Genetic Modification for Improving Seed Vigor Is Transitioning from Model Plants to Crop Plants

Although seed vigor is a complex physiological trait controlled by quantitative trait loci, technological advances in the laboratory are being translated into applications for enhancing seed vigor in crop plants. In this article, we summarize and discuss pioneering work in the genetic modification of seed vigor, especially through the over-expression of protein L-isoaspartyl methyltransferase (PIMT, EC 2.1.1.77) in seeds. The impressive success in improving rice seed vigor through the over-expression of PIMT provides a valuable reference for engineering high-vigor seeds for crop production. In recent decades, numerous genes/proteins associated with seed vigor have been identified. It is hoped that such potential candidates may be used in the development of genetically edited crops for a high and stable yield potential in crop production. This possibility is very valuable in the context of a changing climate and increasing world population.


INTRODUCTION
Seed vigor is a complex physiological trait that is necessary to ensure the rapid and uniform emergence of plants in the field (Ventura et al., 2012), essentially including the seed longevity, the tolerance of environmental stresses by germination, and the ability to withstand prolonged storage and CDT. This trait is controlled by many QTLs that are located on different chromosomes, as found in the model plant Arabidopsis (Clerkx et al., 2004) and in crop plants such as rice (Cui et al., 2002;Miura et al., 2002), Medicago truncatula (Vandecasteele et al., 2011), and maize (Han et al., 2014) and is also affected by environmental factors during seed development, harvest, and storage.
Orthodox seeds, such as cereal seeds, undergo desiccation at the end of the maturation process on the mother plant and maintain their vigor over prolonged time periods (Rajjou et al., 2012). Because of their desiccation tolerance during dry storage, orthodox seeds are most commonly used in agriculture. For example, only three crop species (wheat, rice, and maize) account for more than 50% of all calories consumed by of the global population . In addition to economic and ecological importance, high-vigor seeds are necessary for seedling establishment and sustainable crop productivity, especially under unfavorable conditions (Rajjou et al., 2012;Ventura et al., 2012). High-vigor seeds can improve seed germination and seedling emergence, increase crop yield and reduce the cost of agriculture production. With the widespread application of modern mechanized precision sowing technology for grain (e.g., maize, wheat) production, highvigor seeds have become particularly important. In addition, for seed germplasms conserved in gene banks around the world, seed vigor and longevity may affect the regeneration cycle of accessions stored in seed banks. Seeds in long-term storage, especially under high-temperature and high-moisture conditions, will eventually lose their viability (Ventura et al., 2012). High-vigor seeds can survive a prolonged storage time.
However, the seed vigor trait is often excluded from traditional breeding programs, which are mostly directed toward high yield. To increase the vigor of commercial seed lots, the seed industry practices various invigoration treatments, especially physical priming methods (review in Ventura et al., 2012;Araújo et al., 2016). In fact, the potential of GM technology for enhancing seed vigor has been proposed as the most effective, economical and sustainable approach (Clerkx et al., 2004;Vandecasteele et al., 2011;Han et al., 2014). In the context of classical breeding, the application of GM technology to agriculturally important crops will play an increasingly important role in solving some fundamental challenges that face agriculture, natural resources and the environment. In this article, we summarize and discuss some pioneering work in the GM of seed vigor, especially through the approach of genetic engineering the PIMT (EC 2.1.1.77) in seeds. GM for improving seed vigor is just transitioning from model plants to crop plants. Promisingly, numerous genes/proteins associated with seed vigor, identified over the decades, may be used for the creation of genetically edited crops for a high and stable yield potential in crop production. This possibility is very valuable in the context of a changing climate and increasing world population.

PHYSIOLOGICAL, BIOCHEMICAL, AND GENETIC BASES OF SEED VIGOR
Seed vigor is a complex physiological trait involving regulatory networks that integrate genetic programs, metabolic signals, and hormonal signaling pathways (Rajjou et al., 2012). Many QTLs located on different chromosomes have documented associations with seed vigor. The various candidate genes identified within these QTLs are mainly involved in the glycolytic pathway, protein metabolism and signal transduction (Cui et al., 2002;Miura et al., 2002;Vandecasteele et al., 2011;Han et al., 2014). Seed vigor also has a close relationship with seed maturity degree, harvest time, and storage period: it has a maximum at physiological maturity and then decreases during storage (Sun et al., 2007). Carbohydrates, proteins, and mRNAs stored during seed development on the mother plant assist with hormone signaling pathways, especially the ABA signaling pathway, to regulate seed germination and influence seed vigor (Rajjou et al., 2012). ABA participates in regulating the expression of some seed genes in the mother plant during seed dehydration, such as LEA proteins, and inhibits the germination of developing seeds (Williams and Tsang, 1991). During storage, the seed will always deteriorate through a series of changes, such as the accumulation of ROS, lipid peroxidation, loss of cellular membrane integrity, enzyme inactivation, weak energy metabolism, and DNA degradation (Kibinza et al., 2006;Parkhey et al., 2012;Ventura et al., 2012;Xin et al., 2014;Yin et al., 2014;Kong et al., 2015;Ratajczak et al., 2015).
The loss of seed vigor is a complex normal biological phenomenon. In research on the mechanism of seed vigor change, CDT is the main way to simulate the seed aging process because aging naturally is time-consuming. Proteomics analysis displays a similar proteome characterization between artificial and natural aged Arabidopsis seed (Rajjou et al., 2008). However, a recent study reported substantial differences in scutellum nuclear content and morphology between the viability loss of accelerated and naturally aged wheat seed (Ahmed et al., 2016). Numerous studies have been performed on the process of seed deterioration in various plant species (e.g., Catusse et al., 2008Catusse et al., , 2011Galpaz and Reymond, 2010;Han et al., 2014;Nagel et al., 2014). However, the underlying mechanism remains unclear. It has become increasingly accepted that ROS damage to DNA (Vanderauwera et al., 2011), proteins (Rajjou et al., 2008) and membrane lipids (Roqueiro et al., 2010) plays a role in seed aging. ROS are continuously generated during seed development, storage and germination and exist in a state of dynamic equilibrium in cells under the action of free radical scavenger enzymes. Thus, the accumulation of ROS could be a common mechanism in seed deterioration. As a countermeasure, seed vigor has evolved a sophisticated mechanism (protection, detoxification, and repair) to protect macromolecules from ROS damage (review in Rajjou et al., 2012;Ventura et al., 2012). The possibility of restricting ROS accumulation may be a promising step toward successfully engineering seed vigor in crops.

POTENTIAL CANDIDATES OF GENES/PROTEINS ASSOCIATED WITH SEED VIGOR TRAIT
Under natural conditions, it is very rare to acquire high-vigor seeds through natural variation. Traditional breeding has made great progress in crop improvement; however, the process is time-consuming, and the genetic resources regarding seed vigor are limited. Promisingly, with the development of global omics approaches, such as genomics, transcriptomics and proteomics, numerous potential candidates (genes/proteins) involved in seed vigor have been identified with high efficiency in recent decades (Table 1), though few have been detected in the identified QTLs associated with seed vigor (Cui et al., 2002;Miura et al., 2002;Vandecasteele et al., 2011;Han et al., 2014). These potential candidates may be used in breeding programs and/or in biotechnological approaches to improve seed vigor and crop yields.

Repair Proteins
The formation of isoAsp, arising from both the deamidation of L-asparaginyl residues and the isomerization of L-aspartyl residues, is a frequent chemical modification that alters protein structure and leads to a loss of function (Lowenson and Clarke, 1992). The PIMT counteracts such damage by catalyzing the conversion of isoAsp to normal Asp in a variety of organisms, including plants (reviewed in Clarke, 2003). The PIMT-mediated protein repair mechanism represents a good example that has been successfully engineered for enhanced seed vigor (see below: case of PIMT, Table 2). For orthodox seeds, DNA damage, caused by ROS stress, occurs during seed dehydration and storage, leading to vigor loss. It is generally recognized that enhanced seed vigor and successful priming depend on DNA repair mechanisms activated during imbibition (Ventura et al., 2012). In Arabidopsis, the plant-specific DNA ligase VI (AtLIG6 and AtLIG4) is an important determinant of seed vigor and longevity under adverse germination conditions; atlig6 and atlig6::atlig4 mutants show significant hypersensitivity to CDT, displaying delayed germination and reduced seed vigor (Waterworth et al., 2010). A bifunctional DNA glycosylase/apurinic/apyrimidinic lyase, AtOGG1, is involved in base excision repair for eliminating 8-oxo-G from DNA, and the over-expression of AtOGG1 enhances seed longevity and abiotic stress tolerance . These DNA repair pathways represent potential targets for the generation of crops with improved seed vigor traits.

Protective Proteins
Protective molecules such as LEA proteins and HSPs are generally associated with desiccation tolerance and longevity and are accumulated in the maturation phase during seed development. These stress-related proteins may also play a role in seed vigor. Transgenic Arabidopsis seeds over-accumulating a HSF exhibit enhanced accumulation of HSPs and improved tolerance to aging (Prieto-Dapena et al., 2006). Knockout mutation in ATEM6 of the Arabidopsis group 1 LEA family resulted in a premature phenotype, demonstrating that ATEM6 protein is associated with water retention/loss during seed maturation; however, it might not be required in mature seeds for viability or efficient germination (Manfre et al., 2006(Manfre et al., , 2009. Dehydrins are LEA proteins that accumulate during seed maturation and in response to abiotic stresses The physiological role of AtPIMT1 in seed vigor and longevity has been established in Arabidopsis. The higher PIMT1 amount in pimt1-1 seeds correlates with lower isoAsp accumulation in vivo and increases both seed longevity and germination vigor, and vice versa. Germination % after 8 days storage (40 • C, 15-20% humidity): 52 and 25% for WT seeds; 80 and 50% for the pimt1-1 mutant seeds, monitored at 4 days after sowing.

Ogé et al., 2008
Cicer arietinum Seed-specific Over-expression of CaPIMT1 and CaPIMT2 in Arabidopsis The role of CaPIMT2 in seed vigor and longevity has been elucidated.
CaPIMT2 enhances seed vigor and longevity by repairing abnormal isoAsp in the seed nuclear proteome. Germination % after 4 days of CDT, control seeds, 10-14%; CaPIMT1 and CaPIMT2 transformed seeds, 80-90%. Verma et al., 2013 O. sativa Overexpressing OsPIMT1 lines and OsPIMT1 RNAi lines The role of OsPIMT1 in seed vigor and longevity has been elucidated. Germination % after 21 days of CDT, overexpressing OsPIMT1 transgenic seeds, increased 9-15%; OsPIMT1 RNAi lines, rapid loss of germination. Wei et al., 2015 Transgenic rice and Arabidopsis lines with altered expression of OsPIMT1 and OsPIMT2 The PIMT-mediated protein repair mechanism during seed development and aging in rice has been elucidated, i.e., OsPIMTs repairs antioxidative enzymes and proteins that restrict ROS accumulation, lipid peroxidation, and so on, thus contributing to seed vigor and longevity. Transgenic rice overexpressing OsPIMT1 and OsPIMT2 exhibits improved seed vigor and longevity. Germination % after 4 days of CDT, control seeds, 8% (maximum); OsPIMT1, OsPIMT2, and OsPIMT2 transformed seeds, 43-48%. Petla et al., 2016 in vegetative tissues. A twofold reduction in seed-specific dehydrin (LEA14, XERO1, and RAB18) by RNAi reduced seed longevity and viability in Arabidopsis (Hundertmark et al., 2011). Phospholipase D, which cleaves phospholipids and generates phosphatidic acid (PA), is involved in the early stages of seed deterioration. The accumulation of PA in seeds triggers damage at the level of cellular membranes and storage lipids. Depletion of the Arabidopsis PLDα1 gene, encoding a member of the lipid-hydrolyzing phospholipase D family, resulted in seeds with lower levels of lipid peroxides and increased tolerance to aging (Devaiah et al., 2007).

Detoxification Proteins
This class of proteins performs the degradation and/or elimination of endogenous and exogenous toxins, such as ROS.
In particular, to eliminate ROS, cells develop a number of ROS scavengers such as superoxide dismutase, peroxidase, and vitamins. Enhanced seed longevity has been reported through the elimination of ROS by over-accumulated ROS scavengers in transgenic seeds (e.g., Lee et al., 2010). Three genes (NnMT2a, NnMT2b, and NnMT3) from sacred lotus that encode metallothioneins, cysteine-rich small proteins involved in ROS scavenging, were highly expressed in germinating sacred lotus seeds and dramatically upregulated in response to high salinity and oxidative stresses (Zhou et al., 2012b). Moreover, transgenic Arabidopsis seeds overexpressing NnMT2a and NnMT3 displayed a remarkably improved resistance to accelerated aging treatment, indicating their significant roles in seed germination vigor (Zhou et al., 2012b).
The mitochondrial SSADH is one of the three enzymes involved in the GABA shunt. In plants, the role of the GABA shunt in protection against oxidative stress has been demonstrated (Bouché et al., 2003). The presence of SSADH in dry seeds suggests that the GABA shunt is involved in the control of seed longevity or/and germination. Mutations in the OsALDH7 gene resulted in seeds that were more sensitive to artificial aging conditions and accumulated more malondialdehyde than wild-type seeds, implying that this enzyme plays a role in maintaining seed viability by detoxifying the aldehydes generated by lipid peroxidation (Shin et al., 2009).

GENETIC MODIFIED SEEDS FOR ENHANCED VIGOR: CASE OF PIMT
In seeds, proteins are prone to aging damage during normal aging and CDT. To date, a successful approach to enhanced seed vigor involves enhancing the accumulation of PIMT in seeds. However, no specific proteins have been assigned to the identified QTLs associated with seed vigor. The history of this effort provides an excellent example of how scientific problem solving can be brought to bear on applications in agriculture. Clarke (1993, 1994) first discovered PIMT activity in plants and proposed that PIMT might be involved in seed survival by preventing isoAsp accumulation in the proteins of aging and stressed seeds. PIMT has since been detected in a wide range of plants and cloned in Arabidopsis, wheat, chickpea and rice, and the numbers are still increasing. In plants, PIMT is encoded by two different genes (PIMT1 and PIMT2) (Xu et al., 2004), which display distinct expression patterns but similar biochemical properties (Thapar et al., 2001). Later, Ogé et al. (2008) validated the role of this enzyme in both seed vigor and longevity by altering the expression of PIMT1 in Arabidopsis. Their findings implicate PIMT1 as a major endogenous factor that limits isoAsp accumulation in seed proteins, thereby improving seed traits such as longevity and vigor. Recently, the role of PIMT in seed vigor and longevity has been evaluated in chickpea (Cicer arietinum) (Verma et al., 2013) and rice (Oriza sativa) (Wei et al., 2015;Petla et al., 2016). Notably, transgenic rice constitutively overexpressing OsPIMT1 and OsPIMT2 exhibited improved seed vigor and longevity (Petla et al., 2016).
Although the seed vigor trait depends on a wide range of physical, chemical, molecular and QTLs, the PIMT repair pathway improves seed vigor in rice by restricting the formation of deleterious isoAsp and repairing damaged proteins, not through direct DNA or lipid protection (Petla et al., 2016). This finding implies the efficacy of making high-vigor rice seeds through a target-gene approach. However, it remains to be observed whether this approach can work in the field or whether other single-gene manipulations can also produce such effects. Moreover, the effect of enhanced PIMT expression on other seed traits, e.g., nutrient value, potential health risk as food and feed, and plant phenotypes, must be extensively evaluated. In addition, the exploitation of such PIMT-mediated improvement of seed vigor in other important crops could have a huge impact on the agricultural economy. The successful case of over-expressed PIMT enhancing seed vigor proves a good guide for other potential candidates.

CONCLUDING REMARKS AND PERSPECTIVE
Currently, achieving food supply security with limited arable land is a major global challenge due to the changing climate and increasing global population. The approach of modifying PIMT in seed tissues provides a rational means of creating high-vigor seeds for crop production. Its application to important cereals such as wheat, rice, and maize may have a dramatic impact on global food security. Despite substantial progress, many questions still remain. The possible effect of enhanced seed vigor obtained by the over-expression of PIMT and other proteins on the nutritional value of crops is unclear. It remains to be assessed whether a GM seed with enhanced vigor shares similar health and nutritional characteristics with its conventional counterpart. While numerous potential candidates (genes/proteins) associated with seed vigor are available, their roles in improving seed vigor must be validated by reverse genetics on large-scale samples before translation into application in agriculturally relevant crop species. The rapid development of new genomeediting techniques enables the precise modulation of traits of interest with unprecedented control and efficiency. Among the current genome-editing tools, CRISPR is easy, rapid and inexpensive, exhibiting a broad applicability of plant genome editing for the development of designer crops (review in Khatodia et al., 2016). However, it is important to remember that the safe use of GM food or feed requires an assessment of health risks and environmental effect (Araki and Ishii, 2015).
At present, there are no reports on the application of CRISPR in manipulating seed vigor in plants. Genome-editing techniques represent a promising tool for manipulating the accumulation of proteins associated with seed vigor in a seed-specific manner and should greatly reduce the time needed to obtain valuable crop varieties. Thus, the creation of such transgenic seeds and their subsequent application in agriculture is crucial for better feeding a rapidly growing population in a changing climate.
Seed quality is the basis of agricultural production. Highquality seeds are an unremitting pursuit for every seed producer. GM technology is an effective, economical and sustainable way to improve seed vigor, change seed color or shape, or boost nutrient components and other agronomic traits for crops. The application of GM technology will sharply change the face of agriculture.

AUTHOR CONTRIBUTIONS
WW and XH conceived the article. FN and XW collected references and analyzed the data. XW, FN, and WW revised the manuscript. All authors contributed in manuscript writing, and approved the final manuscript.

FUNDING
The National Natural Science Foundation of China (31371543 to WW); the Program for Innovative Research Team (in Science and Technology) in University of Henan Province (15IRTSTHN015 to WW).