UPP affects chloroplast development by interfering with chloroplast proteostasis

Arabidopsis uracil phosphoribosyltransferase (UPP) is an essential enzyme which appears to have a previously unknown, moonlighting activity. Our analysis of UPP amiRNA mutants has confirmed that this vital function is crucial for chloroplast development and growth. Interestingly, this function appears to be unrelated to nucleotide homeostasis since nucleotide levels were not altered in the studied mutants. Transcriptomics and proteomic analysis suggest that UPP plays a role in chloroplast proteostasis, especially under high light (HL). Immunoblots of mature plants and a de-etiolation experiment with young seedlings revealed PetC, the iron-sulfur protein of the cytochrome b6f complex, as a putative UPP target. In addition, UPP and PetC were identified in a high molecular weight complex. Consistently, we show that PetC is massively reduced in UPP knock-down plants. The block observed in photosynthetic electron transport, as evidenced by reduced high light-induced non-photochemical quenching (NPQ) but increased unregulated energy dissipation (NO), might be therefore a consequence of reduced PetC. After HL treatment, UPP amiRNA mutants showed impaired photosynthesis and reduced carbohydrate contents, resulting in an inability to induce flavonoid biosynthesis. In addition, the levels of the osmoprotectants raffinose, proline and fumarate were found to be reduced. Proteases, including thylakoid filamentation temperature-sensitive 1, 5 (FtsH), caseinolytic protease proteolytic subunit 1 (ClpP1), and processing peptidases, as well as components of the chloroplast protein import machinery, were up-regulated. In sum, our work suggests that UPP assists in stabilizing and protection of PetC during the assembly of the Fe-S cluster and targeting to the thylakoid.


Introduction
Nucleotides are organic molecules that are essential components of genetic material and play important roles in cellular processes.Uracil phosphoribosyl transferase (UPP) catalyses the first committed step in uracil salvage and is located in the chloroplast.Arabidopsis recombinant UPP converts uracil and phosphoribosyl pyrophosphate into uridine monophosphate (UMP).UPP knockout mutants (upp-1) do not exhibit uracil phosphoribosyl transferase (UPRT) activity, indicating that no other protein exhibits UPRT activity (Mainguet et al., 2009;Ohler et al., 2019).The functionality of uracil salvage in vivo was demonstrated by 5-fluorouracil toxicity and consequential resistance of mutants with reduced UPRT activity (Ohler et al., 2019).Seedling lethal upp-1 can be rescued by complementation with nearly catalytic inactive UPP but not with a highly active homolog from Escherichia coli.This agrees with previous results and suggests that an unknown second (moonlighting) function is responsible for the inability of upp-1 to establish photosynthesis (Mainguet et al., 2009;Arrivault, 2019;Ohler et al., 2019).
Photosynthesis is a dominant process in green plant tissues.Chloroplasts have their own machinery for transcription and translation of several photosynthesis genes.Mutants in nucleotide metabolism that exhibit nucleotide limitation also affect photosynthesis.Examples of knock-down mutants targeting enzymes involved in pyrimidine de novo synthesis include aspartate transcarbamoylase (ATC), which catalyses the first committed step, CTP-synthase 2 (CTPS2), which catalyses the final step, as well as thymidine kinase 1b (TK1b) and double knock-out mutants for uridine-cytidine kinase (UCK) 1 and 2 (Clausen et al., 2012;Ohler et al., 2019;Bellin et al., 2021a;Bellin et al., 2021b).The rescue of ATC and CTPS mutants by nucleotide precursors provides evidence that nucleotide limitation is the cause of the observed effects on photosynthesis.
Previous work has provided evidence for a moonlighting function of Arabidopsis UPP in chloroplast development and the establishment of photosynthesis, as reported by Ohler et al. (2019).
Additionally, downstream enzymes such as plastid UMP kinase (PUMPKIN) and nucleoside diphosphate kinase 2 (NDPK2) have also been found to exhibit moonlighting functions (Dorion and Rivoal, 2015;Arrivault, 2019;Schmid et al., 2019).Since knock-out mutants of both enzymes are not lethal to seedlings, as is the case with UPP, these moonlighting functions are likely distinct from those of UPP.
Chloroplast differentiation, which leads to photosynthesizing chloroplasts, occurs in two distinct phases: the Structure Establishment Phase and the Chloroplast Proliferation Phase (Pipitone et al., 2021).The establishment of the structure phase is associated with the formation of thylakoid membranes and the assembly of the photosystems (PSII, PSI, cytochrome b 6 f (Cyt b 6 f)) within the first 24 hours after de-etiolation (Pipitone et al., 2021).However, the biogenesis and assembly of the photosystems, particularly cytb6/f, remain poorly understood.A new assembly factor for this complex was recently identified as de-etiolation induced protein 1 (DEIP1).DEIP1 is a 25 kDa protein located in the thylakoid that interacts with PetA and PetB from cytb6/f during assembly (Sandoval-Ibanez et al., 2022).This work proposes a function of UPP in chloroplast proteostasis, primarily by affecting PetC stability and assembly into cytb6/f.This hypothesis is supported by transcriptomics and proteomics of UPP-amiRNA lines (ami-upp).Additionally, immunoblotting of mature and de-etiolated Arabidopsis plants provides further evidence.Reduced PetC amounts are accompanied by increased levels of chloroplast protein import machinery components and proteases, such as thylakoid filamentation temperature-sensitive (FtsH) metalloproteases involved in the assembly or repair of PSII, PSI, and Cyt b 6 f complexes (Malnoe et al., 2014;Jarvi et al., 2016;Kato et al., 2018).

Results
Although nucleotide levels are wild type like, UPP amiRNA lines exhibit reduced growth.
UPP is an enzyme involved in pyrimidine salvage and has a secondary function that affects the development of UPP mutants (Ohler et al., 2019).To test this finding, we quantified levels of (deoxy)nucleotides, nucleosides, and nucleobases in ami-upp-4 and UPP-OX.There were no significant differences in the levels of these molecules between the two genotypes (Figure 1D, Supplemental Figure S1, 2).It is unlikely that UPP plays a direct role in nucleotide metabolism that is fundamental for the distinct growth phenotypes of amiRNA mutants.The analysis included plants of all genotypes grown at 4°C or in HL (high light, 1000 µmol photons s -1 m -2 ).Comparing between conditions revealed some alterations, such as increased levels of purine nucleotides ATP, ADP and dADP in one day cold-treated plants (see Supplementary Figure S1).In HL-treated plants, both ADP and dADP increased, whereas UDP and GDP decreased (see Supplemental Figures S1 and S2).

Transcriptome analysis points to alterations in the functional category "protein".
RNA-Seq analysis of four-week-old UPP mutants grown under control conditions or under high light for 3 hours revealed a high number of differentially expressed genes (DEGs).We included the high light treatment because we had already observed a high light-dependent lack of anthocyanin accumulation in ami-upp lines (see below).Under medium light conditions (ML, 120 µmol light), 252 differentially expressed genes (DEGs) were found in ami-upp-4 compared to Col-0, with a padj.value of < 0.05.However, after high light (HL, 1000 µmol) treatment, this number increased to 2502 DEGs in the same comparison together with a higher scattering of expression values when including all genes which were identified (Figure 2A-C).Among genes encoding non plastid proteins (padj.< 0.05 and log 2 -fold change <-1 or >1) qua-quine starch (QQS) and cytosolic ribosomal protein P2D (RPP2D) were found downregulated in ML and HL conditions.QQS represents a regulator of carbon and nitrogen allocation to starch and protein (Li and Wurtele, 2015).When focusing on the expression of nuclear genes encoding chloroplast proteins (log fold change <-1 or >1), only three altered genes were identified in ML-grown plants and five genes in HL-grown plants.The upregulation of both early light-induced proteins (ELIP) 1 and 2 was more pronounced under high light conditions compared to moderate light conditions (Figure 2D, E).ELIPs function as regulators of chlorophyll synthesis and protect against light-induced damage by binding free chlorophyll molecules (Tzvetkova-Chevolleau et al., 2007).Additionally, high expression levels of ELIPs are observed 4 to 24 hours after de-etiolation (Pipitone et al., 2021).An analysis of GO terms for significantly differentially expressed genes (DEGs) in HL-grown plants revealed a high abundance of genes related to the categories of 'protein' and 'ribosomal proteins', while the categories of 'stress' and 'abiotic stress' were downregulated (Figure 2F).

Proteomics reveal distinct alterations in proteostasis especially upon HL treatment.
Label-free proteomics was selected as an objective method to investigate the moonlighting function of UPP.Whole rosette material from five-week-old plants was exposed to HL treatment after two hours of illumination with ML (0h HL).Three hours of HL (3h HL), eight hours of HL (8h HL), and subsequent 12 hours of dark recovery followed by 2h HL (recovery) were also tested.Samples from all time points were processed in three independent biological replicates.The untreated samples exhibited 104 significantly (p-value < 0.05 in pairwise comparison) upregulated and 113 downregulated proteins in ami-upp-4 compared to Col-0 (Figure 3A, B).This number increased to 151 upregulated and 209 downregulated proteins after 3h HL, of which 23% were located in the chloroplast (Figure 3A).The most significant increase was observed after 8h HL, with 380 upregulated and 303 downregulated proteins, of which 27% were located in the chloroplast (Figure 3A, C).After recovery, 166 proteins were still upregulated and 245 downregulated compared to Col-0 (Figure 3A).The 8h HL proteome underwent a GO-term analysis using the Mapman tool (Thimm et al., 2004).The analysis showed significant changes in categories 29 (protein), 20 (stress), 21 (redox), and 16 (secondary metabolism) (Figure 3D).Bin 29 (protein) comprises 158 proteins, of which 48% are differentially expressed proteins (DEPs) involved in protein-synthesis, followed by 23% in proteindegradation, 10% in amino acid activation, 8% in protein-targeting, 4.5% in posttranslational modification, 4.5% in protein-folding, and 2% in assembly and cofactor ligation (see Figure 3E).
The protein groups that are markedly affected include those involved in electron transport reactions, chloroplast protein import and folding, and proteases (as shown in Figure 4A-C).Other groups consist of ribosomal proteins and redox processes, as well as the CBB (Calvin-Benson-Bassham cycle) (as depicted in Supplemental Figure S3).It was found that the light harvesting complexes, PSII subunits and Cytb6/f subunits were mostly reduced, whereas those of ATP-synthase, NDH (NAD(P)H dehydrogenase subunits involved in cyclic electron flow) and FNR (Ferredoxin NADP+ reductase) were increased (Figure 4A).Additionally, several components of chloroplast protein import were found to be altered.For instance, the proteins Toc33, Toc159, Toc75-3, and Tic110, which are part of the translocon of the outer and inner chloroplast envelope, respectively, were found to be increased (Figure 4B).Additionally, chaperones from the HSP and CPN60 families were observed to be altered at different time points in HL.OEP (outer envelope protein) 161 and 163 were reduced together with a factor (AKR2) mediating import of outer envelope membrane proteins (Bae et al., 2008).
In contrast, the amounts of OEP64 and OEP80 were increased after 8 hours of HL.
Additionally, spSecA1, a motor protein involved in the cpSec1 pathway for ATP-driven import of nuclear-encoded proteins into thylakoids, was upregulated (Figure 4B).Furthermore, several chloroplast proteases were found to be altered.After 8 hours of HL exposure, stromal and thylakoid processing peptidases SPP, SPPA, and Prep1 (presequence peptidase 1) increased, similar to thylakoid-bound FtsH (filamentation temperature sensitive-H) 1 and 5 (= variegated1, var1) (Figure 4C).DEGP1 (degradation of periplasmic proteins 1) and 2 decreased, whereas members of the Clp (caseinolytic protease) family mostly increased after 8 hours of HL or recovery.A metalloprotease is necessary for grana development and works together with FtsH2 (var2) to regulate chloroplast development.Its expression is significantly reduced after 8 hours of high light exposure (Figure 4C).

PetC levels are reduced and thylakoid located FtsH proteases are up in ami-upp mutants.
As many changes in the chloroplast proteome were detected, most prominent for PetC (Figure 4D) the steady-state levels of selected photosynthesis-related proteins were quantified through immunoblotting.The most significant change in abundance was again observed with PetC, which was greatly reduced in abundance in ami-upp-4 and -7 under both control and 8h HL (Figure 5A).An upregulation of FtsH 1/5 was observed, most notably in HL treated ami-upp-7.Both ami-upp mutants showed upregulation of PsbS under ML and HL conditions, while all other tested proteins remained nearly unchanged (Figure 5A).PsbS, also known as NPQ4, activates non-photochemical quenching (NPQ) and regulates high light acclimation by enhancing reactive oxygen species (ROS) homeostasis (Correa-Galvis et al., 2016).Additionally, proteomics analysis revealed an increase in FtsH1 and FtsH5 after 8h HL and recovery (Supplemental Figure S4).We conducted a de-etiolation experiment to determine if impaired PetC accumulation was observed shortly after illumination.After 7 days of growth in complete darkness, we observed only low accumulation of UPP and PetC.However, after only 24 hours of illumination, PetC levels were strongly induced together with UPP in Col-0, but not in petc-2 (Figure 5B).The amount of other tested proteins was hardly affected.
Next, a Blue Native-polyacrylamide gel electrophoresis was performed on thylakoids of Col-0 plants to examine a putative interaction between UPP and PetC.PetC, which is part of the cytochrome b 6 f complex, is consistently found at higher molecular weights.UPP may exist as a tetramer, with a maximum size of approximately 100 kDa.However, it was discovered that both PetC and UPP occur at the same molecular weight of around 400 kDa, indicating that they are part of the same complex (Figure 5C).
As increased levels of ribosomal proteins were observed (Supplemental Figure S3C), we quantified the translation efficiency in this organelle using two independent setups.Firstly, chloroplast translation was blocked by lincomycin.After overnight incubation in the dark, high light was applied for a further 4 hours, and the photosynthetic efficiency (Fv/Fm) was determined by pulse amplitude modulation (PAM).At all time-points, Fv/Fm was significantly reduced in ami-upp-7, and in ami-upp-4 at 1 h and 4 h HL (Figure 5D).Newly synthesized proteins in leaf discs of five-week-old plants were labelled with 35 S and their activity in total labelled proteins was quantified over 40 min.After 15 min of labelling, both ami-upp-4 and -7 showed significantly lower incorporation, and ami-upp-7 also showed lower incorporation after 40 min (Figure 5E).Both experiments suggest that low UPP levels reduce chloroplast translation, which contradicts the finding of higher levels of ribosomal proteins.This observation indicates altered proteostasis, which may be caused by impaired protease activity.
UPP-G224L was identified as a mutant not able to complement upp-1.When the recombinant, purified protein was subjected to size exclusion chromatography, most of the protein eluted at a size corresponding to an UPP dimer, whereas wild-type UPP appeared as tetramer (Supplemental Figure S5).

Flavonoid biosynthesis is impaired in high light treated ami-upp-4 line
During the testing of UPP mutants in various growth regimes, it was observed that ami-upp-4 and -7 lacked the typical red leaf coloration after several days of HL treatment.The quantification of anthocyanins revealed a 14-fold increase in Col-0 after 4d HL, which was similar in UPP-OX, but only a 7-fold increase in ami-upp-4 (Figure 6A).The quantification of gene expression for early and late flavonoid biosynthetic genes, as well as the corresponding transcription factors (TFs, Araguirang and Richter 2022), revealed a significant reduction in the induction of transcripts essential for anthocyanin biosynthesis in HL-treated ami-upp-1 and -4 mutants (Figure 6B).The degree of transcript reduction correlated well with the reduced UPP expression and growth in the corresponding mutants.Figure 6B shows that the expression of PAP1, a major transcription factor essential for the induction of anthocyanin biosynthetic genes in HL, was reduced.This suggests downregulation of sugar-related signaling pathways (Zirngibl et al., 2022).Quantification of neutral sugars glucose, fructose, and sucrose revealed significant reductions in HL-treated ami-upp-7, but not in upp-ami-4 (Figure 6C).However, both mutants were unable to increase starch levels after HL treatment compared to Col-0 (Figure 6C).Glucose-6-phosphate and fructose-6-phosphate levels were similar to those of Col-0.HL treated ami-upp-4 showed a reduction in phosphoenolpyruvate (PEP), which is a precursor of the chloroplast shikimate pathway, as well as fumaric acid and raffinose, both of which are relevant for stress responses (Kaplan et al., 2007;Dyson et al., 2016) (Figure 6D).In the context of altered carbohydrate levels, a significant reduction in TPT levels was observed in ami-upp-4 after 3, 8, and 24 hours of HL exposure (Figure 6E).TPT is the primary export route for photoassimilates and the most prominent membrane protein in the envelope membrane (Flügge, 1999).Amino acids act as the fundamental constituents of proteins, metabolic precursors, and signaling molecules.Their biosynthesis and degradation undergo changes during plant development and respond to stress levels.Typically, amino acid levels rise under stress conditions (Hausler et al., 2014) including the most abundant amino acids, glutamate, glutamine, and aspartate.The study measured amino acids in ami-upp-4 and UPP-OX mutants under different conditions.It was observed that UPP-OX mutants showed little or no changes in amino acid levels compared to Col-0, while ami-upp-4 mutants showed an increase in seven amino acids after 3 h HL, which became more pronounced after 8 h HL.The highest increases were measured in serine, aspartate, and glutamate (Supplemental Figure S6).During the recovery phase after the night, most amino acids in ami-upp-4 decreased again.The largest increase was observed for alanine after 3h HL (Supplemental Figure S6).

UPP affects chloroplast proteostasis and thereby impairs accumulation of PetC
The findings of this study indicate that UPP plays a crucial role in maintaining chloroplast proteostasis.Proteomics and immunoblot analysis revealed that PetC, a subunit of the cytochrome b 6 f (Cyt b 6 f) complex, is a major target.The Cyt b 6 f complex acts as a central node between the two photosystems in the thylakoid membrane of chloroplasts.Notably, the absence of PetC in petc-2 mutants only allows growth in the presence of sucrose (Maiwald et al., 2003).Reduced PetC levels may account for the failure of ami-upp-lines to establish NPQ.Additionally, PetC was previously identified as an interaction partner of UPP in yeast two-hybrid studies (Arabidopsis Interactome Mapping, 2011).Blue native page followed by immunoblotting revealed that UPP and PetC run at the same molecular weight, which is significantly higher than that of tetrameric UPP.Therefore, it is likely that both proteins interact in the same complex.
Since the PetC transcript remained unchanged in ami-upp-4 compared to Col-0, the observed decrease must take place at the protein level.This could be due to reduced translation, import into chloroplasts and thylakoids, altered turnover, or assembly into the Cyt b 6 f complex.A correlation between UPP and PetC was found in de-etiolation experiments.Here we could demonstrate that PetC levels increase significantly after 24 hours of light in Col-0, but not in ami-upp-7.Both proteins were present in low levels in dark-grown seedlings and increased significantly shortly after the start of illumination in our work in line with observations found in Pipitone et al. (2021).Additionally, a strong correlation was also found between UPP protein levels, PetC levels, and the growth performance of corresponding mature plants (ami-upp and UPP-OX).We interpret this as a hint for a direct role of UPP in regulating photosynthesis at the level of PetC which then affects assimilation and growth.In omics experiments, no apparent link to other metabolic or regulatory pathways that might be affected by UPP was seen.A putative obvious link to nucleotide metabolism is also less likely as nucleotide amounts were not changed in ami-upp.The genetic background also did not influence the increased amounts of ADP and dADP accompanied by decreased UDP levels that we observed upon HL treatment.The previously described significant increase of ATP in the cold (Chen et al., 2022) was also not altered in any of the genotypes.
UPP mutants display noticeable changes in photosynthetic energy dissipation.While ami-upp-4 can compensate for reductions in yield (YII) by increasing NPQ, the strongest line (ami-upp-7) fails to do so and instead increases non-regulated energy dissipation (NO).NO increases when electron transport is blocked between PSII and the Cytb6f complex, as is the case in plants treated with 3-(3',4'-Dichlorophenyl)-1',1'-dimethyl urea (DCMU) (Russell et al., 1995;Vogel et al., 2012).Thus, the NO levels in ami-upp-7 grown under control conditions are like those of DCMU-treated Col-0 plants (see Figure 6).This observation provides additional experimental evidence of a block in electron flow in cytb6/f, which regulates NPQ based on the pmf across the thylakoid (Armbruster et al., 2017) (Figure 6B).Therefore, we conclude that UPP plays a role in stabilising or protecting PetC against degradation or supporting its repair.Little is known about this process.However, recently, de-etiolation induced protein 1 (DEIP1) was identified as an interactor of PetA and PetB subunits of Cyt b 6 f.The Cyt b 6 f core subunits PetA, B and C were missing in deip-1 loss of function mutants (Sandoval-Ibanez et al., 2022).As previously observed for petc-2, deip-1 exhibits similarities with upp-1 and upp-ami lines in terms of phenotype, such as smaller habitus with pale leaves and reduced chloroplast size.
Protein amount alterations were detected in signal processing peptidases, ClpP6 protease complex subunits, and thylakoid FtsH proteases in ami-upp-4 (Figure 4C), all of which are crucial for coordinating the maturation and degradation of the chloroplast proteome (Rowland et al., 2019).In addition, the CPN60 chloroplast chaperone complex is an essential component of the chloroplast protein folding machinery (Ries et al., 2023) and might be involved.How do proteases and chaperones interact with Cyt b 6 f?It was proposed that PetC assembles with Cpn60 in complexes before being incorporated into the Cyt b 6 f complex (Molik et al., 2001).However, this assumption was recently questioned by screening for CPN60 interactors through competition co-immunoprecipitation, where neither PetC nor UPP were identified (Ries et al., 2023).Nevertheless, interactions between the cochaperonin lid and the ClpP protease complex were identified.The amount of ClpP subunits was found to be altered in ami-upp-4, and UPP was identified as a substrate of ClpP6 (Sjogren et al., 2006), suggesting the possibility of indirect interaction between these complexes and UPP.
UPP may play a role in stabilising PetC during its import into the thylakoid.PetC is imported into the thylakoids via the cpTat or delta pH pathway to be assembled into the cytochrome b 6 f complex.However, its import is slower compared to other thylakoid proteins, likely due to the need to incorporate Fe-S cofactor.UPP may bind to PetC during this process in a holdase-like manner.
Holdases are typically small, abundant proteins, like the many small heat shock factors (Reinle et al., 2016) and UPP is similar in this respect.It has been shown that only oligomers, but not dimers, can exhibit holdase function for Get3 (Ulrich et al., 2023).Similarly, the UPP-G224L mutant appears as a dimer in contrast to the wildtype tetramer (Supplemental Figure S5).Although wild-type UPP and other point-mutated versions of UPP were able to complement upp-1 (Ohler et al., 2019), we were unable to achieve complementation with UPP-G224L.This suggests that the oligomeric state of UPP as a tetramer is crucial for its function.

UPP mutants fail to accumulate sugars and anthocyanin in HL
An interesting finding was that ami-upp-4 and -7 were unable to accumulate anthocyanins after HL-treatment.Upon analysing the expression of genes involved in flavonoid biosynthesis, it was observed that all tested biosynthesis genes were less induced by HL in the mutants compared to Col-0, including the transcription factors Myb12 and Pap1 (Figure 6).This implies that ami-upp mutants failed to induce flavonoid biosynthesis.It was demonstrated that upon HL sugars, particularly triosephosphates (TP), exported to the cytosol by TPT, are crucial for this process (Zirngibl et al., 2022).In line with this, the detectable amount of TPT protein in ami-upp-4 was significantly reduced already after 3h HL and even more so after 8h HL.Therefore, it is probable that the reduced export of TP from the chloroplast and the absence of the inducing sugar signal are responsible for the decreased anthocyanin biosynthesis in UPP mutants.
The photosynthetic efficiency of ami-upp-4 is reduced, but this is largely compensated by the reduced growth.Neutral sugars and hexose phosphates were not significantly altered in these plants.
However, starch accumulation in HL was significantly reduced (Figure 6).Reduced TPT amounts may function to retract triosephosphates in the chloroplast to prevent depletion of the CBB or to support the oxidative pentose phosphate pathway in the dark.Protease activity is most likely required to achieve such a fast a reduction of more than 8-fold in TPT, the most abundant protein in the chloroplast envelope (Flügge, 1999).This phenomenon has not been previously described.Future studies should investigate whether the rhomboid-like proteases RBL10 and RBL11.So far, the only known proteases acting in the chloroplast envelope (Lavell et al., 2019) are involved in regulating TPT abundance.
Besides anthocyanins, raffinose, phosphoenolpyruvate and fumarate were significantly reduced in ami-upp-4 under control and HL conditions compared to Col-0 (Figure 6).Increased levels of these metabolites are a typical HL response and this is also reflected in our analysis in Col-0 plants.
Fumarate and starch can act as flexible and alternative carbon sinks in Arabidopsis and FUM2 is a key regulator in this scenario (Chia et al., 2000;Pracharoenwattana et al., 2010).In line with reduced levels of fumarate, we observed decreased levels of FUM2, which may be attributed to the stabilization of starch levels in ami-upp-4.Moreover, the inability of ami-upp-4 to accumulate these osmoprotectants is most likely a consequence of reduced photosynthetic capacity in these mutants.transcription into cDNA.Transcripts related to flavonoid biosynthesis and corresponding primers for qRT-PCR are described in (Zirngibl et al., 2022).Additional primers for UPP are listed in Supplemental Table S2.

Protein extraction, gel-electrophoresis and immunoblotting
Proteins were extracted according to (Ermakova et al., 2019).Therefore, 50-100 mg leaf samples of were grinded in liquid nitrogen with protein extraction buffer (100mM trisaminomethane-HCl, pH 7.8, supplemented with 25 mM NaCl, 20mM ethylenediaminetetraacetic acid, 2% sodium dodecyl sulfate (w/v), 10 mM dithiothreitol and 2% (v/v) protease inhibitor cocktail (Sigma, St Louis, MO).Protein extracts were incubated at 65 °C for 10 min and then centrifuged at 13,000 g for 1 min at 4 °C.Protein extracts were supplemented with 4× Laemmli buffer (BioRad, Hercules, CA), the samples were stored at -80°C until useSDS page was performed with BioRad gels and according to the manufacturers protocol.
Immunoblots were essentially performed as given in (Ohler et al., 2019), here the proteins were separated on SDS-gels and transferred onto a nitrocellulose membrane by the semi-dry Transblot Turbo Transfer System (Bio-Rad, Hercules, CA, USA).Primary antibodies were obtained from Agrisera (Vännäs, Sweden) except for anti-UPP (Ohler et al., 2019).For blue native electrophoresis (BN-PAGE), extraction of thylakoids and was done as given in (Schägger et al., 1994).After electrophoresis, the gels were blotted onto a PVDF membrane as described above.

Photosynthetic activity was measured using pulse-amplitude-modulation
(PAM) fluorometry, as previously described (Bellin et al., 2023).Five-week-old plants grown under standard conditions were transferred to high light, and leaf discs from respective plants were treated with DCMU (5µM).Five leaf discs were incubated for 90 minutes in 5ml buffer with CaCl 2 (Russell et al., 1995).

Statistical analyses
Statistical analysis of qRT-PCR, proteomic and metabolomic analysis data was performed using the Student's t-test (see also legend for details).Asterisks indicate statistically significant differences determined using Student's t-test (*p < 0.05; **p < 0.01, ***p < 0.001).The significance criteria for comparisons within the proteomic dataset were based on adjusted p < 0.05.Changes in protein levels within a given pairwise comparison were expressed as log2 fold change (log 2 FC), or as fold change (FC) where FC equals the second condition divided by the first condition (Pagano et al., 2023).

S pulse labelling of proteins in leaves
Labelling of proteins in leaf discs by 35 S methionine/cysteine was performed as given in (DeTar et al., 2021).).Plants were grown for five weeks grown under a short-day regime on soil (10h light/ 14h dark).Scattered plot of all up-and down-regulated genes found in ami-upp-4 and Col-0 under control (B) and high light (C) conditions.Three biological replicates were analyzed per time point.All genes located in the chloroplast, from both control (D) and high light (E) conditions, were plotted in a separate scatter plot.B-E Labeled genes filtered by padj.< 0.05, log 2 fold change >1, <-1, all units: log 10 FPKM.(F) GO-term analysis of significantly responded Mapman bins after Benjamini-Hochberg correction (padj.< 0.05).

Figure legends Figure 1 .
Figure legends

Figure 2 .
Figure 2. Gene expression analysis showed alterations in ami-upp-4 under high light stress.(A) RNA Seq analysis of different expressed genes (DEGs) in ami-upp-4 compared to Col-0 under three hours of medium light (= control) and three hours of high light (1000 μmol photons m −2 s −1

Figure 3 .
Figure 3. Proteomics reveal distinct alterations in proteostasis under high light stress.(A) Workflow of proteome experiment.HL (1000 μmol photons m −2 s −1 ) was applied for the indicated time period and differentially (p-value < 0.05) up and downregulated proteins were analyzed (numbers are indicated).Three biological replicates were analyzed per time point.DEPs were analyzed in pairwise comparison using t-test.(B, C) Volcano plots show DEPs (p < 0.05) in ami-upp-4 compared to Col-0 under control (B) and after eight hours high light treatment (C).Proteins localized to the chloroplast are highlighted in dark green and plastidic proteases in bright green.(D) GO-term analysis of significantly responded Mapman bins after Benjamini-Hochberg correction (padj.< 0.05) and (E) further subdivision of bin 29, red = upregulated proteins, blue = down-regulated proteins.(* = p < 0.05; ** = p < 0.01; *** = p < 0.001).

Figure 4 .
Figure 4. Proteome mass spec analysis reveals alterations in plastidic proteins in ami-upp-4 mutants upon high light treatment.(A, B, C) Comparison of protein alterations between Col-0 and ami-upp-4 associated to electron transport reactions (A), chloroplast protein import and folding (B) and chloroplast proteases (C) (p < 0.05).Plants were grown for five weeks grown under a short-day regime on soil (10h light/ 14h dark).(D) Proteome mass spec quantification (Log 2 LFQ) of PetC from a HL time course experiment.P-