Photosystem rearrangements, photosynthetic efficiency, and plant growth in far red‐enriched light

SUMMARY

Arabidopsis plants were grown in white light (400–700 nm) or in white light supplemented with far‐red (FR) light peaking at 730 nm. FR‐enriched light induced the typical shade avoidance syndrome characterized by enhanced length of seedling hypocotyl and leaf petiole. FR supplementation also caused a noticeable decrease in the carotenoid and chlorophyll content that was attributable to a block of pigment accumulation during plant development. The carotenoid decrease resulted from a downregulation of their biosynthesis pathway rather than carotenoid degradation. The losses of photosynthetic pigments are part of structural and functional rearrangements of the photosynthetic apparatus. The plastoquinone pool was chronically more oxidized in plants acclimated to white + FR light compared to white light‐grown plants. Growth in FR‐enriched light was associated with a higher photochemical efficiency of PSII compared to growth in white light and with a substantial increase in root and shoot biomass production. Light distribution between the photosystems was modified in favor of PSII by an increase in the PSII/PSI ratio and an inhibition of state transitions. Neither LHCII abundance nor nonphotochemical energy dissipation in the PSII chlorophyll antennae were modified significantly by the addition of FR light. A PSI supercomplex, not previously observed in Arabidopsis, was specifically found in plants grown in FR‐enriched light. This large PSI complex contains a supplementary Lhca1‐4 dimer, leading to a total of 6 LHCI antennae instead of 4 in the canonical PSI. Through those photosystem rearrangements and the synergistic interaction with white light, FR light is photosynthetically active and can boost photosynthesis and plant growth.

Significance Statement

Arabidopsis plants were grown in white light supplemented or not with far‐red light. Far‐red‐enriched light induced both structural and functional modifications of the photosystems, with noticeable changes in photosynthetic pigment content, photosystem stoichiometry, PSI supercomplex organization, and state transition capacity. Those modifications were associated with an enhancement of photosynthetic electron‐transport efficiency and plant biomass productivity. Through photosystem rearrangements and synergistic interaction with white light, far‐red light is photosynthetically active and can boost photosynthesis and plant growth.

INTRODUCTION

Plants harvest and utilize light energy to convert CO₂ and H₂O into organic matter, concurrently releasing O₂, through the photosynthesis mechanism, allowing them to fuel energy‐demanding cellular activities (Blankenship, 2021). This process relies on both light quantity and quality. Based on the work of McCree (1971, 1972), the concept of photosynthetically active radiation (PAR) has been established, which is defined as the light driving photosynthesis within the wavelength range from 400 to 700 nm. Light wavelengths outside this range, either shorter than 400 nm or longer than 700 nm, are considered as unimportant for photosynthesis, due to the low quantum yield of CO₂ assimilation when they are applied to plants as a single energy source (McCree, 1971). However, this does not mean that the non‐PAR light has no effect on photosynthesis. Emerson (1957) showed that the photosynthesis rates of Chlorella cells increase when long‐wavelength light (680–720 nm) was combined with shorter‐wavelength light. This enhancement effect was subsequently observed in several plant species when far‐red (FR) light (>700 nm) was combined with white, blue or red light (e.g. Canaani et al., 1982; Huber et al., 2024; Zhen & van Iersel, 2017). The Emerson effect reflects the unbalanced light energy distribution between PSI and PSII. Optimal excitation wavelengths differ for both photosystems, with FR light preferentially exciting PSI and several shorter wavelengths preferentially exciting PSII. Thus, adding FR light absorbed by PSI on shorter wavelength light favoring PSII excitation can relieve the PSI limitation on photosynthetic electron transport. In addition, there are various regulatory mechanisms that adjust light energy distribution between the photosystems in the short or medium term such as the so‐called state transitions or the transcriptional regulation of the antennae abundance (Rochaix, 2014).

Compared to vascular plants, the ability to use FR light for photosynthesis can be much higher in some photosynthetic microorganisms. In some cyanobacteria, special forms of chlorophyll, named chlorophylls d and f, extend the spectral range that can be utilized effectively for photosynthesis into the FR region at about 700–750 nm (Gan et al., 2014; Wolf & Blankenship, 2019). This feature can be ecologically advantageous since the visible range between 400 and 700 nm spectral region makes only about half of the incident solar spectrum (Zhu et al., 2010). It has been estimated that, in an idealized situation, absorption of all photons of wavelengths between 700 and 750 nm in a crop field would lead to a 19% increase in absorbable photons (Chen & Blakenship, 2011). So, the spectral expansion in FR‐absorbing cyanobacteria can allow them to thrive in niches not available to spectrally more‐limited competitors (Swingley et al., 2008; Duxbury et al., 2009). Some algae are also able to spectrally tune the absorption of chlorophyll a by modifications of pigment–protein interactions in their light‐harvesting antennae, enhancing FR light absorption (Wolf & Blankenship, 2019). These adaptations of specialized microorganisms have been suggested to provide a potential approach to increase photosynthesis and biomass production of vascular plants by creating FR light‐absorbing crops. Furthermore, in shaded environments, such as vegetation proximity or canopy shade, the red‐to‐FR light ratio of the incident light markedly decreases due to preferential reflection and transmission of FR light (Cutolo et al., 2023; Martinez‐Garcia & Rodriguez‐Concepcion, 2023; Roig‐Villanova & Martínez‐García, 2016). Those findings explain the growing interest in exploring the impact of FR‐enriched white light on the morphology, physiology, and development of vascular plants, as well as investigating the efficiency with which plants utilize FR light (Tan et al., 2022).

FR‐enriched white light has been demonstrated to induce the shade avoidance syndrome (SAS) in some plant species, including Arabidopsis thaliana, through the regulation of phytochrome activity, specifically PHYB and PHYA, activating a transduction pathway involving the key regulators PIF7 (PHYTOCHROME INTERACTING FACTOR7) and HY5 (ELONGATED HYPOCOTYL5) (Burko et al., 2022; Li et al., 2012; Ortiz‐Alcaide et al., 2019). PIF7 is a positive regulator of SAS while HY5 acts as a negative regulator (Martinez‐Garcia & Rodriguez‐Concepcion, 2023). SAS responses include stem stretching, internode length and branching patterns. In low red/FR environments, a decrease in photosynthetic pigment content was also observed (Bout‐Torrent et al., 2015; Morelli et al., 2021), correlated with alterations in chloroplast ultrastructure (Morelli et al., 2021). However, studies on the effect of growing plants in FR‐enriched white light on the photosynthetic apparatus are scarce (Huber et al., 2024; Kono et al., 2020; Melis & Harvey, 1981; Morelli et al., 2021; Zhen & Bugbee, 2020). In particular, there is no data available on the organization of the photosynthetic complexes in white + FR light conditions other than the PSII‐PSI stoichiometry. This prompted us to perform a comparative study of Arabidopsis plants grown in white light or in white light supplemented with FR light (730 nm). FR light supplementation caused a decrease in photosynthetic pigments, which was independent of the SAS response and reflected substantial changes in the photosynthetic properties of the leaves, which are the focus of the present study.

RESULTS

Plant growth

Arabidopsis seedlings grown in vitro in white light for 3 days were exposed to FR‐enriched light (W + FR) for 4 days. The PAR irradiance (from 400 to 700 nm) was set to the same value (140 μmol photons m⁻² s⁻¹) for the two conditions. Seedlings in W + FR displayed the expected SAS with a 5‐fold increase in hypocotyl length compared to seedlings kept in white light (W) (Figure 1A,B). This response is typical of shade‐avoider species to favor apical dominance and to escape from shade (Martinez‐Garcia & Rodriguez‐Concepcion, 2023). Remarkably, the shoot and root biomass noticeably increased in W + FR by 54% for the shoot fresh mass (Figure 1C), 40% for the shoot dry mass (Figure 1D) and 50% for the root fresh mass (Figure 1E). Similar phenotypic characteristics were observed in 18‐days old plants grown on soil exposed to W + FR for 7 days, with a 1.8‐fold increase in leaf petiole elongation (Figure 1F,G). The shoot fresh mass and dry mass also increased (+36% and +29%, respectively) compared to plants grown in W conditions (Figure 1H,I). Thus, FR‐light supplementation at constant PAR is beneficial for plant growth.

Figure 1.

Growth phenotype of WT Arabidopsis in response to supplemental FR light.

(A) Arabidopsis seedlings grown in vitro under white light for 7 days (W condition) or grown for 3 days under white light and subsequently transferred to W + FR for 4 days (W + FR condition).

(B) Hypocotyl elongation of 7‐days old seedlings under W or W + FR conditions. Data are mean values ± SD of at least 30 repetitions. Wilcoxon test with P‐value <0.001.

(C) Shoot fresh mass, (D) shoot dry mass, (E) root fresh mass of 7‐days old seedlings grown in vitro under W or W + FR conditions. Data are mean values ± SD of at least 4 independent samples of 10 roots or shoots. Student's t test.

(F) Plants grown on soil for 25 days in W or grown for 18 days in W and transferred for 7 days to W + FR.

(G) Petiole elongation in 25‐days old plants grown on soil under W or W + FR conditions. Data are mean values ± SD of at least 18 repetitions. Wilcoxon test with P‐value <0.001.

(H) Shoot fresh mass and (I) shoot dry mass of 25‐days old plants grown on soil under W or W + FR conditions. Data are mean values ± SD of at least 14 repetitions. Student's t test with *, **, *** indicating statistical difference with P < 0.05, 0.01 and 0.001, respectively.

Photosynthetic pigments

Light with a low red/FR ratio was previously reported to induce a reduction of photosynthetic pigments in some plant species (Bout‐Torrent et al., 2015; Morelli et al., 2021). Accordingly, the W + FR treatment led in average to a 30% decrease in photosynthetic pigment levels in 7‐days old in vitro plants (Figure 2A). The relative decreases in chlorophylls and carotenoids were very similar: around 28% for chlorophyll a, 27.5% for chlorophyll b and 30% for carotenoids. The same phenomenon was observed in 25‐days old plants grown on soil, exhibiting a 10% decrease in pigments in W + FR light (Figure S2). Figure 2B,C shows the time course of the photosynthetic pigment changes after transfer of 3‐days old in vitro plants to W + FR compared to seedlings kept in W light. During the first 48 h after transfer to W + FR, the pigment levels increased similarly in both conditions, between 1.5‐ and 2.2‐fold. However, from 48 to 96 h, the kinetics of accumulation markedly differed. In W conditions, chlorophylls and carotenoids continued to raise, whereas, in W + FR, the levels remained almost constant, resulting in a strong difference in pigment concentrations between the two conditions. This latter effect could be due to a block in pigment biosynthesis, to pigment degradation or to both. We performed experiments to clarify this point.

Figure 2.

Photosynthetic pigments in WT Arabidopsis seedlings in response to supplemental FR light.

(A) Photosynthetic pigments (chlorophyll a, Chl. a; chlorophyll b, Chl. b and total carotenoids, Carot.) in 7‐days old seedlings grown in vitro under W or W + FR conditions. Data are mean values ± SD of at least 6 independent samples (pooled shoots). Wilcoxon test with P‐value <0.01. Time course of the changes in (B) total chlorophyll and (C) total carotenoid in seedlings grown in vitro under W or W + FR conditions. Data are mean values ± SD of at least 3 independent samples (pooled shoots).

(D) Relative expression levels of carotenoid biosynthesis genes in seedlings grown for 3 days in W and then transferred to W + FR for 8–72 h. Colors indicate relative expression levels in W + FR relative to W as indicated by the color bar. Data are mean values of at least 3 independent samples (pools of shoots).

(E) Chloroplast density in leaves of 7‐days old seedlings grown in vitro under W or W + FR conditions. Data are mean values ± SD of at least 30 independent confocal microscopic measurements. Wilcoxon test with P‐value <0.001.

We were particularly interested in the possibility of carotenoid degradation in FR light because enzymatic and non‐enzymatic oxidative carotenoid cleavages lead to carotenoid oxidation products, called apocarotenoids, which can have important biological functions (Havaux, 2020a; Moreno et al., 2021). In particular, short‐chain apocarotenoids, such as β‐cyclocitral or β‐ionone, are volatile compounds that act as growth regulators (Havaux, 2020a; Felemban et al., 2024). Their accumulation during plant acclimation to FR would raise the interesting possibility of a carotenoid control of SAS.

First, we studied the expression levels of some genes coding for carotenoid degradation enzymes: the carotenoid cleavage dioxygenases CCD1, CCD4, CCD7 and CCD8 and the lipoxygenase 2 (LOX2) (Figure 3A). CCDs are enzymes specialized in the oxidative cleavage of carotenoids (Harrison & Bugg, 2014; Ohmiya, 2009), and LOXs are lipid dioxygenases that can also use carotenoid as substrates (Gao et al., 2019). The transcript levels for those genes showed very little sensitivity to the transfer from W to W + FR. No clear induction of any of those genes was observed in W + FR, except a transient increase in CCD4 expression in the first steps (8 h) of acclimation to W + FR.

Figure 3.

Carotenoid degradation in Arabidopsis plants in response to supplemental FR light.

(A) Relative expression levels of carotenoid biosynthesis genes in WT seedlings grown in vitro for 3 days in W and then transferred to W + FR for 8–72 h. Colors indicate relative expression levels in W + FR relative to W as indicated by the color bar. Data are mean values of at least 3 independent samples (pools of shoots).

(B) Apocarotenoid content in 7‐days old WT seedlings grown in vitro under W or W + FR. Putative identifications are based on mass spectra after LC–MS analysis. Data are mean values ± SD of at least 14 independent samples (pooled shoots). Wilcoxon test with P‐value <0.05.

(C) Hypocotyl elongation of 7‐days old in vitro‐grown seedlings of WT Arabidopsis and of ccd and lox2 mutants under W or W + FR conditions. Data are mean values ± SD of at least 50 repetitions. Kruskal‐Wallis test with P‐value <0.001.

(D) Total carotenoids in 7‐days old in vitro‐grown seedlings of WT Arabidopsis and of ccd and lox2 mutants under W or W + FR conditions. Data are mean values ± SD of at least n = 3 independent samples (pooled shoots). Kruskal‐Wallis test with P‐value <0.001. Different letters indicate statistical differences.

We also quantified apocarotenoid levels in leaves by UHPLC–MS. 20 putative apocarotenoids were detected and quantified (Figure 3B). The apocarotenoid levels decreased in W + FR compared to W. In fact, the extent of the apocarotenoid reductions was proportional to the decrease in the carotenoid content. The only exceptions were 3‐OH‐apo‐13‐carotenone‐B and C, which tended to increase in W + FR (although the difference with W was not statistically significant).

Finally, we studied the phenotypic response of ccd and lox2 mutants to W + FR (Figure 3C). All genotypes displayed an increased hypocotyl elongation in W + FR. However, the effect was significantly less marked in the mutants compared to WT. In contrast, the reduction of the carotenoid content in W + FR was similar in WT and in the mutants (Figure 3D). This finding confirms that enzymatic carotenoid oxidation is not a major factor in the pigment decrease in W + FR.

Short‐term exposure of Arabidopsis seedlings (up to 8 h) to W + FR light was previously shown to induce a significant decrease in the transcript levels of the PSY gene encoding phytoene synthase (Bout‐Torrent et al., 2015). We performed a more detailed analysis of the carotenoid biosynthesis pathway: The expression levels of a number of genes coding for carotenoid biosynthesis enzymes (PSY, PDS3, CrtISO, LCYs) were measured by RT‐qPCR in Arabidopsis seedlings after long‐term acclimation to W + FR (up to 72 h, Figure 2D). Their transcript levels were downregulated in W + FR compared to W. The strongest downregulation was observed for the gene coding for PSY, the known rate‐limiting enzyme in carotenoid biosynthesis.

Leaf expansion, photosynthetic pigment biosynthesis and chloroplast development are regulated in a coordinated manner (Gügel & Soll, 2017). Consequently, we analyzed the chloroplast density (number of chloroplasts by unit leaf area) by confocal fluorescence microscopy (Figure 2E; Figure S3). We observed that chloroplast density was lowered in W + FR compared to W. Possibly, this change, albeit rather modest, could partly contribute to the reduction of the photosynthetic pigment content in W + FR.

Taken together, the data of Figures 2 and 3 lead to the conclusion that the lower levels of carotenoids in W + FR relative to W is due to an inhibition of the carotenoid biosynthesis pathway rather than a stimulation of carotenoid degradation.

Phytochrome signaling

PIF7 is a major regulator of the shade response (Martinez‐Garcia & Rodriguez‐Concepcion, 2023). Its activity is controlled by its phosphorylation state in relation to phytochrome PHYB active or inactive forms. PIF7 promotes the expression of auxin biosynthesis genes (de Wit et al., 2015; Li et al., 2012). PIF7‐defective lines displayed a strongly reduced proximity shade response (Jiang et al., 2019; Li et al., 2012; Mizuno et al., 2015). This was observed in Figure 4A showing a markedly reduced hypocotyl elongation in the pif7 mutant relative to WT (3.4‐fold against 1.7‐fold increase in WT and pif7, respectively). However, the shoot biomass increased in a similar way in WT and pif7 (+24%–25% and +21%–23% for the shoot fresh and dry mass, respectively) (Figure 4B,C). Moreover, the pigment decreases in W + FR were similar in WT and pif7 (−30%–31% and –32%–36% for chlorophylls and carotenoids, respectively) (Figure 4D,E).

Figure 4.

Growth and photosynthetic pigment content of WT and pif7 Arabidopsis plants in response to supplemental FR light.

(A) Hypocotyl elongation of 7‐days old seedlings under W or W + FR conditions. Data are mean values ± SD of at least 80 samples. Kruskal‐Wallis test with P‐value <0.001.

(B) Shoot fresh mass and (C) shoot dry mass of 7‐day old seedlings grown in vitro under W or W + FR conditions. Data are mean values ± SD of at least 3 independent samples (pools of 20 shoots). Kruskal‐Wallis test with P‐value <0.05.

(D) Total chlorophyll and (E) total carotenoids in 7‐days old in vitro‐grown seedlings under W or W + FR conditions. Data are mean values ± SD of at least 4 independent samples (pools of shoots). Kruskal‐Wallis test with P‐value <0.05.

(F) Relative expression levels of carotenoid biosynthesis genes in seedlings grown for 3 days in W and then transferred to W + FR for 8–72 h. Colors indicate relative expression levels in W + FR relative to W as indicated by the color bar. Data are mean values of at least 3 independent samples (pools of shoots). Different letters indicate statistical differences.

WT + FR induced a downregulation of the expression of carotenoid biosynthesis genes in pif7 as previously observed in WT (Figure 4F). The FR‐induced decrease in expression of β‐LCY, PSY, and CrtISO appeared to be slightly less pronounced in pif7 leaves compared to WT leaves, but the differences were not statistically significant. Taken together, the data of Figure 4 indicate that FR‐induced enhancement of plant growth and reduced pigment levels are not directly connected to phytochrome. One can also conclude that the carotenoid and chlorophyll reduction does not play a major role in the attenuated SAS in pif7.

Photosystem composition and functioning

The photochemical efficiency of PSII (Φ_PSII) was measured in intact seedlings under the growth light conditions by chlorophyll fluorometry (Figure 5A). Seedlings grown in W + FR displayed the highest Φ_PSII value measured (around 0.72). This value is higher than Φ_PSII measured in W‐grown seedlings (0.67). When chlorophyll fluorescence emission from W‐grown seedlings was measured in W + FR, there was a slight increase in Φ_PSII (to 0.73). This increase corresponds to the so‐called Emerson enhancement effect (Canaani et al., 1982; Emerson, 1957), which is the immediate enhancement of electron transport by FR light by release of the PSI limitation. In contrast, when Φ_PSII of plants grown in W + FR was measured in W conditions, there was a marked drop in Φ_PSII to 0.55. Φ_PSII was also measured in plants grown on soil. Φ_PSII was significantly higher in plants grown in F + W compared to plants grown in W (Figure S4), confirming the results obtained in in vitro grown seedlings.

Figure 5.

Photosynthetic characteristics of WT Arabidopsis grown in white light or in white light supplemented with FR light.

(A) PSII photochemical efficiency Φ_PSII in leaves of 7‐days old plants grown in vitro under W or W + FR conditions. Measurements were done both in W and W + FR light conditions. Data are mean values ± SD of n > 40, test P‐value <0.001. Different letters indicate that values are statistically different.

(B) Plastoquinone‐9 content (reduced and oxidized PQ9) and (C) plastoquinone‐9 oxidation state in 7‐days old plants grown in vitro under W or W + FR conditions. Leaves were harvested in W and W + FR light conditions for both growth conditions. Data are mean values ± SD of n > 4, test P‐value <0.05.

(D, E) qS and qT chlorophyll fluorescence parameters reflecting state transition. Data are mean values ± SD of at least 3 repetitions. Wilcoxon test with P‐value <0.05.

(F) SDS‐PAGE of thylakoid membranes from 7‐day old plants grown in vitro under W or W + FR conditions. A representative lane of 3 biological replicates per growth condition is shown.

(G) PSII/PSI ratio as measured by abundance ratio CP43/‘PSI small (units)’. Protein abundances were measured after densitometric analysis of the gel in F. Note that the CP43/“PSI small” ratio does not represent the real stoichiometric ratio CP43/PSI small subunits since the calculation is based on the SYPRO signal; More important is the variation between the two conditions which represents an increase of about 23% in PSII/PSI in W + FR condition. Data are mean values ± SD of 3 biological repetitions (pools of plantlet shoots).

(H) 77 K chlorophyll fluorescence emission spectra of thylakoid extracts from 7‐day old plants grown in vitro under W or W + FR conditions. Excitation light was at 440 nm. The data are the average of 3 biological replicates per growth condition, each one tested in two technical repetitions. Spectra are normalized on PSII maximum emission (around 685 nm). Standard deviation at the maximum PSI emission (around 730 nm) is shown.

The antenna system of PSI contains the so‐called far‐red chlorophylls, which absorb at wavelengths beyond 700 nm (Gobets & van Grondelle, 2001). Consequently, FR preferentially excites PSI. Not surprisingly, the plastoquinone‐9 (PQ) pool was noticeably more oxidized in W + FR‐grown seedlings than in W‐grown seedlings (Figure 5B).

Plants possess a mechanism called state transitions to balance excitation energy distribution between the two photosystems (Minagawa, 2011; Ruban & Johnson, 2009). When PSII is overexcited compared to PSI, the PQ pool becomes more reduced, triggering phosphorylation of the major light‐harvesting complex of PSII (LHCII) and its migration to PSI to reequilibrate light energy distribution between the two photosystems. This rearrangement is called state 2. On the opposite, overexcitation of PSI leads to the inverse phenomenon, with dephosphorylated LHCIIs moving back to PSII (state 1) to improve light absorption in PSII. The chronic oxidation of PQ in plants acclimated to W + FR is a condition that should maintain the photosystems in state 1 (Figure 5C). Using chlorophyll fluorometry with red and FR actinic lights, we measured the capacity for state transitions by calculating the qT and qS parameters as described by Ruban and Johnson (2009) (Figure 5D,E). Red + FR light was used to induce state 1, and red light alone was used to reach state 2. Both qS and qT were lower in W + FR plants compared to W plants, indicating a reduced capacity for state transition and preferential binding of LHCII to PSII. A high oxidation status of the PQ pool and a lowered capacity for state transitions would maintain the association of LHCII with PSII, thus favoring PSII excitation under W + FR conditions that can lead to PSI overexcitation.

Non‐photochemical energy quenching (NPQ) is another mechanism that can modulate light absorption in PSII antennae by increasing thermal dissipation of the absorbed light energy (Ruban, 2016). NPQ was measured in W and W + FR seedlings illuminated with red light or white light combined or not with FR light (Figure S5). No significant difference was found between the two types of plants in the different light conditions.

We also measured the photosystem stoichiometry by biochemical and fluorometric analyses (Figure 5F–H). Thylakoid complexes were separated by SDS‐PAGE. The SYPRO‐stained bands were used for quantification because of the highly reliable and linear signal. In particular, CP43 was used as an indicator of the PSII reaction center level, whereas the ‘PSI small subunits’ band was used to quantify PSI. PSI small subunits correspond to some or all these PSI core subunits: PsaE, PsaH, PsaO, PsaN, PsaC, and PsaK (9–10 KDa). This region in the gel is practically depleted of PSII bands, as shown in Ksas et al. (2022) and Crepin et al. (2020). Subunit C of ATP synthase (8 kDa) is the only subunit of a major complex that is present in this region and migrates at the bottom of this “PSI small subunits” region (Crepin et al., 2020). However, this band has a minor contribution and moreover ATP synthase did not significantly change with FR‐light supplementation as shown by the gamma subunit signal, which is the band below CP43. Therefore, it causes at worst a minor error in the estimation of the PSII/PSI ratio variations between the two growth conditions. CP43 was identified by SDS‐PAGE and immunoblot using PSII core and PSII membranes as in Ksas et al. (2022) and Crepin et al. (2020). This subunit was chosen since at this position there are no significant interferences from subunits of other photosynthetic complexes (PSI, ATP Synthase, Cyt b6f).

Plants grown in W + FR displayed an increased PSII/PSI ratio (CP43/small PSI) by +23% compared to plants grown in W (Figure 5F,G), as expected (e.g. Chow et al., 1990; Melis & Harvey, 1981). Similar results were found by measuring 77 K chlorophyll fluorescence spectra (Figure 5H), with a 20% decrease in the fluorescence emitted by PSI (at 730 nm) in W + FR. Those photosynthetic apparatus adjustments, although rather modest, allow rebalancing the PSI over‐excitation by FR and increase the photosynthetic efficiency in W + FR.

Photosystem complexes in far red‐enriched light

Photosynthetic complexes were solubilized from thylakoid membranes of plants grown in W or W + FR and separated by Blue Native (BN)‐PAGE and Clear Native (CN)‐PAGE (Figure 6A). The profiles obtained for the two growth conditions are rather similar, except a few differences: The PSI band was decreased, as expected from the data of Figure 5, and a new band was visible in the W + FR profile between the PSI‐LHCII and PSI complexes (Figure 6A). This band was absent in W‐grown plants. The low fluorescence of this new band (Figure 6A, right panel) suggests that it corresponds to a PSI supercomplex. In W + FR, this new band was more intense than the PSI‐LHCII band (Figure 6B). To the best of our knowledge, this PSI band has not been previously observed in Arabidopsis. Considering its position between PSI and PSI‐LHCII and that LHCII has a size of ~120 kDa, this new PSI band (hereafter denoted PSI*) could be a PSI complex with an additional component of about 70–80 kDa. The canonical PSI is known to bind 4 LHCI (Lhca1, a2, a3 and a4, 35–40 kDa each). PSI* could correspond to the PSI complex previously found in green algae (Ozawa et al., 2018; Qin et al., 2019; Suga et al., 2019) which contains an additional Lhca dimer (~80 kDa) which fits very well with the increased size of PSI* compared to PSI. Crepin et al. (2020) reported the existence in Arabidopsis of a PSI‐LHCII supercomplex containing an additional Lhca1‐Lhca4 dimer, which was also found here in the green gels of W– and W + FR‐grown plants. This PSI‐6LHCI‐LHCII band is present in the green gels of Figure 6 above the PSI‐LHCII band, and its abundance was similar in W and W + FR thylakoids.

Figure 6.

Photosynthetic complexes of WT Arabidopsis grown in white light or in white light supplemented with FR light.

(A) Left side: Blue native PAGE (BN‐PAGE) and clear native PAGE (CN‐PAGE) of thylakoid membranes from 7‐day old plants grown in vitro under W or W + FR conditions. The new band specific of W + FR conditions is indicated in red. Right panel: Fluorescence emission of the CN‐PAGE gel. Excitation in the blue region by using an UV conversion screen with max at 455 nm and FHWM of 45 nm; emission in the red region by using a long pass filter >600 nm. The BN‐ and CN‐PAGEs here shown are representative of experiments repeated on 4 biological replicates (and many more technical replicates).

(B) Densitometric analysis of the CN‐PAGE showing the new band in red.

(C) Absorption spectrum of PSI‐6LHCI‐LHCII, PSI and PSI* particles eluted from the BN‐PAGE gels.

(D) Comparison of the absorption spectra of the PSI‐6LHCI‐LHCII and PSI particles previously published in Crepin et al. (2020) with a simulation of the absorption spectrum of a PSI complex containing an additional Lhca1‐a4 dimer (PSI‐6LHCI simulated). See the Experimental Procedures section for the simulation method.

We cut the bands corresponding to PSI, PSI* and PSI‐6LHCI‐LHCII from the native electrophoretic gel and eluted the particles. The absorption spectrum of the new complex was different from the PSI‐6LHCI‐LHCII spectrum, with a decreased absorption in the 450–500 nm regions (Figure 6C). This suggests a decrease in chlorophyll b content in the band, compatible with a loss of LHCII. The PSI* spectrum is also different from the PSI spectrum. In fact, the spectrum of the PSI* particles matches very well the simulation of the absorption of a PSI complex containing an additional Lhca1‐Lhca4 dimer (Figure 6D).

A SYPRO‐stained SDS‐PAGE of the complexes shows an increase in the Lhca1‐a4 band compared to Lhca3 in the PSI‐6LHCI‐LHCII, as expected (Crepin et al., 2020), as well as in the PSI* complex (Figure 7A–C). LHCII was detected in PSI* (Figure 7A,C), but the level was very low compared to LHCII in PSI‐6LHCI‐LHCII. We believe that the presence of LHCII in PSI* is due to a light contamination by small PSII supercomplexes migrating closely or by the near migrating PSI‐LHCII supercomplex (Figure 6A). It should be remembered that LHCII is a highly abundant complex that is present almost everywhere in a native gel since it is bound to several PSII complexes or subcomplexes, which can partly be disrupted during migration. Indeed, some LHCII is also detectable in the PSI band (which is known not to bind LHCII). Also, the PSI‐6LHCI‐LHCII complex shows a great amount of LHCII, much more than a trimer (which is the complex bound to this PSI complex), meaning that the all other LHCIIs are contaminants, in particular by the closely migrating PSII C2S complex (Crepin et al., 2020).

Figure 7.

LHC composition of the PSI* complex.

(A) Sypro PAGE of PSI*, PSI and PSI‐6LHCI‐LHCII eluted from the BN‐PAGE gels (the LHC region of the gel is shown).

(B) Densitometric profiles of the gel shown in A after normalization on the Lhca3 band.

(C) Densitometric profiles of the gel shown in A on the Lhca region. Identification of the bands is according to Crepin et al. (2020).

(D, E) Immunoblots of Lhca4 and Lhca1 in comparison with Lhca3 in PSI and PSI*. The densitometric profiles are shown after normalization on the Lhca3 signal.

In any case, considering the position of the PSI* in the native gel below the PSI‐LHCII (Figure 6A), which indicates the presence in PSI* of additional subunits of smaller size than an LHCII trimer, and a clear increase of the SDS‐PAGE band corresponding to Lhca1‐a4 (Figure 7C), we considered the presence of an additional Lhca1‐a4 dimer as the more likely possibility. To further investigate this point, we performed immunoblots against Lhca1, Lhca4 and Lhca3 (as reference) (Figure 7D,E), which confirmed the increased levels of both Lhca1 and Lhca4 in PSI* relative to Lhca3. The estimated increase in Lhca1 and Lhca4 is around 2 times (Figure 7C–E), as expected if the PSI* complex contains an additional Lhca1‐4 dimer.

Based on this result (the presence in PSI* of one additional Lhca1 and Lhca4 subunit when compared with PSI, which accounts for the entire difference of ~80 kDa between PSI* and PSI), the hypotheses that LHCII detected in the PSI* fraction (Figure 7A) is bound to PSI* would require the loss of PSI subunits for a total of 120 kDa (the MW of LHCII), which is very unlikely and never reported. This would imply indeed the loss of practically all small and very small PSI subunits or the loss of three Lhca complexes. Loss of Lhca has been reported only in Lhca antenna mutants (Wientjes et al., 2009). However, this is not the case for PSI* since Lhca2, Lhca3, and PsaD content does not change when compared with PSI, and Lhca1 and Lhca3 are doubled (Figure 7). The alternative hypothesis that the LHCII detected in the gel is bound to PSI* in a non‐trimeric form (which fits better with the amount that can be estimated in Figure 7B) is even more unlikely since, to maintain the estimated size of PSI*, this would require the concomitant binding of a monomeric (40 kDa) or a dimeric (80 kDa) LHCII to PSI* and the loss of some PSI subunits for the same MWs. However, monomeric or dimeric LHCII have never been shown to bind to stable photosynthetic complexes in Arabidopsis or other vascular plants.

To strengthen our conclusion, we analyzed the composition of the bands separated by CN‐PAGE by performing SDS‐PAGE in the second dimension (Figure S6). As expected, LHCII is clearly present in PSI‐6LHCI‐LHCII and in PSI‐LHCII (both in W and in W + FR conditions) and not in correspondence of PSI. PSI‐6LHCI‐LHCII, PSI‐LHCII, and PSII‐CS complexes are known to contain one LHCII trimer per complex. On the contrary, at the position of PSI*, the signal of LHCII was almost undetectable, supporting our conclusion that PSI* does not bind LHCII.

DISCUSSION

FR‐enriched light inhibits the accumulation of photosynthetic pigments during plant development

Arabidopsis thaliana is a shade‐avoiding species and, accordingly, we observed typical shade avoidance responses, such as hypocotyl and leaf petiole elongation, to a high FR/red light ratio simulating vegetation shade. However, long‐term exposure of Arabidopsis plants to FR‐enriched light not only induced morphological changes through phytochrome regulation but it also brought about changes in the properties of the photosynthetic machinery. Concerning photosynthetic pigments, the responses to changes in the red/FR light ratio are not consistent among plant species and among different varieties of a given species (Tan et al., 2022). As previously reported (Bout‐Torrent et al., 2015; Huber et al., 2024; Morelli et al., 2021), we observed a substantial decrease in photosynthetic pigment contents in Arabidopsis plants grown in W + FR compared to W‐grown plants, and we could attribute this effect to a block of pigment accumulation in developing seedlings. For the carotenoid pigments, this response was associated with a downregulation of several genes of the carotenoid biosynthesis pathway in FR‐enriched light. This phenomenon was disconnected from phytochrome signaling since the inhibition of hypocotyl elongation in the pif7 and ccd mutants in FR light showed similar changes in the carotenoid content as in WT. Moreover, based on the phenotype of the ccd and lox2 mutants, the expression levels of CCD and LOX2 genes and the quantification of apocarotenoids, we can exclude that the decreased content of carotenoids involves an enhancement of oxidative degradation of the carotenoid pigments. Both carotenoids and chlorophylls substantially decreased in W + FR, and this effect was quantitatively similar for both types of pigments (−30% on average), suggesting a coordination process, as discussed below.

Chronic oxidation of plastoquinone in far red‐enriched light

Since FR light is predominantly absorbed by PSI (Croce & van Amerongen, 2013; Gobets & van Grondelle, 2001), FR‐enriched light will inevitably cause an unbalanced absorption of light in favor of PSI. This is illustrated in this study by the enhanced oxidation level of the PQ pool. In fact, a faster oxidation of the PQ pool in FR light will reduce excitation pressure on the PSII complexes enabling them to use absorbed photons more efficiently. Importantly, PQ oxidation state is also a signal that triggers changes in the expression of photosynthesis‐related genes (Havaux, 2020b) and induces post‐transcriptional regulations (Frigero et al., 2007). A highly reduced PQ pool was found to decrease the abundance of antenna proteins, in particular the Lhca antennae (Frigero et al., 2007). The PQ pool is also known to play a role in carotenoid biosynthesis, both directly as a co‐factor of phytoene desaturase (Foudree et al., 2012) or indirectly by modulating the expression of carotenoid biosynthesis genes. A number of Arabidopsis carotenoid biosynthesis genes, including 3 β‐carotene hydroxylases, were upregulated when the PQ pool was over‐reduced while the opposite was found when the oxidation level of the PQ pool was enhanced (Kawabata & Takeda, 2014). Interestingly, a direct connection has also been proposed between the redox state of the PQ pool and the biosynthesis of chlorophyll (Steccanella et al., 2015). PQ might act as an electron donor/acceptor for the cyclase reaction that leads to the chlorophyll precursor Protochlorophyllide. Moreover, there is a tight coregulation of the chlorophyll and carotenoid biosyntheses to ensure balanced pigment production for the synthesis, assembly and maintenance of the photosynthetic apparatus, through the action of OR family proteins (Sun et al., 2023). Interestingly, in Scots pine plantlets, there was a concomitant downregulation of the PSY gene and the chlorophyll biosynthesis PORA gene in far‐red light compared to white light (Pashkovskiy, Ivanov, et al., 2023). Those observations indicate that the parallel decrease in carotenoids and chlorophylls in W + FR light could reflect this coordination process.

Acclimation to FR‐enriched light enhances photosynthetic efficiency

The decrease in pigment content likely reflects a general photosynthetic acclimation induced by FR light. When measured in the growth light conditions, the photochemical efficiency Φ_PSII of W + FR‐grown plants was increased compared to that of W‐grown plants. The efficiency of W + FR‐grown seedlings dropped when measured in white light, indicating an acclimation process that optimizes the photochemical use of W + FR light at the expense of the photosynthetic performance in W light. Although the W + FR‐induced increase in PSII/PSI was not very pronounced, it could cause a PSII acceptor side limitation when W + FR‐grown plants are placed in W light.

In Arabidopsis grown in fluctuating light with high PFD alternating with low PFD, addition of FR light was previously reported to improve the PSII photochemical quantum yield in the low light phase (Kono et al., 2020). In contrast, in lettuce, there was no change in the photosynthetic quantum yield during long‐term growth in FR‐enriched light (Zhen & Bugbee, 2020). Also in Cardamine hirsuta, adding FR light to W light during growth did not increase photosynthetic electron transport (Morelli et al., 2021). In rice, FR light supplementation affects photosynthetic traits very marginally (Huber et al., 2024). However, the instantaneous effect of supplemented FR on photosynthetic rate (Emerson effect) was larger in rice plants pre‐acclimated to W + FR compared to rice grown in W light. So, the effect of acclimation to W + FR on the photosynthetic efficiency appears to be species‐dependent. The spectral characteristics and intensity of the FR light used in the experiments could also modulate the plant responses (Zhen et al., 2019).

The light absorption capacities of PSII are enhanced during plant acclimation to far red‐enriched light

There are several mechanisms that are known to re‐equilibrate light absorption in the chloroplasts as a function of the light quantity or quality, such as changes in the PSII/PSI ratio and adjustment of the light absorption in PSII, which could also be involved in the response of plants to changes in the red/FR light ratio. For instance, in Pisum sativum plants transferred to “PSI light” (580–740 nm), the abundance of the PSI protein PsaA/B decreased while that of the PSII proteins D1 and D2 increased (Kim et al., 1993), resulting in a high PSII/PSI ratio (Chow et al., 1990). Our biochemical and spectroscopic analyses confirmed the decrease in PSI relative to PSII in Arabidopsis seedlings acclimated to W + FR.

Plant adaptation to low light intensity is known to involve an increase in the antenna size of PSII (Rochaix, 2014). Our biochemical analyses (Figures 5E and 6A) indicate that this does not seem to be the case for the adaptation to W + FR light conditions; no significant increase in accumulation of LHCII trimers was observed in seedlings grown in W + FR conditions. However, state transitions were perturbed in those plants reducing significantly the transition to state 2 and hence favoring the maintenance of the association of LHCII to PSII and thus maximizing light absorption in PSII. Light absorption can also be modulated by modifying the capacity of thermal energy dissipation by the NPQ process. Zhen and van Iersel (2017) reported a FR‐induced decrease in NPQ in lettuce leaves, but this was not found here in Arabidopsis seedlings grown in the presence of FR light.

A specific PSI supercomplex in plants acclimated to far red‐enriched light

The effects of far‐red light on PSI were not limited to a modification of the photosystem stoichiometry, but it also affected the organization of a part of this photosystem. The most striking result was the appearance of an unusual PSI complex that contains an additional Lhca1‐Lhca4 dimer, similarly as in the large PSI complexes described in several green algae (Ozawa et al., 2018; Qin et al., 2019; Suga et al., 2019). A very large PSI complex containing a supplementary Lhca dimer was previously reported in Arabidopsis plants (Crepin et al., 2020), but this supercomplex also contains, in close proximity of the additional Lhca1‐Lhca4 dimer, a LHCII trimer in the same position as in the well‐characterized PSI‐LHCII complex (Galka et al., 2012). We also observed this PSI‐6LHCI‐LHCII complex in our electrophoretic separation of thylakoid complexes from W and W + FR Arabidopsis plants at a position above the PSI* complex. In a recent study performed on the CAM plant species Tillandsia flabellate, Hu et al. (2023) also observed PSI supercomplexes larger than the canonical PSI (PSI with 4LHCI). Although the Lhca composition of those complexes was not analyzed in detail, the authors also suggested the presence of additional LHCI which could be similar to the PSI* here described for WT + FR‐grown Arabidopsis plants. Contrary to PSI* in W + FR Arabidopsis, the large PSI in Tillandsia appeared as a doublet. The absorption spectra of our PSI* complex (Figure 6C) and the new PSI supercomplexes reported by Hu et al. (2023) showed increased absorption in the 450–500 nm domain compared to the PSI‐4LHCI spectrum, in line with the presence of extra (chlorophyll b‐containing) LHCs. A detailed biochemical, spectroscopic and structural analysis of PSI* and the two large PSI complexes of Thillandia would be interesting to evaluate their similarity.

It is difficult to say and investigate what the physiological function of larger PSI supercomplexes is. We cannot create plants without PSI*. Indeed, knock‐out mutants of Lhca exhibited important rearrangements of PSI structure also on normal growth conditions (Wientjes et al., 2009) that would hamper any conclusion about the function of PSI*. Interestingly, the CAM PSI containing extra LHCIs was associated with a low PSI/PSII ratio, as it was the case in Arabidopsis grown in W + FR. So, large PSI complexes with extra Lhcas may be associated with conditions that promote a decrease in PSI activity vs. PSII. It is also possible that the increased antenna system by LHCIs that, differently from LHCII, are subunits only bound to PSI, plays a special role in low light. Bailey et al. (2001) have shown that Lhca1 and Lhca4 tend to increase with decreasing light intensity while it is the opposite for Lhca2 and Lhca3: the abundance of the latter antennae increases with increasing light intensity. Since a high FR‐to‐red light ratio simulates proximity shade conditions imposed by vegetation (Martinez‐Garcia & Rodriguez‐Concepcion, 2023), PSI complexes enriched in Lhca1‐Lhca4 could represent an acclimation to low light. An interesting finding in this direction is provided by Morelli et al. (2021) who showed that Arabidopsis seedlings grown in W + FR have a much better photoacclimation capacity to very low light than seedlings grown in W light; FR light somehow primes the photosynthetic system to subsequent low PFDs. Since the additional Lhca dimer can interact with LHCII (Crepin et al., 2020), it is also possible that it could be important for recovery from low light environments by allowing rapid docking of LHCII to PSI upon return to higher light intensities. The exact physiological function of the extra dimeric Lhca complexes would clearly deserve to be studied further in the future.

It should be noted that what is observed in a native gel does not forcedly represent the in vivo situation: if this PSI‐6LHCI clearly exists and is visible only in FR conditions, it could in reality represents the detergent‐resistant part of a larger complex that has some important biological function under these conditions. It is also possible that the PSI* complex might be important and much more abundant in some different species, such as in shade plants. This will require further investigations on plants from different ecological niches.

Note that we exclude the possibility that the PSI* complex is a sort of artifact due to a differential solubilization of W and W + FR thylakoids. First of all, to our knowledge, the PSI* has never been described before in works from different laboratories using Arabidopsis grown under many different conditions (and whose membranes were solubilized using different detergents). Second, our thylakoids were easily and completely solubilized indicating that there were no major differences with respect to the action of the detergent on the thylakoids from the two conditions, which is not the case for instance for comparison of thylakoids from plants grown under very contrasting conditions (low light vs. high light; well‐watered plants vs. drought conditions). Moreover, our biochemical analyses do not indicate significant changes in all other photosynthetic complexes (Figure 6). Third, in a previous work (Crepin et al., 2020), the PSI‐6LHCI complex was specifically and unsuccessfully searched using extremely mild solubilization conditions on different thylakoids (however, FR conditions were not tested at that time). In conclusion, we do believe that this PSI‐6LHCI complex is promoted by FR‐enriched light and could play a role in FR light acclimation.

Far red‐enriched light enhances plant growth

Photosynthetic acclimation to W + FR improved light use efficiency, providing more energy for plant growth. Increased concentrations of different sugars in leaves were previously reported in several plant species exposed to light with FR addition (Driesen et al., 2023; Lercari, 1982). FR light supplementation markedly increased Arabidopsis plant biomass, and this effect was observed for both root and shoot growth. Low red/FR light environments have also been reported to favor plant biomass productivity in several species such as tomato, lettuce, soybean and foxglove (Elkins & van Iersel, 2020; Tan et al., 2022; Yang et al., 2020; Zou et al., 2019). As far as eukaryotic microalgae are concerned, the effects of far‐red light addition to white light are poorly documented. Far‐red light supplementation appears to induce different effects from those observed in vascular plants, with a decrease in photosynthetic activity and chlorophyll concentration and an increase in growth rate and carotenoid content (Sanchez‐Saavedra et al., 1996). In macroalgae, far‐red light enrichment decreased photosynthesis and growth rate (Figueroa et al., 2003). Because of its selective absorption by PSI chlorophylls, pure FR light applied alone does not sustain efficient photosynthesis and causes noticeable alterations of the photosynthetic apparatus (Hu et al., 2021; Pashkovskiy, Khalilova, et al., 2023). In contrast, this work, together with some previous reports, shows that supplementing white light with FR light enhances photosynthetic efficiency and plant growth. These beneficial effects of far‐red light could question the notion of PAR restricted to the 400–700 nm wavelength range. It has been previously proposed that the definition of PAR should be extended to include FR photons and cover the 400–750 nm domain (Zhen et al., 2021). Our results are in line with this proposition since this extended PAR is better correlated with photosynthesis. On a practical point of view, the synergistic action of FR photons and PAR photons on plant photosynthesis and growth could be taken into account in horticultural light fixtures. Our results are also interesting in the context of the intercropping agricultural practice which consists in growing different crops close to each other (Li et al., 2023; Lithourgidis et al., 2011). Because PSII chlorophylls weakly absorb FR photons, the light energy reflected by the neighboring vegetation in high‐density multicrop conditions is enriched in FR light wavelengths (Cutolo et al., 2023; Martinez‐Garcia & Rodriguez‐Concepcion, 2023). The resulting increase in the FR/red ratio of the incident light, with limited reduction of total PFD, reaching plants in intercropped field can therefore have positive effects on photosynthesis.

EXPERIMENTAL PROCEDURES

Plant material and growth conditions

Wild‐type (WT) Arabidopsis thaliana (ecotype Col‐0) and the following mutant lines were used in this study: ccd single and double mutants, lox2 and pif7. ccd single mutants (SAIL_390_C01, SALK_097984C, max3‐9, max4‐1, defective in CCD1, CDD4, CDD7, and CCD8, respectively) were obtained from the NASC seed bank. The pif7 phytochrome mutant and the ccd1 ccd4 double mutant were provided by Dr. M. Rodriguez‐Concepcion (IBMCP, Spain). The seeds were surface sterilized by agitation with 70% ethanol containing 0.05% SDS for 2 min. They were then rinsed with 100% ethanol for 2 min and dried overnight before being resuspended in 500 μL of sterilized water for use.

Plants were grown under controlled conditions using the facilities of the Phytotec platform (BIAM, CEA/Cadarache). Light was delivered by arrays of white LEDs supplemented or not with FR LEDs (peaking at 730 nm), providing two different light growth conditions: white light (W) and FR‐enriched light (W + FR). SpectraPen mini (Photon Systems Instruments, PSI) was used to measure the light conditions. The spectra of the two light conditions are given in Figure S1. The PAR photon flux density (PFD) of W light was 140 μmol photons m⁻² s⁻¹ _, with a red (660 nm)/FR (730 nm) ratio of 4.8. The PAR PFD of W + FR was adjusted to the same value, but the red/FR ratio was much lower, 0.11. In nature, this value can be observed in vegetation shade conditions as defined by Martinez‐Garcia and Rodriguez‐Concepcion (2023). The irradiance of the FR light was 120 μmol m⁻² s⁻¹. It was measured with the PSI SpectraPen Mini by setting the wavelength range for photon detection between 700 and 750 nm. In the W condition, FR irradiance was ~4 μmol m⁻² s⁻¹. Consequently, the total PFD (PAR + 700‐750‐nm FR light) was higher in the W + FR condition (140 + 120 μmol m⁻² s⁻¹) compared to the W condition (140 + 4 μmol m⁻² s⁻¹).

For in vitro experiments, seeds were sown in Petri dishes on a solid medium containing 1/10 MS, 0.5% sucrose and 0.75% agar at pH 5.8. After stratification, the plates were incubated in a growth chamber under continuous W light for 3 days at 22°C. Then, the plates were either left in this condition or transferred to FR‐enriched white light (W + FR) for 4 days. Depending on the experiments, the Petri dishes were placed vertically or horizontally in the growth chambers.

Shoot and root samples were collected, immediately frozen in liquid nitrogen, and stored at −80°C until use for biochemical analyses. For biomass measurements, fresh mass was measured immediately after harvest, and the samples were then dried in an oven at 65°C for 3 days to measure dry mass.

For soil‐grown plant experiments, the pot size (two plants per pot) was 5.5 cm × 5.5 cm × 5 cm. Plants were grown for 18 days on potting soil (Seedlingsubstrat, Klasmann‐Deilmann) under controlled conditions (8 h/16 h, day/night; W light PFD, 140 μmol photons m⁻² s⁻¹; 24/20°C, day/night; relative air humidity, ca. 65%). Plants were then transferred in a growth chamber under either W or W + FR light with a photoperiod of 12 h for 7 days.

Photosynthetic pigments

Photosynthetic pigments were extracted from 15 mg of frozen shoots. Shoots were ground in 400 μL of chilled methanol. After centrifugation at 14000 g for 15 min at 4°C, the supernatants were collected and injected into the HPLC system (Waters, France). Pigments were separated using a Waters Nova‐Pak C18 Column (60 Å, 4 μm, 3.9 mm X 300 mm, 1/pk) by a gradient flow at 1.5 mL min⁻¹ with ethyl acetate as solvant A and acetonitrile/water/triethylamine (90/10/0.1, V/V/V) as solvant B. The gradient was as follows: 0 min, 100% B; 0–1 min, 100% B; 1–16 min, 0% B; 16–17 min, 100% B; 17–22 min, 100% B. The pigments were detected by a photodiode array detector 996 (Waters).

Data were collected and processed using the Empower software (Waters). For quantification, calibration curves were performed by using commercial pigment standards (carotenoid pigment standard purchased from Extrasynthèse, and chlorophyll a and b purchased from Sigma‐Aldrich).

Chloroplast density

Leaves of 7‐days old in vitro plants were harvested and fixed according to the following procedure. They were incubated in a freshly prepared solution of 0.05 M potassium phosphate buffer, pH 6.8 + 1% PFA (paraformaldehyde) for 10 min. The samples were washed four times with 0.05 M potassium phosphate buffer, pH 6.8. Afterwards, the samples were mounted between slide and coverslip in 0.05 M potassium phosphate buffer, pH 6.8 + 50% glycerol. Samples were stored at 4°C prior to confocal observation. Chloroplasts were visualized by observing chlorophyll autofluorescence using a LSM980 confocal microscope (Zeiss) with an excitation at 488 nm and a detection window set at 639–691 nm. Chloroplasts were counted in mesophyll cells and the amounts of chloroplasts were normalized to the measurement area.

RT‐qPCR

Total RNA was extracted from 50 mg of frozen shoot powder using the Direct‐zol RNA Miniprep Plus Kit (Zymoresearch) and treated with the DNase I (Zymoresearch) according to the manufacturer's instructions. Reverse transcription (RT) was performed on 400 ng total RNA using the qScript cDNA SuperMix (Quantabio). Real time quantitative reverse transcription PCR was performed using a LightCycler® 480 real‐time thermocycler (Roche) and the LightCycler® 480 SYBR Green I Master (Roche) with 10 μM primers and 0.1 μL of RT reaction product in a total volume of 5 μL per reaction. The specific primers used for PCR amplification are listed in Table S1. ROC3 (AT2G16600) was used as a reference gene for normalization.

UHPLC–MS analyses

Metabolites were extracted from 50 mg of frozen shoot powder with 250 μL of methanol supplemented with an internal standard (chloramphenicol, 5 mg/L) as described by Rodrigues et al. (2023). The extract was sonicated in an ultrasound bath for 15 min before centrifugation at 12000g at 10°C for 10 min. Supernatants were analyzed using a Vanquish Flex binary UHPLC system (Thermo Scientific) equipped with a diode array detector (DAD). Chromatographic separation was performed on an Acquity HSS T3 Column (100 X 2.1 mm, 1.8 μm particle size, 100 Å pores; Waters) maintained at 30 °C. The mobile phase consisted of acetonitrile/formic acid (0.1%, v/v) (eluant A) and water/formic acid (0.1%, v/v) (eluant B) at a flow rate of 0.33 mL/min. The gradient elution program was as follows: 0–1 min at 85% B; 1–4 min, 85–70% B; 4–5 min, 70–50% B; 5–6.5 min, 50–40% B; 6.5–8.0 min, 40–1% B; 8.0–10 min, 1%B isocratic; 10–11 min, 1–85% B. The injected volume of the sample was 1 μL. The liquid chromatography system was coupled to an Exploris 120 Q‐Orbitrap MS system (Thermo Scientific). The mass spectrometer operated with a heated electrospray ionization source in positive ion mode. The key parameters were as follows: spray voltage, +3.5 kV; sheath‐gas flow rate, 40 arbitrary units (arb. unit); auxiliary‐gas flow rate, 10 arb. unit; sweep‐gas flow rate, 1 arb. unit; capillary temperature, 360°C; auxiliary gas heater temperature, 300°C. The scan modes were full MS with a resolution of 60 000 fwhm (at m/z 200) and ddMS2 with a resolution of 60 000 fwhm; the normalized collision energy was 30 V; the scan range was m/z 85–1200. Continual internal mass calibration was operated using an EASY‐IC internal calibration source, allowing single mass calibration for the full mass range. Data acquisition and processing were carried out with Xcalibur 4.5 and Free Style 1.7 (Thermo Scientific), respectively. Apocarotenoids were identified based on detailed analysis of mass spectra and comparison with published data (Mi et al., 2018).

Chlorophyll fluorescence

In vivo chlorophyll fluorescence from Arabidopsis seedlings was measured with a PAM‐2000 fluorometer (Walz). The quantum yield of PSII (Φ_PSII) was calculated as: Φ_PSII = (F _m′–F _s)/F _m′, where F _m′ is the maximum fluorescence level measured in the light and F _s is steady‐state chlorophyll fluorescence in the light.

State transitions were monitored as described in Crepin and Caffarri (2015). State 1 was reached by illumination with red light (42 μmol photons m⁻² s⁻¹; 660 nm peak) supplemented with FR light (41 μmol photons m⁻² s⁻¹; 730 nm peak) for 15 min. Then, the FR light was turned off, allowing transition to state 2 by illumination for 10 min with red light only. The efficiency of state transition was quantified by the qT and qS parameters, as described in Ruban and Johnson (2009). qT is the decline of the maximal chlorophyll fluorescence level from state 1 to state 2 reflecting the decrease in the LHCII antenna size of PSII. qS indicates how state transitions can cope with the changing quality of light in optimization of the electron transport.

$$\mathit{qT}=\left(F_{\mathrm{m}}\mathrm{I}–F_{\mathrm{m}}\mathrm{II}\right)/F_{\mathrm{m}}\mathrm{I},$$

$$\mathit{qS}=\left(F_{\mathrm{s}}{\mathrm{I}}^{\prime }–F_{\mathrm{s}}{\mathrm{II}}^{\prime }\right)/\left(F_{\mathrm{s}}{\mathrm{I}}^{\prime }–F_{\mathrm{s}}\mathrm{II}\right).$$

F _mI and F _mII are the maximum fluorescence levels in state 1 and in state 2, respectively. FsI′ is the peak fluorescence level in state 1 after switching off the far‐red light. FsII and FsII′ are the steady‐state chlorophyll fluorescence levels in State 2 with and without far‐red light, respectively.

Non‐photochemical energy quenching (NPQ) was measured in red light (660 nm, 61 μmol photons m⁻¹ s⁻¹) or in white light (230 μmol photons m⁻¹ s⁻¹) combined or not with FR light (730 nm, 41 μmol photons m⁻¹ s⁻¹). NPQ was calculated as (F _m/F _m′)‐1, where F _m is the maximal chlorophyll fluorescence level in dark‐adapted leaves before illumination with the actinic red/white light.

Thylakoid preparations

Thylakoids membranes were prepared from 2 g of frozen shoots as described in Caffarri et al. ( 2009). Briefly, frozen shoots were ground in 50 mL of B1 solution (hypertonic solution to preserve chloroplast structure, 20 mM Tricine/KOH, pH 7.8 + 0.4 M sorbitol +5 mM MgCl₂ and the protease inhibitors 0.2 mM benzamidine, 1 mM E caproic acid, 10 mM NaF). After filtration with a nylon filter (mesh 30 micron) and centrifugation at 1400 g for 10 min at 4°C, the pellet was washed in 45 mL of B2 solution (isotonic solution, 20 mM Tricine/KOH, pH 7.8 + 0.15 M sorbitol +5 mM MgCl₂ and protease inhibitors as before). After centrifugation at 4000 g for 15 min at 4°C, the pellet was resuspend in 45 mL of B3 solution (hypotonic solution to burst chloroplasts, 20 mM HEPES/KOH, pH 7.5 + 15 mM NaCl +5 mM MgCl₂ + 10 mM NaF). This solution was centrifuged at 4000 for 15 min at 4°C. Finally, the pellet was resuspend in 500 μL of B4 solution (10 mM HEPES/KOH, pH 7.5 + 0.4 M sorbitol +15 mM NaCl +5 mM MgCl₂ + 10 mM NaF).

Chlorophyll concentration of thylakoid membranes was assessed as described in Chazaux et al. (2022). Thylakoid membranes were stored at −80°C until use. Low temperature fluorescence spectra of thylakoid membranes were measured using a Cary Eclipse fluorimeter (Varian). Samples were stacked thylakoids at 24 μg mL⁻¹ final chlorophyll concentration resuspended in 90% glycerol +10 mM HEPES/KOH, pH 7.5 + 5 mM MgCl₂. The measurements were done on frozen samples in liquid nitrogen with excitation at 440 nm and emission spectra recorded in the range 650–800 nm.

Blue‐native and clear‐native polyacrylamide gel electrophoresis (BN‐ and CN‐PAGE)

BN‐PAGE were realized according to Järvi et al. (2011) with some modification as described below. Thylakoid membranes (15 μg of chlorophylls) were washed in 200 μL of 10 mM BisTris/HCl, pH 7, then centrifuged at 12000g for 10 min at 2°C. The pellet was resuspended in 25 mM BisTris/HCl, pH 7 + 20% glycerol at a 1 mg.mL⁻¹ chlorophyll final concentration. The same volume of detergent solution was added to obtain final concentrations of 0.5 mg.mL⁻¹ of chlorophyll in 0.6% (w/v) digitonin and 0.2% (w/v) n‐dodecyl‐α‐d‐maltoside (α‐DDM). After 20 min of solubilization on ice in the dark, the solution was centrifuged at 12000 g for 10 min at 2°C, then Serva blue G buffer 10x (250 mM BisTris/HCl, pH 7 + 5% glycerol +5% Serva Blue G) was added to the supernatant at a final concentration of 1x. The solubilized complexes were loaded (15 μg of chlorophyll/per lane) on a homemade blue‐native polyacrylamide gels (BN‐PAGE) using a 3.30–12% acrylamide/Bis‐acrylamide (ratio of 29:1) gradient. The anode buffer was 50 mM BisTris/HCl, pH 7 and the cathode buffer was 50 mM tricine +15 mM BisTris/HCl, pH 7 + 0.01% Serva blue G. The migration was done overnight in the dark at 4°C at 25 V.

CN‐PAGE were prepared in a similar way with the exception that Serva blue G buffer 10x was not added to the sample and the cathode buffer was 50 mM tricine +15 mM BisTris/HCl, pH 7 + 0.1% Deriphat and 0.02% α‐DDM.

SDS‐PAGE and immunoblot analyses

Separation of photosynthetic proteins was done on SDS‐PAGE as described in Crepin et al. (2020). Laemmli system with a 14% acrylamide/Bis‐acrylamide (ratio of 29:1) + 2 M urea was used and gels were stained using SYPRO Ruby (Thermo Fisher Scientific). Densitometric analysis was performed with Gel‐Pro Analyzer. Immunoblot analyses were performed using specific antibodies from Agrisera (Lhca1, AS01‐0051, Lot 0512; Lhca3, AS01‐007, Lot 0601; Lhca4, AS01‐008, lot 0508), and revelation was done by fluorescence (secondary antibody conjugated with Alexa Fluor™ 790, Thermo Fisher) using a Fusion FX Spectra revelation system (Vilber).

The spectrum of a PSI complex containing an additional Lhca1‐a4 dimer (therefore a PSI‐6LHCI) was simulated using the spectrum of the purified PSI complex (PSI with 4 LHCI) added with the spectrum of the purified Lhca1‐a4 dimer (Wientjes & Croce, 2011) normalized to have the same stoichiometric amount.

Plastoquinone

Plastoquinone (PQ) was extracted from 30 mg of frozen shoots according to Ksas et al. (2022). Briefly, frozen shoots were ground in 700 μL of ethyl acetate. After centrifugation, the supernatant was evaporated under a stream of nitrogen. The residue was recovered in methanol/hexane (17/1, v/v) and analyzed by HPLC (Waters). The column was a Macherey‐Nagel Nucleosil 100–5 C18 (5 μm, 250 × 4 mm). The separation was done by an isocratic flow at 1.2 mL/min with methanol/hexane (17/1, v/v) as eluent. The PQ pool consists of reduced and oxidized forms of PQ‐9. Quantification was achieved using purified PQ‐9, which was received from J. Kruk (Krakow, Poland). Oxidized PQ‐9 was detected in UV absorbance at 255 nm, reduced plastoquinone was detected by fluorescence at 330 nm with excitation wavelength at 290 nm. To ensure that the reduction state of PQ reflects the native state in planta, samples were frozen with liquid nitrogen while being illuminated.

Conflict of Interest Statement

The authors declare no conflicts of interest.

Supporting information

Figure S1. Spectra of the W and W + FR light conditions.

Figure S2. Pigment content of 25 day‐old Arabidopsis plants grown on soil under W or W + FR light conditions.

Figure S3. Confocal fluorescence imaging of chloroplast density in leaves of W‐ and W + FR‐grown Arabidopsis plants. Chloroplasts are visualized by the fluorescence of their chlorophyll molecules.

Figure S4. Photosystem II photochemical efficiency (Φ_PSII) in leaves of 25 days‐old Arabidopsis plants grown on soil under W or W + FR light conditions.

Figure S5. NPQ in Arabidopsis leaves grown in W or in W + FR.

Figure S6. Two‐dimension electrophoresis of the region of PSI and PSII complexes. The first dimension is by native CN‐PAGE (as in Figure 6) and the second dimension is by SDS‐PAGE.

Table S1. Primers used in this study.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.