Involvement of PHOSPHATE TRANSPORTER TRAFFIC FACILITATOR1 in COPII assembly by interacting with SAR1 GTPase

1 The plant -specific endoplasmic reticulum (ER) -resident PHOSPHATE TRANSPORTER 2 TRAFFIC FACILITATOR1 (PHF1) is structurally related to SEC12, which initiates the coat 3 protein complex II (COPII) assembly as a guanine nucleotide exchange factor (GEF) by 4 activating the small GTPase SAR1. In contrast, PHF1 loses the conserved catalytic residues 5 critical for GEF activity and instead specifically assists the ER exit of the plant PHOSPHATE 6 TRANSPORTER1 (PHT1) family. However, the underlying molecular mechanism remains to 7 be elucidated. In this study, we showed that the overexpression of Arabidopsis thaliana PHT1;1 8 (AtPHT1;1) but not of SUGAR TRANSPORTER 1 ( AtSTP1) caused a portion of AtPHF1 9 distribution into AtSAR1b-, AtSAR1c- and AtSEC24a-labeled ER exit sites in the tobacco 10 transient expression system . Based on the in-planta tripartite split-GFP association, AtPHF1 11 interacted with AtSAR1b and AtSAR1c but not with other COPII -related proteins. By 12 performing the miniTurbo-based proximity labeling in agro -infiltrated tobacco leaves , we 13 verified the interaction of AtPHF1 and AtSAR1b and demonstrated its physiological relevance 14 by co -immunoprecipitation of the endogenous AtPHF1 and AtPHT1;1/2/3 proteins with 15 AtSAR1c-GFP using Arabidopsis transgenic lines. Furthermore, while both the cytosolic and 16 transmembrane domains of AtPHF1 contribute to the interaction with AtSAR1b and AtSAR1c, 17 AtPHF1 preferentially interacted with the GDP-locked inactive form of AtSAR1b. On the basis 18 of these findings, we propose that by interacting with SAR1 GTPase, PHF1 participates in the 19 early step of COPII recruitment for the ER export of PHT1 proteins. 20 21

coat protein complex II (COPII) , PHOSPHATE TRANSPORTER TRAFFIC 22 FACILITATOR1, ER export, PHOSPHATE TRANSPORTER1 23 24 25 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint

1 As sessile organisms, plants adjust the protein and lipid composition of the plasma 2 membrane (PM) to sense and adapt to the ever -changing environment. Such dynamic cellular 3 processes heavily rely on the modulation of the endomembrane trafficking system (Wang et al., 4 2020), including the coat protein complex II (COPII) -mediated endoplasmic reticulum (ER) -5 to-Golgi transport pathway. Although the core COPII machinery composed of five cytosolic 6 factors, SAR1, SEC23, SEC24, SEC13, and SEC31 , is conserved throughout all eukaryotes, 7 the mechanisms underlying COPII recruitment and assembly are most extensively studied in 8 the budding yeast Saccharomyces cerevisiae (Barlowe and Miller, 2013) . In this model 9 organism, the COPII-related components are encoded by single genes and recruited sequentially 10 to the ER exit sites (ERES) (Kurokawa and Nakano, 2019) . The SAR1 GTPase is a ctivated 11 upon the exchange of GDP to GTP by the guanine nucleotide exchange factor (GEF) SEC12, 12 exposing its amphipathic N-terminal helix to be inserted into the ER membrane (d'Enfert, 1991; 13 Paul et al., 2023). SAR1 then interacts directly with SEC23, inducing the formation of the inner 14 coat SEC23–SEC24 complex, which in turn recruits the outer coat SEC13–SEC31 complex (Bi, 15 2002; Stagg et al., 2006) . In addition, SEC16, an initial factor assembled at the ERES, serv es 16 as a scaffold protein by concentrating SEC12 and GTP-bound active SAR1 in proximity (Supek 17 et al., 2002; Connerly et al., 2005) . By comparison, the plant COPII subunit paralogs are 18 encoded by multi-gene families and outnumber those in other organisms (Chung et al., 2016). 19 Multiple COPII paralogues allow the diversity of COPII transport carriers to adapt to various 20 developmental and stress -related cues (Wang et al., 2020; Li et al., 2022) . For instance, the 21 phytohormone abscisic acid triggered the formation of Arabidopsis thaliana SAR1a 22 (AtSAR1a)-dependent COPII vesicles that carry osmotic stress-related carriers (Li, 2021). The 23 AtSAR1a-AtSEC23a pairing was also shown to mediate the ER export of the ER -associated 24 transcription factor bZIP28, which is upregulated under ER stress (Zeng et al., 2015) . By 25 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint contrast, how the COPII machinery regulates protein ER export to cope with nutrient deficiency 1 remains unknown. 2 Inorganic phosphate (Pi) is an essential macronutrient for plant growth. Due to its poor 3 solubility and mobility in soil, Pi bioavailability is limited, thus causing plants to face Pi 4 deficiency (Shen, 2011). Under Pi-limited conditions, plants enhance the external Pi uptake as 5 well as the recycling and remobilization of internal Pi, which involves the upregulation of the 6 PHOSPHATE TRANSPORTER1 (PHT1) gene family at the transcript and post -transcriptional 7 level (Nussaume et al., 2011). While most PHT1 genes are upregulated by Pi deprivation, the 8 newly synthesized PHT1 proteins (PHT1s) must exit the ER for the PM targeting. Similar to 9 the inhibitory effect of SEC12 overexpression on the ER export of the Golgi transmembrane 10 protein ERD2 in tobacco leaves (Hanton et al., 2007), the overexpression of SEC12 impaired 11 the PM targeting of the Arabidopsis thaliana PHT1;2 (AtPHT1;2) (Bayle et al., 2011) . Thus, 12 the ER-to-Golgi transport of PHT1s is dependent on efficient COPII assembly. Interestingly, 13 the plant -specific SEC12 -related p rotein PHOSPHATE TRANSPORTER TRAFFIC 14 FACILITATOR1 (PHF1) was identified to facilitate the ER exit of AtPHT1s (Gonzalez et al., 15 2005; Bayle et al., 2011). Despite sharing the sequence similarity with SEC12, AtPHF1 lost the 16 conserved catalytic residues critical for GEF activity and failed to rescue the growth defect of 17 the yeast sec12 mutant (Gonzalez et al., 2005). Instead, the Arabidopsis loss-of-function phf1 18 mutant showed decreased cellular Pi content and impaired PM targeting of AtPHT1;1 (Gonzalez 19 et al., 2005). As the role of PHF1 in assisting the ER exit of PHT1s resembles that of the yeast 20 Pho86p in the ER exit of Pho84p, the yeast homolog of PHT1s (Lau, 2000) , PHF1 may 21 functionally diverge from SEC12 (Gonzalez et al., 2005). 22 The transient protein expression using tobacco leaves is a well-established plant system to 23 study the dynamic organization of COPII proteins at the ERES (daSilva et al., 2004; Hanton et 24 al., 2009). In this system, the ERES formation can be monitored by the subcellular distribution 25 of fluorescent protein-tagged COPII components changing from cytosolic patterns into punctate 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint structures (daSilva et al., 2004; Hanton et al., 2007; Hanton et al., 2008; Hanton et al., 2009) . 1 Overexpression of COPII -dependent membrane cargoes enhanced the recruitment of YFP -2 AtSEC24a to the ERES and induced the de novo formation of ERES (Hanton et al., 2007) . 3 Likewise, AtPHT1;2-CFP overexpression resulted in the recruitment of YFP -AtSEC24a and 4 YFP-ScSar1 to the ERES puncta (Bayle et al., 2011). Although the co-expression study showed 5 that the ER-resident AtPHF1 failed to co-localize with AtSEC24a, thus leading to the conclusion 6 that AtPHF1 is not involved in COPII recruitment (Bayle et al., 2011) , it is unclear whether 7 PHF1 participates in cargo recognition and selective packaging of PHT1 Pi transporters into 8 COPII vesicles. 9 Since PHF1 shares sequence similarity with SEC12 yet exhibits a distinct function, we 10 speculated that PHF1 may engage the COPII-dependent ER export of PHT1s on a molecular 11 basis different from SEC12. To explore this hypothesis, we aim ed to investigate whether 12 AtPHT1;1 overexpression can induce the distribution of AtPHF1 into ERES puncta and whether 13 AtPHF1 can interact with COPII-related components. Using agro-infiltrated tobacco leaves, we 14 showed that overexpression of AtPHT1;1 triggered a portion of AtPHF1 distribution into 15 punctate structures that partially co-localized with AtSAR1b, AtSAR1c and AtSEC24a. 16 Furthermore, AtPHF1 interacted with AtSAR1b and AtSAR1c but not with AtSEC24a and other 17 COPII-related components based on tripartite split-GFP association. Consistently, we verified 18 the interaction of AtPHF1 and AtSAR1b by the proximity labeling in the transient tobacco 19 system and demonstrated that the endogenous AtPHF1 and AtPHT1;1/2/3 was co-20 immunoprecipitated with the AtSAR1c-GFP in the Arabidopsis transgenic line . More 21 importantly, we showed that like AtSEC12, AtPHF1 preferentially interacted with the GDP -22 locked inactive form of AtSAR1b. Our results unveiled that AtPHF1 acts as an early regulator 23 of COPII assembly for the ER export of PHT1s through interacting with AtSAR1 proteins 24 (AtSAR1s). 25 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint

1 Overexpression of AtPHT1;1 triggers the partial distribution of AtPHF1 into ERES -2 associated punctate structures 3 Both the mammalian and plant SEC12 proteins are dispersed throughout the ER 4 (Weissman et al., 2001; Bayle et al., 2011) . However, the mammalian SEC12 can be 5 concentrated at the ERES by the collagen cargo receptor component cTAGE5 for the ER export 6 of collagen (Saito et al., 2014). We thus wondered whether AtPHF1, which distributes evenly 7 across the ER (Gonzalez et al., 2005; Bayle et al., 2011) , can be induced to concentrate at the 8 ERES when the ER export of AtPHT1;1 is highly demanded. When expressed alone in agro-9 infiltrated tobacco le aves, AtPHF1-GFP showed a reticular ER pattern , while a portion of 10 AtPHF1-GFP showed punctate -like structures when the split -GFP tagged AtPHT1;1 11 (AtPHT1;1-S11) was co-expressed (Fig. 1A). By contrast, the co-expression of AtPHF1-GFP 12 with the sugar transpor ter AtSTP1-S11 did not change the distribution of AtPHF1-GFP (Fig. 13 1A). Statistical analysis further suggested that the overexpression of AtPHT1;1-S11 but not of 14 AtSTP1-S11 significantly triggered the partial distribution of AtPHF1-GFP in to punctate 15 structures (Fig. 1B). 16 In the tobacco transient expression system, the co -expression of membrane cargoes 17 recruited the COPII components, such as AtSAR1s and AtSEC24, to concentrate at the punctate 18 ERES (Hanton et al., 2007; Hanton et al., 2008). To address whether the punctate structures of 19 AtPHF1-GFP induced by the overexpression of AtPHT1;1 might represent ERES, we selected 20 AtSAR1b as the COPII marker to co-localize with AtPHF1-mCherry. This is because AtSAR1b 21 is the most highly expressed AtSAR1 gene in roots and is upregulated by Pi deprivation (Liu et 22 al., 2016) (Supplementary Fig. S1). In addition, the previous proteomic analysis of the Pi 23 overaccumulator pho2 root has revealed the co-induction of AtSAR1b and AtPHT1s (Huang et 24 al., 2013), implying an increased demand for AtSAR1b to promote the ER -to-Golgi traffic of 25 AtPHT1s. Considering that the N-terminal tail of SAR1 is inserted into the ER membranes upon 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint activation (d'Enfert, 1991; Paul et al., 2023), we fused the S11 tag to the C terminus of AtSAR1b 1 to prevent steric hindrance. In tobacco epidermal cells, AtSAR1b-S11 exhibited both a diffuse 2 cytosolic distribution and an ER membrane -associated punctate pattern (Supplementary Fig. 3 S2A). As AtSAR1c is the second most expressed AtSAR1 gene in roots (Liu et al., 2016) 4 (Supplementary Fig. S1 ) and play s an interchangeable role with AtSAR1b in pollen 5 development at the protein level (Liang et al., 2020) , we also included AtSAR1c-S11 for 6 comparison. AtSEC24a is involved in the cargo selection for COPII vesicles and is also widely 7 used as a reliable ERES marker (Miller, 2002; Hanton et al., 2007) , so we generated the 8 construct expressing AtSEC24a-S11 for co-localization with AtPHF1-GFP as well . The 9 subcellular localization s of AtSAR1b-S11 and AtSAR1c-S11, as detected by the 10 complementation of GFP1–10, were similar to the previously reported results (Hanton et al., 11 2008; Zeng et al., 2015) (Supplementary Fig. S2A ). AtSEC24a-S11 also displayed the 12 cytosolic and punctate patterns ( Supplementary Fig. S2A ). While t he co -expression of 13 AtPHF1-mCherry did not affect the subcellular distribution of AtSAR1b and AtSEC24a, the 14 AtPHT1;1 overexpression-induced punctate structures of AtPHF1-mCherry co-localized with 15 AtSAR1b- and AtSEC24a-labeled ERES ( Fig. 2A and 2B). Pearson’s correlation analysi s 16 further revealed that AtPHF1-mCherry puncta co-localized better with AtSAR1b than with 17 AtSEC24a, as indicated by a statistically significant higher coefficient for AtSAR1b (Fig. 2C). 18 Similarly, AtPHF1-mCherry puncta partially co-localized with AtSAR1c (Supplementary Fig. 19 S2B). There was no significant difference in the co-localization coefficient between AtSAR1b 20 and AtSAR1c (Supplementary Fig. S2C). Taken together, these results suggested that when 21 AtPHT1;1 is overexpressed as the COPII -dependent export membrane cargo, AtPHF1 can 22 partially distribute into the puncta associated with the ERES markers. 23 24 The interaction of AtPHF1 with AtSAR1b/c in planta based on tripartite split -GFP 25 association 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint As AtPHT1;1 overexpression triggered the distribution of AtPHF1 partially into punctate 1 structures that are associated with AtSAR1b, AtSAR1c and AtSEC24a, we postulated that 2 AtPHF1 may participate in the COPII recruitment or assembly. To explore this possibility, we 3 examined the interaction of AtPHF1 with several COPII-related components using the tripartite 4 split-GFP assay in agro-infiltrated tobacco leaves (Cabantous et al., 2013; Liu et al., 2018). The 5 expression of split-GFP tagged AtSEC12 and AtPHF1 (S10-AtSEC12 and S10 -AtPHF1) 6 confirmed their localization at the ER when co-expressed with the cytosolic S11-GFP1–9 (Fig. 7 3A). We also generated the constructs expressing the split-GFP tagged AtSEC16a and 8 AtSEC13a (AtSEC16a-S11 and AtSEC13a-S11). As previously reported (Hanton et al., 2009; 9 Takagi et al., 2013), AtSEC16a-S11 predominantly localized to the cytosol, while AtSEC13a-10 S11 displayed both the cytosol and nucleus patterns (Fig. 3A). As expected, S10-AtSEC12 as a 11 positive control for the protein-protein interaction interacted with AtSAR1b and AtSAR1c, 12 yielding the GFP complementation signals (Fig. 3B). Notably, like S10-AtSEC12, S10-AtPHF1 13 interacted with AtSAR1b-S11 but not with AtSEC24a-, AtSEC16a-, and AtSEC13a-S11 (Fig. 14 3B). A similar interaction pattern was also observed for AtPHF1 with AtSAR1c ( Fig. 3B ), 15 reinforcing the specificity of AtPHF1 interaction with AtSAR1s. 16 17 The interaction of AtPHF1 with AtSAR1b in planta based on miniTurbo (mTb) proximity 18 labeling 19 Although the tripartite split-GFP association detects protein-protein interaction without 20 spurious background signals (Cabantous et al., 2013; Liu et al., 2018) , the reconstitution of 21 split-GFP fragments may result from a non-specific irreversible interaction. Therefore, we used 22 the pr oximity labeling method as a complementary approach to validate the interaction of 23 AtPHF1 and AtSAR1b in planta. Proximity labeling starts after the application of biotin and 24 can be reversibly halted by removing biotin, thus allowing the capture of transient and dynamic 25 protein-protein interaction at a ten -nanometer scale (Xu et al., 2023) . Due to the superior 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint feasibility of miniTurbo (mTb) in the tobacco system (Mair et al., 2019) , the mTb-based 1 proximity labeling was used to assess whether AtPHF1 is in close proximity to AtSAR1b. For 2 unknown reasons, our initial attempt to express mTb-EYFP-AtPHF1 in agro-infiltrated tobacco 3 leaves was unsuccessful, so we generated the AtSAR1b-mTb-S11 construct by fusing mTb to 4 AtSAR1b (Fig. 4A). We also used t he cytosol-localized mTb-NES-EYFP, which carries the 5 nuclear export sequence (NES), as a negative control (Mair et al., 2019). Taking the advantage 6 that S11 can self -complement with GFP1–10, we could detect the expression and subcellular 7 distribution of AtSAR1b-mTb-S11 in agro -infiltrated tobacco leaves (Fig. 4B). We then co-8 expressed mTb-NES-EYFP or AtSAR1b-mTb-S11 with AtSEC12-mCherry or AtPHF1-9 mCherry and incubated the harvested leaves in a biotin solution, followed by tissue 10 homogenization and protein extraction. The protein extraction of non-infiltrated tobacco leaves 11 was used for the background comparison ( Fig. 4C). As the e xpression of the cytosolic mTb -12 NES-EYFP was much greater than the membrane-associated AtSAR1b-mTb-S11, we adjusted 13 the ratio of protein amount for AtSAR1b-mTb-S11: mTb-NES-EYFP to 30: 1. In such a way, 14 the protein amount s of mTb -NES-EYFP and AtSAR1b-mTb-S11 were comparable to fairly 15 compete with binding to streptavidin beads . When co-expressed with AtSAR1b-mTb-S11, 16 AtSEC12-mCherry and AtPHF1-mCherry could be immunoprecipitated by streptavidin and 17 detected by immunoblot, indicating that they were biotinylated (Fig. 4C). By contrast, the co-18 expression of mTb-NES-EYFP with AtSEC12-mCherry or AtPHF1-mCherry did not yield 19 similar effects (Fig. 4C). These results again suggested that AtPHF1 interacts with AtSAR1s in 20 planta. 21 22 Co-immunoprecipitation of the endogenous AtPHF1 with AtSAR1c-GFP in Arabidopsis 23 root 24 To answer whether the interaction of AtPHF1 and AtSAR1s is physiologically relevant, we 25 used the Arabidopsis transgenic plants expressing AtSAR1c-GFP or GFP under the UBQ10 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint promoter for co-immunoprecipitation (co-IP) (Zeng et al., 2015) . The Arabidopsis seedlings 1 were subjected to seven days of Pi starvation to mimic the physiological conditions in which 2 the endogenous AtPHF1 is upregulated and the ER export of AtPHT1 proteins is also highly 3 demanded. Total root protein extract was immunoprecipitated with GFP -Trap beads followed 4 by immunoblot using an anti-AtPHF1 specific antibody. Due to the differences in the expression 5 level between the cytosolic GFP and the membrane -associated AtSAR1c-GFP in transgenic 6 plants, we increased the protein amount for the co-IP of AtSAR1c-GFP by 50-fold than that for 7 the co-IP of GFP (Fig. 5A). Immunoblot of the IP suggested that the endogenous AtPHF1 was 8 co-immunoprecipitated with AtSAR1c-GFP but not with GFP (Fig. 5A). Conversely, when we 9 used equal amounts of root total protein for the co-IP experiments, the AtSAR1c-GFP 10 expression was ten times lower than the GFP expression but co-immunoprecipitated more 11 endogenous AtPHF1 (about 3.1-fold) (Fig. 5B), suggesting that AtSAR1c-GFP exhibits a higher 12 binding affinity toward AtPHF1. Moreover, the endogenous AtPHT1;1/2/3 proteins could also 13 be co-immunoprecipitated with AtSAR1c-GFP (Fig. 5A and 5B), indicating that the presence 14 of protein complex containing AtSAR1s, AtPHF1 and AtPHT1s. These data supported the 15 notion that by interacting with AtSAR1s, AtPHF1 participates in the initial phase of COPII 16 assembly for the ER export of AtPHT1s. 17 18 The cytosolic and transmembrane domains of AtPHF1 interact with AtSAR1b and 19 AtSAR1c 20 To determine which region of AtPHF1 is involved in the interacti on with AtSAR1s in 21 planta, we generated four N-terminally S10-tagged AtPHF1 truncated variants as follows: the 22 cytoplasmic domain alone (AtPHF1 N), the cytoplasmic domain with the transmembrane (TM) 23 domain (AtPHF1 N–TM), the TM domain alone (AtPHF1 TM), and the TM domain with the 24 ER-luminal domain (AtPHF1 TM–C) (Fig. 6A). The expression and the subcellular distribution 25 of these AtPHF1 variants suggested that S10-AtPHF1 N was present in the cytosol and nucleus, 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint while the other truncated forms of AtPHF1 localized to t he ER ( Supplementary Fig. S3A). 1 The subcellular localization of the C-terminally S11-tagged truncated forms of AtPHF1 also 2 showed similar results (Supplementary Fig. S3B), suggesting that the TM domain of AtPHF1 3 alone is sufficient for its ER targeting. As a ll the S10-tagged truncated AtPHF1 variants 4 interacted with AtSAR1b-S11 (Fig. 6B) and AtSAR1c-S11 (Fig. 6C), we concluded that both 5 the cytosolic and the TM domains of AtPHF1 contribute to the interaction with AtSAR1. 6 7 AtPHF1 preferentially interacts with the GDP-locked inactive form of AtSAR1b 8 Because the assembly and disassembly of COPII is controlled by the Sar1 GTPase cycle 9 (Van der Verren and Zanetti, 2023), we next asked whether AtPHF1 interacts with AtSAR1s in 10 a specific nucleotide-binding state . The wild -type AtSAR1b-S11 and the GTP-locked 11 AtSAR1b[H74L]-S11 localized to both the ER membrane and cytosol as previously reported 12 (daSilva et al., 2004; Wei and Wang, 2008) (Fig. 7A). The GDP-locked AtSAR1b[T34N]-S11 13 also displayed a similar subcellular pattern (Fig. 7A ). However, based on the biochemical 14 fractionation that distinguishes the soluble and microsomal fractions, the membrane-associated 15 proportion of the wild-type AtSAR1b-S11, the AtSAR1b[H74L]-S11 and the AtSAR1b[T34N]-16 S11 were 64%, 63%, and 51%, respectively, indicating that the GDP-locked form of AtSAR1b 17 was less membrane-associated (Supplementary Fig. S4). These results were in agreement with 18 the finding that upon the binding of GTP, SAR1 is activated to being inserted into the ER 19 membrane (d'Enfert, 1991; Paul et al., 2023) . More importantly, the in-planta tripartite split-20 GFP assay suggested that AtPHF1 interacted with the GDP-locked form of AtSAR1b but not 21 with the GTP-locked form of AtSAR1b (Fig. 7B ), indicating that , like SEC12, PHF1 22 preferentially interacts with the GDP -bound SAR1 and thus may be involved in recruiting 23 COPII components to the ER membranes. 24 25

26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint Although several ER accessory proteins have been identified to participate in the ER exit 1 of membrane proteins, the underlying molecular mechanisms are mostly unclear (Lau, 2000; 2 Kota and Ljungdahl, 2005) . In Arabidopsis, AUXIN RESISTANCE4 (AXR4) was shown to 3 interact with auxin influx carriers AUX1 and LAX2 to promote their PM targeting (Dharmasiri, 4 2006; Tidy et al., 2024). The KAONASHI3 (KNS3) family is recently identified to mediate the 5 ER exit of the boric acid channel AtNIP5;1 (Zhang et al., 2024). While AtPHF1 is required for 6 the PM targeting of AtPHT1s, it was proposed to act as an ER-localized chaperone for the 7 protein stability of AtPHT1s (Gonzalez et al., 2005) and not participate in the COPII vesicle 8 formation (Bayle et al., 2011). On the contrary, our data revealed a role for AtPHF1 in the early 9 step of COPII assembly through the interaction with AtSAR1s. Using the three different 10 methods, i.e., tripartite split-GFP complementation, proximity labeling, and co -11 immunoprecipitation, we showed that AtPHF1 interacted with AtSAR1s in planta. Furthermore, 12 the in-planta tripartite split-GFP assay demonstrated that AtPHF1 preferentially interacted with 13 the GDP-bound AtSAR1s. Our findings uncover that AtPHF1 likely mediates the assembly of 14 COPII vesicles by interacting with AtSAR1s, thus facilitating the ER export of AtPHT1s (Fig. 15 8). 16 The COPII recruitment to ERES is increased in response to the demand for ER export of 17 membrane cargoes (Hanton et al., 2007). Our results showed that the transient overexpression 18 of AtPHT1;1 in N. benthamiana leaves also triggered partial distribution of AtPHF1 into the 19 punctate structures associated with the ERES markers AtSAR1b, AtSAR1c and AtSEC24a (Fig. 20 1, Fig. 2, and Supplementary Fig. S2). In other words, under increased demand for the COPII-21 mediated export of AtPHT1s, AtPHF1 may partially relocate to the ERES to promote this 22 process. Although AtPHF1 lacks the conserved residues required for GEF activity and cannot 23 complement the yeast sec12 growth defect (Gonzalez et al., 20 05), we surmised that upon 24 AtPHT1;1 overexpression, the distribution of AtPHF1 to the ERES may facilitate the COPII 25 recruitment or assembly. We reasoned that w ithout high expression of AtPHT1s, observing 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint partial co-localization of AtPHF1 with ERES is beyond the detection limit of confocal imaging, 1 which may explain the discrepancy between the results from a previous study (Bayle et al., 2 2011) and ours. In support of this notion , we showed that in transgenic plants subjected to Pi 3 starvation, the endogenous AtPHF1 and AtPHT1;1/2/3 were co-immunoprecipitated with 4 AtSAR1c-GFP (Fig. 5 ), suggesting that the interaction of AtPHF1 and AtSAR1s is 5 physiologically relevant. In addition, AtPHF1 interacted with AtSAR1b and AtSAR1c but not 6 with other COPII-related components (Fig. 3B), indicating that AtPHF1 participates in the early 7 step of COPII assembly, thus likely assisting the loading of AtPHT1s into COPII vesicles for 8 the ER export. Furthermore, we demonstrated that AtPHF1 preferentially interacted with the 9 GDP-locked form of AtSAR1b in planta (Fig. 7B). These results indicated that AtPHF1 could 10 link the specific recognition of AtPHT1s with the modulation of AtSAR1s activity for the COPII 11 assembly. Such a mechanism is dissimilar to that reported for SED4, the yeast structural 12 homolog of SEC12 lacking GEF activity (Gimeno, 1995; Kodera et al., 2011) . Genetic and 13 biochemical evidence suggested that SED4 and SEC12 are not functionally interchangeable 14 and that SED4 weakly inhibits the GTPase-activating (GAP) activity of SEC23 toward SAR1 15 (Saito-Nakano and Nakano, 2000) . On the contrary, an in vitro study showed that SED4 had 16 stronger binding to the non-hydrolyzable GTP analog GMPPNP-bound SAR1 and stimulated 17 SEC23 GAP activity , thereby accelerating the dissociation of COPII coats from the ER 18 membrane (Gimeno, 1995; Kodera et al., 2011) . Unlike the interaction of SED4 and SEC16 19 required for the localization of SEC16 to ERES (Gimeno, 1995), the interaction of AtPHF1 and 20 AtSEC16a was not observed (Fig. 3B). As many COPII-related factors regulate SAR1 activity 21 through direct or indirect interaction (Van der Verren and Zanetti, 2023) , we propose that 22 AtPHF1 modulates the recruitment efficiency of AtSAR1s, thereby facilitating the formation of 23 COPII vesicles carrying AtPHT1s (Fig. 8). 24 AtPHF1 is an ER-resident membrane protein not incorporated into COPII vesicles (Bayle 25 et al., 2011) . Interestingly, we recently reported that the Arabidopsis CORNICHON 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint HOMOLOG 5 ( AtCNIH5) acts as an ER cargo receptor that cycles between the ER and the 1 early Golgi and selects AtPHT1s for ER export (Chiu et al., 2024; Liu et al., 2024) . AtCNIH5 2 is preferentially expressed in the outer layers of the root above the apical meristem and interacts 3 with AtPHF1 (Chiu et al., 2024). As AtPHF1 increased in the root of cnih5 and the cnih5 phf1 4 mutant showed more severe growth defects than the phf1 single mutant (Chiu et al., 2024), we 5 suspected that AtPHF1 and AtCNIH5 work interdependently . Namely, AtPHF1 couples the 6 recruitment of COPII proteins and AtPHT1s by simultaneously interacting with AtPHT1s and 7 AtSAR1s, while AtCNIH5 is involved in the selective packaging of AtPHT1 into COPII by 8 interacting with other COPII components. 9 AtPHF1 is structurally related to SEC12 proteins, which adopt a seven-bladed β propeller 10 (Gonzalez et al., 2005; Joiner and Fromme, 2021) . Based on the crystal structure, the 11 catalytically critical K loop within the cytoplasmic domain of yeast SEC12 makes direct contact 12 with nucleotide -free SAR1 (McMahon et al., 2012; Joiner and Fromme, 2021) . Our data 13 showed that both the cytosolic and the transmembrane domains of AtPHF1 interact with 14 AtSAR1b/c in planta (Fig. 6 ), indicat ing that the interaction interface may reside across 15 relatively large areas. Since the catalytic residues required for GEF activity are absent in 16 AtPHF1 (Gonzalez et al., 2005), it awaits to elucidate whether the interaction interface of PHF1-17 SAR1s is different from that of SEC12-SAR1s and how PHF1 modulates the recruitment of 18 SAR1s without GEF activity. 19 20

21 Plant material and growth conditions 22 The seeds of Arabidopsis thaliana wild-type (WT) (Columbia, Col-0) and phf1 were obtained 23 from the Arabidopsis Biological Resource Center (ABRC). Seeds of A. thaliana were surface-24 sterilized and germinated on one -half modified Hoagland agar plates, comprising 2.5 mM 25 Ca(NO3)2, 2.5 mM KNO3, 1 mM MgSO4, 50 μM [NaFe(III) EDTA], 250 μM KH2PO4, 10 μM 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint H3BO3, 0.2 μM Na2MoO4, 1 μM ZnSO4, 2 μM MnCl2, 0.5 μM CuSO4, 0.2 μM CoCl2, 0.5 mM 1 MES (pH 5.7), 1% Sucrose and 1% Bacto agar and grown in the growth chamber at 22℃ under 2 a 16-h-light/8-h-dark cycle. The Pi -sufficient (+P) and Pi -deficient (−P) media were one -half 3 modified Hoagland nu trient solutions containing 250 μM and 0 μM KH 2PO4, respectively. 4 Seeds of Nicotiana benthamiana (N. benthamiana) were surface-sterilized and germinated on 5 Murashige and Skoog medium diluted 2 -fold agar plates and grown in the growth chamber at 6 24℃ under a 16-h-light/8-h-dark cycle. 7 8 Plasmid construction 9 For the self -complementing split -GFP assay (Liu et al., 2018) , the constructs encoding 10 UBQ10:sXVE:GFP1–10 (Addgene plasmid #125663), UBQ10:sXVE:ER-GFP1–10 (Addgene 11 plasmid #125664) , UBQ10:sXVE:S11-GFP1–9 (Addgene plasmid #125665), or 12 UBQ10:sXVE:ER-S11-GFP1–9 (Addgene plasmid #125666) was used. For the tripartite split-13 GFP association, the constructs encoding UBQ10:sXVE:GFP1–9 (Addgene plasmid #108187), 14 UBQ10:sXVE:ER-GFP1–9, or 35S:GFP1–9.pMDC201 was used. For cons tructing the S10 -15 tagged clones, the CDS of AtPHF1 and AtSEC12 were amplified by PCR and subcloned into 16 UBQ10:sXVE:S10-(MCS) (Addgene plasmid #108177). For constructing the S11 -tagged 17 clones, the CDS of AtSTP1 was amplified by PCR and subcloned into UBQ10:sXVE:(MCS)-18 S11 (Addgene plasmid #108179). The CDS of AtPHT1;1 was amplified by PCR and subcloned 19 into UBQ10:sXVE:(MCS)-3xHA-S11 (Addgene plasmid #108180) . The CDS of AtSAR1b, 20 AtSAR1b[H74L], AtSAR1b[T34N], AtSAR1c, AtSEC24a, AtSEC13a, and AtSEC16a were 21 amplified and subcloned into UBQ10:(MCS)-S11. For constructing the GFP fusion clones, the 22 CDS of AtPHF1 was amplified and subcloned into UBQ10:(MCS)-GFP. For constructing the 23 mTb-fused clones, the CDS of mTb was amplified by PCR using R4pGWB601_UBQ10p - 24 miniTurbo-NES-YFP (Addgene plasmid #127369) as the template and subcloned into 25 UBQ10:AtSAR1b-S11, yielding UBQ10:AtSAR1b-mTb-S11. The abovementioned constructs 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint were generated either by restriction enzyme clonin g or by In -Fusion® HD cloning (Takara). 1 Primer sequences and the restriction enzyme cutting sites used are listed in Supplementary 2 Table S1. 3 4 Agro-infiltration of Nicotiana benthamiana leaves 5 Three- to four-week-old of Nicotiana benthamiana plants were used for A. tumefaciens (strain 6 EHA105)-mediated infiltration. The bacterial cultures were grown overnight in Luria -Bertani 7 (LB) medium containing appropriate antibiotics (50 μg/ml spectinomycin, 5 μg/ml rifampicin, 8 or 50 μg/ml kanamycin) at 30℃ with shaking at 200 rpm. After centrifugation, the pellet was 9 resuspended in infiltration buffer containing 10 mM MgCl 2, 10 mM MES (pH 7.5), 150 µM 10 acetosyringone to an OD600 of 1.0 and incubated at room temperature (RT) for 2–4 h. Tobacco 11 leaves were infiltrated with the agrobacterial suspension at a final OD 600 in the range of 0.01–12 1.0. For gene expression under the UBQ10:sXVE inducible promoter, the 9.18 µM ß-estradiol 13 was applied for additional 24 or 48 h. The N. benthamiana leaves at 48 or 72 h post-infiltration 14 were collected for confocal analysis or protein extraction. 15 16 Confocal image analysis 17 Confocal images were captured using the Zeiss LSM 800 microscope (Zeiss) equipped with 18 Plan-Apochromat 10×/0.45 M27, 20×/0.8 M27, and 40×/1.3 Oil DIC M27 objective s. The 19 acquisition was performed in multi-track mode with line switching, and the data were averaged 20 over two readings. Excitation/emission wavelengths were 488 nm/410–546 nm, 561 nm/560–21 617 nm, and 488 nm/656–700 nm for GFP, mCherry , and chlorophyll autofluorescence, 22 respectively. For the quantitative analysis of punctate structures in the epidermal cells of agro-23 infiltrated N. benthamiana leaves, the punctate structures were defined using the ImageJ 24 (Schneider et al., 2012) ‘analyze particles’ function, and the Pearson’s correlation coefficient 25 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint was measured using the Fiji ‘co-localization’ function. The box plots were generated by using 1 the BoxPlotR (Spitzer et al., 2014). 2 3 Proximity labeling assay 4 The N. benthamiana leaf tissues were incubated in 50 µM biotin solution (in DMSO) at RT for 5 30 min and ground with liquid N. The tissue powder was resuspended with two volumes (2 ml 6 per gram tissue) of lysis buffer containing 20 mM HEPES (pH 7.5), 40 mM KCl, 1 mM EDTA, 7 1% Triton X-100, 0.2 mM phenylmethylsulfonyl fluoride (PMSF), and 1× Protease Inhibitor 8 Cocktail (P9599, Sigma-Aldrich). After centrifugation at 4 ,000×g at 4℃ for 5 min, the 9 supernatant was subsequently centrifugated at 20,000×g at 4℃ for 15 min and the resulting 10 supernatant was collected . A 25–750 µg of t he crude extract was incubated with 5 µl 11 Streptavidin-agarose (S1638, Sigma-Aldrich) at 4℃ overnight under 5 rpm end-to-end rotation. 12 The beads collected by centrifugation at 2,500×g and 4 ℃ for 5 min were washed with a lysis 13 buffer containing 0.1% Triton X-100 once. After centrifugation at 2,500×g at 4`℃ for 5 min, 14 the beads were eluted with 2× SDS sample buffer containing 100 mM Tris-HCl (pH 6.8), 20% 15 glycerol, 4% SDS, and 100 mM dithiothreitol (DTT) with additional 0.4 M urea. 16 17 Co-immunoprecipitation assay 18 The 11-day-old Col-0, UBQ10:GFP, and UBQ10:AtSAR1c-GFP seedlings (Zeng et al., 2015) 19 were grown with 7 days of Pi starvation. The roots were harvested for total protein extraction 20 with lysis buffer containing 25 mM HEPES (pH 7.2), 150 mM NaCl, 0.5% IGEPAL CA -630 21 (I8896, Sigma -Aldrich), 2 mM EDTA, 0.5 mM MgCl 2 and 1× Protease I nhibitor Cocktail 22 (P9599, Sigma-Aldrich). The supernatants were obtained by centrifugation at 16,000×g at 4℃ 23 for 5 min and incubated with 2.5 mM dithiobis [succinimidyl propionate] (DSP) at RT for 30 24 min. Unreacted DSP was quenched with 100 mM Tris (pH 7.4) at RT for 15 min. A 15–750 µg 25 of protein mixture was incubated with 5 µl of GFP-Trap® Agarose (ChromoTek) at 4℃ for 2 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint h with end-over-end rotation at 5 rpm, washed with the lysis buffer once, centrifuged at 2,500×g 1 at 4℃ for 5 min, and eluted with 2× SDS sample buffer containing 100 mM Tris-HCl (pH 6.8), 2 20% glycerol, 4% SDS, and 100 mM DTT. 3 4 Immunoblot analysis 5 Protein samples were loaded onto 4 –12% Q -PAGE™ Bis-Tris Precast Gel (SMOBIO) and 6 transferred to polyvinylidene difluoride (PVDF) membranes (IPVH00010 Immobilon -P 7 Membrane, Merck). The membrane was blocked with 1% BSA in 1× PBS solution containing 8 0.2% Tween 20 (PBST, pH 7.2) at RT for 1 h and hybridized with primary antibodies: anti -9 AtPHF1 (Huang et al., 2013) at RT for 2 h, anti-PIP2;7 (1:5,000, AS22 4812, Agrisera), anti-10 mCherry (1:1 ,000, ab167453 , Abcam), and anti -S11 (1:10,000) at RT for 1 h , the anti -S11 11 polyclonal rabbit antibody was raised against the peptide of S11 12 (EKRDHMVLLEYVTAAGITDASC) and produced by LTK BioLaboratories, Taiwan. The 13 membrane was washed four times with 1× PBST for 5 min, followed by hybridization with the 14 horseradish peroxidase-conjugated secondary antibody (1:20,000 –1:40,000, GeneTex) in 15 blocking solution at RT for 1 h. After four washes in 1× PBST for 5 min and a rinse with distilled 16 water, chemiluminescent substrates (WesternBrightTM ECL and WesternBrightTM Sirius , 17 Advansta) for signal detection were applied. 18 19 Chemical treatment 20 Acetosyringone (150 mM stock in DMSO , D134406, Sigma-Aldrich) was diluted to a final 21 concentration of 150 µM in ddH 2O. β-estradiol (36. 5 mM stock in DMSO , E2758, Sigma-22 Aldrich) was diluted to a final concentration of 9.18 μM in ddH 2O. Dithiobis [succinimidyl 23 propionate] (25 mM stock in DMSO, 22586, ThermoFisher) was diluted to a final concentration 24 of 2.5 mM in ddH2O. 25 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint Accession Numbers 1 Sequence data from this article can be found in the Arabidopsis Genome Initiative under the 2 following accession numbers: AtPHF1 (At3g52190), AtSEC12 (At2g01470 ), AtPHT1;1 3 (At5g43350), AtSAR1b (At1g56330), AtSAR1c (At4g02080), AtSEC24a (At3g07100), 4 AtSEC13a (At3g01340), AtSEC16a (At5g47480). 5 6 Supplementary Data 7 Supplementary Figure S1. Expression of AtSAR1 isoforms in Col-0 under Pi deprivation 8 by RNA-seq analysis. Expression of AtSAR1a/b/c/d/e in the shoot and root of 10-day-old WT 9 (Col-0) seedlings under Pi-sufficient conditions (+P0) or under one day (–P1) and three days of 10 Pi starvation ( –P3) as previou sly described (Liu et al., 2016) . RPKM stands for reads per 11 kilobase of transcript per million mapped reads. Numbers represent average RPKM values of 12 two replicates, with an assigned value of 0.25 for readings below this threshold. 13 14 Supplementary Figure S2. AtPHT1;1-induced AtPHF1 puncta co-localize with AtSAR1c 15 in agro-infiltrated N. benthamiana leaves. (A) Expression and distribution of AtSAR1b-S11, 16 AtSAR1c-S11, and AtSEC24a-S11. The confocal images were taken at the peripheral layer of 17 epidermal cells. Scale bars, 5 µm. (B) Co -expression of AtPHT1;1, AtPHF1-mCherry, and 18 cytosolic GFP1 –10 with AtSAR1c-S11. Arrows indicate the punctate structures labeled by 19 AtPHF1-mCherry correspond to those labeled by the ERES marker. The puncta not overlapped 20 are circled. Scale bars, 5 µm. (C) Distribution of the Pearson’s correlation coefficient between 21 AtPHF1-mCherry and GFP signals of the ERES markers. The number of punctate structures 22 used for the quantification analysis is shown in the parentheses. Data were collected from three 23 independent experiments. For the box plot, center lines show the medians; box limits indicate 24 the 25 th and 75 th percentiles; whiskers extend to minimum and maximum values ; crosses 25 represent means. Data used for quantification analysis of AtSAR1b is the same as in Fig. 2C. 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint 1 Supplementary Figure S3. Subcellular distribution of truncated AtPHF1 variants in agro-2 infiltrated N. benthamiana leaves. (A and B) The expression and distribution of the S10 -3 tagged truncated AtPHF1 variants (A) and the S11-tagged truncated AtPHF1 variants (B). Scale 4 bars, 20 µm. 5 6 Supplementary Figure S4. Detection of AtSAR1b/[H74L]/[T34N] in the soluble and 7 microsomal fract ions. Protein expression of AtSAR1b/[H74L]/[T34N]-S11 in the agro -8 infiltrated tobacco leaves. The soluble (S) and microsomal (M) fractions were isolated by the 9 low-speed pellet (LSP) method, as previously reported (Yoshimoto et al., 2004) . Anti-S11 10 antibody was used to detect S11 fusion proteins. The membrane association was calculated as 11 a percentage of the intensity of the M fraction by the sum of the S and M fractions. Detection 12 of the plasma membrane aquaporin PIP2;7 was used as microsomal fraction control. Amido 13 black staining was used as loading control. 14 15 Supplementary Table S1. Oligonucleotides used for plasmid constructs. 16 17 Funding 18 This work was supported by the Ministry of Science and Technology (MOST 108-2311-B-007-19 003-MY3) and the National Science and Technology Council (NSTC 112 -2313-B-007-001-20 MY3). 21 22 Acknowledgments 23 We thank Dr. Liwen Jiang at the Chinese University of Hong Kong, People's Republic of China, 24 for sharing the Arabidopsis seeds of UBQ10:AtSAR1c-GFP and UBQ10:GFP homozygous 25 lines, Dr. Masaki Takeuchi at the University of Tokyo, Japan, for sharing the clone 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint 35S:AtSAR1b[H74L], and Dr. Tzyy -Jen Chiou at Academia Sinica, Taiwan (R.O.C.), for 1 sharing the clone 35S:AtPHF1-mCherry. We acknowledge Ms. Wen -Chun Chou for 2 constructing the plasmid encoding 35S:GFP1–10 and the technical support from the confocal 3 image core, National Tsing Hua University (sponsored by MOST 108-2731-M-007-001, and 4 MOST 110-2731-M-007-001). 5 6 Author Contributions 7 T.-Y. L. conceived the original research plan, designe d and supervised the experiments and 8 wrote the article. H. -F. L. designed and performed the experiments, analyzed the data, wrote 9 the article and contributed to the figure preparation. J.-D. C. designed and performed the 10 experiments. 11 12 Disclosures 13 The authors have no conflicts of interest to declare. 14 15 Figure Legends 16 Figure 1. AtPHT1;1 overexpression triggers partial distribution of AtPHF1 to the punctate 17 structures in agro-infiltrated N. benthamiana leaves. (A) Distribution of AtPHF1-GFP in the 18 absence or presence of the co-expression of AtPHT1;1-S11 or AtSTP1-S11. Arrows indicate 19 punctate structures. Scale bars, 5 µm. (B) Quantification of AtPHF1-GFP-labeled punctate 20 structures. Images were collected from 4–5 independent experiments. The number of regions 21 of interest (ROI) used for quantification is shown in parentheses. For the box plot, center lines 22 show the medians; box limits indicate the 25 th and 75th percentiles; whiskers extend to the 5th 23 and 95 th percentiles; data points are plotted as dots. Data significantly different from the 24 AtPHF1-GFP expression alone are indicated (*** P < 0.001; Student’s t-test, two-tail). 25 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint Figure 2. AtPHT1;1-induced AtPHF1 puncta partially co-localizes with the ERES 1 markers in agro-infiltrated N. benthamiana leaves. (A and B) Co-localization of AtSAR1b-2 S11 (A) and AtSEC24a-S11 (B) with AtPHF1-mCherry in the absence or presence of AtPHT1;1 3 overexpression. Images were taken at the peripheral layer of the epidermis. Arrows indicate the 4 co-localization of punctate structures between the AtPHF1-mCherry and the ERES markers. 5 The non-overlapping puncta are circled. Scale bars, 5 µm. (C) Distribution of the Pearson ’s 6 correlation coefficient between AtPHF1-mCherry and GFP signals of the ERES markers. The 7 number of punctate structures used for the quantification analysis is shown in the parentheses. 8 Data were collected from three independent experiments. For the box plot, center lines show 9 the medians; box limits indicate the 25th and 75th percentiles; whiskers extend to minimum and 10 maximum values; crosses represent means. Data significantly different from the AtSAR1b are 11 indicated (*** P < 0.001; Student’s t-test, two-tail). 12 13 Figure 3. AtPHF1 interacts with AtSAR1b and AtSAR1c in planta based on the tripartite 14 split-GFP assay in agro -infiltrated N. benthamiana leaves. (A) The expression and 15 distribution of S10-AtSEC12, S10 -AtPHF1, and the COPII-related components AtSAR1b-, 16 AtSAR1c-, AtSEC24a-, AtSEC16a-, and AtSEC13a-S11. Scale bars, 10 µm. (B) The interaction 17 of S10-AtSEC12 or S10-AtPHF1 with the COPII-related components AtSAR1b-, AtSAR1c-, 18 AtSEC24a-, AtSEC16a-, and AtSEC13a-S11. Scale bars, 5 µm. 19 20 Figure 4. AtPHF1 interacts with AtSAR1b in planta based on the proximity labeling in 21 agro-infiltrated N. benthamiana leaves. (A) Schematic diagrams of mTb-fused bait protein in 22 the proximity of prey protein. The reactive intermediate biotinyl- 5'-AMP biotinylates proximal 23 prey proteins. The grey region indicates the distance range of biotinylation. (B) The expression 24 and subcellular distribution of mTb-NES-EYFP and AtSAR1b-mTb-S11. Scale bars, 20 µm. (C) 25 The i nteraction analysis of AtPHF1-mCherry and AtSAR1b by proximity labeling. Co-26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint expression of mTb -NES-EYFP or AtSAR1b-mTb-S11 with AtSEC12-mCherry or AtPHF1-1 mCherry. Anti-S11 antibody was used to detect EYFP-tagged and S11-tagged fusion proteins. 2 3 Figure 5. Co-immunoprecipitation of AtSAR1c-GFP with the endogenous AtPHF1 and 4 AtPHT1;1/2/3 in Arabidopsis transgenic plants . (A and B) The interaction analyses of 5 AtSAR1c-GFP with the endogenous AtPHF1 and AtPHT1;1/2/3 in 11-day-old wild-type (Col-6 0), UBQ10:GFP, and UBQ10:AtSAR1c-GFP seedlings subjected to Pi deprivation (0 µM 7 KH2PO4, 7 days of starvation). The protein amounts used for input and IP in (A) and (B) were 8 shown as indicated. Anti-S11 antibody was used to detect GFP fusion proteins. 9 10 Figure 6. AtSAR1b and AtSAR1c interact with the cytosolic and transmembrane domains 11 of AtPHF1 in agro-infiltrated N. benthamiana leaves. (A) Schematic diagram of the truncated 12 AtPHF1 variants. Domain boundaries are indicated by amino acid numbers . (B and C) The 13 interaction analyses of the S10-AtPHF1 truncated variants and AtSAR1b-S11 (B) or AtSAR1c-14 S11 (C). Scale bars, 10 µm. 15 16 Figure 7. AtPHF1 preferentially interacts with the GDP-locked inactive form of AtSAR1b 17 in agro -infiltrated N. benthamiana leaves. (A) The expression and distribution of 18 AtSAR1b/[H74L]/[T34N]-S11. (B) The interaction analyses of S10-AtSEC12 or S10-AtPHF1 19 with AtSAR1b/[H74L]/[T34N]-S11. Chlorophyll autofluorescence was used as a chloroplast 20 marker. Scale bars, 5 µm. 21 22 Figure 8. Hypothetical model depicting a role for AtPHF1 in COPII recruitment for the 23 ER export of AtPHT1 proteins. The GEF SEC12 recruits and activates SAR1 GTPase by 24 exchanging of GDP to GTP. The direct interaction of GTP -bound SAR1 with SEC23 recruits 25 SEC23/24 to form the inner coat, followed by SEC13/31 recruitment to form the outer coat . 26 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint The ER -resident PHF1 concentrates on ERES triggered by increased PHT1 proteins and 1 interacts with GDP-bound SAR1. While the scaffold protein SEC16 at ERES does not interact 2 with PHF1, t he interaction of PHF1 and SAR1 may link to the regulation of SAR1 GTPase 3 cycle, thereby accelerating COPII assembly at ERES to facilitate the ER-to-Golgi transport of 4 PHT1 proteins in the budding COPII vesicles. 5 6

7 Barlowe CK, Miller EA (2013) Secretory protein biogenesis and traffic in the early secretory 8 pathway. Genetics 193: 383-410 9 Bayle V, Arrighi JF, Creff A, Nespoulous C, Vialaret J, Rossignol M, Gonzalez E, Paz -Ares J, 10 Nussaume L (2011) Arabidopsis thaliana high -affinity phosphate transporters exhibit 11 multiple levels of posttranslational regulation. The Plant Cel 23: 1523-1535 12 Bi X, Corpina, R. A., & Goldberg, J. (2002) Structure of the Sec23/24-Sar1 pre-budding complex 13 of the COPII vesicle coat. Nature 419: 271-277 14 Cabantous S, Nguyen HB, Pedelacq JD, Koraichi F, Chaudhary A, Ganguly K, Lockard MA, 15 Favre G, Terwilliger TC, Waldo GS (2013) A new protein -protein interaction sensor 16 based on tripartite split-GFP association. Scientific reports 3: 2854 17 Chiu C-Y, Ts a i C-D, Wang J-Y, Ts a i M-H, McGinness AJ, Lung H-F, Kanno S, Kriechbaumer V, Liu 18 T-Y (2024) Phosphate starvation -induced CORNICHON HOMOLOG 5 as endoplasmic 19 reticulum cargo receptor for PHT1 transporters in Arabidopsis. bioRxiv 06 20 Chung KP , Zeng Y , Jiang L (2016) COPII paralogs in plants: functional redundancy or diversity? 21 Trends in Plant Science 21: 758-769 22 Connerly PL, Esaki M, Montegna EA, Strongin DE, Levi S, Soderholm J, Glick BS (2005) Sec16 23 is a determinant of transitional ER organization. Current Biolog 15: 1439-1447 24 d'Enfert C, Wuestehube, L. J., Lila, T., & Schekman, R. (1991) Sec12p-dependent membrane 25 binding of the small GTP -binding protein Sar1p pro motes formation of transport 26 vesicels from the ER. The Journal of cell biology 114: 663-670 27 daSilva LL, Snapp EL, Denecke J, Lippincott -Schwartz J, Hawes C, Brandizzi F (2004) 28 Endoplasmic reticulum export sites and Golgi bodies behave as single mobile secretory 29 units in plant cells. The Plant Cell 16: 1753-1771 30 Dharmasiri S, Swarup, R., Mockaitis, K., Dharmasiri, N., Singh, S. K., Kowalchyk, M., ... & 31 Estelle, M. (2006) AXR4 is required for localization of the auxin influx facilitator AUX1. 32 Science 312: 1218-1220 33 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint Gimeno RE, Espenshade, P ., & Kaiser, C. A. (1995) SED4 encodes a yeast endoplasmic 1 reticulum protein that binds Sec16p and participates in vesicle formation. The Journal 2 of cell biology 131: 325-338 3 Gonzalez E, Solano R, Rubio V, Leyva A, Paz-Ares J (2005) PHOSPHATE TRANSPORTER TRAFFIC 4 FACILITATOR1 is a plant-specific SEC12-related protein that enables the endoplasmic 5 reticulum exit of a high-affinity phosphate transporter in Arabidopsis. The Plant Cell 17: 6 3500-3512 7 Hanton SL, Chatre L, Matheson LA, Rossi M, Held MA, Brandizzi F (2008) Plant Sar1 isoforms 8 with near -identical protein sequences exhibit different localisations and effects on 9 secretion. Plant Molecular Biology 67: 283-294 10 Hanton SL, Chatre L, Renna L, Matheson LA, Brandizzi F (2007) De n ovo formation of plant 11 endoplasmic reticulum export sites is membrane cargo induced and signal mediated. 12 Plant Physiology 143: 1640-1650 13 Hanton SL, Matheson LA, Chatre L, Brandizzi F (2009) Dynamic organization of COPII coat 14 proteins at endoplasmic reticulum export sites in plant cells. The Plant Journal 57: 963-15 974 16 Huang TK, Han CL, Lin SI, Chen YJ, Tsai YC, Chen YR, Chen JW, Lin WY , Chen PM, Liu TY , Chen 17 YS, Sun CM, Chiou TJ (2013) Identification of downstream components of ubiquitin -18 conjugating enzyme PH OSPHATE2 by quantitative membrane proteomics in 19 Arabidopsis roots. The Plant Cell 25: 4044-4060 20 Joiner AMN, Fromme JC (2021) Structural basis for the initiation of COPII vesicle biogenesis. 21 Structure 29: 859-872 e856 22 Kodera C, Yorimitsu T, Nakano A, Sato K (2011) Sed4p stimulates Sar1p GTP hydrolysis and 23 promotes limited coat disassembly. Traffic 12: 591-599 24 Kota J, Ljungdahl PO (2005) Specialized membrane-localized chaperones prevent aggregation 25 of polytopic proteins in the ER. The Journal of cell biology 168: 79-88 26 Kurokawa K, Nakano A (2019) The ER exit sites are specialized ER zones for the transport of 27 cargo proteins from the ER to the Golgi apparatus. The journal of biochemistry 165: 28 109-114 29 Lau WTW, Howson, R. W., Malkus, P ., Schekman, R., & O'S hea, E. K. (2000) Pho86p, an 30 endoplasmic reticulum (ER) resident protein in Saccharomyces cerevisiae, is required 31 for ER exit of the high -affinity phosphate transporter Pho84p. Proceedings of the 32 National Academy of Sciences 97: 1107-1112 33 Li B, Zeng Y , Jiang L (2022) COPII vesicles in plant autophagy and endomembrane trafficking. 34 FEBS letters 596: 2314-2323 35 Li B, Zeng, Y ., Cao, W., Zhang, W., Cheng, L., Yin, H., ... & Jiang, L. (2021) A distinct giant coat 36 protein complex II vesicle population in Arabidopsi s thaliana. Nature Plants 7: 1335-37 1346 38 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint Liang X, Li SW, Gong LM, Li S, Zhang Y (2020) COPII components Sar1b and Sar1c play distinct 1 yet interchangeable roles in pollen development. Plant Physiology 183: 974-985 2 Liu T -Y, Ts a i M-H, Wang J -Y, L u n g H-F, Chow H -X, Chiu C -Y, L u C-A (2024) CORNICHON 3 HOMOLOG 5 -dependent ER export of membrane cargoes in phosphate -starved 4 Arabidopsis root as revealed by membrane proteomic analysis. bioRxiv 12 5 Liu TY , Chou WC, Chen WY , Chu CY , Dai CY , Wu PY (2018) Detection of membran e protein-6 protein interaction in planta based on dual -intein-coupled tripartite split -GFP 7 association. The Plant Journal 94: 426-438 8 Liu TY , Huang TK, Yang SY , Hong YT, Huang SM, Wang FN, Chiang SF, Tsai SY , Lu WC, Chiou TJ 9 (2016) Identification of plant v acuolar transporters mediating phosphate storage. 10 Nature communications 7: 11095 11 Mair A, Xu SL, Branon TC, Ting AY , Bergmann DC (2019) Proximity labeling of protein 12 complexes and cell -type-specific organellar proteomes in Arabidopsis enabled by 13 TurboID. elife 8 14 McMahon C, Studer SM, Clendinen C, Dann GP , Jeffrey PD, Hughson FM (2012) The structure 15 of Sec12 implicates potassium ion coordination in Sar1 activation. Journal of Biological 16 Chemistry 287: 43599-43606 17 Miller E, Antonny, B., Hamamoto, S., & Schekma n, R. (2002) Cargo selection into COPII 18 vesicles is driven by the Sec24p subunit. The EMBO journal: 6105-6113 19 Nussaume L, Kanno S, Javot H, Marin E, Pochon N, Ayadi A, Nakanishi TM, Thibaud MC (2011) 20 Phosphate import in plants: focus on the PHT1 transporters. Frontiers in plant science 21 2: 83 22 Paul S, Audhya A, Cui Q (2023) Molecular mechanism of GTP binding -and dimerization -23 induced enhancement of Sar1 -mediated membrane remodeling. Proceedings of the 24 National Academy of Sciences 120: e2212513120 25 Saito-Nakano Y , Nakano A (2000) Sed4p functions as a positive regulator of Sar1p probably 26 through inhibition of the GTPase activation by Sec23p. Genes to Cells 5: 1039-1048 27 Saito K, Yamashiro K, Shimazu N, Tanabe T, Kontani K, Katada T (2014) Concentration of Sec12 28 at ER exit sites via interaction with cTAGE5 is required for collagen export. Journal of 29 Cell Biology 206: 751-762 30 Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. 31 Nature methods 9: 671-675 32 Shen J, Yuan, L., Zhang, J., Li, H., Bai, Z., Chen, X., ... & Zhang, F. (2011) Phosphorus dynamics: 33 from soil to plant. Plant physiology 156: 997-1005 34 Spitzer M, Wildenhain J, Rappsilber J, Tyers M (2014) BoxPlotR: a web tool for generation of 35 box plots. Nature methods 11: 121-122 36 Stagg SM, Gurkan C, Fowler DM, LaPointe P , Foss TR, Potter CS, Carragher B, Balch WE (2006) 37 Structure of the Sec13/31 COPII coat cage. Nature 439: 234-238 38 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint Supek F, Madden DT, Hamamoto S, Orci L, Schekman R (2002) Sec16p potentiates the action 1 of COPII proteins to bud transport vesicles. The Journal of cell biology 158: 1029-1038 2 Takagi J, Renna L, Takahashi H, Koumoto Y , Tamura K, Stefano G, Fukao Y , Kondo M, 3 Nishimura M, Shimada T, Brandizzi F, Hara-Nishimura I (2013) MAIGO5 functions in 4 protein export from Golgi-associated endoplasmic reticulum exit sites in Arabidopsis. 5 The Plant Cell 25: 4658-4675 6 Tidy A, Abu Bakar N, Carrier D, Kerr ID, Hodgman C, Bennett MJ, Swarup R (2024) Mechanistic 7 insight into the role of AUXIN RESISTANCE4 in trafficking of AUXIN1 and LIKE AUX1 -2. 8 Plant Physiology 194: 422-433 9 Van der Verren SE, Zanetti G (2023) The small GTPase Sar1, control centre of COPII trafficking. 10 FEBS letters 597: 865-882 11 Wang X, Xu M, Gao C, Zeng Y , Cui Y , Shen W, Jiang L (2020) The roles of endomembrane 12 trafficking in plant abiotic stress responses. Journal of integrative plant biology 62: 55-13 69 14 Wei T, Wang A (2008) Biogenesis of cytoplasmic membranous vesicles for plant potyvirus 15 replication occurs at endoplasmic reticulum exit sites in a COPI -and COPII-dependent 16 manner. Journal of virology 82: 12252-12264 17 Weissman JT, Plutner H, Balch WE (2001) The mammalian guanine nucleotide exchange factor 18 mSec12 is essential for activation of the Sar1 GTPase directing endoplasmic reticulum 19 export. Traffic 2: 465-475 20 Xu SL, Shrestha R, Karunadasa SS, Xie PQ (2023) Proximity labeling in plants. Annual review 21 of plant biology 74: 285-312 22 Yoshimoto K, Hanaoka H, Sato S, Kato T, Tabata S, Noda T, Ohsumi Y (2004) Processing of 23 ATG8s, ubiquitin-like proteins, and their deconjugation by ATG4s are essential for plant 24 autophagy. The Plant Cell 16: 2967-2983 25 Zeng Y , Chung KP , Li B, Lai CM, Lam SK, Wang X, Cui Y , Gao C, Luo M, Wong KB, Schekman R, 26 Jiang L (2015) Unique COPII component AtSar1a/AtSec23a pair is required for the 27 distinct function of protein ER export in Arabidopsis thaliana. Proceedings of the 28 National Academy of Sciences 112: 14360-14365 29 Zhang Z, Nakamura S, Yamasaki A, Uehara M, Takemura S, Tsuchida K, Kamiya T, Shigenobu 30 S, Yamaguchi K, Fujiwara T, Ishiguro S, Takano J (2024) Arabidopsis KNS3 and its two 31 homologs mediate endoplasmic reticulum -to-plasma membrane traffic of boric acid 32 channels. Journal of Experimental Botany 75: 7046-7065 33 34 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint A B AtPHF1-GFP + AtPHT1;1-S11 AtPHF1-GFP + AtSTP1-S11 AtPHF1-GFP Figure 1. AtPHT1;1 overexpression triggers partial distribution of AtPHF1 to the punctate structures in agro-infiltrated N. benthamiana leaves. (A) Distribution of AtPHF1-GFP in the absence or presence of the co-expression of AtPHT1;1-S11 or AtSTP1-S11. Arrows indicate punctate structures. Scale bars, 5 µm. (B) Quantification of AtPHF1-GFP-labeled punctate structures. Images were collected from 4–5 independent experiments. The number of regions of interest (ROI) used for quantification is shown in parentheses. For the box plot, center lines show the medians; box limits indicate the 25th and 75th percentiles; whiskers extend to the 5th and 95th percentiles; data points are plotted as dots. Data significantly different from the AtPHF1-GFP expression alone are indicated (*** P < 0. 001; Student’s t-test, two-tail). AtPHF1 -GFP AtPHF1 -GFP + AtPHT1;1 -S11 AtPHF1 -GFP + AtSTP1 -S11 *** Number of AtPHF1-GFP puncta per 1000 µm2 (n=25) (n=33) (n=25) preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint A B C AtPHF1-mCherry + AtPHT1;1 + AtPHT1;1 + AtPHT1;1 AtSAR1b-S11 + GFP1–10merged AtPHF1-mCherry AtSEC24a-S11 + GFP1–10 + AtPHT1;1 + AtPHT1;1 + AtPHT1;1 merged Pearson’s coefficient (r) AtSAR1b AtSEC24a (n=173) (n=142) Figure 2. AtPHT1;1-induced AtPHF1 puncta partially colocalizes with the ERES markers in agro-infiltrated N. benthamiana leaves. (A and B) Co-localization of AtSAR1b-S11 (A) and AtSEC24a- S11 (B) with AtPHF1-mCherry in the absence or presence of AtPHT1;1 overexpression. Images were taken at the peripheral layer of the epidermis. Arrows indicate the colocalization of punctate structures between the AtPHF1-mCherry and the ERES markers. The non-overlapping puncta are circled. Scale bars, 5 µm. (C) Distribution of the Pearson’s correlation coefficient between AtPHF1-mCherry and GFP signals of ERES markers. The number of punctate structures used for the quantification analysis is shown in the parentheses. Data were collected from three independentexperiments. For the box plot, center lines show the medians; box limits indicate the 25th and 75th percentiles; whiskers extend to minimum and maximum values; crosses represent means. Data significantly different from the AtSAR1b are indicated (*** P < 0.001; Student’s t-test, two-tail). preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint A B GFP1–10S11-GFP1–9 S10-AtSEC12 S10-AtPHF1 AtSAR1b-S11 AtSEC24a-S11 AtSEC16a-S11 AtSEC13a-S11 AtSAR1c-S11 S10-AtSEC12 + GFP1–9 S10-AtPHF1 + GFP1–9 AtSAR1b-S11 AtSEC24a-S11 AtSEC16a-S11 AtSEC13a-S11 AtSAR1c-S11 AtSAR1b-S11 AtSEC24a-S11 AtSEC16a-S11 AtSEC13a-S11 AtSAR1c-S11 Figure 3. AtPHF1 interacts with AtSAR1b and AtSAR1c in planta based on the tripartite split-GFP assay in agro-infiltrated N. benthamianaleaves. (A) The expression and distribution of S10-AtSEC12, S10-AtPHF1, and the COPII-related components AtSAR1b-, AtSAR1c-, AtSEC24a-, AtSEC16a-, and AtSEC13a-S11. Scale bars, 10 µm. (B) The interaction of S10-AtSEC12 or S10-AtPHF1 with the COPII-related components AtSAR1b-, AtSAR1c-, AtSEC24a-, AtSEC16a-, and AtSEC13a-S11. Scale bars, 5 µm. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint mTb-NES-EYFP AtSAR1b-mTb-S11 + GFP1–10 A B C Figure 4. AtPHF1 interacts with AtSAR1b in planta based on the proximity labeling in agro-infiltrated N. benthamiana leaves. (A) Schematic diagrams of mTb-fused bait protein in the proximity of prey protein. The reactive intermediate biotinyl- 5'-AMP biotinylates proximal prey proteins. The grey region indicates the distance range of biotinylation. (B) The expression and subcellular distribution of mTb-NES-EYFP and AtSAR1b-mTb- S11. Scale bars, 20 µm. (C) The interaction analysis of AtPHF1-mCherry and AtSAR1b by proximity labeling. Co-expression of mTb-NES-EYFP or AtSAR1b-mTb-S11 with AtSEC12-mCherry or AtPHF1-mCherry. Anti-S11 antibody was used to detect EYFP-tagged and S11-tagged fusion proteins. mTb-NES-EYFP AtSAR1b-mTb-S11 AtPHF1-mCherry AtSEC12-mCherry - - - - + - - + + - + - - + - + - + + - 57 — AtSEC12-mCherry AtPHF1-mCherry Input: α-S11 Input: α-mCherry 57 — 42 — 72 — AtSAR1b-mTb-S11 mTb-NES-EYFP IP: α-mCherry 57 — 72 — AtSEC12-mCherry AtPHF1-mCherry kDa Protein amount 1x 1x 30x 1x 30x Protein amount 25x 25x 750x 25x 750x preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint A B Figure 5. Co-immunoprecipitationof AtSAR1c-GFP with the endogenous AtPHF1 and AtPHT1;1/2/3 in Arabidopsis transgenic plants. (A and B) The interaction analyses of AtSAR1c-GFP with the endogenous AtPHF1 and AtPHT1;1/2/3 in 11-day-old wild-type (Col-0), UBQ10:GFP, and UBQ10:AtSAR1c-GFP seedlings subjected to Pi deprivation(0 µM KH2PO4, 7 days of starvation). The protein amounts used for input and IP in (A) and (B) were shown as indicated. Anti-S11 antibody was used to detect GFP fusion proteins. Col-0 GFP AtSAR1c -GFP 57 — 42 — 31 — 24 — AtSAR1c-GFP GFP Input: α-S11 42 — Input: α-AtPHF1 AtPHF1 42 — 31 — Input: α-AtPHT1;1/2/3 AtPHT1;1/2/3 AtPHF1IP: α-AtPHF1 42 — 42 — 31 — IP: α-AtPHT1;1/2/3 AtPHT1;1/2/3 kDa Protein amount 1x 1x 50x Protein amount 10x 10x 500x Input: α-S11 57 — 42 — 31 — 24 — Relative intensity: 1.0 0.1 Relative intensity: 1.0 3.1 IP: α-AtPHF1 42 — Input: α-AtPHT1;1/2/3 42 — 31 — IP: α-AtPHT1;1/2/3 42 — 31 — kDa Input: α-AtPHF1 AtPHF142 — Col-0 GFP AtSAR1c -GFP Protein amount 1x 1x 1x AtPHT1;1/2/3 AtSAR1c-GFP GFP Protein amount 100x 100x 100x AtPHF1 AtPHT1;1/2/3 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint AtSAR1c-S11 + GFP1–9 S10-AtPHF1 N S10-AtPHF1 N–TM S10-AtPHF1 TM S10-AtPHF1 TM–C B C A AtSAR1b-S11 + GFP1–9 S10-AtPHF1 N S10 -AtPHF1 N–TM S10-AtPHF1 TM S10-AtPHF1 TM–C S10 N 333 359 398 TM C S10 N S10 N TM S10 TM S10 TM C S10-AtPHF1 S10-AtPHF1 N–TM S10-AtPHF1 TM S10-AtPHF1 TM–C S10-AtPHF1 N Figure 6. AtSAR1b and AtSAR1c interact with the cytosolic and transmembrane domains of AtPHF1 in agro-infiltrated N. benthamiana leaves. (A) Schematic diagram of the truncated AtPHF1 variants. Domain boundariesare indicated by amino acid numbers. (B and C) The interaction analyses of the S10-AtPHF1 truncated variants and AtSAR1b-S11 (B) or AtSAR1c-S11 (C). Scale bars, 10 µm. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint B A AtSAR1b-S11 AtSAR1b[H74L]-S11 GFP1–10 AtSAR1b[T34N]-S11 AtSAR1b-S11 AtSAR1b[H74L]-S11 S10-AtPHF1 + GFP1–9 AtSAR1b[T34N]-S11 S10-AtSEC12 + GFP1–9 Figure 7. AtPHF1 preferentially interacts with the GDP-locked inactive form of AtSAR1b in agro- infiltrated N. benthamiana leaves. (A) The expression and distribution of AtSAR1b/[H74L]/[T34N]-S11. (B) The interaction analyses of S10-AtSEC12 or S10-AtPHF1 with AtSAR1b/[H74L]/[T34N]-S11. Chlorophyll autofluorescencewas used as a chloroplastmarker. Scale bars, 5 µm. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint Figure 8. Hypothetical model depicting a role for AtPHF1 in COPII recruitment for the ER export of AtPHT1 proteins. The GEF SEC12 recruits and activates SAR1 GTPase by exchanging of GDP to GTP. The direct interaction of GTP-bound SAR1 with SEC23 recruits SEC23/24 to form the inner coat, followed by SEC13/31 recruitmentto form the outer coat. The ER-resident PHF1 concentrateson ERES triggered by increased PHT1 proteins and interacts with GDP-bound SAR1. While the scaffold protein SEC16 at ERES does not interact with PHF1, the interaction of PHF1 and SAR1 may link to the regulation of SAR1 GTPase cycle, thereby accelerating COPII assembly at ERES to facilitate the ER-to-Golgi transport of PHT1 proteins in the budding COPII vesicles. preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted January 10, 2025. ; https://doi.org/10.1101/2025.01.09.632146doi: bioRxiv preprint