Differential Expression of Cell Wall Remodeling Genes Is Part of the Dynamic Phase-Specific Transcriptional Program of Conidial Germination of Trichoderma asperelloides

Abstract

The nature of saprophytic and mycoparasitic hyphal growth of Trichoderma spp. has been studied extensively, yet its initiation via conidial germination in this genus is less well understood. Using near-synchronous germinating cultures of Trichoderma asperelloides, we followed the morphological progression from dormant conidia to initial polar growth to germling formation and to evidence for first branching. We found that the stage-specific transcriptional profile of T. asperelloides is one of the most dynamic described to date: transcript abundance of over 5000 genes-comprising approximately half of the annotated genome-was unremittingly reduced in the transition from dormancy to polar growth. Conversely, after the onset of germination, the transcript abundance of approximately a quarter of the genome was unremittingly elevated during the transition from elongation to initial branching. These changes are a testimony to the substantial developmental events that accompany germination. Bayesian network analysis identified several chitinase- and glucanase-encoding genes as active transcriptional hubs during germination. Furthermore, the expression of specific members of the chitin synthase and glucan elongase families was significantly increased during germination in the presence of Rhizoctonia solani-a known host of the mycoparasite-indicating that host recognition can occur during the early stages of mycoparasite development.

Keywords: Congo Red; cell wall remodeling; chitin synthase; chitinase; conidial germination; glucan elongase; glucanase; mycoparasite.

Figure 1

Phenotypic characterization of conidia germinating on PDA. (A) Conidia germinated in PDA were imaged at 1, 5, 7, 10, and 12 h; scale bar: 100 µm for all images). (B) The degree of synchronous progression of germination was monitored and cells were assigned as either dormant conidia, polar growth, doubling of long axis, or first hyphal branching. (C) Average diameters of conidia were assessed during the early stages of germination; whiskers indicate the standard deviation of the mean for the three biological replicates.

Figure 2

Summary of the number of genes significantly (A) down- or (B) up- regulated across stages of germination (Stage 1→Stage 2: dormant conidia to polar growth; Stage 2→Stage 3: polar growth to germinated conidia; and Stage 3→Stage 4: germinated conidia to first branching).

Figure 3

Selected 20 GO terms among the cellular components showing divergent enrichment patterns during phases of conidial germination of T. asperelloides. (A) Differentially expressed genes (DEGs) that were significantly downregulated during the process; (B) DEGs that were significantly upregulated during the process. Enrichment status is color-coded for different stage shifts. The bubble plot was scaled with the percentage of observed DEG for each GO term. Significantly enriched GO terms (Benjamini–Hochberg adjusted p < 0.05) are circled with red.

Figure 4

Relative gene expression and regulatory interactions of chitinases. (A) Log-scaled relative gene expression levels of chitinases across germination of Trichoderma asperelloides. Expression levels are expressed relative to the lowest (set as 1) among genes across the process. (B) Bayesian gene expression regulatory network was inferred from relative gene expression measurements for select chitinase-encoding genes.

Figure 5

(A) Log-scaled relative expression levels of glucanases across conidial germination of Trichoderma asperelloides. Only glucanases with a minimal r value of 1 are presented. (B) Bayesian gene expression network, inferred from relative expression among select glucanase-encoding genes.

Figure 6

Log relative gene-expression levels of chitin synthases across the four-stage germination time course of Trichoderma asperelloides. Gene expression was determined by RT-PCR. Values shown are based on three biological repeats, with whiskers signifying one standard error. Gene expression was normalized to tef1. The threshold cycle (2^−ΔΔCT) method was used to determine fold changes in expression.

Figure 7

Log relative gene-expression levels of Trichoderma asperelloides glucan elongases across four stages of conidial germination. Gene expression was determined by RT-PCR. Values shown are based on three biological repeats, with whiskers signifying one standard error. Gene expression was normalized to tef1. The threshold cycle (2^−ΔΔCT) method was used to determine fold changes in expression.

Figure 8

Germination of T203 conidia imaged at four time points (white numbers). Germination in (A) 0.5 mM Congo Red and (B) PDB. Scale bar: 100 µm for all images.

Figure 9

Expression levels calculated as of gel3 during germination in PDB and in the presence of 500 µM Congo Red. Gene expression was determined by RT-PCR. Values shown are based on three biological repeats, with whiskers signifying one standard error. Gene expression was normalized to tef1. The threshold cycle (2^−ΔΔCT) method was used to determine fold changes in expression. Standard errors (whiskers) were based on three biological replicates.

Figure 10

Phenotypic characterization of conidia germinating on R. solani host. (A) Conidia germinated on R. solani were imaged at different time points, as indicated. Scale bar: 100 µm for all images. (B) Conidial germination as a percentage over time, determined by live-cell imaging. (C) Average germling length, quantified by n = 50 randomly selected conidia for each time point. Error bars describe standard deviations of the mean for three biological replicates.

Figure 11

Expression levels of chs1–chs7 during germination (A) on solid medium or (B) on the host Rhizoctonia solani. Expression levels of gel1–gel4 genes during germination (C) on solid medium or (D) on the host R. solani. Gene expression was determined by RT-PCR. Values shown are based on three biological repeats, with whiskers signifying one standard error. Gene expression was normalized to tef1. The threshold cycle (2^−ΔΔCT) method was used to determine fold changes in expression.