- However, our understanding of the complement, function and origin of lncRNAs – and especially transposon derived lncRNAs (TE-lncRNAs. - By assessing the fraction of lncRNAs that respond to abiotic stresses like heat, cold, salt and drought, we identified 1077 differentially expressed lncRNA transcripts, including 509 TE-lncRNAs. - Much of the additional function is encoded by RNAs, which vary in size from small RNAs (sRNAs) of<. - 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0. - 2 Department of Ecology and Evolutionary, Biology, UC Irvine, Irvine, CA, USA Full list of author information is available at the end of the article. - For example, Kapusta et al. - More than 85% of the maize genome consists of DNA derived from TEs [32], and we therefore expect that many lncRNAs exhibit sequence similarity to TEs.. - Lv et al. - Of the 95,523 assembled tran- scripts, CPC2 annotation identified 31,967 non-coding transcripts (CPC2 score <. - Additional file 1: Fig. - of the total . - Most (61%, or 9341 of 15,312) TE- lncRNAs showed similarity to TEs over ≥90% of their length (Fig. - Hereafter we refer to lncRNAs with se- quence similarity to TEs as “TE-lncRNAs”.. - We further investigated the superfamily of TEs that were similar to the 15,312 TE-lncRNAs. - We found that 86% had sequence similarity to Long Terminal Repeat (LTR) retrotransposons of the Gypsy and Copia super- families (Table 2) and also that some of these TE- lncRNAs exceeded 3750 bp in length (Fig. - A much smaller proportion of TE-lncRNAs were derived from DNA transposons (Table 2). - the longest of these were shorter than the longest TE-lncRNAs with similarity to Gypsy and Copia elements (Fig. - These observations raise an interesting question: Do LTR/Gypsy and LTR/Copia elements give rise to lncRNAs more often than expected, given their propor- tion of the genome? To address this question, we estimated the proportion length among all annotated TEs that were attributable to LTR/Gypsy , LTR/Copia and other element superfamilies, based on RepeatMas- ker analyses. - We then compared these percentages to the proportion length among inferred TE-lncRNAs (Table 2). - We found, for example, that LTR/Gypsy ele- ments produced TE-lncRNAs at roughly the expected. - However, LTR/Copia elements contributed TE-lncRNAs at a lower proportion than their proportion length among annotated TEs (22% vs.. - Particularly notable is the fact that class II DNA elements produced TE-lncRNAs in our dataset at ~ 2- fold higher rate (12% vs. - Altogether, our results verify that the most maize lncRNAs derive from TEs, but they also indicated that different TE superfamilies give rise to TE-lncRNAs at different rates.. - levels, on average, than non-TE-lncRNAs (average FPKM:. - 4.30) and six-fold higher maximum levels, on average, than TE-lncRNAs (average FPKM: 12.57 vs. - c The length distribution of TE-lncRNAs organized by their inferred TE superfamily of origin. - Interestingly, 529 non-redundant TE-lncRNAs were differentially expressed under one or more conditions.. - The proportion of differentially expressed TE-lncRNAs was lower than the proportion of all lncRNAs. - TE- lncRNAs were 65% of the total proportion of lncRNAs, but constituted only 45 and 56% of up- and down- regulated lncRNAs. - Most of the differentially expressed. - TE-lncRNAs had similarity to LTR/Gypsy and LTR/. - Copia , as expected, but other TE families also contrib- uted to differentially expressed TE-lncRNAs. - 2b), suggesting that lncRNAs may either be involved in cis regulation or are subject to some of the same cis regulatory features as their neighboring genes.. - Co-expression modules associated with stress responses Compared to coding genes and microRNAs, the function of lncRNAs in abiotic stress response remains largely Table 2 The proportion of base pairs attributable to different TE superfamilies based on the total length of inferred TE-lncRNAs and the B73 reference genome. - TE Class Number of TE-lncRNAs % total length of TE-lncRNAs % total length of TEs in the genome Retroelements. - 2 Features of the expression dynamics of lncRNAs. - b The correlation (log2(Fold Change)) of TE -lncRNAs and the closest neighboring gene under different stress conditions. - The blue curve indicates the best fit across all of the plotted points and clearly indicates a strong positive correlation from when log2(Fold Change) varies between roughly − 5 and 5. - Table 3 Numbers of differentially expressed genes, lncRNAs and TE-lncRNAs in maize seedlings under abiotic stress. - Computational construction of gene co- expression networks can be a valuable tool for linking lncRNAs and coding RNAs and also for beginning to infer potential biological functions, because co-expressed genes are often members of the same pathway or protein complexes, are often either functionally related, or are controlled by the same transcriptional regulatory pro- gram . - Of the 40 inferred mod- ules, 16 were significantly correlated with stress treat- ments (Fig. - S4, Additional file 4:. - These 16 contained 7221 transcripts including 408 lncRNAs, of which 171 were TE- lncRNAs. - Most of the 16 modules were associated with a single stress and developmental stage, but some were cor- related with two or more stresses or stages (Fig. - 1e-49) and contained 147 lncRNAs and 65 TE-lncRNAs. - 3 A visual representation of the 16 modules that were significantly correlated with abiotic stress responses. - All of the modules were associated with one stress condition and developmental stage, such that they exhibit a temporal cascade of stress responsiveness under different stresses and across V3 to V6 developmental stages. - The scale of the heat map reflects the level of correlation (r) among genes in an expression module for a specific abiotic stress (i.e., Heat, Cold, Drought, Salt, UV, Ozone) at a specific development stages (i.e., V3 to V6). - These two modules contained 52 lncRNAs and 16 TE-lncRNAs, and their eigengenes exhibited significant enrichment of the GO terms ‘intracellular ribonucleoprotein complex’, ‘HslUV protease complex’, ‘cytoplasmic translation’ and ‘intracel- lular membrane-bounded organelle’ (Additional file 6:. - Overall, GO-inferred functions helped verify that the modules reflect aspects of the stress response.. - An interesting facet of the 16 stress-associated modules is that each contained both lncRNAs and TE-lncRNAs.. - We have mentioned that the ME_yellow module con- tained the most lncRNAs of the 16 modules, with 147 lncRNAs and 65 TE-lncRNAs, but other modules were similar in containing lncRNAs. - For example, the ME_tan module, which is associated heat stress in V3, contained 26 lncRNAs and 9 TE-lncRNAs. - 2b), is that some of the lncRNAs in modules are co-expressed with genes due to cis interactions. - Bottom) The bar plots below the heat graph are eigen-lncRNA expression values selected from the top three overrepresented TE-lncRNAs with high interconnectivity. - The x-axis is the same as the heat map, and the id of the TE-lncRNAs is provided by the color key. - This graphs shows, again, that the TE-lncRNAs tend to be more highly expressed under heat stress, but with some dependence on developmental stage. - high connectivity (i.e., intramodular connectivity within the top 10% of all members of the module), membership >. - Based on these filters, we identified 670 hubs that included 39 lncRNAs from different stress-responsive modules (Additional file 5: Table S4), of which 18 were TE-lncRNAs.. - Considering the heat-responsive modules as an example, the 3 associated modules had 27 lncRNAs as hubs, out of 225 total lncRNAs, with 12 of the 27 categorized as TE- lncRNA. - The 27 hub lncRNAs included transcript TE- lncRNAs such as MSTRG.32907 (TE-lncRNA, LINE/L1, p <. - 4, we illustrate the ex- pression patterns of three of the top-ranked hubs within the heat-stress associated modules, with the top-ranked hubs for the other five abiotic stresses in Additional file 1:. - These results suggest the possibility that lncRNAs – and more specifically, some TE-lncRNAs – act to regulate abi- otic stress responses. - We tested this idea by blasting each of the 39 hub lncRNAs to an evolutionary gradient of genomes that in- cluded sorghum, rice and Arabidopsis (Additional file 9:. - Of the 39, 16 had strong hits (e <. - 10 − 15 ) to sor- ghum, a close relative to maize, and 4 of these 16 were TE-lncRNAs. - Moreover, three of the hub lncRNAs had hits to rice, but zero TE-lncRNAs had rice hits, and none of the 39 hub lncRNAs had significant hits to Arabidopsis . - For example, if many lncRNAs are derived from genic regions, then our estimate of the proportion of lncRNAs that are derived from TE-lncRNAs is an overestimate. - It is worth noting, however, that our estimate of the proportion of TE-lncRNAs (65%) is similar to a previous, smaller maize study that estimated 68% of lncRNAs were derived from TEs [8]. - Most lncRNAs are TE-lncRNAs. - As we noted above, our estimates of the proportion of TE- lncRNAs could be too high, based on our search strat- egy. - In each network diagram, the green circles represent TE-lncRNAs. - The size of the dot represents intramodular connectivity, with larger sizes representing higher connectivity. - From these networks, we can infer that lncRNAs and TE-lncRNAs are sometimes as or more interconnected than transcription factors. - of the genome estimated to consist of DNA derived from TEs [32]. - We also investigated the TE families from which TE- lncRNAs originated. - Most of the TE-lncRNAs were de- rived from LTR/Gypsy and LTR/Copia families (Table 2), reflecting their preponderance in the maize genome [32, 53]. - Of the 13,822 polyA+ lncRNAs, we found that were differentially expressed under stress condi- tions, including 529 TE-lncRNAs. - These TE-lncRNAs provided an opportunity to assess whether they could be linked to the expression of nearby genes, indicating some sort of cis -regulatory pattern, as has been observed in other species . - TE-lncRNAs were significantly correlated (r 2 = 0.48. - 2b), suggesting that TE-lncRNAs may either be involved in cis regulation or are subject to some of the same cis regulatory features as their neighboring genes, such as open chromatin structure.. - These experiments were based on leaf tissue from seedlings of the V3, V4 and V6 stages. - Of the 16 modules that were significantly associ- ated with stress responses, we identified 670 hubs, many of which corresponded to genes from known transcription factor families (Fig. - One somewhat surprising finding is that 18 of the 39 lncRNA hubs are related in sequence to – and perhaps derived from - TEs. - The location of TE-lncRNAs as hubs, along with their connectedness to known TFs, suggests that a small subset of TE-derived lncRNAs may function as trans- acting regulatory factors in maize. - Clearly additional work is required to prove that these TE-lncRNAs function as hypothesized in abiotic re- sponse, but their centrality in co-expression modules is nonetheless an intriguing result that is consistent with previous findings showing that most lncRNAs are de- rived from TEs [7] and that lncRNAs can play central regulatory roles in plant and animal development [63].. - Of the Illumina datasets, 89 represented polyA+. - All of the 129 datasets were used for lncRNA dis- covery. - A subset of 71 of the 91 RNAseq datasets were employed for differential gene expression analyses (Add- itional file 2: Table S1). - Finally, all of the 89 polyA+ Illumina RNAseq datasets were used for inferring gene co- expression networks. - Finally, we retained transcripts as high confidence lncRNAs if they passed all of the pre- vious four steps, if they were longer than 200 bp, and if. - 20, based on SAM- Tools [78]. - The correlation values (r) between each of the inferred coexpression modules and the specific traits (e.g., Heat, Cold) at different development times (e.g., V3 to V6). - The x-axis is the same as the heat map, and the id of the lncRNAs is provided by the color key.. - Differentially expressed lncRNAs and TE- lncRNAs under different abiotic stresses across developmental stages. - The funding body has no roles in in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.. - The long intergenic noncoding RNA (lincRNA) landscape of the soybean genome. - Genome-wide identification, splicing, and expression analysis of the myosin gene family in maize (Zea mays). - Unveiling the complexity of the maize transcriptome by single-molecule long-read sequencing. - Transcriptomic profiling of the maize (Zea mays L.) leaf response to abiotic stresses at the seedling stage
Xem thử không khả dụng, vui lòng xem tại trang nguồn hoặc xem
Tóm tắt