Lasergene v15

Sequences were assembled using Lasergene v15 (DNASTAR, Inc., Madison, WI, USA), and combined with sequences obtained from Genebank (Supplementary Materials Table S1). For the Beauveria strains, we used a backbone tree from Imoulan et al. [48 (link)], and for the wider Cordycypitaceae phylogeny we used a backbone tree from Zhou et al. [49 (link)] with sequences of Paecilomyces hepiali Q.T. Chen and R.Q. Dai ex R.Q. Dai et al. (Cordycipitaceae, Hypocreales) added. Multiple sequence alignments were conducted using MAFFT v7.271 [50 (link)] and trimmed with trimAI 1.2rev59 [51 (link)] or, for protein coding sequences, Gblocks 0.91b [52 (link)]. The trimmed alignments were partitioned according to molecular markers, and, in case of protein-coding loci, the codon position was also partitioned. Maximum likelihood trees were generated with RAxML-NG v. 0.8.0 [53 (link)].

The NtFT5 genomic locus including ~2.74 kb of the promoter (P_NtFT5) was amplified from N. tabacum cv. SR1 genomic DNA in five overlapping fragments. The parts were separately amplified by PCR using gene-specific primers (Supplementary Table S1). The primer sequences were designed in silico based on the NtFT5 locus previously identified in the published N. tabacum cv. Basma Xanthi genome (Sierro et al., 2014 (link); Beinecke et al., 2018 (link)). The resulting PCR products were adenylated using MangoTaq DNA polymerase (Bioline, London, UK), transferred to pCRII-TOPO using the TOPO TA Cloning kit (Thermo Fisher Scientific) and sequenced. The full-length genomic sequence of NtFT5 was then assembled in silico using SeqManPro and SeqBuilder Pro in Lasergene v15 (DNASTAR, Madison, WI, USA). The same software was used to determine the gene structure by aligning the genomic clone with the previously-described coding sequence: GenBank KY306470.1 (Beinecke et al., 2018 (link)).