Omics analyses and biochemical study of Phlebiopsis gigantea elucidate its degradation strategy of wood extractives

Wood extractives, solvent-soluble fractions of woody biomass, are considered to be a factor impeding or excluding fungal colonization on the freshly harvested conifers. Among wood decay fungi, the basidiomycete Phlebiopsis gigantea has evolved a unique enzyme system to efficiently transform or degrade conifer extractives but little is known about the mechanism(s). In this study, to clarify the mechanism(s) of softwood degradation, we examined the transcriptome, proteome, and metabolome of P. gigantea when grown on defined media containing microcrystalline cellulose and pine sapwood extractives. Beyond the conventional enzymes often associated with cellulose, hemicellulose and lignin degradation, an array of enzymes implicated in the metabolism of softwood lipophilic extractives such as fatty and resin acids, steroids and glycerides was significantly up-regulated. Among these, a highly expressed and inducible lipase is likely responsible for lipophilic extractive degradation, based on its extracellular location and our characterization of the recombinant enzyme. Our results provide insight into physiological roles of extractives in the interaction between wood and fungi.

For example, a set of genes encoding an efflux ABC transporter in Grosmannia clavigera is thought to confer monoterpene-tolerance 14 , and a glutathione S-transferase in Phanerochaete chrysosporium provides antioxidant activity 15 . Other genes are suggested to function in lipid metabolism by P. gigantea but our knowledge of specific genes involved in the metabolism(s) of coniferous extractives by this fungus remains uncertain in large part due to the complexity of the woody substrate 13 . In short, it is not clear whether the coniferous extractives specifically induced the observed metabolic pathways in P. gigantea.
Free and esterified fatty acids together with resin acids are major components of softwood extractives, and basidiomycetes are known to secrete lipases that hydrolyze glycerides to release fatty acids 13 . However, how this group of fungi regulates the genes involved in utilizing lipids in the wood extractives remains unknown. Here, we attempt to elucidate the extractive-induced mechanism(s) of P. gigantea. We cultivated the fungus on a defined culture medium containing microcrystalline cellulose as a carbon source and supplemented with different amounts of the Pinus taeda sapwood extracts, which contained high levels of the lipids. Metabolites of lipophilic extractives in the culture supernatant were determined by GC-MS. Transcriptomic analyses of the grown mycelia and proteomic analyses of the supernatants were performed by RNA-seq and LC-MS/MS, respectively, to identify genes encoding enzymes related to transformation or degradation of the extractives by P. gigantea. A putative extracellular lipase was highly expressed and regulated in response to the extractives. This enzyme was produced in the heterologous expression system of Pichia pastoris, and functionally characterized as a novel basidiomycete lipase.

Results
Chemical composition of P. taeda wood extractives during the growth of P. gigantea. The chemical composition of the 70 v/v% acetone-treated extractives from the grounded powder of P. taeda sapwood was analyzed by GC-MS ( Fig. 1; Fig. S1). In this GC-MS analysis, a constant amount of microcrystalline cellulose (Avicel) was supplemented with the following series of P. taeda extractives, in the absence of extractives (AV0X), the equivalent amount of extractives present in the P. taeda wood (AV1X), twofold (AV2X), and fourfold P. taeda extractives (AV4X), which were also used as the culture medium of P. gigantea. The 20 representative compounds ( Fig. 1; AV0X, AV1X, AV2X, and AV4X) were determined. The fatty acids present in control sample (AV0X) were thought to be contaminants arisen during sample preparation. The respective concentration of sample components generally increased with more concentrated extracts except for some fatty acids, which was probably due to contaminants. As expected, this included most of the detectable fatty and resin acids, diglycerides, free sterols and sterol esters, and triglycerides 16 . Culture media with the same composition were also analyzed after 5 days growth ( Fig. 1; AV0X Phlgi, AV1X Phlgi, AV2X Phlgi, and AV4X Phlgi). When the medium was inoculated with P. gigantea, it grew on media containing microcrystalline cellulose, and the levels of di-and triglycerides appeared to decrease (Fig. 1) whereas fatty and resin acids, and sterols were unaffected after the 5-days cultivation. For some compounds, amounts per total amounts of lyophilized biomass seems to increase likely due to the efficient consumption of microcrystalline cellulose by P. gigantea. Decrease of di-and triglycerides during growth of P. gigantea was hypothesized to be a response to the extractives, especially lipid compounds.
Among the genes encoding lignocellulose-degrading enzymes, only seven genes were up-regulated, and two were down-regulated when the extractives were present in the media (AV1X, AV2X, and AV4X), compared to that of no extractives (AV0X), respectively (P < 0.02, > twofold, RPKM ≥ 10) (Table 1). Under the less stringent threshold (P < 0.05, > twofold, RPKM ≥ 10), 16 and 10 transcripts were up-, and down-regulated, respectively, relative to the AV0X (Table S2). Dye peroxidase (Phlgi_78526) and AA5_1 copper radical oxidase (Phlgi_128606) were significantly down-regulated, while AA3_2 GMC oxidase (Phlgi_101126), GH18 chitinases (Phlgi_88507 and Phlgi_27927), GH11 xylanase (Phlgi_21241), GH12 glucanase (Phlgi_126344), and GH28 pectinase (Phlgi_28251 and Phlgi_36341) were significantly up-regulated. These results indicated that the CAZymes, which play central roles in lignocellulose degradation, such as GH7 and GH6 cellobiohydrolases/endo-glucanases, and AA2 manganese peroxidases (MnP) are not regulated by the presence of extractives. In contrast, a few genes encoding www.nature.com/scientificreports/ glycoside hydrolase (GH) families 11 endo-xylanase, GH12 endo-glucanase, and GH28 polygalacturonase were significantly up-regulated, suggesting that pectin and other polysaccharide hydrolysis is likely a first step of wood decay in response to extractives coated on wood material. The functional distribution of the highly up-regulated genes in the presence of the increasing amounts of extractives were involved in amino acid, lipid, xenobiotic, and carbohydrate metabolism based on GO classifications ( Fig. 2B; Supplemental data file S2). The 121 significantly up-regulated genes included 12 genes in lipid metabolism, 8 genes in biodegradation of xenobiotics, 6 genes in carbohydrate metabolism, and several genes in tryptophan metabolism and the biosynthesis of secondary metabolites, including phenylpropanoids and alkaloids. Twelve genes classified into lipid metabolism such as β-oxidation-related proteins, included Phlgi_130767, Phlgi_27604, and Phlgi_27649 β-ketothiolases (KTs), Phlgi_126556 long fatty acid CoA ligase, Phlgi_18116, and Phlgi_91676 acyl-CoA dehydrogenases (ADH), and Phlgi_29221 and Phlgi_27759 enoyl-CoA hydratases (EHs). Among the 8 genes involved in biodegradation of xenobiotics such as cytochrome P450s, interestingly, four substantially up-regulated genes were annotated as cytochrome P450 monooxygenases, Phlgi_20511, Phlgi_20514, Phlgi_82811, and Phlgi_68457, (> fourfold, P < 0.02, RPKM values ≥ 10) in response to supplementation with pine extractive (Fig. 2C; Supplemental data file S2). The induced genes belonged to P450 families, including Chemical composition of lipophilic compounds in lyophilized Avicel microcrystalline cellulose medium containing none (AV0X) or increasing amounts of loblolly pine extract (AV1X, AV2X, AV4X) without and with P. gigantea inoculation by GC-MS analyses (n = 1). The culture media were harvested five days after the inoculation (AV0X Phlgi, AV1X Phlgi, AV2X Phlgi, AV4X Phlgi) (n = 1). Numbers in parentheses correspond to peaks shown in GC-MS chromatograms ( Figure S1). *The fatty acids present in control samples (AV0X and AV0X Phlgi) were thought to be contaminants arisen during sample preparation. www.nature.com/scientificreports/ Highlighted transcripts corresponding to four cytochrome P450s and lipase protein model #19028 exhibiting significant accumulation (> fourfold) even at relatively low extract addition (AV1X). www.nature.com/scientificreports/ families CYP52 (CYP5150A), CYP53 (CYP53C), CYP67 (CYP5035A), and CYP503 (CYP512B), respectively 13 . Among the 41 significantly down-regulated genes, many were categorized as amino acid metabolism, but most were not assigned by GO terms. Tolerance to terpenes may be mediated in part by a putative ABC efflux transporter (Phlgi1_130987, Supplemental data file S2). Of the 51 ABC transporters of P. gigantea, this protein is most closely related to the GcABC-G1 gene of the ascomycete Grosmannia clavigera, a pathogen of Pinus contorta 14 . The GcABC-G1 gene is upregulated in response to various terpenes and appears to be a key element against the host defenses. Consistent with a similar function, our analysis showed the P. gigantea homolog to be upregulated 11-to 20-fold relative to microcrystalline cellulose without extractives (Supplemental data file S2). Differential regulation also implicated glutathione S-transferase (Phlgi_101998) in the transformation and detoxification of extractives (Supplemental data file S2). Homologs of upregulated genes encoding aldehyde dehydrogenase (Phlgi_121047), and aryl-alcohol dehydrogenase (Phigi_89048) are induced by aromatic compounds in P. chrysosporium 17,18 . Lipid metabolism in P. gigantea. An integrated pathway of lipid metabolism was proposed in a previous study 13 . Based on our current transcript analysis, we mapped the genes on each reaction in the pathway (Fig. 3). The genes involved in lipid metabolism were mostly up-regulated in the presence of extractives including AV1X, AV2X, AV4X and LPAS compared to microcrystalline cellulose (AV0X) as a sole carbon source. However, there were some difference in expression patterns between extractive-coated microcrystalline cellulose and pine wood substrate such as lipase Phlgi_129172 and CoA ligase Phlgi_2959. Among nine lipases, four lipase-encoding transcripts were significantly up-regulated (Fig. 3). Of these, lipase Phlgi_19028 transcripts were most abundant (RPKM = 1767 in AV4X) and accumulated 4.64-fold over AV0X. These results strongly suggest that the gene encoding lipase, which contains a signal peptide, was induced for the extracellular lipid degradation. Further, genes involved in β-oxidation, TCA cycle, and GLOX cycle were activated by the presence of extractives as previously observed in pine wood powder 13 . Thus, the fatty acid existing in the extractives or generated from diglycerides, and/or triglycerides were immediately metabolized through β-oxidation together with the glyoxylate shunt in the cells. Conversely, genes encoding oxaloacetase (OXA), which converts oxaloacetate in the TCA cycle into oxalate, were not up-regulated in the presence of extractives. This observation was different from the comparative transcriptome results of pine wood powder relative to glucose 13 . Considered together with the limited CAZy gene regulation, use of extractive-coated microcrystalline cellulose seems well suited for investigating genetic regulation of degradative processes without the complexities inherent in woody substrates.
Activities and sequences of P. gigantea enzymes in culture filtrates. To monitor lipid degradation by P. gigantea in response to increasing amounts of extractives, the time course of lipase activity was measured in culture supernatants from AV0X, AV1X, AV2X, and AV4X, using p-nitrophenol-dodecanoate (pNPD) as the C12 substrate which is one of model compounds of lipase activity. Significantly increased lipase activities were observed in the culture supernatant of AV1X, AV2X, and AV4X media during time course, compared to AV0X (Fig. 4A). To identify the specific enzyme(s) responsible for this lipase activity, AV0X and AV4X culture filtrates together with AV1X and AV2X were analyzed by LC-MS/MS ( Fig. 4B; Supplemental data file S1). A total of 196 and 295 proteins were identified in the AV0X and AV4X proteomes, respectively. Identified proteins met our criteria that a minimum of two unique peptides per protein be identified. Exponentially modified protein abundances (emPAI) 19 were calculated (summarized in Supplemental data file S1). Among a total of 327 proteins, 32 and 131 proteins were identified only in AV0X and AV4X, respectively, and most of AV4X specific proteins were intracellular proteins including TCA cycle and fatty acid metabolism probably due to the stress of wood extractives. Among 164 common proteins in AV0X and AV4X, 35 were substantially upregulated in the presence of extractives (> threefold). They included Phlgi_19028 lipase (emPAI value = 0.9 and 3.0 in AV0X and AV4X, respectively) together with other secreted enzymes such as Phlgi_28251 GH28 polygalacturonase (emPAI value = 25 and 138 in AV0X and AV4X respectively). The result indicated that Phlgi_19028 lipase likely plays an important role in the degradation of extracellular lipid.

Structure, phylogenetic analyses and heterologous expression of lipase Phlgi_19028.
To further characterize Phlgi_19028, the recombinant lipase was produced in a Pichia pastoris expression system. Previous reports demonstrated that E. coli expressed lipases localized in inclusion bodies 20 , but Pichia produced functional lipases in the soluble fractions 21,22 . Of nine P. gigantea lipases, Phlgi_32096 and Phlgi_99110 were the most distantly related but featured the conserved motif, GQSAG and GESAG, found in basidiomycetes lipases 22 (Fig. 5A). Clustal analyses showed other lipases formed separate, well-defined clades with the GHSLG motif [23][24][25] . The predicted amino acid sequences of the highly expressed Phlgi_19028 featured the conserved pentapeptide and catalytic triads (Ser180, Asp237, and His251) (Fig. S4). A secretion signal, three N-glycosylation, and one O-glycosylation sites were also predicted.
The recombinant lipase Phlgi_19028 was successfully produced as a major protein in P. pastoris culture supernatants (Fig. S5, lane 1) relative to the pPICZα vector control (lane 3). After deglycosylation, the lipase matched the calculated molecular weight (minus glycosylation and signal peptide) of approximately 30.0 kDa as estimated by SDS-PAGE (lane 2). Indeed, significantly higher lipase activity (85.41 ± 3.75 U/mg) toward model compound pNPD was detected compared to a vector control (2.55 ± 0.06 U/mg) (Fig. 5B). We herein named this enzyme PgLip19028. Because some lipases have been reported to increase their activities in the presence of metal ions 26 , we examined the effect of metal ions on recombinant PgLip19028. One mM of nine different metals was added to the assays, and no increases were observed (Table S3). Using the pNPD substrate, the optimal reaction temperature and pH were determined to be 25 °C and pH 4.5, respectively (Fig. 5C,D). Notably, the optimal pH 4.5 is apparently lower than other lipases 22 www.nature.com/scientificreports/ and the reaction end products were analyzed by TLC and GC-MS analyses. As a result, a peak area of triolein appeared to decrease after the reaction, while peaks of diolein, monoolein, and oleic acid were identified as the end products ( Fig. 6A; Fig. S6A). Under the conditions tested, 2.2 mg (≒7.8 µmol) of oleic acid was released from 10 mg (≒11.3 µmol) of triolein after 17 h of reaction. In addition, PgLip19028 released oleic acid, linoleic acid, linolenic acid, and palmitic acid from triglycerides and diglycerides in the coniferous extractives ( Fig. 6B;  Fig. S6B). In contrast, 0.10 mg (≒0.35 µmol) of oleic acid and 0.16 mg (≒0.58 µmol) of linoleic acid were released from approximately 1 mg of triglycerides included in 10 mg extractives after 17 h of reaction. These indicated that lipase activity toward extractives was lower than that toward triolein probably because concentration of triglycerides was low and other substances of the extractives inhibited the activity. Additional measurements over time under various conditions may clarify the precise mechanism. Overall, these results showed that this lipase produces a broad range of fatty acids with different carbon chain lengths and unsaturated bonds distinct from lipids and other lipophilic compounds in coniferous extractives, and that it could play a central role in the degradation of extractives.

Discussion
The white rot basidiomycete P. gigantea used in this study was isolated from P. taeda (loblolly pine), which is one of the most common conifer species and heavily used as lumber and pulpwood. P. gigantea is known to spread fast and deeply on softwood stumps preventing growth of softwood pathogens. Previous work on P. gigantea revealed a distinctive repertoire of genes involved in a unique and efficient system for degrading all components of softwood. Transcriptome analyses had identified hundreds of potential CAZymes, many of which were upregulated on P. taeda wood relative to media containing glucose as the sole carbon source 13 . Extending transcript analyses to pine devoid of acetone soluble extractives, comparative analyses suggested that unknown extractives in P. taeda influenced the regulation of a diverse array of P gigantea genes 13 . In this study, P. gigantea was shown to utilize microcrystalline cellulose as a sole carbon source and induced various CAZymes like other white-rot fungi 37,38 . To more precisely identify the genes involved in the response to the extractives, we examined the P. gigantea transcriptome and proteome when grown on defined microcrystalline cellulose media supplemented differing amounts of P. taeda sapwood extractives. Transcripts encoding protein Phlgi_19028 accumulated sevenfold in www.nature.com/scientificreports/ non-extracted wood (LPAS) relative to microcrystalline cellulose (AV0X) in this study (Supplemental data file S2). Also, pathways involved in the intracellular metabolism of fatty acids, such as fatty acid β-oxidation, the TCA cycle, and the glyoxylate shunt were up-regulated (Fig. 3). Furthermore, extracellular lipase activities were induced by the extractives (Fig. 4A). These results suggest that lipase activity plays a key role in the metabolism of softwood extractives and that PgLip19028 is responsible for the lipase activity. On the other hand, OXA, which interacts with the TCA cycle and the glyoxylate shunt, was up-regulated by softwood materials in the previous study 13 , but was not induced in this study with microcrystalline cellulose and acetone softwood extractives (Fig. 3). This suggests that OXA associated with oxalic acid accumulation may be induced by unidentified wood components such as cellulose, hemicellulose, lignin, and their degradation products, but not by extractives. In addition to extracellular and intracellular lipid metabolism, CAZymes involved in the degradation of pectin and other polysaccharides (members of GH families 11, 12, 28) were significantly up-regulated by the presence of extractives. This suggested that removal of pectin and other polysaccharides by these enzymes is a first step for efficient deconstruction of extractives-coated wood materials. Optimal pH (C) and temperature (D) activities of recombinant PgLip19028. pNPD was used as a substrate. Reaction temperature for optimal pH was 25 °C, and reaction pH for optimal temperature was acetate buffer (pH 4.5 www.nature.com/scientificreports/ P. gigantea's gene expression patterns reveal multiple strategies of this fungus for overcoming the challenging composition of resin-coated microcrystalline cellulose, i.e. ABC transporter (Phlgi1_130987) possibly involved in tolerance to terpenes, glutathione S-transferase (Phlgi_101998), aldehyde dehydrogenase (Phlgi_121047), and aryl-alcohol dehydrogenase (Phigi_89048) in the transformation and detoxification of extractives (Supplemental data file S2). Notably, the significantly upregulated P450s may oxidize aliphatic and/or aromatic chemicals such as those occurring in plant defense systems and thus may be involved in the transformation and degradation of wood extractives. Cytochrome P450 transcripts, especially Phlgi_20511, Phlgi_20514, Phlgi_82811, and Phlgi_68457, were substantially induced by pine extractives (Fig. 2C; Supplemental data file S2). These upregulated transcripts belonged to the reported P450 families CYP52 (CYP5150A), CYP53 (CYP53C), CYP67 (CYP5035A), and CYP503 (CYP512B), respectively 13 . CYP52 transcripts which were the most abundant among the extractive-induced P450s are known to oxidize alkyl chains such as in n-alkanes, fatty acids, and alkyl-substituted aromatics 27,28 . In biochemical studies, members of CYP5150A and CYP5150B, upregulated in this study, are known to hydroxylate n-alkyl-substituted benzoic acid and a number of plant defense chemicals including resin acids (e.g. dehydroabietic acid), flavonoids (e.g. flavone) and coumarins (e.g. 7-ethoxycoumarin) 29 . Likewise, induced P450 genes CYP512B (CYP503 family), and CYP5035A (CYP67 family) were shown to catalyze oxidation of these plant defense chemicals. CYP53 family represented is known to hydroxylate benzoate and its derivatives 30 . Beyond these simpler aromatic compounds, P450 monooxygenases CYP512, CYP5141, and CYP5150 have also been shown to oxidize more complex aromatic compounds including polyaromatics and heteroaromatics 31 . Considering their regulation and catalytic potential, P. gigantea P450s likely play a role in the metabolism of P. taeda extractives.
Based on the sequence annotation, Phlgi1_19028 was predicted to encode lipase in the triacylglycerol hydrolases (E.C. 3.1.1.3) (Fig. 5A). The recombinant PgLip19028 showed a different molecular weight, optimal pH and temperature activities, and substrates specificities than the previously characterized lipases from basidiomycetes [20][21][22]26,32,33 . Especially, the optimal pH 4.5 is apparently lower than other lipases, but it seems consistent with P. gigantea's external environment, which is usually acidic due to secreted organic acids such as oxalic acid 34 . Additionally, from our biochemical characterization, this novel lipase releases various unsaturated fatty acids from Pinus extractives at ambient temperature under acidic conditions (Fig. 6). Hence, PgLip19028 is thought to efficiently produce unsaturated fatty acids from lipophilic extractives, allowing P. gigantea to proficiently degrade softwood. www.nature.com/scientificreports/ Lipases, α/β hydrolases (abH), are classified into 38 abH subfamilies based on amino acid sequences and 3D structures (http:// www. led. uni-stutt gart. de/). PgLip19028 is categorized in abH23 subfamily, which includes lipases reported from various pathogens, i.e. Aspergillus flavus and Fusarium graminearum, and some of these lipases are involved in penetration to the host organisms 23,24 . In conifers, extractives occur in resin canals and the surrounding parenchyma cells in sapwood, which structurally prevent fungal developments within cell lumens 35 . In addition, it has been reported that larger amounts of triglycerides were extracted from sapwood, compared to heartwood in Pinus species 9 . The growth rate of P. gigantea greatly varies depending on softwood species, i.e., it grows well on Pinus species but not on spruce, larch and fir 36 . Therefore, PgLip19028 is thought to play important roles in the penetration into Pinus sapwood by P. gigantea, and there would be value in examining P. gigantea responses to different extractives from various softwood species in future studies. Considering the wide sequence conservation (Table S4), orthologs of this lipase may generally function in the degradation of lipophilic extractives by wood decay fungi. Further genetic studies, comparative expression, and biochemical studies of this lipase within wood decay fungi will be needed for elucidating more physiological roles in the interaction between wood and fungi.

Methods
Culture conditions. The P. gigantea isolate 11061-1 was collected from loblolly pine (P. taeda) in Macon County, Alabama and sequenced (https:// mycoc osm. jgi. doe. gov/ Phlgi1/ Phlgi1. home. html) as described 13 . For each treatment, a single-basidiospore derivative of 11061-1 strain 5-6 was cultivated in triplicate on 250 mL basal salts medium 37,38 containing: (i.) 1.25 g ground (1 mm mesh) debarked loblolly pine sapwood that had been suspended in 10 mL acetone and subsequently dried in a roto-evaporator (LPAS); or (ii.) 1.25 g microcrystalline cellulose (Avicel PH101, 50uM, Fluka Chemika, Switzerland) that had been suspended in 10 mL acetone and subsequently dried in a roto-evaporator (AV0X); or (iii.) 1.25 g microcrystalline cellulose that had been suspended in 10 mL of an acetone soxhlet extract of 1.25 g, 2.5 g or 5 g loblolly pine and subsequently dried to reconstitute the extractive content at the same (AV1X), two times (AV2X) and four times (AV4X) the native loblolly pine extract concentration, respectively. All loblolly pine was ground from the same 22 cm cross sectional bolt containing 16 annual rings at waist height. After debarking, the material was primarily composed of sapwood. The use of plant parts in the present study complies with international, national and/or institutional guidelines. The composition of basal salts medium contained, per liter, 2 g of NH 4  The culture filtrate and mycelia were separated after the cultivation, and total RNA were purified from the mycelia by RNeasy Mini Kit (Qiagen) with DNase treatment (Qiagen) as previously described 40,41 . For GCMS-based metabolome analyses, the harvested entire cultures including the substrates and fungal mycelium from all the treatments were frozen in liquid nitrogen and lyophilized.
Time course of lipase activities in the culture filtrates. P. gigantea isolate 11061-1 strain CR5-6 was cultivated in triplicates in the culture medium of AV0X, AV1X, AV2X and AV4X, the culture supernatants were collected after 3, 5 and 7 days of cultivation, and protein concentration in the culture supernatants was measured by Protein Assay (Bio-Rad). To determine lipase activities in the secretomes, p-nitrophenol-dodecanoate (pNPD, Sigma-Aldrich) was used as substrate. Two microliter of 75 mM pNPD in dimethylsulfoxide (DMSO) was mixed with 50 µL of culture supernatant and 25 µL of 100 mM acetic acid buffer (pH 5.0) in a total volume of 100 µL. After the incubation at 37 °C for 30 min, 25µL of 100 mM Na 2 CO 3 was added to stop the reaction. The released pNP amount was measured at 405 nm using the standard curve for pNP (Sigma-Aldrich). One unit of lipase activity was defined as the amount of enzyme releasing one micromole of pNP per a minute.
RNA analysis methods. The transcriptomes were analyzed using Illumina RNA-Seq. Plate-based RNA sample prep was performed on the PerkinElmer Sciclone NGS robotic liquid handling system using Illumina's TruSeq Stranded mRNA HT sample prep kit utilizing poly-A selected mRNA as outlined by Illumina (https:// suppo rt. illum ina. com/ seque ncing/ seque ncing_ kits/ truseq-stran ded-mrna. html). Total RNA starting material was 100 ng per sample and 10 cycles of PCR was used for library amplification. Total RNA qualities were examined by using Fragment Analyzer (Agilent Technologies), and RNA quality numbers (RQN) scores were 7.2 ± 0.3, 7.2 ± 0.8, 8.0 ± 1.7, 6.4 ± 0.2, 6.9 ± 2.7 for AV0X, AV1X, AV2X, AV4X and LPAS, respectively. The prepared libraries were quantified using KAPA Biosystem's next-generation sequencing library qPCR kit and run on a Roche LightCycler 480 real-time PCR instrument. The libraries were then multiplexed with other libraries and sequencing was performed on the Illumina NovaSeq sequencer using NovaSeq XP V1 reagent kits, S4 flow cell, following a 2x150 indexed run recipe. Sequence data for the libraries were deposited in SRA and assigned accession numbers SRP245000 through SRP245010 and 245022 through 245025.
Illumina reads were filtered and trimmed using for artifacts, RNA spike-in reads, PhiX reads and reads containing any Ns. Quality trimming was performed using the phred trimming method set at Q6. Finally, following trimming, reads under the length threshold were removed (minimum length 25 bases or 1/3 of the original read length-whichever is longer). Filtered reads from each library were aligned to the reference genome using HISAT2 version 2.1.0 42 . Strand-specific coverage was determined using deepTools v3.1 43 and gene counts were generated using feature Counts 44 . Raw gene counts were used to evaluate the level of correlation between biological replicates using Pearson's correlation and determine which replicates would be used in the differential gene www.nature.com/scientificreports/ expression analysis. DNAStar module ArrayStar ver.16 (Madison, Wisconsin) was used to visualize the normalized reads and compute differential expression in pairwise comparisons. Unless otherwise indicated, parameters used to call genes differentially expressed between culture conditions were P value < 0.02 or P value < 0.05 when the stringency is relaxed, with > twofold change and RPKM values > 10. Differential expression of all genes were illustrated in volcano plots in Fig. S3.

Proteome analysis.
With minor modification, NanoLC-MS/MS analysis identified extracellular proteins in culture filtrates as described [45][46][47] . Filtrates from microcrystalline cellulose cultures, with (AV1X, AV2X, AV4X) or without (AV0X) addition of loblolly pine acetone extract, were filtered after 5 days and analyzed. Filtered proteins were precipitated with 10% (wt/vol) trichloroacetic acid and washed three times in cold acetone before air drying. Total proteins from the pellets were further purified via methanol/chloroform/water partitioning, where chloroform and methanol were added to pellets first, followed by water and allowed to partition with a protein interphase formed between polar and non-polar fraction. After multiple methanol washes, these purified protein preps where ultimately resolubilized in 8 M urea/50 mM NH 4 HCO 3 (pH8.5)/1 mM TrisHCl. For NanoLC-MS/MS protein identification of samples, equal amounts of total protein per sample were trypsin/LysC digested, OMIX C18 SPE purified (Agilent Technologies), and finally 2 µg loaded for nanoLC-MS/MS analysis using an Agilent 1100 nanoflow system (Agilent Technologies) connected to a hybrid linear ion trap-orbitrap mass spectrometer (LTQ-Orbitrap ELITE, ThermoFisher Scientific) equipped with an EASY-SPRAY electrospray source. Chromatography of peptides prior to mass spectral analysis was accomplished using a capillary emitter column (PEPMAP C18, 3 µM, 100 Å, 150 × 0.075 mm, ThermoFisher Scientific) onto which 2 µL of purified peptides was automatically loaded. The nanoHPLC system delivered solvents A: 0.1% (v/v) formic acid , and B: 99.9% (v/v) acetonitrile, 0.1% (v/v) formic acid at 0.50 µL/min to load the peptides (over a 30 min period) and 0.3 µL/min to elute peptides directly into the nano-electrospray with gradual gradient from 3% (v/v) B to 20% (v/v) B over 154 min and concluded with 12 min fast gradient from 20% (v/v) B to 50% (v/v) B at which time a 5 min flash-out from 50 to 95% (v/v) B took place. As peptides eluted from the HPLC-column/electrospray source, survey MS scans were acquired in the Orbitrap with a resolution of 120,000, followed by MS2 fragmentation of 20 most intense peptides detected in the MS1 scan from 380 to 1800 m/z; redundancy was limited by dynamic exclusion. Raw MS/MS data were converted to mgf file format using MSConvert (ProteoWizard: Open Source Software for Rapid Proteomics Tools Development) for downstream analysis. Resulting mgf files were used to search against forward and decoyed-reversed P. gigantea protein database via the JGI portal (https:// genome. jgi. doe. gov/ portal/ Phlgi1/ Phlgi1. downl oad. html) with a list of common lab contaminants (available at ftp:// ftp. thegpm. org/ fasta/ cRAP) to establish False Discovery Rate (23,858 total entries) using in-house Mascot search engine 2.2.07 [Matrix Science] with variable methionine oxidation, asparagine and glutamine deamidation, plus fixed cysteine carbamidomethylation. Scaffold (version Scaffold_4.7.5, Proteome Software Inc., Portland, OR) was used for spectral based quantification and to validate MS/MS peptide and protein identifications. Peptide identifications were accepted if they could be established at greater than 80.0% probability to achieve an FDR less than 1.0% by the Scaffold Local FDR algorithm. Protein identifications were accepted if they could be established at greater than 99.0% probability to achieve an FDR less than 1.0% and contained at least 2 identified peptides. Protein probabilities were assigned by the Protein Prophet algorithm 48 . Each detected protein abundance was calculated as exponentially modified protein abundances (emPAI) 19 . Proteins that contained similar peptides and could not be differentiated based on MS/MS analysis alone were grouped to satisfy the principles of parsimony. Sequence data was deposited in PRIDE (PRoteomics IDEntifications Database) and assigned identifier PXD021332 (https:// www. ebi. ac. uk/ pride/ archi ve/ proje cts/ PXD02 1332).
GC-MS analysis. About 0.4 g of the lyophilized culture samples (n = 1) of microcrystalline cellulose coated with acetone extracts of loblolly pine were Soxhlet-extracted with acetone for 6 h, dried and redissolved in chloroform for GC-MS analyses. GC-MS analyses were performed with a Varian 3800 chromatograph coupled to an ion-trap detector (Varian 4000) using a medium-length fused-silica DB-5HT capillary column (12 m × 0.25 mm internal diameter, 0.1 µm film thickness) from J&W Scientific, enabling simultaneous elution of the different lipid classes 4 . The temperature program started at 100 °C (1 min), raised to 380 °C at 10 °C min −1 , and held for 5 min. The transfer line was kept at 300 °C, the injector was programmed from 120 °C (0.1 min) to 380 °C at 200 °C min −1 , and helium was used as carrier gas at a rate of 2 mL min −1 . Compounds were identified by mass fragmentography, and by comparing their mass spectra with those of the Wiley and NIST libraries and standards. Quantification was obtained from total-ion peak area, using response factors of the same or similar compounds (palmitic acid, abietic acid, sitosterol, 1,3-dipalmitin, cholesteryl linoleate and tripalmitin). N, O-Bis(trimethylsilyl)trifuluoroacetamide (BSTFA) in the presence of pyridine was used to prepare trimethylsilyl derivatives.
Synthesis of the recombinant P. gigantea lipase Lip19028. The DNA encoding the class 3 lipase (PgLip19028), minus the native secretion signal, was synthesized by JGI service (Fig. S4). The codon-optimized gene was subcloned into an expression vector of pPICZα (Invitrogen) by using primers for PIPE cloning 49 , PgiLip19028F_PIPE_Pic, TCT CTC GAG AAA AGA CTC CCA TCG CCT GTC CAC , PgiLip19028R_PIPE_Pic, AGT TTT TGT TCT AGA TTA AGT ACA ATA AAT AGT TCCG, PicVector_PIPE_F, TCT AGA ACA AAA ACT CAT CTC AGA AGA GGA TCT GAA TAG CG, PicVector_PIPE_R, TCT TTT CTC GAG AGA TAC CCC TTC TTC TTT AGC AGC AAT GCTG (underlined sequences were overlapped). The recombinant lipase was produced by transformed Pichia pastoris overexpressing Lip19028 following the manufacturer's instructions (Invitrogen). The culture supernatant of Pichia pastoris was applied to saturated ammonium sulfate precipitation, and the resulting protein precipitate was dissolved in 100 mM acetate buffer (pH 5.0). Ten micrograms of crude protein were
The biochemical assay of the recombinant PgLip19028 from P. gigantea. Prior to lipase activity determinations, the protein concentration of crude proteins was diluted to 10 µg/mL and then 1 µg/mL of crude protein was mixed with 1.5 mM pNPD in tert-butyl methyl ether (tBE), 5% DMSO and 25 mM buffer in total volume of 160 µL. The final concentration of tBE was kept 15%, which showed the highest activity compared to 5%, 10%, 15%, 20%, 25%, and 30% tBE. After the reaction mixture was incubated at the appropriate temperature for 10 min, 40 µL of 100 mM Na 2 CO 3 was added to stop the reaction. One unit was calculated as described above. For optimal pH determination, tartaric buffer ranging from pH 3.0 to pH 3.5, sodium acetate buffer ranging from pH 3.5 to pH 5.5 and potassium phosphate buffer ranging from pH 6.0 to pH 6.5 were assessed.