Cell growth and concentration of extracellular and intracellular protein.
To generate an effective secretome with active extracellular enzymes in Pseudomonas putida, P. putida cells were first grown on M9 medium with 5 g·L-1 glucose and 1 g·L-1 NH4Cl for 8 days. The optical density (OD600) and extracellular and total protein (intracellular + extracellular) concentrations were measured.
Figure 1 depicts the cell density (OD600), and extracellular and total protein concentrations over time. As shown in Fig. 1a, with 5 g·L-1 glucose, cells grew fast for the first 72 h of fermentation, and the optical density reached 2.1 and maintained a similar level until the end of the fermentation.
A distinct pattern was revealed for protein concentration. As shown in Fig. 1b (blue line), the extracellular protein concentration reached 1.1 mg·mL-1 at 24 h and dropped to 0.6 mg·mL-1 at 48 h, then gradually increased to 0.9 mg·mL-1 at 144 h and further decreased to 0.4 mg·mL-1 at the end of fermentation. The same trend was also observed for total protein concentration (red line). The total protein concentration reached 1.2 mg·mL-1 at 24 h and dropped to 0.7 mg·mL-1 at 48 h, then gradually increased to 1.0 mg·mL-1 at 144 h and further decreased to 0.7 mg·mL-1 at the end of fermentation. The total protein concentrations were higher than the extracellular protein, including the lysate intracellular protein. These results indicated that the P. putida cells secreted protein to extracellular space and the extracellular protein concentration fluctuated over time. However, further enzyme activity assays are still required to exam whether these proteins are secreted as active form or not.
Extracellular vs. intracellular enzyme activity.
P. putida cells secreted proteins to extracellular space when grown on glucose. However, whether the secretome harbors active ligninolytic enzymes still requires specific investigation. The phenolic substrate 2,6-dimethoxyphenol (DMP) is usually used to track laccase-like oxidases and peroxidase activities [17]. Therefore, oxidases, peroxidases, and Mn2+-oxidizing peroxidases were tracked in the culture supernatants and intracellular over 8-day incubations. In Fig. 1c, extracellular oxidase activity was detected at a low level at 24 h, and gradually increased to 0.25 U/L at 120 h. In Fig. 1d, the profiles for intracellular enzymes are similar to extracellular enzymes. The oxidase activity was maintained at a low level at 24 h and then gradually increased to 0.12 U/L at 120 h.
The peroxidase activity (e.g., Dyp) was also measured with DMP and H2O2 addition [17]. In Fig. 1c, extracellular peroxidase activity was detected at a low level at 24 h and gradually increased to 1.8 U/L at 120 h. In Fig. 1d, the profiles for intracellular enzymes are similar to extracellular enzymes. The peroxidase activity was maintained at a low level at 24 h and then gradually increased to 1.6 U/L at 120 h.
DMP was also used to measure the reaction’s Mn2+-peroxidase activity with H2O2 and Mn2+ [17]. In Fig. 1c, extracellular peroxidase activity was detected at a low level at 24 h, and gradually increased to 2.5 U/L at 120 h. In Fig. 1d, the profiles for intracellular enzymes are similar to extracellular enzymes. The Mn2+-peroxidase activity was maintained at a low level at 24 h and then gradually increased to 1.7 U/L at 120 h.
Collectively, the results in Fig. 1c demonstrate that P. putida secreted the enzyme to the extracellular space, and the maximum enzyme activity appeared at 120 h of fermentation. The intracellular enzyme activity showed a similar pattern as extracellular in Fig. 1d, with the maximum enzyme activity at 120 h. Therefore, the extracellular enzyme at 120 h of fermentation was selected for the following lignin degradation reaction.
Amount of alkali lignin degraded.
The lignin degradation reaction solution was set up with secretome and 2 g·L-1 alkali lignin with or without 0.1 mM H2O2 and 0.1 mM Mn2+ addition and reacted for 5 days. Apart from the secretome alone treatment, H2O2 and/or Mn2+ were added separately to activate the capacity of peroxidase and Mn2+-peroxidase performance on lignin degradation. In Fig. 2, when individual secretome was used for lignin degradation reaction, the amount of alkali lignin degraded only reached 8.1%, which was substantially lower than that for either H2O2 or H2O2 with Mn2+ addition. When H2O2 was added to the reaction, the amount of alkali lignin degraded reached 13.5 %. With the presence of both H2O2 and Mn2+, the alkali lignin degradation rate reached a maximum value of 14.5%. On the other side, the effectiveness of H2O2 alone has also been tested for comparison, with the alkali lignin degradation rate as of 7.8% only.
GC-MS analysis of lignin breakdown products.
GC-MS analysis was performed to identify the lignin breakdown products with secretome and the addition of H2O2 and Mn2+. The NIST library was used to identify and assign the chromatographic peaks. The amount of each aromatic degradation product was determined from the corresponding peak area in the chromatogram.
The GC-MS chromatogram identified aromatic compounds, furan, aldehydes, esters, organic acids, and alkanes among different samples (Figs S2-S4, and Table S1-S2). These identifications should be interpreted as unvalidated candidates; some may be incompatible with known or plausible lignin degradation mechanisms.
Peak area for all observed compounds was analyzed and shown in Fig. S4, which were further separated as four groups after being compared with their alternatives in lignin control (Table S2). Notably, when lignin was treated with H2O2, a wide range of aromatic monomers and dimers (No. 1, 4, 5, 6, 12, 14, 16-18, 20-21) were significantly increased. These compounds were only present in limited amounts or even not detected when treated with secretome alone or in the presence of H2O2 and Mn2+, suggesting the non-specific oxidative cleavage (e.g., H2O2) showed stronger products release ability compared to enzymatic cleavage. Moreover, the peak area of identified compounds among all the secretome treatments showed a similar distribution pattern (compounds No. 3, 7, 8, 9, 10, 13, 14, 25, 30, 31, 33) but with difference abundance, suggesting the increased compounds might be associated with different activated enzymes. For example, phenol, 2-methoxy-4-propyl- and 3-benzofurancarboxylic acid, 2,3-dihydro-2-methoxy-, methyl ester, trans- (compound 8 and 13) were increased in secretome alone treatment and maintained the similar level or decreased by H2O2 and Mn2+ addition, indicating these products might be associated with oxidases behavior. Similarly, the peak area of phenol, 4-ethyl-2-methoxy- and acetovanillone (compounds 3 and 9) were enhanced by H2O2, and vanillic acid (compound 11) was enhanced by Mn2+ addition, suggesting these compounds might be associated with peroxidase behavior. How lignin structural changes were impacted by these treatments was further explored by NMR analysis.
2D HSQC NMR spectra from the hydrolyzed lignin.
The cleavage of lignin C-O-C linkages is vital in the bacterial lignin degradation process. The quantification of each linkage is based on the volume integration of cross-peak contours in the HSQC spectra. In this study, lignin linkages such as β-O-4, β-β, β-5, and β-1 were cleaved with H2O2 or secretome treatments (Figs 3a-j and S5, Tables 1 and S3). As shown in Table 1, in general, the cleavage of β-5 was less extensive than the cleavages of β-O-4, β-β, and β-1, which indicated that β-5 bonds were relatively more stable. The cleavage of these linkages was increased when lignin was treated with H2O2. Among all the detected linkages, β-O-4 bonds exhibited the most extensive cleavage, demonstrating the non-enzymatic chemical reaction with H2O2 mainly cleavage β-O-4 bonds. In contrast, when lignin was treated with secretome alone, β-O-4 bonds only exhibited limited cleavage compared to that with H2O2. While the cleavage of β-β, β-5, and β-1 showed increased cleavage compared to that with H2O2, suggesting the secretome alone may only exhibit oxidase ability and cleavage the C-C bond. When introducing the H2O2 with the secretome, β-O-4 bonds were extensively decreased, β-β and β-1 were further decreased compared to secretome alone. Hydrogen peroxide activates the peroxidase and further cleavages the β-O-4 bonds. In addition, the presence of Mn2+ further enhanced the peroxidase performance and resulted in further enhanced β-O-4, β-β and β-1 bonds cleavage.
Secretome and cellular proteome profiles in P. putida.
GC-MS and NMR results demonstrated that P. putida secretome is able to depolymerize lignin. Enzyme assay revealed that the secretome exhibited the oxidase and peroxidase activity. However, isoenzymes with similar activities cannot be distinguished based on simple activity analysis. To deeply characterize the secretome, mass spectrometry-based global proteomics was utilized to profile the proteome of secretome and intracellular extracts in P. putida KT2440. Bacterial secretome and cell pellet were harvested at 120 h of fermentation with the highest oxidase and peroxidase activity. Only those proteins presented in more than three replicates were considered as reliable detection and quantitation. Results showed that 1312 proteins were identified in the secretome sample, which was lower than that in the intracellular extracts for 2388 identified proteins. Around 94.8% of the secretome proteins were shared with the intracellular protein, and only 68 proteins were exclusively detected in the secretome (Fig. 4a).
Secretome proteins were then organized by function and presented in Fig. 4b. Despite the other function and unknown function groups, the oxidoreductase group was the most abundant among the rest of the functional groups, which accounted for 14.1% in secretome proteome. Moreover, six glutathione S-transferases were detected and grouped in the transferase group. A deeper analysis of oxidoreductase is presented in Fig. 4b as well. The dehydrogenase group was the most abundant (42.2%), followed by reductase (17.3%), peroxidase (7.6%), oxidase (6.5%), monooxygenase (4.9%), and dioxygenase (2.2%). Notably, there are 14 peroxidases detected, such as Dyp-type peroxidase (PP_3248), cytochrome c551 peroxidase (PP_2943), and alkyl hydrogeperoxidase reductase (e.g., ahpC). Besides, 12 oxidases were detected, including three multicopper oxidases (copA/B and CumA). Besides, a couple of dehydrogenases were detected in the secretome, including NAD(P)H dehydrogenase (e.g., PP_1644), choline dehydrogenase (e.g., betA), aldehyde dehydrogenases (e.g., aldB-I), and alcohol dehydrogenases (e.g., PP_2827). In addition, some hydrogen peroxide alleviating enzymes were also detected, such as catalase (e.g., katG), superoxide dismutase (sodB), and thioredoxin (e.g., trx). Overall, proteomics analysis revealed the oxidoreductase enzymes in the secretome, which not only confirmed the enzyme assay results but also revealed the specific ability to selectively depolymerize lignin.
Lignin degradation pathways of P. putida secretome.
Based on the enzyme activity analysis, GC-MS, NMR, proteomics results, the proposed lignin degradation reaction pathways are shown in Fig. 5 [21, 27, 37, 40-46].The enzymes identified that might be involved in the lignin catabolism pathway were summarized in Table S4. Overall, the proposed lignin degradation pathways include cleavage of β-aryl ether (β-O-4), resinol (β-β), phenylcoumaran (β-5), and spirodienone (β-1) linkages. Figure 5 demonstrated that vanillin (compound 7) and vanillic acid (compound 11) are the most important intermediates for all these linkages.
Limited β-O-4 bond cleavage (Table 1) was observed when lignin was treated with secretome alone, possibly catalyzed by the NAD(P)H dehydrogenases (similar to LigD/N/L in Sphingobium sp. SYK-6) [44] and multicopper oxidase (e.g., CopA) in the enzyme system through α-OH oxidation and Cα-Cβ cleavage forming vanillic acid (compound 11). Besides, proteomics results also revealed that glutathione S-transferases (GST), glutathione reductases (gor), and choline dehydrogenase (betA) were presented in the secretome, which functioned as LigE/F, LigG, and HpvZ in Sphingobium sp. SYK-6 [42, 44, 47].Thus, in parallel to the multicopper oxidase route, oxidized β-O-4 linkage could also be converted to vanillic acid with the assist from GST, betA, and gor. In addition, NMR results revealed that the presence of H2O2 with secretome and lignin enhanced the β-O-4 cleavage and GC-MS demonstrated the peak abundance of vanillic acid (compound 11) increased with the H2O2 and Mn2+ addition. Proteomics results detected the Dyp-type peroxidase (PP_3248) in the secretome. Therefore, the β-O-4 linkage was degraded by Dyp-type peroxidase, forming vanillic acid as the degradation product (Fig. 5).
NMR results revealed the presence of secretome alone with lignin degraded the resinol linkage (β-β). Based on the homologous alignment of the amino acid sequence in NCBI database (blastp), FAD-binding oxidoreductase (PP_5154) exhibited the 62.8% similarity to pinoresinol α-hydroxylase in Pseudomonas sp. SG-MS2 [48].Therefore, resinol linkage might be converted to vanillin and vanillic acid by FAD-binding oxidoreductase (Fig. 5). Besides, the addition of H2O2 stimulated the resinol linkage cleavage, suggesting the peroxidase might be involved in β-β bond degradation. Therefore, peroxidase might be involved in the resinol linkage cleavage and forming vanillic acid as the intermediate (Fig. 5).
NMR results revealed the presence of secretome alone with lignin degraded the phenylcoumaran linkage (β-5). Notably, choline dehydrogenase (betA) exhibited 40% of similarity to phcC/D in Sphingobium sp. SYK-6 based on blastp analysis in NCBI for P.putida KT2440 [44, 49, 50]. Proteomics results demonstrated the aldehyde dehydrogenase (aldA/aldB-I) presented in P. putida secretome (Table S4). However, lignostilbene α,β-dioxygenase (lsdD) in Sphingobium sp. SYK-6 did not exhibit the homology protein in P. putida after the blast analysis. There might be other degradation mechanisms in P. putida KT2440 for stilbene structure which requires future investigation. Therefore, β-5 linkage degradation pathway of P. putida KT2440 is only partially consistent with the route of Sphingobium sp. SYK-6 (Fig. 5) [44, 49, 50]. In addition, the presence of H2O2 only slightly increased the β-5 linkage compared to secretome alone, suggesting the peroxidase might not involve in β-5 cleavage.
NMR results demonstrated the secretome alone contributed to spirodienone linkage cleavage, suggesting the β-1 bond cleavage capacity in the secretome. NCBI blast revealed there was no homologous protein for lsdE and lsdA in P. putida KT2440 [41].Therefore, there might be other mechanisms for β-1 linkage in P. putida. Besides, the presence of H2O2 stimulating the β-1 cleavage, and Mn2+ addition further enhanced cleavage, suggesting the peroxidase might be involved in β-1 cleavage as well. Therefore, the β-1 linkage degradation pathway for P. putida KT2440 was presented in Fig. 5.