Introduction

Diatoms are photosynthetic microorganisms found in many environments, and account for up to 40% of global primary productivity (Serôdio and Lavaud 2020). They are a major group within chromalveolate algae and their cell walls are predominantly composed of silica that, together with other nutrients, are important for photosynthetic efficiency (Broddrick et al. 2019). Moreover, they are a natural source of many compounds, including a variety of pigments, such as chlorophylls and carotenoids, which have specific functions in cellular metabolism and have attracted biotechnological interest in recent years (Sharma et al. 2021).

Diatoms have eight main types of carotenoids which can be classified into two groups: carotenes, e.g., β-carotene (β-car); and xanthophylls, represented by zeaxanthin (Zx), antheraxanthin (Ax), violaxanthin (Vx), neoxanthin (Nx), diadinoxanthin (Ddx), diatoxanthin (Dtx) and fucoxanthin (Fx) (Kuczynska et al. 2015; Bauer et al. 2019). The marine diatom Phaeodactylum tricornutum has been recognized as a model organism for physiological and molecular studies, mainly due to being one of the first microalgae to have its genome sequenced (Bowler et al. 2008; Nymark et al. 2013; Falciatore et al. 2020). This microalga is also an interesting candidate for carotenoid commercial production, particularly Fx, that has nutraceutical properties and can be used against a variety of inflammation and cancer-related diseases (Liu et al. 2016; Andrade et al. 2018; Lopes et al. 2020).

Fucoxanthin biosynthesis in P. tricornutum was described by Lohr and Wilhelm (2001) to occur from lycopene (a precursor of carotenoids) via hydroxylation and epoxidation reactions, and two alternative routes of xanthophylls production were described. In the first pathway, Vx acts as a precursor of Ddx and Fx, while in the second pathway, Dambek et al. (2012) suggest that Vx leads to Nx, Ddx and Fx synthesis. Regarding this second alternative pathway, Dautermann et al. (2020) proposed a new hypothesis of carotenoid biosynthesis in chromalveolate algae that shows Vx as a precursor of latoxanthin and Nx, while Nx leads to Ddx and dinoxanthin, being both a putative intermediate in the Fx biosynthesis. Despite the sequencing of the genomes of Thalassiosira pseudonana by Armbrust et al. (2004) and P. tricornutum by Bowler et al. (2008), some unknown enzymes need to be identified in order to understand even more this metabolic pathway (Bauer et al. 2019).

The regulation of this biosynthetic pathway is influenced by the photosynthetically active radiation (PAR), causing diatoms to alter their content of photosynthetic pigments like chlorophylls and carotenoid profile, as they respond to changes in light intensity and light energy doses (Falciatore et al. 2020). Diatoms have a light-harvesting complex called fucoxanthin-chlorophyll a/c-binding antenna protein (FCP) which contains chlorophyll-c (Chl-c) as accessory of chlorophyll-a (Chl-a) and Fx as the major carotenoid (Gelzinis et al. 2015). This complex is homologous to light-harvesting complex II (LHCII) of vascular plants, performing similar roles (Büchel et al. 2022). In fact, diatoms have evolved various mechanisms and strategies such as changes in pigmentation, in order to enhance photosynthetic responses and growth rate (Ragni and d’Alcalà 2007; Lepetit et al. 2017; Conceição et al. 2020). An important regulatory mechanism in photosynthesis is the cyclic electron transport around Photosystem II (PSII) which can dissipate the excessively absorbed light energy under high illumination (Wagner et al. 2016).

According to Lohr and Wilhelm (2001), two xanthophyll cycles are found in P. tricornutum: the violaxanthin cycle (Vx-Ax-Zx) and the diadinoxanthin cycle (Ddx-Dtx). In both cycles, three epoxidation reactions are catalyzed by zeaxanthin epoxidase enzymes (ZEP1, ZEP2 and ZEP3) under low irradiance/dark conditions, as well as three reverse de-epoxidation reactions are catalyzed by violaxanthin de-epoxidase enzymes (VDE, VDL1 and VDL2) at higher irradiance. However, the Ddx-Dtx cycle is considered more efficient in diatoms, since the de-epoxidation rate can reach fourfold higher than that of the violaxanthin cycle (Lohr and Wilhelm 2001). In addition, similarity research of the de-epoxidation was found in two clonal strain of the haptophyte Tisochrysis lutea with a tenfold higher rate of Ddx-Dtx cycle in comparison to violaxanthin cycle as reported by Pajot et al. (2023). One of the main functions of these carotenoids is to be involved in the photoprotection mechanism, which can quickly dissipate the excess energy through non-photochemical quenching (NPQ) that is linked to the de-epoxidation state of carotenoids involved in the xanthophyll cycle (Goss and Lepetit 2015). For instance, under high illumination, the synthesis of these pigments increases to protect the photosynthetic apparatus from a potential photo-oxidation damage (Lepetit et al. 2017; Broddrick et al. 2019). Due to the antioxidant activity of these xanthophylls, cells are shielded from the damaging effects of free radicals and reactive oxygen species (ROS). This protection not only preserves cellular integrity but also enhances the equilibrium and efficiency of the photosynthetic apparatus (Lepetit et al. 2010).

The conclusion of P. tricornutum genome sequencing led to the proposal of new strategies in order to provide a more in-depth understanding of some factors that regulate important cellular processes in diatoms (Bowler et al. 2008). It is important to emphasize that the regulation of gene transcription also depends on certain culture conditions, however, the mechanisms behind this physiological plasticity are not well understood (Depauw et al. 2012; Kuczynska et al. 2015; Falciatore et al. 2020). Despite the different mechanisms evolved by photosynthetic organisms, a general rule in photochemistry is well-known, the Bunsen-Roscoe reciprocity law claims that the extent of photochemical effects is directly proportional to the total energy dose by cumulative irradiance regardless of the type of illumination regime administered (Bunsen and Roscoe 1859). This law has not been described in the diatom P. tricornutum, which is important for understanding the mechanisms of photoprotection and photoacclimation to changing light environments in marine phytoplankton. Thus, the aim of this study was to analyze whether the photoprotection of P. tricornutum is linked to light energy doses or radiation intensity.

Material and methods

Culture conditions

The diatom Phaeodactylum tricornutum strain CCAP1055/1 was maintained in the Laboratory of Algae Cultivation (LCA) at the Federal University of Santa Catarina – Brazil.

Before starting experimental assay, we focused to determine what would be irradiances to be applied. Cultures were exposed to increasing levels of irradiance (0, 25, 50, 100, 150, 600 and 1200 µmol photons m−2 s−1) and these values were obtained by placing the diatom vessels in different distances from a LED-light source (Taschibra 70W TL Slim) at 21 ± 1 °C. In addition, this light source spectrum (Supplementary Fig. S1) was measured with a spectroradiometer (Avantes Starline, AvaSpec-2048). In each case, the irradiances were measured by a quantometer (LI-250A, LI-COR). Samples were incubated in triplicate during 1 h and oxygen concentrations were measured in the water before and just after the incubation procedure with an oximeter (YSI, Model ProODO). From the photosynthesis-irradiance (P-I) curve (Supplementary Fig. S2) of P. tricornutum, one non-saturating irradiance (< EK = 100 µmol photons m−2 s−1), one irradiance at onset of saturation (EK = 200 µmol photons m−2 s−1) and one saturated irradiance (> EK = 400 µmol photons m−2 s−1) were determined.

Prior to the beginning of the experimental period, cultures of all treatments were acclimated for 7 days and maintained in a concentration of 70 mg of algal biomass per L, by daily dilution every 24 h to minimize self-shading effects. The assay was carried out in 2 L borosilicate flasks in LCA-AM medium (Sales et al. 2019), under constant agitation by bubbling with atmospheric air with an addition of 0.5% of CO2 (v/v) and at 21 ± 1 °C. The experiment was conducted combining three irradiances with two photoperiods, resulting different light energy doses: irradiance × light: dark (L:D) cycle, totaling six treatments as summarized in the Table 1, in triplicate. Culture samples were collected at different time intervals (at the beggining, i.e. 0, 24, 48 and 72 h after starting the experiment) to evaluate growth, and for the measurements of chlorophyll-a fluorescence in PSII, pigment content and qPCR analysis in P. tricornutum.

Table 1 Treatments applied to Phaeodactylum tricornutum cultures based on the light energy doses (irradiance × L:D cycle)
Full size table

Growth evaluation and light absorption

Before the experiment we conducted measurements of turbidity (NTU) with a turbidimeter (HI 93703 Hanna, at 890 nm) and dry biomass by gravimetric method using glass fiber microfilters (APHA 2011) (mg L−1). These measurements were used to follow the growth rates of P. tricornutum. Then, we took samples with different algal concentration and associated the data of turbidity with algal biomass (regression analysis) in each sample according to Eq. 1 (R2 = 0,99):

$$Biomass \left(mg \ {L}^{-1}\right)={ABS}_{890}\times 2.20+0.01$$
(1)

Before the experimental procedures, these data were used to calculate the specific growth rate (µ, day−1) according to Eq. 2, as proposed by Ye et al. (2018):

$$\mu =\frac{\mathit{ln}({X}_{f} -{ X}_{i})}{\Delta t}$$
(2)

where Xf and Xi are biomass concentrations at the period “Δt” (days).

The light attenuation measurements were performed in the center of 2 L borosilicate flasks containing P. tricornutum cells in suspension at a horizontal depth of 6.8 cm. Irradiance data were collected inside the vessel for each treatment with the aid of a quantometer (LI-250A, LI-COR) and the light attenuation coefficient (k, m−1) was calculated according to Eq. 3, as proposed by Krause-Jensen and Sand-Jensen (1998):

$${I}_{z}= {I}_{0}{e}^{-kz}$$
(3)

where Iz = irradiance at depth (µmol photons m−2 s−1); I0 = irradiance emitted by the LED (µmol photons m−2 s−1) at the level of the medium surface and z = horizontal depth of 2 L borosilicate flasks (m).

After this step the light attenuation coefficient of the experiment beggining was normalized by the time intervals of 24, 48 and 72 h to determine how much light was attenuated in 24 h. The new value of light attenuation (AT) was used to calculate the light absorption (AB) according to Eq. 4:

$$AB \left(\%\right)=\left(1-AT\right)\times 100$$
(4)

Chlorophyll-a fluorescence measurements

The photosynthetic parameters were measured by pulse amplitude modulation fluorescence (WATER-PAM—Heinz Walz GmbH, 2000) according to Malapascua et al. (2014) (Table 2).

Table 2 Photosynthetic parameters calculated from fluorescence measurements. Fo, Fv and Fm = minimum, variable and maximum fluorescence in the dark-acclimated condition. F and Fm’ = steady-state and maximum fluorescence in the light-acclimated state. EPAR = irradiance levels according to Table 1. ABSt = absorptance. FII = fraction of total chlorophyll a in PSII
Full size table

During the acclimation, minimum (Fo) and maximum (Fm) fluorescence levels were measured after dark acclimation for 10 min to calculate the maximum photochemical quantum yield of PSII (Fv/Fm), while stable (F) and maximum (Fm') fluorescence were measured after light acclimation at the experimental irradiances (without any actinic light from the PAM) for 10 min to provide a saturating pulse and then obtain fluorescence values to calculate the effective quantum yield of PSII (YII). Fluorescence measurements were collected at 1-h intervals to determine the maximum yields and, after seven days, when the YII remained constant, the experiment began, as this indicates the acclimatization of P. tricornutum cells. To allow a proper estimation of ETR, it is necessary to take into account only the ratio of chlorophylls associated to each photosystem, and in this case, these ratios were determined previously by Johnsen and Sakshaug (2007) considering different algal species containing diverse pigmentary composition. In this way, the fraction of cellular chlorophyll-a in PSII (FII) of P. tricornutum acclimated to low light and high light were provided by these authors. The absorptance (ABSt) can be estimated from the transmittance values (T) as proposed by Enríquez and Borowitzka (2010) according to Eq. 5, based on the attenuation coefficient calculations.

$${ABS}_{t}=1- T$$
(5)

Determination of chlorophyll concentrations and pigment analysis

For each sample, 5 mL was filtered through a 0.45 μM glass fiber filter and stored at -20 °C for further analysis. The photosynthetic pigments Chl-a and Chl-c were extracted as in Strickland and Parsons (1972) and quantified according to Jeffrey and Humphrey (1975) using a UV–Vis spectrophotometer (Genesys 10 vis, Thermo). The content of both pigments is reported as mg g−1 biomass dry weight (DW).

Ddx, Dtx, and Fx were quantified using a high-performance liquid chromatography (HPLC) of P. tricornutum extracts. An aliquot of each sample (10 μL, n = 3) was injected into a liquid chromatograph (Shimadzu LC-10 A) and the metabolite quantification was performed according to Schmitz et al. (2022). The readings were taken in triplicate and the results were reported in mg g−1 DW. The de-epoxidation state (DES, %) was calculated as described in Ruban et al. (2004) according to Eq. 6:

$$DES= \frac{Dtx}{Dtx+Ddx} \times 100$$
(6)

where, Dtx = diatoxanthin content (mg g−1 DW) and Ddx = diadinoxanthin content (mg g−1 DW).

RNA purification and cDNA synthesis

Culture samples were taken from each experimental unit according to Lopes et al. (2019). The culture was centrifuged (2950 × g, 5 min), washed 1 × with sterile marine water, and the pellets were immediately frozen and stored at -80 °C. Total RNA was purified from 5 mg biomass using 1 mL of Qiazol Lysis Reagent (Qiagen) according to the manufacturer’s instructions to the final volume of 20 µL. The RNA concentration was estimated at 260 nm, with purity checked at ratios of 260/280 (> 1.8) and 260/230, on a Nanodrop 1000 (Thermo).

Then, cDNA was synthesized with the Quantitect Reverse Transcription Kit (Qiagen) according to the manufacturer’s instructions. Prior to reverse transcription, 1 μg of total RNA was treated with the provided gDNA wipeout buffer for 2 min at 42 °C to eliminate gDNA contamination. The cDNA concentration and purity were then checked on a Nanodrop as previously described.

Quantitative real-time PCR analysis (qPCR)

Primers were designed using the OligoAnalyzer software (IDT, http://www.idtdna.com) based on sequences of mRNAs from the National Center for Biotechnology Information (NCBI) and the resulting primer pairs are listed in Table 3.

Table 3 List of primers used for amplification of target genes-related to fucoxanthin biosynthesis in the diatom Phaeodactylum tricornutum
Full size table

Transcription levels of selected genes were analyzed by qPCR using the QuantiNova SYBR Green PCR kit (Qiagen). Real-time reactions were performed from 100 ng cDNA and 1 µM of each primer per reaction and quantified using the Rotor-Gene Q thermocycler (Qiagen). The PCR product was analyzed and submitted to melting analyses as in Lopes et al. (2019). The efficiency of the reaction (E) was determined for each pair of primers and confirmed through the cDNA calibration curve from serial dilutions (1:2) with 800; 400; 200 and 100 ng (in duplicates) from a pool of all cDNA samples (90). All curves showed a correlation coefficient above 0.99 and efficiency between 95 and 105%. Normalization for each target gene was performed according to Rhinn et al. (2008), based in the cDNA concentrations in the samples. All data were calibrated by the mRNA levels of the control group (D12h, n = 3) and the results were analyzed by the 2−ΔCq method (Schmittgen and Livak 2008).

Statistical analysis

All data points are presented as the mean of each treatment (n = 3) plus the standard deviation (SD). Data normality and homoscedasticity were evaluated by the Shapiro–Wilk and Levene tests, respectively. One-way ANOVA coupled with Tukey post hoc test was performed to study the significance between light energy doses (six levels) and also the isolated factor irradiance. Two-way ANOVA coupled with Tukey post hoc test was performed to study the interaction between the same irradiances (three levels) on both L:D cycles (two levels). Figures were generated using GraphPad Prism (version 9.0), and statistical analyses were conducted using Statistica software (version 7.0). Moreover, a Pearson correlation analysis (RStudio software—Version 3.1.1; https://www.rstudio.com/) was performed to determine the degree of correlation between growth, photosynthetic parameters, pigment content and gene transcript levels. For all analyses, a level of significance of 5% was adopted.

Results

Growth performance

Phaeodactylum tricornutum cultures were diluted every 24 h and maintained in a biomass of 70 mg L−1 to minimize any self-shading effects. Figure 1 shows the effect of different light energy doses on specific growth rate. The different treatments showed similar photophysiological conditions between the same light energy doses of 2D12h/2D24h and 4D12h/4D24h with a specific growth rate of 0.18 ± 0.01 and 0.19 ± 0.01 day−1, respectively. Also, a higher specific growth rate of 0.21 ± 0,01 day−1 in 8D24h dose was found as compared to D12h dose (0.12 ± 0.01 day−1). For the treatments that received the same irradiance on both L:D cycles, there was significant differences (p < 0.05) for the irradiances 100 and 400 µmol photons m−2 s−1 in 24:00 h L:D cycle, reaching a specific growth rate of 1.5- and 1.2-fold higher than 12:00 h L:D cycle, respectively.

Fig. 1
figure 1

Specific growth rates in Phaeodactylum tricornutum cells under different light energy doses (D = 4.32 mol photons m−2). Each value is represented as mean ± standard deviation of biological triplicates (n = 3). Different capital letters represent the differences caused by the interaction between the same irradiance on both L:D cycles obtained by two-way ANOVA (post hoc Tukey, p < 0.05). Different small letters represent the differences caused by the light energy doses obtained by one-way ANOVA (post hoc Tukey, p < 0.05)

Full size image

Regarding light absorption by P. tricornutum cells (Supplementary Fig. S3), about 90% of the light was absorbed by cultures exposed to the 8D24h dose, while the treatment with a lower dose (D12h) presents a limitation in the amount of energy within the culture system, reaching 30% absorption. Treatments with the same light energy doses of 2D12h/2D24h and 4D12h/4D24h showed light absorption rates of 60 and 80%, respectively. For the treatments that received the same irradiance on both L:D cycles, there were significant differences (p < 0.05) for the irradiances 100, 200 and 400 µmol photons m−2 s−1 in 24:00 h L:D cycle, reaching a light absorption of 2-, 1.4- and 1.2-fold higher than 12:00 h L:D cycle, respectively.

Chlorophyll-a fluorescence in PSII

As can be seen in Fig. 2A, all treatments showed significant effects (p < 0.05) on the maximum quantum yield of PSII in P. tricornutum. The Fv/Fm of the photosynthetic apparatus decreased as the light intensity increased, in both L:D cycles. The cells treated with the 2D12h and 4D12h doses showed lower values (0.58 ± 0.04 and 0.45 ± 0.02) in relation to the same doses applied in the 24:00 h L:D cycle (0.69 ± 0.06 and 0.54 ± 0.04). However, the lowest Fv/Fm value (0.33 ± 0.01) was reported for cells exposed to the 8D24h dose, while the dose D12h presented two fold this quantum potential. For the treatments that received the same irradiance on both L:D cycles, there were no significant differences (p < 0.05) for the irradiances 100 µmol photons m−2 s−1, while the irradiances of 200 and 400 µmol photons m−2 s−1 in 12:00 h L:D cycle were 1.1- and 1.4-fold higher than 24:00 h L:D cycle, respectively.

Fig. 2
figure 2

Chlorophyll-a fluorescence parameters as Fv/Fm (A), NPQ (B) and ETR (C and D) in Phaeodactylum tricornutum cells under different light energy doses (D = 4.32 mol photons m−2) and irradiances. Each value is represented as mean ± standard deviation of biological triplicates (n = 3). Different capital letters represent the differences caused by the interaction between the same irradiance on both L:D cycles obtained by two-way ANOVA (post hoc Tukey, p < 0.05). Different small letters represent the differences caused by the light energy doses and different italic letters represent the differences caused by the isolated factor irradiance, both obtained by one-way ANOVA (post hoc Tukey, p < 0.05)

Full size image

Regarding non-photochemical quenching (NPQ, Fig. 2B), the maximum values (1.05 ± 0.04 and 1.37 ± 0.08) were found in the 4D12h and 8D24h doses respectively, reaching a dissipation of 2- and 2.7-fold higher than in the lowest light energy doses on both L:D cycles. However, there were no significant differences (p > 0.05) between the lowest irradiances, regardless of the irradiance of 400 µmol photons m−2 s−1. The electron transport rate (ETR, Fig. 2C and D) in P. tricornutum cultures, after receiving distinct light energy doses reached maximum values of 70 and 60 µmol electrons m−2 s−1 for the doses of 2D12h/4D24h and 4D12h/8D24h, respectively in 200 and 400 µmol photons m−2 s−1. Minimum values of 40 µmol electrons m−2 s−1 were found for the irradiance of 100 µmol photons m−2 s−1, regardless of the L:D cycles and the doses applied. In cultures that received the same light energy dose, only the 2D12h dose showed a significant difference (p < 0.05) with an ETR 1.5-fold higher in comparison to the 2D24h treatment.

Pigment content

Chlorophylls concentration in P. tricornutum cells changed significantly (p < 0.05) in all treatments (Fig. 3). The content of Chl-a (Fig. 3A) increased significantly (p < 0.05) at irradiances of 100 µmol photons m−2 s−1 (26.63 ± 0.61 mg g−1 DW) and 200 µmol photons m−2 s−1 (21, 34 ± 2.26 mg g−1 DW), regardless of L:D cycles. However, the level of these pigment decreased with the irradiance of 400 µmol photons m−2 s−1 at the dose of 4D12h reaching values of 16.54 ± 1.54 mg g−1 DW. The Chl-a content dropped even more in the 8D24h treatment, presenting values of 10.37 ± 0.62 mg g−1 DW. There was no significant difference (p > 0.05) between the irradiances of 2D12h/2D24h and 4D12h/4D24h doses. Chl-c content (Fig. 3B) was influenced by irradiance, since higher concentrations were found in D12h, 2D12h as well as in 4D24h and 8D24h doses, reaching values of 4 mg g−1 DW.

Fig. 3
figure 3

Chl-a (A) and Chl-c (B) concentrations in Phaeodactylum tricornutum cells under different light energy doses (D = 4.32 mol photons m−2) and irradiances. Each value is represented as mean ± standard deviation of biological triplicates (n = 3). Different capital letters represent the differences caused by the interaction between the same irradiance on both L:D cycles obtained by two-way ANOVA (post hoc Tukey, p < 0.05). Different small letters represent the differences caused by the light energy doses obtained by one-way ANOVA (post hoc Tukey, p < 0.05)

Full size image

In the case of Fx productivity (Fig. 4A and B), maximum concentrations were found at lowest light energy doses in both irradiances of 100 and 200 µmol photons m−2 s−1, reaching values of 21.87 ± 1.2, 20.72 ± 2.82, 19.92 ± 0.77 and 18.40 ± 0.83 mg g−1 DW in the doses of D12h, 2D12h, 2D24h and 4D24h, respectively. Moreover, the level of Fx decreased with the irradiance of 400 µmol photons m−2 s−1 at the dose of 4D12h and 8D24h reaching values of 15.03 ± 0.64 and 11.95 ± 1.34 mg g−1 DW, respectively.

Fig. 4
figure 4

Fx (A and B), Ddx (C and D) and Dtx (E) concentrations in Phaeodactylum tricornutum cells under different light energy doses (D = 4.32 mol photons m−2) and irradiances. Each value is represented as mean ± standard deviation of biological triplicates (n = 3). Different capital letters represent the differences caused by the interaction between the same irradiance on both L:D cycles obtained by two-way ANOVA (post hoc Tukey, p < 0.05). Different small letters represent the differences caused by the light energy doses and different italic letters represent the differences caused by the isolated factor irradiance, both obtained by one-way ANOVA (post hoc Tukey, p < 0.05)

Full size image

Regarding Ddx (Fig. 4C and D) and Dtx (Fig. 4E), it is evident that these two xanthophylls have an inverse relationship with each other. The decrease in irradiance resulted in an increase in Ddx production at doses D12h (0.90 ± 0.14 mg g−1 DW), 2D12h (0.81 ± 0.29 mg g−1 DW), 2D24h (0.83 ± 0.27 mg g−1 DW), and 4D24h (0.76 ± 0.13 mg g−1 DW), while the increase in irradiance reflected in a higher concentration of Dtx in the 4D12h (2.00 ± 0.09 mg g−1 DW) and 8D24h (5.46 ± 0.03 mg g−1 DW) doses. The accumulation of Dtx in the 8D24h dose was approximately 12-fold higher than in the lowest irradiances suggesting a joint resulted of conversion Ddx to Dtx and the enhancement of the xanthophyll biosynthesis.

The de-epoxidation state (DES, Fig. 5) at irradiances of 400 µmol photons m−2 s−1 reached approximately twofold the irradiances of 100 and 200 µmol photons m−2 s−1, which induces a rapid conversion from Ddx to Dtx as a necessity to photoprotect the photosynthetic apparatus. The was no significant differences (p > 0.05) between the same light energy doses of 2D12h/2D24h presenting a de-epoxidation of 52%, while significant differences (p < 0.05) were found at the 4D24h dose being 1.2-fold higher than 4D12h dose.

Fig. 5
figure 5

De-epoxidation state in Phaeodactylum tricornutum cells under different light energy doses (D = 4.32 mol photons m−2) and irradiances. Each value is represented as mean ± standard deviation of biological triplicates (n = 3). Different capital letters represent the differences caused by the interaction between the same irradiance on both L:D cycles obtained by two-way ANOVA (post hoc Tukey, p < 0.05). Different small letters represent the differences caused by the light energy doses obtained by one-way ANOVA (post hoc Tukey, p < 0.05)

Full size image

Carotenoid biosynthesis genes transcript levels

In the present study all treatments showed different transcription profiles (p < 0.05) in the regulation of light-dependent genes and with reactions related to the role of photoprotection. In the first group, which includes the ZEP1, ZEP2 and ZEP3 genes (Fig. 6A-C), a similar pattern between ZEP1 and ZEP2 was observed, characterized by the induction of relative gene overexpression ranging from 1.7-fold-change at dose 4D12h to 3.4- and 4.0-fold-change at 4D24h and 8D24h doses, respectively. In opposition to the ZEP1 and ZEP2 genes, a smaller variation in the ZEP3 gene expression was observed, in which only the 2D24h dose presented a significant (p < 0.05) lower gene expression from the other treatments.

Fig. 6
figure 6

Gene transcript levels of ZEP1 (A), ZEP2 (B) and ZEP3 (C) in Phaeodactylum tricornutum cells under different light energy doses (D = 4.32 mol photons m−2) and irradiances. Each value is represented as mean ± standard deviation of biological triplicates (n = 3). Different capital letters represent the differences caused by the interaction between the same irradiance on both L:D cycles obtained by two-way ANOVA (post hoc Tukey, p < 0.05). Different small letters represent the differences caused by the light energy doses obtained by one-way ANOVA (post hoc Tukey, p < 0.05)

Full size image

The second group comprising the VDE, VDL1 and VDL2 genes (Fig. 7A-D), showed a transcription profile significantly (p < 0.05) similar to the ZEP1 and ZEP2 genes, with an increase in the transcripts level in the higher light energy doses and irradiances, while VDE expression had a lower stimulation than in VDL1 and VDL2 genes. Regarding VDL1 and VDL2, higher values of relative gene overexpression were found at 4D12h/4D24h doses with a variation from 3.4- to 4.2-fold-change in VDL1 and from 2.4- to 4.1-fold-change in VDL2. There was no significant difference (p > 0.05) between the 4D24h and 8D24h doses of both genes.

Fig. 7
figure 7

Gene transcript levels of VDE (A and B), VDL1 (C) and VDL2 (D) in Phaeodactylum tricornutum cells under different light energy doses (D = 4.32 mol photons m−2) and irradiances. Each value is represented as mean ± standard deviation of biological triplicates (n = 3). Different capital letters represent the differences caused by the interaction between the same irradiance on both L:D cycles obtained by two-way ANOVA (post hoc Tukey, p < 0.05). Different small letters represent the differences caused by the light energy doses and different italic letters represent the differences caused by the isolated factor irradiance, both obtained by one-way ANOVA (post hoc Tukey, p < 0.05)

Full size image

Pearson correlation

In order to improve the interpretation of the data set and correlate the different responses obtained in this study, a Pearson correlation analysis was performed, in which the data were characterized according to their respective correlation coefficients and presented in a heatmap. As can be seen in Fig. 8, Ddx showed a significant positive correlation with Chl-a (r = 0.93), Fx (r = 0.93) and Fv/Fm (r = 0.89). On the other hand, the same carotenoid showed a significant negative correlation with the parameters involved in the photoprotection mechanism, including the Dtx (r = -0.70), DES (r = -0.84), NPQ (r =—0.77) and the genes VDE (r = -0.52), VDL1 (r = -0.67), VDL2 (r = -0.62), ZEP1 (r = -0.72) and ZEP2 (r = -0.72). In addition, Dtx showed a strong positive correlation with DES (r = 0.86), NPQ (r = 0.93) and the genes VDE (r = 0.50), VDL1 (r = 0.80), VDL2 (r = 0.60), ZEP1 (r = 0.77) and ZEP2 (r = 0.80). There were no significant correlations for Chl-c or ZEP3. ETR and the specific growth rate (µ) showed a slightly negative correlation with the parameters involved in light capture and photoinhibition, while both showed a slight positive correlation with DES and the genes involved in the energy dissipation process.

Fig. 8
figure 8

Heatmap based on Pearson's correlation coefficients between dependent variables of Phaeodactylum tricornutum cells exposed to different light energy doses. The abbreviations used are: µ: specific growth rate; Chl-a: chlorophyll-a; Chl-c: chlorophyll-c; Ddx: diadinoxanthin; Dtx: diatoxanthin, Fx: fucoxanthin, DES: de-epoxidation state; Fv/Fm: PSII maximum quantum yield; NPQ: non-photochemical quenching; ETR: electron transport rate; ZEP1, ZEP2 and ZEP3: zeaxanthin epoxidase gene; VDE: violaxanthin de-epoxidase gene; VDL1 and VDL2: violaxanthin de-epoxidase-like gene. Asterisks indicates significant differences with weak (*, p < 0.05), moderate (**, p < 0.01) and strong correlations (***, p < 0.001)

Full size image

Discussion

Light energy is a crucial environmental parameter in phototrophic microalgae cultures as it has a significant influence on the growth and biochemical composition of these organisms. Therefore, its adequate supply is considered one of the greatest challenges in photoautotrophic cultures, which demand adequate amounts to avoid photolimitation and photoinhibition phenomena (Broddrick et al. 2019). In addition, efficient conversion of light energy is also required, as most diatoms grow in environments where light availability fluctuates drastically throughout the year requiring adaptation of the photosynthetic apparatus (Wagner et al. 2016; Zhou et al. 2021). The dissipation of this amount of absorbed energy occurs both in PSII and in the diatom-FCP in the form of heat or in a way to redirect electrons, so the carbon fixation process is not overloaded (Ramanan et al. 2014). Even so, our growth data reported that the highest irradiance tested did not affect the growth performance of P. tricornutum, probably due to the fact that the highest irradiance tested is within the photo-saturation zone, as indicated by the P-I curve.

The Bunsen-Roscoe reciprocity law states that the light energy absorbed by a given species of photosynthetic organism is proportional to the light energy it emits (Bunsen and Roscoe 1859). In this case, P. tricornutum has been shown to adjust well to light energy doses and irradiances. In fact, the irradiances are an amount of photons offered per each second, while the doses depend on the exposure time as well. This ability enables P. tricornutum to achieve a balance between light absorption and energy dissipation, thereby protecting itself from potential photodamage.

The Fv/Fm values reported in the present study are within the expected range at the lowest energy doses, since they generally reach a value close to 0.7 in non-stressed cultures (Masojídek et al. 2013). This ratio is often used as an estimate of the photochemical yield of PSII and also as an indicator of photoinhibition, since it is associated with high irradiance, possibly indicating that the increase in light intensity has caused stress to P. tricornutum cells. In particular, the reason for this yield decline is that cells absorb more light that can be utilized during photosynthetic process, resulting in the need for energy dissipation through the NPQ to avoid photo-oxidative damage. Seydoux et al. (2022), for example, reported a similar relationship between NPQ and Fv/Fm during a high to low light relaxation/transition. However, in the present study, an inversely proportional relationship was found between these indicators, suggesting that PSII did not maintain the same photochemical capacity during this variation.

Comparing the NPQ results with the Fv/Fm and ETR values at the highest light energy doses, it can be observed that there is a difference between the potential of ETR with the Fv/Fm, since ETR remained stable, even with a possible photoinhibition in the irradiance of 400 µmol photons m−2 s−1. Although the Fv/Fm represents the maximum functioning capacity of the photosystem in terms of energy use, the destination of this energy is more associated with the ETR, according to the data presented in this study. In fact, electron transport is not only used for carbon fixation and biomass production during photosynthesis, but it is also necessary for energy to be destined to activate other functions of the biochemical metabolism of microalgal cells (Depauw et al. 2012). One of the possible alternative routes is the use of this electron flow in the reduction of ferredoxin enzyme to activate nitrite reductase enzymes during nitrogen assimilation (Hippmann et al. 2022). Furthermore, it can also be used to activate the thioredoxin enzyme (Blommaert et al. 2021), that acts in the Calvin cycle and in Mehler reaction (Lepetit et al. 2022) aiming to eliminate potential ROS and other free radicals. This explains why ETR behaves differently than Fv/Fm and why it may not be positively correlated with Chl-a and Fx.

The most common feature of the acclimatization response to irradiance is a change in the level of pigmentation, typically associated with changes in the abundance of the antenna complex. The high amount of the light-harvesting pigments (Chl-a and Fx) observed in this study are in agreement with Kuczynska et al. (2020), who also identified a significant increase of these parameters in low light in comparison to high light conditions. In addition, given the differences obtained in chlorophylls and Fx in comparison with Dtx and NPQ, it is possible to say that this antenna protein complex is involved in the acclimation regulation in P. tricornutum (Conceição et al. 2020).

Photoprotective mechanisms are induced to protect the antenna complex from overexcitation of the photosystems when the light intensity is too high or increases too rapidly for the photochemical phase to utilize the absorbed energy. Excess of excitation energy in the PSII antenna complex is dissipated as heat through many cellular reactions, which are referred to as non-photochemical quenching of chlorophyll fluorescence emission (Goss and Lepetit 2015).

Diatoms have developed efficient photoprotective mechanisms to minimize photoinhibition (Costa et al. 2013; Valle et al. 2014). Among the short-term defenses that are activated by a sudden increase in light intensity, photoprotective dissipation of excess absorbed light energy is known as an important mechanism, where the xanthophylls Ddx and Dtx are involved. NPQ is mainly controlled by the interconversion between epoxidation and de-epoxidation of carotenoids involved in Ddx-Dtx cycle and most of the excess energy is discarded as heat by a regulatory system involving the transtylakoid ΔpH, which are of great importance in the photoregulation of photosynthesis (Lavaud et al. 2012; Seydoux et al. 2022). This energy-dependent system reduces the quantum efficiency of PSII, helping to prevent excessive PSII reduction and photoinhibitory damage (Fisher et al. 2020). Efficiently in effect, this system provides a means of dumping excess energy before it can damage the photosynthetic apparatus.

Our results showed that the physiological and molecular processes are connected with energy dissipation in P. tricornutum. The regulation of VDL1 and VDL2 genes at high irradiance is associated with an increase in NPQ and Dtx concentration while lower levels of Ddx and Fx were observed. In this scenario, violaxanthin de-epoxidase enzyme has the function of converting Ddx to Dtx, which is of great importance for the photoprotection mechanism in high irradiance.

On the other hand, ZEP1 and ZEP2 genes did not show an inverse reaction in relation to VDL1 and VDL2 genes, although it is known that the enhancement of Dtx synthesis requires the enhancement of epoxidases activity. Furthermore, the expression of these genes often show variation between samples collected at the end of the light phase and at the end of the dark phase. This difference in sampling time could potentially influence the results, since diurnal fluctuations of the mentioned genes typically increase in amplitude during prolonged high light exposure, suggesting that their products may play a crucial role in long-term high light acclimation as reported by Kuczynska et al. (2020).

These genes have been demonstrated, through transcriptomic and photophysiological analyses, as global regulators of photosynthetic acclimation to irradiance by participating in the xanthophyll cycle, which is involved in the acclimation of plastids to changes in environmental light conditions (Bowler et al. 2008; Eilers et al. 2016; Dautermann and Lohr 2017; Dautermann et al. 2020). Furthermore, Blommaert et al. (2021), concluded that ZEP enzyme is inhibited at light irradiances lower than the light saturation (EK) and the onset of NPQ. Also, according to the same authors, this regulation of ZEP genes during high irradiance may be associated with the ZEP enzymes reactivation, since this is linked to chloroplast redox regulation with thioredoxin proteins. In this case, the overexpression of these genes at higher irradiance may be related to the excess of electrons destined for the photosynthetic process. Moreover, the balance adjustment of the Ddx-Dtx cycle and the relaxation of the NPQ seem to be affected by the size of the pool of these xanthophylls in relation to the regulation of the VDL and ZEP genes in P. tricornutum.

According to Bai et al. (2022), there is an interaction between the isoforms of VDL2 and ZEP1, VDL1 and ZEP2 as well as VDE and ZEP3 genes, which were evolved by the duplication of enzyme-encoding genes with important roles in the Fx biosynthesis, preferably related to the photoprotection mechanism. However, the activation and localization of VDL proteins may differ from those of VDE proteins, as the Glu-rich domain in VDL1 and VDL2 sequences has been entirely swapped with a neutral C-terminal domain, as reported by Coesel et al. (2008). In this case, our results suggest that ZEP1, ZEP2, VDL1 and VDL2 genes may be more efficiently and eventually related to the photoprotection, since VDE and ZEP3 genes were not. In addition, data from this study have shown that these genes were similar to each other in a way that their expression was not correlated to the photoperiod, which are in agreement with Kuczynska et al. (2020). An interesting discovery was reported by Dautermann et al. (2020) where the VDL enzymes play a crucial role as it utilize Vx, a major substrate, to direct carotenoids into either the photoprotective or light-harvesting branch of carotenoid biosynthesis, in the same way as Nx serves as the precursor for both Ddx and Fx. It is of great importance to explore and characterize the Fx biosynthetic pathway to identify the enzymes involved that have not yet been discovered and the specificity of their reactions, which can highlight even more the diversification of carotenoid functions in chromalveolate algae.

Conclusion

In conclusion, this study demonstrated the Bunsen-Roscoe reciprocity law was valid, since the use of light energy doses and irradiances affected the photoacclimation and photoprotection responses in P. tricornutum. The specific growth rate was 1.9-fold higher in 8D24h dose with threefold higher light absorption in comparison to dose D12h, even though there was more light stress as indicated by the photosynthetic parameters. The higher content of the pigments involved in the light-harvesting complex were influenced by low light intensity, while the regulation of the xanthophyll cycle was potentiated by the highest light energy doses. The transcriptional profile of the ZEP1, ZEP2, VDL1 and VDL2 genes was influenced by the highest light energy doses, except for VDE and ZEP3 genes, which were poorly regulated by light. In addition, a similar transcription pattern was found between two isoforms of ZEP genes as well as in VDL genes, which suggests the importance of both transcription products on photoprotection in P. tricornutum.