Introduction

Soybean (Glycine max) is one of the most important sources of high-quality protein and edible vegetable oil for human consumption, livestock feed, and aquaculture (Masuda and Goldsmith 2009). Canadian production has followed the general trend of increasing demand in national and international markets (Ritchie and Roser 2021). Soybean acreage in Canada reached 2.15 million ha in 2021, with a total production of 6.27 million mt. Soybean ranks fourth among Canada's major crops and has become an important crop of Canadian prairies. In 2021, Manitoba produced approximately 1 million mt soybean from 532900 ha (Statistics Canada 2021). This westward expansion of soybean production can largely be attributed to the development of short-season cultivars that can mature before the autumn frosts (Voldeng et al. 1997; Morrison et al. 1999; Cober and Voldeng 2012). In Manitoba, the typical date of the last spring frost falls between 13 and 31 May and the first fall frost occurs between 11 and 26 September, resulting in a range of 75–135 frost-free days in a season (Manitoba Agriculture). Although duration of frost-free season has generally increased on the prairies with a trend towards earlier last spring frosts (Cutforth et al. 2004), temperatures can vary significantly from year to year making crops vulnerable to cold damage, particularly during the germination, seedling, and flowering stages (Fig. 1) (Hatfield and Egli 1974; Brown and Blackburn 1987; Wang et al. 1997; Lukatkin et al. 2012; Jeschke et al. 2020). To reduce the risk of yield loss, Manitoba Pulse and Soybean Growers (MPSG) guidelines recommend planting soybean in the second and third weeks of May with subsequent harvesting in September (Mohr 2018; MPSG 2019). The choice of seeding date is based on the minimum recommended soil temperature for soybean germination (i.e., 10 °C) and to reduce the risk from cold damage during imbibition, the process in which seeds absorb water for the initiation of germination (MPSG 2019). Variable spring weather can often significantly delay planting on the prairies resulting in potential yield losses. A Manitoba and North Dakota study showed a yield reduction of ∼16 kg ha⁻¹ for each 1-day delay in the seeding of soybean (Tkachuk 2017). MacMillan and Gulden (2020) reported a yield reduction of 15%–36% between the normal (24 May–12 June) and the late (6–24 June) planted soybeans in three out of eight Manitoba environments. Early planting however increases the risk of injury from low temperatures and excessive moisture in sensitive cultivars (Cober et al. 2013). The year-to-year variability in the early spring weather in Canada can potentially expose both early and timely planted soybeans to freezing or low temperatures (Fig. 1). Low temperature can slow down the rate of seed imbibition and germination (Tully et al. 1981; Nykiforuk and Johnson-Flanagan 1997; Booth and Bai 1999; Cheng et al. 2010; Yu et al. 2015) and damage the cell membranes (Bramlage et al. 1978; Yu et al. 2015). Excess water causes hypoxia and limits aerobic respiration (Sarlistyaningsih et al. 1995). Cold and wet seedbeds can result in poor plant stands in Western Canada and in areas with similar climates, which offsets any potential gains from early planting. For early planting to be economically beneficial, it is important to develop cultivars that thrive better under these conditions. The aims of this article are to briefly review soybean germination and the impact of cold and excess moisture stress on this process, followed by a description of the latest research methods being used to address the problem along with the latest discoveries, focusing on soybean. Finally, the future perspectives section will discuss the possible role of translational genomics in discovering future solutions.

Fig. 1.

Spring weather in soybean growing areas of Eastern and Western Canada shows significant year-to-year variability increasing a chance of both cold and frost injury in both early and timely planted soybeans. Monthly minimum temperature (A) and total precipitation (B) data are from weather stations located at Morden, MB (49°11″N, 98°05″W) and London, ON (43°02″N, 81°09″W). Historical weather data were downloaded from  https://climate.weather.gc.ca/historical_data/search_historic_data_e.html

Seed germination

Rapid and robust seed germination is crucial for good plant development, growth, and ultimately crop yield. During germination, the seed can be exposed to many factors that may encourage or inhibit seedling development and plant productivity. Germinability of seeds, ability of seed to overcome dormancy also known as seed vigor, is determined by both genetic and environmental factors (Finch-Savage and Bassel 2016) such as temperature (Torabi et al. 2013), water availability (Holdsworth et al. 2008a), oxygen (Magneschi and Perata 2009), and light quality and quantity (Flores et al. 2006). Those environmental factors regulate the synthesis and metabolism of hormones (Finch-Savage and Leubner‐Metzger 2006), helping vigorous seeds overcome dormancy. When the ratio of abscisic acid (ABA)/gibberellin (GA) is high, seeds remain dormant; lower ratios promote the germination process (Holdsworth et al. 2008b). Shuai et al. (2017) reported that dry seeds of soybeans have an ABA concentration of approximately 32 ngg⁻¹ and undetectable low amounts of GA; however, after 6 h of imbibition in water, levels of active GA₁ and GA₄ increase to 0.6 and 0.4 ngg⁻¹, respectively, while ABA concentration drops to 14 ngg⁻¹.

Seed germination encompasses physical and physiological processes that begin with imbibition by dry seeds and end with the emergence of the radicle (Bewley 1997), namely water absorption, cell elongation, and increased cell number. Depending on the rate of water absorption, the germination and immediate post-germination period is characterized by three distinct phases (Fig. 2) (Bradford 1986; Nonogaki et al. 2010; Rajjou et al. 2012) The first phase involves rapid water absorption by the cells of dry seeds (dead or alive), which is largely driven by matrix potential (Krishnan et al. 2004). Metabolically, this phase is accompanied by DNA damage repair (Macovei et al. 2011), glycolysis, and the restoration of the pentose phosphate oxidation pathway (Howell et al. 2006). This phase is also associated with leakage of some organic materials initially, which declines rapidly as the imbibition progresses (Simon and Harun 1972; Watson and Morris 1987; Bhattacharya 2022). Next is the plateau phase, when the rate of water absorption slows down, and the embryo axis elongates and breaks through the covering layer of the seed to complete germination. Seed dormancy, low or high temperatures, lack of water, or presence of ABA usually prolong the second phase of water absorption, and the factors that promote germination generally shorten this period. The second phase is considered to be a metabolically active period. During this phase of water absorption, repair and multiplication of mitochondria, translation of stored mRNA occurs (Dinkova et al. 2011) and mobilization of metabolite reserves begins. Mobilization of reserves is one of the most critical events in the germination process, providing both precursors and the energy source for biosynthetic processes. Although the mobilization of reserves may not be a prerequisite for germination (Pinfield‐Wells et al. 2005; Kelly et al. 2011), it is essential for improving the germination efficiency and seedling establishment (Eastmond et al. 2000). Once the radicle has penetrated the seed coat and starts growing, germination is complete and seedling growth begins (Bradford 1990). The third phase accompanies rapid water absorption and is the post-germination stage in which the radicle begins to grow (Manz et al. 2005).

Fig. 2.

Three phases based on the rate of water uptake during germination and post-germination period (Nonogaki et al. 2010; Iqbal et al. 2022). The first phase is critical to cold injury. The figure is modified from Nonogaki et al. (2010).

Effects of cold temperature stress during germination

Temperature affects seed germination during the following three physiological processes. First, during seed storage, the effect of excess moisture content and low temperature can cause deterioration, resulting in death (Rani et al. 2013). Second, high temperatures generally induce dormancy in seeds, but low temperatures can also induce dormancy in some cases (Roberts 1988). The optimal temperature for soybean seed germination is 25 °C (Khalil et al. 2010); however, the base temperature (below which the germination rate is zero) is 4 °C (Lamichhane et al. 2020). Soybean germination is usually studied under controlled constant temperature regimes, but alternating temperature usually results in more robust germination (Totterdell and Roberts 1980). Nleya et al. (2005) demonstrated that temperature was not the only factor controlling germination, and the length of time that the seeds were exposed to, a specific temperature is also important. Finally, the germination rate of seeds that have lost their dormancy demonstrated a positive linear relationship between the base temperature and the optimum temperature, and a negative linear relationship between the optimum temperature and the ceiling temperature (at and above which the rate is again zero) (Roberts 1988). Temperatures below or above the optimal temperature result in varying degree of reduction in seed germination (Bhattacharya 2022). Actual commercial production, often requires planting under suboptimal low-temperature soil conditions such as those experienced under the short-season growing conditions in Western Canada (Fig. 1).

The first phase of rapid water absorption is crucial for reorganization of membranes, which lose their structures in the dried seeds (Simon 1974, 1978). Both normal and freeze-induced dehydration causes lamellar-to-hexagonal-II (H_II) phase transition in the plasma membranes (Gordon-Kamm and Steponkus 1984; Steponkus and Lynch 1989) which contributes to demixing of plasma membrane components and cytoplasmic leakage upon imbibition (Bramlage et al. 1978; Duke et al. 1983). Under normal circumstances, membranes reorganize from an H_II phase to a lamellar phase soon after hydration to regain their proper function as a semipermeable barrier (Simon 1974; Steponkus 1984). The imbibition of cold water disrupts the reorganization of cells during rehydration and reduces germination rates (Obendorf and Hobbs 1970; Bramlage et al. 1978; Leopold 1980). The main effect of low temperatures during germination is the damage to the cell membranes and the changes in plasma membrane composition that help to retain or regain the bilaminar structure are crucial for a high rate of survival after low-temperature exposure (Steponkus 1984; Lynch and Steponkus 1987; Yoshida and Uemura 1989; Uemura and Steponkus 1999; Degenkolbe et al. 2012; Takahashi et al. 2016; Takahashi et al. 2018; Sanchez et al. 2019). Noblet et al. (2017) using lipidomics in maize (Zea mays) seeds showed that the ability to germinate under cold stress is associated with phospholipid remodeling. The cold acclimation process, which increases the degree of freezing tolerance in response to low, non-freezing temperatures, involves changes in the plasma membrane composition that confers tolerance and reduces leakage of cellular contents (Steponkus 1984; Posmyk et al. 2001; Uemura et al. 2006; Ruelland et al. 2009; Yu et al. 2015; Takahashi et al. 2018; Bhattacharya 2022). In addition to changes in membrane composition, several other mechanisms, such as the accumulation of sugars and dehydrins, also contribute to freeze tolerance as a result of acclimation (Yamada et al. 2002; Janska et al. 2010; Ruelland and Collin 2012; Takahashi et al. 2018). The interaction between temperature and genotype on germination percentage is also significant (Borowski and Michalek 2014), and some aspects of plasma membrane composition contribute to the inherent differences in cold-tolerant vs. cold-sensitive genotypes. For example, a recent report suggested that accumulation of phospholipase-D-mediated phosphatidic acid plays a role in chilling injury in sensitive soybean cultivars (Yu et al. 2015). The accumulation of saturated or poorly unsaturated lipids during seed imbibition was also shown to be associated with the seed sensitivity to cold temperatures in maize (Noblet et al. 2017).

Exposure to cold stress during germination causes the membranes to lose their selective permeability and release many cellular components, including lipids, proteins, and ions (Watson and Morris 1987; Bhattacharya 2022). The deeper cell layers are exposed to some secondary stresses, such as drought and microbial attack, due to the functional loss of the epidermis. Nutrients leaked from the damaged cells promote the growth of microorganisms, which subsequently causes further adverse effects on germination (Nijsse et al. 2004). Yin et al. (2009) reported that the ultrastructure of soybean axes organelles at low temperatures was significantly changed compared to that under the optimum temperature, resulting in the absence of mitochondrial cristae, cytoplasmic layer irregularities, and underdeveloped endoplasmic reticulum structures. Improvement in mitochondrial membrane function through osmopriming of soybean seeds leads to improved response to chilling injury (Sun et al. 2011). This demonstrates how low-temperature swelling induces a barrier to membrane repair. Therefore, the metabolic activity of seeds may be severely affected following low-temperature swelling. Normally, seeds have mechanisms that respond to limit these damages caused by imbibition (Doria et al. 2019). Physiological dysfunction caused by hypothermia can be restored if the tissue returns to a normal temperature before injury occurs. These disturbances do not result in visible damage or changes in growth and developmental rates because disturbances in physiological processes can be reversed before they become stable (Lyons et al. 1979); however, irreversible damage caused by prolonged hypothermia may result from the accumulation of toxic metabolites (Graham and Patterson 1982).

The most critical time for imbibition is within 24 h of planting (Pollock and Toole 1966). Therefore, imbibitional chilling effects are more severe when seeds are planted in colder compared to warmer soils followed by a drop in temperature. The initial moisture content of seeds at planting also affect the degree of injury from chilling, in which higher moisture content plays a protective role (Obendorf and Hobbs 1970). After emergence, soybeans are more susceptible to frost injury when compared to corn and wheat (Triticum aestivum) because the soybean growing points are above the ground. Cotyledons can provide some protection initially to the growing point immediately after emergence (Hume and Jackson 1981); however, as the primary leaves expand and the growing points are exposed, low temperatures at 0 °C or below become lethal. Cold injury can also make seedlings more susceptible to damping-off (Serrano and Robertson 2018; Jeschke et al. 2020). Unfortunately, environmental stresses rarely occur alone and this is also true for cold stress. In Canada, cold temperatures in the spring often occur in combination with excess moisture due to the condensation of atmospheric moisture and rains (Fig. 1). Cold directly exacerbates the damage from waterlogging (Hobbs and Obendorf 1972; Wuebker et al. 2001). For example, in terms of loss in germination, 1 h of flooding at 15 °C was equivalent to nearly 36 h of flooding at 25 °C. Later in the crop season, cold autumn temperatures also play a role in ripening and yield (Gass et al. 1996; Kurosaki and Yumoto 2003). The average seed yield in warmer regions is generally higher than in the cold regions (Zhao et al. 2020). Consequently, cold-tolerant cultivars reduce the economic risks in Canada (Balasubramanian et al. 2004). Previous reports indicate that genetic variation in soybean for cold tolerance can be assessed based on the time to seed germination and seedling emergence (Littlejohns and Tanner 1976) and the ability to form pods at low temperatures (Hume and Jackson 1981). The cold germinability of soybean seed is also a potential physiological index for judging the survival rate of seedlings at low temperatures (Robison et al. 2017).

Effects of waterlogging stress during germination

Water availability is critical in the seed germination process. Water is required for optimal enzyme activity, dissolution, and transportation of reactants and as a reactant in chemical processes. Lack of water will hinder the germination of seeds and even cause death during emergence. Excess moisture or waterlogging is a common abiotic stress encountered in soybean production in Canada, which significantly impacts crop yield. For germination, both the duration and timing of waterlogging are important. Waterlogging for just 1 h has the potential to result in a significant loss of germination capacity, and the loss increases with both the duration and time after imbibition when the stress is applied (Wuebker et al. 2001). Wu et al. (2017a) reported that under field conditions, flood stress for 1 day can reduce the germination rate by 50%. Excess water impedes aerobic respiration by oxygen deprivation and increases the likelihood of soil-borne diseases further reducing the germination rate (Tian et al. 2005). These factors may cause the overall breakdown of seed metabolism and result in the inability of the seeds to develop into normal seedlings. Water uptake in the seed depends on the water potential gradient between the seed and the external environment. The rate of water uptake by seeds is affected by temperature (Watson and Morris 1987; Cheng et al. 2010) and seed characteristics (Hou and Thseng 1992; Tian et al. 2005; Muramatsu et al. 2008; Rajendran et al. 2019; Sato et al. 2019). The seed coat plays a protective role since cultivars with black or brown seed coats show better waterlogging tolerance than cultivars with a yellow seed coat (Hou and Thseng 1992; Rajendran et al. 2019). Quantitative trait loci (QTL) mapping studies have shown a genetic link between seed coat pigments (Sayama et al. 2009; Otobe et al. 2015), surface roughness (Otobe et al. 2015), and water impermeability. Yellow seed coat cultivars do not accumulate anthocyanins (anthocyanin glycosides) or proanthocyanidins (polymeric anthocyanins) (Todd and Vodkin 1993). Therefore, seed color and pigment content are important factors affecting water absorption and the waterlogging resistance of seeds (Zhang et al. 2008), although it is not relevant to most commercial production since the seed coats of most cultivars are yellow. The thickness of the aleurone layer and responsiveness to low oxygen concentration has also been shown to play a role in seed germination tolerance to flooding stress (Tian et al. 2005; Sato et al. 2019). High surface-to-volume ratios in small-seeded cultivars reportedly increase access to oxygen and lead to high germination under flooding stress (Rajendran et al. 2019; Wiraguna 2022). Many factors have been implicated in different studies; however, it is still unclear how the combined effect of these seed structural and chemical features may contribute to the rate of water absorption in soybean seeds (Chachalis and Smith 2001; Shao et al. 2007; Muramatsu et al. 2008; Nakayama and Komatsu 2008) and possibly flooding tolerance during germination. One hypothesis suggests that damage from flooding may be caused by rapid water absorption in the seeds. For example, osmotically slowing water uptake by soaking seeds in 30% polyethylene glycol (osmopriminig) protects the seeds from damage caused by flooding (Woodstock and Taylorson 1981). Interestingly, osmopriming was also shown to strongly improve chilling resistance in a sensitive cultivar (Sun et al. 2011). Just as cold stress, both genetic and environmental factors determine the extent of damage caused by waterlogging. Temperatures stress can exacerbate the extent of damage from excess moisture. Seeds exposed to a germination temperature of 15 °C are more susceptible to flood stress than seeds exposed to a germination temperature of 25 °C (Wuebker et al. 2001); however, moderately low temperatures can have a protective role during flooding stress (Hou and Thseng 1991; Nguyen et al. 2021).

Cold signaling and mechanism of tolerance

Recent molecular genetic studies in both model and crop species have led to a greater understanding of the processes involved in cold perception, acclimation, and tolerance pathways (Ruelland et al. 2009; Chinnusamy et al. 2010; Survila et al. 2010; Aslam et al. 2022). Many components of the pathways have been identified (Fig. 3). Downstream, the cold signaling pathways activate the myriad responses that lead to tolerance. Glycine-rich RNA-binding proteins (GR-RBPs) have been implied to play roles in plant response to cold stress, and a GR-RBPs protein designated atRZ-1a was illustrated to enhance cold tolerance in Arabidopsis (Kim et al. 2005; Kim and Kang 2006). Dogras et al. (1977) reported that the cold resistance of seeds depends on their ability to synthesize large amounts of unsaturated fatty acids during the initial stages of germination; however, the difference in cold sensitivity between peas (Pisum sativum) and soybeans was not related to any difference in the composition of the major lipid components of the seed membranes (Priestley and Leopold 1980). Roskruge and Smith (1997) showed that cold damage causes lipid peroxidation in seeds resulting in the loss of fatty acid unsaturation, so the ability of seeds to control lipid peroxidation reflects their tolerance to cold damage. The species-specific mechanisms may necessitate parallel studies for cold tolerance in different crops. Discovery and evaluation of existing genetic variations are often the catalyst for such approaches which may include QTL mapping, RNAseq analysis, and proteomics.

Fig. 3.

Overview of cold signaling in plants. Cold signaling in plants begins with induction of Ca²⁺ transients and ROS (reactive oxygen species) due to exposure to cold stress (Carpaneto et al. 2007; Yuan et al. 2018). Intracellular changes in Ca²⁺ concentration and ROS are sensed by Ca²⁺ sensors such as Ca²⁺-dependent kinases (CPKs), Ca²⁺-responsive protein calmodulins proteins (CaMs), Calcineurin-B Like (CBL) proteins and a group of serine/threonine protein kinases, named CBL-interacting protein kinases (CIPKs) triggering the Ca²⁺ signaling cascade (Xiang et al. 2007; Albrecht et al. 2003; Hwarari et al. 2022; Iqbal et al. 2022). The signaling pathway results in activation of transcription factors in the nucleus, such as the NAC (NAM/ATAF/CUC) family (Hu et al. 2006; Jeong et al. 2010), MYB (myeloblastosis) family (Lippold et al. 2009), DREB (dehydration-responsive element-binding) family (Yang et al. 2011), bHLH (basic Helix-loop-helix) family (Zhou et al. 2009), and WRKY family (bind W-box promoter elements in downstream genes) (Wang et al. 2012). The ultimate end result of the signaling pathway is the expression of cold-regulated genes (COR) (Hwarari et al. 2022), which in turn affect physiological changes such as membrane integrity, cell growth, osmotic homeostasis, and cryoprotection (Cober and Voldeng 2012) leading to cold acclimation and tolerance.

Genetic analysis and QTL-mapping approaches to study cold and waterlogging tolerance

Traditionally, the development of soybean cultivars for cold and waterlogging tolerance can be accomplished by breeding selections under field or growth chamber conditions (Wu et al. 2017b; Alsajri et al. 2019). Molecular marker-assisted breeding can increase the efficiency of field selection by reducing the number of genotypes that need to be evaluated in the field. The identification of the QTL or discovery of the candidate genes that regulate the traits in highly tolerant cultivars is needed to develop molecular markers for marker-assisted selection or gene-editing applications (Kaur et al. 2014; Zhao et al. 2017; Gantait et al. 2019). Until recently, QTL mapping was mainly performed in biparental populations such as F₂ or recombinant inbred lines (RILs). Biparental populations can be used for high-resolution mapping and cloning; however, the amount of variation for a trait of interest can be limited by the choice of the parents (Korte and Farlow 2013). Genome-wide association studies (GWAS) overcome this limitation by using a diverse panel of genotypes to find associations between abundant genomic markers and the trait of interest (Elshire et al. 2011; Bohra 2013). GWAS has been successfully applied to the study of abiotic stress tolerance in soybean (Zeng et al. 2017; Khan et al. 2018; Zhang et al. 2022). Research on abiotic stresses in plants has already received abundant attention due to the extensive variability of the production environments. A large number of QTL associated with waterlogging tolerance and cold tolerance related traits have been detected during seed germination and emergence (Andaya and Tai 2006; Qiu et al. 2007), plant growth stages (Andaya and Mackill 2003; Ballesteros et al. 2015), and seed development and maturing processes (Suh et al. 2010; Xu et al. 2015). Table 1 lists a representative sample of the studies that dealt with low temperature and flooding stress. These studies indicate that cold tolerance during the germination process involves a combination of seed cold tolerance, low-temperature survival, and germinability, whereas seed anoxia tolerance and seed germinability through waterlogging are associated with seed waterlogging tolerance. Zhang et al. (2012a) used Hongfeng 11 as the recurrent parent to construct a population of BC₂F₃ lines with the Canadian soybean variety Harosoy. The lines were screened for seed germination under drought and low-temperature conditions. QTL were identified by the chi test and analysis of variance using genotypic and phenotypic data. A total of 18 QTL for drought tolerance and 23 QTL for low-temperature tolerance were detected, of which 12 QTL (i.e., Satt253, Satt513, Satt693, Satt240, Satt323, Satt255, Satt557, Satt452, Sat331, Satt338, Satt271, and Satt588) were associated with both drought and low-temperature tolerance. Zhang et al. (2012b) subsequently partitioned the self-pollinated progeny of 95 BC₂F₃ lines into a randomly selected group and another group based on cold-tolerance germination phenotype and identified 25 QTL at the germination stage and 13 QTL at the seedling stage that were associated with low-temperature tolerance. Of these, 10 QTL were common to the two stages, indicating that soybean low-temperature tolerance is partially genetically consistent at different stages. Among these 10 common QTL, most of the beneficial alleles for low-temperature tolerance at the germination stage were from Harosoy, except Satt440 and Satt041, whereas most of the beneficial alleles at the seedling stage were from Hongfeng 11, except Satt041.

Table 1.

QTL and candidate genes related to seed-flooding and seed low-temperature tolerance.

Previous studies in both biparental and GWAS genetic mappings have reported QTL for seed-flooding tolerance. Five QTL, i.e., Sft1, Sft2, Sft3, and Sft4 (Sayama et al. 2009), were reported in a biparental population evaluated using a lab-based soaking treatment. Among these, Sft1 and Sft2 showed a significant effect on the germination rate. The QTL Sft1 was located within the interval of markers on chromosome 12, near the candidate QTL region for root development under hypoxia and flooding (Nguyen et al. 2017). A near-isogenic soybean line for this QTL region for root development under hypoxia provided significant waterlogging tolerance during germination as compared to the recurrent parent (Nguyen et al. 2017, 2021). The QTL Sft2 is localized near the I locus on chromosome 8 that is involved in seed coat pigmentation. Antioxidants in the colored seed coat appeared to reduce the damage caused by waterlogging stress. Sft3 and Sft4 were associated with the growth of cotyledons and radicles. A GWAS study using 347 diverse soybean genotypes identified a reliable QTL-designated QTN13 using three seed-flooding related traits (Yu et al. 2019). Through candidate gene prediction of the 1.0 Mb region around QTN13/Gm13_35324537 and qRT-PCR analysis, GmSFT (Glyma.13g248000) was identified as the most likely candidate gene for regulating flooding tolerance in soybean seeds (Yu et al. 2019). GmSFT encodes a B-box type zinc finger protein and likely plays a role in hypoxia tolerance based on the evidence from rice (Oryza sativa) (Pandey and Kim 2012).

A GWAS of 243 soybean lines evaluated using a lab-based flooding treatment identified four major single nucleotide polymorphisms associated with seed-flooding tolerance and predicted eight candidate genes and three pivotal genes that were directly or indirectly associated with stress defense mechanisms based on gene ontology enrichment analysis, gene function annotation, and protein–protein interaction network analysis (Sharmin et al. 2021). The above studies indicate that genes involved in multiple abiotic stress and hypoxia are common among the candidate for flooding tolerance during germination.

Other molecular approaches such as transcriptomics, proteomics, and metabolomics have also been successfully used to understand mechanisms of adaptation and genetic regulation of stress tolerance in soybean (Bohnert et al. 2006; Rasool et al. 2015; Gogoi et al. 2018). The transcriptome represents all RNA molecules transcribed or expressed from the cellular genome at a given point in time. It reveals the responses or factors associated with improving plant tolerance to abiotic stresses, mainly by comparing gene expression in plants with or without the environmental stress conditions (Gokce et al. 2020). The proteome represents all proteins present in a tissue or a cell, and the expression of functional proteins depends on the growing conditions, developmental stages, and (or) environmental stimuli. Therefore, it can be used to examine complex biological mechanisms, including plant responses that confer abiotic stress tolerance. Plants respond to a stress by modulating abundance of candidate proteins associated with defense processes. It is important to note that the proteomic responses of plant stresses can be organ specific (Komatsu and Hossain 2013). Proteomics has been widely applied to investigate the soybean responses to abiotic stress conditions. Most of these studies have shown that common proteins related to redox systems, carbon metabolism, photosynthesis, signal transduction, and amino acid metabolism are associated with various stress responses in soybean (Zhen et al. 2007; Yamaguchi et al. 2010). A brief review of relevant literature is provided in the next sections.

Transcriptomic and proteomic approaches to study germination cold tolerance

Cheng et al. (2008) used cDNA-amplified fragment length polymorphism to identify differentially regulated cDNAs following chilling imbibition of the embryonic axis of soybean seed. GmCHI (Glycine max Chilling-Inducible) was shown to be a highly cold stress-inducible gene involved in an ABA-dependent signaling pathway. Cheng et al. (2010) then performed proteomic analysis of the soybeans exposed to low temperatures during germination. A total of 40 proteins were identified that showed significant changes in response to low temperature during germination, of which 25 were upregulated and 15 were downregulated. This study demonstrates roles of multiple metabolic processes, such as cellular defense, substrate, and energy metabolism, protein synthesis, cell growth and division, transport, storage, and transcription in the seed cold-hardiness responses. Enhanced expression of some key enzymes involved in the tricarboxylic cycle (e.g., malate dehydrogenase and phosphoenolpyruvate carboxylase) seem to improve seed survival under stress. Alcohol dehydrogenase I and RAB21 (responsive to ABA 21) likely improved the ability of soybeans to cope with low-temperature stress by alleviating hypoxic symptoms. Changes in stress-related proteins such as LEA (late embryogenesis abundant) and GST24 (glutathione S-transferase 24) were also noted. Swigonska et al. (2014) also showed that regulation of protein synthesis to stablize polysomes could be an important adaption in the response to cold, osmotic, and their combined stress. Jiang et al. (2014) used cold-tolerant soybean germplasm for comparative proteomic analysis of soybean seed cotyledons at chilling and non-chilling temperatures and identified 29 differentially regulated protein spots. Peptide mass fingerprinting identified several proteins, including the glycerol triphosphate hydrogenase A subunit precursor involved in glycolysis, the protein family of the disease-resistant protein RPP1-WsC (Recognition of P. parasitica-Wassilewskija homolog C), the Ca²⁺/calmodulin-dependent protein kinase family with high signaling capacity, and the proteasome α-subunit type 6 protein family involved in apoptosis, signal transduction, and transcriptional regulation. During soybean seed germination, osmotic stress is often a secondary stressor at low temperatures. Plant responses to a combination of stresses can be complex and may result in similar changes in gene expression, protein translation and processing, and metabolism (Pandey et al. 2015; Lamers et al. 2020). Swigonska and Weidner (2013) demonstrated that an overlapping response in protein abundance was induced during cold water and osmotic stress treatments, which confirms that cold water and osmotic stress were regulated by the same circuit and that certain proteins showed expression changes only when subjected to the combined stress.

Transcriptomics and proteomic approaches to study waterlogging tolerance

Sharmin et al. (2020) studied the transcriptome profiles of root tissues in wild soybean seeds via RNA-sequencing. Compared to the control, seeds immersed in water during germination, the genes related to cell wall activity, antioxidant activity, protein metabolism, hormone signaling, transcription factor activity, protein kinase activity, and signaling showed differential expression. The differentially expressed gene encoding gibberellin 2-β-dioxygenase 2 and DWARF8 were significantly upregulated in submergence-tolerant soybeans, thereby affecting the hormone metabolism associated with germination. The members of the ethylene response factor superfamily (ERFVIIs) (Hinz et al. 2010; Licausi et al. 2011) (i.e., HRE1, HRE2, RAP2.2, RAP2.3, and RAP2.12), which have been identified as key regulators of the response to hypoxia in Arabidopsis, were significantly upregulated in tolerant soybean materials. The receptor-like kinases (RLKs) are signaling components known to regulate plant responses under various abiotic stresses (Zhang et al. 2013; Chen et al. 2017). The RLKs were significantly up- and downregulated in flood-tolerant and sensitive soybean types, respectively (Sharmin et al. 2020). Protein hydrolases, which play a key role in the degradation of storage proteins under anoxic conditions during germination (Han et al. 2013), were found to be upregulated in flood-tolerant soybeans. Sharmin et al. (2020) reported upregulation of proteolytic enzymes namely, E3 ubiquitin-protein ligase, aspartic proteinase, and subtilisin-like proteases in flood-tolerant when compared to flood-sensitive soybean. Thus, under seed submergence/low oxygen stress, tolerant plants may derive energy for seed germination from protein and lipid metabolism (Sharmin et al. 2020).

Komatsu et al. (2009a) used transcriptomic and proteomic techniques to jointly analyze the roots and hypocotyls of soybean seedlings that had germinated for 2 days and then were flooded for 12 h. Hemoglobin, acid phosphatase, and Kunitz trypsin inhibitor were shown to be altered at both the transcriptional and translational levels. Genes associated with alcohol fermentation, ethylene biosynthesis, pathogen defense, and cell wall relaxation were also significantly upregulated, and ROS scavengers and chaperones were altered only at the translational level (Komatsu et al. 2009a). Analysis of total protein in soybean roots and hypocotyls using proteomics indicates that flood stress alters protein levels associated with protein transport and storage, ATP synthesis, metabolism, and signal transduction (Komatsu et al. 2012a). A comparative proteomic analysis of wild type and a mutant that lacked growth-suppression response to flooding confirmed the role of efficient anaerobic metabolism in conferring early-stage flooding tolerance (Komatsu et al. 2013). A time course proteomic study found that flooding causes growth suppression within 3 h of initiation of stress and altered energy metabolism and calcium-related signaling transduction are key events at the early stages of stress response (Yin et al. 2014). Inhibition of cell wall synthesis and protein metabolism-related proteins may cause more severe damage as submersion time increases (Khatoon et al. 2012; Nanjo et al. 2013).

Subcellular proteomics is an emerging field that can help elucidate communication and signaling between organelles during a stress response (Komatsu and Hashiguchi 2018). The first organelle that responds to flooding stress is the cell wall. Komatsu et al. (2010) revealed that under flooding stress, soybean roots and hypocotyls inhibited lignification by the downregulation of ROS and jasmonic acid biosynthesis-related proteins. The plasma membrane is the main boundary connecting the cytoplasm to the extracellular atmosphere and plays an important role in cellular communication (Hossain and Komatsu 2014). Komatsu et al. (2009b) studied the plasma membrane proteome of germinating soybean seedlings during a flooding treatment and showed that cell wall proteins, superoxide dismutase, and heat shock related and signaling proteins were upregulated. It suggests that the antioxidant system, and heat shock proteins may be protecting cells from oxidative damage and protein denaturation and degradation in response to flood stress. Furthermore, regulation of plasma membrane H⁺-ATPase for maintenance of ionic homeostasis is achieved through the synergistic interactions with 14–3–3 protein and serine/threonine protein kinases and band 7 family proteins (Komatsu et al. 2009b). A nuclear proteomics study of early flood response in germinating seeds indicates the involvement of ABA, possibly by altering the phosphorylation of nuclear-localized phosphoproteins such as glycine-rich proteins, zinc finger/BTB domain-containing protein 47, and rRNA processing protein Rrp5 (Yin and Komatsu 2015). Long-term flooding stress suppresses protein translation in soybean root tip nuclei by inhibiting pro-ribosomal biogenesis-associated and mRNA-processing-associated proteins (Yin and Komatsu 2016). Seventeen nuclear proteins associated with chromatin structure were reduced in soybean in response to initial flooding stress. For example, the mRNA expression levels of histone H3 were significantly reduced in abundance due to inundation stress. Some nuclear proteins associated with the protein synthesis, RNA, and DNA were reduced in direct correlation with the length of time of the flooding. The mRNA expression of genes encoding flood-responsive nuclear proteins was repressed by the transcriptional repressor Actinomycin D (Yin and Komatsu 2016). Among the cellular organelles, mitochondria are often targeted for study because the mitochondrial electron transport chain is compromised by most forms of abiotic stress, leading to excessive ROS production (Taylor et al. 2005; Jacoby et al. 2010). Komatsu et al. (2011) used proteomic and metabolomic techniques to study mitochondria in roots and hypocotyls of 4-day-old soybean seedlings that had been flooded for 2 days. Proteins and metabolites associated with the tricarboxylic acid cycle and γ-aminobutyric acid shunt were upregulated under flooding stress. In contrast, endosomal carrier proteins and proteins associated with complexes III, IV, and V of the electron transport chain were downregulated (Komatsu et al. 2011). While NADH and NAD content increased, the content of adenosine triphosphate decreased. The endoplasmic reticulum has various cellular functions, including secretory protein synthesis, protein folding and degradation, lipid synthesis and transfer, and calcium signaling and storage (Chen et al. 2010). The soybean root tip endoplasmic reticulum study by Komatsu et al. (2012b) showed that proteins associated with protein synthesis, post-translational modifications, protein folding, protein degradation, and protein activation in the soybean endoplasmic reticulum were downregulated by flooding stress. The response of different subcellular structures to flooding stress is listed in Table 2. Although these studies elucidate stress responses in specific subcellular regions, intracellular stress signaling and its role in these responses require further investigation.

Table 2.

Proteins involved in soybean seed germination in a waterlogged environment.

Proteomics has also highlighted the role of metabolites in flooding responses. A proteomic study of soybean cotyledons under flooding stress determined that ferritin in soybean cotyledons may play a positive role in plant defense by promoting iron homeostasis and scavenging ROS (Kamal et al. 2015). A nuclear proteomics study of the early flood response in germinating seeds indicates the involvement of ABA, possibly by altering the phosphorylation of nuclear-localized phosphoproteins such as glycine-rich proteins, zinc finger/BTB domain-containing protein 47 and rRNA processing protein Rrp5 (Yin and Komatsu 2015). Another plant hormone melatonin demonstrated a protective role (Hassan et al. 2022) by improving germination and reducing cell death in soybeans during flooding (Wang et al. 2021). Melatonin treatment of soybean seeds promoted their growth potential under flooding stress by controlling protein degradation, RNA modification, and cell wall metabolism (Wang et al. 2021).

Perspectives and concluding remarks

QTL studies show that inheritance can be highly complex due to the involvement of many genes which have both small and medium impacts on cold and flooding tolerance. This is in agreement with the environmental association analyses which also show that genetics underlying environmental adaptation can be highly complex (Bandillo et al. 2017). Genomic selection (GS) approaches can be more suitable where a large number of genes with minor effects are needed for an adequate expression of a trait (Li et al. 2016; Michel et al. 2019). Although the authors are not aware of any study on efficacy of GS for cold tolerance for germination, GS has shown promise for some other traits that have been studied (Stewart-Brown et al. 2019). Alternatively, knowledge obtained from the studies of signaling and molecular mechanisms of cold tolerance can be leveraged to directly design strategies for developing cold- and flooding-tolerant germplasm. Translational genomics, which utilizes knowledge generated from model systems can be used to engineer stress tolerance traits in crops. A homologous gene of AtTCF1 (A. thaliana tolerant to chilling and freezing 1) in soybean (GmTCF1a), greatly enhanced Arabidopsis plant survival when ectopically expressed under freezing conditions (Dong et al. 2021). Another translational study reported that ectopic expression of GmNEK1, a homolog of AtNEK6 (A. thaliana Never In Mitosis Gene A-related kinase 6) conferred salt and cold tolerance in transgenic Arabidopsis (Pan et al. 2017). Overexpression of AtXTH31 (A. thaliana Xyloglucan endotransglycosylases/Hydrolase 31) in transgenic soybean plants resulted in enhanced germination rate (40%–58% compared to 25% for control) and elongated root and hypocotyl under flooding conditions (Song et al. 2018). Upregulation of RNA-binding proteins such as RZ-1a (At3g26420) is hypothesized to enhance the cold tolerance during germination through the stabilization of RNA secondary structure (Kim et al. 2005; Kim and Kang 2006). Overexpression of GmWIN1-5 (Glycine max Wax Inducer1-5), a homolog of AtWIN1 (At1g15360), was recently shown to increase cold tolerance in transformed hairy roots (Cai et al. 2022). GmWIN1-5 encodes a transcription factor that regulates the expression of lipid biosynthetic genes and is preferentially expressed in flowers and developing seeds. Studies which were not focused on germination have also identified interesting genes that improved the responses to drought and (or) salt stresses at various post-germination stages (Liao et al. 2008a, 2008b; Yang et al. 2019). It remains to be seen whether manipulation of these stress-responsive genes can also improve germination under cold and flooding stress. Moreover, the genetic materials generated through these approaches also need to be characterized using the latest techniques discussed in previous sections to correlate the phenotypes with molecular changes for expanding the understanding of mechanisms involved. Still these examples suggest that current knowledge from model systems can be leveraged to directly develop the phenotype of interest. Soybean candidate genes from QTL studies described in the previous sections also need further characterization at both field and molecular levels, and validation in different genetic backgrounds before they can be utilized in cultivar development (Biswas et al. 2017).

Most of the studies on abiotic stress tolerance using transcriptomics have only examined the responses to a single treatment (Zhang et al. 2019; Pan et al. 2020; Saux et al. 2020; Wang et al. 2022); however, field crops often face combinations of two or more forms of stress: cold and waterlogging in the case of soybeans production in the prairies. Therefore, combined effects of these stresses on plants cannot be completely inferred from the effects of a single stress. For example in Arabidopsis, about 61% of the transcriptome changes in response to a combined stress, which could not be predicted from the response to a single stress (Rasmussen et al. 2013). The importance of the combined effects of cold and waterlogging stress on production should receive more attention, especially on the identification and validation of the genetic loci associated with the stress. Climate change presents greater likelihood of adverse temperature and flooding stress in crop production. Future research on climate risk reduction needs to be an important part of variety development.