Read summary and alignment statistics
The capsule and stem tissue libraries from the HN1 cultivar 49,685,786 and from the Munich cultivar 27,967,424 were collected. After filtering the adaptor sequences and low-quality reads, we acquired 42,862,247 and 26,776,297 clean reads, respectively. In the HN1 cultivar, 96% of the reads were mapped to the opium poppy genome for both the capsule and stem. On the other hand, in the Munich cultivar, 96% of the reads were mapped to capsule tissue, whereas 77% were mapped to stem tissue.
In silico identification and characterization of lncRNAs in Opium poppy cultivars
To identify tissue-specific and cultivar-specific lncRNAs in Opium poppy, we conducted in silico analysis of RNA-seq data from the capsule and stem tissues of the Munich and HN1 cultivars. Here, we identified potential lncRNAs in two different cultivars of opium poppy. After filtering for nucleotides shorter than 200 with a single exon, we estimated the coding potentials of all the libraries.
We obtained 2693 potential lncRNAs from the Munich cultivar stem and capsule transcriptome libraries and 6393 potential lncRNAs from the HN1 cultivar. LncRNAs were then classified into five types: sense-intronic, sense-overlapping, antisense, long intergenic noncoding RNAs (lincRNAs) and others (Table S1). Most of the transcripts were lincRNAs in the two cultivars. The chromosomal locations of these lncRNAs were unevenly distributed. In the Munich cultivar, chromosome 11 consisted of the highest number of lncRNA genes, whereas in the HN1 cultivar, chromosome 1 had the highest number of lncRNA genes (Table S1). We then classified the lncRNAs according to their locations, namely, antisense, sense-overlapping, sense-intronic and lincRNA. Within 2693 lncRNAs, 70.9% of the transcripts were identified as lincRNAs in the Munich cultivar. In the HN1 cultivar, 4889 transcripts accounted for 76.47% of the lncRNAs detected as lincRNAs (Table S1, Figs. 1a and 1b).
Analysis of differentially expressed lncRNAs in opium poppy cultivar tissues
Determination of tissue-specific expression profiles provides important data on lncRNAs involved in different biological processes. In this study, we performed tissue-specific gene expression analysis to determine the functions of opium poppy lncRNAs in the capsule and stem tissues of the Munich and HN1 cultivars. A transcript was considered expressed if it had an FPKM ≥ 1 for lncRNAs in Opium poppy. We demonstrated that a great number of transcripts were tissue specific.
In the Munich cultivar, we detected 6551 and 5038 expressed lncRNA transcripts in the capsule and stem, respectively (Table S2, Fig. (2a)). We also observed that 1427 lncRNA transcripts were shared by at least two tissues. Among the capsule tissue transcripts, XR_003334198.1, TCONS_00045571 and XR_003330631.1 were highly expressed, whereas XR_003330631.1, TCONS_00045571 and TCONS_00036556 had the highest expression levels in the stem.
In the HN1 cultivar, we detected 11152 and 10125 potential lncRNAs expressed in capsule and stem tissues, respectively (Table S2, Fig. (2b)). XR_003325865.1 and XR_003330631.1 had the highest expression levels. A total of 7372 lncRNAs were expressed in common between capsule and stem tissues. In this study, a total of 66 lncRNAs were found to be differentially expressed. Among the 66 DELs, 14 lncRNAs were downregulated, whereas 52 lncRNAs were upregulated. The comparison between cultivars revealed that 160 DELs were unique to the Munich cultivar, while 47 DELs were unique to the HN1 cultivar (Fig. 2c). Interestingly, among the shared DELs, almost all showed no expression in the stem.
To compare the expression patterns of lncRNAs between capsules and stems, differentially expressed lncRNAs (DELs) were detected. The DELs of the capsule and stem tissues for each cultivar were analysed individually. Statistical significance was assigned based on a P value ≤ 0.05 and a log2FC ≥ 2 and ≤-2. Differential expression analysis revealed that a total of 179 opium poppy lncRNAs (101 downregulated and 78 upregulated) were differentially expressed between capsule and stem tissues in the Munich cultivar. Among the differentially expressed lncRNAs, the most upregulated transcripts were TCONS_00098806 and XR_003339374.1 in the capsule vs stem comparison, whereas XR_003337750.1 was the most downregulated (Table S3, Fig. 3).
Target gene prediction, annotation and enrichment
It is well known that lncRNAs can act on neighboring genes through cis-regulation (Wang et al. 2015). We predicted the cis-target genes encoding genes lying upstream and downstream of DELs within 100 kb using a python script. The DEL targets were annotated through Gene Ontology (GO) analyses. According to their functions, in addition to genes with unknown functions, we grouped them into three major GO categories.
In total, we detected 3005 putative cis-target genes that might be targets of 179 DELs in the Munich cultivar (Table S3). The analysis revealed that three DELs (XR_003322948.1, XR_003325330.1 and XR_003326843.1) targeted only one cis-target gene, while 41 DELs targeted more than 10 cis-target genes. In the HN1 cultivar, we detected 923 cis-target genes. The distribution of the genes varied from 1 to 160.
The results revealed that a total of 48 significantly enriched GO terms were enriched according to the DAVID online tool. In the Munich cultivar, GO annotation analysis of cis-targets revealed that the top three biological process subgroups were nucleic acid metabolic process (GO:0090304) and intracellular membrane-bounded organelle (GO:0043231), which were the most enriched subcategories in the cellular compartment and substrate-specific transporter activity (GO:0022892) in the molecular function (Table S4, Fig. 4).
In the HN1 cultivar, GO enrichment analysis revealed that 923 cis-target genes of lncRNAs were significantly enriched for 53 processes, such as response to stimulus (GO:0050896) in biological process, the endomembrane system (GO:0012505) in cellular compartment and ion binding (GO:0043167) in molecular function. The top ten GO terms are given in Fig. 5.
According to the KEGG enrichment analysis of the Munich cultivar, the cis-target genes were mainly associated with the biosynthesis of secondary metabolites and the starch and sucrose metabolism pathways (Fig. 6). On the other hand, for the KEGG pathways of the cis-target genes in the HN1 cultivar, the most significant pathways were associated with the biosynthesis of secondary metabolites, phenylpropanoid biosynthesis, and starch and sucrose metabolism (Fig. 6). In the HN1 cultivar, the isoquinoline biosynthesis pathway (psom00950) was also significantly enriched.
BIAs are the main secondary metabolites of opium poppy and are synthesized in stems and accumulate mainly in capsules. Based on the GO and KEGG enrichment analyses, we focused on the DELs and their cis-target genes associated with secondary metabolite production, especially isoquinoline alkaloid biosynthesis. Although we could not detect significant cis-target genes in the Munich cultivar through enrichment analysis, we found 13 genes involved in BIA biosynthesis. In the HN1 cultivar, we identified three significantly enriched cis-target genes involved in isoquinoline alkaloid biosynthesis. The genes encoding polyphenol oxidase, chloroplastic-like (LOC113318085 and LOC113318091) and 3'-hydroxy-N-methyl-(S)-coclaurine 4'-O-methyltransferase 1-like (LOC113294323) (Table S4) Polyphenol oxidase, chloroplastic-like, the target of XR_003354235.1, are involved in the conversion of L-tyrosine to dopamine. XR_003335810 targeted the 3'-hydroxy-N-methyl-(S)-coclaurine 4'-O-methyltransferase 1-like. In addition, we identified 12 cis-target genes related to BIA biosynthesis. The significantly differentially expressed DELs and their targets for the two cultivars are shown in Table 1.
Table 1
Identification of cis-target genes related to BIA biosynthesis in the Munich and HN1 cultivars
lncRNA transcript | Capsule | Stem | Cis target gene_ID | Gene description |
MUNICH | | | | |
XR_003325309.1 | 0,00 | 21,37 | LOC113280036 | salutaridinol 7-O-acetyltransferase-like |
XR_003326480.1 | 8,31 | 6,65 | LOC113282259 | berberine bridge enzyme-like 8 |
XR_003326480.1 | 8,31 | 6,65 | LOC113284641 | salutaridine reductase-like |
XR_003326480.1 | 8,31 | 6,65 | LOC113282258 | tyrosine/DOPA decarboxylase 1-like |
XR_003329687.1 | 0,00 | 9,07 | LOC113285454 | berberine bridge enzyme-like 8 |
XR_003330692.1 | 4,93 | 0,00 | LOC113285898 | tyrosine/DOPA decarboxylase 1-like |
XR_003331952.1 | 1,06 | 0,00 | LOC113294392 | S-norcoclaurine synthase 2-like |
XR_003359776.1 | 0,16 | 1,82 | LOC113351001 | codeine O-demethylase-like |
XR_003359776.1 | 0,16 | 1,82 | LOC113353263 | codeine O-demethylase-like |
XR_003359776.1 | 0,16 | 1,82 | LOC113353268 | codeine O-demethylase-like |
XR_003361372.1 | 6,59 | 0,00 | LOC113353561 | 3'-hydroxy-N-methyl-(S)-coclaurine 4'-O-methyltransferase 2-like |
XR_003361372.1 | 6,59 | 0,00 | LOC113353562 | codeine O-demethylase-like |
HN1 | | | | |
XR_003335810.1 | 0,00 | 11,52 | LOC113294323 | 3'-hydroxy-N-methyl-(S)-coclaurine 4'-O-methyltransferase 1-like |
XR_003324972.1 | 6,49 | 0,00 | LOC113280120 | berberine bridge enzyme-like 15 |
XR_003328885.1 | 31,66 | 0,00 | LOC113285448 | berberine bridge enzyme-like 8 |
XR_003328885.1 | 31,66 | 0,00 | LOC113285451 | berberine bridge enzyme-like 8 |
XR_003328885.1 | 31,66 | 0,00 | LOC113285452 | berberine bridge enzyme-like 8 |
XR_003328885.1 | 31,66 | 0,00 | LOC113286972 | berberine bridge enzyme-like 8 |
XR_003328885.1 | 31,66 | 0,00 | LOC113286971 | berberine bridge enzyme-like 8 |
XR_003335810.1 | 0,00 | 11,52 | LOC113294301 | codeine O-demethylase-like |
XR_003335810.1 | 0,00 | 11,52 | LOC113294305 | codeine O-demethylase-like |
XR_003335810.1 | 0,00 | 11,52 | LOC113294281 | norreticuline-7-O-methyltransferase-like |
XR_003335810.1 | 0,00 | 11,52 | LOC113294312 | norreticuline-7-O-methyltransferase-like |
XR_003353449.1 | 0,00 | 2,63 | LOC113313111 | S-norcoclaurine synthase 2-like |
XR_003343257.1 | 0,00 | 3,80 | LOC113317142 | tyrosine/DOPA decarboxylase 2-like |
XR_003354235.1 | 1,08 | 0,00 | LOC113318085 | polyphenol oxidase, chloroplastic-like |
XR_003354235.1 | 1,08 | 0,00 | LOC113318091 | polyphenol oxidase, chloroplastic-like |
On the other hand, a total of 10 putative cis-target genes were predicted. in the phenylpropanoid biosynthesis pathway: two berberine bridge enzyme-encoding genes (113285448 and 113280120), six peroxidase genes (113318116, 113361174, 113317634, 113283461, 113283463 and 113283462), an agmatine coumaroyltransferase-2-like gene (113280797) and a beta-glucosidase BoGH3B-like gene (113324160).
Interactions between opium poppy DELs and miRNAs
We mined miRNA sequences to identify the DELs that showed complementarity with one of them. In the Munich cultivar, we detected a total of 818 miRNAs (776 unique) that can be the targets or precursors of DELs using psRNATarget. Among 174 DELs, 94 were found to be complementary to miRNAs downloaded from miRbase. Additionally, we used psRNATarget to investigate which opium poppy miRNAs might interact with DELs. We found that pso-miR414 was complementary to XR_003329687.1 and XR_003344042.1, whereas pso-miR13 was complementary to the XR_003324344.1 and XR_003326480.1 DELs. The results obtained for the HN1 cultivar through psRNATarget revealed that 61 DELs were complementary to 478 miRNAs, 469 of which were unique. Interestingly, no opium poppy miRNAs interacted among the 66 DELs (Table S5).
LncRNA-derived SSRs in Opium poppy
As a microsatellite marker, SSRs are widely used for molecular characterization and genetic diversity (Abebaw et al. 2017). The MISA tool was used to identify SSRs in the Munich and HN1 cultivars. In Munich, 25 SSR sequences were detected from DELs. Among the identified SSRs, the A/T motif was the most frequent (48%). Considering the sequence similarity, the only tetra motif was ATTC/AATG. Among the analysed DELs, 10 were found to have 10 SSRs in the HN1 cultivar. The highest frequency was detected for the A/T motif (Fig. 7).