The miR396-GRFs Module Mediates the Prevention of Photo-oxidative Damage by Brassinosteroids during Seedling De-Etiolation in Arabidopsis

BRs play important roles in the transition from skotomorphogenesis to photomorphogenesis by optimizing the biosynthesis of chlorophyll through the BZR1-PIF4-GRF7 transcriptional complex.

Abstract

The switch from dark- to light-mediated development is critical for the survival and growth of seedlings, but the underlying regulatory mechanisms are incomplete. Here, we show that the steroids phytohormone brassinosteroids play crucial roles during this developmental transition by regulating chlorophyll biosynthesis to promote greening of etiolated seedlings upon light exposure. Etiolated seedlings of the brassinosteroids-deficient det2-1 (de-etiolated2) mutant accumulated excess protochlorophyllide, resulting in photo-oxidative damage upon exposure to light. Conversely, the gain-of-function mutant bzr1-1D (brassinazole-resistant 1-1D) suppressed the protochlorophyllide accumulation of det2-1, thereby promoting greening of etiolated seedlings. Genetic analysis indicated that phytochrome-interacting factors (PIFs) were required for BZR1-mediated seedling greening. Furthermore, we reveal that GROWTH REGULATING FACTOR 7 (GRF7) and GRF8 are induced by BZR1 and PIF4 to repress chlorophyll biosynthesis and promote seedling greening. Suppression of GRFs function by overexpressing microRNA396a caused an accumulation of protochlorophyllide in the dark and severe photobleaching upon light exposure. Additionally, BZR1, PIF4, and GRF7 interact with each other and precisely regulate the expression of chlorophyll biosynthetic genes. Our findings reveal an essential role for BRs in promoting seedling development and survival during the initial emergence of seedlings from subterranean darkness into sunlight.

INTRODUCTION

Seedling greening is essential for their survival as they emerge through the soil and into the light. After germination beneath the soil surface, seedlings follow a programmed developmental pattern termed skotomorphogenesis that is characterized by fast growth of the hypocotyl, formation of a closed apical hook, and yellow cotyledons (Von Arnim and Deng, 1996; Chen et al., 2004). Once seedlings break through the soil and reach the light, seedlings switch to light-mediated photomorphogenic development, including cotyledons opening and greening, which allows seedlings to become photosynthetically active and capable of autotrophic growth (Von Arnim and Deng, 1996; Chen et al., 2004). Effective chlorophyll biosynthesis during this initial light exposure is critical for seedling greening, as excess accumulation of protochlorophyllide, the phototoxic chlorophyll precursor, can cause photo-oxidative damage (Reinbothe et al., 2010; Liu et al., 2017a). Thus, plants have evolved sophisticated mechanisms to regulate plastid development during the transition from the dark (below ground) to the light (aboveground).

To date, multiple phytohormones and environmental signaling pathways have been reported to act in the regulation of cotyledon greening (Liu et al., 2017a). Light is the main environmental factor that affects cotyledon greening in etiolated seedlings (Huq et al., 2004). The perception of light signals by the phytochrome family of photoreceptors triggers the degradation of a subfamily of basic Helix Loop Helix transcription factors, PHYTOCHROME-INTERACTING FACTORs (PIFs), which play critical roles in the control of chlorophyll biosynthesis in the dark (Huq et al., 2004; Moon et al., 2008; Shin et al., 2009; Stephenson et al., 2009). Loss of function in multiple PIFs in Arabidopsis (Arabidopsis thaliana), including PIF1, PIF3, PIF4, and PIF5, cause excess accumulation of free protochlorophyllide in etiolated (or dark-grown) seedlings; consequently, they show lethal photobleaching upon exposure to light (Huq et al., 2004; Moon et al., 2008; Shin et al., 2009; Stephenson et al., 2009; Chen et al., 2013). Molecular and genetic analysis revealed that the phytohormone ethylene promotes greening of etiolated seedlings upon light irradiation by activating the transcription factors ETHYLENE INSENSITIVE 3 (EIN3) and EIN3-LIKE 1 (EIL1) to cooperate with PIFs to regulate chlorophyll biosynthesis (Liu et al., 2017b; Zhong et al., 2009). PIFs and EIN3/EIL1 can indeed directly induce the expression of the two genes encoding the key NADPH: protochlorophyllide oxidoreductases (PORs) PORA and PORB to prevent photo-oxidative damage and promote cotyledons greening (Cheminant et al., 2011).

Brassinosteroids (BRs) are a group of plant steroid hormones that are essential for plant growth and development (Clouse and Sasse, 1998). BRs are perceived by the plasma membrane-localized receptor kinase BRASSINOSTEROID INSENSITIVE1 (BRI1) and its coreceptor BRI1-ASSOCIATED KINASE1 (BAK1) or its homologues SOMATIC EMBRYOGENESIS RECEPTOR KINASEs (SERKs; Li et al., 2002; Nam and Li, 2002). Activated BRI1 sequentially phosphorylates and activates downstream BR signaling kinases, such as BRASSINOSTEROID-SIGNALING KINASE1 (BSK1) and CONSTITUTIVE DIFFERENTIAL GROWTH1 (CDG1), as well as the Ser/Thr phosphatase BRI1-SUPPRESSOR1 (BSU1) that dephosphorylates the negative kinase BRASSINOSTEROID INSENSITIVE2 (BIN2) and leads to BIN2 degradation by the F-box protein KINKSUPPRESSED IN bzr1-1D (KIB1; Tang et al., 2008; Kim et al., 2009; Kim et al., 2011; Zhu et al., 2017). The inactivation of BIN2 allows the accumulation of unphosphorylated transcription factors BRASSINAZOLE RESISTANT 1 (BZR1) and BRI1-EMS-SUPPRESSOR 1 (BES1) in the nucleus, where they bind to the promoters of thousands of target genes and regulate their expression (He et al., 2002; Wang et al., 2002; Yin et al., 2002; Sun et al., 2010). BZR1, PIF4, and AUXIN RESPONSE FACTOR 6 (ARF6) are genetically interdependent for the promotion of cell elongation, and they directly interact with each other and coregulate many genes (Oh et al., 2012; Oh et al., 2014). Hydrogen peroxide (H₂O₂) induces the oxidation of BZR1 and increases its transcriptional activity by enhancing the interaction of BZR1 with PIF4 and ARF6 (Tian et al., 2018). DELLAs, the negative regulators of gibberellin signaling pathway, also interact with BZR1, PIFs, and ARF6, thereby inhibiting their DNA binding abilities to repress gene expression and cell elongation (de Lucas et al., 2008; Feng, 2008; Bai et al., 2012; Oh et al., 2014; Chaiwanon et al., 2016). Thus, the BZR-ARF-PIF/DELLA module integrates diverse environmental and phytohormone signals to regulate cell elongation (Bai et al., 2012; Chaiwanon et al., 2016).

GROWTH-REGULATING FACTORs (GRFs) are plant-specific transcription factors that carry out important functions in the regulation of plant growth and development (Omidbakhshfard et al., 2015). Overexpression of GRF1 and GRF2 in Arabidopsis resulted in larger leaves and increased seed weight and seed size (Kim et al., 2003). Similarly, overexpression of the rapeseed (Brassica napus) gene GRF2 (BnGRF2) in Arabidopsis increased seed oil production by upregulating the expression of chlorophyll biosynthetic genes to enhance photosynthetic efficiency (Liu et al., 2012). A gain-of-function mutation in the rice (Oryza sativa) GRF4 gene (OsGRF4) showed significantly longer panicles, larger seeds, increased seed weight and grain yield (Che et al., 2015; Duan et al., 2015). Members of the GRF family are regulated posttranscriptionally by the microRNA miR396 (Curaba et al., 2014; Spanudakis and Jackson, 2014). In Arabidopsis, miR396 is encoded by MIR396a and MIR396b and induces the cleavage of seven GRF mRNAs, with the exception of GRF5 and GRF6 (Liang et al., 2014). The miR396-GRF module is conserved among angiosperms and gymnosperms, and ensures the accurate expression of GRFs for precise regulation of plant growth and development (Omidbakhshfard et al., 2015).

In this study, we show that BRs tightly controls chlorophyll biosynthesis to achieve a delicate balance between absorption of light energy by chlorophyll and the capacity of chlorophyll to channel that energy into productive photochemical reactions, thus preventing lethal photo-oxidation. Further, we discovered that BZR1, PIF4, and GRF7 interact with one another and cooperatively regulate the expression of HEMA1, HEMB1, CHLD, CHLH, and GENOMES UNCOUPLED4 (GUN4), which encode key enzymes in chlorophyll biosynthesis. Our findings illustrate the role of the BZR1-PIF4-GRF7 transcriptional module in mediating the crosstalk between BR- and light-signaling pathways to attain a precise regulation of chlorophyll biosynthesis, thereby enhancing seedling survival during de-etiolation in Arabidopsis.

RESULTS

BRs and Activated-BZR1 Promote Greening of Etiolated Seedlings

The steroid hormone brassinosteroids play key roles in seedling skotomorphogenesis, as the mutants affecting BRs biosysnthesis or signal transduction display constitutive photomorphogenesis in darkness (Li et al., 1996; Li and Chory, 1997). However, the role of BRs during the transition from skotomorphogenesis to photomorphogenesis remains unclear. To this end, we examined the cotyledon-greening phenotype of various BR-related mutants grown in the dark for 4–8 d before exposure to white light for 2 d. The percentage of green cotyledons in wild-type Columbia-0 (Col-0) seedlings gradually declined with increasing days spent in the dark before transfer to white light, and reached ∼37% after 8 d in the dark (Figures 1A to 1C). However, we observed that genetic materials with a stronger BR signal, including plants overexpressing the BR receptor (BRI1-Ox) or the cytochrome P450 DWARF4 involved in BR biosynthesis (DWF4-Ox), displayed a higher greening rate than wild-type seedlings when grown in the dark for 6-8 d (Figure 1C), whereas the greening rates of the BR-deficient mutants det2-1 (de-etiolated2), rot3-2 (rotundifolia3), dwf4, and the BR-insensitive mutants bri1 and bin2-1 were much lower than those of wild-type seedlings after 4 d or more in the dark (Figures 1B and 1C; Supplemental Figures 1A to 1D). Treatment with brassinolide (BL), the most active BRs in plants, increased the greening rates of wild type and det2-1, but had weak effects on bri1-301, indicating that BRs signaling is required for light-induced cotyledon greening (Figure 1D). Not surprisingly, the reduced greening rates of det2-1 and bri1-116 were suppressed by the dominant gain-of-function mutant bzr1-1D (Figures 1A and 1B; Supplemental Figure 1A), which stabilizes the BZR1 protein due to an enhanced interaction with the type 2A Ser/Thr protein phosphatase PP2A. In addition, loss of function mutations in BZR1 and BES1 significantly reduced cotyledon greening rates (Figure 1E). These results suggested that BRs promote cotyledon greening in etiolated seedlings by activating BZR1 and related proteins.

Figure 1.

BRs and Activated-BZR1 Increase the Greening Rate of Dark-Grown Seedlings.

(A) Representative images of 6-d-old etiolated seedlings of wild-type Col-0, det2-1, bzr1-1D, and bzr1-1D det2-1 after 2 d in white light.

(B) to (E) Greening rates of various etiolated seedlings grown in darkness for the indicated number of days before exposure to white light for 2 d on half-strength MS medium with or without 100 nM BL. Data are shown as means ± sd (n = 3). Different letters above bars indicate statistically significant differences between samples (two-way ANOVA followed by post hoc Tukey test, P < 0.05).

(F) to (H) Relative fluorescence of protochlorophyllide of various etiolated seedlings grown on half-strength MS medium for 6 d in darkness.

Overaccumulation of the chlorophyll precursor protochlorophyllide (Pchlide) in the dark results in photobleaching of etiolated seedlings after transfer into light (Chen et al., 2013). To test whether the BR-increased greening rate might be attributed to the reduction in the amounts of Pchlide in seedlings, we analyzed Pchlide levels in various BR mutants. Our results showed that the BR-deficient mutant det2-1, the BR-insensitive mutants bri1-301 and bzr1 bes1 double mutant all accumulated much higher levels of Pchlide when compared to wild-type seedlings (Figures 1F to 1H). By contrast, Pchlide content decreased in seedlings with higher BR signaling such as BRI1-Ox, DWF4-Ox, and bzr1-1D (Figures 1F and 1G). The large amounts of Pchlide accumulating in det2-1 mutant were suppressed by bzr1-1D (Figure 1F). These results indicate that BRs and BZR1 promote the greening of etiolated seedlings by regulating Pchlide biosynthesis. Excess Pchlide induces the production of reactive oxygen species (ROS) upon light exposure, causing photobleaching and cell death in cotyledons (Chen et al., 2013). To examine whether ROS are involved in BR-regulated de-etiolation, we analyzed ROS contents in different BR mutants upon light exposure. Light exposure resulted in strong fluorescent signals from the ROS indicator 2',7'-dichlorodihydrofluorescein diacetate (H₂DCFDA) as well as prominent increase in trypan blue staining indicative of cell death in det2-1, weak signals in wild-type seedlings, and almost no signal in bzr1-1D or bzr1-1D det2-1 (Supplemental Figures 2A and 2B). Together, these results demonstrate that BRs and activated BZR1 reduce the accumulation of Pchlide and ROS to prevent photobleaching and promote seedling greening during de-etiolation.

PIFs Play Critical Roles in BZR1-induced Greening of Etiolated Seedlings.

Previous studies have shown that PIFs play essential roles in the greening of etiolated seedling upon light exposure (Huq et al., 2004; Moon et al., 2008; Stephenson et al., 2009). Since BZR1 and PIFs regulate cell elongation interdependently (Oh et al., 2012), we therefore examined whether PIFs are invovled in BZR1-promoted cotyledon greening. Consistent with previous results, loss of PIF1, PIF3, PIF4, and PIF5 in the pifq quadruple mutant caused a lower greening rate compared with the wild type when grown in the dark for 4 d, and none of the etiolated seedlings turned green when grown in the dark for more than 6 d (Figures 2A and 2B). The bzr1-1D pifq mutant showed a similar greening rate as pifq (Figures 2A and 2B). BR treatment partially restored cotyledon greening in pifq (Figure 2C). Furthermore, dark-grown pifq seedlings overaccumulated Pchlide. The content of Pchlide decreased in bzr1-1D pifq relative to pifq, but was still much higher than that of wild-type seedlings (Figure 2D). These results indicate that PIFs play dominant functions in BR- and BZR1-promoted greening of etiolated seedlings.

Figure 2.

PIFs Play Critical Roles in BZR1-induced Greening of Etiolated Seedlings.

(A) Representative images of 6-d-old etiolated seedlings of wild-type Col-0, pifq, bzr1-1D, and bzr1-1D pifq after 2 d in white light.

(B) and (C) Greening rates of various etiolated seedlings grown in darkness for the indicated number of days before exposure to white light for 2 d on half-strength MS medium with or without 100 nM BL. Data are shown as means ± sd (n = 3). Different letters above bars indicate statistically significant differences between samples (two-way ANOVA followed by post hoc Tukey test, P < 0.05).

(D) Relative fluorescence of protochlorophyllide of wild-type Col-0, pifq, bzr1-1D, and bzr1-1D pifq seedlings grown on half-strength MS medium for 6 d in darkness.

(E) Venn diagram showing the overlap between sets of genes differentially expressed in dark-grown pifq versus wild type, and bzr1-1D versus bzr1-1D pifq, and regulated by white light in wild type.

(F) GO analysis of BZR1, PIFs, and light coregulated genes. Numbers indicate the percentage of genes belonging to each GO category. “Genome” means the random genomic control.

To determine whether PIFs can regulate the BZR1 function to promote the greening process, we analyzed the expression of BZR1 family members in the pifq mutant. Loss of PIF function in pifq resulted in decreased transcript levels for BZR1, BEH4 (BES1/BZR1 HOMOLOG4), and the BRs biosynthesis genes DWF4 and BR6OX2 (BRASSINOSTEROID-6-OXIDASE 2), but had no effects on the expression of BES1 or other related genes (Supplemental Figure 3A). Quantitative chromatin immunoprecipitation PCR (ChIP-qPCR) analysis revealed that PIF4 physically bound to the promoters of all BZR1 family members, as well as to the promoters of DWF4 and BR6OX2 (Supplemental Figure 3B). These results indicate that PIFs increase BZR1 activity not only through direct induction of its transcription rate, but also by increasing the expression of BR biosynthesis genes to increase BR levels in seedlings and promote the dephosphorylation of BZR1.

To determine how BZR1 and PIFs regulate greening in etiolated seedlings, we identified differentially expressed genes in the wild type Col-0, pifq, and bzr1-1D pifq by reanalyzing published deep-sequencing transcriptome (RNA-seq) data sets (Oh et al., 2012). Of the 3345 genes differentially expressed in pifq relative to the wild type, 2178 genes (65.1%) were regulated by pifq in the bzr1-1D background. Among these 2178 common genes, 2076 genes (95.3%) were affected by pifq in the same direction in both wild type and bzr1-1D, consistent with the similar reduced greening rates observed in pifq and bzr1-1D pifq. Among the 2178 common genes, 1052 genes were also regulated by light, of which 934 genes (87.1%) were regulated by light and PIFs in opposite directions (Figure 2E). Gene ontology analysis indicated that chloroplast development and photosynthetic genes were markedly enriched in the genes coregulated by BZR1, PIFs, and light (Figure 2F). These data suggest that BZR1 and PIFs play important roles in the regulation of chloroplast development.

GRF7 and GRF8 are Induced by BZR1 and PIFs

To investigate the molecular mechanisms by which BZR1 and PIFs enhance cotyledon greening upon illumination, we analyzed the genes coregulated by BZR1 and PIFs. GROWTH-REGULATING FACTOR (GRF) genes encode a class of plant-specific transcription regulators that have been reported to regulate leaf development, floral organ formation, seed development, and nitrogen assimilation (Omidbakhshfard et al., 2015; Li et al., 2018). Here, we found that BRs induced, whereas light repressed the expression of multiple GRF family members, including GRF2, GRF7, and GRF8 (Figures 3A and 3B). Consistent with this observation, transcript levels of GRF2, GRF7, and GRF8 were significantly lower in pifq, det2-1,and bri1-5 mutants when compared to the wild type (Figures 3C and 3D; Supplemental Figures 4A). The gain-of-function mutant bzr1-1D partially recovered the decreased expression levels of GRF7 and GRF8 when combined with the pifq or det2-1 mutants, but not in the case of GRF2 (Figures 3C and 3D; Supplemental Figures 4B to 4H). To test whether BZR1 and PIFs cooperatively regulate the expression of GRF7 and GRF8, we generated promoter-luciferase (LUC) reporter constructs with the GRF7 and GRF8 promoters for transient gene expression analysis in Arabidopsis mesophyll protoplasts. We observed that the luciferase activity derived from the GRF7pro:LUC and GRF8pro:LUC constructs increased when BZR1 or PIF4 were transfected alone, and was further induced when BZR1 and PIF4 were coexpressed, suggesting that BZR1 and PIFs cooperatively activated GRF7 and GRF8 transcription (Figures 3E and 3F). Chromatin immunoprecipitation analysis showed that BZR1 and PIF4 both bound to fragment D of the GRF7 promoter, and to fragments F and G of the GRF8 promoter (Figures 3G to 3L). Furthermore, BL and light had no effects on the protein levels of rGRF7, which is a miR396-resistant form of GRF7 (Supplemental Figures 5A and 5B). Together, these results indicate that BRs and light regulate the levels of GRF7 and GRF8 mainly at the transcriptional level through BZR1 and PIFs.

Figure 3.

BZR1 and PIFs Directly Induce the Expression of GRF7 and GRF8.

(A) and (B) RT-qPCR analysis of the effects of light and BRs on the expression of members of the GRF family. Col-0 seedlings were grown on half-strength MS medium in constant light for 5 d before treatment with mock solution (80% ethanol) or 100 nM BL for 3 h (A) or for 4 d in the dark before exposure to white light for 3 h (B). The PP2A gene was used as an internal control.

(C) and (D) RT-qPCR analysis of the expression of GRF7 and GRF8 in wild type and indicated mutants. Seedlings for the wild-type Col-0 and various mutants were grown on half-strength MS medium for 4 d in the dark. PP2A gene was used as an internal control.

(E) and (F) BZR1 and PIF4 synergistically promote the expression of GRF7 (E) and GRF8 (F). Arabidopsis protoplasts were transformed with dual luciferase reporter constructs containing GRF7pro:LUC or GRF8pro:LUC and 35Spro:REN, and constructs overexpressing the indicated effectors. LUC (firefly luciferase) activity was normalized to REN (Renilla luciferase).

(G) to (L) BZR1 and PIF4 directly bind to the promoter regions of GRF7 and GRF8. The 35Spro:BZR1-YFP, 35Spro:PIF4-YFP, and 35Spro:YFP were used to perform ChIP assays. Schematic illustration of different regions in the promoters of GRF7 (G) and GRF8 (J). The levels of BZR1 (H and K) and PIF4 (I and L) binding were calculated as the ratio between 35Spro:BZR1-YFP and 35Spro:YFP control or 35Spro:PIF4-YFP and 35Spro:YFP control, respectively, and then normalized to that of the control gene PP2A.

All data are shown as means ± sd (n = 3). *P < 0.05, as determined by a Student`s t-test. Different letters above bars indicate statistically significant differences between samples (one-way ANOVA followed by post hoc Tukey test, P < 0.05).

GRFs Promote the Greening of Etiolated Seedlings

In Arabidopsis, several members of the GRF family are regulated posttranscriptionally by miR396. To determine whether GRFs are functionally involved in seedling greening during de-etiolation, we generated transgenic lines that overexpressed either miR396a (35Spro:miR396a) or a target mimic of miR396a (35Spro:MIM396), in which miR396a is inactivated. RT-qPCR analysis demonstrated that transcript levels of several members of the GRF family, except GRF5, were significantly reduced in 35Spro:miR396a and increased in 35Spro:MIM396, compared to wild-type seedlings (Supplemental Figure 6A). In light-grown seedlings, overexpression of miR396a led to slightly shorter hypocotyls and roots when compared to wild-type seedlings, in accordance with previous reports (Supplemental Figures 6B to 6D). Surprisingly, we found that 35Spro:miR396a showed a constitutive photomorphogenic phenotype in the dark, including short hypocotyls, an absence of an apical hook, and two open cotyledons. By contrast, 35Spro:MIM396 exhibited an increased hook angle and longer hypocotyls than wild-type seedlings (Figures 4A to 4C), indicating that miR396 plays an important role in photomorphogenesis.

Figure 4.

miR396 and GRFs Play Important Roles During Greening of Etiolated Seedlings.

(A) Representative images of 6-d-old etiolated seedlings of the wild-type Col-0, the grf3 grf4 grf7 triple mutant (grf-t), the grf1 grf4 grf 7 grf8 quadruple mutant (grf-q), 35Spro:miR396a, 35Spro:rGRF7, and 35Spro:MIM396. Scale bar = 0.5 cm.

(B) and (C) GRFs promote seedling de-etiolation. Seedlings for the wild-type Col-0 and the indicated mutants were grown on half-strength MS medium with 1% Suc for 6 d in the dark. Hypocotyl lengths (B) and apical hook curvature (C) were measured from at least 20 seedlings.

(D) and (F) Greening rates of various etiolated seedlings grown in the dark for the indicated number of days before exposure to white light for 2 d.

For (B) to (D) and (F), data are shown as means ± sd (n = 3). Different letters above bars indicate statistically significant differences between samples (two-way ANOVA followed by post hoc Tukey test, P < 0.05).

(E) and (G) Relative fluorescence of protochlorophyllide of various etiolated seedlings grown on half-strength MS medium for 6 d in the dark.

Cotyledon greening assays showed that overexpression of miR396a significantly reduced greening rate. When first grown in the dark for 4 d, the cotyledons of etiolated wild-type and 35Spro:MIM396 seedlings all turned green upon light exposure, whereas only 2% of 35Spro:miR396a seedlings turned green. When the dark growth period was extended to 6 d, 76% of 35Spro:MIM396 seedlings turned green, and 57% of wild-type seedlings turned green, but 35Spro:miR396a had completely lost the ability to green (Figure 4D). In agreement with these results, Pchlide content in 35Spro:MIM396 was slightly lower than that of wild-type seedlings, while 35Spro:miR396a accumulated nearly ten times more Pchlide than wild-type seedlings (Figure 4E). These results indicate that miR396 induces the accumulation of Pchlide and reduces the greening of etiolated seedlings upon exposure to light.

To further verify the roles of GRFs in cotyledon greening and seedling survival, we used single T-DNA insertion grf mutants and higher-order grf mutants to determine the greening rate of their cotyledons following light exposure. Single grf mutants displayed slightly lower greening rates relative to the wild type, but the grf7 grf8 double mutant reduced greening rate by ∼50% compared to the wild type after growth in the dark for 6 d. Higher-order mutants in other members of the GRF family, such as the grf3 grf4 grf7 triple mutant (grf-t) and the grf1 grf4 grf7 grf8 quadruple mutant (grf-q), exhibited significantly lower greening rates relative to the wild type (Figure 4F; Supplemental Figures 7A to 7D). Disruption of GRFs function also resulted in a de-etiolation phenotype in the dark (Figures 4A to 4C; Supplemental Figures 8A to 8C). However, 35Spro:rGRF7 transgenic seedlings, expressing a miR396-resistant form of GRF7 under the control of the constitutive Cauliflower Mosaic Virus 35S promoter, displayed a significantly higher greening rate compared to wild-type seedlings (Figure 4F). We further observed a significant accumulation of Pchlide in the grf-t and grf-q mutants, whereas Pchlide levels dropped in 35Spro:rGRF7 seedlings (Figure 4G). These results indicate that the miR396-GRFs module plays a critical role in photomorphogenesis and greening of etiolated seedlings upon illumination.

GRFs Play Important Roles for BR-mediated Greening of Etiolated Seedlings

Considering that GRF7 and GRF8 are upregulated by BRs through BZR1, we next wished to test whether miR396 and GRF7/GRF8 were involved in BR-regulated plant growth and development. Overexpression of miR396a resulted in the semi-dwarf plants, whereas overexpression of MIM396 led to plants with long petioles (Figure 5A). BR sensitivity tests using hypocotyl elongation assays indicated that 35Spro:MIM396 and 35Spro:rGRF7 seedlings had markedly enhanced BRs responses, whereas 35Spro:miR396a, which contained low levels of GRF proteins, had a significantly reduced BRs response, while the grf-q quadruple mutant showed only a slight reduction in its BRs response (Figure 5B; Supplemental Figure 9A). Sensitivity to propiconazole (PPZ, a specific inhibitor of BRs biosynthesis) showed that 35Spro:miR396a and grf-q were hypersensitive to PPZ, whereas 35Spro:MIM396 and 35Spro:rGRF7 were less sensitive to PPZ (Figure 5C; Supplemental Figure 9B). These results indicate that the miR396-GRF module is involved in BRs signaling to regulate photomorphogenesis.

Figure 5.

GRFs Play Important Roles for BR-mediated Cell Elongation and Greening of Etiolated Seedlings.

(A) Representative plants of the wild-type Col-0, 35Spro:miR396a, 35Spro:MIM396, det2-1, and 35Spro:MIM396 det2-1 grown in soil for 4 weeks. Scale bar = 1 cm.

(B) and (C) Overexpression of miR396a reduces the sensitivity to BRs and increases sensitivity to the BRs biosynthesis inhibitor propiconazole (PPZ). Wild-type Col-0, 35Spro:miR396a, and 35Spro:MIM396 seedlings were grown on half-strength MS medium containing different concentrations of BL or PPZ for 7 d in constant light (for BL treatments) or in the dark (PPZ treatments). Hypocotyl lengths were measured from at least 20 plants. Data are shown as means ± sd.

(D) and (E) Col-0, det2-1, 35Spro:miR396a, 35Spro:MIM396, and 35Spro:MIM396 det2-1 seedlings were grown on half-strength MS medium in the dark for 6 d. Hypocotyl lengths were measured from at least 20 plants. Data are shown as means ± sd. Scale bar = 1 cm.

(F), (J), and (K) Greening rates of wild-type Col-0 and indicated mutant seedlings grown in the dark for the indicated number of days (F), 4 d (J), or 6 d (K) before exposure to white light for 2 d. Data are shown as means ± sd (n = 3).

For (D) to (F) and (K), different letters above bars indicate statistically significant differences between samples (one-way ANOVA followed by post hoc Tukey test, P < 0.05).

(G) Relative fluorescence of protochlorophyllide of various etiolated seedlings grown on half-strength MS medium for 6 d in the dark.

(H) and (I) Inhibition of GRFs functions by overexpression of miR396a (H) or by loss of function in GRFs (I) has no significant effects on BZR1-promoted hypocotyl elongation. Seedlings of the wild type Col-0 and the indicated mutants were grown on half-strength MS medium containing different concentrations of PPZ in the dark for 7 d. Hypocotyl lengths were measured from at least 20 plants. Data are shown as means ± sd.

We next generated 35Spro:MIM396 det2-1 lines by crossing det2-1 with 35Spro:MIM396, and analyzed their growth and greening phenotypes. In this case, 35Spro:MIM396 in the wild-type background showed larger rosette leaves and longer hypocotyls than wild-type plants; conversely, 35Spro:MIM396 det2-1 displayed smaller rosette leaves and shorter hypocotyls, phenotypes that are similar to the dwarf phenotype of det2-1 (Figures 5A, 5D, and 5E). The impaired light-induced cotyledon greening phenotype of det2-1 was not affected by 35Spro:MIM396 in 35Spro:MIM396 det2-1, but the excessive accumulation of Pchlide in det2-1 was partially suppressed by overexpression of MIM396 (Figures 5F and 5G). These results indicate that GRFs are involved in BR-promoted cell elongation and chlorophyll biosynthesis in the dark.

To further elucidate the genetic relationship between BZR1 and GRFs, we crossed bzr1-1D with 35Spro:miR396a and the grf7 grf8 double mutant, and analyzed their growth and greening phenotypes. In agreement with previous results, bzr1-1D was resistant to the BRs biosynthesis inhibitor PPZ, but grf loss-of-function (either by overexpressing miR396 or the grf7 grf8 double mutant) had no significant effects on the PPZ resistance of bzr1-1D (Figure 5H and 5I). Overexpression of miR396a or mutations in GRF7 and its homologous genes both gave rise to a de-etiolated phenotype in the dark and a lower greening rate of etiolated seedlings upon illumination, whereas these phenotypes were partially suppressed by bzr1-1D (Figures 5J and 5K; Supplemental Figures 10A and 10B). These results suggest that miR396 and GRFs are not required for BZR1-promoted cell elongation or greening of etiolated seedlings.

PIFs Play Important Roles for GRF-promoted Cell Elongation and Greening of Etiolated Seedlings

Given the important roles of GRFs in light-regulated greening of seedlings during de-etiolation, we wished to examine if GRFs functioned in the light signaling pathway. We observed that 35Spro:rGRF7 seedlings displayed longer hypocotyls, whereas the 35Spro:miR396a and grf-q seedlings had shorter hypocotyls when compared to the wild type when grown in constant white, red, and blue light (Figures 6A to 6C). Fluence-rate response curves showed that 35Spro:rGRF7 seedlings had a reduced sensitivity to light, while 35Spro:miR396a and grf-q seedlings were more sensitive to light (Figures 6A to 6C). These results indicated that miR396 is a positive regulator in the light signaling pathway.

Figure 6.

PIFs Play Important Roles for GRF-promoted Cell Elongation and Greening of Etiolated Seedlings.

(A) to (C) GRFs-mediated hypocotyl elongation in response to different light conditions. Wild-type Col-0, 35Spro:miR396a, grf-q, and 35Spro:rGRF7 seedlings were grown under different light conditions for 7 d. Hypocotyl lengths were measured from at least 20 plants. Data are shown as means ± sd.

(D) Representative plants of the wild-type Col-0, pifq, 35Spro:MIM396, and 35Spro:MIM396 pifq grown in soil for 4 weeks. Scale bar = 1 cm.

(E), (G), and (H) Greening rates of wild type and the indicated mutant seedlings grown in the dark for the indicated number of days (E), 6 d (G), or 4 d (H) before exposure to white light for 2 d.

(F) Relative fluorescence of protochlorophyllide of etiolated seedlings of wild type, pifq, 35Spro:MIM396, and 35Spro:MIM396/pifq that were grown on half-strength MS medium for 6 d in the dark.

(I) Overexpression of PIF4 recovered the hypocotyl length of 35Spro:miR396a transgenic plants. Seedlings of the wild type Col-0, 35Spro:miR396a, 35Spro:PIF4, and 35Spro:PIF4 35Spro:miR396a were grown on half-strength MS medium in the dark for 4 d. Hypocotyl lengths were measured from at least 20 seedlings.

For (E) and (G) to (I), data are shown as means ± sd (n = 3). Different letters above bars indicate statistically significant differences between samples (one-way ANOVA followed by post hoc Tukey test, P < 0.05).

To test the genetic relationship between miR396 and the master photomorphogenesis regulators PIFs, we crossed 35Spro:MIM396 into the pifq background, and crossed 35Spro:miR396a and grf7 grf8 with 35Spro:PIF4 transgenic lines. Overexpression of MIM396 had no significant effects on hypocotyl length or rosette leaves in the pifq mutant background, nor did it have any strong effects on the greening rate or Pchlide content of etiolated 35Spro:MIM396 pifq seedlings following light exposure (Figures 6D to 6F; Supplemental Figures 11A and 11B). By contrast, the lower greening rates of 35Spro:miR396a and grf7 grf8 seedlings were suppressed by the overexpression of PIF4 (Figures 6G and 6H). Furthermore, 35Spro:PIF4 also suppressed the de-etiolated phenotype and shorter hypocotyls of 35Spro:miR396a seedlings (Figure 6I; Supplemental Figures 11C and 11D). These results indicated that PIFs are required for GRFs function in promoting cell elongation and greening of etiolated seedlings.

GRF7 Interacts with BZR1 and PIF4 in Vivo and in Vitro

Given that BZR1 and PIFs are involved in GRF-promoted cell elongation, and that BZR1 interacts with PIFs to regulate gene expression, we hypothesized that GRFs might interact with BZR1 and PIFs. Indeed, transient bimolecular fluorescence complementation assays showed strong YFP fluorescence in the nucleus of Nicotiana benthamiana leaf cells coexpressing BZR1-nYFP and GRF7-cYFP, or PIF4-nYFP and GRF7-cYFP (Figure 7A). In vitro pull-down assays showed that glutathione S-transferase (GST) fusion protein GST-GRF7 interacted with maltose binding protein (MBP) fusion protein MBP-BZR1, MBP-PIF4, but not with MBP (Figure 7B). Coimmunoprecipitation assays with transgenic plants expressing 35Spro:rGRF7-Myc only or coexpressing 35Spro:BZR1-YFP and 35Spro:rGRF7-Myc, 35Spro:PIF4-YFP and 35Spro:rGRF7-Myc, showed that GRF7 interacted with BZR1 and PIF4 in plants (Figure 7C). These results indicated that GFR7 physically interacts with BZR1 and PIF4 in vivo and in vitro.

Figure 7.

GRF7 Interacts with BZR1 and PIF4 In Vivo and In Vitro.

(A) Confocal images of ratiometric bimolecular fluorescence complementation assays show that GRF7 interacts with BZR1 and PIF4 in Nicotiana benthamiana leaf cells. Scale bar = 20 μm.

(B) GRF7 directly interacts with BZR1 and PIF4 in vitro. Glutathione agarose beads loaded with GST-GRF7 were incubated with equal amounts of MBP, MBP-BZR1, or MBP-PIF4. Proteins bound to GST-GRF7 were detected by immunoblot analysis with an anti-MBP antibody.

(C) GRF7 interacts with BZR1 and PIF4 in vivo. Immunoprecipitation was performed using YFP-trap beads and transgenic Arabidopsis seedlings expressing 35Spro:rGRF7-Myc only or coexpressing 35Spro:BZR1-YFP and 35Spro:rGRF7-Myc, 35Spro:PIF4-YFP and 35Spro:rGRF7-Myc. The coimmunoprecipitation experiments were performed using GFP-Trap agarose beads, and the immunoblots were probed with anti-Myc or anti-YFP antibodies.

miR396 Co-Regulates Many Genes with BZR1 and PIFs

To evaluate the function of the GRF7-BZR1-PIF4 interaction in the regulation of gene expression, we performed RNA-Seq analysis using 4-d-old dark-grown wild-type Col-0 and 35Spro:miR396a seedlings. We identified 2207 differentially expressed genes affected in 35Spro:miR396a compared to wild-type seedlings, and 4923 genes affected by a light treatment consisting of 6 h irradiation in etiolated wild type seedlings (Figure 8A; Supplemental Data Sets 1 and 2). Of the 2207 genes regulated by 35Spro:miR396a, 1173 genes (53.1%) were also regulated by light. Among these coregulated genes, 957 genes (81.6%) were affected in the same direction by miR396a overexpression and light treatment, with a correlation coefficient R = 0.62, consistent with the function of miR396 as a positive regulator of photomorphogenesis (Figure 8B; Supplemental Data Set 3). Gene ontology (GO) analysis indicated that genes involved in chloroplast development and photosynthesis were highly enriched in the light and miR396 coinduced genes, whereas genes involved in ROS homeostasis were the most significantly enriched functional classes of the light and miR396 corepressed genes, suggesting that light and miR396 might regulate ROS homeostasis to control the greening of etiolated seedlings (Figure 8C). These data provided strong evidence that the miR396-GRFs module plays a critical regulatory role in seedling photomorphogenesis.

Figure 8.

GRFs, BZR1, and PIF4 Share Many Coregulated Genes Across the Genome.

(A) Venn diagrams of the number of genes differentially expressed in dark-grown 35Spro:miR396a versus wild-type Col-0, and genes affected by light treatment in wild-type Col-0 plants.

(B) Hierarchical cluster analysis of the genes coregulated by light treatment and 35Spro:miR396a. The numerical values for the yellow-to-blue gradient bar represent the log₂ of the ratio.

(C) Gene ontology analysis of light and 35Spro:miR396a coregulated genes. Numbers indicate the percentage of genes belonging to each GO category. “Genome” mean genome-wide values.

(D) and (F) Overlap of genes coregulated by 35Spro:miR396a and pifq (D) or bzr-h (F). The top black numbers are the numbers of genes coregulated by miR396 and PIFs or BZR1, and the bottom red numbers are the percentage of PIFs or BZR1 targets.

(E) and (G) Scatterplot of log²-fold change values of coregulated genes by miR396 and pifq (E) or bzr-h (G). The red dots represent the target genes of PIFs or BZR1, the block dots mean the nontargets of PIFs or BZR1.

To determine whether miR396-regulated genes are preferentially regulated by BZR1 and PIFs, we grouped genes on the basis of their expression profile in 35Spro:miR396a, pifq, and bzr-h (a hextuple mutant that disrupted the functions of all six members of BZR1 family), and we calculated the percentage of direct PIF4 or BZR1 targets based on published ChIP data for PIF4 and BZR1 (Sun et al., 2010; Oh et al., 2012; Chen et al., 2019). Among the 2207 miR396a-regulated genes, 314 genes (14.2%) were affected by PIFs, of which 239 genes (76.1%) were regulated by miR396a and PIFs in opposite directions, with a correlation coefficient R = 0.4 (Figures 8D and 8E). A similar negative correlation was observed between miR396-mediated and BZRs-mediated gene expression changes (Figures 8F and 8G). Furthermore, 35Spro:miR396a seedlings shared many coregulated genes with a triple mutant in three BR receptors, bri1 brl1 brl3 (bri-t), and more than 60% of common genes were regulated by miR396 and BRIs in opposite directions (Supplemental Figures 12A and 12B). These results strongly argue for the important role of miR396 in mediating the antagonistic effects of BRs and light on gene expression and photomorphogenesis.

GRFs, BZR1, and PIFs Coregulate the Expression of Chlorophyll Biosynthesis-related Genes.

To further understand the functional importance of the interactions between GRFs, BZR1, and PIF4 on cell elongation and greening of etiolated seedlings, we analyzed the effects of the genetic alteration of each component on the expression of cell elongation and chlorophyll biosynthesis genes. We discovered that the expression levels of multiple cell elongation-related genes, such as PACLOBUTRAZOL RESISTANCE5 (PRE5) and IAA INDUCIBLE19 (IAA19), significantly decreased in pifq and det2-1 mutants compared to wild-type seedlings when grown in the dark, but their lower expression levels were partially suppressed by the addition of 35Spro:MIM396 in these mutant backgrounds (Supplemental Figures 13A and 13B). To further test whether BZR1, PIFs and GRFs cooperatively regulate the expression of PRE5, we generated PRE5pro:LUC reporter construct for transient gene expression analysis in Arabidopsis mesophyll protoplasts (Supplemental Figure 13C). Our results showed that the activity of PRE5pro:LUC increased when BZR1, PIF4, or GRF7 was overexpressed, and were further induced when BZR1, PIF4, and/or GRF7 were coexpressed, suggesting that BZR1, PIF4, and GRF7 cooperatively activate the expression of PRE5.

To determine how BZR1, PIFs, and GRFs regulate the biosynthesis of chlorophyll, we systematically analyzed the expression of 25 key genes in chlorophyll biosynthesis in the wild type Col-0, det2-1, pifq, and 35Spro:miR396a with or without light treatment. Among these 25 genes, 17 genes were significantly induced by light in wild-type seedlings, but impaired for some genes in the det2-1, pifq, and 35Spro:miR396a backgrounds (Supplemental Figures 14A and 14B). The chlorophyll biosynthesis-related genes HEMA1, HEMB1, CHLH, GUN4, and CHLD were upregulated in det2-1, but their higher transcript levels were partially suppressed by the gain-of-function mutant bzr1-1D, suggesting that BRs and BZR1 repress chlorophyll biosynthesis by reducing the expression of this subset of chlorophyll biosynthesis-related genes in the dark (Figure 9). Additionally, we also observed that their expression significantly increased in pifq mutants and remained high even in the bzr1-1D pifq background (Figure 9). Furthermore, overexpression of MIM396 partially blocked the increased transcript levels of HEMA1, HEMB1, CHLH, GUN4, and CHLD in the det2-1 background, but not in the pifq background (Figure 9). These results indicated that BZR1, GRFs, and PIFs all reduce the expression of the HEMA1, HEMB1, CHLH, GUN4, and CHLD genes to inhibit chlorophyll biosynthesis in the dark, although PIFs play more dominant roles.

Figure 9.

GRFs, BZR1, and PIF4 Coregulate the Biosynthesis of Chlorophyll.

RT-qPCR analysis of the expression of chlorophyll biosynthesis genes in wild-type Col-0, det2-1, pifq, 35Spro:MIM396, bzr1-1D, bzr1-1D det2-1, bzr1-1D pifq, 35Spro:MIM396 det2-1, and 35Spro:MIM396 pifq. Seedlings of the wild-type Col-0 and various mutants were grown in the dark for 4 d. PP2A was used as the internal control. Data are shown as means ± sd (n = 3). Different letters above the bars indicate statistically significant differences between the samples (one-way ANOVA followed Tukey test, P < 0.05).

DISCUSSION

Seedling greening is essential for the survival of seed plants during the transition from skotomorphogenesis to photomorphogenesis, particularly under high light environmental conditions. Multiple phytohormones are involved in light-induced seedling greening (Liu et al., 2017a). BRs are a class of steroid hormones that play important roles in plant organ and chloroplast development (Clouse and Sasse, 1998). Here, we demonstrated that BRs are indeed crucial for cotyledon greening and seedling survival. The BR-deficient mutant det2-1 and the BR-insensitive mutant bri1-301 exhibited constitutive photomorphogenic phenotypes in the dark and accumulated higher levels of Pchlide. As a consequence, the greening rates of det2-1 and bri1-301 etiolated seedlings were much lower than that of wild-type seedlings. Conversely, BR treatment effectively rescued the bleached-cotyledon phenotype of det2-1, but not that of bri1-301. Photobleaching of cotyledons and the accumulation of Pchlide seen in det2-1 were both suppressed by the gain-of-function mutant bzr1-1D. The loss-of-function mutants in BZR1 and BES1 exhibited a decreased greening ratio and high levels of Pchlide, similar to det2-1. In addition, overexpression of BRI1 or DWF4 caused over 70% of seedlings to become green upon light exposure after growing in the dark for 8 d, whereas only 37% of wild-type seedlings turned green under the same conditions. Furthermore, we showed that HEMA1, HEMB1, CHLD, CHLH, and GUN4, which encode key enzymes in chlorophyll biosynthesis, were more highly expressed in the det2-1 mutant, an effect that was blocked by the gain-of-function mutation bzr1-1D in the bzr1-1D det2-1 double mutant. These results reveal that BRs and BZR1/BES1 suppress chlorophyll biosynthesis in the dark, thus preventing photo-oxidative damage during seedling de-etiolation in Arabidopsis.

The strong photo-oxidative damage and high accumulation of Pchlide observed in the BR-deficient det2-1 mutant were suppressed by the bzr1-1D mutation, indicating that BRs repress photo-oxidative damage through BZR1. Several key factors involved in the BR-mediated regulation of chlorophyll biosynthesis are known to be regulated by BZR1 (Luo, 2010; Fan et al., 2012; Oh et al., 2012; Wang et al., 2012). GOLDEN2-LIKE1 (GLK1) and GLK2 are homologous transcription factors with MYB-like DNA binding domains that have been reported to play important roles in BRs regulation of chloroplast development. GLK1 and GLK2 are BES1/BZR1 targets that exhibit higher expression levels in bri1 mutants and lower transcript levels in bzr1-1D, bes1-1D and following BRs treatment (Wang et al., 2012). GLK1 and GLK2 promote the expression of many photosynthetic genes and induce chloroplast development in diverse plant species. The glk1 glk2 double mutant displays pale green leaves and phenotypes consistent with chloroplast defects (Waters et al., 2008). Mutants bpg2-1 and bpg3-1D in the BRASSINAZOLE INSENSITIVE PALE GREEN2 (BPG2) and BPG3 genes were identified by screening for mutants with pale green leaves when grown on the BRs biosynthesis inhibitor brassinazole (Komatsu et al., 2010; Yoshizawa et al., 2014). BPG2 encodes a chloroplast stroma-localized YqeH-type GTPase required for the translation of chloroplast photosynthetic RNAs under BR-deficient or high salt stress conditions (Komatsu et al., 2010; Li et al., 2016). BPG2 expression is induced by light but repressed by BRs and high salt. The BPG2 protein has been reported to associate with the chloroplast 16S and 23S rRNA, indicating that BPG2 is involved in BRs, salt and light-regulated rRNA processing to control plastid development (Kim et al., 2012). BPG3 encodes a chloroplast protein that is conserved in bacteria, algae, and land plants. Light and brassinazole treatment both increase the expression levels.of BPG3. An inhibition of electron transport in photosystem II (PSII) was detected in bpg3-1D chloroplasts, suggesting that BRs might be involved in regulating electron transport in PSII (Yoshizawa et al., 2014). However, whether GLK1, GLK2, BPG2, and BPG3 are involved in BR-mediated regulation of greening of etiolated seedlings must await elucidation.

GRFs are a class of conserved plant-specific transcription factors that were first recognized for their effects on leaf growth and development. More recent studies have uncovered the important functions of GRFs in other aspects of plant biology, including seed and root development, flowering, and the coordination of growth processes under diverse stress conditions (Omidbakhshfard et al., 2015). Studies in Arabidopsis using promoter reporter fusion transgenic plants have shown that GRFs are generally more expressed in actively growing tissues compared to mature tissues (Liang et al., 2014), which is consistent with the distribution of endogenous BRs (Clouse and Sasse, 1998). Higher expression of GRF1, GRF2, GRF7, or MIM396 causes larger cotyledons, longer petioles, and larger rossette leaves relative to wild-type plants (Kim et al., 2003). By contrast, null mutants in GRF1-GRF3 or overexpression of miR396a resulted in smaller leaves and cotyledons, and shorter hypocotyls and petioles (Kim et al., 2003). The alteration of leaf growth in GRF overexpressing lines and in null grf mutants was due to the increase and decrease in cell size, respectively, but not cell number (Kim et al., 2003). Genetic analysis showed that seedlings overexpressing rGRF7 or MIM396 were hypersensitive to BRs, but less sensitive to the BRs biosynthesis inhibitor PPZ, whereas overexpression of miR396a or inactivation of multiple members of the GRF family in grf-q led to reduced BRs responses and the increased sensitivity to PPZ. In addition, BRs induce the expression levels of GRF7 and GRF8 through BZR1, and then amplify BR signals to regulate the expression of downstream genes. Furthermore, 35Spro:miR396a and grf-q plants both accumulated high levels of Pchlide and suffered strong photo-oxidative damage upon illumination during seedling de-etiolation. The increased expression levels of HEMA1, GUM4 and CHLH in det2-1 were partially repressed by 35Spro:MIM396. These results demonstrate that miR396 and GRFs play important roles in BRs signaling pathways.

Overexpression of MIM396 had no significant effects on the hypocotyls or rosette leaves of det2-1, possibly because BR-activated BZR1 is required for GRFs promotion of cell elongation. GRF7 interacts with BZR1 to coregulate many common target genes. However, in the det2-1 background, where BR levels are lower and most BZR1 is inactivated by phosphorylation, GRF7 failed to promote the expression of cell elongation-related genes, resulting in a dwarf phenotype. Another possibility is that GRFs are phosphorylated by BIN2 and lose their transcriptional activity when BRs levels are low. A recent study showed that a gain-of-function mutation in rice GRF4 (OsGRF4) showed significantly longer panicles, larger seeds, increased seed weight, and grain yield. GSK2, the rice homologue of Arabidopsis BIN2, directly interacts with OsGRF4 and inhibits its transcription activation activity to regulate grain length (Che et al., 2015; Duan et al., 2015). However, whether BIN2 interacts with GRFs to inhibit their activities in Arabidopsis remains to be tested.

The phenotypes of 35Spro:miR396a, grf-q, and 35Spro:rGRF7 seedlings grown in different light conditions indicated the important roles of miR396 and GRFs in photomorphogenesis. Suppression of GRFs functions by overexpressing miR396a or genetically inactivating GRFs both resulted in constitutive photomorphogenesis phenotypes in the dark, and hypersensitivity to blue, red, and white lights. By contrast, 35Spro:rGRF7 seedlings had longer hypocotyls under all light conditions tested. Transcriptome analysis showed that miR396a and light treatment of etiolated seedlings resulted in a gene expression profile with significant overlap. Most miR396a and light coregulated genes were affected in the same direction. A genome-wide analysis showed that the major function of miR396a is to activate genes involved in chloroplast development while to repress the transcription of cell elongation-related genes. Light treatment repressed the expression of GRF7 and GRF8 through PIFs. PIF4 directly bound to the GRF7 and GRF8 promoters to enhance their expression. GRF7 physically interacted with PIF4 in vivo and in vitro. However, PIF4 overexpression reduced de-etiolated and photobleaching phenotypes of 35Spro:miR396a seedlings, whereas 35Spro:MIM396 had only slight effects on the growth phenotypes of pifq. It is likely that GRF7 interacts with PIF4 to control light-responsive genes and developmental processes, although PIFs seem to play more dominant roles in these processes.

Overall, our results provide strong genetic and molecular evidences for the role of the miR396-GRFs module in the antagonistic interactions between the BRs and light signaling pathways (Figure 10). GRF7 and GRF8 are oppositely regulated by BRs and light through BZR1 and PIF4, respectively. BZR1, PIF4 and GRF7 interact with each other, and they share many common coregulated genes. Each transcription factor enhances the transcriptional activities of the others, and their functions in inducing greening of etiolated seedlings and promoting hypocotyl elongation are partially interdependent. The shared corepressed genes by BZR1, PIFs and GRFs are highly enriched with functions in photosynthesis and chloroplast development, consistent with their important roles in the greening of etiolated seedlings. However, mutation of PIFs resulted in a significantly decreased greening rate in either the bzr1-1D mutant background or in 35Spro:MIM396 seedlings, indicating that PIFs play more dominant roles in greening of etiolated seedlings upon light exposure. In addition, the shared coactivated genes by BZR1, PIF4 and GRFs are enriched in cell wall synthesis and loosening, indicating the important roles of BZR1, PIF4 and GRFs in cell elongation. Genetic analysis showed that BZR1 and PIFs are interdependent to promote cell elongation, whereas overexpression of MIM396 had only slight effects on the growth phenotypes of the BR deficient det2- 1 mutant or the pifq mutant. These results indicate that GRFs have a weaker promoting effect in cell elongation relative to BZR1 and PIFs. Together, such cooperative interaction among BZR1, PIF4, and GRFs elegantly explains the genetic functions of BR, light and miR396 for cell elongation and greening of etiolated seedlings.

Figure 10.

A Working Model for the Roles of BZR1, PIF4, and GRF7 during Cell Elongation and Greening of Etiolated Seedlings.

BR-activated BZR1 and light-inactivated PIF4 directly bind to the promoters of GRF7 and GRF8 to induce their expression. Transcript levels of GRF7 and GRF8 are also controlled by the microRNA miR396. BZR1, PIF4, and GRF7 interact with each other and share many common regulated genes. The shared common repressed genes by BZR1, PIFs, and GRFs are highly enriched with functions in photosynthesis and chloroplast development, consistent with their important roles in the greening switch of etiolated seedlings. PIFs play more dominant roles in this greening process. The shared common activated genes by BZR1, PIF4, and GRFs are enriched in cell wall synthesis and loosening, indicating the critical roles of BZR1, PIF4, and GRFs in cell elongation. Such cooperative interaction between BZR1, PIF4, and GRFs elegantly explains the genetic functions of BRs, light and miR396 for cell elongation and greening of etiolated seedlings.

METHODS

Plant Materials and Growth Conditions

Arabidopsis (Arabidopsis thaliana) Col-0 accession was the wild-type used as control for all phenotypic comparisons and for the generation of transgenic plants. The mutants and transgenic lines bzr1-1D, det2-1, bri1-301, BRI1-Ox, DWF4-Ox, pifq, bzr1-1D det2-1, bzr1-1D pifq, bri1-116, rot3-2, dwf4-102, bin2-1, bzr1-1D bri1-116, bzr1, bes1, bzr1 bes1, det2-9, and pDET2:DET2-GFP det2-9 have been previously described by Wang et al. (2002), Bai et al. (2012), Oh et al. (2012), Lv et al. (2018), and Li et al. (2020). The T-DNA insertion mutants grf1 (SALK_069339), grf3 (SALK_026786), grf4 (SALK_077829C), grf7 (SAIL_1256_F08), and grf8 (SAIL_90_H11) were obtained from the Arabidopsis Biological Resource Center. For the measurement of hypocotyl length, root length, and apical hook curvature, the seedlings were scanned on a flatbed scanner and measured using ImageJ software.

Plasmid and Transgenic Plants

We PCR-amplified a 544 bp genomic sequence containing the miR396a foldback from Col-0 genomic DNA and cloned the PCR product into pENTR/SD/D-TOPO vector (Thermo Fisher Scientific). The MIM396 construct used to inactivate miR396 has been described previously (Debernardi et al., 2012). The full-length GRF7 coding sequence was PCR-amplified and cloned into pENTR/SD/D-TOPO vector. The miR396-resistant version of GRF7, rGRF7, was generated using the Quick-change Site-directed Mutagenesis kit (Stratagene) and cloned into pENTR/SD/D-TOPO vector. Each pENTR clone was recombined with Gateway-compatible destination vectors, including pDEST15 (N-GST), pMAL2CGW (N-MBP), pX-nYFP (35Spro:C-nYFP), pX-cYFP (35Spro:C-cYFP), and pX-YFP (35Spro:C-YFP). Primer sequences used for cloning are listed in Supplemental Table 1. All binary vector constructs were introduced into Agrobacterium (Agrobacterium tumefaciens) strain GV3101, and transformed into Col-0 plants by the floral dip method (Clough and Bent, 1998).

Determination of Greening Rate

Seedlings for wild type and the indicated mutants and transgenic lines were grown on half-strength Murashige and Skoog (MS) medium supplemented with 1% (w/v) Suc at 22°C in darkness for the indicated number of days before exposure to white light for 2 d (white light, 90–110 μmol/m²/s). For BL treatment, seedlings were grown on half-strength MS medium with 1% (w/v) Suc and 100 nM BL or equal volume of solvent only (80% ethanol). Greening rate was determined by counting the percentage of dark green cotyledons from at least 50 seedlings of each genotype. At least three independent biological repeats were performed, each time as technical triplicates. Error bars represent standard errors of three independent experiments.

Pchlide Determination

The 50–100 seedlings grown in darkness for 6 d were harvested under a dim green safe light. The extraction step was conducted as described (Tang et al., 2012). The samples were homogenized in 500 μL of ice-cold 80% (v/v) acetone and incubated in the dark for 6 h. After centrifugation at 5000 g for 5 min, 150 μL of the supernatant was mixed with 350 μL of glycol. Pchlide fluorescence was recorded at room temperature with excitation at 440 nm and emission collected from 600 to 700 nm on a PerkinElmer EnSpire plate reader.

Fluorescence Imaging of ROS

The seedlings were grown in the dark for 6 d before exposure to white light (90–110 μmol/m²/s) for 2 d. Seedlings were then transferred to 100 μM H₂DCFDA in 10 mM Tris-HCl, pH 7.2, in the dark for 10 min. Excess H₂DCFDA was removed by washing with 10 mM Tris-HCl, pH 7.2. H₂DCFDA and chlorophyll fluorescence images were captured on a Zeiss LSM700 laser scanning confocal microscope.

Trypan Blue Staining

Seedlings were boiled for 5 min in staining solution (1.8 mL phenol, 2 mL lactic acid, 2 mL glycerol, and 2 mL 1 mg/mL trypan blue stock solution) and stained for 6 h. The staining solution was then poured off and replaced with destaining solution (500 g chloral hydrate, 200 mL H₂O,pH adjusted to 1.2). Tubes were shaken slowly overnight in the dark. Seedlings were mounted onto slides and photographed on a dissecting microscope.

In Vitro Pull-down Assays

GRF7 fused to GST was purified from bacteria using glutathione beads (GE Healthcare). BZR1 and PIF4 fused to MBP were purified using amylose resin (New England Biolabs). Glutathione beads containing 1 µg of GST-GRF7 were incubated with 1 µg MBP, MBP-BZR1, or MBP-PIF4 as indicated in pull-down buffer (20 mM Tris-HCl, pH 7.5; 100 mM NaCl; 1 mM EDTA), and the beads were washed 10 times with wash buffer (20 mM Tris-HCl, pH 7.5; 300 mM NaCl; 0.5% [v/v] TritonX-100; 1 mM EDTA). The eluted proteins were analyzed by immunoblot analysis with an anti-MBP antibody (New England Biolabs, Cat: E8038L, 1:5,000 dilution).

Coimmunoprecipitation Assays

Arabidopsis seedlings expressing 35Spro:rGRF7-Myc only or coexpressing 35Spro:BZR1-YFP and 35Spro:rGRF7-Myc, 35Spro:PIF4-YFP, and 35Spro:rGRF7-Myc were grown on half-strength MS medium supplemented with 1% (w/v) Suc for 12 d for coimmunoprecipitation assays. Plant materials were harvested and ground in liquid nitrogen and then extracted in lysis buffer containing 20 mM HEPES-KOH, pH 7.5; 40 mM KCl; 1 mM EDTA; 0.5% (v/v) Triton X-100; and 1× protease inhibitor (Sigma Aldrich). After centrifugation at 4°C, 12,000 rpm for 15 min, the supernatant was incubated with GFP-Trap agarose beads (Chromotek) at 4°C for 1 h, and the beads were washed four times using wash buffer (20 mM HEPES-KOH, pH 7.5; 40 mM KCl; 1 mM EDTA; 300 mM NaCl; and 1% [v/v] Triton X-100). The proteins were eluted from the beads by boiling with 2× SDS sample buffer, analyzed by SDS-PAGE, transferred to nitrocellulose membrane and immunoblotted with anti-YFP (TransGen Biotech, Cat: N20610,1:5,000 dilution) and anti-Myc (Sigma Aldrich, Cat: M4439, 1:5,000 dilution) antibodies.

ChIP

Seedlings bearing the constructs 35Spro:BZR1-YFP, 35Spro:PIF4-YFP, and 35Spro:YFP were grown on half-strength MS medium with 1% (w/v) Suc for 12 d under a long day photoperiod, harvested and cross-linked in 1% (v/v) formaldehyde for 30 min under vacuum. Immunoprecipitation was performed as previously described by Tian et al. (2018), using GFP-Trap agarose beads (Chromotek). ChIP products were analyzed by qPCR, and the fold enrichment was calculated as the ratio between 35Spro:BZR1-YFP or 35Spro:PIF4-YFP and 35Spro:YFP, and then normalized by PROTEIN PHOSPHATASE 2A (PP2A, At1g13320), which was used as an internal control. The ChIP experiments were performed with three technical replicates, from which the means and standard deviations were calculated.

RT-qPCR Analysis

Seedlings for wild type and various mutants were grown in the dark for 4 d on half-strength MS medium supplemented with 1% (w/v) Suc and harvested to extract total RNA (TransGen Biotech). First-strand cDNAs were synthesized using RevertAid reverse transcriptase (Thermo Fisher Scientific) and used as qPCR templates. Quantitative PCR analysis was performed on a CFX connect real-time PCR detection system (Bio-Rad) using synergy brands (SYBR) green reagent (Roche) with gene-specific primers (Supplemental Table 1).

Deep Sequencing of the Transcriptome

Seedlings for the wild type Col-0 and 35Spro:miR396a were grown on half-strength MS medium containing 1% (w/v) Suc for 4 d in the dark, and treated with or without white light for 6 h. Total RNA was extracted with Trizol RNA extraction kit (Transgene). Construction and sequencing of the paired-end mRNA libraries was performed on a BGISEQ-500 platform at Beijing Geonomics Institute. The sequence reads were mapped to the Arabidopsis genome using the HISAT and Bowtie2 softwares, and differential gene expression was analyzed using the Noiseq software. Differentially expressed genes were defined by a twofold expression difference with a probability >0.8.

ACCESSION NUMBERS

Sequence data of the Arabidopsis genes studied in this article can be found in The Arabidopsis Information Resource (www.arabidopsis.org) under the following accession numbers: GRF1 (At2g22840), GRF2 (At4g37740), GRF3 (At2g36400), GRF4 (At3g52910), GRF5 (At3g13960), GRF6 (At5g10450), GRF7 (At5g53660), GRF8 (At4g24150), GRF9 (At2g45480), miR396a (At2g10606), BZR1 (At1g75080), BES1 (At1g19350), DET2 (At2g38050), BRI1 (At4g39400), BIN2 (At4g18710), DWF4 (At3g50660), PIF1 (At2g20180), PIF3 (At1g09530), PIF4 (At2g43010), PIF5 (At3g59060), PRE5 (At3g28857), IAA19 (At3g15540), HEMA1 (At1g58290), HEMB1/ALAD1 (At1g69740), HEMC (At5g08280), CHLH/GUN5 (At5g13630), GUN4 (At3g59400), CHLD (At1g08520). The RNA-Seq data described in this article was deposited in the Gene Expression Omnibus database under accession number GSE149834.

SUPPLEMENTAL DATA

• Supplemental Figure 1. BR and BZR1 promote greening of etiolated seedlings upon light exposure.

• Supplemental Figure 2. BZR1 reduces the accumulation of ROS and cell death in the etiolated seedlings upon light exposure.

• Supplemental Figure 3. PIF4 directly binds to the promoters of BZR1 and its homologues.

• Supplemental Figure 4. RT-qPCR analysis of the expression of multiple GRF family members in wild type and various mutant seedlings.

• Supplemental Figure 5. Light and BRs do not affect rGRF7 protein levels.

• Supplemental Figure 6. miR396 promotes photomorphogenesis.

• Supplemental Figure 7. GRFs promote greening of etiolated seedlings.

• Supplemental Figure 8. GRFs are required for skotomophorgenesis.

• Supplemental Figure 9. GRFs positively regulate BRs signaling.

• Supplemental Figure 10. The de-etiolation phenotype of 35Spro:miR396a is suppressed by bzr1-1D.

• Supplemental Figure 11. PIFs are required for miR396-mediated cell elongation.

• Supplemental Figure 12. miR396 and BRIs co-regulate many genes across the genome.

• Supplemental Figure 13. BZR1, PIF4 and GRF7 cooperatively regulate the expression of cell elongation-related genes.

• Supplemental Figure 14. The expression of chlorophyll biosynthesis genes is regulated by BR, PIFs and miR396.

• Supplemental Table. Oligonucleotide sequences used in this study.

• Supplemental Data Set 1. Light-regulated genes in wild type seedlings.

• Supplemental Data Set 2. Genes differentially expressed in 35Spro:miR396a.

• Supplemental Data Set 3. Genes co-regulated by light and miR396a.

• Supplemental Data Set 4. ANOVA analysis in this study.

DIVE Curated Terms

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

• PORA Gramene: AT5G54190

• PORA Araport: AT5G54190

• PORB Gramene: AT4G27440

• PORB Araport: AT4G27440

• HEMB1 Gramene: AT1G69740

• HEMB1 Araport: AT1G69740

• pifQ Gramene: pif1 pif3 pif4 pif5

• pifQ Araport: pif1 pif3 pif4 pif5

• CHLH Gramene: AT5G13630

• CHLH Araport: AT5G13630

• GRF1 Gramene: At2g22840

• GRF1 Araport: At2g22840

• GRF2 Gramene: At4g37740

• GRF2 Araport: At4g37740

• GRF3 Gramene: At2g36400

• GRF3 Araport: At2g36400

• GRF4 Gramene: At3g52910

• GRF4 Araport: At3g52910

• GRF5 Gramene: At3g13960

• GRF5 Araport: At3g13960

• GRF6 Gramene: At5g10450

• GRF6 Araport: At5g10450

• GRF7 Gramene: At5g53660

• GRF7 Araport: At5g53660

• GRF8 Gramene: At4g24150

• GRF8 Araport: At4g24150

• GRF9 Gramene: At2g45480

• GRF9 Araport: At2g45480

• MIR396a Gramene: At2g10606

• MIR396a Araport: At2g10606

• BZR1 Gramene: At1g75080

• BZR1 Araport: At1g75080

• BES1 Gramene: At1g19350

• BES1 Araport: At1g19350

• DET2 Gramene: At2g38050

• DET2 Araport: At2g38050

• DWF4 Gramene: At3g50660

• DWF4 Araport: At3g50660

• PIF1 Gramene: At2g20180

• PIF1 Araport: At2g20180

• PIF3 Gramene: At1g09530

• PIF3 Araport: At1g09530

• PIF4 Gramene: At2g43010

• PIF4 Araport: At2g43010

• PIF5 Gramene: At3g59060

• PIF5 Araport: At3g59060

• PRE5 Gramene: At3g28857

• PRE5 Araport: At3g28857

• IAA19 Gramene: At3g15540

• IAA19 Araport: At3g15540

• HEMA1 Gramene: At1g58290

• HEMA1 Araport: At1g58290

• HEMC Gramene: At5g08280

• HEMC Araport: At5g08280

• GUN4 Gramene: At3g59400

• GUN4 Araport: At3g59400

• CHLD Gramene: At1g08520

• CHLD Araport: At1g08520

• HEMA2 Gramene: At1g09940

• HEMA2 Araport: At1g09940

• HEMA3 Gramene: At2g31250

• HEMA3 Araport: At2g31250

• GSA1 Gramene: at5g63570

• GSA1 Araport: at5g63570

• GSA2 Gramene: AT3G48730

• GSA2 Araport: AT3G48730

• HEMB2 Gramene: AT1G44318

• HEMB2 Araport: AT1G44318

• HEMD Gramene: AT2G26540

• HEMD Araport: AT2G26540

• HEME1 Gramene: AT3G14930

• HEME1 Araport: AT3G14930

• HEME2 Gramene: AT2G40490

• HEME2 Araport: AT2G40490

• HEMF1 Gramene: AT1G03475

• HEMF1 Araport: AT1G03475

• HEMF2 Gramene: AT4G03205

• HEMF2 Araport: AT4G03205

• HEMG1 Gramene: AT4G01690

• HEMG1 Araport: AT4G01690

• HEMG2 Gramene: AT5G14220

• HEMG2 Araport: AT5G14220

• CHLI1 Gramene: AT4G18480

• CHLI1 Araport: AT4G18480

• CHLI2 Gramene: AT5G45930

• CHLI2 Araport: AT5G45930

• CHLM Gramene: AT4G25080

• CHLM Araport: AT4G25080

• CRD1 Gramene: AT3G56940

• CRD1 Araport: AT3G56940

• PORC Gramene: AT1G03630

• PORC Araport: AT1G03630

• BR6OX2 Gramene: AT3G30180

• BR6OX2 Araport: AT3G30180

• BEH1 Gramene: AT3G50750

• BEH1 Araport: AT3G50750

• BEH2 Gramene: AT4G36780

• BEH2 Araport: AT4G36780

• BEH3 Gramene: AT4G18890

• BEH3 Araport: AT4G18890

• BEH4 Gramene: AT1G78700

• BEH4 Araport: AT1G78700