Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2022 Oct 27;112(5):1159–1175. doi: 10.1111/tpj.16002

Chrysanthemum MAF2 regulates flowering by repressing gibberellin biosynthesis in response to low temperature

Jing Lyu 1,2, † , Palinuer Aiwaili 1,2, † , Zhaoyu Gu 2, Yanjie Xu 2, Yunhan Zhang 2, Zhiling Wang 2, Hongfeng Huang 2, Ruihong Zeng 2, Chao Ma 1,2, Junping Gao 1,2, Xin Zhao 1,3,✉, Bo Hong 1,2,✉
PMCID: PMC10092002  PMID: 36214418

SUMMARY

Chrysanthemum (Chrysanthemum morifolium) is well known as a photoperiod‐sensitive flowering plant. However, it has also evolved into a temperature‐sensitive ecotype. Low temperature can promote the floral transition of the temperature‐sensitive ecotype, but little is known about the underlying molecular mechanisms. Here, we identified MADS AFFECTING FLOWERING 2 (CmMAF2), a putative MADS‐box gene, which induces floral transition in response to low temperatures independent of day length conditions in this ecotype. CmMAF2 was shown to bind to the promoter of the GA biosynthesis gene CmGA20ox1 and to directly regulate the biosynthesis of bioactive GA1 and GA4. The elevated bioactive GA levels activated LEAFY (CmLFY) expression, ultimately initiating floral transition. In addition, CmMAF2 expression in response to low temperatures was directly activated by CmC3H1, a CCCH‐type zinc‐finger protein upstream. In summary, our results reveal that the CmC3H1–CmMAF2 module regulates flowering time in response to low temperatures by regulating GA biosynthesis in the temperature‐sensitive chrysanthemum ecotype.

Keywords: Chrysanthemum morifolium, low‐temperature flowering, CmMAF2, CmC3H1, GA biosynthesis

Significance Statement

Chrysanthemum (Chrysanthemum morifolium) is well known as a photoperiod‐sensitive flowering plant. However, it has also evolved a temperature‐sensitive ecotype whose regulatory mechanisms of flowering remain largely unknown. We show that the C3H1–MAF2 module contributes to floral transition by regulating the GA–LFY pathway. This study contributes to development of flowering time regulation technologies and breeding of new energy‐saving chrysanthemum cultivars.

INTRODUCTION

Adaptation of flowering plants to environmental conditions to ensure appropriate flowering time is critical for reproductive success (Amasino, 2009). To this end, angiosperms have evolved diverse strategies that vary depending on whether they are exposed to external stimuli (photoperiod and vernalization) or endogenous cues (autonomous, gibberellins [GAs], and aging), which are interconnected and form a complex network that regulates flowering (Song et al., 2013; Srikanth & Schmid, 2011). For vernalization‐requiring plants, prolonged exposure to low temperatures can promote their flowering ability (Ream et al., 2012), and some of the mechanisms underlying low temperature‐induced flowering have been elucidated in the experimental model plant Arabidopsis thaliana (Xu & Chong, 2018). In winter‐annual A. thaliana, FLOWERING LOCUS C (FLC) encodes a MADS‐box transcription factor that acts as a core floral repressor (Amasino, 2004; Michaels & Amasino, 1999). There are three phases of FLC expression: a high level of FLC expression prior to vernalization, silencing of FLC expression during vernalization, and stable maintenance of FLC silencing upon returning to warmer conditions. It has been shown that flowering competence is acquired through a direct upregulation of FLOWERING LOCUS T (FT) (Gu et al., 2013; Whittaker & Dean, 2017). It is also known that in biennial winter varieties of wheat (Triticum aestivum) and barley (Hordeum vulgare), VERNALIZATION 1 (VRN1), a homolog of A. thaliana APETALA 1, is the key floral activator during vernalization (Xu & Chong, 2018) and that in perennial Arabis alpina (a relative of the annual A. thaliana), PERPETUAL FLOWERING1 (PEP1) confers the vernalization requirement (Wang et al., 2009).

MADS AFFECTING FLOWERING 2 (MAF2) is a member of the MADS‐box transcription factor gene family in Arabidopsis, which has five members (MAF1–5). These genes are close homologs of FLC and have also been shown to regulate flowering time in response to low temperatures (Gu et al., 2013; Ratcliffe et al., 2003). MAF2 expression is less sensitive to vernalization than FLC and works either independently or downstream from FLC (Ratcliffe et al., 2003). Nevertheless, the intricate network by which MAF2 regulates flowering time remains unclear.

The Asteraceae family is regarded as the most diverse family of flowering plants (Barreda et al., 2015). Chrysanthemum (Chrysanthemum morifolium) is a perennial herbaceous member of Asteraceae, whose floral induction mainly depends on the photoperiodic pathway (Higuchi et al., 2013). Hence, chrysanthemum has evolved several different ecotypes due to the differences in critical day length for flowering (Kawata, 1987). However, low temperature also has major effects on the flowering of different chrysanthemum ecotypes. In particular, there is a low temperature‐sensitive ecotype whose floral induction requires low‐temperature exposure, which is also referred to as vernalization (Harada & Nitsch, 1959; Hisamatsu et al., 2017; Schwabe, 1950; Schwabe, 1954). Indeed, several chrysanthemum varieties form a rosette without low‐temperature treatments, indicating a typical vernalization requirement (Sumitomo et al., 2009). Nevertheless, it is still unknown how chrysanthemum regulates floral transition in response to low temperatures.

Genetic analyses have identified the plant hormone GA as having a prominent role in flowering time regulation. Among the GA compounds, only those that are bioactive, such as GA1, GA3, and GA4, function in regulating floral transition (Binenbaum et al., 2018; Langridge, 1957; Mutasa‐GÖttgens & Hedden, 2009; Song et al., 2013). In A. thaliana, the application of GAs rescues the flowering mutant phenotype under non‐inductive short‐day (SD) conditions (Langridge, 1957). Flower initiation in A. thaliana under non‐inductive SD conditions depends on the biosynthesis of GA and on the regulation of the expression of the flower meristem identity gene LEAFY (LFY) (Eriksson et al., 2006). Studies have shown that three major oxidase genes regulate GA biosynthesis and inactivation: GA20‐oxidase (GA20ox), GA3‐oxidase (GA3ox), and GA2‐oxidase (GA2ox) (Lee & Zeevaart, 2007; Rieu et al., 2008). Overexpression of GA20ox promotes shoot elongation and early flowering in A. thaliana (Lee & Zeevaart, 2007; Rieu et al., 2008), and similarly, in SD chrysanthemum, GAs can promote flowering under non‐inductive long‐day (LD) conditions (King, 2012; Pharis, 1972; Yang et al., 2014).

Studies in different species have established a physiological connection between low‐temperature conditions and GA metabolism. In the Brassicaceae species field pennycress (Thlaspi arvense L.), vernalization was found to affect GA metabolism by increasing the levels of the GA precursor kaurenoic acid (Hazebroek et al., 1993). Furthermore, in winter canola (Brassica napus), the abundance of active GA1 and GA3 increased following vernalization treatment (Zanewich & Rood, 1995). Finally, in lisianthus (Eustoma grandiflorum), levels of the GA precursor ent‐kaurene were shown to be increased by low‐temperature exposure, leading to higher endogenous GA1 levels (Hisamatsu et al., 2004). However, the molecular mechanisms underlying the connections between low temperature and GA metabolism remain largely unknown.

Members of the Cys3His‐type zinc‐finger protein superfamily are involved in many aspects of plant growth and development (Bogamuwa & Jang, 2014). In Arabidopsis, TZF1 (AtC3H23), a Cys3His‐type family member, can affect plant growth and stress responses (Lin et al., 2011). AtTZF2 (AtOZF1/AtC3H20) and AtTZF3 (AtOZF2/AtC3H49) are involved in ABA, jasmonate, and oxidative stress responses (Huang et al., 2011; Huang et al., 2012; Lee et al., 2012). However, little is known about the roles of Cys3His‐type zinc finger proteins in regulating flowering time.

In the current study, we report that low temperature is a key environmental factor in inducing the floral transition of a temperature‐sensitive chrysanthemum ecotype. We present evidence for a novel mechanism of low‐temperature response in chrysanthemum in which the CmC3H1–CmMAF2 module contributes to the induction of flowering. Our results reveal that in response to low temperatures, CmMAF2 expression was directly induced by CmC3H1, and CmMAF2 directly targeted CmGA20ox1, elevating bioactive GA levels and in turn activating CmLFY expression, ultimately resulting in flowering. Our discovery of this mechanism will contribute to the development of flowering technologies and the annual production of different chrysanthemum ecotypes.

RESULTS

Low temperatures induce the capacity for flowering

To investigate the impact of extended low‐temperature periods on physiological mechanisms that govern flowering in chrysanthemum, we used a cultivar that is sensitive to low temperatures (C. morifolium cv. Summer Yellow). As shown in Figure 1, under LD conditions, plant growth was severely inhibited by 1–5 weeks of low‐temperature treatment (LT1–LT5), and the growth rate increased after returning to normal temperatures for 1–5 weeks (recovery phase LT5 + N1 to LT5 + N5). The treated plants reached the same height as the control plants after returning to normal temperatures for 5 weeks (LT5 + N5) (Figure 1a,b). We then tested the effects of low temperatures on growth under SD conditions. Control plants grew slowly (C8–C13), while plants exposed to low temperature for 8 weeks showed rapid growth and were significantly taller than the control, ultimately reaching the proper height (LT8 + N5) (Figure S1).

Figure 1.

Figure 1

Open in a new tab

Chrysanthemum flowering in response to low temperature. (a) Plants were grown for 5 weeks at low temperature (LT1–5) and then returned to normal temperature for recovery (LT5 + N1–N5). Control plants (C1–10) were grown under normal temperature and LD conditions. The phenotypes were observed daily. Scale bars, 5 cm. (b) Plant heights after 5 weeks of growth at low temperature (LT1–5) and 1–5 weeks after returning to normal temperature (LT5 + N1–N5). (c) Scanning electron microscopy images of apex development in plants grown at low temperatures for 5 weeks or at normal temperatures under LD conditions. C5–10, 5–10‐week control; LT5, 5‐week low‐temperature treatment; LT5 + N1–5, 1–5 weeks after returning to normal temperature; LF, leaf primordium; IN, involucre; BL, bract leaf primordium. Scale bars, 200 μm. (d) Scanning electron microscopy images of inflorescence and flower bud morphology in plants treated for 5 weeks with low temperature (LT + LD) or without low temperature (LD) after 11 weeks and 14 weeks under LD conditions. Scale bars, 5 cm. (e) Blooming phenotype in plants treated for 5 weeks with low temperature (LT5) or without low temperature (LD) after 18 weeks under LD conditions. Scale bars, 5 cm. (f) Days until initial flower bud emergence in plants treated for 5 weeks with low temperature or without low temperature. (g) The phenotype of plants treated for 5 weeks with low temperature (LT + natural environment [NE]) or without low temperature (NE) after 9 weeks and 13 weeks in the field and photographed on 6 June and 30 June, respectively (Beijing, China). Scale bars, 5 cm. (h) Days until initial flower bud emergence in plants treated for 5 weeks with low temperature (LT + NE) or without low temperature (NE). The results are the means of five biological replicates with standard deviations. Asterisks indicate statistically significant differences (Student’s t‐test, **P < 0.01).

Next, we determined the flowering time in response to low temperatures under different day lengths. Under LD conditions, microscopic observation showed that the apical meristem of plants treated with low temperatures began to exhibit hypertrophy at LT5 + N1, entered the early stage of involucral primordium at LT5 + N3, and formed the involucral primordium at LT5 + N5, while the control plants (C5–C10) remained in the vegetative stage (Figure 1c). The time to floret primordium formation, the development of an apical inflorescence, and flower blooming occurred significantly earlier after a 5‐week low‐temperature treatment at 11 weeks (LT5 + N6), 14 weeks (LT5 + N9), and 18 weeks (LT5 + N13) compared to non‐treated plants (Figure 1d,e). There was also a statistically significant difference in initial flower bud formation (diameter = 2 mm). They were observed at 37 ± 1.7 days after transplanting in plants treated with low temperatures and at 96 ± 2.2 days after transplanting in non‐treated control plants (Figure 1f). We then tested the effect of low temperatures on flowering in natural environments by transplanting plants that have been treated for LT5 to the field. We observed that floral buds appeared and bloomed significantly earlier in plants treated with low temperatures than control plants (Figure 1g,h). Under SD conditions, no flower buds were observed for 200 days on the control plants, while the emergence of initial flower buds was observed at 105 ± 5.0 days following an 8‐week low‐temperature treatment (Figure S1). These data indicate that low temperatures induce the capacity for flowering independent of the photoperiod.

Endogenous GA levels are associated with low temperature‐induced flowering

We analyzed the RNA sequencing (RNA‐seq) data from low temperature‐treated and control plants in the floral transition phase. The results showed that GA pathway genes were differentially expressed (Table 1). To test this, we measured endogenous GA levels under low‐temperature conditions and in the recovery phase. We observed that compared with C5, the low‐temperature treatment LT5 caused a significant decrease in the abundance of bioactive GA1, its precursor GA20, and its metabolite GA8. However, compared with C7, GA1, GA20, and GA8 levels were higher at LT5 + N2 (Figure 2a). Similarly, the concentrations of bioactive GA4 and its precursor GA9 were significantly reduced in LT5 but were higher in LT5 + N2 plants (Figure 2a). Thus, growth inhibition due to low temperatures, combined with early flowering in the recovery phase, was correlated with the change in active GA levels.

Table 1.

Differentially expressed genes related to flowering time at 5 weeks at low temperature (LT5), and at 1 and 5 weeks after returning to normal temperature (LT5 + N1, LT5 + N5)

Gene Annotation RPKM (C5) RPKM (LT5) RPKM (LT5 + N1) RPKM (LT5 + N5)
Vernalization pathway
Cluster‐17868.118397 CmMAF2 10.62 ± 1.02bc 24.04 ± 3.34a 15.49 ± 2.97b 8.83 ± 0.52c
Gibberellin pathway
Cluster‐17868.212239 GA20ox1 1.17 ± 0.14a 0.27 ± 0.16b 1.22 ± 0.21a 0.34 ± 0.17b
Cluster‐17868.71469 GA2ox 0.80 ± 0.30b 4.02 ± 0.50a 1.10 ± 0.33b 1.37 ± 0.15b
Cluster‐17868.234688 GA3ox 3.00 ± 0.85b 12.50 ± 4.98a 1.81 ± 0.53b 0.41 ± 0.21b
Cluster‐17868.223832 GID1c 2.91 ± 0.74c 8.50 ± 0.25a 6.29 ± 1.22b 1.57 ± 0.36c
Cluster‐17868.139458 GAI‐like 22.47 ± 1.75b 9.44 ± 1.00c 44.58 ± 12.78a 41.44 ± 0.92a
Cluster‐17868.147300 RGA 31.17 ± 1.11a 13.18 ± 1.09b 30.16 ± 1.29a 31.41 ± 1.05a
Flowering integrator
Cluster‐17868.27582 AFT 0.94 ± 0.03c 0.80 ± 0.28c 4.73 ± 0.40a 2.66 ± 0.42b
Cluster‐17868.32465 AFL2 4.23 ± 0.96a 1.56 ± 0.22b 0.78 ± 0.40b 5.73 ± 0.93a
Cluster‐17868.114531 FDL1 4.01 ± 1.32a 4.04 ± 0.14a 1.82 ± 0.52b 1.42 ± 0.07b
Cluster‐17868.147733 FDL2 9.91 ± 0.16a 7.81 ± 0.91b 10.05 ± 0.85a 7.76 ± 0.77b
Cluster‐17868.233356 FTL2 0.25 ± 0.01c 0.78 ± 0.07b 0.16 ± 0.04c 1.27 ± 0.06a
Cluster‐17868.71935 FTL3 3.25 ± 0.08b 1.10 ± 0.13c 0.27 ± 0.08c 5.88 ± 1.10a
Open in a new tab

Three independent experiments were performed; ‘±’ indicates standard deviation. Different letters indicate significant differences according to Duncan's multiple range test (P < 0.05).

Figure 2.

Figure 2

Open in a new tab

Effects of GA treatments on chrysanthemum flowering under LD conditions. (a) GA content in the shoot tips of plants at LT5 and at LT5 + N2. Plants grown for 5 or 7 weeks under normal temperature and LD conditions were used as controls (C5 and C7). (b) The effects of exogenous GA4/7 treatment and 5‐week low‐temperature treatment (LT5) on plant height of 8‐week‐old chrysanthemum. GA was given twice per week for 2 months. (c) The effects of exogenous GA4/7 treatments and 5‐week low‐temperature treatments (LT5) on initial flower bud appearance in control and treated plants. (d) Phenotypes of control and treated plants were observed after 14 weeks. The results are the means of three biological replicates with standard deviations. Asterisks indicate statistically significant differences (Student's t‐test, **P < 0.01). Different letters indicate significant differences according to Duncan's multiple range test (P < 0.05). Scale bars, 5 cm.

We next verified the relationship between GA levels and low temperature during chrysanthemum flowering by applying exogenous GA. Under LD conditions, plant height significantly increased after application of 100 μm GA4/7 compared to the control. In addition, the time to flowering was earlier in the plants treated with GA4/7 and similar to that in plants treated with low temperatures for 5 weeks (Figure 2b–d).

Under SD conditions, an increase in plant height was observed in plants treated with 100 μm GA4/7 or low temperature (Figure S2a,b). However, no flower buds appeared on the control plants by day 200 after transplanting, while 100 μm GA4/7 and low temperature‐treated plants had flower buds by day 137 ± 1.6 and 105 ± 5.0, respectively (Figure S2a–c). These results indicate that the effects of low temperature on flowering time are associated with GA biosynthesis.

CmMAF2 and CmGA20ox1 expression is regulated by low temperatures

To elucidate the molecular mechanisms underlying floral induction in response to low temperatures in chrysanthemum, we identified differentially expressed genes at LT5, LT5 + N1, and LT5 + N5. We observed that the expression of a putative MADS‐box gene (Cluster‐17868.118397) was 2.3‐fold higher in LT5‐treated plants compared with controls. Its expression in treated plants gradually decreased after returning to normal temperature (Table 1; Data S1). Amino acid sequence alignment showed that Cluster‐17868.118397 had a high degree of sequence homology to A. thaliana MAF2, so this gene was re‐named CmMAF2 (Figure S3a,b). In addition, the expression of the GA biosynthesis gene CmGA20ox1 (Cluster‐17868.212239) decreased 4.3‐fold in LT5‐treated plants compared with controls, while its expression gradually increased after returning to normal temperatures (Table 1; Data S1).

We measured the mRNA expression levels of CmMAF2 in leaves and shoot tips using quantitative real‐time PCR (qRT‐PCR) and found that both of them were upregulated during low‐temperature treatments and downregulated after returning to normal temperatures (Figure 3a,b). However, the expression of the GA biosynthesis gene CmGA20ox1 showed a reverse trend, being repressed by the low‐temperature treatment and increasing after returning to normal temperatures in both leaves and shoot tips (Figure 3c,d). This expression pattern was consistent with the results from an in situ hybridization analysis, showing that CmGA20ox1 expression in apical meristems was lower under low‐temperature conditions than under control conditions. CmGA20ox1 expression was higher at LT + N1 (Figure 3e). Unlike CmGA20ox1, the expression of another GA20ox gene, CmGA20ox2, was slightly and continuously suppressed by low temperatures, and the expression did not recover after returning to normal temperatures (Figure S3c).

Figure 3.

Figure 3

Open in a new tab

CmMAF2, CmGA20ox1, and CmLFY expression in response to low temperature (LT). (a,b) Expression of CmMAF2 was analyzed by qRT‐PCR in leaves (a) and shoot tips (b). (c,d) Expression of CmGA20ox1 was analyzed by qRT‐PCR in leaves (c) and shoot tips (d). (e) In situ hybridization of CmGA20ox1 in chrysanthemum apical meristems at LT5 and at LT5 + N1. Plants grown under normal temperature for 5 and 6 weeks (C5 and C6) were used as controls. Scale bars, 200 μm. (f) CmLFY expression was analyzed by qRT‐PCR in shoot tips. UBIQUITIN was used as the reference gene. The results are the means of three biological replicates with standard deviations. LT1–5, 1–5 weeks of low‐temperature treatment. LT5 + N1, LT5 + N2, and LT5 + N5 indicate 1 week, 2 weeks, and 5 weeks of normal temperature after a 5‐week low‐temperature treatment, respectively.

To understand the effects of low temperatures on flower initiation in chrysanthemum, we determined the expression levels of CmFTL1, CmFTL2, and CmFTL3, three FT‐homologous genes, and CmLFY, a flower meristem identity gene. The results showed no significant differences in the expression of CmFTL1–3 between low‐temperature treatment and control plants in both leaves and shoot tips (Figure S2e–j). CmLFY expression in shoot tips was not significantly changed compared to the control during the low‐temperature treatments, while CmLFY expression was significantly increased at 2 weeks after returning to normal temperatures, which is the time point for floral transition (Figure 3f). We also determined CmLFY expression levels during GA4/7 and paclobutrazol (PAC) treatments, and the results showed that CmLFY expression was significantly induced by GA, while it was inhibited by PAC (Figure S2d). These data suggest that CmGA20ox1 and CmLFY might participate in flowering regulation in low temperature‐induced chrysanthemum ecotypes.

CmMAF2 regulates flowering by regulating GA biosynthesis

To test whether CmMAF2 plays a role in the regulation of flowering time, we used the 3′ region of the gene to specifically silence CmMAF2 expression in chrysanthemum by RNA interference (RNAi) and generated a population of 36 CmMAF2‐RNAi lines. Of these, we selected three for functional analysis based on their reduced CmMAF2 expression levels, as determined by qRT‐PCR analysis (Figure 4a). Since we did not find any other MAF homologs in our chrysanthemum transcriptome and public genome databases, we measured the expression of the other MADS‐box gene family homologs AGL8, AGL9, AGL104, SVP1, and SVP2, whose functions have been associated with floral transition, in transgenic and wild‐type (WT) plants, and confirmed that only the CmMAF2 gene was silenced (Figure 4b).

Figure 4.

Figure 4

Open in a new tab

CmMAF2‐RNAi plant flowering. (a,b) Expression of CmMAF2 (a) and MADS‐box family homologs (b) in WT and CmMAF2‐RNAi plants as determined by qRT‐PCR. UBIQUITIN was used as the internal control. (c,h) Plant heights of 4‐week‐old (c) and 12‐week‐old (h) WT and CmMAF2‐RNAi plants grown under LD and SD conditions. (d,i) The number of stem nodes was recorded after 9 weeks of LD (d) and 18 weeks of SD (i) conditions. (e,j) Representative photographs of WT and CmMAF2‐RNAi plants at flower bud emergence after 9 weeks of LD (e) and 18 weeks of SD (j) conditions. (f,k) Flower blooming in WT and CmMAF2‐RNAi plants after 15 weeks of LD (f) and 20 weeks of SD (k) conditions. (g,l) Days until initial flower bud emergence in WT and CmMAF2‐RNAi plants grown under LD (g) and SD (l) conditions. (m) Phenotypes of WT and CmMAF2‐RNAi plants treated for 5 weeks with low temperature at week 16 after transplanting. (n) Plant heights of WT and CmMAF2‐RNAi plants treated for 5 weeks with low temperature under LD conditions. (o) Days until initial flower bud emergence in WT and CmMAF2‐RNAi plants treated for 5 weeks with low temperature under LD conditions. R24, R31, and R24 correspond to three independent CmMAF2‐RNAi lines. The results are the means of three biological replicates with standard deviations. Asterisks indicate statistically significant differences (Student's t‐test, *P < 0.05, **P < 0.01). Scale bars, 5 cm.

We observed that under LD conditions, the plant heights of lines RNAi‐24, RNAi‐31, and RNAi‐34 were 38 ± 1.0 cm, 34 ± 0.3 cm, and 33 ± 0.6 cm, respectively, which were significantly greater than that of WT plants (23 ± 1.0 cm), but no obvious difference was observed in the numbers of stem nodes between the transgenic lines and WT plants (Figure 4c,d). Plants grown under SD conditions showed a similar trend (Figure 4h,i). We then monitored the flowering time of CmMAF2‐RNAi and WT plants. Under LD conditions, the emergence of initial flower buds was observed at 39 ± 3.4, 55 ± 3.3, and 47 ± 1.8 days after transplanting in lines RNAi‐24, RNAi‐31, and RNAi‐34, respectively, while flower buds emerged in WT plants at 96 ± 2.2 days. Meanwhile, the blooming time of transgenic lines was also earlier than in WT plants (Figure 4e–g). Under SD conditions, initial flower buds emerged at 93 ± 5.3, 111 ± 2.2, and 112 ± 1.6 days after transplanting in lines RNAi‐24, RNAi‐31, and RNAi‐34, respectively, while no flower buds and blooming time were observed in WT plants until the end of the experiment (200 days after transplanting) (Figure 4j–l). We exposed the CmMAF2‐RNAi lines and WT plants to a 5‐week low‐temperature treatment and observed a reduction in plant height, although CmMAF2‐RNAi plants were still taller than the WT plants (Figure 4m,n). In addition, no significant difference in flowering time was observed between the CmMAF2‐RNAi lines and WT plants (Figure 4m,o), suggesting that the low‐temperature treatment mitigated the consequences of RNAi‐mediated suppression of CmMAF2 expression on flowering time.

To investigate whether CmMAF2 affects flowering through the GA flowering pathway, we measured endogenous GA concentrations in CmMAF2‐RNAi plants and observed that bioactive GA levels (GA1 and GA4) were significantly higher in CmMAF2‐RNAi plants than in WT plants (Figure 5a). We then tested the effects of the GA biosynthesis inhibitor PAC on flowering time of CmMAF2‐RNAi and WT plants. After PAC treatment, plant height was significantly reduced in the CmMAF2‐RNAi and WT plants, but the degree of inhibition in CmMAF2‐RNAi plants was lower than in WT plants (Figure 5b,c). Additionally, flowering time was delayed in both CmMAF2‐RNAi and WT plants, and WT plants showed a more substantial delay in flowering than CmMAF2‐RNAi plants, such that we did not observe flower buds in WT plants for 100 days after treatment (Figure 5b–d). These results indicated that GA biosynthesis is involved in regulating CmMAF2 expression and its effect on flowering.

Figure 5.

Figure 5

Open in a new tab

The effects of GAs on flowering in CmMAF2‐RNAi and WT chrysanthemum plants. (a) GA contents in 10‐week‐old WT and CmMAF2‐RNAi plants grown under LD conditions. (b) Phenotypes of CmMAF2‐RNAi lines treated with the GA biosynthesis inhibitor PAC under LD conditions after 6 weeks and of WT plants treated with the GA biosynthesis inhibitor PAC under LD conditions after 15 weeks. Scale bars, 5 cm. (c) The effect of PAC treatment on the height of 5‐week‐old RNAi and WT plants. (d) Effects of PAC treatment on the time until initial flower bud emergence in RNAi and WT plants. R24, R31, and R24 correspond to three independent CmMAF2‐RNAi lines. Mock treatment with 10% ethanol solution was used as a control. (e,f) Expression analyses of genes related to GA biosynthesis under LD (e) and SD (f) conditions, as determined by qRT‐PCR. UBIQUITIN was used as an internal control. The results are the means of three biological replicates with standard deviations. (g) In situ hybridization analysis of CmGA20ox1 expression in the apical meristems of WT and CmMAF2‐RNAi plants grown under LD conditions. The negative control SP6 was hybridized with the sense probe. Scale bars, 200 μm. The results are the means of three biological replicates with standard deviations. Asterisks indicate statistically significant differences (Student's t‐test, **P < 0.01).

CmMAF2 is a direct upstream regulator of the GA biosynthesis gene CmGA20ox1

To clarify whether CmMAF2 directly regulates genes involved in GA biosynthesis, we analyzed the expression of differentially expressed genes related to GA synthesis in the vegetative stage (1 w) and the early stage of floral transition (5 w) in both CmMAF2‐RNAi transgenic and WT plants by qRT‐PCR. We found that in the vegetative and early floral transition stages, the expression of CmGA20ox1 was higher in CmMAF2‐RNAi plants than in WT under both LD and SD conditions, but CmGA3ox expression was lower. We did not detect a significant difference in CmGA2ox expression between transgenic and WT plants under flowering‐inductive LD conditions of the low temperature‐induced chrysanthemum variety (Figure 5e,f; Figure S4a–c). In situ hybridization analysis of CmGA20ox1 expression in apical meristems showed stronger labeling in RNAi plants than in WT plants, although the stages of inflorescence differentiation were different between the RNAi and WT plants (Figure 5g).

We noted that the CmGA20ox1 promoter contains three annotated CArG‐box motifs, while the CmGA2ox promoter does not contain any (Figure 6a; Figure S5). A yeast one‐hybrid (Y1H) analysis to assess the interaction of CmMAF2 with each of the three motifs revealed that it bound to them all (Figure 6b). Next, we carried out an electrophoretic mobility shift assay (EMSA) to validate the interaction of CmMAF2 with the three CArG‐box motifs and observed that a GST‐tagged version of CmMAF2 (CmMAF2‐GST) bound to each of the CArG‐box‐containing fragments that had been biotin‐labeled. In this analysis, increasing the amounts of unlabeled cold probes significantly decreased the levels of CmMAF2 binding to biotin‐labeled probes, whereas unlabeled mutant probes did not compete for CmMAF2 binding (Figure 6d). Finally, we assayed the activity of a dual‐luciferase reporter to evaluate the regulatory effects of CmMAF2 on the CmGA20ox1 promoter in vivo. Tobacco (Nicotiana benthamiana) leaf cells co‐expressing CmMAF2 and the CmGA20ox1 promoter had a significantly lower LUC/REN ratio than leaf cells transformed with CmMAF2 alone, the CmGA20ox1 promoter alone, or a mutated GA20ox1 promoter (Figure 6c), suggesting that CmMAF2 directly represses the expression of CmGA20ox1.

Figure 6.

Figure 6

Open in a new tab

CmMAF2 represses the expression of CmGA20ox1 by binding to its promoter. (a) Schematic representation of three CArG‐box sequences and their positions in the 600‐bp sequence immediately upstream from the CmGA20ox1 transcription start site (TSS). (b) Interaction between CmMAF2 and the three CArG‐boxes in the CmGA20ox1 promoter, shown by a Y1H assay. The p53‐AbAi bait vector and the p53 fragment prey vector were used as positive controls. All interactions were examined on SD/−Leu medium supplemented with 100 mg μl−1 AbA. (c) CmMAF2 affects the transcriptional activity of CmGA20ox1, as shown using a dual‐luciferase reporter assay in Nicotiana benthamiana leaves. A 680‐bp CmGA20ox1 promoter sequence was used. mPro‐GA20ox is the same promoter fragment with the three CArG cis‐elements mutated. LUC indicates the pGreenII 0800‐LUC empty vector containing the REN gene under the control of the 35S promoter. SK indicates the empty pGreenII 0029 62‐SK vector. Samples were infiltrated into N. benthamiana leaves, and LUC and REN activities were assayed 3 days after infiltration. Three independent experiments were performed, and error bars indicate standard deviation. (d) The interaction of CmMAF2 and biotin‐labeled CArG cis‐elements as shown by EMSA. The oligonucleotide sequences of three CArG‐boxes from the CmGA20ox1 promoter were used as probes. The core sequences are underlined. The purified protein (2 μg) was incubated with 40 nmol of labeled WT or mutated probes. Non‐labeled probes at various concentrations (10–100‐fold) were added for the competition test. (e) Transcript abundance of CmGA20ox1 in WT and CmMAF2‐RNAi plants infected with CaLCuV or CaLCuV‐amiR‐CmGA20ox1. (f,h) Representative photographs of WT and CmMAF2‐RNAi plants infected with CaLCuV or CaLCuV‐amiR‐CmGA20ox1 after 15 weeks of growth (f) and at flower blooming (h) under LD conditions. (g,i) Plant heights (g) and days to initial flower bud emergence (i) were recorded (n > 5). Error bars indicate standard deviation. Asterisks indicate significant differences according to a Student's t‐test (*P < 0.05, **P < 0.01). Different letters indicate significant differences according to Duncan's multiple range test (P < 0.05). Red scale bars, 1 cm; white scale bars, 5 cm.

To confirm that CmMAF2 influences flowering by directly regulating the expression of CmGA20ox1, we silenced CmGA20ox1 in WT and CmMAF2‐RNAi plants using a modified cabbage leaf‐curl geminivirus vector (CaLCuV) containing the artificial microRNA‐CmGA20ox1 (CaLCuV‐amiR‐CmGA20ox1) (Figure 6e). We observed that under LD conditions, compared with WT‐CaLCuV plants, silencing CmMAF2 alone significantly increased the plant height, and silencing CmGA20ox1 alone strongly reduced plant height, while double silencing of CmMAF2 and CmGA20ox1 recovered the plant height close to that of WT‐CaLCuV plants (Figure 6f,g). We also observed the influence of flowering under LD conditions compared with early flowering of CmMAF2‐RNAi‐CaLCuV plants; both the time until initial flower bud emergence and the time until blooming were significantly longer in CmMAF2 and CmGA20ox1 double silenced plants, which were close to those of WT‐CaLCuV plants. Silencing CmGA20ox1 alone delayed flowering extremely much, and we did not even observe flower opening (Figure 6h,i).

Collectively, these data indicate that CmMAF2 affects GA levels to regulate floral transition by directly targeting CmGA20ox1.

CmC3H1 directly activates CmMAF2 expression in response to low temperatures

To investigate the upstream regulatory mechanism of CmMAF2, we first used different promoter regions of CmMAF2 to screen the Y1H library. We found that a CCCH‐type zinc‐finger protein‐encoding gene, CmC3H1, was a potential upstream candidate gene. Then, we measured the expression levels of CmC3H1 in leaves and shoot tips. The results showed that as a general trend, the expression pattern of CmC3H1 was consistent with that of CmMAF2 (Figure 3a,b). Namely, the expression of CmC3H1 was upregulated during low‐temperature treatments and then downregulated after returning to normal temperatures in both leaves and shoot tips (Figure 7a,b).

Figure 7.

Figure 7

Open in a new tab

CmC3H1 directly activates CmMAF2 expression in response to low temperatures. (a,b) Expression of CmC3H1 was analyzed by qRT‐PCR in leaves (a) and shoot tips (b). (c) Interaction between CmC3H1 and different regions in the CmMAF2 promoter, as shown by a yeast one‐hybrid assay. All interactions were examined on SD/−Leu medium supplemented with 100 mg μl−1 AbA. (d) ChIP analysis of the indicated fragments (P1–P7) in the CmMAF2 promoter. The chromatin of pSuper::CmC3H1‐GFP chrysanthemum plants was immunoprecipitated with an anti‐GFP antibody and pSuper::GFP chrysanthemum plants served as a negative control. The amount of the indicated DNA fragment was determined by qRT‐PCR and normalized to the pSuper::GFP control (set to 1 for each fragment). (e,f) Interaction between CmC3H1 and the CmMAF2 promoter, as shown using a dual‐luciferase reporter assay in Nicotiana benthamiana leaves. A 774‐bp CmMAF2 promoter sequence was used. Representative photographs of firefly luciferase fluorescence signals are shown in (e) and the relative LUC/REN ratio is shown in (f). (g) Transcript abundance of CmC3H1 in transiently CmC3H1‐silenced chrysanthemum plants. (h) Representative photographs of WT plants infected with CaLCuV or CaLCuV‐amiR‐CmC3H1 after 14 weeks of flower blooming under LD conditions. (i) Days to initial flower bud emergence were recorded (n > 5). (j) Expression of CmMAF2 and CmGA20ox1 in transiently CmC3H1‐silenced chrysanthemum plants. Error bars indicate standard deviation. Asterisks indicate significant differences according to a Student's t‐test (**P < 0.01). Red scale bars, 1 cm; white scale bars, 5 cm.

Next, we performed a Y1H assay to confirm the interaction between CmC3H1 and the CmMAF2 promoter. The results showed CmC3H1 bound to the third fragment of the CmMAF2 promoter (Figure 7c). A chromatin immunoprecipitation assay coupled with PCR (ChIP‐PCR) showed that CmC3H1 independently bound to the P5 or P6 fragment of the CmMAF2 promoter (Figure 7d). We also assayed the activity of a dual‐luciferase reporter to evaluate the regulatory activity of CmC3H1 on the CmMAF2 promoter in vivo. Nicotiana benthamiana leaf cells co‐expressing CmC3H1 and the CmMAF2 promoter had a significantly higher LUC/REN ratio than leaf cells transformed with CmC3H1 or the CmMAF2 promoter alone, suggesting that CmC3H1 directly activated the expression of CmMAF2 (Figure 7e,f).

Finally, to verify the involvement of CmC3H1 in regulation of flowering, we transiently silenced CmC3H1 in chrysanthemum plants (Figure 7g). We observed that under LD conditions, compared with control plants, both the time to initial flower bud emergence and the time to blooming were significantly reduced in CmC3H1 silencing plants (Figure 7h,i). In addition, we evaluated the expression of CmMAF2 and CmGA20ox1. Compared with the control plants, CmMAF2 expression was significantly downregulated, while CmGA20ox1 expression was significantly upregulated in CmC3H1 silenced plants (Figure 7j).

Together, these results demonstrate that CmC3H1 could regulate the expression of CmMAF2 positively and CmGA20ox1 negatively, suggesting that the C3H1–MAF2 module might regulate flowering in response to low temperature by repressing GA biosynthesis in temperature‐sensitive chrysanthemum ecotypes.

DISCUSSION

Over time, chrysanthemum has evolved into multiple ecotypes, some of which are more sensitive to the photoperiod, while others are more sensitive to temperature. On the one hand, the photoperiod influences flowering time based on the critical day length for flowering (Kawata, 1987), mainly through affecting the expression of BBX24 and AFT/FTL3 in the gene regulatory network (Higuchi et al., 2013; Yang et al., 2014). On the other hand, low temperatures affect floral induction via two different pathways. One of the pathways is that low temperatures inhibit floral induction, and low temperature‐treated plants appeared to be less sensitive to SD conditions (Hisamatsu et al., 2017; Vegis, 1964). The other pathway is that prolonged low‐temperature exposure can break the dormant state of the meristem and induce flowering. The low‐temperature requirement for flowering competence is also called ‘vernalization’ (Schwabe, 1950). However, little is known about the underlying molecular mechanisms by which temperature‐sensitive chrysanthemum initiates flowering in response to low temperatures. Our current work provides a novel molecular framework: the C3H1–MAF2 module responds to low temperatures to induce flowering in temperature‐sensitive chrysanthemum. Under low‐temperature conditions, CmMAF2 expression was upregulated by CmC3H1 to repress the expression of CmGA20ox1 and reduced the levels of bioactive GAs. When returning to warm temperatures, CmMAF2 expression was reduced, the expression of CmGA20ox1 was induced, and high bioactive GA levels rapidly induced CmLFY expression, thereby initiating floral transition (Figure 8).

Figure 8.

Figure 8

Open in a new tab

A schematic model describing the C3H1–MAF2 module and its response to low temperatures to induce flowering in temperature‐sensitive chrysanthemum ecotypes. Under low‐temperature conditions, CmMAF2 expression is upregulated by CmC3H1 to repress the expression of CmGA20ox1, reducing bioactive GA levels. When returning to warm temperatures, CmMAF2 expression is reduced to release the expression of CmGA20ox1, and high bioactive GA levels rapidly induce CmLFY expression, thereby initiating floral transition.

Cys3His‐type zinc finger proteins have been reported to play critical roles in a broad range of processes, such as plant growth, development, and stress responses, in Arabidopsis (Bogamuwa & Jang, 2014). However, the mechanisms underlying the roles of C3H1 in regulating flowering time in response to low temperatures remained unknown. In the present work, CmC3H1 expression was significantly upregulated upon exposure to low temperatures in a temperature‐sensitive chrysanthemum ecotype. CmC3H1 activated the expression of the downstream gene CmMAF2 by directly binding to its promoter region. Both CmC3H1 and CmMAF2 expression decreased rapidly after returning to warm temperatures. Hence, we conclude that CmC3H1 acts directly upstream of CmMAF2 and plays important roles in regulating flowering time in response to low temperatures.

Floral transition in the low temperature‐sensitive chrysanthemum ecotype requires low‐temperature exposure, which is also referred to as ‘vernalization’ (Schwabe, 1950). In A. thaliana, AtFLC expression was reported to be gradually downregulated during vernalization, and the flowering process was completed at these low expression levels after low‐temperature treatments (Whittaker & Dean, 2017). AtMAF2 is considered to be one of the close homologs to AtFLC, which regulates flowering time in response to vernalization (Gu et al., 2013; Ratcliffe et al., 2003). However, our present results show that the expression pattern of CmMAF2 in response to low temperatures is completely different from that of AtFLC. In temperature‐sensitive chrysanthemum, CmMAF2 expression was upregulated during low‐temperature treatment and downregulated when normal temperatures were sensed (Figure 3a,b). Besides, CmMAF2 can directly target the GA biosynthesis gene CmGA20ox1 and regulate the biosynthesis of bioactive GAs. After long‐term exposure to low temperatures, the plants showed early flowering due to increased levels of bioactive GAs (Figure 1d,e; Figure 2a). Prolonged exposure to low temperature appears to induce flowering in this ecotype. Application of exogenous GAs can promote flowering in chrysanthemum (Figure 2b–d) (Yang et al., 2014). On the contrary, reducing GA levels with PAC delays flowering (Figure 5b–d). Hence, GA is a prominent phytohormone promoting floral induction in the temperature‐sensitive chrysanthemum ecotype. Meanwhile, GA can circumvent the prolonged low‐temperature requirements for flowering.

The relationship between low temperatures and GA levels has been elucidated in several plant species. For example, in radish (Raphanus sativus) and field pennycress (Thlaspi arvense), the application of exogenous GA3 can substitute for the vernalization requirement (Metzger, 1985; Suge & Rappaport, 1968), and in winter canola, GA levels increase at the end of vernalization (Zanewich & Rood, 1995). In the herbaceous perennial A. alpine, PEP1, another MAF2 homolog, regulates GA metabolism by binding to GA2ox rather than GA20ox. Additionally, in A. alpine, GA does not overcome the vernalization requirement but reduces the duration of the vernalization time needed. Bioactive GA1 levels are unchanged during or after vernalization, whereas GA4 levels are reduced during vernalization and then increase after vernalization (Tilmes et al., 2019). Our experiments showed that in the temperature‐sensitive chrysanthemum ecotype, CmMAF2 regulates GA metabolism under low‐temperature conditions by binding to CmGA20ox1, causing a reduction in the levels of both bioactive GA1 and GA4. Meanwhile, in CmMAF2‐silenced transgenic plants, bioactive GA1 and GA4 levels increased, contributing to early flowering (Figure 5a). When GA levels were reduced by PAC treatment, the transgenic plants showed reversed phenotypes. Hence, in the temperature‐sensitive chrysanthemum ecotype, both GA1 and GA4 levels affect floral induction. Collectively, the GA metabolism in response to low temperatures is distinct between chrysanthemum and other species.

It has been reported that three FT‐like genes (CsFTL1, CsFTL2, and CsFTL3) play important roles in photoperiod‐sensitive chrysanthemum, and their expression is induced by SD signals regulating floral transition via the photoperiod pathway (Higuchi et al., 2013). However, our present results showed that in temperature‐sensitive chrysanthemum, CmLFY is the critical floral gene responsible for integrating the low‐temperature and GA pathways.

MATERIALS AND METHODS

Plant materials and growth conditions

Chrysanthemum (C. morifolium cv. Summer Yellow), which has a low‐temperature sensitivity, was used in this study. Plants were propagated on half‐strength Murashige and Skoog (1/2 MS) medium for 30 days before being transplanted into 9 cm diameter pots containing a 3:1 (v/v) mixture of peat and vermiculite. For LD conditions, plants were grown in a culturing room at 23 ± 1°C, with a relative humidity of 40% and 100 μmol m−2 sec−1 illumination with fluorescent lamps (16 h light/8 h dark). For SD treatments, plants were transferred to a culturing room with an 8 h light/16 h dark regime and the same humidity and light quality.

Treatments

For low‐temperature treatments, plants were transferred to a 4 ± 1°C chamber under either LD conditions for 5 weeks or SD conditions for 8 weeks and then transferred back to normal temperature under SD or LD conditions. To observe the phenotype in the field, the plants were treated with 5 weeks of low temperature (4 ± 1°C) in a chamber and then transferred to the Shangzhuang Experimental Station of China Agricultural University in May (Beijing, China). For GA treatments, plants were sprayed after transplanting with 10 μm or 100 μm GA4/7 (Sigma Aldrich, Saint Louis, MO, USA) and grown under either LD or SD conditions at 23 ± 1°C. GA4/7 was dissolved in 10% ethanol, and the same ethanol concentration was used as a mock control. Plants were sprayed with GA4/7 twice a week for 2 months. For PAC treatment, both CmMAF2‐RNAi transgenic and WT plants were sprayed with 250 mg L−1 PAC (Sigma Aldrich) twice a week for 5 weeks. PAC was dissolved in 1% dimethyl sulfoxide (DMSO), and the same concentration of DMSO without PAC was used as a mock control.

Gene cloning

We obtained full‐length CmMAF2, CmC3H1, and CmGA20ox1 cDNA using the SMART™ RACE cDNA amplification kit (Clontech, Shiga‐ken, Japan) according to the manufacturer's instructions. Genomic DNA was extracted from chrysanthemum leaves using the cetyltrimethyl ammonium bromide method (Huang et al., 2000), and promoters were cloned using inverse PCR and thermal asymmetric interlaced PCR with 2X M5 HiPer plus Taq HiFi PCR mix (Mei5 Biotech, China) (Liu & Chen, 2007). All primers used are listed in Table S1.

Sequence analysis

The conserved MADS‐box and zinc‐finger domains were predicted using InterProScan (Jones et al., 2014). Multiple amino acid sequences were aligned using ClustalW (Thompson et al., 1994) with default parameters, and the result is shown in BioEdit format (Hall, 1999). The phylogenetic tree was constructed using the neighbor‐joining algorithm with 1000 bootstrap replicates using MEGA 5.0 software (Tamura et al., 2011). The protein sequences of CmMAF2 homologs were identified in the National Center for Biotechnology Information (NCBI) database. The cis‐elements in the CmGA20ox1 and CmGA2ox1 promoters were analyzed using the PlantCARE program (Lescot et al., 2002).

qRT‐PCR analysis

Tips of the fourth expanded leaves were collected from five replicates for each sample at Zeitgeber time 10 (ZT10). Total RNA was extracted using the RNAiso Plus reagent (TaKaRa, Japan) according to the manufacturer's instructions. One microgram of total RNA was reverse transcribed to cDNA using the HiScript II Q Select RT SuperMix for qPCR (+gDNA wiper) kit (Vazyme Biotech Co., Ltd, China), and qRT‐PCR (in a volume of 20 μl containing 2 μl cDNA as the template) was carried out using the StepOne Real‐Time PCR System (Applied Biosystems, USA) in standard mode with a KAPA SYBR FAST Universal qRT‐PCR Kit (Kapa Biosystems, USA). The chrysanthemum UBIQUITIN gene (GenBank accession NM_112764) was used as an internal control. Relative expression was calculated using the 2−ΔΔCt method (Livak & Schmittgen, 2001). The gene‐specific primers are listed in Table S1.

Chrysanthemum transformation

To generate the RNAi vector, 305‐bp CmMAF2‐specific sense and antisense fragments were amplified using two pairs of primers containing either AscI/SwaI or PacI/BamHI sites. The resulting PCR products were digested with the abovementioned restriction enzymes and then inserted into the binary vector PFGC1008 digested with the same enzymes to form an intron‐containing ‘hairpin’ RNA construct containing the cauliflower mosaic virus 35S promoter. This plasmid was then introduced into Agrobacterium tumefaciens strain EHA105 by the freeze–thaw method and transformed into chrysanthemum by A. tumefaciens‐mediated transformation. Each explant for transformation (1.0 cm × 1.0 cm square leaf) was derived from a young chrysanthemum leaf and pre‐cultured on MS medium for 3 days. After agro‐infiltration, the explants were placed under dark conditions for 2 days. Next, the infected explants were washed gently in washing medium (4.4 g L−1 MS medium, 30 g L−1 sucrose, 200 mg L−1 cefotaxime, pH 5.8) and transferred to shoot‐inducing medium (4.4 g L−1 MS medium, 30 g L−1 sucrose, 0.3 g L−1 1‐naphthaleneacetic acid [NAA], 3.0 g L−1 6‐benzylaminopurine [6‐BA], 200 mg L−1 cefotaxime, 1.0 mg L−1 hygromycin B, 6.4 g L−1 agar, pH 5.8) for 3 days. The medium was changed every 2 weeks. Elongated shoots were placed into rooting medium (4.4 g L−1 1/2 MS medium, 30 g L−1 sucrose, 0.02 mg L−1 NAA, 200 mg L−1 cefotaxime, 1.0 mg L−1 hygromycin B, 7.8 g L−1 agar, pH 5.8) after 30 days.

Phenotypic measurements

To evaluate the time until initial flower bud emergence, the day of transplanting was set as day 1. The time of the first visible flower buds (2 mm diameter) was then recorded. The shoot apex and inflorescence were dissected from the chrysanthemum under a light microscope (Leica DFC450, Germany). After dissection, samples were immediately observed by scanning electron microscopy (Hitachi S‐4700, Japan) with an acceleration voltage of 2 kV.

Determination of GA contents

The shoot apex of CmMAF2‐RNAi or WT plants was collected in three replicates at ZT10. GA contents were determined by a commercial company (Metware Biotechnology Co., Ltd., Wuhan, China) using the AB Sciex QTRAP 6500 liquid chromatography–tandem mass spectrometry platform. Raw data were analyzed by Analyst 1.6.3 software (AB Sciex, Waltham, MA, USA).

RNA‐seq analysis

Total RNA samples were extracted from the top four expanded leaves of WT plants grown at low temperature for 5 weeks and after 1 week or 5 weeks of returning to normal conditions. Three biological replicates were collected for each time point. RNA‐seq libraries were prepared (Zhong et al., 2011) and sequenced using the HiSeq 2000 (Illumina) platform at Novogene Co. Ltd. (Beijing, China, http://www.novogene.com/). RNA‐seq data were processed, assembled, and annotated as previously described (Wei et al., 2017).

EMSA

EMSA was performed using the LightShift Chemiluminescent EMSA Kit (Thermo Fisher) according to the manufacturer's instructions with minor modifications. Briefly, the PGEX‐4 T‐2‐CmMAF2 recombinant vector was used to express the GST‐CmMAF2 fusion protein before purification as previously described (Liu et al., 2017). Next, 40 nmol of labeled WT or mutated probes was incubated with 2 μg of protein in 20 μl reaction buffer (2 μl of 10× binding buffer, 1 μl of 1 μg μl−1 poly(dI:dC), 1 μl of 50% glycerol, 1 μl of 1% NP‐40, and 10 mm ethylenediaminetetraacetic acid) at 25°C for 30 min. Complementary biotin‐labeled, mutated, and unlabeled 3′ end DNA oligonucleotides were synthesized and annealed. Unlabeled DNA was used as a competitor probe. The probe sequences are shown in Table S1.

Yeast one‐hybrid assay

Protein and DNA interactions in yeast cells were determined using the Matchmaker™ Gold Yeast One‐Hybrid Library Screening System (Clontech, Japan). To test whether CmMAF2 binds to the promoter of CmGA20ox1, the CmMAF2 full‐length sequence was inserted into the pGADT7 vector, and the CmGA20ox1 promoter and mutated fragments were inserted into the pAbAi vector (Clontech, Japan). Interactions were examined on SD/−Leu medium with 100 mg μl−1 aureobasidin A (AbA) (Clontech, Japan). To identify the upstream gene of CmMAF2, we created a Y1H library using high‐quality chrysanthemum cDNA. Then, the promoter sequence of CmMAF2 was divided into three fragments and inserted into the pAbAi vector. After screening the Y1H library, we confirmed the screening results. Primers used are shown in Table S1.

Dual‐luciferase reporter assay in N. benthamiana

To investigate whether CmMAF2 regulates CmGA20ox1 directly in vivo, we used the pGreenII 0800‐LUC and pGreenII 0029 62‐SK vectors (Hellens et al., 2005). A 680‐bp CmGA20ox1 promoter sequence and the same sequence with three mutated cArG‐boxes were inserted into the BamHI/NcoI sites of the pGreenII 0800‐LUC vector. The CmMAF2 coding sequence was amplified by PCR and inserted into the pGreenII 62‐SK vector using BamHI and KpnI. Similarly, to investigate whether CmC3H1 regulates CmMAF2 directly in vivo, a 774‐bp CmMAF2 promoter sequence was inserted into the HindIII/BamHI sites of the pGreenII 0800‐LUC vector, and the CmC3H1 coding sequence was inserted into the pGreenII 62‐SK vector using EcoRI and KpnI.

All constructs were transformed into A. tumefaciens strain GV3101 harboring the pMP90 and pSoup plasmids (Hellens et al., 2000). Nicotiana benthamiana plants with three to five young leaves grown at 22°C under LD conditions were used as materials. Mixtures of A. tumefaciens cultures expressing either the coding sequence or the promoter fragments (v:v, 1:5) were infiltrated into N. benthamiana leaves using a needleless syringe (Wei et al., 2017). LUC and REN activities were measured using dual‐luciferase reporter assay reagents (Promega, USA) and a GloMax 20/20 luminometer (Promega, USA). The ratios of LUC and REN are expressed as activation or repression. The LUC images were taken using an iKon‐L936 imaging system (Andor Tech, Belfast, UK).

Chromatin immunoprecipitation (ChIP) assay

Chromatin immunoprecipitation experiments were carried out following a previously described protocol (Saleh et al., 2008). The full‐length sequence of CmC3H1 without the stop codon was inserted into the pSuper1300 (GFP‐C) vector using XbaI and KpnI. The resulting constructs and empty vector control were separately introduced into A. tumefaciens strain EHA105. Afterward, Agrobacterium cultures were harvested by centrifugation, resuspended in infiltration buffer (10 mm MES, 10 mm MgCl2, and 200 mm AS, pH 5.6) to a final OD600 of 1.0, and infiltrated into chrysanthemum leaves using a needleless syringe. After 3 days, approximately 1.5 g of young leaves was fixed by incubation in 1% formaldehyde under vacuum for 10 min. The reaction was stopped by adding 2.5 ml 2 m glycine (0.125 mm final concentration) for another 5 min under vacuum. The leaves were washed twice with deionized water and frozen in liquid nitrogen. Chromatin was then extracted and sonicated, followed by overnight immunoprecipitation using anti‐GFP (BE2001, Easybio, Beijing, China) and Magna ChIP™ Protein A + G Magnetic Beads (EMD Millipore, USA). The co‐precipitated DNA was purified with a QIAquick PCR Purification Kit (Qiagen GmbH, Germany). qRT‐PCR was conducted to measure the enrichment of DNA fragments. Primers are listed in Table S1.

In situ hybridization

Shoot apices were fixed in 3.7% formalin–acetic acid–alcohol overnight. Specific CmGA20ox1 probes were designed according to the 3′ untranslated region. Sense and antisense probes were synthesized using SP6 and T7 RNA polymerase, respectively. In situ hybridization experiments were performed as previously described (Zhang et al., 2013). Primers are listed in Table S1.

Virus‐induced gene silencing

To silence CmGA20ox1 and CmC3H1 in chrysanthemum, a previously reported virus‐based microRNA expression system was used (Tang et al., 2010; Xu et al., 2020). A modified CaLCuV vector containing pre‐cmo‐GA20ox1 (CaLCuV + GA20ox1) and pre‐cmo‐C3H1 were generated and introduced into A. tumefaciens strain GV3101. The transformed A. tumefaciens cultures were inoculated overnight in LB medium and resuspended in infiltration buffer to a final OD600 of 1.5.

Then, the cultures containing pCVB and CaLCuV‐ GA20ox1 or pCVB and CaLCuV (control) were mixed in a 1:1 ratio (v/v) and incubated in the dark at 28°C for 3–4 h before vacuum infiltration. Fifty‐day‐old WT and RNAi‐CmMAF2 plants immersed in infiltration buffer were vacuum‐treated (−0.7 MPa) for 3 min. Similar to the previous treatment, 50‐day‐old WT plants immersed in infiltration buffer containing pCVB and CaLCuV‐ C3H1 or pCVB and CaLCuV (control) were vacuumed (−0.7 MPa) for 3 min. Then, the plants were placed in the dark at 8°C for 3 days and transplanted into pots filled with a 1:1 (v/v) mixture of peat:vermiculite and grown at 23 ± 1°C under LD conditions. The silenced plants were validated by RT‐qPCR to determine the expression of CmGA20ox1 or CmC3H1. Three independent experiments were performed, and at least six positive plantlets were used to observe the phenotypes.

ACCESSION NUMBERS

Sequences can be found under the following accession numbers: CmMAF2 (MN255845), CmC3H1 (OM963136), the promoter sequence of CmGA20ox1 (MN255846), AtFLC (AT5G10140), AtMAF1 (AT1G77080), AtMAF2 (AT5g65050), AtMAF3 (AT5G65060), AtMAF4 (AT5G65050), AtMAF5 (AT5G65080), RsFLC (AJN00653), AtSVP (AT2G22540), AtFUL (AT5G60910), BvFL1 (DQ189210), BnFLC (AFU61576).

AUTHOR CONTRIBUTIONS

JL, XZ, and BH conceived and designed the experiments; JL performed most of the experiments; YZ and ZW contributed to the chrysanthemum transformation; AP, YX, HH, and RZ contributed to GA treatment experiments; AP observed apex development using scanning electron microscopy; AP and ZG conducted real‐time PCR experiments; CM, CJ, SG, JG, and BH provided technical support and conceptual advice; JL and XZ analyzed the data; BH and JL wrote the manuscript.

CONFLICT OF INTEREST

The authors declare that there is no conflict of interest.

Supporting information

Data S1. Differentially expressed genes in chrysanthemum after 5 weeks of low‐temperature treatment (LT5).

Click here for additional data file. (914.6KB, xls)

Figure S1. Effects of low temperature on flowering under short‐day (SD) conditions and scanning electron micrographs showing apex development under low‐temperature, long‐day (LD) conditions.

Figure S2. Effects of exogenous gibberellin (GA4/7) treatment on flowering under SD conditions.

Figure S3. CmMAF2 sequence alignment and CmGA20ox2 expression.

Figure S4. CmGAox expression at different developmental stages.

Figure S5. cis‐Element analyses of the CmGA2ox promoter (approximately 900 bp upstream from the start codon).

Click here for additional data file. (3.8MB, docx)

Table S1. Primers used in this study.

Click here for additional data file. (16.3KB, xlsx)

ACKNOWLEDGMENTS

We thank PlantScribe (www.plantscribe.com) and Dr. Puneet Paul (University of Nebraska Lincoln) for carefully editing this manuscript. This work was supported by the National Key Research and Development Program of China (Grant No. 2018YFD1000400), the National Natural Science Foundation of China (Grant Nos. 32030096, 31772347, and 31822045), and the Construction of Beijing Science and Technology Innovation and Service Capacity in Top Subjects (CEFF‐PXM2019_014207_000032).

Contributor Information

Xin Zhao, Email: zhaoxin5691@cau.edu.cn.

Bo Hong, Email: hongbo1203@cau.edu.cn.

DATA AVAILABILITY STATEMENT

All data included in this study are available upon request by contact with the corresponding author.

REFERENCES

  1. Amasino, R. (2004) Vernalization, competence, and the epigenetic memory of winter. The Plant Cell, 16, 2553–2559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Amasino, R. (2009) Floral induction and monocarpic versus polycarpic life histories. Genome Biology, 10, 228. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barreda, V.D. , Palazzesi, L. , Telleria, M.C. , Olivero, E.B. , Raine, J.I. & Forest, F. (2015) Early evolution of the angiosperm clade Asteraceae in the cretaceous of Antarctica. Proceedings of the National Academy of Sciences of the United States of America, 112, 10989–10994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Binenbaum, J. , Weinstain, R. & Shani, E. (2018) Gibberellin localization and transport in plants. Trends in Plant Science, 23, 410–421. [DOI] [PubMed] [Google Scholar]
  5. Bogamuwa, S.P. & Jang, J.C. (2014) Tandem CCCH zinc finger proteins in plant growth, development and stress response. Plant & Cell Physiology, 55, 1367–1375. [DOI] [PubMed] [Google Scholar]
  6. Eriksson, S. , Bohlenius, H. , Moritz, T. & Nilsson, O. (2006) GA4 is the active gibberellin in the regulation of LEAFY transcription and Arabidopsis floral initiation. The Plant Cell, 18, 2172–2181. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Gu, X. , Le, C. , Wang, Y. , Li, Z. , Jiang, D. , Wang, Y. et al. (2013) Arabidopsis FLC clade members form flowering‐repressor complexes coordinating responses to endogenous and environmental cues. Nature Communications, 4, 1947. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Hall, T.A. (1999) BioEdit: a user‐friendly biological sequence alignment editor and analysis program for windows 95/98/NT. Nucleic Acids Symposium Series, 41, 95–98. [Google Scholar]
  9. Harada, H. & Nitsch, J. (1959) Flower induction in Japanese chrysanthemums with gibberellic acid. Science, 129, 777–778. [DOI] [PubMed] [Google Scholar]
  10. Hazebroek, J.P. , Metzger, J.D. & Mansager, E.R. (1993) Thermoinductive regulation of gibberellin metabolism in Thlaspi arvense L. (II. Cold induction of enzymes in gibberellin biosynthesis). Plant Physiology, 102, 547–552. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hellens, R.P. , Allan, A.C. , Friel, E.N. , Bolitho, K. , Grafton, K. , Templeton, M.D. et al. (2005) Transient expression vectors for functional genomics, quantification of promoter activity and RNA silencing in plants. Plant Methods, 1, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Hellens, R.P. , Edwards, E.A. , Leyland, N.R. , Bean, S. & Mullineaux, P.M. (2000) pGreen: a versatile and flexible binary Ti vector for agrobacterium‐mediated plant transformation. Plant Molecular Biology, 42, 819–832. [DOI] [PubMed] [Google Scholar]
  13. Higuchi, Y. , Narumi, T. , Oda, A. , Nakano, Y. , Sumitomo, K. , Fukai, S. et al. (2013) The gated induction system of a systemic floral inhibitor, antiflorigen, determines obligate short‐day flowering in chrysanthemums. Proceedings of the National Academy of Sciences of the United States of America, 110, 17137–17142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hisamatsu, T. , Koshioka, M. & Mander, L.N. (2004) Regulation of gibberellin biosynthesis and stem elongation by low temperature in Eustoma grandiflorum. The Journal of Horticultural Science and Biotechnology, 79, 354–359. [Google Scholar]
  15. Hisamatsu, T. , Sumitomo, K. , Shibata, M. & Koshioka, M. (2017) Seasonal variability in dormancy and flowering competence in chrysanthemum: chilling impacts on shoot extension growth and flowering capacity. Japan Agricultural Research Quarterly: JARQ, 51, 343–350. [Google Scholar]
  16. Huang, J.C. , Ge, X.J. & Sun, M. (2000) Modified CTAB protocol using a silica matrix for isolation of plant genomic DNA. BioTechniques, 28, 432–434. [DOI] [PubMed] [Google Scholar]
  17. Huang, P. , Chung, M.S. , Ju, H.W. , Na, H.S. , Lee, D.J. , Cheong, H.S. et al. (2011) Physiological characterization of the Arabidopsis thaliana oxidation‐related zinc finger 1, a plasma membrane protein involved in oxidative stress. Journal of Plant Research, 124, 699–705. [DOI] [PubMed] [Google Scholar]
  18. Huang, P. , Ju, H.W. , Min, J.H. , Zhang, X. , Chung, J.S. , Cheong, H.S. et al. (2012) Molecular and physiological characterization of the Arabidopsis thaliana oxidation‐related zinc finger 2, a plasma membrane protein involved in ABA and salt stress response through the ABI2‐mediated signaling pathway. Plant & Cell Physiology, 53, 193–203. [DOI] [PubMed] [Google Scholar]
  19. Jones, P. , Binns, D. , Chang, H.Y. , Fraser, M. , Li, W. , McAnulla, C. et al. (2014) InterProScan 5: genome‐scale protein function classification. Bioinformatics, 30, 1236–1240. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kawata, J. (1987) The phasic development of chrysanthemum as a basis for the regulation of vegetative growth and flowering in Japan. Acta Horticulturae, 197, 280–287. [Google Scholar]
  21. King, R.W. (2012) Mobile signals in day length‐regulated flowering: gibberellins, flowering locus T, and sucrose. Russian Journal of Plant Physiology, 59, 479–490. [Google Scholar]
  22. Langridge, J. (1957) Effect of day‐length and gibberellic acid on the flowering of Arabidopsis. Nature, 180, 36–37. [Google Scholar]
  23. Lee, D.J. & Zeevaart, J.A. (2007) Regulation of gibberellin 20‐oxidase1 expression in spinach by photoperiod. Planta, 226, 35–44. [DOI] [PubMed] [Google Scholar]
  24. Lee, S.J. , Jung, H.J. , Kang, H. & Kim, S.Y. (2012) Arabidopsis zinc finger proteins AtC3H49/AtTZF3 and AtC3H20/AtTZF2 are involved in ABA and JA responses. Plant & Cell Physiology, 53, 673–686. [DOI] [PubMed] [Google Scholar]
  25. Lescot, M. , Dehais, P. , Thijs, G. , Marchal, K. , Moreau, Y. , Van de Peer, Y. et al. (2002) PlantCARE, a database of plant cis‐acting regulatory elements and a portal to tools for in silico analysis of promoter sequences. Nucleic Acids Research, 30, 325–327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lin, P.C. , Pomeranz, M.C. , Jikumaru, Y. , Kang, S.G. , Hah, C. , Fujioka, S. et al. (2011) The Arabidopsis tandem zinc finger protein AtTZF1 affects ABA‐ and GA‐mediated growth, stress and gene expression responses. The Plant Journal, 65, 253–268. [DOI] [PubMed] [Google Scholar]
  27. Liu, J. , Fan, Y. , Zou, J. , Fang, Y. , Wang, L. , Wang, M. et al. (2017) A RhABF2/ferritin module affects rose (Rosa hybrida) petal dehydration tolerance and senescence by modulating iron levels. The Plant Journal, 92, 1157–1169. [DOI] [PubMed] [Google Scholar]
  28. Liu, Y.G. & Chen, Y. (2007) High‐efficiency thermal asymmetric interlaced PCR for amplification of unknown flanking sequences. Biotechniques, 43, 649–650 652, 654 passim. [DOI] [PubMed] [Google Scholar]
  29. Livak, K.J. & Schmittgen, T.D. (2001) Analysis of relative gene expression data using real‐time quantitative PCR and the 2−ΔΔCT method. Methods, 25, 402–408. [DOI] [PubMed] [Google Scholar]
  30. Metzger, J.D. (1985) Role of gibberellins in the environmental control of stem growth in Thlaspi arvense L. Plant Physiology, 78, 8–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Michaels, S.D. & Amasino, R.M. (1999) FLOWERING LOCUS C encodes a novel MADS domain protein that acts as a repressor of flowering. The Plant Cell, 11, 949–956. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Mutasa‐GÖttgens, E. & Hedden, P. (2009) Gibberellin as a factor in floral regulatory networks. Journal of Experimental Botany, 60, 1979–1989. [DOI] [PubMed] [Google Scholar]
  33. Pharis, R.P. (1972) Flowering of chrysanthemum under non‐inductive long days by gibberellins and N6‐benzyladenine. Planta, 105, 205–212. [DOI] [PubMed] [Google Scholar]
  34. Ratcliffe, O.J. , Kumimoto, R.W. , Wong, B.J. & Riechmann, J.L. (2003) Analysis of the Arabidopsis MADS AFFECTING FLOWERING gene family: MAF2 prevents vernalization by short periods of cold. The Plant Cell, 15, 1159–1169. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ream, T.S. , Woods, D.P. & Amasino, R.M. (2012) The molecular basis of vernalization in different plant groups. Cold Spring Harbor Symposia on Quantitative Biology, 77, 105–115. [DOI] [PubMed] [Google Scholar]
  36. Rieu, I. , Ruiz‐Rivero, O. , Fernandez‐Garcia, N. , Griffiths, J. , Powers, S.J. , Gong, F. et al. (2008) The gibberellin biosynthetic genes AtGA20ox1 and AtGA20ox2 act, partially redundantly, to promote growth and development throughout the Arabidopsis life cycle. The Plant Journal, 53, 488–504. [DOI] [PubMed] [Google Scholar]
  37. Saleh, A. , Alvarez‐Venegas, R. & Avramova, Z. (2008) An efficient chromatin immunoprecipitation (ChIP) protocol for studying histone modifications in Arabidopsis plants. Nature Protocols, 3, 1018–1025. [DOI] [PubMed] [Google Scholar]
  38. Schwabe, W.W. (1950) Factors controlling flowering of the chrysanthemum: I. the effects of photoperiod and temporary chilling. Journal of Experimental Botany, 1, 329–343. [Google Scholar]
  39. Schwabe, W.W. (1954) Factors controlling flowering in the chrysanthemum: IV. The site of vernalization and translocation of the stimulus. Journal of Experimental Botany, 5, 389–400. [Google Scholar]
  40. Song, Y.H. , Ito, S. & Imaizumi, T. (2013) Flowering time regulation: photoperiod‐ and temperature‐sensing in leaves. Trends in Plant Science, 18, 575–583. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Srikanth, A. & Schmid, M. (2011) Regulation of flowering time: all roads lead to Rome. Cellular and Molecular Life Sciences, 68, 2013–2037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Suge, H. & Rappaport, L. (1968) Role of gibberellins in stem elongation and flowering in radish. Plant Physiology, 43, 1208–1214. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Sumitomo, K. , Li, T.P. & Hisamatsu, T. (2009) Gibberellin promotes flowering of chrysanthemum by upregulating CmFL, a chrysanthemum FLORICAULA/LEAFY homologous gene. Plant Science, 176, 643–649. [Google Scholar]
  44. Tamura, K. , Peterson, D. , Peterson, N. , Stecher, G. , Nei, M. & Kumar, S. (2011) MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Molecular Biology and Evolution, 28, 2731–2739. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Tang, Y. , Wang, F. , Zhao, J.P. , Xie, K. , Hong, Y.G. & Liu, Y.L. (2010) Virus‐based microRNA expression for gene functional analysis in plants. Plant Physiology, 153, 632–641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Thompson, J.D. , Higgins, D.G. & Gibson, T.J. (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position‐specific gap penalties and weight matrix choice. Nucleic Acids Research, 22, 4673–4680. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Tilmes, V. , Mateos, J.L. , Madrid, E. , Vincent, C. , Severing, E. , Carrera, E. et al. (2019) Gibberellins act downstream of Arabis PERPETUAL FLOWERING1 to accelerate floral induction during vernalization. Plant Physiology, 180, 1549–1563. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Vegis, A. (1964) Dormancy in higher plants. The Annual Review of Plant Biology, 15, 185–224. [Google Scholar]
  49. Wang, R.H. , Farrona, S. , Vincent, C. , Joecker, A. , Schoof, H. , Turck, F. et al. (2009) PEP1 regulates perennial flowering in Arabis alpina . Nature, 459, 423–427. [DOI] [PubMed] [Google Scholar]
  50. Wei, Q. , Ma, C. , Xu, Y. , Wang, T. , Chen, Y. , Lu, J. et al. (2017) Control of chrysanthemum flowering through integration with an aging pathway. Nature Communications, 8, 829. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Whittaker, C. & Dean, C. (2017) The FLC locus: a platform for discoveries in epigenetics and adaptation. Annual Review of Cell and Developmental Biology, 33, 555–575. [DOI] [PubMed] [Google Scholar]
  52. Xu, S. & Chong, K. (2018) Remembering winter through vernalisation. Nature Plants, 4, 997–1009. [DOI] [PubMed] [Google Scholar]
  53. Xu, Y.J. , Zhao, X. , Aiwaili, P. , Mu, X.Y. , Zhao, M. , Zhao, J. et al. (2020) A zinc finger protein BBX19 interacts with ABF3 to affect drought tolerance negatively in chrysanthemum. The Plant Journal, 103, 1783–1795. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Yang, Y.J. , Ma, C. , Xu, Y.J. , Wei, Q. , Imtiaz, M. , Lan, H.B. et al. (2014) A zinc finger protein regulates flowering time and abiotic stress tolerance in chrysanthemum by modulating gibberellin biosynthesis. The Plant Cell, 26, 2038–2054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Zanewich, K.P. & Rood, S.B. (1995) Vernalization and gibberellin physiology of winter canola (endogenous gibberellin (GA)) content and metabolism of [3H] GA1 and [3H] GA20 . Plant Physiology, 108, 615–621. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Zhang, X. , Zhou, Y. , Ding, L. , Wu, Z. , Liu, R. & Meyerowitz, E.M. (2013) Transcription repressor HANABA TARANU controls flower development by integrating the actions of multiple hormones, floral organ specification genes, and GATA3 family genes in Arabidopsis. The Plant Cell, 25, 83–101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Zhong, S. , Joung, J.G. , Zheng, Y. , Chen, Y.R. , Liu, B. , Shao, Y. et al. (2011) High‐throughput illumina strand‐specific RNA sequencing library preparation. Cold Spring Harbor Protocols, 8, 940–949. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1. Differentially expressed genes in chrysanthemum after 5 weeks of low‐temperature treatment (LT5).

Click here for additional data file. (914.6KB, xls)

Figure S1. Effects of low temperature on flowering under short‐day (SD) conditions and scanning electron micrographs showing apex development under low‐temperature, long‐day (LD) conditions.

Figure S2. Effects of exogenous gibberellin (GA4/7) treatment on flowering under SD conditions.

Figure S3. CmMAF2 sequence alignment and CmGA20ox2 expression.

Figure S4. CmGAox expression at different developmental stages.

Figure S5. cis‐Element analyses of the CmGA2ox promoter (approximately 900 bp upstream from the start codon).

Click here for additional data file. (3.8MB, docx)

Table S1. Primers used in this study.

Click here for additional data file. (16.3KB, xlsx)

Data Availability Statement

All data included in this study are available upon request by contact with the corresponding author.

Articles from The Plant Journal are provided here courtesy of Wiley

RESOURCES