Analysis of the pcz1 gene and deduced protein from P. rubens Wis 54-1255
The sequence of the pcz1 gene from P. roqueforti (Gil-Duran et al. 2015) was used to scan the genome of P. rubens Wis 54-1255 by BLASTN and BLASTX. Both analyses were redundant and yielded as a result that the pcz1 gene corresponds to the gene Pc22g12400 in P. rubens Wis 54-1255 genome.
The pcz1 gene from P rubens Wis 54-1255 is 2,446 bp long and contains a single intron of 70 bp located at the 5´ end of the gene (Fig. 1). This gene encodes a protein of 791 amino acids. Analyses of conserved domains were done by HMMER tool at Interpro, and CDD at NCBI. Both analyses indicated the presence of a Zn(II)2Cys6 fungal-type domain between positions 394–431 of Pcz1, which includes the six cysteines conserved expected in this kind of domains (Fig. 1).
Penicillium rubens Wis 54-1255.
BlastP analysis of Pcz1 indicated that closer orthologues of this gene can be found in many fungal species from the phylum Ascomycota, mainly from the genus Penicillium. The evolutionary relationships of Pcz1 from P. rubens Wis 54-1255 and its closer orthologous proteins can be seen in Fig. 2. This analysis showed that Pcz1 clustered with orthologues within section Chrysogena in Penicillium, in agreement with the described species phylogeny (Houbraken et al. 2020).
Obtainment of genetically modified strains of P. rubens Wis 54-1255 by CRISPR-Cas9
After transformation of P. rubens Wis 54-1255 with plasmid pFC333-Pcpcz1, thirty-five transformants were obtained. They were submitted to passages in Czapek media containing phleomycin, and subsequently, in Czapek media lacking phleomycin. Ten transformants lost phleomycin resistance, and three of them were randomly chosen and their target region was sequenced. All of them showed large insertions of DNA fragments from plasmid pFC333-Pcpcz1 in the target sequence. Specifically, transformant T1 showed a 51-bp insertion which belonged to a fragment of the gene encoding Cas9 protein (Fig. 3). Transformant T16 has a 27-bp insertion from the AMA region (Fig. 3). Finally, transformant T6 showed a very large insertion of 336 bp, which includes different fragments of AMA region (Fig. 3).
The analysis of the deduced protein of Pcz1 from each transformant was done. Interestingly, and despite the large insertions in their target regions, transformants T1 and T16 do not lose frameshift (Fig. 3), so the deduced proteins from these transformants are essentially the same as native Pcz1, except that they are a few larger (Fig. 3). This suggests that despite the insertions, these transformants would contain a functional Pcz1 protein. On the contrary, the large insertion in transformant T6 produces multiple stop codons in pcz1 gene, suggesting that in this transformant, Pcz1 was inactivated (Fig. 3). Considering these results, we selected transformants T6 (inactive), T1 (with insertion but functional), and the wild-type strain for further comparative experiments.
The inactivation of pcz1 reduces growth and conidiation, but promotes conidial germination in P. rubens Wis 54-1255
Apical growth of transformants was analyzed in five different media. As expected, growth of transformant T1 was indistinguishable from wild-type strain of P. rubens. On the contrary, transformant T6, where pcz1 was inactivated, showed a slight delay in apical growth (Fig. 4). Thereby, and depending on the media used, transformant T6 showed a growth rate between 77% and 86% of the wild-type fungus, suggesting that pcz1 is a positive regulator of growth in P. rubens Wis 54-1255.
Concerning conidiation, the tendency observed was the same described for apical growth. Conidiation was measured in two media, the minimal medium Czapek, and rich medium Power, which was designed for enhancing conidiation (Fierro et al. 1996). Transformant T1 showed a similar conidiation pattern that wild-type strain, whereas transformant T6 with inactivated pcz1 gene showed an important decrease in conidia production (Fig. 5). Thus, and depending on the media and day of measurement, transformant T6 produced between 45% and 64% of conidia produced by the wild-type fungus. These results indicate that pcz1 is a positive regulator of conidiation in P. rubens Wis 54-1255.
Finally, conidial germination was measured. As shown in Fig. 6, transformant T6 showed an earlier germination of conidia as compared with the wild-type strain. For example, after 7, 8 and 9 hours in CM medium, transformant T6 showed 56%, 80% and 90% of conidial germinated, respectively, whereas at the same times and the same medium, P. rubens Wis 54-1255 showed significant lower values of 35%, 56% and 69% of conidia germinated. The same tendency can be observed in minimal medium Czapek, although in this case differences are less pronounced. These results suggest that pcz1 has a negative regulator role in conidial germination in P. rubens Wis 54-1255.
The inactivation of pcz1 reduces the production of penicillin in P. rubens Wis 54-1255
Penicillin is the most important secondary metabolite produced by P. rubens, so we studied its production in transformants selected. As shown in Fig. 7, the production of benzylpenicillin is drastically depleted in transformant T6, whose pcz1 gene was inactivated by CRISPR-Cas9. In average, transformant T6 produces 33 µg/mg of penicillin at 72 hrs, and 91 µg/mg of the compound at 96 hrs, whereas the wild-type strain produces 585 µg/mg and 855 µg/mg of penicillin at 72 and 96 hrs, respectively (Fig. 7). As expected, transformant T1, which has an insertion but does not lose the reading frame, produces similar amounts of benzylpenicillin than wild-type strain (Fig. 7). All together, these results indicate that the inactivation of pcz1 decreases the production of benzylpenicillin, suggesting that this gene exerts a positive regulation on the production of this important secondary metabolite in P. rubens Wis 54-1255.