3.1 Fungal growth in the presence of vulcanised rubber particles
The fungal strains were evaluated in terms of radial growth when incubated in PDA and EM media in Petri dishes, and no significant differences were found between the culture media used (data not shown). Figure 1A shows some of the fungi incubated with the vulcanised rubber particles already incorporated. There was a different fungus-rubber interaction depending on the fungal strain used. Three different behaviours stand out, namely total coating, partial coating and zero coating. Of the 10 strains analysed, only P. ostreatus and T. versicolor achieved a total coating of the vulcanised rubber particles, whereas the strains C. multicolor, L. edodes, L. betulinus and P. eryngii showed partial coating. The time required to reach a fully-grown plate was different for each fungus, which was qualitatively analysed by the relative growth rate (Fig. 1B). Based on this, P. ostreatus obtained complete growth in the shortest time (between 24–30 h), followed by P. eryngii and T. versicolor (from 36–42 h). Both the observations of the level of coating and the relative growth rates were related to the percentage mass loss of the vulcanised rubber particles at the end of the incubation periods. The highest mass loss was 7.5 ± 0.3%, obtained after cultivation with T. versicolor, followed by 6.1 ± 0.4% obtained with P. ostreatus. Based on these results, vulcanised rubber particles treated with T. versicolor and P. ostreatus strains were selected for surface analysis.
Bredberg et al. (2002) studied the detoxification of vulcanised rubber by different wood-rotting fungi, using the aromatic polymeric dye polyvinylamine sulfonate anthrapyridone (Poly-R478) as a model compound. The authors showed that only three strains were able to biodegrade Poly-R478, namely Pleurotus sajor-caju, T. versicolor and Recinicium bicolor. Subsequently, they used the treated rubber in cultures with Acidithiobacillus ferrooxidans and reached higher growth rates compared to rubber without fungal treatment. It was assumed that the extracellular enzymes from the selected fungi degraded aromatic structures and that the rubber obtained had a reduced toxicity. As in our study, only the cultures with white rot fungi obtained positive results.
3.2 Surface analysis of rubber particles after fungal incubation
Changes in the rubber surface were analysed by FTIR (Fig. 2). The functional groups observed may indicate the presence of compounds such as black carbon, aromatic oils, antioxidants, accelerators, among others 28. Doublet was observed between 2,926 and 2,853 cm− 1 (stretching of CH2), which was drastically reduced after the treatment, with the greatest reduction in T. versicolor with EM medium and P. ostreatus
with EM medium 37. A loss of a signal was also observed at approximately 1,715 cm− 1 in the treatment with T. versicolor on EM and PDA media and for P. ostreatus on EM and PDA media, corresponding to carbonyl groups (C = O), which are typical of the rubber chain 37. A band at approximately 1,550 cm− 1, associated with carboxylate or conjugated ketone, was totally lost after cultivation of both fungal species 38. After incubation with T. versicolor, a small band close to 1,455 cm− 1, assigned to C-H bending of CH2, was maintained in PDA, but it decreased using EM. However, for P. ostreatus, the band decreased for the assay with PDA and increased significantly in the treatment with EM. Most likely, the sporulation of the fungus was influenced when PDA medium was used, and therefore, the enzymatic metabolism involved in the decrease of the band was influenced by the sporulation process stimulated by PDA 39. We detected a variation in the polymer structure after treatments corresponding to alkene groups (–C = CH) at a wavelength of 874 cm− 1. This band was not present in the control, but it appeared after incubation with T. versicolor and P. ostreatus on both media 40. One reason for this may be the exposure of groups related to isoprene and butadiene present in rubber, which, before treatment, were hidden in more interior areas of the material.
These results show that the treatment using T. versicolor with EM allowed a greater transformation of the surface of the rubber material, suggesting that the EM medium favours the degradation in a better way than PDA for T. versicolor. For P. ostreatus, the PDA medium allowed, in general, a greater decrease in bands versus the control. The EM medium could stimulate the generation of secondary metabolites, allowing a more efficient degradation of rubber compared to the type of metabolites stimulated by the PDA medium.
We used SEM analysis to demonstrate the changes in the rubber surface against mechanical, chemical or biological treatments 15,41. In this study, SEM analysis revealed morphological changes of the analysed rubber surface after biological treatments (P. ostreatus and T. versicolor using PDA medium) in comparison with the control at 500x (Fig. 3). In both treatments, particle size was approximately 500 to 1,200 µm and was therefore not influenced by the treatments, in contrast to a previous study 41. The size reported in this work is compatible with treatments with lower-performance grinding. When comparing the control versus the biological treatments, an increase in the roughness of the material was observed, which was similar for T. versicolor and P. ostreatus. After the treatment, the surface showed cracks, facilitating the breaking of the rubber particles by subsequent treatments, such as mechanical ones, since they can spread rapidly, generating fractures 42. This result is also related to the elemental analysis of C, O, S, Si and Zn on the rubber surface (Table 2), where the carbon content decreased in the rubber particles subjected to the treatments. The opposite was observed for O and S; it should be noted that the PDA medium lacks sulphates, and therefore, the increase in S and O is not a result of the culture medium. We therefore recommend that the rubber particles were washed with ethanol and a solution of chlorine before superficial analysis (see Methodology), and therefore, the increases in Si and Zn after the biological treatment are not due to contamination, as both elements are commonly present in tyres 28,43. Zinc is used in the non-accelerated vulcanisation of rubber, whereas silicium is used as a filler 44. With P. ostreatus, a higher percentage of Si was detected compared to T. versicolor, whereas for Zn, the percentage was higher with the use of T. versicolor. The changes at the surface can be attributed to oxidation-reduction reactions by the enzymes secreted by the microorganisms, exposing a different proportion of the elements initially arranged on the surface. It is suggested that the observed changes were due to enzymatic processes related to the assimilation of microorganisms towards the components present on the surface of the treated material, which is corroborated by the loss of rubber mass after treatment.
Table 2
Number of laccases and manganese peroxidase in the genomes of fungi species studied.
Fungi Species | Laccase | Manganese peroxidase |
Coriolus multicolor | NA | NA |
Ganoderma applanatum | NA | 3 |
Lentinula edodes | 48 | 9 |
Lenzites trabea | NA | NA |
Lenzites betulinus | 7 | NA |
Pleurotus eryngii | 30 | 1 |
Pleurotus ostreatus | 44 | 17 |
Postia placenta | 4 | NA |
Stereum hirsutum | 13 | 2 |
Trametes versicolor | 37 | 14 |
3.3 Genomic insights into the pathways for rubber degradation
Obtaining genome data is the initial step in understanding the biology of P. ostreatus and T. versicolor. Fungi present a wide variety of functional proteins that play different roles in the acquisition of nutrients and protection against nearby organisms or unfavourable environmental conditions 45.
The genomes of P. ostreatus and T. versicolor have been sequenced, and an analysis of the published fungal genomes, sourced from the public database 30–32, revealed that the P. ostreatus haploid genome contains 11 chromosomes, with a length of 34.3 Mbp and a G + C content of ~ 51%. For T. versicolor, there is no information in the public database about the number of chromosomes; however, the total genomic size is 44.8 Mb, with a G + C content of ~ 58% (Table 1). Although there is a large variation in the genome size in fungi, the average genome size of fungal species taken during this study was 40.0 Mb. Fungal species harbour small genomes with highly specific genic regions and reduced non-coding regions 46,47. Fungi belonging to the phylum Basidiomycota have an average genome size of 46 Mb 48. In line with this, the sequenced genomes of Pleurotus taxa were assembled to less than 50 M bp but exhibited large numbers of annotated protein-coding genes and few constituent TEs (Table 1). However, as mentioned before, there were genome size variations among the two species, which could have arisen from unequal TE variations. The number of transposable elements (TEs), including retrotransposons and DNA transposons, accounted for approximately 253 of the P. ostreatus genome and 233 of the T. versicolor genome. For P. ostreatus, it was possible to determine notable genomic differences in the regions corresponding to TEs, where 80 identified TE families represented 2.5 to 6.2% of the genome sizes 49. Transposable elements are undoubtedly an important source of genetic variation in fungi, as previously found for other fungal species 50.
Table 1
General features of the genome of P. ostreatus and T. versicolor.
Feature | Pleurotus ostreatus | Trametes versicolor |
Chromosomes | 11 | N/A |
Genome size (Mbp) | 34,3 | 44,79 |
GC content, % | 50.85 | 57.7 |
DNA scaffolds | 12 and 572 | 283 |
Total No. TE | 253 | 233 |
Genes total number | 12,330 | 14,572 |
Protein coding genes | 12,330 | 14,302 |
Protein coding genes with function prediction | 706 | N/A |
Protein coding genes with enzymes | 1677 | N/A |
Non coding genes | 315 | 234 |
Pseudo genes | 3 | 2 |
Considering the coding gene sequences in fungi, on average, the Basidiomycota group encodes for 15,431.51 genes in their genomes. The average numbers of annotated genes are 12,330 for P. ostreatus per genome and 14,572 for the T. versicolor genome (Table 1). For the noncoding gene, we found scaffolds of the genome assembly, with 315 and 234 genes, respectively. There were three and two pseudogenes predicted, corresponding to 0.24% of the genome assembly of P. ostreatus and 0.014% of the genome assembly of T. versicolor.
Comparative analysis of fungal genomes showed that fungi are highly divergent. The P. ostreatus genome sequence assembly was distributed across of 12 and 572 scaffolds. Of these, 56 were smaller than 1 kb 51,52. The genome of T. versicolor comprised only 283 scaffolds. The average protein coding genes in P. ostreatus were 12,330, with 14,302 for T. versicolor. These coding genes have homologues with known proteins deposited in the NCBI nr, Pfam, SwissProt and TrEMBL databases.
Both fungal species can partially degrade vulcanised rubber, and genome data revealed that both of fungi encode for a large set of enzymes involved in the degradation of rubber materials. The protein coding genes with enzymes for P. ostreatus included 1,677 genes (Table 1). For T. versicolor, such data were not available in the public repositories. However, based on a previous study, T. versicolor has an expansion of the AA2 gene family (26 genes), a feature that is also found in the central polyporoid clade 53. The putative peroxidase genes (PoPOD) were obtained from the Joint Genome Institute (JGI) 31. The gene density and the mean size of the protein-coding genes in both species of fungi used in this study secrete different enzymes, among which are laccases, manganese peroxidases, versatile peroxidases, glycosylhydrolases, peptidases and fungal esterases/lipases 54. These enzymes can degrade complex compounds such as lignin as well as certain industrial pollutants contaminating vulcanised rubber. Despite the advances in comparative genomics, the genera Pleurotus and Trametes are under-exploited, and a variety of potential biotechnological applications in different industries can be elucidated based on their genomes.
3.4 Bioinformatic analysis for rubber bioremediation
3.4.1. Laccase and manganese peroxidase in the fungi genomes
The genomic data of several species evaluated in this work revealed information concerning the presence of laccases and manganese peroxidase enzymes. For the species L. edodes, L. betulinus, P. eryngii, P. ostreatus, P. placenta, S. hirsutum and T. versicolor, database searches revealed the presence of several sequences of laccase already sequenced; the species L. edodes, P. ostreatus, T. versicolor and P. eryngii showed the highest number of laccase sequences (Table 2). In addition, the species G. applanatum, L. edodes, P. eryngii, P. ostreatus, S. hirsutum and T. versicolor also presented several sequences of manganese peroxidase already characterised. Species such as L. edodes, P. ostreatus and T. versicolor presented the highest number of characterised enzymes of manganese peroxidase in the evaluated database (Table 2). In our experiments, these species showed a high ability to degrade rubber. The high number of deposited sequences of laccase and manganese peroxidase for these species could reinforce their role as microorganisms that can degrade rubber and other polymers naturally, using laccase and manganese peroxidase enzymes 55. However, the species C. multicolor and L. trabea presented no sequences of both mentioned enzymes. Also, G. applanatum had no deposited sequence of laccase; L. betulinus and P. placenta had no sequences of manganese peroxidase in the database used in this study.
3.4.2. Conservation among laccases and manganese peroxidase enzymes from P. ostreatus and T. versicolor
Laccase presented three multicopper oxidase-conserved domains in its structure, namely Cu-oxidase, Cu-oxidase 2 and Cu-oxidase 3. For all sequences, these domains were located in the same region. For Cu-oxidase 3 domain, its location in the sequences started around the 31 amino acid position; it had a size of around 119 amino acids. The Cu-oxidase domain started from the 160 to 169 amino acid position and presented about 159 amino acids. The Cu-oxidase 2 domain started in the 366 to 388 amino acid position and generally had a size of 137 amino acids. These domains can use copper ions as cofactors to oxidise a broad range of substrates 55,56.
Manganese peroxidase has two conserved domains: a peroxidase domain with 229 amino acids and another domain, the so-called “peroxidase extension region”, with a size of 79 amino acids. The peroxidase domain in the sequences starts from 49 to 53 amino acid position and the peroxidase extension region from 279 to 286 amino acid position. These enzymatic domains are involved in oxidative reactions that use hydrogen peroxide as the electron acceptor 55–57.
For alignment at the amino acid level of laccase sequences from P. ostreatus, 18 sequences were used, and from T. versicolor, 23 sequences were used. All partial sequences were removed from the alignment analysis. Laccases from P. ostreatus showed a size of around 530 amino acids and a different pattern of the amino acid content; they were divided into four group of laccases. For T. versicolor, the laccases showed an average size of 520 amino acids and showed the same behaviour as those from P. ostreatus regarding the amino acid content; they were divided into three groups. In the evaluated database, the information about the specific name of each laccase in both species was inconsistent, making it difficult to predict the correct name per group. However, based on the amino acid alignment and the similarity, the laccase groups were determined.
In P. ostreatus, the groups of alignment represented for A and B presented high similarity among their amino acid contents of the studied sequences; however, for the groups C and D, the sequences were almost identical (see Supplementary Fig. S1 online). In T. versicolor, group A also presented a high similarity among the amino acid contents of the sequences, and groups B and C were almost identical (see Supplementary Fig. S2 online).
Regarding manganese peroxidase, we also evaluated its amino acid conservation among all sequences deposited in the NCBI database. Both species presented manganese peroxidase with a size of around 360 amino acids. For P. ostreatus, we evaluated 11 sequences and for T. versicolor 10 sequences. Similar to the laccase analysis, all partial sequences were removed from the alignment analysis. The species P. ostreatus presented four groups based on the amino acid contents of the sequences (see Supplementary Fig. S3 online). The groups A, B and C were almost identical, and group D of manganese peroxidase presented high similarity regarding the amino acid content. The species T. versicolor had two groups of manganese peroxidases enzymes (see Supplementary Fig. S4 online), presenting high similarity among the sequences for the group representing A, while for the group B, the sequences are almost identical. When generating an alignment using all sequences together, one analysis for laccase and another analysis for manganese peroxidase, the level of amino acid similarity decreased (data not shown). Only separated regions of conserved domains were maintained. These results indicate the presence of different isoforms of laccases and manganese peroxidase in the genomes of these two studied fungi or the lack of genomic information in the accessed database 58.