**3.3. The metabolome of Arabidopsis sRNA mutants in responses to** *C. higginsianum*  **infection**

To compare the changes in the metabolomic profile of sRNA mutants and wild-type plants induced by *C. higginsianum* infection, an UHPLC-QTOF-based analysis of secondary metabo‐ lites was performed. The metabolomic fingerprinting provided a global view on the metabolic perturbations induced by *C. higginsianum* attack at 4 dpi. Comparison of the metabolome *hen1‐1* and *hyl1‐2* mutants by a principal component analysis (PCA) resulted in a clear separation of both control-treated and -infected mutants and their respective wild types (**Figure 5(a)** and **(b)**).

**Figure 3.** Disease severity of *C. higginsianum* in *A. thaliana* sRNA mutants and wild‐type plants; (a). *hen1‐1* mutant, (b). *hyl1‐2* mutant, (c). *rdr6‐15* mutant, compared to the respective wild-type background. Fungal growth was determined by quantifying the fluorescent area of *C. higginsianum‐*GFP in mm<sup>2</sup> at different time points in all *A. thaliana* mutants compared to wild type. Severity was determined as percentage of leaf area affected. For statistical analysis, a oneway ANOVA was applied; asterisks indicate statistically significant differences (*P* < 0.05). Error bars indicate standard deviation (SD).

Signs of Silence: Small RNAs and Antifungal Responses in *Arabidopsis thaliana* and *Zea mays* http://dx.doi.org/10.5772/intechopen.69795 17

**Figure 4.** Quantification of phytohormones in *A. thaliana* sRNA mutants and wild‐type plants after *C. higginsianum* infection. (a). salicylic acid (SA), (b). jasmonic acid (JA), and (c). abscisic acid (ABA) in *A. thaliana* sRNA mutants (*hen1‐1, hyl1‐2*, and *rdr6‐15*) and wild-type plants (Col-0 and Ler) under two treatments: infected with *C. higginsianum* and control. Statistical significance was determined using one-way ANOVA. Letters indicate statistically significant differences (*P* < 0.05). Error bars indicate standard deviation (SD).

**Figure 3.** Disease severity of *C. higginsianum* in *A. thaliana* sRNA mutants and wild‐type plants; (a). *hen1‐1* mutant, (b). *hyl1‐2* mutant, (c). *rdr6‐15* mutant, compared to the respective wild-type background. Fungal growth was determined

compared to wild type. Severity was determined as percentage of leaf area affected. For statistical analysis, a oneway ANOVA was applied; asterisks indicate statistically significant differences (*P* < 0.05). Error bars indicate standard

at different time points in all *A. thaliana* mutants

by quantifying the fluorescent area of *C. higginsianum‐*GFP in mm<sup>2</sup>

deviation (SD).

16 Plant Engineering

The PCA performed for *rdr6‐15* grouped the mutants and wild type much closer (**Figure 5(c)**). The metabolomic fingerprinting allowed identifying groups of putative antifungal metabolites that were normally induced in the wild type, for which in turn the mutants showed an abnor‐ mal induction pattern. After PCA analysis, the compounds showing the greatest difference between wild type and mutants were selected for further identification (**Table 1**). Compounds were identified by exact mass, fragmentation spectrum and the retention time of the fragments using the online free databases Metlin, MassBank, Kegg and Aracyc and the in-house database from the chemical analytical service of the University of Neuchatel. The metabolomic analysis revealed a group of glucosinolates, flavonols, phenylpropanoids and the phytoalexine camal‐ exine that were differentially induced in mutants and wild-type plants (**Table 1**). In response to *C. higginsianum*, *hen1‐1* mutant showed lower fold induction of some glucosinolates like 7 methylthioheptyl glucosinolate, glucoerucin, glucoiberin, glucoiberverin and glucolesquerellin. Moreover, glucobrassicin was not induced after infection in *hen1‐1* plants. Kaempferol 3-Orhamnoside-7-O-rhamnoside (kaempferol 3-rha-7-rha) and kaempferol 3-O-rhamnoside-7-Oglucoside (kaempferol 3-rha-7-glu), flavonols which are well-described antifungal compounds [35], were down‐regulated in *hen1‐1* and Ler plants as well as the phenylpropanoids sinapoyl malate and 1-O-β-D-glucopyranosyl sinapate. The phytoalexin camalexin was the most induced compound after infection in *hen1‐1* and Ler plants. Ler showed 84.4-fold induction of cama‐ lexin while infected *hen1‐1* contained 10.7 more than mock-treated plants. The *hyl1‐2* mutant exhibited lower fold induction in most of the glucosinolate levels compared to Col-0 (**Table 1**). Glucobrassicin, glucoiberin and glucoiberverin levels were higher in *hyl1‐2* control and infected treatments than in wild type plants. Moreover, the induction of kaempferol 3-rha-7-rha and kaempferol 3-rha-7-glu was higher in Col-0 than *hyl1‐2* plants. Levels of sinapoyl malate and 1-O-β-D-glucopyranosyl sinapate were also lower in *hyl1‐2* control and infected plants com‐ pared to Col‐0. Camalexin was 72.9‐fold induced in Col‐0 and 69.0‐fold induced in *hyl1‐2*. The *rdr6‐15* mutant exhibited lower fold induction of all glucosinolates, flavonols and phenylpro‐ panoids mentioned in **Table 1** compared to Col‐0. The fold induction of camalexin was similar in *rdr6‐15* mutant compared to Col‐0 plants. In summary, sRNA mutant *hen1‐1* exhibited lower levels of pathogen-induced camalexin, whereas the glucosinolates, flavonol and phenylpro‐ panoid compounds were slightly less prominently induced in response to fungal infection in all the mutants compared to their respective wild‐type plants.

#### **3.4.** *C. graminicola***‐infected maize sets up a tissue‐specific miRNA profile which is not directly linked to plant defence**

Using annotated maize miRNAs (zma), known miRNAs were classified in the different maize sRNA libraries. In order to determine biostress-specific miRNAs and to quantify their expres‐ sion level in the treated samples, the fold change expression was determined by calculating the relative difference of sequence reads in treated samples compared to the control librar‐ ies. Selected miRNAs showing a fold change of >2 are summarized in **Table 2**. Comparing biotrophic and necrotrophic fungal infection stages to mock, zma-miR479, zma-miR1318 and zma-miR1432 were found to be up-regulated; however, their fold induction was higher during the necrotrophic stage. Other miRNAs such as zma‐miR393, zma‐miR1120 and zma‐miR2092 showed an altered expression level exclusively during the biotrophic stage. By contrast, the Signs of Silence: Small RNAs and Antifungal Responses in *Arabidopsis thaliana* and *Zea mays* http://dx.doi.org/10.5772/intechopen.69795 19

The PCA performed for *rdr6‐15* grouped the mutants and wild type much closer (**Figure 5(c)**). The metabolomic fingerprinting allowed identifying groups of putative antifungal metabolites that were normally induced in the wild type, for which in turn the mutants showed an abnor‐ mal induction pattern. After PCA analysis, the compounds showing the greatest difference between wild type and mutants were selected for further identification (**Table 1**). Compounds were identified by exact mass, fragmentation spectrum and the retention time of the fragments using the online free databases Metlin, MassBank, Kegg and Aracyc and the in-house database from the chemical analytical service of the University of Neuchatel. The metabolomic analysis revealed a group of glucosinolates, flavonols, phenylpropanoids and the phytoalexine camal‐ exine that were differentially induced in mutants and wild-type plants (**Table 1**). In response to *C. higginsianum*, *hen1‐1* mutant showed lower fold induction of some glucosinolates like 7 methylthioheptyl glucosinolate, glucoerucin, glucoiberin, glucoiberverin and glucolesquerellin. Moreover, glucobrassicin was not induced after infection in *hen1‐1* plants. Kaempferol 3-Orhamnoside-7-O-rhamnoside (kaempferol 3-rha-7-rha) and kaempferol 3-O-rhamnoside-7-Oglucoside (kaempferol 3-rha-7-glu), flavonols which are well-described antifungal compounds [35], were down‐regulated in *hen1‐1* and Ler plants as well as the phenylpropanoids sinapoyl malate and 1-O-β-D-glucopyranosyl sinapate. The phytoalexin camalexin was the most induced compound after infection in *hen1‐1* and Ler plants. Ler showed 84.4-fold induction of cama‐ lexin while infected *hen1‐1* contained 10.7 more than mock-treated plants. The *hyl1‐2* mutant exhibited lower fold induction in most of the glucosinolate levels compared to Col-0 (**Table 1**). Glucobrassicin, glucoiberin and glucoiberverin levels were higher in *hyl1‐2* control and infected treatments than in wild type plants. Moreover, the induction of kaempferol 3-rha-7-rha and kaempferol 3-rha-7-glu was higher in Col-0 than *hyl1‐2* plants. Levels of sinapoyl malate and 1-O-β-D-glucopyranosyl sinapate were also lower in *hyl1‐2* control and infected plants com‐ pared to Col‐0. Camalexin was 72.9‐fold induced in Col‐0 and 69.0‐fold induced in *hyl1‐2*. The *rdr6‐15* mutant exhibited lower fold induction of all glucosinolates, flavonols and phenylpro‐ panoids mentioned in **Table 1** compared to Col‐0. The fold induction of camalexin was similar in *rdr6‐15* mutant compared to Col‐0 plants. In summary, sRNA mutant *hen1‐1* exhibited lower levels of pathogen-induced camalexin, whereas the glucosinolates, flavonol and phenylpro‐ panoid compounds were slightly less prominently induced in response to fungal infection in all

the mutants compared to their respective wild‐type plants.

**directly linked to plant defence**

18 Plant Engineering

**3.4.** *C. graminicola***‐infected maize sets up a tissue‐specific miRNA profile which is not** 

Using annotated maize miRNAs (zma), known miRNAs were classified in the different maize sRNA libraries. In order to determine biostress-specific miRNAs and to quantify their expres‐ sion level in the treated samples, the fold change expression was determined by calculating the relative difference of sequence reads in treated samples compared to the control librar‐ ies. Selected miRNAs showing a fold change of >2 are summarized in **Table 2**. Comparing biotrophic and necrotrophic fungal infection stages to mock, zma-miR479, zma-miR1318 and zma-miR1432 were found to be up-regulated; however, their fold induction was higher during the necrotrophic stage. Other miRNAs such as zma‐miR393, zma‐miR1120 and zma‐miR2092 showed an altered expression level exclusively during the biotrophic stage. By contrast, the

**Figure 5.** Metabolites distribution in sRNA mutants and wild-type plants upon *C. higginsianum* infection and control treatment. Principal component analysis (PCA) score plot of the metabolome. of the sRNA mutants *hen1‐1* (a), *hyl1‐2* (b), *rdr6‐15* (c) and the wild-type Ler and Col-0 upon 4 dpi with *C. higginsianum* infection and control treatment. The PCA analyses were performed using Marvis Filter and Cluster packages, following a Kruskal-Wallis test (*P* < 0.05). Each data point represents one replicate of six independent biological replicates.


**Table 1.** Fold induction of metabolites in sRNA mutants and controls upon *C. higginsianum* infection.

expression of zma‐miR168, zma‐miR2916 and zma‐miR5205 was altered only during the necro‐ trophic stage. Notably, zma-miR1432 and zma-miR2092 were also up-regulated in infected roots, suggesting that some miRNAs are regulated organ independently. Notably, infected roots showed also a distinct expression profile with zma-miR166, zma-miR169 and zma-miR395 that were down‐regulated, whereas zma‐miR909 and zma‐miR2863 were up‐regulated. A dif‐ ferent situation was found in systemic leaves upon leaf infection. Compared to local infected tissues, less miRNAs showed an altered expression. For instance, zma-miR397, zma-miR916 and zma‐miR5169 were up‐regulated. In systemic leaves upon root infection, zma‐miR1877 and zma‐miR2592 were down‐ and up‐regulated, respectively. Interestingly, zma‐mi395 was down‐regulated, and zma‐miR479 showed elevated expression levels; zma‐miR479 was also found to be up-regulated in local leaf infections, whereas the down-regulation of zma-miR395 was also observed in infected roots. In summary, although some miRNAs were commonly regulated in both locally infected leaves and roots, the miRNA transcriptome was specific for a given infection stage and in addition also organ-specific (**Table 2**). To confirm the deepsequencing results, Northern blots of a selected miRNA were performed. Due to the relatively high expression level and the remarkable difference between control and treated samples, zmamiR395 was selected (**Figure 6**).


**Compound** Glucoberteroin

Glucobrassicin

Glucoerucin Glucoiberin Glucoiberverin

Glucolesquerellin

Gluconasturtiin

Glucoraphanin

7-Methylthioheptyl

462.0958

glucosinolate

kaempferol 3-Orhamnoside‐7‐O

578.1552

431.0942, 285.0399,

0.6

1.0

1.4

0.8

1.7

283.0236

rhamnoside

kaempferol

593.1534

447.0905, 285.0410,

0.6

0.9

0.8

0.7

1.4

283.0240

223.0586, 164.0484,

0.7

0.9

1.2

0.8

1.4

149.0245

3‐rhamnoside‐7‐Glu

Sinapoyl malate

1-O-β-Dglucopyranosyl

385.1147

265.0794, 190.0267,

0.8

1.0

0.9

0.5

1.2

175.0030

sinapate

Camalexin

(4 dpi).

**Table 1.** Fold induction of metabolites in sRNA mutants and controls upon

*C. higginsianum* infection.

199.0332

10.7

84.4 Fold induction of identified compounds from the metabolome of the sRNA mutants *hen1‐1, hyl1‐2*, *rdr6‐15* and the wild-type Ler and Col-0 upon *C. higginsianum* infection

69.0

71.5

72.9

 339.0745

 436.0406

372.0467, 178.0225

95.9527, 74.9920

 1.0

 0.7

1.1 1.6

1.2

0.9

1.9

0.7

0.8

1.8

 422.0578

448.0764

96.9590,

1.1

1.6

1.4 0.8

0.8

2.1

1.1

2.0

95.9513,74.9919

406.0301

96.9619, 95.9494,

1.1

2.1

1.0

0.7

1.4

74.9920

422.0219

96.9619,95.9519,

0.8

1.4

1.7

0.7

1.6

74.9923

420.0457

96.9628, 95.9551,

1.0

1.6

0.2

0.8

1.7

74.9943

447.0512

434.0612

**Mass**

**Fragments (M‐H)‐**

96.9603, 95.9523

96.9601, 95.9523,

0.8

1.7

2.0

0.6

2.4

74.9914

 ‐

‐

0.4

0.9

*hen1‐1* **FI**

*Ler FI*

*hyl1‐2* **FI**

*rdr6‐15* **FI**

 **Col‐0 FI**

1.9

20 Plant Engineering


FI = fold induction compared to control libraries. Inf = infected, L = leaf, R = root, sys = systemic, zma = maize miRNAs.

**Table 2.** Maize miRNAs differently regulated upon *C. graminicola* infection.

**Figure 6.** Northern blot analysis of miR395 expression. The signs + indicates *C. graminicola* infection, - control tissue. H= herbivore (*Spodoptera frugiperda*, non-fungus control). The tRNA and 5S rRNA are shown as a control for equal loading and were stained with ethidium bromide.

As expected, zma‐miR395 showed a reduced expression level upon fungal infections in roots. The signal intensity also corresponded to the sequence reads in the different libraries, with the highest number of reads (93) in control roots. To examine the putative role of zma-miR395 during root infections, the maize genome was analysed for putative target genes. Five known target genes were identified: two genes (dienelactone hydrolase and FMR1-interacting) exhibit two mismatch positions for zma-miR395. The other genes, ATP sulphurylase (APS) on chromosomes 1 and 5, and a sulphate anion transporter, perfectly matched to the zma‐ miR395 sequence. To confirm the genotype of a reduced expression level of zma-miR395 in infected maize roots, the gene expression of two zma‐miR395 putative target genes (*ZmSAT* and *ZmATPS*) was analysed (**Figure 7**).

**Figure 7.** Expression profile of *ZmSAT* and *ZmATPS* genes in *C. graminicola*‐infected roots.
