1. Introduction

Turnip and oilseeds, the most important useable Brassica crops in the Qinghai-Tibet Plateau (QTP), are among the few crops that have been domesticated and are widely cultivated by local people (Zhang et al., 2014; Guedes et al., 2015). Oilseeds, which belongs to Brassica rapa, a member of the A genome according to the ancient triangle of U of Brassica species, is widely cultivated in the QTP as the main oil crop. Turnip, another important B. rapa crop, is especially suitable for cultivation in high-altitude and cold regions in the QTP (Qi et al., 2017). Turnip plays an essential role in husbandry and agricultural development, culture and common life of local people in the QTP.

The phenological development of generative and vegetative crops is very important in crop production (Purugganan and Fuller, 2009; Blümel et al., 2015). Oilseeds is a reproductive organ-type crop, and flowering accuracy is important for reproductive success. However, turnip is a vegetative organ-type crop, and lengthening the duration of vegetative growth before flowering helps to achieve maximum turnip production because early bolting can considerably decrease taproot yield and quality (Zheng et al., 2018). Due to human domestication and natural selection, Brassica crops are highly morphologically and phenologically variable, including with regard to flowering time (Jung and Müller, 2009). Brassica can be grouped into winter annuals with a vernalization response and spring annuals with rapid flowering within one growing season. As a winter annual, turnip maintains a prolonged period of vegetative growth to delay its flowering time before vernalization, which requires a long stretch of low temperature to promote flowering in the following spring (Zheng et al., 2018, 2021).

The gene FLOWERING LOCUS C (FLC) is a strong flowering repressor that acts to repress flowering by directly repressing the expression of floral integrator genes (Amasino, 2010; Hepworth and Dean, 2015; He et al., 2020). Winter annuals carry the dominant FLC allele, which confers vernalization requirements and winter-annual habits on plants. During vernalization, FLC expression levels decrease and are epigenetically silenced once plants are exposed to prolonged cold conditions (Searle et al., 2006). Variation in vernalization requirements among Arabidopsis thaliana accessions has been shown to mainly depend on allelic variation at the FLC locus ((Michaels et al., 2003; Shindo et al., 2005, 2006). Variations at the FLC locus have also been shown to contribute to differences in flowering time and ecotype differentiation in Brassica (Zhao et al., 2010; Wu et al., 2012; Tudor et al., 2020; Yin et al., 2020). In addition, dosage effects of repeated flowering trait loci are thought to contribute greatly to the various flowering phenotypes of Brassica crops (Osborn, 2004; Calderwood et al., 2021). The FLC gene has a significant dosage effect, and its dosage sensitivity also affects its function (Michaels and Amasino, 1999; Hepworth et al., 2020).

During field work on the QTP, we observed that turnip plants commonly bolted early. In this study, we used molecular and genetic experiments to determine the cause of early bolting in turnip cultivation. Specifically, we found that differences in flowering time and life habits between turnip and oilseeds were closely correlated with the genetic variations in BraFLC2. We further found that early bolting in turnips was caused by cross-pollination contamination from oilseeds using BraFLC2 as a testing gene. The BraFLC2 alleles in early-bolting turnip was identified as heterozygote which consisted of one BraFLC2 in turnip and another BraFLC2 in oilseeds. Furthermore, we found that the hybridization between turnip and oilseeds affected the expression level of BraFLC2, resulting in the early bolting of hybrid plants.

2. Materials and methods 2.1. Plant material and growth conditions

The materials used in this study included Tibetan turnip (Brassica rapa var. rapa) (KTRG-B58) and oilseeds (B. rapa var. oleifera) (OI2), which were both collected from the Xizang Autonomous Region in China. Turnip seeds were sown in Petri dishes containing two pieces of filter paper and kept in the dark at 22 ℃ for 3 days until germination. Then, seedlings were transferred into small pots filled with soil and vermiculite at a 3:1 ratio and kept under warm and long-day (LD) conditions (22 ℃ and 16 h light: 8 h dark). The location of all plants was randomly changed every week to avoid location effects, and plants were watered once a week.

For the artificial hybridization experiment, late-flowering parent turnips (KTRG-B58) were vernalized to flower quickly, and an F1 population was established from a cross between a parent turnip and an early-flowering parent oilseeds variety (OI2).

2.2. Vernalization treatments and flowering time evaluation

For vernalization, germinated turnip seedlings with two or three true leaves were treated with vernalization conditions (4 ℃, 8 h light: 16 h dark) for 40 days of vernalization. After vernalization, all plants were transferred to warm and long-day conditions. No vernalized plants were directly placed under the same conditions as vernalized plants at the same time. Flowering time was recorded as the number of days from sowing to the opening of the first flower (DTF). All measurements were based on 15 plants.

2.3. RNA extraction, quantitative RT‒PCR and semiquantitative RT‒PCR

For expression analysis of BraFLC2, tissues (10-day-old leaves, 10-day-old roots and flowers) were collected from turnip (KTRG-B58) and oilseed (OI2) plants. Total RNA was isolated using an Eastep® Super Total RNA Kit (Promega), and RNA purity was detected using a Nanodrop Spectrophotometer 2000. First-strand cDNA was synthesized from 1.5 μg of DNase-treated RNA using 5 × All-In-One MasterMix (with AccuRT Genomic DNA Removal Kit). The quantitative real-time polymerase chain reaction (qRT–PCR) experiment was performed using EvaGreen 2 × qPCR MasterMix (Abm) on a StepOnePlus^TM Real-Time PCR System (Applied Biosystems) following the manufacturer's instructions. The BraTUB2 gene was used as an internal control to normalize the expression levels of the BraFLC2 gene, which were calculated using the 2^−ΔΔCt method. Analysis of each sample was completed using at least three technical replicates and three biological replicates. Primers used are listed in Table S1.

Semiquantitative RT‒PCR analysis was performed with the gene-specific primers listed in Table S1, and cDNA was PCR-amplified over 35 cycles of 95 ℃ for 30 s, 56 ℃ for 30 s, and 72 ℃ for 30 s. The BraTUB2 gene was amplified as an internal control. RT‒PCR products were detected on 1.5% agarose gels.

2.4. DNA extraction and BraFLC2 amplification

To further analyze the sequence variations of the BraFLC2 gene between turnip and oilseeds plants, genomic DNA was isolated from samples of young leaf tissue (100 mg) using the CTAB method. The 5.4-kb genomic sequence of BraFLC2 was amplified and cloned with gene-specific primers using DNA as templates. PCR products were sequenced by Tsingke Biotechnology Co., Ltd. using the Sanger sequencing method. Sequence alignments of 5.4-kb BraFLC2 genomic genes of turnip and oilseeds plants were performed using AlignX of Vector NTI Advance 11 software. Primers used are listed in Table S1.

2.5. Identification of alternative splicing variants of BraFLC2 in oilseeds

The full-length cDNA coding region of the BraFLC2 gene of oilseeds was amplified from the cDNA of turnip and oilseeds using Phanta® Max Super-Fidelity DNA Polymerase. The purified PCR products were cloned into the T-vector using the pMD18^TM-T Vector Cloning Kit, and then these constructs were transformed into Escherichia coli DH5α. Positive colonies were selected and sequenced using corresponding primers (Table S1). The sequences of BraFLC2 PCR products obtained by Sanger sequencing were aligned using AlignX of Vector NTI Advance 11 software. Primers used are listed in Table S1.

2.6. Statistical analysis

For flowering time analyses, at least 15 individual plants per each accession were recorded. For RT‒qPCR, 3 individual plants were pooled per sample, and three biological and three technical replicates were used. Ordinary one-way ANOVA followed by Tukey's multiple comparisons test was used to determine significant differences in each measured parameter.

3. Results 3.1. Comparison of growth habits between turnip and oilseeds in the Qinghai-Tibet Plateau

Combining field and laboratory experiments, we analyzed the growth habits of Tibetan turnip and oilseeds. We found that turnip, sown in the early autumn, showed a significant winter-annual habit and developed taproot characteristics after a 2-month period of vegetative growth. The taproots were then harvested and vernalized during a long period of low temperature in winter and flowered in the following spring (Fig. 1a). Under laboratory conditions, Tibetan turnip flowered very late when grown in the greenhouse under warm conditions; it produced taproots after 45 days and maintained a vegetative growth period extending over 180 days to finish the flowering stage in the greenhouse (Fig. 1c). Our previous study revealed that Tibetan turnip exhibited a vernalization requirement. After a 40-day vernalization treatment, Tibetan turnip started flowering quickly upon being transferred to the greenhouse, with no taproot induction (Zheng et al., 2018). In contrast, Tibetan oilseeds showed a spring-annual habit in which after sowing in spring, flowering and fruiting were completed in the autumn (Fig. 1b). We also found that oilseed plants need only a 40-day vegetative period to flower rapidly under laboratory conditions without a vernalization requirement (Fig. 1d). These results indicated that there was a significant difference in flowering time between turnip and oilseeds in the QTP.

3.2. Problems with the cultivation and agricultural production of turnip on the Qinghai-Tibet Plateau

During the process of field inspection and research on turnip resources in local cultivated areas, we noticed that local planting conditions differed greatly. For example, the turnip landraces cultivated in Xinjiang and Shangri-La, one of the important vegetables for daily consumption, are under a very strict planting management and variety selection process. We found that the development and growth cycle of these turnips in the field under agricultural production showed quite consistent traits and a stable yield (Fig. 2a and b; S1a, b). In contrast, as the major winter supplemental feed for local livestock in the QTP, the planting pattern of turnip under local agriculture production activities was extensive. The development and growth cycle of Tibetan turnip showed interindividual instability and heterogeneity. We found some turnip plants started bolting when they were supposed to be in the vegetative growth season (autumn) (Fig. 2c and d; S1c, d). We further identified that these early-flowering plants were almost all turnips because tiny but obvious taproots emerged. Considering the large-scale planting of oilseeds, one of the relatives of turnip, in the QTP, we speculated that extensive human cultivation may cause seed contamination among varieties.

3.3. Analysis of BraFLC2 expression patterns in turnip and oilseeds

BraFLC2 plays an essential role in delaying flowering and responding to vernalization (Zhao et al., 2010; Zheng et al., 2018). Here, we first investigated the expression pattern of BraFLC2 in different tissues of Tibetan turnip, and found that in Tibetan turnip BraFLC2 was strongly expressed in leaves, expressed at lower levels in roots, and barely detectable in flowers (Fig. 3a). In addition, BraFLC2 expression levels significantly differed between Tibetan turnip and oilseeds. Specifically, BraFLC2 was expressed at much higher levels in turnip than in oilseeds, where expression was barely detectable (Fig. 3a). When we visualized the results of semi-quantitative RT-PCR on an agarose gel, we found that PCR fragments of expected sizes were obtained in leaves and roots in turnip. However, not only was oilseed BraFLC2 expressed at lower levels than was turnip BraFLC2, but that oilseed BraFLC2 amplicons appeared as diffuse, not single band, indicating oilseed BraFLC2 transcripts are truncated or may undergo alternative splicing (Fig. 3b).

We subsequently detected the transcription level of BraFLC2 in turnip leaves at different developmental stages. BraFLC2 was expressed at high levels beginning at the seedling stage and was maintained during all vegetative growth stages (from G15 d to G120 d). When turnip plants reached the reproductive growth stage, BraFLC2 transcript levels decreased sharply; at bolting, BraFLC2 transcript levels were nearly undetectable (G180 d) (Fig. 3c). This finding indicated that the decrease in BraFLC2 expression was related to the transition to flowering, which is consistent with previous claims that this decrease is sufficient for the transition to flowering (Michaels and Amasino, 1999). When we examined BraFLC2 expression level during the growth cycle of rapid-flowering oilseeds, we found that BraFLC2 expression was very low and barely detectable after flowering (Fig. 3d). Taken together, these results showed that the differences in growth habits between turnip and oilseeds were closely related to variation in BraFLC2 expression levels.

3.4. Genomic variation of BraFLC2 in turnip and oilseeds

Previous studies have demonstrated that the variation in vernalization requirements depends on allelic variation at the FLC locus and is reflected in the various life history strategies of different cruciferous plants (Wu et al., 2012; Yin et al., 2020). To investigate sequence variation in the BraFLC2 gene of turnip and oilseeds, 5.4-kb genomic fragments of BraFLC2 (including the 2.0-kb promoter region and 3.4-kb gene region) were sequenced from Tibetan turnip and oilseeds. Compared to the BraFLC2 in turnip, five single nucleotide polymorphisms (SNPs) and one short indel were identified in the promoter region of oilseed BraFLC2. The gene regions of oilseed BraFLC2 contained five SNPs, one indel, one short indel and one insert. Of the five SNPs and one short indel, two were located in the exon region (exon 4 and exon 7), including one synonymous mutation and one nonsynonymous mutation, ultimately resulting in one amino acid substitution. One insert (29 bp) was located in intron 6. One indel (55 bp) was located in exon 4 and intron 4, resulting in a deletion of 18 bases in the exon and a deletion of 37 bases in the intron, ultimately causing a deletion of six amino acids in the coding region of exon 4 (Fig. 4a and b).

We hypothesized that the 55-bp indel and 29-bp insert may cause alternative splicing events in oilseed BraFLC2 mRNA transcripts. To test this hypothesis, the mRNA products of BraFLC2 from oilseeds were cloned and sequenced, and at least five abnormal alternative splicing isoforms were detected, all of which occurred downstream of the deletion region in exon 4. Specifically, in addition to a six-amino-acid deletion in the terminus of exon 4, one alternative splicing isoform also caused the deletion of exon 6, and one of the other alternative splicing isoforms resulted in the retention of intron 5 (Figs. 4c and S2). These results implied that the genetic variations in BraFLC2 in oilseeds resulted in both a decline in its expression level and the appearance of some abnormal alternative splicing events that ultimately caused early flowering time and a spring-annual habit in oilseeds.

3.5. Cross-pollination under human planting activities is one of the main reasons for early flowering of turnip in the Qinghai-Tibet Plateau

Our studies have figured out that the dominant variations existed at BraFLC2 loci between turnip and oilseeds, and selecting BraFLC2 as the testing gene can easily identify whether the early flowering turnip was polluted by cross-pollination with oilseeds. Gene-specific primers of BraFLC2 gene were designed to test the heterozygous of collected early-flowering turnip plants, and compared it with turnip and oilseeds. Since these plants had already flowered, we cloned and sequenced the BraFLC2 genomic fragments rather than detecting the expression level of BraFLC2. Interestingly, we observed obvious double peaks in the sequencing peak diagram of most of the early-flowering plants by sequencing alignment (Fig. S3). According to the base sequence obtained in double peaks by Sanger sequencing, we found that one of the BraFLC2 alleles in early-flowering plants was the same as the turnip BraFLC2, and the other BraFLC2 allele was the same as the oilseeds BraFLC2 (Fig. 5). These results suggested that the BraFLC2 sequence in these early-flowering plants was heterozygous. Considering that turnip and oilseeds are the main crops the cultivated in the QTP, we speculated that the early-flowering phenomenon observed in turnip planting may be due to cross-pollination with nearby oilseeds.

To confirm that cross-pollination by oilseeds causes early flowering in turnips, we performed an artificial hybridization experiment under laboratory conditions. Flowered Tibetan turnip and oilseeds were crossed to obtain F1 hybrid seeds (Fig. 6a). We observed that F1 generation plants produced tiny enlarged taproots and quickly flowered (55 days after sowing in soil), which was consistent with the early-bolting phenotype we found in the field in the QTP (Fig. 6b). The BraFLC2 sequence was heterozygous (Fig. S4), which was also detected in the early-flowering turnip plants in the field. We also examined BraFLC2 mRNA transcript levels in oilseed parents, turnip parents, and F1 generation seedlings at the vegetative growth stage (Fig. 6c). Our data showed that in the F1 plants BraFLC2 was expressed at significantly lower levels than in the turnip parents but at higher levels than in the oilseed parents. As FLC has been shown to have a dominant dosage effect (Michaels and Amasino, 1999), our results suggested that after cross-pollination, the decreased expression level of BraFLC2 ultimately shortened flowering time in offspring plants.

4. Discussion

Previous studies have shown that cross-pollination between different varieties affects crop seed purity and even causes gene transfer (Wilkinson et al., 2000, 2003; Heenan and Dawson, 2005). It's difficult to maintain the pure characteristics of potato and rice in the field because of cross-pollination between different varieties under human planting activities (Qiu et al., 2020; Ma et al., 2021). In Brassica rapa, spontaneous hybridization is known to readily occur between morphologically distinct but closely related intraspecific taxa. Specifically, oilseed plants cross easily and spontaneously with cultivated turnip (Davey, 1939; Heenan et al., 2004). In this study, we revealed that cross-pollination between oilseeds and cultivated turnips in the QTP had caused early flowering in turnip plants. We found that BraFLC2, a gene critical to delaying the transition to flowering, was expressed at high levels in turnip during the vegetative growth period and taproot development, but was undetectable at flower bolting. In oilseed plants, BraFLC2 was expressed at low levels throughout development. Furthermore, we found evidence that oilseed BraFLC2 transcripts underwent alternative splicing. Genomic sequence analysis indicated that early-bolting turnips possessed two alleles of BraFLC2, one of which is identical to oilseed BraFLC2. When we crossed turnip and oilseed plants to generate hybrids heterozygous for these alleles, we found that hybrid plants expressed BraFLC2 at low levels, resulting in early-bolting turnips.

FLC2 is a key gene that maintains late bolting and flowering in Raphanus and Brassica crops, such as turnip, cauliflower, and Chinese cabbage (Zhao et al., 2010; Wu et al., 2012; Ridge et al., 2015; Zheng et al., 2018). FLC2 functionality may underpin differences between winter-annual and rapid-cycling Brassica cultivars, especially with respect to flowering time. Nucleotide polymorphisms in the genomic region of the FLC2 gene are also closely related to flowering time variation in B. rapa crops. Previous studies have proposed that allelic variation in the Brassica napus FLC2 gene has contributed to seasonal crop-type divergence (Yin et al., 2020) and variation in vernalization requirement and response (Tudor et al., 2020). Our findings are consistent with these results. Much of the flowering time variation within B. rapa and within other Brassica species may be due to the combined effects of allelic variation at multiple FLC loci (Calderwood et al., 2021). Variations in BraFLC2 genomic sequences and expression levels seem to have caused differences in flowering time of turnip and oilseeds. Our results also indicate that genetic mutation of BraFLC2 in Tibetan oilseeds resulted in low expression levels of BraFLC2 and at least five types of alternative splicing isoforms, ultimately causing the early flowering and rapid-cycling habit of oilseeds. The introduction of oilseeds genome into turnip genome by cross-pollination resulted in the decline of BraFLC2 expression level in hybrid plants. We speculate that this decline might also be caused by changes in the expression of genes that regulate BraFLC2.

We observed that the cultivation pattern of turnip under local agriculture production activities in the QTP was very extensive, resulting in many early-flowering turnips with impure germplasm resources. In addition, the turnip landraces cultivated in Xizang are usually low-quality and low-yielding varieties. Furthermore, local famers generally produce turnip seeds cultivated in Xizang. Consequently, hybrid seeds may be retained without evaluation or screening of crop purity standards. Therefore, it is necessary to improve the level of turnip breeding and speed up the process of hybrid utilization. We recommend that farmers buy commercial seeds with stable quality and pure genetic backgrounds instead of producing their own seeds. Secondly, we recommend that farmers plant turnips independent of other crops and increase isolation distance from other Brassica crops. Finally, the main varieties of cultivated turnips should be derived from excellent high-yielding seed resources and the area on which these varieties are cultivated should be gradually expanded.

In conclusion, our analyses showed that the molecular and genetic factors regulating early flowering in turnip in the growing season are influenced by human activities related to crop planting in the QTP. These findings should enable the development of strategies that improve the cultivation and management of turnips.

Acknowledgments

This research was supported by the National Natural Science Foundation of China (no. 32200306, 32170385 and 32070362), the Postdoctoral Directional Training Foundation of Yunnan Province, and the Second Tibetan Plateau Scientific Expedition and Research (STEP) program (no. 2019QZKK0502).

Author contributions

YPY and XXK conceived and designed the experiments. YZ analyzed the sequencing data and wrote the manuscript. YZ and LDL performed the experiments. XSYL, QC and YY participated in the sample collecting and data analysis. XXK and YWD revised the paper. All authors have read and agreed to the published version of the manuscript.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.pld.2023.04.002.