Motif analysis reveals diurnal utilization of phosphorylation sites
We examined our phosphoproteome data for enrichment of phosphorylation motifs that are connected to known protein kinases using Motif-X (motif-x.med.harvard.edu; Chou & Schwartz, 2011; Schwartz & Gygi, 2005). The significantly changing phosphorylated peptides at each transition were analyzed against all quantified phosphopeptides (P value ≤ 0.05). Motifs corresponding to serine (pS) phosphorylation sites were enriched at each transition, while enrichment of phosphorylated threonine (pT) or tyrosine (pY) motifs was absent (Supplemental Table 7). The lack of pY motif enrichment has also been reported in other studies examining phosphoproteome changes under either ED vs EN (Reiland et al., 2009; Uhrig et al., 2019) or free-running circadian cycles (Choudhary et al., 2015; Krahmer et al., 2019). Only one pT motif (pTP) has been previously associated with ED vs EN phosphoproteome changes (Uhrig et al., 2019). The lack of an enriched pTP motif here is likely due to our stringent multi-time point threshold requirement for each phosphorylation site versus the comparison previously performed between ED and EN only (Uhrig et al., 2019). Furthermore, pS accounts for 84-86% of all phosphorylation events in plants, compared to only 10-12% pT and 1-4% pY (Nakagami et al., 2010; Sugiyama et al., 2008), which makes it less likely to find an enrichment of pT and/or pY motifs in the phosphoproteome. Of the phosphorylation sites (site probability score ≥ 0.8 we quantified, 82.8%, 16.5% and 0.7% were pS, pT and pY respectively, which is similar to previously reported distributions of phosphorylation events (Nakagami et al., 2010; Sugiyama et al., 2008).
At the L-D transition, we found 16 motifs of which 10 correspond to phosphorylation sites previously identified as targets of protein kinases CaMKII, PAK1, extracellular signal-regulated kinase (ERK 1/2), proto-oncogene c-RAF (RAF1), and cell division cycle 2 (CDC2) protein kinase A and B (Supplemental Table 7). Six phosphorylation sites did not correspond to known kinase motifs, therefore likely representing currently uncharacterized and possibly plant-specific motifs considering the large expansion of protein kinases in plants relative to humans (Lehti-Shiu & Shiu, 2012). At the D-L transition, four of five identified motifs are known phosphorylation sites for checkpoint kinase 1 (CHK1), PAK2, calmodulin kinase IV (CaMKIV) and casein kinase (CKII) (Supplemental Table 7). CKII phosphorylates the core circadian clock transcription factors LHY and CCA1 (Lu et al., 2011), which also peak at the D-L transition (Kusakina & Dodd, 2012). Consistent with this, we find enrichment of phosphorylation events corresponding to the CKII phosphorylation motif for these proteins at the D-L transition.
Calcium (Ca2+) has been implicated in circadian regulation (Marti Ruiz et al., 2018), suggesting that calcium-dependent calmodulin (CaM) protein kinase orthologs are interesting candidates for mediating circadian clock signaling. Unlike the enrichment of CKII motifs at only the D-L transition, we find enrichment of Ca2+ related kinases CaMKII (D-L and L-D) and CaMKIV (D-L) phosphorylation motifs at both transitions (Supplemental Table 7). Previous analyses identified other Ca2+ kinase motifs enriched at both ED and EN (CDPK-like motifs; Uhrig et al., 2019). SnRK1-related motifs were identified in the phosphoproteome data from Arabidopsis CCA1-Ox plants growing in a free-running cycle (Krahmer et al., 2019). SnRK1 is a central mediator of energy signaling between different organelles and also phosphorylates CDPKs (Wurzinger, Nukarinen, Nagele, Weckwerth, & Teige, 2018). Together, these studies and the results presented here suggest a broader role for Ca2+ in diurnal plant cell regulation during the L-D and D-L transitions.
Compared to humans, plants have more protein kinases (Lehti-Shiu & Shiu, 2012), but most of their targets remain unknown. Our phosphoproteome results, together with previously reported phosphoproteome datasets (Choudhary et al., 2015; Krahmer et al., 2019; Reiland et al., 2009; Uhrig et al., 2019) provide a compilation of specific phosphorylation motifs that are rapidly modified at light-dark diurnal transitions. When combined with transcriptional expression data for lower abundant proteins kinases (Uhrig et al., 2019), the phosphoproteome changes uncovered here in other studies (Choudhary et al., 2015; Krahmer et al., 2019; Reiland et al., 2009; Uhrig et al., 2019) narrow the protein kinase sub-families to those most likely catalyzing diurnal phosphorylation events.