Soybean, an important legume, is one of the most widely grown food crops in the world due to its valuable seed composition. In 2019, the annual global soybean production was estimated to be above 333 million tones (Faostat, 2019). Soybean provides an inexpensive source of protein and fats, and natural nitrogen fertilisation for the soil (Foyer et al., 2016). Interestingly, the economic benefits derived from soybean cultivation are not just limited to the food supply; it is also an important industrial crop utilised in producing edible oils, wax, paints, dyes and fibre (Rezaei et al., 2002; Raghuvanshi and Bisht, 2010). Also, meat substitutes based on soybean are extensively used by vegan and vegetarian consumers (Messina and Messina, 2010; Raghuvanshi and Bisht, 2010; Tang, 2017).

Soybean is grown mainly in tropical, subtropical and temperate regions (Fao, 2021). It is a water-intensive crop, requiring substantial water to grow and reproduce (Bhardwaj, 1986). Consequently, rising global temperatures and changing precipitation patterns pose a significant threat to soybean production, especially in under-irrigated or rainfed areas (Jin et al., 2017; Cotrim et al., 2021). It is known that under dry conditions or drought, soybean yield can reduce by more than 50%, causing substantial financial losses to farmers and growers (Wei et al., 2018). Hence, drought is a significant climatic risk that calls for effective mitigation strategies to sustain the supply of soybeans worldwide.

Soybean varieties are classified into maturity groups according to their response to the photoperiod. Early maturing varieties belong to groups 0 to 3, whereas late-maturing varieties fall in groups 6 and onwards (Zhang et al., 2007; Yang et al., 2019). Drought impacts soybeans cultivars differently, as some cultivars are more susceptible than others (Oya et al., 2004; Maleki et al., 2013; Du et al., 2020; Dayoub et al., 2021). Also, the timing of drought stress, whether at the vegetative or the reproductive phase, is important in determining yield loss. Desclaux et al. (2000) investigated the drought-induced phenotypes of early maturity soybean varieties grown in France. They reported that drought stress at vegetative stages led to reduced plant height and a decline in seed number in the early reproductive stages and reduced seed weight in late reproductive stages. The water scarcity between flowering and early seed filling stages can affect branches’ vegetative growth, resulting in decreased branch seed number and reduced branch seed yield (Frederick et al., 2001). A report on the effect of drought on soybeans grown in the semi-arid and semi-humid regions of Huaibei regions of China reported a 73–82% decline in yield when drought stress was applied at flowering and seed filling stages (Wei et al., 2018). Recently, Du et al. (2020) reported that long-term drought stress in reproductive stages decreases biomass allocation to reproductive organs, thereby reducing seed weight in soybean. In addition, drought also impacts the symbiotic nitrogen-fixing ability of soybeans by disturbing nitrogenase activity, which can cause carbon shortage and oxygen limitation leading to poor growth and yield (Arrese-Igor et al., 2011; Collier and Tegeder, 2012; Kunert et al., 2016).

Plants use diverse mechanisms to overcome the adverse effects of drought, and the ability of crops to adjust using adaptive traits is termed ‘drought tolerance’ (Basu et al., 2016). Decreased stomatal opening associated with reduced photosynthesis is a typical drought response observed in plants (Liu et al., 2003; Mak et al., 2014). Abscisic acid (ABA) plays a vital role in reducing water loss under dry conditions (Wilkinson and Davies, 2002; Kim, 2014; Wang et al., 2016). Further drought-tolerant soybeans exhibit higher ABA levels than drought-susceptible varieties (Mutava et al., 2015). ABA, synthesised in plant roots, is transported to the guard cells of leaves, where it induces closure of stomatal openings to reduce water loss (Wilkinson and Davies, 2002). Further, there is evidence that ABA synthesised in leaf xylem also contributes to this process (Malcheska et al., 2017). However, reduction in the stomatal opening leads to reduced CO₂ assimilation and photosynthesis, affecting growth and development (Cornic and Briantais, 1991; Ohashi et al., 2006; Mutava et al., 2015; Cohen et al., 2021).

Maintenance of cell turgidity is another essential adjustment to survive drought. Under dehydration conditions, cells induce biochemical changes by synthesising necessary metabolites called osmoprotectants (Silvente et al., 2012). These include soluble and complex sugars, sugar alcohols, organic acids and free amino acids (Ashraf and Iram, 2005; Silvente et al., 2012; Kido et al., 2013). Osmoprotectant accumulation in the cell balances the osmotic difference between cell exteriors and the cytosol help to retain water and maintain the integrity of the cell membrane (Yancey, 2005; Basu et al., 2016). An increase in soluble sugars, such as sucrose and fructose, improves homeostasis under stressed conditions. Further, soluble sugars are required for enhanced carbohydrate metabolism, signal transduction and synthesising enzymes and hormones needed to survive under drought (Gupta and Kaur, 2005; Mak et al., 2014; Du et al., 2020). Metabolite profiling of biochemical compounds synthesised during drought stress revealed an increase in pinitol in the leaves of a drought-susceptible cultivar (Silvente et al., 2012). Pinitol is a common sugar alcohol that acts as an osmoprotectant in legumes (Ford, 1984; Streeter et al., 2001; Dumschott et al., 2019). Further, Silvente et al. (2012) reported increased levels of amino acids, such as proline, during drought stress in flowering stages. Proline helps retain water by adjusting the intracellular osmotic potential of the cells (Heerden van and Krüger, 2002). In addition, reactive oxygen species (singlet oxygen) produced in stressed cells can cause severe oxidative damage under prolonged drought conditions. Alia et al. (2001) suggested that proline acts as a scavenger of singlet oxygen. However, the role of proline in quenching singlet oxygen in stressed plants remains debated as Signorelli et al. (2013) demonstrated that proline could not quench singlet oxygen in an aqueous buffer. Figure 1A summarises the effects of drought on soybean.

On the cellular level, largely overlapping signalling mechanisms and phytohormone cross-talks mediate drought response in plants (Basu et al., 2016). With the advent of high-throughput sequencing techniques, several genes or gene families have been identified and characterised in soybean (Wong et al., 2013; Zhang et al., 2020, 2021). Among these, the CUC (NAC; Hussain et al., 2017), MYB (Chen L. et al., 2021), WRKY (Shi et al., 2018), ABA-responsive element binding (AREB; Fuganti-Pagliarini et al., 2017) and dehydration response element-binding proteins (DREB; Nguyen et al., 2019) transcription factor families are some of the prime regulators that control drought response by regulating the synthesis of drought-responsive hormones, such as ABA, ethylene and other drought signalling compounds including Brassinosteroids (Nguyen et al., 2019; Chen L. et al., 2021).

Comparative transcriptomics have further elucidated the molecular mechanisms underlying drought response in soybean (Ha et al., 2015; Hussain et al., 2017). For example, Hussain et al. (2017) identified 28 drought-responsive GmNAC genes in soybean and reported that only eight GmNAC genes showed high expression levels in drought-tolerant soybean variety; with drought-sensitive cultivars exhibiting lower expression levels. WRKY transcription factors have been highlighted to play vital roles in plant abiotic stress tolerance (Ning et al., 2017; Shi et al., 2018). Shi et al. (2018) identified a drought-responsive soybean WRKY gene, GmWRKY12, whose over-expression in a transgenic hairy root assay led to increased proline levels under drought stress. Further, Wei et al. (2019) characterised another WRKY gene, GmWRKY54, which mediates drought tolerance via ABA and Ca⁺² signalling pathways. The over-expression of GmWRKY54 conferred drought tolerance in soybean (Wei et al., 2019). Other recently identified gene families participating in soybean drought response are AT-hook motif (Wang et al., 2021b), P-type ATPases (Zhao et al., 2021), CCT family (Mengarelli and Zanor, 2021) and GRAS (Wang et al., 2020). Identifying new drought-responsive genes is vital for developing drought-smart soybeans. The term ‘drought-smart’ refers to soybean cultivars that can adapt to diverse types of drought, such as early drought, middle-stage drought, later-drought and seasonal drought, by combining different mechanisms of drought resistance, such as drought avoidance, drought tolerance and drought recovery. Drought avoidance is generally achieved by retaining water in plant tissues, either by restricting water loss or using water judiciously to support different plant functions. However, drought tolerance is achieved using adaptive traits, such as biochemical adjustments to maintain cell turgidity and minimise photosynthetic damage (Basu et al., 2016). Further, cultivars with drought avoidance and drought tolerance features can recover fast upon rehydration after a short-term or seasonal drought (Dong et al., 2019). Hence, smart soybeans are able to produce a stable yield in drought-prone areas. Traits, such as improved water retention, restricted transpiration and stable photosynthesis, are desirable to sustain biomass and yield under water-deficit conditions (Wei et al., 2019; Kunert and Vorster, 2020; Yang et al., 2020). A combination of such traits would make a cultivar suitable for cultivation in arid or semi-arid regions (Liu et al., 2005; Abdel-Haleem et al., 2012; Carter et al., 2016). Also, as genome sequencing of soybean cultivars gains momentum, the new genetic diversity resources will aid in developing drought-smart varieties using advanced breeding or genetic engineering methods (Fuganti-Pagliarini et al., 2017; Golicz et al., 2018; Kajiya-Kanegae et al., 2021). Here, we review genetic improvement approaches for improving drought tolerance in soybean.

A combination of advanced phenotyping, molecular breeding and genetic engineering approaches can be employed to breed drought-smart soybeans (Figure 1B). Using conventional breeding techniques, pre-selected soybean germplasms (donor cultivars) with desired drought-responsive traits can be crossed to introduce favourable alleles in the resulting populations. However, extensive screening of subsequent generations is required to select a line with stable characteristics for continued cultivation (Carter et al., 2016). Further, genetic transformation offers the possibility of targeted gene expression under constitutive or inducible promoters (Ribichich et al., 2020). For example, Arabidopsis thaliana AtMYB44 gene was transformed in soybean using Agrobacterium-mediated transformation which resulted in improved soybeans with better yield under water-deficit conditions in the field (Seo et al., 2012). Improvements in tissue culture regeneration of commercial soybean cultivars and optimization of Agrobacterium-mediated transformation methods will also facilitate engineering soybeans for drought tolerance (Raza et al., 2017, 2019). Recently, more precise gene editing technique, CRISPR/cas9, has shown promising results in modifying soybean’s genome in a targetted manner to obtain more specific gene modifications. CRISPR/cas9 has been successfully employed to characterise soybean drought-responsive genes using knock-down approaches. For example, CRISPR/cas9-mediated mutagenesis of soybean circadian rhythm genes (GmLCLs) generated mutant plants with decreased water loss under dehydration stress conditions (Yuan et al., 2021). Interestingly, breeding and genetic transformation methods have successfully delivered improved soybeans that have been tested in laboratory, glasshouse or field conditions and a few of these varieties have also been approved for commercial production. In the sections below, we review recent examples of soybeans improved for drought tolerance via breeding or transgenic approaches.

Soybean is a crop of immense economic importance (Johnson and Myers, 1995; Messina and Messina, 2010). Vast genetic diversity has been reported in soybean germplasm, and the increasing availability of soybean genetic resources has instigated the development of drought-smart soybeans (Carter et al., 2016; Ribichich et al., 2020; Kajiya-Kanegae et al., 2021; Chen L. et al., 2021). Modern breeding and advanced biotechnology methods have shown promising results, and market-ready drought-tolerant soybeans have been released in some parts of the world (Carter et al., 2016; Ribichich et al., 2020). However, soybean production is still dependent on adequate irrigation facilities in many regions, especially in under-developed and developing nations (Droppers et al., 2021; Suriadi et al., 2021). As genetic and non-genetic improvement methods are tested on more cultivars, the dependency of soybean production on rainfall or heavy irrigation should reduce. With changing precipitation patterns and a hotter climate, drought-tolerant soybeans will play a significant role in ensuring our future food security.

All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.