Induction of wheat straw delignification by Trametes species

Wheat straw is the major crop residue in European countries which makes it the most promising material for bioconversion into biofuels. However, cellulose and hemicellulose are protected with lignin, so delignification is an inevitable phase in lignocellulose processing. The organisms predominantly responsible for its degradation are white-rot fungi and among them Trametes species represent promising degraders due to a well-developed ligninolytic enzyme system. Although numerous studies have confirmed that low molecular weight compounds can induce the production and activity of ligninolytic enzymes it is not clear how this reflects on the extent of delignification. The aim of the study was to assess the capacity of p-anisidine and veratryl alcohol to induce the production and activity of Mn-oxidizing peroxidases and laccases, and wheat straw delignification by six Trametes species. Significant inter- and intraspecific variations in activity and features of these enzymes were found, as well as differences in the potential of lignocellulose degradation in the presence or absence of inducers. Differences in the catalytic properties of synthesized enzyme isoforms strongly affected lignin degradation. Apart from enhanced lignin degradation, the addition of p-anisidine could significantly improve the selectivity of wheat straw ligninolysis, which was especially evident for T. hirsuta strains.

. Activity of Mn-dependent peroxidase (a), Mn-independent peroxidase (b) and laccase (c) in Trametes spp. strains depending on treatment with inducers. Boxplots represent the median, interquartile range, and minimum to maximum range of enzyme activity level, n = 5. significant variability among other strains was noted. Veratryl alcohol significantly induced the activity of laccase in all strains and especially in T. versicolor BEOFB 321 (Fig. 1c), where the increase was highest compared with the control, reaching even 22.7-fold (16291.4 ± 405.9 U L −1 vs. 716.7 ± 121.1 U L −1 ) (Supplementary Table S1).
Two-way analysis of variance of MnP, MnIP and laccase activities showed that strain, inducer type, as well as their interaction were significant factors which affected the activities of these enzymes. The F statistic for MnP was F 32,132 = 38.45, P < 0.001; for MnIP was F 32,132 = 38.45, P < 0.001 and for laccase was F 32,132 = 38.45, P < 0.001.
The two inducers significantly affected total protein production which was reflected in specific activities of the enzymes studied (Fig. 2). Total protein production was higher in the presence of veratryl alcohol and the maximum was noted in T. pubescens BEOFB 330 and T. sualveolens HAI 300 (12.0 ± 1.6 mg mL −1 ) which had a 6-fold higher concentration than under control conditions (2.0 ± 0.1 mg mL −1 ). Under control conditions, the highest specific MnP activity was present in T. gibbosa BEOFB 311 (656.6 ± 5.6 U mg −1 ), slightly lower in T. sualveolens HAI 300 (432.2 ± 63.5 U mg −1 ), while in all other strains MnP activities were below 100.0 U mg −1 . The two inducers had no significant effect, except in the case of p-anisidine which showed a positive effect in T. gibbosa BEOFB 311 (431.9 ± 30.4 U mg −1 ) and T. multicolor HAI 540 and HAI 426 (230.6 ± 46.0 U mg −1 and 154.4 ± 13.4 U mg −1 , respectively).
MnIP was not expressed under control conditions in nearly all strains, except T. multicolor HAI 540, where the maximum activity was found (95.6 ± 3.9 U mg −1 ), T. versicolor BEOFB 322 (41.7 ± 7.5 U mg −1 ) and T. suaveolens HAI 300 (19.4 ± 4.2 U mg −1 ). p-anisidine stimulated specific MnIP activity only in T. multicolor HAI 540 (135.9 ± 21.8 U mg −1 ) which was about 30% higher than in the control, while in all other strains its effect was either absent or inhibitory, as in the cases of T. versicolor BEOFB 322 and T. suaveolens HAI 300. Veratryl alcohol showed similar effects, and expressed the enzyme only in the case of T. hirsuta BEOFB 301, where the specific activity was 60% higher than in the control (54.9 ± 10.3 U mg −1 ) (Fig. 2). In the control and in the presence of p-anisidine, specific laccase activity was almost absent, while veratryl alcohol had a stimulatory effect, especially in T. versicolor BEOFB 321, where the maximum was reached (95.0 ± 1.4 U mg −1 ), T. hirsuta BEOFB 30 and T. versicolor BEOFB 320 (about 90 U mg −1 ) (Fig. 2).
Significant inter-and intraspecific variations were found in the activities of Mn-oxidizing peroxidases and laccase within the genus Trametes in the presence and the absence of inducers. Significant interspecific differences were primarily present in MnP production. Namely, contrary to T. multicolor where this enzyme was the most active, the activity obtained in fermentation of wheat straw/(NH 4 ) 2 SO 4 /wheat bran medium with T. hirsuta was about 30-fold lower, which is in accordance with results reported by Erden et al. 17 for T. versicolor and T. hirsuta.
The effect of inducers on the activities of ligninolytic enzymes has been the subject of numerous studies [9][10][11]15,16,18 . Due to their chemical compositions, wheat straw and wheat bran were found to be excellent substrates for both conversions into fungal biomass and ligninolytic enzymes 8,[19][20][21] . Namely, wheat straw is rich in cellulose (33.7-47.5%), hemicellulose (25-32%) and lignin (11-23%), as well as in macro-and microelements necessary for enzyme synthesis, and wheat bran contains lignin, phenolic acids and minerals [22][23][24][25] . However, the absence or very low activities of Mn-oxidizing peroxidases and laccase in some cases can be explained by specific physiological characteristics of each strain and by factors which regulate the expression of peroxidase and laccase genes at the transcription level 26 . Thus, it is known that inducers can act as transcription factors affecting the synthesis and activity of ligninolytic enzymes. Galhaup and Haltrich 11 and Xavier et al. 15 demonstrated the stimulatory effects of p-anisidine on laccase activity in T. pubescens and veratryl alcohol in T. versicolor. However, these findings were not in accordance with results of Galhaup et al. 16 and Bakkiyaraj et al. 21 who reported their inhibitory effects. Positive effects of veratryl alcohol and other inducers on the synthesis of Mn-oxidizing peroxidases in numerous white-rot species were also demonstrated by Kapich et al. 13 and Hakala et al. 14 . According to these authors, this inducer stimulated the production of MnP during wheat straw fermentation with Phanerochaete chrysosporium and an asparagine-succinate medium with Physisporinus rivulosus regulating the transcription of MnP genes. Variations in isoenzyme profiles. After 19-days of solid-state fermentation of wheat straw/(NH 4 ) 2 SO 4 / wheat bran medium, strains of Trametes species synthesized one MnP isoform of pI around 4.1, except T. suaveolens HAI 300 where the pI was 4.4 (Fig. 3a). The presence of p-anisidine did not cause any change in MnP isoform profiles except in T. hirsuta BEOFB 301, in which it inhibited the synthesis of this enzyme (Fig. 3b). The previously-observed decrease of MnP activity in T. gibbosa BEOFB 311, T. hirsuta BEOFB 30, T. multicolour HAI 540, T. suaveolens HAI 300 and T. versicolor BEOFB 321 is therefore attributed to decreased enzyme production (Fig. 1a). In contrast, in T. pubescens BEOFB 330 there was no difference in enzyme production, while the level of MnP activity increased (Fig. 1a). This, it seems likely that p-anisidine acted as a chemical mediator and thus increased the activity of the same MnP isoform. The effect of veratryl alcohol was strain dependent (Fig. 3c). In T. gibbosa strains, T. hirsuta BEOFB 30, T. suaveolens HAI 300 and T. versicolor BEOFB 322 this inducer led to the synthesis of MnP isoforms with higher pI values (a maximum pI of around 5.6 was obtained in BEOFB 322), in T. pubescens BEOFB 330 veratryl alcohol stimulated the production of an additional isoform (pI 3.8) and completely inhibited the enzyme in T. hirsuta BEOFB 301. The induced isoform of pI 4.6 in T. hirsuta BEOFB 30 and T. suaveolens HAI 300 showed increased levels of enzyme activity, while at the same time less of it was produced (Fig. 1a), which implies that isoforms differ in their catalytic properties. The synthesis of an additional isoform in T. pubescens BEOFB 330 was followed by an increase in MnP activity. However, the role of veratryl alcohol as a chemical mediator cannot be ruled out.
MnIP production in the Trametes spp. strains reacted similarly to the presence or absence of the two inducers ( Fig. 4). Thus, all strains synthesized one MnIP isoform of pI 4.1 except T. hirsuta BEOFB 30 where one additional isoform of pI 4.3 was detected (Fig. 4a). In the case of p-anisidine, production of a pI 4.3 isoform was induced in T. suaveolens HAI 300 and enzyme synthesis was inhibited in T. hirsuta BEOFB 301 (Fig. 4b). The decrease of MnIP activity in strain HAI 300 could be explained by the lower catalytic properties of the newly synthesized isoform (pI 4.3), despite total protein production being increased in the presence of p-anisidine (Figs 1b and 2). Veratryl alcohol stimulated the synthesis of two additional MnIP isoforms in T. gibbosa strains (pI around 3.8 and 4.3), caused the production of an isoform with the highest pI in T. suaveolens HAI 300 (pI 4.8), and an isoform with a lower pI in T. hirsuta BEOFB 301 (pI around 3.8) (Fig. 4c). The increase of MnIP activity noted in the strains that switched isoform production in the presence of veratryl alcohol can be explained not only by increased enzyme production, but also by higher activities of the newly-synthesised isoforms (Figs 1b and 2).
Under control conditions, the picture obtained for laccases was more complex (Fig. 5a). Namely, one isoform of pI 4.2 was detected in all strains except in T. hirsuta BEOFB 30, which produced an isoform with a pI of 4.4, and in T. versicolor BEOFB 322 which synthesized three additional isoforms (pI 4.7; 5.6; 6.3). p-anisidine completely inhibited synthesis of the enzyme in T. hirsuta BEOFB 301 and T. versicolor BEOFB 322, while in T. suaveolens HAI 300 production of an additional isoform (pI 4.4) was stimulated (Fig. 5b). However, the effects of veratryl alcohol on laccase synthesis in the strains of Trametes species were rather different (Fig. 5c). In most strains, 3-5 new laccase isoforms were detected though some of them were observed as wide bands. Both strains of T. gibbosa and T. multicolor, as well as T. versicolor BEOFB 321 synthesized laccase isoforms of high pI values (6.4 and 6.7) in the presence of the inducer. The appearance of wide bands is likely to be the result of point changes in amino acid sequence or some post-translational modifications of laccase molecules. This effect is expected because laccase synthesis is controlled by families of closely-located genes originating from gene duplication [27][28][29] . Thus, isoforms with pI 6.4 and 6.7 indicate higher interspecific similarity between T. multicolor and T. versicolor BEOFB 321. As there were no functional changes, a point mutation could have been the cause of small differences in pI values of laccases encoded by gene families, and this would be reflected through the densely-packed protein bands on the gel (Fig. 5c). In this way, the coloured product of ABTS oxidation could easily have diffused subsequently to make the wide bands. This effect of a single point mutation on the isoelectric point has been theoretically and experimentally shown before for other proteins 30,31 .
Although laccase activities were not found spectrophotometrically in T. gibbosa strains, T. hirsuta BEOFB 301 and T. versicolor BEOFB 322, the enzyme was detected by isoelectric focusing in all strains cultivated in the absence of inducers and in the presence of veratryl alcohol, which could be explained either by trace amounts of synthesized enzyme or alternatively by low enzymatic activity of the isoforms produced.
The number of laccase isoforms was best-fitted with two factors in the GLM: treatment and strain (P values were < 0.001). Tukey's multiple post hoc tests showed a significant difference between veratryl alcohol treatment and the control (P < 0.001), and no difference between p-anisidine treatment and the control (P = 0.987). Only T. versicolor BEOFB 322 showed a significant difference, which was the same with all other strains (P < 0.01).
The level of enzymatic activity was not significantly correlated with all isoforms. The isoform present in the control group (base expression), pI 4.0-4.3, was the one with the weakest effect on the laccase activity detected (Spearman's correlation coefficient ρ = − 0.018, P = 0.02). On the other hand, isoforms synthesized in the treatment with veratryl alcohol (pI 4.3-4.5 and pI 6.7) strongly and positively affected the enzyme activity (P < 10 −10 ).

Variation in capacity of wheat straw depolymerisation. Proportions of lignin, hemicellulose and
cellulose in the wheat straw were 9.7%, 29.5% and 35.5%, respectively. Significant variation in the degradation of these polymers was found and rates of their degradation differed among species and strains and in some cases they were not correlated with the levels of enzyme activities (Fig. 6).
Capacities of all the species and strains to degrade cellulose were significantly lower in comparison with their capacities for hemicellulose and lignin depolymerisation. Under control conditions, the highest extent of cellulose degradation was noted in substrate fermentation with T. multicolor HAI 540 (40.2%) and lowest with T. pubescens BEOFB 330 (17.2%) (Supplementary Table S2). p-anisidine in the medium caused the weak induction of cellulose degradation in T. gibbosa BEOFB 310, T. multicolor HAI 426, T. pubescens BEOFB 330, T. suaveolens HAI 300, T. versicolor BEOFB 320 and BEOFB 321 and decreased degradation in T. gibbosa BEOFB 311, T. hirsuta stains, T. versicolor BEOFB 322 and T. multicolor HAI 540. However, the addition of veratryl alcohol to the medium increased the extent of degradation in all species and strains and the amount of degraded cellulose was higher by 2.5% (T. multicolor HAI 540) to 18.9% (T. pubescens BEOFB 330) (Fig. 6c). T. multicolor HAI 426 was the best degrader of cellulose (48.7%, in the presence of veratryl alcohol) and T. hirsuta BEOFB 301 the weakest one (14.7%, in the presence of p-anisidine).
Two-way analysis of variance of the lignin, hemicellulose and cellulose degradations demonstrated that the strain, inducer and their interaction were statistically significant factors which affected depolymerisation. The overall F statistic for lignin degradation was F 32,132 = 105.3, P < 0.001, for hemicellulose F 32,132 = 118.9, P < 0.001, and for cellulose F 32,132 = 329.7, P < 0.001. Tukey's post-hoc test revealed significant differences (P < 0.05) between strains in all treatments for each polymer degradation, except for the pair veratryl alcohol -p-anisidine in hemicellulose analysis.
Previous studies showed that the amount of degraded lignin during wheat straw fermentation with T. versicolor varied between 12.0% after 16 d of cultivation and 42.0% after 35 d 32,32 , which was significantly lower in comparison with our data. Lignin degradation was mainly attributed to Mn-oxidizing peroxidases as laccase synthesis was either absent or weakly induced by veratryl alcohol. Mn-oxidizing peroxidases have the main role during the initial phase of delignification as they generate Mn 3+ , while laccase has limited diffusion into the Scientific RepoRts | 6:26529 | DOI: 10.1038/srep26529 non-modified plant cell wall due to its molecule size. However, in numerous studies of Trametes spp. strains, the extent of delignification was not correlated with the level of enzyme activity. Thus, a high capacity of lignin degradation in the presence of p-anisidine characterised T. hirsuta BEOFB 30, where a low MnP activity and the absence of MnIP and laccase were noted, while T. multicolor BEOFB 540 was the weakest lignin degrader both under control conditions and in the presence of inducers, though activities of the enzymes were high. The significant degradation of hemicellulose (32.0-54.9%) and cellulose (14.7-48.7%) is an expected result of substrate fermentation with these white-rot fungal species and strains as they simultaneously degrade lignin and the other two polymers 34 .
The best-fitted minimal GLM for the efficiency of ligninolysis by these six Trametes species identified five main factors (strain, type of treatment, and individual activities of MnP, MnIP and laccase) and two significant two-way interactions among them ( Table 1). As our results shown, strains reacted with considerable variation to the inducers, so the resulting significance of strain, treatment and their interaction was not surprising.
Looking overall at the efficiency of ligninolysis, T. pubescens BEOFB 330 was the most selective lignin degrader under control conditions (rates of degraded lignin and cellulose were 44.1% vs. 17.2%). However, in the presence of p-anisidine, both T. hirsuta strains were the most selective in the sense that cellulose degradation was lower than 20% (rates of degraded lignin and cellulose were 56.0% vs. 17.4%, and 41.6% vs. 14.7%, respectively).

Conclusions
Our results clearly showed significant inter-and intraspecific variations in the activities and characteristics of Mn-oxidizing peroxidases and laccases, as well as the potential of lignocellulose degradation within the genus Trametes in the presence or absence of inducers. p-anizidine and veratryl alcohol can be considered as efficient inducers of these ligninolytic enzymes in Trametes species. Differences in the catalytic properties of enzyme isoforms is a factor which strongly affects lignin degradation and this study indicated the most active isoforms. Apart from enhanced lignin degradation, the addition of p-anisidine can significantly improve the selectivity of wheat straw ligninolysis which was especially demonstrated for T. hirsuta strains. This study demonstrates not only a method to optimize the system for the degradation of wheat straw but also clarifies the role of individual enzymes in lignin degradation by different Trametes species and strains.

Methods
Organisms and cultivation conditions. Cultures of six Trametes species with 11 strains were isolated from fruiting bodies collected from Serbia and Russia or obtained from the culture collection of the Institute of Evolution, University of Haifa, Israel (HAI), and maintained on malt agar medium in the culture collection of the Institute of Botany, Faculty of Biology, University of Belgrade (BEOFB) ( Table 2).
Solid-state cultivation was carried out at 25 °C in 100-mL flasks containing 1.8 g of wheat straw as the carbon source, 0.2 g of wheat bran as an organic nitrogen source and 10.0 mL of the modified synthetic medium (without glucose and with (NH 4 ) 2 SO 4 at a previously-defined optimum nitrogen concentration of 10.0 mM). The medium with dH 2 O instead of the modified synthetic medium was used as the control. The obtained inoculum (3.0 mL per flask) was used for inoculation. p-anisidine or veratryl alcohol were added to the medium 72 h after inoculation to final concentrations of 1.0 mM and 0.5%, respectively, to test their potentials to induce ligninolytic enzyme activity and enhance delignification of wheat straw.
Samples were harvested after 19 d of cultivation and further used for ligninolytic enzyme extraction, while the residues were used for determination of hemicellulose, cellulose and lignin contents.
Enzyme activity assay and determination of total protein content. Ligninolytic enzymes synthesised during incubation were extracted after solid-state fermentation of wheat straw/wheat bran medium by stirring samples with 50.0 mL of dH 2 O on a magnetic stirrer at 4 °C for 10 min. The extracts were centrifuged (4 °C, 3000 rpm, 15 min) and the supernatants obtained were used for determining Mn-oxidizing peroxidases and laccases activities spectrophotometrically (CECIL CE2501 (BioQuest)), as well as total protein content. Activities of Mn-oxidizing peroxidases (Mn-dependent peroxidase [EC 1.11.1.13; MnP] and Mn-independent peroxidase [EC 1.11.1.6; MnIP]) were determined with 3.0 mM phenol red (ε 610 = 22,000 M −1 cm −1 ) as a substrate, in a succinate buffer pH 4.5 (succinic acid disodium salt, albumin from bovine serum (BSA) and DL-lactic acid sodium salt). The reaction mixture (V tot = 1.0 mL) contained buffer, sample, 2.0 mM H 2 O 2 and phenol red, with or without 2.0 mM MnSO 4 (for MnP and MnIP, respectively). The reaction was stopped by adding 2.0 M NaOH. The reaction mixture using 2.0 M NaOH added before phenol red was used as the blank.
Enzymatic activity is expressed in U and 1 U is defined as the amount of enzyme that transforms 1.0 μ mol of substrate per min.
The amount of total proteins was determined according to the method of Bradford et al. 35 . The reaction mixture contained 5.0 mL of Bradford's reagent and 100.0 μ L of sample and absorbance was measured after 5 min of incubation at room temperature at 595 nm. The protein content, presented as mg mL −1 , was used for determining specific enzymatic activity (U mg −1 ). ; Na 2 B 4 O 7 × 10H 2 O, 6.81 g L −1 ; pH 6.9-7.1), 0.5 g Na 2 SO 3 and a few drops of 1-octanol was heated to boiling and then refluxed for an hour to remove soluble sugars, proteins, pectin, lipids and vitamins from the sample. The sample was filtered, washed three times with boiled water and twice with cold acetone, dried at 105 °C for 8 h and weighed as neutral detergent fibres (NDF). The samples were further treated with acidic detergent solution (ADS) (CTAB (20.0 g) dissolved in 1000.0 mL of 0.5 M H 2 SO 4 ; pH 6.9-7.1), heated to boiling, refluxed for an hour, filtered and washed with boiled water and a few times with acetone. Thereafter samples were dried at 105 °C overnight and weighed. Weight of acidic detergent fibres (ADF) was determined gravimetrically as the residue remaining after extraction. The hemicellulose content was expressed as ADF-NDF.

Electrophoresis.
Determination of cellulose and lignin contents. Acidic detergent fibres were used for determining cellulose and lignin contents. Lignin content was defined by the Klason or 72% H 2 SO 4 method 38 . 1.0 mL of 72.0% H 2 SO 4 was added for each 100.0 mg of sample. The mixture was incubated at 30 ± 0.5 °C in a water bath for an hour, stirred frequently and after that diluted using 28.0 mL of water for each 1.0 mL of acid. Secondary hydrolysis was performed in an autoclave at 120 °C for 1 h. The hot solution was filtered through a tared Gooch crucible and the Klason lignin residues were washed with hot water. The sample was dried at 105 °C till constant weight and lignin content (LC) is expressed as a percentage of that present in the original sample. Cellulose content is presented as ADF-LC.

Statistical analysis.
The assays were carried out in five replicates for each measurement occasion and results correspond to the mean ± standard error. Two-way analysis of variance (ANOVA) and Tukey's HSD post-hoc test were performed to test the significance of differences among mean values of enzyme activities and polymer degradation rates.
Numerical variation in the isoenzyme profiles, as well as in the efficiency of lignin degradation were analysed by generalized linear models (GLM). The difference between percentage of degraded lignin and hydrolyzed cellulose was taken as a measure of the efficiency of ligninolysis for all strains. The full GLM model for isoenzyme profile contained strain, treatment and their interaction as exploratory variables, while full GLM model for efficiency of ligninolysis contained strain, treatment, and activity of ligninolytic enzymes, as well as interactions among these factors. The minimal best fitting model was determined by backward selection, as the model with the lowest Akaike information criterion (AIC). All statistical analyses were performed in R software, ver. 3.2.1. 39 .