Development of a FungalBraid Penicillium expansum-based expression system for the production of antifungal proteins in fungal biofactories.

Fungal antifungal proteins (AFPs) have attracted attention as novel biofungicides. Their exploitation requires safe and cost-effective producing biofactories. Previously, Penicillium chrysogenum and Penicillium digitatum produced recombinant AFPs with the use of a P. chrysogenum-based expression system that consisted of the paf gene promoter, signal peptide (SP)-pro sequence and terminator. Here, the regulatory elements of the afpA gene encoding the highly produced PeAfpA from Penicillium expansum were developed as an expression system for AFP production through the FungalBraid platform. The afpA cassette was tested to produce PeAfpA and P. digitatum PdAfpB in P. chrysogenum and P. digitatum, and its efficiency was compared to that of the paf cassette. Recombinant PeAfpA production was only achieved using the afpA cassette, being P. chrysogenum a more efficient biofactory than P. digitatum. Conversely, P. chrysogenum only produced PdAfpB under the control of the paf cassette. In P. digitatum, both expression systems allowed PdAfpB production, with the paf cassette resulting in higher protein yields. Interestingly, these results did not correlate with the performance of both promoters in a luciferase reporter system. In conclusion, AFP production is a complex outcome that depends on the regulatory sequences driving afp expression, the fungal biofactory and the AFP sequence.

Introduction

Antifungal proteins (AFPs) secreted by filamentous ascomycetes have lately gained attention as biofungicides since they specifically inhibit fungal growth without affecting plant or mammalian cell viability (Vila et al., 2001; Szappanos et al., 2005, 2006; Palicz et al., 2013). AFPs are small cationic cysteine‐rich proteins (CRPs) that form three or four disulfide bonds and fold into compact tertiary structures, which makes AFPs highly stable against adverse biochemical and biophysical conditions such as pH, temperature and proteolysis (Batta et al., 2009). Similar to other CRPs, AFPs contain a conserved γ‐core motif (Yount and Yeaman, 2004) and are coded with a signal peptide (SP) at the N‐termini that includes a pre‐sequence involved in AFP secretion to the extracellular space and a pro‐sequence (SP‐pro sequence) that has been predicted to inactivate the protein until cleavage (Marx et al., 1995). AFPs exhibit potent antifungal activity and different mechanisms of action against opportunistic human, animal, plant and foodborne pathogenic fungi (Marx et al., 2008; Hegedüs and Marx, 2013; Delgado et al., 2016; Tóth et al., 2020a; Czajlik et al., 2021; Martínez‐Culebras et al., 2021).

The first identified AFPs were the one produced by Aspergillus giganteus (Nakaya et al., 1990) and the so‐called PAF protein secreted by Penicillium chrysogenum (Marx et al., 1995). Both are abundantly secreted, with yields above 50 mg l‐1. This is also the case of the protein PeAfpA from Penicillium expansum, with purification yields reaching up to 125 mg l‐1 (Garrigues et al., 2018). By contrast, the AFPs from Neosartorya (Aspergillus) fischeri, NFAP and NFAP2, are produced in modest amounts (Kovács et al., 2011; Tóth et al., 2016), while PdAfpB from Penicillium digitatum and PAFB from P. chrysogenum remained undetectable in the culture medium although their encoding genes were transcribed at high levels (Garrigues et al., 2016; Huber et al., 2018).

The potential use of AFPs as novel biocidal compounds in crop and postharvest protection (Vila et al., 2001; Theis et al., 2005; Barakat, 2014; Garrigues et al., 2018, 2020; Delgado et al., 2019; Tóth et al., 2020a, 2020b; Gandía et al., 2021), medicine (Garrigues et al., 2018; Tóth et al., 2018; Kovács et al., 2019; Holzknecht et al., 2020) and food preservation (Delgado et al., 2015; Martínez‐Culebras et al., 2021) has been extensively reported, but mainly under laboratory conditions. Undoubtedly, further exploitation of AFPs requires safe, efficient and economic biofactories for their production. P. chrysogenum was demonstrated to serve as an efficient fungal expression factory for AFPs. The expression of recombinant AFPs was achieved with the use of a P. chrysogenum‐based expression system, the paf cassette, which consisted of the strong paf gene promoter (Ppaf), paf SP‐pro sequence for protein processing and secretion, and terminator (Tpaf) (Sonderegger et al., 2016). This approach allowed the overexpression of high amounts of PAF, PAF mutants, NFAP and NFAP2 (Sonderegger et al., 2016, 2018; Tóth et al., 2018), P. expansum PeAfpB and PeAfpC (Garrigues et al., 2018) and P. chrysogenum PAFB (Huber et al., 2018) and PAFC (Holzknecht et al., 2020). The versatility of this system was demonstrated with the successful expression of PAF and PdAfpB in P. digitatum (Sonderegger et al., 2016; Garrigues et al., 2017).

FungalBraid (FB) is a synthetic biology modular cloning platform for the assembly and exchange of DNA elements tailored to fungal biotechnology and adapted from the GoldenBraid (GB) system developed for plants (Sarrión‐Perdigones et al., 2011; Vázquez‐Vilar et al., 2017, 2020; Hernanz‐Koers et al., 2018). Moreover, both GB and FB systems are fully compatible and allow the exchange of suitable DNA parts between plants and fungi. We had previously adapted the DNA elements of the paf cassette to FB to facilitate the straightforward cloning and efficient production of AFPs and rationally designed variants (Heredero et al., 2018; Hernanz‐Koers et al., 2018). The paf cassette was also successful to homologously produce proteins different from AFPs such as Sca, an anionic P. digitatum CRP, in high yields (Garrigues et al., 2020).

Since PeAfpA is a highly secreted protein that reaches yields above 100 mg l‐1 in its wild‐type producing fungus, regulatory elements of afpA gene provide an excellent opportunity to develop a P. expansum‐based expression system for the production of AFPs and likely other CRPs. In this work, we have evaluated the feasibility of the afpA gene promoter and terminator sequences to drive the expression of AFPs. This afpA cassette, together with the previously described paf cassette, have been assessed for PeAfpA and PdAfpB production, using their corresponding native signal peptide (SP)‐pro sequences, in P. chrysogenum and P digitatum. Finally, the strength of afpA and paf promoters was tested and compared in a luciferase‐based reporter system.

Results

Development of a FungalBraid P. expansum‐based expression system

The FB system follows the standards of synthetic biology and enables reusability of genetic parts and direct comparison of experiments among different fungal species. In this study, we used different already available FB elements and generated others to compare the regulatory elements of the afpA and paf genes to produce PeAfpA and PdAfpB in two fungal biofactories, P. chrysogenum and P. digitatum.

For the development of the afpA cassette, three genetic elements were domesticated: the afpA promoter (PafpA) and terminator (TafpA) sequences from P. expansum and the coding sequence (CDS) from the afpA gene including its own SP‐pro sequence (Table 1). In order to compare the P. expansum‐based expression system with the previously described paf cassette (Sonderegger et al., 2016), we used the already available elements Ppaf and Tpaf together with the afpB gene sequence from P. digitatum (Hernanz‐Koers et al., 2018). Former studies from our group showed that P. digitatum produced up to 20 mg l‐1 of PdAfpB when expressed under the control of Ppaf and Tpaf (Garrigues et al., 2017).

Table 1

GoldenBraid (GB) and FungalBraid (FB) elements used and FB vectors generated in this study.

Code	Genetic Element(s)	GB Plasmid	References
GB0096	Luciferase (luc)	pUPD	GB website a
GB3458	3a1_AtS/MAR10 insulator sequence	pDGB3α1	Pérez‐González et al. (2019)
FB002	Ttub	pUPD2	Hernanz‐Koers et al. (2018)
FB003	PtrpC::hph::Ttub	pDGB3α2	Hernanz‐Koers et al. (2018)
FB007	PgpdA	pUPD2	Hernanz‐Koers et al. (2018)
FB009	PtrpC::nptII::Ttub	pDGB3α2	Hernanz‐Koers et al. (2018)
FB029	Ppaf	pUPD2	Hernanz‐Koers et al. (2018)
FB030	Tpaf	pUPD2	Hernanz‐Koers et al. (2018)
FB031	afpB	pUPD2	Hernanz‐Koers et al. (2018)
FB033	Ppaf::afpB::Tpaf	pDGB3α1R	Hernanz‐Koers et al. (2018)
FB036	Ppaf::afpB::Tpaf(←)::PtrpC::hph::Ttub(→)	pDGB3Ω1	Hernanz‐Koers et al. (2018)
FB107	PafpA	pUPD2	This work
FB108	afpA	pUPD2	This work
FB109	TafpA	pUPD2	This work
FB112	PafpA::afpA::TafpA	pDGB3α1R	This work
FB114	PafpA::afpA::TafpA(←)::PtrpC::hph::Ttub(→)	pDGB3Ω1	This work
FB115	PafpA::afpA::TafpA(←)::PtrpC::nptII::Ttub(→)	pDGB3Ω1	This work
FB116	Ppaf::afpB::Tpaf(←)::PtrpC::nptII::Ttub(→)	pDGB3Ω1	This work
FB146	Ppaf::afpA::Tpaf	pDGB3α1R	This work
FB158	Ppaf::afpA::Tpaf(←)::PtrpC::hph::Ttub(→)	pDGB3Ω1	This work
FB159	Ppaf::afpA::Tpaf(←)::PtrpC::nptII::Ttub(→)	pDGB3Ω1	This work
FB230	PafpA::afpB::TafpA	pDGB3α1R	This work
FB244	PafpA::afpB::TafpA(←)::PtrpC::hph::Ttub(→)	pDGB3Ω1	This work
FB245	PafpA::afpB::TafpA(←)::PtrpC::nptII::Ttub(→)	pDGB3Ω1	This work
FB258	Ppaf::luc::TtrpC	pDGB3α2	This work
FB259	PafpA::luc::TtrpC	pDGB3α2	This work
FB261	AtS/MAR10(→)::Ppaf::luc::TtrpC(→)	pDGB3Ω2	This work
FB262	AtS/MAR10(→)::PafpA::luc::TtrpC(→)	pDGB3Ω2	This work
FB310	Nanoluciferase (Nanoluc)	pUPD2	This work
FB312	PgpdA::Nanoluc::Ttub	pDGB3α1	This work
FB316	PgpdA::Nanoluc::Ttub(→)::PtrpC::nptII::Ttub(→)	pDGB3Ω1	This work
FB323	FB316(→)::AtS/MAR10(→)::Ppaf::luc::TtrpC(→)	pDGB3α1	This work
FB324	FB316(→)::AtS/MAR10(→)::PafpA::luc::TtrpC(→)	pDGB3α1	This work

Glyceraldehyde‐3‐phosphate dehydrogenase gene (gpdA); neomycin phosphotransferase gene (nptII) conferring geneticin resistance; hygromycin‐B‐phosphotransferase gene (hph); tryptophan biosynthesis protein C gene (trpC); β‐tubulin gene (tub).

https://gbcloning.upv.es/

To test and compare the production of PeAfpA and PdAfpB under the control of either PafpA and TafpA, or Ppaf and Tpaf sequences, FB multipartite and binary assemblies were successfully performed, and the binary vectors needed for the final transformation of either P. chrysogenum or P. digitatum were generated (Table 1; Fig. 1; see experimental procedures for further details).

Fig. 1

Schematic diagrams of FungalBraid transcriptional units (TU) and binary vectors for the expression of afp genes under the control of either the afpA or paf cassettes.

A. Binary assembly of TUs FB112 and FB146 with hygromycin (hph) resistant marker (FB003) and geneticin (nptII) resistant marker (FB009) to obtain the final binary vectors to transform P. digitatum (FB114 and FB158) and P. chrysogenum (FB115 and FB159) for the production of PeAfpA.

B. Binary assembly of TUs FB230 and FB033 with FB003 and FB009 to obtain the final binary vectors to transform P. digitatum (FB036 and FB244) and P. chrysogenum (FB116 and FB245) for the production of PdAfpB. FB033 is described in Hernanz‐Koers et al. (2018).

PeAfpA production is only achieved using the P. expansum‐based expression system

We used the afpA cassette and the paf cassette for recombinant production of PeAfpA in P. chrysogenum (Table 1; Fig. 2A). Several P. chrysogenum positive independent transformants (Fig. S1) were obtained and protein production was evaluated after 5 days of growth in P. chrysogenum minimal medium (PcMM) (Sonderegger et al., 2016). Figure 2B shows the SDS‐PAGE and western blot analyses of culture supernatants of seven positive clones from each transformation experiment. For recombinant production under the control of the afpA cassette (Fig. 2B, left panel), three out of the seven transformants analysed produced a protein band of apparent molecular mass similar to that of the pure PeAfpA. Moreover, PeAfpA‐specific signals were observed in these three supernatants when analysed by western blot (PCMG11522, PCMG11532 and PCMG11552), confirming the production of the protein. On the contrary, for recombinant PeAfpA production under the control of the paf cassette (Fig. 2B, right panel), no PeAfpA‐producing clones were obtained, suggesting that PeAfpA is not present, or is not present in detectable amounts in supernatants of P. chrysogenum transformants obtained under the control of the paf promoter and terminator sequences. Analysis of the gene copy number by quantitative PCR (qPCR) of genomic DNA indicated that two of the three PeAfpA‐producer clones likely contain two copies of the PafpA::afpA::TafpA construction (PCMG11522 and PCMG11552), while the low producer PCMG11532 contains at least 3 copies (Fig. 2C). Figure 2D shows the growth of the P. chrysogenum PeAfpA producing clones, the reference strain P. chrysogenum Q176 and the parental P. chrysogenum strain used for transformation (Δpaf), in potato dextrose agar (PDA) and PcMM media. The radial growth of P. chrysogenum transformants was indistinguishable from those of the control strains regardless of the medium, although conidia production of PCMG11532 was significantly lower than that showed by control strains and PCMG11522 and PCMG11552 in PDA, which was further confirmed by conidia production measurement (Fig. S2). The clone PCMG11552 was selected as the highest recombinant protein producer for further characterization.

Fig. 2

Analyses of P. chrysogenum transformants for PeAfpA production with either the afpA or the paf cassette.

A. Schematic diagram of the binary vectors FB115 and FB159 used for P. chrysogenum transformation.

B. SDS‐PAGE (top) and western blot analyses (bottom) of pure PeAfpA (2 µg) and growth supernatants of recombinant strains (10 µl of 10× supernatants loaded per lane) obtained by either FB115 (left) or FB159 (right) transformation. SDS‐PAGE analyses were visualized by Coomassie blue staining; M: SeeBlue® Pre‐stained protein standard. Immunoblot analyses were performed using specific anti‐PeAfpA antibody. Parental strain Δpaf was loaded as a negative control. Positive PeAfpA producing strains (PCMG11522, PCMG11532 and PCMG11552) are highlighted in red.

C. Evaluation of afpA gene copy number in the different PeAfpA producing strains by qPCR. The Ct signal of afpA and L18a was normalized to that of β‐tub. Results are presented as mean values ± standard deviation (SD) of three technical replicates. Under this experimental design, the resulting gene copy number is expected to be 1 for afpA in CMP‐1, and ≥ 1 for afpA in P. chrysogenum transformants.

D. Colony morphology of P. chrysogenum PeAfpA producing strains PCMG11522, PCMG11532 and PCMG11552 compared to the wild‐type Q176 and the parental strain Δpaf after 6 days of growth on PDA and PcMM plates.

Thereafter, we tested both expression systems for the recombinant production of PeAfpA in P. digitatum (Fig. 3A). Several P. digitatum positive transformants were obtained from each transformation experiment (Fig. S1) and evaluated after 11 days of growth in P. digitatum minimal medium (PdMM) (Sonderegger et al., 2016). Figure 3B shows that faint PeAfpA‐specific signals were immunodetected in three out of six culture supernatants, corresponding to the recombinant production under the control of the afpA cassette (PDGL11412, PDGL11432 and PDGL11442). However, no signal was observed in the supernatants corresponding to the paf cassette. These results confirm that, in the conditions tested, PeAfpA is only detected in supernatants of transformants obtained under the control of its own regulatory elements. Analysis of the gene copy number revealed that only one copy of the PafpA::afpA::TafpA construction was randomly inserted in these three strains in which mild PeAfpA production was detected (Fig. 3C). The growth in PDA and PdMM of the P. digitatum PeAfpA‐producing clones and the parental P. digitatum strain used for transformation (CECT 20796) are shown in Fig. 3D. P. digitatum transformants did not show phenotypical differences compared to the parental strain in both media.

Fig. 3

Analyses of P. digitatum transformants for PeAfpA production using either the afpA or the paf cassette.

A. Schematic diagram of the binary vectors FB114 and FB158 used for P. digitatum transformation.

B. SDS‐PAGE (top) and western blot analysis (bottom) of pure PeAfpA (2 µg) and growth supernatants of recombinant strains (10 µl of 10× supernatants loaded per lane) obtained by either FB114 or FB158 transformation. SDS‐PAGE analyses were visualized by Coomassie blue staining; M: SeeBlue® Pre‐stained protein standard. Immunoblot analyses were performed using specific anti‐PeAfpA antibody. Parental strain CECT 20796 was loaded as a negative control. Positive PeAfpA producing strains (PDGL11412, PDGL11432 and PDMG11442) are highlighted in red.

C. Evaluation of afpA gene copy number in the different PeAfpA‐producing strains by qPCR. The Ct signal of afpA and L18a was normalized to that of β‐tub. Results are presented as mean values ± SD of three technical replicates. Under this experimental design, the resulting gene copy number is expected to be 1 for afpA in CMP‐1, and ≥ 1 for afpA in P. digitatum transformants.

D. Colony morphology of P. digitatum PeAfpA producing strains PDGL11412, PDGL11432 and PDGL11442 compared to the parental strain CECT 20796 after 7 days of growth on PDA and PdMM plates.

Comparison of production between recombinant PeAfpA‐producing strain P. chrysogenum PCMG11552 and wild‐type P. expansum CMP‐1

To compare the level of PeAfpA production, the selected PeAfpA producer PCMG11552 and P. expansum CMP‐1 were grown in PcMM and, subsequently, analyzed by SDS‐PAGE and western blot in a time‐course experiment (Fig. 4A). As expected, PeAfpA‐specific signals were immunodetected in most of the supernatants evaluated. Both strains produced PeAfpA from 5 to 10 days of growth. Nevertheless, in CMP‐1 supernatants, PeAfpA amount increased throughout the experiment, whereas in PCMG11552, the protein reached a maximum at day 7. With the exception of 5‐day supernatants, higher quantities of PeAfpA were detected in CMP‐1 supernatants.

Fig. 4

Production and identification of PeAfpA in wild‐type P. expansum CMP‐1 and recombinant P. chrysogenum PCMG11552 strains.

A. SDS‐PAGE (top) and western blot analysis (bottom) of 10 µl of 5× supernatants of strains grown in P. chrysogenum minimal medium (PcMM) for 3, 5, 7 and 10 days. One µg of pure PeAfpA was added as control. SDS‐PAGE analysis was visualized by Coomassie blue staining; M: SeeBlue® Pre‐stained protein standard. Immunoblot analysis was performed using specific anti‐PeAfpA antibody.

B. Peptide mass fingerprinting (PMF) of the recombinant PeAfpA protein purified from PCMG11552 grown in PcMM for 5 days. Peptides obtained by PMF covered 49% of PeAfpA primary sequence (top).

C. Dose–response curve comparing the antifungal activity of native (red circles) and recombinant PeAfpA (blue circles) against P. digitatum. Plotted data are mean values ± SD of triplicate samples after 48 h at 25°C.

Next, we characterized the processing of the recombinant PeAfpA. For this purpose, the protein was purified from PcMM supernatant of PCMG11552 grown for 5 days. After one‐step cation‐exchange chromatography, pure protein was subjected to peptide mass fingerprinting (PMF). Figure 4B shows a sequence coverage of 49% and the proper PeAfpA processing in the N‐terminal region, in which the SP‐pro peptide is cleaved out. The yield of recombinant protein after purification reached 4.8 mg l‐1, similar to the quantity produced at the same time point by CMP‐1 (4.2 mg l‐1). No differences in antifungal activity were observed among recombinant and native PeAfpA (Fig. 4C).

The feasibility of the expression system to produce PdAfpB depends on the fungal biofactory

In order to expand and compare their applicability, we evaluated the feasibility of both cassettes to produce a second AFP, PdAfpB from P. digitatum.

First, PdAfpB production was evaluated in P. chrysogenum (Fig. 5A). Several P. chrysogenum positive clones were obtained for each transformation event (Fig. S3). SDS‐PAGE and western blot analyses of 5‐days culture supernatants of positive transformants are depicted in Fig. 5B. No production of recombinant PdAfpB was observed in any of the nine evaluated transformants (Fig. 5B left panel) when using the afpA cassette. Conversely, by using the paf cassette, PdAfpB production was observed in two of the six culture supernatants evaluated, as confirmed by the immunoreaction observed in western blot analysis (Fig. 5B right panel, in green). These results evidence that, in the conditions tested, we could only achieve production of PdAfpB in P. chrysogenum under the control of the paf regulatory elements. Analysis of the gene copy number revealed that only one copy of the Ppaf::afpB::Tpaf construction was present in the two P. chrysogenum transformants producing recombinant PdAfpB (PCMG11612 and PCMG11613) (Fig. 5C). Growth comparison of both PdAfpB producing clones with that of P. chrysogenum Q176 and Δpaf strains in solid PDA and PcMM (Fig. 5D) showed no different phenotype on PDA plates, while a reduction of growth in the recombinant strains was observed in PcMM.

Fig. 5

Analyses of P. chrysogenum transformants for PdAfpB production using either the afpA or the paf cassette.

A. Schematic diagram of the binary vectors FB245 and FB116 used for P. chrysogenum transformation.

B. SDS‐PAGE (top) and western blot analyses (bottom) of pure PdAfpB (2 µg) and growth supernatants of recombinant strains (10 µl of 10× supernatants loaded per lane) obtained by either FB245 (left) or FB116 (right) transformation. SDS‐PAGE analyses were visualized by Coomassie blue staining; M: SeeBlue® Pre‐stained protein standard. Immunoblot analyses were performed using specific anti‐PAFB antibody. Parental strain Δpaf was loaded as a negative control. Positive PdAfpB producing strains (PCMG11612 and PCMG11613) are highlighted in green.

C. Evaluation of afpB gene copy number in the different PdAfpB producing strains by qPCR. The Ct signal of afpB and L18a was normalized to that of β‐tub. Results are presented as mean values ± SD of three technical replicates. Under this experimental design, the resulting gene copy number is expected to be 1 for afpB in CECT 20796, and ≥ 1 for afpB in P. chrysogenum transformants.

D. Colony morphology of P. chrysogenum PdAfpB producing strains PCMG11612 and PCMG11613 compared to the wild‐type Q176 and the parental strain Δpaf after 7 days of growth on PDA and PcMM plates.

Successful PdAfpB production in P. digitatum under the control of Ppaf and Tpaf sequences was previously shown (Garrigues et al., 2017; Hernanz‐Koers et al., 2018). In this study, PdAfpB production under the control of the afpA regulatory elements (Fig. 6A) was evaluated. SDS‐PAGE and western blot analyses of the culture supernatants of five positive P. digitatum clones (Fig. S3) are shown in Fig. 6B. A clear PdAfpB‐specific signal was only immunodetected in the PDAL24444 supernatant. By contrast, very faint immunoreaction was observed for PDAL24425 and PDAL24441 strains. Gene copy number analyses revealed that PDAL24441 presented three copies of the PafpA::afpB::TafpA construction while the other two transformants presented one copy each (Fig. 6C). As it can be seen in Fig. 6D, none of the three clones showed phenotypic differences to that of the parental strain CECT 20796 in both solid media tested. These results suggest that, in the conditions tested, PdAfpB can also be detected in supernatants of P. digitatum transformants obtained under the control of PafpA and TafpA sequences.

Fig. 6

Analyses of P. digitatum transformants for PdAfpB production using the afpA cassette.

A. Schematic diagram of the binary vector FB244 used for P. digitatum transformation.

B. SDS‐PAGE (top) and western blot analyses (bottom) of pure PdAfpB (2 µg) and growth supernatants of recombinant strains (10 µl of 10× supernatants loaded per lane) obtained after transformation with FB244. SDS‐PAGE analyses were visualized by Coomassie blue staining; M: SeeBlue® Pre‐stained protein standard. Immunoblot analysis was performed using specific anti‐PAFB antibody. Parental strain CECT 20796 was loaded as a negative control. Positive PdAfpB producing strains (PDAL24425, PDAL24441 and PDAL24444) are highlighted in red.

C. Evaluation of afpB gene copy number in the different PdAfpB producing strains by qPCR. The Ct signal of afpB and L18a was normalized to that of β‐tub. Results are presented as mean values ± SD of three technical replicates. Under this experimental design, the resulting gene copy number is expected to be 1 for afpB in CECT 20796, and ≥ 2 for afpB in the transformants.

D. Colony morphology of P. digitatum PdAfpB producing strains PDAL24425, PDAL24441 and PDAL24444 compared to the parental strain CECT 20796 after 7 days of growth on PDA and PdMM plates.

Comparative analysis of PafpA and Ppaf in a luciferase reporter system

Our results suggest that the effectiveness of PafpA and Ppaf depends on the AFP produced and the fungal biofactory. To further compare the strength of both promoter sequences, a reporter system was designed and constructed to drive the expression of the firefly luciferase (luc) gene. Concurrent expression of the Nanoluciferase (Nanoluc) gene from the shrimp Oplophorus gracilirostris was used as an internal normalization standard.

For the development of the FB luciferase reporter system, the CDS from the Nanoluc gene was introduced in the FB system by routine protocols (Hernanz‐Koers et al., 2018; Vázquez‐Vilar et al., 2020). Genetic elements already adapted to FB (PgpdA, Ttub and the CDS from the luc gene) were also used (Table 1). Next, single TUs combining either Ppaf or PafpA with the luc CDS were assembled. In parallel, the TU for Nanoluc expression was obtained and combined with the geneticin‐resistant marker (FB009) to generate the binary vector FB316. Finally, vectors FB323 and FB324 were generated and used for fungal transformation (Table 1 and Fig. 7A).

Fig. 7

Luciferase assay for testing PafpA and Ppaf strength in Penicillium.

A. Schematic diagram of TU assemblies to drive the expression of the luciferase (luc) gene under the control of PafpA or Ppaf (FB258 and FB259) and the Nanoluciferase (Nanoluc) gene under the control of PgpdA (FB312). Final vectors obtained with geneticin (nptII) resistant marker (FB009) were used for transformation of P. chrysogenum and P. digitatum (FB323 and FB324). An insulator sequence (GB3458) was used to allow the binary assembly of the plasmids containing the luc TU with the plasmid containing Nanoluc TU and geneticin resistant marker.

B. Luciferase/Nanoluciferase signal ratio of 3 independent transformants for each construct in P. digitatum and P. chrysogenum at 2 days of growth in minimal medium (PdMM or PcMM, respectively). Values are represented as the mean ± standard error (SE). Asterisks (*) denote statistically significant differences in comparison to control values (ANOVA and Tukey’s HSD test, P < 0.05).

C. Luciferase/Nanoluciferase signal ratio of a selected transformant for each construct in P. digitatum and P. chrysogenum at 2, 5 and 10 days of growth in PdMM or PcMM, respectively. Values are the means ± SE from three independent replicates. Asterisks (*) denote statistically significant differences between promoters at each time‐point (P < 0.05, Student’s t‐test).

First, three independent positive clones for luc expression under the control of either PafpA or Ppaf in both P. digitatum and P. chrysogenum (Fig. S4) were analysed in 1 ml of PdMM or PcMM after 2 days of growth. Under these growing conditions, very faint luc expression under the control of PafpA was recorded in all the transformants analysed (Fig. 7B, in red). Although slightly higher in the transformants, luc expression was not statistically significant to that showed by the control strains. On the other hand, under the control of Ppaf, luc expression was significantly higher than that observed with PafpA and in the control strains, regardless of the fungal factory (Fig. 7B, in blue). Subsequently, one independent clone for each condition (PDEM32421, PDEM32311, PCEM32431 and PCEM32331) was chosen to monitor gene expression during time‐course experiments in higher volumes of MM, mimicking the conditions used for AFP production (Fig. 7C). Promoter strength was analysed after 2, 5 and 10 days of growth. Regardless of the fungal biofactory and the promoter used, the highest luc expression was achieved at day 5. In P. digitatum, the strength of PafpA was significantly higher than that of Ppaf to drive luc expression at days 2 and 5, whereas at day 10 no significant differences in luc expression between both promoters were observed. By contrast, in P. chrysogenum, the Ppaf was stronger than the PafpA at days 2 and 5, and no significant differential expression was detected at day 10. Results obtained indicated higher luc expression, and thus higher promoter activation, under the growing conditions mimicking those employed for AFP production.

Discussion

In this study, we explored the feasibility of a new P. expansum‐based expression system adapted to the modular cloning platform FB for AFP production in two Penicillium species, the biotechnologically relevant P. chrysogenum (Jami et al., 2010), and the non‐mycotoxigenic postharvest pathogen of citrus fruit P. digitatum (Marcet‐Houben et al., 2012), which has been already demonstrated to produce some AFPs in high yields (Sonderegger et al., 2016; Garrigues et al., 2017). Combinatorial experiments exchanging the different FB DNA parts allowed accurate comparison of the new afpA cassette with that of the previously developed paf cassette for the production of PeAfpA and PdAfpB. As evidenced by our results, the performance of both promoters depends on the AFP sequence, the growing conditions and the biofactory.

In this work, PeAfpA, the most effective AFP from P. expansum (Garrigues et al., 2018), was only produced with the afpA cassette and resulted in higher yields in P. chrysogenum than in P. digitatum. PeAfpA shows a strong antifungal potency against human and plant pathogens, lack of cytotoxicity and significant in vivo protection against phytopathogenic fungi (Garrigues et al., 2018; Gandia et al., 2020). Thus, PeAfpA can be considered a promising compound for application in agriculture, but also in medicine or food preservation (Garrigues et al., 2018; Martínez‐Culebras et al., 2021). Natural production of PeAfpA is achieved from wild‐type P. expansum CMP‐1 strain, reaching yields higher than 100 mg l‐1 (Garrigues et al., 2018). Despite the high production of PeAfpA, P. expansum is a phytopathogenic fungus of concern due to the production of mycotoxins patulin and citrinin (Tannous et al., 2018), and thus the possibility of an alternative and safer cell factory for PeAfpA production is of relevance. Here, we have demonstrated that P. chrysogenum is an efficient cell factory for PeAfpA production using the new afpA cassette. Although the level of recombinant PeAfpA at day 5 is similar to that produced by P. expansum wild‐type strain at the same time point (4.8 vs. 4.2 mg l‐1, respectively), it is far from the maximum level produced by CMP‐1 after 10 days of growth (Garrigues et al., 2018). Further research to improve recombinant PeAfpA yields is in progress. In this work, we have used the P. chrysogenum ∆paf mutant as recipient strain to avoid co‐expression of the wild‐type PAF. However, the use of P. chrysogenum wild‐type strain as factory is also feasible provided that specific AFP purification protocols are employed, as already described for PAFB purification (Huber et al., 2019).

On the other hand, the production of PdAfpB has also been evaluated in this work. PdAfpB is the only AFP encoded in the fungus P. digitatum, although it is not naturally produced (Garrigues et al., 2016). Its biotechnological production was successfully achieved in P. digitatum with the use of the paf cassette (Garrigues et al., 2017). PdAfpB has shown high antifungal activity against mycotoxin‐producing fungi (Martínez‐Culebras et al., 2021) and protection against Botrytis cinerea infection in tomato leaves and plants (Garrigues et al., 2018; Shi et al., 2019). In this work, we have demonstrated that under the regulation of the Ppaf sequence, P. chrysogenum recombinant strains do not reach the levels of PdAfpB produced by P. digitatum transformants (12‐20 mg l‐1) (Garrigues et al., 2017). Finally, the afpA cassette did not result in PdAfpB production in P. chrysogenum, while in P. digitatum transformants faint levels of PdAfpB were achieved. Thus, in the conditions tested, PdAfpB was only detected in supernatants of P. digitatum transformants, being the Ppaf sequence the most appropriate promoter for PdAfpB production.

In this study, we confirm that P. chrysogenum and P. digitatum can be good biofactories for AFP production, although the election of both the promoter and the fungus is dependent on the specific AFP. The production of these small CRPs in bacterial systems often fails due to incorrect folding and disulphide bridge formation (Kiedzierska et al., 2008; Rosano and Ceccarelli, 2014). Transient production in plants is also an attractive alternative (Shi et al., 2019), although fungal systems in general, and filamentous fungi in particular, stand out for better yields and single‐step purification procedures (López‐García et al., 2010; Virágh et al., 2014; Sonderegger et al., 2016; Garrigues et al., 2017; Tóth et al., 2018). Another CRP group of interest comprises antifungal plant defensins. Their heterologous production has been achieved in both E. coli (Bleackley et al., 2016) and P. pastoris (Hayes et al., 2013). The evaluation of both the afpA‐ and paf‐cassette for the production of these CRPs in filamentous fungi will be considered in the near future.

Efficient systems for homologous and heterologous gene expression in P. chrysogenum (Graessle et al., 1997; Díez et al., 1999; Zadra et al., 2000), as well as new promoters for strain engineering have been described (Polli et al., 2016), and this fungus has been extensively used in biotechnology as cell factory for the production of biomolecules (Jami et al., 2010). For AFP production in P. chrysogenum, the paf cassette turned out to be a perfect tool for the generation of correctly folded and active proteins with purification yields in the range from 3 (N. fischeri NFAP) to 80 mg l‐1 (P. chrysogenum PAF) (Sonderegger et al., 2016). Here, the paf cassette was not successful for PeAfpA production and only minor production of PdAfpB was achieved. The original paf cassette included the paf SP‐pro sequence to warrant the secretion of AFPs, as described for the above‐mentioned NFAP and PAF (Sonderegger et al., 2016), but also for P. chrysogenum PAFB and PAFC (Huber et al., 2018; Holzknecht et al., 2020) and P. expansum PeAfpB and PeAfpC (Garrigues et al., 2018). In this work, the production of both PeAfpA and PdAfpB was evaluated using either the afpA or the paf cassette containing the corresponding native SP‐pro sequence, suggesting that for AFP production in P. chrysogenum the paf SP‐pro sequence might be more appropriate. However, in a previous work, the production of PeAfpA in P. chrysogenum with the paf regulatory elements and the paf SP‐pro sequence failed (Garrigues et al., 2018), indicating that the impossibility to detect PeAfpA in culture supernatants is not due to either the native afpA or paf SP‐pro sequence, but to the regulatory sequences which would to some extent compromise the stability of the resulting mRNA.

With respect to P. digitatum as AFP biofactory, the successful production of PdAfpB under the control of Ppaf and Tpaf sequences (Garrigues et al., 2017) prompted us to evaluate PdAfpB production under the control of PafpA and TafpA. However, the protein was produced in very low levels compared to that obtained with the paf cassette (Garrigues et al., 2017). It should be noted that its native SP‐pro sequence allowed production, secretion and optimal purification of the protein with similar yield to that achieved with the paf SP‐pro sequence (Garrigues et al., 2017), pointing out the importance of the regulatory elements also observed in P. chrysogenum. Interestingly, minor amounts of PeAfpA were produced with the afpA regulatory elements, whereas no protein was detected under the control of Ppaf and Tpaf, suggesting a role of the AFP amino acid sequence in protein production/accumulation.

Special attention was paid to the correct processing of recombinant PeAfpA produced by P. chrysogenum. In our previous work, we demonstrated that only one amino acid difference (Leu1) in PdAfpB determined the correct folding/unfolding capabilities of the protein after denaturation procedures, although its antifungal activity was not affected (Garrigues et al., 2017). Here we show that P. chrysogenum proteases efficiently recognized the P. expansum afpA SP‐pro sequence during protein maturation, since the protein is secreted and the N‐terminal region of recombinant PeAfpA is equal to that of the wild‐type protein (Garrigues et al., 2018).

The dependence of the promoter efficiency on each AFP has also been observed for the constitutive PgpdA sequence from Aspergillus nidulans. Attempts to detect PdAfpB in P. digitatum transformants under PgpdA failed, and transformants showed a drastic reduction of axenic growth, abnormal hyphal morphology and delayed conidiogenesis (Garrigues et al., 2016). By contrast, PgpdA was successful to produce the antifungal protein NFAP from N. fischeri in A. nidulans, although in low levels (1.7 mg l‐1), while negatively affected hyphal growth and germination of transformed strains (Galgóczy et al., 2013). Recently, the side‐by‐side comparison of paf, pafB and xylP promoters to produce PAFB in P. chrysogenum revealed that the paf promoter efficiency was similar to that of the pafB, but superior to that of the xylP promoter (Huber et al., 2019).

Our results highlight the difficulties associated with producing antimicrobial peptides and proteins in fungal factories. The toxicity of AFPs against the host strain might partially explain the differences in production. AFPs tested in this work show antifungal effect towards both P. chrysogenum and P. digitatum. Remarkably, only one of the three P. chrysogenum strains producing recombinant PeAfpA (PCMG11532) had conidia production affected, although this could be also related to gene copy number (3 copies for PCMG11532 vs 2 copies for PCMG11522 and PCMG11552). Moreover, we have not found any relationship between gene copy number and protein overproduction. The clone PCMG11552 selected as the highest recombinant PeAfpA producer contains two copies of the PafpA::afpA::TafpA construction, the same number of copies than the clone PCMG11522 in which PeAfpA production is lower, suggesting the effect of the integration loci. In addition, the integration of three copies of the gene does not improve PeAfpA production (clone PCMG11552 vs PCMG11532). In P. digitatum, the integration of one copy of the PafpA::afpA::TafpA did not result in overproduction of the protein, and detection was only accomplished by western blot. Whether production of recombinant PeAfpA in P. digitatum needs at least two copies of the afpA gene as observed in P. chrysogenum requires further studies. Regarding transformants for PdAfpB overproduction, only one copy of the Ppaf::afpB::Tpaf construction was present in the two P. chrysogenum transformants (PCMG11612 and PCMG11613) producing faint quantities of recombinant PdAfpB while in P. digitatum, the strain producing the highest quantities of protein (PDAL24444) presented three copies of the PafpA::afpB::TafpA construction. All together, these results again indicate the importance of the regulatory sequences, the AFP coding sequence and the fungal biofactory.

Finally, we have constructed a reporter expression system for promoter strength analysis in P. chrysogenum and P. digitatum. The luciferase reporter assay is commonly used as a tool to study gene expression at the transcriptional level since it gives quantitative measurements instantaneously. In the system, the luc gene was placed under the control of either PafpA or Ppaf sequences. For comparison and internal calibration, the gene encoding the Nanoluc under the control of the PgpdA was also integrated into the same expression cassette. Although the use of the ratio luc/nanoluc eliminates potential interferences by variations in growth, gene copy number or genome position of the transferred DNA, differences among biological replicates were observed, as described for other reporter systems with internal calibration (Polli et al., 2016). Our data clearly showed differences in gene expression depending on growing conditions. When fungi were grown in low medium volumes with reduced aeration, luc expression was overall much lower. However, in these conditions the performance of Ppaf was much better than that of PafpA. In the growing conditions used for AFP production, both promoters showed a similar expression trend, and the highest expression was achieved at day 5. Intriguing is the lack of correlation between the strength of promoters in the reporter system and that found during protein production experiments. In P. chrysogenum, the strength of Ppaf was higher than that of PafpA, but PeAfpA production was only achieved under the control of its own regulatory sequences while minor quantities of PdAfpB were achieved under the control of Ppaf and Tpaf. These results strengthen the importance of the target AFP to be produced. In P. digitatum, in which the performance of PafpA was slightly higher than that of Ppaf, minor quantities of PeAfpA were detected only in transformants containing the PafpA::afpA::TafpA construction. As reported earlier, high yields of PdAfpB were obtained under the control of Ppaf and Tpaf (Garrigues et al., 2017; Hernanz‐Koers et al., 2018) whereas very faint production was achieved in transformants containing the PafpA::afpB::TafpA construction.

In summary, this study describes the performance of a new afp promoter, PafpA, for the production of AFPs in two fungal biofactories, in comparison with the well‐known Ppaf. In the conditions tested, our findings suggest that there is neither a universal afp promoter nor universal fungal biofactory for the production of a given AFP. The reporter expression system for promoter strength analysis in P. chrysogenum and P. digitatum developed here strengthen the importance of the target AFP to be produced. Future efforts are directed to analyze new fungal promoters and to clarify the role of the afp regulatory and SP‐pro sequences to optimize protein production.

Experimental procedures

Microorganisms, media and culture conditions

Fungal strains used in this study were P. digitatum CECT 20796 (PHI26) (Marcet‐Houben et al., 2012), P. chrysogenum wild‐type strain Q176 and P. chrysogenum Δpaf (Hegedüs et al., 2011). P. digitatum and P. chrysogenum Δpaf were used as parental strains for fungal transformation. Fungi were cultured on PDA (Difco‐BD Diagnostics, Sparks, MD, USA) plates for 7–10 days at 25°C. For transformation, vectors generated were amplified in E. coli JM109 grown in Luria Bertani (LB) medium supplemented with either 25 µg ml‐1 chloramphenicol, 50 µg ml‐1 kanamycin, 100 µg ml‐1 spectinomycin or 100 µg ml‐1 ampicillin at 37°C depending on the vector. Agrobacterium tumefaciens AGL‐1 strain was cultured in LB medium with 20 µg ml‐1 rifampicin at 28°C.

To compare the growth of the P. chrysogenum and P. digitatum transformants with that of parental strains on solid media, 5 µl of conidial suspension (5 × 104 conidia ml‐1) were deposited on the center of PDA and PcMM or PdMM plates (Sonderegger et al., 2016). Colony morphology was assessed and compared by visual inspection after 6–7 days of growth.

Design, domestication and assembly of genetic elements to generate different expression and AFP production vectors through FB system

All the genetic elements previously available or generated in this study are described in Table 1. GB0096 element codes for the firefly luciferase (luc) gene (https://gbcloning.upv.es/feature/GB0096/), whereas GB3458 element encodes an insulator (AtS/MAR10) from Arabidopsis thaliana (Pérez‐González and Caro, 2019), both obtained as GB plasmids. The regulatory elements PafpA and TafpA, and the CDS for the P. expansum afpA gene (PEX2_042150) were designed according to GB rules (https://gbcloning.upv.es) and provided by an external company (IDT, Integrated DNA Technologies) as synthetic genes (gBlocksTM gene fragments). The CDS for Nanoluciferase (Nanoluc, from the shrimp Oplophorus gracilirostris) was codon‐optimized for fungal expression and ordered as a synthetic gene. Each single genetic element was ligated into the domestication entry vector pUPD2 through protocols described previously (Hernanz‐Koers et al., 2018; Vazquez‐Vilar et al., 2020) to obtain FB107, FB108, FB109 and FB310 elements (Table 1). Positive E. coli clones were confirmed by routine PCR amplifications using external specific primers designed for pUPD2 vectors (Hernanz‐Koers et al., 2018) (Table S1) and confirmed by Sanger sequencing.

Multipartite assemblies to obtain the TUs to drive the production of PeAfpA and PdAfpB with the afpA or paf cassettes (Table 1) were carried out as described previously (Hernanz‐Koers et al., 2018; Vázquez‐Vilar et al., 2020). The TU to produce PdAfpB under the control of Ppaf and Tpaf (Table 1) was generated in a previous work (Hernanz‐Koers et al., 2018). To compare the efficiency of the promoters Ppaf and PafpA to drive the expression of the luc gene, the elements FB258 and FB259 were obtained. The TU unit to drive the expression of the Nanoluc gene under the control of the PgpdA promoter (FB007) and Ttub terminator (FB002) was also obtained (FB312; Table 1).

Binary assemblies (FB114, FB115, FB158, FB159, FB116, FB244 and FB245) (Table 1) were obtained as previously described (Hernanz‐Koers et al., 2018; Vázquez‐Vilar et al., 2020) and combine the TUs for the production of PeAfpA (FB112 and FB146) and PdAfpB (FB230 and FB033) with the TUs used as fungal positive selection markers (hygromycin (hph) for P. digitatum and geneticin (nptII) for P. chrysogenum, which correspond to FB003 and FB009 elements, respectively) (Fig. 1). Binary assembly FB036 that combines FB033 and FB003 was previously obtained (Hernanz‐Koers et al., 2018). FB316, FB323 and FB324 were obtained using binary assembly to combine the Nanoluc TU with FB009 element and with insulator TU.

Positive E. coli clones and correct assembly were confirmed by restriction and/or PCR analyses using combinations of the universal specific primers designed for pDGB3 vectors (Hernanz‐Koers et al., 2018) (Table S1). All binary vectors generated that would later be used for fungal transformation were introduced into A. tumefaciens AGL‐1 strain by electroporation. These binary vectors corresponded to FB115, FB159, FB245 and FB116 for P. chrysogenum transformation; and FB114, FB158 and FB244 for P. digitatum transformation (Table 1).

Fungal transformation

Fungal transformation of P. digitatum CECT 20796 and P. chrysogenum Δpaf with the corresponding FB binary vectors described before (Table 1) was performed following the A. tumefaciens‐mediated transformation (ATMT) protocol previously described (Harries et al., 2015; Vázquez‐Vilar et al., 2020). The A. tumefaciens AGL‐1 strain was used for fungal transformation (Gandía et al., 2014). P. digitatum and P. chrysogenum transformants were selected in 25 µg ml‐1 hygromycin B or 25 µg ml‐1 geneticin (G418) (Invivogen, San Diego, CA, USA), respectively. All transformants were confirmed by PCR (Table S1; Fig. S1, S3 and S4) using genomic DNA isolated as described previously (Khang et al., 2006), and subsequently by 1% agarose gel electrophoresis.

Dual luciferase assays

Penicillium chrysogenum and P. digitatum transformants were grown in triplicate in 2 ml tubes containing 1 ml or in 100 ml flasks containing 25 ml of either PcMM or PdMM for 2, 5 or 10 days at 25°C and 150 rpm. Mycelia were then collected by centrifugation (12 000 g, 10 min at 4°C) and immediately frozen in liquid nitrogen. The frozen mycelia (~ 20 mg) were homogenized in 180 µl of Passive Lysis Buffer (Promega, Madison, WI, USA) with a pestle. Luciferase assay was performed with the Dual‐Glo® Luciferase Assay System (Promega), following the manufacturer’s protocol with minor modifications. Briefly, 10 µl of the fungal homogenized extract were transferred to a white 96‐well plate (Thermo Fisher Scientific, Waltham, MA, USA) and mixed with 40 µl of Luciferase Reagent to measure Luciferase luminescence. Thereafter, 40 µl of Stop&Glow Reagent were added and Nanoluciferase luminescence signal quantified. Luciferase and Nanoluciferase luminescence signals were determined using a CLARIOStar multimode microplate reader (BMG LABTECH GmbH, Quakenbrück, Germany) with a measurement of 10 s and a delay of 2 s. Measurements were repeated three times.

Protein production and western blot analyses

For protein production, 25 ml of PcMM or PdMM were inoculated with a final concentration of 106 conidia ml‐1 of transformants and were incubated at 25°C and 150 rpm for 5 (P. chrysogenum) or 11 days (P. digitatum). Total proteins from supernatants and purified PeAfpA and PdAfpB were separated by SDS‐PAGE (16% polyacrylamide gels) and transferred to Amersham Protran 0.20 µm NC nitrocellulose transfer membrane (GE Healthcare Life Sciences, Chicago, IL, USA) as described (Garrigues et al., 2018). Protein detection was accomplished using anti‐PeAfpA antibody diluted 1:2,500 (Garrigues et al., 2018) or anti‐PAFB antibody diluted 1:1,000 (Garrigues et al., 2017). As secondary antibody, 1 : 20 000 dilution of ECL NA934 horseradish peroxidase donkey anti‐rabbit (GE Healthcare Life Sciences) was used and chemiluminescent detection was performed with ECLTM Select Western blotting detection reagent (GE Healthcare Life Sciences) using Amersham Imager 680 (GE Healthcare Life Sciences). The experiments were repeated at least twice.

PeAfpA purification and peptide mass fingerprinting (PMF)

PeAfpA purification from 5‐day PcMM culture supernatants of P. expansum CMP‐1 and P. chrysogenum PCMG1152 was accomplished by cation exchange chromatography following previously published procedures (Garrigues et al., 2018). Peptide mass fingerprinting (PMF) was performed in the proteomics facility of SCSIE University of Valencia (Spain) as previously described (Garrigues et al., 2018).

Determination of afpA and afpB gene copy number

Evaluation of afpA and afpB gene copy number in the different P. digitatum or P. chrysogenum transformants was performed as previously described (Garrigues et al., 2016) with some modifications. Briefly, qPCR was applied to genomic DNA of the P. digitatum or P. chrysogenum AFP‐producing strains with primers specific for afpB (OJM466/OJM467), afpA (OJM588/OJM589), and for the single copy control genes β‐tubulin (β‐tub) (OJM85/OJM86) and the 60S ribosomal protein‐encoding gene L18a (OJM151/OJM152) (Table S1). The Ct signal of afpB, afpA, and L18a was normalized to that of β‐tub used as internal control and to the signal of the parental P. digitatum CECT 20796 (for afpB), or P. expansum CMP‐1 (for afpA). Results are presented as mean values ± SD of three technical replicates.

Conflict of interest

None declared.

Fig. S1. Molecular characterization of P. chrysogenum and P. digitatum transformants for PeAfpA production. (A) Schematic localization of different primers used for PCR analyses. (B) PCR amplification of genomic DNA of the distinct P. chrysogenum strains with different primer pairs as indicated. Positive transformation strains (in red) showed expected amplicons either with primers OJM620/OJM621 (1.8 kb) or OJM483/484 (2 kb). (C) PCR amplification of genomic DNA of the distinct P. digitatum strains with different primer pairs as indicated. Positive transformants (in red) showed expected amplicons either with primers OJM620/OJM621 (1.8 kb) and OJM483/484 (2 kb). Genomic DNA from P. digitatum CECT 20796 and P. chrysogenum Δpaf strains, FB vectors FB003, FB009 and distilled H2O were used as negative controls. FB vectors FB115 and FB159 were used as positive controls.

Fig. S2. Conidia production on PDA plates of P. chrysogenum PeAfpA producing strains. Conidia production was determined as previously described (Gandía et al., 2014). Data show the mean ± standard deviation (SD) of three replicates of the conidia/cm2 produced per plate. Statistical analyses were conducted using the Excel statistical package. Asterisk shows significant differences with the parental strain Δpaf (t‐student test, P < 0.05).

Fig. S3. Molecular characterization of different P. chrysogenum and P. digitatum transformants for PdAfpB production. (A) Schematic localization of different primers used for PCR analyses. Genetic constructions are drawn to scale. (B) PCR amplification of genomic DNA of the P. chrysogenum strains with different primer pairs as indicated. Positive transformation strains (in red) showed expected amplicons either with primers OJM620/OJM621 (1.8 kb) and OJM483/484 (2 kb). (C) PCR amplification of genomic DNA of the P. digitatum strains with OJM466/OJM621 primer pair. Positive transformation strains (in red) showed the expected amplicon (1 kb). Genomic DNA from P. chrysogenum Δpaf and P. digitatum CECT 20796 strains, FB vector FB003 and distilled H2O were used as negative controls. FB vectors FB245 and FB116 were used as positive controls.

Fig. S4. Molecular characterization of P. chrysogenum and P. digitatum transformants for luciferase reporter system. (A) Schematic localization of different primers used for PCR analyses. (B) PCR amplification of genomic DNA of the distinct P. chrysogenum strains with different primer pairs as indicated. Positive transformation strains (in red) showed expected amplicons either with primers OJM509/OJM555 (1.5 kb), OJM620/522 (2.8 kb) or OJM501/522 (2 kb). (C) PCR amplification of genomic DNA of the distinct P. digitatum strains with different primer pairs as indicated. Positive transformants (in red) showed expected amplicons either with primers OJM509/OJM555 (1.5 kb), OJM620/522 (2.8 kb) or OJM501/522 (2 kb). Genomic DNA from P. digitatum CECT 20796 and P. chrysogenum Δpaf strains, FB vector FB009 and distilled H2O were used as negative controls. FB vectors FB323 and FB324 were used as positive controls.

Table S1. Primers used in this study for the molecular characterization of different transformants and for qPCR analyses.

Click here for additional data file.