A red/blue optoswitch for temporal control of chloroplast transcription and biogenesis in Arabidopsis

TL;DR

Scientists created a light-controlled switch to restore chloroplast development in a mutant plant, revealing new insights into plant biology.

Contribution

A red/blue optoswitch is developed to control chloroplast transcription and biogenesis in Arabidopsis.

Findings

A blue-light-induced PAP7 construct rescues non-viable albino mutants by restoring PEP activity.

Chloroplast biogenesis is only inducible in very young leaf tissues.

Initial PEP formation and function are independent of photosynthesis.

Abstract

Photosynthesis genes in plant chloroplasts are transcribed by the plastid-encoded RNA polymerase called PEP. Consequently, PEP-deficient mutants cannot generate a photosynthetic apparatus and develop non-viable albino seedlings. Inducible complementation of such mutants thus could provide interesting insights in PEP action and chloroplast biogenesis. Here we show the effects of photo-inducible complementation in the albino Arabidopsis mutant pap7-1 using a red/blue optoswitch with monochromatic LEDs. Expression of a blue-light-induced PAP7 construct that is silent under red light reconstitutes PEP at any time point of pap7-1 development resulting in proper chloroplast biogenesis that rescues the non-viable mutant. Induction of chloroplast biogenesis, however, can only occur in very young leaf tissues indicating the existence of a cell-autonomous, biogenic coupling between cell and organelle development. We further uncover that initial PEP formation and function is independent of photosynthesis. The optoswitch termed blue-light valved biogenesis opens experimental avenues to study non-viable plant mutants.

Plants require the plastid encoded RNA polymerase PEP to generate viable photosynthetic apparatus. Here the authors genetically engineer an optoswitch that enables photo-induced complementation of an albino mutant restoring transcription and chloroplast biogenesis on demand.

Figures

d). All green leaves that developed after the shift to B_on_ were photosynthetically competent (Fig. [2d,e](#Fig2)). Thus, the activation capability of the optoswitch is independent of seedling age and not restricted to cotyledons, where highest PEP activity typically is found in wt plants^[8](#CR8),[9](#CR9)^. We further performed light cycling experiments with alternating R_off_ and B_on_ conditions in a 7-day-rhythm (Fig. [2f,g](#Fig2)). These cycles generated plants with white, green and mixed-coloured leaves within one and the same rosette. Mixed-coloured leaves often exhibited white blad

Developmental dependency of chloroplast biogenesis in BVB09.a Developmental timeline of albino BVB09 seedlings grown in red light for 0 to 7 days. Pictures were taken every 24 hours (upper row; scale bars indicate 250 µm). The depicted seedlings at the respective time points (given on top) were then shifted to blue light and grown for additional 10 days as given in left margin (lower row; scale bars indicate 500 µm). b Magnification of a BVB09 cotyledon (white box in a) grown for 3 days in red light prior to the shift to blue light. c Pictures of 7-day-old wt, BVB09 and *pap7-1* seedlings grow

PAP7 complementation induces photosynthesis-independent greening.Phenotypic analysis of greening and determination of photosynthetic capacity by Chl fluorescence imaging of wt and BVB09 seedlings after treatment with DCMU. Pictures were taken after 14 days of growth. Left margin indicates growth light regimes of respective lanes (14 d red, 14 d blue or shifted conditions with 7 d red followed by 7 d blue, all conditions were additionally indicated by red and blue perpendicular bars). DCMU concentrations sprayed on seedlings prior to the blue-light-shift are indicated. Columns wt and BVB09 disp

PAP7 induction enables PEP function independently of photosynthesis.a–i Comparative gene expression analysis in BVB09, *pap7-1* and wt grown under the indicated light conditions, including DCMU (10 µM) treated samples. Growth conditions are indicated by the coloured bars and defined in the inset on bottom. The respective analyzed gene is given on top of each diagram. a Relative expression of *GFP*, green fluorescent protein, b *PRIN2*, plastid redox-insensitive 2 corresponding to *PAP15 (*nuclear encoded), c *GUN1*, genomes uncoupled 1 (nuclear encoded), d *Lhcb1*, light harvesting complex of

Chloroplast biogenesis initiated by the BVB optoswitch.The blue-light inducible *PAP7* gene construct remains silent under either red light or in darkness (left margin). BVB seedlings grown for seven days (rows of arrow heads) develop either seedlings with white cotyledons (under red light) or etiolated seedlings indistinguishable from wt (in the dark), see symbols in the right margin. Blue light activates the *PAP7* gene construct resulting in the formation of the active PEP complex. Depending on the developmental stage of the growing seedling the blue light induction of PEP activates chlorop

Full text

Introduction

Chloroplasts are special organelles in cells of plants and algae where photosynthesis takes place. Their ability to harvest energy from sun light and to convert it into chemical energy is essential for the biosphere of planet Earth^1,2^. Vascular plants inherit chloroplasts to the next generation as undifferentiated proplastids in seeds. These are devoid of photosynthesis and need to build the apparatus after germination. In addition, if germination occurs underground plastids develop into non-photosynthetic etioplasts that rapidly develop into chloroplasts as soon as the seedling reaches light^3^. The initial establishment of photosynthesis in cotyledons requires the simultaneous generation of the thylakoid membrane system, the photosynthetic apparatus and the connected enzymatic pathways that ultimately produce the carbohydrates that drive plant energy metabolism^4^. Detailed knowledge of photosynthetic gene regulation is essential for engineering crop improvement and global food security^5,6^.

A large part of the photosynthetic apparatus is encoded by the plastid genome, the plastome^7^. These genes are transcribed mainly by the plastid-encoded RNA polymerase (PEP) that is comprised of five prokaryotic RPO core subunits encoded in the plastome and 16 PEP-associated proteins (PAPs) encoded in the nucleus. The RPO subunits are remnants of the evolutionary origin of plastids and trace back to the endosymbiosis of a cyanobacteria-like ancestor, while PAPs evolved later likely during the terrestrialization event^8,9^. Recently, 3D cryo-EM structures of the PEP complex have been reported which revealed the overall arrangement and the structural details of the PEP subunits^10–13^. In a nutshell, the RPO core largely retained the ancient bacterial crab claw architecture including all essential sites involved in transcription, while the PAPs are arranged in a belt around this core generating a starfish-like shape. The structural information has advanced our understanding of the PEP complex and provides hypotheses for potential PAP functions. It, however, does not provide proof for the functional roles of PAPs during chloroplast transcription arguing for additional experimental approaches using pap mutants of Arabidopsis^14^.

Homozygous inactivation mutants of PAPs all exhibit severe chlorotic phenotypes. Plastids in these mutants do not develop a thylakoid membrane system for photosynthesis^15–21^ and the seedlings, therefore, are non-viable without external carbon supply. PAPs thus are essential for both PEP function and the initiation of chloroplast biogenesis^22^. These pap mutants generally produce no or only few seeds and must be maintained in a heterozygous state that segregates in the off-spring^15,23,24^ (see Supplementary Fig. 1a for pap7-1) impeding the sampling of homozygous plant material and the application of classical methods used in physiology or genetics. To overcome these constraints we aimed to develop a switchable complementation system allowing us to restore the PEP complex on demand by selectively activating the expression of a given pap gene construct in its corresponding mutant background. This would allow growth of homozygous mutants in a complemented state until flower and seed production, but also to switch off the construct in the next generation in order to obtain non-complemented mutant off-spring that does not segregate anymore. This would provide (i) a source for homogeneous albino material for molecular studies of chloroplast biogenesis and (ii) a tool to study potential interactions of PEP with other processes.

Temporally controlled expression of a defined pap gene construct could be done with chemically induced systems; however, the chemical inductors can cause technical limitations due to low tissue diffusion rates, potential metabolic or toxic side effects or slow removal in shut-off switches^25,26^. Technological alternatives with high spatiotemporal resolution and precision could be optogenetic tools that have been already established in many variations for prokaryotic, fungal or animal model systems. In plants widespread application of optogenetics, however, is lagging behind because of the organisms´ energetic dependency on light, the lack of required co-factors, or the interference of endogenous photomorphogenic photoreceptors eventually causing imprecise or constitutive expression of optogenetic tools^27,28^. Recent studies reported the engineering of sophisticated actuator modules that overcome some of these issues and function even in white light. In the plant-useable light switch element (PULSE) system two combined photoswitches control a synthetic promoter that remains silent in blue light but becomes activated under red light. Because of the blue light portion of white light vegetative growth becomes possible and the gene cassette is expressed only in monochromatic red light^29^. Another system called Highlighter is comprised of a photoswitchable cyanobacterial CcaS-CcaR system that has been reingeneered for the use in plants. The designed CcaS actuator efficiently photoswitches with the plant endogenous phytochromobilin chromophore allowing for repression of target gene expression in blue light or blue light enriched white light while becoming active in other light regimes^26^.

In this study we aimed to establish a tool for photo-inducible complementation (PINC) of the non-viable albino pap mutants, which should provide precise, rapid and reversible activation of the complementation construct by a simple change in growth light colour of the seedlings. Because of the albino phenotype of the pap mutants the properties of the applied optoswitch should (i) allow for a light induced germination of seeds without activation of the complementation construct and (ii) enable a positive selection of complemented individuals upon light induction. Active expression or repression of the photo-induced gene cassette in white light, therefore, was not our prime interest. As regulatory elements we used the blue-light-responsive elements (BLREs) of the chalcone synthase promoter that are specifically recognized by the plant-internal cryptochrome system^30^. The basic idea of this approach was that the BLREs cloned in front of a complementation construct would activate the gene cassette by monochromatic blue-light illumination (B_on_) but would keep it silent in monochromatic red light (R_off_) (Fig. 1a), a condition that enables light induced germination. We expected to monitor the functionality of the system by the greening of complemented plants in B_on_ conditions since the expression of the complementation construct would restore the PEP complex that subsequently would initiate chloroplast biogenesis in the albino mutant background (Supplementary Fig. 1b). This optoswitch therefore was termed blue-light-valved biogenesis (BVB) (Fig. 1a).Fig. 1. Complementation of the pap7-1 mutant by the BVB optoswitch.a Scheme of AtCHS promoter and BLREs used for photo-inducible pap7 complementation constructs and basic principle of PINC. Blue light induces BVB-driven PAP7 expression in pap7-1, yielding green plants. In red light BVB is inactive, leading to albino pap7-1 mutants. In the dark the constructs are silent. cTP: chloroplast transit peptide; NLS: nuclear localisation signal; SET: methyltransferase domain; GFP: green fluorescent protein. b**–e** Phenotypes and relative PAP7 expression in 7- and 21-day-old plants grown under red (R_off_) or blue light (B_on_). Genotypes are indicated on top b, d or bottom c, e. Scale bars in b,d equal 500 μm. Symbols on left margin represent the expected phenotype induced by the optoswitch in the BVB lines. c, e Relative PAP7 expression was determined via qPCR. Expression levels from three independent biological replicates were normalized against EF1α. Error bars depict the standard deviation of the mean. Values give PAP7 expression in complemented lines relative to wt, numbers on top of B_on_ panels indicate degree of complementation. f 3-week-old plants (genotypes in left margin) grown on Petri dishes either in R_off_ or B_on_ conditions (indicated on top). Two left image columns: Plants appearance in standard white light. Two right image columns: Chlorophyll fluorescence imaging of the same plants given as F_v_/F_m_ values in false colour code ranging from black (0) to purple (1) (colour gradient bar on right bottom edge). g Quantification of F_v_/F_m_ values shown in f, numbers on top of B_on_ panel indicate degree of complementation. The first two true leaf pairs of 6 individual plants for each genotype were analysed. Error bars depict standard deviation of the mean and values give the photosynthetic capacity of the complemented lines relative to wt. h Chlorophyll and carotenoid content of plants shown in d, f. Genotypes and growth condition are given on bottom. The error bars give the standard deviation of the mean of three independent biological replicates. Pairwise comparisons between samples were initially evaluated using a two-tailed unpaired Student’s t-test. To control for multiple testing, p-values were subsequently adjusted using the Holm-Bonferroni method. Asterisks indicate the statistical significant differences in total Chl with **: p < 0.01. Statistics are shown only for comparisons between complementation lines and pap7-1 in R_off_ and complementation lines with wt in B_on_. BVB12 vs. pap7-1: p = 0.0036; BVB04 vs. wt: p = 0.0012.

Here we show the application of our strategy to the well characterized albino pap7-1 mutant of Arabidopsis thaliana^16,31^. We demonstrate the successful establishment and application of the BVB optoswitch in pap7-1 enabling us to control PEP formation by blue-light induced PAP7 complementation. We show that this can be done in a reversible and repeatable manner. By testing temporal requirements of the BVB optoswitch, we observe that initiation of chloroplast biogenesis must occur before a specific timepoint during leaf development. We further use the BVB optoswitch in pap7-1 to test whether PEP formation and activity depends on photosynthetic activity since several PEP subunits posses predicted domains with potential redox-related functions. The results indicate that PEP formation and principal function are largely independent from photosynthesis which contrasts some working hypotheses for redox-related PAP functions. Our results demonstrate high functionality of the optoswitch, providing an easy approach for spatio-temporal and qualitative analysis of PEP and its properties. The PINC approach using the red/blue optoswitch as tool further represents an attractive system for the study of other viable and non-viable plant mutants with a wide range of potential biotechnological applications.

Results

A red/blue optoswitch for controlled complementation of PEP

For the establishment of the PINC approach, we used temperature-controlled growth chambers equipped with narrow band-width light-emitting diodes (LEDs) providing high spectral selectivity (Supplementary Fig. 1c). Wild type (wt) seedlings grown under red or blue light developed normal green phenotypes both on agar plates or soil and displayed highly functional photosynthesis, while pap7-1 mutants remained albino under all illumination conditions (Supplementary Figs. 1d,e and 2a) as originally observed^16,31^. To establish the PINC approach we cloned one to four repeats of a 123 base pair (bp) DNA fragment containing the blue-light responsive ACE and MRE elements (BLREs) of the Arabidopsis thaliana chalcone synthase (AtCHS) promoter^30^ in front of a full-length PAP7 gene construct containing the N-terminal chloroplast transit peptide and a C-terminally fused GFP as reporter (Fig. 1a) and stably transformed these constructs into heterozygous pap7-1 mutants. Subsequently, we selected individuals in the off-spring that demonstrated an albino phenotype in R_off_, but a green phenotype in B_on_ and genotyped them. We then isolated lines being homozygous for both the mutant allele and the optogenetic construct. While transformation of constructs with a single BLRE cassette did not yield green plants (Supplementary Fig. 2b), experiments done with constructs containing two, three or four BLRE repeats were successful (Supplementary Fig. 2c, d). In B_on_, representative lines with two (BVB04), three (BVB09) and four (BVB12) BLRE repeats developed green cotyledons (Fig. 1b, Supplementary Fig. 2c), with PAP7 expression of approx. 70–90% of wt levels (Fig. 1c) and green follow-up leaves after 14 days of cultivation (Fig. 1d, Supplementary Fig. 2c). In older plants we observed a promoter dosage effect with increasing PAP7 construct expression the more BLRE repeats were present (Fig. 1e). We further tested the photosynthetic capacity (determined by chlorophyll (Chl) fluorescence parameter Fv/Fm) in BVB lines grown in either R_off_ or B_on_ conditions (Fig. 1f,g) and determined also the accumulation of photosynthetic pigments (Fig. 1h). In R_off_ conditions all lines were unable to become photoautotrophic while in B_on_ conditions the BVB lines developed green photosynthetic leaves. Again we observed a promoter-dosage effect with respect to development of green tissues, photosynthetic capacity and pigment accumulation (Fig. 1f–h). Constructs with four BLRE repeats performed best in Chl fluorescence measurements, but appeared to be slightly leaky under R_off_ since we detected trace amounts of chlorophylls in such lines (Fig. 1h, Supplementary Table 1). Although the Fv/Fm values did not indicate photosynthetic capacity of these plants, we decided to perform all subsequent analyses with lines carrying just three BLRE repeats (i.e. BVB09) since these exhibited complete tightness in R_off_. By this we aimed to avoid any ambiguities arising from an incomplete albinism in BVB12 that could potentially compromise our conclusions in further experiments. In turn, we regarded a complementation degree of around 80% as sufficiently complete to obtain conclusive results for the induction of PEP and chloroplast biogenesis.

Switchability and expression properties of the BVB optoswitch

Seeds of the BVB09 line were germinated and grown in R_off_ for 7, 14, or 21 days before being shifted to B_on_. Resulting plantlets produced white cotyledons as well as albino follow-up leaves if cultivated in R_off_ for long enough time. Upon shift to B_on_, all newly formed leaves of BVB09 plantlets became green regardless of pre-cultivation time in R_off_ indicating successful functional complementation at all time points (Fig. 2a-c). BVB09 plants exhibited retarded growth in R_off_ when compared to wt (Fig. 2d) which corresponds to the growth characteristics of the original pap7-1 line (Supplementary Fig. 1d). All green leaves that developed after the shift to B_on_ were photosynthetically competent (Fig. 2d,e). Thus, the activation capability of the optoswitch is independent of seedling age and not restricted to cotyledons, where highest PEP activity typically is found in wt plants^8,9^. We further performed light cycling experiments with alternating R_off_ and B_on_ conditions in a 7-day-rhythm (Fig. 2f,g). These cycles generated plants with white, green and mixed-coloured leaves within one and the same rosette. Mixed-coloured leaves often exhibited white blades with green tips and edges in variable degrees of expansion (e.g. leaves 3a and 3b in Fig. 2f). The BVB optoswitch thus produces green and white tissues on demand even in older stages of plant development. Its switchability was functional regardless of the starting condition, i.e. it worked both in R_off_-B_on_-R_off_-B_on_ and in B_on_-R_off_-B_on_-R_off_ cycles. Since the BVB09 seedlings grew faster in B_on,_ full synchronicity between light shifts and speed of leaf development was difficult to achieve and we, therefore, were unable to generate plants with perfectly alternating green or white leaves.Fig. 2. Switchability of BVB-controlled chloroplast biogenesis in pap7-1.a-c Blue-light induction of BVB09 for seven days after 7 a, 14 b and 21 days of pre-cultivation in red light c. d Photosynthetic capacity (given as F_v_/F_m_) of wt and BVB09 shifted to blue light after 21 days in red light detected by Chl fluorescence imaging. Fluorescence values are given in false colour. Colour code on bottom right. Note that BVB09 develops slowlier than wt in the red light phase. e Quantification of F_v_/F_m_ values shown in d. The first two true leave pairs of 6 individual plants for each genotype were analysed. The error bars depict the standard deviation of the mean and the values give the photosynthetic capacity of the complemented lines relative to wt. For BVB09, only leaves developed in blue light perform photosynthesis and only these were analyzed for quantification. g, h BVB09 grown in 7-day-repeats of red-blue-red-blue e and blue-red-blue-red f shifts. Pictures document the development of one representative plant for each condition. The orientation of plants may change between single photos. Each pair of leaves is labelled with a number, starting with the cotyledons (1a, 1b and so on) to follow the morphological changes during growth. h Dynamics of relative PAP7 expression during light shifts from R_off_ to B_on_ and back to R_off_. The timeline on top indicates harvesting time points within each light phase. Wt and BVB09 seedlings were grown for 7 days in R_off_ before the shifts. Expression levels were normalized to EF1α and represent the mean of three independent biological replicates, with error bars depicting standard deviation.

In order to understand the action kinetics of the BVB optoswitch we determined its expression dynamics in response to light shifts (Fig. 2h). BVB09 and wt seeds were grown for 7 days in R_off_ conditions and the PAP7 expression level was determined by qRT-PCR. Subsequently, the seedlings were shifted to B_on_ for 24 h followed by a shift back to R_off_ for another 48 h. Multiple samples were taken along this time course to determine activation and deactivation dynamics of the optoswitch. Induction of the construct raised PAP7 expression from zero to wt levels within 30 min and reached a largely stable maximum level after 4 h with more than 2 times the expression level of wt. The shut-off was rapid (reaching less than 10% of maximal level within 1 h and almost zero after 4 h). These data indicate high tightness and rapid activation and deactivation behaviour of the optoswitch. The formation of leaves with mixed green and white areas, therefore, are neither caused by low temporal dynamics or an intrinsic leakiness of the system, but must be caused by constraints in the developmental programmes of the cotyledons and follow-up leaves.

In sum, the results demonstrate that (i) the optoswitch can be activated independent of the age of the plant and (ii) that repeated activation/deactivation cycles are possible. However, only newly emerging tissues became green while older white tissues did not suggesting the existence of a specific cell age threshold for initiation of chloroplast biogenesis within the leaf development programme. In cells younger than this threshold (e.g. in the shoot apical meristem (SAM)), PEP induction can initiate chloroplast biogenesis, while in cells that have already passed it, an induction of PEP has no effect anymore and the cells remain white even in extended B_on_ conditions. This cell age threshold, thus, represents a point-of-no-return in leaf cell development with respect to chloroplast formation.

Temporal constraints of chloroplast biogenesis

We aimed to determine more precisely the observed cell age threshold by performing growth experminents with increasing pre-cultivation times in R_off_ before a switch to B_on_ (Fig. 3a). When the BVB09 line was pre-cultivated for four to seven days in R_off,_ the cotyledons remained albino after the shift to B_on_. They, however, could green after pre-cultivation of only one to two days in R_off_ with an intermediate state after three days, where cotyledons developed patchy green and white leaf areas (Fig. 3b, Supplementary Fig. 3). Follow-up leaves frequently developed green tips and edges in experiments with repetitive light shifts (Fig. 2e, leaves 3a and 3b; further examples in Supplementary Fig. 3). These patterns appear to be different from those observed in the cotyledons and are most likely generated by the shift to R_off_ during their development. Cells that developed in leaf primordia in B_on_ became green, while cells that developed after the switch to R_off_ became white and largely remained white even after a subsequent switch back to B_on_. This strongly suggests that the point-of-no-return acts cell-autonomously coupling the initiation of chloroplast formation to a specific juvenile age of individual cells rather than to the tissue or the plant.Fig. 3. Developmental dependency of chloroplast biogenesis in BVB09.a Developmental timeline of albino BVB09 seedlings grown in red light for 0 to 7 days. Pictures were taken every 24 hours (upper row; scale bars indicate 250 µm). The depicted seedlings at the respective time points (given on top) were then shifted to blue light and grown for additional 10 days as given in left margin (lower row; scale bars indicate 500 µm). b Magnification of a BVB09 cotyledon (white box in a) grown for 3 days in red light prior to the shift to blue light. c Pictures of 7-day-old wt, BVB09 and pap7-1 seedlings grown in dark (top row) or blue light (bottom row). d Relative PAP7 expression in seedlings shown in c. Three independent biological replicates were analyized, values were normalized against EF1α. Error bars represent standard deviation of the mean. e, f Developmental timelines of wt e and BVB09 f seedlings grown in darkness. Pictures were taken after 0 to 8 days in darkness (upper rows). Depicted seedlings at the respective time points were then shifted to blue light and pictures were taken after 7 and 14 additional days (lower rows, respectively). As additional control dark-grown BVB09 seedlings f were shifted to red light for 14 days to demonstrate the blue-light dependency of chloroplast biogenesis. Scale bars equal 500 µm. Growth programmes and genotypes are indicated in the margins.

Since chloroplast formation is integrated into the photomorphogenesis programme, we wondered what would happen if we perform the same experiment with a pre-cultivation in the dark instead of red light. The BVB09 line was morphologically indistinguishable from wt plants either in the dark or in blue light (Fig. 3c), but in contrast to wt it did not express PAP7 in the dark (Fig. 3d). Further, while wt seedlings did green regardless of the length of the dark period before the shift to B_on_ (Fig. 3e), BVB09 seedlings were unable to respond to B_on_ after three days pre-cultivation in the dark. Only follow-up leaves turned green when B_on_ conditions last for 14 days (Fig. 3f). The point-of-no-return thus is not light-dependent, suggesting a fixed check-point in cell development that is important for the acquisition of competence to initiate chloroplast biogenesis in the light.

As alternative hypothesis one could argue that the lack of chloroplast biogenesis in older non-greening cotyledon cells could be simply caused by a developmental block or down-regulation of BVB optoswitch activity at day three and later. In order to test this we performed control experiments in which we determined the gene expression of the PAP7 gene construct in BVB09 seedlings at days two, three, four and five after germination, before and two days after the shift to B_on_. We observed high induction of the construct at all these tested time points (Supplementary Fig. 3g,h). This result excludes the possibility of a time- or development-dependent deactivation of the BVB optoswitch at this critical time point and provides additional support for our hypothesis of a cell-age-dependent block of chloroplast formation in albino leaf cells (Supplementary Fig. 3).

PEP formation and initial activity do not depend on photosynthesis

The PEP subunits PAP6, PAP10, PAP13 and PAP15 were proposed to mediate photosynthetic redox signals towards chloroplast transcription, affecting both assembly and activity of the complex^18,32^. The BVB optoswitch provides a suitable tool to prove these proposals by combining the induction of PEP formation with a pre-treatment of seedlings with 3-(3´,4´-dichlorophenyl) 1, 1´-dimethyl urea (DCMU), a membrane-permeable, highly stable inhibitor of photosystem II that blocks photosynthetic electron transport^33^. We grew wt and BVB09 plants seven days in R_off_, applied the herbicide and put the seedlings for seven days to B_on_ (Fig. 4). Non-treated 14-day-old controls demonstrated high photosynthetic capacity in wt (both in R_off_ and B_on_) and BVB09 (only in B_on_) (Fig. 4, first two top rows). The R_off_-B_on_ shifted control plants demonstrated high Fv/Fm values in all wt leaves and in the B_on_-induced leaves of BVB09 (Fig. 4, third row), while R_off_–grown BVB09 lines remained photosynthetically incompetent. Herbicide treatment before the shift to B_on_, however, blocked photosynthesis in all plants (Fig. 4, bottom two rows). Despite the DCMU-treatment, the BVB09 line started to develop green follow-up leaves when put to B_on_ (Fig. 4, bottom two rows) suggesting that PEP is active in these tissues. To prove this we isolated these new green leaf tissues by micro-dissection. In blue-native polyacrylamide gel electrophoresis (BN-PAGE) and subsequent immuno-blot analyses we analyzed the formation of PEP and of major photosynthesis complexes in these greening tissues of the BVB09 lines and in wt as control (Fig. 5). RubisCO, the CO_2_-fixing enzyme of photosynthesis, photosystems I and II (PSI and PSII) (Fig. 5a) and PEP (Fig. 5b) accumulated to wt-levels in the B_on_-induced BVB09 tissues, but not or only in trace amounts in R_off_ conditions. The DCMU treatment did not hinder the observed PEP complex formation, rather it appeared to promote it, especially in BVB09 (Fig. 5b).Fig. 4PAP7 complementation induces photosynthesis-independent greening.Phenotypic analysis of greening and determination of photosynthetic capacity by Chl fluorescence imaging of wt and BVB09 seedlings after treatment with DCMU. Pictures were taken after 14 days of growth. Left margin indicates growth light regimes of respective lanes (14 d red, 14 d blue or shifted conditions with 7 d red followed by 7 d blue, all conditions were additionally indicated by red and blue perpendicular bars). DCMU concentrations sprayed on seedlings prior to the blue-light-shift are indicated. Columns wt and BVB09 display magnified representative single seedlings of wt and BVB09 lines, the column Petri WL displays Petri dishes with wt and BVB09 seedlings grown in parallel (indicated in right margin and separated by broken line) as they appear under white light. The column Fv/Fm displays the Chl fluorescence of the seedlings shown in column Petri WL. Note that BVB09 lines are devoid of photosynthesis after 14 days in red light; scale bars equal 500 µm.Fig. 5PAP7 induction enables PEP complex assembly independently of photosynthesis.a Coomassie-stained BN-PAGE of total protein extract from 14-day-old seedlings. Genotypes and treatments indicated on top are identical to Fig. 4. Ferritin was used as a running marker (M). Labelling of photosynthesis complexes followed established detection maps for Arabidopsis. b BN-PAGE immuno-blot (same samples as in a) using PAP8 and RPOB antisera (left margin), both detecting the assembled PEP-complex at ~ 1 MDa (indicated by arrows in right margin).

We further tested the transcriptional activity of this PEP complex, and also that of nuclear RNA polymerase (RNAP) II, in the newly formed green leaves of DCMU-treated BVB09 seedlings by determining the transcript accumulation of GFP (as marker for transgene expression) and selected nuclear and plastid encoded genes for plastid proteins (Fig. 6, Supplementary Fig. 4). Overall BVB09 seedlings grown 14 days in R_off_ exhibited very similar expression characteristics as the pap7-1 mutant. Growth of 14 days in B_on_ resulted in a strong promotion of most plastid and nuclear encoded genes in BVB09 reaching expression levels comparable to the wt control under the same condition. We could not find any inhibitory effect of the DCMU treatment on the accumulation of PEP- or RNAPII-transcribed genes in green BVB09 tissues and only minor effects on a minority of genes in WT seedlings. This is notably true for PAP15/PRIN2 (Fig. 6b), which is proposed to mediate PEP formation in response to photosynthetic redox signals^32^, but also for GUN1 (Fig. 6c) encoding a component proposed to be central in retrograde signalling during chloroplast biogenesis^34,35^. However, we observed a slight promotion of rpoB transcripts in BVB09 (Fig. 6e) that is consistent with the PEP immunoblot data (Fig. 5b). The rpo genes are transcribed by a second phage-type nuclear encoded RNA polymerase (NEP)^36^, suggesting a potential connection between photosynthesis and NEP activity. In the case of the rbcL gene (encoding the large RubisCO subunit) we could not observe any DCMU effect on the transcript accumulation, however in the BN-PAGE analyses DCMU negatively affected the accumulation of the full RubisCO complex suggesting a selective inhibitory effect on protein translation (compare Fig. 5a and Fig. 6g). These results were confirmed by testing the expression of additional genes including Lhcb2.2, Lhcb2.3, accD, petD, ycf2.1, atpE, clpP, psaJ, psbB and psbN (Supplementary Fig. 4). Besides the gene specific expression control, we observed that transcript accumulation of plastid genes in BVB09 under R_off_ conditions equals that of the pap7-1 mutant indicating the tightness of the BVB optoswitch. The results further demonstrate an increase in PEP-dependent gene expression in B_on_ conditions (including an increase after a shift) confirming the data obtained for PAP7 expression under the same conditions (compare Figs. 2, 3). In sum, the data indicate that control of PAP7 expression by the BVB optoswitch results in a corresponding regulation of PEP-dependent transcripts and that inhibition of photosynthetic electron flow has no negative effects on PEP formation and its gene expression activity.Fig. 6PAP7 induction enables PEP function independently of photosynthesis.a**–i** Comparative gene expression analysis in BVB09, pap7-1 and wt grown under the indicated light conditions, including DCMU (10 µM) treated samples. Growth conditions are indicated by the coloured bars and defined in the inset on bottom. The respective analyzed gene is given on top of each diagram. a Relative expression of GFP, green fluorescent protein, b PRIN2, plastid redox-insensitive 2 corresponding to *PAP15 (*nuclear encoded), c GUN1, genomes uncoupled 1 (nuclear encoded), d Lhcb1, light harvesting complex of PSII protein 1 (nuclear encoded), e rpoB, β-subunit of PEP core (plastid-encoded, transcribed by NEP), f atpB, β-subunit of plastid ATP synthase (plastid encoded, transcribed by NEP and PEP), g rbcL, large subunit of RubisCO (plastid encoded, transcribed by PEP), h psaB, core protein B of PSI (plastid encoded, transcribed by PEP), i psbA, D1 protein of PSII (plastid encoded, transcribed by PEP). Additional genes tested are given in Supplementary Fig. 4. Relative expression in three independent biological replicates was determined by qRT-PCR and normalized to EF1α expression within each sample. Error bars represent standard deviations of the mean. Pairwise comparisons between samples were initially evaluated using a two-tailed unpaired Student’s t-test. To control for multiple testing, p-values were subsequently adjusted using the Holm-Bonferroni method. The asterisks indicate statistical significant differences with n.s. being not significant, **: p < 0.01 (a p = 0.0083, b p = 0.0092, c p = 0.0075, e p = 0.0053, i p = 0.0072) and ***: p < 0.001 (e p = 0.000015, f p = 0.000018, g p = 0.000033 and p = 0.000608, h p = 0.000568).

Discussion

Molecular, genetic and physiological analyses of pap mutants are an experimental challenge because of their albinism, which generates many handling problems. This includes the sampling of larger amounts of homozygous tissues, the propagation of mutant lines as well as breeding approaches. The BVB optoswitch developed in this study represents a solution for all these problems. In particular, we now can produce albino material on demand and additionally can maintain homozygous pap mutants in a complemented green stage until the flowering stage, allowing us to use established tools for genetic manipulation by floral dip and to collect homozygous seed material that does not segregate. This substantially facilitates the investigation of PAPs and PEP mutants. In addition, the BVB optoswitch allows several rounds of rapid on/off switches, which is an interesting option for detailed developmental and physiological experiments. Repetitions of on/off switches with high frequency are difficult to achieve with chemically inducible systems, because the chemical inductor once applied cannot be easily removed from the tissues and the shut-off dynamic depends largely on its degradation kinetics. Our BVB optoswitch revealed high and rapid expression changes of the gene construct both under promoting and repressing conditions (Fig. 2h) that are comparable to or even faster than chemically induced systems, especially in the shut-off condition. Further, the on- and off-states of the system were complete and did not exhibit any leakiness, neither at the level of PAP7 expression (Figs. 1, 2, 3) nor at the downstream level of plastid transcript accumulation (Fig. 6).

Most importantly, our study identified a point-of-no-return for chloroplast biogenesis in BVB09, indicating a strong coupling between the developmental stage of proliferating cells and their competence to generate chloroplasts (Fig. 7). This corresponds well to the complex differentiation processes of chloroplasts that have been reported for the SAM, the leaf primordia^37^ and during leaf blade development^38^ and explains the patchy cotyledons and follow-up leaves observed in our light shift experiments (Figs. 2, 3, Supplementary Fig. 3). Once this critical check-point in cell development is passed, a subsequent induction of chloroplast biogenesis by PEP is apparently blocked in BVB09 leaves, indicating that effective induction of chloroplast formation can occur only in a relatively short phase of approx. 72 h after the initiation of leaf formation. This suggests the existence of a pre-determined developmental cascade that couples cell age with chloroplast development which likely become desynchronized if PAP7 is not expressed in early stages of cell development. This scenario is in agreement with recent studies analysing the chloroplast biogenesis in cells along the regular developmental gradient in monocot leaves^39^. Current models assume a predominant role of NEP in early plastid development of young (likely meristematic) cells while PEP becomes predominant in later stages of chloroplast formation when leaf development has already progressed^40^. Our data suggest that PEP activity is either earlier required than anticipated so far or an additional, yet unknown factor/influence is involved. It is fully open whether this factor/influence represents a proactive regulator or an indirect impact by metabolites such as sugars or lipids. In sum, understanding of the transition between both transcription systems during early chloroplast biogenesis is still poor^41^ and will require more detailed investigations. The BVB optoswitch provides an useful tool for this experimental avenue.Fig. 7. Chloroplast biogenesis initiated by the BVB optoswitch.The blue-light inducible PAP7 gene construct remains silent under either red light or in darkness (left margin). BVB seedlings grown for seven days (rows of arrow heads) develop either seedlings with white cotyledons (under red light) or etiolated seedlings indistinguishable from wt (in the dark), see symbols in the right margin. Blue light activates the PAP7 gene construct resulting in the formation of the active PEP complex. Depending on the developmental stage of the growing seedling the blue light induction of PEP activates chloroplast biogenesis in a differential way. If PEP activation occurs within the first three days of cell development, seedlings do develop green cotyledons and follow-up leaves. If PEP activation occurs later than this point, cotyledons remain albino and only follow-up leaves become green. This accounts for light- and dark-grown plants indicating a fixed developmental point-of-no-return for chloroplast biogenesis. On and off switches are repeatable (not shown).

Our data further show that the induction of chloroplast biogenesis occurs cell-autonomously resulting in patchy leaf tissues (Figs. 2, 3; Supplementary Fig. 3). Interestingly, this provides a means to identify still proliferating cells in mostly mature tissues allowing to follow the leaf blade development in cotyledons of dicot plants. The identified time range for successful induction by the optoswitch corresponds with that for the action of retrograde biogenic signals from plastids described in wheat or pea^39,42^. The physical nature and mechanistic explanation of this point-of-no-return remains however unknown yet. The expression of GUN1, encoding a protein proposed to play a major role in biogenic retrograde signalling^43,44^, was not disturbed in BVB09 (Fig. 6), suggesting that it is not a problem of GUN1-dependent plastid-nucleus communication. Interestingly, the point-of-no-return was effective also in the dark, indicating that it is not an intrinsic hub of photomorphogenesis. The PEP subunit PAP7, like PAP8, PAP5 and PAP12, is dually located to plastids and nucleus^9,23,45^ and may have an important, yet unknown, function in the nuclear-controlled cell development programme^46^, which may diverge from the well-established photomorphogenic signalling pathways^47^. It is, however, potentially connected to the GLK and MYB-mediated signalling pathways controlling chloroplast biogenesis^48,49^, since pap7-1 mutants exhibit reduced GLK1 expression^50^. More detailed studies of the functional involvements of PAP7 may help to understand these processes in the future.

We further investigated a potential influence of redox signals from photosynthetic electron transport (PET) on PEP build-up and chloroplast biogenesis by DCMU treatment of the plants before a shift to B_on_ conditions. We observed proper greening in newly formed leaves (Fig. 4) corresponding to studies reporting no general inhibition of chloroplast biogenesis by DCMU^51,52^. More importantly neither the formation of PEP (Fig. 5) nor its general transcription activity (Fig. 6) were prevented by the drug. We observed only some minor gene-specific effects that, however, had no impact on chloroplast biogenesis. Interestingly, we observed an enhanced formation of PEP complexes upon DCMU treatment (Fig. 6). The increased transcript accumulation of GFP and rpoB under these conditions (Fig. 6a, e) suggest that the strong inhibition of all newly emerging photosystem II complexes by the DCMU treatment mimics a metabolic or developmental plastid stage that is similar to an initial step of chloroplast biogenesis when NEP promotes PEP accumulation. The effect might be induced indirectly, e.g. by lack of metabolites such as sugars or lipids. Further, it is possible that this effect is transient. The plastids in BVB09 are less well developed than those in wt, because their development started with 7 days delay (after the shift to B_on_). Possibly NEP is more active in BVB09 than in wt in that moment. However, further studies will be necessary to better understand this feedback effect. Proposed redox-related functions of PAP6, PAP10, PAP13, and PAP15, thus are either involved in processes other than transcription or functional under different conditions. For instance, regulation of PEP transcription by redox signals from the plastoquinone pool^53^ are not excluded by our results since the experimental design of our study does not reflect these specific conditions. Apart from its specific physiological aspect, the inhibitor experiments demonstrate the general suitability of the BVB optoswitch for the investigation of environmental influences on chloroplast biogenesis and function. This can include fluctuations in light intensity or temperature, or the influence of endogenous factors like plant hormones that promote chloroplast biogenesis, i.e., cytokinin^54^. The optoswitch thus provides a versatile test system to study (i) the coupling of cell-organelle-development and (ii) the influence of endogenous and exogenous signals on chloroplast biogenesis.

At the biotechnological level, our PINC approach represents a new powerful strategy for inducible complementation of inactivation mutants using a light-regime switch from monochromatic red light (which still supports normal seed germination) to monochromatic blue light, which activates transgene expression via the endogenous cryptochrome pathway. Unlike other optogenetic systems that rely on synthetic repressor or activator modules^55^, our approach harnesses plant-specific, blue-light-responsive cis-elements, with adjustable copy number to optimize induction efficiency. This renders the system simple to apply since no genetic adaptation of modules or provision of additional chromophores are required. The PINC strategy using the BVB optoswitch resembles the previously reported infrared-laser-induced local gene expression system driven by an engineered heat-shock promoter in wt Arabidopsis roots^56^, but offers an accessible alternative that does not require advanced technical Laser equipment and is specifically designed to enable selective complementation of Arabidopsis mutants in standard growth cabinets equipped with narrow band LED light sources. Plants permanently grown under such lights generated well-developed rossettes, flowered and produced viable seeds in amounts comparable to plants grown in white light of equal intensity indicating that the plants are energetically not limited in the red or blue light conditions. The only detectable morphological effect was an elongation of petioles in the red-light condition, a result of absent cryptochrome and phototropin photoreceptors signalling due to the lack of blue light. The application range of the PINC approach is not restricted to albino pap mutants as studied here, but in principle can be used with any Arabidopsis mutant exhibiting either a visible or molecularly trackable phenotype—provided the mutant phenotype is not altered by the light quality shift. If this is the case, a specific PINC tool as presented here by the BVB lines can be constructed. It is important to note that all blue-light-induced BVB lines can be maintained under white daylight illumination since the spectrum of white daylight includes the necessary blue light wavelengths to induce the complementation construct, even though it might be necessary to supplement the white light source with additional blue light due to the observed dosage effects.

The BVB optoswitch presented here differs in experimental goal and working principle from other optogenetic systems reported recently in plants^26,29^ as it is not aimed to keep the complementation construct silent in white light. It rather provides a tool to rescue and characterise mutants that otherwise are not genetically or physiologically accessible. Given the ubiquituous presence of cryptochromes across plant species, the approach is also likely to function in other plant models. It, therefore, offers a valuable addition to the already existing optogenetic toolbox in plant science since it enables precise and selective control of gene expression in plants—particularly in inactivation mutants—expanding the technological potential for molecular plant research^28^.

Methods

Plant material

Arabidopsis thaliana ecotype Columbia-0 (Col-0) was used as the wild type control in this study. The well-characterized PAP7/pTAC14 (AT4g20130) inactivation mutant line SAIL_566_F_06 served as the genetic background for the blue light-inducible complementation lines.

Plasmid construction

All cloned plasmids used in this study are listed in the Supplemental material (Table S2). Constructs were PCR-amplified from Arabidopsis thaliana Col-0 cDNA or genomic DNA using primers synthesized by Eurofins Genomics (listed in Supplemental Data). PCR was carried out using Phusion DNA Polymerase (New England Biolabs) and purified with the NucleoSpin PCR Clean-up Kit (Macherey-Nagel). Golden Gate modular cloning^57^ was used for assembly, employing the type IIS restriction enzymes BpiI and BsaI (New England Biolabs), following the Cut-Ligation protocol as described^58^. Ligation products were transformed into chemically competent E. coli (DH5α strain). Resulting colonies were screened via PCR using Taq DNA Polymerase (ThermoFisher Scientific), and positive clones were subjected to plasmid purification using the GeneJET Plasmid Miniprep Kit (ThermoFisher Scientific). All constructs were confirmed by Sanger sequencing (SeqLab).

Growth and Illumination Conditions

Seeds were surface sterilized by sequential washes in 70% ethanol (twice), followed by a final wash in 90% ethanol. Sterilized seeds were sown on ½ Murashige and Skoog (MS) medium (Duchefa), supplemented with 1% (w/v) sucrose (Roth) and 0.8% (w/v) phytoagar (Duchefa), and stratified for 2 days at 4 °C. Seedlings were grown under controlled conditions in LED-equipped growth chambers (Polyklima) under a 10/14 h light/dark photoperiod at 22/18 °C day/night temperatures with a photon flux density of 100 μmol m⁻² s⁻¹. For experiments requiring constant illumination, plants were grown under the same spectral and intensity conditions at a constant 22 °C. For seed production, plants were transferred to soil after a minimum of two weeks on plates and grown under a 16/8 h light/dark cycle with 100 μmol m⁻²s⁻¹ light intensity. The spectral characteristics of the LED panels (measured with LI-180; LI-COR) are provided in Supplementary Fig. 1.

Generation of stably transformed mutant lines

Heterozygous A. thaliana SAIL_566_F_06 mutants were grown for 5 weeks under white light. The presence of the heterozygous T-DNA insertion was confirmed by PCR-based genotyping using Edwards buffer ^59^ and specific primers (Supplemental Data). Plants were then shifted to long-day conditions, and immature flower buds were transformed using the floral dip method with Agrobacterium tumefaciens strain GV3101 pMP90, following a previously described protocol^60^. T₁ transformants were selected by growth on ½ MS medium supplemented with 50 µg mL⁻¹ kanamycin for 7 days under red light, followed by continued growth in blue light. After 4 weeks, all green seedlings were genotyped to confirm both construct insertion and homozygosity for pap7 gene inactivation. T₂ seeds were screened by growing seedlings in red light for 7 days, during which albino seedlings were marked. Plates were then shifted to blue light, and after another 7 days, the previously marked seedlings were assessed for greening. All selected transformants were genotyped, and lines showing a 3:1 segregation ratio (green:albino) were retained. Lines that were not homozygous for the pap7 gene in T₂ -evident by the presence of green seedlings already under red light- underwent an additional selection round in the T₃ generation. This process was repeated until homozygosity for both T-DNA insertions (construct and pap7 inactivation) was confirmed. These verified lines were used in all further experiments (see Supplementary Fig. 2).

RNA isolation and quantitative reverse transcription PCR

Samples were harvested, immediately frozen in liquid nitrogen, and ground to a fine powder. Total RNA was extracted using the TRIzol method (Invitrogen), following the manufacturer’s protocol. RNA purity was assessed by measuring absorbance at 260 nm and 280 nm, with a ratio of ~2.0 considered indicative of pure RNA. RNA concentration was calculated accordingly, and 1 µg of total RNA was treated with DNase I (ThermoFisher Scientific) to eliminate genomic DNA contamination. cDNA synthesis was performed using MMLV Reverse Transcriptase (New England Biolabs) and a mixture of random nonamer primers and oligo(dT)18-primers (Eurofins Genomics). The resulting cDNA samples were diluted 1:10 with nuclease-free water before quantitative PCR. Gene expression levels were measured using SYBR Green Master Mix (ThermoFisher Scientific) and gene-specific primer pairs on a StepOnePlus Real-Time PCR System (Applied Biosystems). Each reaction was performed in technical duplicate. For quantification, serial dilutions (1:4) of a cDNA mass standard were included for each gene and run in parallel with the samples. EF1α (ATG07940) served as the internal reference gene for normalization of expression levels. All used primer pairs are listed in Supplementary Information (Supplementary Table 2).

Chlorophyll extraction and quantification

Fresh plant material (30–90 mg) was used for chlorophyll extraction in 95% ethanol (1 ml per 50 mg fresh weight). Samples were ground thoroughly and centrifuged for 15 min at 4 °C at 10,000 × g to remove debris. Pigment concentrations were measured using a dual-wavelength spectrophotometer by recording absorbance at 664 nm, 649 nm, and 470 nm. Chlorophyll a, b, and total carotenoid contents were calculated according to established formulas as described previously^61^.

DCMU treatment

3-(3´, 4´-dichlorophenyl)-1, 1´-dimethylurea (DCMU) (Sigma-Aldrich) was applied after 7 days of seedling growth under red light. Seedlings were sprayed with 10 µM or 40 µM DCMU, diluted in water from a 10 mM stock solution prepared in 50% ethanol. Control plants were sprayed with an equivalent concentration of ethanol in water. Following treatment, all seedlings were transferred to blue light and grown for an additional 7 days prior to further analysis.

Chlorophyll fluorescence imaging

Photosynthetic capacity was assessed using the MAXI version of the Imaging-PAM fluorometer (Walz). Prior to measurement, seedlings were dark-adapted for 30 min to ensure accurate determination of photosystem II efficiency. Fluorescence parameters were recorded and analyzed using the ImagingWinGigE software (Walz).

BN-PAGE and western-immuno blot analysis

The assembly of the PEP complex was analyzed by Blue Native PAGE (BN-PAGE) followed by immunodetection according to previously described protocols^62^ with minor modifications^63^. Approximately 50 mg of Arabidopsis tissue, grown under the indicated conditions, was harvested and ground in liquid nitrogen. The resulting powder was extracted with three volumes of protein extraction buffer (100 mM Tris-HCl pH 7.2, 10 mM MgCl₂, 25% (v/v) glycerol, 1% (v/v) Triton X-100, 10 mM NaF, 5 mM β-mercaptoethanol, and 1× cOmplete™ EDTA-free protease inhibitor cocktail (Roche). Protein extracts were adjusted to contain 1% (w/v) β-D-dodecyl maltoside (DDM), 50 mM aminocaproic acid (ACA, pH 7.0), 25 mM BisTris (pH 7.0), and 1 µg/ml recombinant DNase I (Macherey-Nagel), and incubated for 1 h at room temperature. Samples were then supplemented with 0.25% (w/v) Coomassie Brilliant Blue G-250 (CBB-G250) and 20% (v/v) glycerol, followed by centrifugation at 16,000 × g for 10 min at 4°C. 25 µl of the resulting supernatant were loaded onto 4-16% acrylamide BisTris/ACA gradient gels (49% T, 3% C) with a 3% stacking gel. Electrophoresis was carried out using 50 mM Tricine/15 mM BisTris pH 7.0 with 0.01% (w/v) CBB-G250 as the cathode buffer and 50 mM BisTris-HCl pH 7.0 as the anode buffer, at 4 °C. Gels were run initially at 40 V for 4 h, then continued overnight at 30 V after replacing the dark blue cathode buffer with light blue cathode buffer (same composition but 0.005% CBB-G250), and finally switched to clear cathode buffer (no dye) for a final run at 50 V until the dye front exited the gel. Molecular weight markers were included by loading 10 µl of 200-fold diluted horse spleen ferritin (Sigma-Aldrich, F4503), which was later stained using Coomassie Brilliant Blue R-250. For immunodetection, proteins were transferred to PVDF membranes (Roti®-PVDF, 0.45 µm, Carl Roth) over 3 h at 2 mA cm⁻² using transfer buffer (0.5% SDS, 25 mM Tris-HCl pH 7.6, 192 mM glycine, 10% ethanol). Membranes were destained in 100% methanol, blocked in 5% (w/v) skim milk powder in TTBS (50 mM Tris pH 7.5, 150 mM NaCl, 0.1% (v/v) Tween 20), and incubated with either anti-RpoB antibody (PhytoAB, PHY1700, 1:10,000) or custom-made anti-PAP8 antibody (1:10,000) for 2.5 h. Membranes were washed 5 times in TTBS, incubated with HRP-conjugated Goat Anti-Mouse IgG (Agrisera, 1:25,000 in 15 ml) for 1.5 h, and washed again 5 times in TTBS. Chemiluminescence detection was performed using the SuperSignal™ West Atto Ultimate Sensitivity Substrate (Thermo Fisher), and images were acquired with a ChemoStar system (iNTAS) in sequence mode, recording 100 images of 5–10 s depending on signal intensity.

Statistics

Quantitative data given in the figures were subjected to statistical analyses wherever required for comparison. Sample number per experiment is indicated in the corresponding figure or figure legend. Plotting and statistical analysis were performed with Python using SciPy (v1.11.2) or Excel. Pairwise comparisons between relevant samples were initially evaluated using a two-tailed unpaired Student’s t-test. To control for multiple testing, p-values were subsequently adjusted using the Holm-Bonferroni method. Comparisons with adjusted p-values below 0.05 were considered statistically significant. Significance between groups is indicated in the figures with connecting bars and denoted as *p < 0.05, **p < 0.01, and ***p < 0.001.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Supplementary Information Reporting Summary Transparent Peer Review file

Source data

Source Data