CRISPR/Cas9 Gene Editing of NtAITRs, a Family of Transcription Repressor Genes, Leads to Enhanced Drought Tolerance in Tobacco

Tobacco is a cash crop throughout the world, and its growth and development are affected by abiotic stresses including drought stress; therefore, drought-tolerant breeding may help to improve tobacco yield and quality under drought stress conditions. Considering that the plant hormone ABA (abscisic acid) is able to regulate plant responses to abiotic stresses via activating ABA response genes, the characterization of ABA response genes may enable the identification of genes that can be used for molecular breeding to improve drought tolerance in tobacco. We report here the identification of NtAITRs (Nicotiana tabacum ABA-induced transcription repressors) as a family of novel regulators of drought tolerance in tobacco. Bioinformatics analysis shows that there are a total of eight NtAITR genes in tobacco, and all the NtAITRs have a partially conserved LxLxL motif at their C-terminus. RT-PCR results show that the expression levels of at least some NtAITRs were increased in response to ABA and drought treatments, and NtAITRs, when recruited to the Gal4 promoter via a fused GD (Gal4 DNA-binding domain), were able to repress transcription activator LD-VP activated expression of the LexA-Gal4-GUS reporter gene. Roles of NtAITRs in regulating drought tolerance in tobacco were analyzed by generating CRISPR/Cas9 gene-edited mutants. A total of three Cas9-free ntaitr12356 quintuple mutants were obtained, and drought treatment assays show that drought tolerance was increased in the ntaitr12356 quintuple mutants. On the other hand, results of seed germination and seedling greening assays show that ABA sensitivity was increased in the ntaitr12356 quintuple mutants, and the expression levels of some ABA signaling key regulator genes were altered in the ntaitr12356-c3 mutant. Taken together, our results suggest that NtAITRs are ABA-responsive genes, and that NtAITRs function as transcription repressors and negatively regulate drought tolerance in tobacco, possibly by affecting plant ABA response via affecting the expression of ABA signaling key regulator genes.


Introduction
Environmental stresses including abiotic stresses and biotic stresses reduce the yield of plants, including crops [1,2]. It is estimated that biotic stresses such as pathogens and insects cause~20% whereas abiotic stresses such as drought and salt cause~50% of global yield loss for most major crops [3,4]. On the other hand, to survive the abiotic stress conditions they are constantly exposed to, plants have evolved mechanisms in response to abiotic stresses [5], and dissection of the mechanisms underlining plant responses to abiotic stresses may lead to the breeding of crops with enhanced abiotic stress tolerance, therefore increasing crop yield under abiotic stress conditions. It is proposed that plants respond to abiotic stresses via two different signaling pathways, i.e., the plant hormone ABA (abscisic acid)-dependent and -independent pathways [5]. As a key stress hormone, ABA accumulates dramatically in response to abiotic stresses such as drought and salt, leading to the activation/repression of ABA response genes via signaling transduction through several key regulators, including the PYR1/PYL/RCAR (Pyrabactin resistance 1/PYR1-like/regulatory component of ABA receptor) ABA receptors, the ABA signaling negative regulators A-group PP2C (PROTEIN PHOSPHATASE 2C) phosphatases, the ABA signaling positive regulators SnRK2 (NON-FERMENTING 1 (SNF1)-RELATED PROTEIN KINASES) protein kinases, and downstream ABF/AREB/ABI5 type bZIP (basic region leucine zipper) transcription factors, thereby affecting plant responses to abiotic stresses [6][7][8][9][10][11][12][13].
Consistent with this, changes in expression levels of the ABA signaling key regulator genes including the PYR1/PYL/RCAR receptor genes, the PP2C phosphatase genes, the SnRK2 protein kinase genes and the ABF/AREB/ABI5 transcription factor genes affected plant responses to ABA, resulting in changes in abiotic stress tolerance in plants [14][15][16][17]. Similarly, changes in expression levels of protein stability regulator genes of some ABA signaling key regulators, including the E2 ligase gene VPS23A and E3 ligase genes CUL4, RSL1, KEG (KEEP ON GOING), DWA1, DWA2, PUB12 and PUB13, are also able to affect plant response to ABA and thereby plant responses to abiotic stresses [18][19][20][21][22][23][24][25][26]. In addition, changes in expression levels of ABA response genes such as the MYB transcription factor genes MYB44 and MYB71, the bHLH transcription factor gene bHLH112, the WD40 protein genes AIW1 (ABA Induced WD40-repeat 1) and AIW2, and heat shock factor gene HSFA6b also affected plant responses to abiotic stresses [27][28][29][30][31]. Considering that there are hundreds of thousands of ABA-responsive genes remained uncharacterized, it is very likely that some of the genes may be involved in the regulation of plant responses to abiotic stresses.
In an attempt to identify novel regulators of plant abiotic stress responses from unknown-function ABA response genes, we identified AITRs (ABA-induced transcription repressors), a novel family of transcription repressors in Arabidopsis as negative regulators of plant abiotic stress responses [12,13]. We found that the Arabidopsis aitr mutants showed enhanced tolerance to abiotic stresses including drought and salt [12], and knockout of the entire family of AITRs did not affect plant growth and development in Arabidopsis [13]. Considering that AITRs are conserved in angiosperms [12], AITRs may be good candidate genes for molecular breeding to improve abiotic stress tolerance in crops. Indeed, we found that AITRs in several different crops including tomato, rice, soybean and cotton shared similar features with Arabidopsis AITRs, i.e., their gene expression is induced by ABA, and they function as transcription repressors [12,32,33]. In addition, expression of cotton GhAITR-A1 restored the enhanced abiotic stress tolerance in aitr1 mutants [32], indicating that GhAITRs may also function as negative regulators of plant abiotic stress tolerance. Most importantly, knockout of several GmAITR genes by CRISPR/Cas9 gene editing moderately enhanced salt tolerance in soybean [33]. However, we also note that gmaitr mutants showed slightly enhanced ABA sensitivity, a phenotype opposite to that of the Arabidopsis aitr mutants, which showed greatly decreased ABA sensitivity [12,33]. These results suggested that AITRs in different plants may have different functions at least in some aspects.
Tobacco (Nicotiana tabacum) is an economically important crop throughout the world, which is usually grown for smoking, but its biomass can also be a good source of advanced biofuels [34,35]. Similar to other plants, tobacco's growth and development is also affected by environment stresses including abiotic stresses [36][37][38][39][40][41], and some studies have found that under abiotic stress conditions, numerous physiological and molecular changes in tobacco including accumulation of ABA were observed [39][40][41]. Studying the mechanisms of stress tolerances and mining regulatory genes involved in regulating abiotic stress responses, including the ABA response genes in tobacco, are essential for molecular breeding to improve abiotic stress tolerance in tobacco.
In this study, we provide evidence that tobacco AITRs (NtAITRs) are involved in the regulation of plant drought tolerance in tobacco. We found that the expression levels of at least some of the NtAITRs increased in response to ABA and drought treatment, and NtAITRs function as transcription repressors. By generating ntaitr mutants using CRISPR/Cas9 gene editing and examining their responses to ABA and abiotic stresses, we found that NtAITRs negatively regulate drought tolerance in tobacco via affecting the plant responses to ABA; therefore, NtAITRs may be targeted for CRISPR/Cas9 gene editing to improve drought tolerance in tobacco.
Amino acid sequence alignment shows that NtAITRs shared high identity and similarity, especially between each pair of the proteins, i.e., NtAITR1 and NtAITR3, NtAITR2 and NtAITR7, NtAITR4 and NtAITR8, and NtAITR5 and NtAITR6, but all of them have only a partially conserved LxLxL motif at the C-terminus ( Figure 1). Protein domain assays show that there are five conserved domains in all the NtAITRs, including one with the partially conserved LxLxL motif ( Figure S2). Phylogenic analysis of AITRs from tobacco, Arabidopsis and soybean shows that NtAITRs formed one clade with three Arabidopsis AITRs (AITR1, AITR3 and AITR4) and two soybean AITRs (GmAITR3 and GmAITR6), whereas the other three Arabidopsis AITRs (AITR2, AITR5 and AITR6) and the other four soybean AITRs (GmAITR1, GmAITR2, Protein domain assays show that there are five conserved domains in all the NtAITRs, including one with the partially conserved LxLxL motif ( Figure S2). Phylogenic analysis of AITRs from tobacco, Arabidopsis and soybean shows that NtAITRs formed one clade with three Arabidopsis AITRs (AITR1, AITR3 and AITR4) and two soybean AITRs (GmAITR3 and GmAITR6), whereas the other three Arabidopsis AITRs (AITR2, AITR5 and AITR6) and the other four soybean AITRs (GmAITR1, GmAITR2, GmAITR4 and GmAITR5) formed another clade ( Figure 2). In addition, the phylogenic analysis also shows that NtAITR5 and NtAITR6 are closely related, and the other six NtAITRs are closely related ( Figure 2).
NtAITRs were obtained on NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 30 October 2018) and used for multiple sequence alignment by using BioEdit. Identical amino acids are shaded in black and similar amino acids are shaded in gray. Underline indicates the partially conserved LxLxL transcriptional repression motif.
Protein domain assays show that there are five conserved domains in all the NtAITRs, including one with the partially conserved LxLxL motif ( Figure S2). Phylogenic analysis of AITRs from tobacco, Arabidopsis and soybean shows that NtAITRs formed one clade with three Arabidopsis AITRs (AITR1, AITR3 and AITR4) and two soybean AITRs (GmAITR3 and GmAITR6), whereas the other three Arabidopsis AITRs (AITR2, AITR5 and AITR6) and the other four soybean AITRs (GmAITR1, GmAITR2, GmAITR4 and GmAITR5) formed another clade ( Figure 2). In addition, the phylogenic analysis also shows that NtAITR5 and NtAITR6 are closely related, and the other six NtAITRs are closely related ( Figure 2).

Expression Levels of NtAITRs Are Increased in Response to ABA and Drought Treatments
Thus far, it has been shown that all the AITRs characterized from different plants including the model plant Arabidopsis, and crops including tomato, rice, soybean and cotton, are all inducible by ABA treatment, and all the AITRs function as transcription repressors [12,32,33]. To examine if that is also the case with NtAITRs, we first examined their expression in response to ABA treatment. Tobacco seedlings if the K326 wild type were treated with ABA for a short time, and the expression of NtAITRs was examined by RT-PCR. As shown in Figure 3A, it is clear that the expression levels of all the NtAITRs but NtAITR2 were increased in the ABA-treated samples, as a very faint band was obtained for NtAITR2 in the ABA-treated sample, which indicates that it may respond slightly to ABA treatment. We also examined the expression of NtAITRs in response to drought treatment by using PEG treatment to mimic drought condition, and we found that the expression levels of at least some NtAITRs including NtAITR1, NtAITR2, NtAITR3 and NtAITR5 were increased in response to drought treatment ( Figure 3B).
We then examined the transcription activities of NtAITRs by using protoplast transfection assays. We successfully cloned full-length CDS for NtAITR1, NtAITR3, NtAITR4, NtAITR6, NtAITR7 and NtAITR8, and generated constructs with a fused N-terminal GD tag. Plasmids of the GD-NtAITR constructs were co-transfected with plasmids of the transcription activator gene LD-VP and the LexA-Gal4:GUS reporter gene into protoplasts isolated from Col wild type Arabidopsis leaves. As shown in Figure 4, LD-VP-activated GUS expression was repressed when GD-AITRs were co-transfected, suggesting that all the NtAITRs examined function as transcription repressors. spond slightly to ABA treatment. We also examined the expression of NtAITR sponse to drought treatment by using PEG treatment to mimic drought conditio we found that the expression levels of at least some NtAITRs including Nt NtAITR2, NtAITR3 and NtAITR5 were increased in response to drought treatme ure 3B). NtAITRs in response to ABA treatment. Ten-day-old K326 wild type tobacco seedling treated with 50 µM ABA for 4 h, then RNA was isolated and used for RT-PCR analysis to e the expression of NtAITRs. The expression of NtGAPDH gene was used as a control. (B) sion of NtAITRs in response to drought treatment. Twenty-day-old soil-grown K326 w seedlings were watered with 20% PEG6000, seedlings were collected 6 h after and then RN isolated and used for RT-PCR analysis to examine the expression of NtAITRs. The expre NtGAPDH gene was used as a control.
We then examined the transcription activities of NtAITRs by using pro transfection assays. We successfully cloned full-length CDS for NtAITR1, Nt NtAITR4, NtAITR6, NtAITR7 and NtAITR8, and generated constructs with a N-terminal GD tag. Plasmids of the GD-NtAITR constructs were co-transfecte plasmids of the transcription activator gene LD-VP and the LexA-Gal4:GUS report into protoplasts isolated from Col wild type Arabidopsis leaves. As shown in Fi LD-VP-activated GUS expression was repressed when GD-AITRs were co-trans suggesting that all the NtAITRs examined function as transcription repressors. NtAITRs in response to ABA treatment. Ten-day-old K326 wild type tobacco seedlings were treated with 50 µM ABA for 4 h, then RNA was isolated and used for RT-PCR analysis to examine the expression of NtAITRs. The expression of NtGAPDH gene was used as a control. (B) Expression of NtAITRs in response to drought treatment. Twenty-day-old soil-grown K326 wild type seedlings were watered with 20% PEG6000, seedlings were collected 6 h after and then RNA was isolated and used for RT-PCR analysis to examine the expression of NtAITRs. The expression of NtGAPDH gene was used as a control.

Generation of Transgene-Free ntaitr Mutants
The above results show that NtAITRs shared similar features with AITRs c terized from other plants including Arabidopsis, soybean, tomato, cotton and ric their gene expression is induced by ABA, and they function as transcription repr These results indicate that NtAITRs may also be involved in the regulation of AB abiotic stress responses in tobacco. To examine if this is indeed the case, we deci generate ntaitr mutants by using CRISPR/Cas9 gene editing.
Considering that NtAITRs shared high nucleotide sequence identity, especia tween each pair of the genes, i.e., NtAITR1 and NtAITR3, NtAITR2 and NtA NtAITR4 and NtAITR8, and NtAITR5 and NtAITR6 ( Figure S3), we selected four ent target sequences that may simultaneously target several different NtAITRs, i

Generation of Transgene-Free ntaitr Mutants
The above results show that NtAITRs shared similar features with AITRs characterized from other plants including Arabidopsis, soybean, tomato, cotton and rice, i.e., their gene expression is induced by ABA, and they function as transcription repressors. These results indicate that NtAITRs may also be involved in the regulation of ABA and abiotic stress responses in tobacco. To examine if this is indeed the case, we decided to generate ntaitr mutants by using CRISPR/Cas9 gene editing.
Considering that NtAITRs shared high nucleotide sequence identity, especially between each pair of the genes, i.e., NtAITR1 and NtAITR3, NtAITR2 and NtAITR7, NtAITR4 and NtAITR8, and NtAITR5 and NtAITR6 ( Figure S3), we selected four different target sequences that may simultaneously target several different NtAITRs, including NtAITR1, NtAITR2, NtAITR3, NtAITR5 and NtAITR6 ( Figure 5A), and made an FT expression cassette containing the CRISPR/Cas9 construct with four sgRNAs ( Figure 5B). The containing of the FT expression cassette enables the reduction in the time required to generate gene-edited mutants and facilitates the selection of Cas9-free mutants [42]. By using the construct to transform wild-type tobacco, we successfully obtained some transgenic plants with early flowering phenotypes. Gene editing status in these early flowering plants was examined, and transgene-free mutants were obtained in next generation based on flowering phenotype, then confirmed by PCR amplification of the Cas9 fragment, and gene editing status was examined again. Three transgene-free homozygous quintuple mutants, i.e., ntaitr1 ntaitr2 ntaitr3 ntaitr5 ntaitr6-c1 (ntaitr12356-c1), ntaitr12356-c2 and ntaitr12356-c3 were obtained. In all the mutants, either a single-nucleotide insertion/deletion or the deletion of a few nucleotides occurred at the target sites of the NtAITRs, resulting in a frame shift after the edited sites of the corresponding NtAITRs ( Figure 6A). It should be noted that NtAITR2 in ntaitr12356-c2 and NtAITR3 in ntaitr12356-c2 were edited in a bioallelic manner ( Figure  6B), but either way resulted in a frame shift ( Figure 6A); therefore, seeds collected from the mutants were directly used for the following experiments. Three transgene-free homozygous quintuple mutants, i.e., ntaitr1 ntaitr2 ntaitr3 ntaitr5 ntaitr6-c1 (ntaitr12356-c1), ntaitr12356-c2 and ntaitr12356-c3 were obtained. In all the mutants, either a single-nucleotide insertion/deletion or the deletion of a few nucleotides occurred at the target sites of the NtAITRs, resulting in a frame shift after the edited sites of the corresponding NtAITRs ( Figure 6A). It should be noted that NtAITR2 in ntaitr12356-c2 and NtAITR3 in ntaitr12356-c2 were edited in a bioallelic manner ( Figure 6B), but either way resulted in a frame shift ( Figure 6A); therefore, seeds collected from the mutants were directly used for the following experiments.

The ntaitr12356 Mutant Plants Show Enhanced Tolerance to Drought
Expression level changes of AITRs in both Arabidopsis and soybean affected plant response to abiotic stresses [12,13,33]. After showing that NtAITRs shared similar features with AITRs from other plants examined so far, we examined drought tolerance in the ntaitr12356 quintuple mutants. The K326 wild type tobacco and the wild type-I were used as controls. The wild type-I is a non-edited plant segregated from the offspring of the T0 plant from which the ntaitr12356-c3 quintuple mutant was isolated. Seeds of the wild type, the wild type-I and the ntaitr12356 quintuple mutants were germinated in soil pots and grown in a growth room. Ten days after germination, the plants were watered thoroughly, and then watering was withheld for drought treatment. Twenty days after withholding watering, watering was resumed. As shown in Figure 7, the seedlings of the wild type, the wild type-I and the ntaitr12356 quintuple mutants were largely similar before the drought treatment. All the plants continued growing during the treatment, but most of the leaves of the plants nearly dried after the drought treatment, with the three ntaitr12356 quintuple mutants to a less severe degree (Figure 7). Five days after watering was resumed, at least four-fifths of each of the three ntaitr12356 quintuple mutants survived, whereas at least four-fifths of the wild type and the wild type-I plants died, and no big difference was observed among the three ntaitr quadruple mutants or between the wild type and the wild type-I plants (Figure 7), indicating that drought tolerance was enhanced in the ntaitr12356 quintuple mutants.

The ntaitr12356 Mutant Plants Show Enhanced Tolerance to Drought
Expression level changes of AITRs in both Arabidopsis and soybean affected plant response to abiotic stresses [12,13,33]. After showing that NtAITRs shared similar features with AITRs from other plants examined so far, we examined drought tolerance in the ntaitr12356 quintuple mutants. The K326 wild type tobacco and the wild type-I were used as controls. The wild type-I is a non-edited plant segregated from the offspring of the T0 plant from which the ntaitr12356-c3 quintuple mutant was isolated. Seeds of the wild type, the wild type-I and the ntaitr12356 quintuple mutants were germinated in soil pots and grown in a growth room. Ten days after germination, the plants were watered

The ntatir12356 Mutants Are Hypersensitive to ABA
After showing that drought tolerance was enhanced in ntaitr12356 quintuple mutants, we further examined if this may be caused by alternated ABA response in the mutants. Seed germination and cotyledon greening assays were used to examine ABA sensitivity of the ntaitr12356 quintuple mutants, and the K326 wild type tobacco and the wild type-I were used as controls. As shown in Figure 8A, in the seed germination assays, all three ntaitr12356 quintuple mutants showed a clearly increased sensitivity to ABA compared to the wild type and the wild type-I. We also note that little if any difference on the germination rate was observed among the three ntaitr12356 quintuple mutants, and the seed germination rate of the wild type-I was also largely indistinguishable from that of the wild type ( Figure 8A). but most of the leaves of the plants nearly dried after the drought treatment, with the three ntaitr12356 quintuple mutants to a less severe degree (Figure 7). Five days after watering was resumed, at least four-fifths of each of the three ntaitr12356 quintuple mutants survived, whereas at least four-fifths of the wild type and the wild type-I plants died, and no big difference was observed among the three ntaitr quadruple mutants or between the wild type and the wild type-I plants (Figure 7), indicating that drought tolerance was enhanced in the ntaitr12356 quintuple mutants. Figure 7. Drought tolerance of the ntaitr12356 quintuple mutants. Seeds of the K326 wild type, the wild type-I and the ntaitr12356 mutants were germinated in soil trays and grown for 10 days before drought treatment. Drought treatment was performed by withholding watering for 20 days, and then watering was resumed. Pictures were taken before drought treatment, after drought treatment and 5 days after watering was resumed. Plants that survived were counted 5 days after watering was resumed, and survival rate was calculated. Data represent the mean ± SD of five replicates.

The ntatir12356 Mutants Are Hypersensitive to ABA
After showing that drought tolerance was enhanced in ntaitr12356 quintuple mutants, we further examined if this may be caused by alternated ABA response in the mutants. Seed germination and cotyledon greening assays were used to examine ABA sensitivity of the ntaitr12356 quintuple mutants, and the K326 wild type tobacco and the wild type-I were used as controls. As shown in Figure 8A, in the seed germination assays, all three ntaitr12356 quintuple mutants showed a clearly increased sensitivity to ABA compared to the wild type and the wild type-I. We also note that little if any difference on the germination rate was observed among the three ntaitr12356 quintuple mutants, and the seed germination rate of the wild type-I was also largely indistinguishable from that of the wild type ( Figure 8A). . Drought tolerance of the ntaitr12356 quintuple mutants. Seeds of the K326 wild type, the wild type-I and the ntaitr12356 mutants were germinated in soil trays and grown for 10 days before drought treatment. Drought treatment was performed by withholding watering for 20 days, and then watering was resumed. Pictures were taken before drought treatment, after drought treatment and 5 days after watering was resumed. Plants that survived were counted 5 days after watering was resumed, and survival rate was calculated. Data represent the mean ± SD of five replicates.
Clearly increased ABA sensitivity for all three ntaitr12356 quintuple mutants was also observed in the seedling greening assays ( Figure 8B). Quantitative assays show that in the presence of ABA, less than 20% of the seedlings of all the three ntaitr12356 quintuple mutants produced green cotyledons, compared with~60% for wild type and wild type-I seedlings ( Figure 8C). Again, no difference was observed among the three ntaitr12356 quintuple mutants, or between the wild type and the wild type-I ( Figure 8C).
It has been shown that both the Arabidopsis AITRs and GmAITRs regulate ABA response via regulating the expression of some ABA signaling key regulator genes [12,13,33]. We thus examined if the expression levels of the ABA signaling key regulator genes may be changed in the ntaitr12356 quintuple mutants. As the three ntaitr12356 quintuple mutants showed a similar response to ABA in both the seed germination and seedling greening assays, we used only the ntaitr12356-c3 mutant for the experiment. As shown in Figure 8D, the expression level of NtPYL7, an ABA receptor gene [43], was increased, whereas that of NtPP2CA, a PP2C phosphatase gene, was decreased in the ntaitr12356-c3 mutant compared to the wild type seedlings.
Clearly increased ABA sensitivity for all three ntaitr12356 quintuple mutants was also observed in the seedling greening assays ( Figure 8B). Quantitative assays show that in the presence of ABA, less than 20% of the seedlings of all the three ntaitr12356 quintuple mutants produced green cotyledons, compared with ~60% for wild type and wild type-I seedlings ( Figure 8C). Again, no difference was observed among the three ntaitr12356 quintuple mutants, or between the wild type and the wild type-I ( Figure 8C). wild type, the wild type-I and the ntaitr12356 quintuple mutants in response to ABA. Sterilized seeds the K326 wild type, the wild type-I and the ntaitr12356 mutants were kept at 4 °C in darkness for 7 days and then sown on 1/2 MS plates in the presence or absence of 2 µM ABA. The plates were transferred to a growth room and seeds germinated were counted daily after the plates were transferred. Data represent the mean ± SD of three replicates. (B) Seedling greening of the K326 wild type, the wild type-I and the ntaitr12356 quintuple mutants in response to ABA. Pictures were taken 12 days after the plates were transferred. (C) Percentage of green seedlings of the K326 wild type, wild type-I and the ntaitr12356 quintuple mutants in response to ABA. Green seedlings were counted 12 days after the transfer, and the percentage of green seedlings was calculated. Data represent means ± SD of three replicates. (D) Expression of NtPYL7 and NtPP2CA in the K326 wild type and the ntaitr12356-c3 quintuple mutant seedlings in response to ABA. Ten-day-old seedlings were treated with 50 µM ABA or solvent methanol as a control for 4 h. Total RNA was isolated, and RT-PCR was used to examine the expression of the ABA signaling component genes. The expression of NtGAPDH was used as a control.
It has been shown that both the Arabidopsis AITRs and GmAITRs regulate ABA response via regulating the expression of some ABA signaling key regulator genes [12,13,33]. We thus examined if the expression levels of the ABA signaling key regulator genes may be changed in the ntaitr12356 quintuple mutants. As the three ntaitr12356 quintuple mutants showed a similar response to ABA in both the seed germination and seedling greening assays, we used only the ntaitr12356-c3 mutant for the experiment. As shown in Figure 8D, the expression level of NtPYL7, an ABA receptor gene [43], was increased, whereas that of NtPP2CA, a PP2C phosphatase gene, was decreased in the ntaitr12356-c3 mutant compared to the wild type seedlings. Figure 8. ABA sensitivity of the ntaitr12356 quintuple mutants. (A) Seed germination of the K326 wild type, the wild type-I and the ntaitr12356 quintuple mutants in response to ABA. Sterilized seeds the K326 wild type, the wild type-I and the ntaitr12356 mutants were kept at 4 • C in darkness for 7 days and then sown on 1/2 MS plates in the presence or absence of 2 µM ABA. The plates were transferred to a growth room and seeds germinated were counted daily after the plates were transferred. Data represent the mean ± SD of three replicates. (B) Seedling greening of the K326 wild type, the wild type-I and the ntaitr12356 quintuple mutants in response to ABA. Pictures were taken 12 days after the plates were transferred. (C) Percentage of green seedlings of the K326 wild type, wild type-I and the ntaitr12356 quintuple mutants in response to ABA. Green seedlings were counted 12 days after the transfer, and the percentage of green seedlings was calculated. Data represent means ± SD of three replicates. (D) Expression of NtPYL7 and NtPP2CA in the K326 wild type and the ntaitr12356-c3 quintuple mutant seedlings in response to ABA. Ten-day-old seedlings were treated with 50 µM ABA or solvent methanol as a control for 4 h. Total RNA was isolated, and RT-PCR was used to examine the expression of the ABA signaling component genes. The expression of NtGAPDH was used as a control.

Discussion
ABA signaling via several key regulators activates/represses ABA responses genes, thereby regulating plant abiotic stress tolerance [6][7][8][9][10][11][12][13]. By characterizing unknown-function ABA response genes, we previously identified AITRs as a novel family of transcription repressors that negatively regulate abiotic stress tolerance in Arabidopsis [12,13]. As AITRs are conserved in angiosperms [12], we then characterized AITRs in several different crops to see if AITRs may be targeted for CRISPR/Cas9 gene-editing-based molecular breeding to improve abiotic stress tolerance in crops. Thus far, we found that gmaitr mutants generated by CRISPR/Cas9 gene editing of soybean AITRs showed moderately enhanced salt tolerance [33], but ABA sensitivity was slightly enhanced in the gmaitr mutants [33], opposite to that of the Arabidopsis aitr mutants [12], indicating that AITRs in different plants may have different functions, at least in some aspects.
We provide evidence in this study that NtAITRs shared similar features with AITRs characterized so far in other plants including Arabidopsis, tomato, cotton and soybean [12,32,33], i.e., NtAITRs are ABA response genes, as their expression levels were increased in response to ABA treatment (Figure 3), and NtAITRs function as transcription repressors, as the NtAITRs examined repressed the expression of the reporter gene in transfected protoplasts (Figure 4). It should be noted that due to the low expression level of NtAITR2 and high nucleotide identity of the NtAITRs gene pairs ( Figure S3), we failed to generated constructs for NtAITR2 and NtAITR5, therefore transcription repression activities of NtAITR2 and NtAITR5 were not examined in this study. However, considering that NtAITRs shared high amino acid identity and similarity, and all have a partially conserved LxLxL motif (Figure 1), it is reasonable to assume that NtAITR2 and NtAITR5 can also function as transcription repressors, yet experimental evidence are required to examine if this is indeed the case.
Our evidence in this study also show that even though NtAITRs shared similar features with AITRs in other plants, a difference in regulating ABA sensitivity was observed, as examined with the ntaitr12356 quintuple mutants generated by CRISPR/Cas9 gene editing ( Figure 6). Greatly increased ABA sensitivity was observed in the ntaitr12356 quintuple mutants as compared with the wild type tobacco plants (Figure 8), opposite to that observed in the Arabidopsis aitr mutants, which showed greatly decreased ABA sensitivity [12,13]. On the other hand, NtAITRs may have similar functions as GmAITRs in regulating ABA sensitivity, as slightly enhanced ABA sensitivity was observed in the gmaitr mutants [33]. There are six genes in Arabidopsis but eight in cotton encoding AITRs [12,33], and there are also eight in tobacco ( Figure 1). Considering that the maximum decrease in ABA sensitivity in Arabidopsis aitr mutants was observed in aitr256 triple mutants [13], and the ntaitr mutants used this study are ntaitr12356 quintuple mutants ( Figure 6), whereas increased ABA sensitivity observed the gmaitr mutants may be only caused by the loss of function of two GmAITR genes [33], it is very likely that high-order mutants of GmAITRs may also result in greatly increased ABA sensitivity in soybean, i.e., NtAITRs and GmAITRs may have similar functions in regulating ABA sensitivity.
Our results show that so far, all the AITRs examined in different plants negatively regulate abiotic stress tolerance, Arabidopsis aitr236 triple and high-order mutants showed greatly enhanced tolerance to drought and salt [12,13], gmaitr mutants showed moderately enhanced tolerance to salt [33], and ntaitr12356 quintuple mutants showed greatly enhanced tolerance to drought (Figure 7). These results provided evidence that ATIRs in crops indeed could be good targets for CRISPR/Cas9 gene editing to improve abiotic stress tolerance.
As an industrial crop with great economic importance, tobacco's yield and quality are affected by abiotic stresses [36,37,[39][40][41]. Breeding of tobacco plants with enhanced abiotic stress tolerance will increase their yield and quality. As ABA affects plant responses to abiotic stresses via regulating ABA response genes [6][7][8][9][10][11][12][13], characterization of unknownfunction ABA response genes may enable the identification of novel regulators of abiotic stress tolerance. Our evidence in this study indicates that NtAITRs are this kind of regulators. As the ntaitr12356 quintuple mutants showed enhanced sensitivity to ABA (Figure 8), it is very likely that the functions of NtAITRs in regulating abiotic stress tolerance in tobacco are due to its feedback regulating roles of ABA responses. Yet, detailed mechanisms of NtAITRs in regulating abiotic stress tolerance in tobacco need to be further investigated.
Most importantly, our evidence in this study clearly indicates that it is not only possible but also practical to used NtAITRs as targets for CRISPR/Cas9 editing to improve abiotic stress tolerance in tobacco. First, coding sequences of NtAITRs shared high nucleotide identity ( Figure S3), therefore it will be easier to identify target sequences that can target at least a pair of NtAITR genes. As an example, we identified four target sequences that can be used to target five NtAITR genes ( Figure 5). Second, it is not difficult to generate transgenefree gene-edited mutants for NtAITRs, which may avoid the difficulty of using transgenic plants in production. Even though not all the target sites were edited, we obtained mutants with five NtAITR genes being edited (Figure 6), and we successfully isolated transgenefree mutants by using flowering phenotypes caused by the FT expression cassette in the CRISPR/Cas9 construct ( Figure 5). Third, greatly enhanced drought tolerance was observed in mutants with only some of the NtAITR genes being edited. The eight NtAITR proteins shared high amino acid sequence identity as well as conserved domains (Figures 1 and S2), suggesting that they may have redundant functions. However, greatly enhanced drought tolerance was already observed in the ntaitr12356 quintuple mutants (Figure 7), indicating that it may be not necessary to edit all the NtAITRs to obtain tobacco plants with enhance abiotic stress tolerance. Yet, it will be of great interest to examine if further enhanced abiotic tolerance can be achieved if all the NtAITRs are being edited and if the loss of function of all NtAITRs may have any fitness cost.
In summary, our results in this study show that the expression of NtAITRs is induced by ABA and that NtAITRs function as transcription repressors and they positively regulate ABA sensitivity but negatively regulate drought tolerance in tobacco. As a result, NtAITRs can be used as targets for CRISPR/Cas9 gene-editing-based molecular breeding to improve abiotic stress tolerance in tobacco.

Identification of AITRs in Tobacco
To identify NtAITRs in tobacco, the entire amino acid sequences of Arabidopsis AITRs were used for Protein BLASTing on NCBI (https://blast.ncbi.nlm.nih.gov/Blast.cgi, accessed on 30 October 2018) with N. tabacum as selected organism. The full-length amino acid sequences of the retrieved NtAITRs were then used in reiterative BLASTing until no more NtAITRs were identifiable. The coding sequences of the NtAITRs were obtained by using the locus numbers to conduct gene searching on the KEGG GENES Database (https: //www.genome.jp/kegg/genes.html, accessed on 30 October 2018), and corresponding genome sequences were obtained by BLASTing the N. tabacum genome with the coding sequences on the Sol Genomics Network (https://solgenomics.net/organism/Nicotiana_ tabacum/genome, accessed on 30 October 2018). Full-length amino acid sequences of obtained NtAITRs were subjected to phylogenetic analysis (http://hylogeny.fr/phylo_cgi/ index.cgi, accessed on 1 September 2022) with AITRs from Arabidopsis and soybean [12,33], and motif analysis was performed using MEME with default settings [44].

Plant Materials and Growth Conditions
The K326 tobacco was used as wild type for plant transformation and as control for ABA and drought treatment and drought tolerance assays. The ntaitr12356 quintuple mutants were generated by CRISPR/Cas9 gene editing. The non-edited plant wild type-I was segregated from the offspring of the T0 plant from which the ntaitr12356-c3 quintuple mutant was isolated and was also used as a control for ABA and drought treatment. The Col wild type Arabidopsis was used for protoplast isolation. The tobacco seedlings used for RNA isolation and ABA treatments were grown on 1 2 MS plates, and tobacco plants used for plant transformation, drought treatment and drought tolerance assays were grown in soil pots at 22-24 • C in a growth room with a 16/8 h (light/dark) photoperiod, as described previously [45].

ABA and Drought Treatment, RNA Isolation and RT-PCR
To examine the expression of NtAITRs and ABA signaling component genes in response to ABA, 10-day-old tobacco seedlings were treated with 50 µM ABA for 4 h in darkness on a shaker at 40 rpm and frozen in liquid N 2 immediately for RNA isolation. To examine the expression of NtAITRs in response to drought treatment, 20-day-old soil-grown K326 wild type tobacco seedlings were watered with 20% PFG6000 and seedlings were collected 6 h after and frozen in liquid N 2 immediately for RNA isolation.
Total RNA was isolated by using a Plant RNA kit (OMEGA BIO-TEK, Norcross, GA, USA) following the manufacturer's instructions, and 1-2 µg total RNA was subjected to first-strand cDNA synthesis by using an EasyScript First-strand DNA Synthesis Super Mix (TransGen Biotech, Beijing, China). RT-PCR analysis was performed by using the synthesized cDNA as templates. The expression of the NtGAPDH gene was used as an internal reference control [46]. The primers used were listed in Table S1.

Constructs
The reporter construct of LexA-Gal4:GUS and the effector constructs GD and LD-VP were as described previously [47,48].
To generate 35S:GD-NtAITRs constructs for protoplast transfection assays, the fulllength open reading frame (ORF) of the NtAITRs was amplified by RT-PCR using RNA isolated from tobacco seedlings and cloned in frame with an N-terminal GD tag into the pUC19 vector with the 35S promoter [47,48].
In order to generate CRISPR/Cas9 constructs for NtAITRs gene editing, a total of four different target sequences within the exons of NtAITRs that may be simultaneously target five genes were selected on CRISPRscan (http://www.crisprscan.org/ ?page=sequence, accessed on 30 October 2018), and target specificity was evaluated on cas-offinder (http://www.rgenome.net/cas-offinder, accessed on 30 October 2018). The four target sequences are the following: 5 -GCACCTAAACCGGTCGCCGG-3 for NtAITR1 and NtAITR3, 5 -GCTCCTAAACCAGCCGCTGG-3 for NtAITR5 and NtAITR6, 5 -GCCGTGTGATGTGTGGAAGA-3 for NtAITR1, NtAITR2 and NtAITR3, and 5 -GCCGTGT-GATGTGTGGAGAA-3 for NtAITR5 and NtAITR6. The four targets were inserted in to the pHEE vector with an FT expression cassette as reported previously [42]. The primers used were listed in Table S1.

Plasmid Isolation, Protoplast Isolation and Transient Transfection
Plasmid isolation, protoplast isolation and transfection, and GUS activity assays were performed by following the procedure described previously [12,[48][49][50]. Briefly, plasmid DNA of the reporter and effector constructs was isolated by using an EndoFree Plasmid Maxi Kit (OMEGA BIO-TEK). Protoplasts were isolated from leaves collected from~3week-old Col wild-type Arabidopsis and co-transfected with the LexA-Gal4:GUS reporter, the LD-VP and the GD-NtAITRs effector plasmid DNA. The transfected protoplasts were incubated for 20-22 h under darkness at room temperature. GUS activities were measured by using a SynergyTM HT fluorescence microplate reader (BioTEK, Winooski, VT, USA).

Generation of Transgene-Free ntaitr Mutants
Five-to six-week-old K326 tobacco plants were used for Agrobacterium GV3101mediated stable transformation by infiltration, as described previously [45]. T0 transgenic seedlings generated were transferred into soil pots and grown in a growth room. The plants with early flowering phenotypes were selected, and gene editing status was detected by PCR amplification and sequencing of the DNA sequence fragments of NtAITR1, NtAITR2, NtAITR3, NtAITR5 and NtAITR6. Transgene-free homozygous mutant plants were isolated in T1 or T2 progeny based on flowering phenotype and sequencing, as previously described [42,51].

Seed Germination and Seedling Greening Assays
Seed germination and cotyledon greening assays were used to examine ABA sensitivity of the ntaitr12356 mutants obtained, and the K326 wild type and the wild type-I tobacco were used as controls. Healthy and uniform-sized seeds of the K326 wild type, the wild type-I and the ntaitr12356 mutants were surface sterilized and germinated on 1/2 MS plates in the presence or absence of 2 µM ABA. Germinated seeds were counted daily after the transfer. Pictures were taken 14 days after the transfer, and seedlings with green cotyledons were counted.

Drought Tolerance Assays
Drought tolerance assays of soil-grown tobacco were performed by using the procedures described previously [52][53][54] with some modification. Seeds of the K326 wild type, the wild type I and the ntaitr12356 mutants were sown in soil trays and grown for 10 days with sufficient watering. The plants were then subjected to drought treatment by withholding watering for 20 days, and watering was resumed. Pictures were taken before drought treatment, after drought treatment and 5 days after watering was resumed.