The MYB33, MYB65, and MYB101 transcription factors affect Arabidopsis and potato responses to drought by regulating the ABA signaling pathway

Abstract Drought is one of the main climate threats limiting crop production. Potato is one of the four most important food crop species worldwide and is sensitive to water shortage. The CBP80 gene was shown to affect Arabidopsis and potato responses to drought by regulating the level of microRNA159 and, consequently, the levels of the MYB33 and MYB101 transcription factors (TFs). Here, we show that three MYB TFs, MYB33, MYB65, and MYB101, are involved in plant responses to water shortage. Their downregulation in Arabidopsis causes stomatal hyposensitivity to abscisic acid (ABA), leading to reduced tolerance to drought. Transgenic Arabidopsis and potato plants overexpressing these genes, with a mutated recognition site in miR159, show hypersensitivity to ABA and relatively high tolerance to drought conditions. Thus, the MYB33, MYB65, and MYB101 genes may be potential targets for innovative breeding to obtain crops with relatively high tolerance to drought.


| INTRODUCTION
Drought is one of the most visible effects of climate change currently worldwide. These effects are also some of the most drastic factors affecting agriculture and the economy. Via classic and molecular approaches, many genomic loci and genes potentially associated with the response to water shortage have been identified in model and crop species (Chaves et al., 2003;Duque et al., 2013;Joshi et al., 2016;Kulkarni et al., 2017). Potato (Solanum tuberosum L.) is considered a drought-sensitive crop species, although cultivar-dependent differences in tolerance have been described (Boguszewska-Ma nskowska et al., 2018;Dahal et al., 2019;Gervais et al., 2021;Lehretz et al., 2021;Pieczynski et al., 2017;Soltys-Kalina et al., 2016;Zarzy nska et al., 2017). Tolerance to drought is a very complex polygenic trait in potato (Obidiegwu et al., 2015). Several genes, as well as quantitative trait loci, for drought tolerance have been identified on all 12 potato chromosomes (Anithakumari et al., 2011(Anithakumari et al., , 2012Khan et al., 2015;Pieczynski et al., 2018).
Genes involved in RNA metabolism have been found to affect the ABA signaling pathway (Hugouvieux et al., 2001(Hugouvieux et al., , 2002Laloum et al., 2018;Pieczynski et al., 2013;Reyes & Chua, 2007;Roychoudhury et al., 2013). For example, it was found that alternative splicing, affected by Serine and arginine-rich proteins, can produce three alternative protein isoforms of the zinc-induced-facilitator-like 1 (ZIFL1) transporter required for polar auxin transport in Arabidopsis thaliana (Eckardt, 2013). A shortened form of the transporter, ZIFL1.3, is located exclusively in the plasma membrane of leaf stomatal guard cells and is involved in drought tolerance via stomatal closure, most likely by modulating potassium and proton fluxes in plant cells.
Another group of proteins affecting the plant response to abiotic stresses is the glycine-rich RNA-binding proteins (GRPs), of which heterogeneous ribonucleoprotein particle -like proteins are members.
However, the functions of these proteins remain largely unknown. In Arabidopsis, GRP7 negatively affects plant development in response to drought stress during seed germination, seedling growth, and stomatal movement . GRP7, as well as GRP8, binds to mRNA, which affects splicing and microRNA biogenesis (Koster et al., 2014;Streitner et al., 2008).
The Arabidopsis splicing factor protein STABILIZED1, a homolog to the human U5 small nuclear ribonuclearproteins -associated protein, is involved in pre-mRNA splicing and turnover of unstable transcripts. It also affects microRNA biogenesis via pri-miRNA splicing.
This protein is a regulator of plant responses to drought as well as to other abiotic stresses (Ben Chaabane et al., 2013;Kim et al., 2017;Lee et al., 2006;Shin et al., 2011).
This study aimed to confirm the above-described model showing that the level of selected MYB TFs affects plant tolerance to drought.
We show that downregulation of MYB33, MYB65 (as closely related to MYB33 in the sequence and with assumed redundancy in functionality), After selection, the seedlings were transferred to Jiffy pots and grown to seeds, which were collected (Szarzynska et al., 2009). The Agrobacterium-mediated transformation of potato plants was carried out as described in Pieczynski et al. (2013)

| Nucleic acid isolation, cDNA synthesis and PCR
Total RNA from 3-week-old or 4-week-old plant leaves was isolated from Arabidopsis plants and in vitro potato cultures using a Direct-zol RNA MiniPrep Kit (Zymo Research). A TRIzol™ reagent (Invitrogen)based protocol was used for the reverse transcription reaction performed with Superscript™ III Reverse Transcriptase (Invitrogen), and oligo-dT was used as a primer (Szarzynska et al., 2009) in accordance with the protocol provided by the manufacturer. The RNA was then cleaned with Turbo™ DNase (Invitrogen) according to the provided protocol.
Bands containing polymerase chain reaction (PCR) products were cut from the gel, and/or PCR products after the reaction were extracted from the gel and cleaned with a GeneJET Gel Extraction and DNA Cleanup Kit (ThermoFisher Scientific) according to the protocol provided by the manufacturer. The products were subsequently cloned into a pGEM ® -T Easy vector (Promega). Real-time PCR and quantative PCR (qPCR) calculations were performed as previously described by Szarzynska et al. (2009) and Sierocka et al. (2011). The Mann-Whitney U-test was used for statistical analyses. The following p-values were set as statistically significant: p < 0.05; p < 0.01; p < 0.001.
Genomic DNA was isolated from Arabidopsis plants for genotyping with a "fast" protocol by grinding small leaves in a 1.5 ml Eppendorf tube via a small plastic pestle and extraction buffer containing 10% sodium dodecyl sulfate (SDS), EDTA, NaCl, and Tris-HCl, after which the DNA was precipitated with isopropanol. According to the provided protocol, plasmid DNA was isolated with a GenElute™ Plasmid Miniprep Kit (Sigma Aldrich).

| Genetic constructs
Constructs for OE were prepared with a pMDC32 Gateway™ binary plasmid containing the CaMV 35S promoter, the hygromycin B phosphotransferase gene as a selective gene, and kanamycin and rifampicin resistance genes (Curtis & Grossniklaus, 2003). The cDNA sequences of the genes of interest were inserted into pENTR™/D-TOPO™ (Invitrogen) plasmids and transferred to pMDC32 using the Gateway™ LR Clonase™ II (Invitrogen) technique. The nucleotide sequences of A. thaliana MYB33 and MYB65 with mutated miR159 recognition site genes were kindly provided by Prof. A. A. Millar (Millar & Gubler, 2005) and were amplified using the primers kMYB33 a.thATG_F/R and kMYB65 a.th F/R ( Table 1).
tRE_ATG_FLAG_F/S33_ASCI, and kMYB65s.tRE_ATG_FLAG_F/ S65_ASCI were used (Table 1). In addition, a FLAG tag was also introduced into the 5 0 sites of the given MYB cDNA coding sequence. The Agrobacterium tumefaciens AGL1 strain was used for the floral dip transformation of A. thaliana (Clough & Bent, 1998), and the LBA4404:rif R pAL4404 strain was used for S. tuberosum transformation (Millam, 2006). The potato transformation procedure was performed as described by Wyrzykowska et al. (2016).

| RWC measurements
The RWC was measured as described previously (Pieczynski et al., 2013). All measurements for Arabidopsis were obtained as the average of five leaves of similar size from the middle part of the rosette and taken from two or three individual plants. The et al. (1997, 2000) and as described previously by Pieczynski et al. (2013) and Pei et al. (1997). The adaxial side was imprinted in clear nail polish.
Images of the specimens on microscope slides were taken with a Nikon

| Stomatal movement analysis
Stomatal aperture measurements were measured as described previously (Hugouvieux et al., 2001;Pei et al., 2000;Pieczynski et al., 2013;Savvides et al., 2012). Images of the epidermal peels and measurements of stomatal aperture were analyzed similarly to those used for stomatal density (see above). For all ABA concentrations, at least 10 images of Arabidopsis leaves from 2 to 3 plants and average measurements were assessed using the NIS-Elements Advance Research program (Zhang et al., 2016). Statistical analysis was carried out using the Mann-Whitney test.

| Cuticle thickness analysis
The procedure for evaluating leaf cross-sections and measuring cuticles via transmission electron microscopy was carried out according to Krzesłowska and Woźny (1996) and as previously described by Pieczynski et al. (2013). Statistical analysis was carried out using the  Figure S1).  The RWC data are shown as the means ± SDs of n = 3 independent experiments; Mann-Whitney test, p value: *p < 0.05; **p < 0.01; ***p < 0.001. Scale bar: 50 mm. OE, overexpression; RWC, relative water content; WT, wild-type.
properly developed green seeds ( Figure S5). These results indicate that overly high AtMYB101 gene OE negatively affects fertilization and induces embryo lethality. We decided to continue our studies on the role of AtMYB101 in plant drought tolerance using heterozygous plants, which were genotyped before each experiment. We also did not obtain Arabidopsis plants overexpressing potato MYB65. The expression of transgenic cassettes in selected Arabidopsis transgenic lines was tested using RT-PCR and western blotting ( Figure S4B,C).
For subsequent experiments, we selected three transgenic lines with the highest expression of each introduced transgene (A33 2-1, A33 6-4, A33 6-6, A65 3-1, A65 3-2, A65 5-4, A101 2-2-2, A101 2-11-2, A101 2-11-4, S33-1, S33-5, S33 3-11).  Interestingly, the numbers of stomata on the adaxial and abaxial sides of leaf blades, as well as trichome density, were affected in all mutant plants overexpressing MYB TFs in the same way as that in the case of the cbp80/abh1 mutants. When the abaxial leaf surfaces were inspected, the number of stomata on the Arabidopsis mutants showed a general increase compared to those on the WT plants, although this increase was not always statistically significant. The opposite trend, a statistically significant decrease in stomata number, was observed in the case of adaxial leaf surfaces in a majority of the mutant lines ( Figure S6A). The number of trichomes on the adaxial surface of leaf blades also showed a statistically significant increase in the majority of mutant lines compared to WT plants ( Figure S6B). In the case of the cuticle, we did not observe any morphological changes in its density; however, it was thinner in the majority of the mutant lines than in the WT plants ( Figure S7). Thus, OE of the MYB33, MYB65, and MYB101 genes suggests that stomatal and trichome density are at least partially under the control of the signaling pathway involving these genes and CBP80/ ABH1, which acts upstream, while cuticle thickness is not. Interestingly, the OE of the MYB33, MYB65, and MYB101 genes in potato plants did not affect stomatal or trichome density in the same manner in all the transgenic lines, which is in contrast to that which occurred in the Arabidopsis plants overexpressing MYB TFs ( Figure S11).

| DISCUSSION
The results of this work support the model of the plant drought tolerance pathway that is induced by the lack of the CBP80/ABH1 gene presented in our previous study (Pieczynski et al., 2013). As shown previously, the lack of this gene impairs miRNA159 induction, which is known to control MYB33, MYB65, and MYB101 gene expression in . The data are shown as the means± SDs of n = 3 independent experiments, with 30 stomata per data point; Mann-Whitney test, p value: *p < 0.05; **p < 0.01; ***p < 0.001. Stomatal aperture was measured in two out of three independent transgenic lines for all OE constructs. ABA, abscisic acid; OE, overexpression; WT, wild-type.
increased plant tolerance to drought (Pieczynski et al., 2013 (Millar & Gubler, 2005). High similarity in a sequence may indeed suggest some redundancy in function.
It was previously shown that MYB33 and MYB65 are involved in anther and pollen development (Dubos et al., 2010;Millar & Gubler, 2005). Moreover, it was suggested that both proteins act redundantly in these processes. However, MYB33 and MYB65 are not expressed equally in all plant tissues: MYB33 is expressed in all organs and tissues, with the highest expression of this gene occurring in germinating seeds, mature leaves, flowers, and carpels, while MYB65 is also expressed in all organs; however, the highest amount is detected only in germinating seeds and flowers. Generally, MYB33 is expressed at a higher level than MYB65, and MYB101 is expressed at the lowest level in all organs, with a slight increase in flowers and seeds during the first stages of germination (Winter et al., 2007). Arabidopsis MYB101, together with MYB97 and MYB120, has been shown to function as male factors that control pollen tube-synergid interactions during fertilization (Liang et al., 2013). Together, all these data show that subgroup) also revealed that this gene plays a role in the plant response to drought Wu et al., 2019;Zhao et al., 2018). Plant overexpressing MYB44 showed enhanced stomatal closure, which provided drought and salinity tolerance. Moreover, it was shown that AtMYB44 OE resulted in reduced expression of genes encoding protein phosphatase 2C phosphatases that are known to be negative regulators of ABA signaling. Maize transgenic plants overexpressing OsMYB55, which also belongs to the R2R3 subgroup, also exhibited enhanced tolerance to water deficiency (Casaretto et al., 2016). OE of the potato MYB TF StMYB1R-1, which belongs to another subgroup of MYB TFs (R1), also resulted in improved plant responses to drought and relatively rapid stomatal closure under drought (Shin et al., 2011). It was shown that StMYB1R-1 enhanced the expression of genes involved in the regulation of water loss. It is highly probable that the mode of action of MYB33/65/101 in the Arabidopsis and potato responses to drought is involved in the same or other pathways that negatively affect ABA signaling and the regulation of water loss.
Arabidopsis MYB33, MYB65, and MYB101 were also shown to inhibit cell division in vegetative plant tissues and thus inhibit plant growth (Allen et al., 2007;Millar & Gubler, 2005).