A Mitogen-Activated Protein Kinase Tmk3 Participates in High Osmolarity Resistance, Cell Wall Integrity Maintenance and Cellulase Production Regulation in Trichoderma reesei

The mitogen-activated protein kinase (MAPK) pathways are important signal transduction pathways conserved in essentially all eukaryotes, but haven't been subjected to functional studies in the most important cellulase-producing filamentous fungus Trichoderma reesei. Previous reports suggested the presence of three MAPKs in T. reesei: Tmk1, Tmk2, and Tmk3. By exploring the phenotypic features of T. reesei Δtmk3, we first showed elevated NaCl sensitivity and repressed transcription of genes involved in glycerol/trehalose biosynthesis under higher osmolarity, suggesting Tmk3 participates in high osmolarity resistance via derepression of genes involved in osmotic stabilizer biosynthesis. We also showed significant downregulation of genes encoding chitin synthases and a β-1,3-glucan synthase, decreased chitin content, ‘budded’ hyphal appearance typical to cell wall defective strains, and increased sensitivity to calcofluor white/Congo red in the tmk3 deficient strain, suggesting Tmk3 is involved in cell wall integrity maintenance in T. reesei. We further observed the decrease of cellulase transcription and production in T. reesei Δtmk3 during submerged cultivation, as well as the presence of MAPK phosphorylation sites on known transcription factors involved in cellulase regulation, suggesting Tmk3 is also involved in the regulation of cellulase production. Finally, the expression of cell wall integrity related genes, the expression of cellulase coding genes, cellulase production and biomass accumulation were compared between T. reesei Δtmk3 grown in solid state media and submerged media, showing a strong restoration effect in solid state media from defects resulted from tmk3 deletion. These results showed novel physiological processes that fungal Hog1-type MAPKs are involved in, and present the first experimental investigation of MAPK signaling pathways in T. reesei. Our observations on the restoration effect during solid state cultivation suggest that T. reesei is evolved to favor solid state growth, bringing up the proposal that the submerged condition normally used during investigations on fungal physiology might be misleading.


Introduction
The systematic risk inherent to a fossil fuel based economy has drawn worldwide attention, leading to the proposal of replacing these non-renewable energy sources with renewable energy sources including energy from renewable biomass. One such biomass is lignocellulose, which is abundantly available as agricultural or forestry wastes. Conversion of lignocellulose to energy sources that are compatible to modern industry is a complicated process, in which the most important step is the saccharification of lignocellulose, catalyzed by cellulases produced from filamentous fungi [1]. The most widely used and important cellulase hyper-producing organism is Trichoderma reesei. Physiological investigations of this so-called industrial cellulase-producing 'workhorse' are therefore essential in both scientific and technological regards [2].
Cells sense their surrounding environments and react to external signals via signal transduction pathways. Among all known signal transduction pathways, the mitogen-activated protein kinase (MAPK) pathway is ubiquitous in almost all eukaryotic species and is one of the most well characterized pathways [3]. This pathway features a signal relay cascade in which three kinases are involved: the MAPK kinase kinase (MAPKKK), the MAPK kinase (MAPKK) and MAPK [4]. MAPK further phosphorylates downstream elements involved in the regulation of physiological activities. The most extensive research of these pathways was carried out in Saccharomyces cerevisiae, in which six MAPK pathways were identified [5]. In filamentous fungi, three major classes of MAPKs are present, respectively homologous to yeast Hog1p, Slt2p, and Fus3p. These MAPKs have been shown to function in a variety of physiological processes such as fruiting body development [6], polarized growth [7], biosynthesis [8], conidiation [9], pathogenicity [10], circadian rhythmicity [11], stress response [12,13], protein production [9,14] and cell wall integrity maintenance [15].
Investigations carried out on Slt2p homologues in filamentous fungi suggested the involvement of this MAPK in the maintenance of cell wall integrity in A. fumigatus [12,15], Magnaporthe oryzae [23], Trichoderma virens [24], Claviceps purpurea and Fusarium graminearum [25], but not Mycosphaerella graminicola (Mgslt2) and Botrytis cinerea [26]. The involvement of a gene in cell wall integrity maintenance is generally examined by testing for hypersensitivity in the gene deletion strain to cell wall lysing enzymes or cell wall interfering compounds such as calcofluor white (CFW) and Congo red (CR). Interestingly, the deletion of hog1 homologues in T. harzianum and A. fumigatus showed unchanged or even increased resistance to CFW and CR, suggesting the hypersensitivity to improved osmolarity is unrelated to weakening of the cell wall [21,27].
In filamentous fungi, yeast Fus3p homologues are involved in quite diverse processes, one of which is the production of glycoside hydrolases. Cellulase and chitinase production was upregulated in the tmkA (a fus3 homologue) deletion strain of T. virens [9]. Chitinase production was improved in the tmk1 (a fus3 homologue) knockout strain of Trichoderma atroviride [14]. The expression levels of N-acetylglucosaminidaseand chitinase-encoding genes increased when tvk1 (a fus3 homologue) was disrupted in T. virens [28]. This phenomenon is of particular interest in cellulaseproducing filamentous fungi like Trichoderma and Aspergillus species, as they are well-known industrial cellulase hyper-producing strains, out of which T. reesei is the most widely studied and used.
T. reesei (syn. Hypocrea jecorina) was first isolated on the Solomon Islands during World War II, and has received considerable improvements over the last seventy years for industrial applications [29]. It has turned a paradigm for the investigations of cellulases, hemicellulases, and the molecular mechanisms underlying their synthesis and regulation [30]. However, knowledge on  the signal transduction cascades in this fungus is limited to a few aspects: 1) light-modulated cellulase production mediated by G proteins (Gna1 and Gna3), a PAS/LOV domain protein ENVOY and cAMP-dependent protein kinase A signaling [31][32][33][34]; 2) regulation of the sexual development process [35]; 3) cellulose-and cAMP-independent modulation of cellulase production mediated by Ras GTPase TrRas2 [36]. Relatively little is known in other well-known signal transduction pathways such as MAPK pathways, Ca 2+ -signaling pathways and factors such as casein kinase II, germinal center kinases and protein kinase C.
In silico reconstruction of the MAPK signal transduction cascade in T. reesei identified three distinct pathways, in which three putative MAPKs are involved: the yeast Fus3p homologue Tmk1, Slt2p homologue Tmk2, and the Hog1p homologue Tmk3 [37]. None of these MAPKs have been studied in T. reesei yet. In this study, by characterizing the properties of the T. reesei Dtmk3 deletion strain, we attempt to identify the role of Tmk3 in T. reesei, particularly in cell wall integrity maintenance and cellulase production that have not been identified for Hog1type MAPKs in other fungal species. Further comparison between submerged and solid state growth leads to the finding of novel, interesting restoration effects during solid state cultivation. These studies are the first research done on MAPKs in T. reesei, leading up to further in-depth understanding of the regulatory mechanisms of this well-known industrial cellulase hyper-producing workhorse.

Phylogenetic analysis
Phylogenetic tree construction and protein sequence comparison were carried out using the Clustal X2 software [39].

Construction of tmk3 deletion strain
Construction of T. reesei Dtmk3 was carried out essentially as previously described [38], using the pyr4 selection marker which restores uridine biosynthesis capabilities in the pyr4-deficient parent strain. Amplification of tmk3 from T. reesei genome was carried out using Kod FX high fidelity enzymes (TOYOBO CO. LTD. Osaka, Japan).

Southern blotting analysis
Southern blotting experiment was used to confirm whether T. reesei Dtmk3 was successfully constructed. Genomic DNA extracted from T. reesei Dku70 or Dtmk3 strain was digested using HindIII prior to hybridization. Detection of probe-hybridized DNA fragment was carried out using the DIG High Prime DNA Labeling and Detection Starter Kit I (Roche Diagnostics, Mannheim, Germany).

Phenotypic analysis
Approximately 10 6 conidiospores of T. reesei Dku70 or Dtmk3 strain were inoculated on minimal media (MM) plates containing glucose, starch, sucrose, lactose or glycerol, as well as PDA plates. The plates were incubated in an incubator (Model MJX-250, Ningbo Jiangnan Instrument Factory, Ningbo, China) at 30uC for 4 days. Double-layer avicel plates were prepared by first casting a MM agarose bottom layer containing no carbon sources, and a second MM agarose top layer containing 1% avicel. Approximately 10 6 conidiospores of T. reesei Dku70 or Dtmk3 strain were inoculated on the plates, which were incubated at 30uC for 4 days prior to examination.
Images of T. reesei Dku70 or Dtmk3 hyphae were taken with a bright field microscope (Nikon eclipse E100, 400 fold magnification) from cultures grown in submerged minimal media containing glucose as the carbon source.

Biochemical assays
The ATP level was assayed using the Checklite 250 plus ATP kit (Kikkoman Biochemifa Company, Minato-ku, Japan). The concentration of ATP indicates the average ATP concentration in cultures. Five individual replicates were carried out for ATP level analysis in solid state cultures. Three individual replicates were carried out for ATP level analysis in submerged cultures. The DNA content of a T. reesei Dku70 or Dtmk3 culture was also used as a measure of biomass. To measure the DNA content, solid state and submerged cultures were first prepared as is described in this work. For submerged cultures, 1 ml of the culture was drawn from flasks for DNA content determination. For solid state cultures, 25 ml of distilled water was used to suspend the cultures, and 1 ml of suspended culture was drawn for DNA content determination. The cultures were subsequently diluted 5-fold in 15 ml centrifuge tubes, and were subjected to a brief centrifugation to remove supernatant. One milliliter of 10% trichloroacetic acid (TCA) was then added to the tube, thoroughly mixed, and incubated on ice for 3 minutes. The tube was boiled in a water bath (Shanghai Jinghong Laboratory Instrument Co. Ltd., Shanghai, China) for 30 minutes, followed by centrifugation at 10,000 rpm for 10 minutes. Absorbance at 260 nm in the supernatant was subsequently measured using a UV-visible spectrophotometer (Model 2802, UNICO, Dayton, NJ, US). The total DNA content in the cultures was calculated from A 260 . Three individual replicates were carried out for each assay.
Protein concentration was determined using the Lowry method [40]. Filter paperase activity (FPA) was assayed following previously published protocols [41]. Cellobiohydrolase, endoglucanase, b-glucosidase and b-xylosidase activities were respectively assayed by abilities to hydrolyze pNPC, CMC, pNPG and pNPX following published protocols [42]. Six replicates were performed to determine extracellular protein concentrations, pNPCase, CMCase, pNPGase and pNPXase activities in solid state cultures. Three replicates were performed to determine extracellular

Chitin content assay
Chitin contents of T. reesei Dku70 and T. reesei Dtmk3 were assayed similarly to previous reports [43,44]. Approximately 10 6 spores of T. reesei Dku70 or Dtmk3 strain were inoculated in submerged media containing 2% glucose. The mycelia were harvested after 3 days of growth, and subsequently dried by heating at 105uC for 4 hours in an oven (Model DHG-9030, Shanghai Jinghong Laboratory Instrument Co. Ltd., Shanghai, China) for the determination of dry cell weight (DCW). Sixty mg of mycelia (14 mg DCW) of both T. reesei Dku70 or Dtmk3 strain were heated at 80uC for 90 minutes in 1 ml of 6% KOH using a water bath (Shanghai Jinghong Laboratory Instrument Co. Ltd., Shanghai, China), followed by addition of 0.1 ml glacial acetic acid. The treated mycelia were further centrifuged in a microcentrifuge (Centrifuge 5415R, Eppendorf, Hamburg, Germany) at 13,000 rpm for 10 minutes. The pellet was suspended in 0.5 ml phosphate buffer (pH 6.3) and digested using 0.1 U chitinase by incubation at 37uC for 1 hour. The reaction system was centrifuged again at 13,000 rpm for 10 minutes. Zero point two five mg of cytohelicase was subsequently added to 500 ml supernant, followed by incubation at 37uC for 1 hour. The amount of released N-acetylglucosamine was assayed according to previous reported procedures [44].

Real-time PCR reactions and data manipulation
Total RNA was extracted from T. reesei Dku70 or Dtmk3 strain growing in solid state media and submerged media containing 2% avicel and 2% wheat bran for the examination of expression of cellulase-, hemicellulase-, chitin synthase-and b-1,3-glucan synthase-coding genes, as well as from T. reesei Dku70 or Dtmk3 strain growing in submerged media containing 2% glucose and 0 or 0.15 M NaCl for the examination of expression of glycerol-3phosphate dehydrogenase-and a,a -trehalose-6-phosphate synthase-coding genes. cDNA was synthesized using PrimeScript RT reagent kit with gDNA erase (Perfect Real Time) from Takara Bio Inc. (Shiga, Japan).
Real-time PCR reactions were carried out on a LightCycler 480II Real-Time PCR system (Roche Applied Science, Mannheim, Germany) using SYBR Premix EX Taq TM II (Takara Bio Inc., Shiga, Japan) as the dye. Three individual biological replicates and three individual technical replicates for each biological sample (a total of 9 replicates for each reaction) were carried out. The relative abundance of genes was calculated using the 2 2DDCt method as previously described [45].

Results and Discussion
tmk3 encodes a Hog1-type MAPK in T. reesei Phylogenetic analysis of Tmk3 from T. reesei and other previously characterized MAPKs showed Tmk3 apparently cluster with Hog1-type MAPKs from other species (Fig. S1), leading to the suggestion that tmk3 likely encodes a Hog1 type MAPK. Sequence comparison between Tmk3 from T. reesei and Hog1 from S. cerevisiae showed a sequence identity of 66%, further supporting this suggestion (Fig. S2). It can therefore be concluded that Tmk3 is homologous to Hog1-type MAPKs, which function in high osmolarity resistance in S. cerevisiae and other filamentous fungi [3].

Construction and growth patterns of T. reesei Dtmk3
The tmk3 deletion strain was constructed via homologous recombination using T. reesei Dku70 as the parent strain (Fig. S3), in which the non-homologous end joining pathway was defective [38]. Examination of the growth of parent and Dtmk3 strains on plates lead to several interesting findings: 1) The growth of T. reesei Dtmk3 on each minimal media plate containing a tested carbon source (glucose, starch, lactose, sucrose, glycerol) is significantly worse than the parent strain; 2) Only slightly slower growth was observed when T. reesei Dtmk3 was grown on complete media (PDA plates); 3) a significantly larger transparent zone was observed around the colony of T. reesei Dtmk3 in comparison to the parent strain, when the two strains were grown on double layer plates containing avicel (Fig. 1).
The differences between the performance of T. reesei Dtmk3 on minimal and complete media reflect damaged biosynthesis of certain critical compounds in the tmk3-deficient strain. The availability of these compounds in complete media partially  Table 2. Predicted MAPK phosphorylation sites in transcription factors. Cre1  S164, S262, T281, T285, T289   Xyr1  S324, T405, T443   ACEI  S10, S42, T365, T393, T513, T521, S523, S524, T525,  T614, T660   ACEII  T201 doi:10.1371/journal.pone.0072189.t002 compensates for the defects. Our results cannot specifically suggest which pathways are damaged, which remains to be elucidated in further studies. The larger transparent zone around the colony on avicelcontaining double layer plate is an indication of improved avicelase production in T. reesei Dtmk3. The term 'avicelase' here refers to the avicel-hydrolyzing enzymatic activities, which are primarily determined by cellobiohydrolase activities for two reasons: 1) cellobiohydrolase is the most abundant enzyme in secreted T. reesei cellulases [46]; 2) avicel has a highly crystalized cellulose structure and lacks the amorphous regions favored by endoglucanases. Although the contribution of other major cellulases cannot be excluded, the improved avicelase production in T. reesei Dtmk3 can be interpreted primarily as the increase of cellobiohydrolase secretion.

Transcription factor MAPK phosphorylation sites
Growth of T. reesei parent and Dtmk3 strains was also compared during submerged cultivation. The levels of ATP production and DNA content were used as measures of biomass accumulation. As is shown in Fig. 2, growth is significantly hampered in the tmk3deficient strain, suggesting the important role Tmk3 plays in vegetative growth. Participation of Tmk3 in high osmolarity resistance via derepression of glycerol and trehalose synthesis genes We examined the tolerance of T. reesei parent and Dtmk3 strains to elevated osmotic pressure by growing them on plates containing various concentrations of NaCl. It appears that T. reesei Dtmk3 colonies ceased to develop at the NaCl level of 0.5 M, while the parent strain can still grow in the presence of 0.9 M NaCl (Fig. 3A). The apparent higher sensitivity to NaCl for the tmk3 deletion strain is an indication of hampered tolerance to high osmolarity, which is indicative of a Hog1-like function for Tmk3.
It was reported that glycerol and trehalose are involved in high osmolarity resistance by improving intracellular osmolarity in S. cerevisiae [16,47]. In particular, gpd1 that encodes a glycerol-3phosphate dehydrogenase and functions in glycerol biosynthesis was upregulated during exposure to high osmolarity [16]. In T. reesei, we found two glycerol-3-phosphate dehydrogenase-coding genes functioning in glycerol biosynthesis (Trire2_76620 and Trire2_2574) and five a,a-trehalose-6-phosphate synthase-coding genes (Trire2_75295, Trire2_77602, Trire2_121491, Trire2_73 134 and Trire2_48707) functioning in trehalose biosynthesis. Unlike in S. cerevisiae, the upregulation of these genes in response to improved osmolarity is not apparent in T. reesei parent strain (Fig. 4). In T. reesei Dtmk3, however, the expression levels of these genes are downregulated when exposed to 0.15 M NaCl (Fig. 4). These results suggest the Hog1-homologue Tmk3 functions in high osmolarity resistance in T. reesei, similarly to S. cerevisiae. The mechanism of this function, however, appears different: Hog1p of S. cerevisiae functions in stimulation of genes in the osmotic stabilizer biosynthesis; Tmk3 of T. reesei functions in derepression of genes in the osmotic stabilizer biosynthesis.

Involvement of Tmk3 in cell wall structure integrity maintenance
The integrity of T. reesei parent and Dtmk3 strains' cell wall was investigated by testing of sensitivity to cell wall interfering substances CFW and CR. On CFW-and CR-containing plates, T. reesei Dtmk3 colonies are significantly smaller than those of the parent strain (Fig. 3B, 3C). T. reesei Dtmk3 could not grow on plates containing 125 mg/L CR or 30 mg/L CFW, while the parent strain was able to tolerate at least 200 mg/L CR and 40 mg/L CFW. Analysis of colony diameters showed T. reesei Dtmk3 is affected by CFW and CR more severely than T. reesei parent strain (Table 1), further suggesting the deletion of tmk3 leads to hypersensitivity to CFW and CR and reduction of cell wall integrity.
Comparison of the hyphal phenotype of T. reesei parent and Dtmk3 strains further supported the presence of weakened cell wall in the tmk3 knockout strain. T. reesei parent strain hyphae have a smooth cell wall structure (Fig. 5A, 5B), while T. reesei Dtmk3 hyphae apparently adopt a 'budded' appearance, in agreement with having compromised cell wall structure (Fig. 5C, 5D), although other possible defects such as impaired polarity maintenance cannot be excluded [48,49].
The fungal cell wall is primarily composed of proteins and polysaccharides including chitin, b-1,3-glucan and b-1,6-glucan [50]. It has been previously reported that chitin synthase and b-1,3-glucan synthase are involved in the synthesis of chitin and b-1,3-glucan in filamentous fungi [51,52]. Nine chitin synthase coding genes (Trire2_112271, Trire2_58188, Trire2_55341, Trire2_51492, Trire2_124228, Trire2_122172, Trire2_71563, Trire2_67600) and one b-1,3-glucan synthase coding gene (Trire2_78176, fks) are present in the genome of T. reesei. The transcriptional abundance of these genes in T. reesei parent and Dtmk3 strains was investigated with real-time PCR. The transcription of all these genes except for Trire2_67600 was detected, and all the detected genes except for Trire2_55341 are significantly (2,8 folds) downregulated in T. reesei Dtmk3 (Fig. 6), suggesting Tmk3 participates in the cell wall integrity pathway by regulating the synthesis of chitin and b-1,3-glucan. These transcriptional responses were further supported by the observation of higher chitin content in T. reesei parent strain (4.1960.07 percent DCW) than in T. reesei Dtmk3 strain (3.4860.07 percent DCW).
Our results reported in this study clearly suggest the involvement of Tmk3 in the maintenance of cell wall integrity in T. reesei. Interestingly, instead of Hog1-type MAPKs, Slt2-type MAPKs were reportedly involved in cell wall integrity maintenance (as is summarized in the introduction section). The unaltered or even increased resistance to CR and CFW in hog1 deletion strains of T. harzianum and A. fumigatus further suggested Hog1 is not involved in cell wall integrity pathways in these two filamentous fungi [21,27]. The 'budded' appearance of T. reesei Dtmk3 hyphae in minimal media is similar to that of A. fumigatus DmpkA (MpkA is a Slt2 homologue in A. fumigatus) [15], while this phenomenon was only apparent in hog1 deletion strains of M. grisea and A. nidulans when exposed in high salt media due to their incapability to cope with high osmotic pressure [10,19]. All these observations lead to the suggestion that Tmk3, unlike Hog1-type MAPKs in other fungi, functions in the cell wall integrity pathway similarly to Slt2-type MAPKs. This also explains the discrepancy of growth performance in standard low salt media between T. reesei Dtmk3 and other filamentous fungi in which hog1 homologues were inactivated [10,13,21]: Hog1 in other filamentous fungi only responds to highsalt environments, so they grow normally in low salt media; Tmk3 is involved in cell wall integrity maintenance, and there is a growth defect for T. reesei Dtmk3 even under low osmolarity.
The decrease in cellulase and hemicellulase production in T. reesei Dtmk3 cannot be solely explained by the reduced biomass accumulation, because the transcriptional levels of cellulase-and hemicellulase-coding genes are clearly downregulated. These observations lead to the suggestion that Tmk3 is involved in cellulase production and/or induction. In T. reesei, several transcription factors were shown to regulate cellulase production, including the activator Xyr1 [53], repressor Cre1 [54], repressor ACEI [55] and activator ACEII [56]. Using a phosphorylation prediction software KinasePhos [57], we were able to identify MAPK phosphorylation sites on all these transcription factors ( Table 2). Previous reports have suggested that the phosphorylation of some of these transcription factors is the prerequisite for their activities [58,59]. It is therefore reasonable to propose that Tmk3 may regulate cellulase production by phosphorylating, and therefore activating, transcription factors. This proposal, however, should not be interpreted as the activation of ALL these transcription factors by Tmk3, as two more MAPKs (Tmk1, Tmk2) are present in T. reesei, and the phosphorylation specificity of the transcription factors has not been identified.

Restored growth and cellulase production in solid state media
One common problem on the studies of T. reesei physiology is that the way T. reesei is cultured in these investigations does not resemble their natural habitats. In nature, T. reesei grows on solid lignocellulosic particles, instead of in submerged media which are normally used for T. reesei cultivation during physiological studies. There are considerable differences between these two environments: the level of encountered moisture varies significantly; the exposure to oxygen is different; and in the natural environment, secreted proteins do not diffuse like during submerged cultivation. We therefore compared the performance of T. reesei Dtmk3 in submerged media and solid state media, in order to further identify the role of Tmk3 in T. reesei.
A significant improvement in ATP and DNA contents, as measures of biomass accumulation, could be observed when T. reesei Dtmk3 was grown on solid state media (Fig. 9), when compared to submerged cultures (Fig. 2). However, in comparison with submerged cultivation, chitin synthase and b-1,3-glucan synthase transcription showed a similar magnitude of downregulation during solid state cultivation following tmk3 deletion, except for Trire2_55341 (Fig. 6). Therefore, the improved ATP and DNA contents, as measures of biomass production, during solid state cultivation for T. reesei Dtmk3 can unlikely be attributed to recovered cell wall integrity. It has been known for a long time that cell walls serve as the skeletons of fungal cells, and that fungal protoplasts die due to inflow of water unless osmolarity in the growth medium is adjusted [60]. One proposed explanation to the observed improvement in biomass accumulation for T. reesei Dtmk3 is that during solid state cultivation, secreted proteins do not diffuse away from the fungal cells, therefore increasing regional protein accumulation and subsequently regional osmolarity. This 'protein layer' could serve as the osmotic stabilizer for fungal cells with weakened cell wall, and could lead to improved growth of T. reesei Dtmk3 in solid-state media.
In contrary to chitin synthase and b-1,3-glucan synthase encoding genes, cellulase and hemicellulase encoding genes show a significantly different transcription pattern in T. reesei Dtmk3. The transcription of cbh1, cbh2, egl1, egl2, bgl1 was downregulated by 1.2 to 2.6 fold during solid state cultivation (Fig. 8), while the magnitude of downregulation was respectively 93, 80, 93, 73 and 5.5 fold during submerged cultivation. The transcription of bxl1 was downregulated by 4.9 fold during submerged growth, but was upregulated by 4.4 fold during solid state growth. The behavior of bgl1 and bxl1 transcription was not the same as that of cbh1, cbh2, egl1 and egl2, suggesting they were regulated differently.
Extracellular protein level and cellulase activities in T. reesei Dtmk3 cultures were partially restored during solid-state cultivation similar to the transcriptional abundance ( Fig. 10A-E). b-xylosidase activity was not restored, although the transcription of bxl1 was upregulated in the tmk3 deletion strain (Fig. 10F), again suggesting b-xylosidase production is regulated differently from cellulases. Interestingly, the production of pNPCase and pNPGase activities is improved in T. reesei Dtmk3 when comparing with the parent strain (Fig. 10C, E). This is consistent with our observation that a larger clearing zone was present around T. reesei Dtmk3 colonies than T. reesei parent strain colonies when they are grown on double-layer avicel plates, because the surrounding environment of aerial hyphae on plates is similar to solid state cultivation. One proposed explanation to this improvement of cellulase production is that the weaker cell wall in T. reesei Dtmk3 allows easier protein permeation and therefore may benefit secretion and/or release of cell bound enzyme or protein. This beneficial effect could not be offset by the slight downregulation of cbh1 and bgl1 transcription and decreased biomass accumulation, and therefore leads to the improvement of overall cellobiohydrolase and b-glucosidase production.
Providing an explanation for the restoration of transcription for all the cellulase-coding genes during solid state cultivation is difficult before investigations could provide further evidence for the involvement of an additional regulatory mechanism. However, these results do suggest the interplay between Tmk3 mediated and other cellulase/hemicellulase regulatory mechanisms. Indeed, cellulase/hemicellulase production is regulated by a complicated regulatory network that is governed by many signals (such as light, carbon source etc), as well as extracellular and cytoplasmic factors [34,36]. It is therefore not a surprise for Tmk3 mediated regulatory mechanism and other regulatory mechanisms to affect each other.
Our studies of T. reesei Dtmk3 grown during solid state cultivation showed a clear restorative effect in comparison to submerged cultivation. The molecular mechanisms underlying these effects vary on different aspects of cellular physiology, but the unified final restorative effect suggests T. reesei physiology is evolved to favor solid state cultivation. Although secreted proteins or metabolites can be produced more efficiently from filamentous fungi during submerged cultivation, therefore benefiting industrial application, the submerged condition might actually alter fungal growth and metabolism, and care needs to be taken when assumptions are made that growth under certain submerged growth conditions is 'optimum' or 'physiological', particularly during mechanistic studies on fungal physiology.

Conclusions
Results reported herein lead to the suggestion of the physiological roles in high osmolarity resistance, cell wall integrity maintenance and cellulase production of a Hog1-type MAPK Tmk3 in T. reesei. Although sequence comparison suggests Tmk3 is homologous to Hog1p from S. cerevisiae, their functions appear to vary significantly. While Hog1p from S. cerevisiae and Tmk3 from T. reesei are both involved in the resistance to high osmolarity, the mechanism of this function differs: induction of osmotic stabilizer resistance genes in S. cerevisiae and derepression of these genes in T. reesei. Further investigations showed the function of Tmk3 in more novel aspects. Our phenotypic analysis, chemical sensitivity studies and transcriptional profiling all suggested Tmk3 functions in the cell wall integrity signaling pathway similarly to Slt2-type MAPKs, a role never identified for Hog1-type MAPK in filamentous fungi before. We further observed that deletion of tmk3 leads to an apparent decrease in cellulase transcription, and suggested the involvement of Tmk3 in cellulase production regulation by phosphorylating transcription factors.
When grown in submerged media, biomass accumulation and cellulase production of T. reesei were significantly reduced upon tmk3 deletion. The degree of reduction was much smaller after transferring to solid-state media. The mechanism of the restoration of growth likely differs from the mechanism of the restoration of cellulase production, but the similarity between the effects suggests T. reesei is evolved to favor solid state cultivation.

Supporting Information
(TIF) Figure S2 Sequence alignment of Tmk3 from T. reesei and Hog1 from Saccharomyces cerevisiae. Trtmk3, Tmk3 from T. reesei; Schog1, Hog1 from S. cerevisiae. (TIF) Figure S3 Southern blotting analysis of T. reesei parent and Dtmk3 strains. M, DNA molecular size marker; Dtmk3, T. reesei Dtmk3; parent, T. reesei parent strain. Indicated by arrows are the predicted sizes of DNA fragments hybridized with the probe. (TIF)