**3. Discussion**

Ovothiols are natural sulfur compounds mainly occurring in marine organisms and in the microbial world, in three differentially methylated forms: A, B and C [25]. The biological role of these molecules has often been linked to their peculiar antioxidant properties [26,27] mediating defense from oxidative stress during fertilization and development in the sea urchin [47,48], from environmental stressors in fish eye lenses [49], from macrophage-triggered stress in pathogenic parasites during infection [50] and from light-dependent stress in microalgae [38,39]. In addition, ovothiols have been recently suggested to protect the sea anemone from the stress induced by UV radiation [51] and mussels from environmental pollutants during spawning [52]. However, biological functions not strictly related to oxidative stress have also been discovered. For example, ovothiols have been reported to induce egg release in marine polychaetes during sexual reproduction [53]. Interestingly, ovothiols might play important roles also in animals that, although lacking the biosynthetic pathway, might acquire this bioactive compound from external sources [49,54,55].

We recently described the occurrence and distribution of OvoA in diatoms and we identified ovothiol B as the ovothiol form produced by the coastal centric diatom *Skeletonema marinoi* at micromolar concentrations, when grown under moderate light condition [23]. Our hypothesis is that ovothiol biosynthesis could have been evolved and conserved in diatoms to help them to defend from the oxidative stress enhanced by high light, thus contributing to the ecological success of these photosynthetic protists. Following this assumption, we investigated the light-dependent regulation of ovothiol B biosynthesis in *S. marinoi*, mainly examining the expression of the gene *ovoA* encoding the key ovothiol biosynthetic enzyme.

The experimental strategy adopted in this study involved the cultivation of *S. marinoi* cells under different stressful light conditions, following a preacclimation period under moderate light. For each condition, samples were collected at different time points and were analyzed for gene expression of *ovoA* as well as for intracellular concentration of reactive oxygen species (ROS) and nitric oxide (NO), considered key mediators of stress response [11,40–43,56]. To integrate data on NO production and to investigate a possible involvement of the arginine-dependent NO biosynthesis in response to stressful light conditions, we also followed the gene expression of NO synthases in *S. marinoi*, which we named *nos1* and *nos2*. Among all the conditions tested, the high sinusoidal light (Sin600) resulted to be the most efficient in inducing an increased expression of *ovoA*. Indeed, *ovoA* is upregulated after only 2 and 6 h from the light switch with a significant increase in NO and ROS content at 6 h, as well as ovothiol production, which doubles its concentration compared to the control condition. Under this light, also *nos2* is upregulated after 6 h from a light switch together with increased NO levels already at 6 h and maintained high until 24 h, while *nos1* is not modulated at any experimental time. This may indicate a different role for these two nos transcripts. The *in silico* analysis of the Nos1/2 protein primary structure points out an inducible nature of the two Nos isoforms, but our data sugges<sup>t</sup> that only Nos2 responds

to high light, with a concomitant increase of NO. This issue is quite interesting and would deserve further investigation to understand which factors may be involved in Nos1 induction.

The finding that *ovoA* is not modulated in cells under square-wave light conditions suggests that the exposure to fast increases in photon flux density (PFD) does not allow the cells to efficiently modulate their metabolism, as also evident from the very high oxidative stress observed in cells at 24 h from the light switch to Square300 and Square600. Indeed, these unnatural conditions likely compromise the physiological defense mechanisms, finally leading to cell death, as indicated by the observed drop in cell growth. Even at a very early time (0.2 h) the cells do not exhibit any response in terms of *ovoA* modulation and other measured variables. Interestingly, our data are in line with a recent study evaluating the overall response of *S. marinoi* to different light regimes in terms of activated photoprotective and antioxidant systems, including ascorbic acid, one of the most abundant intracellular antioxidants, carotenoids and phenolic compounds, which significantly increase under the Sin600 condition [19]. Thus, the finding that both ovothiol and glutathione double their content in Sin600 condition may indicate that this light regime is able to modulate the diatom metabolism such that all the triggered photoprotective and antioxidant systems may synergically contribute to the defense from the high light stress. Indeed, the increased ovothiol production under the Sin600 condition suggests that this condition not only induces OvoA activity but also all the enzymatic toolkit necessary for ovothiol B biosynthesis, including the lyase OvoB and the still uncharacterized methyltransferase [36,37,55]. Moreover, the uncommon antioxidant properties of ovothiols compared to other thiols, including glutathione [26–28], can partially explain the observed difference in concentration of these two thiols in *S. marinoi*, suggesting that very low levels of ovothiol could be enough to exploit its antioxidant function.

The lack of *ovoA* modulation in Square300 highlights that the daily light dose is not a trigger for the increase of *ovoA* expression, since this condition is characterized by the same daily light dose as Sin600 (around 14.4 mol photons m<sup>−</sup><sup>2</sup> for both Sin600 and Square300), which is conversely able to stimulate ovothiol biosynthesis. Moreover, ovothiol biosynthesis does not react to fast increases of PFD, provided with Square600 condition, but instead to a gradual PFD increase under Sin600 condition.

These results highlight the phenotypic plasticity of *S. marinoi* when submitted to different light conditions, with ovothiol involvement into the photoresponse when cells cope with natural high light condition (sinusoidal), while it does not take part to the response of cells to unnatural high light condition. This is a quite interesting issue being that phenotypic plasticity, defined as the ability of an organism's genotype to display different phenotypes in response to the environmental variability, is a key process explaining microalgal adaptability to natural conditions [57–59]. In particular, microalgae are adapted to a natural light sinusoidal distribution during the day, allowing them to efficiently cope with light variations, e.g., high light environment, through the activation of different kinds of complementary photoresponses, from the fast non-photochemical quenching or xanthophyll cycle operation to changes in carotenoids levels or in antioxidant network activation. When cells are submitted to unnatural light variations (e.g., square-wave distribution or very fast light increase), they undergo physiological/biochemical stress responses, which are different in amplitude and time succession sequence compared to responses to natural light conditions [16–19,60].

The results of our study let us hypothesize that ovothiol synthesis/activation is among the players into the natural physiological response of microalgae to high light, but seems not to be a way for cells to cope with highly stressful (unnatural) light conditions. This might be due to the high-energy cost for cells to synthesize ovothiol, or, more likely, to the lack of biological intracellular signal for the activation of ovothiol biosynthesis.

Yet, the lack of *ovoA* modulation during the stationary growth phase reinforces the finding that the transcriptional level of *ovoA* responds to light and is not modulated by other physiological stressful processes, such as senescence.

The results on *ovoA* modulation under darkness and very low light conditions confirm the key role of light in enhancing ovothiol biosynthesis. Indeed, while there is basically no *ovoA* modulation under prolonged darkness and Sin10 condition (12 h:12 h light:dark), cells grown under continuous Square10 (24 h:0 h light:dark) display a downregulation of the genes *ovoA* and *nos1*/*2* over time, reaching a high significance at 30 h from the light switch, together with an increase of NO and ROS levels, and a drastic decrease in cell growth rate. This may be due to the extremely unnatural character of this condition. By contrast, diatoms can tolerate prolonged darkness and very low sinusoidal light, since they naturally experience these conditions when they sink at the limit or below the photic zone. These results may indicate that the absence of *ovoA* upregulation under darkness and low light conditions leads to very low levels of ovothiol, which are presumably not enough to counteract the enhanced oxidative stress. Interestingly also other antioxidants and protective systems do not increase in dark and low light conditions, thus underlining a crucial role of light to induce the production of photoprotective and antioxidant systems, excluding an internal circadian clock [19].

Overall, the conditions mostly a ffecting the variables here measured are Sin600 and Square10. Indeed, while a too high intensity light could damage the photosynthetic apparatus of diatoms, also a prolonged light, even though at very low light intensity, could cause a chronic accumulation of ROS and lead to cell death, as observed. However, the upregulation of antioxidant systems, including ovothiol biosynthesis, by high sinusoidal light allows the cells to better tolerate the stress, while cells experiencing a low light (or dark) stress are not able to e fficiently respond due to the absence of light modulation of such defense systems.

The present work might provide the basis for a possible eco-sustainable production of ovothiols by diatoms, to be used as "biofactories" for biotechnological purposes through the light modulation of their growth, metabolism and physiology [3,7,19,61–63]. Indeed, despite the increasing interest in the pharmaceutical potential of this class of molecules endowed with antiproliferative, anti-inflammatory and antifibrotic properties [30–35], studies regarding their bioactivities are still limited by the small amounts of ovothiol available for applied research. Currently, they can be extracted by sea urchin eggs, thus obtaining 2.5 mg of pure ovothiol A from 10 g of wet material [25,32]. However, sea urchins cannot provide su fficient amounts of ovothiols for extensive studies, also considering the necessity to preserve natural populations at sea. On the other hand, the light-induced upregulation of ovothiol formation by 2-fold might still be not enough to allow an e fficient exploitation of such system for biotechnological purposes. Therefore, further studies will be necessary to find alternative and more efficient growth conditions, e.g., di fferent nutrient availabilities, that might be able to increase ovothiol levels more e fficiently, or genetic engineering of diatoms to optimize the ovothiol biosynthesis by gene manipulation.
