The phytochrome-interacting factor genes PIF1 and PIF4 are functionally diversified due to divergence of promoters and proteins

Abstract

Phytochrome-interacting factors (PIFs) are basic helix–loop–helix transcription factors that regulate light responses downstream of phytochromes. In Arabidopsis (Arabidopsis thaliana), 8 PIFs (PIF1-8) regulate light responses, either redundantly or distinctively. Distinctive roles of PIFs may be attributed to differences in mRNA expression patterns governed by promoters or variations in molecular activities of proteins. However, elements responsible for the functional diversification of PIFs have yet to be determined. Here, we investigated the role of promoters and proteins in the functional diversification of PIF1 and PIF4 by analyzing transgenic lines expressing promoter-swapped PIF1 and PIF4, as well as chimeric PIF1 and PIF4 proteins. For seed germination, PIF1 promoter played a major role, conferring dominance to PIF1 gene with a minor contribution from PIF1 protein. Conversely, for hypocotyl elongation under red light, PIF4 protein was the major element conferring dominance to PIF4 gene with the minor contribution from PIF4 promoter. In contrast, both PIF4 promoter and PIF4 protein were required for the dominant role of PIF4 in promoting hypocotyl elongation at high ambient temperatures. Together, our results support that the functional diversification of PIF1 and PIF4 genes resulted from contributions of both promoters and proteins, with their relative importance varying depending on specific light responses.

The functional diversification of the phytochrome-interacting factor genes PIF1 and PIF4 arises from both the promoter and protein, with their relative contributions varying depending on light responses.

Introduction

Plants are equipped with sets of photoreceptors to monitor the quality, quantity, and direction of light. Among photoreceptors, members of the phytochrome family (phy; phyA to phyE in Arabidopsis [Arabidopsis thaliana]) are responsible for monitoring red and far-red light and regulating a wide range of light responses, from seed germination to flowering (Franklin and Quail 2010). Phys have the ability to sense red and far-red light, which arises from the reversible absorption of these light wavelengths by their chromophore, accompanied by reversible structural changes in the holoprotein (Rockwell et al. 2006). This results in the reversible cycling between the Pfr and Pr forms when exposed to red and far-red light, respectively. The Pfr form is biologically active and can translocate into the nucleus, where it interacts with diverse proteins and forms a membraneless organelle called photobody (Nagatani 2004; Chen and Chory 2011; Klose et al. 2015; Pardi and Nusinow 2021; Chen et al. 2022a; Kim et al. 2023). Phy promotes light responses in part by repressing the function of their interacting proteins, including PHYTOCHROME-INTERACTING FACTORs (PIFs) and CONSTITUTIVE PHOTOMORPHOGENIC 1/SUPPRESSOR OF PHYA-105 (COP1/SPA) complexes (Bae and Choi 2008; Hoecker 2017; Legris et al. 2019). In addition to their role as photoreceptors, phys also function as temperature sensors, utilizing the property of thermal reversion of Pfr to Pr, which increases with rising temperatures (Jung et al. 2016; Legris et al. 2016).

As bHLH transcription factors, PIFs induce skotomorphogenic gene expression upon binding to G-box (CACGTG) or a related E-box (CACATG) in target promoters (Martinez-Garcia et al. 2000; Hornitschek et al. 2009, 2012; Oh et al. 2009; Shin et al. 2009; Zhang et al. 2013; Pfeiffer et al. 2014). The Pfr form of phytochrome promotes photomorphogenic gene expression, in part, by directly interacting and inhibiting the function of PIFs through protein degradation, sequestration, or masking the transcription activation domain (Bauer et al. 2004; Khanna et al. 2004; Park et al. 2004, 2012; Al-Sady et al. 2006; Lorrain et al. 2008; Qiu et al. 2017). In Arabidopsis, 8 PIFs (PIF1 to PIF8) have been characterized based on the presence of a motif called active phytochrome B-binding (APB) and their abilities to bind to the Pfr of phyB (Khanna et al. 2004; Leivar and Quail 2011). Among them, PIF1 and PIF3 possess an additional motif called active phytochrome A-binding (APA), which confers the binding ability to the Pfr of phyA. PIFs are also found in other plant species. For instance, tomato (Solanum lycopersicum), a eudicot plant diverged from Arabidopsis 112 million yr ago, possesses 8 PILs (PIF-LIKEs) with an APB motif, of which 3 PILs possess an additional APA motif (Rosado et al. 2016). On the other hand, rice (Oryza sativa), a monocot plant diverged from Arabidopsis 200 million yr ago, possesses 6 PIF/PILs with an APB motif, of which OsPIL15 possesses an additional APA motif (Nakamura et al. 2007; Ji et al. 2019). Woody plants like apple (Malus × domestica) and peach (Prunus persica) also have PIFs/PILs that promote hypocotyl elongation when expressed in Arabidopsis (Zheng et al. 2020; Chen et al. 2022b; Liu et al. 2022a). Similar to Arabidopsis PIFs, PIF/PILs in other plants bind to G-box or E-box in target promoters to induce gene expression and are inhibited by the Pfr form (Cordeiro et al. 2016; Llorente et al. 2016; Sun et al. 2020; Hoang et al. 2021; Pan et al. 2021). PIFs are also found in bryophytes, such as Physcomitrium and Marchantia, which diverged from Arabidopsis 400 million yr ago. Physcomitrium has 4 PIFs with APA or APB-like motifs (Possart et al. 2017), while Marchantia has 1 PIF with an APA motif (Inoue et al. 2016). Similar to angiosperm PIFs, bryophyte PIFs regulate the expression of target genes, and their activities are inhibited by red light through unknown mechanisms (Physcomitrium) or by the Pfr form of phy through protein degradation (Inoue et al. 2016; Xu and Hiltbrunner 2017). This supports that phy-PIF signaling modules have emerged in these early land plants.

PIFs have undergone functional diversification, leading to both shared and distinct roles in various light responses (Jeong and Choi 2013). The majority of PIFs redundantly suppress seedling photomorphogenesis, as evidenced by the etiolation observed in various PIF-overexpressing lines even in the presence of light (Park et al. 2004; Khanna et al. 2007) and also by the progressively severe deetiolation observed in pif high-order mutants, such as pifQ (pif1/3/4/5) in the dark (Leivar et al. 2008; Shin et al. 2009) and pifP (pif1/3/4/5/7) in shade conditions (Zhang et al. 2020). On the contrary, PIF1 is the major PIF suppressing seed germination in the dark (Oh et al. 2004), while PIF4 plays a key role in promoting hypocotyl elongation at high ambient temperatures (Koini et al. 2009). Combinations of 2 or 3 PIFs have also been shown to regulate other light responses: PIF1, PIF3, and PIF5 mainly inhibit chlorophyll biosynthesis in dark-grown seedlings (Shin et al. 2009), PIF4 and PIF5 primarily promote leaf senescence in the dark (Sakuraba et al. 2014), while PIF4, PIF5, and PIF7 mainly promote shade avoidance responses (Lorrain et al. 2008; Li et al. 2012) and rhythmic growth during the day (Niwa et al. 2009; Kunihiro et al. 2011). The pifQ mutant also develops reduced hypocotyl adventitious roots in the dark compared to any of the single mutants (Li et al. 2022). Similarly, in other species, PIFs have also undergone functional diversification, resulting in shared and distinct roles. Five rice PIFs (OsPIL11-15), 3 maize PIFs (ZmPIF3-5), 6 Brachypodium PIFs (BdPIF4/5A-C, BdPIF3A-B, and BdPIF8), and 4 Physcomitrium PIFs (PpPIF1-4) promote hypocotyl elongation when expressed in Arabidopsis (Nakamura et al. 2007; Possart et al. 2017; Shi et al. 2017; Xu and Hiltbrunner 2017; Wu et al. 2019; Jiang et al. 2022). Among OsPILs, OsPIL13/OsPIL1 preferentially inhibits leaf senescence (Sakuraba et al. 2017a) and promotes chlorophyll biosynthesis (Sakuraba et al. 2017b), whereas OsPIL15 primarily decreases grain yield (Ji et al. 2019). It is not known, however, which features of PIF genes have been diversified to cause these shared and distinct roles.

The functional diversification of related genes can be achieved by diversifying promoter activities governing mRNA expression patterns or molecular activities of proteins, such as specific enzyme activity. The roles of promoters and proteins for the functional diversification of related genes have been reported for various genes, including 4 Arabidopsis transcription factor genes: WEREWOLF (WER), GLABRA1 (GL1), MYB23, and MYB2. These MYB transcription factor genes regulate various processes, such as root hair development (WER; Lee and Schiefelbein 1999), trichome initiation on leaf surface (GL1; Oppenheimer et al. 1991), trichome initiation at the leaf edge (GL1, MYB23; Kirik et al. 2005), and trichome branching (MYB23; Kirik et al. 2005), as well as other processes like senescence and axillary meristem formation (MYB2; Guo and Gan 2011; Jia et al. 2020). The contributions of promoters and proteins for the functional diversification differ among gene pairs. First, the functional diversification of WER and GL1 genes is caused by the divergence of their promoters rather than their proteins as WER expressed by GL1 promoter could fully restore trichome initiation in the gl1 mutant but not root hair development in the wer mutant, whereas GL1 expressed by WER promoter could fully restore root hair development in the wer mutant but not trichome initiation in the gl1 mutant (Lee and Schiefelbein 2001). Such promoter-derived functional diversifications of related genes have also been reported for various genes, including 2 DELLA genes, REPRESSOR OF GA (RGA) and RGA-Like 2 (RGL2), in regulating stem elongation and seed germination in Arabidopsis (Gallego-Bartolome et al. 2010), 2 ribosomal genes, Ribosomal protein S27 (Rps27) and Rps27-like (Rps27I), needed in different developmental stages in mouse (Xu et al. 2023), and outer membrane porin F (ompF) and ompC in regulating osmolarity in Escherichia coli (Matsuyama et al. 1984). Second, the functional diversification of MYB2 and WER or GL1 is caused by the divergence of proteins rather than their promoters as MYB2 expressed by WER promoter or GL1 promoter cannot rescue wer mutant or gl1 mutant, respectively (Lee and Schiefelbein 2001). Such protein-derived functional diversifications of related genes have also been reported for other Arabidopsis genes. For example, INDOLE-3-ACETIC ACID INDUCIBLE 19/MASSUGU 2 (IAA19/MSG2) protein regulates shoot development differently from IAA14/SOLITARY ROOT (SLR) or IAA7/AUXIN RESISTANT 2 (AXR2) proteins, even when expressed by the same IAA7 promoter (Muto et al. 2007). Similarly, the differences among TRANSPORT INHIBITOR RESPONSE 1/AUXIN SIGNALING F-BOX (TIR1/AFB) family genes are attributable to their proteins (Parry et al. 2009). Third, the functional diversification of GL1 and MYB23 genes in regulating trichome branching requires both promoters and proteins, as neither GL1 expressed by the MYB23 promoter nor MYB23 expressed by GL1 promoter can restore trichome branching in the myb23 mutant (Kirik et al. 2005). Instead, MYB23 can restore trichome branching only when expressed by the MYB23 promoter. Such requirements of both promoters and proteins to the functional diversification of 2 related genes have also been reported for various Arabidopsis genes, including APETALA1 (AP1) and FRUITFUL (FUL) genes in regulating development of floral organs, silique, and cauline leaf (McCarthy et al. 2015), 2 RESPIRATORY BURST OXIDASE HOMOLOGUE genes (RboHD and RbohF) in regulating pathogen responses (Morales et al. 2016), and WRKY DNA-BINDING PROTEIN 12 (WRKY) 12 and WRKY13 in regulating flowering time (Li et al. 2016).

PIFs exhibit distinct mRNA expression patterns and bind to different promoters in vivo, suggesting that the functional diversification of PIFs could also be attributed to the divergence of promoters and proteins. PIF1 mRNA is expressed at higher levels than other PIF mRNAs in imbibed seeds, aligning with the preferential role of PIF1 in seed germination (Lee et al. 2014). During the early seedling stage, PIF1 and PIF3 mRNAs are expressed more prominently than PIF4 and PIF5 mRNAs but decrease in expression during the late stage (Jeong and Choi 2013). These expression patterns also correspond to the more dominant roles of PIF1 and PIF3 in early seedling development, including the repression of cotyledon opening and chlorophyll biosynthesis during seedling emergence from the soil (Shin et al. 2009), and more dominant roles of PIF4 and PIF5 in late development, including flowering (Galvao et al. 2019) and shade avoidance responses (Lorrain et al. 2008). The expression of PIF4 mRNA is induced by high ambient temperature, which also matches with its dominant role in high ambient temperature response (Koini et al. 2009). Moreover, the expression of PIF1, PIF3, PIF4, and PIF5 is observed in both pavement cells and guard cells of leaf epidermis, consistent with their involvement in enhancing drought and salt tolerance (Liu et al. 2022b). Genome-wide DNA-binding analyses suggest that the functional diversification of PIFs could also result from their binding to different promoters in vivo. PIF1 binds to 750 target genes in seeds (Oh et al. 2009) and 1,911 genes in seedlings (Pfeiffer et al. 2014), PIF3 binds to 828 genes (Zhang et al. 2013), PIF4 binds to 1,279 genes in the dark (Pfeiffer et al. 2014) and 4,363 genes in white light (Oh et al. 2012), PIF5 binds to 1,218 target genes (Hornitschek et al. 2012), and PIF7 binds to 975 target genes (Chung et al. 2020). Among all these target genes, only 113 target genes are bound by all 5 PIFs, implying that PIFs bind to both shared and distinct target genes in vivo. These varying mRNA expression patterns and the binding to distinct target genes suggest that PIFs may have functionally diversified due to divergences in their promoters and proteins. However, experimental evidence supporting the relative importance of promoters versus proteins for the functional diversification of PIF genes is lacking.

To experimentally investigate the role of promoters and proteins for the functional diversification of PIF genes, we selected PIF1 and PIF4 genes, 2 functionally divergent PIF genes. We systematically analyzed the contribution of promoter and protein to the functional diversification of these 2 PIF genes by swapping PIF1 and PIF4 promoters. We further examined which domain of PIF protein contributes to the functional diversification of 2 PIF proteins through the analysis of chimeric PIF1 and PIF4 proteins.

Results

PIF1 gene primarily inhibits seed germination, while PIF4 gene mainly promotes hypocotyl elongation under both red light and high ambient temperature conditions (Huq and Quail 2002; Oh et al. 2004; Leivar et al. 2008; Koini et al. 2009; Shin et al. 2009). Such a functional diversification of PIF1 and PIF4 genes could be due to their divergent mRNA expression patterns governed by their respective promoters or to their different molecular functions governed by their proteins. To investigate which component contributes more to the functional diversification of PIF1 and PIF4 genes, we performed the promoter-swapping experiment. We generated transgenic lines expressing PIF1 or PIF4 under the control of PIF1 promoter fragment (proPIF1) or PIF4 promoter fragment (proPIF4) in the pif1 pif4 double mutant background (proPIF1:PIF1, proPIF1:PIF4, proPIF4:PIF1, and proPIF4:PIF4; Fig. 1A). Two independent transgenic lines for each construct, with similar levels of transgene mRNA, were used for subsequent analysis.

Figure 1.

PIF1 promoter primarily confers a dominant role to PIF1 gene in regulating seed germination. A) Schematic illustration of promoter-swapped PIF1 and PIF4 expression cassettes. Arrows indicate promoters (proPIF1, proPIF4), squares indicate PIF1 and PIF4 coding sequences (PIF1, PIF4), hexagons indicate flag tags (F), and boxes indicate NOS terminators (NOS-T). Each construct was transformed into the pif1 pif4 double mutant, and independent homozygous lines were established for subsequent analyses. B) Higher expression of endogenous PIF1 mRNA than PIF4 mRNA in imbibed wild-type seeds. Sterilized wild-type (Col-0) seeds were irradiated with a far-red light pulse (3 μmol m⁻² s⁻¹, 5 min) and then incubated at 22 °C in darkness for 12 h before sampling for RNA extraction. mRNA expression levels were determined by RT-qPCR and normalized by the level of PP2A. Error bars indicate SEM (n = 4, biological replicates). C) Higher expression of proPIF1-driven PIF transgene mRNAs than proPIF4-driven PIF transgene mRNAs in imbibed transgenic seeds. Sterilized transgenic seeds were irradiated with a far-red light pulse (3 μmol m⁻² s⁻¹, 5 min) and then incubated at 22 °C in darkness for 12 h before sampling for RNA extraction. Transgene mRNA levels were determined by RT-qPCR using a primer pair annealing to the transcribed region of the NOS terminator and normalized by the level of PP2A. #n indicates transgenic line numbers. Error bars indicate SEM (n = 3, biological replicates). D) Light scheme for phyB-dependent seed germination assay. Sterilized seeds were irradiated with a 5-min far-red light pulse (FRp, 3 μmol m⁻² s⁻¹) and then either incubated at 22 °C in darkness (phyB_off) or incubated at 22 °C in continuous white light (cWL, 50 μmol m⁻² s⁻¹) for 5 d. E) Representative images of seed germination assay conducted in the phyB_off condition. F) Inhibition of seed germination by proPIF1-driven PIF1 and PIF4. Germination frequencies of the transgenic lines were determined in both the phyB_off condition and the cWL condition, as described in D). Seeds with protruded radicles were counted as germinated seeds. Error bars indicate SEM (n = 3, biological replicates). Alphabet letters above bars indicate statistical differences, as determined by ANOVA with Tukey's HSD post hoc tests for multiple comparisons (P < 0.05). Individual data points are indicated with dots.

PIF1 promoter mainly confers a dominant role to PIF1 gene in regulating light-dependent seed germination

Seed germination has been shown to be inhibited by the PIF1 gene but not by PIF4 gene. However, when overexpressed by the strong 35S promoter, both PIF1 and PIF4 could inhibit seed germination (Oh et al. 2004; Lee et al. 2014), suggesting that the expression patterns of PIF mRNAs might play an important role in the functional diversification of the 2 PIFs. To weigh if mRNA expression patterns could be associated with their functional differences in regulating seed germination, we examined the expression of endogenous PIF1 and PIF4 mRNAs in imbibed seeds. Endogenous PIF1 mRNA was highly expressed in imbibed wild-type seeds whereas endogenous PIF4 mRNA was virtually not (Fig. 1B), implying that PIF1 promoter is more active than PIF4 promoter in imbibed seeds. Consistent with this, the proPIF1 strongly drove the expression of both transgenic PIF1 and PIF4 mRNAs in imbibed transgenic seeds, while the proPIF4 virtually did not, as determined by either a primer pair amplifying a transcribed region of the common NOS terminator (Fig. 1C) or PIF-specific primer pairs (Supplementary Fig. S1). The levels of transgenic PIF1 and PIF4 mRNA driven by the proPIF1 promoter were 1.8- to 3.4-fold of endogenous PIF1 mRNA level in imbibed seeds (Supplementary Fig. S1). These results suggest that the promoter may confer a dominant role to PIF1 gene in regulating light-dependent seed germination.

We experimentally determined whether the promoter indeed confers a dominant role to PIF1 gene by analyzing promoter-swapped PIF transgenic plants. For the assay, seeds were irradiated with a far-red light pulse for 5 min and incubated either in darkness (phyB_off condition) or in white light for 5 d (Fig. 1D). Consistent with the previous report, pif1 pif4 double mutant seeds germinated almost 100% even in the phyB_off condition, while wild-type seeds did not (Fig. 1, E and F; Shin et al. 2009). Among promoter-swapped PIF genes, PIF1 expressed under the proPIF1 strongly inhibited seed germination in the phyB_off condition, whereas PIF1 expressed under the proPIF4 did not (Fig. 1, E and F). The lack of germination in proPIF1:PIF1 seeds was not due to seed viability, as they germinated almost 100% when exposed to white light. These results indicate that PIF1 promoter, but not PIF4 promoter, could support the dominant role of PIF1 gene in regulating light-dependent seed germination. However, the promoter is not the sole factor conferring dominance to PIF1 gene in regulating seed germination. The germination assay further revealed that PIF4 expressed under the proPIF1 could partially inhibit seed germination in the phyB_off condition. The weaker inhibition of seed germination by proPIF1-driven PIF4 compared to proPIF1-driven PIF1 was not due to lower PIF4 mRNA levels than PIF1 mRNA levels in these transgenic lines (Fig. 1C). Together, these results support that the dominance of PIF1 gene in regulating seed germination is mainly attributed to PIF1 promoter and partially to PIF1 protein.

PIF4 protein mainly confers a dominant role to PIF4 gene in regulating hypocotyl elongation in red light

Hypocotyl elongation is strongly promoted by PIF4 gene but only very weakly, if at all, by PIF1 gene in red light (Leivar et al. 2008; Shin et al. 2009). To see whether mRNA expression patterns could be associated with their functional differences in regulating hypocotyl elongation in red light, we determined endogenous PIF1 and PIF4 mRNA levels in red light-grown wild-type seedlings. PIF1 and PIF4 mRNA levels were similar in dark-grown seedlings (Fig. 2A). However, only PIF4 mRNA expression increased strongly in red light, resulting in a higher PIF4 mRNA level compared to the PIF1 mRNA level in red light-grown seedlings. Consistent with these endogenous PIF mRNA expression patterns, the proPIF4 promoter drove the expression of both transgenic PIF mRNAs more strongly than the proPIF1 promoter did in red light (Fig. 2B). However, both promoters similarly drove the expression of transgenic PIF mRNAs in darkness, although the levels of PIF4 mRNA driven by proPIF1 were slightly lower than those of PIF1 mRNA in darkness. These results suggest that the promoter may confer a dominant role to PIF4 gene in regulating hypocotyl elongation in red light.

Figure 2.

PIF4 protein mainly confers the dominant role to PIF4 gene in controlling hypocotyl elongation in red light. A) Higher expression of endogenous PIF4 mRNA than PIF1 mRNA in red light-grown wild-type seedlings. Wild-type (Col-0) seedlings were grown at 22 °C either in darkness (Dark) or under continuous red light (Red, 20 μmol m⁻² s⁻¹) for 4 d before sampling for RNA extraction. mRNA expression levels were determined by RT-qPCR and normalized by the level of PP2A. Error bars indicate SEM (n = 3, biological replicates). B) Higher expression of proPIF4-driven PIF transgene mRNAs than proPIF1-driven PIF transgene mRNAs under red light. Seedlings were grown at 22 °C either in darkness (Dark) or under continuous red light (Red, 20 μmol m⁻² s⁻¹) for 4 d before sampling for RNA extraction. Transgene mRNA levels were determined by RT-qPCR using a primer pair annealing to the transcribed region of the NOS terminator and normalized by the level of PP2A. #n indicates transgenic line numbers. Error bars indicate SEM (n = 3, biological replicates). C) Representative images of dark- (left) and red- (right) grown seedlings (bar = 5 mm). D) Promotion of hypocotyl elongation by PIF4 driven by either proPIF1 or proPIF4 in red light. Seedlings were grown at 22 °C either in darkness (Dark) or under continuous red light (Red, 20 μmol m⁻² s⁻¹) for 4 d. Error bars indicate Sd (n = 30 seedlings). Hypocotyl lengths of individual seedlings are indicated with dots. Alphabet letters above each bar indicate statistical differences as determined by ANOVA with Tukey's HSD post hoc tests for multiple comparisons (P < 0.05). Individual data points are indicated with dots.

We determined hypocotyl elongation in the promoter-swapped PIF transgenic lines in red light. Consistent with the previous report, pif1 pif4 double mutant seedlings had shorter hypocotyl lengths than wild-type seedlings in red light (Fig. 2, C and D). In contrast to the expectation based on mRNA expression patterns, however, among the promoter-swapped PIF genes, PIF4 expressed by either the proPIF1 or the proPIF4 promoted hypocotyl elongation in red light, whereas PIF1 expressed by either the proPIF1 or the proPIF4 did not. The inability of PIF1 to promote the hypocotyl elongation was not due to a general defect in hypocotyl elongation, as all seedlings exhibited similar elongated hypocotyl lengths in the dark (Fig. 2, C and D). Furthermore, this difference in function was not attributed to lower PIF1 mRNA levels compared to PIF4 mRNA levels in these transgenic lines (Fig. 2B). Together, these results support that PIF4 protein rather than PIF4 promoter is mainly responsible for the dominant role of PIF4 gene in regulating hypocotyl elongation in red light. It should be noted, however, that proPIF4:PIF4 seedlings exhibited longer hypocotyl lengths than proPIF1:PIF4 seedlings, corresponding to the higher PIF4 mRNA levels in proPIF4:PIF4 compared to proPIF1:PIF4. Together, the results support that the dominance of PIF4 gene in regulating hypocotyl elongation in red light is mainly due to PIF4 protein, and this effect is further strengthened by the high activity of PIF4 promoter in red light.

Both PIF4 promoter and PIF4 protein are required for PIF4 gene to promote hypocotyl elongation at high ambient temperature

Hypocotyl elongation is promoted mainly by PIF4 gene but not by PIF1 gene at high ambient temperature (Koini et al. 2009). To test whether mRNA expression patterns could be associated with their distinct roles in regulating hypocotyl elongation at high ambient temperatures, we assessed endogenous PIF1 and PIF4 mRNA levels in wild-type seedlings at high ambient temperature. PIF4 mRNA levels were already higher than PIF1 mRNA levels at 22 °C in white light (Fig. 3A). The expression of PIF4 mRNA but not PIF1 mRNA increased further when seedlings were transferred to 28 °C, resulting in a higher PIF4 mRNA level than PIF1 mRNA level at high ambient temperature. Consistent with the endogenous PIF mRNA expression patterns, the proPIF4 drove the expression of both transgenic PIF mRNAs more strongly than the proPIF1 did at 28 °C (Fig. 3B). These results suggest that the promoter confers a dominant role to PIF4 gene in regulating hypocotyl elongation at high ambient temperature.

Figure 3.

Both PIF4 promoter and PIF4 protein are required for PIF4 gene to promote hypocotyl elongation at high ambient temperature. A) Increased expression of endogenous PIF4 mRNA at high ambient temperature. Four-day-old continuous white light-grown wild-type (Col-0) seedlings (50 μmol m⁻² s⁻¹, 22 °C) were either kept at 22 °C or transferred to 28 °C for 6 h before sampling for RNA extraction. mRNA expression levels were determined by RT-qPCR and normalized by the level of PP2A. Error bars indicate SEM (n = 3, biological replicates). B) Increased expression of proPIF4-driven PIF transgene mRNAs at high ambient temperature. Four-day-old continuous white light-grown seedlings (50 μmol m⁻² s⁻¹, 22 °C) were either kept at 22 °C or transferred to 28 °C for 6 h before sampling for RNA extraction. Transgene mRNA levels were determined by RT-qPCR using a primer pair annealing to the transcribed region of the NOS terminator and normalized by the level of PP2A. #n indicates transgenic line numbers. Error bars indicate SEM (n = 3, biological replicates). C) Representative images of seedlings grown at either 22 or 28 °C for 6 d. D) Promotion of hypocotyl elongation by proPIF4-driven PIF4 at high ambient temperature. Seedlings were initially grown at 22 °C under white light (50 μmol m⁻² s⁻¹) for a day and then transferred to either 22 or 28 °C for 6 d. Error bars indicate Sd (n = 30 seedlings). Alphabet letters above each bar indicate statistical differences as determined by ANOVA with Tukey's HSD post hoc tests for multiple comparisons (P < 0.05). Individual data points are indicated with dots.

We experimentally determined hypocotyl elongation in the promoter-swapped PIF transgenic lines at high ambient temperature. As expected from the known role of the PIF4 gene in promoting high ambient temperature-induced hypocotyl elongation, the pif1 pif4 double mutant seedlings had shorter hypocotyl lengths than wild type at 28 °C (Fig. 3, C and D). However, contrary to our expectations based on mRNA expression patterns, among the promoter-swapped PIF genes, PIF4 expressed by the proPIF4 promoted hypocotyl elongation at 28 °C, whereas others, including PIF1 expressed by the proPIF4, did not. The inability of PIF1 to promote hypocotyl elongation at 28 °C was not attributed to lower PIF1 mRNA levels compared to PIF4 mRNA levels at that temperature, as the proPIF4 drove both PIF1 and PIF4 mRNA expressions similarly high at 28 °C (Fig. 3B). Interestingly, PIF4 expressed by the proPIF1 also did not promote hypocotyl elongation at 28 °C (Fig. 3, C and D), which is in contrast to its effect on hypocotyl elongation in red light (Fig. 2D). Together, these results support that the dominance of PIF4 gene in regulating hypocotyl elongation at high ambient temperature is caused by the combined action of PIF4 promoter and PIF4 protein.

The N-terminal domains confer functional characteristics to PIF1 and PIF4 proteins

The analysis of promoter-swapped PIF genes indicated that the functional diversification of PIF1 and PIF4 genes has arisen not only from their promoters but also partly from their proteins. Thus, we further investigated which domains of PIF proteins confer their functional characteristics. Both PIF1 and PIF4 are DNA-binding transcription factors having basic helix–loop–helix motifs at their C-terminal domains. Thus, we divided PIF1 and PIF4 proteins into N-terminal domains (N) and C-terminal domains (C) and swapped them to generate 2 chimeric PIF proteins (1N4C: chimeric protein having the N-terminal domain of PIF1 and the C-terminal domain of PIF4; 4N1C: chimeric protein having the N-terminal domain of PIF4 and the C-terminal domain of PIF1; Fig. 4A, Supplementary Fig. S2). We linked the chimeric genes to the proPIF1 or the proPIF4 and generated transgenic lines expressing chimeric PIFs in the pif1 pif4 double mutant background (proPIF1:1N4C, proPIF1:4N1C, proPIF4:1N4C, and proPIF4:4N1C). Two independent lines for each chimeric gene with similar transgene mRNA levels, except proPIF4:1N4C, were established (Fig. 4, B and C) and subsequently used for the analysis. The proPIF1 drove the expression of both chimeric PIF mRNAs more strongly than the proPIF4 in imbibed transgenic seeds, while the proPIF4 drove the expression of them either similarly or slightly higher than the proPIF1 did in red light-grown seedlings.

Figure 4.

N-Terminal domains of PIF proteins confer functional characteristics of PIF1 and PIF4 genes. A) Schematic illustration of chimeric PIF1 and PIF4 expression cassettes. Arrows indicate promoters (proPIF1, proPIF4), squares indicate chimeric PIFs (PIFXN: N-terminal domain of PIFX; PIFXC: C-terminal domain of PIFX), hexagons indicate flag tags (F), and boxes indicate NOS terminators (NOS-T). Each construct was transformed into the pif1 pif4 double mutant, and independent homozygous lines were established for subsequent analyses. B) Similar transcript levels between proPIF1-driven chimeric PIFs in seeds. Sterilized transgenic seeds were irradiated with a far-red light pulse (3 μmol m⁻² s⁻¹, 5 min) and incubated at 22 °C in darkness for 12 h before sampling for the RNA extraction. Transgene mRNA levels were determined by RT-qPCR using a primer pair annealing to the transcribed region of the NOS terminator and normalized by the level of PP2A. #n indicates transgenic line numbers. Error bars indicate SEM (n = 3, biological replicates). C) Similar transcript levels between proPIF4-driven chimeric PIFs in seedlings under red light. Seedlings were grown at 22 °C under continuous red light (Red, 20 μmol m⁻² s⁻¹) for 4 d before sampling for the RNA extraction. Transgene mRNA levels were determined by RT-qPCR using a primer pair annealing to the transcribed region of the NOS terminator and normalized by the level of PP2A. #n indicates transgenic line numbers. Error bars indicate SEM (n = 3, biological replicates). D) Inhibition of seed germination by proPIF1-driven 1N4C. Germination frequencies of the transgenic lines were determined both in the phyB_off condition and the cWL condition as described in Fig. 1D. Seeds with protruded radicles were counted as germinated seeds. Error bars indicate SEM (n = 3, biological replicates). E) Promotion of hypocotyl elongation by 4N1C driven by both proPIF1 and proPIF4 in red light. Seedlings were grown at 22 °C either in darkness (Dark) or under continuous red light (Red, 20 μmol m⁻² s⁻¹) for 4 d. Error bars indicate Sd (n = 30 seedlings). F) Promotion of hypocotyl elongation by proPIF4-driven 4N1C at high ambient temperature. Seedlings were grown at 22 °C under white light (50 μmol m⁻² s⁻¹) for a day and then transferred to either 22 or 28 °C for 6 d. Error bars indicate Sd (n = 39 seedlings). Alphabet letters above bars indicate statistical differences as determined by ANOVA with Tukey's HSD post hoc tests for multiple comparisons (P < 0.05). Individual data points are indicated with dots.

We investigated which chimeric PIFs behave like PIF1 to inhibit light-dependent seed germination. Of 2 proPIF1-driven chimeric PIFs, the 1N4C almost completely inhibited seed germination in the phyB_off condition, whereas the 4N1C did not (Fig. 4D). Different germination frequencies of chimeric PIF lines were not due to dead seeds, as all seeds germinated well in white light. The results indicated that the N-terminal domain of PIF1 confers the functional characteristics to PIF1 protein in regulating light-dependent seed germination. We next determined which chimeric PIFs behave like PIF4 to promote the hypocotyl elongation in red light. Of 2 proPIF4-driven chimeric PIFs, the 4N1C promoted hypocotyl elongation in red light, whereas the 1N4C did not (Fig. 4E). The proPIF1-driven 4N1C also promoted hypocotyl elongation, whereas the proPIF1-driven 1N4C did not. The inability of the 1N4C to promote hypocotyl elongation was not due to a general defect in hypocotyl elongation, as all seedlings had similarly elongated hypocotyls in the dark. It was not also due to the lower expression level of 1N4C as both the mRNA and protein levels of proPIF1-driven 1N4C were similar to those of proPIF1-driven 4N1C (Fig. 4C, Supplementary Fig. S3). Similarly, while mRNA levels of proPIF4-driven 4N1C were either similar or higher than those of proPIF4-driven 1N4C (Fig. 4C), the protein levels of them were also similar (Supplementary Fig. S3). We also examined which chimeric PIFs behave like PIF4 to promote the hypocotyl elongation at high ambient temperature. Only proPIF4-driven 4N1C promoted hypocotyl elongation at 28 °C, whereas proPIF4-driven 1N4C did not (Fig. 4F). The results indicated that the N-terminal domain of PIF4 protein confers functional characteristics to PIF4 protein in regulating hypocotyl elongation both in red light and at high ambient temperature. Collectively, our results support that the N-terminal domain, rather than the DNA-binding C-terminal domain, confers functional characteristics to PIF1 and PIF4 proteins in regulating seed germination and hypocotyl elongation, respectively.

The N-terminal domain of PIF1 protein imparts higher protein stability compared to that of PIF4 protein in imbibed seeds

Our analysis showed that PIF1 protein inhibited light-dependent seed germination more strongly than PIF4 protein (Fig. 1F), and this strong inhibitory activity of PIF1 protein could be traced to its N-terminal domain (Fig. 4D). Our analysis also showed that PIF4 protein promoted hypocotyl elongation more strongly than PIF1 protein, both in red light and at high ambient temperature (Figs. 2D and 3D), and the strong elongation activity of PIF4 protein could be also traced to its N-terminal domain (Fig. 4, E and F). Given that the overall activity of a protein is generally proportional to its protein level, we determined if functional characteristics of PIF1 and PIF4 proteins partly stem from their differences in protein stability.

The immunoblot analysis showed that proPIF1-driven PIF1 protein accumulated to a higher level than proPIF1-driven PIF4 protein in imbibed seeds (Fig. 5A), despite PIF1 mRNA levels not being higher than PIF4 mRNA levels in these transgenic imbibed seeds (Fig. 1C, Supplementary Fig. S1). This higher accumulation of PIF1 protein was attributable to its N-terminal domain, as proPIF1-driven 1N4C also accumulated more than proPIF1-driven 4N1C (Fig. 5B). The higher accumulation of 1N4C protein than 4N1C protein was not also due to higher 1N4C mRNA levels than 4N1C mRNA levels in these transgenic imbibed seeds (Fig. 4B). These results support that PIF1 protein inhibits seed germination more strongly than PIF4 protein, partly because the N-terminal domain of PIF1 protein allows PIF1 protein to accumulate to a higher level than PIF4 protein in imbibed seeds.

Figure 5.

PIF1 protein is more stable than PIF4 protein in imbibed seeds. A) Higher protein stability of PIF1 than PIF4 in imbibed seeds. Sterilized seeds were irradiated with a far-red light pulse (3 μmol m⁻² s⁻¹, 5 min) and incubated at 22 °C in darkness for 12 h before sampling for the total protein extraction. Flag-tagged PIF proteins and alpha-tubulin were detected by immunoblot using anti-Flag antibody (Flag) and anti-tubulin antibody (Tub). Relative intensities were determined by comparing the intensities of Flag bands to those of Tub bands. Error bars indicate SEM (n = 3, biological replicates). B) Higher protein stability of 1N4C than 4N1C in imbibed seeds. Methods and notations are identical to A). Error bars indicate SEM (n = 3, biological replicates). C and D) Similar protein stabilities of PIF1 and PIF4 in red light C) and at high ambient temperature D). Total proteins were extracted from seedlings grown under continuous red light (Red, 20 μmol ⁻² s⁻¹) for 4 d C) or from seedlings grown under white light (50 μmol m⁻² s⁻¹, 22 °C) for 4 d then transferred to 28 °C for 6 h D). Notations are identical to A). Error bars indicate SEM (n = 3, biological replicates). E and F) Similar protein stability of 1N4C and 4N1C in red light E) and at high ambient temperature F). Total proteins were extracted from seedlings grown under continuous red light (Red, 20 μmol m⁻² s⁻¹) for 4 d E) or from seedlings grown under white light (50 μmol m⁻² s⁻¹, 22 °C) for 4 d and then transferred to 28 °C for 6 h F). Notations are identical to A). Error bars indicate SEM (n = 3, biological replicates). Alphabet letters above bars indicate statistical differences as determined by ANOVA with Tukey's HSD post hoc tests for multiple comparisons (P < 0.05). Individual data points are indicated with dots.

The N-terminal domain of PIF4 protein activates the expression of auxin-related target gene mRNAs higher than that of PIF1 protein does in seedlings

In seedlings, however, proPIF4-driven PIF1 and PIF4 proteins accumulated similarly either in red light or at high ambient temperature (Fig. 5, C and D, Supplementary Figs. S3 and S4). proPIF4-driven 4N1C and 1N4C proteins also accumulated similarly either in red light or at high ambient temperature (Fig. 5, E and F, Supplementary Figs. S3 and S4). These results indicate that PIF4 protein promotes hypocotyl elongation more strongly than PIF1 protein not due to higher accumulation of PIF4 protein over PIF1 protein but because of other properties associated with the N-terminal domain of PIF4 protein in seedlings.

Since PIF4 promotes hypocotyl elongation partly by activating auxin signaling (Franklin et al. 2011; Sun et al. 2012), we investigated whether the strong promotion of hypocotyl elongation by PIF4 is partly associated with its preferential activation of auxin signaling. We examined the mRNA expression of an auxin biosynthetic gene, YUCCA8 (YUC8), and an auxin signaling gene, IAA19, both of which are known to be directly targeted by various PIFs in seedlings (Hornitschek et al. 2012; Oh et al. 2012; Zhang et al. 2013; Pfeiffer et al. 2014). Interestingly, proPIF4-driven PIF4 activated the expression of both YUC8 and IAA19 mRNAs at high ambient temperature, whereas proPIF4-driven PIF1 did not (Fig. 6, A and B). Similarly, proPIF4-driven 4N1C but not 1N4C also activated the expression of both YUC8 and IAA19 mRNAs at high ambient temperature (Fig. 6, C and D). The inability of PIF1 to activate the expression of these genes was not due to its inability to target these gene promoters in vivo, as both PIF1 and PIF4 targeted YUC8 and IAA19 promoters equally well (Fig. 6E), while they did not target PP2A locus (Oh et al. 2009, 2012; Pfeiffer et al. 2014). Nor was it due to the lack of a transcription activation activity in PIF1 protein, as the N-terminal domains of both PIF1 and PIF4, when fused to GAL4 DNA-binding domain, could activate the expression of UAS-driven reporter genes both in yeasts and protoplasts (Fig. 6, F and G). Together, these results suggest that PIF4 protein promotes the hypocotyl elongation more strongly than PIF1 protein partly because the N-terminal domain of PIF4 protein activates the expression of auxin-related target gene mRNAs more efficiently than PIF1 protein does in seedlings.

Figure 6.

PIF4, but not PIF1, has the ability to activate the expression of auxin-related target genes at high ambient temperatures. A and B) Activation of YUC8A) and IAA19B) mRNA expression by proPIF4-driven PIF4 at high ambient temperature. Four-day-old white light-grown seedlings (50 μmol m⁻² s⁻¹, 22 °C) were either kept at 22 °C or transferred to 28 °C for 6 h before sampling for RNA extraction. mRNA expression levels were determined by RT-qPCR and normalized by the level of PP2A. Error bars indicate SEM (n = 3, biological replicates). Asterisks indicate significant differences between samples (Student's t test, *P < 0.05; **P < 0.001; n.s., not significant). C and D) Activation of YUC8C) and IAA19D) mRNA expression by proPIF4-driven 4N1C at high ambient temperature. Methods and notations are identical to A). Error bars indicate SEM (n = 3, biological replicates). E) Similar binding of PIF1 and PIF4 proteins to their target promoters. Seedlings (proPIF4:PIF1, proPIF4:PIF4) were grown under white light (50 μmol m⁻² s⁻¹, 22 °C) for 4 d and then subjected to high ambient temperature (28 °C) or maintained at ambient temperature (22 °C) for 6 h before ChIP assay. Crosslinked and sheared chromatins were immunoprecipitated with anti-Flag antibodies. The enrichment of target promoters (YUC8, IAA19) by proPIF4:PIF1-Flag (PIF1) and proPIF4:PIF4-Flag (PIF4) was determined by qPCR, and the enrichment levels were normalized by input samples. PP2A was used as a negative control. Error bars indicate SEM (n = 4, biological replicates). Alphabet letters above bars indicate statistical differences as determined by ANOVA with Tukey's HSD post hoc tests for multiple comparisons (P < 0.05). F) Similar transcriptional activation activities of PIF1N and PIF4N in yeast. AH109 yeast cells were transformed with vectors expressing either PIF1N (N-terminal domain of PIF1) fused with the GAL4 DNA-binding domain (BD) or PIF4N (N-terminal domain of PIF4) fused with BD. An empty pGBKT7 vector was used as a negative control (−). Transformed yeast cells were grown on SD/-Trp medium (-W) and SD/-Trp/-His medium (-WH). G) Similar transcriptional activation activities of PIF1N and PIF4N in protoplasts. Arabidopsis protoplasts were transformed with vectors expressing either PIF1N (N-terminal domain of PIF1) fused with the GAL4 DNA-binding domain (BD) or PIF4N (N-terminal domain of PIF4) fused with BD as an effector. UAS-fLuc (firefly luciferase) and pro35S:rLuc (Renilla luciferase) were cotransformed as a reporter and the transformation control, respectively. Transformed protoplasts were incubated overnight at 22 °C under dim light (10 μmol m⁻² s⁻¹ s) or additionally incubated at 28 °C for 6 h (28 °C). Error bars indicate SEM (n = 3, biological replicates). Alphabet letters above bars indicate statistical differences as determined by ANOVA with Tukey's HSD post hoc tests for multiple comparisons (P < 0.05). Individual data points are indicated with dots.

Discussion

Both the promoter and the protein contribute to the functional diversification of PIF1 and PIF4 genes, with their relative importance varying depending on the light responses

PIF genes are thought to have emerged in early nonvascular plants and have been functionally diversified after the multiplication by whole genome duplication events and local gene duplication events during the evolution of land plants (Possart et al. 2017; Jiang et al. 2022). However, what confers the shared and distinctive roles to PIF genes remains incompletely understood. The functional diversification of 2 PIF genes can arise from either differences in mRNA expression patterns mainly governed by their promoters or variations in molecular functions of proteins, such as DNA binding (Panchy et al. 2016). Here, we report the functional diversification of 2 PIF genes, PIF1 and PIF4, have arisen not only from a single element but from both promoters and proteins (Fig. 7). However, while both elements contributed to functional diversification, the degrees of their contributions were not constant; rather, the degrees varied depending on the specific light responses. Our analysis showed that the functional diversification of PIF1 and PIF4 genes in regulating seed germination was mainly attributed to the promoter, as the PIF1 promoter mainly conferred dominance to the PIF1 gene by promoting stronger expression of PIF1 mRNA compared to the PIF4 promoter in imbibed seeds. Additionally, PIF1 protein also partly contributed to the dominance of PIF1 gene in regulating seed germination by being more stable than PIF4 protein in imbibed seeds. Conversely, the functional diversification of PIF1 and PIF4 genes in regulating hypocotyl elongation under red light was primarily influenced by the protein, with the PIF4 protein largely assigning a dominant role to the PIF4 gene by promoting hypocotyl elongation more strongly than PIF1 protein. PIF4 promoter also partly contributed to the dominance of PIF4 gene by expressing PIF4 mRNA more strongly than PIF1 promoter in red light. In the context of regulating hypocotyl elongation at high ambient temperature, both PIF4 promoter and PIF4 protein are required for PIF4 gene to promote the elongation. The ability of the PIF4 protein, but not the PIF1 protein, to promote hypocotyl elongation either in red light or at high ambient temperature aligns with its ability, when expressed by the 35S promoter, to promote hypocotyl elongation in white light. Together, our analysis indicates that although both the promoter and the protein contribute to the functional diversification of PIF1 and PIF4, the relative importance of the promoter and the protein varies depending on specific light responses (Fig. 7).

Figure 7.

Functional diversification of PIF1 and PIF4 genes. The functional diversification of PIF1 and PIF4 genes arises from both the promoter and protein, with their relative contributions varying depending on light responses. PIF1 promoter mainly confers a dominant role to PIF1 gene in regulating seed germination, with a minor contribution from the PIF1 protein, whereas PIF4 protein primarily confers a dominant role to PIF4 gene, with a minor contribution from the PIF4 promoter in regulating hypocotyl elongation in red light. In contrast, both PIF4 promoter and PIF4 protein are required for PIF4 gene to promote hypocotyl elongation at high ambient temperature. Arrows are promoters (proPIF1, proPIF4), and squares are PIF1 and PIF4 protein (PIF1, PIF4). Relative sizes of arrows and squares represent relative importance of the promoter and the protein for the functional diversification of PIF1 and PIF4 genes in each light response.

The response-specific contributions of the promoter and the protein have been observed in the functional diversification of other plant genes as well. For instance, GL1 and MYB23 genes are functionally diversified primarily due to the promoter when regulating trichome initiation, whereas they are diversified due to both the promoter and the protein when regulating trichome branching (Kirik et al. 2005). Similarly, IAA7, IAA13, and IAA19 are functionally diversified due to the protein when regulating shoot development, whereas they are functionally diversified due to both the promoter and the protein when regulating gravitropic response (Muto et al. 2007). This form of functional diversification among related genes, wherein the roles of the promoter and the protein vary according to specific responses, might have been advantageous compared to functional diversification caused simply by the promoter, the protein, or the fixed contribution of both, regardless of responses. The differential utilization of both promoter and protein depending on specific responses may have enabled plants to enhance their abilities to regulate a more diverse range of responses with a finite set of related genes.

Molecular phylogenetic analysis suggests that land plant PIFs can be grouped into 3 clades. Among Arabidopsis PIFs, clade A includes PIF1, PIF4, and PIF5; clade B includes PIF7 and PIF8; and clade C includes PIF2, PIF3, and PIF6 (Jiang et al. 2022). However, when overexpressed by the 35S promoter, PIF3, PIF4, PIF5, and PIF7 strongly promote hypocotyl elongation in red light, whereas PIF1 and PIF8 barely induce elongation, and PIF2 and PIF6 inhibit it (Lee and Choi 2017). Similarly, when overexpressed by the 35S promoter, PIF1, PIF3, PIF4, and PIF5 inhibit seed germination (Lee et al. 2014; Oh et al. 2004). These findings indicate that the ability to promote hypocotyl elongation in red light or to inhibit seed germination is not restricted to specific clades. At the protein level, though all PIFs bind to phytochromes, their light stabilities differ, with PIF2, PIF6, and PIF7 being light stable while others are light labile (Lee and Choi 2017; Oh et al. 2020). This suggests that light-induced protein degradation is not confined to specific clades. Similarly, mRNA expression patterns cluster Arabidopsis PIF genes into 3 groups: PIF2/PIF6, PIF1/PIF3, and PIF4/PIF5/PIF7/PIF8, which does not match the 3 clades (Oh et al. 2020). The noncongruence between the molecular phylogenetic clades and physiological functions, protein light stabilities, or mRNA expression patterns suggests that PIFs have diverged both in their promoters and proteins across all clades. Thus, it would be interesting to further determine the causes of the functional diversification of PIFs within the same clade and among different clades.

PIF1 and PIF4 promoters possess different sequence elements

Our analyses support that the functional diversification of PIF1 and PIF4 genes is partly attributed to the divergence of their promoters. Although the specific cis-elements and regulatory factors responsible for the differences in mRNA expression patterns are not fully understood, studies suggest that 2 promoters possess different cis-regulatory elements.

Previous studies have shown that PIF4 promoter possesses cis-regulatory sequence elements regulating its mRNA expression patterns in a diurnal cycle and at high ambient temperatures. First, the expression of PIF4 mRNA is rhythmic peaking at midday, whereas that of PIF1 is not (Soy et al. 2014). This rhythmic expression of PIF4 mRNA is controlled by circadian clock components, including evening-phased transcription factors, Evening Complex (EC) comprised of EARLY FLOWERING 3 (ELF3), ELF4, and LUX ARRHYTHMO (LUX), that bind to PIF4 promoter and repress PIF4 mRNA expression (Nusinow et al. 2011). Consistent with the circadian regulation of PIF4 mRNA expression by these clock components, PIF4 promoter possesses a LUX binding site (LBS, GATWCG) that is directly targeted by LUX, in vivo, whereas PIF1 promoter does not possess the LBS and is not targeted by LUX (Nusinow et al. 2011; Ezer et al. 2017; Silva et al. 2020). Secondly, the expression of PIF4 mRNA is induced by high ambient temperatures, in contrast to PIF1 (Fig. 3A). This induction of PIF4 mRNA expression in response to elevated temperatures is partly associated with the decreased binding of LUX to PIF4 promoter (Mizuno et al. 2014) or the sequestration of ELF3 into speckles (Jung et al. 2020). Such inhibitions of EC components alleviate the repressive function of EC on the expression of PIF4 mRNA, thus leading to the increased expression of PIF4 mRNA at high ambient temperatures. Additionally, other components, such as BRASSINAZOLE-RESISTANT 1 (BZR1) and TCPs, directly bind to PIF4 promoter and enhance the expression of PIF4 mRNA at high ambient temperatures (Ibanez et al. 2018; Han et al. 2019; Zhou et al. 2019). Consistent with the regulation of PIF4 mRNA expression by these regulatory factors at high ambient temperature, the PIF4 promoter possesses a BZR1-binding element (CACGTGTC; Yu et al. 2011), which is absent in the PIF1 promoter. For TCP-binding sites (TBSs; Kosugi and Ohashi 2002), the PIF4 promoter contains multiple copies of various TBSs, including GGTCCAC, GGTCC, and CGGGC, whereas the PIF1 promoter has only a single TBS (GGTCC) within 2 kb promoters, suggesting that the difference in TBSs may also account for the regulation of PIF4 mRNA expression by TCPs at high ambient temperatures.

Unlike the better-studied transcriptional regulators of PIF4, only a few transcriptional regulators of PIF1 are reported, despite the critical function of PIF1 gene in seed germination and its high mRNA expression in imbibed seeds. Two reported transcriptional regulators of PIF1 are REVEILLE1 (RVE1) and EARLY FLOWERING IN SHORT DAYS (EFS). RVE1 is a MYB-like transcription factor with homology to CIRCADIAN CLOCK ASSOCIATED 1 (CCA1) and LATE ELONGATED HYPOCOTYL (LHY) that directly binds to PIF1 promoter and activates the expression of PIF1 mRNA (Yang et al. 2020). However, since the overexpression of REVEILLE 1 (RVE1) also causes the arrhythmic expression of PIF4 mRNA (Rawat et al. 2009), it is not certain if RVE1 specifically regulates the expression of PIF1 mRNA. EFS is a histone methyltransferase reported to be necessary for the high expression of PIF1 mRNA in imbibed seeds by increasing H3K36me3 levels and recruiting RNA polymerase II to the PIF1 locus (Lee et al. 2014). The transcription levels of other PIFs, including PIF3, PIF4, and PIF5, remain unaffected by the efs mutation in imbibed seeds, suggesting that the selective recruitment of EFS to PIF1 locus may contribute to the different mRNA expression patterns of PIF1 and PIF4 in imbibed seeds. It is not known, however, which specific sequence element in the PIF1 promoter is responsible for the recruitment of EFS to the PIF1 locus.

Finally, genome-wide binding site analyses of various transcription factors suggest that PIF1 and PIF4 promoters have both shared and distinct regulatory sequence elements. Transcription factor binding site predictions, based on publicly provided chromatin immunoprecipitation (ChIP)-seq and DAP-seq samples (Chow et al. 2019), indicate that PIF1 and PIF4 promoters are bound by 276 and 339 transcription factors, respectively. Among these, 82 transcription factors that specifically bind to the PIF1 promoter are largely linked to biological processes such as the “ethylene-activated signaling pathway (GO:0009873),” “response to gibberellin (GO:0009739),” and “regulation of seed dormancy process (GO:2000033).” On the other hand, 145 transcription factors that specifically bind to the PIF4 promoter are associated with the “brassinosteroid (BR) mediated signaling pathway (GO:0009742),” “cellular response to auxin stimulus (GO:0071365),” and “regulation of circadian rhythm (GO:0042752). Such the enrichment of different hormone signaling factors suggest that the spatiotemporal regulation of hormone signaling may partly account for the differences in mRNA expression patterns between PIF1 and PIF4. Further experimental analyses of these transcription factor binding sites in PIF1 and PIF4 promoters might provide insight into how PIF1 and PIF4 promoters have diverged to drive the functional diversification of PIF1 and PIF4 genes.

A promoter-dependent spatiotemporal expression of a gene is likely to lead to the formation of spatiotemporal-specific protein complexes. Such complexes would enable a single protein to perform different functions in different tissues at different developmental stages. When expressed by the same promoter, 2 related proteins could potentially form either identical protein complexes, different protein complexes, or a mixture of both identical and different protein complexes. Firstly, if 2 related proteins have the ability to form identical protein complexes, swapping promoters would allow these proteins to form identical spatiotemporal protein complexes, thereby enabling them to perform the same function. In this scenario, we can attribute functional diversification to the divergence of promoters. Secondly, if 2 related proteins cannot form identical protein complexes, swapping promoters would not enable them to perform the same function. Here, functional diversification would be attributed to the divergence of proteins. Thirdly, if 2 related proteins form a mix of identical and different protein complexes, swapping promoters would allow them to perform identical functions for some responses but not for others. In this case, functional diversification would be attributed to the divergence of both promoter and proteins. Since the functional diversification of PIF1 and PIF4 is influenced by both the promoter and the protein, it is plausible that PIF1 and PIF4 proteins may form a mix of identical and different protein complexes. Thus, it would be intriguing to identify such protein complexes.

N-Terminal domains of PIF1 and PIF4 proteins are functionally diversified presumably due to specific interacting proteins

Our analyses support that the functional diversification of PIF1 and PIF4 genes is also partly caused by the divergence of their proteins. The stronger activity of PIF1 protein compared to PIF4 protein in regulating seed germination is partly due to the higher stability of PIF1 protein than PIF4 protein in imbibed seeds. In contrast, the stronger activity of PIF4 protein than PIF1 protein in regulating hypocotyl elongation is not caused by different stabilities of PIF1 and PIF4 proteins in seedlings. It was also not due to different DNA-binding activities of PIF1 and PIF4 proteins or the lack of intrinsic transcription activation activity in PIF1 protein. Rather, it is associated with the transcriptional regulation of a subset of target genes as PIF4 protein, but not PIF1 protein, can activate the mRNA expression of auxin-related target genes in seedlings. Since both PIF proteins have transcription activation activity both in yeast and in protoplasts, the specific ability of PIF4 protein to induce the expression of auxin signaling genes could be associated with PIF4-specific interacting proteins in seedlings. Similarly, the higher stability of PIF1 protein than PIF4 protein in imbibed seeds but not in seedlings could be associated with PIF1-specific interacting proteins in imbibed seeds.

Our analyses with chimeric PIF1 and PIF4 proteins indicate that the divergence of PIF1 and PIF4 proteins can be traced to their N-terminal domains rather than their DNA-binding C-terminal domains. This was surprising because the role of a transcription factor is thought to be determined by its DNA binding to distinct target gene promoters, which, in the case of PIFs, would be attributed to the C-terminal domain. However, our analysis shows that the C-terminal domains of PIF1 and PIF4 are interchangeable, whereas the N-terminal domains are not. This indicates that DNA binding per se is not critical for the functional diversification of PIF1 and PIF4 in regulating seed germination and hypocotyl elongation. Rather, the N-terminal domains of PIF1 and PIF4 confer the characteristics: the N-terminal domain of PIF1 provides higher protein stability for chimeric PIFs in imbibed seeds, while the N-terminal domain of PIF4 imparts the ability of chimeric PIFs to activate the transcriptional regulation of auxin-related target genes.

The N-terminal domains of PIF proteins possess transcription activation domains (Yoo et al. 2021) and active phytochrome binding motifs, called APA and APB (Leivar and Quail 2011). Of 2 active phytochrome binding motifs, the N-terminal domains of all PIFs, including PIF1 and PIF4, possess APB motifs, while the N-terminal domains of a subset of PIFs, including PIF1 and PIF3 but not PIF4, possess APA motifs. In addition to these well-defined motifs, the N-terminal domain of PIF1 contains a stretch of aspartic acid-rich sequence that may act as a strong acidic transcription activation domain (Seipel et al. 1994). On the other hand, the N-terminal domain of PIF4 possesses a stretch of serine-rich sequence that destabilizes PIF4 protein when phosphorylated by BRASSINOSTEROID-INSENSITIVE 2 (BIN2) protein kinase in Arabidopsis (Bernardo-Garcia et al. 2014).

Though it is currently unknown if any of these motifs within the N-terminal domain of PIFs are responsible for the functional diversification of PIF1 and PIF4 proteins, it is tempting to speculate that some of these motifs contribute to diversification through interactions with specific proteins. Previous studies have shown that PIFs interact with various other proteins to regulate gene expression. One interesting example of PIF4 interacting proteins is HOOKLESS1 (HLS1). HLS1 is a transcriptional regulator identified as an ethylene signaling component (Lehman et al. 1996). HLS1 interacts more strongly with PIF4 protein than PIF1 protein, and this interaction is essential for PIF4 to activate the mRNA expression of a subset of target genes, including YUC8 gene (Jin et al. 2020). Although it is not currently known if HLS1 interacts with any motif in the N-terminal domain of PIF4, its preferential interaction with PIF4 protein and its role in activating specific PIF4 target genes make HLS1 a compelling candidate contributing to the functional diversification of PIF1 and PIF4 in regulating hypocotyl elongation in seedlings. Besides HLS1, other PIF4 interacting proteins, including BZR1, SEUSS (SEU), TCP5/13/17, and TIMING OF CAB EXPRESSION 1 (TOC1), have been shown to regulate hypocotyl elongation (Oh et al. 2012; Zhu et al. 2016; Huai et al. 2018; Han et al. 2019). Of these, BZR1, a major downstream transcription factor in the BR signaling pathway, interacts with the N-terminal domain of PIF4 and synergistically promotes hypocotyl elongation by activating the mRNA expression of PIF4 target genes including IAA19 and PACLOBUTRAZOL-RESISTANCE (PRE) genes (Oh et al. 2012). PIF4 and BZR1 also interact with AUXIN RESPONSIVE FACTOR 6 (ARF6) to cooperatively activate the expression of genes that promote elongation (Oh et al. 2014). Since elongation-promoting conditions such as shade and high ambient temperature activate both BR and auxin signaling (Kozuka et al. 2010; Ibanez et al. 2018), the cooperation among these factors might assist PIF4 in activating the expression of a subset of elongation-related target genes. However, since BZR1 has been shown to interact also with PIF1, and other factors (SEU, TOC1, and TCP5/TCP13/TCP17) have been shown to interact with the C-terminal domain rather than the N-terminal domain of PIF4, further experimental analyses are needed to determine whether these interacting proteins contribute to the functional diversification of PIF1 and PIF4 proteins.

Materials and methods

Plant materials and growth conditions

Arabidopsis (A. thaliana) plants were grown and maintained in a growth room set at 22 °C under long-day conditions (16 h of white light/8 h of darkness) for general growth and seed harvesting. To generate promoter-swapped PIF transgenic lines, promoters and coding sequences of PIF1 and PIF4 were amplified using primer sets (Supplementary Table S1) and cloned into a pCAMBIA1300-derived vector expressing a Flag-tag fusion protein (proPIF1:PIF1-Flag, proPIF1:PIF4-Flag, proPIF4:PIF1-Flag, and proPIF4:PIF4-Flag; Nguyen et al. 2015). To generate chimeric PIF transgenic lines, coding sequences corresponding to N-terminal domains and C-terminal domains of PIF1 and PIF4 (Supplementary Fig. S2) were amplified using primer sets (Supplementary Table S1) and ligated to form chimeric PIF genes (1N4C, 4N1C). These chimeric PIF genes were cloned into a pCAMBIA1300-derived vector expressing a Flag-tag fusion protein (proPIF1:1N4C-Flag, proPIF1:4N1C-Flag, proPIF4:1N4C-Flag, and proPIF4:4N1C-Flag). The cloned plasmids were transformed into the pif1 pif4 double mutant (Shin et al. 2009) using the Agrobacterium tumefaciens-mediated floral dip transformation method (Clough and Bent 1998; Zhang et al. 2006). For each construct, 2 independent lines were selected based on similar transgene mRNA levels at T2 bulk seedlings, and subsequently, homozygous lines were established from these 2 selected lines. To avoid any bias, phenotypes were not tested during the selection processes. Amplified homozygous lines were used for the final transgene mRNA expression test and phenotypic analyses.

Phenotype analyses

Seed germination assay

To determine seed germination frequencies, we placed 60 sterilized seeds onto a 1/100 MS agar plate. One hour after the start of sterilization, seeds were irradiated with far-red light (3 μmol m⁻² s⁻¹) for 5 min and incubated in darkness for 5 d at 22 °C (phyB_off). To assess the seed viability, the same set of seeds was spotted onto another 1/100 MS agar plate, identically irradiated with far-red light (3 μmol m⁻² s⁻¹) for 5 min and then incubated under white light (50 μmol m⁻² s⁻¹) for 5 d at 22 °C (cWL). Germination frequencies were determined by counting the number of seeds that had protruded radicles after 5 d of incubation in the dark or under white light.

Hypocotyl length

To measure the hypocotyl lengths of seedlings, around 40 sterilized seeds were spread on a 1/2 MS agar plate and cold stratified in darkness at 4 °C for 4 d. To measure hypocotyl length under red light or in the dark, cold-stratified seeds were exposed to white light (50 μmol m⁻² s⁻¹, 22 °C) for 6 h to promote germination, then grown either under continuous red light (20 μmol m⁻² s⁻¹, 22 °C) or in darkness (22 °C) for 4 d. To measure hypocotyl length at high ambient temperatures, cold-stratified seeds were exposed to 22 °C white light (50 μmol m⁻² s⁻¹) for 1 d, then transferred to either 22 or 28 °C white light (50 μmol m⁻² s⁻¹) for 6 d. The hypocotyl lengths of approximately 30 to 40 seedlings were measured using ImageJ software (http://imagej.nih.gov/ij/).

mRNA expression analysis

To analyze mRNA expression, seeds or seedlings were grown on 1/2 MS agar, collected, flash-frozen in liquid nitrogen, and ground with tungsten beads using TissueLyser (QIAGEN). Total RNAs were extracted according to the manufacturer’s protocol (Spectrum Plant Total RNA Extraction Kit, Sigma-Aldrich). Two micrograms of RNAs was used for the reverse transcription using MMLV-RTase (Promega) and oligo dT primer. mRNA expression level of each gene was determined by qPCR with TOPreal qPCR 2X PreMIX (Enzynomics) and gene-specific primers in CFX Connect machine (Bio-Rad Laboratories). The relative expression levels of each gene were calculated using the delta–delta cycle threshold (CT) method (Rao et al. 2013). All gene expression values were normalized to that of PP2A in the same sample. The primer sequences used for mRNA expression analysis are provided in Supplementary Table S1.

Immunoblot analysis

Total proteins were extracted from seeds or seedlings grown on 1/2 MS plate. Samples were flash frozen in liquid nitrogen and ground with tungsten beads using TissueLyser (QIAGEN). The ground samples were dissolved in a UREA buffer (100 mM NaH₂PO₄, 8 M UREA, and 10 mM Tris-HCl, pH8.0), and debris were removed by centrifuging at 14,621 × g for 10 min at 4 °C. The resulting supernatants were mixed with 5X SDS sample buffer and heated at 95 °C for 3 min. The protein levels were determined by immunoblot analyses using anti-FLAG (Sigma-Aldrich) and anti-TUB (Sigma-Aldrich) antibodies as primary antibodies and goat anti-rabbit IgG-HRP (Cell signaling) and goat anti-mouse IgG-HRP (AbFrontier) as secondary antibodies. The blots were developed with Clarity Western ECL substrate (Bio-Rad Laboratories) and then detected using the ChemiDoc XRS+ System (Bio-Rad Laboratories). Image Lab Software was used for quantifying band intensities. Relative intensities were calculated based on the band intensities of the FLAG protein bands in comparison to the band intensities of the tubulin bands in the same sample.

ChIP

ChIP experiments were conducted as described previously with minor adjustments (Park et al. 2018). One thousand proPIF4:PIF1 and proPIF4:PIF4 seeds were placed on a 1/2 MS plate, cold stratified in darkness at 4 °C for 4 d, and grown under continuous white light (50 μmol m⁻² s⁻¹, 22 °C) for 4 d. Seedlings were then transferred to 28 °C for 6 h or kept at 22 °C for the same duration. The seedlings were crosslinked with 1% formaldehyde using vacuum infiltration and then ground to a fine powder with a mortar and pestle in liquid nitrogen. After removing debris and shear chromatins, protein–DNA complexes were precipitated with anti-FLAG antibodies (Sigma-Aldrich) and Protein A Agarose/Salmon sperm DNA (Sigma-Aldrich). The precipitated DNAs were decrosslinked and purified using QIAquick PCR Purification Kits (QIAGEN) and subjected to real-time qPCR using iQ STBR Green Supermix (Bio-Rad Laboratories) with gene specific primers (Supplementary Table S1). All values were normalized based on their respective input values.

Transcription activation activity assay

To test the transcription activation activity assay in yeast, N-terminal domains of PIF1 and PIF4 were cloned into the pGBKT7 vector. The cloned plasmids were then transformed into the yeast AH109 strain (PT3024-1, Clontech). The transformed yeast cells were spotted on yeast dropout media plates lacking either tryptophan (-W) or both tryptophan and histidine supplemented with 10 mM of 3-aminotriazole (-WH). The yeast cells were incubated at 28 °C for 3 d to allow the formation of colonies for visualization.

To test the transcription activation activity in protoplasts, N-terminal domains of PIF1 and PIF4 were each cloned into a pBI221-drived vector that contains 35S promoter and the Gal4 DNA-binding domain (pro35S:DBD-PIF1N, pro35S:DBD-PIF4N). Each of these cloned effector plasmids was cotransformed with proUAS:fLuc (firefly luciferase) and pro35S:rLuc (Renilla luciferase) into Arabidopsis protoplasts as previously described (Yoo et al. 2007). Transfected protoplasts were incubated under dim white light (10 μmol m⁻² s⁻¹) at 22 °C overnight or additionally incubated at 28 °C for 6 h (28 °C). The expressed luciferases were determined using Dual-Luciferase Reporter assay system (Promega) and immediately measured using Multi-functional Microplate reader (Tecan). The firefly luciferase activity of each sample was normalized to its respective Renilla luciferase activity (fLuc/rLuc).

Statistical analysis

Statistical analyses were performed as described in each figure legend. Statistical data are provided in Supplementary Table S2.

Accession numbers

Sequence data from this article can be found in TAIR (arabidopsis.org) under the following accession numbers: PIF1 (AT2G20180), PIF4 (AT2G43010), PP2A (AT1G13320), YUC8 (AT4G28720), and IAA19 (AT3G15540).

Author contributions

H.K., N.L., and G.C. designed the experiments. H.K., N.L., and Y.K. performed the experiments. H.K. and G.C. wrote the paper.

Supplementary data

The following materials are available in the online version of this article.

Supplementary Figure S1. mRNA expression levels of endogenous and transgenic PIFs (supports Figs. 1 and 2).

Supplementary Figure S2. Sequence alignment between PIF1 and PIF4 proteins (supports Fig. 4).

Supplementary Figure S3. Transgenic PIF protein levels in red light (supports Fig. 5).

Supplementary Figure S4. Transgenic PIF protein levels at high ambient temperature (supports Fig. 5).

Supplementary Table S1. List of primers that are used in this study.

Supplementary Table S2. ANOVA and t test results.

Funding

This research was supported by the National Research Foundation of Korea (NRF-2018R1A3B1052617).

Data availability

All data are incorporated into the article and its online supplementary material.

Dive Curated Terms

The following phenotypic, genotypic, and functional terms are of significance to the work described in this paper:

• BZR1 Gramene: AT1G75080

• BZR1 Araport: AT1G75080

• TOC1 Gramene: AT5G61380

• TOC1 Araport: AT5G61380

• ARF6 Gramene: AT1G30330

• ARF6 Araport: AT1G30330

• BIN2 Gramene: AT4G18710

• BIN2 Araport: AT4G18710

• CCA1 Gramene: AT2G46830

• CCA1 Araport: AT2G46830

• ELF4 Gramene: AT2G40080

• ELF4 Araport: AT2G40080

• MYB23 Gramene: AT5G40330

• MYB23 Araport: AT5G40330

• MYB2 Gramene: AT2G47190

• MYB2 Araport: AT2G47190

• IAA7 Gramene: AT3G23050

• IAA7 Araport: AT3G23050

• AT1G13320 Gramene: AT1G13320

• AT1G13320 Araport: AT1G13320

• PIF3 Gramene: AT1G09530

• PIF3 Araport: AT1G09530

• ELF3 Gramene: AT2G25930

• ELF3 Araport: AT2G25930

• PP2A Gramene: AT1G13320

• PP2A Araport: AT1G13320

• PIF7 Gramene: AT5G61270

• PIF7 Araport: AT5G61270

• IAA13 Gramene: Os03g0742900

• IAA13 Araport: Os03g0742900

• RGL2 Gramene: AT3G03450

• RGL2 Araport: AT3G03450

• RBOHD Gramene: At5g47910

• RBOHD Araport: At5g47910

• WRKY13 Gramene: AT4G39410

• WRKY13 Araport: AT4G39410

• pifQ Gramene: pif1 pif3 pif4 pif5

• pifQ Araport: pif1 pif3 pif4 pif5

• OsPIL1 Gramene: Os03g0782500

• OsPIL1 Araport: Os03g0782500

• RGA Gramene: At2g01570

• RGA Araport: At2g01570