Fungal spores were collected from several locations of central Taiwan area during July-December of 2011 by pumping air into 0.005% Triton-X100 buffer (at a flow rate of 300 L/min for 10 minutes) with a Coriolis^® μ Air Sampler (Bertin Technologies, France). Serially diluted air samples were spread on potato dextrose agar plate (PDA, Difco, USA) and incubated at 25˚C for fungal strain germination and isolation. Isolated fungal strains (numbered from #1 to #92 in this study) were stored at −20˚C for follow-up screening for potential cellulases.

Each of the isolated fungal strains or a benchmark strain of Trichoderma reesei from culture stocks were individually recovered by culturing on PDA plates at pH ~ 5.6 and room temperature ~ 27˚C; and then, a ~5 mm³ 7-day-old mycelia block of each strain was re-cultured at pH ~ 5.3 and ~27˚C on modified MR plate (Mandel- Reese medium without peptone, MR-P) with 2% α-cellulose (each liter had 1.4 g (NH₄)₂SO₄, 0.8 mL of 6 M urea, 2 g KH₂PO₄, 0.34 g CaCl₂, 0.3 g MgSO₄∙7H₂O, 0.00005 g FeSO₄∙7H₂O, 0.000016 g MnSO₄∙7H₂O, 0.000014 g ZnSO₄∙7H₂O, and 0.00002 g CoCl₂∙6H₂O, and 20 g agar). Consequently, 6 of ~5 mm³ 7-day-old mycelia blocks from the MR-P plate were transferred into 100-mL MR-P liquid medium (duplicate n = 2 for environmental isolates and n = 6 for T. reesei) and cultured for another 7 days by shaking at 150-rpm to harvest crude enzyme (secretome) from each of isolated fungal strains and the benchmark strain T. reesei for following cellulolytic activity tests and enzyme characteristics assessments.

The crude enzymes (or secretomes) produced from fungal strains were tested for cellulolytic activities using pNPG (5mM), CMC (1%, w/v), and avicel (1%, w/v) as substrates in order to screen potential cellulases. Additionally, FPase activity was extra conducted for potential exoglucanase using a piece of 3 mg filter paper as substrate (6mm-diameter circle shape cut by paper puncher) in order to compare with other previous studies. The enzymatic pNPGase activity was determined based on amount of p-nitrophenol (pNP) released from 5 mM pNPG after 5-min reaction at 55˚C and pH 5 by measuring absorbance at a wavelength of 405 nm in a microtiter plate reader (SpectraMax M5e, Molecular Devices, USA). Additionally, the CMCase, avicelase, and FPase activities were assessed by measuring reducing sugar concentration at a wavelength of 550 nm using the di-nitro- salicylic acid (DNS) method after reacting at 55˚C and pH 5 for 10 minutes (for CMCase) or 60 minutes (for avicelase and FPase) in 50 mM sodium acetate buffer. All the cellulolytic activities were represented as U/mL (μmol of product, i.e., pNP or reducing sugar, produced within 1 minute per mL of enzyme).

Optimal cultivation conditions for the potential fungal strains (i.e., strains #58, #60, and #71) (original sample names: 13-5-2-2, 13-6-2A, and THU4) were investigated by culturing the fungal strains at two temperature levels of 25˚C or 30˚C (using thermal-control water bath) and three pH values of 5, 6, or 7; and, followed by avicelase activity assessments for the fungal secretomes at six temperature levels (i.e., 30˚C, 40˚C, 50˚C, 60˚C, 70˚C, and 80˚C) and eight pH values (i.e., pH 2, 3, 4, 5, 6, 7, 8, and 9). The avicelase activities were represented as relative activity (%) by comparing each activity with the highest activity measured throughout the assessments and used to determine the optimal cultivation conditions and the optimal avicelase reaction conditions.

In addition, pH stability and thermostability for the most potential crude enzyme were also resolved by incubating the fungal secretome at eight different pH levels (pH 2, 3, 4, 5, 6, 7, 8, and 9) or six temperature levels (30˚C, 40˚C, 50˚C, 60˚C, 70˚C, and 80˚C) for four hours and followed by avicelase activity test under optimal reaction conditions. All measured avicelase activities were also converted to relative activity (%) by deviding each activity with the highest activity measured throughout the assessments for assessing enzyme stability.

Genomic DNA of the potential fugal strain #58 was purified from 6-day cultured fungal mycelia using a commercial extraction kit (Genomic DNA Extraction Kit, Fast ID, Genetic ID NA, IA, USA). Identification of the fungal strain #58 was conducted by comparing gene sequences of internal transcribed spacer (ITS) region that was PCR-amplified using ITS1 forward primer (TCC GTA GGT GAA CCT GCG G) and ITS4 reverse primer (TCC TCC GCT TAT TGA TAT GC). Each 25-μl PCR reaction contained 12.5 μl of Master Mix (EmeraldAmp^® MAX HS PCR Master Mix, TaKaRa, Shiga, JP), 1 μl of each 10 μM forward and reverser primers (final concentrations were 400 nM), 8.5 μl of PCR-grade water, and 2 μl of the fungal DNA extract. A PCR reaction were run in a thermal cycler (MyCycler™, Bio-Rad, Hercules, CS, USA) based on the following program: 5 min at 94˚C; followed by 35 cycles at 94˚C for 1 min, 50˚C for 1 min, 72˚C for 1 min; and finally 10 min at 72˚C.

PCR amplicon was purified from agarose gel and sent for sequencing at MissionBio Co (Taipei, TW). Sequencing result was compared with NCBI database to determine fugal genus and species. The ITS fragments of the fungal strain #58, the first matched fungal species, and some closely related fungal strains were aligned by ClustalX method [20] and phylogenetic trees were constructed by neighbor-joining method and displayed with TreeView program [21] .

Crude enzyme concentration was determined based on Bio-Rad Protein Assay using BSA (bovine serum albumin) as calibration standards (Bio-Rad, USA). For better visualization on PAGE (polyacrylamide gel electrophoresis), the crude enzyme was further concentrated by lyophilisation for 4 hours and followed by dialyses for 3 hours to eliminate salts and metals in the MR culture medium using 3500 MWCO SnakeSkin™ Dialysis Tubing (Thermo Fisher Scientific, Life Technologies, Waltham, MA, USA).

Native PAGE and SDS (sodium dodecylsulfate) PAGE were performed based on the protocols suggested by Hoefer’s Protein Electrophoresis Applications Guide [22] . Briefly, the concentrated crude enzyme was directly loaded in a 10% native-PAGE or first denatured by heating for 3 minutes at 100˚C in a 2× protein treatment buffer (containing 0.125 M Tris-HCl, 4% SDS, 20% glycerol, 0.2 M DTT, and 0.02% bromophenol blue) and then loaded in a 10% SDS-PAGE gels to conduct electrophoresis (90 V for 30 minutes and then 140 V for 90 minutes). Followed by electrophoresis, the gel was stained with Coomassie blue R250 to view the native-PAGE or SDS-PAGE images. For zymographic assessment, the native-PAGE gel was further merged in a MUC solution (pH 5) and incubated at 70˚C and 20 rpm for 30 minutes and subsequently observed under UV light. Exoglucanase activity was visible as clear light bands against a black background.

The bands of target exoglucanase observed on the SDS-PAGE were cut and purified from gel and analyzed for amino acid sequences at a Proteomics Core Laboratory (located in the Agricultural Biotechnology Center of Academia Sinica, Taipei, Taiwan) using Applied Biosystems Procise^® Protein Sequencer model 494 (Applied Biosystems, USA). The sequenced amino acid fragments were compared with sequences from the NCBI protein database using a program of Protein BLAST.

Additionally, proportion of the target exoglucanase within the crude enzyme produced from the fungal strain #58 was estimated by using another 10% SDS-PAGE with different amounts of BSA as standards (i.e., 0.5, 1, 2, 3, 5, and 10 μg BSA). A calibration curve was constructed based on correlations between protein concentrations and image intensities of protein on the PAGE; and, was used to calculate the quantity of the target exoglucanase. Specific FPase and avicelase activities were determined afterward according to the exoglucanase concentration evaluated from the above-mentioned approach.

During 7-day cultivation for the fungal strains isolated from environments together with the benchmark strain T. reesei, avicelase activities of the fungal-secreted crude enzymes (for indicating exo-glucanase titers) increased gradually and reached highest levels mostly around 6 days (Figure 1). CMCase and pNPGase activities also reached highest levels after 5 - 7 days cultivation (time course activity profile data not showed). Among the 92

environmental isolates, most of them had highest avicelase activities less than 0.1 U/mL while only some fungal strain such as #58, #59, #60, and #63 showed more significant avicelase activities (>0.1 U/mL) as illustrated in Figure 1. The avicelase activities for fungal strain #58 increased more significantly (up to about 0.106 U/mL) than other fungal strains only after about 1 day cultivation and reached a highest level of 0.209 U/mL at 6^th-day making it one of the top candidates for finding novel potential exoglucanase.

The highest avicelase, CMCase, and pNPGase activities for each fungal strain during 7-day cultivation were selected for further comparisons to screen potential cellulases (Figure 2). Most of the fungal strains had notable CMCase (endo-glucanase) activities and few of them even had higher CMCase activity than that for T. reesei (Figure 2(b)). Besides, few fungal strains had pNPGase (beta-glucosidase) activities (Figure 2(c)). In addition, six out of 92 fungal strains had avicelase activities higher than 0.1 U/mL (i.e., strains # 26, #58, #59, #60, #63, and #71) as T. reesei had an avicelase activity of 0.27 ± 0.02 U/mL (Figure 2(a)). Beside the results from tube activity assay mentioned above, MUC zymogram was also used to preliminarily screen exoglucanase activity for these six strains. MUC zymogram results showed only strains #58, #60, and #71 had significant MUCase activities (images showed in Figure 2(a)). Therefore, the fungal strain #58, #60, and #71 were selected for further assessing their optimal culturing conditions and optimal avicelase reaction conditions.

Fungal strain #58 grew at a pH of 5 and a temperature of 25˚C exhibited better avicelase activity than other culturing conditions (Figure 3). Higher pH values at 6 or 7 and higher temperature at 30˚C resulted in less avicelase activity. In addition, optimal enzyme reaction conditions for secretome of the fungal strain #58 (produced at pH 5 and 25˚C) were in a pH range of 4 - 6 and a temperature range of 60˚C - 80˚C (Figure 3(a)). Furthermore,

the exo-glucanase activity for the fungal strain #58 remained more stable at acid conditions (pH 3 - 5) than at base conditions (Figure 4(a)); and, could stay rather stable over a wide temperature range even at high temperature of 70˚C - 80˚C (Figure 4(b)). By considering enzyme stability, optimal avicelase reaction conditions were lastly determined as a pH of 5 and a temperature of 70˚C for the secretome of strain #58.

During 7-day cultivations for the 92 isolated fungal strains, the culturing pH was around 5.3 for the MR-P medium and incubation temperature was round 27˚C for room temperature of the laboratory with air conditioner. These culturing conditions were close to those culture conditions determined for the fungal strain #58 (pH 5 and 25˚C); and, preliminarily screening results indicated the strain #58 was the most competitive candidate for exploring potential exo-glucanase. The avicelase activity gained during screening (0.209 U/mL) was even higher than that at pH 5 and 25˚C (0.145 U/mL) for optimal culturing conditions assessment. Furthermore, the screening conditions might not suitable for other fungal strains as the strains #60 and #71 had better avicelase activities when cultured at 25˚C and at pH 7 and 6, respectively (data not showed). Nevertheless, the highest avicelase activities for strain #60 (relative activity 50.3% under optimal culturing and enzyme reaction conditions) and #71 (94.2%) were still lower than that for the stain #58 (100%). Hence, only the results for the strain #58 were showed in the Figure 3. When the fungal strain #58 was cultured at 30˚C, the avicelase activities of harvested secretomes were typically less than 60% of the highest avicelase activities observed at 25˚C and pH5 suggesting lower temperature around 25-27˚C and pH about 5-5.3 were more suitable for exo-glucanase production from the strain #58 than at 30˚C and pH 6-7.

Interestingly, as the fungal strain #58 was cultured at 30˚C, optimal avicelase reaction conditions of the crude enzyme appeared to be shifted from pH 6 and 70˚C (Figure 3(a) and Figure 3(b)) to pH 4 and 60˚C (Figure 3(d) and Figure 3(e)). Besides, when it was cultured at pH 7, the optimal enzyme reaction conditions (i.e., pH 7 - 8 and 30˚C - 50˚C, Figure 3(c) and Figure 3(f)) were different from those for the secretomes harvested at pH 5 and 6 (i.e., pH 4 - 6 and 60˚C - 70˚C, Figure 3(a), Figure 3(b), Figure 3(d), and Figure 3(e)). The results indicated that various types of enzymes (e.g., isozymes) with diverse characteristics might produce from the strain #58 when it was cultured under different pH and temperature conditions. A previous study had suggested that productions of lignolytic enzymes from a basidiomycetous fungal strain were dissimilar under different culture conditions [23] ; and, another report had indicated that Aspergillus fumigates Z5 produced different proteins after

cultivation with various carbon sources (i.e., glucose, avicel, and rice straw) [24] . It was surmised some cellulolytic enzymes might not be produced or might just be produced at very low level under inappropriate culturing conditions.

Given that some other common cellulase-secreting fungi (e.g., Aspergillus sp., Fusarium sp., Penicillium sp., and Trichoderma sp.) typically having optimal culturing pH level of about 5 and optimal temperature of around 28˚C - 30˚C, the fungal strain #58 isolated in this study seemed more suitable to be cultured under low pH of 5 as well but at lower temperature of 25˚C - 27˚C. Besides, crude enzyme harvested from the strain #58 had better avicelase activities when reacted at pH 5 that was similar to most of other exo-glucanases (Table 1); but, had an optimal reaction temperature of 70˚C which was much higher than typical reaction temperature of 50˚C - 60˚C for other exo-glucanases. In addition, the strain #58 secretome remained stable at high temperatures of 60˚C - 80˚C for at least four hours which was seldom found for many of general cellulases from fungi. Some of these unique features inferred that the fungal strain #58 might different from the common cellulase-secreting fungi and some new exo-glucanases might be able to be found from this potential fungal strain.

The internal transcribed spacer (ITS) region of the fungal strain #58 was PCR amplified and further sequenced for fungal type identification. Sequencing result of the ITS fragment indicated that the fungal strain #58 was closely matched with a yeast Meyerozyma caribbica (96% similarity, NCBI accession number HG970748) in the Saccharomycetes class of Ascomycota phylum and formed a noticeable group together with Candia sp. and Pichia sp. in a phylogenetic tree (Figure 5).

The Meyerozyma sp. had been identified in yeast population of olive mill wastes [25] and had been found capable of improving maize productivity and minimizing requisite chemical fertilization [26] . Besides, a yeast strain of Meyerozyma caribbica MG20W had been isolated from plant rhizosphere soil and its complete genome had recently been sequenced [27] . These previous studies well supported that Meyerozyma sp. could ubiquitously present in some environments and might be able to secrete glycosidase, hydrolase, or specific cellulase.

Currently, a beta-1,3 exo-glucanase had been reported for the yeast Meyerozyma caribbica [28] which was belong to the glycosyl hydrolase (GH) 5 family (EC 3.2.1.58). But, no beta-1,4 exo-glucanase has been discovered from this yeast up to the best of our understanding. The avicelase activity observed for the secretomes produced from the fungal strain #58 isolated in this study indicated this Meyerozyma sp. yeast might capable of secreting beta-1,4 exo-glucanase.

The target exoglucanase of the Meyerozyma sp. strain #58 had a molecular size of about 70 kDa based on the SDS-PAGE analyses for the enzyme purified from bands observed and cut from the native-PAGE (Figure 6(a) and Figure 6(c)). Besides, the target exoglucanase showed notable activity on the MUC zymogram like a commercial enzyme C1.5 L did (Figure 6(b)). It appeared that specific activity of C1.5L was higher than that of the

strain #58 while C1.5L required smaller amount of enzyme than the strain #58 to exhibit strong image intensity like observed on the native-PAGE and MUC-zymogram (Figure 6(a) and Figure 6(b)).

The target exoglucanase of the Meyerozyma sp. strain #58 was further extracted from the SDS-PAGE and subsequently analyzed for the amino acid sequences using LC/MS/MS. Two amino acid fragments from sequencing results (i.e., YSGTCDPDGCDFNPYR and AGAKYGTGYCDSQCPR) were identical with 1,4-beta- D-glucan cellobiohydrolases (CBHI) from Aspergillus aculeatus (NCBI accession number O59843) [29] . Two amino acid fragments revealed from LC/MS/MS demonstrated the target Mc-CBHI was beta-1,4 exo-glucanase belonged to glycosyl hydrolase family 7 (GH7). The two fragments were also extensively found in some cellobiohydrolases of other strains including Aspergillus parasiticus SU-1 (KJK66765.1), Acytostelium subglobosum LB1 (GAM28417.1), Chaetomium senegalense (CDF76452), Lophodermium sp. ZY-2014 (AIJ50313.1), Penicillium roqueforti FM164 (CDM37977.1), or Rhizopus stolonifer var. reflexus (AHI87690.1) etc. within the NCBI database. These results demonstrated that the target exoglucanase of the Meyerozyma sp. strain #58 belonged to beta-1,4 exo-glucanase (or 1,4-beta-D-glucan cellobiohydrolases, CBHI) (EC 3.2.1.91) and was lately named Mc-CBHI.

The proportion of the Mc-CBHI within the crude enzyme produced from the strain #58 was further determined by comparing band intensity of the Mc-CBHI on the SDS-PAGE with those for known amount of BSA. Image intensities for different amount of BSA were well correlated with the quantity of BSA (Figure 6(e)) and the target exoglucanase observed on the SDS-PAGE was estimated to be about 1.3 μg (or 1.66 mg/mL) which was approximately 3.6% of total crude enzyme (i.e., 36 μg on SDS-PAGE or 45.93 mg/mL) harvested from the Meyerozyma sp. strain #58.

The specific FPase and avicelase activities of the Mc-CBHI became 0.179 U/mg and 0.126 U/mg calculated from the volumetric activity of 0.376 U/mL and 0.209 U/mL (Table 2); which were close to those for T. reesei CDU-11 (i.e., 0.25 ± 0.05 and 0.11 ± 0.03 U/mg, respectively) and were about 40% - 50% of a commercial cellulase Accellerase-1000 (i.e., 0.44 ± 0.01 and 0.25 ± 0.01 U/mg) [30] . These results indicated that of the newly-found Mc-CBHI was potentially competitive. If enzyme quality and quantity could be further improved by optimizing fermentation conditions or enzyme purification, the Mc-CBHI could be considered as a potential exoglucanase resource for related cellulolytic industrials.