Potential role of long noncodon RNA in soybean for salt tolerance

Vijayata Singha, Jogendra Singha, Zeetendra Singha, Sachin Kumara, Manish Sutharb, Giriraj Kumawatc, SK Sanwala
and PC Sharmaa

a
ICAR-Central Soil Salinity Research Institute, Karnal, Haryana, India
b
ICAR-Directorate of Medicinal and Aromatic Plants Research, Boriavi, Anand, Gujarat-387310, India.
c
ICAR- Indian Institute of Soybean Research, Khandwa road, Indore- 452001

*Authors email: vijayata.singh@icar.gov.in

Introduction
An upshot of the saline soil and water in crops/plants is limit growth, development and productivity,
largely, in arid and semi-arid areas. Thus, plants should evolve special adaptive mechanisms to cope with
salt stress. In India arid and semi-arid areas covered about 66% of total cultivated area affecting greatly
with salt affected soils and water. Soybean is one of the important pulse crops for such area. However its
productivity is greatly affected by salt stress conditions. A clear understanding of salt tolerance
mechanism can be resulted in to development of salt tolerant genotypes of soybean that will have great
potential to convert problematic area in to profitable and help in to enhance livelihood and nutritional
security of the population living in such resource poor areas.
Plants generally followed osmotic tolerance, ion exclusion and compartmentation strategies for mitigating
salt stress effects, while, the molecular mechanisms underlying the response to salt stress in plants mainly
due to function of protein-coding genes, such as HKT1, Na+/H+ exchanger (NHX), and Salt Overly
Sensitive (SOS). The mechanism of salt tolerance in soybean is very limited. The findings on salt
tolerance strategies in other plants/crops have very limited scope in soybean hence there will be some
other regulatory factors also that have potential role in regulating the response to salt stress need to be
identified by extensive study. In recent years, numerous studies have shown that noncoding RNAs have
function as precursors of microRNAs and other small RNAs or as microRNA target mimics, regulatory
molecules in various developmental processes and respond to biotic or abiotic stress in plants and are thus
considered to be potential regulators of plant responses to these stresses. However, Long noncoding
RNAs (lncRNAs) are indulging in various biological regulatory processes. The status of lncRNA in plants
and their roles in the tolerance to salt stress lags behind. For understanding the mechanisms of gene
regulation especially the Long Non Coding RNAs with target proteins and metabolites definitely will gain
momentum to provide a molecular basis for gene regulation and better adaptation to salt stress in soybean.
Long non codon RNAs in soybean work done so far by researchers
LncRNAs refers to a class of RNA longer than 200 nucleotides that do not code for proteins, because they
do not possess an open reading frame (ORF) of >100 amino acids. LncRNAs are commonly transcribed
by RNA polymerase II, and as a result, possess a 5′ cap and a poly-A tail, similar to mRNAs (Marchese
et al., 2017). LncRNAs have not been done in plants but in mammals demonstrated the multidimensional
mechanisms by which they act (Kopp and Mendell, 2018). However, one of the recently characterized
lncRNA genes, GmENOD40 (Yang et al., 1993), was identified in soybean (Glycine max), the onward
motion of lncRNA research in plants has been slower than that in mammalian species.
Although the definition of lncRNA is rather arbitrary, coding potential is regularly used for classification
purposes. Various tools are available to evaluate the coding potential of transcripts and distinguish non-
coding RNAs from protein-coding ones using machine learning approaches. These methods have used a
variety of models, such as support vector machine Vieira et al., 2017), random forest (Hu et al., 2017)
logistic regression and REPTree . The features used by these machine learning approaches are usually
extracted directly from nucleotide sequences. Regardless of their popularity, machine learning approaches
require good-quality training sets to perform accurately. In the case of lncRNA identification, they require
a true set of protein-coding transcripts and a true set of lncRNAs. For many organisms, such as soybean, a
true set of lncRNAs may not be available. As a result, these methods will depend heavily on features
learned from protein-coding genes and may thus miss transcripts that encode small peptides. The method
PhyloCSF (http://github.com/mlin/PhyloCSF/wiki) represents another type of tool, which is based on a
statistical phylogenetic model (Lin et al., 2011). This method is highly dependent on the genome
alignment, which is not computed by PhyloCSF itself, and is quite computationally intensive (Lin et al.,
2011). In addition, it is only applicable to aligned genomic regions and ignores highly diverse or
unaligned ones (Lin et al., 2011). The result generated by PhyloCSF can only suggest whether a region is
more likely to be protein-coding or non-coding, and it is insufficient to predict a transcript as an lncRNA
only based on this result. To compensate for the inadequacy of computational methods in lncRNA
prediction, ribosome profiling and mass spectrometry (MS) data can be used to validate the translation
process and peptide products, respectively (Choi et al., 2018).
In view of the aforementioned gap in knowledge on soybean lncRNAs, we took advantage of a large
collection of in-house transcriptome data from various soybean tissues treated under various conditions to
perform a comprehensive identification of soybean lncRNAs. We also retrieved existing soybean
transcriptome data from public sources that are of sufficient quality and sequencing depth to enrich our
analysis. A total of 332 RNA-sequencing (RNA-Seq) data samples were used for this analysis. An
approach that integrated reference-based and de novo transcriptome assemblies was developed to predict
soybean lncRNAs, thus identifying ∼69,000 lncRNAs. We demonstrated that lncRNAs differ
substantially from protein-coding transcripts in terms of lengths, numbers of exons, transposable element
(TE) compositions, and sequence conservation patterns across legume species. Detected lncRNA gene
loci are not simply genomic background noise as their transcription start sites (TSSs) and expression
levels can be linked to their nearest protein-coding genes. We validated the salt stress responses of some
lncRNA genes by reverse transcription quantitative PCR (RT-qPCR). We also showed that a small
portion of lncRNAs can, in fact, encode peptides, albeit short ones (with fewer than 100 amino acids), as
supported by MS evidence. This reveals their true identity as small peptide genes rather than lncRNAs.
Databases of plant lncRNAs

Mammalian lncRNAs, especially human and mouse lncRNAs, were recorded elaborately in public
databases. In addition to basic annotation information, the expression level and imprinting information of
mammalian lncRNAs were also deposited in specific databases. Unlike mammalian lncRNAs, lncRNAs
identified in plants were not comprehensively and timely recorded in public databases.
Currently, only six databases are available for depositing plant lncRNAs. These include TAIR––
Arabidopsis gene structure and function annotation, PlantNATsDB––a comprehensive database of plant
NATs , lncRNAdb––a reference database for lncRNAs NONCODE––integrative annotation of lncRNAs ,
PLncDB––plant lncRNA database , and PNRD––a plant ncRNA database .
Among these databases, NONCODE and lncRNAdb are comprehensive databases but not specifically
designed for recording plant lncRNAs. PlantNATsDB depos-it’s about 2 million NAT pairs from 70 plant
species, but pro-viding no genomic view. Although initially designed for plants, PLncDB currently
deposits lncRNA information only of Arabidopsis, and aims to contain comprehensive information
including genomic, transcriptomic, and epigenomic information related to plant lncRNAs. PNRD aims to
provide lncRNAs of 150 plant species and now contains 5571 lncRNAs of A. thaliana, Oryza sativa, Zea
mays, and Populus trichocarpa only.
Conclusion
Firstly, Researchers must to explore the conservation of sequences and expression patterns of lncRNAs in
responses to salt stress in soybean. Secondly, to provide insights into the metabolites and proteomics
regulatory pathways that responds to salt stress to regulatory role of lncRNAs in the salt tolerance of
soybean. We must conclude that research about lncRNAs will provide a comprehensive reference for
future computational and experimental studies to uncover the functions of lncRNAs, proteins and
metabolomics in salt tolerance and pave the way for development of salt tolerant soybean genotypes.
Reference:
Marchese, F.P., Raimondi, I. & Huarte, M. (2017). The multidimensional mechanisms of long noncoding
RNA function. Genome Biol 18, 206. https://doi.org/10.1186/s13059-017-1348-2

Kopp F, Mendell JT (2018) Functional classification and experimental dissection of long noncoding

RNAs. Cell 172: 393–40

Yang WC, Katinakis P, Hendriks P, Smolders A, deVries F, Spee J, vanKammen A, Bisseling T, Fransse
n H (1993) Characterization of GmENOD40, a gene showing novel patterns of cell-specific expression
during soybean nodule development. Plant J 3: 573–585.

Vieira LM, Grativol C, Thiebaut F, Carvalho TG, Hardoim PR, Hemerly A, Lifschitz S, Ferreira PCG, W
alter MEMT (2017) PlantRNA_Sniffer: An SVM-based workflow to predict long intergenic non-coding
RNAs in plants. Noncoding RNA 3: 11

Hu L, Xu Z, Hu B, Lu ZJ (2017) COME: A robust coding potential calculation tool for lncRNA

identification and characterization based on multiple features. Nucleic Acids Res 45: e2.

Lin MF, Jungreis I, Kellis M (2011) PhyloCSF: A comparative genomics method to distinguish protein

coding and non-coding regions. Bioinformatics 27: i275–i282

Choi S-W, Kim H-W, Nam J W. (2018) .The small peptide world in long noncoding RNAs. Brief

Bioinformatics 20: 1853–1864