The thylakoid lumen Deg1 protease affects non-photochemical quenching via the levels of violaxanthin de-epoxidase and PsbS

1 2 Non-photochemical quenching (NPQ), the dissipation of excess light energy as heat, has been 3 long recognized as a major protective mechanism that minimizes the potential for oxidative 4 damage to photosystem II (PSII) reaction centers. Two major positive contributors to NPQ are 5 the carotenoid zeaxanthin, generated from violaxanthin by the enzyme violaxanthin de -6 epoxidase (VDE or NPQ1), and the thylakoid protein PsbS (NPQ4). The involvement of the 7 lumenal Deg proteases in the repair of PSII from photoinhibition prompted us to further explore 8 their possible role in other responses of Arabidopsis thaliana to high light. Here we show that 9 upon exposure to high light, the single deg1 and the triple deg158 mutants display different 10 levels and kinetics of NPQ, compared to the deg58 mutant and WT that behave alike . In 11 response to high light, the two genotypes lacking Deg1 over -accumulate NPQ1 and NPQ4. 12 After temporal inhibition of protein translation in vivo, the level of these two proteins in deg1 13 is higher than in WT . Together, the results suggest that Deg1 represents a new level of 14 regulation of the NPQ process through adjusting the quantity of NPQ1 and NPQ4 proteins , 15 probably through their proteolysis. 16 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 3

17 18 Light energy absorbed by photosynthetic pigments and transferred to photosynthetic reaction 19 centers (RCs) drives photosynthesis and hence life on Earth. Only a fraction of the absorbed 20 light is utilized for photochemistry, while the rest is released mostly as h eat, and also as 21 fluorescence. Chlorophyll fluorescence has been traditionally used as a non -destructive tool 22 for monitoring different parameters of photosynthesis (Maxwell and Johnson, 2000) . 23 Accordingly, photochemistry, the conversion of light energy to chemical one, occurring in 24 RCs, has been referred to as ‘photochemical quenching’ of chlorophyll fluorescence, whereas 25 the excess energy dissipated as heat was coined ‘non-photochemical quenching’ (NPQ) (Ruban 26 et al., 2012; Ruban, 2016) . Under natural conditions, the relative amount of energy driving 27 photosynthetic electron transfer (ETR) and that released as heat vary. In fact, channeling excess 28 excitation energy to NPQ has been considered a protective mechanism that minimize s the 29 potential for oxidative damage to RCs, mostly to those of photosystem II (PSII) (Ruban et al., 30 2012; Nicol et al., 2019). Such damage, the phenomenon known as ‘photoinhibition’ (Powles, 31 1984; Adir et al., 2003) , leads to decrease d photosynthetic activity , which can affect plant 32 growth and fitness. 33 NPQ is triggered upon acidification of the thylakoid lumen during photosynthetic ETR. 34 Two major components contribute to NPQ in higher plants: the carotenoid zeaxanthin and the 35 protein PsbS. The precise mechanism of how zeaxanthin mediates NPQ is not clear. However, 36 its level is determined by the balance between the activity of two enzymes of the xanthophyll 37 cycle: violaxanthin de -epoxidase (VDE, also known as NPQ1) that converts violaxanthin to 38 zeaxanthin, and zeaxanthin epoxidase (ZEP or NPQ2) , which mediates the opposite reaction 39 (Niyogi et al., 1998; Jahns and Holzwarth, 2012) . PsbS (also known as NPQ4) is a thylakoid 40 membrane protein that does not bind pigments (Li et al., 2000; Li et al., 2002) . Its activation 41 involves light-induced monomerization , that is accompanied by reorganization of the P SII 42 antenna and binding to LHCII trimers (Correa-Galvis et al., 2016). 43 Despite NPQ and other protective mechanism s, the PSII reaction center and especially 44 its D1 protein are constantly prone to oxidative damage that leads to photoinhibition. 45 Concomitant with this, a repair cycle that involves de -novo synthesis of D1 and other 46 components of PSII operates in chloroplasts . Inherent to this cycle is the proteolytic 47 degradation of oxidatively damaged proteins, which is carried out cooperatively by the 48 thylakoid FtsH protease complex (Lindahl et al., 2000; Bailey et al., 2002; Zaltsm an et al., 49 2005; Kato et al., 2012), facing the stromal side of the membrane , and the lumenal Deg1 and 50 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 4 Deg5-Deg8 proteases (Kapri-Pardes et al., 2007; Sun et al., 2007; Kley et al., 2011) . The 51 apparently redundant function of Deg1 and Deg5 -Deg8 was investigated through analysis of 52 single, double and triple deg mutants, to uncover the physiological dominance of the former 53 due to higher abundance and more potent proteolytic activity (Butenko et al., 2018). Thus, all 54 mutants lacking Deg1 were smaller and more sensitive to stresses , whereas plants missing 55 Deg5-Deg8 were very similar to WT plants under all tested conditions. 56 The involvement of the lumenal Deg proteases in the repair of PSII from photoinhibition 57 prompted us to further explore their possible role in other responses of Arabidopsis thaliana to 58 high light. In this study, we show that the single deg1 and the triple deg158 mutants exhibit 59 different levels and kinetics of NPQ upon exposure to high light, compared to the deg58 mutant 60 and WT that behaved alike. Specifically, upon exposure to high light, they had higher levels of 61 NPQ and slower recovery . Comparative proteomics revealed that the lack of either Deg1 or 62 Deg5-Deg8 complexes was not compensated for by upregulation of the other. However, the 63 two genotypes lacking Deg1 over-accumulated NPQ1 and NPQ4 upon exposure to high light. 64 After temporal inhibition of protein translation in vivo, the level of these two proteins in deg1 65 was higher than in WT. Taken together, the results suggest that Deg1 is involved in the 66 regulation of the NPQ process through an effect on the level of NPQ1 and NPQ4 proteins. 67 68

69 70 Effect of high light on photosynthesis parameters in Deg protease mutants. 71 72 The well -established role of Deg proteases in the response of Arabidopsis plants to 73 photoinhibition (Kapri-Pardes et al., 2007; Sun et al., 2007; Kley et al., 2011; Kato et al., 2012) 74 prompted us to explore possible additional roles of Deg in the response to high light (HL). To 75 this end, we grew WT and Deg mutant plants, lacking either Deg1 (deg1), Deg5-Deg8 (deg58) 76 proteases or both (deg158) in short days, under normal light (NL; ~75 µmol photons m-2 s-1). 77 Five-weeks old plants were transferred to HL (~750 µmol photons m-2 s-1) for 1-8 h, or allowed 78 to recover for additional 24 h in NL, and subjected to pulse-amplitude-modulation (PAM) 79 measurements at each of these time points . The analyzed plants and their PAM images are 80 provided in Supp. Fig. 1A. Consistent with our previous work (Butenko et al., 2018), prior to 81 exposure to HL, deg1 and deg158 demonstrated a lower maximum efficiency of PSII (Fv/Fm) 82 compared to WT and deg58 (Supp. Fig. 1B). Upon exposure to HL, all four genotypes showed 83 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 5 a continuous decrease in Fv/Fm values. However, the decrease in deg1 and deg158 was more 84 pronounced, indicating that they were more sensitive to HL than WT and deg58. After 24 h of 85 recovery under NL , all genotypes regained their original maximum efficiency of PSII. 86 Consistent with this were the light response curves of photosynthetic electron transport rates 87 (ETR) (Supp. Fig. 2), which also correlated to plant size of the four genotypes (Supp. Fig. 1A). 88 Induction and relax ation curves of NPQ measured in the same plants provided some 89 interesting insights. As shown in Fig. 1, NPQ levels in deg1 and deg158 plants were generally 90 higher compared to WT and deg58 plants. The rate of NPQ induction in the Deg1 -lacking 91 genotypes appeared to be faster, and the relaxation slower , than in th e genotypes containing 92 Deg1. Another interesting observation was the relation between the length of HL exposure and 93 NPQ kinetics. In WT and deg58, the longer the exposure to HL was, the lower the NPQ levels 94 were, up to 4 h of HL, suggesting an adaptation to the stress condition. Exposure to HL for 8 h 95 did not result in further decrease in these two genotypes . This behavior is consistent with a 96 previous report showing a lower contribution of NPQ und er prolonged exposure to high light 97 (Saccon et al., 2022). In contrast, no decrease in NPQ was observed in deg1 and deg158 plants. 98 During 1-4 h of HL, the kinetics of NPQ remained the same. Only after exposure to 8 h of HL, 99 the values of NPQ increased in plants lacking Deg1. After 24 h recovery under NL, the capacity 100 for NPQ was restored in WT and deg58, although with somewhat different kinetics than before 101 recovery. In deg1 and deg158 plants, the values of NPQ further increased after the recovery, 102 to values well above those observed at the beginning of the experiment (Fig. 1). These results 103 imply that Deg1 might somehow be involved in the regulation of NPQ. 104 105 Effect of HL on the proteome of WT and Deg protease mutants. 106 107 To further explore the effect of HL on WT and Deg protease mutants, we carried out a 108 comparative proteomic analysis on the same set of four genotypes, at six different time points 109 along the exposure to HL and following recovery at NL, with four biological replicates each. 110 Mass-spectrometry (MS) analysis of the 96 samples allowed identification and quantification 111 of 40,546 peptides, mapped to 4,188 different proteins (Supp. Table 1). Of these, ~1,400 were 112 predicted to be chloroplast proteins. Assuming that detectability is proportional to protein 113 abundance, and that chloroplast proteins account for 10 -15% of the plant proteome, these 114 numbers indicate that chloroplast proteins are more abundant than those located in other 115 compartments of the leaf cells. More than half of the chloroplast proteins were identified, 116 compared to only ~1/5 of the proteins residing in other compartments . Volcano plots of the 117 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 6 data obtained from plants prior to exposure to HL revealed the extent of alterations in protein 118 abundance only due to knocking out the Deg proteases. Cutoffs for significance were 2 -fold 119 change in the level of a given protein and p-value <0.05 in comparisons between the mutants 120 and WT . As shown in Fig. 2, all three mutant genotypes demonstrated alterations in 121 accumulation of proteins when compared to the WT proteome. Interestingly, changes were not 122 limited to chloroplast proteins, as the numbers of over- or under-accumulating proteins among 123 ‘all proteins’ (upper panels) were higher than those considered ‘chloroplast proteins’ (lower 124 panels), suggesting that variations in the chloroplast proteome were transduced to impact other 125 cellular compartments as well. Another interesting observation was that the levels of neither 126 Deg1 nor Deg5-Deg8 were adjusted in the absence of the other Deg proteins, implying that the 127 loss of one was not compensated for by over -accumulation of the other (Fig. 2). Moreover, 128 pairwise Pearson’s correlation analysis on ‘time 0’ samples (NL, prior to exposure to HL) 129 revealed positive high correlation between the proteomes of deg1 and deg158 plants, and lower 130 but still positive correlation between WT and deg58 plants (Supp. Fig. 3). 131 The relatedness of WT and deg58 proteomes and those of deg1 and deg158 was retained 132 also after exposure to HL. As can be seen by the principle component analysis (PCA) presented 133 in Supp. Fig. 4, the different samples were clearly di fferentiated from each other at all time 134 points, and the aforementioned pairs were grouped separately along PC1. 135 136 The response of thylakoid proteins to HL in WT and Deg protease mutants. 137 138 As Deg1 and Deg5-Deg8 proteases are found soluble in the thylakoid lumen , or peripherally 139 attached to the lumenal-side of the thylakoid membrane, they can potentially degrade or cleave 140 lumenal proteins o r lumen-exposed regions of thylakoid membrane integral proteins. W e 141 therefore focused on the fate of such proteins along the exposure of WT plants to HL. Of the 142 ~1,400 identified proteins predicted to localize to chloroplasts, 118 were categorized as 143 thylakoid integral membrane proteins, thylakoid peripheral proteins associated with the lumen 144 side of the membrane, or soluble lumen proteins (Supp. Table 1, analyzed proteins – column 145 ‘GI’). Fifty-five of these proteins were found differentially accumulating (FC> ±2, P value 146 <0.05) in response to HL in WT. Nineteen proteins were up-regulated, including allene oxide 147 synthase, FIB1a and FIB1b, all three involved in jasmonic acid (JA) biosynthesis, two 148 components of the Cyt b6f complex, six proteins of the PSI core complex and its antenna, 149 HCF244 involved in tr anslation initiation of PsbA, PPD6 involved in redox regulation, 150 CURT1A that controls grana architecture, and RIQ1 and RIQ2 involved in LHCII organization, 151 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 7 as well as a number of proteins with unknown function (Supp. Table 1). The numerous down-152 regulated proteins in WT included the following: two subunits of ATP synthase, seven subunits 153 of PSII, two subunits of PSI, 15 proteins involved in the assembly, stability and repair of PSII, 154 the kinase STN8, the TLP18.3 phosphatase , FtsH1,2,5 and 8 proteases, and fi nally, NPQ1 155 (VDE) – the xanthophyll cycle enzyme synthesizing zeaxanthin. 156 To elucidate the impact of the absence of Deg1 and Deg5 -Deg8 on the response of 157 thylakoid membrane and lumenal proteins to HL, we performed a comprehensive co-158 expression clustering analysis. The 118 proteins identified as thylakoid integral membrane or 159 lumenal proteins were subjected to K-means clustering, using Pearson’s correlation as a 160 distance metric, and were grouped into nine clusters of co -regulated proteins in a genotype- 161 and time-dependent manner (Fig. 3, Supp. Table 1). As depicted in the upper dendrogram of 162 Fig. 3, WT and deg58 samples were clearly separated from deg1 and deg158 samples, 163 indicating that the effect of genotype on the accumulation of these proteins was stronger 164 than the effect of HL . Of particular interes t were clusters 2 and 4, containing proteins that 165 over-accumulated in deg1 and deg158 at all time points. As such, at least some of these 166 proteins may be substrates of the Deg1 protease. Proteins in cluster 2 included three subunits 167 of NADH dehydrogenase, four factors associated with PSI or PSII assembly, FtsH1 and 168 FtsH5 proteases, tocopherol synthase (VTE1), and most interestingly, NPQ1 (VDE). Cluster 169 4 proteins included the three aforementioned proteins asso ciated with JA biosynthesis, the 170 light-induced serine protease SppA , and NPQ4 (PsbS). 171 Pairwise comparisons between the three mutant genotypes and WT plants, across the 172 entire dataset , allowed us to focus on the 118 thylakoid proteins that might be in physical 173 contact with the lumenal Deg proteases . As can be seen in Fig. 4 , only very few thylakoid 174 proteins over -accumulated in deg58. In contrast, 43 proteins were found to be up-regulated 175 in many of the 12 remaining pairwise comparisons of deg1 and deg158 to WT, 22 of them 176 in five comparisons or more . These included proteins involved in photosynthesis 177 (plastocyanin, PPD3, PsbO2), PSII assembly/stability/repair (PSB33, HCF136, ALB3, 178 FKBP-20-2, TLP18.3), NADPH dehydrogenase (PnsL1, PnsL4, PnsL5), plastoglobule 179 proteins (FIB1b, FIB4), the protease SppA, u ncharacterized proteins (TL20.3, AT1G32220, 180 AT5G52970, HBP5, ENH1, STR10), and m ost noteworthy - NPQ1 and NPQ4 (Fig. 4). As 181 enhanced expression of the genes encoding NPQ1 and NPQ4 could contribute to their over -182 accumulation, we determined their transcript levels at different time points during the 183 exposure to HL. However, we did not detect differential massive increase in the level of 184 these transcripts in deg1 and deg158 compared to WT (Supp. Fig. 5). Thus, th e over -185 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 8 accumulation of these two proteins in the two genotypes lacking Deg1 during exposure to 186 HL suggests that they might be substrates of this protease . The relatively higher levels of 187 NPQ1 and NPQ4 may lead to the different level and kinetics of NPQ observed in deg1 and 188 deg158 (Fig. 1). 189 190 The effect of cycloheximide on NPQ1 and NPQ4 in WT and the deg1 mutant. 191 192 To further explore the possibility that the level of NPQ1 and NPQ4 is regulated by Deg1, 193 we tested the short-term effect of the protein translation inhibitor cycloheximide (CHX) on 194 WT and deg1 seedlings. To mitigate the size differences between WT and deg1 plants 195 observed in long -term autotrophic growth (see Supp. Fig. 1), seedlings o f both genotypes 196 were grown on Murashige & Skoog agar plates supplemented with 1% sucrose for 15 days. 197 These seedlings, of almost similar size and appearance , were transferred to liquid medium, 198 and incubated with or without CHX for 24 h. Before and after the incubation , total protein 199 extracts were prepared from both genotypes and subjected to MS analysis. More than 5,000 200 different proteins were identified , 4,683 of them with at least two peptides, and these were 201 quantified in the six different groups (two genotypes, before and after incubation with or 202 without CHX; Supp. Table 2). As can be seen in the volcano plots (Fig. 5) and the heatmap 203 (Supp. Fig. 6), the level of more than 700 proteins was altered in the WT in response to the 204 CHX treatment. The level of 442 proteins decreased, whereas that of 268 proteins increased . 205 Nevertheless, NPQ1 and NPQ4 were not among them. In the deg1 mutant, over 1,000 206 proteins were affected by the presence of CHX. 475 proteins were down -regulated, and 578 207 proteins were up -regulated, among them were NPQ1 and NPQ4 (Fig. 5 , Supp. Fig. 6 – 208 cluster no. 3 and Supp. Table 2, analyzed proteins – column ‘BB’ ). 209 Monitoring the level of the highly abundant stromal protei n RBCL revealed no major 210 differences in its level in the different treatments and genotypes (Fig. 6). Similarly, no 211 differences in the level of NPQ1 and NPQ4 in the different samples were observed in the 212 WT. However, after 24 h in the liquid medium, seedlings lacking Deg1 contained apparently 213 more NPQ1, and statistically significant higher level of NPQ4 compared to WT. This trend 214 was even more pronounced in the presence of CHX (Fig. 6). The similar, though somewhat 215 less pronounced trend observed in the absence of CHX suggests that the over -accumulation 216 of these proteins might be relat ed to higher se nsitivity of the deg1 genotype to the transfer 217 from solid to liquid culture, that might induce increased synthesis of the two proteins, before 218 CHX penetrated the cells to exert its inhibitory effect. Nevertheless, the higher accumulation 219 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 9 of both NPQ1 and NPQ4 in plants lacking Deg1 is consistent with the suggestion that the 220 level of the two proteins is under the control of the Deg1 protease , most likely through their 221 degradation by Deg1 . 222 223

224 225 NPQ has been long considered as a prominent mechanism protecting the photosynthetic 226 machinery from oxidative damage through dissipation of excess light energy as heat. It is well 227 established that the xanthophyll zeaxanthin, and the enzyme VDE (NPQ1) synthesizing it, as 228 well as the PsbS (NPQ4) protein, are key components necessary for the induction of NPQ 229 (Niyogi et al., 1998; Li et al., 2000; Li et al., 2002; Jahns and Holzwarth, 2012; Ruban et al., 230 2012; Ruban, 2016). Likewise, the enzyme ZEP (NPQ2), converting zeaxanthin to 231 violaxanthin, and a number of other proteins, including SOQ1, LCNP and ROQH1, are 232 involved in regulating the relaxation of NPQ (Brooks et al., 2013; Malnoe et al., 2018; Amstutz 233 et al., 2020), enabling the funneling of more excitation energy to photochemistry upon the 234 decline in light intensity. Expediting NPQ relaxation has been shown to increase plant biomass 235 under fluctuating light conditions in tobacco (Kromdijk et al., 2016), although this was not the 236 case for Arabidopsis (Garcia-Molina and Leister, 2020). The exact site where NPQ occurs is 237 still a matter under investigation (Sacharz et al., 2017; Nicol et al., 2019). 238 The results of the current study suggest that the thylakoid lumen-located Deg1 protease 239 is also involved in regulating the level of NPQ, through fine-tuning the levels of VDE and PsbS 240 by proteolytic degradation. Three lines of evidence support this: i. The two genotypes that do 241 not contain the Deg1 protease, deg1 and deg158, accumulate higher levels of VDE (NPQ1) 242 and PsbS (NPQ4) not only under optimal growth conditions (Butenko et al., 2018), but also 243 upon exposure to HL (Figure 4); ii. These plants exhibit higher levels of NPQ compared to WT 244 and plants lacking Deg5-Deg8 (Fig. 1). Consistent with this, ETR in the former genotypes is 245 lower than in the latter (Supp. Fig. 2), and consequently, the mutant plants lacking Deg1 are 246 smaller (Supp. Fig. 1; (Butenko et al., 2018)); iii. When protein synthesis is temporally 247 inhibited, the level of NPQ proteins in deg1 is higher than in WT (Fig. 6). This suggestion is 248 also consistent with previous observations: recombinant Deg1 could bind PsbS (as well as a 249 number of other thylakoid proteins from solubilized thylakoids), in a pull-down assay; and 250 when recombinant Deg1 was incubated with solubilized thylakoids, PsbS was slightly 251 degraded (Zienkiewicz et al., 2012). Thus, under natural conditions, when light intensity 252 fluctuates, the balance between synthesis of VDE and PsbS and their proteolytic degradation 253 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 10 determines how much excitation energy is funneled to photochemistry and how much is 254 dissipated as heat (Fig. 7). In mutants lacking the Deg1 protease, the balance between synthesis 255 and degradation of VDE and PsbS is impaired, leading to a higher level of heat dissipation and 256 lower rates of ETR compared to WT, resulting in a decrease in photosynthesis yield (Fig. 7). 257 It is interesting to note that VDE and Deg1 are not only found in the same sub-organellar 258 compartment, and as suggested here, regulate NPQ, but they also share a similar mechanism 259 of activation – through oligomerization that is induced by acidification of the lumen. Although 260 both enzymes have totally different enzymatic activities, they are both monomeric at ne utral- 261 and basic pH. Upon acidification of the lumen VDE dimerizes, and the channel connecting the 262 two active sites can accommodate its substrate, violaxanthin, and de -epoxidize it (Arnoux et 263 al., 2009). Acidification of the lumen is critical for activation of Deg1 as well. Protonation of 264 a specific His residue induces a conformational change in the protein that leads to a change in 265 the orientation of its N-terminal a-helix. This, in turn, results in trimerization of the protein 266 and dimerization of trimers , forming the proteolytically-active hexamers (Kley et al., 2011) . 267 Thus, both VDE and Deg1 are activated under the environmental conditions that necessitate 268 their activity, to protect and regulate the photosynthetic machinery. 269 Although VDE and PsbS accumulate to certain levels under most if not all light 270 conditions, we do not know how they are recognized by the constitutively expressed Deg1 271 protease only when n eeded. One possibility is that under excess light they themselves get 272 oxidized. This oxidation might induce slight conformational changes that expose termini or 273 loops in the substrates , that are long enough to be recognized by Deg1 (Knopf and Adam, 274 2018). This and other possibilities will have to be experimentally challenged in future studies. 275 276

277 278 Plant material – Four genotypes of Arabidopsis thaliana (ecotype Columbia-0) were used in 279 this study: WT, deg1 single mutant, deg58 double mutant and deg158 triple mutant, all of them 280 previously described (Butenko et al., 2018). Seedlings were grown for five weeks under short-281 day conditions (10 h light / 14 h darkness) at 22/180C (day/night). Photon flux density was ~75 282 μmol photons m-2 s−1, representing normal light (NL) conditions. For high light (HL) treatment, 283 the plants were exposed to ~750 μmol photons m-2 s-1 for up to 8 h, starting 1 h after the onset 284 of the light period. For the recovery phase, the HL-treated plants were subjected to additional 285 24 h under NL. Alternatively, seedlings were grown on half strength Murashige & Skoog 286 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 11 agar plates supplemented with 1% sucrose for 15 days at 220C under 100 μmol photons m-2 287 s−1, and then transferred to the same liquid medium, with or without 200 µM cycloheximide, 288 for an additional 24 -h incubation. 289 290 Chlorophyll fluorescence measurements – All measurements were done on whole plants using 291 the Maxi-Imaging PAM Chlorophyll Fluorometer (Walz, Germany), operated by a dedicated 292 software module (ImagingWin 2.3, Walz). Pulse-modulated chlorophyll a fluorescence was 293 recorded as follows: Following dark adaptation for 30 min, the measuring light (9 µmol 294 photons m-2 s-1) was turned on, and the minimal fluorescence in the dark (Fo) was determined. 295 Plants were then exposed to a pulse of saturating light to determine the maximum fluorescence 296 in the dark (Fm). After a short delay (40s), made to enable reoxidation following the saturation 297 pulse (SP), the leaves were illuminated with an actinic light (AL) of ~700 µmol photons m-2 s-298 1 on which successive SPs were superimposed at 20-s intervals, enabling determination of the 299 maximum fluorescence in the light-adapted state (Fm’). Once steady state has been established, 300 the AL was turned off , the SPs were set at 40 -60-s intervals, and the maximal and minimal 301 fluorescence during illumination (Fm’ and Fo’) were determined. Rapid light response curves 302 were recorded using a 10-min protocol of successive 20-s cycles of exposure to increasing light 303 intensities, from 1 to 1250 µmol photons m-2 s-1. All measurements were performed in a dark 304 room at 250C. Parameters calculated included the maximum quantum yield of PSII (Fv/Fm), 305 efficiency of PSII (Y[II]), electron transport rate (ETR) and non-photochemical quenching 306 (NPQ). 307 308 Shotgun proteomic analysis – Frozen leaf samples were crushed in 5% SDS, 0.1M Tris-HCl 309 pH 7.9. The lysates were incubated for 5 min at 96°C, then cleared by centrifugation at 14,000g 310 for 5 min at RT. Proteins were reduced by incubation with 5 mM dithiothreitol (Sigma) for 45 311 min at 60°C, and alkylated with 10 mM iodoacetamide (Sigma) in the dark for 45 min at RT. 312 The lysates were loaded on a S-trap column and washed with 90% methanol. The proteins were 313 proteolytically digested on the column by incubation with trypsin (Promega) overnight at 37°C, 314 at 50:1 protein/trypsin ratio. Peptides were eluted from the column with 50% acetonitrile and 315 0.2% formic acid. The samples were vacuum dried and stored in 80˚C until further analysis. 316 Samples were redesolved and loaded in a random order, using split-less nano-Ultra 317 Performance Liquid Chromatography (nanoUPLC; 10 kpsi nanoAcquity; Waters, Milford, 318 MA, USA). The mobile phase consisted of H2O + 0.1% formic acid (A) and acetonitrile + 0.1% 319 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 12 formic acid (B). Desalting of the samples was perfo rmed online, using reversed -phase 320 Symmetry C18 trapping column (180 µm internal diameter, 20 mm length, 5 µm particle size; 321 Waters). Peptides were separated on a T3 HSS nano -column (75 µm internal diameter, 250 322 mm length, 1.8 µm particle size; Waters) at 0 .35 µL/min. They were eluted from the column 323 into the mass spectrometer using the following gradient: 4% to 30%B in 105 min, 30% to 324 90%B in 5 min, maintained at 90% B for 5 min and then back to 4%B. 325 The nanoUPLC was coupled online through a nanoESI emitter (10 μm tip; New 326 Objective; Woburn, MA, USA) to a tribrid Orbitrap Fusion Lumos mass spectrometer (Thermo 327 Scientific), using a PicoView nanospray apparatus (New Objective). Data was acquired in data 328 dependent acquisition (DDA) mode, using a top speed metho d with 3 second cycle time, 329 according to the manufacturer recommendations . MS1 resolution was set to 120,000 (at 330 200m/z), mass range of 300-1600m/z, Automatic Gain Control (AGC) of 4e5, and the 331 maximum injection time was set to 50msec. Fragmentation was performed in the ion trap, with 332 quadrupole isolation window of 1m/z, AGC of 1e4, dynamic exclusion of 20sec and maximum 333 injection time of 34 msec. 334 Raw data was processed using MaxQuant version 1.6.0.16. The data were searched with 335 the Andromeda search engine against the TAIR proteome, appended with common lab protein 336 contaminants. Quantification was based on the LFQ method (Cox et al., 2014), based on unique 337 peptides for each protein. 338 339 MS data analysis - Individual intensities were log2 transformed and Z-score normalized. 340 Missing values were imputed using normal distribution, with a mean and standard deviation 341 adjusted to resemble low-abundant proteins signals. Proteins were considered for comparative 342 analysis if a protein was identified in at least three out of four replicates. Analysis of Variance 343 (ANOVA), was used to identify significant differences across the biological replicates. The 344 criteria used to denote significantly differentially expressed proteins for further analysis were 345 fold-change >2 and false discovery rate (FDR) correction of 5% (q value < 0.05). The quality 346 of the expressed and differentially expressed data was examined by principal component 347 analysis (PCA). Pairwise Pearson's correlation was performed to evaluate the dynamic (time-348 varying) strength of the association between the variables. Unsupervised Hierarchal clustering 349 analysis was performed using Euclidean distance metric and Ward’s Minimum-Variance-350 linkage agglomerative method. Heat maps were based on K-means clustering, using Pearson 351 correlation coefficient as a distance metric. The optimal number of clusters was computed 352 using the gap statistic method and comprised 1000 Monte Carlo iterations (Tibshirani et al., 353 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 13 2001). Proteins with significantly changed abundances were subjected to gene ontology (GO) 354 annotation using Blast2Go (Conesa and Gotz, 2008) software or the DAVID (Huang da et al., 355 2009) database. Gene set enrichment analysis (GSEA) of differentially expressed proteins 356 (DEPs) based on Fisher exact test was performed on GO cellular compartments (CC), 357 biological processes (BP) and molecular function (MF). Groups with at least 2 proteins 358 compared to the identified proteins background and a p value <0.05 were considered enriched. 359 For identifications involving uncharacterized/unannotated proteins, sequences were retrieved 360 from the NCBI database and were blasted against Plants/Arabidopsis thaliana protein 361 sequences. Analyses were performed using R/Bioconductor (version R.3.4.2), jmp10 software, 362 or Microsoft Excel 2016. 363 364 RT-qPCR analysis – RNA was extracted with the SV Total RNA Isolation System (Promega) 365 then RNA was treated with DNase I (RNase-free) (Ambion, Thermo Fisher Scientific) and 366 purified with PCI (25:24:1) prior to its use in the assays. Reverse transcription was carried out 367 with 200 U of Superscript III reverse transcriptase, in presence of 40 U of RNase OUT (both 368 from Invitrogen, using ~1 μg of total RNA and 100 ng of a mixture of random hexanucleotides 369 (Promega) and incubated for 50 min at 50 °C. Reactions were stopped by 15 min incubation at 370 70 °C and the RT samples served directly for real-time PCR. Quantitative PCR (qPCR) 371 reactions were run on a LightCycler 480 (Roche-Diagnostics, Basel, Switzerland), using 2.5 372 μL of qPCRBIO SyGreen Blue Mix (PCR BIOSYSTEMS, Pennsylvania, USA) and 2.5 μM 373 forward and reverse primers in a final volume of 5 µL. Reactions were performed in triplicate 374 in the following conditions: pre-heating at 95 °C for 10 min, followed by 40 cycles of 10 s at 375 95 °C, 10 s at 58 °C and 10 s at 72 °C. The nucleus-encoded ubiquitin and tubulin were used 376 as reference genes in the qPCR analyses. 377 378 AUTHOR CONTRIBUTIONS 379 380 Z.A. and Z.R. conceived and supervised the project. E.A.-S., L.N. and L.D.S. grew all plants 381 and performed all growth manipulations and treatments. E.A.-S., L.N. and D.C. did all 382 photosynthesis measurements. E.A.-S., L.D.S. and M.K. performed the proteomic and MS data 383 analyses. L.D.S. did RT-qPCR analyses. E.A.-S., L.D.S., D.C., Z.R. and Z.A. analyzed all data. 384 Z.A. wrote the manuscript, with contribution of all the other coauthors. 385 386 387 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 14 SUPPLEMENTAL DATA 388 The following materials are available in the online version of this article. 389 390 Supplemental Figure 1. Imaging PAM measurements of PSII maximum quantum yield 391 (Fv/Fm). 392 393 Supplemental Figure 2. Light response curves of photosynthetic electron transport rates 394 (ETRs). 395 396 Supplemental Figure 3. Pairwise Pearson’s correlation analysis of WT and deg mutant plants 397 under NL conditions. 398 399 Supplemental Figure 4. Principal component analysis (PCA) of the proteomic data of WT 400 and deg mutant plants subjected to NL and HL conditions. 401 402 Supplemental Figure 5. Relative expression of NPQ1 and NPQ4 during high light exposure 403 in WT and deg mutants. 404 405 Supplemental Figure 6. Differential expression of altered proteins in the deg1 mutant and WT 406 in response to cycloheximide treatment. 407 408 Supplemental Table 1. Proteomic raw data and analysis of WT and deg mutants under NL, 409 HL and recovery. 410 411 Supplemental Table 2. Proteomic raw data and analysis of WT and deg1 with or without 412 cycloheximide treatment. 413 414 415 FUNDING 416 This work was supported by grants from The Israel Science Foundation (ISF) no. 1377/18 (to 417 D.C.), 1082/17 (to Z.R.) and 2585/16 and 1167/18 (to Z.A.), and the National Science 418 Foundation United States - Israel Binational Science Foundation no. 2015839 and 2019695 (to 419 Z.R.). 420 421 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 15

422 423 Adir, N., Zer, H., Shochat, S., and Ohad, I. (2003). Photoinhibition - a historical perspective. 424 Photosynth Res 76: 343-370. 425 Amstutz, C.L., Fristedt, R., Schultink, A., Merchant, S.S., Niyogi, K.K., and Malnoe, A. 426 (2020). An atypical short-chain dehydrogenase-reductase functions in the relaxation of 427 photoprotective qH in Arabidopsis. Nat Plants 6: 154-166. 428 Arnoux, P., Morosinotto, T., Saga, G., Bassi, R., and Pignol, D. (2009). A structural basis for 429 the pH-dependent xanthophyll cycle in Arabidopsis thaliana. Plant Cell 21: 2036-2044. 430 Bailey, S., Thompson, E., Nixon, P.J., Horton, P., Mullineaux, C.W., Robinson, C., and Mann, 431 N.H. (2002). A critical role for the Var2 FtsH homologue of Arabidopsis thaliana in the 432 photosystem II repair cycle in vivo. J. Biol. Chem. 277: 2006-2011. 433 Brooks, M.D., Sylak-Glassman, E.J., Fleming, G.R., and Niyogi, K.K. (2013). A thioredoxin-434 like/beta-propeller protein maintains the efficiency of light harvesting in Arabidopsis. 435 Proc Natl Acad Sci U S A 110: E2733-E2740. 436 Butenko, Y., Lin, A., Naveh, L., Kupervaser, M., Levin, Y., Reich, Z., and Adam, Z. (2018). 437 Differential roles of the thylakoid lumenal Deg protease homologs in chloroplast 438 proteostasis. Plant Physiol 178: 1065-1080. 439 Conesa, A., and Gotz, S. (2008). Blast2GO: A comprehensive suite for functional analysis in 440 plant genomics. Int J Plant Genomics 2008: 619832. 441 Correa-Galvis, V., Poschmann, G., Melzer, M., Stuhler, K., and Jahns, P. (2016). PsbS 442 interactions involved in the activation of energy dissipation in Arabidopsis. Nat Plants 443 2: 15225. 444 Cox, J., Hein, M.Y., Luber, C.A., Paron, I., Nagaraj, N., and Mann, M. (2014). Accurate 445 proteome-wide label-free quantification by delayed normalization and maximal peptide 446 ratio extraction, termed MaxLFQ. Mol Cell Proteomics 13: 2513-2526. 447 Garcia-Molina, A., and Leister, D. (2020). Accelerated relaxation of photoprotection impairs 448 biomass accumulation in Arabidopsis. Nat Plants 6: 9-12. 449 Huang da, W., Sherman, B.T., and Lempicki, R.A. (2009). Systematic and integrative analysis 450 of large gene lists using DAVID bioinformatics resources. Nat Protoc 4: 44-57. 451 Jahns, P., and Holzwarth, A.R. (2012). The role of the xanthophyll cycle and of lutein in 452 photoprotection of photosystem II. Biochim Biophys Acta 1817: 182-193. 453 Kapri-Pardes, E., Naveh, L., and Adam, Z. (2007). The thylakoid lumen protease Deg1 is 454 involved in the repair of photosystem II from photoinhibition in Arabidopsis. Plant Cell 455 19: 1039-1047. 456 Kato, Y., Sun, X., Zha ng, L., and Sakamoto, W. (2012). Cooperative D1 degradation in the 457 photosystem II repair mediated by chloroplastic proteases in Arabidopsis. Plant Physiol 458 159: 1428-1439. 459 Kley, J., Schmidt, B., Boyanov, B., Stolt -Bergner, P.C., Kirk, R., Ehrmann, M., Knopf , R.R., 460 Naveh, L., Adam, Z., and Clausen, T. (2011). Structural adaptation of the plant protease 461 Deg1 to repair photosystem II during light exposure. Nat Struct Mol Biol 18 728-731. 462 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 16 Knopf, R.R., and Adam, Z. (2018). Lumenal exposed regions of the D1 prote in of PSII are 463 long enough to be degraded by the chloroplast Deg1 protease. Sci Rep 8: 5230. 464 Kromdijk, J., Glowacka, K., Leonelli, L., Gabilly, S.T., Iwai, M., Niyogi, K.K., and Long, S.P. 465 (2016). Improving photosynthesis and crop productivity by accelerating recovery from 466 photoprotection. Science 354: 857-861. 467 Li, X.-P., Bjorkman, O., Shih, C., Grossman, A.R., Rosenquist, M., Jansson, S., and Niyogi, 468 K.K. (2000). A pigment-binding protein essential for regulation of photosynthetic light 469 harvesting. Nature 403: 391-395. 470 Li, X.P., Muller -Moule, P., Gilmore, A.M., and Niyogi, K.K. (2002). PsbS -dependent 471 enhancement of feedback de -excitation protects photosystem II from photoinhibition. 472 Proc. Natl. Acad. Sci. U S A 99: 15222-15227. 473 Lindahl, M., Spetea, C., Hund al, T., Oppenheim, A.B., Adam, Z., and Andersson, B. (2000). 474 The thylakoid FtsH protease plays a role in the light -induced turnover of the 475 photosystem II D1 protein. Plant Cell 12: 419-431. 476 Malnoe, A., Schultink, A., Shahrasbi, S., Rumeau, D., Havaux, M., and Niyogi, K.K. (2018). 477 The plastid lipocalin LCNP is required for sustained photoprotective energy dissipation 478 in Arabidopsis. Plant Cell 30: 196-208. 479 Maxwell, K., and Johnson, G.N. (2000). Chlorophyll fluorescence --a practical guide. J Exp 480 Bot 51: 659-568. 481 Nicol, L., Nawrocki, W.J., and Croce, R. (2019). Disentangling the sites of non-photochemical 482 quenching in vascular plants. Nat Plants 5: 1177-1183. 483 Niyogi, K.K., Grossman, A.R., and Bjorkman, O. (1998). Arabidopsis mutants define a central 484 role for th e xanthophyll cycle in the regulation of photosynthetic energy conversion. 485 Plant Cell 10: 1121-1134. 486 Powles, S.B. (1984). Photoinhibition of photosynthesis induced by visible light. Annu. Rev. 487 Plant Physiol. 35: 15-44. 488 Ruban, A.V. (2016). Nonphotochemical Chlorophyll Fluorescence Quenching: Mechanism 489 and Effectiveness in Protecting Plants from Photodamage. Plant Physiol 170: 1903 -490 1916. 491 Ruban, A.V., Johnson, M.P., and Duffy, C.D. (2012). The photoprotective molecular switch in 492 the photosystem II antenna. Biochim Biophys Acta 1817: 167-181. 493 Saccon, F., Wilson, S., Morey-Burrows, F.S., and Ruban, A.V. (2022). Quantifying the long -494 term interplay between photoprotection and repair mechanisms sustaining photosystem 495 II activity. Biochem J 479: 701-717. 496 Sacharz, J., Giovagnetti, V., Ungerer, P., Mastroianni, G., and Ruban, A.V. (2017). The 497 xanthophyll cycle affects reversible interactions between PsbS and light -harvesting 498 complex II to control non-photochemical quenching. Nat Plants 3: 16225. 499 Sun, X., Peng, L., Guo, J., Chi, W., Ma, J., Lu, C., and Zhang, L. (2007). Formation of DEG5 500 and DEG8 complexes and their involvement in the degradation of photodamaged 501 photosystem II reaction center D1 protein in Arabidopsis. Plant Cell 19: 1347-1361. 502 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 17 Tibshirani, R., Walther, G., and Hastie, T. (2001). Estimating the number of clusters in a data 503 set via the gap statistic. J. R. Statist. Soc. B 63: 411-423. 504 Zaltsman, A., Ori, N., and Adam, Z. (2005). Two types of FtsH protease subunits are required 505 for chloroplast biogenesis and Photosystem II repair in Arabidopsis. Plant Cell 17: 506 2782-2790. 507 Zienkiewicz, M., Ferenc, A., Wasilewska, W., and Romanowska, E. (2012). High light 508 stimulates Deg1-dependent cleavage of the minor LHCII antenna proteins CP26 and 509 CP29 and the PsbS protein in Arabidopsis thaliana. Planta 235: 279-288. 510 511 512 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 18 513 514 515 516 517 518 Figure 1. Induction/relaxation curves of non -photochemical quenching (NPQ). Recordings were 519 performed on intact leaves before (0h) and following exposure to high-light (HL) (~750 μmol m2 s1) for 520 1, 2, 4 and 8 hours (2h, 4h and 8h, respectively). (32h) Curves recorded on leaves exposed to HL for 8 521 h and then let to recover for 24 h under NL conditions. Values shown represent the mean ±SE obtained 522 from recordings of 25 leaves from at least 3 different plants, at each time point, for each genotype. 523 524 525 526 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 19 527 528 529 530 Figure 2. Volcano plots of proteomic data obtained from plants under normal light (NL) 531 conditions. Volcano plots are depicted with the log2 fold-change (-1, 1; vertical dashed lines) 532 of each protein and the -log10 p-value (p<0.05) (horizontal dashed line). The averages of the 533 proteomic expression data of each mutant (deg1, deg58 N = 4; deg158, N = 3) were compared 534 with the averages of the data for WT (N = 4). Blue circles mark proteins whose level is higher 535 in the mutants, whereas red circles mark those whose level is lower. Gray circles mark proteins 536 whose level is not significantly altered or those which exhibit a log 2 fold-change < 1. Deg1 537 (black circle), Deg5 (black triangle) and Deg8 (black square) are highlighted according to their 538 p-values and fold ch ange. Data are shown for (A) all 4188 identified proteins and (B) ~1400 539 chloroplast proteins. 540 541 542 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 20 543 544 545 Figure 3. Differential expression of thylakoid membrane and lumenal proteins in WT and 546 deg mutants under normal- and high light, and following recovery. The heat map generated 547 for the thylakoid membrane and lumenal proteins identified in our study reveals segregation 548 into nine co-expression groups. 549 550 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 21 551 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 22 Figure 4. Volcano plots of thylakoid integral membrane and lumenal proteins expression 552 levels in deg mutants vs. WT. Volcano plots are depicted with the log 2 fold-change (vertical 553 dashed lines) and the -log10 p-value (p<0.05) (horizontal dashed line) of the 118 proteins 554 localized to the thylakoid membranes and lumen. The averages of the proteomic expression 555 data of each mutant were compared with those of the WT at each time point. Blue and light 556 blue circles m ark lumenal and integral membrane proteins, respectively, whose levels 557 increased more in the mutants relative to WT. Red circles denote proteins whose levels 558 decreased more in the mutants vs. WT. Orange circles denote NPQ1 and NPQ4 proteins and 559 highlight their over-accumulation in deg1 and deg158 plants compared to WT. Gray circles are 560 proteins without any significant differences and/or log2 fold-change <1. 561 562 563 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 23 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 Figure 5. Volcano plot of proteomic data obtained from WT and deg1 plants in response 593 to cycloheximide treatment. Volcano plots are depicted with the log 2 fold-change (vertical 594 dashed lines) of each protein and the -log10 p-value (p<0.05) (horizontal dashed line). The 595 averages of the proteomic expression data of deg1 mutant and WT following 24 -hour 596 incubation with cycloheximide were compared with the averages of the data for deg1 mutant 597 and WT without cycloheximide incubation (N = 5). Blue circles mark proteins whose level was 598 higher following the cycloheximide treatment, whereas magenta circles mark those whose level 599 was lower. Gray circles mark proteins whose level was not significantly altered or those which 600 exhibit a log2 fold-change < 1. NPQ1 and NPQ4 (orange circles) are highlighted according to 601 their p-values and fold change. 602 603 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 24 604 605 Figure 6. Effect of cycloheximide on accumulation of NPQ1 and NPQ4. Boxplots showing 606 the levels of NPQ1, NPQ4 and RBCL proteins in WT and deg1 mutant plants (N=5) before (0 607 h), with or without cycloheximide treatment for 24 h. *p<0.05, **p<0.01. 608 609 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint 25 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 Figure 7. A model for the proposed role of the Deg1 protease in regulation of NPQ. In the 630 presence of the thylakoid lumen-located Deg1 protease, NPQ levels are fine-tuned via the 631 continued proteolytic degradation of VDE (NPQ1) and PsbS (NPQ4), allowing for high 632 electron transfer rate (ETR) and an increased photosynthetic yield. In the absence of the Deg1 633 protease, the degradation of VDE (NPQ1) and PsbS (NPQ4) is impaired, causing their 634 accumulation and, in turn, an increased level of heat dissipation on account of photochemistry 635 (lower ETR). As a result, the photosynthetic yield is decreased and the development and growth 636 of the plants is compromised. 637 638 639 640 641 .CC-BY-NC-ND 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is The copyright holder for this preprintthis version posted May 17, 2024. ; https://doi.org/10.1101/2024.05.14.594122doi: bioRxiv preprint