Detoxification of pulping black liquor with Pleurotus ostreatus or recombinant Pichia pastoris followed by CuO/TiO2/visible photocatalysis

Cellulose-pulping requires chemicals such as Cl2, ClO2, H2O2, and O2. The black liquor (BL) generated exhibits a high chemical oxygen demand (COD), five-day biochemical oxygen demand (BOD5), and chlorophenol content, along with an augmented colour and increased pH. BL is often discharged into water bodies, where it has a negative impact on the environment. Towards that end, laccases are of great interest for bioremediation, since they can degrade aromatic and non-aromatic compounds while reducing O2 to water instead of H2O2. As such, we evaluated Pleurotus ostreatus and Pichia pastoris (which produces rPOXA 1B laccase) in the treatment of synthetic BL (SBL) in an “in vitro” modified Kraft process followed by CuO/TiO2/visible light photocatalysis. Treating SBL with P. ostreatus viable biomass (VB) followed by CuO/TiO2/visible light photocatalysis resulted in 80.3% COD removal and 70.6% decolourisation. Toxic compounds such as 2-methylphenol, 4-methylphenol, and 2-methoxyphenol were eliminated. Post-treated SBL exhibited low phytotoxicity, as evidenced by a Lactuca sativa L seed germination index (GI) > 50%. Likewise, SBL treatment with P. pastoris followed by VB/CuO/TiO2/visible light photocatalysis resulted in 63.7% COD removal and 46% decolourisation. Moreover, this treatment resulted in the elimination of most unwanted compounds, with the exception of 4-chlorophenol. The Lactuca sativa L seed GI of the post-treated SBL was 40%, indicating moderate phytotoxicity.

Effect of SBL concentration on the removal capacity of P. ostreatus and P. pastoris. We found that initial SBL concentration plays a key role in the removal capacity of both native (P. ostreatus) and recombinant strains (P. pastoris). Preliminary tests with P. ostreatus demonstrated that the native fungus efficiently removed more than 50% of the colour and COD (data not shown) of SBL at different concentrations (100, 75, 50, and 25% (v/v)). However, with 10, 5, and 1% (v/v) SBL, higher removal (98, 97, and 90% COD removal) was observed with decolourisation percentages of 84, 87, and 90%, respectively. The laccase activity ranged from 290 to 400 UL −1 at the three evaluated SBL concentrations (Fig. 1A).
When comparing the bioremediation by P. ostreatus at different SBL concentrations, significant differences were observed in decolourisation, COD removal, and laccase activity (p > 0.0001), demonstrating that the native strain could be used to remove organic matter at all evaluated concentrations. Therefore, only the highest SBL concentration (10% v/v), similar to that of Kraft pulping liquor obtained in a mill, was used for the subsequent removal curve assays 1,37 . Similarly, with P. pastoris, significant differences in the decolourisation and COD removal (p < 0.0001) were observed with different concentrations of SBL. 10% SBL (v/v) resulted in the lowest removal percentage (<50%). Therefore, the removal curve assays were only performed with 5 and 1% (v/v) SBL.
The removal capacity of P. ostreatus is associated with the native fungus ligninolytic enzyme system. This process involves different laccases, H 2 O 2 -dependent enzymes (manganese peroxidase, and lignin peroxidase), and hydrogen peroxide-generating systems 24,38,39 . The ligninolytic system can act on BL, white liquor, or lignosulfonates, which are rich in phenolic compounds with various molecular weights, via successive oxidation reactions to form cationic radicals, aliphatic compounds, and possible mineralization leading to CO 2 production 40 . Additionally, P. ostreatus biomass can partially remove SBL by adsorption to the fungal wall, exhibiting an additional physicochemical mechanism not associated with primary metabolism and augmenting the removal efficiency 41,42 .
The genome of filamentous fungi, in particular Basidiomycetes, contains multiple genes encoding laccases. The expression of these genes can be induced by the addition of analogous substrates, copper ions, or lignin-related aromatic compounds. Moreover, adding nutritional factors, such as carbon and nitrogen sources, can induce laccase expression 43 . However, the presence of an inducer does not necessarily guarantee the expression of all encoded genes. The genome of P. ostreatus itself encodes 12 laccase genes. However, it has been determined that in liquid cultures, only two laccase genes are expressed in high proportions, namely LCC1 and LCC2 44 .
In contrast, recombinant P. pastoris only produces laccase rPOXA 1B 25 , which partially acts on phenolic aromatic compounds; other ligninolytic enzymes are required to obtain greater SBL contaminant removal. However, the contact time between SBL and P. pastoris (rPOXA 1B producer) at evaluated concentrations can influence the removal efficiency. The adsorption capacity of P. pastoris (producer of rPOXA 1B) was lower than that of P. ostreatus, most likely as a consequence of the differences in yeast vs. fungus size and morphology. In addition, differences in the cell wall affect removal based on adsorption phenomenon, as they can result in diminished potential active sites.

Figure 1.
Effect of synthetic black liquor concentration on the removal capacity of P. ostreatus and P. pastoris. SBL percentages of 1, 5, and 10% (v/v) were evaluated. (A). Experiment with P. ostreatus. Decolourisation percentages of 90, 87, and 84% (v/v) were observed for 1, 5, and 10% (v/v) SBL, respectively. COD removal values of 90, 97, and 98% were determined for 1, 5, and 10% (v/v) SBL, respectively. Laccase activity ranged between 290 UL −1 to 400 UL −1 for all evaluated concentrations. (B) Experiment with P. pastoris. Decolourisation percentages of 59, 53, and 19% (v/v) were observed for 1, 5, and 10% (v/v) SBL, respectively. COD removal values of 53, 30, and 5% were determined for 1, 5, and 10% (v/v) SBL, respectively. The laccase activities for the three concentrations evaluated were 1,327, 1,346, and 1,399 UL −1 . The letters a,b, and c represent Tukey homogeneous subsets. The letter a corresponds to the best result, followed in order by b and c. Mean ± SD (n = 3). Lignin removal (%) was analysed at 72, 144, and 192 h. A high removal was observed with VB of the two fungi, reaching percentages of 88 and 73% for P. ostreatus and P. pastoris with 10 and 1% (v/v) SBL, respectively. The removal for the two microorganisms ranged from 13 to 30% (Fig. 3) for IB/P.O and IB/P.P. P. ostreatus can modify lignin polymers. Dashtban et al. (2010) reported that white rot fungi can simultaneously perform enzymatic and non-enzymatic degradation of this polymer, producing simultaneous decreases in colour, COD, and TOC in wastewater from the paper industry. Polyphenol oxidases, haemoperoxidases, and accessory enzymes that generate hydrogen peroxide, such as glyoxal oxidase (E.C. 1.2.3.15), aryl alcohol dehydrogenase (E.C. 1.1.1.90), and quinone reductase (E.C. 1.6.5.5) participate in these enzymatic degradations. The combined action of these enzymes allows for partial or total delignification, forming free radicals, which can continue to oxidise the products to give low molecular weight intermediates. In non-enzymatic degradation, during the delignification process, the fungi lower the pH by releasing hydrogen peroxide, which can react with iron salts to generate hydroxyl radicals with high oxidizing power (the biological Fenton reaction). This reaction indirectly favours the oxidation of aromatic rings to produce carboxylic acids and CO 2 . Furthermore, other sources of C and N were added to the SBL as nutritional supplements. The fungus can assimilate these nutritional sources for biomass formation and energy production, resulting in decreased COD, TOC, and TSS 40 . Sharma et al. (2014) reported additional strategies involving fungi with cellulase and hemicellulase production that were useful for effluent bioremediation in the paper industry. Although the enzymes were not quantified in the present study, Rojas-Higuera et al. (2017) reported that when grown in pine, oak, and eucalyptus sawdust, and COD removal of 90 and 37%, respectively. For P. pastoris, VB removal percentages for 5 and 1% SBL increased gradually with time, with higher values for VB than for IB. Decolourisation values of 47.5 and 52% were observed, along with COD removal of 30 and 53% and TOC removal of 20 and 55%. Lower values were observed for IB: 9 and 6.3% for decolourisation, 14 and 13% for COD removal and less than 10% for TOC. The letters represent heterogeneous groups obtained after Tukey test analysis. The letter a corresponds to the best treatment, followed by the letter b. The bars represent the Mean ± SD (n = 3) for consistency. The production of these enzymes could increase COD and TOC removal 45 . Another factor that may favour COD and TOC removal is a high C/N ratio in SBL compounds. In this study, SBL was supplemented with carbon (5.0 gL −1 glucose), favouring fungal metabolism. Similar results were reported by Pedroza-Rodríguez et al. (2013), who found that by gradually increasing the paper effluent C/N ratio during treatment with immobilised Trametes versicolor biomass, COD removal (>80%) and decolourisation (>80%) were favoured 46 .
When comparing the results obtained in this work with those of other studies, differences between white rot fungi of the same genus may be noted. Freitas et al. (2009) evaluated the effect of Pleurotus sajor-caju on COD, lignin, and colour removal in paper effluents. After 10 days of treatment, removal percentages of 72, 79, and 72% were observed. These values were lower than those obtained in this study; moreover, two additional days of treatment were required.
In contrast, the removal capacity of P. pastoris is associated with the production of a single laccase (rPOXA 1B), as well as the use of glucose and yeast extract (nutritional supplementation of SBL) as carbon and nitrogen sources for primary metabolism. Under these conditions, the laccase enzyme was able to generate changes in aromatic rings with -OCH 3 groups. In addition, changes or stretches in the C=O bonds of conjugated carbonyls and acid carboxyl groups were also observed 47 . These modifications could account for the lower decolourisation, decreased COD removal, and diminished TOC removal. Although they did not reach the values achieved by P. ostreatus, they are important considering that only a ligninolytic enzyme (rPOXA 1B) was present.
Laccase activity, glucose consumption, and pH. Different laccase activities were observed between yeast and fungus. For the P. ostreatus removal curves, an increase was observed up to 48 h of processing, where a value of 862 UL −1 was obtained. Subsequently, at 192 h, the activity decreased to 290 UL −1 . This could suggest that phenolic compounds present in the SBL induced the activity of laccases in P. ostreatus. The enzymes could oxidise and demethylate aromatic compounds with subunits containing methoxyl groups at carbons 3 and 5. Other enzymes, such as peroxidases, could have generated aliphatic intermediates and carboxylic acids downstream 47 . Therefore, no positive (significant) correlations between laccase activity and colour removal, COD, TOC, and lignin (p = 0.051, 0.057, 0.067, and 0.071, respectively) were observed with p > 0.05 (Fig. 4A).
Native laccase induction by aromatic compounds alters the enzyme activity as a function of time and relationship with peroxidases. Using T. versicolour, Pedroza-Rodríguez and Rodríguez-Vázquez (2013) showed that when the fungus was in contact with the effluent from paper pulping industries, laccase and manganese peroxidase activity could be quantified. Both enzymes were related to colour removal, COD, and dechlorination of chlorinated aromatic compounds 46 .
Compared to P. ostreatus, the heterologous laccase production by P. pastoris as a function of time was different. Enzyme activity increased up to 192 h, regardless of SBL concentration, with final activities of 1,346 and 1,327 UL −1 with 1 and 5% SBL, respectively. This result was associated with the constitutive expression system used to clone the enzyme 25 . In P. ostreatus, laccase production varied over time and was influenced by the presence of different inducers. P. pastoris rPOXA 1B production is regulated by the P GAP promoter, which controls the expression of glyceraldehyde 3-phosphate dehydrogenase (GAPDH, E.C. 1.2.1.12). GAPDH is involved in one of the most important steps of the glycolytic pathway, and is constitutively expressed in the presence of several carbon sources, such as glucose, glycerol, oleic acid, and methanol 48 . Therefore, the constitutive expression , P. pastoris VB/P.P and IB/P.P with 10%, 5%, and 1% (v/v) SBL. Lignin removal was observed for both microorganisms, reaching levels of 88% for 10% (v/v) SBL with P. ostreatus and 73% for 1% (v/v) SBL with P. pastoris. The removal by adsorption for both cases ranged between 13 and 30%. The letters represent heterogeneous groups obtained after analysis with the Tukey test. The letter a, corresponds to the best treatment, followed by the letters b,c,d and e. The results are presented as the mean ± SD (n = 3). At all time points, the laccase activities of recombinant P. pastoris were higher than those obtained with P. ostreatus. Additionally, the laccase activity showed a significant positive correlation with all variables at both SBL concentrations (R 2 values above 0.9 and p values below 0.05). As confirmation, the only laccase expressed in P. pastoris was involved in the removal of some compounds present in SBL (Fig. 4A).
In relation to glucose consumption and pH variation, both microorganisms gradually consumed the carbon source, resulting in the production of organic acids, thus decreasing the pH. However, an important aspect of P. pastoris rPOXA 1B laccase is the recombinant enzyme stability at different pH values and temperatures. Previously, this enzyme was reported to be stable at pH 4.0 for 1 h (>90% relative activity) 49 . Furthermore, at pH values between 2.0 and 9.0, the recombinant laccase maintained relative enzymatic activity >60% for 1 h. In addition, it was stable for 1 h at temperatures between 10 and 70 °C, with a relative enzymatic activity of >50%. These results confirmed the possibility of using rPOXA 1B laccase with different residues for numerous applications, due to its activity at different pH values and temperatures 49 .
Glucose consumption by P. ostreatus could favour the production of organic acids and hydrogen peroxide, which are necessary for manganese and lignin peroxidase activity (Fig. 4B,C). The production of the P. pastoris recombinant laccase POXA 1B is directly regulated by P GAP , which is associated with glucose metabolism and biomass production (behaviour is depicted in Fig. 4A,C) 25 . Therefore, residual glucose in the medium decreased as a result of yeast metabolism, with a concomitant gradual increase in recombinant laccase production over time.

Gas Chromatography Mass Spectrometry (GC-MS). GC-MS analyses of initial SBL and SBL
post-treated with P. ostreatus or P. pastoris VB are presented in Table 2. The compounds identified in untreated SBL were 2,4-methylphenol RT = 10.47, 2,4-dichlorophenol RT = 11.36, and 2,6-dimethoxyphenol RT = 15.96. The presence of some chlorophenols compounds could be related to the extraction process. Under these conditions, free aromatic rings are chemically unstable and may undergo chlorination by small quantities of free chlorine present in water. Both microorganisms transformed chlorinated and non-chlorinated phenolic compounds. However, the compound peak shape differed between the native and recombinant strains.
The dechlorination of 2,4-dichlorophenol at positions 2 and 4 was also demonstrated. The demethoxylation at position 6 of dimethoxyphenol to form 2-methoxyphenol, which was not present in the initial SBL, was also revealed. This compound is considered one of the final aromatic intermediates in lignin degradation and has reported bactericidal and inhibitory effects on many seeds and plants 45 (Table 2).
A variety of chlorinated and non-chlorinated phenolic compounds were obtained in SBL post-treated with P. pastoris, including 2-chlorophenol, 2,4 dichlorophenol, 2-methylphenol, 4-methylphenol, 2-methoxyphenol, and 2-methoxy-4-ethylphenol. The presence of 2-methylphenol, 4-methylphenol, and 2-methoxyphenol in both post-treated SBL samples suggested the biotransformation of precursors 2,4-dimethylphenol and 2,4-dimethoxyphenol by laccase. These findings can be supported by the fact that only one laccase (rPOXA 1B) was expressed by the yeast P. pastoris. This laccase likely cannot perform other specific modifications to lignin or phenolic compounds present in SBL. In contrast, the possible presence of several laccases in P. ostreatus and other ligninolytic enzymes could be related to the differences in the phenolic compounds in the samples, as determined by GC-MS.
The dechlorination capacity of recombinant P. pastoris was different from that of P. ostreatus in regard to chlorinated aromatic compounds, since 2,4-dichlorophenol and 2-chlorophenol were detected in P. pastoris. This result suggests the following: i) rPOXA 1B laccase might lack a dechlorination cycle; ii) the recombinant laccase requires additional processing time; or iii) this enzyme cannot perform dechlorination at position 4.
The literature reveals that ligninolytic enzymes can perform different processes such as oxidation, demethylation, reduction of benzoquinones, hydroxylation, and reductive dehalogenation of chlorinated aromatic compounds 50 . Moreover, the number and position of chlorine atoms (ortho-, meta-, and para-) can limit the biotransformation capacity and reaction rate 51 .
Determination of GI for Lactuca sativa L seeds. L. sativa is among the 10 plants approved by the United States Environmental Protection Agency (EPA) 52 , the International Organization for Standardization (ISO) 53 , and the Organization for Economic Cooperation and Development (OECD) 54 for the determination of ecological effects of toxic substances. Assays employing GI for Lactuca sativa L seeds are simple, fast, reliable, and inexpensive, and they do not require sophisticated equipment. In addition, these assays are based on plants, which are more sensitive to environmental stress than other organisms 55,56 . This germination assay demonstrates the response of a vegetal biological model in the presence of initial SBL compounds and the final SBL compounds after treatment with P. ostreatus or P. pastoris VB. The GI index serves as a reference to evaluate effluent toxicity in a vegetal organism. Values below 50% indicate that the effluent has high phytotoxicity to the vegetal model 57 .
The untreated SBL had a high phytotoxicity (GI of 18.96%), which was similar to that obtained for the ZnSO 4 positive control (16.9%). The index decreased to 7.47% after treatment with P. ostreatus VB for 192 h, indicating a greater phytotoxicity than at the beginning of the experiment. As the effluent was diluted, germination rates increased, but did not exceed 50% (Fig. 5). Therefore, SBL post-treated with P. ostreatus VB continued to exhibit an important phytotoxic effect.
Furthermore, SBL samples post-treated with P. ostreatus contained compounds such as 2-methylphenol, 4-methylphenol, and 2-methoxyphenol, as detected by GC-MS. These low molecular weight compounds can cross seed cell walls, affecting some biochemical processes related to germination 8 ( Table 2).
The germination assays showed that the phytotoxic effect after SBL treatment with P. pastoris was even higher than after treatment with P. ostreatus. The germination index was zero at all concentrations (100, 75, 50, and 25% v/v), possibly resulting from intermediates, as revealed by GS-MS. As shown in Table 2, the following compounds were detected in SBL treated with P. pastoris: 2-chlorophenol, 2,4-chlorophenol, 2-methylphenol, 4-methylphenol, 2-methoxyphenol, and 2-methoxy-4-ethylphenol. The combination of these compounds could have potentiated the adverse effects on the seeds. On the other hand, it has also been demonstrated that at 50 and 100 mgL −1 , 2,4-chlorophenol exhibits a phytotoxic effect on L. sativa L seeds 58 .  Operating cycles. The results obtained with P. ostreatus demonstrated that VB could be reused for three continuous cycles of operation, equivalent to 576 h, given that no significant differences were observed between cycles for decolourisation, COD, TOC, and lignin removal (p values < 0.0001). The laccase activity and radial growth of P. ostreatus revealed that the fungus was functional for three continuous cycles of operation, favouring removal processes involving biotransformation and adsorption. The decreased enzyme activity and reduced biomass growth in cycles 4 and 5 could be associated with biomass saturation and the formation of intermediates, which were adsorbed on the fungal wall and resulted in an inhibitory effect on growth and enzymatic activity 42,59 ( Table 3). P. pastoris was used for cycle trials with 1% (v/v) SBL. According to the results, recombinant yeast could be used for COD removal for three consecutive cycles, as no significant differences were observed between the three cycles (p > 0.0001). However, colour and lignin removal decreased as the number of cycles increased, which could be related to the accumulation and adsorption of chlorinated compounds (2,4-chlorophenol and 4-chlorophenol). In relation to yeast viability and enzymatic activity, the process could be maintained for four Untreated SBL exhibited high phytotoxicity (GI: 18.96%). A similar response was observed for the positive control with ZnSO 4 (16.9%). After 192 h of P. ostreatus VB treatment, GI decreased to 7.47%, indicating higher toxicity than untreated SBL. As the SBL was diluted, the GI percentages increased without exceeding 50%. In the P. pastoris experiments, the phytotoxic effect was greater than that obtained for P. ostreatus. For this fungus, the GI values were zero for all assayed SBL concentrations (100, 75, 50, and 25% v/v). Germination indexes for P. ostreatus and P. pastoris effluents after treatment with visible-light photocatalysis using Cu/TiO 2 for 5 h (section B of the graph). Experiments with SBL post-treated with P. ostreatus/CuO/TiO 2 photocatalysis with visible light gave GI >50%, indicating that the final phytotoxic effect was reduced. Experiments with P. pastoris/CuO/TiO 2 under visible photocatalysis attained a GI of 40%, indicating a moderate phytotoxic effect. All bars represent the mean ± SD (n = 3).  When performing the photocatalytic removal tests for 5 h, 80.3% COD removal and 70.6% decolourisation were obtained for P. ostreatus-treated SBL. In the experiments with P. pastoris, the COD removal was lower at 63.7%; moreover, a decolourisation of 46% was obtained. Furthermore, COD and CU removal under dark conditions by adsorption to the composite did not exceed 15% for either treatment ( Table 4).

Detection of compounds
The differences between the results were related to concentration, CU, and initial pH values, which were lower for P. ostreatus than for P. pastoris. Under these conditions, the photocatalysis was more efficient with low loads of organic matter. This result is consistent with those reported by Ghaly et al. 60 . In their work, they reported that elevated colour and COD prevented radiation passage, resulted in a decreased photonic efficiency, and saturated the semiconductor surface, resulting in deactivation. On the other hand, the pH determines the surface charge of the composite. Thus, at acidic pH, the composite acquires a positive charge and favours the removal of contaminants with a negative surface charge. In contrast, at alkaline pH, the composite is negatively charged, and the removal of contaminants with the opposite charge is favoured. Both phenomena likely occurred in this experiment, since the SBL had a different pH (Table 1). However, in SBL treated with P. pastoris, the efficiency was also affected by the initial contaminant concentration. The oxidation of chlorinated and non-chlorinated aromatic compounds could be carried out by hydroxyl radicals and other strongly oxidizing species. However, detailed studies are needed to understand the production dynamics and mechanisms involved in the oxidation.
To verify whether the photocatalytic treatment produced any changes in the compounds generated by the two organisms, sample analysis (GC-MS) was performed after 5 h of photocatalysis. Under visible-light photocatalysis, all components detected in post-treated SBL with P. ostreatus (2-methylphenol, 4-methylphenol, and 2-methoxyphenol) were eliminated. A similar result was obtained for P. pastoris, where five of the six initial compounds were not detected. The only compound present was 4-chlorophenol with a retention time of 7.83 min ( Table 2).
Previous reports have demonstrated that conventional treatments are not effective for the degradation of toxic contaminants and the organic complexes that constitute BL. For example, one recent development in this field is the advanced oxidation process (AOP). TiO 2 is a catalyst widely used treats organic pollutants in industrial wastewater 61 . Some authors have evaluated photocatalysis for the treatment of pulping BL (100%) by employing C-Dots/TiO 2 62 as a visible-light photocatalyst for 3 h. Using this methodology, most organic compounds were removed, except for heterocyclic compounds containing nitrogen atoms. A COD removal percentage of 81.2% was obtained, which was superior to the visible-light photocatalytic effect of pure TiO 2 and coal-based C-Dots. It is important to highlight that Bo et al. (2017) did not refer to CU or CU removal. However, decolourisation was important, since they worked with 100% BL. In contrast, in this study the photocatalysis treatment followed microorganism detoxification. In addition, the SBL concentration was adjusted to attain microorganism tolerance, resulting in 70.6 and 46.0% decolourisation by P. ostreatus and P. pastoris, respectively. Doping was carried out with CuO, and visible photocatalysis was employed for COD removal, resulting in removal percentages of 80.3 and 63.7% for P. ostreatus and P. pastoris, respectively.
The GI results also supported the effects of the visible photocatalysis. The SBL post-treated with P. ostreatus/ CuO/TiO 2 /visible photocatalysis exhibited a germination percentage >50%, indicating a low final phytotoxic effect. The germination index increased as SBL was diluted to the extent that it was comparable to the distilled water control. In contrast, P. pastoris/CuO/TiO 2 /visible photocatalysis resulted in a GI of 40% (moderate phytotoxic effect) (Fig. 5). All results indicated that complementary treatment with CuO/TiO 2 /visible photocatalysis improved the final effluent quality and verified this combined approach as a promising alternative that must be further studied. Future studies will optimise the operating conditions and possible scale-up processes will be evaluated.  Even though this combined approach resulted in bioremediation, the limitations include the gradual wear and inactivation of titanium films, as well as the difficulty in scaling up the film production. Since microorganisms are being used instead of their isolated enzymes, depending on the COD, TOC, BOD, and CU concentrations, substantial effluent dilutions may be necessary to reduce toxicity towards the microorganism and avoid system saturation. Reusing treated water is an alternative; however, it is not an attractive solution for engineers. In addition, in this work, only one gene (POXA 1B) was cloned into P. pastoris yeast, and the laccase expression was less effective than that of P. ostreatus, which produced several enzymes, including laccases, manganese peroxidase, and lignin peroxidase.

Conclusions
Sequential SBL treatment using native or recombinant microorganisms diminished the initial toxicity load in terms of CU and COD, allowing further treatment with CuO/TiO 2 /visible photocatalysis. High CU and COD concentrations are the main limitations of photocatalysis when visible light lamps are utilised, since they lack sufficient energy to generate semiconductor photoexcitation (TiO 2 ). Therefore, the semiconductor was modified with CuO to allow treatment with CuO/TiO 2 /visible photocatalysis, improving the visible light results.
It was also established that VB can be reused for several cycles, which is promising since biological systems must maintain their biotransformation potential for multiple cycles to reduce operational costs. Finally, the results obtained on the laboratory scale with both microorganisms demonstrated pulping liquor bioremediation potential for several operational cycles. However, it is necessary to complement the treatment with other technologies to completely eliminate intermediates affecting L. sativa L seed germination.
Even though only one plant model was employed (L. sativa L), the results suggest that the water could be evaluated for irrigation purposes and should therefore be assessed with other plant species. A desirable treatment should guarantee removal parameters associated with irrigation processes and the effect of post-treated effluents on sensitive species.
The results obtained in this study raise the possibility of using this type of sequential technology in other stages of the paper industry to comply with environmental regulations in each country. Examples include biological pulping coupled to chemical pulping. Moreover, high added value by products can be employed in the chemical and pharmaceutical industries. Additionally, BL can be used as a raw material. All of these possibilities should be explored and could help corporations to strengthen cleaner production programs, while maintaining sustainable and environmentally friendly development. Biomass production for studies in liquid medium was carried out in WBE broth. The fungus was cultivated for eight days at 30 °C and 120 rpm. Subsequently, the biomass was recovered by filtration, washed three times with sterile distilled water, and stored at 4 °C until use.

Materials and Methods
For P. pastoris (X33/pGAPZαA-LaccPost-Stop), biomass production vials were collected from the Master Cell Bank (MCB) 63 clone 1. After thawing, they were inoculated in a screw-cap glass tube containing 5 mL of sterile YPG-Z medium (1% (w/v) yeast extract, 2% (w/v) peptone, and 2% (w/v) glucose, supplemented with 40 μg mL −1 of Zeocin). Tubes were incubated overnight (O/N) at 30 °C and 180 rpm. Then a 500 mL Erlenmeyer flask was inoculated with 100 mL (EWV, effective working volume) of fresh YPG-Z medium and cultured under the same conditions for 12 h. To rule out the presence of contaminating morphologies, the resulting culture was verified by Gram stain 64 and used as a 10% (v/v) inoculum for the small-scale production of the crude extract (concentrated supernatant) containing recombinant POXA 1B laccase (rPOXA 1B) 65 .
The production of rPOXA 1B on the laboratory scale was performed under pre-optimised culture conditions: 500 mL Erlenmeyer containing 300 mL culture medium (3/5) Analytical techniques. The chemical oxygen demand 66 was determined using a closed reflux method employing the commercial HACH Kit with a detection range of 15 − 15,000 mgL −1 . The TOC (https://www. hach.com/asset-get.download.jsa?id = 7639983811) concentration was determined using the sodium persulfate oxidation technique (HACH Kit 0-20 mgL −1 ), and the colour units (CU) and percentage of decolourisation were estimated using a previously reported methodology 67 . The laccase activity was quantified following the methodology reported by Tinoco et al. (2001) 68 . The reducing sugars were quantified using the 3,5-dinitrosalicylic acid technique reported by Miller et al. (1959) 69 .

Black liquor collection and characterization after chemical pulping. Synthetic black liquor (SBL)
was prepared by performing an "in vitro" modified Kraft process by adding 50 g of Pinus caribea sawdust to 500 mL of 5% (w/v) NaOH. Extraction was performed at 15 psi and 121 °C for 15 min. Black liquor was separated from the cellulose pulp by centrifugation at 7,000 × g for 20 min. The liquor pH was adjusted with 1.0 M H 2 SO 4 , and the liquor was refrigerated at 4 °C until use. Table 1 shows the initial characterization at 100, 75, 50, 25, 10, 5, and 1% (v/v) SBL. Erlenmeyer flasks containing 50 mL of SBL supplemented with 5 gL −1 glucose, 0.05 gL −1 (NH 4 ) 2 SO 4 , and 1 mM CuSO 4 with a final pH of 7.0 ± 0.2. An inoculum of 5% (w/v) of pelletized biomass was added to each Erlenmeyer flask. The Erlenmeyer flasks were incubated for 192 h at 120 rpm and 30 °C. At the end of the incubation period, the post-treated SBL was recovered by centrifugation at 7,000 × g for 20 min. The percentages of decolourisation, COD removal, and laccase activity (UL −1 ) were determined as response variables. The SBL concentration used in subsequent experiments was selected based on the concentrations with removal >50% and maximum laccase activity (UL −1 ).

Effect of BL concentration on
For the evaluation of P. pastoris, SBL was prepared at concentrations of 10, 5, and 1% (v/v), since the preliminary tests determined that the yeast was inhibited at SLB concentrations of 25, 50, 75, and 100% (v/v). Synthetic black liquor was supplemented with 10 gL −1 glucose, 20 gL −1 peptone, 15 gL −1 yeast extract, 0.1 mM CuSO 4 , and 20 mM ammonium sulfate at a final pH of 7.0 ± 0.2. Experiments were performed in triplicate in 500 mL Erlenmeyer flasks containing 300 mL of SBL supplemented and inoculated with 10% (v/v) P. pastoris culture. The Erlenmeyer flasks were incubated for 192 h at 120 rpm and 30 °C. At the end of the incubation period, the post-treated SBL samples were recovered by centrifugation at 7,000 × g for 20 min. Subsequently, the same response variables were quantified as for P. ostreatus.
In addition, to verify yeast viability, decimal dilutions and microdrop cultures were performed on YPG agar, following a previously reported methodology 70 . The SBL concentration to be used in the removal curves was selected in the same manner as for P. ostreatus.
Decolourisation, chemical oxygen demand, and total organic carbon removal of synthetic black liquor effluent by Pleurotus ostreatus and Pichia pastoris. Removal curves were performed in triplicate using 100 mL Erlenmeyer flasks containing 50 mL of SBL at 10% (v/v) for P. ostreatus and 500 mL Erlenmeyer flasks containing 300 mL of SBL at 5% and 1% (v/v) for P. pastoris. The synthetic black liquor was supplemented according to the nutritional needs of each microorganism. The inoculum VB percentages and operating conditions were the same as those used in the tolerance tests. For these experiments, periodic sampling was carried out for 192 h, and IB was used as an adsorption control (obtained by sterilization in an autoclave for 15 min at 121 °C and 1 psi) 59 , labelled as inactive biomass of P. ostreatus (IB/P.O) and inactive biomass of P. pastoris (IB/P.P). The evaluated response variables were the percentage of decolourisation, COD removal, TOC, residual glucose (gL −1 ) 69 , residual lignin (gL −1 ), pH, and laccase activity (UL −1 ) 68  Subsequently, samples were mixed and sonicated for 15 min. The pH of the resulting suspension was adjusted to 9.0 with 1.0 M NaOH. Afterwards, the solid was separated by centrifugation, washed with distilled deionized water, oven-dried at 40 °C for 12 h, and calcined in air at 450 °C for 1 h. The resulting CuO/TiO 2 composite was then characterised in terms of specific surface area, pore volume, and average pore diameter using nitrogen adsorption-desorption at 77.35 K, using a gas sorption analyser (Quantachrome, NOVA 4200e series, USA). The specific surface area was calculated from isotherms by the Brunauer-Emmett-Teller (BET) equation, and the pore size distribution was determined using the Barrett-Joyner-Halenda (BJH) method. The morphological and surface characteristics of the CuO/TiO 2 composite were observed using a field emission scanning electron microscope (FE-SEM, JEOL, JSM-7401F) operated at 20 kV, equipped with an energy dispersive X-ray spectrometry (EDS) system. The chemical analysis of the CuO/TiO 2 composite was performed via EDS. The analysis of the CuO/TiO 2 composite crystalline phase was performed via X-ray diffraction (XRD) using a PANalytical X'Pert PRO MOD diffractometer, operated with Cu K radiation (wavelength = 1.54050 Å) at 45 kV and 40 mA. XRD patterns were collected in a 2θ range from 10 to 90° with a scan rate of 0.026° per 40 s and a step size of 0.026°. The optical properties of the CuO/TiO 2 composite were studied using UV-Vis spectroscopy in the wavelength range of 240-800 nm.
Preliminary tests of visible photocatalysis with CuO/TiO 2 . The final physical and chemical characteristics of SBL treated with viable biomass obtained from both microorganisms was used to evaluate whether further decreases in certain parameters could feasibly be achieved by performing a complementary treatment with visible photocatalysis using a CuO/TiO 2 -based composite. The intermediate compound elimination and the increase in L. sativa L seed germination rate (GI) resulting from photocatalysis were also assessed 71 .
Experiments were performed in triplicate using 0.5 g of composite (calcined at 450 °C for 1 hour) 71 , which was deposited by evaporation at 50 °C on the base of a 5.0 cm diameter Petri dish. Subsequently, 10 mL of SBL post-treated with VB of P. ostreatus and P. pastoris was added. Each treatment was irradiated for 5 h, using two 15 W fluorescent lamps. The lamps were installed 10 cm above the glass base to avoid sample evaporation 71 . At the end of the treatment, post-irradiated SBL was recovered by centrifugation at 10,000 × g for 10 min, and the percentages of decolourisation and COD removal, the germination index (GI) 59 and the GC-MS spectrum fragmentation patterns were determined.
Germination index determination with Lactuca sativa L seeds. The effects of the initial SBL (T 1 ) and of SBL post-treated with the two microorganisms on the GI of L. sativa L seeds 55,56,72 (SEMICOL ® , www. semicol.co) were tested following a previously reported methodology 59,73 . The controls and evaluated treatments were as follows: distilled water (1), 0.001 M ZnSO 4 (2), initial SBL (3), SBL post-treated with VB/P.O (4), and SBL post-treated with VB/P.P (4) 59,73 . In Fig. 1, numbers 5 post-treated SBL for P. ostreatus and P. pastoris. The response variable was the GI (%) calculated using equation 1, as proposed by Zucconi et al. (1985) 74 : = * * * G L Gc Lc GI% 100 (1) where GI is the germination index (%), G is the average number of germinated seeds per analysed sample, Gc is the average number of germinated seeds in the negative control, L is the average length of the sample radicle (mm), and Lc is the average radicle length (mm) in the negative control. The GI value can vary from 0 to 100%.

Number of operation cycles.
To estimate the number of re-use cycles for P. ostreatus and P. pastoris VB, tests were performed in Erlenmeyer flasks using 10% (v/v) SBL for P. ostreatus and 1% (v/v) SBL for P. pastoris.
In each experiment, the sample was incubated for 5 days, at 30 °C and 120 rpm for P. ostreatus and at 30 °C and 180 rpm for P. pastoris. At the end of the first cycle, the post-treated SBL was removed, and a new batch of SBL was added to the Erlenmeyer flasks. The procedure was repeated until five cycles were completed. In each cycle, the percentage of decolourisation, TOC removal, COD, laccase activity (UL −1 ) 68 , and pH were determined. P. ostreatus was monitored by radial growth on potato dextrose-agar (PDA, 4 gL −1 potato infusion from 200 g, 20 gL −1 dextrose, 15 gL −1 agar-agar). The viability of P. pastoris was determined using a YPG agar microdrop.
Statistical analysis. Data distribution and normality tests were performed using the Shapiro-Wilk and Levene tests followed by non-parametric Kruskal-Wallis to define the removal efficiency between VB and IB for both fungi (removal curves) and the number of operational cycles sustained by either VB or IB for both fungi. To identify differences among the sample means, a Tukey test was performed. The treatments were compared with a distilled water control for the L. sativa L GI assays. A value of p < 0.05 (5%, α = 0.05) was considered to be significant. Additionally, the Pearson coefficient was determined for established significant associations among the different parameters monitored during VB treatments with both P. ostreatus and P. pastoris. For all analyses, SAS 9.0 for Windows was used.