Understanding Omics Driven Plant Improvement and de novo Crop Domestication: Some Examples

In the current era, one of biggest challenges is to shorten the breeding cycle for rapid generation of a new crop variety having high yield capacity, disease resistance, high nutrient content, etc. Advances in the “-omics” technology have revolutionized the discovery of genes and bio-molecules with remarkable precision, resulting in significant development of plant-focused metabolic databases and resources. Metabolomics has been widely used in several model plants and crop species to examine metabolic drift and changes in metabolic composition during various developmental stages and in response to stimuli. Over the last few decades, these efforts have resulted in a significantly improved understanding of the metabolic pathways of plants through identification of several unknown intermediates. This has assisted in developing several new metabolically engineered important crops with desirable agronomic traits, and has facilitated the de novo domestication of new crops for sustainable agriculture and food security. In this review, we discuss how “omics” technologies, particularly metabolomics, has enhanced our understanding of important traits and allowed speedy domestication of novel crop plants.


INTRODUCTION
The process of crop domestication began approximately 12,000 years ago, and was an important milestone during human civilization and led the foundation of modern agriculture. In the 21st century, most of the cultivated crops are domesticated from their wild ancestral progenitors (Meyer et al., 2012). During the domestication process plants were selected based on visible traits guided by needs of the time which led to the selection of only few desired alleles and dilution of the genetic variation present within the crop (Figure 1). For example, in cereals like wheat and rice, traits such as increase in the number of seeds per plant, uniform seed maturation, and non-shattering of seeds were preferred over the size of kernels during early domestication (Si et al., 2016). However, the selection of such traits varies greatly from plant to plant or between crops. For instance, in fleshy fruits or berries such as tomato, eggplant and apple, the size of the fruits and berries were preferred over overall yield . Likewise, in tuber producing plants such as potato the tuber size is one of the preferred traits (Fernie and Yan, 2019). Surprisingly, cultivated plant species represent only 250 fully domesticated species among 2500 species, which have undergone the process of domestication, and represent 160 plant families (Smýkal et al., 2018). This proportion is even starker considering the total plant diversity available for the cultivation or those, which are already being cultivated in different places (semicultivated species). For example, around 400,000 semi-cultivated plant species are currently known which can be utilized for designing future crops (Smýkal et al., 2018;Fernie and Yan, 2019).
The process of domestication of a species is a very slow and steady process. In fact, the modern cultivars available were generated following a long series of events: (a) Neolithic Revolution, (b) Columbian Exchange, (c) Industrial Revolution, (d) Green Revolution, and (e) Genomic Revolutions (Smýkal et al., 2018). Presently, to feed an ever-growing global population in the face of climate change is challenge for agriculture especially due to reduction of the arable lands due consistent conversion of lands into semi-arid areas and nutrient deficient land along with increase in soil pH or salinity. Therefore, a more rapid method of developing elite climate smart cultivars with desired traits is required. This could be achieved through integrated OMICS approaches, which include genomics, transcriptomics, proteomics, metabolomics and phenomics integrated with bioinformatics analyses (Kumar et al., 2017(Kumar et al., , 2018Sharma et al., 2021). Plant OMICs based research have played very important role in deciphering metabolic pathways and their molecular regulators, which govern key traits and several plant developmental processes (Kumar et al., 2017;Razzaq et al., 2019). In the past decade there has been a significant progress in the field of both sequencing (van Dijk et al., 2018;Kumar et al., 2020;Schmidt et al., 2020) and analytical methods for the detection of molecules Gilmore et al., 2019;Macklin et al., 2020), which has not only improved detection throughput but also the accuracy and the sensitivity (Kumar et al., 2017;Chiang et al., 2018;Qi et al., 2019).
In the past, for the selection of traits breeding programs involved phenotypic selection of plants (which are guided by metabolic pathways) (Kiszonas and Morris, 2018). For instance, during the Green Revolution (from 1960 to 1980), development of semi-dwarf high yielding varieties of rice and wheat involved phenotypic selections of improved lines which actually involved selection of gibberellic acid pathway genes including the GA20 oxidase and DELLA protein encoding genes (Silverstone and Sun, 2000). In fact, most of the traits, which were targeted for the Green Revolution, are controlled by one or more metabolic pathways. Therefore, precise editing of these metabolic pathways can help in the development of varieties in a very short time (Rodríguez-Leal et al., 2017;Fernie and Yan, 2019). Previously, most of the reviews on plant omics have focused on the instrumentation involved and results obtained by different researchers (Kumar et al., 2017;Mangul et al., 2019;Misra et al., 2019;Tang and Aristilde, 2020). In this review, we represent how this omics knowledge can be utilized for development of improved cultivars by targeting metabolic pathways and also emphasize the use of this information for de novo domestication of wild ancestral species for sustainable food security.

ROLE OF OMICS DATA IN UNDERSTANDING PLANT TRAITS
Genomics plays an important role in the identification of quantitative trait loci (QTLs) and genes controlling important traits, particularly in domesticated crops (Fernie and Yan, 2019). Moving forward, great insights have been gleaned from genome sequencing and re-sequencing programs examining wild ancestral species of domesticated crops (Unamba et al., 2015). In plant genomics, Next Generation Sequencing (NGS) has played a very important role and provided opportunities in the field of functional genomics due to the availability of reference genomes for several model and crop plant species. These resources combined with high quality re-sequencing offers huge potential for discovery of causal genes and mechanisms associated with the key agronomic traits Chen et al., 2019;Varshney et al., 2019). Re-sequencing also enriched the availability of SNPs data and can be utilized for genomicsbased studies such as GWAS (genome wide association study) and QTL-seq , both of which are useful tools for trait mapping (Rivas et al., 2011;Zhu et al., 2011;Zhang et al., 2021). With the advent of these technologies combined with advances in metabolomics such as integration of GWAS with metabolomics efficient means for dissecting underlying molecular mechanisms involved in the growth and development are available (Table 1; Fang and Luo, 2019).

Sequencing and QTL-seq Based Trait Discovery
Presently, QTL-seq is one of the most successful approach developed for the gene discovery and trait dissection Pandey et al., 2020). This approach offers preliminary idea for positional cloning for linked genetic factors or genes, and it has excellent success in marker-assisted selection for crop improvement programs (Xu F. et al., 2015). With the advancements in NGS technologies new approaches like quantitative trait locus sequencing (QTL-seq) has been utilized for exploring rapid QTL or gene identification (Takagi et al., 2013). In QTL-seq approach, the extreme highest and lowest genotypes are selected from the mapping population for target traits, followed by mixing an equal amount of DNA from each bulk to build up two extreme bulk (High bulk and low bulk). Then, each bulk is sequenced and assembled and gene annotation is performed. This approach combined with Bulked segregant analysis, accompanied by whole genome re-sequencing technologies, is more effective and capable than the previous QTL detection methods (Takagi et al., 2013). Utilizing QTL-seq approach several QTLs and genes for different rice phenotypes (Takagi et al., 2013;Daware et al., 2016;Ogiso-Tanaka et al., 2017;Yang et al., 2017;Kadambari et al., 2018;Qin et al., 2018;Bommisetty et al., 2020;Nubankoh et al., 2020), soybean (Song et al., 2017;Zhang X. et al., 2018), chickpea Deokar et al., 2019), tomato (Illa-Berenguer et al., 2015), groundnut Luo et al., 2019;Zhao et al., 2020), have been effectively identified. This approach has also been deployed across in several crops due to its inherent ability to understand both qualitative and quantitative traits ( Table 2). For instance, Kumar et al. (2020) identified the role of two genes RING-H2 finger protein and zeaxanthin epoxidase in fresh seed dormancy in groundnut; both genes are known to control abscisic acid (ABA) accumulation. Very recently, Ramos et al. (2020) identified three QTLs (genomic regions) viz QtlPC-C04, QtlPC-C11 and QtlPC-C14 (linked to genes R1R2R3) associated with resistance to Phytophthora capsici Leonian which causes crown rot in squash (Cucurbita moschata). The most significant benefit of whole genome sequencing is that it allows the identification of causative mutations in target chromosomal regions. Additionally, this method identifies molecular markers which are located inside the harboring chromosomal region that can also be used to narrow down the genomic region which will help in mining the trait linked genes.

RNA-seq Based Trait Discovery
Advances in RNA sequencing (RNA-seq) technologies and approaches have made significant impact toward trait discovery, and have enabled plant developmental studies characterizing expression patterns of all the functional genes as well as regulatory RNAs (Nayak et al., 2019). RNA-seq is a more robust approach for precise measurement of transcripts and has been widely used for transcript profiling in several plant species (Cloonan et al., 2008;Wang et al., 2009). This data is vital for improving genome annotations, and offers precise gene position information for functional characterization and genome editing. The RNA-seq approach has been deployed for molecular characterization of several important agronomic traits such as seed size (Wan et al., 2017), seed coat color (Wan et al., 2018), seed coat cracking , seed and bud dormancy (Qi et al., 2015;Zhu et al., 2015;Khalil-Ur-Rehman et al., 2017), fatty acid biosynthesis and oil quality (Nayak et al., 2019), nutritional quality traits (Reddy and Ulaganathan, 2015), etc., which can offer precise gene information for developing designer crops for future. Also, RNA-seq have been used to understand the molecular mechanisms associated with salt tolerance in rice Lei et al., 2020); Chinese rye grass (Sun et al., 2013), and maize (Liang and Schnable, 2016). In wheat, RNA-seq study reported the drought responsive molecular pathways along with key candidate genes and molecular markers in the root tissue (Iquebal et al., 2019). RNA-seq has also been shown to be highly useful in combination with other -omics methods for gene discovery and pathway investigations.

Proteomics Enabled Genetic Trait Understanding
Knowledge of proteomics is being used to map the translated genes and loci controlling the expression of respective genes. It helps in identification of proteins responsible for bringing intricate phenotypic variations. High throughput proteomics has gone beyond the identification of individual proteins, to quantitative profiling, post translational modification studies, signaling, protein-protein interaction and many more areas. Photosynthesis plays major role in biomass production and yield. Change in protein profile studies was performed in chlorophyll deficient Brassica napus leaves which established the relationship between chlorophyll biosynthesis and photosynthesis . Xylem sap proteomics has revealed several insights related to cell wall formation (Zhang M. et al., 2014), leaf senescence  biotic and abiotic stress response González et al., 2012), cell to cell communication (Agrawal et al., 2010), and root-shoot communication (Krishnan et al., 2011). The enhanced level of redox proteins, stress and defense related proteins, calcium ion regulation proteins, signaling G-protein and RNA metabolism related proteins were induced in phloem sap study. Recently, proteomics study revealed that low light stress obstructs carbon fixation and OsGAPB overexpression augment tolerance to low light stress conceivably by increasing CO 2 assimilation and chlorophyll accumulation in rice . Interestingly, simultaneous upregulation of both biotic and abiotic stress responsive protein has been observed during bacterial blight infection in rice, which indicate the activation of common pathway (Kumar et al., 2015). Whereas in case of rice-M. oryzae interaction PBZ1, OsPR-10, SalT, Glu1, Glu2, and TLP proteins were up-regulated (Kim et al., 2004). iTRAQ proteomics study of Oryza officinalis provided evidences that proteins related to biosynthesis of secondary metabolites and carbon metabolism were mostly enriched after   planthopper infestation (Zhang et al., 2019c). Several proteomics and transcriptomics study conducted on seed dormancy study revealed the important role of antioxidant mechanism, protein thiol and redox regulation along with hormonal signaling in rice, wheat and barley (Hu et al., 2015). Mass spectrometry (MS) based proteomics study demonstrated the cultivar specific induction of proteins in salt stress condition such as glutathione-based detoxification of ROS was highly induced in tolerant variety whereas proteins involved in iron uptakes were more expressed in salt sensitive variety in barley (Witzel et al., 2009). Similarly, the role of OsCYP2 in moderating the antioxidant enzymes was established in transgenic rice overexpressing cyclophilin during salt stress (Ruan et al., 2011). Seed proteomics of various developmental stages during different stresses have revealed the process involved in seed dormancy, seed germination, and seed development . Proteomics related to environmental changes and abiotic stress focused on water supply responsive proteins in wheat against drought, high temperature, low temperature, frost, salt and heavy metals have been carried out Han et al., 2013;Kosová et al., 2013;  Alvarez et al., 2014;Capriotti et al., 2014;Kang et al., 2015). These studies offered novel insights and better understanding of crop physiology and metabolism during various kinds of stresses.

Metabolomics Based Trait Understanding
Holistic metabolomics based to study trails in plants were started late, particularly this approach was started through the introduction of untargeted metabolome detection (Alonso et al., 2015). Several studies have been reported where a particular metabolic pathways have been mapped (Sharma et al., 2021). For instance, the substantial changes in the various phytohormones was investigated in poplar leaf (Novák et al., 2008), rice various aerial organs (Kojima et al., 2009), rosemary leaves et al. (Müller and Munné-Bosch, 2011), Arabidopsis developing seeds (Kanno et al., 2010), strawberry fruits , potato tuber (Peivastegan et al., 2019), wheat developing seeds (Matsuura et al., 2019), watermelon fruit (Kojima et al., 2021), etc. The targeted approach has been also applied to explore the carotenoid pathway (Fernandez-Orozco et al., 2013;Kim et al., 2016;Mibei et al., 2017;Yoo et al., 2017;Price et al., 2018;Di Lena et al., 2019), flavonoid pathways (Karimi et al., 2011;Torres et al., 2019), amino acids (Torres et al., 2019;Praveen et al., 2020), and fatty acids (Talebi et al., 2013;Vidigal et al., 2018). Such profiling studies has helped in improving several important traits in plants by targeting specific pathways. Almost 10 years back Liu et al. (2011) targeted fatty acids biosynthesis pathways for enhancing biofuel production. Very recently and fatty acid desaturase 2 was targeted in several crops such as canola (Okuzaki et al., 2018), peanut , rice (Abe et al., 2018), false flax (Morineau et al., 2017), and Soybean , for enhanced production of oleic acid (C18:1), respectively. Several un-targeted metabolomics approach has been utilized to target multiple class of metabolites (amines, sugars, organic acids, etc.) from a sample extracted from various tissues of the model and crop plants such as Arabidopsis, apple, groundnut, kiwi fruit, alpine bird's-foot-trefoil, strawberry, grapes, mango, maize, medicago, orange, pear, sunflower, soybean, tomato, rice, white lupin, etc. (Sharma et al., 2021). Now, the targeted and un-targeted metabolomics approach have been coupled with genomics data for carrying out metabolomics-based quantitative trait locus (mQTL) and metabolic genome-wide association studies (mGWAS) studies (Wen et al., 2015;Chen et al., 2016); which simultaneously identifies the genomic region, causal genes and key metabolites and associated metabolic pathways that govern particular trait in plants. Recently, Li K. et al. (2019) identified 65 primary metabolites viz 22 amino acids, 21 organic acids, 12 sugars, four amines and six miscellaneous metabolites in the leaf of teosinte (an ancestor of maize) and identifies advantageous genes present in the wild relative associated with grain yield and shape trait in maize. In tomato, for one of the important trait accumulation of secondary metabolite in fruit was analyzed, and reported several subset of mQTLs-including mQTLs associated with acylsugar, hydroxycinnamates, naringenin chalcone, and a range of glycoalkaloids . Likewise, there are several studies which identified key genomic regions, candidate genes and mQTLs related to important traits through mQTL and mGWAS based studies including some domesticated traits, this was extensively reviewed by Sharma et al. (2021).
Previously, a combined transcriptome, proteome and metabolomics approach was used to investigate the ripening process with a final aim of extending tomato fruit shelf life (Osorio et al., 2011). This study showed a strong relationship between metabolites and their associated transcripts controlling ripening such as sugars, organic acids, and cell wall metabolism pathways. Similar studies have been done for banana which led to identification of genes including ERF1B, fructose-1,6bisphosphatase and polygalacturonase as key regulators of pulp ripening (Li T. et al., 2019). Recently, a combined transcriptome and metabolome study was deployed to study the molecular aspects of resistance and the interaction between Trichoderma harzianum strain T22 with tomato during defense responses against aphids (Coppola et al., 2019). This study demonstrated the importance of plant transcription factor families such as ZIP, MYB, NAC, AP2-ERF, and WRKY in biotic stress resistance. These examples show the potential of the -omics studies, working in tandem to characterize complex molecular interactions. These data have been used for the development of several gene expression/proteome/metabolome atlases to facilitate omics-driven crop improvement (Table 3).

MOLECULAR REGULATIONS OF DOMESTICATION RELATED TRAITS: SELECTED EXAMPLES
Over the past two decades the molecular regulation and the associated metabolic pathways of several agronomic traits has been revealed because of intensive research and the deployment of omics tools (Table 4). For the major domesticated traits their associated genes pathways have been linked with metabolic networks; however, more focused research is required to understand their specific role in particular metabolic pathways.
Here, we review progress in omics-based investigations of several important domestications related traits.

Transcriptional Control for Loss of Seed Shattering Trait in Cereal
From an evolutionary viewpoint, natural selection allows wild plant species to have specific functions to disperse seeds and fruits. Although from the agronomic viewpoint, natural seed dispersal is an undesirable trait in crops as it leads to significant seed loss in harvest. Consequently, natural seed dispersal was strongly chosen against by ancient humans to ensure productive cultivation during the domestication period (Purugganan and Fuller, 2009;Lenser and Theißen, 2013). The non-shattering traits were considered as the landmark of domestication in seed crops, as it makes the domesticated species mostly rely on human activity for propagation and enables the fixation of other domestication traits (Purugganan and Fuller, 2009). Seed crops have established their reduction of seed shattering ability independently and it is a convergent morphological adaptation to artificial selection (Purugganan and Fuller, 2009;Olsen and Wendel, 2013).
In cereal, seed shattering or fruit dehiscence is enacted through an abscission layer in the lemma-pedicel joint. Various transcription factors (TFs) coding genes were found in rice (Oryza sativa), which are involved in decreasing seed shattering. Shattering4 (Sh4) encodes the TF with Myb3 homology and is important for the formation of a functional abscission layer in the pedicle (Li et al., 2006). A single change of amino acid in DNA binding domain of Sh4 is intimately linked to the reduced seed shattering in domesticated rice. Also, the expression of the domesticated allele has been substantially reduced compared to the wild allele (Li et al., 2006). Thus, the combination of coding and regulatory alteration of Sh4 seems to affect the formation of the abscission layer, and consequently tries to weaken the shattering phenotype (Li et al., 2006). qSH1 is a major QTL on chromosome 1 involved in seed shattering in rice. The main gene, qSH1, codes a homeobox transcription factor-like BEL1 which is homologous to AtRPL (Konishi et al., 2006). A single nucleotide polymorphism (SNP) in the 5 -regulatory region effectively nullifies qSH1 expression in the preliminary abscission layer in the early development stage and contributes to non-shattering traits of rice (Konishi et al., 2006). Interestingly, the regulatory SNP in the homologs of RPL promoter are also amenable for distinct structures of seed dispersal based on natural selection of Brassica species with diminished replum development (Arnaud et al., 2011). These studies show a notable convergent mechanism where the same regulatory SNP could describe developmental variations in seed dispersal structures, which are important for both domestication and natural selection in distant species (Arnaud et al., 2011;Gasser and Simon, 2011). SH5 is another homeobox type BEL1 gene with a high qSH1 homology. SH5 has been expressed in the abscission layer (Yoon et al., 2014). Knockout of SH5 inhibits abscission layer formation and prevents seed shattering. Over-expression of SH5 leads to higher seed shattering, a consequence of decreased pedicel lignin levels (Yoon et al., 2014). The regulatory pathway of abscission layer formation has recently been expanded to include Shattering abortion 1 (SHAT1), an AP2 transcription factor encoding gene (Zhou et al., 2012). SHAT1 is needed for seed shattering by specifying abscission layer. Sh4 positively regulates the SHAT 1 expression in the abscission layer. qSH1 expression is lost in abscission layer in both the shat1 and sh4 mutant background, indicating qSH1 acts downstream of the shat1 and sh4 in the abscission layer establishment (Zhou et al., 2012). Intriguingly, qSH 1 is also needed in the abscission layer for expression of SH1 and Sh4. Thus the qSH 1 possibly takes part in a positive feedback loop of SH1 and Sh4 by establishing the SHAT1 and Sh4 expression in the abscission layer (Zhou et al., 2012). While SH5 and SHAT1 play a role in differentiating the abscission layer, the question remains whether both are artificially selected domestication genes. Like rice, decrease of seed shattering in domesticated sorghum is a result of loss of abscission in the joint that connects the seed hull with the pedicel. In sorghum, seed shattering is regulated by a single gene, Shattering1 (Sh1), which encodes a transcription factor YABBY. The nonshattering trait can be accounted for by any one of the three different loss-of-function mutations selected independently during sorghum domestication process . The notable mutations in Sh1 orthologs in rice and maize may be related to the shattering decrease in these crops . Whether Sh1 has been rewired into an SH5-directed seed shattering network in rice remains to be investigated in the future. In a wild relative of sorghum (Sorghum propinquum), seed shattering is conferred by the SpWRKY gene. It is believed that SpWRKY controls cell wall biosynthesis genes negatively in the abscission layer. Even so, SpWRKY was not crafted by artificial selection to contribute to the non-shattering characteristic for domesticated sorghum (Tang et al., 2013). These above studies together have raised a fascinating potential that the convergent domestication of nonshattering crops may have achieved the same underlying genetic goals by parallel selection (Lenser and Theißen, 2013). In domesticated wheat (Triticum aestivum) free-threshing trait (loss of spike shattering tendency) is conferred by important Q gene (Simons et al., 2006). Q-gene encodes the AP2-family transcription factor. The domesticated Q allele is abundantly transcribed than the wild q allele. Besides, both alleles differ in single amino acid, which significantly improves the homodimerization ability of the cultivated allele (Simons et al., 2006). Similar to Sh4, the development of the free-threshing character in cultivated wheat might also have been due to the combination of the coding and regulatory changes in the cultivated gene. The difference of expression between Q and q seems more significant as it can clarify the free threshing character in cultivated wheat (Simons et al., 2006;Zhang et al., 2011). Even though mutation which gives rise to Q has a significant effect on the process of wheat domestication, as it helps farmers to harvest the grain more effectively, the exact cellular cause contributing to free-threshing character is still unclear. Similar research has been progressed in non-cereals crop such as overexpression AtFUL to make the pods shattering resistance in Brassica juncea (Østergaard et al., 2006).

Cross-Talk Between Phytohormones and Related Genes Regulating Seed Shattering and Dehiscence Zones (DZ)
Hormonal homeostasis and interactions have been found recently as direct downstream effects of the core genetic network. As an example indehiscent (IND) expression is involved in the formation of local auxin minimum at the margin of the valve by regulating the auxin efflux in the separation layer cells (Sorefan et al., 2009). Further findings reveal that another b-HLH class SPATULA (SPT) transcription factor, required for carpel fusion early in the female reproductive organ development, may interact physically with IND (Girin et al., 2011). Auxins and cytokinins play an antagonistic role in plant growth and development (Bishopp et al., 2011). This scenario also indicates that the cytokinin signaling pathway is active at the valve margins and such a signaling pathway is interrupted in the shp1/2 and ind mutant. Consequently, local application of cytokinins in the fruit development can help to restore valve margin formation and further enhance dehiscence in shp1/2 and ind mutants, suggesting that cytokinins play a crucial role in valve margin differentiation (Marsch-Martínez et al., 2012). Recent studies reveal gibberellins (GAs) are also involved in the establishment of separation layer cell identity, in addition to auxins and cytokinins (Arnaud et al., 2010). As per the "relief of restraint" model, GA-mediated degradation of DELLA protein is important for GA signaling and also necessary to trigger expression of downstream genes (Harberd, 2003;Sun and Gubler, 2004). GA3ox1, which facilitates the final step in bioactive GAs synthesis, is shown as the direct target of IND. ALC interacts physically with DELLA repressors and local GAs production destabilizes the DELLA protein and relieves ALC to play its role in SL cell specification (Arnaud et al., 2010). In summary, these findings show that many phytohormones participate in the DZ specification and indicate that precise balance between biosynthesis and response is important. Notwithstanding the studies where the function of hormones in the development of DZ have been elucidated, very few studies about how such hormonal signals are coordinated in DZ have been carried out. One of the key challenges is to unravel the complete context of the molecular mechanisms and interactions of plant hormones underlying DZ-specification. There are many ways for minimizing crop losses due to crop shattering ranging from conventional parental selection with minimum shattering to the screening of mutants and gene editing methods. By advancing the next-generation sequencing and the marker traits associations, many genes involved in pod dehiscence were found, and a series of mutations underlying shattering resistance in several crops and their wild relatives have been identified (Fuller and Allaby, 2009;Dong and Wang, 2015). Attempts have been made to improve shattering resistance in Brassica, which include interfering in the dehiscence process by manipulating the molecular and hormonal control pathways (Fuller and Allaby, 2009;Altpeter et al., 2016) and developing transgenic lines with pod-shattering resistance (Liljegren et al., 2000(Liljegren et al., , 2004. In future, studies should focus, alongside gene-editing methods, on fine-tuning of the degree of shatter-resistance with RNA interference or the use of mutated forms of genes related to shattering in various crops.

Key Genes Targeted for Dwarfing of Cereal to Enhance the Productivity
The plant architecture is genetically controlled by a set of genes which subsequent affect yield and productivity of crop plant species. Often, mutation or knockdown of a single gene could also lead to significant change in the overall plant growth and development, subsequently plant architecture (Spielmeyer et al., 2002). In 1960s, the agricultural transformation that increased the production of rice and wheat was via the introduction of cultivars with a genetic predisposition to a short stature due to restricted elongation of stem (Silverstone and Sun, 2000). This phenotype enabled a significant partitioning of photosynthate produced from photosynthesis to sink organs like grains (Sun and Frelich, 2011).
Currently introduction of dwarfing genes is the most important aspect deployed in modern cereal breeding. The stems of tall wheat and rice crops are not strong enough to sustain heavy grains of the high yielding cultivars, which result in significant yield losses. In addition, the proportion of assimilates partitioned in grain increases yields. Genes associated with the semi-dwarf growth of the wheat and rice cultivars have been studied. In wheat, Reduced height (Rht) gene has been identified which is shown to interfere with GA signaling transduction pathway (Peng et al., 1999). Subsequently, three research groups investigated semi dwarf1 (SD1) gene from rice and found that the same hormone impair the biosynthesis (Monna et al., 2002;Sasaki et al., 2002;Spielmeyer et al., 2002). Thus, gibberellin hormone appears to be central to plant stature control.

Wheat Rht Gene and Gibberellin Signaling
The Green Revolution's wheat dwarfing genes originated in Japan (Gale et al., 1985). The Norin 10 dwarfing genes are now available worldwide in 70% of current commercial wheat cultivars. Norin10 contains two dwarfing genes that are semidominant homologous alleles on Chromosomes B and D. These alleles are labeled as Rht-B1b (formerly Rht1) and Rht-D1b (Rht2) to reflect their chromosome position (Boerner et al., 1996). The Rht alleles cause a reduced response to the plant hormone GA class (Gale et al., 1985). These plant hormones are diterpenoid carboxylic acids, that are involved in several processes of development in higher plants, including stem elongation (Hooley, 1994). The Rht gene is an ortholog of Arabidopsis GA-Insensitive (GAI) and maize dwarf 8 genes, for which mutations result in GA-insensitive dwarfs (Peng et al., 1999). Rht-1a/d8/GAI (wild type protein) is a subgroup of the GRAS family of proteins that are thought to act as transcriptional regulators (Pysh et al., 1999). Peng et al. (1999) reported base substitutions in the Rht-B1b and Rht-D1b alleles that insert stop codons within the DELLA region. They mentioned that translational re-initiation at one of several methionines which follow the stop codon could lead to the formation of truncated Rht protein without the DELLA domain, which functions as a constituent (GA insensitive) growth repressor. The D8 (Peng et al., 1999) and GAI mutations (Peng et al., 1997) also lead to partial or complete deletion from one or both of the conserved domains. The Rht-1a/d8/GAI proteins thus function as negative GA signaling regulators and suppress GA function, provided N-terminal domains are present (Harberd et al., 1998;Dill et al., 2001). To support this concept, ectopic expression of GAI (Peng et al., 1999) in rice caused dwarfism and loss of function mutations in Rht-like genes in some cases produces an overgrowth phenotype (Ikeda et al., 2001;Chandler et al., 2002). Besides d8, Rht-1a orthologs were reported in rice (known as OsGAI or SLR1) (Ogawa et al., 2000;Ikeda et al., 2001) and barley (SLN1) . While cereals have a single case of Rht-1a/d8/GAI type proteins, Arabidopsis contains a gene family encoding RGA proteins and three RGA-like proteins (RGL1, -2, -3) in addition to GAI. The Arabidopsis homologues seem to overlap in their function in various GA-regulated developmental processes (Olszewski et al., 2002). It is unknown how a single protein in cereals crops is functionally equivalent to five proteins in Arabidopsis; such variation may indicate major functional redundancy in Arabidopsis or fundamental differences in GA signaling pathways between Arabidopsis and Gramineae members. Recently, some progress was made in understanding the function of Rht-like proteins and their GA repression. RGA (Dill et al., 2001), SLR1 , and SLN1  are found in the nucleus and thus rapidly degraded with GA presence, the DELLA domain needed for this process. Rht's upstream signal transduction pathway is still unknown, but GAinduced degradation is believed to involve ubiquitin-mediated proteolysis .

Rice sd1 Gene and Gibberellin Biosynthesis
Unlike Rht, the sd1 mutation of rice is recessive and normal height can be restored in mutants using GA application showing that they have been impaired in GA production . Three research groups independently isolated the sd1 gene and showed it encodes GA 20-oxidase (GA20ox), an enzyme involved in biosynthesis of GA (Monna et al., 2002;Sasaki et al., 2002;Spielmeyer et al., 2002). Two of these research groups have used positional cloning to detect a GA20ox open reading frame close to the sd1 locus on the long chromosome arm (Monna et al., 2002;Spielmeyer et al., 2002). They also reported mutations in corresponding genes from semi-dwarf varieties. The third group, which had inferred the gene's identity by the effect of GA content mutations, used PCR to amplify DNA fragments, corresponding to two GA20ox genes, one of which mapped to the sd1 loci Ashikari et al., 2002). Semi-dwarf rice cultivars with Dee-geo-woo-gen sd1 allele contain a 383-bp deletion in the GA20ox gene (known as OsGA20ox2), which incorporates stop codon that is likely to result in a highly truncated, inactive enzyme. Gibberellin 20-oxidases are 2-oxoglutarate-dependent dioxygenases catalyzing carbon-20 depletion in the penultimate stage in biosynthesis of GA (Hedden and Phillips, 2000). These oxidases are encoded by small gene families, members of which have partial functional redundancy due to overlapping (but different) expression profiles or because of movement of the intermediates synthesized by enzymes between tissues. Therefore, loss-of-function GA20ox mutants are relatively less GA-deficient and are semi-dwarfs, unlike significant GA-deficient plants, which are extremely dwarfed and sometimes sterile. Two GA20ox genes were defined in rice: OsGA20ox1 (Toyomasu et al., 1997) and OsGA20ox2. Remarkably, selection for semi-dwarfism in rice has consistently yielded mutations in OsGA20ox2 instead of OsGA20ox1 or another GA-biosynthesis gene (for example, GA 3-oxidase is also encoded by a multi-gene family). Mutations in other genes might have a severe developmental impact or have negative impact on yield, and thus have been not selected in breeding programs. Genetic and functional analyses of SLR1/RHT and SD1 genes in rice and wheat have enormously improved the understanding of GA biosynthesis and signals, resulting in a strong methodology for manipulating the plant height of major crops. Both cases illustrate the central role played by GAs in controlling developmental processes. Therefore, GA signaling pathways (biosynthesis and signal transduction) are key aspects for manipulation in pursuit of further crop yield improvements. The yields of existing cereal crops seem to be approaching their limit, and new interventions are required if population is not to outstrip the food supply. Targeted genetic engineering/modification using newly emerged genomics, genome-editing technologies may be part of the next Green Revolution.

Achieving Submergence Tolerance
The incidences of uncertain rain and flood have been increased due to continued climate change. Today, more than 30 percent of the rice-planting land is vulnerable to flooding resulting in crop loss. In 1960s, the development of semi-dwarf variety was one of greatest achievement which significantly addressed the issue of global hunger threat caused due to human population explosion. The suppression of GAs production in the stem reportedly made high yielding semi-dwarf rice varieties susceptible to one of the most important abiotic stress "water logging." These developed semi-dwarf rice varieties lacked submergence tolerance. The lower nodes of these varieties unable to produce enough gibberellins to trigger elongation of the internode.

Genomics Based Discovery of Genomic Regions Associated With Submergence Tolerance
Submergence stress causes several adverse impacts on a plant such as low light intensity, hypoxia, nutrient effusion, physical injury, susceptibility to pathogen and pests attacks (Angaji et al., 2010). Several QTL mapping studies reported number of QTLs controlling submergence tolerance (Xu and Mackill, 1996;Nandi et al., 1997;Toojinda et al., 2003). A major QTL (Sub1) for submergence tolerance has been identified on chromosome 9 with LOD 36 and 69% of phenotypic variance explained (PVE) (Xu and Mackill, 1996). Sequencing of Sub1 genomic region identified three genes which encodes a ERFs (Sub1A, Sub1B, and Sub1C) in which Sub1A has been reported as a key component of submergence tolerance . Further cloning and characterization of Sub1 QTL helping in the detection of responsible genes and also help to discover tightly linked gene-based markers for molecular breeding program Toojinda et al., 2005;Neeraja et al., 2007). Furthermore, in other studies major QTLs namely qAG9-2 on L.G. 9 and qAG7-1 on L.G. 7 were reported (Angaji et al., 2010;Septiningsih et al., 2013). Later on, qAG9-2 QTL has been fine mapped and found a candidate gene OsTPP7 which encodes a trehalose-6-phosphate phosphatase which is responsible to regulate anaerobic generation (Kretzschmar et al., 2015). Both Sub1 and qAG9-2 major QTLs are widely used in rice breeding programs to improve submergence tolerance at germination and vegetative stages. Utilizing genomics resources several breeding efforts are also made in developing submergence tolerance varieties to sustain rice production. Various landraces and traditional genotypes namely, Kurkaruppan, FR13A, Thavalu, Goda Heenati, etc., were reported to be a suitable source of alleles which is associated with submergence tolerance (Miro and Ismail, 2013).

Precise Characterization of Genes Governing Submergence Tolerance
In recent years significant progressed have been made toward understanding the physiological, biochemical and genetic basis of submergence tolerance, to identify the causal gene(s) that are crucial for submergence tolerance (Oladosu et al., 2020). Recently, Kuroha et al. (2018) identified the gene SD1 (SEMIDWARF) responsible for submergence-induced elongation of internode by producing gibberellins mainly GA4. Another study identified genes SNORKEL 1 (SK1) and SK2 which encodes for ERFs, appeared to trigger submergence tolerance via ethylene signaling (Hattori et al., 2009). Both gene products further facilitate the internode elongation through GAs. Previous study identified a submergence tolerance gene SUB1A (an Ethylene-response-factor-like gene) on chromosome 9 which encodes ERFs Fukao et al., 2006). During flash floods, SUB1A inhibits plant elongation at the seedling stage.
Flash floods usually last for a few weeks. Cultivars carrying SUB1A tolerance gene show stunted growth and can survive in submerged conditions for a few weeks. Both SNORKEL 1 and SNORKEL 2 (SK1/2) genes and SUB1A encode ERFs which are associated with GAs, but they act in opposite ways in controlling plant development in response to submergence. Further more research is required to uncover the various pathways associated with SK1; SK2 and SUB1A. Furthermore, recently two genes have been identified ACCELERATOR OF INTERNODE ELONGATION 1 (ACE1) and DECELERATOR OF INTERNODE ELONGATION 1 (DEC1) which are responsible to control stem elongation (Nagai et al., 2020). ACE1 gene encoding an unknown function protein which is associated with internodes elongation via GAs, whereas, DEC1 gene encoding a zincfinger TF, which suppresses internodes elongation. Both the genes influence gibberellin-activated cell division in stem nodes. The expression of ACE1 gene during submergence conditions in rice triggers elongation of internodes within a cell-division zone of the plant. This results in an increased number of elongated internodes and increased plant height. Further gene ACE1C9285 is controlled by SUB1C, a gibberellin-activated TF which is upregulated in response to submergence (Fukao and Bailey-Serres, 2008). SUB1C expression level seemingly low in cultivars that contain the SUB1A-1 regulator gene, a homolog to SUB1C. In short rice cultivars expressing gene SUB1A-1, GAs responsiveness altered, subsequently use carbon pool for leaves elongation, and restrict overall plant development and enter to transient quiescent stage during flooding, an adaptation to overcome deep floods Xu et al., 2006). In semi-dwarf cultivars, internodes elongation only takes place in the upper internodes during growth stage. Nagai et al. (2020) reported a gene ACE1-LIKE1, which triggers upper internodes growth in deep-water. Presently, these omics study based information on the genetic basis of submergence tolerance is the base of rapid improvement of plant architecture to design a high yielding crop tolerant submergence.

TRANSLATION OF OMICs DRIVEN DATA FOR RE-DOMESTICATION AND DE NOVO DOMESTICATION: UTILIZATION OF GENOME/GENE EDITING TOOL
Gene-editing technologies have become choice of a researcher to domesticate neglected crops and wild relatives in a short period (Fernie and Yan, 2019). Traditionally, plant domestication and the development of productive cultivars required decades of breeding, which is also the key reason why so many breeding programs over the last 100 years focused on further improvement of a relatively small number of crops. Recent identification of several major domestication genes and scientific breakthroughs in integrating various genomic changes in plants concurrently with CRISPR/Cas9 editing has allowed re-domestication of existing crop plants and de-novo domestication wild species to be domesticated within a single generation (Figure 2) (Schindele et al., 2020). De-novo domestication has contributed to agrobiodiversity and diet quality, with possible future environmental and nutritional benefits (Singh et al., 2019). In the history of crop domestication amid higher yield selection and breeding, international germplasm exchange; multiple local resistance and resilience genes of wild species have been lost or have never been completely incorporated into breeding lines (Fernie and Yan, 2019). In other words, wild relatives of domesticated plants have significantly higher variable gene pool than that of domesticated ones . As we start to uncover more about the framework of crop genomes and the loci of quality traits, there are chances of incorporating valuable characters into existing crop species and ways to quickly redomesticate new crops. This step can be effectively achieved using breakthrough CRISPR-Cas9 gene-editing technologies, in particular, to introduce beneficial alleles without linkage Auxin biosynthetic and signaling genes affecting plant growth and reproductive organ development Zhou et al., 2018 drag , to produce novel quantitative variations (Rodríguez-Leal et al., 2017), direct deletion of deleterious alleles (Johnsson et al., 2019), and/or higher recombination rates (Mieulet et al., 2018). Recently, gene editing has been shown to enhance plant architecture, flower development, and fruit size in Physalis pruinosa (Lemmon et al., 2018). Gene editing is a promising method to generate diversity and to compensate for the genetic hitchhiking effects in germplasm. For reference, associated selection of traits such as fruit weight and disease resistance altered the tomato metabolome, providing an opportunity for precise breeding to alter nutritional and flavor traits . These hitchhiking effects and others, such as those found in rice and maize, represent promising goals for genetic modification to fettle linkage drag (Palaisa et al., 2004). For instance, African rice landrace Kabre possess superior resistance to pests and tolerance to drought; however, during domestication the plant architecture compromised affecting their overall yield potential. To address this Lacchini et al. (2020) targeted multiples genes which control plant architecture (HTD1) and control seed size and/or yield (GS3, GW2, and GN1A) by generating knockouts through multiplex CRISPR/Cas9. In knockouts, mutation in HTD1 gene caused reduced plant high to diminish lodging and improved tillering, whereas mutations in GS3, GW2, and GN1A resulted increased panicle and length along with improved seed girth.  (Shan et al., 2013;Xie and Yang, 2013;Shan et al., 2014;Xu et al., 2014;Zhang H. et al., 2014;Zhou et al., 2014;Woo et al., 2015;Meng et al., 2017;Nieves-Cordones et al., 2017;Mao et al., 2018;Ma et al., 2018). Likewise, in wheat EDR1, TaMLOA1, TaMLOB1, and TaMLOD1 targeted for resistance to powdery mildew, and GW2 and TaGW2 targeted for increased grain size, weight and protein content (Shan et al., 2014;Wang et al., 2014;Gil-Humanes et al., 2017;Kim et al., 2018;Wang et al., 2018). In orphan crops cassava and flax herbicide resistance has been introduced by targeting a gene EPSPS (Sauer et al., 2016;Hummel et al., 2018); whereas ALS was targeted in soybean (Cai et al., 2015). Similarly, many traits have been introduced or improved by targeting various genes in some economically important crops plants such as maize, tomato, potato, grapes, orange, cucumber, tea, etc. (Adhikari and Poudel, 2020;Bhatta and Malla, 2020).
The wild ancestral species of crop plants such as Solanum pimpinellifolium for tomato; Solanum demissum and S. stoloniferum of potato; Fragaria vesca of strawberry; Teosinte and Tripsacum of maize; Triticum dicoccoides, and T. turgidum L. ssp. Durum of wheat; Oryza rufipogon and O. longistaminata of rice; Manihot glaziovii and M. neosana and Glycine soja of soybean have been used for introgression key agronomic important traits into cultivars though breeding program (Zsögön et al., 2017). Moreover, most of the domesticated related traits and associated genes well characterized and has been linked with the metabolic pathway(s), and/or hormone biosynthesis and signaling (Table 4); therefore, integrated omics approach  Frontiers in Genetics | www.frontiersin.org which also involved metabolomics study has provided insights into the molecular basis of trait domestication. One can target these domesticated genes in wild ancestral plants for their speedy domestication. Now through CRISPR-Cas9 method these wild relative can be directly used for re-domestication or de-novo domestication (Figure 3 and Tables 5, 6). One of the important case study of de novo domestication in tomato has been done by Zsögön et al. (2018) by targeting important domestication related genes through CRISPR-Cas9 in tomato wild ancestral species S. pimpinellifolium. Zsögön et al. (2018) targeted SELFPRUNING (SP, control general plant growth habit), OVATE (O, regulate fruit shape); FASCIATED (FAS), FRUIT WEIGHT 2.2 and CLAVATA3 (CLV3) (control fruit size and locule numbers), MULTIFLORA (MULT, regulate fruit number), and LYCOPENE BETA CYCLASE (CycB). The engineered S. pimpinellifolium lines and achieved remarkable change in the plant overall phenotype with important traits essential for the commercial purpose such as increased lycopene content, enhanced fruit shape and determinant growth of plant; moreover, this was achieved in just single generation. Another study involved editing of multiples genes SP, SP5G (control day-length insensitivity), CLV3, WUSCHEL (WUS) and GDP-Lgalactose phosphorylase 1 (GGP1, control biosynthesis of ascorbic acid) in S. pimpinellifolium . This study clearly showed how selective editing of domesticated related genes can completely alter the plant architecture and improves the nutritional quality of fruits and makes convert wild relative into domesticated crop with retained biotic and abiotic stress tolerance properties . Very recently, in the wild strawberry (Fragaria vesca) few attempts has been made to demonstrate the procedure of the re-domestication or de novo domestication (Zhou et al., 2018;Feng et al., 2019). These attempts involved editing of genes tryptophan aminotransferase of Arabidopsis 1 (TAA1, converts tryptophan to indole-3pyruvic acid), Auxin response factor 8 (ARF8, repressor of auxin signaling) and YUCCA10 (YUC10, family of flavin-containing monooxygenases convert IPyA to IAA), key auxin biosynthetic and signaling pathways genes. Rice has five allotetraploids (BBCC, CCDD, HHJJ, HHKK, and KKLL) wild species which are also valuable genetic resources for improving of elite rice varieties. Among them the CCDD (species from South America genome) possess much stronger biotic and abiotic resistance and larger biomass compared to the cultivated diploid rice.
Recently Yu et al. (2021) demonstrated de novo domestication of wild allotetraploid rice PPR1 (O. alta; CCDD type genome) by improving six agronomically important traits viz nutrition use efficiency, abiotic stress tolerance, grain yield and quality, heading date, biotic stress resistance and sterility by genome editing targeting multiple genes including OaSD1-CC, OaSD1-DD, OaAn-1-CC, and OaAn-1-DD by CRISPR/Cas9 method. This suggests that CRISPR/Cas is a promising approach tool for the domestication of crops (Crews and Cattani, 2018), and is highly important for characters of defined selective sweeps in related species. These achievements were possible due to precise prediction of causal genes and metabolic pathways achieved by interpretation of data generated through genomics, transcriptomics, metabolomics, etc.

CONCLUSION
Omics have helped plant biologists to dissect important developmental clues and gene characterization. Presently, multidimensional omics approach where the biological sample can be analyzed for transcriptomics, proteomics and metabolomics in parallel, etc; offers plant biologists a complete understanding of plant metabolism by revisiting the metabolic pathways or identification of newer pathways. In the past 20 years, plant biologists have gathered significant amount of data relevant to genomes, transcriptome, proteome, and metabolome. Recent attempts are on development of gene-expression and proteome atlas. Altogether, this would strengthen the knowledge of the metabolic pathways, which have played crucial role during domestication of crop as well as trait improvement. Now, this knowledge has been translated to develop designer crops with desired traits by editing metabolic pathways of wild ancestral species (rich resource of genetic variations) called as de novocrop domestication. Domestication of wild or semi domesticated crop (tolerant to stress responses) would be feasible by multi step process were few important traits need to be improved first using genome editing; later the homologous lines can be selected for next level of trait modification. Such approach would be able to deliver a commercial line in 5 to 10 years. The CRISPR/Cas technique need to be explored in full extent by targeting several traits such as bio-fortification of nutrition's; because the current growing population also demand nutritional security. To achieve this, analysis of resequencing data available for the several crops is important; including GWAS which can identify high quality SNPs and haplotypes associated with target trait. Therefore, we expected in next 20 years' omics technology driven de-novo crop domestication will play very important role in the field of plant biotechnology.

AUTHOR CONTRIBUTIONS
RK received the invitation and conceived the plan for the manuscript. RK and VS wrote the manuscript. AK, SS, DR, SK, KP, BH, AV, RK, MP, ST, and GN improved the section and developed the table and figures. ST helped in developing the revised version. All the authors have read the manuscript before submission.