3.2. CBL7 is required for fungus-mediated growth promotion

CBL proteins are plant-specific Ca2+ sensors that decode transient calcium fluctuations through the interaction with CBL-interacting protein kinases (CIPKs) (Tang et al., 2020). It has previously been shown that several CBL proteins, i.e. CBL1, CBL8, CBL9, and CBL10, interact with CIPK23 to regulate the transport activity of HAK5 (Ragel et al., 2015). Hence, we speculate that CBL7 and CIPK13 may physically interact with each other to control HAK5 activity as their downstream target, because of their shared expression profiles in response to the infection of Arabidopsis seedlings withS. indica . To test this hypothesis, we first conducted a yeast two-hybrid analysis (Figure 2 ). The corresponding full-length sequences of CBL7 and CIPK13 were included into the pDEST-22/pDEST-32 yeast 2-hybrid system vectors. The direct interaction of the two proteins was monitored by the ability of transformed Saccharomyces cerevisiae HF7c cells to grow on selection medium (SD/-Trp/-Leu/-His) containing 5 mM 3-amino-1,2,4-triazole. Figure 2A provides evidence of a weak interaction between CBL7 and the protein kinase CIPK13. A GeneMANIA protein-protein interaction query (Warde-Farley et al., 2010), provided additional evidence for the direct interaction of CBL7 with two other CIPK proteins (Figure 2B ), CIPK9 and CIPK24. However, we were not able to verify the predicted interaction of CBL7 with these two CIPKs by yeast two-hybrid studies (Supplemental Image 2 ).
Next, we took a reverse genetics approach to assess the role of CBL7, CIPK13, and HAK5 in the fungus-mediated promotion of plant growth. To do so, we used two independent T-DNA insertion mutants for both CBL7and CIPK13 , as well as the previously described hak5mutant. The mutants were grown for seven days on ½ MS medium alongside with corresponding wild-type control plants, before the seedlings were transferred to PNM medium, where they were either co-cultivated withS. indica or a mock control. After ten days, the vertically grown plants were photographed, and the root system architecture was analyzed using the GiA Roots software. The statistical assessment of the monitored parameters facilitated a quantitative comparison of the growth promoting effect elicited by S. indica in the different genotypes. As shown in Figure 3A , the functional knockout ofCIPK13 had no impact on the total root network length. Bothcipk13 alleles exhibited a significant growth promotion through the co-cultivation with S. indica , similar to the response of the wild-type control plants. The same observation was made for thehak5 mutant (Figure 3C ). In contrast, the loss-of-function mutants of CBL7 , cbl7-1 andcbl7-2 , are characterized by a substantial reduction of the growth promoting effect observed in Col-3 control plants (Figure 3B ). The additional analysis of the total network area of the different groups of seedlings confirmed the obtained results (Table 1 ). Taken together, the results suggest that CBL7 plays an important role in the establishment/maintenance of the symbiosis, because one the one hand the functional inactivation of CBL7 coincides with a loss of the growth promoting effect normally exerted by the fungus. CIPK13 and HAK5, on the other hand, are seemingly likely not vital or replaceable in the plant-fungus interaction, given that the knockout mutants showed no significant alteration of their growth behavior when infected withS. indica .

3.3. A loss of CBL7 interferes with potassium distribution in Arabidopsis

To further investigate the biological functions of CBL7, we conducted another set of RNA-seq analyses comparing the transcriptional profiles of cbl7-2 and Col-3 (wt) plants under control conditions. After applying the same threshold as before (FDR padj. ≤ 0.05, log2FC ≥ 1.25), we identified only 20 induced and 53 repressed DEGs (Figure 4A,B ), respectively. The GO analysis of this reduced number of DEGs did not provide evidence for the significant enrichment of any GO term. As the arbitrary chosen expression threshold provided no significant results, we repeated the analysis with less stringent threshold values (FDR padj. ≤ 0.05, log2FC ≥ 0.35), giving 119 induced and 947 repressed DEGs, respectively (Figure 4A ,B ). Subsequent GO and KEGG enrichment analysis revealed the repression of 33 genes associated with the spliceosome (Figure 4C, Supplementary Data Sheet 3 ). Furthermore, we were able to identify the induction of 23 (padj. = 1·13-12) and 11 (padj. = 3.45·10-2) genes related to cell wall organization and ion transport GO terms, respectively (Supplemental Data Sheet 3 ). Among the latter genes, we found the high-affinity nitrate transporter NRT2.4 as well as the low-affinity nitrate transporter NPF2.9 . The involvement of CBL7 in the transcriptional regulation of nitrate transporters has previously been reported (Ma et al., 2015). Additionally, we again found that the high-affinity potassium transporter gene HAK5 was significantly induced (log2FC = 1.85, padj. = 0.022). Although our previous results questioned the role of HAK5 in the plant-fungus interaction, the reiterated appearance of HAK5 led us to analyze the potassium content in roots and shoots of cbl7-2and Col-3 to determine whether CBL7 could be involved in the regulation of potassium homeostasis in Arabidopsis. To quantify the K+ content in Arabidopsis seedlings, mutant and wt seeds were germinated on ½ MS and then transferred to square plates with modified Hoagland agar medium lacking glucose and containing either no KCl or 1 mM KCl. After three days of acclimatization, the seedling roots were inoculated with S. indica chlamydospores and mock treated, respectively. After another week of co-cultivation, the potassium contents were quantified by atomic emission spectroscopy. As shown inFigure 5 , the potassium content of the cbl7-2 roots grown at 0 mM KCl was similar to that of the Col-3 control plants. However, the previously described decrease in K+contents in response to the inoculation with S. indica (Conchillo et al., 2021) was only observed in control plants, but not incbl7-2 . Interestingly, the K+ level in the shoots of cbl7-2 plants was significantly lower than in the mock-treated wild type, equal to the level observed for Col-3 treated with the fungus. Although K+ levels showed a tendency to decrease in cbl7-2 shoots, there was no significant difference between mock and fungus-treated cbl7-2 plants. Plants grown on plates containing 1 mM KCl showed a similar picture (Figure 5B ), only wild-type Col-3 plants showed the expected decrease in K+ in response to the S. indica infection. Furthermore, we found a significant accumulation of K+in cbl7-2 roots, while the K+ content incbl7-2 shoots appeared to be significantly reduced. We have therefore concluded that CBL7 contributes to the regulation of the distribution of K+ in the plant. The functional loss of CBL7 is unequivocally linked to an altered K+distribution profile, which is characterized by an accumulation of K+ in the roots, most likely due to a lack of transport of K+ to areal plant tissues. At the same time, our data suggest that CBL7 could also play a key role in the observed decrease in K+ levels in plants challenged with S. indica , as cbl7 mutants present a reduced reduction of K+ when co-cultivated withS. indica .

3.4. The cbl7 mutant shows increased plant defense responses

The above results highlighted the transcriptional regulation ofCBL7 in response to the infection of Arabidopsis roots withS. indica and its involvement in K+partitioning in the plant. With the goal of further exploring the role of CBL7 in the establishment and maintenance of the symbiosis between Arabidopsis and S. indica , we compared the RNA-seq data surveyed from cbl7-2 knockout plants and corresponding Col-3 control plants infected with S. indica and mock treated, respectively (Supplementary Data Sheet 3 ). Quantitative analysis of differential gene expression revealed a substantial increase in induced genes in the cbl7 mutant relative to wt, while the number of repressed genes in cbl7 is clearly diminished (Figure 6A ). Furthermore, the examination of the possible logical relations between the two data sets revealed four non-overlapping and two overlapping groups for DEGs in cbl7-2 and Col-3 (S. indicavs mock) (Figure 6C ). Next, we performed a GO and functional network analysis to identify biological processes significantly affected by the loss of CBL7 . The main alterations in the cbl7mutant could be associated with plant defense-related processes, especially with metabolic pathways that are involved in the biosynthesis of secondary metabolites related to plant defense (Figure 6B,D ). The enriched genes in these GO terms included a substantial number of cytochrome P450 enzymes, such as CYP71B3 ,CYP71A12 , CYP71A13 , and CYP71B15 (PAD3), which are known to participate in glucosinolate and camalexin biosynthesis (Glawischnig, 2007; Frerigmann et al., 2016). Furthermore, we found an induction of the myrosinase genes BGLU34 and BGLU35 that are involved in the turnover of glucosinolates (Wittstock and Burow, 2010). The observation of an induction of plant defense-related compounds is further supported by the induction of transcription factors (TFs) NAC042 , WRK33 , and WRK51 , which have been linked with the regulation of camalexin and indole glucosinolate biosynthesis (Birkenbihl et al., 2012; Saga et al., 2012; Frerigmann and Gigolashvili, 2014; Zhou et al., 2020). Furthermore, we also observed an induction of WRKY70 , a TF involved in modulating cell wall-related defense responses (Li et al., 2017). In addition, the functional analysis of the transcriptomics data revealed an enrichment of glutathione S-transferases among the induced genes in cbl7-2 , which included the genes GSTU10 , GSTU12 , GSTF3 ,GSTF6 , and GSTF7 . Glutathione S-transferases are readily induced by a wide range of stress conditions, including bacterial and fungal infections (Dixon et al., 2002; Gullner et al., 2018). Considering that all these processes are observed in the cbl7loss-of-function mutant, it must be concluded that CBL7 is involved in the suppression of multilayered defense responses to facilitate the establishment of S. indica in the root apoplast. However, other processes that appear enriched in S. indica challenged wt plants but not in cbl7 , such as the induction of sucrose transporter genes and WRKY46 orchestrated abiotic stress responses, or the repression of bHLH100 and MYB72 mediated metal ion homeostasis, are likely to contribute to the significant difference in growth promotion between cbl7 and wt triggered by the fungus.
Based on our hypothesis of an increased defense response in thecbl7 knockout mutant, it must be expected that the roots of mutant plants are less well colonized by S. indica than comparable wt roots. To prove this, we analyzed root colonization incbl7 and Col-3 by qPCR and trypan blue staining. Indeed,root colonization of the cbl7 mutant was significantly reduced (Figure 7 ). Our experiments have shown that CBL7 plays a critical role in maintaining the equilibrium between symbiotic interaction and plant defense that keeps the endophytic fungus load under control.