Samples of mixed substrate (30% wheat bran and 70% cottonseed hull) and Cyclobalanopsis log substrate were prepared, and the mixed substrate was put into bottles. The same culture of T. fuciformis and Annulohypoxylon stygium was inoculated into bottles and logs, respectively. Both samples developed at 25°C until the fruit bodies were harvested. Tremella fuciformis SMCC174.01.1 and A. stygium SMCC238.02.34 were isolated and preserved in our lab.

The polysaccharide content of the fungi was determined using a colorimetric method using phenol and sulfuric acid (Dubois et al., 1956). The cellulose, hemicellulose, pectin, and lignin contents of the growth substrates were tested using commercial kits (BC4280, BC4440, BC1400, and BC4200; Beijing Solarbio Science & Technology Co., Ltd.). ELISA tests were also performed using commercial kits (Shanghai Huabang Biotechnology Co., Ltd.) to measure glycosyltransferase activity. The activity of pectin methylesterase (PME) was measured using commercial kits from Shanghai C-reagent Biotechnology Co., Ltd. under the conditions of 37°C and pH 7.8. Xylanase activity was identified using commercial kits from Jiangxi Jiangnanchun Biotechnology Co., Ltd. The samples were incubated at 40°C for 60 min with reagents 1 and 2 and then in a boiling water bath for 5 min. After cooling, distilled water was added to measure the OD value at 540 nm, and finally, xylanase enzyme activity was calculated. The activity of cellulase (endoglucanase) and lignin peroxidase was also investigated using commercial kits (BC2540 and BC1610; Beijing Solarbio Science & Technology Co., Ltd.). The sample was mixed with reagents 1 and 2 evenly, then bathe in a 40°C water bath for 30 min. After taking it out, it was immediately put into boiling water and boiled for 15 min. Then, reagent 3 was added and boiled in boiling water to develop color. After cooling, distilled water was added to measure the absorbance value at 540 nm, and finally, cellulase activity was calculated. Reagents 1, 2, and 3 were mixed in a ratio of 6:2:1 to form a working solution, which was preheated at 37°C. After adding the sample and the working solution, the OD values were measured at 310 nm at different reaction times to calculate lignin peroxidase activity.

Fresh samples of T. fuciformis fruit bodies (30 g) were added to water (2 L) and boiled for 30 min. The T. fuciformis soup was then filtered (using a 40-mesh sieve). The supernatant was used to fill two-thirds of a test cup for texture analysis; the weight of the filtered colloids was also recorded. A texture analyzer was used in compression mode to determine the texture of the T. fuciformis soup. The experimental parameters used were 35-mm diameter A/BE probe, 10 kg load cell capacity, returning to start option, testing speed of 1 mm/s, post-test speed of 2 mm/s, target distance of 25 mm, auto (force) 5 g of trigger type, and data acquisition rate of 250 pps. The firmness, consistency, cohesiveness, and work of cohesion of this soup were tested.

Firmness: This refers to the force corresponding to the largest positive peak on the curve. Firmness indicates the hardness of the surface of the T. fuciformis soup. Consistency: This pertains to the average of the positive peak area of the curve and the yield point to the maximum value. Consistency refers to the internal friction resistance to fluid deformation, reflecting the fluidity of the T. fuciformis soup. The greater the consistency, the worse the fluidity. Cohesion: This refers to the force value corresponding to the largest negative peak on the curve. Cohesion pertains to the size of the internal binding force required to maintain the shape of the sample, reflecting the strength of the intermolecular binding effect of the T. fuciformis soup. The greater the value, the larger the cohesion. Viscosity: This refers to the negative peak area of the curve. Viscosity is related to flow resistance, which reflects the resistance of the T. fuciformis soup when it flows. The larger the value, the greater the viscosity.

Tremella fuciformis fruit bodies (100 mg) were ground using liquid nitrogen. The homogenate was then resuspended with prechilled 80% methanol and centrifuged at 15,000g. The supernatant was diluted with mass spectrometry water until the methanol content was 53%. Then, it was centrifuged at 5,000 rpm and 4°C for 1 min, and the supernatant was transferred into a new centrifuge tube and freeze-dried into a dry powder. A 10% methanol solution was added according to the sampled volume to dissolve the powder and then injected into a UPLC-MS/MS system for analysis.

UPLC-MS/MS analysis was performed at the Novogene facility in Beijing using a Vanquish UHPLC system (Thermo Fisher, Germany) coupled with an Orbitrap Q Exactive™ HF mass spectrometer (Thermo Fisher, Germany). The samples were injected into a Hypersil GOLD column (100 mm × 2.1 mm, 1.9 μm) with a 17-min linear gradient and a flow rate of 0.2 mL/min. In a positive polarity mode, a 0.1% formic acid (FA) solution (eluent A) was used followed by methanol (eluent B); in a negative polarity mode, eluent A (5 mM of ammonium acetate, pH 9.0) was used first followed by eluent B (methanol). The solvent gradients employed were as follows: 2% B for 1.5 min, 2%–85% B for 3 min, 100% B for 10 min, 100%–2% B for 10.1 min, and 2% B for 12 min, successively. The Q Exactive™ HF mass spectrometer was used in a positive/negative polarity mode with a capillary temperature of 320°C, spray voltage of 3.5 kV, sheath gas flow rate of 35 arb, auxiliary gas flow rate of 10 arb, S-lens RF level of 60, and auxiliary gas heater temperature of 350°C. The raw UHPLC-MS/MS data were analyzed using Compound Discoverer 3.1 software (CD3.1, Thermo Fisher) to conduct peak alignment, peak picking, and quantitation of each metabolite. The software packages R (v3.4.3), Python (v2.7.6), and CentOS (v6.6) were used to conduct the statistical analyses. The metabolites were annotated using three databases: KEGG (https://www.genome.jp/kegg/pathway.html), HMDB (https://hmdb.ca/metabolites), and LIPID MAPS (http://www.lipidmaps.org/). Partial least-squares discriminant analysis (PLS-DA) was performed using metaX (Wen et al., 2017). Univariate analysis (t-test) was used to calculate statistical significance (p-value). Metabolites with variable importance in projection (VIP) values >1, p-value <0.05, and fold change (FC) ≥2 or ≤0.5 were deemed to be differentially abundant metabolites (DAMs). The data used to generate clustered heatmaps were first normalized using the z-scores of the intense areas of the differential metabolites. The data were then plotted using the heatmap function in the R package. A p-value <0.05 was considered to be statistically significant, and correlation plots were plotted using the corrplot package in R. The KEGG database was used to identify the functions of these metabolites and their metabolic pathways. Metabolic pathways satisfying the condition x/n > y/N were considered to be enriched (and those with a p-value <0.05 were also deemed to have been significantly enriched).

Additional samples were individually ground in liquid nitrogen lysed with SDT lysis buffer and then subjected to ultrasonication on ice for complete dissolution. The samples were then incubated at 95°C for 8 min and then centrifuged at 12,000g for 15 min at 4°C. The resulting supernatant was collected, 10 mM of DTT was added, and then incubated at 56°C for 1 h. Sufficient IAM was added and incubated at room temperature in the dark for 1 h. Next, four volumes of precooled (–20°C) acetone were added and precipitated at −20°C for at least 2 h. The samples were centrifuged at 4°C and 12,000g for 15 min, and the precipitate was collected. One milliliter of precooled (−20°C) acetone was added during the resuspension and washing of the precipitate. The mixture was centrifuged at 12,000g for 15 min at 4°C. The precipitate was collected and air-dried, and the protein precipitate was dissolved using an appropriate amount of protein dissolution solution (8 M of urea, 100 mM of TEAB, pH = 8.5). BSA protein solution was used as a standard in the protein concentration tests (performed using 12% SDS-PAGE gel electrophoresis). Qualified protein samples were dissolved in DB. After digestion using trypsin and CaCl₂, the supernatant was loaded onto a C18 desalting column and washed with buffer eluent.

More precisely, a 0.1-M solution of tetraethylammonium bromide (TEAB) buffer was supplemented to the samples, along with a tandem mass tag (TMT) labeling reagent dissolved in acetonitrile. All the labeled samples were mixed in equal volumes, desalted, and lyophilized. For separation, a gradient elution was established utilizing mobile phases: phase A comprised 2% acetonitrile, adjusted to pH 10.0 with ammonium hydroxide, and phase B consisted of 98% acetonitrile. Fractionation of the samples was conducted on a Rigol L3000 HPLC system equipped with a C18 column (Waters BEH C18, dimensions 4.6 mm × 250 mm, particle size 5 μm), with the column oven maintained at 45°C. All resultant fractions were dried under vacuum and reconstituted in 0.1% (v/v) aqueous formic acid (FA). These fractions were then subjected to shotgun proteomics analysis using an EASY-nLC 1200 UHPLC system (Thermo Fisher Scientific), interfaced with a Q Exactive HF-X mass spectrometer (Thermo Fisher Scientific) operated in the data-dependent acquisition (DDA) mode.

The samples were injected into a homemade C18 Nano-Trap column (dimensions: 4.5 cm × 75 μm, 3 μm). The subsequent peptide separation was achieved in a homemade analytical column (15 cm × 150 μm, 1.9 μm), using a linear elution gradient. The separated peptides were then analyzed using a Q Exactive™ HF-X mass spectrometer (Thermo Fisher) fitted with a Nanospray Flex™ ion source. The spray voltage was adjusted to 2.3 kV, while the ion transfer capillary temperature was maintained at 320°C. A full scan encompassed the m/z range from m/z 350 to 1,500 with a resolution of 60,000 (at m/z 200). The target value of the automatic gain control (AGC) was set to 3 × 10⁶, and the maximum ion injection time was 20 ms. Within the full scan, the top 40 most abundant precursors were selected for fragmentation via higher-energy collisional dissociation (HCD) and subsequent MS/MS analysis, where the resolution was 45,000 (at m/z 200) for 10 plex. The AGC target was set to 5 × 10⁴, with a maximum ion injection time of 86 ms; the normalized collision energy was set to 32%; the intensity threshold was set to 1.2 × 10⁵; and the dynamic exclusion parameter was configured to 20 s.

Proteome Discoverer (PD) software (v2.4, Thermo Fisher) was used to analyze the spectra obtained in each run and compare the results with the UniProt database. The mass tolerance for the precursor ion was set to 10 ppm, and that for the product ion to 0.02 Da. Carbamidomethyl was specified as a fixed modification, and oxidation of methionine (M) and TMT plex were specified as dynamic modifications. Acetylation, TMT plex, Met-loss, and Met-loss+Acetyl were specified as N-terminal modifications in PD. A maximum of two missed cleavage sites was allowed. The software PD was used to further filter the retrieval results to improve the quality of the analysis results. The protein quantitation results were statistically analyzed using t-tests. Proteins whose quantitation results were significantly different in the experimental and control groups (p < 0.05 and FC > 1.2 or FC < 0.83) were assumed to be differentially expressed proteins (DEPs). The InterProScan program against the non-redundant protein database was used for Gene Ontology (GO) functional analysis (Jones et al., 2014). The KEGG and Clusters of Orthologous Groups (COG) databases were used to study the protein families and pathways. Carbohydrate-active enzyme was annotated using the dbCAN2 meta server (Zhang et al., 2018). Raw omics data have been uploaded to iProX (IPX0012156000) and MetaboLights (REQ20250604210974).

Data met the assumptions of normality (Shapiro–Wilk p > 0.05) and homogeneity of variances (Levene’s p = 0.18). Differences among groups were analyzed by one-way ANOVA with Tukey’s post-hoc test. To account for multiple comparisons across three experimental conditions, p-values were adjusted using the Benjamini–Hochberg procedure (FDR = 0.05).

According to the calculation formula (x/n > y/N), the 21 most enriched metabolic pathways were selected. x represents the number of DEPs associated with this pathway, y represents the number of all proteins associated with this pathway, n represents the number of DEPs annotated by KEGG, and N denotes the number of all proteins annotated by KEGG.

The Pathview package was used to co-analyze the proteome–metabolome data. After supplying Pathview with gene or compound data and specifying the target pathway, the software automatically downloaded the pathway graph data, parsed the data file, mapped the user data to the pathway, and rendered a pathway graph containing the mapped data (Luo and Brouwer, 2013). The software packages R (v3.4.3), ggplot2 (v3.3.4), and grid (v3.4.3) were used to harvest bubble diagrams of the different proteins and metabolites both enriched in the KEGG pathways. The software packages R (v4.0.3), ggplot2 (v3.3.5), and pathview (v1.30.1) were used to identify the different enzymes and metabolisms involved in the same pathways.

All the results presented are expressed in the form: means ± standard deviation (SD). Differences between groups are deemed to be statistically significant if p <0.05.

To investigate the differences between the T. fuciformis cultivated in the two different substrates, the metabolite profiles of TF1- and TY3-grown specimens were analyzed and quantitatively compared using LC-MS/MS (in both positive and negative ionization modes). As a result, 1,221 metabolites were identified.

The PLS-DA results ( Figures 2A, B ) show that the metabolic profiles of TF1- and TY3-grown specimens were significantly different in general. In total, 488 DAMs were screened with a threshold value of FC ≥1.5 (or ≤0.7), VIP >1, and p-value ≤0.05 ( Supplementary Table S1 ). The results of a hierarchical clustering analysis of these DAMs are presented as heatmaps in Figures 2C, D , wherein the shorter the cluster branch, the higher the similarity of the samples. The number of DAMs that were found to be significantly upregulated (colored red in Figures 2C, D ) was 317, while 171 of them were found to be significantly downregulated (colored blue in Figures 2C, D ). These DAMs were subsequently divided into different subclasses ( Supplementary Table S2 ). The metabolites in the TF1-grown samples that were significantly upregulated mainly corresponded to amino acids, peptides and analogs, benzoic acids and derivatives, purine ribonucleotides, pyridinecarboxylic acids and derivatives, and indolyl carboxylic acids and derivatives. Carbohydrates and carbohydrate conjugates, fatty acid esters, fatty amides, and triterpenoids were mostly downregulated in the TF1-grown samples compared with those grown on TY3. In the subclass of “carbohydrates and carbohydrate conjugates,” the number of upregulated DAMs in TY3-grown fungi is twice the number of downregulated DAMs in TF1, which may lead to the polysaccharide content being higher in the TY3-grown fungi. The results of the proteomics analysis were then used to try to identify the enzymes responsible for these metabolic changes.

A TMT-based proteomic approach was implemented to compare T. fuciformis cultivated on the mixed substrate TF1 and the Cyclobalanopsis substrate TY3. A total of 6,736 peptides were identified, which were assembled into 1,685 proteins ( Supplementary Table S3 ). GO, COG, and KEGG analyses were subsequently performed to acquire more information about the function of these proteins, giving the results shown in Figure 3 .

The GO results ( Figure 3A ) demonstrate the function and enrichment levels of 1,167 proteins associated with various biological processes and cellular features. The oxidation–reduction process was found to be the most enriched (115) among the biological processes. These, presumably, generate energy for other processes, such as metabolic processes (59), translation (49), protein phosphorylation (42), and carbohydrate metabolic processes (33). This outcome is consistent with the finding that the number of proteins associated with ATP binding is the most enriched (214) category in the molecular function group, as well as protein binding (121), nucleic acid binding (70), GTP binding (62), oxidoreductase activity (54), and catalytic activity (50). In the category of cellular components, the main differences involve proteins associated with membrane structure, e.g., integral components of the membrane (43), membrane (29), cytoplasm (23), and membrane coat (10).

The KEGG pathway results ( Figure 3B ) show that the 1,685 proteins enrich 21 pathways. The metabolic pathways were the most enriched, especially pathways related to the metabolism of carbohydrates (157) and amino acids (130). Other, less significant, changes were related to genetic information processing, cellular processes, environmental information processing, and organismal systems.

To collect more information regarding the functions of each protein, a COG analysis was implemented ( Figure 3C ). In this case, 1,092 proteins were classified into 25 categories. The proteins most enriched were related to translation, ribosomal structure, and biogenesis (168); post-translational modification, protein turnover, and chaperones (122); and energy production and conversion (93), reflecting a status of hyperactive expression. The following functions were also enriched: amino acid transport and metabolism (105) and carbohydrate transport and metabolism (77). This coincides with the aforementioned changes in metabolic function.

Threshold values of FC ≥1.2 (or ≤0.8) and p ≤0.05 were set to determine the DEPs in the T. fuciformis grown on TF1 and TY3. As a result, 668 DEPs were identified, 345 of which were upregulated, and the remainder (323) were downregulated ( Figure 2E ). To elucidate the main biochemical metabolic pathways and signal transduction pathways the DEPs were associated with, the DEPs were categorized into pathways they were involved with: 66 significantly enriched pathways were thus found ( Supplementary Table S4 ). The DEPs mostly enriched were related to the metabolism of amino acids, carbohydrates, lipids, cofactors, and vitamins.

With regard to carbohydrate metabolism, the DEPs are implicated in diverse pathways, including the pentose phosphate pathway, ascorbate and aldarate metabolism, starch and sucrose metabolism, pyruvate metabolism, inositol phosphate metabolism, propanoate metabolism, and fructose and mannose metabolism (also encompassing the biosynthesis of various N-glycans). With respect to amino acid metabolism, the prominent DEPs are primarily engaged in the biosynthesis and/or metabolic processes of lysine, arginine, tryptophan, tyrosine, histidine, proline, alanine, aspartate, glutamate, and phenylalanine. In terms of lipid metabolism, the DEPs are associated with the synthesis and degradation of fatty acids and ketone bodies, as well as the metabolism of arachidonic acid, ether lipids, steroids, and glycerolipids. Additionally, other notable DEPs play crucial roles in maintaining regular physiological activities (e.g., replication and repair, translation, folding, sorting and degradation, cell growth and death, transport and catabolism, etc.).

SThe CAZyme database offers information about enzymes that degrade, modify, or create glycosidic bonds (and also conjugation to nucleic acid, lipids, polyphenols, proteins, and other compounds) (Ma et al., 2021). In our study, 161 enzymes in proteomes were annotated using the CAZyme database ( Supplementary Table S5 ), and 67 of them were DEPs ( Table 1 ). These DEPs were related to carbohydrate synthesis [via glycosyltransferases (GTs)], carbohydrate degradation [glycoside hydrolases (GHs), carbohydrate esterases (CEs), polysaccharide lyases (PLs), and auxiliary activities (AAs)], and carbohydrate recognition [carbohydrate-binding module (CBM)] (André et al., 2014).

More specifically, 10 enzymes related to carbohydrate synthesis (GTs) were upregulated in TF1-grown specimens compared with TY3-grown specimens, and 18 of them were downregulated. As for carbohydrate degradation, 16 enzymes were upregulated, while 13 were downregulated. Four enzymes related to carbohydrate recognition were found to be increased in TF1-grown fungi, and six were decreased in TY3-grown fungi. The TY3 Cyclobalanopsis substrate thus appeared to have a greater ability to promote carbohydrate synthesis. As a result, the fruit bodies of TY3-grown T. fuciformis and its soup both have higher polysaccharide contents ( Figures 1B, C ).

The polysaccharides are mainly synthesized by GTs. The polysaccharides in fungal cells are mainly cellulose, hemicellulose, pectin, and lignin. GT2 occupies a dominant position in the synthesis of cellulose (Wei et al., 2015). The annotated enzymes of cellulose synthesis (F5HB36, Q5KBD0) belonging to GT2 were both downregulated in TF1-grown T. fuciformis compared with the TY3-grown specimens ( Table 1 ). The synthesis of hemicellulose is more complicated than that of cellulose because of its more complex structure. Therefore, not only GT2 but also GT8, GT34, GT37, GT43, GT47, and GT61 were involved in hemicellulose synthesis (Dawn et al., 2012; Zabotina et al., 2012; Pauly et al., 2013). Xyloglucan β-galactosyltransferase (GT47, J9VI09), which adds a galactose side chain to the xyloglucan skeleton of hemicellulose and also participates in the side chain synthesis of xylose polygalacturonic acid, was found to be downregulated in Table 1 (in TF1-grown fungi compared with TY3) (Zabotina et al., 2012; Wei et al., 2015). TY3-grown specimens having a higher content of polysaccharides may therefore be related to more GTs being upregulated in the TY3-grown T. fuciformis (compared to that with TF1 specimens). ELISA tests of the GTs also verified that TY3 species have higher GT activity than TF1 species ( Figure 4A ).

Fungi possess a variety of enzymatic systems that play different roles in degrading the cell walls of different plants (Sarkar et al., 2009). This includes T. fuciformis when it vegetates on Cyclobalanopsis bed log or a mixture consisting of wheat bran and cottonseed hull. The cell walls of plants mainly comprise cellulose, hemicellulose, and pectin. Cellulose is composed of unbranched glycan chains created by the formation of β-1,4 glycosidic bonds. The CAZyme annotation results indicate that endo-β-1,4-glucanase (GH5, J9VRQ0) hydrolyzes glycosidic bonds in cellulose chains from the inside. Then, β-glucosidase (GH3, J9VZB3) can degrade the glucan chains thus formed. The auxiliary activity enzyme monooxygenase (AA9, Q5KGP3) also contributed to the degradation of lytic cellulose ( Figure 5A ) (Sweeney and Xu, 2012). The annotated endoglucanase (GH5, J9VRQ0), β-glucosidase (GH3, J9VZB3), and monooxygenase (AA9, Q5KGP3) were all upregulated in TF1-grown fungi compared to TY3-grown fungi ( Table 1 ). The test results obtained for the cellulase activity (endo-β-1,4-glucanase, EC 3.2.1.4) also verified that it was elevated in TF1-grown fungi compared to TY3-grown fungi ( Figure 4B ). Hemicellulose has a different skeletal and more complex side chain structure than cellulose. Its main chain structure takes many forms, including xylan, galactomannan, xyloglucan, β-(1,3;1,4)-D-glucan, and other types. The annotated xyloglucan-specific endo-β-1,4-glucanase (GH12, Q5KCX7) can degrade the xyloglucan skeleton (Master et al., 2008) ( Figure 5B ). The hemicellulose skeleton of β-(1,3;1,4)-D-glucan, like cellulose, is composed of β-D-glucose; however, due to the differences between the bond types formed during their synthesis, except endo-β-1,4-glucanase (GH5, J9VRQ0), α-1,3-glucanase (GH71, F5H989) is also needed in the degradation of β-(1,3;1,4)-D-glucan (Mc et al., 2003; Boyce and Walsh, 2007) ( Figure 5C ). The enzymes required for the degradation of the main hemicellulose chain—xyloglucan-specific endo-β-1,4-glucanase (GH12, Q5KCX7) and α-1,3-glucanase (GH71, F5H989)—are both downregulated in TF1-grown fungi compared to those in TY3-grown fungi ( Table 1 ). There are five annotated degradation enzymes of hemicellulose belonging to GH5, and two of them (J9VRQ0, A0A226BGM6) were upregulated, while three of them (Q5KGL9, A0A225YW44, F5HHP5) were downregulated ( Table 1 ). The hemicellulose in nature is mainly xylan, and the activity of xylanase (EC 3.2.1.8) was detected to be upregulated in TF1-grown fungi compared to TY3-grown fungi ( Figure 4C ).

Pectins possess the most complex polysaccharide structures in plant cell walls. Four main types are based on polygalacturonic acid (HG), xylose polygalacturonic acid (XG), rhamnose polygalacturonic acid I (RG-I), and rhamnose polygalacturonic acid II (RG II) (Himmel et al., 2007). More than 60% of pectin is based on HG. The calcium bridges that occur between HG (Willats et al., 2001) and araban or galactan units in RG-I side chains (Zykwinska et al., 2005) and also boron bridges between RG-II side chains (Tavares et al., 2015) are the main factors influencing the pore sizes of plant cell walls. The pore size controls the level of exposure that occurs between cellulose, hemicellulose, and the hydrolytic enzyme GH, thereby affecting the efficiency of biomass degradation (Fleischer et al., 1999; Ridley et al., 2001; Tavares et al., 2015). HG-modifying enzymes (including PMEs, pectin acetylesterases, pectate lyases, and polygalacturonases) change the pectin content or branching pattern and thereby reduce the content of HG (which limits the solubility of the major components of cell walls). Thus, they increase the solubility of the pectin, xylan, and other hemicellulose components, which improves the efficiency of the enzymatic hydrolysis process used to degrade biomass (Biswal et al., 2014; Sénéchal et al., 2014; Tomassetti et al., 2015). PME (EC 3.1.1.11) excises the methoxyl groups to form free carboxyl groups on the galacturonic acid chain and methanol ( Figure 5D ). TF1-grown T. fuciformis was found to have higher PME (CE8, A0A225ZTU9) activity than the TY3-grown fungi according to proteomics results ( Table 1 ). However, based on the kit tests, the PME activity in TY3-grown fungi was found to be higher than that in TF1-grown fungi ( Figure 4D ). PMEs were encoded by a super-large family of genes in several species, and CE8 was one of them. This single protein (CE8, A0A225ZTU9) on its own cannot represent the activity of the entire PME family.