Single-Cell Transcriptomic Analysis of Flowering Regulation and Vernalization in Chinese Cabbage Shoot Apex

Chinese cabbage development, the interplay between shoot apex activity and vernalization is pivotal for flowering timing. The intricate relationship between various cell types in the shoot apex meristem and their roles in regulating flowering gene expression in Chinese cabbage is not yet fully understood. A thorough analysis of single-cell types in the Chinese cabbage shoot apex and their influence on flowering genes and vernalization is essential for deeper insight. Our study first established a single-cell transcriptomic atlas of Chinese cabbage after 25 days of non-vernalization. Analyzing 19,602 single cells, we differentiated them into 15 distinct cell clusters using established marker genes. We found that key genes in shoot apex development and flowering were primarily present in shoot meristematic cells (SMC), companion cells (CC), and mesophyll cells (MC). The MADS-box protein FLOWERING LOCUS C 2 ( BrFLC2 ), a gene suppressing flowering, was observed in CC, mirroring patterns found in Arabidopsis . By mapping developmental trajectories of SMC, CC, and MC, we elucidated the evolutionary pathways of crucial genes in shoot apex development and flowering. The creation of a single-cell transcriptional atlas of the Chinese cabbage shoot apex under vernalization revealed distinct alterations in the expression of known flowering genes, such as VERNALIZATION INSENSITIVE 3 ( VIN3 ), VERNALIZATION 1 ( VRN 1 ), VERNALIZATION 2 ( VRN 2 ), BrFLC , and FLOWERING LOCUS T ( FT ), which varied by cell type. Our study underscores the transformative impact of single-cell RNA sequencing (scRNA-seq) for unraveling the complex differentiation and vernalization processes in the Chinese cabbage shoot apex. These insights are pivotal for enhancing breeding strategies and cultivation management of this vital vegetable.


Introduction
Chinese cabbage (Brassica rapa L. ssp.pekinensis), also known as heading or wrapping cabbage, is a leafy vegetable from the crucifer family with significant economic value as a globally cultivated crop [1].This vegetable is known for its rapid growth and development, making it crucial in agriculture.Its flowering timing is closely linked to environmental factors, especially the interaction between shoot apex activity and vernalization.Vernalization, the exposure to prolonged cold temperatures to trigger spring flowering, is vital for many temperate plants [2].This process is key for optimal crop production and highlights the complexity of flowering time regulation mediated by vernalization.Understanding this mechanism in Chinese cabbage involves dissecting the cellular and molecular dynamics influenced by vernalization.
In the realm of plant developmental biology, the shoot apical meristem (SAM) of Chinese cabbage is a dynamic and complex structure that orchestrates the vegetative growth phase and transition to flowering.The molecular mechanisms of vernalization have been well-studied in model plants, including Arabidopsis, identifying key regulators such as FLC (FLOWERING LOCUS C) and VRN1 (VERNALIZATION 1) [3,4].However, the intricate processes and regulatory networks in Chinese cabbage are not yet fully understood.Single-cell RNA sequencing (scRNA-seq) provides a unique opportunity to dissect the vernalization-mediated mechanism of flowering time control.By providing a high-resolution view of the cellular complexity and regulatory networks, scRNA-seq enables a distinctive opportunity to evaluate the response to vernalization.Although substantial research has explored the SAM regulatory mechanisms in model plants, including Arabidopsis, the processes in Chinese cabbage are less understood.Our study addresses this gap by employing innovative scRNA-seq techniques.Previous research has shown that vernalization involves the complex interplay of genetic and epigenetic modifications, leading to the stable repression of floral repressors and the activation of flowering pathways [2,5].However, these findings need to be expanded and validated in species, such as Chinese cabbage to understand their universal applicability and species-specific variations.
The emergence of scRNA-seq technology has transformed our grasp of cellular complexity and regulatory mechanisms in plants, making it an ideal tool to explore vernalization-mediated flowering time control.Single-cell studies in plants are no longer limited to the leaves [6], shoot apex [7], stomata [8], and roots [9][10][11] of Arabidopsis but have gradually been extended to other species.scRNA-seq has been employed to uncover the cellular map of Populus xylem roots [12], to identify strong cell-type markers and specific regulatory programs in legume root cells [13], and to characterize key transcription factors (TFs) in allotetraploid peanut leaves [14].The use of single cells has provided unique insights into organ growth and development, differentiation processes, tissue-specific responses to abiotic stress, cell-type-specific inheritance patterns, responses to biotic stress, distinct cell-type reactions to genetic changes, and the dynamics of cell cycle regulation dynamics.The ability of scRNA-seq to reveal cell-type-specific responses and regulatory programs makes it particularly suited for investigating the complex developmental processes influenced by vernalization in Chinese cabbage.This novel application of scRNA-seq is starting to shed light on these processes, suggesting that vernalization affects flowering time by modulating the activity and interaction of specific cell types within the shoot apex.
In this study, we aimed to bridge this knowledge gap by employing scRNA-seq to dissect the cellular landscape of the Chinese cabbage shoot apex.Our objective was to create a comprehensive single-cell transcriptomic atlas of Chinese cabbage that

Single-cell RNA sequencing of cells from the Chinese cabbage shoot apex
To perform scRNA-seq on the Chinese cabbage shoot apex, we collected approximately 200 shoot apexes from 25-day-old non-vernalized (N25) Chinese cabbage and observed them under a stereomicroscope (M125, Leica, Germany).The cells were converted into protoplasts (Chinese cabbage shoot apex cells without cell walls) and analyzed using scRNA-seq on the 10 × Chromium platform (10 × Chromium) (Fig. 1A).A total of 19,602 individual cells were obtained after the cell filtering process (Supplemental Fig. S1, Supplemental Table S1), and 48,407 genes were identified (Supplemental Table S2).Different cell populations were identified using known marker genes and classified into 15 distinct clusters (Fig. 1B, Supplemental Table S3 and Table S4).
The t-SNE and UMAP algorithms were used to illustrate local similarities and overall cell population structures (Fig. 1B and Supplemental Fig. 1B).We used reported marker genes to identify the cell types of different clusters of Chinese cabbage shoot apexes and revealed seven broad populations: primordia cells (PCs), mesophyll cells (MCs), vascular cells (VCs), epidermal cells (ECs), companion cells (CCs), shoot meristematic cells (SMCs), and guard cells (GCs) (Fig. 1B and 1C).The

Characterization of the novel marker gene in each cell cluster
To identify marker genes, we compared the associated upregulated differentially expressed genes (DEGs) across one cluster with those in other clusters, and the DEGs were distributed at a range from 156 to 1797 per cluster (Supplemental Table S5).We analyzed the annotated clusters using Gene Ontology (GO) and found that the PC population with the largest share of DEGs was mainly annotated in 'response to stimulus', 'response to biotic stimulus', and 'response to stress'.Consistent with the cell cluster classification results, GO term analysis of the MC population was mostly organelle related (Supplemental Fig. S3).Only a few plants, including Arabidopsis, rice, and peanut, have databases for cell type identification.The marker genes from these databases were not entirely sufficient for cell type identification in Chinese cabbage, thus making it a critical step in Chinese cabbage to determine cell type-specific expression of new genes.After analysis of the DEGs, we identified and designated the five most highly expressed genes in each cluster as new marker genes for those clusters (Supplemental Table S5, Supplemental Fig. S4).In total, we identified 60 new marker genes, which were visualized in a heatmap (Fig. 2A).The top marker gene for each cluster was determined using t-SNE mapping (Fig. 2B).
These new marker genes will be helpful in the identification of different cell types in future shoot apex studies in Chinese cabbage and even cruciferous plants.

Arabidopsis
Arabidopsis shoot apex scRNA-seq datasets were already available [7].We investigated the conservation and differentiation of Chinese cabbage and Arabidopsis shoot apex cell types in various genera within the same family.We also explored the distribution of conserved and specifically expressed genes in conserved cell types.We used the published Arabidopsis shoot apex shoot 4 scRNA-seq data (BioProject PRJCA003094) and reperformed cell clustering and cell group identification on this dataset using the same analysis pipeline for Chinese cabbage and identified 19 clusters (Fig. 3A and 3B, Supplemental Table S6).We then identified one-to-one homologous genes between Chinese cabbage and Arabidopsis using OrthoMCL [36] to compare gene expression at the cellular level (n=19,145, Supplemental Table S7).A pairwise comparison of the Chinese cabbage and Arabidopsis cell populations revealed significant correlations between the clusters corresponding to CC.The correlation of MC, SMC, and VC was also found in the corresponding clusters of Chinese cabbage and Arabidopsis (Fig. 3C).Interestingly, the MC and SMC clusters of Chinese cabbage also correlated with other cell clusters in Arabidopsis (Fig. 3C).
The results offer valuable insights into evolutionary conservation and cell divergence in Chinese cabbage and Arabidopsis.
Next, we integrated the shoot apex scRNA-seq data of Chinese cabbage and Arabidopsis for cell clustering analysis, which yielded 44,430 cells and 19,145 genes and revealed 20 cell clusters (Supplemental Fig. S5, Supplemental Table S8).As Chinese cabbage and Arabidopsis shoot apexes were correlated with CC, MC, and SMC cell populations (Fig. 3D and 3E), we explored genes associated with shoot development and flowering in these three cell populations.For the three cell populations, SMC, CC, and MC (Fig. 3F, Supplemental Table S9, Table S10, and Table S11), gene clustering analyses identified a set of conserved expression patterns and specific expression genes in Chinese cabbage and Arabidopsis.We found that genes related to shoot development and flowering were primarily present in conserved expression regions (A = C) in Chinese cabbage and Arabidopsis (Fig. 3F).Many previous studies in maize (Zea mays), Arabidopsis, and other species have revealed the importance of the KNOXI transcription factor SHOOT MERISTEMLESS (STM) in the establishment and maintenance of the SAM [37][38][39][40].Here, the STM genes of Chinese cabbage and Arabidopsis belonged to the conserved expression genes of the SMC cell population, and it was evident that the STM gene also played a significant role in the SMCs of Chinese cabbage (Fig. 1C).In Arabidopsis, MADS-box protein FLOWERING LOCUS C (FLC) acts by repressing a series of flowering genes to Downloaded from https://academic.oup.com/hr/advance-article/doi/10.1093/hr/uhae214/7723780 by guest on 02 August 2024 suppress rapid flowering [41].Four homologous FLC genes (BrFLC1, BrFLC2, BrFLC3, and BrFLC5) were identified and validated in Chinese cabbage compared to Arabidopsis [42,43].It was evident that, in contrast to other BrFLC types, BrFLC2 (BraA02g00340.3C) was the most similar class with respect to AtFLC and was present in multiple cell populations.In contrast, BrFLC2 was not only present in the conserved cell population CC but also in the PC and SMC populations in Chinese cabbage and in the EC, MC, and VC populations in Arabidopsis (Fig. 3G).This suggests that BrFLC2 is important in both Chinese cabbage and Arabidopsis.The inter-species comparisons analysis revealed conserved expression genes between the two species, suggesting that the shoot apexes of the different species have conserved characteristics.
Here, we discovered that the BrFLC2 (BraA02g00340.3C)gene is highly conserved in both Arabidopsis and Chinese cabbage and is present in numerous cell populations.We focused on understanding the mechanism and function of BrFLC2, a gene encoding a MADS-box protein involved in flowering.The coding sequence of Chinese cabbage BrFLC2 was 591 bp and contained 7 exons and 6 introns (Fig. 4A).
Phylogenetic analysis showed that BrFLC2 was similar to RsFLC, AtFLC, and CgFLC in group II (Fig. 4B).Experiments revealed that BrFLC2 proteins localize in the nucleus when expressed in tobacco leaves (Fig. 4C).We further explored the function of BrFLC2 by overexpressing it in Arabidopsis flc mutants (Fig. 4D).This led to slower growth and delayed flowering compared to the flc mutants, with the growth rate and flowering time being similar to those of the WT (wild type) (Fig. 4E   and 4F).The expression of BrFLC2 in the BrFLC2-OX line was significantly higher compared to the flc and WT lines, where BrFLC2 expression was negligible, approximately zero.Additionally, the expression of the flowering genes AtFT and AtSOC in the BrFLC2-OX line was significantly lower than in the flc line, tending towards the levels observed in the WT.This indicates that the ectopic expression of BrFLC2 affected the expression of Arabidopsis-related flowering genes (Fig. 4G-4I).Simultaneously, we discovered that the BrFT gene in Chinese cabbage was exclusively expressed in the CC population, while BrSOC was expressed in all cell groups (Fig. 4M).This suggests that BrFLC2 influenced the expression of BrFT in the CC population of Chinese cabbage.At the same time, BrFLC2 overexpression delayed the days to bolting, flowering, and seed setting, similar to the WT (Fig. 4J-4L).Further experiments, including yeast two-hybrid (Y2H) and luciferase complementation (LUC) assays, identified an interaction between BrFLC2 and BrMSI4 (Fig. 4N and 4O).Initially, we conducted yeast two-hybrid screening and identified 23 proteins that interacted with BrFLC2 (Supplemental Table S12).
Following a comprehensive literature search and functional annotation, we found that BrMSI4 was associated with flowering regulation.BrMSI4 acts as a DDB1-and CUL4-associated factor that interacts with the CLF-polycomb repressive complex 2 (PRC2) to repress FLC expression [44].Consequently, we selected BrMSI4 for further validation due to its potential role in flowering control.This study enhances our understanding of BrFLC2's function in the flowering process of both Chinese cabbage and Arabidopsis, indicating that the CC cell group might be crucial in plant flowering and merits additional research.

Analysis of pseudo-time trajectories in Chinese cabbage shoot apex cells
To examine the spatial and temporal distribution of individual shoot apex cells in Chinese cabbage, we used Monocle 2 software [45] for pseudo-time trajectory analysis.This approach allowed us to illustrate the placement of each cell cluster along the main stem (Fig. 5A and 5B, Supplemental Table S13).The pseudo-time analysis also identified five key marker genes (OPR1, GRP3, AFP4, RBCS and BraA03g00670.3C)(Fig. 5D, Supplemental Table S14) that can categorize all individual cells into five unique states (Fig. 5C) of shoot apex development and differentiation.
Based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) presentation, the pathways enriched in branch 1 were primarily involved in 'ribosome,' 'glutathione metabolism,' 'oxidative phosphorylation,' and the 'MAPK signaling pathway-plant,' while branch 2 was highly concentrated in ribosome (Supplemental Fig. S6C and   S6D).The branching provided a more intuitive observation for visualizing changes in cell expression.A total of 32 DEGs associated with the flowering and shoot development in the two differently branches were identified (Supplemental Fig. S6E).
Interestingly, nearly all of these DEGs (31) were in branch 1, and 1 DEG (1-sep) was present in both branches (Supplemental Fig. S6E).The CC population was also enriched for several flowering and shoot development genes that were highly expressed in the CC population compared to other cell populations, with correspondingly high expression in state 5 (Fig. 5F-H).The main pseudo-time trajectory pathway for shoot development and flowering-related genes was shown to be the CC population differentiation pathway.This also corroborated our conclusion drawn from cross-species analysis that a large number of genes associated with flowering might have existed in the CC population.

Differentiation trajectories of shoot and flowering genes
Many flowering and shoot development genes existed in the MC, SMC, and CC populations, and we speculated that these three cell populations had interdependent developmental relationships.To test this hypothesis, the developmental trajectories of MC, SMC, and CC cell clusters were analyzed (Fig. 6A).By studying the differentiation orientation of the cells along the timeline, pseudo-time trajectory data revealed that SMC and MC accompany each other (Fig. 6B).However, SMC are Downloaded from https://academic.oup.com/hr/advance-article/doi/10.1093/hr/uhae214/7723780 by guest on 02 August 2024 essentially undifferentiated cells within the plant growth point capable of differentiating into MC, suggesting that MC can arise from SMC at any moment and location.Positioned at the initial segment of the timeline, MC may also originate from other cell types, such as PC.Simultaneously, CC have the potential to differentiate from both SMC and MC (Fig. 6B).The backbone along the pseudo-time trajectory was divided into five states and two branches (Fig. 6C).Monocle 2 [45] is adept at harnessing the gene expression signals present across all cells, and by analyzing the gene expression profiles characteristic of cells in distinct differentiation states, it adeptly identifies DEGs across various states and branches of differentiation.As a result, the pseudo-time trajectory of the main stem revealed 3862 DEGs (Fig. 6D, Supplemental Table S16), and 2109 DEGs were identified in the two branches, with 485 DEGs in branch 1 and 1624 DEGs in branch 2 (Supplemental Table S17).Sixty genes related to flowering and shoot development were screened from the DEG profile, and most of the related genes were highly expressed in the CC population, i.e., state 4 (Fig. 6E).We then selected six representative flowering genes for description in state 4 (companion cell population) (Fig. 6F).The six representative genes were not only highly expressed in state 4, but with the pseudo-time trajectory, their expression was also more skewed towards branch 2 (states 1, 3, 4) (Fig. 6G).The SMC population, representative of SAM tissue cells in Chinese cabbage, was theoretically presumed to predominantly encompass genes related to flowering and shoot development.These flowering and shoot development genes were primarily observed in the CC population, exhibiting high expression levels.This observation led to the speculation that the hypothesized differentiation trajectory of the CC population might not exclusively originate from the mesophyll cell (MC) but could also differentiate from SMCs, as depicted in Fig. 6H.Overall, this trajectory-based analysis delineates a hypothetical model for the differentiation of MC, SMC, and CC cell types, simultaneously reaffirming the importance of the CC population in regulating genes crucial for flowering and shoot development.

Vernalization-induced transcriptomic changes vary among cell types
To investigate the cellular heterogeneity of the Chinese cabbage shoot apex in response to vernalization, we isolated protoplasts from the shoot apex vernalized for 25 days (V25), as detailed in Supplemental Table S18, with N25 serving as the control (Fig. 7A).Notably, a significant disparity in the number of cells captured between N25 and V25 prompted a cell frequency analysis and sample cluster recognition.This analysis confirmed regular cell frequency and essentially identical cell types in both samples (Supplemental Fig. S7).Furthermore, differential expression analysis of genes between the groups revealed no excessive deviation (Supplemental Fig. S7A).This finding aligns with previous studies [46,47], which noted a substantial difference in the number of cells among single-cell groups, allowing our comprehensive analysis to proceed as expected.Utilizing subpopulation data from both samples, we conducted an intergroup differential analysis.DEGs were identified for each cell type under the control and vernalization treatments, considering a difference in the mean expression level of |log2FC| ≥ 0.36 and P<0.05 (Fig. 7B, Supplemental Table S19).
Excluding overlap and unknown cell clusters, compared with N25, the PC cluster of V25 exhibited the highest number of DEGs (2635 upregulated, 3300 downregulated), followed by the MC cluster (1100 upregulated, 670 downregulated), with the GC cluster displaying the fewest DEGs (10 upregulated, 16 downregulated).
Our focus was specifically on a cohort of genes intimately associated with vernalization, namely VERNALIZATION INSENSITIVE 3 (VIN3), VERNALIZATION 1 (VRN 1), VERNALIZATION 2 (VRN 2), BrFLC, and FLOWERING LOCUS T (FT) [41,[48][49], across various cell types in two samples.We identified three VIN3 genes, two of which (BraA03g012460.3C and BraA06g040160.3C) exhibited high expression in the V25 samples of the EC population and one (BraA02g012310.3C)present in the SMC population of the V25 samples (Fig. 7C).Additionally, BraA01g033970.3C (VRN1) and BraA05g028310.3C (VRN1) showed high expression in the SMC and CC populations of the V25 samples, respectively.In contrast, the Downloaded from https://academic.oup.com/hr/advance-article/doi/10.1093/hr/uhae214/7723780 by guest on 02 August 2024 expression of BraA05g038610.3C (VRN1) was marginally higher in the N25 samples of the MC, VC, and EC populations than in V25.Of the two VRN2 genes, only BraA01g019670.3C was highly expressed in the EC population of the NV samples.
Among the detected FLC genes, only FLC2 and FLC3 (BraA03g004170.3C)adhered to vernalization-regulated expression, which was substantially expressed in the CC population of the V25 sample.Conversely, the flowering gene FT (BraA02g016700.3C)was notably expressed in the CC of V25 due to vernalization, which aligns with the notion that vernalization fosters the expression of flowering genes.
We compared DEGs across six cell populations-MC, PC, VC, EC, CC, and SMC-and identified 60 DEGs in all populations between N25 and V25 (Fig. 7D, Supplemental Table S20).GO and KEGG analyses of these genes predominantly revealed significant enrichment in categories related to environmental adaptation and 'stimulus' responses (Supplemental Fig. S8A and S8B), signifying vernalization's notable influence across various cell types.Among these shared differential genes, we identified five TFs: three ERFs, one C3H, and one NAC.The NAC gene was more highly expressed in the V25 samples, whereas the other four TFs showed higher expression in the N25 samples (Fig. 7E).Additionally, we conducted a detailed screening of DEGs in each cell type of Chinese cabbage for genes associated with shoot development and flowering, followed by GO annotation.Our findings revealed that genes linked to 'reproductive shoot system development (GO:0090567)' and 'shoot system development (GO:0048367)' were differentially expressed in all populations except GC.Moreover, two genes with a GO enrichment of 'negative regulation of flower development (GO:0009910)'-BraA04g023000.3C(SOBIR1) and BraA04g023010.3C (SOBIR1)-exhibited significant differences in the PC, MC, and VC clusters of the two samples.Investigating their reduced expression in other cell populations due to vernalization, it appeared that SOBIR1 may function similarly

Discussion
This study provides a detailed single-cell atlas of the vegetative shoot apex in Chinese cabbage, revealing the complex cellular architecture of this non-model plant.
Employing advanced scRNA-seq technology, we explored the cellular diversity within and across cell types in the plant.This innovative approach marks a significant stride in the study of complex biological processes, especially in non-model species.
We meticulously characterized numerous cell clusters and state variations in the shoot apex, driven by a wide range of cell marker genes.This investigation not only demonstrates the feasibility of applying scRNA-seq to non-model organisms but also expands the potential for functional studies in the shoot apex of such species.Our findings underscore the pivotal role of vernalization in its developmental processes.
The insights gained are crucial in enhancing our understanding of vernalization, opening up potential avenues for developing varieties with optimized flowering times and increased environmental adaptability.
We evaluated the pivotal role of the shoot apex in the growth and development of Chinese cabbage, a widely cultivated leafy vegetable.To this end, we utilized shoot apex data from the N25 treatment, analyzing 19,602 individual cells and 48,407 genes.These were classified into 15 distinct cell clusters based on specific marker genes.
Our analysis identified various cell types, namely PCs, MCs, VCs, ECs, CCs, SMCs, and GCs, which are instrumental for studying the developmental dynamics of Brassica species' shoot apex tissues (Fig. 1).We conducted a cross-species comparative analysis using published Arabidopsis vegetative shoot apex scRNA-seq data alongside our own findings, leading to the identification of highly correlated cell populations and conserved gene expressions between these two species (Fig. 3 and Supplemental Fig. S5).Notably, the CC population exhibited a significant correlation across both species, with substantial parallels observed in the MC and SMC clusters.Furthermore, by integrating and comparing the scRNA-seq data from Chinese cabbage and Arabidopsis, we discovered numerous conserved genes involved in flowering and shoot development, particularly within the MC, SMC, and CC populations.This suggests a high degree of evolutionary conservation of these genes, reinforcing the value of cross-species studies in understanding plant development.For instance, the STM gene, a hallmark SMC marker in Arabidopsis [34], also plays a significant role in the SMC of Chinese cabbage, validating our cell group categorization.Similarly, the FLC gene, known as a classical flowering suppressor in Arabidopsis and involved in multiple flowering regulatory pathways [41].The FLC functions as a floral repressor gene, delaying the transition from vegetative to reproductive growth in plants.The role of AtFLC in Arabidopsis has been extensively characterized, particularly its primary targeting of three genes that influence flowering time: FLOWERING LOCUS T (FT), SUPPRESSOR OF OVEREXPRESSION OF CONSTANS 1 (SOC1), and FLOWERING LOCUS D (FD) [3,50].Furthermore, at the level of epigenetic regulation, vernalization is controlled by H3K4me3 and H3K36me3 modifications, which form protein complexes that negatively regulate AtFLC in response to vernalization.Research on AtFLC has been progressively advancing.However, in Chinese cabbage, which possesses four BrFLC paralogs [42], a comprehensive annotation of BrFLC remains elusive [41].Our cross-species analysis revealed that BrFLC2 (BraA02g003340.3C) in the CC population of Chinese cabbage had a highly conserved expression pattern with the FLC gene in Arabidopsis.
Functional validations of BrFLC2 (Fig. 4) confirmed its role as a negative regulator of flowering in Chinese cabbage.Single-cell RNA sequencing provided a detailed perspective on BrFLC2 expression in specific cell types within the shoot apex, highlighting its regulatory mechanisms.The conservation between BrFLC2 and FLC underscores BrFLC2's crucial role in vernalization and flowering regulation, presenting significant potential for breeding programs to develop cultivars with desired flowering traits.We evaluated the pseudo-time trajectory of shoot apex development in Chinese cabbage (Fig. 5).In this process, DEGs among various cell populations were classified into five distinct states and bifurcated into two branches, reflecting the dynamic nature of the developmental trajectory.Intriguingly, our analyses of these states and branches revealed that a substantial majority of genes associated with flowering and shoot development were predominantly located in branch 1 and state 5, corresponding to the CC population (Supplemental Fig. S6E).This observation aligns with findings in Arabidopsis, where FT, a key component of florigen, has been shown to transmit photoperiodic flowering signals from leaf companion cells to the shoot apex [51].This led us to hypothesize that the enrichment of flowering-and shoot development-related genes in the CC population of Chinese cabbage might similarly signal the translocation of specific flowering cues from companion cells to the shoot apex and promote bolting and flowering.To substantiate our hypothesis, we conducted additional pseudo-time trajectory analyses focusing on the MC, SMC, and CC groups.These analyses further supported our initial speculation.Additionally, examining the distribution of relevant flowering and shoot development genes (Fig. 6) allowed us to conclude confidently that our hypothesis was accurate.During the seed germination stage, most of the genes pertinent to Chinese cabbage are concentrated in the SMC population.Subsequently, as Chinese cabbage growth progresses, these genes likely transition to MC and CC, thereby facilitating rapid bolting and flowering.This comprehensive analysis not only elucidates the developmental dynamics within the shoot apex of Chinese cabbage but also underscores the critical role of cellular heterogeneity in the orchestration of key developmental processes.Additionally, our comparative analysis shows the transformative impact of scRNA-seq in revealing the cellular mechanisms of vernalization in Chinese cabbage.This novel application of single-cell technology in the study of vernalization provides new opportunities for crop improvement and breeding strategies, especially in terms of manipulating flowering times and enhancing stress resistance.In our comparative Downloaded from https://academic.oup.com/hr/advance-article/doi/10.1093/hr/uhae214/7723780 by guest on 02 August 2024 analysis of Chinese cabbage shoot apexes under varying vernalization treatments, we observed that while vernalization did not significantly alter the intrinsic characteristics of cell types, it caused significant alterations in the relative proportions of cell-type-specific gene expression.Notably, the count of single cells in the vernalized samples was substantially lower than that in the non-vernalized samples (Supplemental Fig. S7).This discrepancy, alongside the observed consistency in cell types and frequencies, aligns with findings from previous studies [46,47], suggesting that the reduced cell number in vernalized samples is likely attributable to low temperature treatment.Thus, vernalization treatment adversely affects the status and quantity of single cells in the Chinese cabbage shoot apex.By conducting scRNA-seq on samples subjected to different treatments, we found heterogeneity and consistency in gene expression patterns post-vernalization.A greater number of DEGs were identified in the PC and MC populations compared to the others, with a notably lower number of DEGs in the GC population (Fig. 7B).Furthermore, the expression patterns of the five key genes involved in the vernalization pathway varied across different cell types in the two sample sets (Fig. 7C).Intriguingly, the CC population was enriched with DEGs related to the VRN1, FLC, and FT genes.This finding corroborates our previous observation that the CC population is a hub for genes involved in flowering, highlighting its critical role in the flowering pathway of Chinese cabbage.Previous studies have indicated that vernalization does not affect the expression of VRN1 and VRN2, but under vernalizing conditions, VRN1 and VRN2 can repress the expression of FLC genes [4,48].However, our findings suggest that the expression patterns of VRN1 and VRN2 varied across different cell populations in the NV and V25 samples.This indicates that while vernalization may not significantly alter the overall expression of VRN1 and VRN2 genes in the entire sample, it might affect the expression of VRN1 and VRN2 in specific cell types.Additionally, among the 60 DEGs identified across various cell populations, excluding GC, five TFs were prominent (Fig. 7E).Cell differentiation is often governed by transcriptional This finding highlights the importance of the CC population and suggests that it warrants more in-depth exploration in future studies.Additionally, the pseudo-time trajectory and DEG analyses of samples treated under different conditions have provided invaluable resources.These insights are pivotal for elucidating the function and evolutionary pathways of vernalization-related genes at the single-cell level.
Moreover, our findings have significant implications for the breeding of Chinese cabbage.By understanding the intricacies of gene expression associated with key developmental stages and processes, breeders can develop more efficient strategies for cultivating Chinese cabbage varieties with desired traits.This research not only contributes to the fundamental understanding of plant developmental biology but also offers practical applications in agricultural biotechnology, potentially leading to the enhancement of crop quality and yield.

Plant materials and growth conditions
Germination and plumule-vernalization treatments were applied to the bolting-resistant Chinese cabbage DH line 'Ju Hongxin' (JHX) [52].Eight hundred uniformly healthy JHX seeds were carefully selected, sterilized with water, and arranged in petri dishes lined with two layers of filter paper.Each dish contained 40 seeds, and there were a total of 20 dishes, equally divided for vernalization and non-vernalization treatments.To expedite germination, the seeds were incubated in a climate chamber set at a constant temperature of 25°C with 16 hours of light per day for two days.Once the radicles emerged, the dishes were split into two groups: one group was moved to a vernalization chamber at 4°C with a 22/2 hour light/dark cycle and 150 µmol m −2 s −1 light intensity for 25 days (V25 treatment), and the other group was kept in an artificial climate chamber at 25°C with the same light/dark cycle and light intensity for 25 days (N25 treatment) (Fig. 1A).S2).Red signifies high expression levels, while blue denotes low expression levels.(B) Expression patterns of 12 new marker genes distributed on the t-SNE map.The gradient color in each t-SNE plot represents the expression level of the gene, with darker points indicating higher expression and lighter points indicating lower expression.
Downloaded from https://academic.oup.com/hr/advance-article/doi/10.1093/hr/uhae214/7723780 by guest on 02 August 2024 has undergone non-vernalization and to clarify the roles of distinct cell types in the regulation of flowering genes.By comparing our findings with available data from model species, such as Arabidopsis, we seek to uncover conserved and unique aspects of shoot apex development and flowering gene regulation.Additionally, we have constructed a single-cell map of Chinese cabbage during vernalization, highlighting the transformative potential of scRNA-seq in understanding the complex interplay of cellular responses during vernalization.This study not only aims to advance our understanding of the developmental biology of Chinese cabbage, but also seeks to provide valuable insights for the effective management and breeding of this important crop.
Downloaded from https://academic.oup.com/hr/advance-article/doi/10.1093/hr/uhae214/7723780 by guest on 02 August 2024to FLC genes, inhibiting flowering under normal conditions and promoting flowering when expression is reduced due to vernalization.
Downloaded from https://academic.oup.com/hr/advance-article/doi/10.1093/hr/uhae214/7723780 by guest on 02 August 2024 regulation, and the expression of these five TFs varied among samples and cell populations under different vernalization treatments, providing valuable insights for future vernalization research.This observation highlights the importance of transcriptional regulation in determining cell differentiation and modulating the developmental trajectory of individual cell types [14].This analysis shows that vernalization differentially affects cell types within the shoot apex, thus advancing our understanding of the cellular mechanisms governing flowering genes in Chinese cabbage.Our findings not only provide a deeper understanding of the molecular mechanisms of vernalization in Chinese cabbage but also pave the way for future research focused on optimizing crop development and flowering through advanced single-cell technologies.This study serves as a stepping stone for further exploration into the functional interactions of key transcription factors and shoot apex development under diverse environmental conditions.The exploration and GO annotation of DEGs related to shoot development and flowering revealed distinct enrichment patterns in various cell types.Among these findings, the identification of the flowering-suppressor gene, SOBIR1 (Fig. 7F), opens up a new avenue for research on Chinese cabbage and its flowering traits.These insights are invaluable for future agricultural practices and breeding programs, as they provide a deeper understanding of the genetic and cellular foundations of plant development and vernalization.In summary, our study successfully established a novel and comprehensive gene expression profile of the Chinese cabbage shoot apex at a single-cell resolution.This achievement not only enhances precision in cell type identification and characterization using various cell markers but also lays a foundational framework for future research in cellular biology and genomics.Furthermore, our comparative analysis of the shoot apex in both Chinese cabbage and Arabidopsis has shed light on the conserved and divergent aspects of cell type function and development in these species.This comparison offers a fresh perspective on the role and evolutionary dynamics of cell types in Chinese cabbage.One of the pivotal discoveries of our Downloaded from https://academic.oup.com/hr/advance-article/doi/10.1093/hr/uhae214/7723780 by guest on 02 August 2024research is the proposal that the CC population in plants, particularly in Chinese cabbage, may be a crucial reservoir of a large number of genes involved in flowering.

Figure 1 .
Figure 1.Generation of a cell atlas for the Chinese cabbage shoot apex.(A) This schematic illustrates the isolation of protoplast cells from the Chinese cabbage shoot apex and their subsequent placement on the 10 × Genomics platform.(short scale bar represents 200 μm; long scale bar represents 1000 μm).(B) t-SNE visualization shows 15 identified cell clusters in the Chinese cabbage shoot apex.Each dot represents an individual cell, with colors indicating the corresponding clusters.(C) The bubble plot demonstrates expression patterns and distributions of cluster-specific genes, aiding in cell type identification within the Chinese cabbage shoot apex.These plots show both the average expression level (by color) and the proportion of cells expressing each gene (by dot size).

Figure 2 .
Figure 2. Discovery of novel marker genes in cell-type clusters.(A)The heatmap displays the top five differentially expressed genes (DEGs) with the highest log 2 TPM expression levels in each sub-cluster (Supplemental TableS2).Red signifies high expression levels, while blue denotes low expression levels.(B) Expression patterns of 12 new marker genes distributed on the t-SNE map.The gradient color in each t-SNE plot represents the expression level of the gene, with darker points indicating higher expression and lighter points indicating lower expression.

Figure 3 .
Figure 3.Comparison of Chinese cabbage and Arabidopsis shoot apexes at single-cell resolution.(A) t-SNE visualization shows 19 identified cell clusters in the Arabidopsis shoot apex.Each dot represents an individual cell, with colors indicating the corresponding clusters.(B) The bubble plot demonstrates expression patterns and distributions of cluster-specific genes, aiding in cell type identification within the Arabidopsis shoot apex.These plots show both the average expression level (by color) and the proportion of cells expressing each gene (by dot size).(C) Pairwise correlations of Chinese cabbage (top) and Arabidopsis (left) shoot apex cell clusters are shown, with dots indicating statistically significant correlations.Abbreviations: CC (companion cell), EC (epidermal cell), GC (guard cell), MC (mesophyll cell), PC (primordia cell), SMC (shoot meristematic cell), VC (vascular cell), UC (unknown cell).(D) t-SNE plot depicting cell clusters in Chinese cabbage and Arabidopsis shoot apex cells, with dotted circles marking common MC, CC, and SMC clusters.(E) Sankey diagrams showing the similarity of Chinese cabbage to Arabidopsis across cell clusters.All clusters were generated after merging the Arabidopsis and Chinese cabbage scRNA data on the left (Supplemental Fig. S5B).Cluster numbers for Chinese cabbage (Fig. 1B) and Arabidopsis (Fig. 3A) shoot apex cells are given on the right.(F) Gene clustering of the SMC, CC, and MC clusters.A=C indicate genes

Figure 4 .
Figure 4. Preliminary analysis of the molecular function of Chinese cabbage BrFLC2.(A) Illustration of the BrFLC2 genome structure.(B) Phylogenetic tree of BrFLC2 homologs in various plant species.(C) Subcellular localization of BrFLC2 protein in the tobacco nucleus (scale bar, 25 μm).(D) BrFLC2 coding sequences from BrFLC2-OX lines cloned by PCR.(E-F) Phenotypes of BrFLC2-OX, flc and WT lines grown in medium for 10 and 25 days after planting.(G-I) Relative expression of BrFLC2, AtFT, and AtSOC in BrFLC2-OX, flc and WT lines.Error bars indicate SE (n=3).(J-L) Days to bolting, flowering, and seed setting in BrFLC2-OX, flc and WT lines.Error bars indicate SE (n=10).(M) The expression pattern of BrFT and BrSOC in Chinese cabbage, as plotted on t-SNE.(N) Transcriptional activation function of BrFLC2 and BrMSI4 in yeast, with DDO representing SD/−Trp/−Leu, and TDO/X representing SD/−Trp/−His/−Leu medium supplemented with X-α-gal.(O) BrFLC2 and BrMSI4 interaction in tobacco epidermal cells, demonstrated by the luciferase complementation assay.

Figure 5 .
Figure 5. Pseudo-time trajectory analysis of cell types in the Chinese cabbage shoot apex.(A) Development trajectory of all shoot apex cells and the placement of each cell cluster in the trajectory map.(B) Placement of each individual cell cluster in the trajectory map.(C) Trajectory analysis divided single cells into five differentiation states.(D) Pseudo-time trajectory analysis of five key marker genes' expression patterns across five states.(E) Cell distribution within each cluster and pseudo-time trajectory.(F) Heatmap showing the average expression levels of flowering genes across all cell trajectories.Red represents a high expression level.(G) Heatmap showing average expression levels of flowering genes in five states, with red representing high expression and blue representing low expression.(H) The distribution of expression of eight flowering genes in the t-SNE map and heatmap.

Figure 6 .
Figure 6.Developmental trajectory of companion cells from mesophyll cells and shoot meristematic cells.(A-C) Distribution of cell clusters, differentiation states, and branches along the pseudo-time trajectory of mesophyll development.(D) Clustering and expression dynamics of DEGs along the main stem of the pseudo-time trajectory.(E) Heatmap showing the average expression of relevant genes across five cell differentiation states, with red representing high expression and blue denoting low expression.(F) Expression distribution of six representative flowering genes in the cell differentiation state.(G) Expression distribution of six representative flowering genes in different branches.(H) A putative model for the developmental and differentiation patterns of companion cells from mesophyll cells and shoot meristematic cells.