Introduction

Wheat (Triticum aestivum L.) is the second most important staple food in of South Africa. Due to its economic importance, wheat needs to be protected from pathogen infections that can potentially cause epidemics, leading to food insecurity (Lorrain et al. 2019). Pucciniales species, such as Puccinia triticina (Pt: leaf rust), P. graminis f. sp. tritici (Pgt: stem rust), and P. striiformis f. sp. tritici (Pst: stripe rust) are economically important pathogens of grain crops, with Pt infecting wheat and triticale, Pgt infecting wheat, triticale and barley and Pst infecting wheat (Boshoff et al. 2019; Bolton et al. 2008; Duplessis et al. 2011; Saunders et al. 2019; Zhang et al. 2021). The economic importance of these pathogens is attributed to factors such as their ability to cause yield loss, and ease of spread via airborne spores, which both contribute to increased risks for epidemic outbreaks. Their ability to mutate leads to the bridging of resistance gene(s) deployed in previously resistant host plants (Kloppers and Pretorius 1997; Kolmer et al. 2019; Visser et al. 2019; Zhang et al. 2021). Breeding programmes develop resistant wheat cultivars to mitigate the effects of the rust-causing pathogens, but the emergence of races with wider virulence mandates the implementation of other control strategies. Thus, we argue that studying carbohydrate metabolism during the interaction between the rust fungus and wheat could shed light on new rust control strategies.

Rust fungi of wheat have very complex life cycles because they need two unrelated host plant species to complete their life cycle (Bolton et al. 2008; Prasad et al. 2020). As obligate biotrophic pathogens, they exclusively feed on living host tissue (Duplessis et al. 2011; Garnica et al. 2014; Prasad et al. 2020). During infection and penetration, the rust fungi produce different kinds of Carbohydrate-Active Enzymes (CAZymes) (Duplessis et al. 2011). These CAZymes are involved in a range of metabolic processes, including cell wall degradation, biosynthesis or modification of polysaccharides, oligosaccharides, and glycoconjugates (Duplessis et al. 2011; Garnica et al. 2013; Chebli and Geitmann 2017; Lorrain et al. 2019). There are four functional classes of CAZymes identified in the Puccinia spp. secretome, which include glycoside hydrolases (GHs), glycosyltransferases (GTs), polysaccharide lyases (PLs), and carbohydrate esterases (CEs). These groups are classified based on the similarity of amino acid sequences and 3D conformational structure (http://www.cazy.org). Three classes (GHs, CEs, and PLs) are secreted by fungi during the infection process and are often referred to as cell wall degrading enzymes (CWDEs) (Garnica et al. 2013; Cosgrove 2016; Silva-Sanzana et al. 2020).

Plants have evolved mechanisms to cope with pathogen attacks (Huai et al. 2020); including modifying the plant cell wall, which serves as the first barrier against pathogen invasion. The plant cell wall is composed of a polymeric network of cellulose micro-fibrils cross-linked by hemicellulose and lignin and embedded in pectin (Cosgrove 2000; Mnich et al. 2020; Silva-Sanzana et al. 2020). Cellulose micro-fibrils consist of β-D-1,4-glucan chains, which are the largest component of the cell wall (Mafa et al. 2021). The intra- and intermolecular hydrogen bonds formed between the parallel or overlap** layers of cellulose fibres lead to microcrystalline cellulose (MCF) formation. In the cell wall, the MCFs are arranged to create a strong yet extensible network layer that protects the plant cell (Chebli and Geitmann 2017; Cosgrove 2022). This matrix polymer (i.e., pectin and hemicellulose) contributes to basal defence responses by forming papillae or callose to limit pathogen penetration or perforation (Vorwerk et al. 2004). In addition, hemicellulose is one of the complex cell wall polymers that is composed of β-D-1,4-glycosidic bonds linking glucose (for xyloglucan), mannose (for mannan), or xylose (for xylan) forming the main chain (Mendis and Simsek 2015; Mnich et al. 2020). Xyloglucan is the major hemicellulose in dicotyledonous plants, while for most grass species arabinoxylans predominate (Mnich et al. 2020). Moreover, cross-linking of xylan to lignin by ferulic acid restricts CWDEs such as xylanases and cellulases from degrading arabinoxylan and MCF, respectively (Mnich et al. 2020; Mafa et al. 2021).

Cellulases produced by pathogenic fungi are part of CWDEs classified under the GHs, which catalyse the hydrolysis of the β-1,4-glycosidic bond in cellulose (Cosgrove 2016; Silva-Sanzana et al. 2020; Mafa et al. 2021). Cellulases can be divided into three groups according to their mode of hydrolysis and substrate specificity, i.e., endoglucanases (EGs), cellobiohydrolase (CBHs), and β-glucosidases (Mafa et al. 2021). Generally, β-glucosidases produce glucose from soluble cellooligosaccharides released by the catalytic activity of CBHs and EGs on the cellulose substrates (Kuhad et al. 2016). Several studies have shown that wheat rusts possess CWDEs and many genes coding for cellulases and other enzymes capable of breaking down polysaccharides to produce hexose sugars such as glucose, fructose, galactose or mannan (Heiler et al. 1993; Xu and Mendgen 1997; Garnica et al. 2013; Heller 2020). Production of hexose sugars is essential for the survival of rust fungi as they play a role during carbon metabolism and energy generation (Garnica et al. 2014; Morkunas and Ratajczak 2014). They enter the haustoria through hexose transporter proteins (HXT) (Voegele et al. 2001). As part of their invasion mechanism, biotrophic fungi also produce invertases (Chang et al. 2017; Tauzin and Giardina 2014). The invertase enzyme catalyses sucrose into glucose and fructose, which are transported into Pt haustoria via the HXT. The plant generally produces higher quantities of sucrose in the leaves during photosynthesis and transports it to the sink tissues. Haustoria create a second sink competing with the plant sink tissues for sucrose in the apoplastic region (Heiler et al. 1993; Voegele et al. 2001; Morkunas and Ratajczak 2014). Thus, understanding the role of cellulases and invertase could shed light on develo** new ways/tools to mitigate wheat infection by Pt.

Our previous study demonstrated that at the neutral pH range of 6–7, the CWDE enzyme activity levels were significantly higher in a Pt-inoculated susceptible cultivar (IS) compared to a resistant cultivar (IR) (Mafa et al. 2023). Additionally, reduced cell wall integrity in the IS samples could be explained by increased CWDEs activities at a neutral pH range. In contrast, Cosgrove argues that plant cell wall synthesis or re-enforcement occurs under acidic conditions, with the pH ranging from 4.5 to 5.5 (Cosgrove 2000, 2022). Hence, in this follow-up study we investigated whether acid pH conditions (pH 5) affected the CWDEs and invertase enzyme in Pt-inoculated IR Thatcher + Lr9 wheat plants compared to those found in the susceptible Thatcher control.

Materials and methods

Materials

Wheat seed stocks for the susceptible cultivar Thatcher and resistant NIL (Thatcher + Lr9) and urediniospores of Pt isolate UVPt9 (South African race annotation 3SA133; North American race annotation PDRS) were obtained from the wheat germplasm, and rust isolate collections kept by the Plant Pathology Division in the Department of Plant Sciences, University of the Free State. The Avicel, p-nitrophenol (pNP)-acetate, pNP-butyrate, 3,5-Dinitrosalicylic (DNS) acid, xylanase and cellulase were purchased from Sigma-Aldrich (Missouri State, USA). Wheat arabinoxylan (WAX), pNP-α-D-maltoside and glucose-oxidase/peroxidase-kit (GOPOD-kit) were purchased from Megazyme (Megazyme, Wicklow, Ireland). Invertase was purchased from BDH Biochemicals (England), while sucrose was purchased from Merck (Darmstadt, Germany). Other analytical chemicals were purchased from Sigma-Aldrich, except when stated otherwise.

Growth conditions and inoculation of wheat seedlings

Wheat cultivar Karee was used for the multiplication of UVPt9 urediniospores, according to the method of Castelyn et al. (2015). Seedling germination, growth and inoculation with UVPt9 urediniospores were done as described (Mafa et al. 2020, 2023). Mock-inoculated wheat seedlings were sprayed with water. Leaves of both mock-inoculated (control susceptible (CS) and control resistant (CR)) and Pt-inoculated (susceptible (IS) and resistant (IR)) seedlings were sampled at 1, 7, and 14 dpi in liquid nitrogen. The sampled leaves were ground to a fine powder in liquid nitrogen and stored at -80℃.

Protein extraction and concentration determination

About 300 mg fresh leaf tissue of mock (CS and CR) and Pt-inoculated (IS and IR) samples were extracted on ice with 5 mL 50 mM sodium phosphate buffer pH 7 containing 1% (w/v) Polyvinylpolypyrrolidone (PVPP). The mixture was vortexed vigorously and then centrifuged at 10 000 g for 10 min. The supernatant was used as a total protein extract containing enzymes of interest. The Bradford method was used to determine the total protein concentration (Bradford 1976). Briefly, 190 µL Bradford reagent (Bio-Rad) was mixed with 10 µL of each extract, followed by incubation at 25℃ for about 20 min and absorbance measured at 590 nm with a microplate reader (Anthos Zenyth 3100). Bovine serum albumin (BSA) was used as a suitable standard.

Enzyme activity determination

Holocellulose modifying enzyme activity

The activities of CWDEs, in particular xylanase and cellulase under acidic conditions, were investigated using the total protein extract. For the activity assay for cellulase, 300 µL of 1% (w/v) Avicel dissolved in 50 mM sodium acetate buffer (pH 5.5) was mixed with 100 µL of the extract. A similar approach was used for measuring xylanase activity, except that 1% (w/v) WAX was used as the substrate. The reaction time for both enzymes was 5 h at 37℃. The total reducing sugars released by the enzymes from the Avicel and WAX substrates were determined with an ultraviolet-visible spectrophotometer (Varian Cary 100 Bio UV-Visible) set at 540 nm according to the modified DNS method (Miller 1959). Glucose and xylose were suitable standards for these assays.

Disaccharide cleaving enzyme activity

Enzymatic assays for β-glucosidase, amyloglucosidase, and invertase activity were performed under acidic conditions. The invertase reaction mixture consisted of 250 µL 1% (w/v) of sucrose dissolved in 50 mM sodium acetate buffer (pH 5) and 100 µL protein extract. The positive control consisted of similar components, except 50 µL (38 U/mg protein activity levels) commercial invertase was used. All reactions were incubated at 37℃ for 1 h. Similar enzyme activity assays for β-glucosidase and amyloglucosidase were employed. However, 1% (w/v) cellobiose was included as substrate in the case of β-glucosidase, and 1% (w/v) pNP-α-D-maltoside or maltose for amyloglucosidase. The blanks were created by excluding the enzymes from the reactions. After terminating the invertase, amyloglucosidase and β-glucosidase reactions, the GOPOD reagent was used to determine the glucose concentration following the manufacturer’s instructions. Briefly, 100 µL of each reaction mixture was added into 1000 µL GOPOD reagent, the reaction mixture was incubated at 25℃ for 20 min, and absorbance of samples measured at 510 nm using a spectrophotometer (Varian Cary 100 Bio UV-Visible).

Peroxidase and carbohydrate esterase activity

Peroxidase activity in all samples was also determined under acidic conditions. The reaction was initiated by mixing substrate solution (0.5% (v/v) H2O2, 5.0 mM guaiacol in 50 mM sodium acetate buffer (pH 5.5)) with 100 µL protein extract. The reaction was measured at a wavelength of 470 nm with a spectrophotometer (Varian Cary 100 Bio UV-Visible). The extinction coefficient (26.6 mM-1 cm-1) of guaiacol was used to determine the peroxidase activity using ‘Beer’s law.

For CEs, activity assays were performed by mixing 5.0 mM pNP-Acetate, and pNP-butyrate dissolved in 50 mM sodium acetate buffer (pH 5.5) with 50 µL enzyme extract. The reactions were incubated for 15 or 30 min at 37℃. Absorbance was measured at 410 nm using spectrophotometry (Varian Cary 100 Bio UV-Visible), and the enzyme activity determined using a pNP extinction coefficient = 17 500 M-1 cm-1.

LC-MS analysis of the soluble sugars in Pt-inoculated samples

The resistant (IR) and susceptible (IS) wheat inoculated with Pt were selected for liquid chromatography–mass spectrometry (LC-MS) analysis after 14 days. The soluble sugars were extracted according to methods described by Zhao et al. (2010) and modified by Mafa et al. (2023). Samples were analysed using an ABSCIEX 4000 QTRAP hybrid triple quadrupole ion trap mass spectrometer with a Shimadzu high-performance liquid chromatography (HPLC) stack as a front end. All data acquisition and processing were performed using Analyst 1.5 (AB SCIEX) software. Twenty microliters of each sample were separated on a C18 (Carbohydrate 4.6 × 250 mm, Agilent) column at a flow rate of 500 µL/min using a water (solvent A) and acetonitrile (solvent B) gradient from 100% B to 60% B over 10 min followed by column re-equilibration steps with a total run time of 20 min to allow for column re-equilibration. Eluting analytes were ionised in negative electrospray mode in the TurboV ion source with a 400℃ heater temperature to evaporate the excess solvent, 30 psi nebuliser gas, 30 psi heater gas and 20 psi curtain gas. Declustering potential was set at 350 V. The eluting analytes were mass analysed on the mass spectrometer in a Q1 scan mode ranging from 150 Da to 1700 Da with a dwell time of 3 s.

Data analysis

Experiments were designed in the split-plot randomized fashion with two biological repeats. The values presented in the graphs represented the means ± standard deviation generated from three experimental data points except if stated otherwise. The statistical difference between inoculated Thatcher and Thatcher + Lr9 relative to mock-inoculated controls was established with ANOVA and Tukey HSD homogeneity set at p < 0.05 using Statistica software (version 13).

Results

Phenotypic analysis of Pt-inoculated IR and IS

The phenotype of Pt-inoculated leaves of the susceptible Thatcher and resistant Thatcher + Lr9 were significantly different, because after 14 dpi, the IS samples showed pustules indicative of advanced infection (Fig. 1). The IR phenotype was not much different from the mock-inoculated samples, except for a few small chlorotic flecks, which represent the classical hypersensitive response. In contrast, the IS samples showed the typical rust mosaic on the leaves where the cell wall/ epidermis was probably hydrolysed by CWDEs as the epidermis rolled up to expose the urediniospores. Pustules were surrounded by light-yellow leaf chlorosis, which could theoretically represent fungal biomass beneath the leaf epidermis or what could also be a combination of fungal growth and the disruption of plant cellular functions.

Fig. 1
figure 1

The phenotypes of infected resistant Thatcher + Lr9 (IR) and susceptible Thatcher (IS), as well as controls (CS or CR) 14 dpi with Puccinia triticina isolate UVPt9. The white arrows indicate the epidermal peeling effect of cell wall degrading enzymes (CWDEs) on the IS samples, and the white ovals potential fungal biomass accumulation under the leaf epidermis. The yellow arrows indicate a few chlorotic flecks on the IR leaf samples

Full size image

Holocellulolytic enzyme activity levels in Pt-inoculated wheat samples

The holocellulolytic enzyme activity under acidic conditions (pH 5) was determined by measuring the exoglucanase and xylanase activities of total protein extracts from the CS, CR, IS and IR samples (Fig. 2). The exoglucanase activity was not significantly different at 1 dpi in all samples (Fig. 2a). However, the IR exoglucanase activity levels were significantly higher at 7 and 14 dpi compared to the CR samples, showing 2-fold and 4-fold increases respectively. The exoglucanase activity levels in the IS samples at 7 dpi were significantly decreased, but significantly increased by about 2-fold at 14 dpi. For xylanase activity levels, both the IR and IS samples showed similar patterns to exoglucanase relative to the CR and CS samples (Fig. 2b). After a significant 2-fold reduction in the IS samples at 7 dpi, activity levels significantly increased at 14 dpi to about 2-fold difference compared to the CS results. In contrast, xylanase activity in the IR samples at 7 dpi were 2-fold higher than in the CR sample, followed by a 5-fold increased activity levels at 14 dpi compared to CR.

Fig. 2
figure 2

The in planta cellulase (exoglucanase) and hemicellulase (xylanase) activity in Puccinia triticina (Pt) inoculated susceptible Thatcher (IS) and resistant Thatcher + Lr9 (IR) relative to mock-inoculated plants (CS or CR) at 1, 7 and 14 days post inoculation (dpi). Avicel (a) and wheat arabinoxylan (b) were respectively used as substrates. The experiments were performed in triplicate, and the values represent means ± SD. Alphabet annotations demonstrate the significant difference between samples within a particular time point of infection (tested by Homogenous Groups Tukey HSD). The * represents significant differences between the control and the Pt-inoculated samples (p < 0.05)

Full size image

Esterase activity levels in Pt-inoculated wheat

Carbohydrate esterases also form part of the supergroup of CWDEs, which aid in catalysing cell wall modification to achieve reinforcement or accelerate degradation. To test the esterase activity levels in Pt-inoculated wheat samples under acidic conditions (pH 5), two substrates with shorter (pNP-acetate (C2)) and longer carbon chains (pNP-butyrate (C4)) were respectively used (Fig. 3) During the hydrolysis of the pNP-acetate (Fig. 3a), IR samples showed slightly increased esterase activity levels compared to the control at 1 dpi. At 7 dpi, the esterase enzyme activity levels were significantly reduced in IR and IS samples relative to CR and CS, respectively, except that IS was about 1.5-fold lower relative to CS. However, IS displayed a 1.6-fold increased esterase enzyme activity level compared to the CS samples at 14 dpi, while the IR samples still significantly reduced this enzyme activity relative to CR. On the pNP-B substrate (Fig. 3b), the IR samples showed higher esterase activity levels at 1 and 7 dpi compared to the controls, followed by a significant reduction at 14 dpi in the IR. The IS esterase displayed a similar pattern when it hydrolysed the pNP-acetate and pNP-butyrate, except that esterase enzyme activity levels on pNP-butyrate were about 1.9-fold in IS compared to CS. In general, the IR ‘samples’ esterase enzyme activity results indicated the enzymes were not effective on short-chain substrates over a prolonged time. However, on the longer carbon chain substrate, the esterase activity was significantly increased during the early to mid-period of infection. On the other hand, the esterase enzyme activity was significantly increased at the later stages of infections in the IS samples.

Fig. 3
figure 3

The in planta cell wall cross-linking enzyme (esterase) activities in Puccinia triticina (Pt) inoculated susceptible Thatcher (IS) and resistant Thatcher + Lr9 (IR) relative to mock-inoculated plants (CS or CR) after 1, 7 and 14 days post inoculation (dpi). The pNP-acetate (a) and pNP-butyrate (b) were used as substrates. The experiments were performed in triplicate, and values represent means ± SD. Alphabet annotations demonstrate the significant difference between samples within a particular time point of infection (tested by Homogenous Groups Tukey HSD). The * represents significant differences between the control and the Pt-inoculated samples (p < 0.05)

Full size image

Peroxidase activity level in Pt-inoculated wheat plants

Lignin peroxidase is a versatile peroxidase that modifies the lignin component of the cell wall. To investigate the peroxidase activity under acidic conditions (pH 5), guaiacol and 8-aminoquinoline were used for the reactions, while the total protein extracts were used as peroxidase sources (Fig. 4). Peroxidase activity levels in the IR samples were not significantly different to the ones detected in the CR samples at 1 and 7 dpi when guaiacol was used as a substrate (Fig. 4a). At 14 dpi, the IR samples showed a 4-fold significant increase in peroxidase activity. In IS, the peroxidase activity levels were lower at 1 and 7 dpi, but were about 2-fold higher at 14 dpi. The IR samples displayed significantly higher peroxidase activity levels at all tested time points when using 8-aminoquinoline as substrate (Fig. 4b). It is important to note that the peroxidase activity levels in the IR samples at 1 and 14 dpi were about 1.5-fold and 2-fold significantly higher than the peroxidase activity displayed by the control. The IS samples also showed significantly increased peroxidase activity levels at 1 and 14 dpi but showed a significant reduction at 7 dpi. The peroxidase activity was thus significantly increased in IR samples for all the tested time points using 8-aminoquinoline as substrate, while the IS displayed a reduction in the early infection period (guaiacol) and at 7 dpi (8-aminoquinoline).

Fig. 4
figure 4

The in planta peroxidase activity in Puccinia triticina (Pt) inoculated susceptible Thatcher (IS) and resistant Thatcher + Lr9 (IR) relative to mock-inoculated plants (CS or CR) at 1, 7 and 14 days post inoculation (dpi). Guaiacol (a) and 8-aminoquinoline (b) were used as substrates. The experiments were performed in triplicate, and the values represent means ± SD. Alphabet annotations demonstrate the significant difference between samples within a particular time point of infection (tested by Homogenous Groups Tukey HSD). The * represents significant differences between the control and the Pt-inoculated samples (p < 0.05)

Full size image

Hexose-producing enzymes/oligosaccharide hydrolysis

β-glucosidase and amyloglucosidase activity

Oligosaccharides produced during cell wall hydrolysis by cellulases or starch hydrolysis by amylase, are further hydrolysed into glucose by β-glucosidase and amyloglucosidase, respectively. Generally, the β-glucosidase cleaves β-1,4-glycosidic bonds linking cell-oligosaccharides, while amyloglucosidase cleaves α-1,4-glycosidic bonds linking maltooligosaccharides. The β-glucosidase activity on cellobiose substrate was significantly increased in IR samples at 1, 7 and 14 dpi compared to the controls (Fig. 5a). At 14 dpi, the IR samples displayed a 3-fold higher activity. The IS and CS had similar β-glucosidase activity levels at 1 dpi, however, at 7 and 14 dpi, IS showed a significant 2-fold activity increase. The IR samples possessed significantly higher amyloglucosidase activity levels at 1, 7, and 14 dpi using both maltose (Fig. 5b), and pNP-maltopyranoside (Fig. 5c) as substrates. The amyloglucosidase activity was higher in IS compared to CS at 1 and 14 dpi, while activity was significantly reduced at 7 dpi using either of maltose or pNP-maltopyranoside as substrates.

Fig. 5
figure 5

The in planta hexose producing enzymes (β-glucosidase and amyloglucosidase) activity level in Puccinia triticina (Pt) inoculated susceptible Thatcher (IS) and resistant Thatcher + Lr9 (IR) relative to mock-inoculated plants (CS or CR) at 1, 7 and 14 days post inoculation (dpi). The cellobiose (a), maltose (b) and pNP-α-D-maltoside (c) were used as substrates. The experiments were performed in triplicate, and the values represent means ± SD. Alphabet annotations demonstrate the significant difference between samples within a particular time point of infection (tested by Homogenous Groups Tukey HSD). The * represents significant differences between the control and the Pt-inoculated samples (p < 0.05)

Full size image

Invertase activity in Pt-inoculated wheat

The enzyme that catalyses sucrose to its monomeric constituents is invertase. Invertase activity was thus measured using sucrose as substrate (Fig. 6). The IR and CR samples showed similar invertase activity profiles at 1 and 7 dpi, followed by a significant reduction in IR at 14 dpi. The IS plants displayed a gradual increase in the invertase activity from 1 to 14 dpi. IS invertase activity levels were increased by 1.36-fold at 7 dpi, while at 14 dpi the enzyme activity was increased by 3-fold in the IS compared to the control.

Fig. 6
figure 6

The in planta hexose producing invertase activity levels in Puccinia triticina (Pt) inoculated susceptible Thatcher (IS) and resistant Thatcher + Lr9 (IR) relative to mock-inoculated plants (CS or CR) at 1, 7 and 14 days post inoculation(dpi). The sucrose was used as substrate. The experiments were performed in triplicate, and the values represent means ± SD. Alphabet annotations demonstrate the significant difference between samples within a particular time point of infection (tested by Homogenous Groups Tukey HSD). The * represents significant differences between the control and the Pt-inoculated samples (p < 0.05)

Full size image

LC-MS profiles of the oligosaccharides

The enzymological studies showed that the CS and CR responses were similar, while the Pt inoculation induced differential phenotypic and biochemical responses in the IS and IR samples at 14 dpi. The 14 dpi IS and IR samples were selected for LC-MS analysis to identify and quantify carbohydrates and their conjugates (Table 1). The identified carbohydrates were divided into five categories, namely carbohydrates that are part of metabolites, those that are potentially part of the wheat or Pt cell wall (cellulose), those that are part of the hemicellulose content (xylan), those that are associated with the Pt cell wall (chitin) and conjugates of carbohydrates potentially sourced from Pt’s sialic acid groups.

Carbohydrates associated with metabolic processes in plant tissue were significantly higher in the IR compared to the IS sample. For instance, the IR samples had an additional 40 000 cps of fucose relative to IS samples, and for glucose, the IR had about 2.6-fold increased glucose content compared to IS samples. For sucrose, IR displayed a 1.307-fold increase compared to the IS sample (Table 1). The significant reduction in these important metabolites suggests there could be a competition for resources between the fungus and plant in the Pt-inoculated susceptible plants.

In the case of oligosaccharides associated with the cell wall, it was clear that cellooligosaccharides were significantly higher in the IR compared to the IS samples (Table 1). Worth noting is that both cellotriose and cellopentaose were not detected in the IS samples. IR samples had about 5-fold increased cellotetraose compared to the IS. Hemicellulose is another component of the plant cell wall, consisting of arabinoxylan in wheat plants. The xylooligosaccharide levels were significantly higher in the IS than in IR samples. The IS samples had 1.7-fold increased quantities of xylobiose, followed by 2.1-fold of xylotetraose compared to the IR samples. In addition, the IS samples had xylotriose that was not detected in the IR samples. The IR sample had longer xylooligosaccharides with a degree of polymerisation of 5 (DP of 5) with an intensity value of 6 000 cps, while this oligosaccharide was absent in the IS samples. In addition, the IR sample had acetylated xylotriose and acetylated xylotetraose, with the former having similar quantities in both IR and IS, while acetylated xylotetraose (8 900 cps) was detected only in the IR samples. In general, the results reveal there was incomplete hydrolysis or degradation of the hemicellulose in the IR sample because it possessed longer and decorated xylan oligosaccharides.

The chitooligosaccharide mass fragments were also identified in both samples 14 dpi. Glucosamine was about 1.4-fold higher in the IR compared to the IS plants (Table 1). The shorter chitooligosaccharides like chitobiose had the same levels in the IR and IS; however, longer chitooligosaccharides such as acetylated tetraacetyl-chitotetraose levels were 3.4-fold higher in the IR compared to IS samples; additionally, acetylated penta-N-acetylchitopentaose was only detected in the IR plants. Since plants lack chitin in their cell wall, the results suggest that the IR sample had a mechanism of degrading the Pt cell wall releasing the chitooligosaccharides.

The accumulation of the four different sialic acid products (part of fungal cell surface and cell wall carbohydrates) in the IR plants confirmed that the Pt cell wall was degraded mostly in the IR compared to the IS samples (Table 1). Keto-deoxy-nonulonic acid and N-glycolylneuraminic acid were detected in the IR samples, with about 8 700 and 5 100 cps intensities, respectively. However, both N-acetyl-neuraminic acid, and 9-O-acetyl sialic acid were detected in both the IR and IS samples. The IS had about 2.75- and 3.18-fold increased content of N-acetyl-neuraminic acid and 9-O-acetylsialic acid compared to IR samples.

Table 1 The carbohydrates and their conjugates identified in Pt-inoculated Thatcher + Lr9 (IR) and Thatcher (IS) 14 days post inoculation (dpi) using their mass profiles obtained from liquid chromatography–mass spectrometry (LC-MS).
Full size table

Discussion

The current study is a continuation of our previous study in which we investigated the potential defensive role of CAZymes and carbohydrates during Pt-wheat interaction (Mafa et al. 2023). However, several studies that dealt with the cell wall reinforcement topic propose that cell wall loosening and relaxation occur under acidic conditions (Durachko and Cosgrove 2009; Cosgrove 2016; Chebli and Geitmann 2017). Cell wall loosening is defined as stress relaxation due to polymer rearrangement and detachment of molecular bonds between polymers due to the application of force, such as turgor pressure or enzymatic action (Chebli and Geitmann 2017). The enzymatic action of the expansins, cellulases, pectins and hemicellulases are some of the enzymes demonstrated to induce plant cell wall loosening (Durachko and Cosgrove 2009; Cosgrove 2016; Chebli and Geitmann 2017). To get a fuller picture of how the CAZymes involved in cell wall loosening are affected during the Pt-wheat interaction, we performed all the enzymatic activity studies under acidic conditions in the current study. This approach gave us a comprehensive perspective on how the CWDEs can loosen the plant cell wall.

Important to note is that both Pt and wheat plants possess CWDEs. In plants, increased activity of CWDEs is involved in cell wall loosening, while pathogens mostly secrete these enzymes to degrade the plant cell wall; thus, one should apply restraint when analysing and interpreting the CWDEs’ functions during the plant-pathogen interaction. Several studies by Cosgrove demonstrated that at a neutral pH, plant cell wall modifying agents, such as expansins and cellulases lose their activity resulting in less or negligible enzyme activity, and by extension the cell wall physicochemical changes do not occur (Cosgrove 2000, 2016; Durachko and Cosgrove 2009).

Urediniospores of the broad bean rust fungus Uromyces viciae-fabae formed infection structures on artificial membranes (Heiler et al. 1999). The authors indicated that this rust fungus secreted cellulase enzymes with gradually increasing activities at different stages of its differentiation stages from the germ tube (4 h post inoculation (hpi)) to the formation of haustoria (18 hpi). Schmidt and Wolf (1999) also observed that the apoplastic fluid of the bean plant inoculated with U. appendiculatus had increased cellulase activity over the infection period (10 days). These studies only confirm that the bean rust fungus has several acidic cellulase isozymes but did not investigate how CWDEs/CAzymes levels are affected during rust fungus interaction with resistant and susceptible plants.

Our study investigated the CWDE activity levels in the Pt-inoculated resistant Thatcher + Lr9 and Thatcher cultivars and the implication on the integrity of the plant cell wall. Results showed exo-glucanase and xylanase activity levels under acidic conditions were significantly increased in the IR compared to CR throughout the infection period. However, the activity levels of both enzymes were significantly lower in the IS compared to control samples at 7 dpi. Phenotypic findings showed that the IR and CR leaf tissues were comparable, except for a few small chlorotic flecks. These flecks are generally associated with HR responses (Seifi et al. 2021). In addition, these observations suggest that the IR significantly increased cellulase and xylanase to loosen and reinforce the cell wall. Cosgrove and other researchers proposed that the cellulases and hemicellulases (such as xylanases) play an important role during cell wall loosening, which allows new wall material to be loaded (Durachko and Cosgrove 2009; Cosgrove 2016; Chebli and Geitmann 2017; Mafa et al. 2023) demonstrated that cell wall physiochemical changes in the IR samples corroborate the cell wall reinforcement proposal. The reduced enzyme activity levels may indicate that the tools necessary for cell wall loosening/modifying in the IS sample were not as efficient as in IR. Hence, the phenotypic findings of IS showed that the fungal spores were exposed on the leaf surface. The use of the CWDEs by Pt to expose the spores to the leaf surface is supported by a similar observation reported by Heller (2020). Also, the cowpea rust fungus uses CWDEs to penetrate the epidermis of the leaf (Xu and Mendgen 1997). Garnica et al. (2013) showed through transcriptomic studies that Pst had abundant CAZymes including CWDEs. Though there is a thin line in how one interprets the increased CAZymes activity during plant-pathogen interaction, the data suggest the functions of CWDEs, such as cellulases and xylanases in the IR sample under acidic conditions, are linked to cell wall loosening and reinforcements (Mafa et al. 2023).

The peroxidase activity levels were significantly increased in the IR compared to CR samples at 7 and 14 dpi. However, the activity level of this enzyme was reduced in IS at 1 and 7 dpi. At 14 dpi, the peroxidase activity levels in the IS increased, but they were still lower compared to those found in IR. Mnich et al. (2020) showed that the number of genes encoding putative lignin peroxidases is higher in grasses (e.g., rice has 138 of these genes). The authors argued that class III or lignin peroxidase is involved in the phenolic compound cross-linking process that produces higher-order molecules such are lignin (Mnich et al. 2020). Johnson and Cunningham (1972) reported an increased peroxidase activity in wheat samples infected with the leaf rust fungus compared to healthy wheat samples. Samples showing rust symptoms generally display lower peroxidase activity levels relative to resistant ones (Seevers et al. 1971; Johnson and Cunningham 1972). The lower activity of the peroxidase in the IS compared to control at 1 and 7 dpi relative to control suggests the Pt-infected wheat plant has lower levels of lignin, as demonstrated in our earlier study (Mafa et al. 2023). Altogether, an early increase in the peroxidase activity is a mode of lignin cross-linking and reinforcing the plant cell wall. Seevers et al. (1971) reached a similar conclusion regarding elevated activities of peroxidase in a rust resistant wheat cultivar. These observations support the phenotypic findings that indicated that the pathogen growth in the IR samples was reduced/stopped, but in the IS, Pt bridged the cell wall and develop to a point where it successfully formed sporulating colonies.

For the β-glucosidase activity level, IR samples had the highest activity over the infection period. The IS only showed higher β-glucosidase activity compared to the control at 7 and 14 dpi. A similar pattern for amyloglucosidase, except that at 7 dpi, the level of this enzyme was reduced compared to the CS samples. In contrast, the invertase activity levels were significantly higher in the IS relative to CS throughout the inoculated period. However, there was no difference in the activity levels of invertase in the IR and CR, except that at 14 dpi, its activity level decreased. The three enzymes act on different oligosaccharides or disaccharides to produce hexose sugars used in plant or pathogen metabolism for growth and energy. Li et al. (2019; 2021) indicated hexose sugars accumulation and enzymes involved in the metabolism of hexose sugars are essential for the growth and development of wheat plants. Huai et al. (2020) used the wheat sugar transporter protein 13 (TaSTP13) to show that transport of hexose sugars from plant to Pst infection structures led to the development of rust symptoms. In particular, wheat samples overexpressing the TaSTP13 gene were affected, but after silencing the gene, a significant reduction in the fungal biomass and symptoms was observed (Huai et al. 2020). These observations support our studies, which showed resistant wheat cultivars infected with Pt had a significantly higher concentration of glucose and sucrose relative to the controls and susceptible cultivars. Using LC-MS the current study also showed that the IR sample had significantly higher quantities of the hexose sugars, such as glucose, fucose and sucrose, compared to the IS sample at 14 dpi. We propose that the higher activity of β-glucosidase and amyloglucosidase in IR samples can be attributed to the accumulation of hexose sugars. However, in the IS sample, invertase activity levels were significantly higher than in IR samples. Other studies showed increase in fungal-specific invertase secretion by Pst during wheat infection or U. fabae during bean infection resulted in the depletion of sucrose and hexose sugars in the host plant (Voegele et al. 2001; Chang et al. 2017). Thus, our results show a link between increased invertase activity level, reduced hexose sugars (mono- and disaccharides) and successful fungal infection (phenotypic results).

Both longer and complex celloligosaccharides and xylooligosaccharides were observed in the IR samples in higher concentrations, while the IS samples had higher concentrations of the xylobiose, xylotriose and xylotetraose. We propose that complex oligosaccharides with a higher degree of polymerisation in the IR sample indicate that the plant cell wall was not completely hydrolysed. Perhaps, these complex oligosaccharides are used as building blocks for cell wall reinforcement processes. Also, the acetylated xylooligosaccharides and the acetylated chitooligosaccharides are known to induce defence responses against the pathogen in plants (Mélida et al. 2020). Rebaque et al. (2021) demonstrated that 33-α-L-arabinofuranosyl-xylotetraose (XA3XX) is one of the complex xylooligosaccharides that induced defence responses in tomato plants, which conferred resistance against Sclerotinia sclerotium. The mixed-linked glucans (MLG) oligosaccharides and the chitohexose triggered immune and disease responses against Hyaloperonospora arabidopsidis in Arabidopsis (Rebaque et al. 2021). Generally, chitin is part of the fungal cell wall. Hence, the increased concentration of chitooligosaccharides in the IR sample suggests that the ‘pathogen’s cell wall was degraded by the IR defence systems. The increased fungal cell surface carbohydrates, i.e., fucose, glucosamine, and sialic acids (Alviano et al. 1999; Vanhooren and Vandamme 1999; Konopka 2012; Ghosh 2020), were abundant in the IR compared to IS samples. Altogether, the oligosaccharides concentration and diversity reveal that the IS samples cell wall could be hydrolysed and absorbed by Pt, whereas the higher concentration and more complex oligosaccharides in the IR could be used as building blocks for reinforcing the cell wall.

Conclusion

Enzymological activity studies showed that the CAZymes, such as exoglucanase and xylanase were significantly increased in IR at the acidic pH range. Altogether, phenotypic observations and increased CAZymes activities under acidic conditions in the IR suggest that cell wall loosening occurred in these samples. This is a process that allows the wall to loosen so that new material can be loaded during plant growth which require cell wall reinforcements (Durachko and Cosgrove 2009; Cosgrove 2016; Chebli and Geitmann 2017). Higher peroxidase activity in the IR samples could be responsible for the phenolic-crosslink processes, supporting the claim of cell wall reinforcements. The reduced invertase activity and accumulation of the hexoses and complex oligosaccharides in the IR samples indicate the presence of sufficient building blocks to reinforce the cell wall; however, the depletion of the hexoses and oligosaccharides can be explained by the increased infection in IS as shown by the observed phenotypes and LC-MS analysis. Therefore, studying the CAZymes under acidic conditions reveals that the IR exoglucanase and xylanase don’t have neutral pH range isozymes, but IS samples had higher activity than the control at neutral pH ranges.