Introduction

Auxin is required for flower initiation in Arabidopsis and other plants (Cheng and Zhao, 2007; Yamaguchi et al., 2013). Several Arabidopsis mutants and mutant combinations fail to initiate flowers on inflorescences, despite the fact that they can undergo the normal transition from vegetative growth to reproductive growth, resulting in the formation of pin-like inflorescences (Bennett et al., 1995; Przemeck et al., 1996; Galweiler et al., 1998; Cheng et al., 2007). All of the known pin-like Arabidopsis mutants are caused by defects in some aspects of auxin biology. PIN-FORMED 1 (PIN1) was the first characterized gene that causes the formation of pin-like inflorescences when compromised (Galweiler et al., 1998). PIN1 is an auxin efflux carrier that transports auxin from the cytosol into the extracellular space (Yang et al., 2022). Plants grown on media containing the polar auxin transport inhibitor N-1-naphthylphthalamic acid (NPA) phenocopy pin1 mutants (Okada et al., 1991). It has been shown that NPA can directly bind to PIN1 to inhibit auxin transport (Yang et al., 2022). Disruption of PINOID (PID), which encodes a Ser/Thr protein kinase, also leads to the formation of pin-like inflorescences (Smyth and Alvarez, 1995; Christensen et al., 2000). Overexpression of PID leads to phenotypes similar to the well-characterized auxin resistant mutants, linking PID functions to auxin (Christensen et al., 2000). NAKED PINS IN YUC MUTANTS 1(NPY1) (Cheng et al., 2007), which is also called ENHANCER OF PINOID (ENP) (Treml et al., 2005) and MACCHI-BOU 4 (MAB4) (Furutani et al., 2007), was isolated from genetic screens for pid enhancers and for enhancers of yuc1 yuc4 double mutants (Cheng et al., 2006), which are defective in auxin biosynthesis (Cheng et al., 2007). The yuc1 yuc4 npy1 triple mutants produce pin-like inflorescences similar to those in pin1 and pid (Cheng et al., 2007). Inactivation of NPY1 and its close homologs NPY3 and NPY5 also causes the failure of floral initiation from inflorescences (Cheng et al., 2008). The yuc1 yuc2 yuc4 yuc6 quadruple mutants also develop small pin-like structures (Cheng et al., 2006).

Although the aforementioned genes are known to participate in auxin-mediated flower development, the mechanisms by which the genes regulate floral development and the relationships between them are not fully understood. The predominant model regarding the relationship among PIN1, PID, and NPY genes is centered on regulating PIN polarity, localization, and activity by PID and NPY proteins. It was reported that PID directly phosphorylates the hydrophilic loop of PIN proteins, resulting in changes in PIN polarity and activation of PIN-mediated auxin export (Friml et al., 2004; Michniewicz et al., 2007). NPYs were suggested to regulate PIN internalization through the endocytosis pathway (Furutani et al., 2011). Recently, it was reported that NPY1, PID, and PIN1 form a protein complex at the plasma membrane and that the recruitment of NPYs to the plasma-membrane by PIN limits the lateral diffusion of PINs, thus maintaining PIN polarity (Glanc et al., 2021). Phosphorylation of PIN by PID enhances the recruitment of NPY to plasma membrane, which then promotes PIN phosphorylation by recruiting or interacting with AGC kinases, thus forming a self-reinforcing loop to maintain PIN polarity (Glanc et al., 2021). Other protein kinases including D6 PROTEIN KINASE are also involved in phosphorylating PIN proteins and activating PIN-mediated auxin transport (Willige et al., 2013; Barbosa et al., 2014).

Recently, we discovered that pid null mutant phenotypes were suppressed when one copy of the PIN1 gene is inactivated (heterozygous pin1 mutants) (Mudgett et al., 2023). The pid null mutants were also suppressed by the PIN1-GFPHDR fusion in which GFP was precisely inserted into the PIN1 gene via CRISPR/Cas9-mediated homologous recombination (Mudgett et al., 2023). PIN1-GFPHDR is less active than wild type (WT) PIN1, indicating that pid is suppressed by lowering either PIN1 activity or PIN1 gene dosage. This new type of genetic interaction between pid and pin1, called haplo-complementation, renders PID unnecessary for flower development when expressing only 50% of PIN1, whereas the presence of 0% or 100% PIN1 makes PID essential. The observed genetic interaction between pid and pin1 suggests a much more complex relationship between the two genes than the current model of PIN1 phosphorylation by PID directly (Mudgett et al., 2023). Moreover, it was known that pid mutants produce more cotyledons and true leaves than wild type plants whereas pin1 mutants make fewer cotyledons and fewer true leaves compared to wild type plants (Bennett et al., 1995). The opposite cotyledon/true leaf phenotypes of pid and pin1 cannot be accounted for by PID-mediated direct phosphorylation/activation of PIN1.

In this study, we analyzed the dosage effects of NPY1 on pid phenotypes. We show that overexpression of NPY1 (NPY1 OE) completely suppressed pid null mutants, demonstrating that NPY1 and PIN1 have opposite dosage effects on pid mutants. Furthermore, we find that NPY1 OE led to an increase of PIN phosphorylation, suggesting that NPY1 OE -triggered PIN phosphorylation actual inhibits PIN functions. Moreover, we discovered that the C-terminal motif of NPY1, which contains 30 amino acid residues with 50% Ser/Thr, is required for suppression of pid by NPY1 OE. Overexpression of NPY1 that lacked the C-terminal motif (NPY1ΔC) did not suppress pid, but caused auxin resistance and a complete loss of gravitropism. Overexpression of NPY1ΔC also increased phosphorylation of PIN proteins including PIN2, which causes agravitropic roots when mutated, further suggesting that phosphorylation of certain residues/regions of PIN proteins inhibits PIN functions. Our results demonstrated that NPY1 functions downstream of PID during flower development. In addition, our results indicate that NPY1 affects PIN phosphorylation, which inhibits PIN function and which may account for the suppression of pid by NPY1 OE.

Results

Protein null pin1 mutants further confirm haplo-complementation of pid mutants by pin1

We previously showed that pid null mutants were rescued by either PIN1-GFPHDR fusion or the presence of only one copy of functional PIN1 gene, suggesting that a reduction of PIN1 activity or PIN1 gene dosage is sufficient to restore the fertility of pid (Mudgett et al., 2023). Because the pin1 mutants in our previous study had the potential to produce truncated PIN1 proteins, we could not completely rule out the possibility that the predicted truncated PIN1 proteins might have played a role in rescuing pid (Mudgett et al., 2023). To further clarify the mechanisms of rescuing pid by heterozygous pin1 mutants, we used CRIPSR/Cas9 to delete the coding region of PIN1 so that no truncated PIN1 proteins would be produced (Figure 1A). We obtained two pin1 mutants, which lacked almost the entire coding region of PIN1 and which were unlikely to produce any truncated PIN1 proteins (Figure 1A). The two pin1 mutants displayed strong pin-like phenotypes (Figure 1B & 1C). In the heterozygous pin1 mutant backgrounds, the pid null mutants no longer produced pin-like inflorescences and were able to set seeds (Figure 1B & 1C). Rescuing pid by heterozygous pin1 mutants was not specific to a particular pid allele. Both T-DNA insertion (Figure 1B) and CRISPR deletion pid mutants (Figure 1C), which were previously described as pid null mutants (Mudgett et al., 2023), were rescued by the heterozygous pin1 full deletion mutants (Figure 1), demonstrating that pid mutants were rescued by a reduction of PIN1 gene dosage and that suppression of pid mutant phenotypes by heterozygous pin1 mutants in our previous study was not caused by the predicted truncated PIN1 proteins.

Figure 1.

NPY1 and PIN1 have different effects on pid mutants - a reduction of NPY1 gene dosage cannot rescue pid

Inactivation of NPY1 and its close homologs NPY3 and NPY5 leads to the development of pin-like inflorescences (Cheng et al., 2008). Moreover, npy1 pid double mutants fail to develop any cotyledons, a phenotype that was also observed in pin1 pid double mutants (Furutani et al., 2004; Treml et al., 2005; Furutani et al., 2007; Cheng et al., 2008). Therefore, we hypothesized that a reduction of NPY1 gene dosage might also be able to rescue pid mutants. We genotyped F2 plants from a cross of pid-TD1 and npy1-2 to identify plants with either npy1-2 pid-TD1^+/- or npy1-2^+/- pid-TD1 genotype to investigate the effects of gene dosage on genetic interactions between the two mutants. It was very clear that heterozygous npy1-2 did not rescue pid-TD1 (Supplemental Figure 1). Interestingly, heterozygous pid-TD1 further enhanced npy1-2, which is fertile and which does not produce pin-like structures (Supplemental Figure 1). Plants with the npy1-2 pid-TD1^+/- genotype were essentially sterile and produced pin-like structures (Supplemental Figure 1). Our results suggest that gene dosage effects of NPY1 and PIN1 on pid are different.

Dosage effects of NPY1 and PIN1 on pid are opposite - Increased NPY1 gene expression is sufficient to suppress pid null mutants

We placed NPY1 gene under the control of the UBIQUITIN 10 promoter to strongly express NPY1 (named as NPY1 OE). We directly transformed the construct into a pid-c1 segregation population through Agrobacterium-mediated transformation. Among the 84 T1 plants we genotyped, we obtained 15 homozygous pid-c1 mutants. Interestingly, none of the pid-c1 plants produced any pin-like inflorescences and all were fully fertile, demonstrating that increases of NPY1 gene expression were sufficient for the suppression of pid-c1. We genotyped T2 progenies from two pid-c1 heterozygous T1 plants (#68 and # 83) for the presence of pid-c1 and for pid-c1 zygosity. We used mCherry signal, which was included in the NPY1 OE construct, as a proxy to determine the presence and absence of the NPY1 transgene. For each line, we identified T2 plants without the NPY1 transgene and without the pid-c1 mutation (called WT-68 and WT-83, respectively). We also isolated T2 plants that contained the NPY1 overexpression construct, but did not have the pid-c1 mutation (called NPY1 OE #68 in WT, and NPY1 OE #83 in WT). Finally, we identified T2 plants that were pid-c1 homozygous and that had the NPY1 transgene (called NPY1 OE #68 in pid-c1 and NPY1 OE #83 in pid-c1). These genetic materials enabled us to compare the same NPY1 OE transgenic event in different genetic backgrounds.

As shown in Figure 2A and Supplemental Figure 2, both NPY1 OE #68 and #83 lines rescued pid-c1, which lacks almost the entire PID coding region and which is completely sterile under our laboratory conditions (Mudgett et al., 2023). Overexpression of NPY1 in pid-c1 completely eliminated the development of pin-like inflorescences and led to the production of plenty of fertile siliques (Figure 2A & Supplemental Figure 2). NPY1 OE lines had notable developmental phenotypes: NPY1 OE plants had reduced petiole length and had shorter stature compared to WT plants (Supplemental Figure 2). At the young adult stage, NPY1 OE in WT and in pid-c1 appeared very similar, suggesting that overexpression of NPY1 made PID unnecessary for plant development (Supplemental Figure 2). NPY1 OE lines, regardless of the presence or absence of the pid-c1 mutation, developed plenty of flowers with normal appearance (Supplemental Figure 2). Most flowers from NPY1 OE in pid-c1 plants were normal, but occasionally, some flowers had minor defects such as missing a stamen, or developing fused petals with a stamen-like structure (Supplemental Figure 2). In comparison to the suppression of pid null mutants by PIN1-GFPHDR or heterozygous pin1, suppression of pid-c1 by NPY1 OE appeared to be more complete (Mudgett et al., 2023).

Figure 2.

We also transformed the NPY1 OE construct into pid-TD1 and pid-TD2, which are T-DNA insertional null alleles (Mudgett et al., 2023). Both pid T-DNA mutants were rescued by overexpression of NPY1, demonstrating that NPY1 OE could rescue different types of pid mutants (Supplemental Figure 3). Some of the siliques in the rescued pid-TD plants had only one valve, a phenotype that was likely caused by a partial suppression of pid-TD (Supplemental Figure 3). Our results demonstrated that overexpression of NPY1 rescued pid mutants and the suppression of pid by NPY1 OE was not caused by any background mutations, a particular T-DNA insertion site, or dependence on special pid mutations. Our results clearly showed that NPY1 and PIN1 had opposite dosage effects on pid mutants.

NPY1 protein accumulates less in the absence of PID

We conducted proteomic analysis to determine the expression levels of NPY1 in the transgenic lines (Figure 2B). In NPY1 OE line #68 in WT, NPY1 protein level increased 8.0-fold compared to non-transgenic WT-68 (Figure 2B). The NPY1 protein level was slightly higher in line #83 than in line #68 (9.7-fold vs 8.0-fold) (Figure 2B). The same NPY1 transgenic events also led to elevated NPY1 protein levels in pid-c1 compared to non-transgenic WT (Figure 2B). Interestingly, NPY1 protein levels in the NPY1 OE in pid-c1 plants were significantly lower than NPY1 OE in WT. For line #68, NPY1 level was 23% lower in pid-c1 than in NPY1 OE in WT. In line #83, NPY1 protein level was 18.5% lower in pid-c1 than in WT. Our results suggest that PID likely plays a role in regulating NPY1 protein accumulation.

NPY1 is phosphorylated and the phosphorylation of NPY1 does not require PID

We identified 15 phospho-peptide of NPY1 (Supplemental Table 1). Among the phosphorylation sites identified, most of them were located in the peptide ANHSPVASVAASSHSPVEK, which has five serine residues and which is located immediately downstream of the conserved NPH3 domain (Supplemental Figure 4). Previous studies have identified two phosphorylation sites in NPY1: S514 and S553 (Matthes et al., 2024). We also detected that S553 was phosphorylated (Supplemental Table 1 and Supplemental Figure 4). Moreover, we identified a couple of phospho-serine residues outside of the C-terminal domain (Supplemental Figure 4). Interestingly, S181 is located in a highly conserved WSYT motif and is located in the region between the BTB domain and the NPH3 domain (Supplemental Figure 4). The S181 residue is conserved among all NPY proteins, but in NPH3, the serine residue is replaced with an alanine (Supplemental Figure 4).

In the absence of PID, all of the 15 phospho-peptide of NPY1 were still phosphorylated (Supplemental Table 1). Quantitatively, all of the peptides except LHEASVK had lower phosphorylation levels in NPY1 OE in pid plants than in NPY1 OE in WT plants. But given that in pid, NPY1 protein concentrations were about 20% lower than in WT (Figure 2B), the observed differences in phosphorylation levels could be caused by a decrease of NPY1 protein concentration in pid.

Overexpression of NPY1 leads to increases in phosphorylation of PIN proteins

Overexpression of NPY1 caused significant changes in the phospho-proteome (Figure 2C). Among the phospho-peptides that were highly enriched in NPY1 OE lines compared to WT, the majority of them were NPY1 peptides (Figure 2D). Interestingly, the next highly enriched peptides in NPY1 OE lines were from PIN proteins (Figure 2D).

The phospho-peptides AGLQVDNGANEQVGKsDQGGAK from PIN7 and AGLNVFGGAPDNDQGGRsDQGAK from PIN3 were about 6-fold higher in NPY1 OE lines than in WT (Table 1). The two peptides mapped to the same region of the two PIN proteins (Supplemental Figure 5). Phospho-peptides in PIN1 and PIN4 were also up-regulated in NPY1 OE lines compared to WT (Table 1). We did not detect phospho-peptides from PIN2, probably because PIN2 is not expressed in flowers. The two phosphorylated sites in PIN1 (S271 and S282) are conserved within the PIN family and have been previously identified and characterized (Supplemental Figure 5). Two of the PIN7 phosphorylation sites were highly conserved while the majority of the phosphorylation sites in PIN3, PIN4, and PIN7 were located in the less conserved regions of their respective hydrophilic loops (Supplemental Figure 5).

Table 1.

The absence of PID hardly affected PIN phosphorylation triggered by NPY1 OE (Table 1), which was consistent with our previous analysis that phosphorylation levels of PIN proteins in pid PIN1-GFPHDR plants were not affected compared to WT plants (Mudgett et al., 2023). We noticed that previous studies uncovered four highly conserved phosphorylation sites in PIN proteins, which are named as S1, S2, S3, and S4 sites and which correspond to S231, S252, S290, and S271 in PIN1, respectively (Barbosa et al., 2018) (Supplemental Figure 5). Phosphorylation of S1, S2, S3, or S4 leads to activation of PIN auxin transport and changes in PIN polarity (Lanassa Bassukas et al., 2022). Given that NPY1 OE suppressed pid mutants and that decreases in PIN1 gene dosage and PIN1-GFPHDR suppressed pid, we concluded that phosphorylation at the newly identified sites in PIN proteins leads to the inhibition of PIN activities.

The NPY1 C-terminal domain is required for the suppression pid mutants

The NPY1 C-terminal domain is Ser/Thr rich (Supplemental Figure 6), suggesting that the region might be important for regulating NPY1 activity and/or interactions between NPY1 and its partners. The NPY1 phospho-peptides we have identified were largely concentrated in this domain (Supplemental Figure 4). We overexpressed a truncated NPY1 that lacked the 30 C-terminal amino acid residues (called NPY1ΔC), of which 50% are Ser/Thr (Supplemental Figure 6), to determine whether the C-terminal tail of NPY1 was important for NPY1 functions and whether overexpression of NPY1ΔC was sufficient for suppression of pid mutants. We transformed the NPY1ΔC construct into a pid-TD1 segregating population. Among the T1 plants that were identified to be pid-TD1 homozygous, none were suppressed by overexpression of NPY1ΔC. We observed that overexpression of NPY1ΔC actually enhanced the pid-TD1 phenotypes (Supplemental Figure 7).

Overexpression of NPY1ΔC disrupts normal responses to auxin and gravity

Overexpression of NPY1ΔC caused obvious developmental phenotypes (Figure 3). The NPY1ΔC OE plants were smaller with epinastic and darker leaves compared to WT plants (Figure 3A). Overexpression of NPY1ΔC led to smaller plant stature and slower development, but NPY1ΔC OE plants were eventually able to set seeds (Figure 3B). We noticed that NPY1ΔC OE plants often developed small pin-like structures (Figures 3C & 3D). At the seedling stage, NPY1ΔC overexpression roots were completely agravitropic (Figure 3E). Moreover, we observed that NPY1ΔC OE plants were resistant to auxin (Figure 3F). In media containing 100 nM 2,4- Dichlorophenoxyacetic acid (2,4-D), roots of WT plants stopped growing while roots of NPY1ΔC OE plants were able to elongate (Figure 3F). Interestingly, the elongated roots of the NPY1ΔC OE plants were still agravitropic (Figure 3F).

Figure 3.

Overexpression of NPY1ΔC is sufficient to trigger phosphorylation of PIN proteins

NPY1ΔC protein levels were increased in the analyzed NPY1ΔC overexpression lines (#4 and #35) (Figure 4A). The phenotypes of NPY1ΔC OE lines appeared to be correlated with the expression levels of NPY1ΔC protein. Line #35 accumulated more NPY1ΔC proteins than line #4 (Figure 4A). Line #35 had stronger phenotypes than line #4 (Figure 3A & B). Overexpression of NPY1ΔC had a profound impact on the phospho-proteome (Figure 4 B). In the NPY1ΔC OE lines, we detected 18 NPY1 phospho-peptides (Supplemental Table S2, Figure 4C) and the majority of the sites were located in the ANHSPVASVAASSHSPVEK peptide, which was also the main phospho-peptide in the NPY1 OE lines (Supplemental Figure 4, Supplemental Table S1).

Figure 4.

Overexpression of NPY1ΔC also led to an increase of PIN protein phosphorylation (Figure 4C, Supplemental Table S2). We detected phospho-peptides from all of the long PINs (PIN1, PIN2, PIN3, PIN4, and PIN7) (Supplemental Table S2). We also detected a phospho-peptide from PIN6, which has a shorter hydrophilic loop than the long PINs (Supplemental Table S2).

Many of the identified phosphorylation sites of PIN proteins in the NPY1ΔC OE lines were also detected in the NPY1 OE lines (Supplemental Figure 8). However, NPY1ΔC lines and NPY1 OE lines also had their unique phosphorylation sites (Supplemental Figure 8).). A caveat of this comparison is that the phospho-proteomic data were generated using different tissues. We used flowers and inflorescence heads from NPY1 OE lines and whole seedings of NPY1ΔC OE lines for the analysis. Overall, most of the phosphorylation sites were clustered to regions close to trans-membrane domain 6 (TMD6) (Supplemental Figure 8). Another noticeable feature is that some of the phosphorylated residues were not highly conserved among the PIN proteins (Supplemental Figure 8).

NPY1 ΔC-mediated phosphorylation of PINs correlates with phenotypes similar to those of loss-of-function pin2 mutants

Very intriguingly, we detected three PIN2 phospho-peptides, which were at least 2-fold more abundant in the NPY1 ΔC lines than in WT (Figure 4D, Supplemental Table S2). The level of two of the PIN2 phospho-peptides were higher in the stronger line #35 than in line #4 (Figure 4D). PIN2 is known to play a critical role in gravitropism and pin2 mutants have agravitropic roots (Chen et al., 1998; Shin et al., 2005). The NPY1ΔC OE lines displayed complete agravitropic roots (Figure 3E) and developed small pin inflorescences (Figure 3F), suggesting that NPY1 ΔC-mediated phosphorylation of PINs led to their inactivation.

Discussion

In this paper, we demonstrate that PID becomes dispensable for flower initiation when NPY1 is overexpressed. We further show that overexpression of NPY1 increases phosphorylation of PIN proteins at sites mostly different from the previously characterized PIN phosphorylation sites. Phosphorylation of PIN proteins at the identified sites described here leads to the inhibition of PIN functions, rather than activation of PINs by phosphorylation at the previously identified sites (Lanassa Bassukas et al., 2022). Moreover, we show that the suppression of pid phenotypes by overexpressing NPY1 requires the C-terminal motif of NPY1. Overexpression of NPY1ΔC causes a loss of gravitropism, auxin-resistance in roots, and phosphorylation of PIN proteins. Our results demonstrate that the main function of PID is not to directly phosphorylate PIN proteins and that PID affects the accumulation of NPY1 proteins.

Our findings that overexpression of NPY1 completely suppressed pid mutants demonstrate that NYP1 functions downstream of PID in regulating flower initiation in Arabidopsis (Figure 2). Our data presented in this study are consistent with our previous hypothesis that PID/NPY1-mediated flower initiation uses a mechanism that is analogous to that of phototropism (Figure 5) (Cheng et al., 2007, 2008). The hypothesis was based on the observation that three signaling components in phototropism have their homologous counterparts in the flower initiation pathway (Figure 5). In phototropism, blue light is perceived by phototropins (PHOT1 and PHOT2), which have two LOV domains in the N-terminal part and a kinase domain in the C-terminal part (Huala et al., 1997; Christie et al., 1998; Harper et al., 2003). LOV domains function as photo-receptors that regulate the kinase activities of PHOT1 and PHOT2. The kinase domain of phototropins is highly homologous to PID (Cheng et al., 2007). Downstream of the phototropins is NPH3, which is homologous to NPY1 (Motchoulski and Liscum, 1999; Cheng et al., 2007). Phototropins are the starting point of phototropic signal transduction, and are undoubtedly upstream of NPH3. Our genetic results unambiguously demonstrated that NPY1 functions downstream of PID (Figure 2), further strengthening the hypothesis that phototropism and PID/NPY1 use analogous mechanisms (Figure 5). In dark, NPH3 is phosphorylated and forms a complex with PHOT1 (Pedmale and Liscum, 2007). Upon receiving light, NPH3 is dephosphorylated and transiently dissociated from PHOT1 (Christie et al., 2018; Sullivan et al., 2021; Haga and Sakai, 2023). Continuous light exposure leads to re-constitute PHOT1-NPH3 complex and phosphorylation of NPH3 by PHOT1, suggesting that NPH3 phosphorylation by PHOT1 is rather complicated. We found that NPY1 is phosphorylated regardless of the presence of PID (Figure 2 and Supplemental Table 1). Moreover, we identified a predominant NPY1 phospho-peptide, which is located right after the NPH3 domain (Supplemental Table 1 and Supplemental Figure4). It will be interesting to determine the biological consequences of the phosphorylation sites in NPY1 through mutagenesis. At present, we do not have evidence that PID uses NPY1 as a substrate. However, it is clear that PID has an impact on NPY1 accumulation (Figure 2). Without PID, NPY1 accumulates significantly less (Figure 2).

Figure 5.

Unexpectedly, overexpression of NPY1 leads to phosphorylation of PIN proteins and likely inhibition of PIN functions. The most abundant phospho-peptides in NPY1 OE lines are from either NPY1 or PIN proteins (Figure 2). The increased NPY1 protein levels in the NPY1 OE lines can partially account for the observed enrichment of phospho-peptides of NPY1 in the overexpression lines. In contrast, overall PIN protein levels were not changed in NPY1 OE lines compared to WT, indicating that NPY1 OE triggered phosphorylation of PIN proteins. Most of the phosphorylation sites in PIN proteins are not the same as those previously characterized (Lanassa Bassukas et al., 2022). We hypothesize that NPY1-triggered phosphorylation of PINs actually inhibits PIN1 functions, because NPY1 OE suppressed pid (Figure 2) and because pid is suppressed by decreasing PIN1 activity or PIN1 gene dosage (Mudgett et al., 2023). Our hypothesis is further supported by our results from overexpression of NPY1ΔC, which resulted in PIN2 phosphorylation (Figure 4) and agravitropic roots (Figure 3), a phenotype that was observed in pin2 null mutants (Chen et al., 1998). It is not clear whether NPY1 OE-triggered phosphorylation of PINs inhibits auxin transport activity per se or disrupts protein-protein interactions between PINs and their partners or both.

Both NPY1 OE and NPY1ΔC caused great changes in PIN phosphorylation and the phosphorylation led to inhibition of PIN functions (Figures 2 & 4). However, NPY1 OE suppressed pid mutants whereas NPY1ΔC could not suppress pid (Figure 2 & 3), suggesting that inhibition of PIN functions per se is not sufficient for suppression of pid. It is known that PIN and NPY proteins physically interact to form protein complexes (Glanc et al., 2021; Matthes et al., 2024). Our results suggest that the C-terminal domain of NPY1 likely interacts with an unknown protein, which is required for normal PID and PIN functions and which needs to be recruited to the NPY-PIN complexes.

In previous models regarding the mechanisms of PID, PIN1, and NPY1, the first step is direct phosphorylation of PINs by PID (Friml et al., 2004; Michniewicz et al., 2007). Phosphorylated PINs recruit NPY1 to the plasma membrane to maintain PIN polarity by limiting lateral diffusion of the PIN complex (Glanc et al., 2021). However, that overexpression of NPY1 eliminates the need of PID in flower initiation and that overexpression of NPY1 leads to phosphorylation of PIN proteins in the absence of PID are not consistent with the model that PIN1 needs to be directly phosphorylated by PID to be activated. The fact that pin1 and pid displayed opposite cotyledon/leaf phenotypes (Bennett et al., 1995) and that pid is suppressed by heterozygous pin1 mutants (Figure 1) and PIN1-GFPHDR (Mudgett et al., 2023) demonstrated that PID and PIN1 function in opposite directions. Our results firmly established the relative positions of PID and NPY1in the pathway responsible for flower initiation in Arabidopsis (Figure 5). Moreover, we show that PIN1 functions downstream of NPY1, which recruits unknown kinase(s) to phosphorylate PIN1 and to inhibit PIN1 (Figure 5).

Materials and Methods

Plant Growth

Plants used in this study were the Arabidopsis thaliana Columbia-0 ecotype. Sterilized seeds were sown on a 0.7% agar-agar medium containing Murashige and Skoog basal salts at ½ concentration. Sown seeds were subjected to a two-day dark stratification period at 4°C and then moved to a growth chamber with a 16-hour day / 8-hour night cycle. After one week, seedlings were transferred to soil and maintained under the same light conditions.

Plant Transformation

Transgenic plants were created by performing the floral dipping procedure on adult plants with unopened flowers using published protocols (Clough and Bent, 1998). The Agrobacterium tumefaciens strain GV3101 was used to perform all transformations in this study. Seeds were harvested from transformed plants and transformants were selected via the expression of the mCherry fluorescent marker present on the transformed vectors (Gao et al., 2016).

Plasmid Construction

Vectors for gene knockouts were created by cloning two guide RNAs into the pHEE401E backbone as described in (Mudgett et al., 2024). Overexpression vectors were assembled via Gibson Assembly in the pHDE backbone (Gao et al., 2016). Primers used in this study are listed in the Supplemental Table 3.

Proteomics Method

The following tissues were collected and were immediately frozen in liquid nitrogen for proteomic analysis. Inflorescence heads, which include flower buds and meristems from WT (Columbia background)-68, NPY1 OE in WT-68, NPY1 OE in pid-c1-68, WT-83, NPY1 OE in WT-83, NPY1 OE in pid-c1-83 were collected. Whole seedlings of 7-day old Col-WT and NPY1ΔC #35, and #4 were used in this study.

About 0.5 gram of frozen tissue was ground in liquid nitrogen by a mortar/pestle for 15 minutes to fine powders, and then transferred to a 50ml conical tube. Proteins were precipitated and washed by 50 ml -20 °C acetone three times, then by 50 ml -20 °C methanol three times. Samples were centrifuged at 4,000x g, 4 °C for 10 minutes. Supernatant was removed and discarded.

Protein pellets were suspended in extraction buffer (8 M Urea/100mM Tris/5mM TCEP/phosphatase inhibitors, pH 7). Proteins were first digested with Lys-C (Wako Chemicals, 125-05061) at 37 °C for 15 minutes. Protein solution was diluted 8 times to 1M urea with 100mM Tris and digested with trypsin (Roche, 03708969001) for 12 hours. Cysteines were alkylated by adding 10 mM iodoacetamode and incubating at 37 °C for 30 minutes in dark.

Digested peptides were purified on a Waters Sep-Pak C18 cartridges, eluted with 60% acetonitrile. TMT labeling (Supplemental Table S4) was performed in 60% acetonitrile/100mM Hepes, pH 7. TMT labeling efficiency was checked by LC-MS/MS to be greater than 99%. Labeled peptides from different samples were pooled together. 150 µg of pooled peptides were analyzed by 2D-nano-LC-MS/MS for total proteome profiling and 1 mg of total peptides was used for phosphopeptide enrichment.

Phosphopeptide enrichment was performed using CeO2 affinity capture. 20% colloidal CeO2 (Sigma, 289744) was added to the acidified peptide solution (2% TFA/2M lactic acid/60% acetonitrile). After brief vortexing, CeO2 with captured phosphopeptides was spun down at 5,000 g for 1 minute. Supernatant was then removed and the CeO2 pellet was washed with 1 mL of 2% TFA/2M lactic acid/50% acetonitrile. Phosphopeptides were eluted by adding 200 µL eluting buffer (50mM (NH4)2HPO4, 2M NH3.H2O, 10mM EDTA, pH 9.5) and vortexed briefly. CeO2 was precipitated by adding 200 µL acetonitrile. Samples were centrifuged at 16,100 g for 1 minute. Supernatant containing phosphopeptides was removed and dried in a Speedvac. Phosphopeptides were resuspended in 100 mM citric acid and ready for mass spec analysis.

A Thermo Scientific Vanquish Neo UHPLC system (Buffer A: Water with 0.1% formic acid; Buffer B: 80% acetonitrile with 0.1% formic acid) was used to deliver a flow rate of 500 nL/min to a self-packed 3-phase capillary (360 µm OD/200 µm ID) chromatography column. Column phases were a 10 cm long reverse phase (RP1, 5 µm Zorbax SB-C18, Agilent), 6 cm long strong cation exchange (SCX, 3 µm PolySulfoethyl, PolyLC), and 20 cm long reverse phase 2 (RP2, ReproSil-Pur 120 C18-AQ, 1.9 µm), with the electrospray tip of the fused silica tubing pulled to a sharp tip.

Peptide mixtures were loaded onto RP1 using an off-line pressure chamber; and the 3 sections were joined for on-line 2D LC separation. Peptides were eluted from RP1 section to SCX section using a 0 to 80% acetonitrile gradient for 60 minutes, and then were fractionated by the SCX column section by injecting a series of ammonium acetate solutions (5 µL of 10 mM, 20 mM, 30 mM, 40 mM, 50 mM, 60 mM, 70 mM, 80 mM, 90 mM, 100 mM, and 1 M), followed by high-resolution reverse phase separation on the RP2 section of the column using an acetonitrile gradient (0–0.1 min 1% B to 5% B, 0.1–110.1 min 5% B to 35% B, 110.1–120.1 min 35% B to 60% B, 120.1–120.3 min 60% B to 95% B, 120.3–126 min 95% B).

Mass Spectra were acquired on a Thermo Exploris 480 mass spectrometer operated in positive ion mode with a source temperature of 300 °C and spray voltage of 2.1 kV. Automated data-dependent acquisition was employed of the top 20 ions with an isolation window of 0.7 Da and collision energy of 35. Precursor Fit was set to 70% of fit threshold and 0.7 Da fit window. The mass resolution is set at 100,000 for MS and 30,000 for MS/MS scans, respectively. TurboTMT was enabled. Dynamic exclusion of 30 seconds was used to improve the duty cycle.

The raw data was extracted and searched using Spectrum Mill vBI.07 (Broad Institute of MIT and Harvard). MS/MS spectra with a sequence tag length of 1 or less were considered to be poor spectra and were discarded. The remaining high-quality MS/MS spectra were searched against Arabidopsis TAIR11 protein database. A 1:1 concatenated forward-reverse database was constructed to calculate the false discovery rate (FDR). Common contaminants such as trypsin and keratin were included in the protein database. There were 96,562 protein sequences in the final protein database. Search parameters were set to Spectrum Mill’s default settings with the enzyme parameter limited to full tryptic peptides with a maximum mis-cleavage of 1. Cutoff scores were dynamically assigned to each dataset to obtain the false discovery rates (FDR) of 0.1% for peptides, and 1% for proteins. Phosphorylation sites were localized to a particular amino acid within a peptide using the variable modification localization (VML) score in Spectrum Mill software. Proteins that share common peptides were grouped using principles of parsimony to address protein database redundancy. Total TMT reporter intensities were used for relative protein quantitation. Peptides shared among different protein groups were removed before TMT quantitation. Isotope impurities of TMT reagents were corrected using correction factors provided by the manufacturer (Thermo). Median normalization was performed to normalize the protein TMT reporter intensities in which the log ratios between different TMT tags were adjusted globally such that the median log ratio was zero.

Proteomics data deposition

The raw spectra for the proteome data have been deposited in the Mass Spectrometry Interactive Virtual Environment (MassIVE) repository (massive.ucsd.edu/ProteoSAFe/static/massive.jsp, accession ID MSV000098241). FTP download link before publication: ftp://MSV000098241@massive-ftp.ucsd.edu; FTP download link after publication: ftp://massive-ftp.ucsd.edu/v10/MSV000098241/.

Acknowledgements

We thank the UCSD Goeddel Family Technology Sandbox for providing the Thermo Scientific Vanquish Neo UHPLC system and Exploris 480 mass spectrometer used in this research to generate the proteomics and phospho-proteomics data. This work was partially supported by the NIH Cell and Molecular Genetics Training Program (5T32GM007240-43) to M. M., NSF Grant 1546899 to S.P.B., and Tata Chancellor’s Endowed Professorship to Y. Z.

Additional files

Supplementary file 1.