Bioinformatic analyses

S1 domain sequences have been identified using InterPro online tool (https://www.ebi.ac.uk/interpro/). Multiple sequence alignment was performed with MUSCLE online tool (https://www.ebi.ac.uk/Tools/msa/muscle/) and represented as phylogenetic tree employing PhyML (https://toolkit.tuebingen.mpg.de/tools/phyml) and iTOL (https://itol.embl.de/).

Plant material and growth conditions

The PRPS1/prps1-2 heterozygous mutant lines were generated by targeting the first exon of PRPS1 locus in Arabidopsis thaliana wild-type (Col-0) genetic background, using the pDe-CAS9 vector earlier described (guide RNA sequence is listed in S1 Table) [40]. PRPS1/prps1-2 heterozygous plants, devoid of CAS9 T-DNA, were selected based on the mutation in PRPS1 sequence. prps1-2 pPRPS1::PRPS1 complemented lines were obtained by introgressing PRPS1 genomic locus, including the pPRPS1 promoter region, in PRPS1/prps1-2 heterozygous plants and by selecting prps1-2 viable plants carrying the pPRPS1::PRPS1 construct. oePRPS1 and oePRPL4 lines were obtained by Agrobacterium-mediated transformation of Arabidopsis Col-0 genetic background with the coding sequences of PRPS1 and PRPL4 genes, under the control of CaMV35S promoter (pB2GW7 plasmid; https://gatewayvectors.vib.be/). indPRPS1 and indPRPL4 lines were obtained by cloning the PRPS1 and PRPL4 coding sequences in the Dexamethasone (DEX)-inducible pOp/LhG4 system [41] and by Agrobacterium transformation of Arabidopsis Col-0 plants. prps1-1 (SAIL_560_B02), prpl11-1 (GABI_380H05), clpc1-1 (SALK_014058C) and clpd-1 (SALK_110649C) T-DNA lines were described in previous works [27,42–44] and manually crossed for obtaining prps1-1 prpl11-1, prps1-1 clpc1-1, prps1-1 clpd-1 double mutants. Primers required for gene cloning and mutant line isolation are listed in S1 Table. Wild-type and mutant seeds were grown on soil in climate chambers under long-day (150 μmol m⁻² s⁻¹ 16 h/8 h light/dark cycles) and short-day (150 μmol m⁻² s⁻¹ 8 h/16 h light/dark cycles) conditions. For growth experiments on Dexamethasone, seeds were surface-sterilised and grown for 16 days (80 μmol m⁻² s⁻¹ on a 16 h/8 h light/dark cycle) on Murashige and Skoog medium (Duchefa) supplemented with 1% (w/v) sucrose and 1.5% Phyto-Agar (Duchefa), Dexamethasone was added at the final concentration of 2 μM. Growth rate was determined by ImageJ software (https://imagej.nih.gov/).

Whole-mount preparation and optical microscopy

To analyse defects in embryo development, siliques of Col-0 and heterozygous PRPS1/prps1-2 plants were manually dissected and cleared as reported previously [45]. Developing seeds were observed using a Zeiss Axiophot D1 microscope equipped with differential interface contrast optics. Images were documented with an Axiocam MRc5 camera (Zeiss).

E. coli strains

PRPS1 coding sequence was cloned into pQE31-pREP4 plasmid system (primers are listed in S1 Table), under the control of bacteriophage T5 promoter fused upstream to lacO operator sequences. PRPS1-pQE31-pREP4 and rpsA-pQE31-pREP4 plasmids [33] were then transferred into araBp-rpsA conditional expression strain C-5699, in which the rpsA gene is expressed in presence of 1% (w/v) arabinose and repressed in presence of 0.4% (w/v) glucose. To obtain the PRPS1 overexpressing strain, PRPS1-pQE31-pREP4 plasmids were introduced into the E. coli C-1a strain [46], while rpsA-pQE31-pREP4 strain C-5691 was used as control [16,33].

Chlorophyll a fluorescence measurements

The Imaging Chl a fluorometer (Walz Imaging PAM; https://www.walz.com/) was used to determine Chl fluorescence in vivo. Eight plants for each genotype and condition were analysed at 18 days after sowing (DAS). Average values plus-minus standard deviations were then calculated. 20 min dark-adapted plants were exposed to blue measuring light (intensity 4) and a saturating light flash (intensity 4) was used to calculate the maximum quantum yield of PSII, Fv/Fm.

Nucleic acid analyses

For qRT-PCR analyses, 1 μg of total RNA was treated with iScript™ gDNA Clear cDNA Synthesis Kit (Bio-Rad; https://www.bio-rad.com/) for genomic DNA digestion and first-strand cDNA synthesis. qRT-PCR analyses were performed on a CFX96 Real-Time system (Bio-Rad; https://www.bio-rad.com/), using primer pairs listed in S1 Table. PP2AA3 (AT1G13320) transcripts were used as internal reference [47]. Data obtained from three biological and three technical replicates for each sample were analysed with the Bio-Rad CFX Maestro 1.1 (v 4.1) (Bio-Rad; https://www.bio-rad.com/).

In vivo Translation Assay

The in vivo translation assay was performed essentially as previously described [48]. indPRPS1 leaf discs (6 mm in diameter) were vacuum-infiltrated in liquid MS medium supplemented with 1% (w/v) sucrose and, where indicated, 2 μM Dexamethasone. After 6-hours exposure to 80 μmol photons m⁻² s⁻¹ white light, leaf discs were incubated with a buffer containing 1 mM K₂HPO₄–KH₂PO₄ (pH 6.3) and 0.05% (v/v) Tween-20, supplemented with 20 μg/ml cycloheximide, to inhibit cytosolic translation. [³⁵S]methionine was then added (0.1 mCi/ml) and leaf discs were vacuum-infiltrated and exposed to light (80 μmol photons m⁻² s⁻¹). 5 leaf discs were collected at each time point (15 and 30 min). Total protein extraction and Tris-glycine SDS-PAGE fractionation is described below. Signals were detected using the Phosphorimager GE Healthcare Life Sciences (https://www.gehealthcare.com/).

Isolation of PRPS1-containing protein complexes

The isolation of PRPS1-containing complexes was performed according to previous works [49,50]. 100 mg of leaf fresh weight were ground in liquid nitrogen and resuspended in 1 ml 0.2 M Tris-HCl, pH 9, 0.2 M KC1, 35 mM MgC1₂, 25 mM EGTA, 0.2 M sucrose, 1% Triton X-100, 2% polyoxyethylene-10-tridecyl ether, supplemented with 500 μg/ml heparin, 100 μg/ml chloramphenicol and 25 μg/ml cycloheximide. The extract was then solubilised with 0.5% (w/v) sodium deoxycholate for 5 min on ice. After centrifugation (15 min at 10000g), 800 μl of supernatant was loaded onto 3.6 ml 15–55% (w/v) sucrose gradients in polysome gradient buffer (40 mM TrisHCl pH 8, 20 mM KCl, 10 mM MgCl₂, 100 μg/ml chloramphenicol and 500 μg/ml heparin). Sucrose gradients were centrifuged in SW60 rotors (Beckman) for 18 h at 180000 g at 4°C. 9 fractions of 400 μl each were collected from the top of the tube and subjected to SDS-PAGE fractionation. E. coli ribosome fractionation was performed as described [16]. Quantification of PRPS1 protein accumulation in Low Molecular Weight (LMW; fractions 1–5) and High Molecular Weight (HMW, fractions 6–9) fractions was evaluated by using the Image Lab software on representative blots obtained from three biological and nine technical replicates. Each lane has been quantified as absolute value. The sum of all signals was set to 1. The values reported in the graph are relative to the total sum for each condition (DEX -/+).

Transmission electron microscopy (TEM)

TEM analyses were performed as described previously [51]. Plants were grown for 16 days on MS synthetic medium supplemented with 1% (w/v) sucrose and, where indicated, 2 μM Dexamethasone. Plant material was vacuum-infiltrated with 2.5% glutaraldehyde, in 100 mM sodium cacodylate buffer, for 4 h at room temperature and incubated overnight at 4°C. Samples were rinsed twice with 100 mM sodium cacodylate buffer for 10 min each, and post-fixed in 1% (w/v) osmium tetroxide in 100 mM cacodylate buffer for 2 h at 4°C. After washings, samples were counterstained with 0.5% (w/v) uranyl acetate overnight at 4°C, in the dark. The tissues were then dehydrated by increasing concentrations of ethanol (70%, 80%, 90%; v/v), 10 min each. Samples were then dehydrated with 100% ethanol for 15 min and permeated twice with 100% propylene oxide for 15 min. Epon-Araldite resin was prepared mixing properly Embed-812, Araldite 502, dodecenylsuccinic anhydride (DDSA) and Epon Accelerator DMP-30. Samples were infiltrated first with a 1:2 mixture of Epon-Araldite and propylene oxide for 2 h, then with Epon-Araldite and propylene oxide (1:1) for 1 h and left in a 2:1 mixture of Epon-Araldite and propylene oxide overnight at room temperature. Samples were then incubated in pure resin before polymerisation at 60°C for 48 h. Ultra-thin sections of 70 nm were then cut with a diamond knife (Ultra 45°, DIATOME) and collected on copper grids (G300-Cu, Electron Microscopy Sciences). Samples were observed by transmission electron microscopy (Talos L120C, Thermo Fisher Scientific) at 120 kV. Images were acquired with a digital camera (Ceta CMOS Camera, Thermo Fisher Scientific).

Protein sample preparation and immunoblot analyses

For immunoblot analyses, total proteins were prepared as described [52]. Plant material was homogenised in Laemmli sample buffer [20% (v/v) glycerol, 4% (w/v) SDS, 160 mm Tris–HCl pH 6.8, 10% (v/v) 2-mercaptoethanol] to a concentration of 0.1 mg μl⁻¹ (leaf fresh weight/Laemmli sample buffer). Samples were incubated for 15 min at 65°C and, after a centrifugation step (10 min at 16 000 g), the supernatant was incubated for 5 min at 95°C. Protein samples corresponding to 4 mg (fresh weight) of seedlings were fractionated by SDS–PAGE 10% (w/v) acrylamide [53] and then transferred to polyvinylidene difluoride (PVDF) membranes (0.45 μm pore size). Replicate filters were immunodecorated with specific antibodies. Antibodies directed against AtHsp90-1 (AS08 346) and ClpB3 (AS09 459) were obtained from Agrisera (https://www.agrisera.com/), AtHsc70-4 antibody was obtained from Antibodies-online (https://www.antibodies-online.com/), antibodies directed against plastid ribosomal proteins (PRPS1, PRPL4 and PRPS5) were obtained from Uniplastomic, while polyclonal S1 antibody was kindly donated by U. Bläsi (University of Vienna).

Transcriptome analysis

Total RNA was extracted from leaf discs harvested from indPRPS1 and indPRPL4 plants and vacuum infiltrated in either the absence or presence of DEX for 6 hours for a total of 5 biological replicates for each group. Total RNA extraction was performed using RNeasy Mini Kit (Qiagen), according to the manufacturer’s instructions. RNA concentrations and integrity were determined via NanoDrop One C (ThermoFisher Scientific) and agarose-gel electrophoresis. Extracted RNA samples were sent to Novogene for sequencing via high throughput Illumina NovaSeq platform which employ a paired-end 150 bp sequencing strategy. The RNA-seq library was prepared by enriching for the PolyA-containing transcripts, therefore, no plastid-encoded genes are represented in the library. Raw data were processed and mapped to the Arabidopsis genome TAIR10 using STAR-RSEM software [54]. Differentially Expressed Genes (DEGs) were identified through the R package EdgeR (v 3.15) [55]. Called DEGs were statistically filtered via Benjamini-Hochberg False Discovery Rate method (FDR < 0.05). GO enrichment analyses were performed using agriGO v2.0 online tool and further processed by REVIGO [56,57]. The RNA-seq data were deposited in the Gene Expression Omnibus data repository under the dataset identifier GSE205271.

Proteome analysis

Proteins were extracted from 1 g of plantlets following SDS/phenol method as described [58]. Proteins were then digested with trypsin via Filter Aided Sample Preparation (FASP) [59]. Peptides were analysed by LC-MS/MS as described [60]. Briefly, after LC separation peptides were sprayed into the mass spectrometer and eluting ions were measured in an Orbitrap mass analyser set at a resolution of 35000 and scanned between m/z 380 and 1500. Data dependent scans (top 20) were employed to automatically isolate and generate fragment ions by higher energy collisional dissociation (HCD); Normalised collision energy (NCE): 25% in the HCD collision cell and measurement of the resulting fragment ions was performed in the Orbitrap analyser, set at a resolution of 17500. Peptide ions with charge states of 2⁺ and above were selected for fragmentation. Raw data were searched against the Arabidopsis thaliana TAIR protein database (2010 version) with MaxQuant program (v.1.5.3.3), using default parameters. For the quantitative analysis, the “ProteinGroups” output files were filtered to retain only protein groups detected with at least two peptides in at least three of the four biological replicates, and in at least one analytical group. The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium via the PRIDE [61] partner repository under the dataset identifier PXD034479. Missing values were replaced with the R package imputeLCMD (v.2.1) using Hybrid imputation method: imputation of left-censored missing data (missing values ≥ 50% of number of replicas) was done using QRILC method, instead missed at random data (< 50% of replicas) were imputed using KNN method. Log2 transformed LFQ intensities were centred by Zscore normalisation method of Perseus (https://www.maxquant.org/perseus/) and then subjected to Student’s Tests (S0 = 0.1, FDR<0.05) in order to discover Differentially Abundant Proteins (DAPs). Hierarchical clustering analysis was carried out using Perseus software and default parameters. DAPs categorisation was achieved using TAIR GO annotation tool (https://www.arabidopsis.org/tools/bulk/go/index.jsp). GO term enrichment analysis was performed using the PANTHER classification system (http://geneontology.org) [62].

Depletion of PRPS1 gene leads to embryo lethality

Previous studies reported on the Arabidopsis prps1-1 knock-down mutant phenotype, characterised by pale-green leaves, reduced growth rate and photosynthetic performance, as result of hampered plastid protein synthesis [27,31]. However, the consequences of the complete inhibition of PRPS1 protein accumulation in Arabidopsis plastids has not yet been investigated. To fill this gap, the nuclear PRPS1 gene was edited by targeting the first exon of PRPS1 coding sequence using the CRISPR/Cas9 technology. Plant lines devoid of the Cas9 gene were obtained in T2 generation and the DNA region complementary to the designed guide RNA was sequenced (primers and guide RNA sequences are listed in S1 Table). Only PRPS1/prps1 heterozygous plants, with wild-type-like plant size, leaf pigmentation and photosynthetic performance, could be isolated and the resulting prps1-2 allele showed the deletion of a Cytosine in the first exon (+80 bp from the transcription starting site), right downstream the ATG translation start codon (+ 4 bp from ATG, Fig 1A and 1B). This event disrupts the PRPS1 reading frame and introduces a premature Umber STOP codon in place of Leu-11 (+108 bp from the transcription starting site) (Fig 1A). Furthermore, only PRPS1/prps1-2 heterozygous and PRPS1/PRPS1 homozygous plants could be identified within the progeny of the self-fertilised PRPS1/prps1-2 heterozygous line, showing the 2-to-1 mendelian segregation ratio typical of mutations causing embryo lethality, as in the case of other plastid ribosomal protein knock-out mutants [27,35,63]. Accordingly, the observation of PRPS1/prps1-2 developing siliques at 10 Days After Fertilisation (DAF) revealed one-quarter of the seeds to be albino (Fig 1C), indicating that PRPS1 is essential during early stages of embryogenesis and seed development. In particular, optical section of cleared, whole-mount seeds from PRPS1/prps1-2 siliques at 3 DAF showed that around 25% of the embryos were arrested at the globular development stage, displaying a disorganised cell division pattern similar to the ones previously described for other knock-out mutants in essential plastid ribosomal proteins (Fig 1D) [27]. Furthermore, the defect in embryo development was fully rescued in prps1-2 pPRPS1::PRPS1 plants, obtained by introducing the PRPS1 genomic DNA into the PRPS1/prps1-2 genetic background and isolating wild-type homozygous prps1-2 plants carrying the pPRPS1::PRPS1 construct (Fig 1C and 1D).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Molecular description of prps1-2 allele and the corresponding embryo-lethal phenotype.

A) Schematic representation of PRPS1 gene. Exons are indicated as numbered white boxes, and introns as black lines. Positions of start and stop codons with respect to the transcription initiation site, as well as the CRISPR/Cas9-induced deletion (prps1-2), are indicated. The PRPS1 reading frame and corresponding amino acids in wild-type and prps1-2 mutant are also reported. B) Sequencing electropherogram showing the single nucleotide deletion (coding strand) in prps1-2 hetero- and homozygous mutants with respect to Col-0. The red arrowhead indicates the position of the CRISPR/Cas9-induced deletion. “N” readings in PRPS1/prps1-2 lines result from double peaks, generated by Cas9-induced deletion in one of the two homologous chromosomes. C) Morphological characteristics of developing seeds in siliques at 10 Days After Fertilization (DAF) of Col-0, heterozygous PRPS1/prps1-2 and prps1-2 complemented with pPRPS1::PRPS1 genomic locus. Around one quarter of white seeds are clearly distinguishable among the green seeds of PRPS1/prps1-2 siliques. Bars = 2 mm. D) Cleared whole mount of Col-0, heterozygous PRPS1/prps1-2 and prps1-2 pPRPS1::PRPS1 complemented seeds containing embryos at the globular stage (3 DAF). Bars = 20 μm.

https://doi.org/10.1371/journal.pgen.1010344.g001

PRPS1 over-expression impairs chloroplast activity and biogenesis

Besides the embryo lethal phenotype caused by the prps1-2 allele and the slightly pale cotyledons and true leaves, together with hampered photosynthetic efficiency (F_V/F_M), typical of plants carrying the knock-down prps1-1 allele (Fig 2A; see also [27,31]), the CaMV35S-mediated over-expression of PRPS1 gene in the Arabidopsis Col-0 background (oePRPS1), also resulted in a visible phenotype, characterised by virescent young leaves with a marked drop in F_V/F_M values (Fig 2A). Interestingly, oePRPS1 seedlings were incapable of over-accumulating the PRPS1 protein (Fig 2B), showing an accumulation level lower than the one observed in prps1-1 leaves, despite the PRPS1 transcript level having been about two-fold the Col-0 control leaves (Fig 2C, see also [28]). On the contrary, Col-0 plants carrying the CaMV35S::PRPL4 construct (oePRPL4), here used as the control, were able to accumulate up to 25–30 fold more PRPL4 transcripts and almost double the amount of PRPL4 protein without affecting chloroplast biogenesis and activity, as shown by oePRPL4 lines indistinguishable from Col-0 (Fig 2A–2C).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Visible and molecular phenotypes of plantlets carrying altered amount of PRPS1 and PRPL4 proteins.

A) Visible phenotypes and photosynthetic efficiency (F_V/F_M) of 16 Days After Sowing (DAS) plantlets grown on MS medium devoid (- DEX) or supplemented (+ DEX) with 2 μM dexamethasone. On the left (bright field), the visible phenotypes shown by Col-0, the knock-down prps1-1 mutant, oePRPS1 and oePRPL4 constitutive over-expressing lines, indPRPL4 and indPRPS1 inducible lines. On the right panel (F_V/F_M), the photosynthetic efficiency of each genotype is shown both as false colour imaging and average values ± standard deviations. Statistical significance calculated via Student’s t-test (*** indicates P < 0.001; N.S. not significant). Scale bar = 1 cm. B) Immunoblots of total protein extracts from Col-0, prps1-1, oePRSP1, indPRPS1, oePRPL4 and indPRPL4 16 DAS plants grown on MS medium devoid (- DEX) or supplemented (+ DEX) with 2 μM dexamethasone. PRPS1 and PRPL4 specific antibodies were used for immuno-decoration. Coomassie Brilliant Blue (C.B.B.) stained gels are shown as loading control. Numbers show the relative protein abundance with respect to Col-0 (indicated as 100). Standard deviation was below ± 15%. One filter out of three biological replicates is shown. C) Relative expression values of PRPS1 and PRPL4 genes determined by qRT-PCR analyses of total RNA extracted from Col-0, prps1-1, oePRSP1, indPRPS1, oePRPL4 and indPRPL4 16 DAS plants grown on MS medium devoid (- DEX) or supplemented (+ DEX) with 2 μM dexamethasone. Results of one out of three biological replicates are shown. Error bars indicate standard deviations of three technical replicates. Phenotypic and molecular data of other three independent transgenic lines for each genotype shown here are reported in S1 Fig.

https://doi.org/10.1371/journal.pgen.1010344.g002

In order to investigate this aspect further, PRPS1 and PRPL4 coding sequences were cloned into the pOp/LhG4 vector, which allows the inducible over-expression of the two genes once the glucocorticoid analogue dexamethasone (DEX) is provided. The two constructs were introduced into Arabidopsis Col-0 genetic background, via Agrobacterium-mediated transformation, resulting in the dexamethasone-inducible lines indPRPS1 and indPRPL4. In the absence of DEX, the indPRPS1 line was virtually indistinguishable from Col-0 when grown on MS medium under sterile conditions for 16 days (Fig 2A). Conversely, when the growth medium was supplemented with 2μM DEX, indPRPS1 seedlings showed a leaf virescent phenotype and a drop in photosynthetic performance, together with a reduced accumulation of PRPS1 protein resembling the phenotype of prps1-1 and oePRPS1 lines (Fig 2A and 2B). This was despite PRPS1 transcripts accumulated to levels higher than Col-0 control leaves (Fig 2C). On the other hand, the inducible overexpression of PRPL4 resulted in an increased accumulation of PRPL4 transcripts and protein, similar to oePRPL4 seedlings (Fig 2B and 2C), without any impact on chloroplast activity and leaf greening (Fig 2A). Additional independent over-expressor and inducible lines for both PRPS1 and PRPL4 genes displayed comparable visible and molecular phenotypes, as shown in S1A–S1C Fig.

To understand how chloroplast ultrastructure organization is affected by the increased expression of PRPS1 gene, thin sections of emerging young leaves from 16 DAS seedlings were observed under Transmission Electron Microscopy (TEM; Fig 3). As expected, Col-0, prps1-1, indPRPS1 − DEX, indPRPL4 ± DEX, and oePRPL4 mesophyll cells displayed properly developed chloroplasts with the typical organization in grana stacks and stroma lamellae (Fig 3). However, both the induction (indPRPS1 + DEX) and the constitutive increased expression of PRPS1 caused the formation of miss-shaped and swollen chloroplasts containing enlarged plastoglobuli in the stroma (Fig 3). Furthermore, large budding vesicles with electron dense material were detectable, suggesting ongoing chloroplast degradation, resembling the fission-type ATG-independent micro-autophagy [34,51,52,64]. In some cases, entire round-shaped chloroplasts, detached from the plasma membranes and with still recognizable grana stacks, were observed inside the vacuole, compatible with the ATG-dependent micro-autophagy process (Fig 3J) [34,65]. To further prove that chloroplasts are indeed undergoing vacuole-mediated degradation, we investigated the relative expression of genes associated with either the chloroplast ATG-dependent or ATG-independent chloroplast quality-check and degradation pathways (S2 Fig). Strikingly, both ATI1 [66] and ATG8f [67] transcripts were highly enriched in plantlets with increased PRPS1 transcript accumulation, driven by either CaMV35S- or DEX-induced promoters (S2A Fig), whereas NPC1 and VPS15 [68] were the only genes of the ATG-independent pathway significantly up-regulated in oePRPS1 plantlets (S2B Fig). On the contrary, transcripts of SAG12 gene involved in senescence-associated vacuoles (SAV) pathway [69] could not be detected, in agreement with data from BAR Toronto ePlant [70], most probably as consequence of the fact that the SAV pathway is mainly activated during senescence rather than upon stress conditions.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3.

TEM micrographs of chloroplasts in mesophyll cells of Col-0 (A) control, prps1-1 (B), indPRPS1 (C, G, H, I, J), indPRPL4 (D, E), oePRPL4 (F) and oePRPS1 (K, L) leaves obtained from 16 DAS plantlets grown on MS medium supplemented, or not, with DEX. The young portion of leaves, proximal to the petiole, showing the yellow to pale-green phenotype in oePRPS1 and indPRPS1 lines, was used for the analyses. S: starch granules; White arrowhead: thylakoid membranes; BV budding vesicles; P: Plastoglobules.

https://doi.org/10.1371/journal.pgen.1010344.g003

PRPS1 short-term increased expression inhibits plastid protein translation

To gain a dynamic view on PRPS1 function, the kinetics of PRPS1 transcript and protein accumulation was monitored in leaf discs (6 mm in diameter) from young leaves of indPRPS1 plantlets infiltrated with 2 μM DEX. Leaf discs were sampled at 0, 3, 6 and 24 hours after infiltration (HAI) and transcript and protein accumulation were monitored by RT-qPCR (Fig 4A) and immunoblotting (Fig 4B), respectively. As control, the same experimental set-up was used to monitor the induction of PRPL4 expression (Fig 4A and 4C). The accumulation of PRPL4 and PRPS1 mRNAs indicated that both inducible lines were able to specifically express the related genes with comparable kinetics and transcript amounts (Fig 4A). Both lines reacted to the presence of DEX showing a high expression level of the related transcripts at 3 HAI, reaching the peak at 6 HAI and a significant decrease at 24 HAI. Nevertheless, the induction of PRPS1 expression failed to yield the over-accumulation of PRPS1 protein. Indeed, PRPS1 protein level remained stable until 3 HAI, while diminished to almost undetectable levels from 6 to 24 HAI (Fig 4B). Interestingly, other plastid ribosomal proteins, such as PRPL4 and PRPS5, showed a marked decreased over time, indicating a general alteration of plastid ribosome accumulation, and possibly of plastid translation. On the contrary, the 24-hour induction of PRPL4 resulted in more than two-fold accumulation of PRPL4 protein with respect to time 0 (Fig 4C). In particular, the increase in PRPL4 protein accumulation was observed over time, starting from 3 HAI and reaching the largest amount at 6 HAI, while the accumulation of PRPS1 and PRPS5 plastid ribosomal proteins remained unaltered.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Molecular features of indPRPS1 and indPRPL4 lines upon DEX-induced gene expression.

A) Relative expression values of PRPS1 and PRPL4 genes determined by qRT-PCR analyses of total RNA extracted from indPRPS1 (circles, blue) and indPRPL4 (squares, green) leaf discs sampled at 0, 3, 6 and 24 Hours After Infiltration (HAI) with 2 μM dexamethasone (DEX). Data from one out of three biological replicates are shown. Error bars indicate standard deviations of three technical replicates. B) Immunoblots of total protein extracts from indPRPS1 leaf discs sampled at 0, 3, 6 and 24 HAI with 2 μM DEX. Filters were incubated with antibodies raised against PRPS1, PRPL4 and PRPS5 plastid ribosomal proteins. C.B.B.-stained gels are shown as loading control. Numbers show the relative protein abundance with respect to 0 HAI (indicated as 100). Data from one out of three biological replicates are shown. Standard deviation was below 15%. C) Immunoblots of total protein extracts from indPRPS4 leaf discs, sampled and analysed as in B, used as control. D) Total protein extracts from Col-0 and indPRPS1 leaf discs previously incubated in DEX-containing MS medium for 6 hours and then infiltrated with ³⁵S-Methionine in the presence of cycloheximide, an inhibitor of cytosolic protein synthesis. The leaf proteins were sampled 15 and 30 minutes after labelling (MAL). RbcL and D1/D2 proteins signals are indicated. The ponceau-stained filter is shown as loading control. E) indPRPS1 ± DEX (6 HAI) leaf material was subjected to sucrose gradient fractionation aimed to isolate PRPS1-contaning complexes. Filters were immuno-decorated with PRPS1 antibody to show the accumulation of PRPS1 in Low Molecular Weight (LMW) and High Molecular Weight (HMW) fractions. C.B.B.-stained gels are shown as loading control. F) Quantification of PRPS1 accumulation in LMW (1–5) and HMW (6–9) fractions as evaluated by Image Lab software on representative blots. Each lane has been quantified as absolute value (see also Materials and Methods). The sum of all signals has been set as 1. The values reported in the graph are relative to the total sum for each condition (DEX -/+).

https://doi.org/10.1371/journal.pgen.1010344.g004

To investigate the possible negative effect of PRPS1 inducible expression on plastid protein translation, indPRPS1 leaf discs were incubated in MS medium (± DEX) for 6 hours and then infiltrated with ³⁵S-Methionine and cycloheximide, allowing for the detection of de novo synthesized plastid-encoded proteins, while blocking the cytosolic translation. As shown by the pulse-labelling experiment, the synthesis rate of RbcL and D1/D2 proteins were markedly reduced in indPRPS1 + DEX, over 15 and 30 minutes, when compared to indPRPS1 leaf discs in the absence of dexamethasone (- DEX; Fig 4D), proving that PRPS1 inducible over-expression leads to plastid translation inhibition. This aspect was investigated further by isolating the PRPS1-containing complexes and monitoring the PRPS1-to-ribosome stoichiometry and the accumulation of “free” PRPS1 fraction, as previously reported in E. coli [16]. To do so, indPRPS1 leaf material (6 hours ± DEX) was subjected to sucrose gradient fractionation and probed with the PRPS1 antibody (Fig 4E). As observed in Chlamydomonas [50], PRPS1 protein was found in two distinct populations, as “Low Molecular Weight (LMW, free PRPS1 fraction)”, bound solely to the mRNA, and as “High Molecular Weight (HMW)”, corresponding to the S1 fraction bound to the ribosome core. The increase of PRPS1 presence in the LMW fractions of indPRPS1 + DEX line, compared to the − DEX counterpart (Fig 4E and 4F), supports further the inhibition of plastid translation observed in Fig 4D. Consistent with these data, we observed that PRSP1 was almost absent in HMW fraction in oePRPS1 samples (S3 Fig).

The over-accumulation of Arabidopsis PRPS1 protein inhibits Escherichia coli cell growth

Similarly to Arabidopsis, the depletion of the ribosomal protein S1 (Rps1) in E. coli cells leads to lethality [11]. Moreover, the down-regulation as well as the over-accumulation of Rps1 impair protein translation by altering the stoichiometry between Rps1 and the ribosome core, leading to bacteriostatic effects ([16,33]; see also Fig 5). The E. coli S1 protein is 557 aa long with a molecular weight of 61.2 kDa and possesses six S1 domains ([18]; Fig 5A) that are typical of several RNA binding proteins [71–73]. In A. thaliana, the S1 homologous PRPS1 is markedly smaller, with 373 aa and a molecular weight of 40.5 kDa, as mature form. The in silico analysis of PRPS1 amino acid sequence identified three S1 domains (Fig 5A), in agreement with early analyses of spinach S1 protein [74]. In addition, PRPS1 protein shows a high degree of identity (i.e. about 50%) with the S1 ribosomal proteins from cyanobacteria, which possess three S1 domains as well [19,75]. In order to investigate the possible relationships between the three S1 domains identified in PRPS1 and the six S1 domains in E. coli S1, the amino acidic sequences of each S1 domain were aligned and clustered in a phylogenetic tree (S4 Fig). The resulting tree showed that domains 1 and 2 of PRPS1 are more similar to the corresponding domains of S1, while the PRPS1 domain 3 clusters together with the domains 3, 4 and 5 of S1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Comparison of PRPS1 and S1 protein activities in E. coli cells.

A) Schematic representation of S1 and PRPS1 proteins. S1 domains are depicted as pink boxes, while the chloroplast transit peptide (cTP) of PRPS1 protein as green box. Numbers indicate the position of amino acid residues. Domain predictions are based on InterPro online tool (https://www.ebi.ac.uk/interpro/). B) OD₆₀₀ measurements of E. coli indrpsA (white symbols) and indPRPS1 (blue symbols) strains grown under Depletion (0.4% glucose; squares), Complementation (0.4% glucose + 0.01 mM IPTG; triangles) or Overexpression (1% arabinose + 1 mM IPTG; circles) conditions. Strains were sampled at 0, 45, 90, 135, 180 and 225 Minutes After Induction (MAI). Data from one out of three biological replicates are reported. C) Immunoblots of total protein extracts from indrpsA and indPRPS1 E. coli strains, grown as described above, sampled at 0, 45, 90, 135, 180 and 225 MAI, using S1 and PRPS1 specific antibodies. C.B.B.-stained gels are added as loading controls. Numbers show the relative protein abundance with respect to 0 MAI. Data from one out of three biological replicates are shown. Standard deviation was below 15%.

https://doi.org/10.1371/journal.pgen.1010344.g005

The possible functional homology between the two proteins was then investigated by introducing the PRPS1 gene in E. coli cells, for over-expression and complementation assays. To experimentally test the ability of PRPS1 to complement S1 functions and to repress cell growth when over-accumulated in bacterial cells, both rpsA (encoding the S1 protein) and PRPS1 coding sequences were cloned into pQE31-pREP4 plasmid system under the control of pT5-lacO promoter in pQE31. The plasmids were then introduced into the arabinose-dependent strain C-5699 [33], in which the chromosomal rpsA gene is transcribed from the araBp promoter, generating indrpsA and indPRPS1 strains. Such systems provide a mechanism to deplete cells of the endogenous S1 protein in absence of arabinose and presence of glucose [16,33] and to modulate the expression of either rpsA or PRPS1 genes cloned in pQE31 depending on IPTG concentration.

Both indrpsA and indPRPS1 strains were cultured in three different conditions: Depletion (0.4% glucose and no IPTG), Complementation (0.4% glucose and 0.01 mM IPTG) and Overexpression (1% arabinose and 1 mM IPTG). The growth conditions for the complementation assay were experimentally optimized based on the growth of indrpsA E. coli strain. Cell growth was measured every 45 minutes up to 225 minutes after the induction (MAI, Fig 5B). At each time point cells were sampled, normalized on the optical density (OD₆₀₀) value, and the total protein extract was used to detect the accumulation of S1 and PRPS1 proteins in both strains (Fig 5C). Under Depletion conditions, both strains showed almost completely impaired growth rate, due to the limited accumulation/absence of S1 protein. When grown in Complementation conditions, indrpsA cells were able to actively replicate and showed comparable amounts of S1 protein at each time point, indicating that the experimental conditions were properly set up to induce a wild-type-like S1 protein accumulation. On the other hand, indPRPS1 cells failed to sustain growth in such conditions despite the gradual accumulation of PRPS1 protein, proving the inability of PRPS1 to complement the S1 function in E. coli. As expected, under Overexpression conditions, indrpsA strain growth was inhibited due to the excessive amount of S1 protein (Fig 5B and 5C). Strikingly, the over-accumulation of PRPS1 protein was effectively repressing the bacterial growth and led to S1 protein depletion, too (Fig 5B and 5C). It is worth noting, that the overaccumulation of PRPS1 Arabidopsis protein was able to repress the growth of E. coli cells, comparably to S1 overaccumulation, even when the overexpression of either rpsA or PRPS1 genes was achieved in the C-1a E. coli strain, devoid of the conditional depletion system of the endogenous S1 protein (S5A Fig). To better understand whether PRPS1 protein is capable of interacting with E. coli ribosomes under overexpression conditions, we sampled both indrpsA and indPRPS1 cells at 90 MAI. Cell lysates were then fractionated into ribosome-unbound (supernatant, SN) and ribosome-bound (pellet, P) fractions and analysed via immunoblotting to detect either S1 or PRPS1 protein localisation. Interestingly, PRPS1 was retrieved in both ribosome-bound and -unbound fractions, similarly to S1 from E. coli, suggesting that the Arabidopsis PRPS1 can compete for the ribosome core with E. coli S1 protein (S5B Fig). Taken together, these data indicate that PRPS1 protein is able to inhibit E. coli growth when over-expressed in addition to the endogenous S1, whereas it is unable to functionally replace the E. coli endogenous S1 protein.

PRPS1 accumulation is negatively regulated by chloroplast CLP protease complex

The phenotypes observed in Arabidopsis plants upon PRPS1 overexpression indicate that the abundance of PRPS1 protein must be kept under a strict post-transcriptional control to prevent inhibition of protein synthesis and chloroplast damage (see Figs 2–4). This notion is supported further by the inhibitory role of PRPS1 protein over-accumulation on E. coli cell growth (Figs 5 and S5). In order to investigate the molecular mechanism responsible for controlling PRPS1 protein abundance, leaf discs were infiltrated with DEX supplemented with lincomycin (LIN), a specific inhibitor of plastid 70S ribosomes. Strikingly, a large accumulation of PRPS1 protein, most probably as result of PRPS1 degradation suppression, could be observed even after 24 hours from DEX infiltration, unlike the control sample (Fig 6A). Intriguingly, the most relevant chloroplast stromal protease is represented by the CLP complex, which is composed by several nuclear-encoded subunits and one plastid-encoded component, ClpP1, that is part of central proteolytic core [39,76]. To investigate whether the CLP complex could indeed be responsible for maintaining PRPS1 protein below levels that would otherwise cause damages to the chloroplast, we crossed prps1-1 with mutants altered either in chloroplast protein translation, prpl11-1, or lacking two plastid chaperones required to feed CLP protease with protein substrates, clpc1-1 and clpd-1 ([44]; Fig 6B). prps1-1 prpl11-1 double mutant showed reduced growth rate (S6 Fig) and a slight decrease in photosynthetic efficiency (F_V/F_M) with respect to prpl11-1 parental line. Similar genetic interactions were observed in prps1-1 clpc1-1 double mutant, which showed a severe reduction in growth rate with respect to prps1-1 and clpc1-1 single mutants. Interestingly, prps1-1 clpd-1 double mutant showed partially restored growth rate, close to wild-type-like levels, and a slight recovery of F_V/F_M parameter. Strikingly, the accumulation of PRPS1 protein increased about two-fold in all the double mutants tested (Fig 6C) with respect to prps1-1, further supporting the notion that plastid translation and the CLP protease complex play an important role in controlling PRPS1 abundance in the stroma of chloroplasts. To further verify the role of CLP protease complex in the regulation of PRPS1 protein abundance in plastids, the clpc1-1 and clpd-1 mutants were manually crossed with indPRPS1 lines to yield the indPRPS1 clpc1-1 and indPRPS1 clpd-1 genotypes. Leaf discs harvested from these lines were then infiltrated with DEX and PRPS1 accumulation was monitored over 24 hours. Importantly, PRPS1 protein was able to accumulate in large amount upon DEX induction in both lines altered in CLP activity, unlike in indPRPS1 leaf discs (Fig 6D). Similar results were observed in the crosses obtained with indPRPS1 #2 line (S7A Fig). Additionally, clpc1-1 and clpd-1 mutants were also crossed with oePRPS1 lines. Interestingly, the obtained oePRPS1 clpc1-1 and oePRPS1 clpd-1 lines sensibly recovered the photosynthetic phenotype of oePRPS1 (S7B Fig), most probably as a consequence of the increased PRPS1 abundance (S7C Fig), further supporting the notion that the CLP protease complex plays an important role in controlling PRPS1 abundance in the stroma of chloroplasts. Finally, no accumulation of PRPS1 precursors were detected outside the chloroplasts under overexpression conditions by immunoblot analyses on plastid-enriched fraction and soluble extra-plastid fraction isolated from Col-0, prps1-1 and oePRPS1 16 DAS plantlets grown on soil, indicating that the reduction of PRPS1 protein upon its overexpression is mainly due to the activity of CLP protease (S8 Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. PRPS1 protein accumulation and the stromal CLP protease.

A) Immunoblots of total protein extracts from indPRPS1 leaf discs sampled at 0, 3, 6 and 24 HAI with 2 μM dexamethasone alone (+ DEX) or supplemented with 550 μM lincomycin (+ DEX + LIN). Filters were incubated with antibodies raised against PRPS1 protein. C.B.B.-stained gels are shown as loading control. Numbers below the immunoblots show the relative protein abundance with respect to 0 HAI (indicated as 100). Data are derived from one out of three biological replicates. Standard deviation was below 15%. B) Visible phenotypes of Col-0 and prps1-1, prpl11-1, clpc1-1, clpd-1, prps1-1 prpl11-1, prps1-1 clpc1-1 and prps1-1 clpd-1 16 DAS seedlings grown on soil and their photosynthetic efficiency (Fv/Fm). Average values ± standard deviations. Scale bar = 1 cm C) Immunoblot of total protein extracts from Col-0 and prps1-1, prpl11-1, clpc1-1, clpd-1, prps1-1 prpl11-1, prps1-1 clpc1-1 and prps1-1 clpd-1 mutants obtained by using a PRPS1 specific antibody. The immunoblot incubated with Histone H3 specific antibody and a C.B.B.-stained gel are shown as loading control. Numbers below the immunoblot show the relative protein abundance with respect to Col-0. Standard deviation was below 15%. D) Immunoblots of total protein extracts from indPRPS1, indPRPS1 clpc1-1 and indPRPS1 clpd-1 leaf discs sampled at 0, 3, 6 and 24 HAI with 2 μM DEX. Filters were incubated with antibodies raised against PRPS1 protein. C.B.B.-stained gels are shown as loading control. Numbers below the immunoblots show the relative protein abundance with respect to 0 HAI (indicated as 100). Data are derived from one out of three biological replicates. Standard deviation was below 15%.

https://doi.org/10.1371/journal.pgen.1010344.g006

Short-term induction of PRPS1 and PRPL4 gene expression induces different nuclear gene expression responses

The opposite behaviour of indPRPS1 and indPRPL4 plants in terms of protein pattern accumulation upon DEX-mediated induction of gene expression make them the ideal genetic material to investigate the primary nuclear gene expression response to changes in plastid ribosomal protein content. To this aim, a transcriptome analysis was performed on leaf discs harvested from indPRPS1 and indPRPL4 and vacuum infiltrated in either the absence or presence of DEX. In particular, total leaf RNA was extracted after 6 hours from infiltration, i.e., at the stage of maximal accumulation of PRPS1 and PRPL4 transcripts (see Fig 4A) and at the beginning of evident changes in protein pattern accumulation (see Fig 4B and 4C), but without major visible and photosynthetic defects, reducing at minimum the consequences of pleiotropic effects (S9 Fig), and subjected to Illumina sequencing. Principal component analyses (PCA) of indPRPS1 and indPRPL4 samples showed a clear separation between the untreated and treated samples (S10 Fig), despite the short induction time. The EdgeR package was used to identify differentially expressed genes (DEGs, listed in S2 Table), obtained by comparing data from indPRPS1-DEX with indPRPS1+DEX, and indPRPL4-DEX with indPRPL4+DEX (Fig 7A). Overall, 431 DEGs in indPRPS1 upon DEX induction and 328 in indPRPL4 were identified. Among them, 124 DEGs were in common between the two datasets. Next, the obtained DEGs were divided into up and down regulated genes both in indPRPS1 and indPRPL4 datasets, obtaining indPRPS1UP, indPRPL4UP, indPRPS1DOWN and indPRPL4DOWN lists, which were further compared to isolate unique DEGs in each group (Fig 7B, see also S2 Table). As a result, among the up-regulated DEGs, 161 were uniquely found in indPRPS1 (Sheet A in S2 Table) and 159 DEGs in indPRPL4 (Sheet B in S2 Table), while 107 DEGs were up-regulated in both lines (Sheet C in S2 Table). Among the down-regulated DEGs, 146 were unique for indPRPS1 line (S2 D), 45 unique DEGs for indPRPL4 (Sheet E in S2 Table) and only 14 were commonly down-regulated (Sheet F in S2 Table). Furthermore, among the 149 down-regulated DEGs in indPRPS1, 3 were upregulated in indPRPL4 (Sheet G in S2 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. Comparison of RNAseq transcriptomes from indPRPS1 and indPRPL4 lines infiltrated or not with DEX.

A) Venn diagram showing the number of DEGs found in indPRPS1 and indPRPL4 upon induction. B) Venn diagram showing the distribution of DEGs resulted from the comparison of indPRPS1UP, indPRPL4UP, indPRPS1DOWN and indPRPL4DOWN lists. Unique DEGs in each group are in bold. Group A: Unique up-regulated DEGs found in indPRPS1; Group B: Unique up-regulated DEGs found in indPRPL4; Group C: Common up-regulated DEGs found in indPRPS1 and indPRPL4; Group D: Unique down-regulated DEGs found in indPRPS1; Group E: Unique down-regulated DEGs found in indPRPL4; Group F: Common down-regulated DEGs found in indPRPS1 and indPRPL4; Group G: up-regulated DEGs found in indPRPL4 but down-regulated in indPRPS1.

https://doi.org/10.1371/journal.pgen.1010344.g007

To investigate the biological functions activated or repressed by the overexpression of either PRPS1 or PRPL4 genes, the unique DEGs found up- or down-regulated in indPRPS1 or indPRPL4 were analysed with agriGO v2.0 online tool [56]. Next, further analysis by REVIGO [57] was preformed to retrieve Biological Process Gene Ontology terms and group them into functional categories (Fig 8 and S3–S5 Tables). The 161 DEGs up-regulated in indPRPS1 produced strongly enriched GO terms associated with responses to endogenous factors or involvement in flowering and seeds production, such as “photoperiodism, flowering” (GO:0048573), “response to karrikin” (GO:0080167), “vegetative to reproductive phase transition of meristem” (GO:0010228) or “cellular response to hormone stimulus” (GO:0032870) (Fig 8A; see also S3 Table). On the other hand, the repressed biological functions found in indPRPS1 were related to the production of secondary metabolites, defence against herbivores and oxidative stress such as “glucosinolate biosynthetic process” (GO:0019761), “sulfur compound biosynthetic process” (GO:0044272), “cell redox homeostasis” (GO:0045454) and “response to wounding” (GO:0009611), to cite a few of them (Fig 8B; see also S3 Table). Interestingly, the high enrichment in GO terms associated with flowering and reproductive phase transition is in good agreement with the anticipated flowering phenotype, calculated as number of rosette leaves at bolting, observed when PRPS1 transcripts are overexpressed (Fig 9). In particular, in the case of plants cultivated on soil under growth chamber conditions, Arabidopsis Col-0 bolted with an average of about 10–11 rosette leaves while the oePRPS1 line showed an anticipated flowering time, bolting at the stage of 8 rosette leaves, comparable to prps1-1 behaviour (Fig 9A and 9B). The same anticipated flowering time occurred under short day growth conditions, with the wild-type bolting at 54 leaves while the overexpressing line flowered at 37 rosette leaves on average (Fig 9C). A similar observation was made when indPRPS1 line was grown on MS medium supplemented with 1% (w/v) sucrose and DEX, in which its number of leaves at bolting was 5, similarly to oePRPS1 (Fig 9D). Instead, when indPRPS1 was grown on MS medium without DEX, the number of leaves at bolting was about 6–7, comparable to what observed in indPRPL4 and Col-0 plants.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Analysis of the biological functions activated or repressed by the short-term overexpression of PRPS1 and PRPL4 genes.

Significantly enriched GO terms identified through agriGO v2.0 and REVIGO online tools retrieved from A) unique up-regulated DEGs found in indPRPS1; B) unique down-regulated DEGs found in indPRPS1; C) unique up-regulated DEGs found in indPRPL4.

https://doi.org/10.1371/journal.pgen.1010344.g008

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 9. Flowering time determination in Col-0, prps1-1, oePRPS1, indPRPS1 (± DEX) and indPRPL4 (± DEX) lines.

A) Pictures of representative plants at bolting grown on soil in long day conditions. B) Average numbers ± standard deviations of leaves at bolting observed in Col-0, prps1-1 and oePRPS1 plants grown on soil in long day conditions. C) Average numbers ± standard deviations of leaves at bolting observed in Col-0, prps1-1 and oePRPS1 plants grown on soil in short day conditions. D) Average numbers ± standard deviations of leaves at bolting observed in Col-0, oePRSP1, indPRPS1 and indPRPL4 plants grown on MS medium devoid (- DEX) or supplemented (+ DEX) with 2 μM dexamethasone in long day conditions. Statistical significance was calculated via Student’s t-test (*** indicates P < 0.001).

https://doi.org/10.1371/journal.pgen.1010344.g009

Conversely, the analysis of the activated biological processes in indPRPL4 showed a strong enrichment in GO terms associated with abiotic stress responses and protein homeostasis such as “response to hydrogen peroxide” (GO:0042542), “response to high light intensity” (GO:0009644), “response to heat” (GO:0009408), “response to oxidative stress” (GO:0006979) and “protein folding” (GO:0006457) (Fig 8C; see also S5 Table). These activated cellular pathways are in agreement with the observed increased levels of CLPB3, HSP90-1 and HSC70-4 proteins in indPRPL4 line upon induction with DEX (see S11 Fig). In particular, CLPB3 protein abundance and even more the abundance of HSP90-1 and HSC70-4 proteins largely increased already at 3 HAI in indPRPL4 leaf disks, unlike in indPRPS1 samples, possibly generated by the overload of the plastid folding machinery caused by PRPL4 abundance and the consequent activation of chloroplast UPR. Interestingly, the unique down-regulated DEGs found in indPRPL4 leaf disks resulted in no significantly enriched GO terms.

The RNA-seq library was prepared to enrich the PolyA-containing transcripts only, therefore, no plastid-encoded genes are represented. Therefore, to evaluate if the plastid transcription was affected by the DEX-mediated induction of either PRPL4 or PRPS1 genes, the very same samples used for transcriptome analyses were analysed via RT-qPCR to quantify the expression of a subset of PEP- and NEP-dependent plastid-encoded genes (S12 Fig). Interestingly, no significant differences in the expression of plastome genes in both lines after DEX treatment could be detected.

Short-term induction of PRPS1 and PRPL4 gene expression induces different changes in proteomic profiles

To further investigate the molecular responses following the DEX-mediated induction of PRPS1 and PRPL4 genes, a proteomic analysis was performed on leaf tissue harvested from indPRPS1 and indPRPL4 plants and vacuum infiltrated either in the absence or presence of DEX and sampled, as in the case of transcriptome analysis, after 6 hours of induction. PCA analysis revealed that dexamethasone-mediated induction led to significant changes also at proteomic level (S13 Fig). Comparing induced (+DEX) and control groups (-DEX), the abundances of 58 and 77 proteins were altered in indPRPS1 and indPRPL4 samples, respectively (T Student’s test, FDR < 0.05; listed in S6 Table). The minimal overlapping between the two datasets (only two common proteins) confirms the different cellular responses following the induction of PRPS1 and PRPL4 genes (Fig 10A). In particular, 27 and 31 proteins were up- and down-accumulated in indPRPS1 leaf discs, whereas 45 and 32 proteins were up- and down-accumulated in indPRPL4 samples, respectively (Fig 10B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 10. Comparison of proteome from indPRPS1 and indPRPL4 lines infiltrated or not with DEX.

A) Venn diagram showing the number of DAPs found in indPRPS1 and indPRPL4 upon induction. B) Venn diagram showing the distribution of DAPs resulted from the comparison of indPRPS1UP, indPRPL4UP, indPRPS1DOWN and indPRPL4DOWN lists.

https://doi.org/10.1371/journal.pgen.1010344.g010

GO enrichment analysis of differentially abundant proteins (DAPs) found in indPRPS1 samples revealed over-represented GO terms only among the down-accumulated proteins, probably due to the small size of the data set. The enriched Biological Process terms were “translational elongation” (GO:0009658) and “chloroplast organization” (GO:0006414) (S7 Table). Among the enriched Cellular Component terms, “ribosome-associated quality control (RQC) complex” (GO:1990112), “transcriptionally active chromatin” (GO:0035327) and “chloroplast stroma” (GO:0009570) were found (S8 Table). All DAPs were grouped based on their GO annotations (Fig 11). Proteins up-regulated in response to PRPS1 induction were mainly involved in “anatomical structure development” (GO:0048856), “catabolic process” (GO:0009056) and “response to light stimulus” (GO:0009416) (Fig 11A). In addition, the up-accumulated proteins in indPRPS1 plantlets were mainly located in “nucleus” (GO:0005634) and “cytoplasm” (GO:0005737) (Fig 11B). Down-regulated DAPs produced GO annotations such as “biosynthetic processes” (GO:0009058), “RNA binding” (GO:0003729), “translation” (GO:0006412) and “post-embryonic development” (GO:0009791) (Fig 11A). Among down-regulated proteins detected in indPRPS1 samples, we found the regulator of fatty-acid composition 3 (RFC3; AT3G17170), which plays an important role in the plastid rRNA processing [77]. Many of these down-regulated proteins were located in chloroplast and nucleus (Fig 11B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 11. Protein categorization based on GO terms.

Protein categorization based on Biological process GO terms of DAPs found in indPRPS1 (A) and indPRPL4 (C) samples and based on Cellular Content GO terms found in indPRPS1 (B) and indPRPL4 (D) samples. Blue and red bars correspond to proteins that were up or down accumulated, respectively.

https://doi.org/10.1371/journal.pgen.1010344.g011

The significantly enriched GO terms, retrieved by analysing the up-regulated proteins detected in indPRPL4 samples, were mainly associated with mitochondrial electron transport chain and mitochondrial ribosomes (S9 Table). Moreover, GO annotation revealed that several DAPs in indPRPL4 seedlings belong to “response to stress” (GO:0006950), as also detected by the transcriptome analysis (Fig 11C). The GO annotation tool allowed to locate DAPs in indPRPL4 plants especially in the nucleus, chloroplast and cytoplasm (Fig 11D). Overall, our results, in agreement with transcriptomic analysis, point out very specific proteomic changes following the alteration of the two ribosomal proteins PRPS1 and PRPL4. In fact, the induction of PRPS1 leads to a general repression of translation, transcription and chloroplast organization. Conversely, the increased amount of PRPL4 protein promotes the cellular stress response, leading to a positive regulation of proteins involved in RNA metabolism and translation, with a major contribution also given by components of the mitochondrial metabolism.

Finally, to verify whether the differential gene expression overserved in indPRPS1+ DEX leaves was due to either increased PRPS1 transcripts or decreased PRPS1 protein amount, we analysed the accumulation of TULP5 and SWEET13 (S14A Fig) transcripts (highly up-regulated in indPRPS1 + DEX leaf discs) by comparing their expression in indPRPS1 +DEX, in which the up-regulation of PRPS1 transcripts is followed by PRPS1 protein degradation, to indPRPS1 clpd-1 + DEX, in which PRPSI transcript upregulation is observed but PRPS1 protein degradation is inhibited (Figs 6 and S7). Interestingly, TULP5 and SWEET13 transcripts did not over-accumulate in the absence of a functional CLP protease, indicating that the re-orchestration of nuclear gene expression requires the accumulation/degradation of PRPS1 protein. Similarly, transgenic lines in which the prps1-2 null allele is either under the control of CaMV35S promoter or the inducible promoter (see Materials and Methods) in Col-0 wild-type background did not show any virescent phenotype, supporting the notion that, in order to trigger its own degradation, the transient over-accumulation of PRPS1 protein is required (S14B Fig).

Altered PRPS1 expression impairs chloroplast biogenesis

PRPS1 is a nuclear gene encoding the S1 protein of the plastid 30S small ribosomal subunits. Reduced expression of PRPS1 gene results in diminished plastid translation and decreased photosynthetic efficiency [27,31], while its disruption arrests embryo development at the globular stage in Arabidopsis (see Fig 1) and prevents greening of seedlings in rice (albino seedling lethality 4; [78]). Clearly, these observations indicate that PRPS1 is an essential gene required during early stages of chloroplast biogenesis, as reported in the case of other essential nuclear genes coding for plastid ribosomal proteins [27,35,79]. The apparent discrepancy between the early block of embryo development in Arabidopsis and the arrest of seedling greening in rice is in agreement with previous findings [80–83] and compatible with the essential nature of fatty acid biosynthesis during chloroplast biogenesis. In particular, the lack of the plastid-encoded accD-subunit of the multimeric acetyl-CoA carboxylase required for fatty acid biosynthesis is responsible for the lethality of Arabidopsis embryos defective in plastid translation [35]. In contrast, grass species contain a plastid-located monomeric acetyl-CoA carboxylase that, differently from Arabidopsis, is encoded in the nucleus and translated in the cytosol [84,85]. Therefore, fatty acid biosynthesis (and embryogenesis) can continue even when plastid protein synthesis is affected in these species.

Intriguingly, the reduced accumulation of S1 protein can also be obtained when PRPS1 gene expression, and the consequent transcript accumulation, is both constitutively increased in oePRPS1 and indPRPS1 + DEX seedlings (Fig 2) and induced in indPRPS1 leaf discs for 24 hours (Fig 4). This finding, together with the very limited increase in transcript accumulation observed in oePRPS1 and indPRPS1 + DEX seedlings, indicates the existence of post-transcriptional regulatory mechanisms aimed to prevent PRPS1 over-accumulation, as shown previously [31]. The strict control on PRPS1 gene expression and protein accumulation appears to be rather specific and certainly it does not apply to PRPL4, a nuclear gene encoding a core subunit of the 50S plastid ribosome, also essential for embryo development and plant viability [27,35]. In fact, oePRPL4 and indPRPL4 + DEX lines were able to accumulate 25–30 times more transcripts, together with the double amount of PRPL4 protein, in comparison to Col-0 and indPRPL4 − DEX controls (Figs 2 and 4).

Furthermore, the constitutive over-expression of PRPS1 gene, driven in 16 DAS oePRPS1 and indPRPS1 + DEX plantlets, resulted in the impairment of chloroplast differentiation and physiology, leading to a virescent phenotype visible in the emerging leaves and in the younger portions of the leaves, corresponding to the tissue proximal to the petiole, but not in cotyledons, supporting the existence of major differences in the molecular mechanisms at the basis of plastid biogenesis, plastid protein homeostasis and degradation between cotyledons and leaves [51,52,86,87]. In particular, transmission electron microscopy (TEM) observations of the chlorotic tissues revealed the presence of cells displaying misshapen chloroplasts with altered thylakoid membrane ultrastructure and large plastoglobuli in the stroma (Fig 3), reported to be associated with the degradation of chlorophylls and thylakoids in response to abiotic and biotic stresses or during senescence [88–90]. Moreover, the presence of vesicles budding from chloroplasts and containing electron dense material, together with the observation of entire round chloroplasts inside the vacuole (Fig 3), indicates the ongoing chloroplast degradation, similarly to what reported in literature [51,52,64,65]. Coherently, the relative expression of ATI1 and ATG8f genes, involved in ATG-dependent micro-autophagy, was enhanced upon PRPS1 attempted over-accumulation (Figs 3 and S2A), whereas the expression of genes involved in ATG-independent micro-autophagy was mildly stimulated in oePRPS1 (S2B Fig) [68]. Overall, these observations indicate that the reduced accumulation PRPS1 severely jeopardizes chloroplast integrity during leaf development, leading to the dismantling of damaged and misshapen chloroplasts with the final aim to remove reactive oxygen species–producing chloroplasts and redistributing nutrients to other tissues [34].

A fully functional proteostasis machinery is needed to control PRPS1 accumulation in chloroplast stroma

All our attempts to increase the abundance of PRPS1 protein failed (Figs 2 and 4; [28]) and, together with that, the accumulation of other plastid ribosomal proteins, such as PRPL4 and PRPS5, was reduced, as observed in indPRPS1 + DEX leaf discs (Fig 4), indicating that the decreased abundance of PRPS1 protein has a deleterious effect on plastid ribosome stability. In agreement with that, the pulse-labelling experiment conducted on indPRPS1 leaf material infiltrated with DEX for 6 hours showed a severe inhibition of plastid protein translation and a larger fraction of PRPS1 protein freely associated to mRNA and not bound to actively translating ribosomes (Fig 4). These findings explain the defects in leaf greening observed in the different Arabidopsis lines (Figs 1 and 2) and support the role of PRPS1 as a stringently regulated translation factor rather than a “real” ribosomal protein, given its weakly and reversible association with the 30S subunit, similarly to previous observations in E. coli [16].

S1 protein is the closest PRPS1 homologue in E. coli, and it is encoded by the essential gene rpsA [11]. To gain possible insights on PRPS1 function, we attempted to rescue E. coli cell lethality due to S1 depletion and to phenocopy the bacteriostatic effects of S1 overaccumulation, by modulating the level of PRPS1 protein in E. coli cells (Fig 5). Our data clearly showed that PRPS1 is not able to functionally replace the endogenous S1, as cells depleted of rpsA but with moderate amount of PRPS1, failed to grow (Fig 5). According to previous studies, S1 protein exerts its functions in relationship with the specializations of its six S1 domains [19]. The interactions with the ribosome relies on S1 domains 1 and 2 [20], whereas the ability to bind mRNAs has been associated with the S1 domains 3, 4 and 5 [14]. As for domain 6, if removed together with S1 domain 5, the initiation of translation is hampered [19,25] and together with domain 4, it is implicated in ribosome dimerization and hibernation in stationary phase [26]. According to our in silico analysis, S1 domains 1 and 2 of PRPS1 appeared to be more similar to the respective domains 1 and 2 of S1, while domain 3 clustered together with domains 3, 4 and 5 of S1 (S4 Fig). The fact that PRPS1 is a smaller protein and possesses three out of the six S1 domains identified in S1, of which none could be associated with those required for translation initiation in E. coli, could explain the inability of PRPS1 to functionally replace S1 protein. Nonetheless, PRPS1 overaccumulation blocked cells growth as much as S1 (Figs 5 and S5). It has been shown that S1 over accumulation inhibits translation since the excess of “free” S1 interacts with mRNAs, preventing the ribosome loading [16,17,25,91]. However, PRPS1 was found both in ribosome-bound and -unbound fractions, suggesting that, although PRPS1 can interact with the ribosome core in E. coli cells, this interaction is rather unfruitful. In this scenario PRPS1 would be able to inhibit E. coli growth by competing with the endogenous S1 protein and inhibiting ribosome activity (S5 Fig). Nevertheless, PRPS1 is likely capable of binding E. coli mRNAs with the S1 domains 2 and 3, since E. coli and plastid mRNAs share similar 5’ UTRs with AU-rich sequences [92], making E. coli transcripts inaccessible to ribosomes and acting as a negative modulator of translation initiation. The inhibitory role of PRPS1 protein on E. coli protein synthesis is corroborated further by the fact that the accumulation of PRPS1 upon overexpression reaches very rapidly the plateau level and concomitantly leads to the decreased accumulation of S1 protein over time (Fig 5C), mimicking the role of S1 as feedback effector of its own regulation at the translational level [17].

Such regulatory mechanism has a different spatial constraint in photosynthetic eukaryotes due to the physical separation of the nuclear/cytosolic compartments, where the PRPS1 transcripts and the precursor protein are synthesized, and the chloroplast stroma where the mature PRPS1 protein plays its functions. While the limited accumulation of PRPS1 transcripts observed upon constitutive expression allows us to hypothesize the existence of a cytosolic post-transcriptional regulatory mechanism [32], our data strongly support the activation of a chloroplast stroma post-translational regulatory mechanism, in response to PRPS1 overexpression, mediated by the plastid proteostasis machinery (Fig 6). In chloroplasts, the major soluble stromal protease is the CLP complex, composed of multiple nuclear-encoded subunits with the addition of ClpP1 subunit, the only one to be plastid-encoded [76]. Consequently, ClpP1 abundance is susceptible to genetic defects in plastid gene expression or to drugs inhibiting plastid translation [39]. The co-infiltration of DEX with the plastid translational inhibitor lincomycin allowed a larger accumulation of PRPS1 protein (Fig 6A). Accordingly, an increased PRPS1 protein accumulation (Fig 6C) could be achieved by introgressing the prps1-1 knock-down allele into prpl11-1 genetic background, in which the chloroplast translation is reduced [42]. This is also in line with the restoration of PRPS1 protein accumulation observed in prps1 gun1 and prps1 rh50 double mutants, as GUN1 stimulates the activity of the Nuclear-Encoded Polymerase (NEP) that, among other plastid house-keeping genes, is responsible of the transcription of clpP1, while RH50 is involved in plastid ribosome assembly and plastid translation [28,52,93–95]. Intriguingly, GUN1 was also found to physically interact with the plastid protein homeostasis machinery, including CLPC subunits [28]. Furthermore, a comparable increase in PRPS1 levels were also observed in the double mutants prps1-1 clpc1-1 and prps1-1 clpd-1 (Fig 6C), in indPRPS1 clpc1-1 and indPRPS1 clpd-1 (Fig 6D) and in oePRPS1 clpc1-1 and oePRPS1 clpd-1 (S7C Fig), in which the two plastid chaperones required by CLP protease to interact with the substrates are missing [44]. These pieces of evidence strongly point towards CLP protease complex as one of the main regulators of PRPS1 protein abundance in Arabidopsis chloroplasts.

Our findings are also in agreement with previous reports. In particular, by combining transcriptomic and proteomic analyses, Wu et al. were able to show that plastid ribosomal proteins are regulated post-translationally, suggesting a protein degradation-based mechanism [96]. Moreover, in Chlamydomonas, CreS1 expression is induced by light, while CreS1 protein levels remain constant, pointing also in this case to the post-translational regulation of CreS1 abundance [50].

On the other hand, the constitutive over-accumulation of PRPL4 protein upon induction didn’t affect neither the plastid translation nor the chloroplast ultrastructure (Figs 2–4). Interestingly, chaperones both resident in the cytoplasm and in plastids were progressively up-regulated as PRPL4 levels increased, indicating the activation of protein homeostasis mechanisms to cope with the increased protein amount (S11 Fig). These different responses to the overexpression of two plastid ribosomal proteins could be due to their intrinsic features and activities, having PRPS1 mRNA-binding properties which, if not kept under strict control, would have dramatic consequences on chloroplast functionality.

PRPS1 overexpression promotes early flowering while PRPL4 overaccumulation triggers cpUPR

The switch from vegetative growth to reproductive growth is a pivotal event in plant development, mostly dependent on the environmental stimuli such as day length (photoperiodic flowering) and temperature (vernalization) that plants perceive to determine the proper timing to assure successful reproduction, and ultimately the survival of the species [97]. However, plants subjected to a variety of stressful conditions can anticipate their flowering through a new category of flowering response, known as stress-induced flowering, aimed to guarantee species survival when they cannot adapt to unfavourable conditions [98]. Although the mechanistic details behind the stress-induced flowering are still not fully understood, hormones seem to be at least partially involved in these pathways as they are produced under stress and regulate gene expression to cope with it [98,99]. This seems to be the case of plants subjected to prolonged PRPS1 overexpression and characterised by an early flowering phenotype regardless of day length (Fig 9). In these plants, the stress seems to be caused by the reduced chloroplast protein synthesis and the consequent accumulation of damaged proteins and, possibly, by low sucrose concentration due to the reduced photosynthetic performance (see Fig 2; [100]). Consistently, the transcriptomic profile of Arabidopsis lines characterised by the short-term induction of PRPS1 gene highlighted that the early nuclear gene expression response to PRPS1 imbalance in chloroplasts is characterised by a robust enrichment of up-regulated genes associated with the reproductive phase transition and the cellular response to hormones, such as “photoperiodism, flowering”, “vegetative to reproductive phase transition of meristem” and “cellular response to hormone stimulus” (Fig 8 and S3 Table), and by the concomitant repression, amongst others, of genes involved in “cell redox homeostasis”, “response to wounding”, “response to oxidative stress”, “response to water deprivation”. Accordingly, the circadian clock regulators RVE8 and CO, the IDD8 transcription factor and the Gibberellic acid enzyme ga3ox1, all positive modulators of flowering, are among the up-regulated genes. On the other hand, the tetratricopeptide thioredoxin-like protein TTL4, required for osmotic stress tolerance, BASS5 and BCAT4, involved in glucosinolate biosynthesis, PTR3, required for defences against pathogens, and the transcription factor NAC019, involved in the response to dehydration, were found to be down-regulated upon PRPS1 induction [101–105]. Interestingly, the mass spectrometry analysis of the tissue overexpressing PRPS1 detected the increased accumulation of far-red insensitive 219 (FIN219; AT2G46370) and the phytochrome associated protein phosphatase 2C (PAPP2C; AT1G22280), both directly involved in the response to red, or far red light and in flower development [106,107]; while chloroplast-located proteins involved in “RNA metabolism” and “translation” where down-regulated.

In contrast, the results obtained from the transcriptomic and proteomic analyses performed on indPRPL4 + DEX leaf discs were markedly different (Figs 7,8,10 and 11). The overaccumulation of PRPL4 promoted abiotic stress responses and protein homeostasis such as “response to hydrogen peroxide”, “response to high light intensity”, “response to heat”, “response to oxidative stress” and “protein folding” (Fig 8C). In particular, transcription factors involved in stress responses such as DREB2A, DREB2C and WRKY26 [108–110] were found to be upregulated, together with proteins directly involved in proteostasis maintenance, such as small HSPs, CLPB1 cytosolic unfoldase and the co-chaperone HOP3 [111–114]. Interestingly, GOLS1 gene coding for a key enzyme of raffinose family sugar synthesis, shown to enhance plant resistance to oxidative damage, was also up-regulated [115,116]. These findings together with the increased accumulation of CLPB3, HSP90-1 and HSC70-4 proteins over time (see S11 Fig), corroborate the hypothesis that PRPL4 over-accumulation contributes to generate pressure on the plastid folding machinery that, in turns, activates retrograde signalling pathways aimed at triggering a chloroplast-related UPR. It can be envisaged that the activated cpUPR contributes to the limited increase of PRPL4 protein abundance in Arabidopsis chloroplasts, i.e., two folds higher upon 24 HAI with respect to 0 HAI, despite the 30-fold increase of PRPL4 transcripts (Fig 4A and 4C).

Taken together these observations highlight the activation of different nuclear and cellular responses upon up-regulation of PRPS1 and PRPL4 gene expression and the consequent alteration of plastid protein homeostasis. Whereas the PRPS1-related response mainly relies on chloroplast breakdown with degradation and loss of chloroplast proteins, nucleic acids, pigments, lipids, and polysaccharides, aimed to promote the remobilization of resources, such as carbon and nitrogen, in favour of the anticipated reproductive phase, i.e. an escape strategy from a stress condition, the PRPL4-related response consists in the activation of a stress-adaptive pathway, compatible with the chloroplast UPR.