2022年1月3日，中国农业科学院蔬菜花卉研究所生物技术实验室在the Plant journal（2023IF=7.2）上发表了名为A cluster of transcripts identifies a transition stage initiating leafy head growth in heading morphotypes of Brassica的研究论文。论文第一作者为张亢和杨印庆，通讯作者为程锋研究员和王晓武研究员。

结球是许多园艺作物的重要性状，具有重要的经济价值。叶球的形成是一种特殊的叶形态发生，其特征是出现扩大的弯曲叶片。然后，转变到结球期的转录调控机制尚不清楚。本文通过对大白菜和甘蓝主要营养生长阶段的大规模时间转录组分析，以不结球白菜Taicai为对照。发现在大白菜中，从莲座叶生长到结球期的发育转换经历了一个转录调控过程。这个过程在不结球白菜中是不存在的，但是在结球甘蓝中也同样保守。我们提出，大白菜通过驯化获得了一个多叶的球状物，和其他卷心菜的起源不同，并且系统发育树分析也同样支持所有卷心菜的祖先都是不结球的。此外，我们还发现这个过渡极端的发生和环境温度有关，而且在过渡阶段，乙烯途径特别活跃。

Figure 1. Overview of the experimental design of the microRNA-sequencing assay.
(a) Dendrogram showing the phylogenetic relationships among heading morphotypes Chiifu and HN53, and non-heading morphotype Taicai WS. Gray fonts
denote other phylogenetically distant Brassica rapa morphotypes.
(b) Leaf sampling process. The third visible leaf (count from inside out, yellow color) was collected.
(c) Schematic overview of the experimental design for three morphotypes.

Figure 2. Identification of the transition stage in Chinese cabbage (Chiifu).
(a) Time-series transcriptomes of Chiifu leaves were clustered into four clades (indicated by different colors). PCC, Pearson’s correlation coefficient.
(b) Gene Ontology enrichment analysis revealed distinct functional enrichments of genes highly expressed in the stage T5–8 (II) and in the stage T9–13 (III). The
color of the circle indicates the statistical significance (here as P value), and circle size indicates the percentage of genes carrying certain GO term.
(c) Clustering of the time-series transcriptomes of the non-heading WS using the same method as (a).
(d) Gene expression heatmap for those genes that were significantly upregulated at least one time point among T5–13 as compared with both T1–4 and T14–18.
A gene set (marked with yellow lines in the left dendrogram) was specifically transcriptionally activated in the hypothetical transition stage.
(e) Activation of hormone-related genes in the transition stage (highlighted in yellow) is exemplified by several genes.
(f) Scenario of Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment at different time points. Enriched KEGG pathways were determined based on the
representative genes at each time point. The transition stage and corresponding KEGG pathways are marked by bold fonts on an orange background.

Figure 3. Conservation of the heading transition stage in heading Brassica rapa HN53 and in cabbage JQ1.
(a) Line chart illustrating the Pearson’s correlation coefficient distribution of different pairs (Chiifu vs. WS, HN53 vs. WS, and Chiifu vs. HN53).
(b) Plant hormone-associated genes exhibit different expression patterns in Chiifu and HN53. Kyoto Encyclopedia of Genes and Genomes (KEGG) genes were
upregulated during the transition stage of Chiifu as compared with WS. Their fold-changes (red) relative to WS in the transition and heading growth stages were
calculated. Their performances in HN53 (green) were also collected using the same method. DEGs, differentially expressed genes.
(c) Expression patterns of hormone-associated genes in Chiifu (left) and in HN53 (right) are exemplified by typical cases. Differences in expression patterns are
consistent with the tendency reflected by the boxplots (d). TPM, transcript per million.
(d) Snapshot of the heading cabbage JQ1.
(e) Sampling strategy for cultivar JQ1. In total, 12 leaf samples were collected with the sampling interval indicated, using days as the unit.
(f) KEGG gene enrichment based on the DEGs between heading stages and the rosette stage of JQ1. Upper panel shows the enriched KEGG pathways (orange)
based on genes that were downregulated in the heading stages compared with the rosette stage; bottom panel shows the pathways (green) based on genes that
were upregulated in the heading stages compared with the rosette stage. Orange highlights (in c left and f) denote the transition stage of Chiifu (T5–8) and JQ1
(C3–6), respectively, while the green highlight (c right) shows the potentially longer transition stage of HN53 (T5–12).

Figure 4. Delayed activation of heading regulatory genes in a higher temperature environment.
(a) Temperatures in the plastic shed and greenhouse. Chiifu in the shed was named CFP, while Chiifu in the greenhouse was named CFG.
(b) Delayed heading process of CFG as shown by photos.
(c) Histograms showing the variation in expression peak of genes in CFP and CFG. Upper panel shows the frequency of gene expression peaks locating in the
transition stage of CFP, while the bottom panel shows their corresponding expression peaks in CFG. The time point with the largest number of genes shifted
from T6 in CFP to T7 in CFG, indicating a delay in expression peak. (d) The vast majority of transition-activated Chiifu genes as identified above were delayed in
CFG. According to the dendrogram, these genes were divided into five clades. (e) Expression patterns of genes in five clades as identified in (d). Transition stage
is marked by bold fonts on an orange background. TPM, transcript per million.

Figure 5. Co-expressed gene clusters and pathways in the co-expression network built for Chinese cabbage.
(a) Snapshot of the co-expression network for Chiifu with genes highly expressed during different stages highlighted by different colors. Connections among
genes (here as edge) and the isolated gene pairs were omitted. Visualization was conducted using the “preferred layout” in Cytoscape. (b) Snapshot of the co-
expression network for WS, the non-heading control. The color of node was determined according to whether the genes marked with color in (a) were still
highly expressed in corresponding stage in WS. If they did, these genes were colored same as in (a). If not, these genes were colored as gray. (c) Two triple-
hormones-involved co-expressed clusters are indicated by circles with different colors (green and blue). The color of genes in the MAPK and microtubule clus-
ters were assigned according to the color of the hormonal cluster to which they connected in the network. Edge color indicates the value of the shortest path
(distance) between two genes in the co-expression network. Distance = 1 means that two genes are directly connected, while distance = 2 or 3 means that they
were indirectly connected via one or two bridging intermediate gene(s), respectively. Numbers labeled beside the intra?/inter-connections indicate the average
distance between two gene clusters. Smaller numbers indicate closer connections. The gene prefix “Br” was omitted from the gene name. Owing to the abun-
dance of multi-copy genes in B. rapa, being descended from a recent hexaploidy event, genes marked with same name indicate different copies, such as the
BES1s and EIN3s. BR, brassinosteroid

Figure 6. Comparison of the performances of the ethylene pathway in the Chinese cabbage and cabbage.
(a) The majority of the Kyoto Encyclopedia of Genes and Genomes-annotated ethylene-related genes were dramatically activated in Chiifu, while no similar phe-
nomenon was found for WS. Right-hand panel, small heatmap showing the frequencies of differentially expressed genes. The gene prefix “Br” was omitted
from the gene name. The transition stage is marked by bold fonts on an orange background.
(b) Expression heatmap of the cabbage genes that were orthologous to those shown in (a). TPM, transcript per million.
(c) Phenotypes caused by the ethylene treatment on the Chinese cabbage Chiifu. The variations in leaf angle are highlighted by the red acute angles, while the
incurved leaves are indicated by the yellow arrows. Photos were taken at 20 and 50 days after the ethylene treatment.
(d) Phenotypes caused by the ethylene treatment on cabbage JZS. The variations in internode length are indicated by the yellow lines. Photos were taken a t
15 days after the ethylene treatment

Figure 7. Hypothetical regulatory model to help explain the transcriptional dynamics during the heading differentiation process in Chinese cabbage.The rosette
stage is accompanied by enhanced photosynthesis and many other primary metabolisms. The transition stage (our primary contribution to this discussion), in
which the development of the folding and erect leaf begins, is considered a particularly important stage. Exogenous stimuli, such as low temperatures, are,
hypothetically, sensed by sensor proteins. Plant hormones and second messenger Ca2+ are then activated, and the signals of which are transmitted and magni-
fied through the MAPK signaling pathway and other kinases such as CaMs. These enhanced signals triggered many biological processes and promoted the
asymmetric growth of leaves in which essential adaxial and abaxial identities have already been set. In the heading growth stage, many growth-related pro-
cesses were activated (indicated by the purple ribbon). In the heading mature stage, the re-enhanced hormonal pathways (indicated by the light orange ribbon),
promoted the formation of the leafy head and boost biomass. BR, brassinosteroid.