Abstract
Carrot (Daucus carota L.) is a biennial plant requiring vernalization to induce flowering, but long days can promote its premature bolting and flowering. The basic genetic network controlling the flowering time has been constructed for carrot, but there is limited information on the molecular mechanisms underlying the photoperiodic flowering response. The published carrot genome could provide an effective tool for systematically retrieving the key integrator genes of GIGANTEA (GI), CONSTANS-LIKE (COL), FLOWERING LOCUS T (FT), and SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1) homologues in the photoperiod pathway. In this study, the bolting time of wild species “Songzi” (Ws) could be regulated by different photoperiods, but the orange cultivar “Amsterdam forcing” (Af) displayed no bolting phenomenon. According to the carrot genome and previous de novo transcriptome, 1 DcGI, 15 DcCOLs, 2 DcFTs, and 3 DcSOC1s were identified in the photoperiod pathway. The circadian rhythm peaks of DcGI, DcCOL2, DcCOL5a, and DcCOL13b could be delayed under long days (LDs). The peak value of DcCOL2 in Af (12.9) was significantly higher than that in Ws (6.8) under short day (SD) conditions, and was reduced under LD conditions (5.0). The peak values of DcCOL5a in Ws were constantly higher than those in Af under the photoperiod treatments. The expression levels of DcFT1 in Ws (463.0) were significantly upregulated under LD conditions compared with those in Af (1.4). These responses of DcCOL2, DcCOL5a, and DcFT1 might be related to the different bolting responses of Ws and Af. This study could provide valuable insights into understanding the key integrator genes in the carrot photoperiod pathway.
Similar content being viewed by others
Introduction
The floral transition from a vegetative to a reproductive state is a critically important stage in the lifecycle of plants. Vernalization, photoperiod, ambient temperature, autonomous, gibberellin, and age pathways comprise a sophisticated regulatory network containing multiple endogenous and external factors that control flowering in Arabidopsis1,2. These pathways have their own unique initiations, but eventually converge to downstream key floral integrator genes, such as FLOWERING LOCUS T (FT) and SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1)1,2,3. The FT protein, as a mobile florigen, moves from leaves to the shoot apical meristem (SAM) and interacts with FLOWERING LOCUS D (FD) to form a complex, which upregulates SOC1 to induce the meristem identity genes to reprogram the primordia to form reproductive organs4,5,6,7. SOC1 can interact with the MADS-box transcription factor AGAMOUS-LIKE 24 (ALG24) to provide a positive feedback loop and activate LEAFY (LFY) expression by directly binding to its promoter8,9,10. Through the effect of these integrator genes, the floral meristem-determining genes are consequently activated to control the flowering time8,11. This progress in the model plant Arabidopsis provides an important reference for researching floral molecular mechanisms in other crops11,12. In fact, there are usually more than two or three interconnected pathways involved in regulating flowering under natural conditions3,12,13,14.
In the photoperiod pathway, the functional hierarchy GIGANTEA (GI)–CONSTANS (CO)–FT has been identified, and plays a key role in regulating flowering7,11,15,16,17,18. During the late afternoons of long days (LDs), GI forms a complex with the FLAVIN-BINDING, KELCH REPEAT, F BOX 1 (FKF1) to degrade CYCLING DOF FACTORs (CDFs), which contains the major regulators of CO transcription, and consequently, CO protein accumulates to promote flowering by activating the transcription of FT and TWIN SISTER OF FT (TSF) in the leaf vasculature5,7,8,19,20,21. The degradation complex cannot be formed during short days (SDs) due to GI and FKF1 expression not coinciding20,22. In rice, HEADING DATE 1 (Hd1), which is the homologue of CO, promotes flowering by activating the expression of the FT-like gene HEADING DATE 3a (Hd3a) on SDs, and represses flowering on LDs12,15,23,24. By integrating light and circadian clock signals to regulate the downstream florigen gene, CO plays a central role in the mechanism of photoperiod flowering in Arabidopsis3,11,18. Extensive gene duplication events have occurred in this gene family, resulting in 16 other CO-like (COL) genes with different functions25,26. Among them, COL1 and COL2 have little effect on the flowering time, while COL3, COL4, and COL9 represent flowering repressors27,28,29. The overexpression of COL8 delays flowering under LDs, whereas COL5 promotes flowering30,31. Multiple CO/COL homologues have been identified in different species: 16 in rice, 9 in barley, 13 in sugar beet, and 28 in soybean32,33,34,35. GmCOL5 can rescue the late-flowering phenotype of the co mutant33. OsCOL13 negatively regulates flowering under LD and SD conditions in rice36. There is little evidence to prove that CO homologues are potential regulators of FT-like genes12.
The flowering habits of many domesticated crops have changed greatly following human selection, which provides an important route for studying flowering mechanisms by comparing cultivars with their ancestors and wild relatives37,38,39,40. Carrot (Daucus carota L.) is a biennial plant requiring vernalization to induce flowering, and the seedling is usually not responsive to a low temperature (between 0 and 10 °C) until it has 8–12 leaves41. Moreover, long days can promote premature bolting in carrot42. With a change of cultivation system, carrots can be harvested and abundantly supplied all year round in markets. The premature bolting of carrot has been a serious risk during the winter–spring period or spring cultivation. The seedlings grow under a low temperature for a long time at the early stage and grow under long days at the late stage, this is why some varieties which are not tolerant to bolting are prone to bolting. Selection tolerance to premature bolting has been a constant concern in carrot breeding43,44,45,46. The wild species D. carota subsp. carota Songzi (Ws) is sensitive to flower induction by vernalization and photoperiod, and orange cultivar D. carota var. sativus Amsterdam forcing (Af) is tolerant42. Through a de novo transcriptome comparison of Ws and Af, a basic genetic network controlling the flowering time was constructed, including photoperiod, vernalization, and gibberellin pathways42. Furthermore, an FT-like gene was identified in carrot47. There is substantial evidence that FT homologues have a conserved role in promoting flowering and reflect variation in the copy number in divergent angiosperms12. Despite carrot being one of the 10 most important vegetables, understanding on the molecular mechanisms of COL and FT homologues underlying the photoperiodic flowering response is limited. In this study, new homologues of COL, FT, and SOC1 were retrieved based on the carrot genome48. The structure, phylogenetic relationship, and molecular evolutionary rate variation of GI, COL, FT, and SOC1 homologues were investigated using 21 D. carota var. sativus accessions (as DCS), 4 D. carota subsp. gummifer species and 9 D. carota subsp. carota species (as DCC), and 5 Daucus species (as Dau) (Supplementary Table S1). The circadian rhythms of GI, COL, FT, and SOC1 homologues under different photoperiods and their expression trends during growth in different seasons were analyzed in Ws and Af. DcCOL2 and DcCOL5a were transformed into Arabidopsis thaliana plants to analyze their function in regulating flowering. This study provides an improved understanding of the regulatory network of photoperiodic flowering in carrots.
Results
Phylogenetic and nucleotide diversity of DcGI, DcCOLs, DcFTs, and DcSOC1s
We previously reported a preliminary study of the photoperiod pathway in carrot42. Only one DcGI sequence was retrieved from the carrot genome database and localized at chromosome (Chr.) 1, like Dct293 (Supplementary Table S2). The nucleotide diversity (π) of DcGI in 21 DCS accessions and 13 DCC accessions was significantly lower than that in 5 Dau accessions (Fig. 1, Supplementary Table S1). DcGI endured selective pressures during evolution, according to the neutral tests of Tajima’s D (TD)49 and Fu and Li’s F (FF)50. Protein sequence alignment was performed to further explore their relationships with 27 GI homologues from other species. The tree was divided into four groups, and DcGI was clustered within group I and closely associated with ChGIL, SiGIL, and EgGIL (Fig. 2A).
Although 28 CO/COL sequences were retrieved, only 16 sequences had complete zinc finger B-box 1 and CCT domains25 (Supplementary Table S2). The size of the DcCO sequence was only 297 bp, without complete B-box 1 and CCT domains. DcCOL2, like Dct43207, had two tandem duplications and was localized at Chr. 2. DcCOL4, like Dct43377, was localized at Chr. 1. DcCOL5 had two homologues and was localized at Chrs. 3 and 6, like Dct7859 and Dct20940, respectively. DcCOL13 also had two homologues and was localized at Chrs. 1 and 3, but only DcCOL13b was found in the transcriptome, like Dct3283. DcCOL15, like Dct39974, was localized at Chr. 7. Moreover, 1 DcCOL9, 3 DcCOL10, and 4 DcCOL16 homologues were retrieved, which had not previously been reported in the transcriptome42. DcCOL9 was localized at Chr. 3. Three DcCOL10 homologues with 83.05% identity were localized at Chrs. 4, 5, and 6, respectively. Four DcCOL16 homologues with 55.61% identity had only B-box 1 and CCT domains and were localized at Chrs. 3, 4, 7, and 9, respectively. Except for DcCOL13b, the nucleotide diversity (π) of the other DcCOLs in DCS and DCC was lower than that in Dau (Fig. 1). According to the neutrality tests of TD and FF, DcCOL10a/c, DcCOL13a, and DcCOL16a/d had significantly negative values, which indicated that these genes might be selected with a higher pressure during evolution. Additionally, it was interesting that DcCOL5b, DcCOL11, and DcCOL15 in DCS had significantly positive values, which indicated that these genes might be strongly selected during carrot evolution. Based on the phylogenetic analysis, 15 DcCOLs were divided into three groups (Fig. 2B). DcCOL2, DcCOL4, and DcCOL5a/b were clustered into group I, which were grouped with AtCO/AtCOL1/2, AtCOL3/4, and AtCOL5, respectively. DcCOL16a/b/c/d were clustered into group II and were closely associated with AtCOL6 and AtCOL16. DcCOL9, DcCOL10a/b/c, DcCOL13a/b, and DcCOL15 were assigned to group III and were closely associated with AtCOL9/10, AtCOL13, and AtCOL14/15, respectively.
Four DcFT homologues were retrieved and annotated as Hd3a48, but only two sequences had the complete phosphatidylethanolamine-binding protein (PEBP) domain and were localized at Chrs. 1 and 7, respectively (Supplementary Table S2). Additionally, DcFT1 had a 23.66 kbp large intron by compared its cDNA sequence with carrot genome48 (Fig. 3B), and DcFT2 was the same as DcFT KY768910 (GenBank number)47. The nucleotide diversity (π) of DcFT2 in DCS was significantly lower than that in DCC and Dau, and it might be selected during evolution, according to the neutrality tests of TD and FF (Fig. 1). Twenty-eight FT/Hd3a protein sequences of other species were aligned to explore these gene relationships. The phylogenetic tree was divided into four groups. DcFT1 was assigned to group I and associated with EuHd3a, LsFT, and DlFT1, while DcFT2 was assigned to group II and was closely associated with VvFT, InFTL, and NtFTL (Fig. 2C).
Six DcSOC1 homologues were retrieved, but only four sequences had the complete MADS-box domain, K-box domain, and SOC1/MOTIF (Supplementary Table S2). DcSOC1a has not previously been reported and was localized at Chr. 2. DcSOC1b, like Dct4069, had two tandem duplications and was localized at Chr. 2. DcSOC1c was localized at Chr. 4, like Dct34200. For DcSOC1a, the nucleotide diversity (π) of DCS was significantly lower than that of DCC and Dau, while similar to that of DcSOC1b/c (Fig. 1). Twenty-two SOC1 protein sequences of other species were aligned to explore the relationships. The phylogenetic tree was divided into four groups. DcSOC1a/b/c were assigned to group III and were associated with HaSOC1L (Fig. 2D).
Circadian rhythm of DcGI, DcCOLs, DcFT1, and DcSOC1s under different photoperiods
Based on the sequences and phylogenetic analysis, DcGI, DcCOL2, DcCOL4, DcCOL5a/b, DcCOL13b, DcCOL15, DcFT1, and DcSOC1b/c were screened for a further study of the circadian rhythm. When the seedlings of Ws and Af had 3–4 leaves, the plants were subjected to SD and LD treatments for 27 days, respectively, and half of the seedlings under SD conditions were then subjected to the above LD conditions for 7 days (as SD-LD). Under LD conditions, Ws plants began to bolt for 67 days after germinating, and the bolting ratio reached 100% after 108 days. Under SD-LD and SD conditions, the bolting of Ws was delayed for approximately 5 and 22 days, and the bolting was 81% and 40% after 108 days, respectively. No bolting plant was found in Af under the above treatments (Fig. 3A). Except that of DcCOL13b, which was only expressed in Ws, the other gene open reading frame (ORF) regions were separately cloned from both Ws and Af (Fig. 3B).
The circadian patterns of DcGI expression in Ws and Af were similar to those described for AtGI51 and OsGI15 (Fig. 3C). Under SD conditions, the expression of DcGI began to increase from zeitgeber time (ZT) 0 to 8 and reached trough levels 4 h later, and its peak value was similar for Ws and Af (323.1/334.8). Under SD-LD conditions, the expression of DcGI peaked at ZT 12 and reached trough levels 8 h later, while the peak value (322.3) of DcGI in Ws was higher than that in Af (237.6). The DcGI expression pattern under LD conditions was similar to that under SD-LD conditions, and its peak value was similar for Ws and Af (256.7/281.8).
Under SD conditions, the expression of DcCOL2 peaked at ZT 0 and then reached trough levels at ZT 8 in Ws and Af. Under SD-LD and LD conditions, the expression of DcCOL2 peaked at ZT 4 and then declined to trough levels 4 h later. The peak values (6.8/3.2) of DcCOL2 in Ws were lower than those in Af under both SD and SD-LD conditions (12.9/5.6), while they were similar under LD conditions (4.8/5.0). Under SD conditions, the expression patterns of DcCOL4 were similar in Ws and Af (11.3/10.9) and peaked at ZT 0. Under SD-LD and LD conditions, the expression levels of DcCOL4 in Ws (5.1/5.9) were significantly reduced, but were still higher in Af (12.4/7.7).
Under SD conditions, the circadian patterns of DcCOL5a/b were similar in Ws and Af and peaked at ZT 8. Under SD-LD and LD conditions, the peaks of DcCOL5a in Ws (14.2/17.1) shifted from ZT 8 to 12 and retained a similar level, but were reduced in Af (7.0/8.0). The peaks of DcCOL5b in Ws were advanced from ZT 8 to 4 under SD-LD and LD conditions, while those in Af were similar to those under SD conditions. DcCOL13b was only expressed in Ws, and not in Af. Under SD conditions, the expression of DcCOL13b peaked at ZT 16, but it peaked at ZT 0 under SD-LD and LD conditions. Under SD conditions, the expression level of DcCOL15 in Af (10.5) was higher than that in Ws (5.0) and peaked at ZT 16. Under SD-LD and LD conditions, the peaks of DcCOL15 in Ws (5.0/7.2) were shifted at ZT 12, while those in Af (5.7/6.7) reduced and shifted at ZT 20.
Under SD-LD and LD conditions, the expression levels of DcFT1 in Ws (626.0/463.0) were about 179 and 132 times higher than that under SD conditions (3.5), while its expression in Af (1.4/0.7/1.4) constantly remained at low levels. The expression pattern of DcSOC1b was significantly different from that of DcSOC1c. The expression levels of DcSOC1b in Ws (322.3/290.0/599.0) were about 500, 610, and 371 times higher than those in Af under SD, SD-LD, and LD conditions (0.6/0.5/0.9), respectively. The expression patterns of DcSOC1c in Ws and Af were similar under the treatments.
Trends of DcGI, DcCOLs, DcFT1, and DcSOC1s expressed in spring and autumn
Plants are usually induced to flower through two or three interconnected pathways under natural conditions3,13,14. In order to further understand the functions of DcGI, DcCOLs, DcFT1, and DcSOC1s during carrot growth, their expression levels were analyzed in Ws and Af in spring and autumn. The seeds were sown directly in the field under natural photoperiods and temperature conditions on 19 March (spring) and 4 August (autumn), respectively. In spring, Ws plants began to bolt for about 65 days after germinating and the bolting ratio rapidly reached 80.3% after 79 days after germinating, while in autumn, the bolting plants initiated 39 days after germinating and peaked at 93.3% after 74 days. No bolting plants were observed for Af during these two seasons (Fig. 4A).
The expression level of DcGI in Ws constantly increased, but later decreased in Af during spring, and fluctuated during autumn (Fig. 4B). DcCOL2 in Af maintained a higher expression level than that in Ws at the beginning and fluctuated during spring, while it maintained a constantly higher level in Af than that in Ws during autumn. The expression of DcCOL4 in Ws remained at low levels, but was high in Af during spring, while it fluctuated during autumn. The expression of DcCOL5a in Ws maintained a higher level than that in Af during both spring and autumn, but that of DcCOL5b was similar. DcCOL13b was only expressed in Ws and peaked at the middle stage. The expression patterns of DcFT1 in Ws and Af were different and fluctuated over the course of the two seasons. DcSOC1b in Ws was expressed at a significantly higher level than that in Af during the two seasons, but DcSOC1c showed a similar expression level.
Effect of the overexpression of DcCOL2 and DcCOL5a on the flowering time of Arabidopsis
To further understand the functions of DcCOL2 and DcCOL5a, their sense cDNAs with the CaMV35S promoter were transformed into wild-type Columbia (WT) Arabidopsis plants (Supplementary Fig. S1A). In the 14 35 S::DcCOL2 plants (T1), 13 were detected in the DcCOL2 cDNA products, but one was not detected (Supplementary Fig. S1B). All 12 35 S::DcCOL5a plants (T1) were detected in the DcCOL5 cDNA products. The bolting time of 35S::DcCOL2 plants (T2) was significantly delayed by about 18.1 days, but the rosette leaf number was similar to that of the WT (Fig. 5A,B). Conversely, the bolting time of 35 S::DcCOL5a plants (T2) was slightly accelerated by 3.0 days and the rosette leaf number decreased by about 1.8.
Discussion
For cultivated carrot, it takes almost a year for it to complete its lifecycle. Many endogenous and external factors can affect carrot bolting and flowering during growth41,42,43. Plants with 8–12 leaves are required to respond to a low temperature in carrot41, but some landraces and cultivars can initiate bolting after a short vernalization period52,53. During the winter–spring period or spring cultivation, the premature bolting of carrot might be interconnected by vernalization and photoperiod pathways3,12,13,14. A basic and putative flowering time network was suggested in carrot42, but knowledge of its photoperiodic regulation is limited. In order to further understand the molecular mechanisms of the photoperiod pathway, the establishment of a set of carrot materials with a sensitivity to different photoperiods is required. In this study, Ws plants began to bolt 67 days after germinating and reached 100% after 108 days under LD conditions (Fig. 3A). In spring, under natural conditions, Ws plants similarly began to bolt about 65 days after germinating, but the bolting ratio increased rapidly (Fig. 4A). Ws plants initiated bolting 39 days after germinating and rapidly reached 93.3% after 74 days in autumn (Fig. 4A), but there was no low temperature recorded during the growth. In addition, the bolting time of Ws could be delayed for 22 days and its bolting ratio was significantly reduced to 40% about 108 days after germinating (Fig. 3A), which was similar to a previous report showing that the floral initiation of carrot can be inhibited by SDs42. Meanwhile, the orange cultivar Af showed no bolting phenomenon, with tolerance of flowering characterization under different photoperiods (Figs. 3A and 4A). This result suggests that Ws and Af can be regarded as suitable materials for further studying photoperiodic flowering regulations.
According to the carrot genome48, previously reported de novo transcriptome genes42 could be localized to specific chromosomes and some new photoperiod genes were identified in this study, including 9 DcCOLs, 1 DcFT, and 1 DcSOC1, but there is no complete structure of CO6,11,15,16,17,18 (Supplementary Table S2). GI plays a general role in controlling circadian rhythms for flowering and is highly conserved in seed plants16,19,54. One DcGI was identified in Daucus species and was strongly selected during evolution (Fig. 1). The circadian pattern of DcGI is similar to that described for AtGI19,54 and OsGI15, and its peak could be delayed by a long photoperiod (Fig. 3C). However, the response for Af and Ws was different that the peak value was quickly reduced in Af under SD-LD conditions, but was maintained in Ws, which was consistent with the trend of expression in spring (Fig. 4B). According to the bolting ratio and time of Ws (Figs. 3A and 4A), DcGI might promote flowering in the carrot photoperiodic pathway.
CO/COL plays a central role in regulating the downstream florigen gene by integrating light and circadian clock signals3,11,18,25,26, and multiple homologues have been identified in many species32,33,34,35. Only some CO homologues in other species can be regarded as potential regulators of FT-like genes12. In this study, 15 DcCOLs were identified with complete B-box 1 and CCT domains in carrot (Supplementary Table S2, Fig. 2B). According to the nucleotide diversity (π) and neutrality tests, 8 DcCOLs might be selected during evolution, especially DcCOL5b, DcCOL11, and DcCOL15 in cultivars (Fig. 1). DcCOL2 showed a similar morning-phase rhythm in Ws and Af, which was comparable to AtCOL228 and ClCOL255. In Ws, the peak of DcCOL2 was advanced and higher under SD conditions than under SD-LD and LD conditions (Fig. 3C), but the bolting time was significantly delayed and the ratio was reduced under SD conditions (Fig. 3A). In Af, the patterns of DcCOL2 were similar and the peak value was higher than that in Ws (Fig. 3C). In the two seasons, the expression of DcCOL2 in Af showed higher levels than in Ws, but no bolting plants were found in Af (Fig. 4A). The bolting time of 35 S::DcCOL2 Arabidopsis plants (T2) was also significantly delayed, but its rosette leaf number was similar to that of the WT (Fig. 5). DcCOL4 also had a morning expression pattern and peaked at dawn under SDs, similar to DcCOL2. It was strongly repressed under a long photoperiod in Ws, but it could maintain a constantly high level in Af (Fig. 3C), even during spring (Fig. 4B). A recent study showed that an increase of COL4 expression causes flowering delay in Arabidopsis29, but COL2 has little effect on the flowering time28. Maintaining a higher expression of DcCOL2 and DcCOL4 might be related to the bolting tolerance in Af, but this claim needs more supporting evidence.
Two DcCOL5 homologues were identified and their circadian rhythms under SDs were similar to that of ClCOL555, but their responses to a long photoperiod were different (Fig. 3C). It was obvious that the peak of DcCOL5a was postponed and maintained a similar level under a long photoperiod, but that of DcCOL5b was advanced and reduced in Ws. In Af, the expression of DcCOL5a was downregulated, but there was no DcCOL5b response under SD-LD and LD conditions. In the two seasons, the expression level of DcCOL5a in Ws was constantly higher than that in Af, but the expression levels of DcCOL5b fluctuated (Fig. 4B). The bolting time of 35 S::DcCOL5a Arabidopsis plants (T2) was only slightly accelerated (Fig. 5). DcCOL13b was only expressed in Ws and maintained similar expression levels under different photoperiods (Fig. 3C), but its circadian rhythm presented contrasting patterns under SD and SD-LD/LD conditions, similar to that of OsCOL1336. Whether DcCOL13b represses flowering or not requires further investigation.
FT is the major primary target of CO in leaves4,6,7,12. In the carrot genome, there are two FT homologues localized at Chrs. 1 and 7, respectively48 (Supplementary Table S2). According to the phylogenetic analysis, DcFT1 was assigned to group I and associated with EuHd3a, LsFT, and DlFT1, while DcFT2 was assigned to group II and closely associated with VvFT, InFTL, and NtFTL (Fig. 2C). It is interesting that DcFT1 had a 23.66 kbp large intron by compared its cDNA sequence with the carrot genome48 (Fig. 3B). The circadian rhythm of DcFT1 in Ws was similar to that of AtFT and its expression levels were significantly upregulated under SD-LD/LD conditions, but constantly low in Af4,7 (Fig. 3C). The expression patterns of DcFT1 in Ws and Af were different, but the levels were similar during the two seasons (Fig. 4B). The expanded studies show that FT homologues have a conserved function in promoting flowering12. Different responses of DcFT1 in Ws and Af might be related to the different bolting response, which could give a clue for further studying its function in flowering regulation.
SOC1 integrates multiple flowering signals to interact with multiple MADS-box proteins and regulate the expresssion of flowering genes8,9,21,56,57. Three DcSOC1s with complete structures were identified and found to be highly conserved during evolution (Fig. 1). The circadian rhythm patterns of DcSOC1b/c were different under the three photoperiod treatments (Fig. 3C). The expression of DcSOC1b in Ws maintained significantly higher levels than those in Af, while DcSOC1c showed similar levels. The same results were observed in the two studied seasons (Fig. 4B). These results suggest that DcSOC1b might play a role in regulating downstream flowering genes, like AtSOC19,20 and ZmSOC157.
In general, Ws plants could initiate bolting 39 days after germinating without low-temperature treatment, which suggests that it can be regarded as a suitable material for studying photoperiodic flowering regulation. Based on previous studies, 1 DcGI, 15 DcCOLs, 2 DcFTs, and 3 DcSOC1s were identified in the photoperiod pathway. The circadian rhythm peaks of DcGI, DcCOL2, DcCOL5a, and DcCOL13b could be delayed under LD conditions. The peak value of DcCOL2 in Af was significantly higher than that in Ws under SD conditions, and their values could be reduced under SD-LD/LD conditions. Under the photoperiod treatments, the peak values of DcCOL5a in Ws were constantly higher than those in Af, even during the two seasons. The expression levels of DcFT1 in Ws were significantly upregulated under SD-LD/LD conditions compared with those in Af. These responses of DcCOL2, DcCOL5a, and DcFT1 might be related to the different bolting responses of Ws and Af. This study could provide valuable information that increases our understanding of the key integrator genes in the carrot photoperiod pathway. Future work focusing on characterizing the function of these candidate genes will be helpful for screening those accessions with tolerance to premature bolting.
Methods
Retrieval and phylogenetic analysis of DcGI, DcCOL, DcFT, and DcSOC1 sequences in carrot
GI, COL, FT, and SOC1 are the key transcript regulators in the photoperiodic pathway4,6,7,11,16,17. Based on the carrot genome48, the sequences of 1 DcGI, 28 DcCOL, 4 DcFT, and 6 DcSOC1 homologues were retrieved and aligned with transcriptome unigenes42 using DNAMAN software (Supplementary Table S1). Twenty-seven GI, 22 CO/COL, 28 FT/Hd3a, and 22 SOC1 homologue protein sequences of other species were retrieved from the NCBI database and used to perform a phylogenetic analysis with the abovementioned genes through MEGA5.0 using the neighbor-joining method, respectively. The phylogenetic tree architecture was validated using the probabilistic bootstrap test with 1000 replicates.
Nucleotide diversity and selection of DcGI, DcCOLs, DcFTs, and DcSOC1s in Daucus species
Twenty-one D. carota var. sativus accessions (as DCS), 4 D. carota subsp. gummifer species and 9 D. carota subsp. carota species (as DCC), and 5 Daucus species (as Dau) were used to perform the evolutionary analysis of GI, COL, FT, and SOC1 homologues (Supplementary Table S1). The genome of cultivar Af, breeding lines 17166 C, 170P2A, 17P25A, and wild species Ws were re-sequenced with each leaf sample in the Illumina HiSeq. 2500 sequencing platform (Biomarker Technologies Co., Ltd. Beijing, China). The sequence data of 34 accessions were downloaded from NCBI (https://www.ncbi.nlm.nih.gov/genome/?term=carrot)48. All genome sequence reads from 39 individual accessions were trimmed and filtered prior to analysis, and aligned to the carrot genome48 using BWA software58. Alignment files were converted into SAM/BAM files and simple nucleotide polymorphism (SNP) calling was performed using SAMtools58. Low-quality SNPs with a base quality value of <20 and a read depth of <4× or those with >32× coverage from the sequences were excluded because these SNPs may be false positives. The FASTA sequences of 39 individuals were extracted using BCFtools59. Nucleotide diversity (π) for the genomic sequence of each data group and neutral tests of TD49 and FF50 were estimated based on the neutral model prediction by DnaSP 660.
Gene cloning and circadian rhythm analysis of DcGI, DcCOLs, DcFT1, and DcSOC1s
For the circadian rhythm analysis, Ws and Af seeds were sown in flats (90 cm × 30 cm × 30 cm) filled with a 1:2:3 mixture of soil, vermiculite, and turf in the greenhouse at a temperature of 13–16/25–28 °C night/day in late February. The average maximum light intensity per day in the greenhouse (55.4 ± 3.7 W ∙ m−2) was only 30.6% of that in the field (180.9 ± 14.4 W ∙ m−2). When the seedlings had 3–4 leaves and 37 days after germinating, the plants were moved to the field under a natural temperature and subjected to SD (8 h light/16 h dark cycle) and LD (16 h light/8 h dark cycle, supplemented with white fluorescent light of 30 μmol/m2/s) treatments for 27 days. Then, half of the seedlings that had 6–7 leaves under SD treatment were subjected to the above LD conditions for 7 days (as SD-LD), and half remained under SD conditions. During the treatment, there were about 12 days with a low temperature (below 10 °C)41. The leaves of each treatment were sampled from five plants with three biological replicates at 4 h intervals after dawn at zeitgeber time 0 (ZT 0). All samples were immediately frozen in liquid nitrogen and stored at −80 °C until use. After this, all plants were grown under natural conditions, and the number of bolting plants was counted every 2–4 days until 108 days after germinating.
The candidate genes of DcGI, DcCOL2, DcCOL4, DcCOL5a/b, DcCOL13b, DcCOL15, DcFT1, and DcSOC1b/c were selected to study the circadian rhythm under different photoperiod treatments. The specific primers for the open reading frame (ORF) sequence clone and real-time qPCR were designed using Primer Premier 5.0 based on the abovementioned gene sequences (Supplementary Table S3). The operation procedures for RNA extraction, cDNA synthesis, reverse transcription PCR, and real-time qPCR were carried out as described by Wang et al.61. Tublin was amplified along with the target genes as an endogenous control to normalize the expression levels between samples62. Three biological repetitions were performed for each expression data point. The comparative CT (2−ΔΔCt) method was applied for calculation63. The lowest expression level of each gene was used for calibration. Values are reported as the mean ± SE from the replicates.
Expression analysis of DcGI, DcCOLs, DcFT1, and DcSOC1s in spring and autumn
The expression levels of DcGI, DcCOLs, DcFT1, and DcSOC1s were further analyzed to understand their functions during carrot growth under natural conditions. The seeds of Ws and Af were sown with 20 cm row spacing in the field under natural photoperiods and temperature conditions on 19 March (spring) and 4 August (autumn), respectively. The cultivation and management of carrot referred to that of Rubatzky et al.45. There were about 15 days with a low temperature (below 10 °C) after the seeds had germinated in spring and no low temperature was recorded before the plants had bolted in autumn. When the plants had about five leaves, the leaves from five plants were sampled with three biological replicates from 08:00 to 09:00 at 55, 63, 71, and 79 days after germinating in spring, and at 32, 39, 46, and 62 days after germinating in autumn. The number of bolting plants was also investigated. Sample collection, total RNA extraction, cDNA synthesis, gene expression, and data analysis were performed as described by Wang et al.61.
Flowering time of transgenic Arabidopsis with DcCOL2 and DcCOL5a
To further understand the functions of DcCOL2 and DcCOL5a, their full-length cDNAs were amplified using gene-specific primers and subcloned into the vector PBI121 plasmid and agrobacterium GV3101 (Tiangen, China) to generate 35 S::DcCOL2 and 35 S::DcCOL5a, which were introduced into the wild-type Columbia (WT) Arabidopsis by the floral dip method64. Transgenic plants were screened on Murashige and Skoog (MS) agar plates supplemented with 50 mg/L kanamycin. After being left in darkness at 4 °C for 3 days, kanamycin-resistant seedlings were transferred to soil after 10 days and grown in a controlled growth room (~22 °C, 16 h light/8 h dark). The seeds from T0 plants were harvested and sown on MS plates supplemented with 50 mg/L kanamycin and treated under the above-described conditions. The seedlings (T1) were transferred and grown under the above-described conditions. Fifteen 35 S::DcCOL2 and 13 35 S::DcCOL5a plant leaves were independently sampled and tested using PCR. The PCR program was followed as described by Wang et al.61. WT and T2 transgenic plants were grown under the abovementioned conditions to investigate the bolting time and the number of rosette leaves30. Values are shown as the mean ± SE from the replicates. Duncan’s post-hoc multiple comparison was used to detect significant differences in the number of rosette leaves and bolting time between wild-type and transgenic Arabidopsis plants (T2) using SPSS (version 10.0).
References
Fornara, F., de Montaigu, A. & Coupland, G. SnapShot: Control of flowering in Arabidopsis. Cell 141, 550 (2010).
Matsoukas, I. G., Massiah, A. J. & Thomas, B. Florigenic and antiflorigenic signaling in plants. Plant. Cell Physiol. 53, 1827–1842 (2012).
Andres, F. & Coupland, G. The genetic basis of flowering responses to seasonal cues. Nat. Rev. Genet. 13, 627–639 (2012).
Abe, M. et al. FD, a bZIP protein mediating signals from the floral pathway integrator FT at the shoot apex. Science 309, 1052–1056 (2005).
Corbesier, L. et al. FT protein movement contributes to long-distance signaling in floral induction of Arabidopsis. Science 316, 1030–1033 (2007).
Moraes, T. S., Dornelas, M. C. & Martinelli, A. P. FT/TFL1: Calibrating Plant Architecture. Front. Plant. Sci. 10, 97 (2019).
Wigge, P. A. et al. Integration of spatial and temporal information during floral induction in Arabidopsis. Science 309, 1056–1059 (2005).
Immink, R. G. et al. Characterization of SOC1’s central role in flowering by the identification of its upstream and downstream regulators. Plant. Physiol. 160, 433–449 (2012).
Lee, J., Oh, M., Park, H. & Lee, I. SOC1 translocated to the nucleus by interaction with AGL24 directly regulates leafy. Plant. J. 55, 832–843 (2008).
Liu, C. et al. Direct interaction of AGL24 and SOC1 integrates flowering signals in Arabidopsis. Development 135, 1481–1491 (2008).
Golembeski, G. S., Kinmonth-Schultz, H. A., Song, Y. H. & Imaizumi, T. Photoperiodic flowering regulation in Arabidopsis thaliana. Adv. Bot. Res. 72, 1–28 (2014).
Ballerini, E. S. & Kramer, E. M. In the light of evolution: a reevaluation of conservation in the CO-FT regulon and its role in photoperiodic regulation of flowering time. Front. Plant. Sci. 2, 81 (2011).
Ream, T. S. et al. Interaction of photoperiod and vernalization determines flowering time of Brachypodium distachyon. Plant. Physiol. 164, 694–709 (2014).
Weller, J. L. & Ortega, R. Genetic control of flowering time in legumes. Front. Plant. Sci. 6, 207 (2015).
Hayama, R., Yokoi, S., Tamaki, S., Yano, M. & Shimamoto, K. Adaptation of photoperiodic control pathways produces short-day flowering in rice. Nature 422, 719–722 (2003).
Nohales, M. A. et al. Multi-level modulation of light signaling by GIGANTEA regulates both the output and pace of the circadian clock. Dev. Cell 49, 840–851 (2019).
Song, Y. H., Ito, S. & Imaizumi, T. Flowering time regulation: photoperiod-and temperature-sensing in leaves. Trends Plant. Sci. 18, 575–583 (2013).
Suarez-Lopez, P. et al. CONSTANS mediates between the circadian clock and the control of flowering in Arabidopsis. Nature 410, 1116–1120 (2001).
Mishra, P. & Panigrahi, K. C. GIGANTEA-an emerging story. Front. Plant. Sci. 6, 8 (2015).
Song, Y. H., Smith, R. W., To, B. J., Millar, A. J. & Imaizumi, T. FKF1 conveys timing information for CONSTANS stabilization in photoperiodic flowering. Science 336, 1045–1049 (2012).
Yoo, S. K. et al. CONSTANS activates SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 through FLOWERING LOCUS T to promote flowering in Arabidopsis. Plant. Physiol. 139, 770–778 (2005).
Sawa, M., Nusinow, D. A., Kay, S. A. & Imaizumi, T. FKF1 and GIGANTEA complex formation is required for day-length measurement in Arabidopsis. Science 318, 261–265 (2007).
Kojima, S. et al. Hd3a, a rice ortholog of the Arabidopsis FT gene, promotes transition to flowering downstream of Hd1 under short-day conditions. Plant. Cell Physiol. 43, 1096–1105 (2002).
Yano, M. et al. Hd1, a major photoperiod sensitivity quantitative trait locus in rice, is closely related to the Arabidopsis flowering time gene CONSTANS. Plant. Cell 12, 2473–2484 (2000).
Robson, F. et al. Functional importance of conserved domains in the flowering-time gene CONSTANS demonstrated by analysis of mutant alleles and transgenic plants. Plant. J. 28, 619–631 (2001).
Simon, S., Ruhl, M., de Montaigu, A., Wotzel, S. & Coupland, G. Evolution of CONSTANS regulation and function after gene duplication produced a photoperiodic flowering switch in the Brassicaceae. Mol. Biol. Evol. 32, 2284–2301 (2015).
Cheng, X. F. & Wang, Z. Y. Overexpression of COL9, a CONSTANS-LIKE gene, delays flowering by reducing expression of CO and FT in Arabidopsis thaliana. Plant. J. 43, 758–768 (2005).
Ledger, S., Strayer, C., Ashton, F., Kay, S. A. & Putterill, J. Analysis of the function of two circadian-regulated CONSTANS-LIKE genes. Plant. J. 26, 15–22 (2001).
Steinbach, Y. The Arabidopsis thaliana CONSTANS-LIKE 4 (COL4)-A Modulator of Flowering Time. Front. Plant. Sci. 10, 651 (2019).
Hassidim, M., Harir, Y., Yakir, E., Kron, I. & Green, R. M. Over-expression of CONSTANS-LIKE 5 can induce flowering in short-day grown Arabidopsis. Planta 230, 481–491 (2009).
Takase, T. et al. Characterization and transgenic study of CONSTANS-LIKE8 (COL8) gene in Arabidopsis thaliana: expression of 35S: COL8 delays flowering under long-day conditions. Plant. Biotechnology 28, 439–446 (2011).
Chia, T. Y., Muller, A., Jung, C. & Mutasa-Gottgens, E. S. Sugar beet contains a large CONSTANS-LIKE gene family including a CO homologue that is independent of the early-bolting (B) gene locus. J. Exp. Bot. 59, 2735–2748 (2008).
Fan, C. et al. Conserved CO-FT regulons contribute to the photoperiod flowering control in soybean. BMC Plant. Biol. 14, 9 (2014).
Griffiths, S., Dunford, R. P., Coupland, G. & Laurie, D. A. The evolution of CONSTANS-like gene families in barley, rice, and Arabidopsis. Plant. Physiol. 131, 1855–1867 (2003).
Nemoto, Y., Kisaka, M., Fuse, T., Yano, M. & Ogihara, Y. Characterization and functional analysis of three wheat genes with homology to the CONSTANS flowering time gene in transgenic rice. Plant. J. 36, 82–93 (2003).
Sheng, P. et al. A CONSTANS-like transcriptional activator, OsCOL13, functions as a negative regulator of flowering downstream of OsphyB and upstream of Ehd1 in rice. Plant. Mol. Biol. 92, 209–222 (2016).
Itoh, H. & Izawa, T. The coincidence of critical day length recognition for florigen gene expression and floral transition under long-day conditions in rice. Mol. Plant. 6, 635–649 (2013).
Kloosterman, B. et al. Naturally occurring allele diversity allows potato cultivation in northern latitudes. Nature 495, 246–250 (2013).
Larson, G. et al. Current perspectives and the future of domestication studies. Proc. Natl Acad. Sci. USA 111, 6139–6146 (2014).
Murphy, R. L. et al. Ghd7 (Ma6) Represses sorghum flowering in long days: Ghd7 alleles enhance biomass accumulation and grain production. The Plant Genome 7 (2014).
Alessandro, M. S., Galmarini, C. R., Iorizzo, M. & Simon, P. W. Molecular mapping of vernalization requirement and fertility restoration genes in carrot. Theor. Appl. Genet. 126, 415–423 (2013).
Ou, C. G. et al. Characterising genes associated with flowering time in carrot (Daucus carota L.) using transcriptome analysis. Plant. Biol. 19, 286–297 (2017).
Atherton, J. G., Basher, E. A. & Brewster, J. L. The effects of photoperiod on flowering in carrot. J. Pomology Horticultural Sci. 59, 213–215 (1984).
Dickson, M. H. & Peterson, C. E. Hastening greenhouse seed production for carrot breeding. J. Amer Soc. Hort. Sci. 71, 412–415 (1958).
Rubatzky, V. E., Quiros, C. F. & Simon, P. W. Carrots and related vegetable Umbelliferae. (CABI publishing, 1999).
Sakr, E. S. & Thompson, H. C. Effect of temperature and photoperiod on seedstalk development in carrots. Proc. Amer. Soc. Hort. Sci. 41, 343–346 (1942).
Zhan, Z. et al. Molecular cloning, expression analysis, and subcellular localization of FLOWERING LOCUS T (FT) in carrot (Daucus carota L.). Mol. Breed. 37, 149 (2017).
Iorizzo, M. et al. A high-quality carrot genome assembly provides new insights into carotenoid accumulation and asterid genome evolution. Nat. Genet. 48, 657–666 (2016).
Tajima, F. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123, 585–595 (1989).
Fu, Y. X. & Li, W. H. Statistical tests of neutrality of mutations. Genetics 133, 693–709 (1993).
Fowler, S. et al. GIGANTEA: a circadian clock-controlled gene that regulates photoperiodic flowering in Arabidopsis and encodes a protein with several possible membrane-spanning domains. EMBO J. 18, 4679–4688 (1999).
Dickson, M., Rieger, B. & Peterson, C. A cold unit system to evaluate bolting resistance in carrots. J. Amer Soc. Hort. Sci. 71, 401–405 (1961).
Alessandro, M. S. & Galmarini, C. R. Inheritance of vernalization requirement in carrot. J. Amer Soc. Hort. Sci. 132, 525–529 (2007).
Mizoguchi, T. et al. Distinct roles of GIGANTEA in promoting flowering and regulating circadian rhythms in Arabidopsis. Plant. Cell 17, 2255–2270 (2005).
Fu, J., Yang, L. & Dai, S. Identification and characterization of the CONSTANS-like gene family in the short-day plant Chrysanthemum lavandulifolium. Mol. Genet. Genomics 290, 1039–1054 (2015).
Jaeger, K. E., Pullen, N., Lamzin, S., Morris, R. J. & Wigge, P. A. Interlocking feedback loops govern the dynamic behavior of the floral transition in Arabidopsis. Plant. Cell 25, 820–833 (2013).
Zhao, S. et al. ZmSOC1, a MADS-box transcription factor from Zea mays, promotes flowering in Arabidopsis. Int. J. Mol. Sci. 15, 19987–20003 (2014).
Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–1760 (2009).
Danecek, P. & McCarthy, S. A. BCFtools/csq: haplotype-aware variant consequences. Bioinformatics 33, 2037–2039 (2017).
Rozas, J. et al. DnaSP 6: DNA sequence polymorphism analysis of large data sets. Mol. Biol. Evol. 34, 3299–3302 (2017).
Wang, H., Ou, C. G., Zhuang, F. Y. & Ma, Z. G. The dual role of phytoene synthase genes in carotenogenesis in carrot roots and leaves. Mol. Breed. 34, 2065–2079 (2014).
Fuentes, P. et al. Light-dependent changes in plastid differentiation influence carotenoid gene expression and accumulation in carrot roots. Plant. Mol. Biol. 79, 47–59 (2012).
Livak, K. J. & Schmittgen, T. D. Analysis of relative gene expression data using real-time quantitative PCR and the 2− ΔΔCT Method. Methods 25, 402–408 (2001).
Clough, S. J. & Bent, A. F. Floral dip: a simplified method for Agrobacterium-mediated transformation of Arabidopsis thaliana. Plant. J. 16, 735–743 (1998).
Acknowledgements
This work was partially supported by the National Key Research and Development Program of China (2016YFD01000204), the China Agriculture Research System (CARS-23), the Science and Technology Program of Qinghai (2017-NK-150), the Innovation Program of the Chinese Academy of Agricultural Sciences (CAAS-ASTIP-IVFCAAS), and the Central Public-interest Scientific Institution Basal Research Fund (IVFS201901).
Author information
Authors and Affiliations
Contributions
F.Z. conceived and designed the research. L.L., C.O., S.C., Q.S., B.L., M.L., Z.Z., X.K. and X.Y. performed the experiments. L.L. and C.O. analyzed the data and wrote the manuscript. L.L. and C.O. share the first authorship. All authors reviewed the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Liu, L., Ou, C., Chen, S. et al. The Response of COL and FT Homologues to Photoperiodic Regulation in Carrot (Daucus carota L.). Sci Rep 10, 9984 (2020). https://doi.org/10.1038/s41598-020-66807-y
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-020-66807-y
This article is cited by
-
Interactive Effects of Iron and Photoperiods on Tomato Plant Growth and Fruit Quality
Journal of Plant Growth Regulation (2023)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.