Fine mapping of QTL and identification of candidate genes responsible for tuberous stem formation in Kohlrabi (Brassica oleracea var. gongylodes L.)

doi:10.21203/rs.3.rs-4696240/v1

Download PDF

Research Article

Fine mapping of QTL and identification of candidate genes responsible for tuberous stem formation in Kohlrabi (Brassica oleracea var. gongylodes L.)

https://doi.org/10.21203/rs.3.rs-4696240/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Tuberous stem is a key determinant of yield and quality in kohlrabi, and it is a prime target of artificial selection for breeding and domestication. To dissect the genetic basis of tuberization, a previously identified major QTL, REnBo2 on chromosome C03 was fine mapped to a 42.8-kb interval using the recombinant populations of three different backcross generations derived from the double haploid lines of kohlrabi and broccoli. Three genes namely Bol012972, Bol012973 and Bol012974 encoding a CBL-interacting protein kinase 6 (CIPK6), fatty acid desaturase 6 (FAD6) and WRKY transcription factor 32 (WRKY32), respectively, were annotated in the fine-mapped interval in Brassica oleracea reference genomes. FAD6 is unlikely to be the responsible gene due to its function as a fatty acid desaturase. Based on the gene structure and expression analyses of the remaining two candidates along with our previously conducted transcriptome and endogenous phytohormone profiles, Bol012972 (CIPK6) was identified as a promising candidate for REnBo2. Since the knockdown mutant of CIPK6 was known to exhibits swollen hypocotyls in Arabidopsis, we speculate that CIPK6 may regulate kohlrabi tuber development through regulating the auxin levels. The result will provide a foundation to elucidate the genetic mechanism for tuber formation in kohlrabi and other Brassica crops.

Tuberous stem

Kohlrabi

Fine mapping

QTL

Candidate genes

CIPK6

Brassica and closely related species, morphologically share radial enlargement in different plant organs: hypocotyl in turnip (B. rapa), stem in kohlrabi (B. oleracea var. gongylodes) and tuber mustard (B. juncea var. tumida Tsen et Lee), root in rutabaga (B. napus), and taproot in radish (Raphanus sativus). These traits have evolved independently during the domestication process of each species, but possibly arose from mutations sharing some of the same orthologous genes, providing an excellent model to study the biological functionality and morphological evolution of agronomic traits (Hovav et al. 2008; Lenser and Theiben 2013; Fuller et al. 2014; Cheng et al. 2016). Therefore, the exploration of the genetic processes underlying kohlrabi stem formation is the foundation for understanding the shared mechanism of organ tuberization and advancing tuber yields through molecular breeding.

In the past decades, quantitative trait locus (QTL) mapping approach has been widely used to explore the genetic mechanism underlying tuberous organ formation in Brassica and other tuber crops. So far, nearly 50 QTLs of tuber and tuber related traits have been identified in Brassica and related species, which is a powerful illustration of the complex quantitative traits of multi-gene control. In turnip, one important QTL (TuQTL-1) for tuber formation (Lou et al. 2007), eighteen QTLs for three different root traits (Lu et al. 2008), four QTLs for turnip size and weight (Kubo et al. 2010) and two QTLs (FR1.1 and FR7.1) for swollen root formation (Wu et al. 2021) were separately detected in various studies using different segregating populations. Similarly, three QTLs for root shape and two QTLs for root diameter in radish (Tsuro et al. 2008), nine QTLs for tuberous root yield traits in sweet potato (Li et al. 2014) and five QTLs for stem expansion traits in B. juncea (Li et al. 2016) were also identified using QTL mapping approach.

Regarding kohlrabi, the stem swelling trait of kohlrabi is expressed in the recessive manner, as the F₁ populations derived from the cross of kohlrabi and broccoli was close to that of broccoli (Baggett and Kean 1989; Hoque et al. 2017). These authors also observed that the F₂ populations showed a narrow range of stem diameter closed to that of the F₁ populations, indicating that the swelling trait of kohlrabi stem is under oligogenic control by several recessive genes. On the other hand, using the BC₁ populations derived from the broccoli double haploid (DH) line (GCP04) and a commercial kohlrabi variety ‘seine’, four significant QTLs for radial enlargement of stem, namely, REnBo1, REnBo2, REnBo3 and REnBo4 were detected on four different chromosomes; C02, C03, C05 and C09, respectively (Hoque et al. 2017). However, the detected QTLs are widespread, and it is difficult to determine the candidate genes involved in the formation of tuberous stem in kohlrabi. So, it is necessary to narrow down the QTLs region.

Thus, in this study, we report the fine mapping of REnBo2 using the recombinant populations of BC₂F₂, BC₃F₂ and BC₃F₃ and the identification of the candidate genes responsible for tuberous formation in kohlrabi. Additionally, we propose a regulatory mechanism explaining the involvement of one of the candidate gene, CIPK6 in tuber formation by combining our previously conducted anatomical, hormonal and transcriptome results (Nuruzzaman et al. 2022, 2023). The resulting data of this study could provide a significant insight into the molecular mechanism of tuberous organ formation in Brassica and other related species.

Development of mapping populations

A schematic diagram of development of the plant population is shown in Fig. S1. An F₁ was developed from the cross between GCP04, a double haploid (DH) line of broccoli and a kohlrabi DH, KDH line (Kr13-511) as the seed and pollen parent, respectively. The mapping populations were developed by several rounds of backcrossing with marker-aided selection. A BC₁F₁ population was produced by backcrossing the F₁ with the KDH parent and the plants having heterozygous alleles at the target QTL (REnBo2) and kohlrabi homozygous alleles at the other three QTLs were selected and selfed to produce BC₁F₂. Then, the individuals from BC₁F₂ populations having homozygous broccoli alleles at REnBo2 and kohlrabi homozygous alleles at other QTLs were selected and backcrossed with the KDH parent to produce BC₂F₁. The segregating BC₂F₂ populations consisting of 800 plants was developed by selfing the selected BC₂F₁ plants having heterozygous alleles at REnBo2 and kohlrabi homozygous alleles at other four QTLs. Again, a BC₂F₁ plant was backcrossed with the parent KDH to produce BC₃F₁ populations. A BC₃F₁ plant having heterozygous alleles at REnBo2 and kohlrabi homozygous alleles at other QTLs was selected and self-pollinated to produce BC₃F₂ populations consisting of 592 plants. Furthermore, to reconfirm the fine mapping, twelve recombinants from BC₃F₂ having heterozygous alleles at the fine-mapped region to check the segregation pattern in the subsequent generations, as well either kohlrabi or broccoli homozygous type to check the consistency of the genotypes, were selected and self-pollinated to produce BC₃F₃ populations comprising of 216 plants.

Growth conditions and recombinant identification

The seedlings of backcross populations were grown in 25°C temperature with 16h of light and 8h of dark conditions in a controlled room environment for up to 2 weeks. The total genomic DNA was extracted from the young leaves of each plant using the modified CTAB method (Murray and Thompson, 1980), and genotyped using the REnBo2 flanking markers (Bol017491 and Bol035216). The plants having either kohlrabi- or broccoli-homozygous alleles at one flanking side and heterozygous alleles at other side; or, kohlrabi homozygous alleles at one flanking side and broccoli homozygous alleles at other side were selected from the BC₂F₂ and BC₃F₂ populations and termed as recombinant plants. The BC₂F₂ recombinant plants were grown in plastic house and the recombinants of BC₃F₂ and BC₃F₃ were grown in the incubator at constant temperature (18°C) with 16:8 h of light (100 µmol/m²/s) and dark conditions. A sufficient amount of water was supplied regularly and liquid compound fertilizer (Hyponex, N-P-K = 6-10-5) was applied at the young stage of plant growth. After that, commercial organic fertilizer mixed with compound chemical fertilizer (Best Match Vegetable, N-P-K-Mg-B = 20-8-16-1-0.15) was applied, and extensive pest protection measures were taken to provide optimal conditions for growth and the swelling of the stem.

DNA polymorphism analysis

The genome DNA of GCP04 and KDH was sequenced by Genebay Co., Ltd. (Kanagawa, Japan) using the HiSeq 2000 device. The obtained reads were mapped to B. oleracea reference genome sequence by BWA (Li and Durbin 2010). A total of seventeen InDel and CAPS (cleaved amplified polymorphic sequence) markers were designed using the genome re-sequence data of broccoli (GCP04) and kohlrabi (KDH) to amplify the different positions of the target QTL (Table S1). For genotyping, each 10 µl polymerase chain reaction (PCR) mixture contained 1 µl DNA, 0.25 µl forward primer (30 pmol) and 0.25 µl reverse primer (30 pmol), 5 µl PCR EmeraldAmp GT PCR master mix (Takara Bio, Tokyo, Japan) and 3.5 µl dH₂O. The detailed PCR procedure was previously described by Hoque et al. (2017).

Measurement of stem swelling traits

The height and maximum diameter of the enlarged stem were measured at 90 days after germination (DAG) for BC₂F₂ and 75 DAG for BC₃F₂ and BC₃F₃ recombinant populations as they were grown in incubator with a limited space. However, 75 days is enough for studying segregation of stem swelling related traits. Based on the degree of stem swelling, the recombinant populations were divided into four qualitative grading in classes (grade – 4 = individual showing a complete spherical enlargement like kohlrabi, grade – 3 = enlarged individual but does not show a complete spherical shaped as kohlrabi, grade – 2 = enlarged individual but the whole stem is uniform, and grade − 1 = individual without hypertrophy) by naked eye observation (Fig. S2). The ratio of the height and the maximum diameter was determined, dividing the height by the maximum diameter of the tuberous stem, and is termed the stem enlargement index (SEI) as described by Hoque et al. (2017). The calculated SEI of tuberous stem was then used to confirm the qualitative grading in classes. The grading of plants based on radially enlarged stems was performed as follows: grade 1, SEI > 1.80, no stem swelling (like broccoli type); grade 2, SEI 1.80–1.51, long-elliptical shape; grade 3, SEI 1.21–1.50, trophy to hypertrophy shape; and grade 4, SEI < 1.21, maximum enlargement of stem (like kohlrabi type).

Data analysis

A linkage map was created using the genotypes and phenotypes data of BC₂F₂ recombinant populations (n = 158) using Haldane map function and LL ordering criterion of Ant map Ver.1.2 (Iwata and Ninomiya 2006). QTL analysis was performed using composite interval mapping model of Windows QTL Cartographer v. 2.5 (Wang et al. 2012). The genomic, coding (CDS), amino acid (AA) and promoter sequences of the candidate genes were aligned using Clustal W (http://www.ebi.ac.uk/clustalw/). MEGA software (version 4.0) (Tamura et al. 2007; http://www.megasoftware.net/index.html) was used to construct the phylogenetic trees. The neighbor-joining method was used, with the pairwise deletion option, Poisson correction model and 1000 bootstrap replicates. The conserved domains and motifs were detected using the NCBI CD-search tool (Lu et al. 2020) and MEME tool (Bailey and Elkan 1994), respectively, and visualized by using TB tools (Chen et al. 2020). The phytohormone related cis-acting elements in the promoter regions of the gene Bol012972 in kohlrabi and broccoli were identified using the PlantCARE tool (Lescot et al. 2002).

Sequence analysis of candidate genes

The genomic sequences of the three candidate genes were extracted from our previously conducted genome resequencing data of KDH and GCP04. Besides, the 3 kb promoter and coding region (CDS) of Bol012972 was amplified and sequenced in the two parents by designing the gene-specific primers (Table S1). In addition, the genomic sequences along with the CDS and amino acid (AA) sequences of the three candidate genes in three different B. oleracea reference genomes including cabbage (Braol_{_}JZS_V2.0, Cai et al. 2020), broccoli (Braol_{_}HDEM_{_}V1.0, Belser et al. 2018) and Chinese kale (TO1000, Lv et al. 2020) were collected from the brassica database.

Analysis of gene expression

Quantitative real-time PCR (qRT-PCR) was conducted to analyze the expression of two potential candidates in REnBo2 fine-mapped region at three different growth stages (10, 20 and 40 days after germination, DAG) in the parents. Total RNA was extracted from the enlarged part of the stems using RNeasy Mini Kit and RNase-Free DNase Kit (QIAGEN). According to published protocol, the amplification reactions were run in an iCycler iQ real-time PCR detection system (Takara Bio). All reactions were performed in three replicates, and the equation ratio = 2^−ΔΔCT was applied to calculate the relative expression level of the candidates using Bra Actin7 as the internal control gene.

The candidate gene sequence data have been submitted to the DDBJ database (https://www.ddbj.nig.ac.jp) under the accession number LC829494, LC829495, LC829496, LC829497, LC829498, and LC829499 for CIPK6, FAD6, and WRKY32 of kohlrabi (KDH) and broccoli (GCP04), respectively. NGS data of KDH and GCP04 have been submitted to the DDBJ database under the accession number (in submission), respectively.

Fine mapping of REnBo2 locus using BC₂F₂ population

The phenotypes of one hundred and fifty-eight plants of BC₂F₂ populations (Fig. S1) were investigated in terms of stem height, maximum diameter and stem grade. A wide ranged distribution of the mean diameter for stem swelling was observed in the recombinant populations (Fig. S2). The phenotypic data of stem swelling related trait fit a normal continuous distribution, but was slightly skewed toward the less radially enlarged grades. The plants were categorized into four grades (1–4) according to the degree of stem swelling, showing a uniform segregation of the stem swelling trait from grade 1; slim-cylindrical broccoli type, to grade 4; round-spherical kohlrabi type. The recombinant plants were further genotyped using the additional nine CAPS and InDel markers within the REnBo2 interval (Table S2). Consequently, different types of recombinants were observed based on their marker genotypes in this region, and graphical genotypes of the recombinants showing grade 1 and grade 4 stem swelling were created (Fig. S3). However, narrowing down the REnBo2 region proved challenging, as we observed a few individuals whose graphical genotypes did not match their respective phenotypes. This discrepancy may be attributed to the heterogeneous genetic background of BC₂F₂ plants. Thus, we did a QTL analysis on C03 chromosome using the phenotypic and genotypic data of all the recombinant plants. A linkage map with a total length of 59.5cM was created using the AntMap ver. 1.2 (Iwata and Ninomiya 2006) with an average marker distance of 6.59Mbp (Fig. 1A). Consequently, a significant LOD peak with a LOD value of 4.8 was detected between the markers Bol012989 and Bol017480 (Fig. 1B). On the linkage map, the range was 10.9cM, and the physical distance was 0.73Mbp in B. oleracea (Braol_JZS_V1.1) reference genome. Thus, through QTL mapping, we were able to narrow down the REnBo2 region, flanked by the markers Bol012989 and Bol017480.

Fine mapping of REnBo2 using BC₃F₂ and BC₃F₃ populations

To further delimit the target region, we utilized a total of 592 BC₃F₂ plants derived from the DH lines of kohlrabi and broccoli for fine mapping. Two flanking markers Bol017491 and Bol035216 were used for preliminary screening. Sixty-seven recombinants from the BC₃F₂ populations were further genotyped and used to refine the target locus. In addition to the above-mentioned nine markers, another six CAPS and InDel markers were designed at the target region (Table S1); and a total of fifteen markers were used to genotype all these recombinants. Stem swelling related traits of all the recombinants, which were grown in the incubator controlled at the temperature of 18°C, were assessed at 75 DAG. A wide range of segregation was observed among the recombinant populations for all the stem swelling traits (Fig. S4). We created graphical genotypes for the recombinants that had either kohlrabi or broccoli homozygous alleles at the target region, along with the phenotype data (Fig. 2, details in table S3). The REnBo2 locus was restricted in a 42.8-kb interval flanked by the markers Bol012971 and Bol012974. The correlation between the phenotypes and genotypes of a marker (Bol012972) within the fine-mapped region was observed (Fig. S5). All the recombinants revealed a distinct correlation, where the individuals having kohlrabi-homozygous alleles showed a maximum stem diameter with kohlrabi (grade 4)-type phenotype. On the other hand, the recombinants showing broccoli-type (grade 1) stem with a maximum diameter of 20 mm, possessed broccoli-homozygous alleles at the target locus (Fig. S5). The heterozygous individuals exhibited a wider variation in stem diameter and grade, ranging from grade 1 (broccoli type) to grade 4 (kohlrabi type) swelling pattern and thus the recombinants having heterozygous alleles at the target region were excluded from the graphical genotypes in Fig. 2.

Furthermore, we reconfirmed the REnBo2 fine mapping using the self-progeny (BC₃F₃) of twelve selected recombinants from the BC₃F₂ populations (Fig. S1). The criteria for the selection of BC₃F₂ recombinants were mentioned in Materials and methods section. A total of 216 BC₃F₃ plants were developed and genotyped using the all fifteen markers within the REnBo2 interval (Table S4). The segregation patterns of these selected plants in the BC₃F₃ populations for stem swelling were shown in Fig. S6. Graphical genotypes were created using the recombinants having either kohlrabi or broccoli homozygous alleles at the target region (Fig. 3A). The graphical genotypes, together with the phenotypic data of the recombinants of each line suggested the confirmation of the previously delimited region flanked by the marker Bol012971 and Bol012974, and the physical distance was 42.8 kb.

Table 1

Prediction and annotation of candidate genes in the fine-mapped region of *REnBo2* in three different *Brassica oleracea* reference genomes. The details of the candidates are shown in the Supplementary table 5.
Gene name in Braol JZS V1.1	Gene name in Braol HDEMV1	Gene name in ‘TO100’	Orthologue in A. thaliana	Gene annotation
Bol012972	BolC3t21148H	Bo3g175290	AT4G30960	CBL-interacting protein kinase 6, CIPK6
Bol012973	BolC3t21150H	Bo3g175310	AT4G30950	Fatty acid desaturase 6, FAD6
Bol012974	BolC3t21152H	Bo3g175320	AT4G30935	WRKY transcription factor 32, WRKY32

Candidates of REnBo2

According to the B. oleracea reference genome (Braol_JZS_V1.1, Liu et al. 2014), the fine-mapped interval contained four predicted genes, designated as Bol012971 through Bol012974. In the corresponding region, there are three, four and five genes in the other B. oleracea reference genomes, Braol_{_}JZS_V2.0 (B. oleracea var. capitate, Cai et al. 2020), TO1000’ (B. oleracea var. alboglabra, Lv et al. 2020) and Braol_{_}HDEM_{_}V1.0 (B. oleracea var. italica, Belser et al. 2018), respectively (Fig. 3B). Among the four genes on Braol_JZS_V1.1, Bol012971 exist only in the Braol_{_}JZS_{_}V1.1 but not in Braol_{_}JZS_V2.0 and in two other B. oleracea genomes (TO1000 and Braol_{_}HDEM_{_}V1.0), indicating that it might be a mistake in gene prediction on the Braol_{_}JZS_{_}V1.1, which is the older version of ‘Braol_{_}JZS_V2.0’ genome. Thus, we speculated that Bol012971 might not exist in REnBo2 and discarded it from further analyses. The rest three genes, Bol012972, Bol012973 and Bol012974 encodes CBL interacting protein kinase 6 (CIPK6), omega-6 fatty acid desaturase protein (FAD6) and WRKY DNA binding transcription factor 32 (WRKY 32) protein, respectively (Table 1). Besides, three small-sized annotated genes, BolC3t21149H and BolC3t21151H in ‘Braol_{_}HDEM_{_}V1.0’ and Bo3g175300 in ‘TO1000’ genome were also detected (Table S5).

Structure analyses of the candidates

To further assess the candidates for the REnBo2, the coding (CDS) and amino acid sequences of the three candidates were compared between the parents. Genomic DNA sequence for the candidate Bol012972 (CIPK6) was 1309 bp, encoding 435 amino acids in both KDH and GCP04. A total of 13 SNPs between the parents were detected, of which six showed non-synonymous substitutions resulted in the change in four amino acids at the kinase domains (STKc_SnRK3 and SPS1) (Fig. 4A). To further study the function of CIPK6, the 3-kb region upstream of the start codon in parental lines, Arabidopsis and three B. oleracea subspecies were analyzed. The hormone related cis-elements including auxin, gibberellin and ABA responsiveness were repeatedly detected in promoter of KDH and GCP04 (Fig. S7A). A single non-synonymous mutation was observed in second exon of Bol012973 (FAD6) causing an amino acid substitution (Proline to Leucine) between the parents (Fig. 4B). A total of six SNPs between the parents in the first, second, third and fourth exons of Bol012974 (WRKY32) were observed. Among them, five were non-synonymous type causing five AA substitutions including one substitution (Alanine to Threonine) at WRKY domain (Fig. 4C).

To analyze the conserved regions of the candidate genes in the parents, A. thaliana and three other Brassica subspecies, their conserved domains and motifs were examined, and their putative protein sequences were aligned. The AA sequences of CIPK6 were analyzed with the CD-Search software, showing that they each had four domains including three kinases (STKc_SnRK3, SPS1 and CIPK_C) and one NAF domain (Fig. S7B). Three conserved domains including DesA, Delta12_FADS like and FA desaturase in FAD6 and two WRKY domains in WRKY32 were identified in the parents, Arabidopsis and three other Brassica subspecies (Fig. S8). Ten conserved motifs along the whole protein sequences for each of CIPK6, FAD6 and WRKY32 of parental lines, Arabidopsis and other three B. oleracea subspecies were shown in Fig. S8. Regarding the other three small-sized annotated genes in fine-mapped region, we did not find any homologous sequence of Bo3g175300 in parental lines, and the AA sequences of BolC3t21149H and BolC3t21151H were similar between the parents and in a B. oleracea subspecies, broccoli (HDEM) (Fig. S9). Therefore, these small-sized genes are not candidate genes for the REnBo2. In addition, Bol012973 (FAD6) is unlikely to be a candidate gene due to its function as a fatty acid desaturase.

Homology and phylogeny analyses of CIPK6 and WRKY32

To reveal the evolutionary relationships of two candidate genes, the global homology and phylogenetic analyses of these genes in Arabidopsis and different Brassica species including B. oleracea, B. rapa, B. napus, B. juncea, B. nigra and B. carinata were performed with the putative protein sequences using MEGA (version 5) software. (Fig. S10). Three homologs of CIPK6 located on C01, C03 and C07 and two homologs of WRKY32 on C03 and C07 were detected in each B. oleracea subspecies (cabbage, broccoli and Chinese kale), respectively. In overall, the phylogeny analysis revealed that the parents, kohlrabi and broccoli were close to the cabbage and broccoli reference genomes, respectively, in regard to CIPK6 and WRKY32.

Expression pattern analysis of candidate genes

To determine if the mRNA expression levels in the candidate genes might lead to functional differences, their expression patterns in early developmental stages were analyzed by both transcriptome and qRT-PCR (Figs. 5 and S11). Based on qRT-PCR data, CIPK6: C03 at chromosome C03 was up-regulated in kohlrabi stem from 10 to 20 DAG with a significantly lower expression at 10 DAG compared to the broccoli. On the contrary, in broccoli stem, CIPK6: C03 was constantly highly expressed at all the growth stages tested as revealed by qRT-PCR analysis. In the transcriptome analysis using RNA-seq, CIPK6: C03 showed significantly lower expression in kohlrabi stem at both growth stages compared to the broccoli (Fig. 5A). Regarding the homologs, one homolog of CIPK6: C01 showed up-regulation in kohlrabi stem at 10 DAG, but at 20 DAG, there was no significant difference between the stem of kohlrabi and broccoli (Fig. S11A). Another homolog of CIPK6: C07 was down-regulated in kohlrabi stem at both growth stages compared to broccoli (Fig. S11B). WRKY32: C03 was up-regulated in kohlrabi stem at all the growth stages tested in both qRT-PCR and RNA-seq, but there was no significant difference in expression of its only homolog on chromosome C07 (Bo7g114840) (Figs. 5B and S11C). FAD6 showed significantly higher expression in broccoli stem at 20 DAG, but non-significant at 10 DAG (Fig. S11D).

Among the diversified morphotypes of Brassica sub-species, tuberous stem is a characteristic feature in kohlrabi, determining its quality and economic value. In the last decades, efforts have been made to dissect the genetic basis of tuber formation in Brassica including turnip (Lou et al. 2007; Lu et al. 2008; Kubo et al. 2010; Wu et al. 2021), stem mustard (Li et al. 2016) and related species like in radish (Tsuro et al. 2008; Zaki et al. 2010) and sweet potato (Li et al. 2014) through QTL mapping. The identified QTLs in those researches were broad having several numbers of genes which made it difficult to propose the potential candidates for this trait. So far, only a single QTL in turnip (FR7.1) was fine mapped leading to the identification of four candidate genes related to cell cycle, glycolysis and plant hormone signal transduction (Wu et al. 2021). But, this type of information is still unavailable in kohlrabi as well in other related species.

Among the four detected QTLs for radial enlargement of kohlrabi stem, REnBo2 was repetitively identified in two different backcross populations (Hoque et al. 2017). The REnBo2 region includes 391 genes according to the B. oleracea reference genome, making it difficult to screen the possible candidates for tuber formation in kohlrabi. In this study, to narrow down the REnBo2 region, we selected plants that displayed a kohlrabi-type phenotype with the heterozygous alleles at the target REnBo2 QTL from each backcrossed progeny (BC₂F₁ and BC₃F₁), so that good allelic segregation for stem diameter and grade was observed in the mapping populations. The REnBo2 was initially delimited to a 0.73Mbp interval consisting of 83 genes based on the B. oleracea reference genome using QTL analysis (Fig. 1). Finally, the REnBo2 locus was fine-mapped to a 42.8 kb interval flanked by the markers Bol012971 and Bol012974 using the graphical genotypes of BC₃F₂ recombinants (Fig. 2) and further reconfirmed by using the BC₃F₃ populations (Fig. 3).

In the self-pollinated progenies (BC₃F₃), the phenotypes of heterozygous recombinants were unstable, showing a wide variation in stem diameter and grade, ranging from broccoli (grade 1) to kohlrabi type (grade 4), while the phenotypes of homozygous recombinants were completely correlated with their respective genotypes (Fig. S5). The unsteady trait in stem swelling in the heterozygous condition is a common phenomenon in kohlrabi, as previously described by Baggett and Kean (1989) and Hoque et al. (2017). It is speculated that under the heterozygous REnBo2 allele, the variable phenotypes may be caused by a complex interplay of home- and hetero-dimer formation of the broccoli and kohlrabi gene products associated with the REnBo2 allele. Additionally, there may be unstable epistatic interactions between the heterozygous REnBo2 allele and other stem enlargement QTLs. Further specific investigation will be required to explore the exact role of the REnBo2 allele in stem swelling in the heterozygous condition.

Candidate genes predicted for tuberous formation in kohlrabi

REnBo2 QTL was narrowed down to a 42.8 kb region with the most likely candidate genes being Bol012972, which encodes a CBL (Calcineurin B-like)-interacting protein kinase 6 (CIPK6), and Bol012974, a gene encoding WRKY transcription factor 32 (WRKY32) (Table 1). Bol012973 (FAD6) is unlikely to be a candidate gene due to its function as a fatty acid desaturase. WRKY transcription factors are a class of sequence-specific DNA-binding proteins found mainly in plants, and there are 74 members in Arabidopsis (Eulgem et al. 2000; Rushton et al. 2010). WRKY TFs are involved in plant growth and developmental processes, including seed dormancy and germination, morphogenesis, flowering, gametophyte, and seed development. Among the WRKY TFs, AtWRKY32 and AtWRKY36 control hypocotyl length in Arabidopsis by targeting a b-ZIP type transcription factor ELONGATED HYPOCOTYL 5 (HYP5) (Zhou et al. 2022; Yang et al. 2018). Zhou et al. (2022) reported that in Arabidopsis, loss-of-function mutants (wrky32-2) exhibited longer hypocotyls than the wild type (Col-0), and overexpression of WRKY32 resulted in shortened hypocotyl phenotypes, similar in shape to Col-0 but not like the swollen hypocotyl of turnip. Thus, we presume that the higher expression of WRKY32 observed in the kohlrabi stem (Fig. 5) might suppress the elongation of kohlrabi epicotyls to some extent, resulting in shortened stems, suggesting that kohlrabi WRKY32 might regulate the expression of other genes involved in biological and physiological processes related to tuberization. In this study, WRKY32 showed five amino acid sequence variations and differential expression between the parents (Figs. 4C and 5). However, this data is not enough for speculating on the association of WRKY32 with epicotyl enlargement, and further specific investigation will be required to explore the exact role of WRKY32 in tuber formation in kohlrabi.

Bol012972 is one of the potential candidate gene that encodes a CBL-interacting kinase protein, CIPK6. The CIPK6 is required for the growth and development of stem hypocotyl and root in plants (Tripathi et al. 2009b). Calcineurin B-like proteins (CBL) are a group of calcium-sensing proteins that specifically target a group of sucrose non-fermenting-related serine/threonine kinases (SnRK3), named CBL-interacting protein kinases (CIPK), to mediate the sensed calcium signal during plant development (Batistic and Kudla 2004; Kolukisaoglu et al. 2004). In Arabidopsis, the CBL-CIPK network of signal transduction depends on differential expression of ten CBLs and 25 partner CIPKs in response to various stimuli in a tissue-specific manner (Kim et al. 2000; Guo et al. 2001). In this study, CIPK6 showed four amino acid substitution between the parents at kinase domains (STKc_SnRK3 and SPS1). Among these, the AA substitution proline to serine (P > S) and threonine to serine (T > S) only occurred in KDH, while T (threonine) and P (proline) is conserved in GCP04, Arabidopsis and other three B. oleracea subspecies (Fig. 4A). This amino acid substitution might have significant effect on stem morphology of kohlrabi and broccoli.

Our previous transcriptome assay using kohlrabi (KDH) and broccoli (GCP04) seedlings revealed that CIPK6 was down-regulated in kohlrabi stem along with the down-regulation of other auxin transporters including PIN4, PIN5, WAT1, Aux1, ABCB4 and ABCB21 (Nuruzzaman et al. 2022). In addition, our previous phytohormones analyses revealed a significantly lower level of IAA in the early growth stages of the kohlrabi stems and down-regulation of auxin transporters and signal transduction-related genes (Nuruzzaman et al. 2023). In Arabidopsis, the knockdown mutant of CIPK6 significantly reduced acropetal shoot-to-root and root basipetal auxin transport, and the plants exhibited swollen hypocotyls and compromised lateral root formation in conjunction with reduced expression of a number of genes involved in auxin transport (PIN2, PIN5) and signaling (ARF16, IAA17, IAA20) compared to the wild type (Tripathi et al. 2009a). In another research, it was further reported that the CIPK6 is involved in polar auxin transport which might function through an ABC integral membrane auxin transporter PGP4, as the mutant line of pgp4-1, an Arabidopsis line mutated in PGP4 also showed reduction in the root tip basipetal and shoot-to-root acropetal auxin transport resulting in the same root phenotype as in atcipk6 (Tripathi et al. 2009b). Furthermore, the CaCIPK6 overexpressed tobacco seedlings exhibited a significantly increased basipetal auxin transport resulting in two times longer root and hypocotyls than the wild genotype (Tripathi et al. 2009a) and CaCIPK6 functionally complements AtCIPK6 in Arabidopsis. The lower IAA might accelerate the higher expression of vascular cambium-related genes, including HB8, REV, and WOX4, which might enhance cambial activity in kohlrabi pith. In fact, the expression of a vascular cambium marker, RsWOX4 in radish, was significantly induced at low IAA concentration, while the transcript level of RsWOX4 was significantly lower at higher IAA during tap root thickening in radish (Dong et al. 2023). Combining the results of our fine-mapping, RNA-seq, and phytohormone profiles, we speculate that Bol012972 (CIPK6) is the most probable candidate for REnBo2 among the three predicted genes in the fine-mapped interval.

Based on this speculation, the probable way of involvement of CIPK6 in the formation of tuberous stem in kohlrabi is proposed, as shown in Fig. S12. The kohlrabi-CIPK6 homolog might promotes tuber formation through maintaining the lower levels of auxin (IAA) by down-regulating the auxin transporters like PIN4, PIN5, WAT1, Aux1, ABCB4 and ABCB21, which might restrict the acropetal/basipetal transport of auxin towards the stem.

The REnBo2 was successfully narrowed-down to an interval of 42.8-kb through a three-step fine mapping using the recombinant plants derived from three different backcross populations, leading to the identification of three annotated genes, Bol012972 (CIPK6), Bol012973 (FAD6) and Bol012974 (WRKY32). Among them, CIPK6 is the most probable candidate as revealed by the gene structure, RT-qPCR, comparative RNA-seq and hormonal assays. Further studies focusing on the verification of this causal gene using CRISPR/Cas9 gene editing technology are warranted. The results of this study will be a useful reference for further research to completely explore the complex biological mechanism behind the formation of tuberous organ in kohlrabi as well as in other Brassica crops.

Acknowledgements The authors sincerely thank the NARO-National Agriculture and Food Research Organization (Japan) for kindly providing the DH line of broccoli (GCP04) and Tohoku Seed Co. LTD. for the DH line of kohlrabi (Kr13-511). The authors also thank Dr. N. Minato and Dr. M. Okada, Niigata University, for their technical assistant. This research was supported by the Monbukagakusho scholarship of the Ministry of Education, Culture, Sports, Science and Technology, Japan (MEXT scholarship) for Nuruzzaman, Md. This work was partially supported by Grant-in-Aid for Challenging Exploratory Research (17 K19262) (JSPS, Japan) to K. Okazaki.

Author contributions M. N., E. F. and K. O. conceived and designed the research. K. O. and E. F. supervised the experiments. M. N. and M. S. conducted the experiments and prepared samples. M. N., E. F. and M. Sh. analyzed the data. M. N. and K. O. wrote the manuscript. E.F. and M. Sh. assisted in revising the manuscript. K. O. contributed to the planning, provided funding and managed the whole project. All authors contributed to the development of this manuscript.

Funding This research was supported by the Monbukagakusho scholarship of the Ministry of Education, Culture, Sports, Science and Technology, Japan (MEXT scholarship) for Nuruzzaman, Md. This work was partially supported by Grant-in-Aid for Challenging Exploratory Research (17 K19262) (JSPS, Japan) to K. Okazaki.

Data availability statement The data that supports the findings of this study are available in the supplementary material of this article. The candidate gene sequence data have been submitted to the DDBJ database (https://www.ddbj.nig.ac.jp) under the accession number LC829494, LC829495, LC829496, LC829497, LC829498, and LC829499 for CIPK6, FAD6, and WRKY32 of kohlrabi (KDH) and broccoli (GCP04), respectively. NGS data of KDH and GCP04 have been submitted to the DDBJ database under the accession number (in submission), respectively.

Competing Interests The authors have no relevant financial or non-financial interests to disclose.

Baggett JR, Kean D (1989) Segregation for heading date and stem enlargement in kohlrabi x broccoli crosses. Euphytica 42:171–176
Bailey TL, Elkan C (1994) Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proceedings of the Second International Conference on Intelligent Systems for Molecular Biology. 28–36
Batistic O, Kudla J (2004) Integration and channeling of calcium signaling through the CBL calcium sensor/CIPK protein kinase network. Planta 219:915–924
Belser C, Istace B, Denis E, Dubarry M, Baurens FC, Falentin C, Genete M, Berrabah W, Chèvre AM, Delourme R, Deniot G (2018) Chromosome-scale assemblies of plant genomes using nanopore long reads and optical maps. Nat plants 4(11):879–887
Cai X, Wu J, Liang J, Lin R, Zhang K, Cheng F, Wang X (2020) Improved Brassica oleracea JZS assembly reveals significant changing of LTR-RT dynamics in different morphotypes. Theor Appl Genet 133:3187–3199
Chen C, Chen H, Zhang Y, Thomas HR, Frank MH, He Y et al (2020) TBtools: An Integrative Toolkit Developed for Interactive Analyses of Big Biological Data. Mol Plant 13:1194–1202
Cheng F, Sun R, Hou X, Zheng H, Zhang F, Zhang Y et al (2016) Subgenome parallel selection is associated with morphotype diversification and convergent crop domestication in Brassica rapa and Brassica oleracea. Nat Genet 48:1218–1224
Dong J, Wang Y, Xu L, Li B, Wang K, Ying J, He Q, Liu L (2023) RsCLE22a regulates taproot growth through an auxin signaling-related pathway in radish (Raphanus sativus L). J Exp Bot 74:233–250
Eklof S, Astot C, Sitbon F, Moritz T, Olsson O, Sandberg G (2000) Transgenic tobacco plants co-expressing Agrobacterium iaa and ipt genes have wild-type hormone levels but display both auxin and cytokinin overproducing phenotypes. Plant J 23:279–284
Eulgem T, Rushton PL, Robatzek S, Somssich IE (2000) The WRKY superfamily of plant transcription factors. Trends Plant Sci 5:199–206
Fuller DQ, Denham T, Arroyo-Kalin M, Lucas L, Stevens CJ, Qin L, Allaby RG, Purugganan MD (2014) Convergent evolution and parallelism in plant domestication revealed by an expanding archaeological record. Proc Nat Acad Sci USA 111:6147–6152
Guo Y, Halfter U, Ishitani M, Zhu JK (2001) Molecular characterization of functional domains in the protein kinase SOS2 that is required for plant salt tolerance. Plant Cell 13:1383–1399
Hoque M, Shea DJ, Asada M, Asad-ud-doullah M, Shimizu M, Fujimoto R, Fukai E, Okazaki K (2017) QTL mapping for tuberous stem formation of kohlrabi (Brassica oleracea var. gongylodes L). Mol Breed 37:109
Hovav R, Chaudhary B, Udall JA, Flagel L, Wendel JF (2008) Parallel domestication, convergent evolution and duplicated gene recruitment in allopolyploid cotton. Genet 179:1725–1733
Iwata H, Ninomiya S (2006) AntMap: constructing genetic linkage maps using an ant Colony optimization algorithm Mapping. Breed Sci 56:371–377
Kim KN, Cheong YH, Gupta R, Luan S (2000) Interaction specificity of Arabidopsis calcineurin B-like calcium sensors and their target kinases. Plant Physiol 124:1844–1853
Kolukisaoglu U, Weinl S, Blazevic D, Batistic O, Kudla J (2004) Calcium sensors and their interacting protein kinases: genomics of the Arabidopsis and rice CBL–CIPK signaling networks. Plant Physiol 134:43–58
Kubo N, Saito M, Tsukazaki H, Kondo T, Matsumoto S, Hirai M (2010) Detection of quantitative trait loci controlling morphological traits in Brassica rapa L. Breed Sci 60:164–171
Kubo N, Saito M, Tsukazaki H, Kondo T, Matsumoto S, Hirai M (2010) Detection of quantitative trait loci controlling morphological traits in Brassica rapa L. Breed Sci 60:164–171
Lenser T, Theiben G (2013) Molecular mechanisms involved in convergent crop domestication. Trends Plant Sci 18:704–714
Lescot M, Dehais P, Thijs G, Marchal K, Moreau Y, Van De Peer Y, Rouze P, Rombauts S (2002) PlantCARE, a database of plant cis-acting regulatory elements and a portal to tools for in silico analysis of promoter sequences. Nucleic Acids Res 30:325–327
Li H, Durbin R (2010) Fast and accurate long read alignment with Burrows-Wheeler transform. Bioinformatics 26(5):589–595
Li H, Zhao N, Yu X, Liu Y, Zhai H, He S et al (2014) Identification of QTLs for storage root yield in sweet potato. Sci Hortic 170:182–188
Li J, Zou X, Zhang L, Cao L, Chen L (2016) Linkage map construction using SSR markers and QTL analyses of stem expansion traits in Brassica juncea. Sci Hortic 209:67–72
Liu S, Liu Y, Yang X, Tong C, Edwards D, Parkin IA, Zhao M, Ma J, Yu J, Huang S, Wang X (2014) The Brassica oleracea genome reveals the asymmetrical evolution of polyploid genomes. Nat commun 5(1):3930
Lou PJ, Zhao JS, Kim S, Shen DP, Carpio D, Song X, Jin M, Vreugdenhil D, Wang X, Koornneef M, Bonnema G (2007) Quantitative loci for flowering time and morphological traits in multiple populations of Brassica rapa. J Exp Bot 58:4005–4016
Lou PJ, Zhao JS, Kim S, Shen DP, Carpio D, Song X, Jin M, Vreugdenhil D, Wang X, Koornneef M, Bonnema G (2007) Quantitative loci for flowering time and morphological traits in multiple populations of Brassica rapa. J Exp Bot 58:4005–4016
Lu G, Cao J, Yu X, Xiang X, Chen H (2008) Mapping QTLs for root morphological traits in Brassica rapa L. based on AFLP and RAPD markers. J Appl Genet 49:23–31
Lu G, Cao J, Yu X, Xiang X, Chen H (2008) Mapping QTLs for root morphological traits in Brassica rapa L. based on AFLP and RAPD markers. J Appl Genet 49:23–31
Lu S, Wang J, Chitsaz F, Derbyshire MK, Geer RC et al (2020) CDD/SPARCLE: the conserved domain database in 2020. Nucleic Acids Res 8:265–268
Lv H, Wang Y, Han F et al (2020) A high-quality reference genome for cabbage obtained with SMRT reveals novel genomic features and evolutionary characteristics. Sci Rep 10:12394
Murray MG, Thompson WF (1980) Rapid isolation of high molecular weight DNA. Nucleic Acids Res 8:4321–4326
Nordstrom A, Tarkowski P, Tarkowska D, Norbaek R, Astot C, Dolezal K, Sandberg G (2004) Auxin regulation of cytokinin biosynthesis in Arabidopsis thaliana: a factor of potential importance for auxin-cytokinin-regulated development. Proc Natl Acad Sci 101:8039–8044
Nuruzzaman M, Kojima M, Sato M, Takebayashi Y, Hoque M, Okamoto S, Shea DJ, Fujimoto R, Fukai E, Sakakibara H, Okazaki K (2023) Comparative anatomical and hormonal analyses between kohlrabi and broccoli seedlings: Relevance to kohlrabi stem tuber initiation. Sci Hortic 316:112002
Nuruzzaman M, Sato M, Okamoto S, Hoque M, Shea DJ, Fujimoto R, Shimizu M, Fukai E, Okazaki K (2022) Comparative transcriptome analysis during tuberous stem formation in Kohlrabi (B. oleracea var. gongylodes) at early growth periods (seedling stages). Physiol Plant 174(5):e13770
Palni LM, Burch L, Horgan R (1988) The effect of auxin concentration on cytokinin stability and metabolism. Planta 174:231–234
Rushton PJ, Somssich IE, Ringler P, Shen QJ (2010) WRKY transcription factors. Trends Plant Sci 15:247–258
Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: molecular evolutionary genetics analysis (MEGA) software version 4.0. Mol Biol Evol 24:1596–1599
Tripathi V, Parasuraman B, Laxmi A, Chattopadhyay D (2009a) CIPK6, a CBL-interacting protein kinase is required for development and salt tolerance in plants. Plant J 58:778–790
Tripathi V, Sayed N, Laxmi A, Chattopadhyay D (2009b) Role of CIPK6 in root growth and auxin transport. Plant Signal Behav 4:663–665
Tsuro M, Suwabe K, Kubo N, Matsumoto S, Hirai M (2008) Mapping QTLs controlling root shape and red pigmentation in radish, Raphanus sativus L. Breed Sci 58:55–61
Tsuro M, Suwabe K, Kubo N, Matsumoto S, Hirai M (2008) Mapping QTLs controlling root shape and red pigmentation in radish, Raphanus sativus L. Breed Sci 58:55–61
Wang S, Basten CJ, Zeng ZB (2012) Windows QTL cartographer 2.5. Department of Statistics, North Carolina State University, Raleigh
Wu Y, Zhang S, Zhang H, Li F, Li G, Fan C, Sun R, Zhang S (2021) QTL mapping and candidate gene identification of swollen root formation in turnip. Int J Mol Sci 22:653
Yang Y, Liang T, Zhang L, Shao K, Gu X, Shang R, Shi N, Li X, Zhang P, Liu H (2018) UVR8 interacts with WRKY36 to regulate HY5 transcription and hypocotyl elongation in Arabidopsis. Nat Plants 4:98–107
Zaki HEM, Yokoi S, Takahata Y (2010) Identification of genes related to root shape in radish (Raphanus sativus) using suppression subtractive hybridization. Breed Sci 60:130–138
Zhou H, Zhu W, Wang X, Bian Y, Jiang Y, Li J, Wang L, Yin P, Deng XW, Xu D (2022) A missense mutation in WRKY32 converts its function from a positive regulator to a repressor of photomorphogenesis. New Phytol 235:111–125

FigureS1.pptx
Supplementary Figure 1. Flow-diagram of development of plant populations for fine mapping of REnBo2. The rectangular boxes indicate the genotypes of the individuals selected from the respected back cross populations to produce sub-sequent generations. The flanking markers for each QTLs were used for Marker Assisted Selection (MAS), which are listed in Table S1.
FigureS2.pptx
Supplementary Figure 2. Phenotypic segregation of the BC₂F₂ recombinant populations for stem swelling related traits. (A-C) Frequency distribution of BC₂F₂ recombinant populations for the phenotypic traits, stem height, maximum stem diameter and grading of tuberous stem, respectively. (D) Phenotypes of four different grades (grade 1 to grade 4) of stem. SEI, Stem enlargement index.
FigureS3.pptx
Supplementary Figure 3. Graphical genotypes of BC₂F₂ recombinants showing grade 1 and grade 4-type stem swelling. The markers, Bol035216 and Bol017491 are the flanking markers of REnBo2 QTL. Kohlrabi-homo (K), broccoli-homo (G) and heterozygous alleles (H) are colored by orange, green and blue, respectively. The physical position of the markers designed within the REnBo2 QTL including the flanking markers in the B. oleracea reference genome (Braol JZS V1.1) are shown in above grey line. SEI: Stem enlargement index.
FigureS4.pptx
Supplementary Figure 4. Phenotypic segregation of BC₃F₂ recombinant populations for stem swelling related traits. (A-C) Frequency distribution of BC₃F₂ recombinant plants for the phenotypic traits, stem height, maximum stem diameter and grading of tuberous stem, respectively. (D) Phenotypes of four different grades (grade 1 to grade 4) of stem. SEI, Stem enlargement index.
FigureS5.pptx
Supplementary Figure 5. Correlation between the genotypes and phenotypes including stem diameter (A) and stem grade (B) for one of the markers (Bol012972) at the fine-mapped regions of REnBo2. Kohlrabi-homo, broccoli-homo and heterozygous alleles for Bol012972 marker are indicated by orange, green and blue colored bar, respectively. The number inside the bar indicate the number of recombinant plants under each category.
FigureS6.pptx
Supplementary Figure 6. Relationship between the phenotypes and genotypes at the recombination site of the REnBo2 fine-mapped region. Stem grade and chromosome segregation in the BC₃F₃ populations derived from self-pollination of selected BC₃F₂ recombinant plants are shown. The REnBo2 flanking markers (indicated by the black dotted line) and the fine-mapped interval (indicated by the black double ended arrow) are shown in the upper position.
FigureS7.pptx
Supplementary Figure 7. Structure analyses of Bol012972 (CIPK6) in parents, Arabidopsis and three other B. oleracea subspecies. (A) Hormone-related cis acting elements on promoter region of CIPK6. (B) Details conserved domain analysis. (C) Motifs locations of CIPK6were shown with combined block diagrams. (D) Sequences of ten identified motifs. AT4G30960, BolC03g081960.2J, BolC3t21148H and Bo3g175290are the homologs of CIPK6 in Arabidopsis, cabbage, broccoli (HDEM) and Chinese kale, respectively.
FigureS8.pptx
Supplementary Figure 8. Structure analyses of Bol012974 (WRKY32) and Bol012973 (FAD6) in parents, Arabidopsis and three other B. oleracea subspecies. Details conserved domains, motifs locations and sequences of ten identified motifs of WRKY32(A-C) and FAD6 (D-F) were shown. AT4G30935, BolC03g081980.2J, BolC3t21152H and Bo3g175320 are the homologs of WRKY32, and AT4G30950, BolC03g081970.2J, BolC3t21150H and Bo3g175310 are the homologs of FAD6 in Arabidopsis, cabbage, broccoli (HDEM) and Chinese kale, respectively.
FigureS9.pptx
Supplementary Figure 9. Alignment of amino acid (AA) sequences of BolC3t21149H (A) and BolC3t21151H (B) in parents (KDH and GCP04) and a B. oleracea reference genome (Broccoli, HDEM). The AA sequences are same in all these three B. oleracea sub-species. KDH and GCP04 are the DH lines of kohlrabi and broccoli, respectively.
FigureS10.pptx
Supplementary Figure 10.Phylogeny analysis of CIPK6 and WRKY32. Phylogenetic relationship of CIPK6 (A) and WRKY32 (B) among different Brassica species including B. oleracea, B. rapa, B. napus, B. juncea, B. nigra and B. carinata, and A. thaliana. Phylogeny of CIPK6 (C) and WRKY32 (D) and their respective homologs in parental lines, Arabidopsis and three different B. oleracea sub-species, namely cabbage, broccoli and Chinese kale. Trees were constructed using NJ method by the MEGA (version 5.0) program using the amino acid sequences.
FigureS11.pptx
Supplementary Figure 11. Differential expression of candidate genes along with their homologs. The transcripts expression of two homologs of CIPK6 at C01 (A) and C07 (B), only homologs of WRKY32 at C07 (C) and FAD6 (D) in stems of kohlrabi and broccoli at 10 and 20 DAG are shown in fragments per kilobase of transcript per million mapped reads (FPKM). Error bars represent the standard error of the mean of three replicated measurements. Statistically significant difference in expression between the parents at each DAG are denoted by an asterisk (*), NS indicate non-significant (P<0.05).
FigureS12.pptx
Supplementary Figure 12. A proposed regulatory mechanism of CIPK6 controlling tuberous stem formation in kohlrabi. The arrows and blunted arrows indicate the up- and down-regulation of the genes in the respective yellow colored boxes, respectively. The green colored rectangular boxes denote the endogenous hormone profiles in kohlrabi stem. This pathway is speculated based on the combined results of this research and our previously conducted phytohormone and transcriptome assay using the kohlrabi and broccoli seedlings (Nuruzzaman et al. 2022; 2023).
TableS1new.xlsx
Supplementary Table 1 (S1). List of primers used in this study. The primer pairs Bol031107 & Bol020154, Bol035216 & Bol017491, CB10065 & BoSF317, Bol019306 & Bol032093, and BoSF1163 & IGF5707b were used as flanking markers for the QTLs REnBo1, REnBo2, REnBo3, REnBo4 and REnBo5, respectively. The primers CIPK6-CDS and CIPK6-Pro was used to amplify the coding sequences and 3 kb promoter regions of the parents kohlrabi (KDH) and broccoli (GCP04).
TableS2.xlsx
Supplementary Table 2 (S2). Genotypes and phenotypes data including stem height (mm), stem diameter (mm) and stem swelling grade of 158 recombinant plants selected from the BC₂F₂ populations for the fine mapping of REnBo2. Kohlrabi-homo, broccoli-homo and heterozygous alleles at each locus are indicated by K, G and H, respectively and colored by orange, green and blue, respectively. Stem height and diameter were measured at 90 days after germination. SEI: Stem enlargement index.
TableS3.xlsx
Supplementary Table 3 (S3). Genotypes and phenotypes data including stem height, stem diameter and grades of sixty-seven recombinant plants selected from the BC₃F₂ populations for fine mapping of REnBo2. Kohlrabi-homo, broccoli-homo and heterozygous alleles at each locus are indicated by K, G and H, respectively and colored by orange, green and blue, respectively. Stem height and the maximum diameter of stem were measured at 75 days after germination. SEI: Stem enlargement index.
TableS4.xlsx
Supplementary Table 4 (S4). Genotypes and phenotypes data including stem height, stem diameter and grades of two hundred and sixteen plants from BC₃F₃ populations for the confirmation of the fine mapping of REnBo2 QTL. Kohlrabi-homo, broccoli-homo and heterozygous alleles at each locus are indicated by K, G and H, respectively and colored by orange, green and blue, respectively. Phenotypic data were collected at 75 days after germination. Stem height and the maximum diameter of stem were measured in millimeter (mm). SEI, Stem Enlargement Index.
TableS5.xlsx
Supplementary Table 5 (S5). Annotated genes in the fine-mapped region of REnBo2 QTL detected in four different Brassica oleracea reference genomes. The reference genomes Braol JZSV1.1, Braol JZSV2.0, Braol HDEMV1 and TO100 represents cabbage (old version), cabbage (updated version), broccoli and Chinese kale, respectively. Bol012971 is only detected Braol JZS V1.1 but not in three other reference genomes. It might be a mistake in gene prediction in Braol JZS V1.1 reference genome as it is older version of Braol JZSV2.0. Thus, it is speculated that Bol012971 might not exist in REnBo2 locus and was discarded from further analyses.

Download PDF

Editorial decision: Major revisions
15 Sep, 2024
Reviewers agreed at journal
25 Jul, 2024
Reviewers invited by journal
25 Jul, 2024
Editor assigned by journal
08 Jul, 2024
First submitted to journal
08 Jul, 2024

You are reading this latest preprint version

Fine mapping of QTL and identification of candidate genes responsible for tuberous stem formation in Kohlrabi (Brassica oleracea var. gongylodes L.)

Status:

Version 1

Abstract

Figures

Background

Materials and methods

Data analysis

Results

Discussion

Conclusion

Declarations

References

Supplementary Files

Status:

Version 1