see for instance Malatesti et al. [22]). The discovery of natural bioproducts with these properties, as in the present case, is, therefore, an important biotechnological asset.

The two very distinct toxicity-testing procedures yielded divergent, albeit complementary results, in any case the presence or absence of light being a significant factor. The duration of the assays (i.e., total time of exposure) is inferred to be a key to explain the differences. Indeed, the shorter-term (1 h) assays with mussel gills revealed a trend for higher DNA-damage effects under light, whereas the *Daphnia* immobilization assay yielded the opposite tendency after 24 and 48 h of exposure following the initial 1 h treatment under light or dark conditions. Consequently, despite the differences between model and endpoint, time of exposure is seemingly a major player in the modulation of toxicity because the generation of singlet oxygen by photodynamic substances occurs swiftly after irradiation with visible light [23]. Evidently, this depends on oxygen supply, which is certainly more reduced in the test medium of *Daphnia*, especially after 48 h, even assuming minimal degradation of the pigments in the test medium or by the cladocerans' own mechanisms of porphyrinoid catabolism and elimination. Consequently, the photodynamic e ffect of porphyrinoids was more evident in DNA damage determined by the standard alkaline Comet assay, which is directly sensitive to singlet oxygen radicals in the short term, with particular respect to the formation of oxidized purines [24,25]. On the other hand, after 24 and 48 h of exposure, low oxygen and photosensitizer exhaustion seemingly reduced the photodynamic e ffects of pigments, leaving dark toxicity of phorphyrinoids as the most evident e ffect. This e ffect has already been described for a range of porphyrin-based photosensitizers in vitro [25,26]. Even though dark toxicity for these compounds in vivo is not well understood, the results sugges<sup>t</sup> significant e ffects in *Daphnia*. However, details on the toxicity, metabolism, and elimination of heme and its secondary metabolites are scarce (refer, for instance, to Lamkemeyer et al. [27] and references therein on *Daphnia* hemoglobin). Still, Fabris et al. [28] found *Daphnia magna* particularly sensitive to both light and dark toxicity of certain porphyrinoids after short term (1 h) and 24–48 h of incubation. It is noteworthy that the same authors employed the standard *Daphnia* immobilization assay as well. These results are thus in good alignment with the current study and show that the toxicodynamics and toxicokinetics of photosensitizers is complex as it depends on the ratio between light:dark toxicity and how it changes with time. Even though the sensitivity of our models and endpoints cannot be directly compared, the findings also demonstrate the importance of testing both immediate and longer-term effects of exposure to isolate light and dark toxicity of porphyrinoids. This subject is of particular relevance for further studies to investigate the potential of pigments in PDT, as dark toxicity can cause significant e ffects during the period of elimination, after the e ffective photodynamic capability of pigments has been exhausted.

There are also very noticeable di fferences between the e ffects caused by the two extracts that ultimately enable shortlisting the pigments with the most promising properties as photosensitizers. The findings show that the extracts from the proboscis, where the yellow pigment with 4.40 min retention time is by far more abundant than in the skin (which makes it responsible for the di fference in color between the two extracts), hold higher phototoxicity. This can be ascertained from the comparison between Figures 6 and 8. Most likely, the majority of the phototoxicity causing the increment of DNA damage relatively to the dark treatment shown in Figure 6 is due to this same yellow pigment, despite inconclusive dose-response e ffects. On the other hand, the mixture of pigments in the skin extracts, which includes the same compounds albeit in much lower proportion, caused higher dark toxicity. Altogether, the results indicate that this yellow pigment is an interesting candidate for further research as an e ffective photosensitizer. It must be noted, though, that the lack of a clear dose-response is in part due to significant inter-replica variation and reduced number of concentrations tested. Nonetheless, we must emphasize that at this stage we were testing crude extracts and not the purified pigments, which also contributed to the overall variation. Further toxicity testing involving the isolation of this candidate pigment are needed, as well a complete chemical characterization. It must also be noticed that the extracts from the skin caused DNA damage and sugges<sup>t</sup> a potential dose-response and light-mediated toxicity as well. However, as pointed out in our previous work, this organ holds a much wider variety of pigments than the proboscis [20], which complicates, at this stage, isolating specific substances of interest and obtaining su fficient amounts for analyses.

It can be inferred that the light toxicity of porphyrins is certain to cause deleterious e ffects to the worm itself. This consequence has been noted before and has, inclusively, been suggested as one of the reasons why Polychaeta are so di fficult to rear in captivity [29], which explains the need to add shelter and provide dimmed light to *Eulalia* in the laboratory [18]. Considering that *Eulalia* is a diurnal forager of the rocky intertidal, its pigments likely have two important roles. First, they are either toxic or repellent to predators and parasites that cause rupture to the pigment cells where the pigments are "safely" stored in granules [19,20]. Second, they warn of excessive exposure to daylight by causing physiological stress, which can be considered a form of non-nervous sensorial arrangement. The fact that *Eulalia* uses its proboscis as the main organ for sensing and preying [19,21] explains the natural di fference between the pigmentation of the skin as a major adaptive trait. These particular features, which are probably not circumscribed to *Eulalia* or even to the Order Phyllodocida, show that the Polychaeta are very promising targets for the bioprospecting of photoactive tetrapyrrolic pigments.

In conclusion, the present work revealed di fferent porphyrin-like pigments in the body of the Polychaeta *Eulalia*, distributed di fferentially between the proboscis and the skin. The di fferential pigmentation pattern between the two organs is also reflected in distinct light:dark toxicity ratios that are influenced by time of exposure, as light-dependent toxicity occurs swiftly whereas light-independent will occur in a longer timeframe until the substances are metabolized and finally eliminated from the recipient organism. With this respect, the two bioassay procedures with mussel gills and *Daphnia* (1 and 48 h, respectively), showed the importance of performing short and longer tests to evaluate the unique toxicodynamics of photosensitizers. From the aggregate findings we were able to detect a specific pigment, yellowish in color, present in both organs albeit much more concentrated in the proboscis of the worm. This pigment, in particular, is a serious candidate as photosensitizer: (i) it is likely amphiphilic and compliant with PBS, which is an advantage for distribution during therapeutics; (ii) its light toxicity is considerably higher than its dark toxicity, and (iii) its reduced toxicity with time in vivo may be an indicator of rapid exhaustion and facilitated elimination. These highly promising results also confirm that marine invertebrates are a prolific source of natural porphyrinoids, a class of substances with high value for therapeutics and whose research is still mostly based on synthetics.

### **4. Materials and Methods**
