Fig 1.
Design of deletion constructs and U. maydis insertional mutants.
(a) Plasmid backbones containing a Spec and an ARS were combined with an hpt resistance cassette and specific borders (LB & RB) via Golden Gate Cloning [23]. The hpt resistance cassette is flanked by UPSs (magenta arrows). (b) Plasmids were linearized with AscI and combined with haploid SG200 protoplasts. Transformants were selected on plates supplemented with hygromycin. (c) Schematic overview of PCR verification of transformants. Three independent fungal transformants were verified for each mutant locus via PCR. PCR products from primer-pair A targeting insertional mutant X was absent in positive transformants and detectable in SG200 control strains. A control primer-pair B gave a product in both insertional mutant X and SG200. ARS, autonomous replication sequence; hpt, hygromycin phosphotransferase; LB, left border; RB, right border; Spec, Spectinomycin resistance cassette; UPS, unique primer binding site.
Fig 2.
iPool-Seq library preparation workflow features tagmentation and UMIs.
(a) Library preparation was carried out for the input mutant collection and for the output after infection. For the output, we harvested infected areas of the second and third maize leaves and isolated gDNA. (b) Extracted gDNA was fragmented with Tn5 Transposase loaded with custom adapters containing an SBS (green), 12-bp UMI, and Tn5 hyperactive MEs (blue). Genome–hpt resistance cassette junctions were PCR-amplified with biotinylated primers directed against UPSs (magenta) and adapter-specific primers directed at the SBS. (c) Biotinylated PCR products were streptavidin-affinity–purified and Illumina-compatible P5 (purple; NGS1) and P7 (purple; NGS2) ends were introduced by nested PCR. Final products were subjected to Illumina PE sequencing on a MiSeq platform. gDNA, genomic DNA; hpt, hygromycin phosphotransferase; iPool-Seq, insertion Pool-Sequencing ME, mosaic end; PE, paired-end; ROI, region of interest; SBS, sequencing primer binding site; UMI, unique molecular identifier; UPS, unique primer binding site.
Fig 3.
Quality control of iPool-Seq library.
(a) Bioinformatic workflow of iPool-Seq analysis. Input and output read percentage after validation, mapping, and UMI analysis shows the mean ± SEM of 3 biological replicates and 2 independent infections. (b) Distribution of reads per individual UMI (bars) and model prediction (dots) over all insertional mutants of 1 representative replicate for input and output. Here, the error-correction threshold was set to 1 for the input and 5 for the output. Predicted true and lost UMIs are indicated. (c) Correlation plot of UMI counts for 5′- and 3′- genomic junctions of the hpt resistance cassette. One representative replicate of input and output is depicted. Each circle represents an insertional mutant. Missing up or downstream reads are marked with x. hpt, hygromycin phosphotransferase; iPool-Seq, insertion Pool-Sequencing; M, mean number of reads per UMI in the predicted distribution; R, correlation value; T, threshold; UMI, unique molecular identifier.
Fig 4.
iPool-Seq identifies significantly depleted mutants after pooled infection.
(a) Log2-fold changes between normalized output abundances and internal reference set for mutants with known phenotypes. p-Values were calculated with Mann–Whitney U tests. p = 5e−9 for neutral versus reduced and p = 3e−4 for reduced versus lost virulence with ***p < 0.001; **** p < 0.0001 (S3 Table). (b) Log2-fold change of output over input abundances for 1 representative replicate. Each circle represents 1 insertion mutant. Internal references are marked in green, significantly depleted in red (tested against reference set using negative binomial test; S1 Data; S1 Supporting methods), unaffected mutants in gray; Insig. area is also highlighted in gray. (c) Heatmap of log2-fold changes of input normalized UMI counts of all insertional mutants sorted by mean level of abundance. Infection A and B are two independent experiments and 1, 2, and 3 are three biological replicates, which were clustered according to similarity. Mutants without detectable reads in output libraries are displayed in black (S1 Data; S1 Supporting methods). FDR, false discovery rate; Insig., insignificant; iPool-Seq, insertion Pool-Sequencing; Sig., significant; UMI, unique molecular identifier.
Fig 5.
Virulence factor mutants identified by iPool-Seq cause reduced disease symptoms on maize.
(a) Disease rating of insertional mutant strains 7 dpi. Mean standard deviation of relative counts from 3 replicates are displayed. Only positive error bars are shown. p-Values were calculated by Fishers exact test. Multiple testing correction was done by Benjamini-Hochberg algorithm. **** p < 0.0001. (S2 Data) (b) Growth assay of insertion mutants on (A) Cm-medium, or Cm-medium supplemented with (B) 75 μg/mL Calcofluor (cell wall stress), (C) 45 μg/mL Congo red (cell wall stress), and (D) Charcoal (b-filament inducing). (c) Confocal microscopy of maize infected with indicated insertional mutant strains 7 dpi. Infected plant tissue was stained with propidium iodide (red) and fungal hyphae with lectin binding WGA-AF488 (green). One representative picture of 9 infected plants is shown. Cm-medium, control Complete medium; dpi, days post infection; iPool-Seq, insertion Pool-Sequencing; ref, reference; wt, wild-type.