The fungal strains used in this study were P. digitatum CECT 20796 (isolate PHI26) (Marcet-Houben et al., 2012) and P. chrysogenum wild-type ATCC 10002 (Q176) (Hegedüs et al., 2011). Fungi were routinely cultured on potato dextrose agar (PDA, Difco-BD Diagnostics) plates for 7 days at 25°C. For transformation, the generated vectors were amplified in the bacterium Escherichia coli JM109 and grown in Luria–Bertani (LB) medium at 37°C with either 50 μg/mL chloramphenicol, 50 μg/mL kanamycin, 100 μg/mL spectinomycin, or 100 μg/mL ampicillin depending on the vector. A. tumefaciens AGL-1 was cultured in LB medium at 28°C with 20 μg/mL rifampicin and the corresponding antibiotic depending on the vector used.

For growth profiles, 5 µL of conidial suspension (5 × 10⁴ conidia/mL) was deposited on the center of PDA plates, and colony morphology was assessed and compared daily by visual inspection.

All the genetic elements that have been incorporated on the FB platform are listed in Table 1. New DNA parts were domesticated according to GB rules and tools (https://gbcloning.upv.es) and ordered from an external company as synthetic genes (gBlocks™, IDT). In the case of the coding sequence (CDS) of pyr4, the gene from Trichoderma reesei was codon-optimized according to the optimal codon frequency of Penicillium genera prior to GB/FB domestication. Domesticated elements were ligated into the pUPD2 entry vector via the restriction–ligation protocol, as previously described (Hernanz-Koers et al., 2018; Vazquez-Vilar et al., 2020). Positive E. coli clones were confirmed by routine PCR amplifications and Sanger sequencing using external specific primers OJM524 and OJM525, which are designed for pUPD2 vectors (Hernanz-Koers et al., 2018) (Table 2). Multiple assemblies into pDGB3α vectors of the DNA parts contained in pUPD2 vectors were carried out to obtain different TUs, and binary assemblies were subsequently performed to combine different TUs into multigenetic constructs within these pDGB3α or pDGB3Ω vectors, as previously described (Hernanz-Koers et al., 2018; Vazquez-Vilar et al., 2020).

The sequence of the single-guide RNA 1 (gRNA1) required to activate GB_SynP promoters was checked for the absence of off-target mutations in the P. digitatum genome using Geneious Prime software (https://www.geneious.com/), and was cloned into the pAMA18.0 vector, as described previously (Mózsik et al., 2021). Briefly, a primer pair was designed (OJM698 and OJM699, Table 2) which contained the target sequence of gRNA1, the hammerhead ribozyme, the inverted repetition of the 5′-end of the spacer sequence, and the recognition sites of BsaI. The resulting PCR product was purified (Wizard SV Gel and PCR Clean-Up System, Promega) and inserted into pAMA18.0 via the restriction–ligation reaction with BsaI and T4 ligase. Correct assemblies of the resulting pAMA18.0_gRNA1 vector were confirmed by Sanger sequencing.

Transformation of P. digitatum CECT 20796 (PHI26) and P. chrysogenum ATTC 10002 (Q176) with the corresponding FB binary vectors described in Table 1 was performed through ATMT, as previously described (Khang et al., 2007), with some modifications (Harries et al., 2015; Vazquez-Vilar et al., 2020).

In the case of P. digitatum uridine-auxotrophic ΔpyrG mutants, in which the pyrG gene (gene ID PDIG_38390) was deleted by homologous recombination without the insertion of any positive selection marker, mutants were selected on PDA supplemented with 1.22 g/L uridine (Sigma-Aldrich) and 1.25 g/L of 5-fluoroorotic acid (5-FOA, Formedium). For the ectopically complemented P. digitatum ΔpyrG:pyr4 strains, mutants were selected on PDA plates. P. digitatum and P. chrysogenum ectopic transformants containing ble TU (phleomycin^R) were selected on PDA plates supplemented with 35 μg/mL and 25 μg/mL phleomycin (InvivoGen), respectively. Finally, P. chrysogenum ectopic transformants carrying ergA TU (terbinafine^R) were selected on PDA plates supplemented with 0.5 μg/mL terbinafine hydrochloride (Sigma-Aldrich).

P. digitatum and P. chrysogenum transformants carrying different luciferase reporter constructs to test the constitutive, inducible, and GB_SynP promoters were selected on PDA plates containing 25 μg/mL geneticin (G418) (InvivoGen).

All transformants were molecularly confirmed by PCR reactions using NZYTaq II DNA polymerase (NZYTech) (Figures 1–3; Supplementary Figures S1–S3) from genomic DNA isolated with the NZY Tissue gDNA isolation kit (NZYTech) and primers purchased from IDT (Table 2).

For the validation of GB_SynP promoters, P. digitatum protoplasts from strains carrying luciferase reporter vectors for each of the three synthetic promoters tested (FB398, FB399, and FB400) (Table 1) were transformed with the self-replicative AMA1-based dCas9-containing plasmid pAMA18.0_gRNA1 (Mózsik et al., 2021), as previously described (Garrigues et al., 2022). Transformants were selected on PDA plates containing 0.95 M sucrose and 35 μg/mL phleomycin. Regarding the reusability of the system, the loss of the pAMA18.0_gRNA1 plasmid was confirmed after three consecutive streaks of the transformants in non-selective PDA plates, as previously described (Garrigues et al., 2022).

P. digitatum and P. chrysogenum strains carrying the luciferase reporter vectors for each of the tested promoters (FB433, FB434, FB435, FB436, FB438, FB439, and FB441) and P. digitatum strains with the luciferase reporter for the GB_SynP synthetic promoters (FB398, FB399, and FB400), carrying or not pAMA18.0_gRNA1, were grown in duplicate for 2 days in 100 mL flasks with 25 mL of liquid potato dextrose broth (PDB, Difco-BD Diagnostics) at 25°C, with shaking (150 rpm). For induction assays, transformants were grown in duplicate for 4 days in 100 mL flasks with 25 mL of either P. digitatum or P. chrysogenum minimal medium (PdMM or PcMM, respectively) (Sonderegger et al., 2016) using 2% D-glucose (PanReac), 2% maltose (Sigma-Aldrich) or 2% D-xylose (Sigma-Aldrich) as the sole carbon source. Grown mycelia were filtered, and a sample of 20 mg was collected and immediately frozen in liquid nitrogen. Luciferase and nanoluciferase measurements were performed using the Dual-Glo^® Luciferase Assay System (Promega), as previously described (Gandía et al., 2022). Briefly, frozen samples were homogenized in 180 µL passive lysis buffer with a pestle and centrifuged (12,000 ×g, 10 min at 4°C). Then, 10 µL of the supernatant was transferred to a white 96-well plate (Thermo Fisher Scientific) and mixed with 40 µL of the luciferase reagent to measure luciferase luminescence in a CLARIOStar microplate reader (BMG LABTECH GmbH) with a measurement of 10 s and a delay of 2 s. Nanoluciferase luminescence was quantified thereafter by adding 40 µL of the Stop & Glo reagent and measured in the same way.

The luciferase/nanoluciferase ratio was determined for each sample, and normalized luminescence was calculated as the mean value of the ratios obtained from each duplicate. Statistical analyses were performed with GraphPad Prism 8.0.1 software. Differences between the strains were analyzed using one-way ANOVA, followed by post hoc multiple comparison Tukey’s test (p < 0.05). For induction experiments, we analyzed the differences in the growth of each strain in the presence of different carbon sources using Student’s t-test (p < 0.05).

P. digitatum parental and mutant strains were inoculated on freshly harvested oranges (Citrus sinensis L. Osbeck cv Lane late), as previously described (González-Candelas et al., 2010). Briefly, three replicates of five orange fruits were inoculated with 5 μL of fungal conidial suspension (10⁴ conidia/mL) at four equidistant wounds around the equator. Control mock inoculations were performed with 5 μL of sterile Milli-Q H₂O. Once inoculated, fruits were maintained at 20°C and 90% relative humidity for up to 6 days. Each inoculated wound was scored daily for infection symptoms on consecutive days post-inoculation (dpi). We repeated the experiments twice. Differences in the percentage of infection for each strain compared to the control CECT 20796 were analyzed using Student’s t-test (p < 0.05) for each individual dpi.

The FB platform already contains some commonly used positive fungal selection markers based on antibiotic resistance, such as hph (hygromycin^R, FB003) or nptII (geneticin^R, FB009) (Hernanz-Koers et al., 2018). However, in the case of integrative approaches, multiple genetic modifications often depend on the availability of different antibiotic resistance genes for transformant selection, which can be a bottleneck for the exploitation of filamentous fungi. In this study, we expand the range of selection markers available in the FB platform by including two alternative antibiotic resistance-inducing genes, the ble resistance gene from the bacterial transposon Tn5 and the squalene epoxidase ergA gene from P. chrysogenum. The expression of the ble gene provides the selection of the antibiotic phleomycin (Austin et al., 1990), whereas the expression of the ergA gene provides resistance against the antibiotic terbinafine in a broad range of filamentous fungi (Austin et al., 1990; Sigl et al., 2010).

In order to include ble resistance in the FB platform, a functional TU was generated. For this, we assembled the ble coding sequence (FB413), together with the promoter of isopenicillin N synthase (PpcbC) from Penicillium rubens (FB411) (Polli et al., 2016) and the terminator from the acetamidase (Tamds) from Aspergillus nidulans (FB416) (Kelly and Hynes, 1985) into the pDGB3α2 vector to obtain FB430 (Table 2; Figure 1A) via restriction–ligation reactions. To functionally validate the resulting construct, we transformed P. chrysogenum and P. digitatum wild-type strains with the same FB430 via ATMT for the ectopic integration of the ble TU. P. chrysogenum transformants grown in the presence of 25 μg/mL phleomycin were selected and analyzed by PCR for the presence of the ble cassette (Figure 1B). The positive transformants, PCEM43053 and PCEM43062, showed growth on phleomycin-containing plates when compared to the parental ATCC 10002 (Figure 1C), further demonstrating the functionality of FB430. In parallel, FB430 was also validated in P. digitatum (Figures 1D–G). P. digitatum transformants grown in the presence of 35 μg/mL phleomycin were selected and confirmed by PCR (Figure 1D). The positive transformants PDZS43023, PDZS43041, and PDZS43051 were able to grow on phleomycin-containing plates (Figure 1E) and showed the same pathogenicity as the parental CECT 20796 in orange fruits (Figures 1F,G).

Similarly, to incorporate the terbinafine resistance-inducing gene in the FB platform, a functional TU for ergA was generated and validated in P. chrysogenum (Figure 2). We assembled the ergA coding sequence (FB414), together with the PgpdA promoter (FB007) and the Tamds terminator (FB416), into the pDGB3α2 vector to obtain the FB431 construct (Table 2; Figure 2A). P. chrysogenum transformants grown on 0.5 μg/mL terbinafine were chosen and confirmed by PCR (Figure 2B). The positive transformants PDZS43122, PDZS43131, and PDZS43142 could grow on phleomycin-containing PDA plates in contrast to the parental ATCC 10002 (Figure 2C), demonstrating the functionality of FB431.

Overall, both resistances were transformed ectopically to avoid any bias regarding the targeting of specific loci, and these experiments validated the use of FB430 and FB431 as standardized TUs for conferring positive selection in the transformation of different fungal species, expanding the antibiotic resistance selection markers currently available in the FB system.

To date, no auxotrophic markers have been included in the FB platform despite the fact that they are sustainable alternatives for the use of antibiotics in transformant selection. The orotidine 5′-phosphate decarboxylase pyr4 gene from T. reesei is widely used as an auxotrophic selection marker that can be counter-selected using 5-FOA or fully supplemented using uridine (Díez et al., 1987; Derntl and Kiesenhofer, 2015). In this study, we set up several experiments to design, test, and validate pyr4 as a selection marker in pyr4/pyrG-deficient fungal strains. As a first step, uridine-auxotrophic P. digitatum ΔpyrG mutants were generated through ATMT using FB372 as the template for homologous recombination at the pyrG locus (Table 2; Supplementary Figure S4). Transformants were selected on PDA plates supplemented with 1.22 g/L uridine and 1.25 g/L 5-FOA, and were molecularly and phenotypically characterized (Supplementary Figure S4). Growth profiles showed that after pyrG deletion, P. digitatum mutants could no longer grow on PDA plates unless supplemented with uridine, confirming their auxotrophic condition. Additionally, these mutants could also grow in the presence of uridine and 5-FOA in contrast to the parental CECT 20796, further confirming pyrG deletion. Finally, infection assays on orange fruits revealed that P. digitatum ΔpyrG mutants showed highly reduced pathogenesis compared to the control (Supplementary Figure S4). Once the pyrG deletion mutants were obtained, a functional TU for the T. reesei pyr4 gene was generated. For this, we assembled the Penicillium codon-optimized and extensively domesticated pyr4 coding sequence (FB272), promoter (FB271), and terminator (FB276) into the pDGB3α2 vector to obtain FB293 (Table 2; Figure 3A) via restriction–ligation reactions. To functionally validate the resulting construct FB293, we transformed the P. digitatum ΔpyrG mutant PDSG37213 with FB293 via ATMT for the ectopic integration of pyr4 TU. Transformants grown on PDA plates were assessed by PCR (Figure 3B) and phenotypically analyzed to confirm pyrG:pyr4 complementation and, therefore, the absence of the auxotrophy. As shown in Figure 3C, complemented mutants PDSG29312, PDSG29321, and PDSG29333 were all able to grow on PDA plates without uridine, in contrast to the auxotrophic parental PDSG37213. Remarkably, P. digitatum complemented mutants that fully recovered their original pathogenicity (Figures 3D,E), which validates pyr4 as an auxotrophic selectable marker also for (phyto)pathogenic fungi, in which the deletion of pyrG orthologs has been demonstrated to decrease pathogenicity and virulence in the corresponding fungi (Zameitat et al., 2007; Higashimura et al., 2022).

In order to further expand and characterize the promoter catalog available in the FB platform, a series of promoters from different fungal species were included in the collection (Table 1) and were functionally validated using a luciferase reporter system, as previously described (Gandía et al., 2022). This reporter consists of two TUs: the nanoluciferase coding sequence (Nluc, FB310) under the regulation of the PgpdA promoter that serves as an internal standard for normalization, and the luciferase sequence (Luc, GB0096) under the regulation of the promoter to be tested. Promoter strength is expressed as the ratio of the Luc signal divided by the Nluc internal standard. The promoters to be evaluated included the previously characterized strong pyruvate kinase gene promoter (PpkiA) from Aspergillus niger (FB389) (de Graaff et al., 1992). Novel promoters from P. digitatum included the antifungal protein AfpB gene promoter PafpB (FB404) (Garrigues et al., 2017), and two promoters with high expression levels were reported in a previous transcriptomic study: the elongation factor 1α gene promoter (Pef1α) (FB407) and the ubiquitin ligase PDIG_07760 gene promoter (P07760) (FB408) (Ropero-Pérez et al., 2023). The inducible promoters included in this study are the endo-1,4-β-xylanase A gene promoter (PxlnA) from A. nidulans (FB291, xylose-responsive) (Orejas et al., 1999), the glucoamylase gene promoter (PglaA) from A. niger, and the TAKA-amylase A gene promoter (PamyB) from Aspergillus oryzae (FB405 and FB406, respectively, both maltose/starch-responsive) (Fowler et al., 1990; Tsuchiya et al., 1992b). The widely used glyceraldehyde-3-phosphate dehydrogenase promoter PgpdA from A. nidulans, which was already available in the FB collection (FB007) (Hernanz-Koers et al., 2018), was also included in the analysis, as well as the luciferase reporter construct for the P. chrysogenum antifungal protein PAF promoter (Ppaf) from previously published data (FB323) (Gandía et al., 2022) to serve as references. To facilitate the cloning of new luciferase reporter constructs, the nanoluciferase reference gene and the nptII resistance gene were cloned into a pDGB3α1 vector (FB367) to be combined in a single reaction with the luciferase TUs cloned into pDGB3α2 vectors that included the promoters to be tested (Figure 4A). Due to the requirements of the FB binary assembly, an insulator sequence (GB3458) was also included at the 3’ end of the FB367 vector, which also helps prevent the interaction between Nluc and Luc TUs. The luciferase reporter constructs for each of the assayed promoters (FB432 to FB441) showed different normalized luciferase expression levels in P. digitatum after 2 days of growth in PDB (Figure 4B). The lowest expression levels were observed for the inducible promoters PxlnA, PamyB, and PglaA, in this order, due to the lack of inducers in this medium. Their expression levels, together with that driven by PafpB, were slightly above the basal signal observed in the control strain, but these were not statistically significant. The expression driven by PpkiA, PgpdA, and the new Pef1α were similar to those observed for Ppaf, while P07760 showed intermediate expression values between these expression and the inducible promoters.

The selected constructs carrying PxlnA, PpkiA, PglaA, PamyB, Pef1α, and P07760 were also transformed into P. chrysogenum, in which the luciferase expression level was about 10 times lower than the overall levels observed in P. digitatum (Figure 4C), except for Pef1α and PamyB, which showed a similar signal in both fungal chassis (0.27 and 0.1 for Pef1α in P. digitatum and P. chrysogenum, respectively, and 0.0007 and 0.0004 for PamyB). Unlike P. digitatum, the signal driven by PamyB was above the basal signal in P. chrysogenum and showed similar expression levels to that of the P7760 promoter. Relative expression levels among the other promoters were nevertheless maintained in both fungi, with PpkiA and Pef1α signals similar to that of the Ppaf reference promoter and with PxlnA-, PglaA-, and PamyB-driven signals similar to that of the control strains.

In order to further characterize the inducible promoters included in this study, we analyzed the induction of luciferase expression directed by PglaA, PamyB, and PxlnA promoters in P. digitatum (Figure 5A) and P. chrysogenum strains (Figure 5B) after 4 days of growth in minimal medium (PdMM for P. digitatum and PcMM for P. chrysogenum) using different inducers as the sole carbon source (2% maltose for PglaA and PamyB, and 2% xylose for PxlnA). When the fungi were grown in the presence of the inducer, the expression levels driven by all three promoters were significantly higher than those observed in the reference media with 2% glucose, which increase by 8x, 4x, and 10x for PglaA, PamyB, and PxlnA, respectively, in P. digitatum (Figure 5A); and by 2x, 9x, and 8x in P. chrysogenum (Figure 5B). Signals observed for the PxlnA promoter in the reference media (MM + glucose) were similar to the basal signal of the reference strains in both fungi, while PamyB and PglaA promoters showed higher basal expression in the same media, especially in P. chrysogenum. Remarkably, expression levels in the reference Ppaf promoter were found to increase significantly in P. chrysogenum when grown in the PcMM supplemented with maltose or xylose (0.22 a.u. on average) when compared to the medium supplemented with glucose (0.015 a.u. on average, 15 times lower). The same was observed, but to a lesser extent, in P. digitatum when Ppaf was expressed in PdMM glucose (average values of 0.6 a.u.) compared to its expression in PdMM xylose (0.15 a.u. average values, four times higher).

Finally, we tested the recently developed GB_SynP (Moreno-Giménez et al., 2022) in our fungal chassis in combination with the pAMA18.0_gRNA1 plasmid, which delivers the CRISPRa system necessary to activate GB_SynP promoters in a non-integrative manner (Mózsik et al., 2021). To this end, we developed luciferase reporter constructs following the same procedure as for natural promoters (Figure 6A). In these constructs, luciferase expression was regulated by synthetic promoters consisting of an A1 distal promoter part formed by a random sequence (GB2815), and an A2 proximal promoter part including the target sequence for gRNA1 that was repeated once (GB2878), twice (GB2885), or three times (GB3276), and the minimal promoter mPAF (GB3423) derived from the native Ppaf from the fungus P. chrysogenum, which was previously found to drive a strong induction when exposed to the dCas9 system loaded with gRNA1 in plants (Moreno-Giménez et al., 2022). The resulting constructs containing one gRNA1 target (FB395, 1xLuc), two targets (FB396, 2xLuc), or three targets (FB397, 3xLuc) were stably transformed into P. digitatum via ATMT. Protoplasts obtained from these strains were re-transformed with the CRISPRa expression vector pAMA18.0_gRNA1 (Figure 6B). The expression levels of 1xLuc and 2xLuc constructs were not significantly higher than the basal signal of the control strain despite the presence of pAMA18.0_gRNA1, except for one of the 1xLuc re-transformants, which showed a low, but statistically significant increase in the luciferase expression when compared to the same strain in the absence of the CRISPRa system. A higher and significant increase of approximately 16 times on average in the luciferase signal was observed for the 3xLuc construct (Figure 6C) in all tested re-transformants. Although the expression driven by this promoter was approximately 10 times lower than that observed with the reference Ppaf promoter, it is comparable to that achieved with PglaA and PamyB , showing the functionality of these dCas9-activated synthetic promoters in fungi.