Maize—Azospirillum brasilense interaction: accessing maize’s miRNA expression under the effect of an inhibitor of indole-3-acetic acid production by the plant

Abstract

MicroRNA (miRNA) is a class of non-coding RNAs. They play essential roles in plants’ physiology, as in the regulation of plant development, response to biotic and abiotic stresses, and symbiotic processes. This work aimed to better understand the importance of maize’s miRNA during Azospirillum-plant interaction when the plant indole-3-acetic acid (IAA) production was inhibited with yucasin, an inhibitor of the TAM/YUC pathway. Twelve cDNA libraries from a previous Dual RNA-Seq experiment were used to analyze gene expression using a combined analysis approach. miRNA coding genes (miR) and their predicted mRNA targets were identified among the differentially expressed genes. Statistical differences among the groups indicate that Azospirillum brasilense, yucasin, IAA concentration, or all together could influence the expression of several maize’s miRNAs. The miRNA’s probable targets were identified, and some of them were observed to be differentially expressed. Dcl4, myb122, myb22, and morf3 mRNAs were probably regulated by their respective miRNAs. Other probable targets were observed responding to the IAA level, the bacterium, or all of them. A. brasilense was able to influence the expression of some maize’s miRNA, for example, miR159f, miR164a, miR169j, miR396c, and miR399c. The results allow us to conclude that the bacterium can influence directly or indirectly the expression of some of the identified mRNA targets, probably due to an IAA-independent pathway, and that they are somehow involved in the previously observed physiological effects.

Supplementary Information

The online version contains supplementary material available at 10.1007/s42770-023-01236-3.

Introduction

MicroRNA (miRNA) is a class of non-coding RNA with a length of about 20–24 nt [1, 2]. miRNA coding genes (miR) are transcribed by RNA polymerase II (Pol II) into pri-miRNAs, which are cleaved by a class of RNase-III nucleases called Dicer-like proteins, and then combined with members of ARGONAUTE (AGO) family proteins to form the RNA-induced silencing complexes (RISCs). The RISC complex cleaves mRNAs or represses their translation, thereby controlling gene expression [1, 3, 4]. It has been known for a long time that miRNAs play an important role in plants’ physiology [2–11]. They are involved in the regulation of plant development [2–5], response to biotic and abiotic stresses [4, 6, 9], and symbiotic processes [4].

Plant growth–promoting bacteria (PGPB) are a group of beneficial microorganisms that can colonize the rhizosphere, the phyllosphere, the root’s surface, and the plant’s internal tissues, stimulating plant growth [12, 13]. It is believed that PGPB can promote plant growth by combining several abilities, like the production of phytohormones, especially indole-3-acetic acid (IAA) [12, 14, 15]. One of the most well-known PGPBs is Azospirillum brasilense, widely used in South America as a cereal crop inoculant. Among this bacterium’s plant growth–promoting traits, the most studied is the ability to fix nitrogen and produce phytohormones (IAA, gibberellins, ethylene, and polyamines) [16–19]. One of these phytohormones, auxins (mainly IAA), is the most studied among those produced by both PGPB and plants [20–23]. In bacteria, IAA is used as a signaling molecule (stimulates root growth and carbohydrate exudation by plants), and it is involved in the quorum-sensing process [20, 22, 24]. In plants, auxins regulate various aspects of their development, such as cell growth and differentiation, the establishment of apical dominance, differentiation of xylem, suppression of abscission, and formation of apical and root meristem [20, 25].

In a previous study, the gene expression during the interaction between Azospirillum brasilense FP2 and maize was investigated under co-cultivation in the presence of the TAA/YUC pathway inhibitor, yucasin. Our main conclusion was that A. brasilense was able to revert the phenotype caused by yucasin through an IAA-independent pathway [26]. In the present work, the presence of differentially expressed maize’s miRNA and their predicted gene targets was investigated in the cDNA libraries constructed in that previous Dual RNA-seq study. This work aimed to verify if A. brasilense can also influence the expression of maize’s miRNA coding genes (miR) through a similar IAA-independent pathway, bringing some light on the role of miRNAs in cereal-bacteria interactions.

Materials and methods

Background

In a previous work [26], the effects of yucasin [5-(4-chlorophenyl)-4H-1,2,4-triazol-3-thiol], an inhibitor of IAA production through the TAA/YUC pathway [27], were studied in the maize transcriptome during a plant-bacterium interaction experiment. This interaction was with the plant growth–promoting bacterium Azospirillum brasilense FP2. In that study, maize seeds were inoculated or not with A. brasilense FP2, germinated, and then transferred to pots with sterilized sand. Yucasin was added 10 days after inoculation (DAI, 10 mL of an aqueous solution at 50 μM final concentration), and the root’s plantlets were sampled after 5 h of treatment and instantly frozen. Total RNA was extracted from roots, used to prepare cDNA libraries, and then sequenced. Root’s IAA concentration was also assessed, as well as the physiological effect of yucasin in the plantlets, which was evaluated by daily administration of this compound to the plantlets from 10 to 15 DAI. Concerning the IAA concentration in the plantlet’s roots after yucasin treatment, we observed that roots from plantlets from the AzoYuc group (which received yucasin and were inoculated with A. brasilense) presented lower IAA concentration than those from the Ctr (control plants that did not receive either yucasin or A. brasilense) and Azo (plants that were inoculated with A. brasilense) groups and that roots from plantlets from the Azo group presented higher IAA concentration than those from the Yuc (plants that received only yucasin) one.

Regarding the physiological data, we observed that plantlets from the AzoYuc group presented longer roots than those from the Ctr and Yuc ones. Plantlets from the AzoYuc group also presented longer aerial parts than those from the Yuc group. Finally, we observed that plantlets from the Yuc group presented a lower number of lateral roots than those from all the other groups. From the transcriptome analysis, we identified differentially expressed genes that are part of the IAA signaling and abscisic acid (ABA) response pathways and involved in the cell division in cDNA libraries obtained from roots of plants inoculated with the bacterium, treated with yucasin, or both [26]. In the present work, we continue to analyze the cDNA libraries obtained in that previous work, now focusing on the miRNAs and the expression of their target genes in response to the applied treatments.

RNA-seq libraries

Twelve cDNA libraries obtained from a previous Dual RNA-seq experiment [26] were investigated concerning the miRNA presence and expression. Libraries were composed of RNA extracted from four experimental groups, with three replicates: Ctr (control plantlets), Yuc (plantlets that received 50 μM of yucasin), Azo (plantlets inoculated with A. brasilense FP2), AzoYuc (plantlets inoculated with A. brasilense FP2 that received 50 μM of yucasin). The plantlet’s endogenous IAA concentrations and their physiological aspects (roots and aerial parts length and number of lateral roots) were also previously evaluated.

The 12 cDNA libraries were deposited in GenBank under the numbers SAMN12391479 to SAMN12391490.

Data analysis, differential gene expression, and miRNA gene target prediction

The reference genomes and the respective annotations were downloaded from the National Center for Biotechnology Information (NCBI) site (https://www.ncbi.nlm.nih.gov/). All the cDNA libraries obtained and the reference genomes with annotations were uploaded into the CLC Genomics Workbench (v. 8.0). The analysis was made according to our previous work [26] with the following modifications. Reads smaller than 18 nucleotides (nt) longer with low quality were removed from libraries using the standard setups of the CLC Genomics workbench. After mapping the reads and removal of the A. brasilense FP2 reads from Azo and AzoYuc groups (using A. brasilense Sp7 (GCA_001315015.1) genome as reference), maize mapped reads were further separated into two groups: (a) reads longer than (or equal to) 18 nt and smaller than (or equal to) 26 nt were separated from the libraries and counted against maize’s miRNA sequences available at miRBase (https://www.mirbase.org/index.shtml) [28] downloaded by CLC workbench; (b) reads longer than 26 nt were counted using the Zea mays cv. B73 (GCF_000005005.2) genome as a reference.

Count files were analyzed with the DESeq2 v1.36.0 [29] package of R software v 4.2.0 (R Development Core Team). Genes (including the miR genes) with p-values < 0.05 [30, 31] and log2fold-change [Lg2(FC)] ≥ |1| were considered differentially expressed (DEGs) [30]. Metabolic pathways were identified using the KEGG’s database (https://www.genome.jp/kegg/). Annotations for the DEGs were made with the assistance of the MaizeMine, online version 1.3 (http://maizemine.rnet.missouri.edu:8080/maizemine/begin.do) and NCBI.

Genes targeted by differentially expressed miRNAs were predicted by the online software psRNATarget [32] (https://www.zhaolab.org/psRNATarget/) with a modified parameter (expectation =3). To identify the predicted targets, mature miRNA sequences were downloaded from miRbase and provided to psRNATarge. Predict targets were manually identified and annotated with the aid of MaizeGDB (https://www.maizegdb.org/) and NCBI (https://www.ncbi.nlm.nih.gov/) databanks. After identifying the predicted targets, a search among the differentially expressed genes previously identified was made to identify among them those which are also predicted targets of at least one miRNA.

Results

Transcriptome analysis

The 12 cDNA libraries from a previous transcriptome study were analyzed to investigate the presence of miRNA, their expression pattern, and their probable targets. Table S1 shows a summary of library mapping for each experimental group.

Reads that mapped to the A. brasilense and Zea mays reference genomes were extracted from libraries of Azo and AzoYuc groups, and reads counting was performed separately using the respective annotated genomes. Since Ctr and Yuc groups were not inoculated, reads mapped to the Z. mays reference genome were extracted and counted using the Z. mays annotated genome. All reads that aligned in the intergenic regions, tRNA, and rRNA sequences were eliminated, and only the reads that aligned to coding sequences (CDS) or miRNA (maize) were further analyzed (Tables S1 and S2).

In maize, excluding the miR genes, at least 1150 differentially expressed genes (DEGs) were identified in pairwise comparisons between the experimental conditions (Table S3, bottom). Concerning the miR genes, 11 differentially expressed miRNA genes (miR-DEGs) were identified in pairwise comparisons between the experimental conditions (Table 1).

Table 1.

Maize differentially expressed miRNA genes (miR - DEGs) in all experimental conditions. Genes that presented |Log2(FC) ≥ 1| and p-value ≤ 0.05 were considered differentially expressed. Ctr control plantlets, Yuc plantlets that received 50 μM of yucasin, Azo plantlets inoculated with A. brasilense FP2, AzoYuc plantlets that received 50 μM of yucasin and were inoculated with A. brasilense FP2

	Yuc vs. Ctr		Azo vs. Ctr		AzoYuc vs. Ctr		AzoYuc vs. Yuc		AzoYuc vs. Azo
MIRs	log2(FC)	p-value	log2(FC)	p-value	log2(FC)	p-value	log2(FC)	p-value	log2(FC)	p-value
MIR156h					3.462804903	0.01863182
MIR159f							−3.721807222	0.045749755
MIR164a							2.726542956	0.013449108	2.362478512	0.030097325
MIR166a					−1.126579589	0.037841388
MIR167b									2.787527267	0.045541705
MIR169h									2.601441653	0.027581789
MIR169j							1.266823612	0.006570097	1.317809453	0.004436383
MIR171f					−3.117462872	0.049628457			−3.146622035	0.041811938
MIR396c					−4.023020363	0.019167993	−4.444370451	0.007687667	−4.529777917	0.006275422
MIR398a									2.895446205	0.017735649
MIR399b							2.525607218	0.001566492	1.726168541	0.020666171

A. brasilense and yucasin regulated the expression of miR genes in maize roots

The comparison between data from the AzoYuc group with data from the Ctr one shows that miR156h was upregulated (log2(FC) 3.46, Table 1), while miR166a, miR171f, and miR396c were downregulated (log2(FC) −1.13, −3.12, and −4.02, respectively, Table 1). When comparing data from the AzoYuc group with the Yuc one, miR164a and miR399b were upregulated (log2(FC) 2.73 and 2.52, respectively, Table 1). At the same time, miR159f and miR396c were downregulated (log2(FC) −3.72 and −4.44, respectively, Table 1). At last, when comparing data from the AzoYuc group with data from the Azo one, miR164a, miR167b, miR169h, miR398a, and miR399b were upregulated (log2(FC) 2.36, 2.79, 2.60, 2.89, and 1.73, respectively, Table 1). At the same time, miR171f and mir396c were downregulated (log2(FC) −3.14 and −4.52, respectively, Table 1).

A. brasilense, yucasin, and miRNAs regulated the expression of miRNA-predicted target genes

The miRNA targets were predicted using the online tool psRNATarget [32] (https://www.zhaolab.org/psRNATarget/). To make this prediction, we used the miRNA mature sequences (Table S4) of the 11 miR-DEGs (Table 1). Approximately 280 targets were predicted for the differentially expressed miRs observed in our analyses (Table S5). Among them, 17 genes were differentially expressed, and most of them were identified as transcription factors (Table 2) and responded to the presence of miRNAs, A. brasilense, yucasin, or IAA levels.

Table 2.

miR predicted targets differentially expressed in all experimental conditions. The marked areas indicate whether the corresponding miR gene is up-regulated (red) or down-regulated (blue). Ctr control plantlets, Yuc plantlets that received 50 μM of yucasin, Azo plantlets inoculated with A. brasilense FP2, AzoYuc plantlets that received 50 μM of yucasin and were inoculated with A. brasilense FP2

When comparing data from the Yuc group with data from the Ctr one, dcl4 (GeneID 100381550) was downregulated (log2(FC) −1.98, Tables 2 and S3), and myb115 (GeneID: 100383604) was upregulated (log2(FC) 2.18, Tables 2 and S3). Comparing data from the Azo group with data from the Ctr one, sbp3 (GeneID: 100278824) was upregulated, while dcl4 and nactf84 (GeneID: 100381550 and 606437, respectively) were downregulated (log2(FC) 3.69, −2.20 and −3.17, respectively, Tables 2 and S3). Comparing data from the AzoYuc group with data from the Ctr one, sbp27, bhlh133, myb122 and -87, abi8, and morf3 (GeneID: 100279234, 103641784, 103627314, 103638687, 103638691, 103631056, and 100286015, respectively) were upregulated (log2(FC) 2.27, 1.97, 2.45, 7.22, 4.19, 3.19, and 1.53, respectively, Tables 2 and S3), and dcl4 and myb22 (GeneID: 100273260) were downregulated (log2(FC) −2.19 and −1.67, respectively, Tables 2 and S3). On its turn, myb122, nactf108 (GeneID: 103650490), and morf3 were upregulated (log2(FC) 5.05, 1.72, and 1.61, respectively, Tables 2 and S3), and grftf1 and – 4 (GeneID: 100285734 and 100125643, respectively) were down-regulated (log2(FC) −1.29 and −3.67, respectively, Tables 2 and S3) when comparing data from the AzoYuc group with data from the Yuc one. Finally, when comparing data from the AzoYuc group with data from the Azo one, sbp27, bhlh133, myb122, and rld2 (GeneID: 100037826) were upregulated (log2(FC) 2.56, 2.68, 8.32, and 1.31, respectively, Tables 2 and S3), and myb138 (GeneID: 100191731), myb22, and hsftf14 (GeneID: 100191641) were downregulated (log2(FC) −2.06, −2.09, and −2.17, respectively, Tables 2 and S3).

Discussion

miR genes and miRNAs predicted target regulation

MicroRNAs (miRNAs) are important expression regulators, as they are involved in developmental regulation, defense against biotic and abiotic stresses, and interaction with beneficial organisms [2–6]. When analyzing our data, we noticed that 11 miR genes were differentially expressed in response to the presence of A. brasilense, yucasin, or IAA concentration (Table 1). Except for miR398a, all the miRNA genes observed belong to miRNA families already related in the literature as having important effects on plant development. According to Yan et al. [2], miRNAs belonging to miR156, miR159, miR167, miR169, and miR399 families are involved in repressing root elongation, while those belonging to the miR171 family promoted it. miRNAs from miR156 and miR528 families repressed lateral root formation, while those belonging to miR164 and miR171families promoted it [2]. miRNAs from the miR156 family repressed adventitious root formation, while those belonging to the miR171 family promoted it [2]. Kang et al. [8] observed that miRNA belonging to the miR families miR166, miR171, and miR319 were highly expressed during seed development in maize. Li et al. [9] observed that miRNAs from miR156, miR159, miR166, miR167, miR319, miR398, miR399, and miR528 families were involved in maize’s drought stress response.

Approximately 280 targets were predicted for the differentially expressed miRNAs observed in our research (Table S5). Some targets presented more than one alternative transcript as a predicted target, and some were predicted as the target of more than one miRNA. Seventeen predicted targets were identified among the differentially expressed genes (Table 2 and Table S5). Most of these predicted target’s code for transcription factors, which are master regulators of gene expression. Among the 11 differentially expressed miRs observed in our study, only five miRNAs had targets being differentially expressed in at least one pairwise comparison. Among the differentially expressed predicted targets, only eight were differentially expressed at the same time as the corresponding miRNA (Table 2).

In our previous study [26], when comparing the data from the AzoYuc group with data from the Ctr one, we observed that plantlets from the AzoYuc group presented longer roots than those from the Ctr group and that the IAA concentration in plantlets from the AzoYuc group was lower than those from the Ctr one. In the same comparison, we observed that miR156h was upregulated and miR166a, miR171f, and miR396c were downregulated (Table 1). The observed expression pattern, together with the data from our previous study, allows us to conclude that miR156h, miR166a, miR171f, and miR396c expression patterns were caused by the lower IAA concentration and the bacterium and that these miRNAs are involved in the physiological effect observed in this comparison. Concerning miR156h, miR166a, miR171f, and miR396c predicted targets (Table S5), we observed that only miR156h and miR166a predicted targets were differentially expressed in at least one pairwise comparison (Table 2). We observed three predicted targets of each gene being differentially expressed (Table 2). Among miR156h differentially expressed predicted targets, there were two members of the SQUAMOSA-PROMOTER BINDING PROTEIN (SBP) or SQUAMOSA-PROMOTER BINDING PROTEIN-LIKE (SPL) family. SBP/SPL is a family of transcription factors that critically regulate plant development and architecture in diverse plant species [33]. In our study, sbp3 and sbp27, which belong to SBP/SPL subgroups I and X, respectively [33], were both upregulated (Azo vs. Ctr and AzoYuc vs. Ctr, respectively (Table 2)), but only sbp27 was differentially expressed in the same comparison as miR156h (AzoYuc vs. Ctr). The third miR156h predicted target differentially expressed was a family member of the DICER-LIKE PROTEINS family. This protein family is involved in the biogenesis of the siRNAs (small interfering RNAs) [34]. In our study, dicer-like 4 (dcl4) gene was downregulated in three pairwise comparisons (Table 2), including the same as miR156h (AzoYuc vs. Ctr). Since miR156h was upregulated (AzoYuc vs Ctr), both gene targets were expected to be downregulated, but this happened only to dcl4. Observing the expression patterns of sbp27 and dcl4 together with our previous data [26] allows us to make some assumptions. Concerning sbp27, we can assume that the lower IAA concentration caused by the yucasin and the presence of the bacterium (AzoYuc vs Ctr) were responsible for this gene upregulation. On the other hand, dcl4 downregulation was directly caused by the upregulation of miR156h and indirectly by the lower IAA concentration caused by the yucasin and the presence of the bacterium. We can also assume that sbp27 upregulation and dcl4 downregulation are involved in the physiological effect observed when comparing plantlets from AzoYuc vs Ctr groups. Concerning miR166a predicted targets, three of them were differentially expressed in at least one pairwise comparison (Table 2), but only multiple organellar RNA editing factor 3 (morf3) gene was in the same comparison as miR166a (AzoYuc vs. Ctr). According to Luo et al. [35], the multiple organellar RNA editing factor/RNA-editing factor interacting proteins (MORFs/RIPs) gene family was first identified in Arabidopsis thaliana. The members of this gene family are required for RNA editing in mitochondria and several plastids [35–37]. In plants, they form six groups, and morf3 is part of group II together with A. thaliana and Oryza sativa morf3 genes [38]. Members of this gene family are involved in maize seed development [37], and mutations in these genes lead to abnormal development of plants [35]. Our study observed that morf3 was upregulated, while miR166a was downregulated. These facts together allow us to assume that morf3 upregulation was directly caused by the lower IAA concentration caused by the yucasin and the presence of the bacterium. Since miR166a was downregulated, it did not influence morf3 expression.

When comparing the AzoYuc group with the Azo one in our previous study [26], we observed that the IAA concentration in the AzoYuc group was lower than the Azo one. In this comparison, we now observed that miR164a, miR167b, miR169h, miR169j, miR398a, and miR399b were upregulated (Table 1). We also observed that miR171f and miR396c were downregulated (Table 1). The observed expression pattern together with the data from our previous study allows us to conclude that miR164a, miR167b, miR169h, miR169j, miR171f, miR396c, miR398a, and miR399b were responding to the lower IAA concentration caused by yucasin. Concerning these miRNAs’ predicted targets, we observed that only one of miR164a’s predicted targets was differentially expressed in this comparison (AzoYuc vs. Azo (Table 2)). When comparing AzoYuc vs. Azo, we observed that myb22 was downregulated. myb22 is part of the MYB transcription factor family. Members of this transcription factor family are involved in plant development, secondary metabolism, hormone signal transduction, and response to abiotic stresses [39, 40]. Based on this, our data suggest that myb22 was directly downregulated by miR164a in response to the lower IAA concentration caused by yucasin.

When comparing the AzoYuc group with the Yuc one in our previous study [26], although we did not observe any difference in IAA concentration, AzoYuc plantlets presented longer roots and aerial parts and a higher number of lateral roots than those from the Yuc group. In this comparison, we now observed that miR164a, miR169j, and miR399b were upregulated (Table 1). We also observed that miR159f and miR396c were downregulated (Table 1). The observed expression pattern and the data from our previous study allow us to make some assumptions (Fig. 1). First, miR159f, miR164a, miR169j, miR396c, and miR399b responded to the presence of the bacterium. Second, they were probably involved in the physiological effects observed in this comparison (AzoYuc vs. Yuc). Regarding the predicted targets of miR159f, miR164a, miR169j, miR396c, and miR399b (Table S5), we observed that only the genes myb122 (miR159f), nactf108 (miR164a), grftf4, and grftf1 (miR396c) were differentially expressed in this comparison (AzoYuc vs Yuc (Table 2)). myb122 and nactf108 were upregulated, and grftf1 and grftf4 were downregulated (Table 2). As myb22, myb122 is also a member of the MYB transcription factor family, which is involved in plant development, secondary metabolism, hormone signal transduction, and response to abiotic stresses [39, 40]. In its turn, nactf108 is a member of the NAC family. Members of this transcription factor family are plant-specific transcription factors and regulate diverse plant processes, such as nutrient acquisition, root growth, plant height, secondary wall synthesis, and cell division [41, 42]. The members of NAC family are also involved in the response to biotic and abiotic stress [42]. Finally, grftf1 and graftf4 are members of the growth-regulating factor (GRF) family. It is a small, plant-specific transcription factor family whose members are important in leaf and stem development as well as in flowering, seed, and root development, in controlling growth under stress conditions, and in regulating plant longevity. This transcription factor family expression is also regulated at the transcription level by the miR396 [43–45]. According to Omidbakhshfard et al. [43], miR396 and GRFs interact and module diverse developmental aspects of the plant. Taking the miR159f and myb122 expressions together, we were allowed to assume that the observed upregulation of the myb122 was indirectly caused by the bacterium through the repression of the miR159f. On the other hand, when we analyzed the expression patterns of nactf108 and miR164a, we observed that both were upregulated. The same behavior was observed when analyzing grftf1, grftf4, and miR396c expressions. All three were downregulated in the same comparison (AzoYuc vs. Yuc). This fact allows us to conclude that nactf108 upregulation and grftf1 and grftf4 downregulation were caused by the bacterium, despite the presence of miR164a or miR396c (Table 2). When analyzing together the expressions of these miRNAs and their predicted targets, the physiological effects, and the fact that there was no difference in IAA concentration in this comparison (AzoYuc vs. Yuc), we were allowed to conclude that these genes’ expression patterns (Fig. 1, Tables 1 and 2) were caused by A. brasilense through an IAA-independent pathway and these genes were involved in the physiological effects observed (AzoYuc vs Yuc).

Fig. 1.

Model to represent the interference of Azospirillum brasilense in the expression of maize’s miRNA genes and their predicted targets that leads to plant growth promotion. This model refers to the comparison between AzoYuc group vs. Yuc one. Red arrows indicate upregulation (genes) or increase in physiological traits, blue arrows indicate downregulation (genes). Black arrows indicate that A. brasilense was interfering with that gene’s expression (right arrows) through an IAA-independent pathway, which leads to the increasing in the physiological traits (left arrows). miRNA genes and their predicted targets are aligned side by side. miR159 = miRNA 159, miR164 = miRNA 164, miR169 = miRNA 169, miR396 = miRNA 396, miR399 = miRNA 399, nactf108 = NAC transcription factor 108, myb122 = MYB transcription factor 122, grftf1 = growth-regulating factor–transcription factor 1, grftf4 = growth-regulating factor–transcription factor 4. Created with BioRender.com

Conclusion

Gene expression analyses showed several miR genes responding to the bacterium, yucasin, the IAA concentration, or all together. Among them, we highlight the miRNAs miR159f, miR164a, miR169j, miR396c, and miR399c that were differentially expressed in response to the presence of the bacterium (AzoYuc vs. Yuc). This observation leads us to assume that A. brasilense can trigger the up- or downregulation of miR genes through an IAA-independent pathway.

Furthermore, gene targets of these miRNAs were predicted, and some were observed being stimulated or repressed in response to the respective miRNA, the bacterium, yucasin, the IAA concentration, or all together. Among them, when the bacterium was the only treatment present, we observed that the bacterium could regulate these gene expressions directly or indirectly (through miRNA). Most of the genes identified as miRNA targets are transcription factors.

Overall, this study contributes to a better understanding the plant–bacteria relationship and the complex interplay between miRNAs and their target genes in response to A. brasilense inoculation. These findings could potentially be exploited to develop more efficient plant–bacteria interactions and consequently increase plant growth promotion.

Data availability

The Data concerning the cDNA libraries were provided in the Materials and Methods section, RNA-seq libraries subsection.

Declarations

Conflict of interest

The authors declare no competing interests.