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

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.


Introduction
Spores are reproductive units of prokaryotic and eukaryotic organisms, including algae and protozoa, lower vascular plants, and even a subset of animals, produced to propagate their genetic material [1]. Many members of the fungal kingdom can produce sexual and/or asexual spores in an astonishing variety of shapes, sizes, and other features adapted to specific lifestyles and the requirements for dispersal or persistence in different environments [2]. The fungal asexual spore-or conidium-is a nonmotile, walled, predominantly-haploid cell, usually generated by mitosis from a parent cell. Asexual spore production is a critical stage in the life cycle of fungi, serving as the primary means for dispersion, infection, and survival under adverse conditions [1,3].
Conidial germination is composed of a sequence of events that convert the resting propagule into a rapidly growing germ tube from which the mycelium will be formed by subsequent elongation and branching [4]. It differs from later hyphal growth in the marked dependence of each conidium on its stored reserves for metabolism. Physiologically, it is characterized by transformation of a conidium from a stage of low to one of high metabolic point of mycoparasite contact occurs at the cell wall of the host, which modulates fungal interactions [23]. The cell wall is an essential component of the fungal cell, providing strength, structure, and protection against environmental stress [29,30]. The conidial cell wall is composed of polysaccharides, proteins, lipids, and pigments. Predominant components include glucans, chitin, chitosan, and glycosylated proteins. Glucan and chitin are synthesized by plasma-membrane-associated glucan synthases and chitin synthases, respectively, while cell wall glycoproteins are synthesized by ER-associated ribosomes. In concert with the biosynthetic protein machinery, an array of glycosyl hydrolases participate in shaping and remodeling the fungal cell wall [31].
During the transition from isotropic growth to polarized growth, and then to germtube formation and elongation, the conidial cell wall undergoes significant modifications [9,32]. To perform these modifications, most fungal genomes encode multiple chitin synthases. Some chitin synthases have been demonstrated to function in specific cell types or during restricted phases of the cell cycle [32]. This multiplicity and specificity is exemplified by the eight chitin synthases of T. atroviride, which have been identified and demonstrated to be differentially expressed during germination and hyphal elongation, as well as differentially affected by the intercalating dye Congo Red [33]. The presence of multiple glycosyl hydrolases-many of them conserved-is also common in fungal species. Members of glycosyl hydrolase family 72 are among the best-characterized cell wall remodeling enzymes. These GPI-anchored enzymes execute an essential role in cell wall biogenesis; for example, β-(1,3)-glucanosyltranferases (GELs) are a major functional component of β-(1,3)-glucan processing [34][35][36][37], and their involvement in conidial germination has also designated some of them as potential targets for antifungals [38].
One of the intensively studied Trichoderma spp. strains is T. asperelloides T203. This strain was formerly identified as T. asperellum, and prior to that as T. harzianum, and is a cryptic sister species to T. asperellum [39]. These two species belong to the section Trichoderma, which contains several other prominent mycoparasitic species, such as T. atroviride, T. gamsii, and T. hamatum. The host range of T. asperelloides T203, along with traits involved in host recognition and some of the interactions between this mycoparasite and its hosts, has been analyzed [40,41]. These studies were largely devoted to characterizing the activity of enzymes involved in cell wall degradation/remodeling. Importantly, the potential to use T. asperelloides T203 for biocontrol has been extensively tested in the greenhouse, as well as in the field [42].
In this study, we analyzed the transcriptional profile of the recently sequenced T. asperelloides strain T203 [43] (whose genetic attributes are summarized at https://mycocosm. jgi.doe.gov/vista_embed/?organism=Triasper1 (accessed on 5 July 2022). Moreover, we assessed the degree of dynamism of transcript abundance during germination of this species. Examining the changes occurring in genes encoding cell wall remodeling proteins, we elucidated how the transcription of at least some members of the chitin synthase and glucan elongase families are affected by the presence of R. solani, a potential host of the mycoparasite.

Strains and Culturing Conditions
Trichoderma asperelloides strain T203 [43] and Rhizoctonia solani TP6 [44] were cultured on Potato Dextrose agar or broth (PDA or PDB, respectively. Difco Laboratories, Detroit, MI, USA) at 28 • C. Conidia of T. asperelloides strain T203 were harvested from 5-7-day-old cultures. Conidia were harvested from three 90 mm diameter Petri dishes by adding 5 mL of cold sterile double deionized water (DDW) to each dish. Conidia were released by brief shaking with glass beads and collected by filtration through a gauze pad layered over a funnel. The conidial suspension was centrifuged at 4000 rpm for 5 min, and the conidial pellet was washed twice with 30 mL cold DDW. Conidial suspensions were incubated overnight at 4 • C and then conidial density was calculated using a hemocytometer and adjusted to a concentration of 5 × 10 6 conidia/mL. To facilitate the harvesting of conidia and germlings cultured on PDA, presterilized 40-µm pore cellulose membranes (Life Technologies, Carlsbad, CA, USA) were placed over the solid medium prior to spreading the inoculum. Plates were inoculated by spreading 50 µL of the adjusted conidial suspension and incubated at 28 • C for the required periods. For growth in PDB, conidial suspensions were first rehydrated as described above. The PDB medium (50 mL) was inoculated to a final concentration of 5 × 10 6 conidia/mL. Cultures were incubated at 28 • C, at 160 rpm, for the required time. When necessary, Congo Red (Sigma, St. Louis, MO, USA),) was added to PDB medium at the desired concentration prior to autoclaving. Cultures were sampled every hour and examined by light microscopy (EVOS FL Auto Cell Imaging System-Life Technologies, Carlsbad, CA, USA) for timing fresh conidia, polar growth, doubling of the long axis, and first hyphal branching. The timing experiment was repeated three times.
To analyze germination in the presence of a potential host of the mycoparasite, R. solani was first cultured on PDA for 5-7 days. R. solani plates were overlaid with sterile cellulose membranes, inoculated with 50 µL of adjusted conidial suspensions of T. asperelloides, and incubated at 28 • C. Cultures were sampled after a range of intervals (within 0-48 h) and examined by light microscopy.

Isolation of Nucleic Acids, RT-PCR, and RNASeq
RNA was isolated from T. asperelloides cultures after freezing in liquid nitrogen and grinding to a powder using a mortar and pestle, using the TRIzol reagent and the 1-Bromo-3-chloropropane (BCP, Sigma Aldrich) method, as previously described [45]. RNA samples were further purified with the RNA Clean and ConcentratorTM-5 kit (Zymo Research, Irvine, CA, USA) and treated with DNase (RNase-Free DNase Set, Qiagen, Hilden, Germany) according to the manufacturer's instructions. RNA yield and quality were assessed using the Synergy™ HTX Multi-Mode Reader (Biotek, Winooski, VT, USA) and verified using gel electrophoresis.
For RT-PCR, RNA was reverse transcribed using the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems, Waltham, MA, USA). Transcript abundance was determined using a Step One Plus RealTime PCR (Applied Biosystems, Waltham, MA, USA) apparatus using gene-specific primers (Table 1) under standard reaction conditions as previously described [46]. To estimate the relative amounts of transcripts in each sample, we calculated 2 −∆∆CT , where ∆∆CT is the ∆CT of the test sample and −∆CT of the calibrator sample, and ∆CT is the difference between the cycle threshold (CT) of the gene of interest and the reference gene. Translation elongation factor 1α (tef1α) was used as a reference gene. To prepare cDNA samples for RNAseq, mRNA was purified from approximately 200 ng of total RNA with oligo-dT beads and sheared by incubation at 94 • C in the presence of Mg (Roche Kapa mRNA Hyper Prep Catalog # KR1352). Further, first-strand cDNAtailing was performed with dUTP to generate strand-specific sequencing libraries. Indexed libraries were quantified by qRT-PCR using a commercially available kit (Roche KAPA Biosystems Cat # KK4854). The quality of cDNA samples was verified with a bioanalyzer (Agilent Technologies 2100; Agilent Technologies, Santa Clara, CA, USA). In total, 12 libraries were constructed from the four developmental stages, each with 3 biological replicates. Ten libraries passed initial quality control and were sequenced. Basic statistics, such as total reads sequenced and mapping rates, are summarized in Table S1. The cDNA samples were sequenced at the Yale Center for Genomics Analysis (YCGA). The libraries underwent 101-bp single-end sequencing using an Illumina HiSeq 2500 in Rapid Run Mode according to Illumina protocols. Adapter sequences, empty reads, and low-quality sequences were removed. Reads were trimmed using fastp v0.21.0 [47] using default parameters. Trimmed reads were aligned to the Trichoderma asperelloides T203 v1.0 genome from GenBank [43] using HISAT2 v2.1, indicating that reads correspond to the reverse complement of the transcripts and reporting alignments tailored for transcript assemblers. Alignments with a quality score below 20 were excluded from further analysis. Reads were counted for each gene with StringTie v1.3.3 and the Python script prepDE.py provided in the package. StringTie was limited to report reads that matched the reference annotation. Transcriptomics data and experiment details were made available (GSE207066) at the GEO database (https://www.ncbi.nlm.nih.gov/geo/ (accessed on 5 July 2022). Genome-wide gene expression was profiled across all four sampled conidial germination stages using De-seq2 [48] LOX v1.6 [49]. Differentially expressed genes (DEGs), including downregulated and upregulated genes, were identified between stages based on the LOX measurement, and a gene expression difference with a LOXP value < 0.05 was deemed statistically significant. Functional enrichment analysis for DEGs significantly up-or downregulated during the conidial germination was based on the GO term annotation of the genome T. asperelloides using T. asperellum as a key reference. Significant enrichment of GO terms was identified using Fisher's Exact Test, adjusted with the Benjamini-Hochberg procedure [50] to p < 0.05.

Bayesian Network Prediction for Functional Genes Groups during Conidial Germination
Functional annotation was based on the JGI fungal genome database. Additional functional annotations were obtained by comparing T. asperelloides with the extensively studied and annotated genome of T. asperellum. T. asperelloides homologs were identified using OrthoLoger [51]. Bayesian gene networks were generated to reveal gene interactions, associations, and regulatory modules involved in conidial germination in T. asperelloides. Biological networks were modeled for selected genes of interest using the Bayesian Network Web Server [52]. Input files consisted of fold changes between adjacent sample points across the four time points of the experiment. Fold changes between stages were calculated from LOX measurements of conidial germination on PDA using change in expression over the minimum expression between the stages, as previously described [53,54]. Global structure-learning settings were retained at default settings. The models depicted were the 50% majority consensus of 100 models (selection threshold set to 0.5; the 100 highestscoring networks were averaged), without imposing any structural constraints. Edges with posterior probability higher than 0.5, based on the consensus of 100 models, were present in the figures.

Phenotypic Characteristics of Germinating T. asperelloides Strain T203
Germination of rehydrated conidia of T. asperelloides on PDA overlaid with a cellophane membrane was monitored over time. Conidial swelling-characterized by an isotropic increase in size-continued for five to seven hours in a near-synchronous manner ( Figure 1A,B). During the course of this transition between resting and fully swollen conidia, the average conidium diameter increased from 2.4 µm to 4.75 µm ( Figure 1C). Once fully swollen, germ tubes emerged from the conidia. By nine to twelve hours, most conidia (90%) produced hyphae that exceeded the diameter of the conidium in length. The first branching of the hyphae was observed after twelve hours.

Changes in Gene Expression during Germination of T. asperelloides Strain T203
We used RNA sequencing to determine the extent of changes in transcript abundance during the four phases of conidial germination. Among the levels of expression of a total of 11,166 genes that were measured ( Figure 2; Table S2), 7932 genes exhibited significant downregulation (p < 0.05; exhibiting lower transcript abundance) and 2152 genes were significantly "upregulated" (exhibiting higher transcript abundance) from stage one to stage two (dormant conidia to polar growth). From stage two to stage three (polar growth to germinated conidia), 3184 genes exhibited significant downregulation and 2174 genes exhibited significant upregulation. From stage three to stage four (germinated conidia to first branching), 2182 genes exhibited significant downregulation and 4305 genes exhibited significant upregulation.
Among the 2152 genes whose expression was upregulated from dormant conidia to commencement of polar growth, only 1214 genes were stage-specific to this transition ( Figure 2). Similarly, 862 genes were stage-specifically upregulated in the transition from polar growth to elongation of the germ tube, and 2961 genes were stage-specifically upregulated during elongation to the first branching of the germlings. As germination progressed, a larger number of genes were upregulated than downregulated: 2961 genes were upregulated between elongation of the germ tube and first branching compared to 534 genes that were downregulated in this interval.
to germinated conidia), 3184 genes exhibited significant downregulation and 2174 g exhibited significant upregulation. From stage three to stage four (germinated conid first branching), 2182 genes exhibited significant downregulation and 4305 genes ex ited significant upregulation.
Among the 2152 genes whose expression was upregulated from dormant conid commencement of polar growth, only 1214 genes were stage-specific to this trans ( Figure 2). Similarly, 862 genes were stage-specifically upregulated in the transition f polar growth to elongation of the germ tube, and 2961 genes were stage-specifically regulated during elongation to the first branching of the germlings. As germination gressed, a larger number of genes were upregulated than downregulated: 2961 genes w upregulated between elongation of the germ tube and first branching compared to genes that were downregulated in this interval. Based on their gene ontology (GO), 5313 of the genes that were well measured w attributed as components of biological processes, 2537 as cellular components, and as exhibiting molecular function. The majority of the genes expressed during germina were classified to transcription factor activity (411 genes), metabolic process (642 gen and nucleic-acid binding (331 genes), all of which are apparently required for proper mination of this strain of T. asperelloides.

Changes in Cellular Component Gene Expression Pattern during Conidial Germ nation
Genes belonging to the cellular-component functional group exhibited both ("upregulated genes") and low abundance of transcript ("downregulated gen throughout germination ( Figure 3). Many functional categories-for example, nucle cytoplasm-, and membrane-related genes-included genes that were both upregul and downregulated in serial stages of germination. Some functions appeared to be st specific as transcript abundance was present at certain germination stages and was detected in other stages (Table S3). Others, such as roles in the septin, proteasome c and ribonucleoprotein complexes, appeared to be predominantly "upregulated" ac stages of germination. In other cases-such as cytoskeletal genes-we determined Based on their gene ontology (GO), 5313 of the genes that were well measured were attributed as components of biological processes, 2537 as cellular components, and 4633 as exhibiting molecular function. The majority of the genes expressed during germination were classified to transcription factor activity (411 genes), metabolic process (642 genes), and nucleic-acid binding (331 genes), all of which are apparently required for proper germination of this strain of T. asperelloides.

Changes in Cellular Component Gene Expression Pattern during Conidial Germination
Genes belonging to the cellular-component functional group exhibited both high ("upregulated genes") and low abundance of transcript ("downregulated genes") throughout germination ( Figure 3). Many functional categories-for example, nucleus-, cytoplasm-, and membrane-related genes-included genes that were both upregulated and downregulated in serial stages of germination. Some functions appeared to be stage-specific as transcript abundance was present at certain germination stages and was not detected in other stages (Table S3). Others, such as roles in the septin, proteasome core, and ribonucleoprotein complexes, appeared to be predominantly "upregulated" across stages of germination. In other cases-such as cytoskeletal genes-we determined that downregulation of some genes occurred from stage 1 to 2, and then upregulation occurred from stage 2 to 3 and from 3 to 4.
It was apparent that some cell-wall-related genes were "downregulated" from stage 1 to stage 2, yet others were "upregulated" throughout all stages of germination. Cell-wallrelated genes and genes expressed to the outside of the plasma membrane were examples of functional subgroups that exhibited this pattern of expression, which implies the occurrence of de novo synthesis of the relevant components from the initial establishment of polarity through the entire germination process until first branching. from stage 2 to 3 and from 3 to 4.
It was apparent that some cell-wall-related genes were "downregulated" from stage 1 to stage 2, yet others were "upregulated" throughout all stages of germination. Cellwall-related genes and genes expressed to the outside of the plasma membrane were examples of functional subgroups that exhibited this pattern of expression, which implies the occurrence of de novo synthesis of the relevant components from the initial establishment of polarity through the entire germination process until first branching.

Expression of Genes Associated with Cell Wall Remodeling during Germination
Differential expression of genes associated with cell wall remodeling was found to occur during germination of T. asperelloides. Four gene families associated with cell wall remodeling and mycoparasitism were chosen as candidates for possible germination markers: chitinases and glucanases that participate in the degradation of the cell wall (self and host), as well as chitin synthases and glucan-elongases that function in the synthesis processes of cell wall remodeling. Among the 31 chitinase genes identified in the course of this study, only the transcripts of 17 (Tables S2 and S4) were detected during germination. Nine of these chitinases were differentially expressed, to different extents, throughout the germination process, and eight were classified as early genes that were detected only in dormant conidia ( Figure 4). Gene 3261 appears to be the most prominent "late

Expression of Genes Associated with Cell Wall Remodeling during Germination
Differential expression of genes associated with cell wall remodeling was found to occur during germination of T. asperelloides. Four gene families associated with cell wall remodeling and mycoparasitism were chosen as candidates for possible germination markers: chitinases and glucanases that participate in the degradation of the cell wall (self and host), as well as chitin synthases and glucan-elongases that function in the synthesis processes of cell wall remodeling. Among the 31 chitinase genes identified in the course of this study, only the transcripts of 17 (Tables S2 and S4) were detected during germination. Nine of these chitinases were differentially expressed, to different extents, throughout the germination process, and eight were classified as early genes that were detected only in dormant conidia (Figure 4). Gene 3261 appears to be the most prominent "late germination gene"; its transcript levels were highest following the onset of first branching. The presence of an extensive co-regulatory interaction between some chitinase-encoding genes was supported by Bayesian networks inferred from the transcriptional data, highlighting an upstream "core" of four chitinase genes (10,302, 4268, 6105, and 3261). germination gene"; its transcript levels were highest following the onset of first branching. The presence of an extensive co-regulatory interaction between some chitinase-encoding genes was supported by Bayesian networks inferred from the transcriptional data, highlighting an upstream "core" of four chitinase genes (10,302, 4268, 6105, and 3261). Similar to that observed in the case of the chitinases, we were able to detect the presence of about half of the glucanases along the timeline of germination ( Figure 5A; Tables  S2 and S4). Nonetheless, many of the glucanase transcripts were present mainly in dormant conidia, while three of them (28,3949, and 1833) appeared to be preferentially transcribed from the onset of polar growth. Interestingly, the abundance of transcripts of "early genes" (those whose transcripts are predominantly detected in dormant conidia) can be, in many cases, approximately 10-100-fold higher than that found in genes that are Similar to that observed in the case of the chitinases, we were able to detect the presence of about half of the glucanases along the timeline of germination ( Figure 5A; Tables S2 and S4). Nonetheless, many of the glucanase transcripts were present mainly in dormant conidia, while three of them (28,3949, and 1833) appeared to be preferentially transcribed from the onset of polar growth. Interestingly, the abundance of transcripts of "early genes" (those whose transcripts are predominantly detected in dormant conidia) can be, in many cases, approximately 10-100-fold higher than that found in genes that are expressed later in the developmental process. To examine co-regulation among glucanaseencoding genes, Bayesian networks were inferred from relative glucanase expression, highlighting a "core" of hub-regulatory glucanases composed of four genes (28, 3949, 11,091, and 11). expressed later in the developmental process. To examine co-regulation among glucanase encoding genes, Bayesian networks were inferred from relative glucanase expression highlighting a "core" of hub-regulatory glucanases composed of four genes (28, 3949 11,091, and 11). In addition to the chitinases and glucanases, we also compared the relative expres sion of genes encoding proteins that are involved in biosynthetic aspects of cell wall re modeling. These genes included chitin synthases and glucan elongases (Figures 6 and 7 Seven of the eight T. asperelloides chitin synthases were differentially expressed during th four stages of germination (Figure 7; Tables S2 and S4). Transcripts of all the transcribed chs genes were detected in dormant conidia; chs2 appeared to be the most abundantly present in that developmental phase. In contrast, chs1 is an apparent "late" germination gene within the chs gene family, suggesting that it functions mainly after the onset of ger mination. In addition to the chitinases and glucanases, we also compared the relative expression of genes encoding proteins that are involved in biosynthetic aspects of cell wall remodeling. These genes included chitin synthases and glucan elongases (Figures 6 and 7). Seven of the eight T. asperelloides chitin synthases were differentially expressed during the four stages of germination (Figure 7; Tables S2 and S4). Transcripts of all the transcribed chs genes were detected in dormant conidia; chs2 appeared to be the most abundantly present in that developmental phase. In contrast, chs1 is an apparent "late" germination gene within the chs gene family, suggesting that it functions mainly after the onset of germination.
Four β-(1,3)-glucanosyltranferase genes were identified in the T. asperelloides genome. The members of this gel gene family were differentially expressed during germination, from dormancy to first branching. Transcripts of the genes gel1 and gel2 were most abundant in dormant conidia. However, they were also substantially transcribed along the time course of germination. Conversely, the gel3 gene was highly expressed "late": even though its transcripts were present in conidia, they were the most abundant compared to the other gel genes in the other phases of germination (Figure 7).

Congo Red Affects Germination and Alters Expression of gel3
Given the changes occurring in the transcript abundance of gel3 during germination and the documented significance of GEL3 homologues in cell wall integrity [35,36], we monitored gel3 expression during germination in liquid medium (PDB) cultures challenged with a sub-lethal concentration (500 µM) of Congo Red. While initiation of polarity did not seem to be affected by the presence of the dye (Figure 8, 7 h), only about 50% of the conidia progressed along the germination timeline beyond that initial stage. Those conidia that did germinate produced normal germ-tubes. However, they failed to branch within the time course of this experiment. Moreover, the characteristic branching and clumping observed in control culture germlings was not evident in the treated cultures. These phenotypic effects of Congo Red on germling development were accompanied by increases in the abundance of gel3 transcripts that were most evident at later stages of germination. After 10 h of incubation, the expression of gel3 was two-fold higher in the treated conidia. The difference increased to five-fold after 12 h of incubation with the dye (Figure 9). These gene expression effects of the inhibitor further support the presence of phase-specific checkpoints during germination and the ability of the developing germlings to respond to cell wall remodeling insults by altering gene expression of key components of this process.
normalized to tef1. The threshold cycle (2 −ΔΔCT ) method was used to determine fold changes in expression.
Four β-(1,3)-glucanosyltranferase genes were identified in the T. asperelloides genome. The members of this gel gene family were differentially expressed during germination, from dormancy to first branching. Transcripts of the genes gel1 and gel2 were most abundant in dormant conidia. However, they were also substantially transcribed along the time course of germination. Conversely, the gel3 gene was highly expressed "late": even though its transcripts were present in conidia, they were the most abundant compared to the other gel genes in the other phases of germination (Figure 7).

Congo Red Affects Germination and Alters Expression of gel3
Given the changes occurring in the transcript abundance of gel3 during germination and the documented significance of GEL3 homologues in cell wall integrity [35,36], we monitored gel3 expression during germination in liquid medium (PDB) cultures challenged with a sub-lethal concentration (500 µM) of Congo Red. While initiation of polarity did not seem to be affected by the presence of the dye (Figure 8, 7 h), only about 50% of the conidia progressed along the germination timeline beyond that initial stage. Those conidia that did germinate produced normal germ-tubes. However, they failed to branch within the time course of this experiment. Moreover, the characteristic branching and clumping observed in control culture germlings was not evident in the treated cultures. These phenotypic effects of Congo Red on germling development were accompanied by increases in the abundance of gel3 transcripts that were most evident at later stages of germination. After 10 h of incubation, the expression of gel3 was two-fold higher in the treated conidia. The difference increased to five-fold after 12 h of incubation with the dye (Figure 9). These gene expression effects of the inhibitor further support the presence of phase-specific checkpoints during germination and the ability of the developing germlings to respond to cell wall remodeling insults by altering gene expression of key components of this process.

Changes in Expression of Chitin-Synthase-and Glucan-Elongase-Encoding Genes in the Presence of Rhizoctonia solani
Genes of the chs and gel families were differentially expressed during germination on PDA. Therefore, we examined to what extent the expression pattern of these genes was dependent on the germination substrate. We determined the level of expression of these genes in the presence of a potential host. First, we observed the morphological dynamics of conidial germination of T. asperelloides conidia plated on membranes that had been placed on a live culture of R. solani. In contrast to the transition from dormancy to swelling that was observed during germination on PDA, the transition from dormancy to swelling on live culture was very subtle and challenging to quantitate. The emergence of germ tubes from the conidia was observed after ten hours, and elongation of the germ tube proceeded for about 40 h, with the average germling length reaching 25 µm after 30 h (Figure 10). Most conidia (75-80%) took almost twice as long (20 h) to germinate as what was measured on PDA. Based on these observations, and on the observation that T. asperelloides was able to germinate at normal rates even on water agar (data not shown), we concluded that the marked delay, as well as alteration in the synchronized progression of the germination program, was most likely linked to the presence of R. solani. The difference in the germination timeline was accompanied by marked changes in the transcript abundance of genes in the chs and gel families, as determined by RT-PCR ( Figure 11). While the expression patterns of the chs gene family on PDA, as determined here, were similar to those determined by RNASeq, the relative expression of members of the chs gene family was ten-fold greater in conidia that germinated on the host R. solani when compared to the expression measured on PDA ( Figure 11A,B). In addition to the increase in chs transcript abundance, the pattern of the chs transcript during germination on the two substrates also varied. In fact, the expression pattern of the chs gene family in the presence of a host appeared as a "mirror image" of the expression patterns under conditions representing a saprotrophic lifestyle of the mycoparasite (on PDA). Most chitin synthase genes, excluding chs1 and chs6, were highly expressed during germination on the host and shared a similar expression pattern.

Discussion
Here, we have shown that a functional classification of differentially expressed transcripts of germinating T. asperelloides conidia is characterized by an increase in various metabolic processes and electron transport, as well as enhancements in transcription factor activity, nuclear activity, nucleic-acid binding activity, and protein-kinase activity, all accompanying a striking and morphologically dynamic developmental process. The transcriptional profile of germinating T. asperelloides featured 11,166 genes whose transcript abundance changed during the fungus germination process. About half (5550 genes) were uniquely downregulated in the transition from dormancy to polar growth, and about a quarter (2961) were exclusively upregulated in the transition from elongation of the germ tube to first branching.
This dynamism of expression indicates the presence of an abundant repertoire of mRNA species that are likely to be produced during conidiogenesis. This was followed by the utilization of the pre-existing RNAs for the onset of germination and subsequent de novo synthesis of RNAs that were used in more advanced stages of germination. RNAs have been shown to be synthesized in the "dormant" conidia of A. nidulans, A. fumigatus, and Talaromyces marneffei [6]. Therefore, it is likely that such processes occur prior to germination in Trichoderma spp. as well. As opposed to loading of RNAs into forming conidia, the extent to which RNA synthesis within dormant conidia contributes to the extremely high representation of detectable RNA species (10,725, comprising over 95% of the predicted gene number in this species) has yet to be determined. However, the possible physiological/ecological implications of such a comprehensive diversity of transcripts are vast. Among the four gel genes, gel3 was the most highly expressed in both germinating conidia and after the onset of the first hyphal branching in cultures growing on PDA. As in the case of the chs transcripts, the relative expression of members of the glucanelongase family was ten-fold greater in conidia that germinated on the host R. solani when compared to their expression on PDA ( Figure 11C,D). Excluding gel4, all the gel genes were upregulated during germination in the presence of R. solani. Among them, gel3 was the most highly expressed in the germinating conidia.
Based on these results, we concluded that, even though most of the cell wall remodeling gene transcripts are present in dormant conidia, the dynamics of their transcriptional abundance can be significantly altered during the germination process in a manner that is dependent on the presence of this potential host.

Discussion
Here, we have shown that a functional classification of differentially expressed transcripts of germinating T. asperelloides conidia is characterized by an increase in various metabolic processes and electron transport, as well as enhancements in transcription factor activity, nuclear activity, nucleic-acid binding activity, and protein-kinase activity, all accompanying a striking and morphologically dynamic developmental process. The transcriptional profile of germinating T. asperelloides featured 11,166 genes whose transcript abundance changed during the fungus germination process. About half (5550 genes) were uniquely downregulated in the transition from dormancy to polar growth, and about a quarter (2961) were exclusively upregulated in the transition from elongation of the germ tube to first branching. This dynamism of expression indicates the presence of an abundant repertoire of mRNA species that are likely to be produced during conidiogenesis. This was followed by the utilization of the pre-existing RNAs for the onset of germination and subsequent de novo synthesis of RNAs that were used in more advanced stages of germination. RNAs have been shown to be synthesized in the "dormant" conidia of A. nidulans, A. fumigatus, and Talaromyces marneffei [6]. Therefore, it is likely that such processes occur prior to germination in Trichoderma spp. as well. As opposed to loading of RNAs into forming conidia, the extent to which RNA synthesis within dormant conidia contributes to the extremely high representation of detectable RNA species (10,725, comprising over 95% of the predicted gene number in this species) has yet to be determined. However, the possible physiological/ecological implications of such a comprehensive diversity of transcripts are vast.
A similar, albeit less pronounced, trend in transcript abundance was observed in germinating conidia of N. crassa conidia [7]. There, transcripts of about a third of the N. crassa genome were present in the dormant conidium, and about 14% were exclusively upregulated during first branch formation. When comparing the expression patterns between the stages of germination in N. crassa, it was observed that most of the transcripts present in the dormant conidia were absent at later stages of germination and that 40 times as many genes were downregulated in dormant conidia compared to germinated conidia [7]. In the example of A. niger, transcripts of 33% of the annotated genes were present in the dormant conidia, and the same fraction was detected in germinating conidia [10]. In Rhizopus delemar, transcripts representing over 70% of the genome were identified in resting spores. Only a slight decrease in the percentile of the genome represented was evident once germination had initiated. However, a shift in the abundance of different functional groups was observed [55]. Some of the transcripts that were unique to the resting spores of R. delemar included those that have predicted roles in lipid storage and localization and were suggested to be involved in dormant spore maintenance.
Interestingly, we were also able to identify several such lipid-related gene transcripts in dormant conidia of T. asperelloides. Among them were plasma-membrane proteolipid 3 and cyclopropane-fatty-acyl-phospholipid synthase (gene IDs 2212 and 8825, respectively; Table S2), whose transcript levels dropped to less than 0.1% once germination was initiated. In contrast, transcriptional profiling of germinating F. graminearum conidia showed similarity in both the number and regulation of genes expressed throughout different stages of germination. Of the 13,969 genes annotated, 1805 genes were upregulated and 1616 genes were downregulated in dormant conidia, while, in germinated conidia, 954 and 947 genes were upregulated and downregulated, respectively [12]. Taken together, it is clear that dynamic changes in transcript abundance of a substantial portion of the coding genome is a hallmark of fungal conidial germination. However, the relative proportions of the coding genomes involved vary among fungal species. In T. asperelloides, the changes we observed were among the most dynamic described to date in terms of the percentage of the genome involved. Members of the genus Trichoderma can be found in almost every conceivable niche in nature, and, in addition to their saprophytic nature, many species can be associated with a multitude of living fungi and plant hosts [14,17,20]. Therefore, these fungi are likely to have evolved to produce conidia with a repertoire of available transcripts that provide the metabolic preparation to germinate and thrive on diverse substrates.
Cell wall remodeling is a critical process involving many genes in various functional groups during germination. Changes in the abundance of transcripts encoding proteins involved in cell wall remodeling were especially striking. The abundances of many of these transcripts were reduced during the early germination (stage 1 < stage 2) but were synthesized de novo during late germination (stage 3 < stage 4). Cell wall remodeling includes both the biosynthesis as well as the degradation of cell wall components [9]. Degradation of the rigid conidial cell wall is a prerequisite for progression of the germination process and may depend on the activities of a range of hydrolytic enzymes. Several genes accounting for these enzymes were identified in transcriptomic and proteomic studies during the germination of Aspergillus sp. [9,11,56,57]. Among them, there were genes for several chitin synthases and 1,3-β-glucanosyltransferases. Regulatory associations among genes in the functional groups of interest were reconstructed via Bayesian gene networks using expression fold-change collected from the four major morphological stages in conidial germination in T. asperelloides. Such networks provide preliminary working hypotheses that can productively guide future gene manipulation experiments that efficiently and rigorously assess regulatory networks for these functional groups. Here, Bayesian networks of the chitinases and glucanases identified a selection of genes that comprise a core of chitinases and glucanases featuring the most marked stage-specific expression changes during germination. We identified several potential "core" genes with marked changes observed in expression. These core genes may serve as functional stage-specific markers for further elucidation of the genetics and regulation of conidial germination in this species and its close (if not distant) relatives.
For efficiency of resource use, the degradation of existing cell wall material must be coordinately orchestrated with de novo synthesis of the cell wall material in the developing germling. Much attention has been devoted to the involvement of chitin synthases in cell wall biosynthesis in a variety of fungi [31,56,[58][59][60][61]. For example, the differential expression of the eight-member T. atroviride chitin synthase gene family during germination and vegetative growth has been characterized [33]. However, under the conditions tested here, the transcripts of only seven of the eight T. asperelloides chitin synthases were above the conventional threshold for detection ( Figure 6). The detectable transcript abundance of all chs genes was highest in dormant conidia, with the exception of chs1, whose transcript was present only from stage 2 of germination. While the N. crassa homologue of this gene, chs-2, was found to be non-essential [62,63], it was found to localize to the Spitzenkörper, which would certainly be expected in a germling, especially once polarity has been established. In contrast, while we found that chs7 transcript levels were highly abundant in conidia, the mRNAs of this gene were almost undetectable once germination had been initiated but were found at high levels in mature hyphae [64]. As mentioned, the transcript of chs8 was below detection levels under the conditions tested. The transcript levels of this gene were reported as being very low during vegetative development yet highly abundant when the fungus was exposed to osmotic or oxidative stress [64].
Another major polymer in the fungal cell wall is β-1,3-glucan, which can be remodeled by members of the GH72 of the glycoside hydrolase family of proteins, also known as glucan elongases (GELs) [34,35]. Most fungi have several GEL-encoding genes, and the expression of each gene may be differentially regulated during various growth stages and/or environmental conditions. N. crassa has five gel genes, among which gel-3 was most strongly expressed both in germinating conidia and during hyphelongation. Deletion of gel-3 affected vegetative growth, causing slow growth and abnormal morphology [65]. To date, 1,3-β-glucanosyltransferases have not been studied in Trichoderma. In this study, the expression of four GEL-encoding genes was detected along the germination timeline. At the same time, gel1-3 exhibited lower, albeit differential, expression during all stages of germination. Among these three gel genes, that of gel3 appeared to be lowest in the dormant conidia, with higher levels of expression observed after the onset of conidiation. To determine whether gel3 expression is not only dependent on the developmental program of germination but could also be induced following external cell-wall-related aggravation, we measured gel3 transcript abundance in cultures challenged with cell wall perturbing agent Congo Red [66]. During the early phases of germination, the effect of Congo Red on the gel3 transcript levels was minor. However, after 12 h of incubation with the dye gel3, the transcript levels were almost five-fold higher when compared to the control. One possible explanation for the late increase in transcript abundance is that the intact thicker conidium cell wall acts as an effective physical barrier against the dye during the pre-germling phase. Alternatively, it is possible that the drug must be taken up into the cell and processed to be effective and that the machinery involved is expressed only at later stages of germination. Gene expression analysis during exposure of growing mycelial cultures of A. fumigatus to Congo Red revealed overexpression of Gel4 and Gel7, suggesting that the presence of Congo Red activates the reorganization of β-1,3-glucans through processing enzymes [67].
To date, the mycoparasitism of T. asperelloides strain T203, as well as other Trichoderma spp. strains, has been extensively studied in advanced developmental stages, mainly during hyphal coiling around its hosts [68,69]. This study has focused on the processes underlying early developmental and pre-infection stages. Germination of T. asperelloides was morphologically and temporally distinct when cultured on PDA versus when cultured in the presence of the host R. solani (Figures 1 and 10). Germination in the presence of R. solani was ten-fold slower, and the germlings appeared shrunken. Furthermore, they elongated for extended periods of time (ten to twenty hours) prior to branching. The germination pattern observed could be attributable to the potential host, whose structure or exudates affect the germination process. In concert with slower germination, the transcript abundance of several genes involved in cell wall remodeling was significantly (up to 10fold) higher. These changes in transcript abundance are in agreement with recent research demonstrating a prey-sensing, recognition, and directional chemotropic approach during the early phases of interaction [70,71]. Therefore, the high transcript levels observed may be related to the mentioned processes. Interestingly, the highest levels of chs gene expression in the presence of the host were observed in the case of chs4, chs5, and chs7, especially at the later stages of germination. Homologues of chs4, chs5, and chs7 in T. atroviridae were highly expressed during mycoparasitism [64]. Our data support the presence of a transcriptional response to the presence of a host during germination. However, the presence of such a response does not rule out the possibility that slower mRNA turnover could also contribute to the higher transcript levels observed.
In this study, we have demonstrated that conidial germination in a member of the highly ubiquitous genus Trichoderma is among the most dynamic, in terms of changes in transcript abundance, compared to other species analyzed to date. The expression changes observed here are likely accompanied by many additional changes in the biochemical and physiological state of the germinating fungus, as has been shown in other species [72][73][74]. The extent to which transcriptional hallmarks of germination are conserved under varying conditions in diverse fungi has yet to be determined. Comparative transcriptional analyses are likely to provide both evolutionary insights concerning fungal development and lifestyles, as well as means for identifying and eventually manipulating key regulatory elements in this fundamental process.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jof8080854/s1, Table S1: RNAseq statistics for 4 stages of conidial germination of T. asperelloides on PDA.; Table S2: LOX measurements of T. asperelloides transcriptomics with detailed functional annotations.; Table S3: GO terms enrichment analysis during T. asperelloides condial germination. Genes that are significantly differentially expressed were annotated for each pair of stage comparisons; Table S4: Gene designations and links for Chitinases, Glucanases, Chitin Synthases and Glucan Elongases.  . We thank the Yale Center for Genome Analysis for the next generation sequencing support. This work was also supported by the HPC facilities operated by the Yale Center for Research Computing and the Yale Center for Genome Analysis, as well as NIH grant 1S10OD018521, which helped fund the cluster.

Conflicts of Interest:
The authors declare no conflict of interest.