BnMLO2_2 is associated with SSR resistance
To detect if loci for SSR resistance were existed in natural populations, a GWAS was conducted in a panel of natural population consisting of 222 accessions. Phenotypic measurements were defined as stem lesion length at 17 days post-inoculation (dpi) with S. sclerotiorum in the natural environment. Phenotypic data showed the extreme variations as it ranges from 1.5 cm to 22.8 cm. A total of 2,779,265 SNPs were used to perform GWAS analysis using general linear model. Our results showed that, one significant block from 17.35–17.45 million base pairs (Mbs) on chromosome A08 was identified targeting SSR resistance (Fig. 1A and Table S1). Among the 0.1 Mb block, most of the significant SNPs were within the region of a MLO family gene, named BnMLO2_2 (BnaA08g25340D), and the significant SNPs showed strong LD (Fig. 1B). To insight in the sequence variations in the region nearby BnMLO2_2 coding sequence, SNPs in all accessions were thoroughly checked and then were classified into three haplotype groups based on the variant of eight SNPs. Each haplotype group contained at least 9 accessions (Fig. 1C). Similarly, in the corresponding phenotypic data, haplotype 3 was significantly different from haplotype 1 and haplotype 2, and accessions in haplotype 3 were more resistant to SSR (Fig. 1C and Table S2). Therefore, we speculated that BnMLO2_2 was associated with SSR resistance in B. napus.
BnMLO2 positively regulates SSR resistance in B. napus
Generally, promoter regions played a huge role in the regulation of gene expression at mRNA levels. As shown in Fig. 1, the BnMLO2_2 promoter regions contain eight types of variations distributed across our natural population, and among, eight SNPs were highly associated with SSR resistance. Therefore, the cis-acting regulatory elements in the promoter regions of the three haplotypes were further investigated using related data retrieved from the “PlantCARE” database. Specifically, 1.5 kb upstream coding region of BnMLO2_2 was analyzed, and a total of 30 cis-acting regulatory elements were obtained, and of these, 28 were found in three haplotypes (Fig. 2A, Table S3), except GA-motif and ERE, which was only detected in accessions of haplotype 1 and 3, respectively (Fig. 2A, Table S3). Moreover, majority of cis-acting regulatory elements appeared multiple times across the population of each haplotype. For example, TATA-box appeared 32 times in each haplotype. Interestingly, there were 15 such single cis-acting regulatory elements which were identified at least once in each haplotype (i.e., MYB and MTB-like). The above regulatory elements were categorized into three groups including development, hormonal and stress response. Besides, we also noted that, ERE was unique and appeared only in haplotype 3, and the GA-motif was unique and appeared only in haplotype 1 (Fig. 2B). Lastly, we analyzed the expression level of BnMLO2_2 in the three haplotypes, and noted that expression level of BnMLO2_2 in haplotype 3 was significantly higher than in the other two (Fig. 2D). So, we speculated that the unique cis-acting regulatory element ERE may be involved in the increment of higher expression level in haplotype 3. To test the stability of the three haplotypes, a leaf inoculation experiment was performed using the accessions of the three haplotypes. As expected, haplotype 3 showed less necrosis than haplotype 1 and haplotype 2 at both 36 hpi and 48 hpi (Fig. 2C). Additionally, the lesion length also showed a significantly difference between haplotype 3 and the other two haplotypes (Fig. 2E, Table S4). In summary, these results indicated that promoter difference led to the expression difference of BnMLO2_2, and this difference in expression level of BnMLO2_2 was associated with SSR resistance in B. napus.
BnMLO2 is expressed higher but not induced by S. sclerotiorum
Many genes expressed in different times and spaces and performed specific functions together to regulate the growth and development of plant organisms. There are seven Arabidopsis orthologous genes of AtMLO2 reported in B. napus [51]. To understand their functions during the entire growth period, we used transcriptome data of 35 different tissues and stages of B. napus cultivar ZS11 retrieved from B. napus transcriptome information resource (BnTIR), a recently available database [52]. The fragments per kilobase of transcript per million mapped reads (FPKM) were used to evaluate the expression level of seven genes belongs to BnMLO gene family. Results showed that, BnMLO2_1 and BnMLO2_7 were relatively expressed lower in all tissues, and BnMLO2_2 was relatively higher expressed in leaf and silique tissues (Fig. 3A and Table S5). Moreover, BnMLO2_2 was a co-orthologs of the gene AtMLO2, which involved in PM susceptibility [53, 54]. Thereby, BnMLO2s in the BnMLO gene family may be significantly associated with PM or other diseases in B. napus.
To investigate the expression difference of BnMLO2 in the response to S. sclerotiorum, leaves of two B. napus cultivar i.e., 888-5 (susceptible, S line) and M083 (resistant, R line) were inoculated. Both, control and under 18 hpi leaves of two cultivars were collected for RNA sequencing, and then downstream analyses. FPKM value was used to evaluate the expression level of the seven members in control and treated leaves. Among the seven genes, BnMLO2_5 and BnMLO2_6 was not expressed in both lines, however, BnMLO2_2, BnMLO2_3, and BnMLO2_7 genes were relatively expressed higher in both lines but with no significant difference in both control and treated samples (Fig. 3C and Table S6). The transcriptome data were later verified using qPCR (Figure S1). Alternatively, we repeated the same experiment in other two cultivars called Wester (S line) and ZY821 (R line) and found a similar result as mentioned above (Fig. 3C) [55]. These results illustrated that SSR resistance may result from high expression level of MLO2, and BnMLO2 mediated resistance may be inherent to the plants.
AtMLO2 positively regulates SSR resistance in Arabidopsis
AtMLO2 was an orthologous gene of BnMLO2_2, their protein conserved domains showed high similarity (Fig. 4A). Similarly, protein sequence analysis of AtMLO2 and BnMLO2_2 also showed a high similarity i.e., 73.12% (Fig. 4B). So, AtMLO2 gene was speculated to be involved in positively regulating SSR resistance in Arabidopsis. A construct carrying AtMLO2 coding sequence with 35S promoter was transformed into wild-type (WT) Arabidopsis. In the transgenic plants, the AtMLO2 expression level was significantly increased compared to WT, and no expression was detected in the mlo2 mutant plants (Fig. 4C). The WT, mlo2, and transgenic lines with 35S:MLO2 2# and 35S:MLO2 16# were inoculated with S. sclerotiorum in the leaf, and the disease condition of mlo2 and transgenic lines showed a significant difference in gene expression levels as compare with WT at 24 hpi, 36 hpi, and 48 hpi (Fig. 4D). Moreover, the lesion length of mlo2 plants was longer than WT, and in transgenic lines lesion length was less than WT (Fig. 4E and Table S7). These results indicated that the higher expression level of AtMLO2 was positively related to SSR resistance in Arabidopsis. To further explore the resistance mechanism of AtMLO2, the leaves of 36 hpi were stained with trypan blue and observed cell death. As expected, in the transgenic plants exhibited less cell death than WT, and mlo2 exhibited more cell death than WT (Fig. 4F). Additionally, the expression level of pathogenesis-related gene PR1 was also checked, and results showed that, expression level of PR1 was decreased in the transgenic plants and higher in mlo2 plants as compared with WT before inoculation, however at 24 hpi, the expression was increased, but with no fold-change in the relative levels (Fig. 4G).
Phylogenetic and Syntenic Relationship of MLO Gene Families in Brassica species
B. napus was formed by the event of allopolyploidy between the recent ancestors of B. rapa and B. oleracea, which shared the ancient ancestor of Arabidopsis [56, 57]. During the brassicas genome duplications and genome mergers, MLO genes were embedded from Arabidopsis into the brassicas including B. napus (Fig. 5A). Using 15 Arabidopsis MLO protein sequences as query, 23, 28, and 57 genes were identified in B. rapa (BaMLO), B. oleracea (BoMLO) and B. napus (BnMLO), respectively through the reciprocal BLASTP between the protein sequences retrieved from their respective genome databases (Table S8). The 57 BnMLO genes were distributed across all the chromosomes of B. napus, and their gene length and physio-chemical properties were also different (Figure S2, Table S9). Among the AtMLOs, AtMLO3, AtMLO9, and AtMLO10 genes have no orthologous genes in B. rapa, B. oleracea, and B. napus, however, AtMLO12, AtMLO6, and AtMLO2 have the most of orthologous genes in B. napus. This replication model may imply their unusual biological functions. The AtMLO genes and their orthologous genes in B. rapa, B. oleracea, and B. napus were all used to analyze the evolution of MLO genes through the phylogenetic tree analysis. A total of 123 MLO proteins were used to construct the phylogenetic tree (Fig. 5B). As we had shown in the Fig. 5B, MLO gene families were classified into 4 groups, group I to group IV. Among, Group II was the largest group, which containing 39 MLO genes. The members with similar functions were classified into the same group, such as AtMLO4 and AtMLO11 were both associated with root morphogenesis and were classified in group II (Fig. 5B) [58, 59].
Many angiospermic species have experienced at least one round of whole-genome duplication (WGD) event through hybridization or subsequent chromosome doubling, this phenomenon is called a polyploidization event [48]. In case of B. napus, the gene function divergence in different subgenomes is becoming more significant. So, the enormous syntenic blocks between A and C subgenomes were identified based on aligned syntenic chromosomal regions. Each chromosome had several syntenic blocks in the corresponding A and C subgenome, but not all the genes or blocks had a syntenic region (Fig. 5C). In total, 24 syntenic gene pairs of BnMLO were identified with a higher identity according to their reference protein sequences, however, 5 pairs of BnMLO syntenic gene were mapped on the unknown chromosomal regions, so they were not displayed (Fig. 5C). The other 9 BnMLO genes were also not shown in the figure because none of any syntenic gene was found between the A and C subgenomes (Fig. 5C). Owing to the complex genomic structure of hetero-tetraploid and the agricultural importance of oilseed rape, B. napus is becoming a model species for the research of evolutionary consequences following polyploidy.
Gene Structure and Conserved Domains of BnMLO Genes
To analyze the gene structure and conserved domains of the 15 AtMLO and 57 BnMLO genes, a new phylogenetic tree was constructed containing five classified clades (Fig. 6A). The genes with high sequence similarity were clustered together. A wide difference was recored in the transcript length of the 15 AtMLO and 57 BnMLO genes with a minimum of 141 base pair (bp) and a maximum of 7303 bp, and the peptide length was ranged from 46 to 672 amino acid (aa), and the number of exons varied from 1 to 16 (Fig. 5b and Table S5). In which, BnMLO12_2, BnMLO8_3, BnMLO8_4, and BnMLO8_5 were consisted of only one exon, and with no untranslated region (UTR) either, but 40.4% (23/57) of the genes contained UTR region either in the upstream or downstream of the coding sequence. For some genes, such as BnMLO2_7, and BnMLO13_1 the exons were extended by a large intron. Some of the syntenic genes, for example, BnMLO5_1 and BnMLO5_2 have similar structure, however BnMLO14_2 and BnMLO14_4 were significantly different (Fig. 6B). Although, the protein length of the BnMLO genes were varied, however all 57 proteins contained the MLO domain or MLO superfamily domain, and no other conserved domain was detected (Fig. 5c). Most of the MLO conserved domains were distributed in the N-terminal, and some of the proteins only contains a MLO conserved domain, such as BnMLO12_6, and BnMLO12_10 (Fig. 6C).
Because of the co-existence of short and long fragment in BnMLO genes as well as the large intron inside the exons of several genes (such as BnMLO12_7 and BnMLO13_1), it is obvious that a small-scale duplication, a large fragment missing and transposons insertion occurred during the evolution. So, the transcriptional efficiency as well as biological function may be altered.