Trichoderma atroviride 102 C 1 : A promising mutant strain for the production of a β-glucosidase ,-xylosidase and-L-arabinofuranosidase activities using agroindustrial by-products

1 Departamento de Microbiologia Geral, Avenida Carlos Chagas Filho Universidade Federal do Rio de Janeiro (UFRJ), Centro de Ciências da Saúde (CCS), Instituto de Microbiologia Paulo de Góes, 373, Bloco I, Laboratório 055, CEP: 21941-902. Rio de Janeiro, RJ, Brazil. 2 Departamento de Bioquímica, Avenida Athos da Silveira Ramos, Universidade Federal do Rio de Janeiro (UFRJ), Centro de Ciências Matemática e da Natureza (CCMN), Instituto de Química, 149, Bloco A, sala 539, CEP: 21941-909. Rio de Janeiro, RJ, Brazil. 3 Departamento de Engenharia Bioquímica, Avenida Athos da Silveira Ramos, Universidade Federal do Rio de Janeiro (UFRJ), Centro de Tecnologia (CT), Escola de Química, 149, Bloco E, sala 108, CEP: 21941-909. Rio de Janeiro, RJ, Brazil.


INTRODUCTION
Plant cell walls are a source of renewable carbon present in nature as cellulose, hemicelluloses and lignin.Cellulose is an abundant linear polymer worldwide and is composed of glucose residues linked by -1,4-glucosidic bonds.Hemicellulose is a branched heteropolymer of pentoses and/or hexoses and various types of uronic acids residues linked by -1,4 / -1,3 glucosidic bonds.The term hollocellulose comprises the cellulose and hemicelluloses of plant cell wall linked together (Gottschalk et al., 2010;Zampieri et al., 2013).
Worldwide attention has focused on the major biotechnological uses of the carbohydrates in agroindustrial by-products, as biomass syrups that have sugars with five or six carbons (derived from xylan and glucan).They can be used as carbon sources in industrial fermentations producing antibiotics, industrial enzymes, and bulk chemicals, including ethanol.For these purposes, however, the polysaccharides in the biomass must first be hydrolyzed (Gottschalk et al., 2010;Shinozaki et al., 2015).
Trichoderma atroviride 676, isolated from Amazon forest soil, was capable to produce cellulases and endoxylanases when agro-industrial by-products were used as substrates (Grigorevski-Lima et al., 2013).After Silva et al. 1221 mutation procedures in this strain, a new mutant, T. atroviride 102C1, producer of high titers of endoglucanase, exoglucanase and endoxylanase was obtained (Oliveira et al., 2014).The objective of the present study was to investigate the production of some hollocellulolytic accessory enzymes of T. atroviride 102C1, including β-glucosidase, β-xylosidase and α-Larabinofuranosidase, using submerged fermentation and, as main substrates, sugarcane bagasse in natura and corn steep liquor.

Maintenance of microorganism
Mutant strain T. atroviride 102C1 was obtained from the wild strain T. atroviride 676 (Grigorevski-Lima et al., 2013), after successive exposures to U.V. radiation and nitrosoguanidine, according to Oliveira et al. (2014).Spore suspensions were prepared according to Hopwood et al. (Hopwood et al., 1985), after cultivation at 28°C for 7 days in potato dextrose agar medium (Hanada et al., 2002).Spores were maintained in 20% (v/v) glycerol at −20°C.Concentration of the spore suspension was determined in a Neubauer chamber.

Enzyme production
Enzyme production was performed in submerged fermentation in Erlenmeyer flasks (125 ml) containing 25 ml of a modified culture medium (Mandels and Weber, 1969), in g.L -1 : urea, 0. Response surface methodology (RSM) was used as a tool for the optimization of SCB and CSL concentrations (independent variables) in the range indicated in Table 1.Enzymes activities, glucosidase, -xylosidase and -L-arabinofuranosidase (U⋅mL −1 ) were the dependent variables.Medium start pH was adjusted to 4.8.The growth media were inoculated with 25 µl of a spore suspension (10 11 spores⋅mL −1 ) and incubated at 28°C, under agitation (200 rpm).Based on preliminary tests, a two days period was chosen.After this, the whole content of a shake flask was filtered through a glass microfiber filter (Whatman GF/A) in duplicate, and the culture supernatants obtained were used to determine the all enzymatic assays.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License determined by mixing 50 µl of the enzyme preparation to 100 µl of p-nitrophenyl-β-D-glucopyranoside or p-nitrophenyl-β-Dxilopyranoside (10 mM) in 200 µl of 100 mM sodium acetate buffer, pH 5.0 at 50°C for 10 min, respectively, supplemented with 650 µl of distilled water.ARF activity was determined by mixing 100 µl of the enzyme preparation to 100 µl of p-nitrophenyl-α-Larabinofuranoside (10 mM) in 200 µl of 100 mM sodium acetate buffer, pH 5.0 at 50°C for 10 min, supplemented with 600 µl of distilled water.After incubation time, the enzyme reaction was stopped by addition of 500 µl of Na2CO3 (1.0 M), pH 10.0 and reading in spectrophotometer at 420 nm.One unit of BGU, BXU and ARF (IU) corresponded to formation of 1.0 µmol of pnitrophenol at 50°C per minute.The results were analyzed using Statistica Statsoft 7.0®.

Determination of physico-chemical properties of enzymes: pH, temperature, thermal stability and metal ions effect on enzymatic activities
Temperature and pH effect on BGU, BXU and ARF activities was investigated using standard assay methods at various temperatures and pH range, according to CCRD 2².In the 12 experiments which were carried out, the temperature ranged from 40 to 70°C and the pH values from 3.0 to 7.0 as shown in Table 3. Citrate buffer (50 mM) was used for pH 3.0, 3.6 and 5.0 and citrate-phosphate (50 mM) for pH 6.4 and 7.0.Statistical analysis was performed using the software Statistica Statsoft 7.0®.Temperature stability range was determined by incubating the crude enzyme at 50 and 70°C.Residual enzyme activity was determined at different time interval up to 8 h.Influence of sodium, calcium, potassium, manganese, and barium ions in the chloride form, and copper, magnesium, zinc, and iron in the sulfate form, on the BGU, BXU and ARF activities were performed by the addition of the relevant salts at 2 mM final concentration in the enzyme activity assay using the previously determined optimal conditions for pH and temperature.

Enzyme production
In this study the mutant strain T. atroviride 102C1 have produced some holocellulolytic accessory enzymes when using sugarcane bagasse (SCB) and corn steep liquor (CSL) as substrates in submerged fermentation.The use of RSM and CCRD tools for the optimization resulted in enzyme activities accumulation in the range of 26.11 to 55.38 U.mL −1 for β-glucosidase, 1.01 to 10.75 U.mL -1 for β-xylosidase and 70.30 to 159.24 U.mL −1 for α-Larabinofuranosidase after two days fermentation (Table 1).The fitted response surface for the production of the three enzymes is given in Figure 1 and, as can be seen, best conditions for each one, were different.For βglucosidase, the best enzyme production (55.38 U.mL -1 ) was observed when 2.5% (w/v) SCB and 1.26% (w/v) CSL were used.For β-xylosidase (10.92 U.mL -1 ) best concentrations were 1.5% (w/v) SCB and 1.1% (w/v) CSL, and for α-L-arabinofuranosidase (159.24U.mL -1 ), 3.5% (w/v) SCB and 0.3% (w/v) CSL.However, for studying the BGU, BXU and ARF concomitantly, the run 8 [2.5% (w/v) SCB and 1.26% (w/v) CSL] was considered as the best for further analysis.Under these conditions the enzyme activities were 55.38 U.mL -1 (BGU), 10.75 U.mL -1 (BXU) and 143.23 U.mL -1 (ARF).The relevant regression equations, resulting from the analysis of variance (ANOVA) (    The model  values of 28.32, 93.82 and 133.28 imply that the models are significant at a high confidence level.The probability value was also very low (<0.05) indicating the significance of the model.

Determination of physico-chemical properties of enzymes
With respect to pH and temperature effects, the maximum enzyme accumulation were, as expected, influenced by pH and temperature.According to CCRD, the analysis of resulting surface response plots revealed that maximal β-glucosidase relative activity (Figure 2A) occurred in pH range of 4.5-5.5 and temperature of 50-60°C, while the maximal β-xylosidase (Figure 2B) and α-L-arabinofuranosidase (Figure 2C) relative activities occurred in pH range of 3.0-4.0and temperature of 50-60°C / 40-50°C, respectively.So, characterization of the crude enzyme showed that the best values for enzyme production from T. atroviride 102C1 were: 55°C and pH 5.0 for BGU, 55°C and pH 3.0 for BXU, and 44°C and 3.6 for ARF (Table 3).These results indicate an acidic condition favoring all enzymes activities, whereas best temperature varied between 44 and 55°C (Table 3).
Considering the concomitant production of the enzymes, one should choose pH 5.0 and temperature of 55°C, for best results (assay number 10, Table 3).
The model was tested for adequacy by ANOVA ( Thermal stability constitutes also an important property when studying the industrial importance of an enzyme. The results of -glucosidase, β-xylosidase and α-Larabinofuranosidase thermal stability are shown in Figure 3.When the enzyme crude extract was incubated at 60°C, the relative activities of -glucosidase (Figure 3A) and β-xylosidase (Figure 3B) decreased around 50% within 30 min, however, a strong decreased (95%) was  observed for α-L-arabinofuranosidase.When tests were performed at 50°C the -glucosidase relative activity decreased 35% after 30 min and remained stable up to 8 h incubation, whereas for α-L-arabinofuranosidase a decrease of more than 80% after 1 h of incubation was observed (Figure 3C).Inversely -xylosidase relative activity was practically not influenced by this temperature, being completely stable even after 8 h incubation.
The effect of metal ions on enzymes activities from T. atroviride 102C1 was also studied (Table 5).In general the tested ions have caused a marked inhibition on enzyme relative activities.The presence of Ba 2+ , Fe 2+ , Cu 2+ , Ca 2+ Zn 2+ and Mn 2+ have totally inhibited the three enzymes.The exceptions were for ion K + , Na + and Mg +2 .

Ion K
+ had no effect on -xylosidase, but caused a decrease of 40 and 82% on -glucosidase and α-L- Trichoderma reesei and T. atroviride however, some studies using other fungal species have been described.Also, in general, differently from our research, the enzymes production has been studied in separate.
Considering the Trichoderma genus, a mutant strain of T. reesei RUT-C30 (ATCC 56765) was studied by Gottschalk et al. (2010).They described a very low titer of -glucosidase (0.15 U.mL -1 ) when the strain was grown in the presence of lactose, after 5 days fermentation.An interesting result was obtained when an heterologous expression of a β-glucosidase gene from Penicillium decumbens was inserted in T. reesei RUT-C30, the maximal activity observed being improved from 4.4 to 34.3 U.mL -1 , when using 2% (w/v) wheat bran and 3% (w/v) microcrystalline cellulose, after 7 days fermentation (Ma et al., 2011).As to T. atroviride, Kovács et al. (2008), developed a mutants strain, F-1505, which was considered the best mutant for β-glucosidase activity, endoglucanase activity and filter paper cellulase activity (FPase).The maximal activity for β-glucosidase (11.7 U.mL -1 ) was observed when pretreated sugarcane bagasse was used, after 3 days fermentation.Also, in previous studies, our group verified a very low βglucosidase production (0.17 U.mL -1 ) in sugarcane bagasse, after 4 days fermentation, by T. atroviride 676, the wild strain of mutant T. atroviride 102C1 (Grigorevski-Lima et al., 2013).The results here obtained, of 55.4 U.mL -1 of β-glucosidase for T. atroviride 102C1 is favorable.
Other fungal species have also been studied for βglucosidase production, some with very high titres, such as Aspergillus niger strain NII 08121 which was cultivate using 1.0% (w/v) wheat bran for 4 days, a maximal βglucosidase activity of 1,400.0U.mL -1 being observed (Singhania et al., 2011).Also Aliyah et al. (2017) observed a maximal β-glucosidase activity (91.67 and 85.01 U.mL -1 ) when A. niger was cultivate using sugarcane bagasse and corn cob for 6 days, respectively.However, Aspergillus and Penicillium species, are considered greater β-glucosidase producers (Zampieri et al. 2013).
Studies on -xylosidase production by Trichoderma strains such as Aspergillus, Penicillium and Talaromyces have shown very low results, compared to those here described, were values of 10.8 U.mL -1 were obtained.Trichoderma reesei RUT C30, for instance, when grown on cellulose 1% (w/v) for 7 days, have shown maximal enzyme activity of 0.25 U.mL -1 (Jiang et al., 2011).Trichoderma virens CTGxAviL, also grown on -cellulose + xylan beechwood, after 3 days, have produced activities of 0.38 U.mL -1 (Tarayre et al., 2015).Other fungal strains, as Aspergillus awamori (Paredes et al., 2015), for instance, have also shown very low results when grown on agro-industrial residues.
However, some studies are comparable with ours, such as those using Lichtheimia ramose by Garcia et al. (2015) which have shown a maximal β-xylosidase activity (11.57U.mL -1 ) in the presence of wheat bran, after 4 days fermentation.Ceratocystis fimbriata RM 35 when grown on wheat bran for 7 days was also able to produce 14.40 U.mL -1 of β-xylosidase (Martins et al., 2018).Even though, our results were obtained after 48 hrs, which can be considered a great advantage.Concerning AFR, as far as we are concerned, there are no studies in literature describing this enzyme activity for Trichoderma strains.However, some studies have shown a low production by other fungi.Penicillium janczewskii, for instance, was tested first with oat spelt xylan, during 7 days fermentation, producing 0.8 U.mL -1 , and then, using 1% (w/v) brewer's spent grain and 1% (w/v) orange waste after 10 days, with maximal values of 0.7 U.mL -1 (Temer et al., 2014).In the same way, production of ARF by A. awamori 2B.361 U2/1 was low, of 0.7 U.mL -1 , when using media containing wheat bran, after 5 fermentation days (Paredes et al., 2015).Other fungal strains, such as Acremonium zeae EA0802 and Talaromyces thermophilus, gave also very low values of α-Larabinofuranosidase, when grown in oat spelts xylan (0.045 U.mL -1 ) after 18 days (Almeida et al., 2011), and wheat bran (0.85 U.mL -1 ) after 100 h (Guerfali et al., 2011).Our strain, T. atroviride 102C1, was able to produce 143.23 U.mL -1 , which, in comparison, is a very high titre.
Data from literature have shown that, in general, BGU activities produced by different fungi were also acidic with optimal temperatures from 54 to 70°C.Trichoderma harzianum IOC-4038 showed as optimal conditions for enzyme relative activity pH 5.0 and 54°C (Castro et al., 2010).Bonfa et al. ( 2018) observed a maximal relative activity at pH 5.0 and 60ºC from thermophilic fungus Myceliophthora thermophila M.7.7. A. niger NII-08121 showed an acidophilic and thermophilic profile, with pH 4.8 and 70°C (Singhania et al., 2011).
Concerning BXU, it is common to observed in literature an acidophilic and thermophilic profile, with optima of pH and temperature around 4.0-5.0,and between 65-70°C, as observed for Aspergillus (Díaz-Malváez et al., 2013;Wakiyama et al., 2008) and Ceratocystis strains (Martins et al., 2018).Terrasan et al. (2011) observed the best condition for β-xylosidase when the enzymatic extract was incubated at 75°C and pH 5.0 from P. janczewskii.These results concerning optimal temperature are compatible with those showed by T. atroviride 102C1, which showed a maximal enzyme relative activity at 55°C.However, concerning pH, the best pH was more acidic (3.0), close to that observed by Knob and Carmona (2009) for Penicillium sclerotiorum, which have shown an even lower optimum pH, 2.5, and an optimum of temperature at 60°C.As for ARF data on literature are scarce.Enzyme relative activity from P. janczewskii was considered acidophilic and thermophilic, with best production at pH 4.0 and 60°C (Temer et al. 2014) whereas Guerfali et al. (2011) have found optimal ARF activity from Talaromyces thermophilus on more mesophilic conditions, pH, 6.0-7.0 at 55°C.In the present study, ARF from T. atroviride 102C1, was acidophilic for best pH, 3.6, and mesophilic for best temperature, 44°C.
There are some studies in literature which report inhibitory effects of some metal ions on β-glucosidases activities, mostly decreasing the enzyme relative activity by several degrees.For instance, Bonfa et al. ( 2018) observed a decreased of 30% for Na 2+ , 50% for Mn 2+ , 100% for Fe 2+ and 37% for Zn 2+ on BGU enzymatic relative activity from My. thermophila M.7.7, but a different result was observed for Ba 2+ and Mg 2+ , where no inhibition was detected.Some studies have reported the apparent activation of fungal BXU by Mn 2+ and Ca 2+ , suggesting that these ions activate and protect the active site (Yang et al., 2014;Pereira et al., 2015;Martins et al., 2018).Terrasan et al. (2011) 75%) inhibited the enzyme relative activity from Alicyclobacillus spp.A4.As one can seen, all these results from literature are compatible to the results obtained for T. atroviride 102C1.

Conclusion
As a conclusion of our study, we can say that the accessory enzymes here studied, produced by the mutant fungus T. atroviride 102C1, can be interesting for various industrial applications.It is important to stress that the three enzymes were obtained concomitantly, in only two days, in expressive amounts.Furthermore, the use of sugarcane bagasse and corn steep liquor, the main energy sources for microbial growth and enzymes production, allows for optimum use of these low cost agriculture residues.Activity at high temperatures is interesting, considering the fungus is mesophilic and can be very important for a future application in processes that are carried out at high temperatures.By this way, the environment will be highly favored due to the use of this system by the bioenergy plants for biofuel production, especially the production of bioethanol in Brazil, increasing sustainability and generating less environmental pollutants.

Table 2 .
Statistical ANOVA for the model of β-glucosidase, β-xylosidase and α-L-arabinofuranosidase activities by T. atroviride 102C1 at different concentrations of SCB and CSL values.

Table 4 )
. The model  values of 11.64 (BGU), 45.21 (BXU) and 14.68 (ARF) indicates that the model is significant at a high confidence level.The probability  value was also very low (<0.05) indicating the significance of the model.

Table 4 .
Statistical ANOVA for the model of β-glucosidase, β-xylosidase and α-L-arabinofuranosidase relative activities at different levels of pH and temperature values.