**2. Results**

### *2.1. Molecular Response to High Light Conditions*

We evaluated the molecular response of the diatom *Skeletonema marinoi* under a moderate light condition, used as a non-stressful control (low sinusoidal light, having midday light intensity peak at 150 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup>2, Sin150), as well as under high light stressful conditions, including high sinusoidal light peaking at 600 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup><sup>2</sup> (Sin600), high square-wave light peaking at 300 (Square300) and 600 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup><sup>2</sup> (Square600). These light conditions vary in both light intensity distribution over time (sinusoidal *vs*. square-wave), in the midday light intensity peak (150 *vs.* 300 *vs.* 600 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup>2) and in the daily light dose (3.6 mol photons m<sup>−</sup><sup>2</sup> d−<sup>1</sup> for Sin150, 14.4 mol photons m<sup>−</sup><sup>2</sup> d−<sup>1</sup> for Sin600 and Square300, 28.8 mol photons m<sup>−</sup><sup>2</sup> d−<sup>1</sup> for Square600).

Under the different experimental conditions, we assessed the expression of *ovoA* and *nos* genes to highlight their eventual light modulation at different time points. Samples were taken at 0, 6 and 24 h under all conditions, while additional samples were included in the high stressful light conditions. An earlier time point (2 h) was taken under Sin600 and two additional samplings (0.2 and 2 h) were carried out under Square300 and Square600 conditions, to evaluate the early response under high light stress. In *S. marinoi* two nos transcripts have been previously identified but no data are available on their functional significance [44]. Interestingly, our *in silico* analysis showed that both *S. marinoi* Nos protein sequences lack the inhibitory loop (Supplementary Figure S1), which is a feature of inducible Nos, making it Ca2+-independent [46]. To detect the redox status of the cells, in all experimental conditions we also measured the levels of reactive oxygen species (ROS) and nitric oxide (NO) through biochemical assays.

Under the control light condition (Sin150), no modulation of *ovoA* expression was observed at any experimental time both in exponential and in stationary growth phases (Figure 1A). The same was observed for *nos1* and *nos2*, except for a *nos1* down-regulation at the midday light intensity peak in exponential growth phase (6 h), with a decreasing trend also at the midday peak in the stationary phase (30 h, Figure 1A). Both NO and ROS levels were higher in stationary compared to the exponential growth phase, while they did not vary during the day in the exponential growth phase, confirming the non-stressful nature of this control condition (Figure 1A). Cultures exposed to Sin600 showed an upregulation of *ovoA* gene expression after 2 and 6 h from the light switch (Figure 1B). Additionally, *nos2* upregulation was observed after 6 h, while no variation was observed for *nos1* (Figure 1B). Both NO and ROS levels increased at 6 h and 24 h from light switch to Sin600 (Figure 1B).

To investigate whether the *ovoA* gene upregulation observed at 6 h of Sin600, compared to Sin150, resulted also in increased production of the molecule, we measured the concentration of ovothiol both under Sin600 and Sin150, at the midday light intensity peaks 600 and 150 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup>2, respectively. We also evaluated the content of the other major intracellular thiol glutathione. In agreemen<sup>t</sup> with the upregulation of *ovoA*, the cellular content of ovothiol B increased by two-fold from 50 μM under Sin150 to 110 μM under Sin600 (Figure 1A; Table 1). Similarly, glutathione also doubled its concentration from 1 mM under control condition (Sin150) to 2.3 mM under high sinusoidal light (Sin600; Table 1).

Cells moved to Square300 did not show any significant modulation in the expression of *ovoA* nor of *nos*, neither at a very early time point (0.2 h). The concentrations of NO and ROS did not increase, with the exception of a significant ROS overproduction at 24 h, following the night phase (Figure 2A). Similarly, no gene modulation was observed in the Square600 condition, at any experimental time, with ROS significantly increasing at 24 h (Figure 2B).

**Figure 1.** *S. marinoi* under control and high sinusoidal light conditions. Scatter plot representing light condition, gene expression data, NO and ROS levels were reported. (**A**) Control condition with midday peak at 150 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup><sup>2</sup> (low sinusoidal light, Sin150); (**B**) High sinusoidal light with midday peak at 600 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup><sup>2</sup> (Sin600). The light:dark photoperiod cycle was 12 h:12 h. Sampling times were highlighted in the light scatter plot by red circles. Fold gene expression data were analyzed by the pairwise fixed reallocation randomization test by REST and are here reported as 2-log scale mean ± standard error and. NO and ROS data were analyzed by Kruskal–Wallis with a Dunn's post hoc test and are reported as mean ± standard deviation. \* *p* < 0.05, \*\* *p* < 0.01 and \*\*\* *p* < 0.001 represent significance compared to 0 time.

**Table 1.** Thiols determination in *S. marinoi.* The concentrations of ovothiol B and glutathione in *S. marinoi* under control (Sin150) and high sinusoidal (Sin600) light conditions are reported as mean ± standard deviation.


**Figure 2.** *S. marinoi* under high square-wave light conditions. Scatter plot representing light condition, gene expression data, NO and ROS levels were reported. (**A**) High square-wave light with midday peak at 300 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup><sup>2</sup> (Square300); (**B**) High square-wave light with midday peak at 600 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup><sup>2</sup> (Square600). The light:dark photoperiod cycle was 12 h:12 h. Sampling times were highlighted in the light scatter plot by red circles. Fold gene expression data were analyzed by the pairwise fixed reallocation randomization test by REST and are here reported as 2-log scale mean ± standard error and. NO and ROS data were analyzed by Kruskal–Wallis with a Dunn's post hoc test and are reported as mean ± standard deviation. \* *p* < 0.05 and \*\* *p* < 0.01 represent significance compared to 0 time.

Interestingly, cells under both square-wave lights (Square300 and Square600) displayed a very high oxidative stress at 24 h, following the night phase (ROS content about 45 fmol DCF/cell; Figure 2A,B), compared to cells under Sin600 (ROS content about 5 fmol DCF/cell; Figure 1B). Indeed, this latter condition, although being characterized by a high intensity of light, mimicks a natural climate, progressively increasing the light intensity until the midday peak. Conversely, under the square-wave distribution, cells are suddenly exposed to the maximal light intensity, not allowing the acclimation process, which occurs in natural conditions. The detrimental effect of the square-wave light conditions

was also visible by the drop in cell growth following the stress received (24 h), differently to what was observed in cells exposed to Sin600 (Figure S2 and Table S1).

### *2.2. Molecular Response to Prolonged Darkness and Low Light Conditions*

In a next step, we assessed the molecular response to prolonged darkness (0 h:24 h light:dark photoperiod cycle) and very low light conditions (midday light intensity peak at 10 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup>2), i.e., Sin10 and Square10, varying in both light:dark photoperiod cycles (12 h:12 h and 24 h:0 h, respectively) and in light distribution over time (sinusoidal *vs.* square-wave, respectively). In these conditions, samplings were carried out at dawn and midday peak of two consecutive days (0, 6, 24 and 30 h) to follow two consecutive light:dark photoperiod cycles.

Cells exposed to a continuous dark condition showed a significant *nos2* upregulation after 24 h without any *ovoA* modulation (Figure 3). NO levels increased at 24 h, while the concentration of ROS was upregulated at 30 h (Figure 3).

**Figure 3.** *S. marinoi* under dark condition. Scatter plot representing the dark condition, gene expression data, NO and ROS levels were reported. Dark was kept constant for all the experiment (light:dark photoperiod cycle 0 h:24 h). Sampling times were highlighted in the scatter plot by red circles. Fold gene expression data were analyzed by the pairwise fixed reallocation randomization test by REST and are here reported as 2-log scale mean ± standard error and. NO and ROS data were analyzed by Kruskal–Wallis with a Dunn's post hoc test and are reported as mean ± standard deviation. \*\* *p* < 0.01 and \*\*\* *p* < 0.001 represent significance compared to 0 time.

Cells shifted to Sin10 did not modulate *ovoA* nor *nos1* gene expression, while *nos2* was upregulated after 24 h from the light switch (Figure 4A). NO and ROS variations under Sin10 condition resembled the pattern obtained under prolonged darkness. Indeed, levels increased after 24 h and 30 h, for NO and ROS respectively (Figure 4A).

Under continuous Square10 condition, *nos1* was downregulated after 24 h from the light switch, while all the target genes were significantly downregulated at 30 h. In parallel, NO and ROS increased levels were observed at the same times, with ROS reaching very high concentrations (about 100–150 fmol DCF/cell, Figure 4B).

Among these three conditions, Square10 resulted to be the most stressful for the cells, as also highlighted by the drop in cell growth observed at 24 h (Supplementary Table S1).

Pairwise Pearson correlation analyses were performed between the different variables in three clusters of light conditions: low light (Sin150, Sin10 and Square10), high light (Sin600, Square300 and Square600) and dark. In the high light cluster, *ovoA* was correlated to *nos2*, and ROS levels were positively correlated to NO and *nos1* (Table 2). Similarly, under low light NO was correlated with ROS while *ovoA* with both *nos1* and *nos2*, and *nos1* with *nos2* (Table 2). In the dark condition, only NO-*ovoA* and ROS-*nos1* pairs were correlated (Table 2). These results displayed the different pattern and potential interactions between these variables shaped by the light quantity harvested by cells.

**Figure 4.** *S. marinoi* under very low sinusoidal and square-wave light conditions. Scatter plot representing light condition, gene expression data, NO and ROS levels were reported. (**A**) Very low sinusoidal light with a midday peak at 10 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup><sup>2</sup> (Sin10). The light:dark photoperiod cycle was 12 h:12 h; (**B**) Very low square-wave light was kept constant for all the experiment at 10 μmol photons s<sup>−</sup><sup>1</sup> m<sup>−</sup><sup>2</sup> (Square 10). The light:dark photoperiod cycle was 24 h:0 h. Sampling times were highlighted in the light scatter plot by red circles. Fold gene expression data were analyzed by the pairwise fixed reallocation randomization test by REST and are here reported as 2-log scale mean ± standard error and. NO and ROS data were analyzed by Kruskal–Wallis with a Dunn's post hoc test and are reported as mean ± standard deviation. \* *p* < 0.05, \*\* *p* < 0.01 and \*\*\* *p* < 0.001 represent significance compared to time 0.


**Table 2.** Pairwise Pearson correlation analyses conducted on all variables (fold gene expression ratios, NO and ROS levels). Different colors refer to different light conditions (red = high light (HL), blue = low light (LL) and green = dark (D)). HL, LL and D indicate the light intensity at which the indicated pair is positivelycorrelated.\* *p* < 0.05,\*\* *p* < 0.01and\*\*\* *p* < 0.001representsignificanceofpositivecorrelation.
