Inositol polyphosphates and target of rapamycin kinase signalling govern photosystem II protein phosphorylation and photosynthetic function under light stress in Chlamydomonas

Stress and nutrient availability inﬂuence cell proliferation through complex intracellular signalling networks. In a previous study it was found that pyro-inositol polyphosphates (InsP 7 and InsP 8 ) produced by VIP1 kinase, and target of rapamycin (TOR) kinase signalling inter-acted synergistically to control cell growth and lipid metabolism in the green alga Chlamydomonas reinhardtii . However, the relationship between InsPs and TOR was not completely elucidated. (cid:1) We used an in vivo assay for TOR activity together with global proteomic and phosphopro-teomic analyses to assess differences between wild-type and vip1-1 in the presence and absence of rapamycin. (cid:1) We found that TOR signalling is more severely affected by the inhibitor rapamycin in a vip1-1 mutant compared with wild-type, indicating that InsP 7 and InsP 8 produced by VIP1 act independently but also coordinately with TOR. Additionally, among hundreds of differentially phosphorylated peptides detected, an enrichment for photosynthesis-related proteins was observed, particularly photosystem II proteins. The signiﬁcance of these results was underscored by the ﬁnding that vip1-1 strains show multiple defects in photosynthetic physiology that were exacerbated under high light conditions. (cid:1) These results suggest a novel role for inositol pyrophosphates and TOR signalling in coordinating photosystem phosphorylation patterns in Chlamydomonas cells in response to light stress and possibly other stresses.

In Chlamydomonas, a mutation in one of two VIP paralogues (Cre03.g185500),vip1-1, was isolated in a screen for increased sensitivity to the TOR-specific inhibitor rapamycin (Rap) (Couso et al., 2016).vip1-1 was also hypersensitive to other TOR inhibitors, torin1 and AZD8055, indicating a specific interaction between the TOR and InsPs signalling pathways (Couso et al., 2016).Interestingly, vip1-1 showed misregulation in carbon assimilation and partitioning, displaying irregular levels of tricarboxylic acid (TCA) cycle intermediates and an overaccumulation of storage lipids.This overaccumulation was exacerbated in the presence of Rap and under nitrogen starvation conditions, both of which downregulate TOR, further supporting an interaction between these two signalling pathways (Couso et al., 2016).However, more work is needed to understand this interaction, as PP-InsPs and TOR crosstalk has only been reported in Arabidopsis and Chlamydomonas (Couso et al., 2016;Van Leene et al., 2019).
In this study, we monitored the phosphorylation levels of RPS6 (a downstream target of TOR signalling) and the autophagy marker ATG8 in vip1-1 compared with wild-type after Rap treatment.These results indicated a positive downstream synergy of PP-InsPs and TOR kinase on the regulation of these two well known TOR targets.To further explore which processes are either shared or specifically regulated by these signalling pathways, we performed global/phosphoproteomic analysis of vip1-1 and wild-type before and after Rap treatment.Markedly, the proteomic analysis indicated differential abundance of proteins and decrease in phosphorylation of annotation terms related to photosynthesis between wild-type and vip1-1.These results led us to evaluate photosynthetic capacity in vip1-1 by measuring chlorophyll fluorescence and comparing InsPs levels in vip1-1 and photosynthetic deficient mutants under low and high light conditions.These data uncovered a novel relationship between TOR and PP-InsPs signalling compounds in governing photosystem II (PSII) and photoprotection that provide new insights in the study of photosynthetic control in the model green alga Chlamydomonas reinhardtii.

Cell culturing and rapamycin treatment
Chlamydomonas reinhardtii strain CC-1690 wild-type MT+ (Sager 21gr) (Sager, 1955) was used as parental strain to be compared with vip1-1.This strain was isolated in an insertional mutant screen in CC-1690 using the hygromycin resistance gene aph7 (Couso et al., 2016).All cultures were maintained on TAP (Tris acetate phosphate) agar plates and grown in 350 ml TAP liquid cultures at 25°C.Experiments were performed using five replicate cultures grown to exponential phase (1-2 9 10 6 cells ml À1 ) for control and Rap treatment, and quenched with cold 40% methanol stored at À80°C before harvesting by centrifuging at 4000 g for 5 min and discarding the supernatant as described in Werth et al. (2019).For rapamycin-treated (LC Laboratories, Woburn, MA, USA) cultures, the drug was added to a final concentration of 500 nM from 1 mM stocks in dimethyl sulfoxide (DMSO) for 15 min before harvesting.For control replicates, just drug vehicle (DMSO) without a chemical inhibitor was added to each replicate culture for 15 min before harvesting.

Proteomic analysis
Global protein extraction and phosphopeptide enrichment were performed using frozen pellets as described in Werth et al. (2019) (Fig. S2).Phosphopeptide samples were resuspended in 20 µl of 5% acetonitrile, 0.1% TFA while global samples were resuspended to a concentration of 0.2 µg µl À1 in 5% acetonitrile, 0.1% TFA.Global and enriched samples were analysed using an Acquity M-Class UPLC system (Waters, Milford, MA, USA) coupled to a Q-Exactive HF-X Hybrid Quadrupole Orbitrap mass spectrometer (ThermoFisher, Waltham, MA, USA).Raw data were processed using PROGENESIS QI for Proteomics (Nonlinear Dynamics; Waters) and parsed using custom R scripts (https://github.com/hickslab/QuantifyR)as described previously (Ford et al., 2019).The mass spectrometry proteomics data were deposited to the ProteomeXchange Consortium via the PRIDE partner repository (Vizca ıno et al., 2014(Vizca ıno et al., , 2016) ) and can be accessed with the Identifier PXD023085.Detailed methods are available in the Supporting Information (Methods S1) including immunoblot analysis.

Determination of chlorophyll fluorescence
Fluorescence of chlorophyll a was measured at room temperature using a pulse-amplitude modulation fluorometer (DUAL-PAM-100; Walz, Effeltrich, Germany) using mid-log phase cultures growing at control (50 µmol m À2 s À1 ) and high light (800 µmol m À2 s À1 ) conditions.Samples were normalised to 15 µg ml À1 Chla after extraction with 80% acetone (Finazzi et al., 1999).The maximum quantum yield of PSII was assayed after incubation of the algal suspensions in the dark for 15 min by calculating the ratio of the variable fluorescence, F v , to maximal fluorescence, F m (F v /F m ).The parameters Y(II) and nonphotochemical quenching (NPQ) were calculated using DUAL-PAM-100 software according to the equations in Kramer et al. (2004) and Klughammer & Schreiber (2008).Measurements of linear electron transport rates (ETR II) were based on chlorophyll fluorescence of dark-adapted samples applying stepwise increasing actinic light intensities up to 1250 lmol m À2 s À1 .Error bars indicate standard deviation (SD) of the values obtained from experiments performed in triplicate.Imaging-PAM M-series Maxi (Walz) was used to monitor F v /F m in Chlamydomonas under 50 and 800 lmol m À2 s À1 .

ATP levels
ATP was assessed via LC-MS/MS analysis and 10 mg of fresh weight powder was extracted with trichloroacetic acid (TCA; Thermo Fisher Scientific, Waltham, MA, USA) as described in Weiner et al. (1987).Recovery experiments were carried out by adding analyte standards of ATP (Merck KGaA, Darmstadt, Germany) to the frozen tissue before the extraction and the analysis was performed as described in Lunn et al. (2006).

InsPs extraction and analysis
For InsP 7 and InsP 8 extraction, 300 ml of mid-log phase culture per sample was collected at a cell density of 2 9 10 6 cells ml À1 .Samples were extracted as reported in Couso et al. (2016) and 1 µM 3-fluoro-InsP 3 (Enzo Life Sciences, Farmingdale, NY, USA) was used as internal standard for normalisation for relative quantification of the same InsPs species.
LC-MS/MS data were acquired using a Q-Exactive mass spectrometer (Thermo Fisher Scientific) equipped with a 1200 Capillary LC system (Agilent, Santa Clara, CA, USA) and a 0.5 9 150-mm 5-µm BioBasic AX Column (Thermo Fisher Scientific) using the conditions reported in Couso et al. (2016).Mean data and SD were calculated from three biological replicates, each of which had three technical replicates.

Results
TOR activity is misregulated in vip1-1 after rapamycin inhibition The insertional mutant vip1-1 is PP-InsPs deficient and displays hypersensitivity to TOR inhibition by Rap (Couso et al., 2016).To further investigate the regulation of TOR in the vip1-1 mutant, RPS6 phosphorylation was monitored over time in WT and vip1-1 cells treated with Rap.We previously demonstrated that phosphorylation of RPS6 on Ser245 is a readout of TOR activity in Chlamydomonas (Couso et al., 2020).The basal phosphorylation level of RPS6 Ser245 (P-RPS6) was similar in both strains.However, vip1-1 showed a significant decrease in P-RPS6 compared with WT (Fig. 1a,b) after 30 min of Rap treatment that was more pronounced after 60 min (Fig. 1b,c) indicating a faster de-phosphorylation in vip1-1 compared with WT.The P-RPS6/RPS6 levels were similar in WT and vip1-1 cells only after 90 min.
The detection of lipidated ATG8 (ATG8-PE) is an effective method to monitor autophagy; ATG8-PE accumulates under autophagy-inducing conditions including TOR inhibition (P erez-P erez et al., 2010).The levels of ATG8 and ATG8-PE were similar in WT and vip1-1 cells under control conditions (Fig. 1d).However, both ATG8 and ATG8-PE were more highly accumulated in vip1-1 after 30 min Rap treatment compared with WT (Fig. 1d), indicating a faster and stronger activation of autophagy in vip1-1.RPS6 was previously demonstrated to be rapidly turned over by autophagy in Chlamydomonas (Couso et al., 2018), therefore explaining the reduced abundance of this protein in vip1-1 after 60 min (Fig. 1b).
Gene ontology analysis in vip1-1 proteomics reveals an especial enrichment for PSII Quantitative proteomics were performed in WT and vip1-1 cells under control conditions and following 15 min of Rap treatment as previously reported in Werth et al. (2019), through which 2460 proteins were quantified (Table S1).No proteins significantly changed in abundance between control and Rap-treated conditions (Fig. S3A), confirming minimal protein turnover after 15 min Rap treatment.However, we observed significant differences in basal levels of proteins between the noninhibited samples for each strain (Fig. S3B).Despite having similar growth rates as the WT strain (Couso et al., 2016), 545 proteins from vip1-1 were differentially abundant compared with the parent strain, with 373 increased and 172 decreased (Table S1).
Quantitative phosphoproteomic analysis of vip1-1 and WT cells identified 3986 unique phosphorylated phosphopeptides, referred to as identifiers, derived from 1935 proteins (Table S2).Given the lack of significant changes in protein abundance within a given strain in the global proteomic analysis following Rap treatment (Fig. S3A; Table S1), changes in phosphopeptide abundance is likely to correspond to changes in phosphorylation state rather than changes in total protein abundance, enabling robust analysis of phosphorylation signalling pathways.Following Rap treatment, 1029 identifiers significantly changed in vip1-1, with 228 decreasing and 801 increasing (Fig. 2a), while 217 identifiers significantly changed in the parent strain, with 129 decreasing and 88 increasing (Fig. 2b).Comparison of the two strains yielded 1625 identifiers differentially abundant before Rap treatment (Fig. 2c) and 346 identifiers following treatment (Fig. 2d).
Gene ontology (GO) enrichment analysis (Ashburner et al., 2000) of the global proteomic dataset of untreated cultures revealed that vip1-1 was enriched over WT in biological functions related to stress responses, including protein folding, photosystem II (PSII) repair, cellular response to oxidative stress, and protein refolding (Fig. S4A).By contrast, vip1-1 was deficient in GO terms related to cellular metabolism, including the tricarboxylic acid cycle and electron transport in PSII, among others (Fig. S4B).
Gene ontology analysis of significantly changed phosphopeptides in vip1-1 following Rap treatment uncovered an enrichment of identifiers involved in RNA processing as well as chromatin and DNA binding (Fig. 3a).By contrast, identifiers significantly decreasing in vip1-1 under the same treatment were enriched in photosynthesis-related GO terms, including PSII assembly, PSII stabilisation, and the PSII oxygen evolving complex (Fig. 3b), indicating an important role of PP-InsPs in the regulation of photosynthetic-related processes in Chlamydomonas that has not been reported in green organisms therefore far.Notably, enrichment of photosynthesis-related proteins was not detected in WT (Fig. S5).

Differential phosphorylation of known and putative TOR substrates are found in vip1-1
In total, 48 phosphorylated identifiers from 22 proteins with homology to known TOR signalling-related proteins were identified in this study (Table S3).Under Rap treatment, 11 of these identifiers significantly increased and two significantly decreased in vip1-1 while one identifier significantly increased and four significantly decreased in the parent strain, with no overlap between strains (Table S3).One of these identifiers was an uncharacterised phosphosite, S2598 on TOR (Cre90.g400553.t1.1), that was significantly increased in the vip1-1 mutant following Rap treatment (log 2 FC: 2.31) but did not change in the parent strain (Table S3).

Phosphorylation of PSII core components are downstream PP-InsPs and TOR signalling
In this study, global protein analysis uncovered 155 proteins related to photosynthesis, photorepair, and chlorophyll biosynthesis
We performed PAM fluorometry under a light induction curve to further investigate PSII defects in vip1-1.F v /F m was decreased in both the vip1-1 and WT cells treated with Rap (Table 1) similar to the results reported using the TOR inhibitor AZD 8055 (Ford et al., 2019;Upadhyaya & Rao, 2019).However, we observed that the electron transfer rate of PSII (ETRII) in vip1-1 was significantly decreased compared with WT after 600 µmol m À2 s À1 , with the ETRII of vip1-1 comprising only 12-19% of the rate of the WT (Fig. 4b).Furthermore, at high fluences Rap treatment decreased ETRII of WT that shows similar levels than vip1-1 cells, which did not change after Rap addition (Fig. 4c).This result suggests that PP-InsPs are involved in the maintenance of electron transfer during high light stress and work constructively with TOR to provide positive regulation of the photosynthetic apparatus under noninhibited conditions, as Rap treatment results in the same decreased ETRII in WT as it is observed in control conditions of vip1-1.In addition, the abundance of several photosynthetic repair-specific proteins were upregulated in vip1-1, including vesicle-inducing protein in plastids 1 (VIPP1, Cre13.g583550.t1.2; log 2 FC 1.62) (Table S4), a multifunctional protein involved in the maintenance of photosystems (Nordhues et al., 2012;Theis & Schroda, 2016;Gupta et al., 2021), and chloroplast DNAJ-like protein 2 (CDJ2, Cre07.g316050.t1.2; log 2 FC 3.54) (Table S4), which interacts with VIPP1 to regulate thylakoid biogenesis (Liu et al., 2005).
Cyclic electron flow is activated after high light stress and rewires energy to increase ATP yield.We identified significant increases in two ATP synthase subunits of vip1-1 (Table S1): subunit E (gi|4117902|ref|NP_958379.1; log 2 FC 1.67), which forms the connection between the lumal and stromal hemispheres, and subunit II (Cre11.g481450.t1.2; log 2 FC 1.37), which regulates ATP synthesis based on the proton gradient across the thylakoid membrane (Lemaire & Wollman, 1989;Richter et al., 2000;Hahn et al., 2018).To investigate the effect of the higher abundance of these ATP synthase subunits in the vip1-1 mutant, we determined the level of ATP in WT, vip1-1 and complemented cells.Our results indicated that ATP was highly increased in the vip1-1 mutant (Fig. 5c).While the activation of CEF seems to be dependent on the coordination of PP-InsPs and TOR controlling PGRL1 phosphorylation, PP-InsPs must also be independently controlling the levels of ATP as no significant changes were seen following Rap treatment (Table S1).

Nonphotochemical quenching and PP-InsPs biosynthesis are likely to be connected by a feedback loop
In Chlamydomonas, LHCSR3 (Cre08.g367400.t1.1) mediates the induction of NPQ (Peers et al., 2009) and can work together or independently with PGRL1 during high light (HL) acclimation (Chaux et al., 2017).By contrast with the observed downregulation of CrPGRL1-S50, three identifiers of LHCSR3 (Cre08.g367400.t1.1; -S165, -T161, -Y170) (Table S5) were upregulated in vip1-1 (log 2 FC 1.48, 1.70 and 2.67, respectively).The phosphorylation of two of these phosphosites was decreased following Rap exposure, while Y170 remained upregulated in the mutant under the same conditions (Table S5).To investigate this further, we compared NPQ in WT, vip1-1 and a complemented line using light response curves and PAM fluorometry in the presence or absence of Rap (Fig. 5d).NPQ was highly downregulated in the vip1-1 mutant curve compared with the other two strains, reaching a difference of 72% under very HL (1250 µmol m À2 s À1 ) (Fig. 5d).Although, WT and the complemented line keep NPQ levels above 0.4 after reaching HL (750 µmol m À2 s À1 ), vip1-1 further downregulates NPQ reaching 0.16, indicating that NPQ is not supported by vip1-1 photosynthetic machinery under HL.In the presence of Rap, NPQ did not change significantly in any of the strains compared with control conditions.
Additionally, F v /F m and Y(II) were analysed in WT and vip1-1 cells under control light (50 µmol m À2 s À1 ), HL (800µmol m À2 s À1 ) and Rap (control light conditions) conditions (Fig. S6; Table 1).Despite few differences in F v /F m under control conditions, vip1-1 showed significantly reduced F v /F m and Y(II) values when subjected to HL, further indicating a misregulation of the light stress compensation mechanisms (Fig. S6A,B; Table 1).

Research
New Phytologist InsP 8 were also monitored following Rap treatment (30 min) in all strains.There were no significant differences in NPQ after Rap treatment in any of these strains compared with control conditions (Fig. 6a).After Rap treatment, InsP 7 an InsP 8 levels in npq2 were comparable to the level detected under control light conditions (Fig. 6b,c).These data indicate a novel connection between PP-InsPs levels and NPQ that may control HL stress compensation in Chlamydomonas that acted independently of TOR.

Pyro-inositol polyphosphates modulate TOR activity in Chlamydomonas
The macrolide rapamycin partially arrests cell growth in Chlamydomonas (Crespo et al., 2005;Couso et al., 2016), suggesting that it does not completely inhibit TOR activity.This is conserved in mammals (Thoreen et al., 2009), in which differential sensitivity of mTORC1 phosphorylation sites to Rap has been reported as well (Kang et al., 2013).We isolated vip1-1 in a genetic screen due to its hypersensitivity to Rap compared with WT and have demonstrated a genetic link between pathways by restoring this phenotype in the vip1fkbp12 double mutant (Couso et al., 2016).In this study, we found that vip1-1 accelerates TOR inactivation under Rap conditions (Fig. 1a-c), indicating that these signalling pathways are likely to be controlling similar downstream targets either coordinately and/or independently of each other.
InsP 7 is a phosphate donor for different nucleolar proteins in yeast (NSR1 and SRP40) and mammals (Nopp140 and TCOF1) (Saiardi, 2004(Saiardi, , 2012b(Saiardi, , 2016;;Bhandari et al., 2007).The InsP 7derived phosphate is thought to be attached to a pre-existing phosphorylation (b-pyrophosphorylation), which may provide a unique mode of signalling to proteins.However, the impact of PP-InsPs on phosphorus-signalling networks in photosynthetic organisms has not yet been reported.We used differential analysis of global proteomics and phosphoproteomics in WT and vip1-1 in the presence or absence of Rap to reveal both the overlap with the TOR signalling network as well as the mechanistic connection between PP-InsPs and PTMs that has not been reported in green organisms therefore far.
While the number of identifiers shown to significantly change in phosphorylation in WT cells following Rap treatment mirrored previous studies (Fig. 2b) (Werth et al., 2019), vip1-1 yielded a larger change in phosphopeptide abundance following TOR inhibition than the WT (Fig. 2a).However, the number of identifiers found to be differentially phosphorylated was significantly lower when comparing WT and vip1-1 before (1625 identifiers) and after (346 identifiers) TOR inhibition (Fig. 2c,d).These results strongly suggested that TOR kinase and PP-InsPs operated in the same signalling cascade in agreement with their previously observed genetic interaction (Couso et al., 2016).We cannot disregard that they could directly interact, as we found an upregulation of S2598 on TOR in vip1-1 after Rap treatment (Table S3) when TOR activity is faster inhibited in the mutant (Fig. 1a-c).Previously, this phosphosite was identified in Chlamydomonas WT, but no change was detected in response to different TOR inhibitors (AZD8055, Torin 1 and Rap) (Werth et al., 2019).Our results suggest TOR phosphorylation at S2598 may be regulated by PP-InsPs, although the impact on TOR activity is still unknown.
We also identified a novel connection between PP-InsPs and established processes under TOR control (Table S3).Autophagy is inhibited by active TOR signalling in diverse eukaryotes including algae and plants (D ıaz-Troya et al., 2008;Yu et al., 2018).Recently, VIP1 has also been connected to the regulation of autophagy by modulating the level of ATG proteins in Candida albicans (Ma et al., 2020).We found an overaccumulation of lipidated ATG8, therefore indicating an overactivation of the recycling process of autophagy in vip1-1 under Rap conditions (P erez-P erez et al., 2010) (Fig. 1d).We also found an ATG11 phosphosite (ATG11-S1176) (Table S3) that was significantly downregulated in vip1-1 under control conditions.ATG11 is a well known scaffold protein that interacts with phosphorylated ATG29 or ATG32 in yeasts to induce mitophagy and the organisation of the phagophore assembly site (Aoki et al., 2011;Mao et al., 2013) and encourages starvation-induced phosphorylation of ATG1 in Arabidopsis, a downstream TOR target (Li et al., 2014).Taking these results together, we conclude that autophagy is regulated coordinately with TOR and PP-InsPs signalling pathways in Chlamydomonas, but that PP-InsPs may also act independently of TOR to potentiate this recycling process.
LARP1 proteins are direct effectors of mTORC1 in mediating mRNA translation (Thoreen et al., 2012) and are found in many eukaryotes (Deragon & Bousquet-Antonelli, 2015).However, none of the significantly changing phosphorylated identifiers found in Chlamydomonas LARP1 (Table S3) was conserved in yeasts or humans.AtLARP1 showed significantly decreased phosphorylation on S644 and S649 following TOR inhibition (Van Leene et al., 2019); the former is conserved in Chlamydomonas (LARP1-S810) and identified as a phosphosite in Werth et al. (2017), but did not significantly change following TOR inhibition (Werth et al., 2019).However, LARP1-S817 phosphorylation decreased following TOR inhibition in the same study (Werth et al., 2019), which was also observed here in WT after Rap inhibition however no significant differences were seen in vip1-1 (Table S3).Instead, three different phosphosites were significantly decreased in vip1-1 compared with WT (LARP1-S529, -T668 and -S670) and only -S670 was upregulated after Rap treatment (Table S3).These results suggest that PP-InsPs act partly through the TOR signalling cascade and partly through TORC1-independent mechanisms to effect LARP1 phosphorylation.

PP-InsPs and TOR control phosphorylation of photosynthetic apparatus
Gene ontology analysis revealed an unexpected enrichment of photosynthetic targets (Fig. 3) that were not described in previous studies reporting protein phosphorylation patterns under TOR inhibition in either Chlamydomonas or Arabidopsis (Roustan & Weckwerth, 2018;Van Leene et al., 2019;Werth et al., 2019).
However, proteomic analysis of reversible cysteine oxidation, a PTM known to crosstalk with phosphorylation, indicated that photosynthesis is regulated by TOR in Chlamydomonas (Ford et al., 2019).Similarly, our data here showed significant differences in photosynthesis-related proteins in global and phosphoproteomic data between WT and vip1-1 under control conditions (Tables S4, S5).These results indicated reduced levels of the catalytic subunits D1 and D2 and reaction centres psbC (CP43) and psbB (CP47).Also, psbC-S456 and psbB-T501 were highly downregulated in the mutant under control conditions but not after Rap treatment (Table S5).D1, D2 and psbC are primarily phosphorylated in response to light stress and endogenous circadian rhythm at their N-terminal threonine residues (Elich et al., 1992;Booij-James et al., 2002) by the Ser/Thr kinase STATE TRANSITION8 (STN8) (Bonardi et al., 2005;Vainonen et al., 2005;Rochaix et al., 2012).The Chlamydomonas paralogue of STN8, STL1 (Cre12.g483650.t1.2) showed a significant downregulation on T126 in vip1-1 that was compensated in the presence of Rap (Table S5).Although this kinase is regulated by the redox state of the PQ pool (Bennett, 1991;Fristedt et al., 2009), it has also been reported to be subjected to phosphorylation (Reiland et al., 2011).Additionally, STL1-T126 was previously reported in Chlamydomonas (Bergner et al., 2015).The phosphorylation of PSII core proteins is part of the PSII repair cycle that proceeds before the proteolytic degradation of damaged D1 protein, preventing its degradation by proteases (Koivuniemi et al., 1995).Additionally, STN8 is thought to control the transitions from linear to CEF by controlling the phosphorylation of PGRL1 in Arabidopsis (Reiland et al., 2011).Our data could indicate that the downregulation of D1 found in vip1-1 is a consequence of the differential phosphorylated identifiers in STL1 and core components found in this mutant and that PP-InsPs and TOR act coordinately upstream of this process controlling PSII photochemistry, possibly facilitating the transition between linear and CEF under high irradiances.
We also determined the light harvesting antenna protein LHCBM5 (Cre03.g156900.t1.2) has decreased abundance in vip1-1.Although LHCBM5 has no homologue in plants, it has been suggested that this protein plays a similar role to the CP24 protein in Arabidopsis (Takahashi et al., 2006).CP24-deficient plants displayed altered kinetics of state transitions (Kov acs et al., 2006).The downregulation of LHCBM5 detected in the vip1-1 mutant could link PP-InsPs to state transitions in Chlamydomonas as LHCBM5 is the more abundant LHCII (type II) polypeptide found in PSI-LHCI under state 2 (Takahashi et al., 2006).
Phosphorylation of LHCB4 and LHCB5 (CP26) is considered fundamental for the detachment of LHCBM polypeptides from PSII during the transition from state 1 to state 2 (Iwai et al., 2008).Three identifiers belonging to LHCB4 (T11, T17 and S39) were significantly downregulated in the mutant under control conditions.LHCB4-T7 and T11 have been previously found to be connected with the phosphorylation state of LHCSR3 that is involved in photoprotective NPQ in Chlamydomonas (Scholz et al., 2019).Another identifier found in this study, LHCB4-S103 (Table S5), has been previously linked to kinase STT7 activity in this alga (Bergner et al., 2015), but we could not find any significant difference in the conditions tested.We also found four identifiers in LHCB5 (T10, -S63, -T187, -S202) (Table S5), although S63 was the only one significantly changing in vip1-1 under control conditions.The upregulation of S63 was then prevented after Rap treatment indicating that TOR and PP-InsPs share this target.Also, LHCB5-T10 has previously been identified after mapping in vivo phosphorylation sites in integral and peripheral membrane proteins (Vener, 2007).These data suggest that PP-InsPs and TOR control state transitions at different levels, with LHCBM5 protein abundance and the phosphorylation state of LHCB4 and LHCB5 influencing the transition to state 2 when light compensation mechanisms are required.Additionally, repair-specific proteins (VIPP1 and CDJ2) were upregulated in vip1-1 mutant.In Chlamydomonas, vipp1 amiRNA knockdown strains are sensitive to HL, which is likely to be due to structural defects in PSII (Nordhues et al., 2012).The upregulation of this protein in vip1-1 suggests that PSII repair mechanisms are more active in this mutant, and are likely to be as a consequence of PSII malfunction.This is supported by decreased F v /F m (Table 1) and ETR II (Fig. 4b) values detected in the mutant.

PP-InsPs and TOR are involved in the maintenance of cell energy levels
In Arabidopsis, PGRL1-T62-T63 has been reported to be a possible target of STN8 kinase (Reiland et al., 2011).We found one identifier in PGRL1 (S50) to be downregulated in vip1-1 under control conditions and in WT after Rap treatment.The levels of PGRL1-S50 were alleviated in the presence of Rap, but only in the mutant, therefore indicating a fine modulation mediated by PP-InsPs and TOR that correlates to the tight control of the initiation of CEF mediated by PGRL1 (Johnson et al., 2014).
Under stress, CEF provides ATP for CO 2 fixation (Lucker & Kramer, 2013), balances overreduction of PSI, and readjusts the ATP poise, leading to increased lumen acidification that is important for photoprotection (Alric, 2010;Peltier et al., 2010;Leister & Shikanai, 2013;Shikanai, 2014).ATP synthase subunits E and II are increased in vip1-1 (Table S1) that corresponds with the high ATP levels found in this mutant (Fig. 5c).Although PP-InsPs have been previously linked to the control of intracellular ATP in yeast kcs1D mutants (Szijgyarto et al., 2011), vip1 mutants have not previously been directly connected with this phenotype.While no IP6K homologue is found in algae or plants, the detection of InsP 8 (1,5(PP) 2 -InsP 4 ) suggests the presence of a functional IP6K enzyme (Desai et al., 2014;Laha et al., 2015;Couso et al., 2016), or a noncanonical ITPK function (Cridland & Gillaspy, 2020) that could be regulating ATP levels in coordination with VIP1.Recently, VIP1 was also reported to have a bifunctional kinase/pyrophosphatase activity that produces and destroys 1-PP-InsPs at the expense of consuming ATP in yeast (Dollins et al., 2020).This could contribute to the increase in ATP observed in the Chlamydomonas vip1-1 mutant.Additionally, mTOR is a homeostatic ATP sensor that adjusts ribosome biogenesis to ATP intracellular levels (Dennis, 2001).

Research
New Phytologist Although vip1-1 displays higher ATP levels than WT, we did not detect any difference in TOR activity before Rap treatment, suggesting that this activation pathway may not be conserved in this alga.

NPQ, PP-InsPs and TOR coordinate to protect cells to excessive irradiance
The photoprotective process NPQ is activated in almost all photosynthetic organisms in PSII antenna to dissipate excess light as heat (Ruban, 2016).NPQ is catalysed by Light Harvesting Complex Stress Related (LHCSR) subunits, with a major role observed for LHCSR3 (Girolomoni et al., 2019).We found three identifiers of LHCSR3 (-S165, -T161, -Y170) (Table S5) increased in vip1-1 and only Y170 remained upregulated in the presence of Rap.LHCSR3 phosphorylation has also been reported as nonessential for NPQ activation in Chlamydomonas (Bonente et al., 2011).However, that work compared phosphorylated with dephosphorylated LHCSR3 that contrasts with the over-phosphorylated identifiers found in vip1-1.Protein levels of LHCSR3 have previously been connected to the regulation of NPQ (Peers et al., 2009) but we could not detect any significant difference in vip1-1 compared with WT (Table S4), suggesting that PP-InsPs and possibly TOR are regulating NPQ at posttranslational level.LHCSR3 phosphorylation at the N-terminus (-S26, -S28-T32 and -T33) has been reported to operate as a molecular switch modulating LHCB4 phosphorylation, which in turn is important for PSII-LHCII disassembly before state transitions (Bergner et al., 2015;Scholz et al., 2019).Also, Arabidopsis koLhcb4 mutants present lower activation of NPQ (de Bianchi et al., 2011).These results indicated a tight control of photoprotective mechanisms mediated by PP-InsPs that act independently (-Y170) and coordinately with TOR (-S165, -T161) over the phosphorylation of LHCSR3.We also need to consider that the deregulation of LHCB4 (CP29) phosphorylation in vip1-1 is suggesting that NPQ is finely controlled by PP-InsPs at different levels in the photosynthetic machinery.In vivo measurements of NPQ levels in the WT, vip1-1, and complemented line revealed highly deceased NPQ levels after HL in vip1-1 that were not observed following Rap treatment (Figs 5d, 6a).Although NPQ does not seem to be affected by TOR inhibition in any of the strains, we cannot disregard that LHSR3-S165 and -T161 levels were compensated after Rap treatment (Table S5).This is another example of the multiplex regulation mediated by PP-InsPs and TOR kinase over the same targets that are mechanistically difficult to delineate.
Nonphotochemical quenching, InsP 7 , and InsP 8 levels were also compared among WT, vip1-1 and the NPQ defective mutant npq2.We found an important decrease of PP-InsPs in npq2 that does not respond to Rap treatment (Fig. 6b,c).These data indicated a possible feedback loop in the regulation of NPQ and PP-InsPs biosynthesis that is independent of TOR signalling (Fig. 6b,c).We also found that PP-InsPs levels and NPQ were highly different from WT levels under HL conditions, indicating that this feedback regulation may be especially relevant under stress conditions where NPQ is activated.

Conclusion
Overall, our data indicated a strong relationship between TOR kinase and VIP1/PP-InsPs that impacted autophagy and TOR activity in Chlamydomonas.We have uncovered that PP-InsPs share common targets with TOR in controlling photosynthetic machinery and compensation mechanisms including state transitions and CEF (Fig. 7).We have also identified PP-InsPs as key components of the signal transduction machinery that can act independently of TOR controlling NPQ under high irradiance or energy levels (Fig. 7).We have begun unravelling the influence of these essential signalling compounds in Chlamydomonas' PTMs; however, several questions still remain.Future work should address the role of PP-InsP signalling in specific cell compartments as well as the conditions in which PP-InsPs signal transduction act independently and/or coordinately with TOR signalling over different targets to maintain cell homeostasis.

Fig. 1
Fig. 1 Immunoblot analysis of P-RPS6/RPS6 as a readout of target of rapamycin kinase activity in Chlamydomonas WT (a) and vip1-1 (b) in the presence of rapamycin (Rap) along a time course.Relative quantitation of P-RPS6/RPS6 was made using three biological replicates.Error bars represent standard deviation of the mean values.Asterisks represent significant differences (P < 0.05) evaluated using Student's t-test.(c).Immunoblot analysis of ATG8 in the presence or absence of rapamycin for 30 min.FKBP12 was used as loading control (d).

Fig. 2
Fig. 2 Phosphoproteomic data represented in volcano plots of two-tailed equal variance t-tests between each Chlamydomonas strain with or without rapamycin (Rap) treatment (a, c) and between strains (b, d).

Fig. 3
Fig. 3 Chlamydomonas vip1-1 mutant phosphoproteomic gene ontology (GO) analysis.(a) Count of the number of proteins in the top five biological process (black), cellular component (green) and molecular function (purple) GO terms with a foldchange enrichment of at least 1.5 from identifiers significantly more abundant in vip1-1 with rapamycin treatment.Cells are shaded to reflect fold-change for each GO term.(b) Count of the number of proteins in the top five biological process (black), cellular component (green) and molecular function (purple) GO terms with a fold-change enrichment of at least 1.5 from identifiers significantly less abundant in vip1-1 with rapamycin treatment.Cells are shaded to reflect fold-change for each GO term.

Fig. 4
Fig. 4 (a) Differential phosphorylation of the photosynthetic apparatus in the vip1-1 mutant compared with the Chlamydomonas parent strain.Proteins coloured green show proteomic coverage in the dataset while proteins coloured grey do not.Each significantly changing phosphosite was localised on a unique phosphopeptide.Nontransformed fold changes are reported.Pink arrows represent significant upregulated phosphorylation of the indicated identifiers while blue arrows represent significant downregulated phosphorylation of the indicated identifiers (b) Electron transfer rate of PSII (ETR II) measured in a light induction curve using photosynthetically active radiation (PAR) from 0 to 1250 µmol m À2 s À1 under control conditions and rapamycin treatment (c) in WT, vip1-1 and the complemented line (vip1-1:VIP1).Errors bars indicate standard deviation of the mean values from three biological and three technical replicates.

Fig
Fig.5(a) Immunoblot analysis of LHCBM5 under control conditions and rapamycin treatment (30 min) in Chlamydomonas wild-type (WT), vip1-1 and the complemented line (vip1-1:VIP1).(b) Quantitation of LHCBM5 was made using three biological replicates.Error bars represent standard deviation (SD) of the mean values.Asterisks represent significant differences (P < 0.05) evaluated using Student's t-test.(c) Quantitation of ATP in WT and vip1-1 mutant under control and rapamycin conditions.Errors bars represent SD of the mean values from three biological replicates.Asterisks represent significant differences (P < 0.05) evaluated using Student's t-test.(d) Nonphotochemical quenching (NPQ) was measured in WT, vip1-1 and the complemented line (vip1-1:VIP1) on an induction curve using photosynthetically active radiation (PAR) from 0 to 1250 µmol m À2 s À1 in the presence or absence of 500 nM rapamycin (Rap).Errors bars indicate SD of the mean values from three biological and three technical replicates.

Fig. 6
Fig.6Nonphotochemical quenching (NPQ) monitoring and pyro-inositol polyphosphates analysis in Chlamydomonas WT, vip1-1 and the npq2 mutants.(a) NPQ was measured on a light induction curve using photosynthetically active radiation (PAR) from 0 to 1250 µmol m À2 s À1 in the presence or absence of 500 nM rapamycin (Rap).Errors bars indicate standard deviation (SD) of the mean values from three biological and three technical replicates.Inositol polyphosphate-7 (InsP 7 ) (b) and inositol polyphosphate-8 (InsP 8 ) (c) were analysed by liquid chromatographymass spectrometry (LC-MS) in the three strains under control (50 µmol m À2 s À1 ), high light (HL) (800 µmol m À2 s À1 ) and in the presence of 500 nM Rap at light control conditions.Bar graphs scaled in arbitrary units (AU) and normalised with a standard, 3-fluoro-InsP 3 , showing relative levels of inositol polyphosphates (InsPs) species extracted from indicated strains and measured using mass spectrometry.This allows relative quantification of the same InsPs species.Error bars indicate SD from at least three biological replicates.Asterisks represent significant differences (P < 0.05) evaluated using Student's t-test.

Fig. 7
Fig.7Summary figure of proposed relationships between target of rapamycin (TOR), and pyro-inositol polyphosphates (PP-InsPs), inositol polyphosphate-7 (InsP 7 ) and inositol polyphosphate-8 (InsP 8 ) produced by VIP1 and processes that are under their influence found in this study.PP-InsPs/VIP1 can directly regulate high light stress and energy levels (on the left) but they also coordinate with TOR (on the right) in regulating photosynthesis and previously known TOR targets such as autophagy.Dashed arrow indicates that PP-InsPs/VIP1 and TOR have a complex interaction that rebound TOR kinase activity and InsPs biosynthesis.Under favourable conditions, VIP is proposed to stimulate TOR kinase activity and TOR is proposed to stimulate PP-InsPs biosynthesis catalysed by VIP.Our data also indicate that complex feedback loops may be regulating PP-InsPs outside TOR signalling such as in nonphotochemical quenching (NPQ).

Table 1 F
v /F m in Chlamydomonas reinhardtii WT, vip1-1 and complemented line under different conditions.