Engineering drought tolerance in plants through CRISPR/Cas genome editing

Abstract

Drought stress is primarily responsible for heavy yield losses and productivity in major crops and possesses the greatest threat to the global food security. While conventional and molecular breeding approaches along with genetic engineering techniques have been instrumental in developing drought-tolerant crop varieties, these methods are cumbersome, time consuming and the genetically modified varieties are not widely accepted due to regulatory concerns. Plant breeders are now increasingly centring towards the recently available genome-editing tools for improvement of agriculturally important traits. The advent of multiple sequence-specific nucleases has facilitated precise gene modification towards development of novel climate ready crop variants. Amongst the available genome-editing platforms, the clustered regularly interspaced short palindromic repeat-Cas (CRISPR/Cas) system has emerged as a revolutionary tool for its simplicity, adaptability, flexibility and wide applicability. In this review, we focus on understanding the molecular mechanism of drought response in plants and the application of CRISPR/Cas genome-editing system towards improved tolerance to drought stress.

Introduction

Improvement of agriculturally important traits in crops to make them resistant to biotic and abiotic stresses has been an uninterrupted activity globally for a very long time. Adapting effective approaches and novel technologies is the need of the hour to sustain crop yield and minimize the effects of global warming and climate change. Of among the various environmental cues, drought is the main cause of agricultural loss globally, and represents a major threat to food security (Lobell and Gourdji 2012). It differs territorially, provisionally and in effects. Thus, plants have varied response and have improved to display several structural and physiological performance and such behaviour comprise of various degrees of drought escape, avoidance and tolerance (Fahad et al. 2017). Traditional breeding and transgenic methods were successful to improve drought tolerance in crops including rice, wheat, maize and soybean (Ashraf 2010). However, most of these promising lines are not able to produce high yields in water-deficit conditions. It indicates that there is still a lot of scope for novel approaches that can uncouple drought stress from plant growth.

All these years, conventional breeding has been the most effective way of developing crops and has facilitated their growth in drought conditions. However the methods are time consuming, laborious and cost intensive. Molecular marker technology played a tremendous role in characterizing the wide array of genetic sections of plants under stress environments (Rao et al. 2016). Several quantitative trait loci (QTLs) responsible for drought tolerance in different crops have been identified (Khan et al. 2016). However, certainty and correctness in QTL recognition are complicated. Genetic engineering was quite effective in improving the crops against biotic and abiotic stresses. However, the social and ethical issues related to it and the biosafety regulations create a hindrance in the acceptance of these modified crops (Prado et al. 2014). Thus, there is a need for novel technologies that can improve the stress conditions in plants and make them climate ready. The advent of genome-editing technologies has revolutionized the field of agriculture. Genome editing tools use sequence-specific nucleases to make precise modifications in genome (Costa et al. 2017). Amongst the available genome-editing platforms, CRISPR/Cas system is widely accepted because of its simplicity and adaptability. The CRISPR/Cas system make use of a complex consisting of a Cas9 endonuclease and a single guide RNA (sgRNA) that moves along the DNA strand and makes a double-stranded break (DSB) on the DNA. The breaks are subsequently repairs by the cells endogenous repair mechanisms leading to the development of novel mutants (Voytas and Gao 2014). In the recent times, CRISPR/Cas technology has been efficiently used for the realization of tolerance against multiple abiotic stresses including salinity, submergence and drought in major crops (Shi et al. 2017; Zhang et al. 2019). The present review focuses on the applications of CRISPR/Cas9 system in crops towards drought tolerance and discuss on the future prospects of this technology towards the development of drought-tolerant crop varieties.

Plants response to drought stress

Drought stress is a climatic situation that occurs due to low level of water in the soil characterized by low or lack of water and continuous water loss through rapid evaporation. The dehydrated condition in turns affects plant growth and development. Plants adapt multiple morphological, biochemical, physiological and molecular mechanisms to survive under water-deficit conditions (Fang and Xiong 2015). Four possible mechanisms govern the plants resistance to drought stress-drought avoidance (DA), drought tolerance (DT), drought escape (DE) and drought recovery (DR) (Fang and Xiong 2015). Drought avoidance is a condition in which plants are able to maintain relatively higher tissue water content and normal physiological processes by adjusting morphological structures despite low water content in the soil (Luo 2010). DA is basically achieved through stomatal closure, wax accumulation, limited vegetative growth, including number and size of leaves and enhanced water uptake through a well-developed root system to avoid dehydration. Drought escape is a situation where the plants naturally or artificially adapt to water stress through modified life cycle or growth period before the onset of drought. Similarly, DR involves plant’s ability to restore growth and vigour after severe exposure to drought stress. On the contrary, drought tolerance denotes plant’s ability to carry out physiological activities under severe drought conditions through the regulation of stress responsive genes and signalling pathways (Fang and Xiong 2015). Drought tolerance is realized by the exertion of one or a combination of these mechanisms in diverse plants under different developmental stages.

Drought tolerance is a complex quantitative polygenic trait and it is crucial to understand the molecular and physiological mechanisms underlying it. Morphological and physiological traits, including water potential, proline content, abscisic acid (ABA) content, root traits, leaf traits and osmotic adjustment capabilities have been used as indicators to evaluate the drought resistance of plants (Fang and Xiong 2015). Additional mechanism such as osmotic adjustment, antioxidation and osmo-protection also enable plants to tolerate the low water content situation (Luo 2010). The morphology of the leaf is also an important agronomic trait in generating drought-tolerant genotypes (Walter et al. 2009). Previous studies have confirmed that rolled leaf genotypes results in reduced rate of water loss and enhance drought tolerance (Xiang et al. 2012). A generalized mechanism of drought tolerance is depicted in Fig. 1. Various signalling molecules, including reactive oxygen species (ROS), calcium, ABA and allied phytohormones and cross talk between different factors play a critical role in signal transmission in response to dehydration stress (Hu and Xiong 2014) (Fig. 1). The phytohormone biosynthesis and subsequent signalling results in the expression of multiple dehydration responsive genes encoding ion transporters, calcium dependent protein kinases (CDPKs), calcineurin interacting protein kinases (CIPKs), mitogen-activated protein kinases (MAPKs), sucrose non fermenting protein (SNF1)-related kinase 2 (SnRK2), calcineurin B-like interacting protein kinase (CIPK) and transcription factors (Fang and Xiong 2015). Up-regulated expression of OsCDPK7 and OsCIPK23 positive regulate drought tolerance in rice (Yang et al. 2008). The MAP kinase cascade, especially the OsMPK5 and MAP kinase kinase kinase (M3K) gene DSM1 in rice have been identified as a crucial molecule implicated in regulating drought tolerance (Sinha et al. 2011). Arabidopsis SnRK2C confers drought tolerance by regulating the expression of stress responsive genes (Umezawa et al. 2004). Various transcription factors, including the AP2/EREBF, AREB/ABFs, MYB, NAC and zinc-finger transcription factors are predicted to be involved in exhibiting tolerance to dehydration in plants (Joshi et al. 2016) (Fig. 1). AREB1, AREB2 and AREB3 have demonstrated a coordinated ABA-mediated positive regulation of drought tolerance in Arabidopsis (Yoshida et al. 2010). Likewise, an AP2/EREBF TF SHN have reported enhanced drought tolerance by activating the biosynthesis of wax in Arabidopsis (Aharoni et al. 2004).

Fig. 1.

A generalized molecular mechanism of drought tolerance in plants. Water scarce conditions activates reactive oxygen species (ROS), calcium ions (Ca⁺), and phytohormones ABA, ET and JA. The phytohormone biosynthesis and subsequent signalling results in the expression of ion transporter genes, protein kinases and transcription factors. Expressive TFs activates the transcription of downstream effector proteins to induce drought tolerance. ABA and Ca⁺ reduce water loss through stomatal closure. Antioxidant enzyme activity causes ROS detoxification. ABA abscisic acid, JA jasmonic acid, ET ethylene, CDPKs calcium dependent protein kinase, CIPK calcium induced protein kinase, MAPKs mitogen-activated protein kinase, PP2Cs protein phosphatase 2C, PYLs pyrabactin resistance 1-like proteins, EIN2 ethylene insensitive 2, SCF SCF complexes, COI1 coronatine insensitive 1, JAZ, SnRK2 sucrose non fermenting 1 (SNF 1)-related protein kinase 2

Among the phytohormones associated with signalling cues, ABA is the most closely linked to drought stress. ABA-mediated drought tolerance in plants is regulated by the coordinated interaction between three classes of proteins, including, (a) the Pyrabactin Resistance 1 (PYR1) and/or PYR1-like protein (PYL) and/or Regulatory component of the ABA receptor (RCAR) (herewith referred as PYLs), (b) Protein phosphatase 2C (PP2C) and (c) SnRK2s (Joshi et al. 2016). In the absence of ABA, PP2Cs are associated with SnRK2s keeping the kinases inactive by dephosphorylating the activation loop. Under dehydrated condition, ABA binds with PYLs thereby inhibiting the phosphatase activity of PP2C. As a result, PP2C releases the SnRK2s which are autophosphorylated and in turn phosphorylate many downstream effectors leading to drought tolerance (Fig. 1). Modification of key enzymes in the ABA biosynthetic pathways has resulted in induced plant growth and enhanced drought resistance (Park et al. 2008). ABA and Ca⁺ together with the osmolytes (sorbitol, proline, mannitol and glycine betaine) and osmoprotective proteins (late embryogenesis abundant protein) reduce water loss through stomatal closure. The effector proteins in turn activate the antioxidant enzymes such as superoxide dismutase (SOD), ascorbate peroxidase (APX) and glutathione peroxidase (GPX) that facilitate ROS detoxification. Understanding of these molecular mechanisms has enabled the improvement of drought resistance in plants through conventional breeding and transgenic technologies. In the last few decades, classical breeding, molecular marker technology and genetic approaches have played a tremendous role in deciphering the molecular mechanism of drought resistance in plants (Kulkarni et al. 2017; Cao et al. 2017; Sahebi et al. 2018). QTLs linked to root and leaf structures and physiological traits have been mapped, hundreds of genes responsive to drought stress have been identified through RNA sequencing and many of them have been validated for conferring drought resistance or tolerance in plants (reviewed in Hu and Xiong, 2014). However, conventional breeding being labour intensive and time consuming and the transgenic plants demands stringent regulatory clearances, plant breeders are now increasing incline at using new technologies such as genome editing that could facilitate efficient and rapid development of drought resistance in crops.

An overview of genome-editing technology

Genome editing refers to a group of advanced molecular techniques that brings about precise, targeted modification at specific genomic loci (Zhang et al. 2018). These techniques make use of sequence-specific nucleases (SSNs) that recognizes specific DNA sequences and introduce double-stranded breaks (DSBs) at the specific targeted sites. The DSBs are fixed through the plants endogenous DNA repair systems-non-homologous end joining (NHEJ) and homologous recombination (Chen and Gao 2014). While the error prone NHEJ method produces insertion or deletion of nucleotides causing gene knockouts, the HR pathway in the presence of donor DNA template results in precise base modification or gene replacement (Chen and Gao 2014). For a successful genome-editing experiment, first a suitable genome-editing tool is selected and the target sequences are designed and introduced into appropriate vectors (Fig. 2). The applicable genome-editing cargo consisting of DNA, RNA or ribonucleoproteins (RNPs) is introduced into the plant using suitable delivery methods. After entering into the plant cell, the genome-editing payload will modify the target sequences. Subsequently, the transformed calli will be screened for desired mutants which will eventually give rise to edited plants. Several gene knockout mutants as well as some insertion and gene replacement mutants have been developed through the genome-editing technologies in a wide variety of plants, most of which have shown to be highly useful for crop improvement (Zhang et al. 2018). Unlike the genetically modified crops, the genome edited mutants have only a few nucleotide modifications similar to the variations found in the naturally occurring populations. After the segregation of the plants containing the Cas9 or other genome-editing agents, the gene edited lines are indistinguishable from the naturally occurring mutations. Therefore, genome-editing technologies have become new players of the current breeding programmes for the precise improvement of crops against multiple biotic and abiotic stresses.

Fig. 2.

Strategy for CRISPR/Cas genome editing in plants. Target site(s) within the target gene(s) are used to design 20 nt oligonucleotide crispr RNA (crRNA) which binds with trans activating crispr RNA (tracrRNA) to form the single guide RNA (sgRNA). The sgRNA together with sequence-specific nuclease like Cas9 constitute the gene editing construct or the ribonucleoprotein complex. The expression construct is introduced into cell/plants using a suitable delivery system. Post transformation, the mutants are generated through tissue culture approach and target specific mutations are analysis through sequencing, restriction enzyme, target genome re-sequencing or T7E1 assay. Subsequently, the gene edited plants are subjected to other analysis

ZFNs consists of zinc-finger DNA binding modules fused with the non-specific DNA cleavage domain of the FokI restriction enzyme. An archetypical ZFN contains approximately 30 amino acids in a conserved ββα configuration. A pair of ZFNs are assembled to bring in two FokI cleavage domain together to create a DSB at the target locus (Bogdanove and Voytas 2011). ZFNs have been used to alter multiple target genes in several plants including the crops like rice, maize and soybean (Martínez-Fortún et al. 2017). However, the complicated chemistry of designing this SSN together with high off-target activity and multifaceted interaction with the target sequence has limited its usage. Similar to ZFNs, TALENs are also a fusion structure of the FokI cleavage domain and the transcription activator-like effector (TALE) repeats acting as the DNA binding domain (Boch et al. 2009). However, the nucleotide specificity of individual TALE repeats provides greater flexibility in target design and selection as compared to ZFNs. TALENs have been successfully utilized in the improvement of many plant species, including rice, maize, barley, soybean, potato, wheat, etc. (Martínez-Fortún et al. 2017). Nevertheless, the comprehensive application of TALENs in targeted editing is highly limited due to the complex protein chemistry involved in the designing of the extensive repeat structure of TALE DBD and variable efficiency of gene targeting. Unlike ZFNs and TALENs, the simplicity of the SSN design, high efficiency and the ability to target multiple sites has made the CRISPR/Cas technology a widely accepted genome-editing tool by the scientific community.

CRISPR/Cas genome-editing system

This CRISPR/Cas system is based on an adaptive immune system found in bacterial and archaeal genomes to protect against the invasion of foreign plasmids or viral DNA (Marraffini and Sontheimer 2010). It is a two component editing system comprising of a Cas9 endonuclease derived from Streptococcus pyogenes and a single guide RNA (sgRNA) molecule that confers target specificity (Hsu et al. 2014) (Fig. 3). The binding of the Cas9-sgRNA complex to the target DNA and the subsequent cleavage depends on the presence of a protospacer adjacent motif (PAM) sequence (5′-NGG-3′) in the downstream of the target region. Thus, the requirement of only different spacer sequences makes it a very simple, rapid, inexpensive and highly efficient editing tool that has been greatly exploited in recent times for crop improvement (Zhang et al. 2018). As of now, the CRISPR/Cas9 system has been used for editing of several agronomically important traits in myriads of plant species, including rice, wheat, maize, barley, sorghum, potato, soybean, lettuce, grapes, cucumber and watermelon (reviewed in Jaganathan et al. 2018). While, trait improvement through gene knockouts and replacement have been highly successful using CRISPR/Cas9, the specificity of 5′-NGG-3′ PAM sequence has restricted its use to potential target sites. To overcome this problem, multiple Cas variants with different PAM specificities have been identified for engineering beyond CRISPR/Cas9 (Wrighton 2018). SpCas9-NAG and xCas9 with broadened PAM compatibility and higher specificity than SpCas9 have been effectively used in rice (Meng et al. 2018; Hu et al. 2018). Among others, the type V CRISPR/Cas12a system has greatly broadened the horizon for CRISPR/Cas editing. The Cas12a or Cpf1 endonuclease derived from Francisella novicida (FnCpf1) and its ortholog from a Lachnospiraceae bacterium (LbCpf1) identifies a T-rich (5′-TTTN-3′) PAM and creates DSBs with cohesive ends of four to five nucleotide overhangs and greatly overcomes the target restriction associated with the CRISPR/Cas9 system (Zetsche et al. 2015) (Fig. 3). The CRISPR/Cas12a system has demonstrated efficient, targeted mutagenesis in Arabidopsis, tobacco, rice and soybean (Endo et al. 2016; Tang et al. 2017). Recently, new Cas12a variants have been developed that recognizes TYCV PAM sequences and facilitate multiplex editing of target genes (Li et al. 2018). For efficient implementation of targeted genome editing, the CRISPR/Cas complex could be delivered into the plants in the form of DNA, RNA or ribonucleoproteins. In the recent times, the dead Cas9 (dCas9) and Cas12a has also been used to expand the scope of genome editing in the development of mutant libraries, gene regulation and epigenetic modifications (Adli 2018). The readers may refer to many recent reviews for a detailed account of the structure and mechanisms of different CRISPR/Cas genome-editing platforms and their application in multiple crop trait improvement (Mishra et al. 2018; Zhang et al. 2018; Jaganathan et al. 2018; Sedeek et al. 2019).

Fig. 3.

The CRISPR/Cas platforms used for genome editing. a The CRISPR/Cas9 recognizes a G-rich (5′-NGG-3′) protospacer adjacent motif (PAM) site at the proximal end and cleaves the DNA by creating blunt ended double-stranded break (DSB). The CRISPR/Cas12a recognizes a T-rich (5′-TTTN-3′) PAM site at the distal end and creates a cohesive ended DSB. b The DSBs are fixed by plant’s endogenous repair system through non-homologous end joining (NHEJ) or homologous recombination (HR). NHEJ process of repairing DSB introduces insertions/deletions (InDels) which results in gene inactivation either by premature stop codon or frame shift mutations. HR results in gene replacement and insertions in presence of a donor DNA template

Genome editing strategies towards drought tolerance

Several abiotic stresses have significantly limited crop yields worldwide by reducing plant growth and development (Pandey et al. 2017). Owing to the complex nature of drought stress, the usage of genome editing towards drought tolerance has been demonstrated only recently (Table 1). Over expression of several genes and transcription factors associated with drought signalling facilitate the accumulation of signalling molecules and metabolites and enhance drought tolerance in plants (Fang and Xiong 2015). Alternatively, expression of sensitive (S) genes, enhances drought conditions in plants through induced ROS production, reduced antioxidant activity and hormonal imbalance. For example, Oryza sativa stress related ring finger protein 1 (OsSRFP1), drought induced SINA protein 1 (OsDIS1) and drought and salt tolerant protein 1 (OsDST) all functions as negative regulator of drought stress and their silencing resulted in induced drought tolerance through enhanced antioxidant enzyme activity and lower H₂O₂ levels (Huang et al. 2009; Ning et al. 2011; Fang et al. 2015). Therefore, natural tolerance to drought could be demonstrated by genome-editing approaches to target drought sensitive (S) or negatively regulating genes that control abiotic stresses. As a first proof of concept study, CRISPR/Cas9 system was used to introduce novel alleles in the gene encoding OPEN STOMATA 2 (OST2), a prominent plasma membrane H+ ATPase responsible for stomatal response in Arabidopsis (Osakabe et al. 2016). Plasma membrane proton (H+) ATPases (AHAs) are implicated in the generation of proton gradients to initiate stomatal opening (Merlot et al. 2007). Under dehydration stress, the ABA binds to the C-terminus leading to inhibition of H+ ATPase causing stomatal closure. Interestingly, tow dominant mutations have been detected in the ost2 locus which obliterate stomatal response to ABA leading to constitutive activity of the proton pump causing necrotic lesions (Merlot et al. 2007). Using a modified CRISPR/Cas9 system having a truncated sgRNA (tru-sgRNA) and Cas9 combination, mutations in the transgenic plants were detected with high mutation efficiency (> 32%) with no off-target modifications. Evaluation of stomatal response under ABA induced conditions revealed that ost2_cripspr mutants had significantly high degree of stomatal closure coupled with low level of transcriptional water loss as compared to the wild type. This indicated that, CRISPR/Cas9 induced mutation at the OST2 locus facilitated drought tolerance through enhanced stomatal response.

Table 1.

List of genes targeted by CRISPR/Cas genome editing system towards drought tolerance

Plant	Target gene	Target trait	Type of edit	References
Arabidopsis	OST2	Stomatal response	CRISPR/Cas9 (loss of function)	Osakabe et al. (2016)
Maize	ARGOS8	Drought tolerance	CRISPR/Cas9 (HDR based insertion of GOS2 promoter)	Shi et al. (2017)
Tomato	SIMAPK3	ABA dependent kinase signalling	CRISPR/Cas9 (loss of function)	Wang et al. (2017a, b)
Tomato	SINPR1	Drought resistance	CRISPR/Cas9 (loss of function)	Li et al. (2019)
Arabidopsis	AREB1	ABA signalling-mediated drought tolerance	CRISPR/dCas9HAT (epigenome editing)	Roca-Paixão et al. (2019)
Rice	OsSAPK2	ABA signalling-mediated drought tolerance	CRISPR/Cas9 (loss of function)	Lou et al. (2017)
Rice	OsSRL1, OsSRL2	Leaf rolling	CRISPR/Cas9 (loss of function)	Liao et al. (2019)

Most recently, CRISPR/Cas9 system was used to generate mutant lines for the tomato non-expresser of pathogenesis-related gene 1 (NPR1) to confirm its role in tomato drought tolerance (Li et al. 2019). While NPR1 is a key regulator of plant defence mechanism, its role in abiotic stress is highly limited. Whereas, a reduced expression of MdNPR1 has been reported in drought responsive apple trees (Bassett et al. 2014), overexpression of AtNPR1 in rice has resulted in hypersensitive response against drought stress (Quilis et al. 2008). The CRISPR/Cas9 induced loss of function s1npr1 mutants demonstrated reduced drought tolerance, wider stomatal aperture, greater electrolytic leakage, higher levels of malondialdehyde (MDA) and hydrogen peroxide and decreased levels of antioxidant enzymes as compared to the wild type (WT) tomato plants. The down-regulation of the drought responsive genes such as SIGST, SIDHN and SIDREB further confirmed the sensitivity of s1npr1 mutants to drought stress. Thus, it can be concluded that S1NPR1 plays a key role in controlling drought stress and multiple SlNPR1 variants could be developed through genome editing towards conferring broad spectrum drought tolerance in tomato and other crops.

Comprehensive molecular analysis has shown that ABA act as a primary factor of drought response in plants by regulating the expression of stress related genes and controlling stomatal closure to prevent water loss (Osakabe et al. 2014). The bZIP group of TFs, termed as ABA responsive element binding protein/ABRE binding factors (AREBs/ABFs) are the crucial element of ABA signalling (Nakashima et al. 2014). Up-regulated expression of AREB1 has demonstrated enhanced drought tolerance while knocking out AREB1 resulted in high sensitivity to drought stress (Yoshida et al. 2010; Singh and Laxmi 2015). AREB1 significantly regulate a huge set of genes downstream of the ABA-signalling pathway and act as a major determinant of ABA biosynthesis, antioxidant signalling and osmotic protection (Barbosa et al. 2013; Li et al. 2013). Therefore, AREB1 could be used as an effective target towards improving drought tolerance in crop plants. In a recent study, a modified CRISPR/Cas9 system consisting of a dead Cas9 (dCas9) fused with the catalytic domain of an Arabidopsis histone acetyl transferase (HAT) enzyme together with the sgRNA was used to target the promoter region of the AREB1 gene in Arabidopsis (Roca-Paixão et al. 2019). In this proof of concept study, the CRISPR/dCas9^HAT system generated stable Arabidopsis transgenic lines with a dwarf phenotype. The binding of the catalytic domain of Arabidopsis HAT enzyme resulted in acetylation of core histone leading to greater exposure of AREB1 promoter region to the transcriptional machinery. Molecular and physiological analysis of the mutants revealed higher chlorophyll content, faster stomatal aperture together with higher expression of the AREB1 and AREB1 regulated RD29A genes under water-deficit conditions. The CRISPR modified lines showed better survival rate under drought stress. Taken together, this suggests that the CRISPR/Cas system can be used for inducing efficient epigenetic modification to improve drought stress tolerance through the positive regulation of drought responsive genes.

SNF 1-related protein kinase 2 (SnRK2), a family of plant-specific protein kinases acts as a key regulator of abscisic acid (ABA)-dependent hyper-osmotic stress signalling and development in plants (Kobayashi et al. 2004). Members of the SnRK2 are variably implicated in ABA signalling, ABA-mediated stomatal closure, hyper-osmotic response, drought tolerance, seed germination and seedling growth (reviewed in Kulik et al. 2011). In particular, a distinctive stress regulatory network of SnRK2 has been identified in Arabidopsis with AtSnRK2.8 demonstrating positive regulation of drought tolerance coupled with up-regulated expression of stress responsive genes (Umezawa et al. 2004). While a clear difference in stomatal damage and survivability is not observed between the wild type and snrk2.8 mutant, microarray analysis has shown that SnRK2 control the regulation of AREB/ABF and their targets (Mizoguchi et al. 2010). Likewise, the members of the subclass I and III of SnRK2 family have demonstrated enhanced abiotic stress response and development in rice (Kulik et al. 2011). For instance, the constitutive expression of OsPYL present upstream of OsSAPK2 reported ABA-hypersensitive phenotype during seedling growth in rice (Kim et al. 2012). Also, OsSAPK9 has been reported has a positive regulator of ABA-mediated stress signalling in rice (Dey et al. 2016). Another recent study has shown that OsSAPK2 induced drought tolerance in rice through phosphorylation of rice genes OsbZIP23 and OsbZIP46 (Zong et al., 2016). This indicated that, OsSAPK2 is possibly fundamental for the ABA-mediated stress tolerance in rice. Therefore, CRISPR/Cas9 technology was used to develop loss of function mutants of OsSAPK2 to ascertain its role in the development of drought-tolerant rice genotypes (Lou et al., 2017). While the sapk2 mutants were more sensitive to drought stress and ROS, the wild type plants with functional SAPK2 increased drought tolerance through accumulation of compatible solutes to prevent water loss, early stomatal closure, induced expression of stress responsive genes and reduced expression of antioxidant enzyme genes to promote ROS scavenging. Overall, these findings imply that SAPK2 could be a significant molecule in the development of drought tolerance in future crop breeding programmes.

Among the different phytohormones involved in the physiological network underlying abiotic stresses, ethylene plays a crucial role in regulating water-deficit conditions and high temperature (Kawakami et al. 2010). Silencing of ethylene biosynthetic gene, 1-aminocyclopropane-1-carboxylic acid synthase 6 (ACS6) in maize has resulted in higher grain yield under water-deficit conditions (Habben et al. 2014). A previous study has also reported that the AUXIN REGULATED GENE INVOLVED IN ORGAN SIZE (ARGOS) family genes are negative regulators of ethylene signalling and confers enhanced drought tolerance and higher grain yield under drought stressed conditions (Shi et al. 2017). The expression of endogenous ARGOS8 transcript is relatively low and spatially fluctuating in Maize. Therefore, a group of scientists at the Dupont Pioneer (presently Corteva Agrosciences) has utilized the CRISPR/Cas9 approach to create novel ARGOS8 variants to confer beneficial traits, including drought tolerance in maize (Shi et al. 2017). The researchers used the HDR pathway to insert the maize native GOS2 promoter into the 5′ untranslated region of ARGOS8, which resulted in drought-tolerant maize lines. A diverse population of above 400 maize inbreds were analysed for ARGOS8 mRNA expression to investigate the use of difference in expression of ARGOS8 in drought tolerance and found that all the inbreds showed lower level of expression compared to the ARGOS8 mutant lines. Additionally, the ARGOS8 variants reported enhanced grain yield and no yield loss under water scarce conditions. This indicate that CRIPSR/Cas9 could be a significant player in the development of novel allelic variation for breeding of drought-tolerant crop varieties.

Accumulated evidences have confirmed that MAPK signalling consisting of the three-tiered phosphorelay cascade of MAPK, MAPK kinases and MAPK kinase kinases perceive extracellular signals and regulate various physiological and biochemical responses including abiotic stress tolerance in plants (Sinha et al. 2011). AtMPK3 from Arabidopsis, ZmMPK3 from maize and OsMSRMK2 and OsMPK5 have shown significant response to dehydration stress, suggesting that they could act as an active target molecule for realizing drought tolerance. In a recent study, CRISPR/Cas9 editing of SlMAPK3 gene demonstrated enhanced response to drought stress in tomato plants (Wang et al. 2017a, b). The slmapk3 mutants displayed severe wilting symptom, higher accumulation of hydrogen peroxide, lower antioxidant enzyme activities, and suffered from significant membrane damage under water deficient conditions as compared to the wild type (WT) plants. Likewise, the SlMAPK3 knockout resulted in significant differential expression of SlLOX, SlGST, SlDREB and other important drought responsive genes. The findings from this work suggest that SlMAPK3 is crucial for drought response in tomato and CRISPR/Cas editing could be used to develop novel MAPK variants in other plants as well for modulating downstream genes for conferring drought tolerance.

A significant feature of drought adaptability in water-deficit areas involves leaf rolling which decreases stomatal conductance and reduces water loss due to transpiration (Fang and Xiong 2015). Therefore, it is imperious to develop leaf rolling genotypes to have enhance yield under drought conditions. Among the various genes controlling leaf phenotypes in rice, Semi-rolled leaf 1 (SRL1) and SRL2 encoding the glycosyl phosphatidyl inositol-anchored proteins are major determinants of the number, size and arrangement of bulliform cells (BCs) in the leaf tissues (Liu et al. 2016). Loss of function in the srl1 mutants has resulted in lesser lignin content and cellulose in BCs while srl2 mutants have demonstrated severely impaired cuticular development (Liu et al. 2016; Li et al. 2017). Therefore, CRISPR/Cas9 system has been recently used to develop rolled leaf mutants in rice plants by modifying the SRL1 and SRL2 genes (Liao et al., 2019). Homozygous mutant lines for SRL1 and SRL2 reported reduced stomatal number, stomatal conductance, transpiration rate, chlorophyll content, vascular bundles and other agronomic traits as compared to wild type (WT). When subjected to drought stress, the mutant plants showed a higher survival rate supported by higher ABA content, super oxide dismutase (SOD), catalase activities and grain filling than the wild plants. Proteomic analysis of CRISPR/Cas9 induced mutants revealed the up regulation of 107 proteins majority of which were abiotic stress responsive. Further, the hybrids developed from the mutants displayed semi-rolled leaves and enhanced agronomic traits, including increased panicle number, grain number and yield per plant. Overall, this study revealed the inherent importance of genome editing and its exploitation in drought tolerance through the development of leaf rolling genotypes.

Limitations and future perspectives

CRISPR/Cas technology has been extensively adopted for modifying specific genes and associated molecular mechanisms for abiotic stresses as required for crop improvement programs. However, a wide number of gene edited mutants being transgenic in nature are still scrutinize by the ethical and regulatory authorities. Therefore, transgene-free integration to minimize the probability of off-target mutations is still a challenging task for effective utilization of genome editing. Preassembled CRISPR/Cas9 ribonucleoproteins has the potential to develop DNA free mutants with reduced frequency of off-target activity in plants (Woo et al. 2015). CRISPR–Cas9 RNPs based transgene-free genome editing has been successfully demonstrated in a large number of plants, including maize (Svitashev et al. 2016), wheat (Liang et al. 2017) and rice (Toda et al. 2019). As no foreign DNA is introduced in RNP, the mutants developed are transgene free and do not come under the ambit of GMO regulations. Most recently, researchers combined a fluorescence-dependent transgene monitoring module to the genome-editing tool box for identification of transgene-free CRISPR-/Cas9-edited plants of rice, tomato and wheat (Aliaga-Franco et al. 2019; Okada et al. 2019). The use of nanoparticles to deliver the CRISPR/Cas RNP complex into the meristematic cells of the plants has also been proposed (Glass et al. 2018). Therefore, CRISPR/Cas RNPs has the potential for rapid improvement of complex crop traits like drought in the ever changing climatic conditions.

Engineering drought tolerance involves the concerted manipulation of multiple genes associated with complex metabolic pathways. Therefore, the practical realization of drought tolerance depends upon the utilization of molecular techniques which has the ability to manipulate many genes concurrently. In the recent times, multiple sgRNAs driven by independent promoters has been multiplexed into single CRISPR/Cas9 expression vector using Golden Gate cloning or the Gibson assembly method (Silva and Patron 2017). In plants, an engineered endogenous tRNA processing system has led to the development of multiplex CRISPR/Cas9 or Cas12a platforms which could efficiently modify multiple target sites within a single polycistronic gene (Xie et al. 2015). Taking it further, Wang et al., (2017a, b) used a single crispr RNA array separated by mature direct repeats (DRs) in a CRSPR/Cas12a system to induce mutations in at least six target sites within three endogenous genes in rice. Hence, a CRISPR/Cas driven multi-editing system could be more aggressively explored as it has the potential to significantly modify multiple genes as required to confer drought and other abiotic stress tolerance.

Several cis-regulatory sequences found in the promoter region of the genes serves as the negative regulatory module for abiotic stress response by acting as binding sites for multiple transcription factors (Liu et al. 2014). For instance, Arabidopsis thaliana ANAC069 is responsible for inhibiting several genes responsible for ROS scavenging and osmotic stress tolerance by binding with a specific CACGT sequence in their promoter region (He et al. 2017). Therefore, in addition to the sensitive (S) or tolerant (T) genes, these cis-elements could also serve as potential target sites to achieve tolerance towards abiotic stress. Recently, a CRISPR–Cas9 genome-editing approach has been used to generate more than hundred regulatory mutations in the tomato SlCLV3 promoters to create novel phenotypic variations for quantitative traits related to yield and vigour (Rodríguez-Leal et al., 2017). Therefore, CRISPR/Cas system could be used to develop novel promoter variants through an HDR driven gain of function mutations for quantitative traits leading to drought tolerance. Although, HDR is still technically challenging in plants due to low efficiency, new and emerging CRISPR/Cas toolboxes could be explored for developing broad spectrum drought tolerance. On the other hand, while some of these cis-elements could act as specific targets for gene editing owing to their negative regulation of gene expression, they also acts as a binding site for multiple transcription factors which facilitate positive regulation of other stress response. For instance, while the rice OsMYB2 increases drought and salinity tolerance by binding to the MYBR cis-regulatory element of the downstream stress responsive genes (Yang et al. 2012), other OsMYBs also binds to the same sequence to negatively regulate stress tolerance and metabolic activities. Therefore, a careful alteration of cis-regulatory sequences coupled with overexpression of complementary stress responsive genes to mitigate the negative effect of cis-mutation could be qualitatively effective in realizing drought tolerance.

Large scale genomic studies have revealed that novel variations in elite traits of the crop plants are attributed to single base modification or polymorphism (Henikoff and Comai, 2003). For instance, 77 SNPs has been associated with 10 drought responsive transcription factors in maize (Mittal et al. 2017). However, it is difficult to induce single base modification using a CRISPR/Cas system as the efficiency of the template DNA dependent HDR approach is much lower compared to template free NHEJ repair process in plants. Hence, novel gene editing strategies to produce precise point mutations in plants in the need of the hour. The CRISPR/Cas mediated base editing technology has emerged as an advanced tool for precise nucleotide substitution without any requisite of a DSB or donor template (Komor et al. 2016). The base editors consist of a catalytically-inactive CRISPR–Cas9 domain (dCas9 or Cas9 nickase) for sequence recognition and a deaminase domain for base substitution. While a cytosine base editor (CBE) results in C-G to T-A conversion, the adenine base editors facilitate A-T to G-C modification. Furthermore, an RNA base editor has been developed using a catalytically-inactive Cas13 (dCas13) in connotation with adenosine deaminase acting on RNA (ADAR) for adenosine to inosine conversion in RNA sequences (Cox et al., 2017). The base editing approach has been successfully demonstrated in many crop plants, including rice, wheat, maize and tomato (reviewed in Mishra et al., 2020). More recently, this technology has been used in tomato, potato, rice and wheat for the development of herbicide tolerant plants (Zong et al. 2018; Veillet et al. 2019). In another study, rBE5 (hAID*∆-XTEN-Cas9n-UGI-NLS) base editor was used to target the blast resistance gene Pi-d2 for modulating broad spectrum disease resistance (Ren et al. 2018). However, there are no reports on the usage of these novel tools for the alteration of complex traits related to drought or salinity tolerance. In the recent times, several CBE and ABE variants have been developed that could be used to expand the scope of base editing in the improvement of abiotic stress tolerance in plants (Mishra et al. 2020). Overall, the base editing technology provides new opportunities for making precise nucleotide specific modifications and could be comprehensively explored in future for improving crops towards drought tolerance.

Conclusions

The ever changing climatic conditions, including high temperature and drought stress are the greatest threat to global food security. Over the past several decades, traditional breeding coupled with molecular markers and genetic engineering approaches has made significant contributions in alleviation of these stresses. However, given that these stress conditions are highly spontaneous and dynamic, the existing techniques are not efficient enough to generate desirable alleles in an effective way and within a short period of time. In contrast, the advent of CRISPR/Cas technology with its ability for precise genetic modification has negated these drawbacks and revolutionized the field of plant science and agriculture. Different variants of CRISPR/Cas technology has made it possible to develop drought tolerance through targeted alteration of important sensitive and tolerant genes in diverse plant species (Fig. 4). Besides, CRISPR/Cas based alteration of cis-regulatory sequences hold great promise in the regulation of gene expression that are crucial for drought tolerance. Further, the arrival of novel tools such as multiplex editing, RNP based DNA free editing and base editing provide a broader resource of CRISPR technologies for effective mitigation of drought and other abiotic stress conditions (Fig. 4). Currently, the inefficient regeneration ability of the edited plants and off-target mutations are the major limitations for CRISPR/Cas application in plant improvement. Nevertheless, these problems could now be overcome through the use of edited pollens and immature embryo that can surpass the traditional tissue culture route (Kelliher et al. 2019) and the application of stress inducible CRISPR/Cas technique resulting in negligible off-target activity (Nandy et al. 2019). Overall, the CRISPR/Cas based technologies will undoubtedly continue to transform the breeding programs towards the development of climate resilient crops.

Fig. 4.

Genome editing strategies towards drought tolerance in plants. Multiple genome editing platforms could facilitate drought tolerance though, a knockout of drought sensitivity (S) genes, b knock in of drought-tolerant (T) genes, c multiplex editing of tolerant and sensitive factors for broad spectrum drought tolerance, d promoter enhancement of tolerance genes through epigenome editing and e precise cis-regulatory modification of S/T gene expression through base editing

Author contributions

SSB and RM conceived and drafted the manuscript. RKJ collaborated in the manuscript preparation and critically reviewed the manuscript. All the authors have read, revised and approved the final manuscript.

Compliance with ethical standards

Conflict of interest

The authors declare that there is no conflict of interest.