Arabidopsis cysteine-rich receptor-like protein kinase CRK33 affects stomatal density and drought tolerance

ABSTRACT

Cysteine-rich receptor-like protein kinases (CRKs) are transmembrane proteins containing two domains of unknown function 26 (DUF26) RLKs in their ectodomain. Despite that CRKs control important aspects of plant development, only few proteins have functionally been characterized. In this work, we analyzed the function of CRK33 by characterizing two insertional lines. The stomatal density and stomatal index were decreased in crk33-2 and crk33-3 plants in comparison to wild-type plants, correlating with a decreased transpiration in transgenic plants and a higher drought tolerance. Furthermore, photosynthesis and stomatal conductance changed. Finally, all four stomata cell fate genes were upregulated, especially the expression of TMM and SPCH in the mutant background, suggesting a role for CRK33 in stomatal spacing.

Introduction

Stomata are specialized pores in the leaf epidermis that consist of two guard cells that function as turgor-operated valves, thereby regulating water and gas exchange between plant and atmosphere, required for photosynthesis and respiration. These cells modulate water loss in plants by control of the transpiration level.^1–3

In Arabidopsis thaliana leaves, the development of stomata relies on a series of oriented stem-cell-like asymmetric divisions followed by a single symmetric division.^1–4 The density, distribution and differentiation of stomata are crucial for stomatal function, and for optimal growth and survival of the plant. Altering stomatal distribution impacts CO₂ uptake, evaporation, and internal leaf temperature in much the same way as pore opening and closing.^5,6

The Arabidopsis genome encodes for 158 proteins carrying the structural motif basic helix-loop-helix (bHLH), which are classified in 26 subfamilies. The subfamily 1a contains three proteins encoded by the genes SPEECHLESS (SPCH: At5g53210), MUTE (At3g06120), and FAMA (At3g24140), which share conserved motifs also outside the bHLH region.⁷

The initiation and proliferation of stomatal precursors are driven by SPCH, the master transcription factor, which regulates hundreds of genes to promote divisions and fate transitions, inducing the first asymmetric division that gives rise to the meristemoid cell, then depending on how much and in which tissue it is expressed, it will determine the stomatal pattern and density.^8–10 The MUTE protein is required to terminate the asymmetric self-renewing divisions and to induce the differentiation of meristemoids into guard mother cells (GMCs), while FAMA protein is required for final transitions from GMC to guard cells (GCs).^9,11–13 These three transcription factors are unable to functionally replace one another during stomatal development due to distinct features of each protein, a mutation in any of these bHLH genes leads to a reduction in the ability to form stomata.^9–12

Stomatal pattering and density depend largely on cell–cell communication, through secreted peptides belonging to the EPIDERMAL PATTERNING FACTOR (EPF) family, plant hormones, and environmental stimuli.^6,14–16

The receptor-like protein (RLP), TOO MANY MOUTHS (TMM), detects the extracellular EPFs, transmits the signal into the cell, and restricts stomatal development. EPF1-ERL1 (ERECTA family-LIKE1)/TMM peptide-receptor kinase signaling is required to orient the polarity of the asymmetric spacing divisions.^17–19

The perception of the signal through cell surface receptors is a common feature among living organisms. Plants have developed sophisticated signaling systems that include cell surface receptors that perceive various signals and stimuli from the environment, in addition to external stimuli, plants use intra and intermolecular messenger molecules (peptides, hormones and reactive oxygen species) to control growth, development and survival.^20–23 Receptor-like kinases (RLKs) harbor different extracellular domains for perception of distinct ligands and play key roles in mediating perception of extracellular signals.^23,24 In addition, expression analyses have linked a large number of RLKs to many different physiological processes and signaling networks in plant development, pathogen defense, and abiotic stress response.^25–27

One of the largest RLK groups is formed by the cysteine-rich receptor-like kinases (CRKs), which are transmembrane proteins containing the Domain of Unknown Function 26 (DUF26). CRKs participate in the regulation of plant development and stress adaptation in response to biotic and abiotic stimuli in a non-redundant fashion.^25,28–30 Signaling pathways mediated by CRKs, have been characterized in hormone perception and pathogen response. Reports have shown that over-production of CRK5 and CRK13 enhances plant resistance to Pseudomonas syringae.^30,31

Furthermore, CRK4, CRK5, CRK19 and CRK20 have been shown to be involved in the hypersensitive response-associated cell death in Arabidopsis.^32,33

CRK7 is involved in the regulating of the responses to extracellular ROS production.³⁴ Overexpression of CRK13 enhances defense responses, by increasing H₂O₂ levels.³¹ CRK36 and a receptor-like cytosolic kinase (RLCK), ARCK1, interact with each other and negatively regulate ABA and osmotic stress signal transduction.³⁵ Overexpression of CRK4 and CRK5 enhances ABA sensitivity of stomatal movement and drought tolerance.³⁶ However, how CRKs enhance drought tolerance remains unknown.

In this study, we show that misregulation of CRK33 increases water-use efficiency (WUE) by affecting the stomatal density. Moreover, up-regulation of the stomata cell fate genes, especially SPCH and TMM in the crk33 genetic background suggests an alteration of stomata development.

Materials and methods

Plant materials and growth conditions

The Arabidopsis thaliana ecotype Columbia (Col-0) was used as a control. The T-DNA insertional lines crk33-2 (SALK_061061) and crk33-3 (SAIL_128_E04; CS806236) were obtained from the Arabidopsis Biological Resource Center (ABRC).

The seeds were sterilized with absolute ethanol (96°) and placed on Murashige Skoog (MS) medium at 4°C for 2 d. Then, they were exposed to photoperiod (16:8 h light/darkness) in a Biotronette growth chamber (Lab-Line Instruments, ILL, USA). At cotyledon stage, they were transplanted to the substrate (peat moss: vermiculite: perlite; 3:1:1) and grown under controlled conditions in the greenhouse at 24°C during the day and 20°C at night. The T-DNA insertions in CRK33 were confirmed by PCR; using CRK33 specific primers and the T-DNA LBb1.3 primer in the case of crk33-2, and the T-DNA LB3 primer in the case of crk33-3 (Table S1).

RNA isolation, cDNA synthesis and qRT-PCR analysis

Total RNA was isolated from different plant tissues following the TRIzol method³⁷ and the recommendations of the manufacturer (Invitrogen). The RNA concentration was estimated from 260/280 nm and 260/230 nm absorbances, which were measured in a spectrophotometer NanoDrop 2000 UV-Vis (Thermo Scientific, OH, USA). The integrity of RNA was verified by agarose gel electrophoresis. The genomic DNA was removed by a DNase treatment (RNase-free DNase set, Zymogen).

The cDNA was synthesized from total RNA using the First-Strand cDNA Synthesis kit (Thermo Scientific, USA) following instructions of the manufacturer. The SYBR Green mix (Radiant Green Lo-ROX qPCR) (Alkali Scientific, FL, USA) was used for qRT-PCR experiments. Specific oligonucleotides for CRK33, FAMA, MUTE, SPCH, TMM, and ACT2 were used for qRT-PCR analysis (Table S1).

The gene expression was normalized to Actin2 gene expression, the endogenous gene, in the same sample. This was done by using the expression (C_T values) levels in analyzed tissue for the different lines according to comparative 2^−ΔΔCt method (Livak and Schmittgen, 2001).³⁸

The amplification was performed in a Real-Time PCR System (CFX96 C1000 Touch (Bio-Rad, USA)). The amplification of a specific product was corroborated by melting curve analysis. Three biological samples were analyzed, each with a technical duplicate.

The amplification conditions were 95°C for 10 s and 30 cycles, 60°C for 30 s, and 30 s at 72°C for semi qRT-PCR and qRT-PCR. For all pairs of oligonucleotides, the T_m was 60°C and their efficiency was not affected during the analysis. The results were analyzed using the CFX Manager program (Bio-Rad).

Construct and plant transformation

The full-length CRK33 coding sequence was amplified from cDNA from inflorescences of Col-0 using specific oligonucleotides (Table S1) and cloned in pENTR™/D-TOPO® vector (Invitrogen). The plasmid was recombined via a Gateway LR reaction with the binary vector pGD625³⁹ resulting in the 35S::CRK33 overexpression construct. Arabidopsis thaliana Col-0 plants were transformed with the construct using the floral dip method⁴⁰ using Agrobacterium tumefaciens strain GV3101, followed by selection on plate.⁴¹

Stomatal density measurement in cotyledons

The number of stomata per unit of area in the adaxial surface of the Col-0, crk33-2 and crk33-3 lines was determined. For this, we used a microscope (Leica DM5000B), equipped with a digital camera (Leica DFC450C). The seeds were previously sterilized and grown on MS medium (see above). Fifteen days after their germination, 100 cotyledons were collected, placed on a slide with ultrapure agarose (Invitrogen) at 3% to make surface imprints, protected with coverslips. The number of stomata in an area of ~161,500 µm² was counted.

Scanning electron microscopy (SEM)

SEM images were taken with a Zeiss EVO40 environmental scanning electron microscope (Carl Zeiss, Oberkochen, Germany) with a 20 kV beam, and the signal was collected using the BSD detector.

Stomatal index (IE) and stomatal density (SD) determination

SEM images were used for the determination of the leaf stomatal index (IE) (stomata/epidermal cells) and the stomatal density (SD) (number of stomata per unit area of the leaf). The IE was calculated using the formula suggested by Wilkinson (1979). IE = (NE/(CE+NE)) × 100 where NE = stomata number per unit area of leaf and CE = epidermal cell number within unit area of leaf.

$$IE=\frac{NE}{CE+NE}x100$$

where

NE: Stomatal number

CE: Epidermal cell number

The SD was obtained by determining the number of stomata observed in an area of ~50,000 µm².

The SEM images were used to quantify the number of stomata and epidermal cells of the Col-0, crk33-2, and crk33-3 lines in basal and cauline leaves, taking as a reference 3 different points on the adaxial surface, respectively.

Photosynthesis, stomatal conductance and transpiration

The physiological carbon assimilation, water transpiration, light capture, and energy in Col-0, crk33-2, and crk33-3 lines were analyzed and quantified using an Infrared Gas Analyzer (IRGA; LI-6400 Portable Photosynthesis System) (Li‐Cor, Lincoln, NE, USA) by measuring the photosynthesis, expressed as µmolCO₂m⁻²s⁻¹; stomatal conductance, expressed as mmol H₂Om⁻²s⁻¹; and transpiration, expressed as mmol H₂Om⁻²s⁻¹. These parameters were determined in three plants from each line in triplicate, after being under greenhouse conditions for 10 h. Parameters were analyzed and quantified using the IRGA equipment in rosette leaves, in plants grown under normal conditions.

Analysis of leaf number in crk33 lines

The effect of CRK33 disregulation on the number of leaves was analyzed. Ten plants at reproductive stage, for each line were used. The number of rosette and cauline leaves was determined by quantifying each leaf. Then, the results were graphed and analyzed.

Statistical analysis

The measurements of photosynthesis, transpiration, and stomatal conductance in Col-0, crk33-2, and crk33-3 lines, were analyzed using the ANOVA, the comparison of means by least significant difference (LSD), and the Kruskal–Wallis test by ranks. The stomatal density measurements were processed using the variance analysis ANOVA and the comparison of means by Tukey.

The Statgraphics Centurion XVI version program (StatPoint Inc, Herndon, VA) was used for these analyses. Results of qRT-PCR were analyzed with one-way ANOVA and the comparison of means by Tukey. The Minitab 16 program (Minitab, LLC, Chicago, IL) was used for these analyses.

Results

CRK33 expression during plant development

In order to know how specific CRK33 is expressed during plant development, we analyzed its expression in different tissues by qRT-PCR. The expression analysis in Arabidopsis thaliana Col-0 wild-type plants indicates that CRK33 gene is barely detectable in floral buds and fruits; however, it showed a higher expression in cauline leaves, rosette leaves and cotyledons (Figure 1A). Interestingly, these are all tissues with an abundant number of stomata. A previous report on CRK analyses also reported difficulties detecting expression of various CRK genes, including CRK33.²⁸ In a complementary expression analysis, we used 20,000 publicly available RNA-seq libraries⁴² to analyze CRK33 expression. This analysis confirmed that CRK33 is very lowly expressed in tissues of Arabidopsis thaliana (Figure 1B and C). Interestingly, CRK33 expression is induced upon drought and nutrient deficiency conditions (Figure 1D).

Figure 1.

Relative expression levels of CRK33 gene in different plant tissues in Arabidopsis thaliana Col-0. (A). The expression of CRK33 was estimated by qRT-PCR. Nine plants for each line were used as the experimental unit. The experiment was performed in triplicate. Stem was used as reference tissue, as it was the sample with the lowest expression. Error bars represent standard error of the mean. (B)-(D) Expression of CRK33 based on 20,000 RNA-seq libraries.⁴² Expression is expressed in FPKM

Identification of CRK33 misregulation plants

The CRK33 coding region contains seven exons, and to study its function, we used two independent T-DNA insertional lines. In the crk33-2 mutant, the T-DNA is located in the last exon (Figure 2B), while in the crk33-3 line, the T-DNA is located in the fifth exon (Figure 2C). In both insertional lines, the T-DNA insertion is located in the kinase domain (Figure 2D), likely affecting its function. To determine if the insertion affects the expression of CRK33, we performed an expression analysis by RT-PCR in cotyledons using primers upstream to the insertions, but we noticed that the transcripts levels increased in both lines (Figure S1; Figure S2(A)). Then, we performed a second analysis, now with primers that align in a region downstream of the T-DNA insertions, and the result was an increase in transcript level in both lines, although, more in the crk33-3 genetic background (Figure S2(B)). Finally, qRT-PCR analysis was performed for all tissues in wild type and in the crk33-2 mutant, showing now the expected decreased CRK33 transcript levels in some tissues (stem and fruit tissue), but again, increased expression in various tissues such as cotyledons, rosette and cauline leaves (Figure 2). Increased expression, instead of expected decreased expression, has been reported before for CRK mutants,²⁸ and thus apparently is something common for CRK genes.

Figure 2.

Relative expression levels of the CRK33 gene in different plant tissues in Arabidopsis thaliana Col-0 and the crk33-2 line. The data obtained in the wild-type background were used to analyze changes in the expression of CRK33 in the crk33-2 background by qRT-PCR. Nine plants for each line were used as the experimental unit. (A) Insertional T-DNA crk33-2 and crk33-3 lines. Localization of the T-DNA insertion in the CRK33 gene (At4g11490) in the fifth exon (crk33-3); (B) and seventh exon (crk33-2). (C) Data represent the mean ± SD of three replicates (n = 3). (D). Protein domain simulation by the SMART program (Smart.embl) and NCBI database; showing that the T-DNA insertions in both mutant lines locate into the kinase domain of CRK33; in the case of crk33-2 (441aa; Ile) and crk33-3 (566aa; Arg). The wild-type tissue was used independently for each analysis

To analyze whether the observed phenotype in crk33-2 and crk33-3 lines is due to gain of function or loss of function of CRK33, we decided to analyze stomatal density in fruit where we did not detect CRK33 expression in any of the crk33-2 or crk33-3 lines (Figure S3). The results showed a decrease in the number of stomata in the valves of fruit of crk33-2 and crk33-3 lines compared to wild type. These data show that the observed phenotype about stomata density is due to an absence of CRK33 function.

In addition to the obtained crk33 mutant alleles, we intended to generate CRK33 constitutive overexpression plants. We generated a construct with the double 35S CaMV enhancer that will result in strong constitutive CRK33 expression. However, despite various attempts to generate transgenic plants, we were not able to obtain 35S::CRK33 overexpression plants. This suggests that constitutive overexpression of CRK33 might be lethal.

CRK33 misregulation affects stomatal density

Analysis of the cotyledon phenotype in the crk33-2 and crk33-3 mutants showed a lower stomatal density compared to wild-type plants (Figure 3A–C and J). An increase in the stomatal spacing was observed in the crk33-2 and crk33-3 lines in the adaxial surface of cotyledons compared to wild type. This may be due to the larger pavement cell size observed (Figure 3). The crk33-2 mutant showed a stronger phenotype than crk33-3, which was confirmed by stomatal index (Figure 3J–H). We also analyzed rosette and cauline leaves, where we also found a decreased stomatal density and stomatal index.

Figure 3.

Phenotypes of the CRK33 insertional lines during stomata development. Samples were taken from Col-0 (a, d, g); crk33-2 (b, e, h); and crk33-3 (c, f, i). Cotyledons agarose impressions (A–C); SEM micrographs of epidermal adaxial surface of rosette leaves (D–F) and cauline leaves (G–I). crk33 mutants showed a decrease in the number of stomata. Stomatal density (J) and index (K) of crk33 lines and Col-0. Scale bars represent 100 µm

Second, we observed stomata development. In wild-type cotyledons, stomata from the meristemoid to the mature stomata stage were observed (Figure 3A), while in crk33-2 most stomata were at the mature stage (Figure 3B), and in crk33-3 early-stage stomata were frequent (Figure 3C). In rosette leaves, in wild type it was observed that although most of the stomata were at the mature stage, some were still in earlier stages (Figure 3D); however, in the crk33-2 (Figure 3E) and crk33-3 (figure 3F), all stomata were at the mature stage. In addition, more cells separate the stomata in the crk33 lines. In cauline leaves, of wild type (Figure 3G) and crk33 lines (Figure 3H and I), stomata were all at the mature stage, but in the crk33 lines there is again more spacing between stomata.

In summary, these results suggest that CRK33 is required for the correct density of stomata, but also to complete their development.

CRK33 misregulation affects the expression of stomatal cell fate markers

To further analyze whether stomata development is affected in the crk33 lines, we opted to analyze the expression of important cell fate marker genes for stomata development. We analyzed by qRT-PCR the expression of SPEECHLESS (SPCH), TOO MANY MOUTHS (TMM), MUTE and FAMA in cotyledon tissue of wild type and crk33-2 (Figure 4). We observed that the expression of all four genes was statically significantly increased, especially SPCH and TMM in crk33-2 compared to wild-type plants. These results can be directly related to the lower number of stomata (per unit area) on the adaxial surface of the cotyledons and leaves.

Figure 4.

Relative expression of SPCH, MUTE, FAMA and TMM in wild type and crk33-2 plants. qRT-PCR results revealed increased expression of all four genes in the crk33-2 mutant compared to wild-type plants. Nine plants for each line were used as the experimental unit; qRT-PCR was performed in triplicate. Data represent the mean ± SD of three replicates (n = 3). Different letters indicate a statistically significant difference

CRK33 misregulation improves water stress tolerance

The photosynthetic capacity of the leaves in crk33 lines was evaluated. According to the analysis of variance, there were no significant differences between the crk33 lines and wild-type plants (Figure 5A), suggesting that the photosynthetic capacity is similar (<2 µmol CO₂m⁻²s⁻¹) in both lines, allowing them to carry out their metabolic functions.

Figure 5.

Photosynthesis, transpiration, stomatal conductance, and Water Use Efficiency (WUE) in crk33 mutants. (A) Photosynthesis levels, (B) Stomatal conductance, (C) Transpiration levels and (D) Water use efficiency in crk33 lines and wild-type control plants

Because photosynthesis has been reported to be related to stomatal conductance, we analyzed it in crk33 lines. The crk33 lines showed a decreased stomatal conductance (mmolH₂Om⁻²s⁻¹) to that of wild-type plants, according to the analysis of variance (Figure 5B).

To analyze if the water release is affected in the crk33 lines, the transpiration was measured. The analysis of transpiration indicated that there is a significant difference between the crk33 lines and wild-type plants (Figure 5C). Plants of the crk33-2 and crk33-3 lines presented a lower level of transpiration than wild-type plants.

We analyzed whether the number of leaves is affected in the crk33 lines. Indeed, both the number of rosette leaves and cauline leaves showed an increase in both alleles (Figure S5).

To analyze whether the ratio of water used in plant metabolism to water lost by the plant through transpiration is affected, water-use efficiency (WUE) was evaluated. In the case of crk33-3 (Figure 5D), the analysis of variance indicated a greater WUE (>80 µmol CO₂/molH₂O) with respect to wild-type plants (<50 µmol CO₂/molH₂O), suggesting that it optimizes the stomatal closure and water loss during transpiration or has defects in the stomata development. No statistical difference was observed for crk33-2 compared to wild-type plants.

In order to know if CRK33 participates in the response of the plant to drought stress conditions, wild-type plants and crk33 lines were subjected to water stress for 15 d (without irrigation) and were compared with well-watered plants (control) (Figure 6A and B). On d 5 after the onset of the drought stress condition (Figure 6C and D), the plants appeared the same, with the same tonality of green. But on d 10, we noticed that crk33 lines were still quite green and even flowering, while the wild types were more than 60% dying and less than 30% still with green areas (Figure 6E and F). On d 15, 100% of the wild-type plants looked dry, while some plants of crk33 lines were still a bit green and about 30% still flowering (Figure 6G and H). On the other hand, compared to the control plants (well-watered for 15 d), crk33 lines showed a slower development and a delayed flowering.

Figure 6.

Plants under water stress. Plants of Col-0, crk33-2 and crk33-3 lines subjected to water stress. D 1 (control; a and b), d 5 (c: stress; and d: control), d 10 (e: stress; and f: control) and d 15 (g: stress; and h: control). Three plants for each line were used as the experimental unit. Throughout the experiment, the effect of water stress on the physiological development of plants was observed

In summary, although additional experiments are required, these results suggest that mutations in crk33, especially in crk33-3, tend to be more tolerant to water stress conditions.

Discussion

Despite their large number and high sequence conservation, individual CRKs have intriguingly distinct functions in different aspects of plant life. This makes the CRKs promising candidates for future studies of their biochemical function. Previous reports have suggested that the DUF26 domain could recognize different ligands, which has allowed it to acquire new functions, such as protein–protein interactions on the cell surface. Therefore, proteins containing these domains may be involved in the perception of environmental signals.⁴³

The RLK subfamily of cysteine-rich receptor-like protein kinases (CRKs) includes 46 members in Arabidopsis and their functions have been related to different processes, among them, photosynthesis, development, stomata, and biotic and abiotic stress.^22,28,29 Understanding how stomata work can allow strategies to be developed to help plants become more tolerant to drought stress.

A detailed study of most of the CRKs in Arabidopsis thaliana was conducted previously.²⁸ For some CRKs (including CRK33) the presence of the transcript could not be detected by semi-quantitative RT-PCR analysis in wild-type plants, and no further analysis was conducted, neither was it possible to observe a phenotype in the analyzed insertional line (SALK_148136 C or crk33-1).²⁸ We also failed to detect the transcript by semi-quantitative RT-PCR analysis (Figure S1), and only by qRT-PCR, we were able to detect expression in some of the analyzed tissues (Figure 1). This low expression was confirmed by analyzing 20,000 RNA-seq libraries (Figure 1). Furthermore, this analysis showed that CRK33 is induced upon drought and nutrient deficiency conditions. Interestingly, when we analyzed the expression of CRK33 in two insertional lines crk33-2 and crk33-3 (Figure 2), we found that the abundance of the transcript increased in cotyledons, rosette and cauline leaves. This is a similar result as has been reported before where it was found that CRK15, CRK30 and CRK26 are upregulated when they have a T-DNA insertion in the promoter region.²⁸ A possible hypothesis is that because in crk33-2 and crk33-3 the transcript is not translated or does not give a functional protein, it causes a self-regulatory system that increases the transcript synthesis of CRK33.

The crk33-2 and crk33-3 lines contain both a T-DNA insertion in the kinase domain. It has been shown that affecting the kinase domain causes the loss of the kinase function of the CRKs.³⁶ We performed analysis by aligning sequences of CRK33 with four other CRKs, CRK5, CRK4, CRK19 and CRK28. These have a significant degree of similarity, CRK33 is 45% similar to CRK4, CRK5 and CRK19, while 39% similar to CRK28. It is important to mention that the greatest similarity was in the kinase domain from amino acid 355 (Figure S5). So, we can assume that the T-DNA insertions in the kinase domain affect the kinase activity of CRK33. On the other hand, constitutive overexpression of CRK33 using the strong 35S promoter seems to be lethal.

Stomata density and correct spacing is determined by the correct development of stomata, which is regulated by genes related to cell fate. We analyzed the expression of four stomata genes related to the initiation (lineage-specific stem cells), proliferation, and differentiation into specialized cell types.⁶ During stomatal development, the participation of SPCH, MUTE and FAMA genes is important because they are required to establish the cell lineage.⁴⁴

SPCH contributes to the specification and proliferation of stomatal lineage cells by inducing the first asymmetric division that gives rise to the meristemoid.^9,13 Overexpression of the SPCH gene confers excess asymmetric cell division and generates highly divided small cells.^3,9,45–47 The TMM gene is required to orient the polarity, because it acts as a negative regulator of asymmetric divisions in neighbor cells, and it is a positive regulator of divisions in meristemoids. In this way, it is the major determinant of stomatal formation as well as patterning,¹⁸ perhaps affecting the asymmetric division and cellular spacing during stomatal development inhibiting the production of stomata.² In the crk33 lines, all four stomata cell fate genes showed increased expression compared to wild type, with TMM the strongest, which correlates with the low stomatal density and increased spacing observed. We also observed protodermal cells at the meristemoid stage, perhaps this explains the increase in the expression of SPCH.^45,47

Phenotypic analyses in crk33 lines showed that photosynthesis is not affected, and this suggests that CO₂ assimilation is not affected. The crk33-2 and crk33-3 lines showed that stomatal density and spacing are affected due to the larger size of the pavement cells and due to the higher number of epidermal cells per stoma, so the distance between them is greater (Figure 3). In Arabidopsis, the number and distribution of stomata changes according to the type of organ.^48–50 Based on our results, CRK33 is required for correct stomatal density which influences plant transpiration (Figure 5C). Stomatal conductance can be optimized in short and long times, through the opening and closing of stomata and through changes in the development of stomatal density. However, if the closure of stomata is prolonged, CO₂ absorption and assimilation may decrease.^51–53

By transpiration, open stomata allow the release of water vapor while absorbing nutrients and water through roots.^54,55 It is suggested that an open stomata allows the balance between photosynthesis and transpiration, later, the plants regulate stomata development in coordination with various stimuli.^14,56

Overexpression of CRK5 enhances abscisic acid sensitivity affecting the opening and closing mechanism of the stomata, conferring tolerance to drought stress.³⁷ Although, in the case of CRK33, it seems that it is not the movement of the stomata that is affected but the stomata density, both of these appear to influence drought stress tolerance. Interestingly, CRK33 expression in wild-type plants is induced by drought conditions, supporting a role for CRK33 in drought stress tolerance.

Mechanisms controlling the abundance of stomatal complexes are important in determining photosynthetic performance and water use efficiency.^57–59 Based on the evaluation of the WUE, it indicates lower stomatal conductance in limiting water conditions.⁶⁰ The crk33-2 and crk33-3 lines were subjected to drought stress, and slightly more drought tolerant could be observed. Although drought stress tolerance was phenotypically not so evident, perhaps because the mutant plants have a mechanism to try to counteract low transpiration by forming more leaves. Our results suggest that CRK33 is required for proper plant transpiration, but its misregulation may allow the plant to withstand more drought stress conditions.

In summary, our results show that CRK33 is involved in the determination of stomatal density and correct stomata development. Future research should shed more light on how CRK33 affects these processes.

Funding Statement

This work was supported by the National Technology of Mexico (TecNM) [5584.19-P]; CONACYT [CB-2017-2018-A1-S-10126 and FC-2015-2/1061].

Disclosure statement

No potential conflicts of interest were disclosed.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.