Engineering Crops for Enhanced Drought Stress Tolerance: A Strategy for Sustainable Agriculture

Abstract

Drought stress can reduce agricultural production, which is a challenge considering the food demand due to the increase in world population. To face this challenge, the design of plants with a phenotype of drought stress tolerance is needed. Conventional breeding has been widely used with this goal, but it requires considerable time and resources. Drought stress response and tolerance are complex issues influenced by numerous environmental and genotypic factors. In this review, experiments involving novel biotechnological tools to improve plant breeding are described and discussed. These experiments involve the use of techniques to accelerate breeding cycles and to enhance the selection of superior genotypes. Furthermore, experiments carried out to elucidate the molecular mechanism of drought stress tolerance and to engineer crops to achieve drought stress tolerance using recombinant DNA technology are described. The main traits associated with drought-tolerant genotypes and the response to drought stress at the morphological, physiological, metabolic, and biochemical levels are analyzed. To cope with the complexity of plant drought response, conventional breeding needs to be integrated with novel tools. It is hoped that this will help to achieve sustainable agriculture development; however, the implications of the use of these biotechnological tools, both alone and coupled, need to be analyzed.

1. Introduction

Sustainable development involves the responsibility of the present generation to meet its needs without compromising the capability of future generations to do the same. It includes the economic, social, and environmental dimensions. This last dimension entails avoiding the degradation of the environment to keep the production of agricultural commodities, the availability of clean water, and the preservation of plant and animal biodiversity. That is why water sustainability in a scenario of sustainable agriculture is a world-level relevant subject [1]. Drought stress is a significant drawback for the transition to sustainable agriculture [2]. In developing countries, drought stress induces significant economic losses, which suggests the need to have more efficient water management to improve agricultural sustainability under conditions of climate change [3].

To manage the increased scarcity of water for agriculture in the world and the increasing world food demand, it is necessary to identify or develop novel drought-tolerant crop varieties [4,5], which can be carried out by developing modern breeding lines, wild varieties, landraces, or the product of their crosses [6,7]. Conventional breeding has been widely used with this goal [8]. However, it requires a significant amount of time and resources, mainly because the main crops have long reproductive cycles, and it requires many cycles of crossing, selection, and testing for the development of the enhanced varieties [4].

Drought stress response and tolerance are complex issues involving intricate mechanisms at all levels of an organism’s metabolism and are influenced by numerous environmental and genotypic factors. Plant drought stress tolerance is a highly polygenic trait and displays phenotypic plasticity determined by management and environmental conditions [6]. To cope with the complexity of plant drought response, conventional breeding needs to be integrated with novel biotechnological tools to accelerate the development of high-yielding and drought-tolerant varieties necessary for the next decades [9]. To develop novel drought-tolerant genotypes, a multidisciplinary breeding approach that includes many of the novel available tools is the appropriate approach to enhance drought tolerance in plants [6].

In this review, experiments involving novel biotechnological tools to improve plant breeding are described and discussed. These experiments involve the use of techniques to accelerate breeding cycles, such as speed breeding (SB), and to enhance the selection of superior genotypes, such as high-throughput phenotyping (HTP) and marker-assisted selection (MAS). Furthermore, experiments carried out to elucidate the molecular mechanism of drought stress tolerance and to engineer crops to achieve drought stress tolerance are described, which involve the use of recombinant DNA technologies, such as Agrobacterium infection, clustered regularly interspaced short palindromic repeats with Cas 9-associated endonuclease (CRISPR/Cas 9), and RNA interference (RNAi). The main traits associated with drought-tolerant genotypes are described. Moreover, a common response to drought stress at the morphological, physiological, metabolic, biochemical, and transcriptional levels is described and discussed. The advantages, disadvantages, and use limitations of each biotechnological tool and their integrated use are also discussed.

2. Development of Resistant Plants with a Drought-Tolerant Phenotype

In order to manage the increased drought and world food demand, conventional breeding still represents a valuable tool for the development of drought-tolerant crops [4,5]. For this goal, plant breeders and scientists can have access to modern breeding lines, wild relatives, landraces, and the products of their crosses [6,7]. Crop wild relatives and landraces can harbor alleles preserved over time that disappeared in domesticated lines, conferring on the plant the capacity to adapt and survive in conditions of water scarcity [10]. Landraces growing in dry areas for thousands of years are a valuable germplasm source to breed drought-tolerant lines [11]. In rice, drought-tolerant traits have been found among both wild and cultivated varieties [12].

Improvement of drought tolerance by conventional breeding has been carried out successfully for maize, potato, rice, sugarcane, wheat, and cassava [10]. Furthermore, 49 rice varieties tolerant to drought stress were developed by traditional breeding [13]. For pearl millet, 15 drought-tolerant landraces have been identified for use in plant breeding [11]. Adapted landraces and elite genetic material can be crossed to enhance adaptation to drought while high productivity is preserved [14]. Nevertheless, some disadvantages of the use of wild relatives in plant breeding are the limited germplasm resource, the low germplasm characterization, and the presence of poor agronomic traits in wild genotypes [6].

In conventional breeding, desirable traits in parental lines are identified, and then, recombination is carried out to obtain a favorable combination of traits for developing an improved line [15]. Drought-stress traditional plant breeding consists of the recombination of parent lines’ alleles by the Mendelian approach to obtain a convenient combination of traits for developing improved varieties [7]. Plant phenotype is the product of complex interactions of dynamic variables involving genotype, environment, and management interactions (GxExM). Thus, an accurate final screening of the descendant lines must be carried out [6].

2.1. Screening for Natural Resistance

Drought-tolerant genotype screening in crops is mainly based on morpho-physiological traits, with seed yield being the most common selection criterion [6,16]. In wheat, the evaluation of thousand-grain weight has led to the identification of good-performing cultivars under prolonged drought stress [5,17]. Nevertheless, in drought stress conditions, yield is influenced by many factors inherent to GxExM interactions. To select the best drought-tolerant genotypes, an evaluation based on more adaptive traits in response to water deficit at the initial stages of plant growth has been proposed [7].

Drought-response constitutive traits are those present in plants growing under optimal conditions, while responsive traits are those expressed under drought stress [16]. These traits can be morphological, physiological, and biochemical. A list of the traits evaluated or proposed in drought tolerance screening is shown in

Table 1. Three different groups of traits have been most associated with the phenotype of drought stress tolerance.

Physical or Agronomic Traits	Physiological Traits	Biochemical Traits
Germination rate	Flowering time	Abscisic acid biosynthesis
Seedling vigor	Osmotic adjustment	Auxin biosynthesis
Plant height	Internal water status	Osmolytes biosynthesis
Harvest index	Cell-membrane stability	Proline biosynthesis
Seed yield	Cuticle deposition	Heat shock protein biosynthesis
Leaf area	Photosynthetic capacity	Dehydrins biosynthesis
Root architecture	Leaf water content
Root mass	Chlorophyll stability
Root volume	Stomatal conductance
	Water-use efficiency
	Canopy temperature
	Cuticular wax leaf content
	Stomatal density

.

Root architecture traits related to drought tolerance include fine roots, root length, root area, root angle, and root weight [16]. These traits influence yield in water-deficient conditions [18]. In the field, they determine the soil area available for nutrient and water uptake for the plant. Roots denser, deeper, and with a high presence of fine roots are convenient traits for good crop performance under drought stress conditions [19]. Deeper and widely distributed roots influence drought tolerance in chickpeas [20]. In potatoes, dry root mass has been used to identify drought-tolerant genotypes [21].

Early flowering and early seedling vigor are traits related to drought tolerance [11]. A reduced crop cycle is advantageous in water-scarce conditions in the field because early flowering leads to the plant ending its reproductive cycle before soil water reserves are depleted [7]. In legumes, short-duration varieties are less prone to terminal drought stress caused by water depletion in the soil [6].

Stomatal conductance and transpiration efficiency are traits used during drought genotype screening [22,23]. Genotypes with drought stress tolerance show an efficient regulation of stomatal activity in order to reduce transpiration rate while carbon dioxide assimilation is maintained. The capacity to perform an efficient osmotic adjustment is another relevant trait for drought tolerance screening, which consists of increasing the intercellular solutes to maintain cell osmotic potential under low water potential conditions [16].

Canopy temperature is also a relevant trait for maintaining plant homeostasis during drought stress conditions. At the most active phases of its reproductive stage, sorghum increases the epicuticular wax leaf content to regulate canopy temperature and reduce cuticular transpiration, leading to drought stress resistance [24]. The epicuticular wax deposition is a valuable indicator during the screening for drought-resistant varieties related to water use efficiency in Sorghum [16].

Proline, heat shock proteins (HSPs), and dehydrin (DHN) synthesis protect the plant from oxidative damage in drought-tolerant cultivars [25]. Proline has been associated with an enhancement of the photosynthetic capacity under drought stress conditions [23,26]. The reduction of oxidative stress facilitates a normal leaf function during water deficit stress, which could lead to a good grain yield under drought-stress field conditions [16].

Reduced leaf area to reduce transpiration is a relevant trait in plants exposed to prolonged drought episodes. This leads to the crops retaining soil water for use during the grain-filling phase [7]. In rice, leaf area, cell–membrane stability, chlorophyll stability index, leaf rolling, days to flowering, root volume, root biomass, and proline content have been associated with drought stress tolerance, being good selection criteria for rice breeding [27,28].

Abscisic acid (ABA) and auxin production are biochemical traits associated with drought stress response in Sorghum. The increase in auxin production in roots during drought conditions has been related to an increase in root systems [16]. High root biomass, deeper roots, leaf area index, leaf cuticular waxiness, higher stomatal conductance, and low canopy temperature have been considered relevant traits for drought avoidance in pigeon pea breeding [6,7,29].

2.2. Genetic Improvement by Traditional Genetics

For the development of drought-tolerant lines, conventional breeding techniques, such as pedigree selection, recurrent selection, backcrossing, and induced mutation, are used. Pedigree selection consists of selecting plants with desirable traits from segregating generations based on their phenotype. By pedigree selection, the combination of many genes controlling abiotic stress can be obtained [8]. It has been helpful for the development of drought-tolerant varieties of rice obtained from rice landraces. The rice variety ‘Nerica’, was developed by interspecific hybridization of African and Asian rice species, Oryza glaberrima and O. sativa, respectively, which present abiotic stress tolerance and high-yielding traits [13]. Nevertheless, pedigree selection requires a high level of knowledge of the breeding materials and periodic evaluation of many lines across reproduction cycles, which makes it time-consuming [8].

In recurrent selection, plants with interesting traits are repeatedly selected and interbred for various generations to enhance specific traits by gradually increasing the frequency of favorable alleles. For instance, 12 drought-tolerant rice lines were obtained through the recurrent selection of 31 rice introgression lines [30]. Backcrossing consists of the transfer of a desirable trait from a donor parent into an adapted plant, known as the recurrent parent. Repeated crosses are carried out to maintain the desirable traits of the recurrent parent while retaining the desirable phenotypic characteristics from the donor parent. This technique has led to the development of drought-tolerant varieties in sorghum and rice [8].

Mutation breeding consists of the intentional mutagenesis in plants induced by exposure to radiation [4]. Induced mutation offers the advantage of creating novel gene alleles that are not found in nature, which has led to the achievement of high-yielding and drought-tolerant mutants [31]. MK-D-2 and MK-D-3 are drought-tolerant rice mutant lines obtained through irradiation with 300 Gy of gamma rays [32]. A drought-tolerant phenotype was obtained through gamma irradiation of the rice landrace “Tarom Mahalli” [33]. Nevertheless, induced mutation can lead to the introduction of undesirable agronomic traits, necessitating the periodic evaluation of many lines.

Genetic improvement has contributed to around 75% of the increase in productivity in agriculture [13]. To cope with the increased scarcity of water for agriculture in the world, it is necessary to identify or develop drought tolerance traits in the existing plant varieties. Nevertheless, conventional breeding requires a significant amount of time and resources. It takes around 10–20 years to develop a new drought-tolerant line by conventional breeding. This amount of time is needed because the main crops have long reproductive cycles, and it requires many cycles of crossing, selection, and testing for the development of the enhanced varieties [4]. For this, conventional breeding approaches can be complemented by novel biotechnological tools.

2.3. Novel Tools to Speed up Plant Breeding

Despite conventional breeding having led to the development of numerous superior varieties and lines, it cannot cope with the increasing drought stress and human population. Thus, a substantial increase in the development of high-yielding and drought-tolerant varieties is necessary [9].

Speed breeding (SB) is a new technique that reduces the reproductive cycle of plants mainly by controlling the photoperiod regime [34]. SB does not require specific equipment for in vitro culturing and can be used with several germplasm. A disadvantage of SB is that daytime length, optimum light intensity, and wavelength need to be standardized for each plant species [35]. SB has been allowed to obtain four to six generations per year of Sorghum, canola, chickpea, barley, and wheat [36,37,38,39,40]. Other conditions that are controlled in speed breeding are temperature, humidity, and CO₂ levels. It is expected that SB will help to strengthen the drought breeding programs [13]. Nevertheless, SB has only been utilized in a limited number of crops. An effort to standardize an SB protocol for other relevant crop species is needed.

The foundation of SB consists of increasing the photosynthesis rate, reducing the vegetative phase, and reducing the flowering period by adjusting the light intensity and duration. In this context, the source of light is a key factor in SB [40]. For plants, lights with low photo-synthetically active photons (PPFs) appear to be more suitable. In rice and barley, the glaucousness drought-tolerant trait can be obtained by SB [9]. In chickpeas, four generations of working seeds can be obtained in a year, leading to the improvement of drought-stress commercial cultivars in half of the time usually required for variety development and release [34].

Single seed descent (SSD) consists of using one seed per plant in each consecutive generation during plant breeding to achieve an essentially homozygous population. On the other hand, double haploid (DH) technology consists of the use of plants with a double set of haploid chromosomes, which accelerates the development of entirely homozygous lines to develop improved varieties and hybrids. It has been demonstrated that it is helpful for maize, wheat, and rice breeding [16,41]. SB can be complemented with DH and SSD to develop enhanced varieties in a short time. Nevertheless, SB required the standardization of a specific protocol for each plant species. Additionally, some plant species are sensitive to high temperature, and suitable facilities and trained staff are required, which could limit their use for some important crops [9].

2.4. Early Molecular Tools

In conventional breeding, the identification of some traits by phenotype observation can be difficult to visualize, expensive, or time-consuming. The use of molecular markers enables the precise, efficient, and faster identification of desirable genotypes, saving time and economic resources. In molecular marker-assisted selection (MAS), an association between specific nucleotide sequences and a target phenotypical trait is established [13]. Restriction fragment length polymorphism (RFLP) and random amplified polymorphic DNA (RAPD) belong to the first molecular markers used for plant breeding. These are based on the visualization of a specific pattern of bands on electrophoresis or southern blot techniques, and have been shown to be helpful to identify drought-resistant genotypes in breeding programs [4,16].

RFLP and RAPD are relatively easy to perform as they do not require sophisticated equipment or extensive technical expertise. Compared with the conventional screening methods, MAS led to a faster and more accurate selection [8]. SB could be easily assisted with MAS to reduce the time for the development of enhanced varieties [4,16]. Nevertheless, the accuracy of RFLP decreases in some plant species with low genetic diversity and non-specific amplification can be present in RAPD. To deal with these drawbacks, the analysis of a rather large plant genomic sequence needs to be carried out in the best scenario, which is possible thanks to the development of next-generation high-throughput sequencing (NGS) platforms.

2.5. High-Throughput Genotyping to Boost Plant Breeding

NGS can assist conventional breeding to accelerate the development of enhanced breeding lines adapted to drought conditions [10]. Allelic diversity of germplasm can be characterized by genomics to identify specific combinations of alleles that develop desirable traits. Efforts to develop genomic breeding have been made in rice, wheat, and maize [7]. Genomic selection (GS), which consists of the identification of breeding lines with high genetic potential for desirable traits, such as drought adaptation, has led to an increase in the efficiency of best individual selection. ‘AQUAmax’ is a drought-tolerant maize hybrid that presents a high yield under drought conditions, developed by GS [42,43].

DNA markers based on quantitative trait loci (QTL) permit the establishment of the relationship between drought tolerance-related phenotypic traits and genomic regions, gene clusters, or genetic polymorphism, which provides significant support for drought stress resistance breeding programs in plants [13]. The enhancement of rice yield under drought conditions was carried out by a MAS-backcrossing breeding (MAB) program, which consists of the introgression of the QTLs qDTY1.1, qDTY2.1, and qDTY3.1 into ‘Swarna’ background [44]. Genome-wide association analysis (GWAS) identified the genomic regions that contribute to a target trait. GWAS has led to the identification of QTLs and single-nucleotide polymorphisms (SNPs) associated with higher yields under drought conditions in rice [45,46].

QTL application in MAS breeding programs is still limited because some inconsistencies, such as non-repeatability, can be present. In drought stress, the phenotypic response is highly influenced by many environmental factors. For this reason, only a few of the QTLs identified experimentally are consistent when they are implemented in MAS programs with different environments and water regimens. Furthermore, identified QTL can present low stability after several plant generations [13].

Meta-analysis of QTLs (MQTL) consists of compiling information from many independent studies to detect reliable and stable QTLs, and can also be carried out between different species. Stable QTL related to drought tolerance have been identified in rice, barley, and wheat by MQTL [47,48,49]. In rice, the meta-analysis of many studies carried out in different environments and different genetic backgrounds led to the identification of 61 stable MQTLs for major agronomic traits for adaptation and survival under drought conditions [50].

A haplotype is an allelic combination of genes that are inherited together. Instead of identifying individual allelic combinations, which MAS carries out, haplotype-based breeding led to identifying superior allelic combinations, which makes it more promising for the improvement of drought-tolerant plants [13]. By haplotype-based breeding (HB), superior haplotypes can be identified in silico by genomics, then, ideal crop varieties can be developed in the actual breeding programs [51], for example, a favorable combination of alleles conferring drought tolerance traits, such as early flowering, deeper root system, and higher yield [7].

2.6. High Throughput Phenotyping to Boost Plant Breeding

Conventional traits screening is time-consuming, labor-intensive, and prone to errors. Furthermore, some evaluations require the destruction of the samples. To overcome the limitation of conventional trait screening, high-throughput phenotyping (HTP) offers an automated, faster, multifunctional, more effective, and reliable alternative [8].

HTP or plant phenomics is a robust method used in plant precision phenotyping that enables a fast and consistent non-destructive evaluation of valuable traits in whole-plant [13]. These tools allow automated data collection and processing carried out by digital imaging, remote sensing, and high-performance computing to analyze a high number of morpho-physiological traits, such as biomass, leaf color, leaf area, seed size, and plant height [16,52]. Even though root structure plays a relevant role in drought stress tolerance, most of the screening methods for traits test only plant aerial traits, mainly because of the difficulty of screening below the ground [8]. Novel non-destructive techniques that lead to an appropriate evaluation of below-ground traits need to be developed.

‘TERRA-MEPP’ is a phenotyping robot that carries out the high-throughput evaluation of architectural traits in sorghum [53]. In wheat, a novel HTP platform that uses normalized difference vegetation index (NDVI) data led to the identification of quantitative trait loci (QTLs) based on NDVI related to drought resistance [54]. Nevertheless, HTP technologies are only applied to a few relevant crops, mainly in developed countries. To ensure food security, the implementation of these innovative techniques focusing on neglected and underutilized crops should be stimulated. Furthermore, significant financial investment should be encouraged to increase the application of these techniques in developing countries [4].

The implementation of these fast and precise genotyping and phenotyping tools can reduce the time and improve the identification and development of novel drought-tolerance breeding lines [13]. A summary of the tools available to improve plant breeding programs and their interactions is illustrated in Figure 1. SB can be complemented with these other biotechnological tools to reduce the time and improve the development of high-yielding and drought-tolerant varieties. Nevertheless, their application is still limited due to the lack of infrastructure, technicians, and trained breeders, mainly in developing countries [9].

3. Recombinant DNA Technology

Recombinant DNA technology consists of the insertion of DNA fragments from an exogenous source into a host organism or a host DNA molecule. A unique gene, multiple genes, or regulatory elements can be inserted. These techniques also involve the increasing, decreasing, or blocking of the expression of endogenous genes from the host, as well as the modification of the functionality of genetic regulatory elements, genes, or the proteins they code [55].

Conventional breeding methods involve transferring a large number of genes from a donor plant to a recipient plant to obtain a desirable trait. Nevertheless, it can cause the introduction of genes that do not contribute to the desirable trait or even affect other traits of agronomic relevance. Recombinant DNA technology allows the precise and fast insertion of a specific gene or a specific regulatory element into the plant genome [55]. Agrobacterium infection, clustered regularly interspaced short palindromic repeats with Cas 9-associated endonuclease (CRISPR/Cas 9), and RNA interference (RNAi) are plant gene editing technologies that have the potential to improve plant breeding by enhancing the resistance to drought stress [56,57,58,59]. Several experimental approaches have been developed in this regard. This review focuses on research articles published from 2019 to the present; interested readers in older research articles can consult the following reviews [60,61,62].

3.1. Agrobacterium Infection

Through Agrobacterium infection, the drought tolerance improvement of relevant economic plants, such as cotton [63,64], potato [65], and rice [66], can be carried out in experimental conditions. These experiments include the overexpression of endogenous genes [65,66], the insertion of genes from a wild relative species [64], or from a bacterial source [63,67].

The transformation of Gossypium hirsutum cotton plants with the universal stress protein GUSP1 from the abiotic stress-tolerant species G. arboretum produces an increase in relative water content (RWC), biomass, and number of closed stomata in drought stress conditions in vitro [64]. Transgenic cotton plants transformed with the Na+/H+ antiporter gene K2-NhaD from a halophilic bacterium showed higher drought tolerance than the wild types, which appears to be associated with an increase in water content in leaves, number of lateral roots, levels of chlorophyll, soluble sugar, proline, and the activity of antioxidant enzymes [63].

It has been suggested that superoxide dismutase (SOD) is the primary antioxidant enzyme playing a role in the resistance to abiotic stress in plants. Rose plants (Rosa hybrida) transformed with the SOD2 gene from Escherichia coli showed lower relative ion leakage and higher water content than isogenic lines after seven days without water. Wilted and brown leaves were shown in isogenic lines, whereas transgenic roses showed a healthy leaf phenotype [67].

The overexpression of the Osr4oc1 gene, coding for a lectin, causes lower rolling and brown leaves, higher RWC, and a higher chlorophyll level in transformed rice in soil with a 45–50% moisture level during seven days [66]. Potato (Solanum tuberosum) plants overexpressing the RING-finger ubiquitin ligase E3 gene (StRFP2) showed lower levels of malondialdehyde (MDA) and higher levels of proline, catalase, and RWC than the potato isogenic lines during seven days under drought stress [65].

As shown, relevant insights into the understanding of drought tolerance response in plants have been possible with Agrobacterium infection. Altogether, these studies show that the activation of the enzymatic antioxidant system and the production of osmoprotectants are relevant and common mechanisms in the alleviative response to drought stress in plants. Nevertheless, the development and deployment of drought-tolerant genotypes developed by Agrobacterium transformation to breeding programs is still limited. Vector construction and transformation, as well as in vitro tissue culturing, are time-consuming and need to be standardized for each specific plant species. Additionally, transformation and regeneration are relatively inefficient for some plant species [9]. Furthermore, the use of crops genetically modified by Agrobacterium infection is still controversial, and policy and biosafety considerations need to be discussed.

3.2. Gene Edition by Clustered Regularly Interspaced Short Palindromic Repeats with Cas 9 Associated Endonuclease (CRISPR/Cas 9)

Compared to other genetic editing technologies, CRISPR/Cas9 is considered more efficient and simpler. Due to that, it is expected that it has more social acceptability, which could facilitate the adoption of genetically edited crops for human consumption worldwide [57]. The facility of CRISPR/Cas9 has led to the development of several recent experiments carried out to both elucidate the drought stress response in plants and the design of plants with a drought stress-tolerant phenotype (DS-TP).

CRISPR/Cas9 gene editing has been applied to relevant crop plants, such as Brassica napus, Glycine max, Nicotiana tabacum, Oryza sativa, Solanum lycopersicum, and Zea mays, achieving a drought stress tolerance response in experimental conditions [68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91].

Table 2. Experimental approaches to study the effect on the plant resistance to drought stress by knock-out through CRISPR/Cas9 technology.

Species	Gene Studied	Phenotypic Effect		Ref.
		Increase	Decrease
Brassica napus	BnPUB18 and BnPUB19	Antioxidant activity and survival percentage.	Water loss, MDA content, and electrolyte leakage percentage.	[72]
Glycine max	GmpPLA-IIe and GmPLA-IIz	Survival rate, plant height, plant weight, shoot length, and root length.		[88]
	GmHdz4	Root dry weight, number of root tips, proline and soluble sugar content, and enzymatic activity of catalase, SOD, and peroxidase.	Hydrogen peroxide, superoxide, and malondialdehyde content.	[91]
Nicotiana tabacum	NtAITR	Sensitivity to ABA, expression of an ABA receptor gene, and survival percentage.		[70]
	NtPIF	Survival rate, RWC, proline content, photosynthetic rate, expression of catalase, peroxidase, and SOD genes.	MDA and hydrogen peroxide content. Degree of wilted leaves.	[73]
	NtNAC028	Survival rate, expression of the senescence-related cysteine proteinase 1 genes and the stress-responsive genes RD29A, RD26, DREB1B, and NCED3-2. Activity of SOD, catalase, and peroxidase.	MDA content and velocity of leaf senescence.	[87]
	NtPOD63l9	Activity of peroxidase and SOD. Proline and lignin content. Expression of the stress-related genes RD29A and DREB2A. Recovery and survival rate. Number of interfascicular fibers in stems.	Water loss rate and stomatal opening.	[89]
Oriza sativa	OsDIRH2	Chlorophyll content, shoot fresh weight, and survival rate.	Stomatal density.	[68]
	OsPUB7	Proline content and recovery rate.	Percentage of ion leakage and damage level.	[69]
	OsPUB16	Stomatal density. Sensitive to methyl jasmonic acid and ABA.	Water loss in leaves, the size of stomata, and plant growth.	[76]
	OsERF109	Expression of SOD, catalase, and ascorbate peroxidase genes. RWC in leaves, photosynthesis, concentration of intercellular CO2, proline content, shoot length, root length, and survival rate.	Canopy temperature, velocity of leaf senescence, water loss in leaves, stomatal conductance, and transpiration. MDA, superoxide, and hydrogen peroxide content.	[77]
	OsCKX2	RWC, photosynthesis rate, proline content, and survival rate. Expression of catalase, SOD, and ascorbate peroxidase genes.	Water loss, transpiration, stomatal conductance, and internal CO2 concentration. MDA and endonucleases associated with programmed cell death content.	[78]
	OsNACO92	RWC, chlorophyll content, and survival rate. Expression of SOD, peroxidase, and catalase genes.	Wilting. MDA, superoxide, and hydrogen peroxide content.	[82]
Solanum lycopersicum	SlGT30	Photosynthetic CO2 assimilation rate, fruit weight, and fruit size. Size and number of fruit pericarp cells.	Water loss, number of stomata, electrolyte leakage, wilting symptoms, and leaf rolling. Superoxide and hydrogen peroxide content.	[75]
	SlWOX4	Plant growth, survival percentage, proline content, and root length.	Area stomatal open, water loss, electrolyte leakage, and MDA content.	[71]
	SlHyPRP1 and SlDEA1	Survival rate, chlorophyll content, and proline content.		[80]
	SlALKBH10B	RWC, chlorophyll, proline, soluble sugars, and starch content. Expression of ascorbate peroxidases, catalases, SOD, and proline synthase genes.	Leaves wilting, water loss, MDA, and hydrogen peroxide content.	[81]
	Slald1	Activity of SOD, peroxidase, catalase, ascorbate peroxidase, dehydroascorbate reductase, and glutathione reductase. Stomatal conductance and transpiration rate.	Electrolyte leakage, leaf rolling, and wilting. MDA, hydrogen peroxide, and superoxide ion content.	[86]
Zea mays	ZmGA20ox3	Jasmonic and ABA content. Plant height. Expression of transcription factor genes related to drought stress	Gibberellin GA1 content.	[74]
	ZmPL1	Survival percentage, RWC, root length, root branches, and proline content. Expression of SOD, peroxide, and catalase genes. Expression of stress-related genes ZmLTP3, ZmRD22, and ZmCBF4.	Degree of wilted leaves. Superoxide ion, hydrogen peroxide, and MDA content.	[85]
	ZmASR1	Survival rate, RWC, chlorophyll content, and net photosynthetic rate. Activity of peroxidase, catalase, and SOD.	Stomatal openness.	[90]

Abbreviations: U-box E3 ubiquitin ligases 18 and 19 (BnPUB18 and BnPUB19, respectively); malondialdehyde (MDA); phospholipases A (GmpPLA-IIe and GMPLA-IIz); homeodomain leucine zipper transcription factor (GmHdz4); superoxide dismutase (SOD); ABA-induced transcription repressors (NtAITR); abscisic acid (ABA); phytochrome-interacting factor (NtPIF); relative water content (RWC); NAC transcription factor (NtNAC028); peroxidase-like 63l9 (NtPOD63l9); RING E3 ligase (OsDIRH2); U-box E3 ubiquitin ligase (OsPUB7); class-II U-box E3 ubiquitin ligase gene (OsPUB16); ethylene response factor gene (OsERF109); cytokinin oxidase 2 gene (OsCKX2); NAC transcription factor gene (OsNACO92); trihelix transcription factor (SlGT30); Wuschel homeobox transcription factor (SlWOX4); hybrid proline-rich protein gene (SlHyPRP1); differentially expressed in response to arachidonic acid 1 gene (SlDEA1); AlkB dioxygenase gene (SlALKBH10B); Pipecolic acid biosynthetic pathway gene (Slald1); gibberelin bioactive GA1 biosynthesis gene (ZmGA20ox3); phylloplanin-like gene (ZmPL1); and ABA stress ripening induced gene (ZmASR1).

summarizes the main findings obtained in these experiments carried out from 2019 to the present. In these experiments, a common response mechanism can be identified at the morphological, physiological, metabolic, biochemical, and transcriptional levels.

At the agronomic and morphological level, drought-tolerant plants' response involves the increase of the stomatal density, root length, root weight, and root branches, which appears to have a direct effect in the survival rate, recovery rate, weight, and size of the plants, as well as the weight and size of shoots and fruits. In tomato, the edition of the genes SlHP2 and SlHP3, which play a role in cytokinin signaling, induced better adaptation to drought stress, with larger root length and lower levels of malonaldehyde (MDA), electrolyte leakage, and water loss [79]. The overexpression of gene coding for the heat shock protein 40 (ZmDnaJ) by CRISPR/Cas9 in maize improves drought tolerance by inducing stomatal closure, increasing antioxidant capacity, and photosynthesis [84]. In tomato, the induced mutation of a gene encoding for a polyamine oxidase protein by CRISPR/Cas9 diminishes the wilting symptom, the leaf temperature, and increases the percentage of RWC under drought stress. The results of this experiment suggest that the resistance of the tomato mutants to drought stress was due to the lower transpiration rate, lower stomatal conductance, and reduced tendency to xylem embolism developed [83].

Due to its high adaptability and simplicity, CRISPR/Cas9 is the genome editing technology most used in plant breeding [8]. Nevertheless, it can cut only in a single target genomic site and has been reported that it can present low editing efficiency and off-targeted mutation [9]. Unlike CRISPR/Cas9, CRISPR/Cpf1 and CRISPR/C2c1 are novel developed alternative systems with the advantage of targeting multiple genomic sites and generating staggered ends of DNA double-stranded breaks [7]. Nevertheless, they also present the limitations of off-targeting.

3.3. RNAi Technology

The phenomena of RNA interference (RNAi) in plants were first documented in 1989 during the creation of tobacco transgenic plants with Agrobacterium. It was observed that the inactivation of a transgene encodes a selectable marker [92]. In the following years, it was observed that the inactivation of the chalcone synthase gene in petunia was engineered to overexpress the gene. This phenomenon was designated as co-suppression or co-silencing [93]. After that, the phenomena were observed in other transgenic plants. Furthermore, it was discovered that the phenomenon was induced by double-stranded RNA [94]. It is known that the mechanism is a homology-dependent gene silencing in which a small interfering RNA induces RNA degradation, RNA translation blocking, or DNA methylation [95]. The vectors that can be utilized to carry out the RNAi phenomena can be constructed in different forms, such as hairpin RNA vectors, intron-containing hairpin RNA vectors, or artificial microRNA [96]. Out of these, the intron-containing hairpin RNA vectors had been shown to be most effective in plants [97]. Furthermore, several methodologies have been utilized to deliver the construct into the plant cell [98].

In model species with an optimized protocol, RNAi is quite reproducible, but silencing efficiency varies between target genes, plant species, and uptake mechanisms [99]. Some delivery systems have been developed to increase the silencing success of RNAi assays, such as spray-induced gene silencing (SIGS) and nano-carrier-mediated delivery. SIGS has been successfully applied to both monocots and dicots [100]. Nevertheless, its efficacy depends on the double-stranded RNA structure, plant species surface characteristics, and environmental factors. Large-scale application of RNAi is difficult due to the variability between different environmental and biological systems. SIGS have the potential to facilitate the scalability of RNAi, but environmental dsRNA stability, successful delivery, and cost issues should be addressed [99].

Table 3. Experimental approaches to study the effect on the plant resistance to drought stress by RNA interference (RNAi).

Species	Gene Targeted	Method	Phenotypic Effect		Ref.
			Increase	Decrease
Arabidopsis thaliana	GmSAP5	RNAi hairpin	Degree of leaf wilting and malondialdehyde content.	Survival rate, leaf RWC, proline, and ABA content.	[105]
Betula platyphylla	BpMYB123	RNAi hairpin	Superoxide, hydrogen peroxide, and MDA content. Electrolyte leakage and degree of wilting in leaves.	Proline content. Activity of peroxidase and SOD.	[116]
Brassica rapa	BrDHC1	RNAi intron-containing hairpin	MDA content. Leaf water loss. Yellow color and curled leaves.	RWC, roots length, seedling fresh weight, and stomatal aperture. Proline, soluble sugars, and chlorophyll content. Activity of SOD, catalase, and peroxidase. Expression of the stress-related genes CAO, CHLG, POD, P5CR, MYB44, and ABRE2.	[115]
Glycine max	GmTRAB1	Antisense RNAi intron-containing hairpin	Leaf wilting. MDA and hydrogen peroxide content.	Survival rate and root fresh weight. Proline content. Activity of catalase and peroxidase. Expression of ABA-responsive, antioxidant-related, and drought-induced genes.	[118]
	GmCIPK29	RNAi hairpin	Leaf wilting. MDA, superoxide, and hydrogen peroxide content.	RWC and survival rate. Activity of catalase and peroxidase. Expression of stress-related genes.	[122]
Gossypium hirsutum	GhCYSTM5_A	Antisense RNAi hairpin	Leaf wilting. MDA content.	Proline content. Activity of SOD.	[101]
Malus domestica	MdRAD23D1	RNAi intron-containing hairpin	MDA, superoxide, and hydrogen peroxide content.	RWC, water utilization efficiency, plant height, trunk diameter, biomass accumulation, and photosynthetic activity. Root weight and root hydraulic conductivity. Activity of SOD, peroxidase, and catalase	[111]
	MdWRKY50	RNAi hairpin	Leaf wilting. Reactive oxygen species and MDA content.	. Photosynthetic rate, fresh weight, and leaf area. Expression of the anthocyanin biosynthetic genes MdCHS, MdCHI, MdF3H, MdF3’H, and MdDFR.	[114]
Medicago sativa	MsDIUP1	RNAi hairpin	MDA content.	Proline, soluble sugar, and chlorophyll content.	[112]
	MsLEA1	Antisense RNAi intron-containing hairpin	Wilting and yellowish color in leaves. Stomatal aperture and stomatal conductance. Electrolyte leakage and MDA content.	Fresh weight, water use efficiency, photosynthetic rate, transpiration rate, rubisco activity, and hydrogen peroxide content.	[121]
Medicago truncatula	MtKNOX3-like	RNAi intron-containing hairpin	Senescence symptoms, water loss, and ion leakage in leaves.	Chlorophyll content in leaves. Proline content and expression of the proline dehydrogenase gene.	[125]
Nicotiana tabacum	NtNCED3	RNAi *		Photosynthesis. Shoot and root development.	[104]
Oriza sativa	MYB (Os04g0583900)	RNAi intron-containing hairpin	Shoot height, root length, and proline content.	MDA content and relative electric conductivity.	[108]
	OsYABBY6	Antisense RNAi	Water loss and survival rate. Soluble sugar and proline content. Expression of a receptor-like kinase that activates the antioxidant system.	MDA, superoxide ion, and hydrogen peroxide content. Expression of three NADPH oxidase genes.	[117]
Populus alba x P. glandulosa	PagTPS1 and PagTPS10	RNAi hairpin	Leaf wilting.	Trehalose content and expression of the genes TPS and trehalose-6-phosphate phosphatase. Expression of drought-responsive genes.	[120]
Populous tremolo × P. Alba	ERF194	RNAi *	MDA content.	RWC, number of leaves, and root length. Activity of catalase, peroxidase, and SOD. Expression of stress-related genes.	[107]
Populus trichocarpa	DREB46	RNAi hairpin	Plant height and leaf wilting.	RWC and survival rate. Activity of SOD, catalase, and peroxidase. Chlorophyll, proline, and lignin content.	[123]
	PdNCED3	RNAi intron-containing hairpin	Stomatal aperture. Curling and rolling in leaf. Hydrogen peroxide content. Expression of the Abscisic acid-insensitive 5-like protein 2 gene.	ABA content and expression of ABA signaling transcription factor genes.	[106]
Solanum tuberosum	StMAPKK5	RNAi by amiRNA	Leaf wilting and MDA content.	Leaf RWC, plant growth velocity, and proline content. Activity of catalase, SOD, and peroxidase.	[103]
	StERF79	RNAi by amiRNA	Water loss, leaf wilting, and MDA content.	RWC and proline content. Activity of SOD, peroxidase, and catalase. Expression of the dehydrin gene StDHN-2.	[109]
	StJAZ23	RNAi miRNA		Height, number of leaves, root area, root length, and root tips. Jasmonic acid and ABA content. Activity of SOD, peroxidase, and catalase.	[110]
Triticum aestivum	TaFBA1	Antisense RNAi hairpin	Water loss, leaf wilting, and electrolyte leakage. Proline, MDA, superoxide, and hydrogen peroxide content. Activity of the stress-inducible delta 1-pyrroline-5-carboylate synthetase.	RWC, plant height, weight of seeds, net photosynthetic rate, transpiration, stomatal conductance, chlorophyll content, soluble sugars content, and aquaporin activity. Activity of SOD, catalase, ascorbate peroxidase, peroxidases, glutathione peroxidase, monodehydroascorbate reductase, and dehydroascorbate reductase. Expression of antioxidant-related and stress-induced genes.	[119]
Vitis vinifera	VyUSPA3	RNAi intron-containing hairpin	Leaf wilting and the number of dead leaves. MDA, hydrogen peroxide, and peroxide ion content.	Activity of SOD, catalase, and peroxidase. Expression of the stress-related genes RD22, RD29B, DREB2A, and NCED1.	[102]
Zea mays	bHLH	RNAi *		Starch content, activity of NAD-malic enzyme, and RUBISCO.	[113]
	CgbHLH00	Antisense RNAi hairpin	Leaf wilting. Hydrogen peroxide, superoxide, and MDA content.	Proline and soluble sugar content. Activity of catalase, peroxidase, and SOD. Expression of drought-inducible genes.	[124]

Abbreviations: Stress associated zinc finger protein 5 (GmSAP5); relative water content (RWC); abscisic acid (ABA); R2R3-MYB transcription factor 123 (BpMYB123); malondialdehyde (MDA); superoxide dismutase (SOD); DEAD-box helicase 1 (BrDHC1); basic leucine zipper transcription factor gene 1 (GmTRAB1); calcineurin B-like protein interacting protein kinase 29 (GmCIPK29); cysteine-rich transmembrane module gene 5 (GhCYSTM5_A); ubiquitin-like-ubiquitin-associated gene 23D1 (MdRAD23D1); WRKY transcription factor 50 (MdWRKY50); drought-induced unknown 1 (MsDIUP1); late embryogenesis abundant protein 1 (MsLEA1); KNOTTED1-LIKE homeobox transcription factor (KNOX3-like); 9-cis-epoxycarotenoid dioxygenase 3 (NtNCED3); transcription factor family YABBY (OsYABBY6); trehalose-6-phosphate synthase genes 1 and 10 (PagTPS1 and PagTPS10, respectively); AP2/ERF transcription factor 016 (ERF016); ethylene responses factor 194 (ERF194); drought-responsive element binding gene 46 (DREB46); mitogen-activated protein kinase 5 (StMAPKK5); artificial microRNA and microRNA (amiRNA and miRNA, respectively); AP2-ERF transcription factor 79 (StERF79); Jasmonate-ZIM domain gene 23 (StJAZ23); F-Box gene A1 (TaFBA1); universal stress protein A3 (VyUSPA3); basic helix-loop-helix transcription factor (bHLH); and bHLH from Chenopodium glaucum (CgbHLH001). * Description of construct not provided.

summarizes the main findings obtained in experiments involving RNAi gene silencing carried out from 2020 to the present to elucidate the drought stress response mechanisms in plants [101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125]. A concordance can be observed between the findings previously described for CRISPR-Cas9 experiments (Table 2) and the findings reported for RNAi. Most of these RNAi experiments are carried out in order to silence the expression of genes tentatively having a role in drought stress tolerance in plants, which results in a drought stress-sensitive phenotype (DS-SP). Like CRISPR-Cas9 experiments, the results show a common mechanism that involves morphological, physiological, metabolic, biochemical, and transcriptional changes in response to drought stress, which supports a common mechanism of response to stress existing in plants.

At the biochemical and metabolic levels, a decrease has been observed in the metabolites proline, soluble sugar (such as trehalose), starch, and lignin [111,123,124], as well as a decrease in the phytohormones jasmonic acid and ABA associated with the DS-SP [105,106,110]. Additionally, a decrease in the activity of the enzymatic antioxidant system is shown. In RNAi studies, plant drought stress sensitivity has been mainly associated with a decrease in the activity of SOD, catalase, and peroxidase [115]. However, the reduction of ascorbate peroxidase, glutathione peroxidase, monodehydroascorbate reductase, and dehydroascorbate reductase has also been reported [119].

MDA is widely reported as an indicator of oxidative stress and cell membrane damage in plants. As can be seen in Table 3, the increase in MDA content is frequently associated with the DS-SP. An increase in the degree of wilting in leaves, electrolyte leakage, and the content of superoxide, hydrogen peroxide, and other reactive oxygen species are frequently present with the increase in MDA content in DS-SP [116,118,122].

At the transcriptional level, an association between the DS-SP and a decrease in the expression of drought-responsive and drought-inducible genes has been reported [118]; to these belong the genes RD22, RD29B, DREB2A, NCED1, CAO, CHLG, POD, P5CR, MYB44, and ABRE2 [102,115]. Additionally, results show that the DS-SP could be associated with a reduction in the expression of ABA signaling transcription factors and ABA-responsive genes [106,118], as well as the anthocyanin biosynthetic genes CHS, CHI, F3H, and DFR [114].

It has been shown that chlorophyll content also increases in DS-TP compared with the sensitive genotypes [112,115,125]. However, it could be an indirect effect of the homeostasis maintenance carried out by other crucial mechanisms. Data suggest that the DS-TP is determined by the capacity of the plant to maintain its growth, RWC, and photosynthetic capacity in water-limited conditions. Antioxidant mechanisms, osmoprotectant accumulation, phytohormone regulation, transpiration regulation, and the enhancement of the root system appear to be the crucial mechanisms that determine the capacity to maintain a physiological water status and the photosynthetic capacity. Consequently, the healthy growth and development of the plant and its organs are achieved.

Drought stress response and tolerance are complex issues involving intricate mechanisms at all levels of an organism’s metabolism and are influenced by numerous environmental and genotypic factors. Thus, the engineering of plants using molecular tools should focus on improving not one but several of these mechanisms to successfully develop a drought stress tolerance phenotype. Moreover, recombinant DNA technologies should be integrated with novel breeding tools to get a successful release of drought stress-tolerant genotypes to the field.

3.4. Other Novel Genetic Engineering Tools

The utilization of ribonucleoprotein (RNP) delivered by physical methods holds the advantage of creating an edited plant transgene-free, which can help to avoid the endless discussion about the potential dangers of transgenic plants [126]. Additionally, it has been demonstrated that the use of CRISPR-RNP complex reduces toxicity and minimizes off-target effects, which makes the use of RNPs a promising tool for plant genetic engineering [127]. In tobacco (Nicotiana tabacum), the editing of the actin depolymerizing factor gene (ADF) by an RNP complex increases RWC under drought stress, suggesting that ADF acts as a key regulator of osmotic stress tolerance in plants [128]. Nevertheless, the implementation of RNPs is still limited to some species because plant regeneration after editing remains challenging, and there is a need to improve the editing efficiency [127].

Long non-coding RNAs (lncRNAs) are transcripts larger than 200 nucleotides that do not code for proteins, but have relevant regulatory roles in plant responses to abiotic stress, including drought [129]. Competitive endogenous RNAs (ceRNAs) are lncRNAs that regulate gene expression through competing binding to regulatory miRNAs [130]. Circular RNAs (circRNAs) are ceRNAs formed by head-to-tail binding of lncRNAs [131]. It has been reported that circRNAs have a more stable structure and higher sequence conservation than linear lncRNAs. Additionally, the inhibition of miRNAs by ceRNAs can be transitory [132]. Due to this, ceRNAs are promising candidates for the development of novel transgene-free crop treatments [133].

The enhancement of drought tolerance by the activity of ceRNAs has been reported in Arabidopsis thaliana [134], wheat [135], and rice [136]. In Arabidopsis, guard cell outward-rectifying K+-channel (circGORK) enhances drought tolerance by regulating the expression of genes related to ABA signaling [134]. In wheat, lncRNA35557 inhibits the activity of miR6206, increasing the abundance of TaNAC018 and promoting drought tolerance, tentatively, by increasing cuticular wax content [135]. In rice, TCONS_00021861 prevents the repression of YUCCA7 by miR528-3p, increasing the biosynthesis of indole-3-acetic acid and auxin and conferring resistance to drought stress [136]. Nevertheless, plant lncRNA research is in its beginning, and the function of only a few lncRNAs has been well characterized [133]. Additionally, an appropriate delivery system needs to be designed for the success of the ceRNAs treatment in plants.

4. Limitations and Future Directions

A field is a complex system that involves many simultaneous biotic and abiotic stresses [6]. Most of the experiments carried out in edited plants to identify genes that improve drought stress tolerance are conducted in laboratory chambers or a glasshouse. Some authors argue that a single laboratory experiment is insufficient to determine the ability of a gene to confer resistance to water stress [8]. It has been shown that genes identified experimentally to improve drought tolerance have only minor effects in field conditions [137].

In field conditions, drought stress is a dynamic and unpredictable factor that varies in severity and timing [13]. Studies in which a single gene is modified to improve drought resistance in experimental conditions are much simpler than the phenomena that can be observed in the field, mainly because other environmental conditions are present with drought stress, such as heat and salinity stress. Assays including simultaneously heat and salt stress must be performed to reliably determine whether transgenic plants can avoid drought in field conditions [8].

Plant drought stress tolerance is a complex polygenic trait. Additionally, plant phenotypes display plasticity determined by the specific environmental conditions [6]. In the field, drought could range from mild to severe and can occur at any stage of plant development. Furthermore, response to drought stress depends on management practices, soil constitution, and microbiome [7]. Therefore, plant drought resistance improvement should be conducted in field conditions and should include a broad spectrum of drought stress over an extended period [8]. The identification of common traits associated with drought-tolerant genotypes growing in different field conditions could help to identify target traits for a custom design of drought-tolerant crops in the field [7].

To cope with the complexity of plant drought response, the analysis of QTLs and genes related to drought stress tolerance should be carried out in the target environment [13]. Nevertheless, QTL application in MAS breeding programs is still limited because inconsistencies, such as non-repeatability, can be present [8]. Furthermore, identified QTL can present low stability over plant generations and years. Instead of identifying individual allelic combinations, such as those carried out by MAS, haplotype-based breeding is more promising for improving drought-tolerant plants. The phenotypic analysis could be carried out to validate the haplotype-based breeding by a process called haplo-pheno-analysis. Furthermore, markers linked to superior genotypes based on haplotypes can be carried out through haplo-based selection [13].

In drought-tolerant plant breeding programs, GS can assist SB to improve and accelerate the selection of desirable candidates by a method named Speed-GS [13]. Moreover, HB can be integrated with GS in SB programs to boost the identification of crop varieties with high performance under drought conditions [7]. However, SB requires the optimization of specific protocols for each plant species. Additionally, SB still lacks equipment and trained technicians [9], which limits its use on a commercial level and in developing countries.

Compared to traditional agricultural methods, genome editing offers a simpler, efficient, and specific alternative to plant breeding. Speed breeding can be coupled with gene editing tools to reduce the amount of lab work and shorten the generation development [4,9]. Although genomic editing by CRISPR/Cas9 does not involve the introduction of foreign DNA, the legal status of CRISPR-edited crops varies globally, and their use still requires risk evaluation in many countries [98]. Furthermore, the possibility of cross-pollination with non-edited relatives and their effect on non-target organisms needs to be assessed [127]. Regulatory concerns, low public acceptance, and a lack of specialized technicians limit the deployment of CRISPR/Cas9 edited drought-tolerant plants, mainly in developing countries [138].

To develop novel drought-tolerant genotypes, multidisciplinary breeding is the appropriate experimental approach. Conventional breeding needs to be integrated with modern breeding tools for accelerating the improvement of drought-tolerant high-yielding crops in the near future [6]. Nevertheless, plant drought tolerance research has mainly been carried out in the main crop species. For crop diversification, future studies should be carried out in wild relatives, neglected, and underutilized species [10], which still lack an efficient method for drought tolerance screening. High-throughput phenotyping and genotyping platforms integrated with the SB technique should be encouraged to exploit the genetic diversity of wild relatives, landraces, neglected, and underutilized species [113]. Drought stress is a significant drawback for the transition to sustainable agriculture in developing countries, inducing significant economic losses [2,3], which suggests the need to have more efficient water management to improve agricultural sustainability under conditions of climate change. A holistic focus by combining socioeconomic policies, agrotechnological, and biotechnological practices to address water scarcity should be carried out. In developing countries, support at academic, industrial, and institutional levels should be given to increase the application of these innovative techniques [4], which could encourage the development of low-cost, precise phenotyping and genotyping platforms coupled with SB to accelerate the identification of promising genotypes [113]. Nevertheless, how the developed tolerant plant will evolve over generations after genetic alteration is still an essential issue to be resolved [137].

5. Concluding Remarks

In this review, the efforts to design plants with a phenotype of water stress resistance by using traditional genetics, screening of water stress traits, and the latest developed breeding tools were analyzed. To cope with the complexity of plant drought response, conventional breeding needs to be integrated with novel biotechnological tools for accelerating the improvement of drought-tolerant, high-yielding crops in the near future. In this review, efforts involving novel tools to accelerate breeding cycles, enhance the selection of superior genotypes, and engineer crops to achieve drought stress tolerance were analyzed. Findings suggest that the drought stress tolerance is determined by the capacity of the plant to maintain both water content and photosynthetic capacity for growth and survive in water-limited conditions. Furthermore, this homeostasis appears to be mainly determined by the plant’s capacity to activate an effective antioxidant and osmoprotectant defense, regulate its transpiration efficiently, and develop an effective root system. Furthermore, phytohormone regulation could play a pivotal role in the effective activation of these mechanisms.

Despite the actual technologies developed, the production of drought stress-resistant high-yielding plants is still insufficient to cope with the increasing human population and drought stress predicted for the next decades. To achieve sustainable agriculture, more efficient water management should be carried out, which requires a holistic focus combining socioeconomic policies, agrotechnological, and biotechnological practices to face the low water availability. Furthermore, efforts should be made to improve the techniques for the plant phenotype design by increasing the knowledge about the different genetic elements playing a role in regulatory mechanisms. Transgenic-free engineering tools, such as RNAi and ceRNA mediated by sprays and other environmentally friendly delivery systems, could be integrated with speed breeding to encourage the exploitation of the genetic diversity of wild relatives, landraces, as well as neglected and underutilized species, which still lack an efficient method for breeding, mainly in developing countries. It is hoped that this will help to achieve the development of sustainable agriculture in the future.

As the amount of scientific data generated increases, soon, it will likely be possible to find the set of genetic elements that need to be modified for the design of a water stress-resistant plant. In turn, all these accomplishments mentioned will be a great help in improving the current developmental phase of sustainable agriculture.