Introduction

Polyploidy or whole genome duplication (WGD) has played a major role in higher plant evolution and speciation (Doyle et al. 2008; Soltis et al. 2009; Jiao et al. 2011). Genomic studies and comparative genomics indicate that ancient WGDs are ubiquitous and recurrent in the history of angiosperms (Van de Peer et al. 2017). Polyploidy in angiosperms frequently occurs in association with allopolyploidy, which combines the diploid nuclear genomes from two or more different ancestral species or genera to form new species. Autopolyploids originate from doubling of structurally similar and homologous genomes (Jackson 1982; Ramsey and Schemske 2002). Autopolyploids exhibit superior features and allow plants to adapt to environmental changes (Vergara et al. 2017). For example, autotetraploid Lolium develops larger leaves by increasing its cell elongation rate, and autotetraploid Morus also has greater height, breast-height diameter, and fruits than their diploid progenitors (Sugiyama 2005; Dai et al. 2015). Autotetraploid rice were found causing low pollen fertility and low seed setting (He et al. 2011; Wu et al. 2015, 2017); however, rice breeders have developed neo-tetraploid rice with high fertility and heterosis using autotetraploid rice (Guo et al. 2017).

Polyploidy induction using colchicine is a popular breeding method in various crops (Majdi et al. 2010; Tang et al. 2010). Polyploidy is particularly of significance in plants that are reproduced asexually such as Chinese cabbage, where leaves serve as the primary photosynthetic organs. Chinese cabbage mutants are produced by isolated microspore culture, which not only affects leaf color but also influences photosynthetic efficiency, thereby altering overall plant growth and development (Huang et al. 2014).

Recent reports have indicated that leaf development and dwarfism are regulated by the interaction of various phytohormones. Autotetraploid Salix viminalis leaves contain significantly higher contents of gibberellins (GAs) and cytokinins (CTKs) compared to diploids, whereas abscisic acid (ABA) and indoleacetic acid (IAA) contents do not significantly differ (Dudits et al. 2016). Additionally, trans-zeatin (TZ), an active cytokinin, with levels varying among individual tetraploid lines, is involved in leaf development (Dudits et al. 2016). Dwarfism has been valuable in enhancing the harvestable yield of various crops (Hedden 2003; Muangprom and Osborn 2004). In oilseed Brassica napus, dwarf genes also may be useful in improving yield because many cultivars readily undergo lodging, leading to yield loss and difficulty in harvesting (Muangprom and Osborn 2004). That same study identified a dwarf gene that is involved in the GA pathway and that influenced plant growth. However, hormone levels of IAA and brassinosteroids (BRs) are remarkably lower in dwarf autotetraploid apples, whereas the GA content has no significant difference (Ma et al. 2016). Wang et al. (2017) found that the DSG1 gene positively regulates cell division and may be involved in multiple hormone pathways in a dwarf rice mutant. Hou and Niu (2017) found that the CsDET2 gene leads to BR deficiency and induces a compact phenotype in a dwarf cucumber mutant.

Next-generation sequencing technology has been extensively applied to various polyploid plants (Stupar et al. 2007; Zhou et al. 2015; Braynen et al. 2017). Transcriptome analysis of Arabidopsis thaliana autotetraploids has shown that gene expression strongly depends on the parental genome and is associated with DNA methylation (Yu et al. 2010). Similarly, some key hormone-related genes that regulate leaf development have been found in woody autotetraploids such as birch (Mu et al. 2012) and mulberry (Dai et al. 2015). Cytological observations and transcriptomic analysis of floral buds in autotetraploid B. rapa have identified key genes that are involved in meiosis (Braynen et al. 2017). In addition, researchers have utilized transcriptome analysis of neo-tetraploid rice to reveal a large number of specific differentially expressed genes (DEGs) associated with heterosis and fertility (Guo et al. 2017). Comparative small RNA analysis of pollen and embryo sac development in autotetraploid rice has identified differentially expressed miRNAs (DEMs) and siRNAs that result in low fertility and poor seed set (Li et al. 2016, 2017). Because miRNAs inhibit their mRNA targets via base pairing, mRNA targets can be predicted using bioinformatics software, and integrated mRNA and miRNA analysis can help in unraveling the molecular network at the transcriptional and post-transcriptional levels (Yang et al. 2016; Chen et al. 2017; Qiao et al. 2017).

To better understand the growth and development of the edible part of Chinese cabbage, namely, the leaves in autotetraploids, we used a Chinese cabbage DH line ‘FT’, which was generated by isolated microspore culture and created autotetraploid Chinese cabbages that had exactly the same genetic background as DH. Furthermore, we applied RNA-seq for both transcriptomic and miRNA analysis in this study. We also determined the DEGs and DEMs between the doubled haploids (2x) and autotetraploids (4x) libraries and identified critical microRNA-target mRNAs that are involved in hormone regulation, which may result in the dwarf autotetraploid Chinese cabbage.

Materials and methods

Plant materials and ploidy identification

Isolated microspore culture was performed using a Chinese cabbage DH line ‘FT’, which exhibits heat resistance, white flowers, and leafy heads. The microspore culture protocol was based on our previously described method (Huang et al. 2014).

The regenerated plants (M0 generation) were transplanted in the field. The M1 generation seeds were harvested after self-crossing. The ploidy of the M1 generation plants was identified using a FACSCalibur flow cytometer (Huang et al. 2014), and 2× and 4× plants were obtained, which exhibited stable inheritance after multiple generations.

Histochemical analysis

Samples were the longitudinal sections of the central part of the mature leaves that were collected from the same area of the 2× and 4× plants. Paraffin sections were prepared according to the method of Chen et al. (2015) and were stained using safranin’ O/Fast Green. The slides were observed under an optical microscope (Nikon ECLIPSE 80i, Japan).

Measurement of hormone content

At the bolting stage, mature leaves from the main stem below the first branch were collected from the 2× and 4× plants. The leaves collected from five 2× or 4× plants were pooled into one biological replicate. Three biological replicates of each ploidy were selected for hormone content measurements, including IAA, GA, TZ, ABA, and BR. Endogenous IAA, ABA, TZ, and ABA were quantified as described elsewhere (Chen et al. 2012), whereas endogenous BR was quantified following the method of Ding et al. (2013).

RNA extraction

The samples of measuring hormone contents were the same as those used for transcriptome and miRNA sequencing. Similarly, leaves collected from five 2× or 4× plants were pooled into one biological replicate for sRNA-seq, and three biological replicates were prepared for transcriptome sequencing. Total RNA extraction was performed using TRIzol reagent (Invitrogen, USA) according to the manufacturer’s instructions. The integrity of the RNA samples was assessed using the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, CA, USA).

Library construction, transcriptome sequencing, and analysis

For the construction of RNA libraries, 3 µg of RNA per sample was used to generate sequencing libraries using the NEBNext® Ultra™ RNA library prep kit for Illumina® (NEB, USA), following manufacturer’s recommendations. mRNA was purified from total RNA using poly-T oligo-attached magnetic beads, followed by fragmentation. First- and second-strand cDNAs were synthesized as described by the manufacturer. The remaining overhangs were converted into blunt ends using an exonuclease/polymerase. Hybridization was conducted by ligating the NEBNext adaptor with a hairpin loop structure. cDNA fragments of preferentially 150–200 bp in length were selected by purifying the library fragments with AMPure XP system (Beckman Coulter, Beverly Hills, CA, USA). Then, 3 µL of a USER enzyme (NEB, USA) was employed with size-selected, adaptor-ligated cDNA at 37 °C for 15 min, followed by 5 min at 95 °C before PCR. The PCR products, which were obtained using Phusion High-Fidelity DNA polymerase, universal PCR primers, and an index (X) primer, were purified, and library quality was assessed using the Agilent Bioanalyzer 2100 system.

Raw data were first processed using in-house Perl scripts. Paired-end clean reads were aligned to the reference genome (http://brassicadb.org/brad/datasets/pub/Genomes/Brassica_rapa/V2.0/) using TopHat (v2.0.12). HTSeq v0.6.1 was used to count the number of reads mapped to each gene. Then, the fragments per kilobase per million mapped reads (FPKM) of each gene was calculated based on the length of the gene and the number of reads that were mapped to this gene. Differential expression analysis of the two groups was performed using the DESeq R package (1.18.0). Genes with an adjusted P value of < 0.05 were assigned as differentially expressed. GOseq and KOBAS were used to test the statistical enrichment of differentially expressed genes in Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. The GO terms and KEGG pathways with a corrected P value < 0.05 were considered significantly enriched.

Library preparation for small RNA (sRNA) sequencing and data analysis

The sRNA libraries were prepared using the NEBNext® Multiplex Small RNA Library set for Illumina® (NEB, USA), following the manufacturer’s recommendations. Approximately 3 µg of total RNA per sample was used as input, and the purified small RNAs were ligated with 5′- and 3′-adaptors at both ends. The ligated sRNAs were further purified and reverse-transcribed into cDNAs. These cDNAs were finally amplified using the LongAmp Taq 2X Master Mix, SR Primer for Illumina, and an index (X) primer. PCR products were purified on an 8% polyacrylamide gel (100 V, 80 min). DNA fragments corresponding to 140–160 bp (the length of small noncoding RNA plus the 3′ and 5′ adaptors) were recovered and dissolved in 8 µL of elution buffer. Single-end sequencing of sRNA libraries was performed on coded samples using a cBot Cluster Generation System and a TruSeq SR Cluster Kit v3-cBot-HS (Illumina). After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2500 platform, and 50-bp single-end reads were generated.

Raw reads were processed through custom Perl and Python scripts. The small RNA tags were mapped to the reference sequence using Bowtie (Langmead et al. 2009) without mismatches to analyze their expression and distribution on the reference. miRNA expression levels were estimated by transcript per million (TPM). For samples without biological replicates, differential expression miRNA analysis of two samples was determined using the DEGseq (Wang et al. 2010). P values were adjusted using q values (Storey 2003). Q values < 0.01 and |log2 (fold-change)| > 1 were set as the threshold for significant differential expression by default. Mapped small RNA tags were used to search for known miRNAs. miRBase 20.0 was used as reference, and modified software mirdeep2 (Friedlander et al. 2011) and srna-tools-cli were used to obtain the potential miRNA and draw secondary structures. The target genes of the miRNAs were predicted using psRobot_tar in psRobot (Wu et al. 2012).

qRT-PCR to validate the transcriptomic and MicroRNAomic data

qRT-PCR was performed to validate the expression level of selected DEGs and DEMs of interest. mRNAs were reverse-transcribed with oligo(dT), and miRNAs were reverse-transcribed with specific stem-loop RT primers (Table S1). The qRT-PCRs were performed on an ABI7300 cycler using FastStart Universal SYBR Green Master (Roche). The reaction conditions were as follows: 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. Three technical replicates of each reaction were performed, and U6 snRNA was used as endogenous control. The sequences of primers used for qRT-PCR of DEGs and DEMs are listed in Table S1. Primers used for reverse transcription were labeled as -RT. Forward primers for qPCR were labeled as -A, and reverse primers were labeled as -S.

Results

Ploidy analysis and phenotypic characterization

Ploidy analysis of novel leaves and control diploid leaves was performed using flow cytometry analysis. Fig. S1 shows that the DNA content of 2× plants reached a major peak at channel 200. Therefore, the DNA content peak at channel 400 was defined as 4×. At the vegetative growth stage, the 4× plants showed significantly smaller leafy head size with shorter length and lower weight than the 2× plants (Fig. 1a; Table 1). At the reproductive growth stage, the 4× plants showed slower growth and more delayed bolting than the 2× plants (Fig. 1b). We selected mature leaves from the main stem below the first branch at the bolting stage for sequencing because through observation of the entire growth period, there were apparent differences in leaf thickness based on touch and size between the 2× and 4× plants (Fig. 1c; Table 1). Moreover, this stage indicated the transition from vegetative to reproductive growth.

Fig. 1
figure 1

Plant morphology of 2× and 4× plants. a Variation in leafy head size in 2× and 4× plants. b Flowering was delayed in 4× plants compared to the 2× plants. c Variations in leaf size at the bolting stage in 2× and 4× plants. Leaves at this stage were sequenced for transcriptomic mRNA and miRNA analyses

Full size image
Table 1 Phenotypic measurement of 2× and 4× plants (mean ± standard deviation)
Full size table

Histological observation of leaf segments

To explore the cause of autotetraploid thickness, we dissected the leaves of the 2× and 4× plants. The longitudinal sections of the leaf central portion showed that the thickness of the 4× leaves was about 100 µm wider than that of the 2× plants (Fig. 2). Moreover, the cell sizes in the 4× leaves were generally larger than those in the 2× leaves.

Fig. 2
figure 2

Longitudinal sections of mature leaf cortices from 2× and 4× plants (×100 magnification)

Full size image

Differences in the endogenous hormone levels

We detected common plant hormones that may be associated with leaf development, including IAA, GA, TZ, ABA, and BR. GA was not detected in 2× or 4× plants. We compared the relative levels of endogenous hormones IAA, TZ, ABA, and BR between the 2× and 4× plants (Table 2). Compared to the 2× plants, IAA and TZ contents in the 4× plants were slightly higher, whereas those of ABA and BR were significantly lower.

Table 2 Levels of endogenous phytohormones (ng/g)
Full size table

Transcriptome analysis

To investigate the genes involved in plant dwarfism and leaf development, transcriptome sequencing was performed. A high degree of reproducibility was observed among the three biological replicates in terms of the number of reads obtained, and approximately 44–54 million reads were obtained in the 2× samples, whereas 48–51 million were generated from the 4× samples (Table S2). After removing the low-quality reads and masking adaptor sequences, about 86.79% of the total reads were mapped to the Brassica reference genome. Among these, uniquely mapped reads represented approximately 79.53% of the total reads and were used in the subsequent analysis.

In this study, we identified 13,225 DEGs, of which 6614 were upregulated and 6611 were downregulated (Table S3). To classify the distribution of gene functions in general, a total of 9863 DEGs (74.58%) were assigned to 4157 terms by GO enrichment. Among these, 301 GO terms were significantly enriched based on corrected P value < 0.05 in the DEGs. The 30 most significantly enriched GO terms are shown in Fig. 3. There were 24 terms belonging to the biological processes category. “Biological process” (7291 DEGs) was the predominant group, followed by “metabolic process” (5537 DEGs) and “cellular process” (5368 DEGs). In the molecular function category, “oxidoreductase activity” (1055 DEGs) was the most represented GO term. In the cellular component category, “photosystem II oxygen evolving complex” (33 DEGs) and “oxidoreductase complex” (52 DEGs) were the significantly enriched GO terms.

Fig. 3
figure 3

The 30 most significantly enriched GO terms among 13,225 differentially expressed genes in doubled haploids and autotetraploids

Full size image

Pathway-based analysis can further our understanding of the biological functions and gene interactions. All DEGs were mapped to 116 KEGG pathways (Table S4). The most represented was “metabolic pathways” (1,357 DEGs), followed by “biosynthesis of secondary metabolites” (741 DEGs), “ribosome” (262 DEGs), “carbon metabolism” (237 DEGs), and “biosynthesis of amino acids” (235 DEGs). Among these, “carbon metabolism” was the unique significantly enriched KEGG pathway.

mRNA-miRNA integration analysis

Transcriptome analysis identified a large number of DEGs. To further investigate the regulatory gene network associated with plant dwarfism and leaf development, we measured miRNA levels from sRNA-seq data in the 2× and 4× plants. A total of 4,443,855 and 8,670,601 sRNA reads were obtained, which represented 964,954 and 1,781,718 unique reads in the 2× and 4× libraries, respectively (Table S5). This showed that nearly twice the amount of sRNA was active in the 4× plants compared to the 2× plants. For miRNA analysis, a total of 56 DEMs were detected, including 25 known and 31 novel miRNAs (Table S6).

To explore the biological significance of DEGs and DEMs, comprehensive analysis was performed to elucidate their relationship. An integrated analysis showed only 102 DEGs corresponding to 35 DEMs (Table S7). To elucidate the traits involved in leaf growth and development from our sequencing data and the associated biological processes, functional classification of differentially expressed (DE) miRNAs-target mRNAs was performed using KEGG pathway enrichment analysis (Fig. 4). KEGG analysis indicated that a total of 36 pathways were enriched, and the top 20 terms were mainly related to the metabolic pathway, glycolysis/gluconeogenesis, fructose and mannose metabolism, pentose phosphate pathway, BR biosynthesis, and inositol phosphate metabolism. These key pathways contain several critical genes that are possibly involved in leaf development in Chinese cabbage.

Fig. 4
figure 4

The top 20 enriched KEGG pathways among 102 differentially expressed genes in doubled haploids and autotetraploids

Full size image

Using the Basic Local Alignment Search Tool (BLAST) and gene function annotation, we found that 13 of the 102 DEGs are related to leaf development, including those for BR synthesis, dwarfing, regulation of flowering time, ABA regulation, cell size, leaf development, and leaf tissue formation (Table 3). Furthermore, we identified 13 DEGs that were regulated by 12 DEMs. Among these, two DEMs (novel_15 and novel_54) were upregulated, and their three target DEGs (BraA01000252, BraA09000469, and BraA05004386) were downregulated, which are involved in BR synthesis and cell elongation, thereby possibly resulting in dwarfism. Another two DEMs (novel_42 and novel_44) targeted two DEGs (BraA07001177 and BraA09003100, respectively), which are probably related to leaf shape and metabolite changes along with dwarfism (Fig. 5). Five DEMs (bra-miR5720, novel_167, bra-miR172a, bra-miR172c-3p, and bra-miR172d-3p) targeted seven DEGs (BraA05002597, BraA10001056, BraA10002300, BraA06001479, BraA03001002, BraA10001990, and BraA07001666) that are involved in the ABA metabolic pathway, leaf formation, and flowering regulation (Fig. 6). Three DEMs (bra-miR171e, novel_35, and novel_103) downregulated one DEG (BraA03002481), which may be a transcription factor that is involved in plant development.

Table 3 Expression profiles of 13 differentially expressed genes with related 12 differentially expressed miRNAs
Full size table
Fig. 5
figure 5

Possible miRNA-dependent regulatory pathways that participate in dwarfism. Solid lines with arrows indicate direct actions, and dotted lines indicate indirect actions. Red represents miRNAs upregulated in 4× vs. 2× plants, and green represents downregulated miRNAs downregulated in 4× vs. 2× plants. (Color figure online)

Full size image
Fig. 6
figure 6

Possible miRNA-dependent regulatory pathways that participate in leaf development and slow growth. Solid lines with arrows indicate direct actions, and dotted lines indicate indirect actions. Red represents miRNAs upregulated in 4× vs. 2× plants, and green represents miRNAs downregulated in 4× vs. 2× plants. (Color figure online)

Full size image

qRT-PCR validation of mRNA and miRNA levels

To confirm the results of transcriptome and miRNA analyses, seven DEGs and seven DEMs that were involved in leaf development and dwarfism were selected. The observed changes in the RNA expression patterns by qRT-PCR coincided with those of RNA-seq analyses, indicating the accuracy of our transcriptomic and microRNAomic data (Figs. S2a, b). However, BraA10002300 and BraA03001002 did not show significant differences in gene expression between the 2× and 4× plants.

Discussion

In this study, we created dwarf autotetraploid Chinese cabbages by isolated microspore culture. Because autotetraploids are generated from the spontaneous chromosome doubling of DHs, their genes originated from the same genome, and their genetic backgrounds are theoretically identical. Therefore, this is the most ideal material for investigating ploidy differences. Alterations in gene expression levels absolutely result from ploidy differences, which cause changes in the observed traits between DHs and autotetraploids. Morphological and histochemical analyses showed that the leaves of 4× plants were smaller and thicker than those of 2× plants, and the weight of 4x plants was lower than that of 2× plants, which differs from the findings of Shahid et al. (2013), who reported that panicle length and leaf area significantly increased, but plant dry weight significantly decreased in autotetraploid rice. The reason for this observed discrepancy requires further investigation. In addition, our physiological and transcriptome-miRNA integrative data indicate that the BR and ABA levels in the 4× plants were significantly lower than those in 2× plants, and 13 DEGs with 12 DEMs regulated the hormone levels that influence leaf development, possibly resulting in dwarf autotetraploid Chinese cabbages.

Dwarfism is an apparent character in our autotetraploid plants. We selected some critical related genes to interpret this phenomenon, combined with data on hormone levels. Two novel miRNAs, novel_15 and novel_54, regulate two critical genes (BraA01000252 and BraA05004386) that are involved in BR synthesis. Functional annotation identified one gene as 3-epi-6-deoxocathasterone 23-monooxygenase (CYP90C1) and the other as cytochrome P450 (CYP734A). BraA01000252 is the homolog of ROTUNDIFOLIA3 (ROT3) gene in Arabidopsis. ROT3 encodes for the enzyme CYP90C1, which regulates leaf length in A. thaliana and is required for later steps in the BR biosynthesis pathway. ROT3 appears to be required for the conversion of typhasterol to castasterone, an activation step in the BR pathway (Kim et al. 2005). CYP734A1 inactivates brassinolide and castasterone by hydroxylating carbon-26 (C-26) of the substrate side-chain (Turk et al. 2003) and has been identified as a BR-inactivating enzyme that plays an important role in steroid-mediated signal transduction in A. thaliana (Thornton et al. 2010). In addition, one DEG (BraA09000469) is regulated by novel_15, which is annotated as a helix-loop-helix (bHLH) transcription factor that plays an important role in the control of cell elongation in A. thaliana (Ikeda et al. 2012). BRs are plant hormones that regulate various physiological processes in plants such as cell expansion and leaf growth and development (Fariduddin et al. 2013). In our study, the downregulation of BraA01000252 and BraA05004386 transcripts caused by the upregulation of novel_15 and novel_54 may lead to the inhibition of BR synthesis, which possibly affects leaves and cause dwarf autotetraploid plants.

Another two DEGs (BraA07001177 and BraA09003100) are regulated by two DEMs (novel_42 and novel_44, respectively), and were annotated as F-box-only protein 6 and fructose-1,6-bisphosphatase. An F-box gene (LEAF CURLING RESPONSIVENESS, LCR) regulates leaf morphology, and loss-of-function lcr mutants exhibit pleiotropic defects with semi-dwarfism, altered leaf shape, and a shorter stem in Arabidopsis thaliana (Song et al. 2012). Fructose-1,6-bisphosphatases (FBPases) are involved in starch and sucrose synthesis and also disrupt both chloroplastic and cytosolic FBPase genes, resulting in starch and metabolite changes along with a dwarf phenotype (Rojas-González et al. 2015). Thus, these two DEGs with their DEMs may indirectly influence dwarf autotetraploid formation. Similarly, the BR content was significantly lower in the 4× plants compared to the 2× plants. These findings suggest that miRNA-dependent regulatory pathways contribute to dwarfism.

Along with the dwarf phenotype, the autotetraploids also showed slow growth. We utilized sequencing data to identify related genes that are responsible for this characteristic. ABA accumulates in senescent tissues during plant growth. In this study, the 4× plants showed slower growth and looked much younger than the 2× plants. Thus, compared to the 2× plants, the 4× plants exhibited significantly lower ABA levels at the same reproductive stage. Similar results were observed in autotetraploid tomatoes (Tal 1977). In our study, three ABA-related target genes were upregulated. We identified bra-miR5720 as a transcriptional regulator of the TAP46 gene, which is related to ABA sensitivity in seed germination and protein phosphatase 2A (PP2A) activity. TAP46 is implicated as a positive factor in ABA-regulated gene expression in Arabidopsis (Hu et al. 2014). Plants overexpressing TAP46 exhibit increased ABA sensitivity in seed germination and reduced PP2A activity (Hu et al. 2014). Overexpression of TAP46 also leads to stimulation of plant growth and increased nitrogen assimilation (Ahn et al. 2015). Another novel miRNA, novel_167, regulates the BraA10001056 gene, which is annotated as the uncharacterized protein At5g02240 and is also involved in responses to ABA. We also found a DEG (BraA03001002) that was annotated as mitochondrial pyruvate carrier 1 (MPC1). MPC negatively regulates ABA-induced guard cell signaling, and its product is transported to the mitochondria of A. thaliana leaves (Li et al. 2014). Therefore, BraA03001002 might be a predominant DEG that significantly decreases ABA levels in 4x plants. ABA regulates approximately 200 loci and thousands of genes, which in turn influence various processes and are implicated in crosstalk with other developmental signals. Thus, ABA-signaling mechanisms may vary among tissues or developmental stages (Finkelstein 2013) and should be studied in more detail in autotetraploids to determine the specific role of ABA in mechanisms regulating plant growth and development.

A known miRNA, bra-miR5720, which is related to the varicose (VCS) and HUA2-like protein 1 (HUAL1) in Arabidopsis and their two orthologs (BraA05002597 and BraA10001056) in Brassica, was downregulated in autotetraploids in our study. Because of the negative correlation between mRNA and miRNA, bra-miR5720 may not be the main miRNA that regulates both DEGs. VCS, a WD-domain protein, is required during leaf development and plays an essential role in leaf blade formation. Arabidopsis with a mutation in vcs shows leaf and shoot apical meristem defects (Deyholos et al. 2003). In A. thaliana, the HUA2 gene is required for proper expression of FLOWERING LOCUS C (FLC), which is a critical regulator of flowering time (Jali et al. 2014). Lack of different HUA2-like genes in mutants results in different flowering times (early-flowering and late-flowering). Thus, downregulation of VCS and HUAL1 is probably responsible for the observed reduction in leaf size and delayed flowering in autotetraploids.

Bra-miR172d-3p not only indirectly regulates MPC1, but also regulates AP2-like ethylene-responsive transcription of bra-miR172a and bra-miR172c-3p in A. thaliana. TARGET OF EAT (TOE) is one of the AP2-like target genes. Different kinds of TOE may also cause different flowering times (Aukerman and Sakai 2003). The regulatory mechanism of flowering time regulation is relatively complicated. Based on the identified DEGs and DEMs that are related to the ABA levels and flowering time regulation, we propose a pathway that depicts this relationship.

Functional and pathway assignments of the DE mRNA- and DE miRNAs-mediated targets using KEGG classification reveal numerous physiological and developmental processes, which may play vital roles in response to polyploidy in B. rapa. The 13 critical DEGs that we found are involved in several key pathways such as metabolic pathways, biosynthesis of secondary metabolites, BR biosynthesis, RNA degradation, glycolysis/gluconeogenesis, fructose and mannose metabolism, pentose phosphate pathway, carbon fixation in photosynthetic organisms, and carbon metabolism, which may offer more information in elucidating the mechanism involving the generation of autotetraploid Chinese cabbage.

Our study provides a theoretical basis for the traits of dwarfism and leaf development in Chinese cabbage and proposes two possible miRNA-target mRNA regulatory pathways that participate in trait formation. DEGs and DEMs were identified in this study; however, their roles in different growth periods require further studies. Moreover, identification of novel miRNAs and their corresponding mRNA in the leaf tissues may help in better understanding the mechanisms employed by autotetraploid plants.