Requalification of a Brazilian Trichoderma Collection and Screening of Its Capability to Decolourise Real Textile Effluent

Water contamination with large amounts of industrial textile coloured effluents is an environmental concern. For the treatment of textile effluents, white-rot fungi have received extensive attention due to their powerful capability to produce oxidative (e.g., ligninolytic) enzymes. In addition, other groups of fungi, such as species of Aspergillus and Trichoderma, have also been used for textile effluents treatment. The main aim of the present study was to requalify a Brazilian Trichoderma culture collection of 51 Trichoderma strains, isolated from different sources in Brazil and preserved in the oldest Latin-American Fungal Service Culture Collection, The Micoteca URM WDCM 804 (Recife, Brazil). Fungal isolates were re-identified through a polyphasic approach including macro- and micro-morphology and molecular biology, and screened for their capability to decolourise real effluents collected directly from storage tanks of a textile manufacture. Trichoderma atroviride URM 4950 presented the best performance on the dye decolourisation in real textile effluent and can be considered in a scale-up process at industrial level. Overall, the potential of Trichoderma strains in decolourising real textile dye present in textile effluent and the production of the oxidative enzymes Lac, LiP and MnP was demonstrated. Fungal strains are available in the collection e-catalogue to be further explored from the biotechnological point of view.


Introduction
Textile industries waste large water quantities that in some cases return to environment as untreated or incorrectly treated effluents [1]. Due to the presence of non-fixed dyes during the dyeing process, these wastewaters are highly coloured [2,3]. Water contamination with large amounts of coloured effluents is a serious environmental problem.
There are mainly three types of treatment for textile effluents: physical, chemical and biological. In recent years, relevant methods based on biological textile effluents degradation have been developed using bacteria and fungi in aerobic and anaerobic processes [1,2,[4][5][6][7][8][9][10]. A recent review on current technologies for biological treatment of textile wastewater discusses in deep these bioprocesses [11]. The use of oxidative enzymes, mainly Lac and peroxidases, in textile industry has increased quickly. It has been due to both: (a) the ability of these oxidative enzymes to bleach textiles; and (b) the biological treatments involving these enzymes seem to be an attractive solution mainly because most existing treatments (e.g., coagulation/flocculation, adsorption, ion exchange, and electrochemical methods) of dye wastewater utilize ineffective and uneconomical processes [12]. In addition, other biomimetic oxidative bleaches, like metalloporphyrins, are very promising catalysts for synthetic textile dyes. However, they also show some limitations and drawbacks [13].
For the treatment of textile effluents, white-rot fungi have been extensive assessed due to their powerful capability to produce oxidative, including ligninolytic (e.g., laccase (Lac), lignin peroxidase (LiP) and manganese peroxidase (MnP)) enzymes [6]. Other groups of fungi have also been evaluated for textile effluents treatment (e.g., species of ascomycetes like Aspergillus and Trichoderma) [6,11]. In addition, although the oxidative system is frequently produced during fungal secondary metabolism, different microorganisms produce different enzymes depending on cultivation conditions [4][5][6]14].
Trichoderma is a very genetically diverse genus with interesting features, which make some of its species of agricultural and industrial interest. Species of Trichoderma are widely used in the production of enzymes for both pulp treatment and food production [15]. In addition, there are also different reports showing the potential of these fungi to remediate the pollution of soil and water [14,[16][17][18].
All these features make Trichoderma an important fungal taxonomic group for biotechnological application. As a matter of consequence, for microbial culture collections that have Trichoderma in their catalogues it is important to have these assets identified and characterised according to the current state-of-the-art. Considering this, the main aim of the present study was to requalify a Brazilian Trichoderma culture collection of 51 Trichoderma strains, isolated from different sources in Brazil and preserved in the oldest Latin-American Fungal Service Culture Collection, The Micoteca URM WDCM 804 (Recife, Brazil). The requalification work was based on the fungal identification through a polyphasic approach including macro-and micro-morphology, molecular biology and screening their capability to decolourise real effluents collected directly from storage tanks of a textile manufacture.

Microorganisms and Culture Conditions
Fifty-one Trichoderma strains previously isolated from different substrates of geographic regions of Brazil (see Table 1

Morphological Fungal Identification
In order to confirm the taxonomical identification of the 51 Trichoderma strains, macro-and micro-morphological analyses were performed (see Table 1). For macro-morphological analyses, the strains were grown on PDA plates at 30, 35 and 40 • C, and on Synthetic Low-Nutrient Agar plates (SNA, for 1 L water: 1.0 g KH 2 PO 4 , 1.0 g KNO 3 , 0.5 g MgSO 4 ·7H 2 O, 0.5 g KCl, 0.2 g glucose, 0.2 g sucrose and 20 g agar) at 35 • C, in the dark without humidity control and under aerobic conditions.
For micro-morphological analysis, strains were inoculated on Cornmeal Agar Medium supplemented with dextrose plates (CMD, for 1 L water: 30 g cornmeal, 20 g dextrose and 20 g agar). Fungi were incubated at 25 • C for 3 days in the same conditions described above. Afterwards, 3 of 16 6 mm diameter plugs of each CMD fungal culture were inoculated on PDA plates and grown for 4 days in the same conditions but with a regime of 12:12 h light:dark.
Conidia and phialides, presence of chlamydospores and sterile hyphae were observed using 90% lactic acid slides on a light microscopy. Final morphological identification followed the taxonomic keys and guides available for Trichoderma [19,20].
The thermal cycling parameters consisted of initial denaturation at 94 • C for 2 min followed by 30 cycles of denaturation at 94 • C for 30 s, primer annealing at 55 • C for 30 s, primer extension at 72 • C for 1 min and a final extension for 10 min at 72 • C [20,21]. Amplification products were visualised on 1% agarose gel electrophoresis, stained with 0.5 µg/mL of GelGreen TM under UV light and then photographed and purified using Fermentas ® kit (Cambridge, UK). The amplified DNA fragments were sequenced in the Sequencing Platform of the Central Laboratory of the Centre of Biological Sciences, Federal University of Pernambuco, Recife, Brazil.
Electropherograms were analysed and edited in the software Staden Package 2.0. Sequences obtained were used to search the most similar sequences deposited in the GenBank by using the Basic Local Alignment Search Tool (BLAST). The experimental sequences of ITS1-5.8S-ITS2 and tef1 gene regions were aligned and edited together with the retrieved from the database using the software MEGA 5.0. The phylogenetic tree was constructed using the Neighbour Joining method and the Maximum Parsimony (MP), with 1000 bootstrap resampling using the software MEGA 5.0 [25].

Decolourisation Assay
Samples of textile effluent were collected directly from storage tanks in a textile manufacture at the municipality of Toritama, State of Pernambuco, Brazil. The effluent samples were kept at 4 • C and transferred to the Laboratory of Environmental and Quality Engineering (Department of Chemical Engineering, Federal University of Pernambuco, Recife, Brazil) to perform the physical-chemical analyses and decolourisation assay.
The physical-chemical analyses involved the Chemical Oxygen Demand (COD) by colorimetric method, Biochemical Oxygen Demand (BOD) by azide modification of a Winkler titration method, Colour (HAZEN scale), Turbidity (NTU scale), pH using the potentiometric measurement principle, and Sedimentable Solids (SD) by gravimetric through a 0.45 µm filter membrane, using the standard methods for the examination of water and wastewater [26].
The screening of textile dye decolourisation by Trichoderma strains was assessed by using 6 mm diameter plugs of fungal cultures previously grown at 30 • C for 7 days on PDA plates. Plugs were transferred to 125 mL Erlenmeyer flasks containing 30 mL of Malt Broth (MB, for 1 L water: 20 g malt extract, 1 g peptone and 20 g dextrose) supplemented with 1.9 × 10 −2 mM Indigo Carmine dye (Sigma-Aldrich, St. Louis, MO, USA).
Fungal cultures were incubated in the dark at 28 • C for 8 days. Fungal biomass was then removed by filtration through a 0.45 µm Millipore membrane and absorbance of filtrates was assessed in a Hitachi U-5100 spectrophotometer at 650 nm which is the maximum absorption of Indigo Carmine (ε 650 nm = 3.1 × 10 3 M −1 ·cm −1 ). The Indigo Carmine was used as standard and MB free of inoculum containing dye was used as abiotic control. The decolourisation percentage of Indigo Carmine dye was calculated according to Miranda et al. [27].
Previous to evaluate the decolourisation of the dye present on the real textile effluent, this was sterilised by autoclave at 121 • C during 15 min. Three different experiments were performed. In each case, 3 plugs of 6 mm diameter of fungal cultures grown for 7 days at 30 • C on PDA plates were added to 125 mL Erlenmeyer flasks containing 30 mL of: (1) textile effluent; (2) textile effluent supplemented with 0.5% (w/v) wheat bran, and (3) textile effluent supplemented with 0.05% (w/v) extract yeast. Cultures were incubated under static conditions, in the dark at 28 • C for 8 days. Afterwards, 2 mL aliquots were transferred into tubes and centrifuged at 1035 g for 15 min at 4 • C. Supernatants were analysed at 650 nm in a Hitachi U-5100 spectrophotometer.
The decolourisation results were submitted to analysis of variance and means compared by Friedman test at 5% probability in the ASSISTAT software [28].

Determination of Enzymatic Activities
In order to determine the activity of Lac, LiP and MnP, 6 mm diameter plugs of fungal cultures previously grown for 7 days at 30 • C on PDA plates were transferred into 125 mL Erlenmeyer flasks containing 50 mL of Basal Culture Medium (BCM, for 1 L water: 0.1 g glucose; 0.15 g yeast extract; 4.5 g wheat bran; and 0.05 g NH 4 Cl, pH 6.0). Fungal incubation occurred under static condition in the dark, at 28 • C for 8 days. The biomass was then separated by filtration through a 0.45 µm Millipore membrane. In order to determine the activity of each enzyme, absorbance of filtrates was assessed in a spectrophotometer Hitachi U-5100 and one unit (U) of enzyme was defined as the release of 1.0 µmol product formed per min under the assay conditions.
Overall, the enzymatic results were submitted to analysis of variance and means compared by Tukey test at 5% probability in the ASSISTAT software [28].

Phenotype Data Analysis
A cluster analysis was applied and a dendrogram of phenotype relatedness was constructed. The analysis was performed using the phenotype traits that showed at least high association with each Trichoderma species, such as: conidia shape, ornamentation, colour, length and width, and phialide length and base width, and presence of sterile hyphae and chlamydospore, and colony diameter on PDA at 30, 35 and 40 • C and SNA at 35 • C. The analysis was made using the hierarchical clustering with the complete linkage method and was performed with the statistical package jmp 8.0.2 for Macintosh (SAS Institute Inc., Cary, NC, USA).

Morphological and Molecular Identification
Fifty-one fungal strains obtained from the Micoteca URM culture collection and belonging to the genus Trichoderma were evaluated. Data obtained from the morphological analysis (Table 1 and Figure 1) show a clear difficulty to cluster the strains by morphological traits when a cluster analysis was used. It is well known that, due to homoplasy (i.e., convergent evolution that creates analogous traits with similar form but were not present in the last common ancestor) of morphological characters, it is often impossible to discriminate species of Trichoderma. Notwithstanding this, for other fungal taxonomic groups, the statistical grouping of phenotype traits such as conidia shape, ornamentation and dimensions (length and wide) are normally very good data for species identification when combined with other traits [31,32].
It is well known that the overall range of variation in conidial dimensions in Trichoderma is not great. However, related species can often be differentiated by slight but consistent differences in size.
The conidial surface appears smooth in most species in light microscope observations, although many species with apparently smooth-appearing conidia are delicately ornamented when examined by SEM. In most species, terminal phialides tend to be more elongate and narrower, and frequently more or less subulate. Chlamydospores are common in many species, although they tend to be uniformly globose or ellipsoidal, terminal and intercalary, smooth-walled, colourless, yellowish or greenish, and 6-15 um diameter in most species. Vegetative hyphae show few characters useful for identification.
This means that the identification based on the classical morphology techniques can be used as presumptive approach. However, to obtain a sound and reliable identification, molecular biology must be performed [33]. Consequently, in the present study the taxonomic information was revised throughout molecular biology analysis based on the 500 and 600 base pairs fragments obtained from the ITS1-5.8S-ITS2 rDNA and tef1 gene regions, respectively.
Based on the former identification available in the records of the Micoteca URM culture collection, the 51 fungal isolates were distributed into six Trichoderma species, as shown in the Table 2. Based on the molecular biology analysis presented herein, it was possible to separate them into 11 different species: Trichoderma asperelloides (8 strains), T. asperellum (4), T. atroviride (6) Table 1.  Table 1.  indicates nucleotide substitutions per site. Hypomyces subiculosus (EU280093 ITS sequence) was used as out group.

Screening of Indigo Carmine Decolourisation in Malt Broth
Among the 51 Trichoderma strains assessed for decolourisation of Indigo Carmine in MB, 18 of them presented over 70% decolourisation, which corresponds to 0.57 × 10 −2 mM or less residual dye (Table 3). From these fungal strains, three presented values greater than 90% decolourisation: T. atroviride URM 3270, URM 3735 and URM 6625, which presented dye decolourisation of 96.86%, 94.61% and 93.57%, respectively. These results corroborate those previous obtained by Adnan et al. [14] for a strain of T. atroviride isolated from tree bark in a Malaysian forest. In that case, authors obtained similar results with 91.1% decolourisation for the dye Reactive Black 5. In contrast, according to the data obtained in the present study, the strains T. asperelloides URM 3086, URM 3934 and URM 5007; T. asperellum URM 6656; T. hamatum URM 4722; T. koningiopsis URM 3606; and T. virens URM 4466 were not efficient in decolourising Indigo Carmine. These results were due to the fact strains belonging to these species require more time to be adapted on the polluted environment to become able to decolourise the effluent. Gajera et al. [34] obtained similar results when assessed the dye decolourisation capability of T. viride and T. harzanium strains isolated from contaminated rhizosphere with an industrial textile effluent in India. In this case, strains belonging to both species required a period longer than 12 days to decolourise completely reactive azo dyes, such as Red HE7B, Violet Reactive 5, Black Red-B, Dark Blue and Light Blue H2GP HEG.

Activity of Oxidase Enzymes
Fungal strains that presented decolourisation percentages of textile dye Indigo Carmine above 70% were selected for assessment of Lac, LiP and MnP activities (18 strains, Table 3). Based on the data obtained for Lac, LiP and MnP activities and after statistical analysis of averages, a significant difference between the means of LiP and MnP activities was observed (Table 3). In addition, the interaction among the averages of Lac, LiP and MnP indicated that all of the assessed fungal strains were better producers of LiP than the other two enzymes (Lac and MnP) ( Table 3).
In a previous study performed by Adnan et al. [14], T. atroviride showed no activity for LiP and MnP. In contrast, in the present study, it was observed that the best LiP activity was obtained for the assays performed with T. atroviride URM 3735, which presented a production of 1307.0 U·L −1 , followed by T. harzianum URM 2842 with an activity of 924.0 U·L −1 and T. virens URM 4996 with an activity of 844.0 U·L −1 . In addition, herein, it was observed that for the MnP, the highest activity was obtained for the strain T. erinaceum URM 3881 (875.3 U·L −1 ), followed by T. atroviride URM 4950 and URM 3735, 680.0 U·L −1 and 617.3 U·L −1 , respectively.
No significant difference between the averages of Lac activity was observed. However, the strain T. virens URM 4996 presented a Lac activity of 15.7 U·L −1 which is numerically greater than those observed for the other species assessed (Table 3). In a previous study, Adnan et al. [14] observed a Lac activity of 5.8 U·mL −1 after four days of experiment for a T. atroviride strain. Authors attested that Lac production was directly related to the degradation of textile dye Reactive Black 5. In the present work, the Lac production by the four fungal strains belonging to the species T. atroviride (URM 3270, 3735, 4950 and 3881) was assessed. After eight days of the experiment, Lac activities assessed were 1.3, 2.0, 8.3 and 2.3 U·L −1 , respectively (Table 3).
Overall, based on the results obtained herein for activities of Lac, LiP and MnP, the following six strains were selected as the best producer of these oxidases and then used in the screening assays of effluent decolourisation: Trichoderma asperellum URM 4926, T. atroviride URM 3735, URM 4950, T. erinaceum URM 3881, T. harzianum URM 2842 and T. virens 4996 (Table 3).

Treatment of Real Textile Effluent
The six above mentioned Trichoderma strains, which were the best enzymatic producers were then selected to decolourise real textile effluent which is characterised by high organic load since the COD and BOD parameters are higher than the normal accepted for inland surface water disposal (COD < 250 mg·L −1 and BOD < 30 mg·L −1 ) ( Table 4). Based on the data obtained for: (1) textile effluent; (2) textile effluent supplemented with 0.5% (w/v) wheat bran; and (3) textile effluent supplemented with 0.05% (w/v) extract yeast, for each strain, and after statistical analysis of averages, a significant difference between the means was observed (Figure 4). According to the results obtained, T. atroviride URM 4950, T. erinaceum URM 3881 and T. virens URM 4996 showed the best performance in decolourising of textile effluent (84.6%, 36.1% and 61.3%, respectively). In addition, T. asperellum URM 4926 (84.6%), T. erinaceum URM 3881(66%) and T. harzianum URM 2842 (60.6%) presented a good performance of decolourisation when textile effluent supplemented with extract yeast was used.
According to Saravanakumar and Kathiresan [17], a large decolourisation rate (89%) of Malachite Green dye after 10 days of incubation of T. harzianum TSK8 (JQ809340) was observed when this strain was grown in supplemented medium with 5.8 mg·L −1 yeast extract. In addition, Adnan et al. [14] evaluated the effect of different sources of carbon and nitrogen in the decolourisation of textile dye Reactive Black 5 by T. atroviride. Authors observed that yeast extract was a good source of nitrogen for growth T. atroviride and to stimulate it to decolourise Reactive Black 5 up to 91.1%. Furthermore, authors evaluated dye decolourisation by using ammonium nitrate and ammonium chloride and, in contrast, observed only a decolourisation rate of 29.1% and 71.7%, respectively. Based on the data obtained for: (1) textile effluent; (2) textile effluent supplemented with 0.5% (w/v) wheat bran; and (3) textile effluent supplemented with 0.05% (w/v) extract yeast, for each strain, and after statistical analysis of averages, a significant difference between the means was observed (Figure 4). According to the results obtained, T. atroviride URM 4950, T. erinaceum URM 3881 and T. virens URM 4996 showed the best performance in decolourising of textile effluent (84.6%, 36.1% and 61.3%, respectively). In addition, T. asperellum URM 4926 (84.6%), T. erinaceum URM 3881(66%) and T. harzianum URM 2842 (60.6%) presented a good performance of decolourisation when textile effluent supplemented with extract yeast was used.
According to Saravanakumar and Kathiresan [17], a large decolourisation rate (89%) of Malachite Green dye after 10 days of incubation of T. harzianum TSK8 (JQ809340) was observed when this strain was grown in supplemented medium with 5.8 mg·L −1 yeast extract. In addition, Adnan et al. [14] evaluated the effect of different sources of carbon and nitrogen in the decolourisation of textile dye Reactive Black 5 by T. atroviride. Authors observed that yeast extract was a good source of nitrogen for growth T. atroviride and to stimulate it to decolourise Reactive Black 5 up to 91.1%. Furthermore, authors evaluated dye decolourisation by using ammonium nitrate and ammonium chloride and, in contrast, observed only a decolourisation rate of 29.1% and 71.7%, respectively.

Conclusions
The Brazilian Trichoderma asset preserved at the Micoteca URM culture collection was reidentified and made available in the culture collection e-catalogue to be further explored from the biotechnologically point of view. In the specific case of dye decolourisation screening using real

Conclusions
The Brazilian Trichoderma asset preserved at the Micoteca URM culture collection was re-identified and made available in the culture collection e-catalogue to be further explored from the biotechnologically point of view. In the specific case of dye decolourisation screening using real textile effluent, the strain T. atroviride URM 4950 gave the best results and can be considered in a scale-up process at industrial level. This strain is herein indicated to be used as a major target for future studies related with the treatment of industrial textile effluent and bioremediation of environments contaminated with dyes. Overall, the data obtained suggest that the dye decolourisation in real textile effluent is not a one-step process. More than one oxidative enzyme can be involved in this complex process. The contribution of these enzymes for decolourisation of textile effluent differs from fungus to fungus. The potential of Trichoderma strains in decolourising real textile dye present in textile effluent and the production of the oxidative enzymes Lac, LiP and MnP were successfully demonstrated. Finally, in this study, the strains available through Micoteca URM culture collection are identified according the current state-of-the-art and better characterised to underpinning life sciences and biotechnology.