Plant growth, pathogen inoculation, and leaf sample collection

Plant materials consisted of six common wheat cultivars whose seeds were obtained from the National Small Grains Collection (Agricultural Research Station in Aberdeen, WA, USA). Five of the common wheat cultivars tested were found to be highly valuable sources of Pt resistance genes: spring cultivars Artigas (“PI 192535”), NP846 (“PI 322263”), Glenlea (“CItr 17272”), Lerma Rojo (“CItr 13651”), and one winter wheat cultivar TX89D6435 (“PI 584759”). The susceptibility control (lack of molecular markers linked to Lr34, Lr46, or Lr67) was the spring cultivar Artigas* (“PI 73046”) [20, 40]. Identification of molecular markers for the Lr46/Yr29 locus of Artigas* and the listed Pt-resistant wheat cultivars analyzed in this paper was recently carried out by our research team [20]. The csLV46G22 marker linked to the Lr46/Yr29 locus (E. Lagudah, unpublished data) was identified in all five resistant wheat genotypes [20].

The experiment was conducted in a growth chamber under controlled conditions: the temperature was set at 18 °C during the day and 16 °C at night with a 16-hour photoperiod, a relative humidity of 60 to 70% was established. In addition, the emission spectrum of the light source was fixed with a photon flux of 572 μE. One month after setting up the experiment, the temperature in the phytotron was raised to 20 °C and 17 °C during the day and night, respectively. Plants were grown to the adult plant stage, i.e., the formation of the flag leaf; plants were inoculated at 34 days after sowing. Cultivars at the adult plant stage were inoculated with Pt by spraying a suspension of urediniospores at a concentration of approximately 5 × 10⁵ spores/mL. Inoculum suspension was an equimolar mixture of four Pt strains suspended in water with 1% v/v Tween 20 reagent and prepared immediately before inoculation [40, 41]. Fungal spores were collected from infected field experiments located in various parts of Poland [40]. Samples of leaf tissue fragments were collected at five time points: 0 (before inoculation), 6, 12, 24, and 48 hours post inoculation (hpi), in three biological replicates. Leaf tissue samples were immediately frozen in liquid nitrogen and stored at -80 °C.

Candidate gene expression analysis using Quantitative Reverse Transcription PCR (qRT-PCR)

Based on the work of Cobo et al. [29], 10 candidate genes for the Lr46/Yr29 gene (Table 1) whose sequences were found in the Ensembl Plants database for common wheat (https://plants.ensembl.org/Triticum_aestivum/Info/Index) were selected and downloaded in FASTA format. These sequences were used to design primers for qRT-PCR reactions, using the Primer3Plus tool (https://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi). The Lr46/Yr29 region described by Cobo et al. [29] involves four copies of the RECEPTOR-LIKE PROTEIN KINASE (RLK) gene, three copies of GLUCAN ENDO-1,3-BETA-GLUCOSIDASES (Glu), as well as one copy each of SORTING NEXIN (Snex), SUGAR TRANSPORTER (STr), and a TRANSCRIPTION FACTOR CARRYING THE WRKY DOMAIN (WRKY). Of all the candidate genes listed, only the STr gene has no other close analogues in the reference wheat genome [42]. The sequences of the two candidate genes Lr46-Snex and Lr46-WRKY were similar to one another, which allowed universal primers to be designed to detect any transcript including homologues in other chromosomes. For the Lr46-Snex gene, a pair of universal primers was designed to allow detection of potential homologue transcripts on chromosomes 1A, 1B, and 1D. Similar was true for the Lr46-WRKY gene, for which a pair of universal primers was designed to allow the detection of potential gene transcripts of homologues on chromosomes 1B and 1D. Regarding the RLK and Glu genes, very similar gene arrangements are found on the other two chromosomes: 1A and 1D, in the region similar to 1B chromosome. Therefore, a comparison was made of only those few copies of genes found in the QYr.ucw-1BL region described, on chromosome 1B. These genes are so highly differentiated that it is not possible to design universal primers for them. Based on the T. aestivum reference genome sequence (taxid: 4565), primers were designed. Isolation of total RNA from leaf tissue samples collected as biological replicates and at variable time points (0, 6, 12, 24, and 48 hpi) was performed using the Maxwell RSC Plant RNA Kit isolation (Promega, Madison, WI, USA). The concentration and purity of isolated total RNA were assessed using a NanoDrop spectrophotometer and A₂₆₀/A₂₈₀ ratio. First-strand cDNA synthesis was performed using the iScript^™ Reverse Transcription Supermix kit for RT-qPCR (Bio-Rad, Hercules, CA, USA), according to the protocol provided by the manufacturer. A temperature gradient PCR was performed for each investigated amplicon (each primer pair) to optimize the thermal profile applied. The gradient was set in the temperature range from 50 °C to 58 °C. The optimized amplification temperature was set at 53.5 °C after obtaining a specific product in a 2% agarose gel. In order to perform amplification standard curves, the amplicons obtained were individually purified using the QIAquick PCR Purification Kit (Qiagen Inc., Hilden, Germany) according to the protocol provided by the company. In the next step, a series of dilutions were performed for each purified amplicon to obtain standard curves of qRT-PCR efficiency (%E) and R² values (Table 2). The standard curves were developed according to the protocol described previously [43].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Presentation of 10 genes located in the candidate region for Lr46/Yr29 gene and QYr.ucw-1BL region, predicted function and their localization on reference wheat genome TGACv2 (GCA_900241085.1).

https://doi.org/10.1371/journal.pone.0309944.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Name and sequences of designed primers based on candidate gene sequences for Lr46/Yr29.

The table also shows qPCR efficiency (%E) and R² values by performing standard curves.

https://doi.org/10.1371/journal.pone.0309944.t002

qRT-PCR was performed using iTaq Universal SYBR Green Supermix and CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). A negative No Template Control (NTC) was also performed for each analyzed gene. The composition of the qRT-PCR reaction mix was as follows: supermix—5 μL, primers (10 μM)—0.5 μL each, nuclease-free water—3 μL, and cDNA template -1 μL. Following current MIQUE recommendations on the validity of real-time PCR analyses, three biological replicates were used for each gene type/timepoint for a given wheat cultivar, and three technical repeats were prepared for each [44]. The following temperature profile was used: initial denaturation for 3 min at 95 °C; then 40 cycles: denaturation for 10 sec at 95 °C, annealing of primers for 30 sec at 53.5 °C. Melt stage (melt curve): melting temperature range 65 °C to 90 °C; every 5 seconds the temperature was increased by 0.5 °C and a fluorescence measurement was taken. Data analyses were performed using Bio-Rad CFX Maestro software and the Gene Study tool (Bio-Rad Laboratories, Inc., Hercules, CA, USA), which allowed to compare the expression of all genes at different time points of each cultivar.

Reference genes for Real-Time Quantitative Reverse Transcription PCR (qRT-PCR)

Based on literature data, four reference genes were selected for stable expression: TUBβ, ARF, RLI [45], and EF2-1 [46]. After analyzing the standard curves for these reference genes, two genes, TUBβ and ARF, were selected and qRT-PCR analyses were performed for them on a test cDNA template [40]. The selection and testing of reference genes were developed according to a previously reported protocol [40]. The analysis of reference genes for the test samples was essential for statistical calculations because the expression of the reference gene was compared to the expression of the gene analyzed for each sample using the Gene Study tool (CFX Maestro from Bio-Rad Laboratories, Inc., Hercules, CA, USA).

Preparation of candidate gene amplicons for Sanger sequencing

Sanger sequencing was performed to confirm that the desired amplicons were obtained for the candidate genes for Lr46/Yr29 and to characterize the possible sequence polymorphism(s) among the resistant genotype (Glenlea cultivar) and the susceptible genotype (Artigas* cultivar). The majority of amplicons for sequencing were amplified from cDNA. However, Lr46-RLK1, Lr46-RLK2, and Lr46-RLK4 were exceptions that resulted from obtaining a low quantity and/or quality PCR products using cDNA; in these cases, genomic DNA was used as template, isolated from leaves using the GeneMATRIX Plant and Fungi DNA Purification Kit (EURx Ltd, Gdańsk, Poland), according to the protocol provided by the manufacturer. For sequence confirmation of the analyzed candidate genes in the Lr46/Yr29 locus, PCR reactions were performed for the respective primer pairs. GoTaq G2 Flexi DNA Polymerase reagents (Promega, Madison, WI, USA) were used to prepare the PCR reaction [40], otherwise identical in composition to qRT-PCR described above. The following amplification temperature profile was used: initial denaturation for 2 min at 94°C; followed by 35 cycles: denaturation for 1 min at 95 °C, annealing of primers for 30 s at 53.5 °C, extension for 30 s at 72 °C, extension for 30 s at 72 °C; final extension for 5 min at 72 °C and cooling to 4 °C. Purification of amplicons was performed using the QIAquick PCR Purification Kit (Qiagen Inc., Hilden, Germany), according to the protocol provided by the manufacturer. After purification, samples were resuspended in an elution buffer in a volume of 30 μL. Purified amplicons were directly sequenced using the Sanger method with the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems) and a 3730xl capillary DNA analyzer (Applied Biosystems) by Genomed (Warsaw, Poland). The results obtained were assembled using Geneious v8.1 software [47].

Validation of miRNAs and analysis of their expression using stem-loop RT-PCR and ddPCR

In this study, the expression of four miRNAs was analyzed: three related to the candidate gene Lr46-Glu2 (tae-miR5384-3p, tae-miR9780, tae-miR9775, tae-miR9780 also being complementary to Lr46-RLK2), and tae-miR164 complementary to Lr46-RLK3. The coding sequences of these candidate target genes downloaded from the Ensembl Plants database (https://plants.ensembl.org/Triticum_aestivum/Info/Index), were analyzed in the psRNATarget database (http://plantgrn.noble.org/psRNATarget/). The sequences of the four miRNAs analyzed were found in the miRBase database (https://www.mirbase.org/) and downloaded in FASTA format and using the Integrated DNA Technologies website (https://eu.idtdna.com/pages), stem-loop primers were designed for miRNA reverse transcription and ddPCR reactions according to the protocols [48–50]. The mirVana^™ miRNA Isolation Kit with phenol (Thermo Fisher Scientific, Waltham, MA, USA) was used to extract the RNA fraction containing miRNA, the concentration and purity of which were measured spectrophotometrically using NanoDrop. The extracted miRNA fraction was then reversely transcribed using SuperScript IV Reverse Transcriptase (Thermo Fisher Scientific, Waltham, MA, USA), using stem-loop primers and the following protocol: incubation at 16 °C for 30 min; 60 cycles at 30 °C for 30 sec, 42 °C for 30 sec, and 50 °C for 1 sec; incubation at 85 °C for 5 min [51]. To quantify the number of miRNA molecules in the plant samples, a ddPCR mixture composed of 10 μL of ddPCR Super Mix Eva Green, primers (the final concentration of each primer was 200 nM), template (reversely transcribed, elongated miRNA), and RNase-free H2O was used. A 20 μL reaction mixture was used to generate the reaction droplets in an 8-well cartridge using a QX100 droplet generator (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The droplets were carefully transferred to a 96-well ddPCR plate and heat-sealed with plastic covers with thermal foil (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Then, cDNA was amplified in a T100 PCR thermal cycler (Bio-Rad Laboratories, Inc., Hercules, CA, USA) under the following cycling conditions: denaturation at 95 °C for 5 min, followed by 40 cycles with a three-step thermal profile of denaturation at 95 °C for 30 sec, annealing at 58 °C for 30 sec, and extension at 72 °C for 45 sec. Subsequently, the products were kept at 72 °C for 2 min for the final extension. After amplification, the products were cooled to 4 °C for 5 min and then heated to 90 °C for 5 min and finally cooled again to 12 °C. The droplets were quantified in a QX100 droplet reader (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The data acquisition and analysis were conducted using QuantaSoft version 1.7 software (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Positive droplets containing amplification products were distinguished from negative droplets by setting the fluorescence amplitude threshold to the lowest value of the positive droplet cluster. Yeast tRNA-Thr (TGT) molecules with the encoded sequence were added to each RT reaction as an internal control.

Statistical analysis

The Kolmogorov–Smirnov test was used to evaluate the null hypothesis that a set of normalized expression data for a particular gene and cultivar comes from a normal distribution. To assess the homogeneity of variances, Bartlett’s test was used. Two-way analyses of variance (ANOVA) were carried out to determine the main effects of cultivar and time point as well as cultivar × time point interaction on the variability of the individual expression profiles of the analyzed genes of candidate genes for Lr46/Yr29. For comparison of means of expression in each examined time point after inoculation to expression before inoculation, the elementary contrasts were performed, conducted for each cultivar independently. The expression of the analyzed candidate genes is shown as heatmaps. Tukey’s Honestly Significant Difference tests (post-hoc factorial one- or two-way ANOVA, depending on ANOVA’s results) were performed in R v.4.4.0 [52] using package agricolae v.1.3–7 [53] at α = 0.05. We referred to the expression values of the candidate genes miRNAs analyzed as ’relative expression’ for the expression level, as our results represent differences in expression compared to the values before inoculation. For visualization of gene expression results, cluster analysis (UPGMA method) with Euclidean distances was performed. All these analyses were conducted using the GenStat statistical software package [54].

Sanger sequencing results

The Lr46/Yr29 region described in Cobo et al. [29] contains as many as four copies of the RECEPTOR-LIKE PROTEIN KINASE (RLK) gene and three copies of GLUCAN ENDO-1,3-BETA-GLUCOSIDASES (Glu) genes, and one copy each of the SORTING NEXIN (Snex) and the WRKY TRANSCRIPTION FACTOR (WRKY). The RLK and Glu genes occur in up to hundreds of copies throughout the wheat genome (https://www.ncbi.nlm.nih.gov/datasets/genome/GCF_018294505.1/, https://plants.ensembl.org/Triticum_aestivum/Info/Index). In addition, very similar gene arrangements are found on the two homoeologous chromosomes 1A and 1D in the same region as on 1B (Fig 1). Sequence similarity analysis of the Lr46-Glu3 amplicon after Sanger sequencing also revealed the presence of copies on chromosomes 2A, 2B, and 2D. Therefore, a comparison was made of only those few copies of genes found in the region described, on chromosome 1B. These genes are highly differentiated, and it was not possible to design universal primers for them. Therefore, specific primers were designed, for each copy of the gene.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Localization of 10 candidate genes for Lr46/Yr29 on chromosome 1B and homologues on wheat chromosomes 1A, 1D, 2A, 2B and 2D.

Candidate genes for Lr46/Yr29, including the QYr.ucw-1BL region, have been highlighted by colors as per their biological functions/family identified in the legend (bottom). The selected candidate gene homologues were selected based on comparative analysis following Sanger sequencing of the obtained amplicons. The position of marker csLV46G22 has also been indicated [29, 55]. ‘Ch1A’ and similar stand for the name of the chromosome and the abbreviation ‘Mb’ next to the scale stands for the Megabase unit referring to the physical length of the chromosomes. The genetic map was constructed using the MG2C_v2.1 tool [56].

https://doi.org/10.1371/journal.pone.0309944.g001

Analytical sequencing succeeded in obtaining the sequences for the majority of the target amplicons (Table 3). The full sequences of the obtained amplicons, location, and characterization of candidate genes for the Lr46/Yr29 gene can be found in S1 File. For the first candidate, Lr46-Glu1, the TraesCS1B02G453900.1 sequence was amplified, but with a single nucleotide polymorphism (G) at position 40 compared with the TraesCS1A02G422500.1 sequence (G). For the ‘Glenlea’ cultivar, we observed a C+G at position 40, yielding a comparable level of amplification of TraesCS1B02G453900.1 and TraesCS1A02G422500.1. The Lr46-Glu2 gene in the Artigas* cultivar amplified the TraesCS1A02G422900.1 sequence (with polymorphic A at position 38) and TraesCS1B02G454200.1 (with polymorphic G at position 38) (Fig 1). In the Glenlea cultivar, the exact TraesCS1A02G422900.1 homolog sequence was amplified (Fig 1). We failed to obtain a sequence that could be read correctly for Lr46-RLK1 despite exhaustive attempts. For the Lr46-RLK3, the sequence of TraesCS1B02G454400.1 gene was amplified, with a small level of homolog TraesCS1D02G431000.1 (position 52) (Fig 1). The last candidate gene analyzed using the Sanger method was Lr46-WRKY, for which we obtained an amplicon (118 bp) of the TraesCS1D02G431600.1 gene, thus a homologue of the candidate gene TraesCS1B02G455000.1(Fig 1, Table 3, S1 File).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Description of the resulting amplicons and candidate genes for Lr46/Yr29 by Sanger sequencing.

https://doi.org/10.1371/journal.pone.0309944.t003

Expression of candidate genes in analyzed varieties at different time-points (hpi)

The qRT-PCR analyses of candidate genes in all studied wheat cultivars and at different time points are summarized as heatmap graphs (Fig 2). The raw data obtained from candidate gene expression analyses for Lr46/Yr29 are presented in S1 Table. The elemental contrast values of the candidate genes tested are shown in S2 Table. Regarding the gene Lr46-Glu1, its expression varied greatly among the analyzed wheat cultivars as well as among the time points, therefore we did not observe any major common expression pattern (Fig 2). A decrease in the expression level of Lr46-Glu2 gene was observed immediately after inoculation (6 hpi) for all cultivars except Glenlea. At 12 hpi, a further decrease in Lr46-Glu2 expression levels was observed for all resistant cultivars (Fig 2). The inflection point occurred at 24 hpi when the defense response was activated for all cultivars. The expression level of Lr46-Glu2 far exceeded the baseline level (0 hpi). At 48 hpi, the expression level of Lr46-Glu2 decreased in all cultivars but still exceeded the baseline in Artigas, Glenlea, Lerma Rojo, and TX89D6435 cultivars. For the Lr46-Glu3 gene, expression before as well as after inoculation was at very low levels (Fig 2). The only exception was the Glenlea cultivar at 24 hpi, where gene expression increased significantly. In the genes Lr46-RLK1 and Lr46-RLK2, the observed expression increased at 6 hpi, shortly after inoculation, then decreased at 12 hpi, and we observed a renewed increase at 24 hpi. The highest expression levels at 6 hpi and 24 hpi, were observed in the NP846 cultivar and at 6 hpi in the ‘Glenlea’ (Fig 2). For Lr46-RLK3, the highest increases in expression after inoculation were observed in the cultivars TX89D6435 and NP946 at 24 hpi; whereas, in ‘Glenlea’ at 6 hpi and 24 hpi. However, this gene was expressed at relatively low levels (Fig 2). A special case is the expression of the Lr46-RLK4 gene, whose highest expression values were observed in the Artigas* cultivar, in which, according to our team’s study, no molecular markers associated with the resistance conferred by Lr34, Lr46, or Lr67 were identified [20, 40]. Expression of the Lr46-RLK4 gene in the Artigas* cultivar was very high already before inoculation at 0 hpi (Fig 2). The Lr46-Snex gene showed low expression with two exceptions: at 12 hpi in the Artigas* cultivar and at 24 hpi in ‘Glenlea’ (Fig 2). In the Lr46-WRKY gene, we observed the highest values in ‘Artigas*’ at all time points except 48 hpi and in the Glenlea cultivar at 24 hpi (Fig 2). The qRT-PCR analyses of the Lr46-STr gene obtained many non-specific PCR products, which was particularly evident in the melt curve; as such, we decided to discard this gene’s data from further analyses (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Expression heatmap of the nine Lr46/Yr29 candidate genes in leaf tissues under Puccinia triticina infection.

The graphs were generated based on the elemental contrasts carried out for each cultivar independently, and separately for each candidate gene. Tukey’s HSD test post ANOVA was carried out and the groups are shown on the heatmap as various letters, for statistically different groups at α = 0.05. Groups indicated on the heatmap denote significant interaction of cultivar × time point (hpi), whereas groups along horizontal axis (time points [hpi]) and/or vertical axis (cultivars) denote significance for single factors but not their interaction. The evaluation included reference cultivars of common wheat (Artigas*, Artigas, Glenlea, Lerma Rojo, TX89D6435, and NP846) indicated at the vertical axes at various time points (0, 6, 12, 24, and 48 hpi) identified at horizontal axes. Red indicates lower normalized expression and yellow—higher expression.

https://doi.org/10.1371/journal.pone.0309944.g002

A two-way analysis of variance (ANOVA) was performed and identified the differences in the individual expression profiles of the analyzed genes for the cultivars at specific time points (Table 4). An additional source of variation was the cultivar × time point interaction. As each gene has a specific expression level, the analysis of variance was performed for each gene separately (Table 4). Cultivars exhibited unique patterns of gene expression in response to Pt infection. For example, ’Glenlea’ showed particularly high expression levels for several genes at 24 hpi, whereas ’Lerma Rojo’ exhibited relatively low expression levels across the majority of analyzed time points (S1 Table; Fig 2). For the majority of the Lr46/Yr29 candidate genes, their expression significantly varied in time across the analyzed cultivars. Each gene responded differently to pathogen infection and varied in its expression levels across cultivars and time points. Notably, Lr46-Glu2 consistently showed high expression levels across all cultivars and time points, indicating its potential importance in the defense response against Pt. For the Lr46-Glu2 gene, we obtained comparably the highest values of mean squares compared to the other candidate genes (Table 4). The results obtained may thus suggest that the Lr46-Glu2 gene harbors a special role in the Pt resistance response. In contrast, Lr46-Snex showed relatively low expression levels compared to other genes and exhibited variation across the analyzed time points but not among the cultivars. The analysis across the candidate genes for Lr46/Yr29 revealed significant effects of both cultivar and time on gene expression in wheat. Notably, the susceptible cultivar Artigas* consistently exhibited lower expression levels compared to the resistant cultivars for various analyzed genes, which indicated its distinct gene expression profile under the Pt stress. For several analyzed genes, including Lr46-Glu1 and Lr46-Glu2, significant interactions between cultivar and time underscored the dynamic nature of gene expression in response to temporal factors and genetic variation. Post-hoc analysis further elucidated these findings and revealed distinct expression patterns among wheat cultivars across different time points (Fig 2). Resistant cultivars often exhibited comparatively higher expression levels compared to the susceptible control cultivar, thus emphasizing their potential role in mediating wheat’s defense response against Pt.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Mean squares from two-way analysis of variance (ANOVA) for the individual expression profiles of the analyzed genes of candidate genes for Lr46/Yr29.

https://doi.org/10.1371/journal.pone.0309944.t004

A comparison of expression profiles between genes was also attempted by correlation analysis (Fig 3). Among the candidate genes tested, the highest positive correlation was observed for the expression profile between candidate genes: Lr46-Glu2 and Lr46-RLK2 (0.794; p < 0.001) (Fig 3). There was also a positive correlation for the Lr46-Glu1 gene between the genes: Lr46-Glu2, Lr46-RLK1, Lr46-RLK2 and Lr46-RLK3, where there were high correlations between the Lr46-Glu1 gene and the genes: Lr46-Glu2 and Lr46-RLK3 (0.726 and 0.730, respectively; p < 0.001) (Fig 3). The Lr46-Glu2 gene showed positive correlations (p < 0.001) with the other candidate genes tested except for genes: Lr46-RLK4, Lr46-Snex, and Lr46-WRKY, where the correlation was close to 0 with p > 0.05 (Fig 3). The Lr46-Glu3 gene showed a positive and statistically significant correlations with the majority of candidate genes except Lr46-RLK1 and Lr46-RLK2 (0.108 and 0.093, respectively, not statistically significant). Interestingly, for the genes: Lr46-RLK1, Lr46-RLK2, and Lr46-RLK3, we observed a general presence of correlation with Glu or RLK candidate genes, but general absence of correlation with other candidate genes (Fig 3). However, a distinct trend was observed for Lr46-RLK4, where we observed correlations with only two genes: Lr46-Snex and Lr46-WRKY (0.495 and 0.702, respectively; p < 0.001). Interestingly, there was a tendency between the Lr46-RLK4, Lr46-Snex, and Lr46-WKRY genes to correlate with each other (Fig 3). The exception is Lr46-Snex, where there is a high correlation also between the Lr46-Glu3 gene (0.735, p < 0.001).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Heat maps of Pearson’s linear pairwise correlation coefficients between the observed candidate genes for Lr46/Yr29 based on the values of their individual expression profiles.

Significance of pairwise correlations is marked: * p < 0.05, ** p < 0.01, *** p < 0.001. Heatmap color scheme (labeled at the bottom) informs of more positive (red hue) or negative (blue hue) correlations.

https://doi.org/10.1371/journal.pone.0309944.g003

Expression of miRNAs complementary to target candidate genes

Using stem-loop RT-PCR and ddPCR, we performed expression analysis of miRNAs selected from databases as complementary to several candidate genes for Lr46/Yr29. Contrast analysis was performed for the expression data of miRNAs, and the resulting elemental contrast values for the expression of miRNAs complementary to the candidate genes can be found in S3 Table. We compared the expression results of the candidate genes with the respective values for the miRNAs complementary to them. According to the databases, tae-miR9780 is complementary to two candidate genes, Lr46-Glu2 (TraesCS1B02G454200.1) and Lr46-RLK2 (TraesCS1B02G454100.1). Towards the Lr46-Glu2 gene, tae-miR9775 is also complementary. Unfortunately, the expression of tae-miR9775 was at a low level and did not allow for confident inferences. In the Lr46-Glu2 gene, we analyzed the expression of two miRNAs: tae-miR9780 and tae-miR5384-3p. The expression of tae-miR9780 in the majority of the tested wheat cultivars was at a constant level, both before and after inoculation. The exceptions were the TX89D6435 and Lerma Rojo cultivars where values higher than at 0 hpi were observed (Fig 4). Especially for the TX89D6435 cultivar, the results suggested an up-regulation. The tae-miR5384-3p expression showed fairly constant expression levels at all time points. Here, again in the Lerma Rojo cultivar, the miRNA was expressed with values higher than at 0 hpi. At 6 hpi, expression increased with the Lr46-Glu2 gene, whereas it decreased slightly at 12 hpi with a large increase in its complementary candidate gene expression. On the other hand, at 24 hpi, tae-miR5384-3p expression increased again together with an increase in its complementary candidate gene, whereas at 48 hpi we observed a large decrease in Lr46-Glu2 expression together with an increase in tae-miR5384-3p (Fig 4). In the case Lr46-RLK2, to which tae-miR9780 is also complementary, relationships similar to those of Lr46-Glu2 were observed. The expression of both Lr46-RLK2 and tae-miR9780 showed very low values (Fig 5). In contrast, tae-miR164, which is complementary to the candidate gene Lr46-RLK3 (TraesCS1B02G454400.1), showed differential expression at pre- and post-inoculation. The expression of the Lr46-RLK3 gene and its complementary tae-miR164 is shown in Fig 6 and suggests that tae-miR164 is a molecule that may be down-regulating its target gene’s expression. In the majority of the tested wheat cultivars, when candidate gene expression increased, then tae-miR164 was down-regulated, and vice versa: when miRNA expression increased then candidate gene expression was down-regulated (Fig 6).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Analysis of complementary miRNAs expression compared to Lr46-Glu2 expression patterns during Puccinia triticina infection.

The numbers 6, 12, 24, and 48 stand for hours post inoculation (hpi). The graphs show relative expression values compared to the 0 hpi time point. The error bars indicate the standard error (SE). Data were obtained from the average of three biological replicates, and each biological replicate had three technical replicates, according to the MIQUE protocol.

https://doi.org/10.1371/journal.pone.0309944.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Analysis of complementary miRNA expression compared to Lr46-RLK2 expression patterns during Puccinia triticina infection.

The numbers 6, 12, 24 and 48 stand for hours post inoculation (hpi). The graphs show relative expression values compared to the 0 hpi time point. The error bars indicate the standard error (SE). Data were obtained from the average of three biological replicates, and each biological replicate had three technical replicates, according to the MIQUE protocol.

https://doi.org/10.1371/journal.pone.0309944.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Analysis of complementary miRNA expression compared to Lr46-RLK3 expression patterns during Puccinia triticina infection.

The numbers 6, 12, 24 and 48 stand for hours post inoculation (hpi). The graphs show relative expression values compared to the 0 hpi time point. The error bars indicate the standard error (SE). Data were obtained from the average of three biological replicates, and each biological replicate had three technical replicates, according to the MIQUE protocol.

https://doi.org/10.1371/journal.pone.0309944.g006