Skip to main content

Comparative transcriptomics identifies candidate genes involved in the evolutionary transition from dehiscent to indehiscent fruits in Lepidium (Brassicaceae)

Abstract

Background

Fruits are the seed-bearing structures of flowering plants and are highly diverse in terms of morphology, texture and maturation. Dehiscent fruits split open upon maturation to discharge their seeds while indehiscent fruits are dispersed as a whole. Indehiscent fruits evolved from dehiscent fruits several times independently in the crucifer family (Brassicaceae). The fruits of Lepidium appelianum, for example, are indehiscent while the fruits of the closely related L. campestre are dehiscent. Here, we investigate the molecular and genetic mechanisms underlying the evolutionary transition from dehiscent to indehiscent fruits using these two Lepidium species as model system.

Results

We have sequenced the transcriptomes and small RNAs of floral buds, flowers and fruits of L. appelianum and L. campestre and analyzed differentially expressed genes (DEGs) and differently differentially expressed genes (DDEGs). DEGs are genes that show significantly different transcript levels in the same structures (buds, flowers and fruits) in different species, or in different structures in the same species. DDEGs are genes for which the change in expression level between two structures is significantly different in one species than in the other. Comparing the two species, the highest number of DEGs was found in flowers, followed by fruits and floral buds while the highest number of DDEGs was found in fruits versus flowers followed by flowers versus floral buds. Several gene ontology terms related to cell wall synthesis and degradation were overrepresented in different sets of DEGs highlighting the importance of these processes for fruit opening. Furthermore, the fruit valve identity genes FRUITFULL and YABBY3 were among the DEGs identified. Finally, the microRNA miR166 as well as the TCP transcription factors BRANCHED1 (BRC1) and TCP FAMILY TRANSCRIPTION FACTOR 4 (TCP4) were found to be DDEGs.

Conclusions

Our study reveals differences in gene expression between dehiscent and indehiscent fruits and uncovers miR166, BRC1 and TCP4 as candidate genes for the evolutionary transition from dehiscent to indehiscent fruits in Lepidium.

Peer Review reports

Background

Flowering plants (angiosperms) form fruits to protect and disperse their seeds. Fruits come in many different types with different morphologies and different properties such as dry or fleshy, and dehiscent or indehiscent [1]. There is a tremendous variation in fruit types both across and within different plant lineages [2]. However, the evolutionary mechanisms that enabled such dramatic shifts to occur, often in a relatively short period of time, remain largely unknown.

The crucifer family (Brassicaceae) includes a number of economically important plants such as cabbage, broccoli, mustard, radish, and turnips. The model plant Arabidopsis thaliana is also a member of this family [3]. Typical fruits of Brassicaceae species are dehiscent, i.e. that the fruits open upon maturation to release the seeds. Dehiscent fruits also likely represents the ancestral fruit type of Brassicaceae [4]. However, indehiscent fruits, i.e. fruits that only release the seed upon decomposition of the fruit, are found in many tribes distributed across the Brassicaceae phylogeny [5]. The scattered distribution of indehiscent fruits indicates that this property evolved independently several times. This situation is mirrored in the genus Lepidium belonging to Brassicaceae: Species of this genus typically produce two‐seeded dehiscent fruits, but the genus also includes species with indehiscent fruits [6].

Brassicaceae fruits are composed of two fruit valves that are connected to the replum and enclose the developing seeds. Dehiscent fruits, such as those of A. thaliana and Lepidium campestre (also known as field pepperwort or field cress), form a well‐defined dehiscence zone (DZ) at the valve margin [7]. The DZ consists of the lignified layer, a stripe of lignified cells, and a separation layer, a region of small thin‐walled cells [8, 9]. During fruit ripening, the whole fruit dries and shrinks. Only the lignified structures stay rigid. Thereby a spring‐like tension is created within the fruit. At the same time, the middle lamellae of the separation layer cells degenerate to form a pre‐determined breaking zone at which the pressure tears the valves apart from the replum. Consequently, the fruit bursts open to release the seeds [9,10,11]. In contrast, the indehiscent fruits of the closely related Lepidium appelianum do not form a DZ. Instead, a continuous ring of lignified cells surrounds the seeds such that the fruit cannot open [7].

Much of the gene regulatory network underlying the proper formation of the fruit valves, replum and DZ has been elucidated in A. thaliana (reviewed in Ballester and Ferrándiz [12]). Establishment of the DZ requires expression of the two redundant MADS-box genes, SHATTERPROOF1 (SHP1) and SHATTERPROOF2 (SHP2). The SHP1 and SHP2 proteins act as transcription factors and activate the basic helix‐loop‐helix protein‐encoding genes INDEHISCENT (IND), ALCATRAZ (ALC) and SPATULA (SPT), and also autonomously contribute to DZ development [8, 13,14,15].

For correct fruit patterning, it is crucial that the expression of the SHP genes is restricted to the DZ. Three genes encoding transcription factors contribute to this process: The MADS box gene FRUITFULL (FUL) which is expressed in the fruit valves [16, 17], the BEL1‐like homeobox gene REPLUMLESS (RPL) [18], also known as PENNYWISE [19], BELLRINGER [20], VAAMANA [21], and BLH9 [22] which is expressed in the replum, and the floral homeotic gene APETALA2 (AP2) which negatively regulates RPL [23].

Transcription factors controlling the expression of these regulators have also been determined. High levels of the C2H2 zinc finger proteins JAGGED (JAG) and the two closely related YABBY1 group proteins FILAMENTOUS FLOWER (FIL) and YABBY3 (YAB3) activate the expression of FUL [24]. In contrast, lower levels of JAG/FIL/YAB3 expression promote expression of SHP genes. The expression of RPL is activated by the KNOTTED1-like homeobox protein BREVIPEDICELLUS (BP) [25] whose gene is in turn activated by the C2H2 zinc finger protein NO TRANSMITTING TRACT (NTT) [26]. AP2 is negatively regulated by the microRNA miR172 [27].

Additionally, other factors which influence the size and the position of the DZ have been identified. The WUSCHEL-RELATED HOMEOBOX gene 13 (WOX13) controls replum width and negatively regulates JAG/FIL/YAB3 [28]. The auxin-response factors ARF6 and ARF8, which are regulated by miR167 [29], activate miR172 together with FUL [27]. The MYB protein ASYMMETRIC LEAVES 1 (AS1), likely in collaboration with the leucine zipper protein ASYMMETRIC LEAVES 2 (AS2), negatively regulates BP [25].

In general, proteins encoded by genes expressed in the replum often negatively regulate genes expressed in the valves and vice versa. Apart from the already mentioned interactions, this includes negative regulation of the replum gene BP by the valve proteins encoded by JAG/FIL/YAB3, and negative regulation of JAG/FIL/YAB3 by the replum protein RPL [30].

In a previous study, we have shown that orthologues of the valve margin genes are expressed in a similar way in L. campestre (dehiscent fruits) as in A. thaliana fruits but that expression of the respective orthologues is abolished in the corresponding tissues of indehiscent Lepidium appelianum fruits [7]. As parallel mutations in different genes are unlikely, we concluded that the changes in gene expression patterns are probably caused by changes in upstream regulators such as FUL, RPL or AP2.

To conduct a more unbiased approach to identify the genetic changes that lead from dehiscent to indehiscent fruits than the analysis of candidate genes, we have sequenced the transcriptomes of floral buds, flowers and fruits of both, L. campestre and L. appelianum in the present study. We have identified differentially expressed genes (DEGs) and differently differentially expressed genes (DDEGs) where the latter refers to genes for which the change in expression level between two structures is significantly different in one species than in the other. More DEGs were identified in flowers than in fruits and floral buds and a higher number of DDEGs was found in fruits versus flowers than in flowers versus floral buds. Cell wall synthesis and degradation are important processes for fruit opening as revealed by gene ontology (GO) analysis. The fruit valve identity genes FRUITFULL and YABBY3 were identified as DEGs such that the possible cause for the evolutionary transition from dehiscent to indehiscent fruits in Lepidium may even be an upstream factor of these genes. Possible candidates are BRANCHED1 (BRC1), an ortholog of which may determine whether dehiscent or indehiscent fruits develop on the dimorphic plant Aethionema arabicum, and TCP FAMILY TRANSCRIPTION FACTOR 4 (TCP4) which may regulate YABBY3. These two genes were found to be DDEGs. Our study elucidates differences in gene expression patterns between dehiscent and indehiscent fruits and reveals BRC1 and TCP4 as possible causes for the evolutionary transition from dehiscent to indehiscent fruits in Lepidium.

Results

Overview of the RNA-seq analysis and transcriptome assembly

Sequencing resulted in an average number of reads per library of 56 Mio. for the mRNA and 12 Mio. for the small RNA (Table 1). An initial analysis of the data revealed contamination with sequences from thrips, likely due to infestation of the plants by these insects. Hence, we removed reads matching to the genome of the thrips Frankliniella occidentalis [31] as well as uncorrectable and unpaired reads and reads corresponding to organelle sequences. After this filtering step, 42 Mio. were retained for further analyses for the mRNA sample. For the small RNA sample many reads seem to be derived from organelle RNA. Hence, after removing uncorrectable reads and those matching to the Frankliniella occidentalis genome and organelle sequences, only 1.5 Mio. reads remained on average for the small RNA sample (Table 1).

Table 1 Number of reads obtained after sequencing and after correction and pruning steps

Assembly using Trinity [32] resulted in a total of 56,413 transcripts for L. campestre and 70,380 transcripts for L. appelianum after removing putative contaminant sequences but including potential splice variants or fragmentary sequences. The assemblies also contained chimeric sequences composed of two different transcripts which were likely a result of mis-assembly [33]. Separation of chimeric sequences increased the number of transcripts to a total of 57,209 for L. campestre and 71,332 for L. appelianum. We used the Benchmarking Universal Single-Copy Orthologs (BUSCO) tool [34] with the dataset eudicotyledons_odb10 as reference to assess completeness of our transcriptomes. The BUSCO analyses revealed that 94.6% of the expected eudicotyledonous “near-universal single-copy orthologs” are present in our assembly of the L. campestre transcriptome while 94.3% of these BUSCOs are present in our L. appelianum transcriptome (Fig. 1). It is common that some genes are fragmented in de novo assemblies. Hence, we analyzed the length distribution of our assemblies. For both species there are two peaks (Fig. 2). One peak appears at a length of about 240 nucleotides (log 2 length of about 7.9) and probably represents fragments. The other peak was found at a length of about 1,450 nucleotides (log 2 length of about 10.5) which indicates that there are also a number of full-length transcripts.

Fig. 1
figure 1

BUSCO completeness analysis. Transcripts from the L. campestre and L. appelianum assemblies were compared to 2121 Eudicotyledons reference orthologs for completeness assessment

Fig. 2
figure 2

Transcript length distribution of the assembled transcripts of L. campestre and L. appelianum

To detect conserved miRNAs, we mapped the small RNA reads onto the mature miRNAs of A. thaliana as provided by miRBase [35]. We found reads for 64 mature miRNAs belonging to 32 miRNA families in the L. campestre small RNA data (Table 2). Using ShortStack [36] and the L. campestre genome as available from NCBI, we identified three novel miRNAs. However, no putative target genes could be identified in the transcriptome of L. campestre using targetfinder (https://github.com/carringtonlab/TargetFinder). Our L. appelianum small RNA data contained reads of 60 mature miRNAs belonging to 30 miRNA families (Table 2). No novel miRNAs could be identified for L. appelianum using ShortStack and our transcriptome as reference “genome”.

Table 2 miRNAs identified in short read data of L. campestre and L. appelianum  by mapping to A. thaliana mature miRNAs or using ShortStack with transcriptome or genome data as reference. Highlighted in bold are conserved miRNAs according to Chavez-Montez et al., 2014 [37], bold and italic indicate moderately conserved miRNAs according to Chavez-Montez et al., 2014 [37]

To assess completeness of our small RNA data, we compared our results to the set of conserved and moderately conserved miRNA families as identified by miRNA sample sequencing of vascular plants [37]. For both species, we identified reads for all 16 miRNA families that were found to have originated before the emergence of eudicots and to be conserved across virtually all corresponding species. Furthermore, we found reads for 6 miRNA families in our L. campestre and 7 miRNA families in our L. appelianum small RNA data out of 21 miRNA families which were classified as conserved, although missing in a few corresponding species.

Differential gene expression analysis

To conduct differential expression analysis, we identified putative ortholog pairs between the transcripts of the two Lepidium species. Thereby, we excluded ortholog pairs in which the shorter sequence was less half in length than the longer one and we kept only one transcript isoform per gene as described in the methods section. To make sure not to lose genes of interests using this conservative approach, we checked transcripts that were excluded from the ortholog transcriptome and that showed similarity to A. thaliana genes annotated to have “DNA-binding transcription factor activity” (GO:0003700) (Supplemental Table 1). Most of these transcripts do not seem to be involved in fruit development. The only transcripts which may have some relation to fruit development are BRANCHED2 (BRC2), MYB26, KANADI 2 (KAN2) and MYB85. BRC2 has similar but weaker effects on branching than BRC1 [38] and may have similar effects on dehiscence as will be discussed for BRC1 later. MYB26 has been shown to have a role in anther dehiscence [39] and hence a role for this TF also in fruit dehiscence is conceivable. KAN2 has been shown to repress ASYMMETRIC LEAVES2 (AS2) in A. thaliana [40]. AS2 itself is part of the fruit development network [25]. MYB85 has a role in lignin biosynthesis in A. thaliana [41]. As lignification is important for fruit dehiscence, MYB85 may also have a role in this process. Future transcriptome analyses with a deeper sampling may help to also include these genes in the analyses. However, in general, we do not seem to miss many factors with potential functions in fruit development.

Finally, we attained two transcriptome datasets, one for L. campestre and one for L. appelianum, each containing 17,755 transcripts and where each transcript in one species has exactly one putative orthologous transcript in the other species. We will refer to these transcriptome datasets as our ortholog-transcriptomes in the following. We reassessed completeness of our ortholog-transcriptomes and found that 89.2% of the BUSCOs remained in our ortholog-assembly for L. campestre while this value was slightly lower at 89.1% for our L. appelianum ortholog-transcriptome.

Reads were mapped independently to the corresponding ortholog-transcriptome and counted using HTSeq-count [42]. A principal component analysis was conducted based on the normalized number of reads mapping to the ortholog-transcriptomes. As expected, the replicates from the same species and structure clustered together (Fig. 3). The species are separable based on first component which explains 54% of the variance while the structures are separable based on second component which explains 30% variance (Fig. 3).

Fig. 3
figure 3

Principal component analysis of gene expression profiles of all samples. Samples from L. campestre are shown in red, samples from L. appelianum are shown in blue. Samples from floral buds are depicted by circles, samples from flowers by triangles and samples from fruits by squares. PCA shows separation of the two species and the different structures

To learn more about the differences in fruit development between the L. campestre and L. appelianum, we analyzed expression in our ortholog-transcriptomes using the programs DESeq2 [43] and edgeR [44]. We used a multi-factor design to not only be able to identify differentially expressed genes (DEGs) between the species in the same structure and between structures in the same species, but also to identify genes where the change in expression between the structures is different between the two species. We will refer to the genes identified in the latter analyses as differently differentially expressed genes (DDEGs).

DESeq2 generally identified more DEGs and DDEGs than edgeR, but there is a great overlap of genes identified by both programs (Fig. 4, Supplemental Fig. 1). Only this overlap between the two methods will be considered in the following. More DEGs were observed between the same structure of the different species as compared to different structures of the same species. In L. campestre, there are similar numbers of DEGs between flower and bud as compared to fruit and flower. In L. appelianum, there are more than twice as many DEGs in flowers versus buds as compared to fruits versus flowers (Fig. 4). When looking at DEGs in the same structure of the different species, the highest number of DEGs is observed in flowers, followed by fruits and buds.

Fig. 4
figure 4

DEGs and DDEGs between different species and different structures. DEGs and DDEGs called by both of  the two programs edgeR and DESeq2 are shown. Numbers between floral buds, flowers and fruits, respectively, of L. campestre and L. appelianum represent differentially expressed genes (DEGs) between the two species in the corresponding structure. Numbers between different structures of the same species represent DEGs between those structures in the corresponding species. Numbers in the two lavender circles indicate differently differentially expressed genes (DDEGs) between flower and floral buds and between fruits and flowers, respectively, when comparing the two species. Black numbers correspond to all DEGs or DDEGs while red numbers represent up- and blue numbers represent downregulated genes. Supplemental Fig. 1 contains separate numbers for DEGs and DDEGs called by edgeR and DESeq2

We also analyzed DDEGs in our dataset, i.e. genes which had a significantly different change in expression in flowers versus buds and in fruits versus flowers, respectively, in L. appelianum as compared to L. campestre. These genes may have a significantly stronger up- or downregulation in L. appelianum as compared to L. campestre or these genes may be downregulated in one species and upregulated in the other species. We found 70 DDEGs in flowers versus buds and 158 DDEGs in fruits versus flowers when comparing the two species (Fig. 4).

We applied the same methods for the identification of DEGs and DDEGs encoding miRNAs. First, we determined orthologs between the miRNAs based on the A. thaliana miRNAs they mapped to. For 56 mature miRNAs belonging to 28 miRNA families reads were found in the small RNA data for both species and these mature miRNAs could thus be used for differential expression analyses (Table 2). We will refer to this dataset as our ortholog-miRNAs. All 16 miRNA families that are conserved across virtually all species according to [37] and 6 out of 21 miRNA families which were classified as moderately conserved belong to our ortholog-miRNAs dataset. Mapping of small RNA reads, counting and differential expression analyses were done as described for the differential expression analysis of the ortholog-transcriptomes.

Only one miRNA was found to be encoded by a DEG or DDEG by both programs DESeq2 and edgeR. The miRNA homologous to miR165a-3p, miR165b, miR166a-3p, miR166b-3p, miR166c, miR166d, miR166e-3p, miR166f and miR166g of Arabidopsis thaliana [45] (they all only differ by one nucleotide), which we will refer to as miR165a-3p, was found to be encoded by a DDEG when comparing fruits and flowers. Targets of miR165a-3p are HD-Zip transcription factors like PHABULOSA, REVOLUTA and PHAVOLUTA [46]. However, the expression of these target genes does not change much in our transcriptome analyses (Supplemental Fig. 2). Hence, the role of miR165a-3p for fruit dehiscence remains to be clarified.

Gene Ontology and transcription factor analyses

A number of gene ontology (GO) terms [47, 48] of the category molecular function are significantly over- or underrepresented in the DEGs and DDEGs (Table 3). Among them, the terms protein binding (GO:0005515) and RNA binding (GO:0003723) were underrepresented in two datasets of DEGs. Interestingly, several GO terms related to cell wall synthesis and degradation, i.e. pectinesterase activity (GO:0030599), cellulose synthase (UDP-forming) activity (GO:0016760), polygalacturonase activity (GO:0004650) and hydrolase activity, hydrolyzing O-glycosyl compounds (GO:0004553) were overrepresented in different sets of DEGs.

Table 3 Gene ontology (GO) terms significantly over- or underrepresented in DEGs and DDEGs. Terms of the category molecular function of GO were analysed. FDR, false discovery rate

As we were interested in differences in the gene regulatory network involved in fruit dehiscence in the two species, known to be largely composed of transcription factors in Arabidopsis thaliana (reviewed in Ballester and Ferrándiz [12]), we analyzed genes annotated to have “DNA-binding transcription factor activity” (GO:0003700) in more detail. This set includes transcription factors and transcriptional regulators. For simplicity, we will refer to this dataset as genes encoding transcription factors (TFs).

When comparing flowers and buds, 21 and 28 TFs were DEGs in L. campestre and in L. appelianum, respectively. Among them, there are 13 TFs that were DEGs comparing these structures in both species, including four genes with known functions in flower development, AGAMOUS-LIKE 104 (AGL104) [49], SPOROCYTELESS (SPL, also termed NOZZLE) [50], ORESARA1 (ORE1, also termed ANAC092, ATNAC2, ATNAC6) [51] and ZINC FINGER PROTEIN 2 (ZFP2) [52] (Table 4). Between fruits and flowers, there are 12 TFs in L. campestre and 23 TFs in L. appelianum that are DEGs. Five of these genes are DEGs in fruits versus flowers in both species (Table 4). TFs with differential expression between structures in both species are probably those TFs with common functions for flower and fruit development.

Table 4 Genes that are differentially expressed in different structures in both species and that are  annotated as “DNA-binding transcription factor activity” (GO:0003700). Reg., regulation; L.c., L. campestre; L.a., L. appelianum

When comparing the two species, 43 TFs were DEGs in buds, 68 in flowers and 49 in fruits. Among these TFs, 19 were DEGs in all structures (Table 5). Interestingly, four genes involved in flowering time determination, SQUAMOSA PROMOTER BINDING PROTEIN-LIKE 4 (SPL4) [53], NUCLEAR FACTOR Y-B2 (NF-YB2) [54], NUCLEAR FACTOR Y-B10 (NF-YB10) [55] and FLOWERING LOCUS C (FLC) [56], as well as the fruit development genes FRUITFULL (FUL) [17] and YABBY3 (YAB3) [24, 30] (Fig. 5) were on this list.

Table 5 DEGs between the two Lepidium species annotated as “DNA-binding transcription factor activity”. (GO:0003700). Reg., regulation
Fig. 5
figure 5

Gene regulatory network for the development of valve, valve margin and replum of a fruit. The network is based on what has been determined in A. thaliana and is modified after Chavez-Montez et al., 2015 [57]. Relative expression levels of genes in L. campestre (dark grey bars) and L. appelianum  (light grey bars) in floral buds, flowers and fruits (shown from left to right in the bar plots)  as determined in this study by transcriptome analysis are shown. Significant differences between L. appelianum and L. campestre are indicated by asterisks (P ≤ 0.05). Larger depictions of expression levels are shown in Supplemental Fig. 3

Two TFs were found to be DDEGs when comparing flowers and buds in the two species (Table 6), among them MASSUGU 2 (MSG2, also known as INDOLE-3-ACETIC ACID INDUCIBLE 19) [58] which has been shown to be involved in stamen filaments development [59]. Comparing fruits and flowers, seven TFs, PHY-INTERACTING FACTOR 1 (PIF1, also known as PHYTOCHROME INTERACTING FACTOR 3-LIKE 5) [60], MYB DOMAIN PROTEIN 57 (MYB57) [61], TCP FAMILY TRANSCRIPTION FACTOR 4 (TCP4, also known as MATERNAL EFFECT EMBRYO ARREST 35) [62], BRANCHED 1 (BRC1, also known as TCP FAMILY TRANSCRIPTION FACTOR 18) [38], REVEILLE 6 (RVE6) [63], TRIPTYCHON (TRY) [64] and OBF BINDING PROTEIN 4 (OBP4, also termed DOF5.4) [65] are DDEGs in L. appelianum as compared to L. campestre (Supplemental Fig. 4).

Table 6 DDEGs in different structures annotated as “DNA-binding transcription factor activity” (GO:0003700)

Extension of the gene regulatory network for fruit development

We next investigated how the TFs shown to be DDEGs between fruits and flowers may be involved in the gene regulatory network controlling fruit development (Fig. 5). Therefore, we searched for binding sites of the seven TFs identified by chromatin immunoprecipitation followed by sequencing (ChIP-seq) experiments in the promoters of the genes known to be involved in fruit development. On ChIP-Hub [66], no ChIP-seq data is available for BRC1 and for TRY.

Binding of OBP4 was found in the promoter of all but one of the 18 fruit development genes (Table 7). Binding of RVE6, MYB57, PIF1 and TCP4 was detected in the promoters of 11, 7, 5 and 2 fruit development genes, respectively. PIF1 predominantly binds to the promoters of valve identity genes, with binding to four out of eight valve identity gene promoters and apart from that only binding to one of five valve margin genes. ARF8 is the only fruit development gene for which none of the DDEGs was found to bind to its promoter. To the promoters of YAB3 and FUL, which were found to be differentially expressed in all structures between L. campestre and L. appelianum, binding of TCP4, RVE6 and OBP4 and of MYB57, RVE6 and OBP4, respectively, was found. It has to be noted, however, that for the ChIP-seq experiments analyzed here, material from young leaves (MYB57, RVE6 and OBP4) or seedlings (PIF1 and TCP4) was used. Hence, as to whether these factors bind to the promoters of fruit development genes in reproductive tissues has still to be determined.

Table 7 Number of binding sites of TFs found to be DDEGs to the promoters of known fruit development genes

Discussion

Transcriptomes and small RNA datasets of L. campestre and L. appelianum are nearly complete

We have sequenced the transcriptomes of floral buds, flowers and fruits of L. campestre and L. appelianum. Benchmarking of Universal Single-Copy Orthologs (BUSCO) analysis revealed that the transcriptome assemblies of the two species contain more than 94% of the eudicotyledonous “near-universal single-copy orthologs”. This number is similar to or more than that for transcriptome assemblies of other Brassicaceae [67,68,69]. Furthermore, we found members of all 16 miRNA families that were found to have originated before the emergence of eudicots and conserved in eudicots [37]. These findings reveal that our transcriptome and small RNA data includes most of the expected transcripts and miRNAs.

Differences in gene expression mainly between floral structures

We identified more DEGs when comparing the same structure between the two species than comparing different structures in the same species (Fig. 4). This indicates that gene regulation has diverged between the two species. This is different to what has been observed other flowering plant species, where the correlation of gene expression is higher in the same structure of different species than in different structures of the same species [70]. However, in this case microarray expression data was analyzed which may select for conserved genes.

The highest number of DEGs was observed in flowers, followed by fruits and buds, and the highest number of DDEGs was found in fruits versus flowers as compared to flowers versus floral buds. This indicates that the differences in gene expression between L. campestre and L. appelianum are most pronounced between flowers and in the transition from flowers to fruits. This is expected as the developmental program leading to fruit dehiscence or indehiscence needs to be initiated before the fruits are formed. Supportingly, in Aethionema arabicum, a plant that develops dehiscent and indehiscent fruits on the very same individual, differences between the fruit types start to occur early, two days after anthesis [71].

A number of GO terms related to cell wall synthesis and degradation, e.g. pectinesterase activity, cellulose synthase activity and polygalacturonase activity were overrepresented in different sets of DEGs. It has been recognized that secondary cell wall formation at the valve margins [72] and degeneration of cell walls in the separation layer are essential processes for fruit dehiscence after the DZ is correctly specified [73]. Hence, the overrepresentation of GO terms related to cell wall synthesis and degradation is not surprising.

Confirmation of previous expression study

In a previous study, we have compared expression of the valve margin genes as well as the valve gene FUL and the replum gene RPL between L. campestre and L. appelianum by in situ hybridization [7]. We showed that their orthologues from L. campestre (dehiscent fruits) are similarly expressed as in A. thaliana while expression of the respective orthologues is abolished in valve margins of indehiscent L. appelianum fruits. Analysis using qRT-PCR revealed that the valve margin genes IND and SHP1 are expressed at a significantly higher level in flowers and early fruits (the fruit stage for which the transcriptome was sequenced here) of L. campestre than in L. appelianum. Significantly higher expression was confirmed in the present study for SHP1 in flowers (Fig. 5). qRT-PCR analysis revealed significantly higher expression of SHP2 in flowers and of ALC in early fruits in L. appelianum  [7]. Expression was not significantly different in the present transcriptome analysis, but expression was also found to be higher for SHP2 in flowers and for ALC in fruits. Like qRT-PCR analysis, our transcriptome analysis also found significantly higher expression of FUL in fruits of L. campestre. Similarly, RPL was found to be expressed at a higher level in the flowers of L. campestre though the difference was only found to be significant in qRT-PCR but not in transcriptome analysis. AP2 was found to be expressed at a lower level in flowers and fruits of L. campestre by both analyses. Again, the difference was significant in qRT-PCR analysis but not in transcriptome analysis. Hence, our transcriptome analysis is in good agreement with the previous qRT-PCR analyses, but in our transcriptome analysis  differences were less often significant  than in the qRT-PCR analyses.

Known flower and fruit development genes are differentially expressed

To identify differences in the regulation of flower and fruit development between L. campestre and L. appelianum, we focused on differentially expressed or differently differentially expressed genes encoding transcription factors (TFs). Among 19 genes encoding TFs which were found to be differentially expressed in all three examined structures, four TFs are involved in flowering time determination. L. appelianum and L. campestre have different flowering periods according to the Jepson Herbarium (Jepson Flora Project (eds.) 2021, Jepson eFlora, https://ucjeps.berkeley.edu/eflora/, accessed on May 25, 2021), which may be caused by differences in the expression of the identified flowering time genes.

As mentioned above, the fruit development genes SHP1 and FUL were found to be differentially expressed. FUL is expressed at a significantly lower level in L. appelianum in all three structures. In A. thaliana, the ful knockout mutation causes indehiscence [16, 17]. FUL represses the expression of SHP1 and SHP2. Hence, lower expression of FUL in L. appelianum should lead to higher expression of SHP1 and SHP2 in this species. However, only for SHP2 a non-significantly higher expression has been found in flowers of L. appelianum (Fig. 5). SHP1 is unexpectedly expressed at a significantly lower level in L. appelianum flowers. This may be due to the fact that YAB3 [24, 30] (Fig. 5) was expressed at a significantly lower level in L. appelianum. YAB3 does not only activate expression of FUL but also that of SHP1 and SHP2 together with JAG and FIL [24]. Hence, a lower level of YAB3 does not only lead to a lower expression of FUL and to less repression of SHP1 and SHP2 by FUL but also to less activation of SHP1 and SHP2 by YAB3. The contrasting effects of YAB3 and FUL on the expression of SHP1 and SHP2 may lead to the observed overall similar expression of the SHP genes in L. campestre and L. appelianum but may still lead to differences in dehiscence due to different expression domains as described in [24]. yab3 single mutants do not have any major defects in dehiscence but fil yab3 double mutants are largely indehiscent [24]. Hence, decreased expression of YAB3 in L. appelianum as compared to L. campestre may have been an important factor for the evolutionary shift from dehiscent to indehiscent fruits in L. appelianum. This also shows that there was not only a change in the control of valve margin identity genes but also of the valve identity genes and shifts the causative mutation further upstream in the gene regulatory network of fruit development.

MiR165 is differently expressed in fruits versus flowers

Our smallRNA sequencing revealed that the miRNA homologous to miR165a-3p, miR165b, miR166a-3p, miR166b-3p, miR166c, miR166d, miR166e-3p, miR166f and miR166g [45] is encoded by a DDEG when comparing fruits and flowers. Targets of miR165 and miR166 are the mRNAs of HD-Zip transcription factors like PHABULOSA (PHB), REVOLUTA and PHAVOLUTA [46]. Recently, a function of the miR166-PHB module in anther dehiscence has been elucidated [74]. Upregulation of miR166 in the jba-1D mutant leads to downregulation of its target gene PHB which results in increased expression of SPOROCYTELESS/NOZZLE (SPL/NZZ). jba-1D mutants do not develop a dehiscence zone in anthers, i.e. overexpression of miR166 leads to indehiscence of anthers. Expression of miR166 in fruits is much higher in L. appelianum (indehiscent fruits) than in L. campestre (dehiscent fruits), while the opposite is the case in flowers (Fig. 6). Hence, miR166 may have a role in the development of indehiscent fruits in L. appelianum though the details of the regulation remain to be elucidated.

Fig. 6
figure 6

Expression data plot of the miRNA homologous to miR165a-3p of A. thaliana. miR165a-3p is identical  or differs only by one nucleotide to miR165b, miR166a-3p, miR166b-3p, miR166c, miR166d, miR166e-3p, miR166f and miR166g such that they cannot be distinguished and hence are summarized here as miR165-3p. Bars indicate mean normalized count values of reads mapping to miR165-3p in the corresponding structure and species. Dark and light grey bars represent the mean values for L. campestre (Lc) and for L. appelianum (La), respectively. The error bars indicate the standard deviation

BRC1 and TCP4 as candidate genes for the evolutionary shift from dehiscent to indehiscent fruits

Our transcriptome analysis also identified seven genes encoding TFs belonging to DDEGs when comparing flowers and fruits (Table 6). PIF1 is a basic helix-loop-helix (bHLH) transcription factor that negatively regulates chlorophyll biosynthesis [60]; it is involved in a variety of biological processes such as the repression of light-induced seed germination and chlorophyll accumulation in light [75]. RVE6 is a MYB protein that controls the pace of the circadian clock together with its close homologs RVE4 and RVE8 [63]. The zinc finger protein OBP4 functions in cell cycle progression and cell expansion [65] and is involved in root development [76, 77]. So far, involvement of these three factors in flower and fruit development has, to the best of our knowledge, not been reported.

Two other MYB genes, MYB57 and TRY have also been found to be DDEGs (Table 6). MYB57 functions redundantly with MYB21 and MYB24 to regulate stamen development [78]. TRY controls the spacing pattern of trichomes, which are single-celled hairs [64]. Recently it has been found that TRY and other MYB genes of the regulatory network for trichome patterning have been modulated to trigger trichome development in fruits [79]. Hence, these two genes are known to function during flower and fruit development but association with fruit dehiscence is not known so far.

More interestingly, the genes encoding for the two TCP transcription factors BRC1 and TCP4 are DDEGs between fruits and flowers when comparing L. campestre and L. appelianum. Expression of BRC1 correlates with bud inhibition [38, 80] but recently, it has been shown that BRC1 is neither necessary nor sufficient for bud inhibition [81]. Noticeably, it has been hypothesized that BRC1 may guide fruit morph determination in the dimorphic Brassicaceae plant Aethionema arabicum [71]. Ae. arabicum produces two fruit morphs on the same plant, one of which is dehiscent and the other one is indehiscent. qRT-PCR analyses showed that the expression of BRC1 in Ae. arabicum is high in flowers and decreases strongly in  fruits of  the indehiscent morph but remains at a low level in flowers and fruits of the dehiscent morph. We observe a very similar pattern in our transcriptome analysis for the indehiscent morph in L. appelianum and the dehiscent morph in L. campestre (Fig. 7). In Arabidopsis thaliana buds, BRC1 controls a transcription factor cascade that results in abscisic acid (ABA) accumulation [82]. It has been proposed that this cascade also plays a role in the development of indehiscent fruits in Ae. arabicum [83]. Thus the effect of BRC1 on fruit indehiscence in L. appelianum may be indirect via ABA.

Fig. 7
figure 7

Expression data plot of BRC1 in L. campestre (Lc) and L. appelianum (La). Bars indicate mean normalized count values of reads mapping to BRC1 in the corresponding structure and species. Dark and light grey bars represent the mean values for L. campestre (Lc) and for L. appelianum (La), respectively. The error bars indicate the standard deviation

TCP4 has been found to be involved in leaf and flower development as well as in seed oil biosynthesis in A. thaliana [62, 84, 85]. Furthermore, TCP4 directly activates the expression of miR167 which targets the TFs ARF6 and ARF8 [86]. This regulation has been hypothesized to be important for flower maturation, but may also be involved in fruit dehiscence as ARF6 and ARF8 are part of the gene regulatory network of fruit development [29] (Fig. 5). Another study found physical interaction of TCP4 and AS2 in yeast-two-hybrid experiments [87]. AS2 has also previously been found to be involved in fruit patterning [25] (Fig. 5). Our analysis of ChIP-seq data on ChIP-Hub [66] additionally revealed that TCP4 binds to the promoter of YAB3 (Table 7), which has been found to be differentially expressed between L. campestre and L. appelianum in all structures examined. In flowers, TCP4 is expressed at a higher level in L. appelianum than in L. campestre while the expression pattern is the other way round for YAB3. Hence, it is conceivable that TCP4 represses YAB3 in flowers.

Conclusions

Taken together, our study provides insights into the gene regulatory differences in fruit development between L. campestre producing dehiscent fruits and L. appelianum forming indehiscent fruits. We confirm differences in the expression of the fruit development genes SHP2 and FUL between the two species and reveal the importance of the valve identity gene YAB3 for fruit indehiscence in L. appelianum. We uncover the microRNA miR166 and the TCP transcription factors BRC1 and TCP4 as new candidates for causing the evolutionary transition from dehiscent to indehiscent fruits in L. appelianum.

Methods

Plant material, RNA extraction and sequencing

Seeds of Lepidium appelianum (KM 1754) were obtained from J Gaskin, USDA, Fremont County, Wyoming, USA and seeds of L. campestre (KM 96) were aquired from the Botanical Garden of the University of Zürich. All seeds were subsequently mass propagated in the Botanical Garden of Osnabrück University, Germany. Seeds from these mass propagations were sawn on a mixture of seedling substrate (Kammlott, Kammlott GmbH, Erfurt, Germany)/sand/vermiculite (1–3 mm) (8:1:1) which was supplemented with 1 g L−1 each of Osmocote mini (Scotts Miracle‐Gro Company, Marysville, OH, USA) and Triabon (http://www.compo-expert.com, COMPO Expert GmbH, Münster, Germany). The seeds were placed for 4 days at 4 °C for stratification and then put in the greenhouse under a light–dark cycle of 16 h light, 8 h dark of artificial light, plus daylight. After 5 weeks in the greenhouse, the plants were vernalized for at least 13 weeks at 4 °C with a light–dark cycle of 8 h light, 16 h dark. After vernalization, the plants were put back in the greenhouse. Plant material was harvested from two batches of independently grown plants 3 to 5 weeks after the end of vernalization. Plant material was harvested between 12 and 4 pm to minimize the effect of circadian rhythm. Late flower buds, flowers and early fruits were harvested and immediately frozen in liquid nitrogen. Three samples were taken for each species and each structure resulting in 18 samples in total. For each sample, about 100 mg plant material was pooled from four individual plants. The material was pulverized in liquid nitrogen using a mortar and pestle. RNA was extracted using Qiazol (Qiagen) according to the manufacturer’s instructions. RNA quantity and quality was checked by gel electrophoresis. The samples were sent to the Vienna BioCenter Facility for Next Generation Sequencing where they were quality checked and sequenced on a HiSeqV4. For mRNA sequencing, 125 bp paired-end reads were produced and 50 bp single-end reads were generated for small RNA sequencing.

Preprocessing of RNA-seq data

Raw reads were corrected using Rcorrector [88] with default settings. Uncorrectable reads were excluded using a python script obtained from https://informatics.fas.harvard.edu/best-practices-for-de-novo-transcriptome-assembly-with-trinity.html which was slightly modified for excluding uncorrectable reads from smallRNA libraries. The remaining reads were trimmed with Trim Galore! (https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/) using the following settings: --clip_R1 12, --clip R2 12, --paired, --retain_unpaired, --phred33, --length 36, -q 5, --stringency 5, -e 0.1 for transcriptome reads and the following settings --phred33, --length 18, --max_length 26, -q 5, --stringency 5, -e 0.1, -a adapter for small RNA reads where adapter was replaced by the corresponding adapter sequence identified using FastQC (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Thereafter, Poly-A and Poly-T tails were removed from transcriptome reads with PrinSeq [89] using the settings --trim_tail_left 5 and --trim_tail_right 5. Reads that mapped to the genome of Frankliniella occidentalis (GenBank: GCF_000697945) or to rRNAs (GenBank: X52320.1), mitochondrial (GenBank: Y08501.2) or chloroplast (GenBank: AP000423.1) sequences from A. thaliana as determined using bowtie2 (settings: --very-sensitive-local, --phred33) [90] were excluded from further analyses from both, the transcriptome and the smallRNA libraries.

De novo assembly

To simplify de novo assembly, also duplicate reads, i.e. reads with the exact same length and sequence, were removed. The remaining reads were assembled using Trinity [32] with default settings for the two species L. campestre and L. appelianum separately. To identify remaining contamination in the transcriptome, a BLASTn search was conducted against the nucleotide sequence database of NCBI (nt) using the transcripts in the assembly as query with the option -max_target_seqs 5. Transcripts for which the best BLASTn result came from a non-plant species and had an eValue of E < 10–10 were removed from the transcriptomes. The completeness of the assembled transcriptomes was evaluated using the Benchmarking Universal Single-Copy Orthologs tool BUSCO [34] and their accompanying dataset for eudicotyledons with 2121 groups (odb10).

Separation of chimeras in the assemblies

The initial assemblies contained chimeras composed of two different transcripts. As these chimeras often result from misassembly [33], we sought to separate chimeras into their separate transcripts. To recognize chimeras, we first conducted a BLASTn search [91] using the transcripts from the Lepidium transcriptomes as query and the cDNA sequences of the representative A. thaliana gene model as provided by TAIR10 (TAIR10_cdna_20110103_representative_gene_model_updated.fasta) as database saving the best two subjects (i.e. A. thaliana cDNAs) for each query (i.e. for each transcript from the Lepidium transcriptomes) (Fig. 8). Using a customized perl-script (Supplementary Data S1) we searched for instances where the two subjects fitted to different regions of the query (i.e. one part of the Lepidium transcript has a BLAST hit corresponding to one A. thaliana cDNA while another part of the same transcript has a BLAST hit corresponding to another A. thaliana cDNA). These instances likely indicate chimeric Lepidium transcripts. To identify the best position to split the chimeras, we considered at which position and to what extend the A. thaliana cDNAs matched to the Lepidium transcripts as shown in Fig. 8. Chimeras were split if the overlap was less than 150 nucleotides either in the middle of the overlap or at the positions corresponding to the corrected end and beginning of the involved transcripts (Fig. 8).

Fig. 8
figure 8

Schematic presentation of the detection and separation of chimeric transcripts in the Lepidium transcriptomes. The procedure is slightly different depending on whether the positions of the best two BLAST hits in A. thaliana cDNAs overlap on the Lepidium transcript or not. If the positions of the best two BLAST hits overlap by less than 150 nucleotides, the Lepidium transcript is split in the middle of the overlap. Otherwise, the beginning and end of the involved transcripts was determined based on the total length of the fitting A. thaliana cDNAs

Identification of miRNAs in smallRNA libraries

SmallRNA reads were mapped to mature miRNAs from Arabidopsis thaliana as downloaded from miRBase [35] employing bowtie2 with the settings -N 1, -L 18. If a read mapped to a specific miRNA from A. thaliana this miRNA was considered to be present in the corresponding Lepidium species. Mature miRNAs only differing by one nucleotide were combined to avoid multiple mapping during read-counting. To identify novel miRNAs in Lepidium, we used ShortStack [92] with the parameters -- foldsize 500, --dicermin 18 and the trinity transcriptome assembly of the corresponding Lepidium species as reference “genome”. For L. campestre, ShortStack was run a second time, this time using the genome sequence of L. campestre as available at the National Centre for Biotechnology Information [93] as reference. The stem-loop sequences classified as “N15” or “Y” by ShortStack were used as query sequences for BLAST searches of pre-miRNAs of A. thaliana as downloaded from miRBase [35] to distinguish known from novel miRNAs. The stem-loop sequences classified as “Y” by ShortStack without similarity to pre-miRNAs of A. thaliana were classified as novel Lepidium miRNAs.

Determination of orthologs

For transcriptome data, putative ortholog pairs were determined using a reciprocal best hit approach as follows. BLASTn searches were conducted using the transcriptome assembly with chimeras separated of L. campestre as query and the transcriptome assembly with chimeras separated of L. appelianum as subject and vice versa. For each transcript of L. campestre the best BLAST hits (i.e. all hits having the same eValue, score and alignment length) in L. appelianum were recorded and vice versa. If a transcript Tc from L. campestre had the transcript Ta from L. appelianum among its best BLAST hits and transcript Ta from L. appelianum had transcript Tc from L. campestre in its list of best BLAST hits, these were considered as best reciprocal BLAST hit. Best reciprocal BLAST hits with an alignment length of more than 250 nucleotides and where the length of the shorter sequence was at least 50% of that of the longer sequence were considered as putative ortholog pairs. Additionally, another BLASTn search was conducted using the transcripts in the transcriptomes as query and the Arabidopsis thaliana TAIR10 cDNA dataset as database. The set of putative orthologous transcript pairs was pruned such that only one transcript isoform was kept for each species unless different isoforms fitted to different A. thaliana genes. The isoform with the longest alignment length between the two species was chosen to be kept. This way, for each transcript in the one transcriptome exactly one transcript in the other transcriptome was kept. We refer to this dataset as the ortholog-transcriptome. The transcripts in the ortholog-transcriptome dataset were named using the TAIR10 identifier of the best BLAST result or numbered if no BLAST result was obtained this way.

For the miRNA data, orthologous miRNAs of the two Lepidium species were defined as those miRNAs fitting to the same miRNA from A. thaliana. Comparison of the novel Lepidium miRNAs revealed that none of these was found in both species.

Read mapping and feature counting

Preprocessed transcriptome and small RNA reads were mapped against ortholog-transcriptome and mature miRNAs, respectively, using bowtie2 (settings: --very-sensitive-local --phred33 and settings: --phred33 -N 1 -L 18, respectively) [90]. A custom GFF was generated with one feature for each transcript and miRNA. Mapped reads per feature were then counted using HTSeq-count [42] with the settings -s no -t transcript -m union.

Differential gene expression analysis pipeline

Differentially expressed genes were identified using R (https://www.r-project.org/) and the Bioconductor packages edgeR [44] and DESeq2 [43]. Transcript counts were normalized with respect to transcript length. Lowly expressed transcripts with normalized counts and lowly expressed miRNAs with raw counts of less than 19 were discarded. Considering the two species L. campestre and L appelianum and the structures bud, flower and fruit, the following multi-factor design was used: species + structure + species:structure. A Likelihood Ratio Test (LRT) and a quasi-likelihood F-test were conducted in DESeq2 (command: DESeq(object, test=”LRT”, reduced=~species + structure)) and EdgeR (command: glmQLFit(object, design)), respectively to identify differentially expressed and differently differentially expressed genes. Only transcripts and miRNAs having a log-fold change to the base of 2 of more than 1 were considered. For DESeq2 the false discovery rate threshold α was set to 0.001.

For principal component analysis, count data was normalized using regularized logarithm with the option blind=FALSE in DESeq2 and the principal components were plotted using the plotPCA function in R.

GO enrichment analysis

Gene Ontology (GO) enrichment analysis was conducted on the GO website (http://geneontology.org/) using the PANTHER Overrepresentation Test [94]. The TAIR10 identifiers of the transcripts in the ortholog-transcriptome were provided as reference list. The TAIR10 identifiers of the transcripts which were identified as significantly differently expressed genes by both programs, DESeq2 and EdgeR, were provided as analyzed list. Arabidopsis thaliana was chosen as organism and “GO Molecular function complete” was selected as annotation data set. Enriched GO categories were determined using the Fisher's Exact Test with False Discovery Rate correction.

GO categories and terms were also determined using the AnnotationDbi in R. Transcripts associated with the term “DNA-binding transcription factor activity” were analyzed further.

Promoter analyses

Binding of transcription factors to the promoters of genes involved in fruit opening was analysed using ChIP-Hub (http://www.chip-hub.org/). ChIP-Hub provides access to data on binding sites determined using chromatin immunoprecipitation followed by sequencing (ChIP-seq). On the ChIP-Hub website, A. thaliana was chosen as species and binding data was visualized on the WashU EpiGenome Browser. For each fruit development gene, 1,500 nucleotides upstream of the translation start codon were investigated and each occurance of binding of one of the transcription factors found to be DDEGs was noted.

Statement

The study complies with relevant institutional, national, and international guidelines and legislation for plant ethics in the methods section.

Availability of data and materials

The datasets supporting the conclusions of this article are available at NCBI under the BioProject identifier PRJNA769250 https://www.ncbi.nlm.nih.gov/bioproject/769250.

Abbreviations

AGL104:

AGAMOUS-LIKE 104

ALC:

ALCATRAZ

AP2:

APETALA2

ARF:

Auxin-response factor

AS1:

ASYMMETRIC LEAVES 1

AS2:

ASYMMETRIC LEAVES 2

bHLH:

Basic helix-loop-helix

BP:

BREVIPEDICELLUS

BRC1:

BRANCHED1

BUSCO:

Benchmarking of Universal Single-Copy Orthologs

ChIP-seq:

Chromatin immunoprecipitation followed by sequencing

DDEG:

Differently differentially expressed gene

DEG:

Differentially expressed gene

DZ:

Dehiscence zone

FIL:

FILAMENTOUS FLOWER

FLC:

FLOWERING LOCUS C

FUL:

FRUITFULL

GO:

Gene ontology

IND:

INDEHISCENT

JAG:

JAGGED

LRT:

Likelihood Ratio Test

MSG2:

MASSUGU 2

MYB57:

MYB DOMAIN PROTEIN 57

NF-YB2:

NUCLEAR FACTOR Y-B2

NF-YB10:

NUCLEAR FACTOR Y-B10

nt:

Nucleotide sequence database of NCBI

NTT:

NO TRANSMITTING TRACT

NZZ:

NOZZLE

OBP4:

OBF BINDING PROTEIN 4

PHB:

PHABULOSA

PIF1:

PHY-INTERACTING FACTOR 1

RPL:

REPLUMLESS

RVE6:

REVEILLE 6

SHP1:

SHATTERPROOF1

SHP2:

SHATTERPROOF2

SPL:

SPOROCYTELESS

SPL4:

SQUAMOSA PROMOTER BINDING PROTEIN-LIKE 4

SPT:

SPATULA

TCP4:

TCP FAMILY TRANSCRIPTION FACTOR 4

TF:

Transcription factor

TRY:

TRIPTYCHON

WOX13:

WUSCHEL-RELATED HOMEOBOX gene 13

YAB3:

YABBY3

ORE1:

ORESARA1

ZFP2:

ZINC FINGER PROTEIN 2

References

  1. Lorts CM, Briggeman T. Evolution of fruit types and seed dispersal: a phylogenetic and ecological snapshot. J Syst Evol. 2008;46(3):396–404.

    Google Scholar 

  2. Dardick C, Callahan AM. Evolution of the fruit endocarp: molecular mechanisms underlying adaptations in seed protection and dispersal strategies. Front Plant Sci. 2014;5:284.

    PubMed  PubMed Central  Article  Google Scholar 

  3. The Arabidopsis Genome Initiative Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature. 2000;408(6814):796–815. https://doi.org/10.1038/35048692.

  4. Hall JC, Sytsma KJ, Iltis HH. Phylogeny of Capparaceae and Brassicaceae based on chloroplast sequence data. Am J Bot. 2002;89(11):1826–42.

    CAS  PubMed  Article  Google Scholar 

  5. Appel O, Al-Shehbaz I, Kubitzki K, Bayer C. The families and genera of vascular plants. Cruciferae. 2003;5:75–174.

    Google Scholar 

  6. Al-Shehbaz I, Mummenhoff K. Stubendorffia and Winklera belong to the expanded Lepidium (Brassicaceae). Edinb J Bot. 2011;68(2):165.

    Article  Google Scholar 

  7. Mühlhausen A, Lenser T, Mummenhoff K, Theißen G. Evidence that an evolutionary transition from dehiscent to indehiscent fruits in Lepidium (Brassicaceae) was caused by a change in the control of valve margin identity genes. Plant J. 2013;73(5):824–35.

  8. Rajani S, Sundaresan V. The Arabidopsis myc/bHLH gene ALCATRAZ enables cell separation in fruit dehiscence. Curr Biol. 2001;11(24):1914–22.

    CAS  PubMed  Article  Google Scholar 

  9. Spence J, Vercher Y, Gates P, Harris N. ‘Pod shatter’ in Arabidopsis thaliana, Brassica napus and B. juncea. Journal of Microscopy. 1996;181(2):195–203.

  10. Meakin PJ, Roberts JA. Dehiscence of Fruit in Oilseed Rape (Brassica nap us L.) II. THE ROLE OF CELL WALL DEGRADING ENZYMES AND ETHYLENE. Journal of Experimental Botany. 1990;41(8):1003–11.

    CAS  Article  Google Scholar 

  11. Meakin PJ, Roberts JA. Anatomical and biochemical changes associated with the induction of oilseed rape (Brassica napus) pod dehiscence by Dasineura brassicae (Winn.). Annals of Botany. 1991;67(3):193–7.

    CAS  Article  Google Scholar 

  12. Ballester P, Ferrándiz C. Shattering fruits: variations on a dehiscent theme. Curr Opin Plant Biol. 2017;35:68–75.

    PubMed  Article  Google Scholar 

  13. Groszmann M, Paicu T, Alvarez JP, Swain SM, Smyth DR. SPATULA and ALCATRAZ, are partially redundant, functionally diverging bHLH genes required for Arabidopsis gynoecium and fruit development. Plant J. 2011;68(5):816–29.

    CAS  PubMed  Article  Google Scholar 

  14. Liljegren SJ, Ditta GS, Eshed Y, Savidge B, Bowman JL, Yanofsky MF. SHATTERPROOF MADS-box genes control seed dispersal in Arabidopsis. Nature. 2000;404(6779):766–70.

    CAS  PubMed  Article  Google Scholar 

  15. Liljegren SJ, Roeder AH, Kempin SA, Gremski K, Østergaard L, Guimil S, Reyes DK, Yanofsky MF. Control of fruit patterning in Arabidopsis by INDEHISCENT. Cell. 2004;116(6):843–53.

    CAS  PubMed  Article  Google Scholar 

  16. Ferrandiz C, Liljegren SJ, Yanofsky MF. Negative regulation of the SHATTERPROOF genes by FRUITFULL during Arabidopsis fruit development. Science. 2000;289(5478):436–8.

    CAS  PubMed  Article  Google Scholar 

  17. Gu Q, Ferrándiz C, Yanofsky MF, Martienssen R. The FRUITFULL MADS-box gene mediates cell differentiation during Arabidopsis fruit development. Development. 1998;125(8):1509–17.

    CAS  PubMed  Article  Google Scholar 

  18. Roeder AH, Ferrándiz C, Yanofsky MF. The role of the REPLUMLESS homeodomain protein in patterning the Arabidopsis fruit. Curr Biol. 2003;13(18):1630–5.

    CAS  PubMed  Article  Google Scholar 

  19. Smith HM, Hake S. The interaction of two homeobox genes, BREVIPEDICELLUS and PENNYWISE, regulates internode patterning in the Arabidopsis inflorescence. Plant Cell. 2003;15(8):1717–27.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  20. Byrne ME, Groover AT, Fontana JR, Martienssen RA. Phyllotactic pattern and stem cell fate are determined by the Arabidopsis homeobox gene BELLRINGER. Development. 2003;130(17):3941–50.

    CAS  PubMed  Article  Google Scholar 

  21. Bhatt AM, Etchells JP, Canales C, Lagodienko A, Dickinson H. VAAMANA—a BEL1-like homeodomain protein, interacts with KNOX proteins BP and STM and regulates inflorescence stem growth in Arabidopsis. Gene. 2004;328:103–11.

    CAS  PubMed  Article  Google Scholar 

  22. Cole M, Nolte C, Werr W. Nuclear import of the transcription factor SHOOT MERISTEMLESS depends on heterodimerization with BLH proteins expressed in discrete sub-domains of the shoot apical meristem of Arabidopsis thaliana. Nucleic Acids Res. 2006;34(4):1281–92.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  23. Ripoll JJ, Roeder AH, Ditta GS, Yanofsky MF. A novel role for the floral homeotic gene APETALA2 during Arabidopsis fruit development. Development. 2011;138(23):5167–76.

    CAS  PubMed  Article  Google Scholar 

  24. Dinneny JR, Weigel D, Yanofsky MF. A genetic framework for fruit patterning in Arabidopsis thaliana. Development. 2005;132(21):4687–96.

    CAS  PubMed  Article  Google Scholar 

  25. Alonso-Cantabrana H, Ripoll JJ, Ochando I, Vera A, Ferrándiz C, Martínez-Laborda A. Common regulatory networks in leaf and fruit patterning revealed by mutations in the Arabidopsis ASYMMETRIC LEAVES1 gene. Development. 2007;134(14):2663–71.

    CAS  PubMed  Article  Google Scholar 

  26. Marsch-Martínez N, Zúñiga-Mayo VM, Herrera-Ubaldo H, Ouwerkerk PB, Pablo-Villa J, Lozano-Sotomayor P, Greco R, Ballester P, Balanzá V, Kuijt SJ, Meijer AH, Pereira A, Ferrándiz C, de Folter S. The NTT transcription factor promotes replum development in Arabidopsis fruits. Plant J. 2014;80(1):69–81. https://doi.org/10.1111/tpj.12617.

    CAS  Article  PubMed  Google Scholar 

  27. Ripoll JJ, Bailey LJ, Mai Q-A, Wu SL, Hon CT, Chapman EJ, Ditta GS, Estelle M, Yanofsky MF. microRNA regulation of fruit growth. Nature plants. 2015;1(4):1–9.

    Google Scholar 

  28. Romera-Branchat M, Ripoll JJ, Yanofsky MF, Pelaz S. The WOX 13 homeobox gene promotes replum formation in the Arabidopsis thaliana fruit. Plant J. 2013;73(1):37–49.

    CAS  PubMed  Article  Google Scholar 

  29. Zheng L, Nagpal P, Villarino G, Trinidad B, Bird L, Huang Y, Reed JW. miR167 limits anther growth to potentiate anther dehiscence. Development. 2019;146(14).

  30. González-Reig S, Ripoll JJ, Vera A, Yanofsky MF, Martínez-Laborda A. Antagonistic gene activities determine the formation of pattern elements along the mediolateral axis of the Arabidopsis fruit. PLoS Genet. 2012;8(11):e1003020.

  31. González VL, Devine AM, Trizna M, Mulcahy DG, Barker KB, Coddington JA. Open access genomic resources for terrestrial arthropods. Curr Opin Insect Sci. 2018;25:91–8. https://doi.org/10.1016/j.cois.2017.12.003.

  32. Grabherr MG, Haas BJ, Yassour M, Levin JZ, Thompson DA, Amit I, Adiconis X, Fan L, Raychowdhury R, Zeng Q. Trinity: reconstructing a full-length transcriptome without a genome from RNA-Seq data. Nat Biotechnol. 2011;29(7):644.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  33. Yang Y, Smith SA. Optimizing de novo assembly of short-read RNA-seq data for phylogenomics. BMC Genomics. 2013;14(1):1–11.

    Article  CAS  Google Scholar 

  34. Simão FA, Waterhouse RM, Ioannidis P, Kriventseva EV, Zdobnov EM. BUSCO: assessing genome assembly and annotation completeness with single-copy orthologs. Bioinformatics. 2015;31(19):3210–2. https://doi.org/10.1093/bioinformatics/btv351.

  35. Kozomara A, Birgaoanu M, Griffiths-Jones S. miRBase: from microRNA sequences to function. Nucleic Acids Res. 2019;47(D1):D155-62.

    CAS  PubMed  Article  Google Scholar 

  36. Axtell MJ. ShortStack: comprehensive annotation and quantification of small RNA genes. RNA. 2013;19(6):740–51.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  37. Montes RAC, De Paoli E, Accerbi M, Rymarquis LA, Mahalingam G, Marsch-Martínez N, Meyers BC, Green PJ, de Folter S. Sample sequencing of vascular plants demonstrates widespread conservation and divergence of microRNAs. Nat Commun. 2014;5(1):1–15.

    Google Scholar 

  38. Aguilar-Martínez JA, Poza-Carrión C, Cubas P. Arabidopsis BRANCHED1 acts as an integrator of branching signals within axillary buds. Plant Cell. 2007;19(2):458–72.

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  39. Yang C, Xu Z, Song J, Conner K, Vizcay Barrena G, Wilson ZA. Arabidopsis MYB26/MALE STERILE35 regulates secondary thickening in the endothecium and is essential for anther dehiscence. Plant Cell. 2007;19(2):534–48.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  40. Wu G, Lin W-c, Huang T, Poethig RS, Springer PS, Kerstetter RA. KANADI1 regulates adaxial–abaxial polarity in Arabidopsis by directly repressing the transcription of ASYMMETRIC LEAVES2. Proc Natl Acad Sci. 2008;105(42):16392–7.

  41. Geng P, Zhang S, Liu J, Zhao C, Wu J, Cao Y, Fu C, Han X, He H, Zhao Q. MYB20, MYB42, MYB43, and MYB85 regulate phenylalanine and lignin biosynthesis during secondary cell wall formation. Plant Physiol. 2020;182(3):1272–83.

    CAS  PubMed  Article  Google Scholar 

  42. Anders S, Pyl PT, Huber W. HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics. 2015;31(2):166–9.

    CAS  PubMed  Article  Google Scholar 

  43. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15(12):1–21.

    Article  CAS  Google Scholar 

  44. Robinson MD, McCarthy DJ, Smyth GK. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139–40.

    CAS  PubMed  Article  Google Scholar 

  45. Reinhart BJ, Weinstein EG, Rhoades MW, Bartel B, Bartel DP. MicroRNAs in plants Genes & development. 2002;16(13):1616–26.

    CAS  Article  Google Scholar 

  46. Rhoades MW, Reinhart BJ, Lim LP, Burge CB, Bartel B, Bartel DP. Prediction of plant microRNA targets. Cell. 2002;110(4):513–20.

  47. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM, Sherlock G. Gene ontology: tool for the unification of biology. Nat Genet. 2000;25(1):25–9. https://doi.org/10.1038/75556.

  48. Gene Ontology Consortium T. The Gene Ontology resource: enriching a GOld mine. Nucleic Acids Res. 2021;49(D1):D325-d334. https://doi.org/10.1093/nar/gkaa1113.

  49. Adamczyk BJ, Fernandez DE. MIKC* MADS domain heterodimers are required for pollen maturation and tube growth in Arabidopsis. Plant Physiol. 2009;149(4):1713–23. https://doi.org/10.1104/pp.109.135806.

  50. Balasubramanian S, Schneitz K. NOZZLE regulates proximal-distal pattern formation, cell proliferation and early sporogenesis during ovule development in Arabidopsis thaliana. Development. 2000;127(19):4227–38.

    CAS  PubMed  Article  Google Scholar 

  51. Gao Z, Daneva A, Salanenka Y, Van Durme M, Huysmans M, Lin Z, De Winter F, Vanneste S, Karimi M, Van de Velde J, Vandepoele K, Van de Walle D, Dewettinck K, Lambrecht BN, Nowack MK. KIRA1 and ORESARA1 terminate flower receptivity by promoting cell death in the stigma of Arabidopsis. Nat Plants. 2018;4(6):365–75. https://doi.org/10.1038/s41477-018-0160-7.

  52. Cai S, Lashbrook CC. Stamen abscission zone transcriptome profiling reveals new candidates for abscission control: enhanced retention of floral organs in transgenic plants overexpressing Arabidopsis ZINC FINGER PROTEIN2. Plant Physiol. 2008;146(3):1305–21. https://doi.org/10.1104/pp.107.110908.

  53. Jung JH, Lee HJ, Ryu JY, Park CM. SPL3/4/5 Integrate Developmental Aging and Photoperiodic Signals into the FT-FD Module in Arabidopsis Flowering. Mol Plant. 2016;9(12):1647–59. https://doi.org/10.1016/j.molp.2016.10.014.

  54. Cao S, Kumimoto RW, Gnesutta N, Calogero AM, Mantovani R, Holt BF 3rd. A distal CCAAT/NUCLEAR FACTOR Y complex promotes chromatin looping at the FLOWERING LOCUS T promoter and regulates the timing of flowering in Arabidopsis. Plant Cell. 2014;26(3):1009–17. https://doi.org/10.1105/tpc.113.120352.

  55. Wenkel S, Turck F, Singer K, Gissot L, Le Gourrierec J, Samach A, Coupland G. CONSTANS and the CCAAT box binding complex share a functionally important domain and interact to regulate flowering of Arabidopsis. Plant Cell. 2006;18(11):2971–84. https://doi.org/10.1105/tpc.106.043299.

  56. Michaels SD, Amasino RM. FLOWERING LOCUS C encodes a novel MADS domain protein that acts as a repressor of flowering. Plant Cell. 1999;11(5):949–56.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  57. Chavez Montes RA, Herrera-Ubaldo H, Serwatowska J, de Folter S. Towards a comprehensive and dynamic gynoecium gene regulatory network. Current Plant Biology. 2015;3:3–12.

    Article  Google Scholar 

  58. Tatematsu K, Kumagai S, Muto H, Sato A, Watahiki MK, Harper RM, Liscum E, Yamamoto KT. MASSUGU2 encodes Aux/IAA19, an auxin-regulated protein that functions together with the transcriptional activator NPH4/ ARF7 to regulate differential growth responses of hypocotyl and formation of lateral roots in Arabidopsis thaliana. Plant Cell. 2004;16(2):379–93. https://doi.org/10.1105/tpc.018630.

  59. Tashiro S, Ce T, Watahiki MK, Yamamoto KT. Changes in growth kinetics of stamen filaments cause inefficient pollination in massugu2, an auxin insensitive, dominant mutant of Arabidopsis thaliana. Physiol Plant. 2009;137(2):175–87.

    CAS  PubMed  Article  Google Scholar 

  60. Huq E, Al-Sady B, Hudson M, Kim C, Apel K, Quail PH. Phytochromeinteracting factor 1 is a critical bHLH regulator of chlorophyll biosynthesis. Science. 2004;305(5692):1937–41. https://doi.org/10.1126/science.1099728.

  61. Bender RL, Fekete ML, Klinkenberg PM, Hampton M, Bauer B, Malecha M, Lindgren K, J AM, Perera MA, Nikolau BJ, Carter CJ. PIN6 is required for nectary auxin response and short stamen development. Plant J. 2013;74(6):893–904. https://doi.org/10.1111/tpj.12184.

  62. Nag A, King S, Jack T. miR319a targeting of TCP4 is critical for petal growth and development in Arabidopsis. Proc Natl Acad Sci. 2009;106(52):22534–9.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  63. Hsu PY, Devisetty UK, Harmer SL. Accurate timekeeping is controlled by a cycling activator in Arabidopsis. Elife. 2013;2.

  64. Schnittger A, Folkers U, Schwab B, Jürgens G, Hülskamp M. Generation of a spacing pattern: the role of TRIPTYCHON in trichome patterning in Arabidopsis. Plant Cell. 1999;11(6):1105–16.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  65. Xu P, Chen H, Ying L, Cai W. AtDOF5. 4/OBP4, a DOF transcription factor gene that negatively regulates cell cycle progression and cell expansion in Arabidopsis thaliana. Scientific reports. 2016;6(1):1–13.

  66. Fu LY, Zhu T, Zhou X, Yu R, He Z, Zhang P, Wu Z, Chen M, Kaufmann K, Chen D. ChIP-Hub provides an integrative platform for exploring plant regulome. Nat commun. 2022;13(1):1–15.

  67. Chandler JO, Haas FB, Khan S, Bowden L, Ignatz M, Enfissi E, Gawthrop F, Griffiths A, Fraser PD, Rensing SA. Rocket Science: The Effect of Spaceflight on Germination Physiology, Ageing, and Transcriptome of Eruca sativa Seeds. Life. 2020;10(4):49.

    PubMed Central  Article  Google Scholar 

  68. Fernandez-Pozo N, Metz T, Chandler JO, Gramzow L, Mérai Z, Maumus F, Scheid OM, Theißen G, Schranz ME, Leubner-Metzger G . Aethionema arabicum genome annotation using PacBio full-length transcripts provides a valuable resource for seed dormancy and Brassicaceae evolution research. The Plant Journal:- 2021.

  69. Yao S, Liang F, Gill RA, Huang J, Cheng X, Liu Y, Tong C, Liu S. A global survey of the transcriptome of allopolyploid Brassica napus based on single-molecule long-read isoform sequencing and Illumina-based RNA sequencing data. Plant J. 2020;103(2):843–57.

    CAS  PubMed  Article  Google Scholar 

  70. Chanderbali AS, Yoo MJ, Zahn LM, Brockington SF, Wall PK, Gitzendanner MA, Albert VA, Leebens-Mack J, Altman NS, Ma H, dePamphilis CW, Soltis DE, Soltis PS. Conservation and canalization of gene expression during angiosperm diversification accompany the origin and evolution of the flower. Proc Natl Acad Sci U S A. 2010;107(52):22570–5. https://doi.org/10.1073/pnas.1013395108.

  71. Lenser T, Tarkowská D, Novák O, Wilhelmsson PK, Bennett T, Rensing SA, Strnad M, Theißen G. When the BRANCHED network bears fruit: how carpic dominance causes fruit dimorphism in Aethionema. Plant J. 2018;94(2):352–71.

    CAS  PubMed  Article  Google Scholar 

  72. Mitsuda N, Ohme-Takagi M. NAC transcription factors NST1 and NST3 regulate pod shattering in a partially redundant manner by promoting secondary wall formation after the establishment of tissue identity. Plant J. 2008;56(5):768–78.

    CAS  PubMed  Article  Google Scholar 

  73. Ogawa M, Kay P, Wilson S, Swain SM. ARABIDOPSIS DEHISCENCE ZONE POLYGALACTURONASE1 (ADPG1), ADPG2, and QUARTET2 are polygalacturonases required for cell separation during reproductive development in Arabidopsis. Plant Cell. 2009;21(1):216–33.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  74. Li X, Lian H, Zhao Q, He Y. MicroRNA166 monitors SPOROCYTELESS/NOZZLE for building of the anther internal boundary. Plant Physiol. 2019;181(1):208–20.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  75. Castillon A, Shen H, Huq E. Phytochrome interacting factors: central playersin phytochrome-mediated light signaling networks. Trends Plant Sci. 2007;12(11):514–21.

    CAS  PubMed  Article  Google Scholar 

  76. Rymen B, Kawamura A, Schäfer S, Breuer C, Iwase A, Shibata M, Ikeda M, Mitsuda N, Koncz C, Ohme-Takagi M. ABA suppresses root hair growth via the OBP4 transcriptional regulator. Plant Physiol. 2017;173(3):1750–62.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  77. Ramirez-Parra E, Perianez-Rodriguez J, Navarro-Neila S, Gude I, Moreno-Risueno MA, Del Pozo JC. The transcription factor OBP 4 controls root growth and promotes callus formation. New Phytol. 2017;213(4):1787–801.

    CAS  PubMed  Article  Google Scholar 

  78. Cheng H, Song S, Xiao L, Soo HM, Cheng Z, Xie D, Peng J. Gibberellin acts through jasmonate to control the expression of MYB21, MYB24, and MYB57 to promote stamen filament growth in Arabidopsis. PLoS Genet. 2009;5(3):e1000440.

  79. Arteaga N, Savic M, Méndez-Vigo B, Fuster-Pons A, Torres-Pérez R, Oliveros JC, Picó FX, Alonso-Blanco C. MYB Transcription Factors Drive Evolutionary Innovations in Arabidopsis Fruit Trichome Patterning. The Plant Cell. 2021.

  80. Braun N, de Saint GA, Pillot J-P, Boutet-Mercey S, Dalmais M, Antoniadi I, Li X, Maia-Grondard A, Le Signor C, Bouteiller N. The pea TCP transcription factor PsBRC1 acts downstream of strigolactones to control shoot branching. Plant Physiol. 2012;158(1):225–38.

    CAS  PubMed  Article  Google Scholar 

  81. Seale M, Bennett T, Leyser O. BRC1 expression regulates bud activation potential but is not necessary or sufficient for bud growth inhibition in Arabidopsis. Development. 2017;144(9):1661–73.

    CAS  PubMed  PubMed Central  Google Scholar 

  82. González-Grandío E, Pajoro A, Franco-Zorrilla JM, Tarancón C, Immink RG, Cubas P. Abscisic acid signaling is controlled by a BRANCHED1/HD-ZIP I cascade in Arabidopsis axillary buds. Proc Natl Acad Sci. 2017;114(2):E245–54.

  83. Arshad W, Lenser T, Wilhelmsson PK, Chandler JO, Steinbrecher T, Marone F, Pérez M, Collinson ME, Stuppy W, Rensing SA, Theißen G, Leubner-Metzger G. A tale of two morphs: developmental patterns and mechanisms of seed coat differentiation in the dimorphic diaspore model Aethionema arabicum (Brassicaceae). Plant J. 2021;107(1):166–81.

  84. Kong Q, Singh SK, Mantyla JJ, Pattanaik S, Guo L, Yuan L, Benning C, Ma W. Teosinte branched1/cycloidea/proliferating cell factor4 interacts with wrinkled1 to mediate seed oil biosynthesis. Plant Physiol. 2020;184(2):658–65.

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  85. Palatnik JF, Allen E, Wu X, Schommer C, Schwab R, Carrington JC, Weigel D. Control of leaf morphogenesis by microRNAs. Nature. 2003;425(6955):257–63.

    CAS  PubMed  Article  Google Scholar 

  86. Rubio-Somoza I, Weigel D. Coordination of flower maturation by a regulatory circuit of three microRNAs. PLoS Genet. 2013;9(3):e1003374.

  87. Li Z, Li B, Shen WH, Huang H, Dong A. TCP transcription factors interact with AS2 in the repression of class-I KNOX genes in Arabidopsis thaliana. Plant J. 2012;71(1):99–107.

    CAS  PubMed  Article  Google Scholar 

  88. Song L, Florea L. Rcorrector: efficient and accurate error correction for Illumina RNA-seq reads. GigaScience. 2015;4:48. https://doi.org/10.1186/s13742-015-0089-y.

  89. Schmieder R, Edwards R. Quality control and preprocessing of metagenomic datasets. Bioinformatics. 2011;27(6):863–4. https://doi.org/10.1093/bioinformatics/btr026.

  90. Langmead B, Salzberg SL. Fast gapped-read alignment with Bowtie 2. Nat Methods. 2012;9(4):357–9. https://doi.org/10.1038/nmeth.1923.

  91. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990;215(3):403–10. https://doi.org/10.1016/S0022-2836(05)80360-2.

  92. Johnson NR, Yeoh JM, Coruh C, Axtell MJ. Improved Placement of Multi-mapping Small RNAs. G3. 2016;6(7):2103–11. https://doi.org/10.1534/g3.116.030452.

  93. Sayers EW, Beck J, Brister JR, Bolton EE, Canese K, Comeau DC, Funk K, Ketter A, Kim S, Kimchi A, Kitts PA, Kuznetsov A, Lathrop S, Lu Z, McGarvey K, Madden TL, Murphy TD, O’Leary N, Phan L, Schneider VA, Thibaud-Nissen F, Trawick BW, Pruitt KD, Ostell J. Database resources of the National Center for Biotechnology Information. Nucleic Acids Res. 2020;48(D1):D9–16. https://doi.org/10.1093/nar/gkz899.

  94. Mi H, Muruganujan A, Ebert D, Huang X, Thomas PD. PANTHER version 14: more genomes, a new PANTHER GO-slim and improvements in enrichment analysis tools. Nucleic Acids Res. 2019;47(D1):D419–26.

    CAS  PubMed  Article  Google Scholar 

Download references

Acknowledgements

We thank Klaus Mummenhoff for seeds of L. appelianum and L. campestre. We are grateful to Heidi Küster for skillful technical assistance as well as Tobias Horst Rogalla and Georgia Daraki for preliminary analyses.

Funding

Open Access funding enabled and organized by Projekt DEAL. Parts of this work were supported by grants from the Deutsche Forschungsgemeinschaft TH 417/6–2 (to G.T.) and FZT 118, 202548816 (iDIV, to M.M.).

Author information

Authors and Affiliations

Authors

Contributions

GT conceived the project. LG and GT designed the experiments. KK performed the experiments. LG, NFP and MH analyzed the transcriptome data. LG, GT, MM and SR wrote the manuscript. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Günter Theißen.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1:

 Supplementary Data 1. Perl script for the detection of chimeric transcripts in the trinity assembly of the Lepidium transcriptomes.

Additional file 2: Supplementary Figure 1.

 

Additional file 3: Supplementary Figure 2.

Additional file 4: Supplementary Figure 3.

Additional file 5: Supplementary Figure 4.

Additional file 6: Supplementary Table 1. 

Genes that were excluded from the ortholog transcriptome and which are annotated as “DNA-binding transcription factor activity” (GO:0003700).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gramzow, L., Klupsch, K., Fernández-Pozo, N. et al. Comparative transcriptomics identifies candidate genes involved in the evolutionary transition from dehiscent to indehiscent fruits in Lepidium (Brassicaceae). BMC Plant Biol 22, 340 (2022). https://doi.org/10.1186/s12870-022-03631-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12870-022-03631-8

Keywords

  • Transcriptome
  • Fruit development
  • Dehiscence
  • Differentially expressed genes
  • Lepidium appelianum
  • Lepidium campestre