Deciphering lignocellulose deconstruction by the white rot fungus Irpex lacteus based on genomic and transcriptomic analyses

The white rots are a large group of fungi that can efficiently decompose plant cell wall and utilize all its components including lignin, cellulose, and hemicellulose. The white rot fungi depend mainly on oxidoreductases and glycoside hydrolases for lignin and polysaccharides degradation, respectively [1]. In contrast, brown rot fungi selectively degrade carbohydrates, owing to the lack of ligninolytic enzymes such as class II peroxidases and laccases [2–4]. Due to their superior capability in destructing lignocellulose, white rot fungi can be widely used in many biotechnological industries including biofuel and biorefinery, especially in the biological pretreatment of lignocellulosic biomass feedstock [5].

With the advancement in next generation sequencing, whole genome sequencing is widely used to explore the underlying molecular mechanisms of lignocellulose degradation by white rot fungi [6, 7]. To date, dozens of white rot fungal genomes, such as those from Phanerochaete chrysosporium and Ceriporiopsis subvermispora, have been sequenced and are publicly available [8]. Comparative genomic studies have revealed that class II peroxidases, dye-decolorizing peroxidases (DyPs), and multiple pathways for hydrogen peroxide (H₂O₂) production collectively constitute the ligninolytic enzyme system in white rots [9]. With these achievements, multi-omics analyses are emerging as an integrated approach to deepen our understanding of enzymatic degradation of lignocellulose by white rots. In recent years, there have been reports of using multi-omics to analyze P. chrysosporium grown on aspen, pine, and spruce [10–12], C. subvermispora on aspen [13], Pycnoporus coccineus on aspen and pine [14], Phlebia radiata on spruce [15], and Dichomitus squalens on aspen, spruce, wheat bran, and cotton seed hulls [16]. These studies mainly focused on gene expression patterns in response to different lignocelluloses or times.

White rot fungi are further grouped into simultaneous and selective lignocellulose degraders [17]. The former class, represented by P. chrysosporium, simultaneously removes cellulose, hemicellulose, and lignin. The selective degraders, typified by C. subvermispora, instead preferentially destroy lignin rather than cellulose and hemicellulose at the early stage [18]. Correspondingly, cellulase is expressed at a constant but high level, while lignin-degrading enzymes gradually accumulate in P. chrysosporium confronting lignocellulose [12]. In contrast, C. subvermispora secrets mainly peroxidases oxidizing lignin initially, then switches to carbohydrate active enzymes (CAZymes) acting on cellulose and hemicellulose at the advanced stage [13].

The simultaneous and selective paradigms of white rots are defined primarily based on their degradation patterns associated with woody lignocellulose. Interestingly, Irpex lacteus CD2, a white rot basidiomycete with potent lignocellulose-degrading ability [19], displays a degradation pattern similar to the selective paradigm in pretreating the corn stover. I. lacteus CD2 preferentially degrades lignin at the early stage, followed by sharply elevated cellulose degradation rate at the advanced stage [19]. The same pattern has also been reported recently for degradation of corn stover by another I. lacteus strain Fr. 238 617/93 [20]. Recently, the genome of an I. lacteus strain F17 was reported [21]. The secretome of I. lacteus strain Fr. 238 617/93 grown on wheat straw has also been studied [22]. These investigations undoubtedly deepen our understanding of lignocellulose degradation by I. lacteus. However, with these achievements, the molecular mechanism underlying the selective-like mode of lignocellulose degradation by I. lacteus and its implication remain unknown. The genomic information of I. lacteus has not yet been linked to transcription of genes relevant to lignocellulose degradation. There are no systematic analyses available for I. lacteus CD2 (and other strains) grown on corn stover, which is an agricultural residue produced in large amounts and can be used as a feedstock for second-generation biofuels [23]. Herein, the genome of I. lacteus CD2 was sequenced and analyzed, combing with biochemical and transcriptomic analyses to elucidate the molecular mechanism on its efficient deconstruction of lignocellulose in corn stover.

Previous to our study, there was only one brief report on the genome sequence of an I. lacteus strain F17 [21]. In this study, we found that the genome of I. lacteus CD2 is 43.16 Mb, similar to but slightly smaller than that of formerly sequenced I. lacteus F17 (44.36 Mb). Despite this similarity, however, I. lacteus CD2 encodes higher numbers of plant cell wall polysaccharide-degrading enzymes, with 182 GHs compared to 161 GHs in I. lacteus F17 [21]. In addition, I. lacteus CD2 does not have a laccase, but have 7 MnPs, 2 LiPs, and 4 DyPs, while I. lacteus F17 bears one laccase, one LiP, 13 MnPs but no DyP. The study of strain F17 focused on detoxification of synthetic dyes but not in pretreatment of lignocellulose [56, 57]. Therefore, it is not known at this time whether the difference in the numbers and amino acid sequences of encoding enzymes by the two strains would impact their efficiency in lignocellulose degradation. The 18S rRNA phylogenetic analysis suggested that I. lacteus F17 is closest to C. subvermispora, a selective lignin degrader [21]. Taxonomy based on 18S rRNA gene showed a limitation for certain order level [58–60], while the phylogenomic approach could improve phylogenetic accuracy. Our comparative analysis at a phylogenomic level revealed that I. lacteus CD2 is very close to the selective white rot fungus P. carnosa.

The sequenced and annotated genome of I. lacteus CD2 presented here provides an excellent platform for subsequent biochemical and transcriptomic analyses of this fungus grown on lignocellulose. The carbon source used for submerged shake flask fermentation in this study was ball-milled corn stover, whose natural physical structure has been disrupted to some extent. However, the expression pattern of extracellular lignocellulose-degrading enzymes was still very similar to that of this fungus grown in solid-state fermentation (SSF) on corn stover [19]. Both the current submerged fermentation and previous SSF revealed roughly identical expression patterns, with high lignin-degrading enzymes appearing at the early stage and high expression of cellulase at the late stage.

The large accumulation of MnP transcripts and enzymatic activities at the early stage highlights their importance involved in lignin degradation. Indeed, two recombinant MnPs from I. lacteus CD2 exhibited Mn²⁺-dependent oxidation ability for phenolic lignin model compounds [55]. These facts are consistent with the finding that supplementing with Mn²⁺ greatly improved the pretreatment efficiency of corn stover by I. lacteus CD2 [61]. MnPs are enzymes that can directly attack phenolic lignin but not non-phenolic lignin compounds [62]. However, phenolic lignin represents only a minor part of lignin (~ 10%), the more recalcitrant and high redox-potential non-phenolic lignin constitutes the major part. The apparent paradox could be explained by our recent discovery that certain carboxylates such as malonate, and more importantly those produced in the life cycle of the fungus such as oxalate, can be used by the above-mentioned MnPs from I. lacteus to generate free radicals, which can mediate oxidation of non-phenolic lignin compounds [55]. As a supporting evidence, the genes controlling oxalate production and metabolism experienced regulation favoring accumulation of oxalate on LC3d (Additional file 5). The oxalate concentration was in the range of 0.11 ~ 0.13 mg/ml (1.22 ~ 1.44 mM) from LC3d to LC9d, which can be used by MnPs for oxidation of non-phenolic lignin. Since there was no detectable LiP and DyP activity during cultivation, MnPs, in combination with GLOX and AOX, appeared to be main components of the extracellular lignin-degrading enzyme system for I. lacteus CD2 growth on corn stover.

Similar to solid-state biological pretreatment of corn stover by I. lacteus CD2 [63], Fe³⁺-reducing activity was detected during all the culture periods. It was noteworthy that the peak of Fe³⁺-reducing activity appeared at the very early stage (day 2) even before the MnPs (day 3). Fe²⁺ can react with H₂O₂ via Fenton reaction and produce hydroxyl radical, attacking both lignin and plant cell wall polysaccharides [44]. Hydroxyl radical demethoxylated aromatic groups in lignin and introduces hydroxyl groups in non-phenolic lignin [26]. This process is important, which results into cleavages of lignocellulose barrier, thus facilitating subsequent penetration of MnPs and glycoside hydrolases [64]. Interestingly, the Fe³⁺-reducing activity rapidly decreased to normal state from day 3 to day 9 in lignocellulose medium, which was at a level similar to that in glucose medium (Fig. 2i). However, despite the down-regulated Fe³⁺-reducing activity during this period, degradation of lignocellulose by free radical generated through reaction of Fe³⁺ with H₂O₂ cannot be excluded. At the early stage (LC3d), a range of AA3 and AA5 auxiliary enzymes were highly enriched (Table 1), providing H₂O₂ for both MnPs and Fe²⁺. At the late stage (LC6d), part of AA3 and AA5 enzymes was still highly expressed (e.g. 0806.696 and 0808.462, Fig. 3c). At this stage, cellulase was highly expressed, hydrolyzing cellulose into cellobiose and cellooligosaccharides. These sugars serve as substrates of CDH, which was also highly induced in LC6d (Fig. 3c), and could generate sufficient amounts of H₂O₂ for Fe²⁺. As another important source of hydroxyl radicals [65], the transcript of glycopeptide was highly expressed on LC3d and LC6d, as well.

In accordance with biochemical analysis, the transcriptomic study revealed that the transcripts for MnP isoenzymes were highly expressed at the early stage, but decreased over time in lignocellulose culture. The transcripts for cellulase and hemicellulase gradually increased with the time going on. The special expression pattern of genes encoding lignocellulose-degrading enzymes was similar to that of the selective ligninolytic fungus C. subvermispora during growth on aspen wood [13]. However, the transcriptomic and biochemical analyses also revealed significant difference in lignin degradation between I. lacteus CD2 and C. subvermispora. For lignin decomposition, I. lacteus CD2 mainly uses MnPs which are assisted by the H₂O₂ producers GLOX and AOX, while C. subvermispora utilizes laccases and MnPs with AAO. Moreover, the substrate scope of MnPs from I. lacteus CD2 expands to the predominant non-phenolic lignin with the aid of oxalate [55], while peroxyl radical, generated through interplay of fatty acid desaturases and MnPs, is implicated to be responsible for cleavage of non-phenolic lignin in C. subvermispora [38]. The expression of important genes involved in lignocellulose degradation is also different between I. lacteus CD2 and another model white rot fungus P. chrysosporium. In I. lacteus CD2, the gene encoding glycopeptide related to Fe³⁺ reduction was highly expressed on LC3d and LC6d. The CDH gene involved in Fe³⁺ reduction was also up-regulated on LC6d compared to LC3d (Table 1), which may affect the rate of Fenton reaction. In contrast, the transcript of CDH remains abundant, while those of glycopeptides are negligible in P. chrysosporium during all the culture periods [12].

Taken together, based on biochemical, genomic, and transcriptomic analyses, we could propose a model for lignocellulose degradation by I. lacteus CD2 in this submerged fermentation. At the very early stage (day 2 and before), when MnPs are to be expressed, I. lacteus CD2 up-regulates the Fe³⁺-reducing power. Fe²⁺ reacts with low level of H₂O₂ to produce hydroxyl radical for an initial attack and breakage of interlinking lignocellulose network. At the early stage (days 3–4), MnPs are predominantly expressed and directly oxidize phenolic lignin compound. With assistance of mediators such as oxalate, non-phenolic lignin components are also oxidized by MnPs. Cellulase and xylanase are beginning to accumulate, but still not to the highest level. At the late stage (days 5, 6, and after), MnPs disappear, but part of enzymes responsible for H₂O₂ generation are highly expressed, which allow production of H₂O₂ and subsequent hydroxyl radical. Cellulase and xylanase are abundantly expressed at this stage, decomposing cellulose and hemicellulose at the maximal speed. In a model described previously [66], the lack of enzymes (such as β-glucosidase and β-xylosidase) cleaving oligosaccharides into simple sugars is key for high saccharification efficiency of grass biomass after I. lacteus pretreatment. Indeed, our findings support this hypothesis, because in LC3d, the transcription of β-glucosidase and the enzymatic activity were very low. With this in mind, our findings corroborate the idea that selective expression of lignin-degrading enzymes at the early stage could be an additional key factor for high saccharification efficiency of lignocellulose after I. lacteus CD2 biopretreatment. Selective removal of lignin at the early stage exposes cellulose ready for enzymatic hydrolysis, while low expression of extracellular β-glucosidase keeps cellulose in long-chain, non-fermentable form. These two characters enable I. lacteus a suitable white rot for pretreating grass feedstock for biofuel production with high efficiency. We also notice that the regulation of Fe³⁺-reducing activity, as well as hydroxyl radical generation, is complex. Since high concentrations of free radicals and H₂O₂ are harmful for both enzymes and microbial hosts, this implies that, any attempts to harness the power of free radicals for formulating novel, robust lignocellulose-degrading enzyme cocktails should be delicately designed.

In this study, we sought to decipher the molecular mechanism underlying efficient lignocellulose degradation by I. lacteus CD2. The genome annotation indicated that I. lacteus CD2 has a full array of enzymes for lignin, cellulose, and hemicellulose degradation. Both biochemical and transcriptomic analyses indicated that the fungus employed a selective strategy for degradation of lignocellulose in this submerged fermentation, a phenomenon reminiscent to that in solid-state fermentation on corn stover. This strategy, in combination with low extracellular glycosidase activity, forms the basis for its high efficiency in biological pretreatment of corn stover for saccharification.