Phytochrome-interacting factors orchestrate hypocotyl adventitious root initiation in Arabidopsis

Adventitious roots (ARs) are important components of root system architecture. Multiple endogenous and environmental factors regulate adventitious rooting (Bellini et al., 2014). Previous studies have shown that auxin accumulation is a crucial step in triggering the initiation of root primordia and is mainly regulated by auxin biosynthesis and transport (Verstraeten et al., 2014; Xu, 2018). In Arabidopsis thaliana, 11 YUCCA (YUC) gene family members encoding flavin monooxygenases catalyze the rate-limiting step in auxin biosynthesis and the conversion of indolepropionic acid (IPA) to indoleacetic acid (IAA) (Zhao et al., 2001; Cheng et al., 2006). YUC-mediated endogenous auxin biosynthesis is required for biological processes; e.g. YUC1 and YUC4 are induced by wounding, and the expression of YUC5, YUC8 and YUC9 increases marginally in response to darkness, while YUC2 and YUC6 primarily contribute to maintenance of basal auxin levels (Chen et al., 2016). Previous studies have shown that YUC6 is expressed in hypocotyl adventitious root (HAR) primordial (Della Rovere et al., 2013; Veloccia et al., 2016); however, the roles of YUC genes in HAR formation remain unknown.

Auxin transport also plays a crucial role in adventitious rooting (Lakehal and Bellini, 2019). PIN-FORMED (PIN) and P-glycoprotein (ABCB/PGP) transport proteins are major auxin efflux carriers, while AUXIN1/LIKE-AUX1 (AUX/LAX) proteins are major auxin influx carriers (Swarup and Peret, 2012; Grones and Friml, 2015). PIN1 and PIN6 have been reported to play opposite roles in HAR formation (Della Rovere et al., 2013; Simon et al., 2016). Moreover, PIN1 and ABCB19 are involved in adventitious rooting from plants whose roots have been removed; in particular, the transcription of ABCB19 is rapidly induced upon root excision, which increases auxin transport and accumulation in the basal region of the hypocotyl, and ultimately drives the formation of ARs (Sukumar et al., 2013). In addition, some studies have reported that the coordinated activity of PIN1, AUX1 and LAX3 causes auxin to accumulate and maintains auxin at maximum levels during HAR formation (Della Rovere et al., 2013, 2015; Veloccia et al., 2016). However, how these factors are regulated in HAR formation remains unclear.

Light is an important environmental factor that suppresses HAR formation; specifically, photoreceptors interact with and stabilize Aux/IAA14 proteins to inhibit the HAR-positive regulators AUXIN RESPONSE FACTORS (ARF) 7/ARF19, LATERAL ORGAN BOUNDARIES DOMAIN (LBD) 16/LBD29, and WUSCHEL-RELATED HOMEOBOX (WOX) 5/WOX7 (Li et al., 2021). However, whether other light signaling components are also involved in HAR initiation is unknown. Phytochrome-interacting factors (PIFs) are basic helix-loop-helix (bHLH) transcription factors (TFs) that accumulate in the dark to promote skotomorphogenesis (Leivar et al., 2008; Zhang et al., 2013). Several studies have reported that PIFs regulate YUC5, YUC8 and YUC9 expression to increase auxin biosynthesis during the shade-avoidance response (Hornitschek et al., 2012; Li et al., 2012; Sun et al., 2012; Di et al., 2016; Pucciariello et al., 2018). However, whether and how PIFs regulate HAR formation is still unknown.

In this study, we reported that PIFs are central factors involved in darkness-induced HAR development. We found that PIFs improved HAR formation by directly activating YUC2 and YUC6, but not YUC5, YUC8 or YUC9. In addition, PIFs mediate the darkness-induced expression of auxin influx carrier-related genes and TF-encoding genes involved in root priming in the hypocotyl to initiate HAR. These findings demonstrate that PIFs regulate HAR formation in darkness by coordinating auxin biosynthesis, auxin transport and TFs involved in root priming. Additionally, these findings reveal a new regulatory mechanism controlling HAR formation, therefore improving our understanding of how darkness induces HAR initiation in Arabidopsis.

Darkness is necessary for HAR formation because photoreceptors inhibit this process by directly interacting with and stabilizing IAA14 proteins (Li et al., 2021). PIF proteins accumulate in the dark and promote skotomorphogenesis in etiolated seedlings (Castillon et al., 2007). However, to address whether PIFs are also involved in the regulation of HAR formation, we grew various Arabidopsis pif mutants and wild-type Columbia (Col-0) plants on half-strength Murashige and Skoog (1/2 MS) media for 14 days in the dark. We evaluated the HARs of these plants and found that the HAR formation in the pif mutants was defective. Among the single mutants, pif1-1 and pif3-7 developed fewer HARs than did the wild-type Col-0 plants, while pif5-3 developed slightly more HARs than the wild type (Fig. 1A,B). These results indicated that PIF1 and PIF3 contribute to HAR formation, but there might be some compensatory mechanism at play in the pif5-3 mutants. Quantitative reverse transcription-polymerase chain reaction (qRT-PCR) confirmed that the expression of PIF1, PIF3 and PIF4 was upregulated in the pif5-3 mutants (Fig. S1). Furthermore, pif double and triple mutants developed significantly fewer HARs than the corresponding single mutants did, and the pif1 pif3 pif4 (pif134) triple mutant developed the fewest HARs among the triple mutants (Fig. 1C-F). In addition, the pif1 pif3 pif4 pif5 quadruple mutant (pifq) did not develop any HARs. Taken together, these data indicate that PIFs are positive regulators and act redundantly in HAR formation.

To confirm this conclusion, we overexpressed Flag-tagged PIFs in Col-0 plants. Overexpression of PIF1-Flag significantly increased HAR numbers in both darkness and red light (Fig. 2A,B and Fig. S2A,B). Moreover, overexpression of PIF3-Flag, PIF4-Flag and PIF5-Flag also promoted HAR formation in red light but not in the dark (Fig. 2A,B and Fig. S2C-H). In contrast, overexpression of PIF4-Flag and PIF5-Flag suppressed HAR formation in the dark (Fig. 2A,B). We performed immunoblot assays to detect the levels of PIF-Flag proteins in all four PIF-Flag overexpression lines under darkness and red light, and found that PIF4-Flag and PIF5-Flag overexpression lines produced PIF4-Flag and PIF5-Flag levels that were notably higher than the PIF1-Flag and PIF3-Flag levels in PIF1-Flag and PIF3-Flag overexpression lines, respectively (Fig. 2C,D). Interestingly, we noticed that ∼80% of the hypocotyls of PIF4-Flag and PIF5-Flag overexpression lines were abnormal (Fig. 2E) and hardly produce HARs under darkness. Moreover, Trypan Blue staining showed that cell death occurred in the stele of the hypocotyls of both the PIF4-Flag and PIF5-Flag plants (Fig. 2F). Such hypocotyl defects might be a result of abnormal xylem differentiation, given that auxin promotes xylem differentiation (Ursache et al., 2014; Smetana et al., 2019) and PIF4 and PIF5 activate expression of auxin synthesis genes (Li et al., 2012; Sun et al., 2012).

To further investigate the roles of PIFs in AR formation, we further evaluated adventitious rooting of leaf explants of the wild type and pifq mutant. Leaf explants of Col-0 and pifq were cultured on B5 media for 12 days under darkness. Rooting was significantly reduced in the pifq mutant, and the mutant phenotype could be rescued by supplementation with IAA (Fig. S3). Taken together, these results suggest that PIFs also can promote adventitious rooting in leaf explants by affecting auxin levels in the leaves.

Given that PIFs have been reported to directly regulate the expression of some YUC genes (YUC5, YUC8 and YUC9) to promote hypocotyl elongation (Hornitschek et al., 2012; Li et al., 2012; Sun et al., 2012), we were interested in whether PIFs control HAR formation through regulation of these YUC genes. To address this question, we examined HAR formation in yuc mutants. Surprisingly, the yuc589 (yuc5 yuc8 yuc9) triple mutant and yuc14 (yuc1 yuc4) double mutant developed HARs similar to those of the wild type (Fig. 3A,B), suggesting that YUC1, YUC4, YUC5, YUC8 and YUC9 are not necessary for HAR development. In contrast, compared with the wild type, the yuc26 (yuc2 yuc6) double mutant developed significantly fewer HARs (Fig. 3A-D). Compared with the wild type, the yuc2 and yuc6 single mutants also developed significantly fewer HARs, while yuc26 developed fewer HARs than did yuc2 or yuc6 (Fig. 3C,D). These results indicate that YUC2 and YUC6 act redundantly in HAR formation.

Supporting the conclusion that YUC genes are required for HAR formation, we found that Col-0 seedlings were not able to develop HARs by yucasin treatment (Fig. S4A,B). Furthermore, when we introduced the DR5:β-glucuronidase (GUS) expression construct into pifq by crossing with a DR5:GUS transgenic line in the Col-0 background, we found that GUS signals were obviously weaker in pifq than in Col-0 (Fig. S5), indicating that PIFs are required for auxin homostasis. Together, these data indicated that PIFs might regulate HARs through the regulation of YUC2 and YUC6. To confirm this, we examined the expression of YUC2 and YUC6 in the hypocotyls of Col-0 and pifq mutants by using qRT-PCR. We found that the expression of YUC2 and YUC6 significantly decreased in the pifq mutant (Fig. 3E). In addition, we also generated promoter-GUS reporter lines for the YUC2 and YUC6 genes in the wild-type and pifq background. GUS staining assays showed that the expression of YUC2 and YUC6 significantly decreased in the hypocotyls of the pifq mutant (Fig. S6). Interestingly, YUC2 and YUC6 were also induced in the hypocotyls by darkness (Fig. S7A,B), which is consistent with the accumulation of PIFs in the darkness-treated hypocotyls (Lorrain et al., 2008; Shen et al., 2008; Soy et al., 2012; Ni et al., 2013).

To test whether PIFs directly regulate the expression of YUC2 and YUC6, we performed a dual-luciferase (dual-LUC) assay in Arabidopsis protoplasts. We found that the expression of pYUC2-LUC and pYUC6-LUC increased when PIFs were co-expressed (Fig. 3F,G). To further test whether PIFs could directly bind the promoters of YUC2 and YUC6, we generated recombinant His-PIF1 proteins and performed a gel electrophoretic mobility shift assay (EMSA) with the YUC2 or YUC6 promoter fragments containing the E-box motif or PBE-box motif, respectively. Our data showed that PIF1 could directly bind to the promoter fragments of YUC2 and YUC6 (Fig. 4A,D). Furthermore, using PIF1-Flag transgenic plants, we performed a chromatin immunoprecipitation (ChIP)-qPCR assay, the result of which confirmed that PIF1 binds to the promoters of YUC2 and YUC6 in vivo (Fig. 4B,E). We also performed ChIP-qPCR with the PIF4-Flag overexpression line and found that the predicted PIF-binding fragments of the YUC2 and YUC6 promoters precipitated to a much greater degree in this line than in the PIF1-Flag transgenic line (Fig. 4C,F), most likely the result of the high level of PIF4 (Fig. 2C,D). Together, these results demonstrated that PIFs activate the expression of YUC2 and YUC6 by directly binding to cis-elements of these two genes.

To address whether the defect in auxin biosynthesis is the only reason for the inability of pifq mutants to develop HARs, we applied IAA to Col-0 plants and to yuc6, yuc26 and pifq mutants. Surprisingly, we found that IAA could promote HAR formation in Col-0, yuc6 and yuc26 mutants, but not in the pifq mutant (Fig. S8A), indicating that auxin biosynthesis is not the only cause of the defective HAR formation in pifq. High concentration of IAA also could not rescue the defective HAR formation in pifq mutants (Fig. S8B,C).

As exogenous auxin could not complement the defective HAR formation in the pifq mutant, we were interested in whether increased levels of endogenous auxin could rescue the phenotype of the pifq mutant. To investigate this, we knocked out SUR1 in the pifq mutant background via the clustered, regularly interspaced, short palindromic repeat (CRISPR)/CRISPR-associated 9 (Cas9) technology to increase endogenous auxin concentrations (Boerjan et al., 1995). We found that sur1 pifq quintuple mutants presented high numbers of ARs, and their phenotype was similar to that of sur1 (Fig. S8D). Taken together, these results indicated that PIFs might regulate HAR formation by affecting auxin influx.

Given that endogenous but not exogenous auxin could complement the defective HAR formation in the pifq mutant, we hypothesized that PIFs are also required for auxin transport. To test this hypothesis, we examined whether the expression of AUX1 and LAX3, two genes encoding auxin influx carriers, is regulated by PIFs. qRT-PCR showed that the transcript abundances of AUX1 and LAX3 were lower in the hypocotyls of pifq than in those of the wild type in darkness (Fig. 5E,F). These results suggested that PIFs promote the transcription of AUX1 and LAX3 in darkness. Moreover, GUS staining assays showed that the expression of AUX1 and LAX3 significantly decreased in the hypocotyl of the pifq mutant (Fig. S9), the results of which were consistent with the qRT-PCR results. Additionally, the AUX1_pro:AUX1-GFP reporter construct revealed that AUX1 was high expressed in the stele and the HAR primordia of wild type, but was apparently low in that of pifq (Fig. 5G). Similarly, the reduced expression of LAX3_pro: LAX3-GFP was also observed in the stele tissues in the hypocotyl of pifq mutant (Fig. 5H). Furthermore, both AUX1 and LAX3 were induced in hypocotyls by darkness (Fig. S7C,D), which is consistent with the role of PIFs in enhancing AUX1 and LAX3 expression.

Given that AUX1 and LAX3 are regulated by PIFs, we speculated that influx of auxin into the cells is required for HAR initiation. Consistent with this hypothesis, we found that the number of HARs was suppressed by treatment with the auxin influx inhibitor 1-naphthoxyacetic acid (1-NOA) (Fig. S4C,D). In addition, we found that the number of HARs of the aux1 and lax3 mutants were decreased compared with those of the wild type, and aux1 developed fewer HARs than lax3 did, suggesting that both AUX1 and LAX3 play roles in HAR formation, while AUX1 plays a stronger role. Moreover, compared with the two single mutants, the aux1 lax3 double mutant developed significantly fewer HARs, indicating that AUX1 and LAX3 act redundantly in terms of HAR formation (Fig. 5A,B). In contrast to aux1 and lax3, compared with the wild type, the lax2 mutant developed similar numbers of HARs, indicating that LAX2 might not regulate HAR formation (Fig. 5C,D). These results, together with the expression evidence, showed that PIFs promote HAR formation through regulation of AUX1 and LAX3.

To test whether PIFs directly activate AUX1 and LAX3 to improve HAR formation, we assessed the promoters of AUX1 and LAX3. We identified several typical PIF-binding sites (G-box, PBE-box and E-box motifs) in these two promoters. Yeast one-hybrid (Y1H) experiments showed that all four PIFs could bind the E-box motif in the promoter of AUX1 (Fig. 6A,B). In contrast, PIFs could not bind these promoters when the E-box was mutated, suggesting that the binding was specific (Fig. 6A,B). Furthermore, we performed transient transcription assays in protoplasts to analyze whether PIFs could activate AUX1 and LAX3 transcription in vivo. Our data showed that co-expression of PIF proteins strongly promoted the activity of LUC reporters driven by the promoters of AUX1 and LAX3 (Fig. 6C,D), suggesting that PIFs directly activate the expression of AUX1 and LAX3.

To further confirm this binding, we generated a recombinant His-PIF1 protein and performed EMSAs together with AUX1 and LAX3 promoter fragments containing the E-box or G-box. Our data showed that PIF1 could bind to the promoter fragments of AUX1 and LAX3 (Fig. 6E,F). Furthermore, we carried out ChIP-qPCR assays with PIF1-Flag and PIF4-Flag transgenic plants, and the results confirmed that PIF1 and PIF4 bound to the promoters of AUX1 and LAX3 in vivo (Fig. 6G,H). Together, these results demonstrated that PIF proteins can bind to the promoters of AUX1 and LAX3, and activate their expression.

To further address the role of AUX1 in PIF-mediated HAR formation, we overexpressed 35S-AUX1 in pifq. We generated three independent transgenic lines overexpressing AUX1 (Fig. S10D) and observed that a few 35S-AUX1/pifq lines were able to develop HARs in the dark (Fig. S10A-C). These data indicated that overexpression of AUX1 could only partially complement the HAR defect of pifq. Interestingly, the application of IAA increased the number of HARs in Col-0 but did not affect the number of HARs in the AUX1 overexpression lines in the pifq background. These observations supported that AUX1 acts downstream of PIFs in regulating HAR initiation but that PIFs are likely required for other processes involved in HAR formation.

Naphthaleneacetic acid (NAA) is an auxin analog of IAA and is able to enter cells by passive diffusion (Delbarre et al., 1996). We found that NAA was able to induce the development of greater numbers of HARs in Col-0, aux1, lax3 and aux1 lax3 but not in the pifq mutant (Fig. S10E,F). These data further confirm that PIF is also required for processes involved in HAR formation other than auxin biosynthesis and transport.

Given that we have reported that, by stabilizing IAA14 proteins, phytochrome represses the initiation of HARs (Li et al., 2021), we further wondered whether PIFs regulate phytochromes or directly affect auxin signaling components to regulate HAR initiation. To address these questions, we first investigated the genetic relationship between PHYB and PIFs. Our data showed that, like the pifq quadruple mutant, the phyB pifq quintuple mutant did not develop HARs either in darkness or in red light. In contrast, phyB was able to develop HARs either in darkness or red light, while Col-0 developed HARs only in the dark (Fig. S11A,B). These data indicate that PIFs are epistatic to PHYB in terms of HAR formation.

We further performed immunoblot assays with anti-IAA14 antibodies to determine the IAA14 protein level in the hypocotyls of wild-type, phyB, pifq and phyB pifq mutant seedlings in darkness and red light. The results showed that accumulation of IAA14 protein was hardly detected in any of the seedlings grown in the dark (Fig. S11C), indicating that the inability of pifq to develop HARs is not related to steady IAA14 levels. In addition, although phyB pifq did not develop any HARs under red light, the abundance of IAA14 protein in this mutant was also reduced (Fig. S11D). These data further confirmed that PIFs regulate HAR formation not through affecting the stability of IAA14.

We also examined whether PIFs interact with AUX/IAA14-ARF7/19 modules. Yeast two-hybrid (Y2H) experiments showed that PIFs did not interact with either IAA14 or ARF19 proteins that lacked an activation domain (Fig. S12). Moreover, the expression of ARF7 and ARF19 was not affected in the hypocotyls of pifq compared with the wild type grown in darkness (Fig. S13). Taken together, these data together support the observation that PIFs do not regulate AUX/IAA14-ARF7/19 modules.

LBD16, LBD29, WOX5 and WOX7 encode four essential TFs required for the formation of lateral roots (LRs) and HARs (Liu et al., 2014; Hu and Xu, 2016; Sheng et al., 2017; Li et al., 2021). We examined the expression of these four genes and observed that they were induced by darkness (Fig. S7E-H) but were downregulated in the pifq hypocotyls (Fig. 7A-D). Furthermore, WOX5_pro:GUS and LBD16_pro:LBD16-GUS reporter lines also showed that expression of WOX5 and LBD16 were not detectable in pifq mutant (Fig. 7E,F). Similarly, the confocal images showed that LBD16-GFP was expressed in HAR primordia and localized in the nuclei, whereas no GFP fluorescence was observed in pifq (Fig. 7G). Such downregulation might be caused by defective auxin biosynthesis or transport in the pifq mutant, but it is also possible that PIFs could directly regulate LBD16/29 and/or WOX5/7. To test this hypothesis, we tested whether PIFs could bind the promoter of LBD16, LBD29, WOX5 and WOX7 . Y1H assays showed that PIF1 is indeed able to bind to the promoter fragments of LBD16 and LBD29 and WOX5/7 that contain a potential PIF-binding sequence (Fig. 7H-K). We further performed EMSAs, the results of which confirmed that PIF1 is able to directly bind to the promoter sequences of LBD16, LBD29, WOX5 and WOX7 in vitro (Fig. 8A-D). Furthermore, ChIP-qPCR assays demonstrated that PIF1 and PIF4 bound to the promoters of LBD16, LBD29, WOX5 and WOX7 in vivo (Fig. 8E-H). Together, these data showed that PIFs directly activate the expression of LBD16, LBD29, WOX5 and WOX7.

Multiple components regulating HAR formation have been identified, including Aux/IAAs, ARFs, photoreceptors, Ago1 and other proteins involved in hormone homeostasis (Sorin et al., 2005; Gutierrez et al., 2009, 2012; Lakehal et al., 2019; Lee et al., 2019; Li et al., 2021). These studies indicated that the initiation of HARs involves a complex process and is initiated with the integration of numerous factors and environmental signals. In this study, we identified PIFs as being essential TFs for the initiation of HARs and ARs. Further studies established that PIFs coordinate multiple processes, including auxin biosynthesis, auxin transport and the transcriptional activation of LBD16, LBD29, WOX5 and WOX7 to trigger HAR initiation (Fig. 9). These findings not only revealed essential regulators involved in HAR formation but also revealed a new mechanism underlying darkness-induced HAR formation.

Previous studies have reported that PIFs are expressed in the hypocotyl and that PIF proteins are key components in skotomorphogenesis, including hypocotyl elongation (Bae and Choi, 2008; Zhang et al., 2013). Given that darkness is necessary for HAR formation, we thus speculated that PIFs might be involved in HAR formation. As expected, we found that different pif mutants showed various defects in adventitious rooting from hypocotyls; pifq quadruple mutants were not able to develop any HARs, verifying that PIFs play both redundant and essential roles in HAR formation (Fig. 1). This conclusion was further supported when PIFs were overexpressed, which resulted in significantly more HAR than the wild type under red light. Surprisingly, compared with the wild type, the PIF4 and PIF5 overexpression lines developed fewer HARs in darkness, although the PIF4 overexpression line developed more HARs than the PIF1 overexpression line in red light (Fig. 2). Further investigation revealed that the relatively low numbers of HARs of the PIF4- and PIF5-overexpressing lines might be caused by defects of xylem differentiation, as we observed that ∼80% of the hypocotyls overexpressing PIF4 or PIF5 exhibited abnormal cell death in the stele (Fig. 2). Previous studies have showed that auxin biosynthesis and influx promote xylem differentiation (Ursache et al., 2014; Smetana et al., 2019), which is reversely related to AR formation (Della Rovere et al., 2015). These data indicate that PIFs might be involved in many more processes than expected, which needs further exploration.

The initiation of HARs requires the integration of multiple processes, including the perception of environmental signals and the activation of internal regulators. Among them, auxin biosynthesis, transport and signaling are central processes, but it is unknown how these processes are integrated. Previous studies showed that PIFs are able to activate the expression of YUC5, YUC8 and YUC9 (Hornitschek et al., 2012; Li et al., 2012; Sun et al., 2012). However, our data showed that YUC5, YUC8 and YUC9 are not required for HAR formation. Instead, we observed that YUC2 and YUC6 are essential for HAR formation. Our evidence also indicated that PIFs directly activate the expression of YUC2 and YUC6 by binding to the promoters of these genes. A previous study reported that YUC2 and YUC6 are involved in adventitious rooting of leaf explants and contribute to basal auxin levels (Chen et al., 2016). Consistently, we found that PIFs are also required for adventitious rooting of leaf explants. Taken together, these data indicated that PIFs could promote auxin synthesis in multiple pathways.

Directional auxin transport in plants is achieved through a combination of auxin transport proteins, including PINFORMED (PIN) efflux carriers, ATP-binding cassette group B (ABCB) auxin transporters and AUX/LAX auxin uptake permeases (Blakeslee et al., 2005; Swarup and Peret, 2012; Grones and Friml, 2015). Previous studies have shown that AUX/LAX auxin uptake permeases contribute to LR and HAR formation (Swarup et al., 2008; Della Rovere et al., 2013, 2015), but how these genes are regulated remains unknown. The identification of PIFs as direct regulators of these genes revealed a new mechanism underlying the regulation of HAR formation.

LBD16/29 and WOX5/7 encode essential TFs involved in HAR formation, and their expression is controlled by upstream regulators involved in auxin signaling, including IAA14 and ARF7/19. We have reported that PHYB suppresses HAR formation by stabilizing IAA14 proteins. Given that PIFs are PHYB-interacting proteins, it was not unreasonable to speculate that PIFs might also regulate HARs by affecting steady-state IAA14 levels. However, our genetic and molecular evidence demonstrated that PIFs act downstream of PHYB in terms of regulating HAR formation and did not affect steady-state levels of IAA14. Instead, we found that PIFs are able to directly activate the expression of LBD16/29 and WOX5/7 to promote HAR formation. These findings indicated that the essential TFs for priming of HAR primordia are not only responsive to auxin signaling but also can be directly regulated by PIFs.

As PIFs could directly activate auxin biosynthesis, transport and TFs for HAR priming, these essential skotomorphogenesis regulators essentially function as coordinators of HAR formation. Multiple regulatory points of HARs controlled by a single type of TF reflect that HAR formation is an orchestrated process. Such regulation of HARs by PIFs might facilitate the tight coupling of HAR formation and skotomorphogenesis, which ensures the development of ARs from the hypocotyls embedded in soil.

We have reported that the initiation of LRs and that of HARs share common regulatory components, including the IAA14-ARF7/19 module, AUX1, LAX3, LBD16, LBD29, WOX5 and WOX7. However, LRs are developed programmatically from roots, whereas HARs emerge from hypocotyls in response to environmental stimuli. Such a difference suggests that there must be a point of divergence in the regulatory mechanism underlying HAR and LR formation. We found that PIFs do not affect LR development, although they do contribute to the regulation of auxin biosynthesis, auxin transport and activation of LBD16, LBD29, WOX5 and WOX7. The different requirements of PIFs in HAR and LR formation explain the divergence of HAR and LR development. The most likely reason for the different requirements of PIFs could be that PIF expression can be highly induced in hypocotyls but be barely induced in roots.

YUC2, YUC6, LBD16, LBD29, WOX5 and WOX7 have also been reported to be involved in AR formation in leaf explants (Chen et al., 2016; Hu and Xu, 2016; Sheng et al., 2017), and our results showed that PIFs are also required for this process. This evidence demonstrated that HARs and wound-induced ARs share similar regulatory mechanisms, which might be mediated by PIFs. However, the involvement of PIFs in HAR and wound-induced AR formation is different, given that pifq does not develop any HARs but that leaf explants still produce ARs, albeit to a lesser degree than the wild type. These differences might also be attributed to low expression of PIFs in leaf explants.

All of the A. thaliana lines used in this work are in the Col-0 background. The pif1-1, pif3-7, pif4-2, pif5-3, phyB-9, pifq and cop1-6 mutants were obtained from Dr Hongtao Liu (Chinese Academy of Sciences Center for Excellence in Molecular Plant Sciences, Shanghai, China). phyB pifq were kindly provided by P. Quail (University of California, Berkeley, USA), and yuc1 yuc4, yuc2 yuc6, aux1 (GABI_459H07), lax2 (GABI_345D11), lax3 (SAIL_855_F07), lax2 lax3 and WOX5_pro:GUS were obtained from Dr. Lin Xu (Chinese Academy of Sciences Center for Excellence in Molecular Plant Sciences, Shanghai, China) (Cheng et al., 2006). The aux1 mutant was crossed with lax3 to generate the aux1 lax3 double mutant. The DR5:GUS plant was crossed with pifq to generate the DR5:GUS/pifq reporter line. The WOX5_pro:GUS plant was crossed with pifq to generate the WOX5_pro:GUS/pifq reporter line. The full-length coding DNA sequences (CDSs) of PIF1, PIF3, PIF4 and PIF5 were cloned into an entry vector (pDONR/Zeo) and subsequently recombined into a destination vector pCAMBIA1300-gateway (Liang et al., 2018) to generate 35S:PIF1-Flag (PIF1-OX), 35S:PIF3-Flag (PIF3-OX), 35S:PIF4-Flag (PIF4-OX) and 35S:PIF5-Flag (PIF5-OX), respectively. For construction of the YUC2_pro:GUS, YUC6_pro:GUS, AUX1_pro:GUS and LAX3_pro:GUS transgenic plants in the wild-type and pifq mutant background, the YUC2 promoter (3 kb), the YUC6 promoter (3.3 kb), the AUX1 promoter (2.6 kb) upstream of the translation start site plus the first exon and intron (270 bp) (Marchant et al., 1999) and the LAX3 promoter (1.9 kb) plus a partial segment of the first exon (132 bp), were subcloned into pRR00 vectors, respectively. For construction of the LBD16_pro:LBD16-GUS transgenic plants and the LBD16_pro:LBD16-GFP transgenic plants in wild-type and pifq mutant backgrounds, the LBD16 promoter (4.8 kb) plus full-length gene sequence was subcloned into pCAMBIA1300 and pCAMBIA1301 vectors, respectively (Hu and Xu, 2016). AUX1_pro:AUX1-GFP and LAX3_pro: LAX3-GFP constructs were generated according to previous studies (Swarup et al., 2004, 2008). For construction of 35S:AUX1 transgenic plants in the pifq mutant background, the CDS of AUX1 was cloned and subsequently inserted into a pCambia1301 vector. Stock seeds of pif1 pif3 (N66045), pif1 pif4 (N68811), pif1 pif5 (N68094), pif3 pif4 (N66046), pif3 pif5 (N68812), pif1 pif3 pif4 (N66500), pif1 pif3 pif5 (N66047), pif1 pif4 pif5 (N68095), pif3 pif4 pif5 (N66048) and yuc5 yuc8 yuc9 (N69942) were purchased from the European Arabidopsis Stock Centre. T-DNA lines of yuc2 (N659779) and yuc6 (N663363) were obtained from AraShare (China).

To initiate the formation of ARs from hypocotyls, seeds were surface sterilized, stratified for 1 day at 4°C and then sown on plates containing half-strength Murashige and Skoog (1/2 MS) media, which were incubated in white light for 6 h and then transferred to red light (40 μmol·m⁻²·sec⁻¹) or darkness for the required duration as indicated. The numbers of HARs were determined using a dissecting microscope. Similarly, induction of the formation ARs from leaf explants was performed as previously described (Li et al., 2021).

Immunoblot assays were performed as previously described (Liu et al., 2008). For immunoblots, the hypocotyls of 5-day-old seedlings were harvested and ground in liquid nitrogen. The samples were subsequently solubilized with NEB buffer, after which the resulting protein extracts were centrifuged at 14,000 g for 10 min. The supernatants were collected, mixed with 5× SDS-PAGE loading buffer and then analyzed using SDS-PAGE and immunoblotting with anti-Flag (Sigma-Aldrich), anti-actin antibodies (Sangon) and anti-IAA14 antibodies (Li et al., 2021).

Trypan Blue staining was performed as previously described (Zhao et al., 2018). Briefly, seedlings were submerged in Trypan Blue staining solution (30 ml of ethanol, 10 g of phenol, 10 ml of H₂O, 10 ml of glycerol, 10 ml of lactic acid and 10 mg of Trypan Blue) and boiled for 2-3 min, after which they were allowed to cool to room temperature for 1 h. The stained seedlings were then decolorized in chloral hydrate solution. Trypan blue staining was observed using a Nikon Eclipse Ni light microscope.

Seedlings were submerged and maintained in GUS staining solution [2 mM K₄Fe(CN)₆, 2 mM K₃Fe(CN)₆, 0.5 mg ml⁻¹ X-Gluc, 10 mM EDTA, 0.1% Triton X-100, 50 mM Na₂HPO₄-NaH₂PO₄ (pH 7.0)] at 37°C for 5-10 h. The stained seedlings were then decolorized in 70% (v/v) ethanol and incubated in chloral hydrate solution (200 g of chloral hydrate, 20 g of glycerol and 50 ml of water) at room temperature for ∼12 h to induce tissue transparency (Liu et al., 2014). GUS staining was observed using a Nikon Eclipse Ni light microscope.

Confocal microscopy was performed using a Leica TCS SP8 confocal laser scanning microscope. Briefly hypocotyls of the seeds grown on 1/2 MS medium in the dark for 6 days were analyzed with an excitation wavelength of 488 nm and an emission wavelength of 505-550 nm. For propidium iodide (PI) staining, seedlings were incubated in a fresh solution with 15 mM (10 mg/ml) PI dissolved in water in the dark for 2 h, followed by three washes with water. The fluorescence excitation and emission wavelengths of PI were 561 nm and 610-630 nm, respectively.

For gene expression analysis, total RNA was extracted from the hypocotyls of seedlings grown in the dark. RNA isolation and qRT-PCR analysis were performed according to previously described methods (Gao et al., 2017) in conjunction with gene-specific primers (Table S1). The UBC21 gene was used as an internal control, and at least three biological replicates were tested.

Y1H assays were performed using a Matchmaker Gold Yeast One-Hybrid System Kit (Clontech) according to the manufacturer's protocol. Fragments of AUX1, LBD16, LBD29, WOX5 and WOX7 promoters containing PBE-box, G-box, E-box or mutant E-box motifs were synthesized, annealed and ligated into pAbAi vectors digested with XhoI to generate pAbAi-pAUX1-P3, pAbAi-pAUX1-mP3, pAbAi-pLBD16-P1, pAbAi-pLBD29-P2, pAbAi-pWOX5-P2 and pAbAi-pWOX7-P3. Similarly, the CDSs of PIF1, PIF3, PIF4 and PIF5 were subcloned and ligated into pGADT7 vectors. The oligonucleotide sequences of the promoters and the primers used to clone the CDSs of PIF1, PIF3, PIF4 and PIF5 are listed in Table S1. The pAbAi vectors were linearized and transformed into Y1H Gold yeast strains. Transformants were selected on SD/-Ura dropout plates, and the minimal concentration of AbA that completely suppressed strain growth was identified. The pGADT7 constructs were subsequently transformed into Y1H Gold strain cells harboring different pAbAi vectors, after which the transformants were screened on SD/-Leu plants with AbA.

The CDS of IAA14 and cDNA fragments without the activation domain of ARF19 (ARF19-AD, of which amino acids 550 to 724 were deleted) were subcloned and inserted into pGBKT7 vectors, and the CDSs of PIF1, PIF3, PIF4 and PIF5 were subcloned and inserted into pGADT7 vectors. For interaction analysis, two combinatorial constructs were co-transformed into the yeast strain AH109 via the lithium acetate transformation procedure, as described in the Yeast Protocols Handbook (Clontech). Empty pGADT7 or pGBKT7 vectors were used as negative controls. The transformants were grown on SD/-Leu-Trp-His-Ade dropout plates.

For transient transcription dual-LUC assays of A. thaliana protoplasts, mesophyll protoplasts were isolated from 4-week-old Col-0 plants grown under long-day conditions (16 h of light/8 h of darkness). The promoters of YUC2, YUC6, AUX1 and LAX3 were cloned and inserted into pGreenII 0800-LUC vectors as reporters, and the CDSs of PIF1, PIF3, PIF4 and PIF5 were cloned and inserted into pA7 vectors as effectors. Protoplast isolation and polyethylene glycol (PEG) transformation were carried out as described previously (Yoo et al., 2007). Approximately 2×10⁴ protoplasts were transfected with 10 µg of reporter DNA and 10 µg of effector DNA and then incubated in red light for 16 h. Then, these protoplasts were harvested by centrifugation for dual-LUC assays with a Dual-Luciferase Reporter system (Promega) according to the manufacturer's instructions. The REN activity for each reaction was used as an internal control.

To generate recombinant His-PIF1 proteins, the CDS of PIF1 was inserted into the pETDuet-M2 vector and expressed in the Escherichia coli Rosetta (DE3) strain. For probes, biotin-labeled complementary oligonucleotides of the YUC2, YUC6, AUX1, LAX3, LBD16, LBD29, WOX5 and WOX7 promoter fragments containing PBE boxes, G boxes or E boxes were synthesized and annealed. EMSAs were performed according to the manufacturer's instructions of the LightShift Chemiluminescent EMSA Kit (Thermo Fisher Scientific). The migration of the biotin-labeled probe was determined using a Chemiluminescent Nucleic Acid Detection Module (Thermo Fisher Scientific) and a Tanon-5200 Chemiluminescent Imaging System (Tanon Science and Technology). The oligonucleotide sequences are listed in Table S1.

ChIP experiments were performed as described previously (Liang et al., 2018) by using transgenic plants that expressed PIF1-Flag or PIF4-Flag, and that were grown under red light for 2 days and then transferred to darkness for 6 days. The seedlings were crosslinked with 1% formaldehyde (Sigma-Aldrich) under vacuum, which was stopped with glycine. The seedlings were then[ground and homogenized in extraction buffer 1 (0.4 M sucrose, 10 mM Tris-HCl (pH 8.0), 5 mM β-mercaptoethanol, 0.1 mM phenylmethylsulfonyl fluoride (PMSF) and 1X cOmplete Protease Inhibitor Cocktail tablets (Roche)]. Nuclei were precipitated by centrifugation at 2000 g for 20 min, washed with extraction buffer 2 [0.25 M sucrose, 10 mM Tris-HCl (pH 8.0), 10 mM MgCl₂, 1% Triton X-100, 5 mM β-mercaptoethanol, 0.1 mM PMSF and 1X cOmplete Protease Inhibitor Cocktail tablets (Roche)] and then lysed in nuclear lysis buffer [50 mM Tris-HCl (pH 8.0), 10 mM EDTA, 1% SDS, 0.1 mM PMSF and 1X cOmplete Protease Inhibitor Cocktail tablets (Roche)]. The chromatin was sheared by sonication, after which the chromatin solution was diluted 10-fold with ChIP dilution buffer [1.1% Triton X-100, 16.7 mM Tris-HCl (pH 8.0), 1.2 mM EDTA, 167 mM NaCl, 0.1 mM PMSF and 1× cOmplete Protease Inhibitor Cocktail tablets (Roche)]. Anti-Flag Sepharose and the chromatin solution were mixed together and then incubated at 4°C overnight. The beads were washed sequentially with a low-salt buffer [150 mM NaCl, 0.1% SDS, 1% Triton X-100, 2 mM EDTA and 20 mM Tris-HCl (pH 8.0)], a high-salt buffer [500 mM NaCl, 0.1% SDS, 1% Triton X-100, 2 mM EDTA and 20 mM Tris-HCl (pH 8.0)], LiCl washing buffer [0.25 M LiCl, 1% NP40, 1% sodium deoxycholate, 1 mM EDTA and 10 mM Tris-HCl (pH 8.0)] and TE buffer [10 mM Tris-HCl (pH 8.0) and 1 mM EDTA]. The resulting immunocomplexes were eluted with elution buffer (1% SDS and 0.1 M NaHCO₃) and the crosslinks reversed overnight at 65°C. The mixture was subsequently treated with Proteinase K, after which the precipitated DNA was recovered and subjected to qPCR. The primers used in the ChIP-qPCR experiments are listed in Table S1.

We thank H. T. Liu, L. Xu and P. Quail for the Arabidopsis seeds. We thank Y. R. Wu for providing the pAbAi vector. We thank J. W. Wang for providing the pRR00 vector.

The peer review history is available online at https://journals.biologists.com/dev/article-lookup/doi/10.1242/dev.200362