Introduction

Penicillium rubens strain 212 (PO212, ATCC201888), formerly Penicillium oxalicum, (Villarino et al., 2016) is an effective biocontrol agent (BCA) against several pathogens in horticultural crops (De Cal et al., 2008; Larena et al., 2003b; Martinez-Beringola et al., 2013) including Fusarium oxysporum f.sp. lycopersici (FOL) (Sacc.) W. C. Snyder & H. N. Hans, causing the vascular wilt disease in tomato plants (De Cal et al., 1995).The application of PO212 conidia (Pascual et al., 2000) and conidial contact with roots (De Cal et al., 2000) make to PO212 effective as a BCA. PO212 is applied as a conidial suspension by watering seedlings in seedbeds seven days before transplanting (De Cal et al., 1999, 2000; Larena et al., 2003a). The main PO212 mode of action is the induction of resistance in tomato plants (De Cal et al., 1997, 1999, 2000). However, the mechanism by which PO212 induces plant resistance is still unknown. Investigations have been made to determine whether its biocontrol capacity is due to a genetic component. Thus, recent genomic comparative analysis between the PO212 and S27 genomes, the latter strain lacking biocontrol activity (BA), revealed a high conservation of their sequences, and therefore, the absence of a specific genetic basis for BA in PO212 (Requena et al., 2023).

The induction of resistance in plants occurs through a mechanism by which plants activate defense responses against pathogens. This activation occurs through the recognition receptors (PRRs) perceiving the presence of pathogen- and microbe-associated molecular patterns (PAMPs and MAMPs, respectively) (Jones & Dangl, 2006). Pathogen compounds or plant- and microorganism derived molecules acting as elicitors can trigger plant defense responses. In this context, and even though xylanases have been classified as enzymes with an essential role in the degradation of xylan, some studies identified endo-xylanases as active elicitors in plant defense systems (Dean et al., 1989). Enzymatic activity of xylanases is dedicated to xylan degradation. Xylan is the main component of hemicellulose, which is a heterogeneous polysaccharide consisting of a chain of β-1,4-units linked by arabinose, methyl glucuronic acid and acetate groups in its side chains (Lei et al., 2016). For xylan degradation, the presence of several endo-xylanases and β-D-xylosidases in microorganisms is essential (Sunna & Antranikian, 1997). Endo-β-xylanases cleave the main chain of xylan and liberate xylo-oligosaccharides with diverse degree of polymerization and then, β-D-xylosidases degraded to xylose (Tanaka et al., 2014). On the other hand, some studies have focused on the elicitor activity xylanases. An endo-xylanase from Trichoderma viride, referred to as the ethylene-inducing xylanases (EIX), has been used to study elicitor activity of endo-xylanases (Dean et al., 1989). A number of studies presented insights into the possible mechanisms for this endo-xylanases-inducing elicitor activity (Enkerli et al., 1999; Furman-Matarasso et al., 1999; Noda et al., 2010). Elicitor activity of Trichoderma reesei and Botrytis cinerea endo-xylanases remains active even when the enzymatic activity was inactivated (Frías et al., 2019). Frias and collaborators, showed that a short 25-residue peptide (named Xyn25) from the B. cinerea xylanase BcXyn11A was able to produce the elicitation as well as the whole BcXyn11A protein and, in addition, they found that two regions were essential to this process (Frías et al., 2019).

Xylanases have been extensively studied in Trichoderma and Aspergillus spp. In Trichoderma strain SY, a well-known xylanase producer isolated from soil, it was found that this strain produced 27-fold of xylanases, and mRNA was highly expressed when this fungus was grown on cellulose or xylan as a sole carbon source (Min et al., 2002). In Aspergillus spp., endo-xylanases were found in A. tubingensis and A. niger, both saprophytic filamentous fungi. A. tubingensis genome encodes three endo-xylanases (XlnA, XlnB and XlnC) whereas A. niger present only two endo-1,4-β-xylanases (XlnB and XlnC) being the xlnA homologue of A. tubingensis absent in A. niger (de Graaff et al., 1994). A. nidulans secretes at least three xylanases of molecular masses 22 (XlnA), 24 (XlnB) and 34 kDa (XlnC) growing on xylan or xylose as carbon sources (Fernández-Espinar et al., 1992, 1993, 1994, 1996; Piñaga et al., 1994). For the complete hydrolysis of xylan in A. niger, xylosidase activity, encoded by xlnD, is also essential (van Peij et al., 1997). According to Tamayo et al. (2008), transcription of the xlnA, xlnB and xlnD genes are regulated by CreA and XlnR involved in a transcription factor cascade regulated by carbon catabolite repression. In Aspergilli, xlnR gene encodes a Zn2Cys6 transcription factor necessary for the synthesis of the main xylanolytic enzymes (Tamayo et al., 2008; van Peij et al., 1998a, b).

In Penicillium spp. multiple studies have focused on the study of xylanolytic enzyme production depending on the carbon source tested (Chávez et al., 2006). This work distributed the xylanolytic enzymes into three families of Glycosyl Hydrolases (GH): GH7, GH10 and GH11 and one xylosidase. Only one endo-xylanase gene from P. chrysogenum (xylP) belonging to family 10, has been studied in greater detail (Haas et al., 1993). The work of Yang et al. (2018) revealed that strain P33 of P. chrysogenum secreted three xylanases, including two GH10 and one GH11 family xylanases, in presence of wheat bran plus microcrystalline.

In this work, we focus on the identification of the genes required for xylan catabolism in PO212 and S27 strains belonging to P. rubens and its possible role in BA. We verified the integrity of the xylanolytic system in the biocontrol strain PO212 and the non-biocontrol strain S27. The expected positive regulation of xylanases encoding genes by xylan or tomato root extracts, and the strong dependence in the activity of transcriptional regulator XlnR. By deleting xlnR in PO212, we explored the effects of downregulating xylanase expression in the BA of PO212.

Materials and methods

Fungal strains and growth conditions

Penicillium rubens strains used in this work are listed in Table 1. PO212 (ATCC 201888), PO212_18.2, and S27 are from strain collection at INIA-CSIC. Dr Mayo (Instituto de Productos Lácteos de Asturias-CSIC, Spain) provided the strain IPLA 33001. This strain was isolated from cheese and identified as P. chrysogenum (Flórez et al., 2007). Table 1 also contains the strains generated by transformation using PO212_18.2 as recipient strain. Conidia from these strains were long-term stored in 20% glycerol at -20 ºC except for PO212, which were stored at 4 ºC as dried conidia. For the maintenance of strains potato dextrose agar (PDA; Difco, Detroit, MI, USA) was used and cultures were stored at 4 ºC until further uses. Conidia were obtained growing mycelium on PDA in Petri dishes and incubated at 22–25 ºC for 7 days. Dried conidia of PO212 were produced in a solid-state fermentation system and dried as previously described by (Larena et al., 2003a). Minimal Medium (MM) (Espeso et al., 2019) was used for growth tests, using D-glucose 1% (w/v) and 5 mM ammonium tartrate as carbon and nitrogen sources, respectively (Villarino et al., 2018). When required D-glucose was substituted by 1% xylan as indicated in the text. Uracil and uridine (uu) were added to solid or liquid media to allow the growth of strains carrying pyrG1 mutation.

Table 1 Penicillium strains used in this work
Full size table

The pathogenic isolate 1A of F. oxysporum f. sp. lycopersici (Sacc) Snyder and Hansen (FOL) was used to test BA of P. rubens strains. FOL strain was provided by Dr Cristina Moyano from the Laboratory for Assessment of Variety, Seed and Nursery Plants, INIA-CSIC (Madrid, Spain). FOL was stored at 4 ºC in tubes containing sterile sand. For mycelial production, conidia from FOL were germinated on Czapek Dox agar (CDA) (Difco Laboratories, Detroit, MI, USA) and cultivated in darkness at 25 ºC for seven days. Microconidial inoculum of FOL was produced in 250-ml flasks containing 150 ml of sterile Czapek Dox broth (Difco). Flasks were inoculated with three mycelial plugs (1 cm diameter) from the 7-day-old cultures on CDA and incubated for 5 days at 25 °C in a rotary shaker (model 3527; Lab-Line Instruments, Inc.) at 150 rpm. Microconidia were separated from mycelial mass by filtration through glass wool. The final conidial concentration in biocontrol assays was adjusted to 5 × 106 microconidia ml−1 (De Cal et al., 1995).

Database search and alignment

Target genes included three β-1,4-endo-xylanases (xlnA, xlnE and xylP), one β-xylosidase (xlnD) and two transcriptional regulators (xlnR and araR) (see Table 2). These genes were identified using homologues from Aspergillus nidulans (FungiDB) and P. rubens Wisconsin 54–1255 (abbreviated as PrWis) (NCBI) in local BlastP searches using PO212 and S27 proteomes (Requena et al., 2023). Nucleotide sequences were aligned with the Multiple Sequence Alignment tool of Clustal Omega (EMBL-EBI).

Table 2 Genes encoding β-1,4-endo-xylanases, β-xylosidase and regulators, identified in PO212 and S27 genomes
Full size table

Expression assays

In order to study the expression of target genes, independent experiments in presence of lyophilized tomato roots or xylan as main carbon source were carried out with the strains under study. Using an inoculum of 5 × 106 conidia ml−1 (final concentration in each flasks), mycelium from each strain was growth in 2-L flasks containing 400 ml of MM with 1% glucose as carbon source. Mycelia were collected by filtration using Miracloth after incubation in a rotary shaker (GFL, Orbital Shaker 3020) at 150 rpm and 25 ºC for 24 h in darkness. Filtered mycelia were washed with sterile distilled water (SDW) and equally divided into four aliquots. Sample indicated as T0 was immediately frozen in liquid nitrogen and used as control. The other three samples designated as T30, T60 and T90 were used to inoculate fresh medium in 250 ml flasks containing 1% of lyophilized powdered roots (tomato roots were lyophilized in a Cryodos at -80 °C during 24 h (Telstar, S.A. Spain)) or 1% of xylan (Xylan from beechwood, Sigma-Aldrich). Mycelium from these samples in flasks was collected at 30 (T30), 60 (T60) and 90 (T90) minutes of incubation at 25 ºC, and was frozen using liquid nitrogen and stored at -80 ºC until processing.

RNA extraction and RT-qPCR

For total RNA extraction, mycelial samples, previously frozen in liquid nitrogen, were pulverized using a pestle and then, we followed the protocol described in Picazo et al. (2020). The quality and quantity of the RNA was checked using the NanoDrop 2000 (Thermo Scientific™). The integrity of RNA was visualized by 1.2% agarose gel electrophoresis. Total RNA samples were flash-frozen in liquid nitrogen and stored at 80 °C until use. cDNA was synthetized using the SuperScript™ First-Strand Synthesis System for RT-PCR kit (Invitrogen, Thermo Fisher Scientific Inc.), following the manufacturer's instructions. Samples were pre-treated with Deoxyribonuclease I, Amplification Grade (Invitrogen), according to the manufacturer's instructions. First-Strand was synthesized using oligo-dT primer. Dilutions of the cDNA were prepared with nuclease-free water (Promega Biotech Ibérica, S.L.) to a final concentration suitable for RT-qPCR. The samples were stored at -20 °C. RT-qPCRs were performed on the 7500 Fast real-time PCR System thermocycler (Applied Biosystems) with Master Mix SYBR Green (Promega). For each RT-qPCR specific primers were designed using Vector NTI™ Suite 8 (Supplementary Table S1). Three technical replicates were done per sample. RT-qPCR conditions were as follows: a 5 min pre-incubation cycle at 95 °C and 40 cycles of 10 s at 95 °C and 30 s at 65 °C. At the end of the program, a cycle consisted of lowering the temperature from 95 °C to 65 °C at a rate of 4.4 °C every 10 s was added. The expression levels were normalized using as housekeeping the tubulin (benA), from PO212 and was calculated following the 2−∆CT or 2−∆∆CT model specified for each case.

Construction of transformation cassettes

For the cassette construction, genomic fragments were obtained from PO212 strain and the selection marker pyrG of Aspergillus fumigatus (pyrGAf) was obtained from plasmid pFNO3 (Villarino et al., 2018). Transformation cassettes were obtained using the fusion PCR protocol (Markina-Inarrairaegui et al., 2011; Nayak et al., 2006). The oligonucleotides used in generating fragments and transformation cassettes are listed in Supplementary Table S1. The proof-reading enzyme PrimeStar (Takara) was used in PCR procedures following manufacturer's instructions. PCR conditions were, an initial denaturation cycle 98 ºC/2 min, 30 denaturation cycles 98 ˚C/10 s, oligonucleotide binding 55 ºC/5 s and extension 72 ºC/1 min/kb. Samples remained at 4 ºC. Fragments were purified using the PCR clean-Up kit (Macherey–Nagel), and following manufacturer’s instructions.

Generation of mutant strains

To improve the yield of homologous recombination in transformation of PO212 derived strains we generated a strain lacking the ku70 gene. The transformation procedure is described in Villarino et al. (2018). For this purpose, PO212_18.2 strain (Espeso et al., 2019; Villarino et al., 2018) which is an auxotrophic strain for pyrG (pyrG1) was used as a recipient strain (Step 1). This transformation was done by locus replacement, using the transformation cassette 1 (TC 1) consisting of the 5' UTR region of the ku70 gene (1.5 kb), the CDS region of the selected marker pyrGAf gene (1.9 kb) and the 3' UTR region of ku70 gene (1.5 kb). The whole cassette had 5 kb. The strain obtained was genotyped to confirm the insertion of the cassette. Once confirmed as ∆ku70, this strain, named 1.1∆ku70, was selected as recipient strain for the next transformation.

The next transformation (Step 2) was done with a transformation cassette 2 (TC 2) consisting of the 5’ UTR fused to 3’ UTR region of ku70 gene (3.1 kb). For the transformation, it was used 5-fluororotic acid (5-FOA) in the medium, necessary to promote the recombination and expulsion of the marker (Villarino et al., 2016). An agarose gel electrophoresis confirmed that the amplified fragment matched to the expected insert size (3.1 kb), in contrast to the recipient strain (1.1∆ku70), which had an amplicon size of 4.9 kb. The obtained strain named T2.7 was a strain ku70 null and auxotrophic for pyrG.

Finally, from T2.7 strain used as recipient strain, we obtained two xlnR null strains named T1.2 and T2.9. It was achieved by the insertion of a transformation cassette (4.9 kb), consisting of the 5’ UTR region of xlnR gene (1.5 kb), the marker pyrGAf (1.9 kb) and the 3’ UTR region of xlnR gene (1.5 kb). Genomic integration of this cassette complements the pyrimidine auxotrophy of recipient strain T2.7, allowing positive selection of transformants with this genetic marker.

These two strains were obtained using alternative strategies for gene deletion. The transformant ∆xlnR T1.2 was obtained using a complete transformation cassette of 4.9 kb (TC 3) integrated in the recipient strain (T2.7) by homologous recombination, favored by the absence of the ku70 gene. The transformant ∆xlnR T2.9 strain was obtained through a co-transformation with two overlapping fragments. One fragment of 2.8 kb (TC 4) consisted in 5’ UTR region of xlnR gene fused to a part of the pyrGAf gene and the other fragment of 2.4 kb (TC 5) consisted in the other part of the pyrGAf gene fused to 3’ UTR xlnR region. Both fragments contained an overlapping pyrGAf gene region. Once the strains were obtained, PCR genotyping and digestion enzymes validated them.

Biocontrol efficacy assays

Assays were performed on tomato plants in growth chambers in order to study the effect of the xlnR gene on PO212 BA against FOL such as described in (Larena & Melgarejo, 2009; Villarino et al., 2018). The previously described transformants and PO212 were used for these assays. Tomato seeds from cv “San Pedro”, which is susceptible to races 1 and 2 of FOL, were used in the efficacy experiments. Tomato seeds were sown in sterile trays (27 × 42 × 7 cm) with autoclaved (1 h at pressure of 1 kg cm−2 and temperature of 121 ºC, during three consecutive days) vermiculite (Termita, Asfaltex, S.A., Barcelona, Spain), and watered with sterile Hoagland solution (Hoagland & Arnon, 1950) and tap water. The trays were maintained in a growth chamber at 24–26 ºC with fluorescent light (100 μE m−2 s−1, 16 h photoperiod) and 80–100% relative humidity during 2 weeks. Tomato seedlings (with at least two true leaves) were treated with an aqueous conidial suspension of each PO212 strain (6 × 106 conidia g−1 vermiculite) 7 days before be transplanted. Conidial suspensions of PO212 and derivates for treatments were prepared as follows. Dried conidia of PO212 were rehydrated in SDW for 2 h using a rotatory shaker at 150 rpm (CERTOMAT® RM). Conidia of the remaining strains were collected from colonies grown on PDA for 7 days in the dark at 25 °C. The day before treatment, conidial viability of all strains was checked by counting germinated conidia according to previously described (Larena et al., 2003a). Tomato seedlings were transplanted into 100-ml flasks containing 100 ml sterile Hoagland solution (Hoagland & Arnon, 1950) with roots in contact with the solution as described by De Cal et al. (1997). At this time, flasks were inoculated with an aqueous (SDW) suspension of conidia of FOL at 5 × 106 conidia ml−1. Plants inoculated with FOL and non-treated with any strain of PO212, were used as the control. Each treatment consisted of five flasks with four plants per flask. Flasks were placed in growth chamber during 4 weeks under the conditions described earlier in this subsection. Disease severity was graded on days 7, 14, 21 and 28 days after transplanting. Disease severity followed a 1 to 5 index scale: 1, healthy plants (0 to 24%); 2, yellow lower leaves (25 to 49%); 3, dead lower leaves and some yellow upper leaves (50 to 74%); 4, dead lower leaves and wilted upper leaves (75 to 99%); and 5, dead plants (100%) (De Cal et al., 1995). At 28 days after transplant, roots were placed in humidity chambers at 25 ºC during 5 days, in order to determine the presence of the pathogen (incidence) in the crowns of the roots. The complete assay was repeated at least two times. If the replicate confirmed the previous results assay, only results from one assay was reported and discussed.

Data analysis

Data were analyzed by analysis of variance (ANOVA) with the Statgraphics program (XVII Centurion. v. 17.2.00). When variances were not homogeneous, data were transformed to log10 to improve the homogeneity before analysis. The Duncan Test was used for comparisons of media, when the F test was significant at p ≤ 0.05 (Duncan, 1955). When homogeneity of variances was not obtained, data analysis was performed using the Kruskal–Wallis nonparametric test, with a confidence level of 5% (Kruskal & Wallis, 1952).

Results

Identification of genes encoding β-1,4-endo-xylanases in PO212 and S27 genomes

To study the role of xylanases in PO212 BA we searched for genes encoding β-1,4-endo-xylanases. Using amino acid sequences of A. nidulans and PrWis xylanases coding genes (xlnA, xlnE and xylP); we performed local BlastP searches in PO212 and S27 proteomes. Table 2 summarizes these searches. In these proteomes, we identified putative homologues for all PrWis xylanases but not for XlnB (AN9365) from A. nidulans. In addition, we searched for additional xylanolytic machinery finding a β-xylosidase coding gene, xlnD, and those genes coding for putative transcriptional regulators XlnR and AraR (a putative arabinolytic regulator). Supplementary Fig. S1 indicates the domains for each protein in PrWis. Hereafter, we designate genes in PO212 and S27 as in Aspergillus putative homologues except for xylP (Table 2).

Xylanolytic system in PO212, S27 and IPLA 33001

To understand the regulation of genes encoding β-1,4-endo-xylanases identified in PO212 and S27 genomes, during colonization of tomato roots, we analyzed the effect in their transcription of adding tomato lyophilized roots to culture media. RT-qPCR results showed that PO212 and S27 present a similar expression pattern of xlnE, xylP, xlnR and araR genes. Because PO212 and S27 strains were isolated from soil, we included in this test the IPLA 33001 strain isolated from cheese as a control. IPLA 33001 presented a basal expression for these genes in root-extract medium. The xlnA gene conserved a similar pattern in the three strains, and PO212 and IPLA 33001 strains showed a similar expression for xlnD gene (Fig. 1d). Notably expression levels for most of the xylanolytic genes were higher in PO212 (Fig. 1).

Fig. 1
figure 1

Relative expression (2−∆∆Ct) of endo-xylanases genes xlnA (a), xlnE (b) and xylP (c), β-xylosidase xlnD (d) and regulators xlnR (e) and araR (f) in the expression assays with PO212, S27 and IPLA 33001 strains in MM with lyophilized tomato roots as carbon source. Mycelial samples were collected at 30, 60 and 90 min. Time 0 indicates mycelium grown in glucose medium for 24 h prior to transfer to medium containing lyophilized tomato roots. Means with the same letter for each strain and for each gene are not significantly different from each other (p ≥ 0.05) according to the Duncan test. Vertical bars represent the standard error of the mean of three biological replicates with three technical replicates

Full size image

Effect of xylan in the regulation of xylanolytic system of PO212 and S27

To determine the effect of xylan in the regulation of xylanolytic system of PO212 BCA and non-biocontrol strain S27, we proceed to culture mycelia from these strains in medium containing xylan as main carbon source. PO212 and S27 are closely related strains (Requena et al., 2023) and, as expected from that study, no differences in the nucleotide sequences of endo-xylanases were found (Supplementary Fig. S2). Presence of xylan in the media caused a variation in transcriptional levels of target genes between the two strains. xlnA, xlnE (Fig. 2a, b), xlnR and araR (Fig. 2e, f) showed higher levels in PO212 compared to S27. The expression of xylP was similar between both strains (Fig. 2c). The xlnD expression was variable in PO212 with a notable increase at 90 min, but in S27 xlnD levels remained constant in an intermediate state compared to PO212 (Fig. 2d). The observed differential expression between PO212 and S27 strains, prompted us to study how deletion of the transcriptional regulator XlnR in PO212 might alter this regulation and its BA.

Fig. 2
figure 2

Relative expression (2−∆∆Ct) of endo-xylanases genes xlnA (a), xlnE (b) and xylP (c), β-xylosidase xlnD (d) and transcriptional regulators xlnR (e) and araR (f) in the expression assays with PO212 and S27 strains in MM with xylan as carbon source. Mycelial samples were collected at 30, 60 and 90 min. Time 0 indicates mycelium grown in glucose medium for 24 h prior to transfer to medium containing xylan. Data from xlnD (d) (PO212) were subject to log10 transformation to improve the homogeneity of variances before analysis. Means with the same letter for each strain and for each gene are not significantly different from each other (p ≥ 0.05) according to the Duncan test. Vertical bars represent the standard error of the mean of two biological replicates with three technical replicates

Full size image

Generation of a null allele of xlnR in PO212

Generation of genetic tools

To understand the implication of PO212 endo-xylanases in BA, we deleted the transcriptional regulator xlnR. For this purpose, we generated a series of strains lacking the gene encoding the non-homologous joining helicase (∆ku70) and able to accept the construct for deleting xlnR.

The first step towards obtaining a null xlnR strain was the transformation of recipient strain PO212_18.2 to generate ∆ku70 allele. We replaced ku70 coding sequence by the auxotrophic selectable marker pyrG from A. fumigatus. A linear DNA fragment of 4.9 kb was constructed by fusion PCR procedures (TC 1). Figure 3a step 1 describes the procedure for constructing the transformation cassette and the expected recombination events leading to the gene replacement of ku70 CDS.

Fig. 3
figure 3

Generation of null ku70 strains. a. Scheme of transformation cassettes used in Step 1 and Step 2. In step 1, the locus of the recipient strain PO212_18.2, the transformation cassette 1 (TC 1) composed by 5’ UTR-ku70, pyrGAf gene and 3’ UTR-ku70 fused (4.9 kb) and the final product after the locus replacement, resulting in a new strain, 1.1∆ku70. In step 2, the locus of the recipient strain 1.1∆ku70, the TC 2 composed by 5’ UTR and 3’ UTR of ku70 gene fused (3.1 kb) and the final product after locus replacement by adding 5-FOA. The new strain was T2.7. b. Growth test of T2.6 and T2.7 in MM in presence ( +) or absence (-) of uracil and uridine (uu) as selected marker. The strains were grown for 5 days at 25 °C. c. PCR product. The order from left to right are: T2.6 and T2.7 are two transformants with the correct amplified fragment size (3.1 kb), 1.1∆ku70 is the recipient strain, with the size of the fragment of 4.9 kb, corresponding to 5’ UTR fused to pyrGAf and 3’ UTR region and PO212, with wild type locus. C- Corresponds to the negative control of PCR. Mw: Molecular weight marker

Full size image

Further to obtaining primary transformants for ∆ku70 allele, we selected strain 1.1∆ku70 and proceeded to reutilize the pyrG1 mutation present in recipient strain PO212_18.2 by a second round of transformation but using a new linear fragment (TC 2, 3.1 kb) comprising the promoter and terminator fused together (Fig. 3a, step 2). Positive transformants were selected on regeneration medium containing 5-FOA. Among transformants, we selected mutant pyrG1 strains T2.6 and T2.7 (Fig. 3b) and PCR analyses showed that the precise deletion of ku70 CDS was carried out without remains of selectable marker (Fig. 3c). The recombinant strain T2.7 was selected as the recipient strain for future transformations.

Generation of ∆xlnR strains

The xlnR null allele was generated through precise locus replacement using mutant T2.7 strain as recipient. Two approaches were followed to increase the probability of obtaining the null xlnR allele. A single transformation cassette designated as TC 3 was constructed by fusion of fragments 5’ UTR xlnR, pyrG gene of A. fumigatus and 3’ UTR xlnR using PCR techniques. Alternatively, we constructed two fragments overlapping truncated versions of pyrGAf selectable marker, TC 4 and TC 5 (Fig. 4a). Both strategies should render similar results and we were able to obtain transformants in both approaches (Fig. 4b).

Fig. 4
figure 4

Generation of null xlnR strains. a. Scheme of the transformation cassettes used for T1.2 and T2.9 strains generation. b. Cultures on selective medium allowing purification of homokaryotic xlnR null colonies. T1: transformation with the whole transformation cassette (TC 3). T2: co-transformation with the overlapping cassettes (TC 4 and TC 5). c. PCR product. The order from left to right are: different transformants obtained from different ways (T1 or T2). T2.7 is the recipient strain with a PCR product of 5.9 kb as well as PO212, as expected. C- Corresponds to the negative control of PCR. Mw: Molecular weight marker. Yellow triangles indicate the selected strains for the following assays. d. Growth test of xlnR null strains (T1.2 and T2.9) in comparison to the recipient strain T2.7 and PO212 in MM supplemented with glucose (1%) or xylan (1%) as carbon source. The strains were grown for 5 days at 25 °C. The addition of uracil and uridine (uu) in order to complement the auxotrophy of T2.7 caused by the lack of pyrG marker was necessary

Full size image

The correct insertion of these fragments was confirmed by PCR for most transformants (Fig. 4c). Strains lacking xlnR gene were unable to grow healthy on medium containing 1% xylan as carbon source but grew normally when the medium was supplemented with glucose. Only a residual growth was possible to appreciate due to there must be other enzymes able to degrade xylan (Fig. 4d).

Regulation of xylanolytic system in PO212

To determine the effect of the absence of XlnR function we measured the expression levels of three β-1,4-endo-xylanases (xlnA, xlnE and xylP), one β-xylosidase gene (xlnD) and the regulator gene (araR) coding genes in T1.2 and T2.9 strains and compared to PO212. Expression levels of xlnR in PO212 elevated along time in mycelium cultivated in medium containing 1% xylan (Fig. 5e). As expected xlnR transcript was not detected in samples from null xlnR T1.2 and T2.9 (Fig. 5e).

Fig. 5
figure 5

Relative expression (2−∆Ct) of endo-xylanases genes xlnA (a), xlnE (b) and xylP (c), β-xylosidase xlnD (d) and regulators xlnR (e) and araR (f) in the expression assays with PO212, T1.2 and T2.9 strains in MM with xylan as carbon source. Samples were collected at 30, 60 and 90 min. Time 0 indicates mycelium grown in glucose medium for 24 h prior to transfer to medium containing xylan. The Kruskal–Wallis test of independence was used for statistical analysis. Vertical bars represent the standard error of the mean of three biological replicates each one including three technical replicates

Full size image

After confirmation of the absence of xlnR gene expression in null strains, we studied the gene expression of β-1,4-endo-xylanases xlnA xlnE and xylP (Fig. 5a-c), and β-xylosidase xlnD (Fig. 5d) under the same conditions described previously and were compared to PO212. The xlnA expression levels in PO212 increased in presence of xylan (T90); whereas in null strains, xlnA showed a basal expression either glucose or xylan during the studied times (Fig. 5a). Expression levels of xlnE, xylP and xlnD (Fig. 5b-d) genes showed similar profiles. These genes had higher expression in PO212 after 90 min in xylan compared to null strains, where there was hardly any increase in the level of these genes over time. When determining the expression levels of araR (Fig. 5f), this putative transcriptional regulator involved in the catabolism of arabinose branches from xylan, showed a significant higher expression in null xlnR strains compared to PO212.

Efficacy assays

To assess the effect of ∆xlnR allele on BA we performed biocontrol assays in which we compared T1.2 and T2.9 ∆xlnR strains to PO212. We also included in these experiments the intermediate strains 1.1∆ku70 and T2.7 as control of the possible effect of deleting ku70 gene. Thus, 1.1∆ku70, T2.7, T1.2 and T2.9 strains were used in these assays and compared to PO212 efficacy on reducing Fusarium wilt disease in tomato plants. Conidia suspensions from all strains had a germination rate higher than 95% before the treatment. Plants treated with conidia from PO212, 1.1∆ku70 and T2.7 strains showed a reduction of the disease severity, indicating that 1.1∆ku70 and T2.7 strains had not lost the efficacy of control against Fusarium wilt. However, the treatment with T1.2 and T2.9 conidia did not reduce either disease severity or incidence (Table 3).

Table 3 Effect of PO212 and derive strains on the severity (%) and incidence (%) of the disease caused by Fusarium oxysporum f.sp. lycopersici in tomato plants cv. “San Pedro” at 28 days after inoculation in growth chamber
Full size table

Discussion

The ability of PO212 as an effective BCA against several pathogens in horticultural crops such as FOL in tomato plants has been widely demonstrated (De Cal et al., 1995, 2008; Larena et al., 2003a, b; Martinez-Beringola et al., 2013). It is documented that PO212 confers resistance against FOL in tomato plants via the host by triggering induced resistance (De Cal et al., 1997, 1999, 2000). However, the molecular basis of biocontrol phenotype displayed by PO212 is still poorly understood.

Since β-1,4-endo-xylanases have been previously described as possible elicitors of the plant defense response (Dean et al., 1989), we considered studying the effect of these xylanolytic enzymes on the BA of PO212. Based on the extended literature in Aspergillus spp. and the availability of protein sequences of endo-xylanases in A. nidulans and PrWis proteomes, we retrieved and defined the xylanolytic system in PO212 and S27 strains composed by three endo-xylanases (xlnA, xlnE and xylP), one β-xylosidase (xlnD) and two transcriptional regulators (xlnR and araR). The nucleotide sequence of each endo-xylanase coding genes was compared between PO212 and S27 strains, and no differences were observed between both strains. These results confirmed those published in Requena et al. (2023), since these genes were not within those genes that presented differences. In PO212 and S27 genomes, the xlnB coding gene was not found, unlike in A. nidulans and P. purpurogenum. In each of these fungi, two endo-xylanases were found (named XlnA and XlnB in A. nidulans) which were regulated by ambient pH via PacC, being differentially expressed in alkaline and acidic growth conditions (Belancic et al., 1995; MacCabe et al., 1998). The fact that PO212 present difficulties to grow well on alkaline growth conditions may explain the presence of a single gene encoding these β-1,4-endo-xylanases (Pascual et al., 1997).

To study the enzymatic regulation of endo-xylanases in PO212 BCA, in vitro assays with lyophilized tomato roots and xylan were carried out. Understanding the regulation of PO212 endo-xylanases could give us an insight into their involvement in biocontrol activity. In presence of xylan, the increase of xlnA, xlnE and xylP genes in PO212 appears to be correlated with the increase of xlnR. This analysis revealed high differences in the expression patterns of endo-xylanases genes between PO212 and S27 strains, since S27 presented a lower expression of xlnR and therefore, a lower expression of xlnA and xlnE was observed. A similar regulation was observed in A. niger by van Peij et al. (1998a, b), regulating the expression of xlnB, xlnC and xlnD genes. The low expression of xlnR in presence of glucose was also observed confirming that xlnR promoter is repressed by glucose via CreA, and when the repression is abolished, the expression of xlnA, xlnB and xlnD stops being repressed as in A. nidulans (de Graaff et al., 1994; Tamayo et al., 2008).

The use of xlnR null strains is a powerful tool which has allowed us to evaluate the role of the transcriptional regulator on the BA of PO212. Moreover, we have developed a homologous transformation system in PO212 to delete any gene and be able to study its function on the biocontrol phenotype background. In this study, two different strains were obtained following two different strategies, and having the similar behavior. These two xlnR null strains were unable to grow on xylan as well as they did when MM was supplemented with glucose. A residual growth was possible to appreciate since, due to the xylan heterogeneity, a complex set of enzymes is required for its degradation (Biely, 1985).

In vitro assays with null xlnR strains and PO212 showed that in absence of xlnR, the expression level of xlnA, xlnE, xylP and xlnD are very low and did not experience a rise at time T90 as does PO212. This confirms what it was already seen in A nidulans, in which XlnR regulates the transcription of the endo-xylanases xlnB and xlnC, and the β-xylosidase xlnD coding genes (van Peij et al., 1998a, b). Moreover, an increase of araR gene expression in the ∆xlnR background was observed in presence of xylan as carbon source. This may be related to the catabolism of the arabinose residues present in the xylan side chains, as is the case with A. niger, responding to arabinose by AraR (Battaglia et al., 2011). The residual growth observed in XlnR null strains on media contained xylan as carbon source, could be due to a higher expression of this araR gene. In addition, efficacy assays showed that ∆xlnR strains lost their ability to reduce Fusarium wilt in tomato plants, demonstrating the involvement of this system in PO212 BA.

These results strongly supports a model in which the activity of the xylanolytic system driven by XlnR or the predicted extracellular presence of xylanases are key elements in mediating the biocontrol activity of PO212. This work showed that the search of key elements in biocontrol, which is a complex interaction, involves exploring pathways that may be differentially expressed, and supports the conclusions reached with the sequencing of the two strains used in this study, PO212 and S27, that show a differential phenotype in terms of their capacity as BCAs.