Background

Lignocellulosic biomass is an easily available, low cost and renewable feedstock for biofuel which is an important direction for the development of renewable energy. Currently, the process for pretreatment of lignocellulosic biofuel requires removal or delocalization of lignin, which might generate aromatic compounds that inhibit enzymatic hydrolysis and fermentation [1]. Due to its rich aromatic content, lignin is a valuable waste from the biomass industry [2]. Researchers worldwide are focusing on lignin and its components for conversion into value-added products. Lignin is a complex aromatic heteropolymer derived from the condensation of hydroxyphenylpropane monomers and comprises various ether bonds and carbon–carbon bonds [3]. It is mainly composed of three basic monomers: guaiacyl (G) units, syringyl (S) units and p-hydroxyphenyl (H) units [4, 5]. Microorganisms that degrade lignocellulose are widely distributed in nature, and known examples include bacteria and fungi. Among them, fungi and some bacteria have mainly been studied, and white-rot fungi, and brown rot fungi in particular have obvious degradation effects on lignin [6, 7], but the development of industrial applications has been difficult [8]. Some bacteria also participate in lignin degradation. Several typical lignin-degrading bacteria, such as Rhodococcus, Pseudomonas, Sphingobium and Sphingomonas, have been identified [9,10,11,12]. Streptomyces are among the identified bacteria capable of degrading lignin [13]. Pasti et al. [14] isolated 11 strains of actinomycetes from the intestines of termites and analysed their ability to degrade lignocellulose, lignin and carbohydrates. Other researchers have also screened streptomyces from the soil that degrade lignin [15,16,17].

There are many kinds of enzymes involved in the degradation of lignin, including laccase (Lac), lignin peroxidase (LiP), manganese peroxidase (MnP), multifunctional peroxidase (VP), and dye decolouring peroxidase (DyPs) [18]. The specific mechanism for biological degradation of lignin needs to be studied further. Lignin is rich in high-value degradation intermediates, such as vanillin, guaiacol, catechin and protocatechin [19]. At the same time, the study of the metabolic mechanism of lignin and its transformation and utilization is an important part of the research needed for production of “second-generation biofuels” [20].

Hundreds of lignin derivatives have been identified in studies of bacterial degradation of lignin [21]. Because of the complex degradation mechanism, the interpretation of metabolic pathways and intermediate products is an important part of understanding the degradation of lignin. With the development of sequencing technology and bioinformatics, genomics research has become an important method for studying degradation mechanisms. Researchers are paying increasing attention to analysing metabolism of intermediate products through annotations of related degradation genes. Pseudomonas putida is a lignin-degrading bacterium that has been studied earlier. Lin et al. [22] identified several lignin-degrading enzymes, including haem peroxidase, from its genome and constructed five lignin metabolic pathways. Masai et al. [23] conducted a series of studies on lignin degradation and related genes in Sphingomonas paucimobilis SYK-6 and established a relatively complete lignin degradation metabolic pathway. Niewerth et al. [24] determined and analysed the whole genome sequence of Arthrobacter sp. Rue61a in soil, and the results showed that it has an aromatic degradation pathway that utilizes the characteristic products of lignin degradation, reflecting the saprophytic capabilities and nutritional diversity of organisms.

Several Streptomyces strains, such as Streptomyces viridosporus T7A [25] and Streptomyces setonii 75Vi2 [26], have been reported to degrade lignin. However, the actual catabolic pathways of lignin derivatives and the responsible enzymes and genes have not been investigated using molecular methods with Streptomyces. In our present study, a novel isolate, DF3-3, identified as Streptomyces thermocarboxydus, was found to degrade lignin. Alkali lignin is a model compound with a structure similar to that of lignocellulose and is often used as a raw material for lignin degradation studies [27]. This study used alkali lignin to investigate the characteristics of alkali lignin degradation by Streptomyces thermocarboxydus strain DF3-3. GC–MS combined with genomics was used to identify the genes responsible for lignin degradation and explore the metabolic pathway for lignin degradation by Streptomyces thermocarboxydus strain DF3-3.

Results

Morphological and physiological characteristics of DF3-3

A strain of actinomycete named DF3-3 was isolated from the greening litter of Bei**g University of Agriculture. The scribing form of the plate shows a rough white surface, the upper and lower surfaces are inconsistent in colour, and the hyphae in the base are obvious on Gause's medium (Fig. 1a). When DF3-3 was inoculated on Gause’s guaiacol medium, a clear colour reaction appeared. When inoculated on Gause’s–Azure B medium, a transparent fading circle appeared (Fig. 1b, c). This shows that DF3-3 has the ability to degrade lignin [28]. Microscopy was used to observe the morphology of DF3-3 (Fig. 1d). DF3-3 grew luxuriantly on the plate, hyphae were developed, and aerial hyphae were slender. The spore filaments were spiral-shaped with obvious characteristics of Streptomyces. Figure 1d shows a scanning electron micrograph of DF3-3 grown on Gause's medium. Mature spore chains were moderately long, with 40 to 80 spores per chain. The single spores were oval or cylindrical with diameters of 0.5 to 0.7 μm and lengths of 1.1 to 1.3 μm and have rough surfaces.

Fig. 1
figure 1

DF3-3 colour reaction and morphological characteristics. a Colony morphology streaked on a plate of DF3-3. b Transparent fading circle on Gause’s–Azure B medium. c Colour reaction on Gause’s–guaiacol medium. d Scanning electron micrograph of DF3-3

Full size image

While kee** the conditions of the basal medium otherwise unchanged, different nitrogen sources and carbon sources were added to observe the growth of DF3-3 with different nitrogen sources. The results are shown in Table 1. DF3-3 could grow on six common nitrogen sources including ammonium chloride, potassium nitrate, ammonium sulfate, ammonium tartrate, acrylamide and peptone, and could grow on sixteen common carbon sources including glucose, mannose, melibiose, starch, maltose, α-D-methyl glucoside, trehalose, cellobiose, xylose, salicin, glycerol, sodium malate, sodium succinate, sodium malonate, sodium tyrosinate,Amylase. However, it could not grow on other eleven common carbon sources (Table 1). DF3-3 was inoculated onto the Gause’s liquid medium, and incubated, respectively, at different temperatures and pH conditions, the optimum temperature range of 30–35 °C and optimum pH 7.5–8.5 were found by examining the strain weights after 7 days (Additional file 1: Fig. S2).

Table 1 Utilization characteristics of the nitrogen source and carbon source of DF3-3
Full size table

Molecular identification

Genome de novo sequencing was performed on DF3-3 cells using second- and third-generation sequencing methods, namely, Illumina HiSeq + PacBio, and the gene location and sequence information of the samples were obtained through de novo assembly and gene prediction. According to the whole genome sequencing results (Table 2), strain DF3-3 has a chromosome with a total genome length of 7,311,713 bp. GeneMarkS predicted and annotated a total of 6929 coding sequences (CDSs), with a G + C content of 72.24%.

Table 2 Genome-wide characteristics of DF3-3
Full size table

At present, 95% of the average nucleotide identity (ANI) is often used as the standard for species classification and species clustering [29]. The whole genomes of eight strains with high homology to strain DF3-3 were selected and compared and analysed with DF3-3 using ANI. The results are shown in Fig. 2. The similarity between DF3-3 and Streptomyces thermocarboxydus reached 98.96%, and it can now be identified as Streptomyces thermocarboxydus.

Fig. 2
figure 2

ANI analysis results for DF3-3 and other bacteria

Full size image

Biodegradation of alkali lignin by DF3-3

To investigate lignin degradation by strain DF3-3, cells were incubated at 30 °C in medium with alkali lignin as the carbon source. The growth curve and degradation rate of alkali lignin are shown in Fig. 3. The growth rate of DF3-3 increased significantly in 1–7 days, then, maintained a small fluctuation in 8–14 days. The degradation rate of alkaline lignin maintained a relatively uniform increase in days 1–14, eventually, the degradation rate reached a maximum value of 31% on day 15. The degradation of lignin by microorganisms requires a relatively slow process to reach a significant level. Fungi are more efficient in the breakdown of lignin than bacteria in which delignification is slower and more limited [30], the white-rot fungus Phanerochaete chrysosporium was used for degradation of lignin, and the efficiency reached approximately 20% on day 15 [31], so DF3-3 exhibited a better performance. This result is also similar to the degradation results seen with some Streptomyces strains, such as S. viridosporus T7A (lignin loss 30.9%) and S. setonii 75Vi2 (lignin loss 34.1%) [26, 32].

Fig. 3
figure 3

Growth curve and alkaline lignin degradation curve for DF3-3. Average values of three replicates are shown with the standard error of the mean as error bars

Full size image

Analysis of lignin-degrading enzymes

Lignin molecules are not easily taken passively into the cell; therefore, Streptomyces thermocarboxydus DF3-3 might produce extracellular enzymes for synergistic degradation. Three major types of enzymes responsible for the degradation of lignin are lignin peroxidase (LiP), manganese peroxidase (MnP) and laccase (Lac) [33]. MnPs oxidize Mn(II) to Mn(III), and Mn(III) oxidizes phenolic compounds and generates phenoxy radicals that in turn undergo a variety of reactions, resulting in depolymerization. In the presence of Mn(II), MnP oxidizes nonphenolic lignin model compounds via peroxidation of unsaturated lipids. LiP is the most effective peroxidase and can oxidize phenolic and nonphenolic compounds, amines, aromatic ethers, and polycyclic aromatics [34]. Lac is a copper oxidoreductase [35] that can degrade refractory polyphenols and nonphenolics in lignin, and the expression of its coding genes in bacteria has also been reported [36]. The activities of these three enzymes from DF3-3 are shown in Fig. 4. MnP and Lac activity increased constantly during the initial 6 days, with maxima of 1821.66 U/L and 1265.58 U/L seen at day 6, followed by slight decreases from day 7. Lip activity was maintained at a low level, with a maximum of 480.33 U/L on day 4. These results indicated that MnP and Lac play crucial roles during the entire process of alkaline lignin degradation by DF3-3.

Fig. 4
figure 4

Enzyme activities of LiP, Lac, and MnP of DF3-3 during 11 days of incubation. Average values of three replicates are shown with the standard errors of the mean shown as error bars

Full size image

Aromatic intermediates identified by GC–MS analysis

Using ethyl acetate as the solvent for GC–MS determination [71]. Organic acids, phenoxy free radicals and various transition metal coordination complexes can act as diffusion media of electrons and react directly with lignin, causing bond breakage [72]. A large number of similar low molecular weight compounds were detected in the lignin degradation products of DF3-3, which may be involved in both degradation agents and products during the degradation process. Low molecular weight compounds, containing phenolic hydroxyl, methoxy, amine, ketone, aldehyde, carboxyl and other functional groups, can generate phenoxy radicals to oxidize the non-phenolic residues of lignin in the presence of redox enzymes [73]. The roles of these compounds in the degradation of lignin by DF3-3 need further investigation.

Conclusions

Based on the above data and analyses, we isolated a bacterial strain identified as Streptomyces thermocarboxydus strain DF3-3 from greening litter and concluded that it degraded alkaline lignin, and the degradation efficiency reached 31% within 15 days. In total, 19 alkaline lignin degradation intermediates were identified by GC–MS, and 107 possible lignin-degrading enzyme encoding genes in the DF3-3 genome were annotated, 7 pathways for metabolism of lignin and its intermediates were predicted, including the β-ketoadipate pathway and peripheral reactions, gentisate pathway, anthranilate pathway, catabolic pathway for resorcinol, homogentisic pathway, phenylacetate–CoA pathway, and the 2,3-dihydroxyphenylpropionic acid pathway. Intermediates in the first five metabolic pathway were detected by GC–MS. The degradation products and genomics analyses show that DF3-3 has a relatively complete lignin degradation pathway.

Methods

Sampling, isolation and screening of bacterial strain

Samples were collected from greening litter of Bei**g University of Agriculture (40.0947°N, 116.3151°E). One gram of sample was placed in a 50-mL sterile centrifuge tube containing 10 mL of sterile distilled water and shaken at 120 rpm overnight. Next, 10−1 and 10−2 serial dilutions of each sample suspension were spread as 0.1-mL aliquots on Gause’s synthetic medium with the formula (g/L): 0.5 NaCl; 1 KNO3; 0.5 K2HPO4·3H2O; 0.5 MgSO4·7H2O; 0.01 FeSO4·7H2O; 20 soluble starch [8]. The plates were incubated at 30 °C for 1 week, and distinct colonies were picked and subcultured for further analysis. Gause’s guaiacol medium and Gause’s–Azure B medium used for lignin degradation screening contained 0.1% guaiacol and 0.1% aniline blue, respectively, added to Gause’s medium. Different external nitrogen sources (20 g/L), such as acrylamide and potassium nitrate, and additional carbon sources (1 g/L), such as glucose and mannose, were used to replace soluble starch culture strains in studies of their utilization of nitrogen sources and carbon sources. The culture medium for detecting lignin degradation and enzyme activity was kraft lignin–MSM medium (3 g of kraft lignin L, 2 g of [NH4]2SO4, 1 g of K2HPO4, 1 g of KH2PO4, 0.2 g of MgSO4, 0.1 g of CaCl2, 0.05 g of FeSO4, and 0.02 g of MnSO4 in 1 L distilled water, pH 7.0).

Scanning electron microscope observations

The shapes of the bacteria were observed by scanning electron microscopy. A cover glass was inserted into the solid medium to cultivate the strain, and the insert was removed after the bacterial body climbed onto the glass slide. The precipitate was washed by adding a phosphate buffer solution (pH 7.2), added to 2.5% glutaraldehyde, fixed at room temperature for 2–4 h, and then placed in a refrigerator at 4 °C overnight. After elution with a 30–95% ethanol gradient, the material was rinsed with tert-butanol, then 20 μL of tert-butanol was added and the mixture was put into a refrigerator at − 20 ℃ until it froze and solidified. Using critical point drying (HITACHI HCP-2 Critical Point Dryer) and gold sputter coating (Eiko IB-3 ion plating machine), the sample was observed by scanning electron microscopy (SEM, JSM-6360LV, JEOL, Japan) [74].

Strain growth curve determination

To assess the growth of bacteria, an equal quantity of bacteria was inserted into Gause’s liquid medium and cultured on a shaker. The culture solution was removed and centrifuged every 24 h. The supernatant was discarded, and the filter paper was placed into an oven. The mixture was dried to a constant weight, and the filter paper and the bacteria were weighed. The weight of the filter paper was compared with the weight of the bacteria. All assays were performed with three replicates.

Biodegradation of alkali lignin

To determine the lignin loss from alkaline lignin cause by various strains, samples (1.5 mL) were centrifuged at 12,000 ×g for 10 min. One millilitre of supernatant was diluted by adding 2 mL of phosphate buffer (pH 7.6). The lignin concentration was determined by measuring the absorbance at 280 nm with a UV–Vis spectrophotometer (Shimadzu UN-1900i) [75]. The calculated standard curve for lignin was y = 0.0786x–0.0245, R2 = 0.9988.

Enzyme assay

Samples were centrifuged at 12,000 rpm for 5 min, and the supernatant was used for lignin peroxidase (Lip), laccase and manganese peroxidase (MnP) enzyme assays. Laccase activity was determined by monitoring the oxidation of ABTS at 420 nm (ε420 = 36,000 M−1 cm−1). A lignin peroxidase assay was carried out using peroxidase oxidation of Azure B. LiP activity was determined by measuring the absorbance at 651 nm (ε651 = 48.8 M−1 cm−1). Manganese peroxidase activity was determined from the change in absorbance occurring when Mn2+ is oxidized to Mn3+ and forms a complex with malonate, which produced absorbance at 270 nm (ε270 = 11,590 M−1 cm−1) [76].

Genome sequencing and functional annotation

The genome of DF3-3 was sequenced at Major Biomedical Technology Co., Ltd. (Shanghai, China). Genomic DNA was extracted using a Wizard® Genomic DNA Purification Kit (Promega). Purified genomic DNA was quantified by a TBS-380 fluorometer (Turner BioSystems Inc., Sunnyvale, CA). The genome was sequenced by adopting the second-generation + third-generation sequencing method of Illumina HiSeq + PacBio, with a shotgun library of 400 bp insertion size. Assembly software canu, SPAdes, etc. was used for three-generation sequence assembly [77], and GeneMarkS software was used to predict the coding sequence (CDS) in the genome [78]. The prediction and annotation of genes were carried out using Prodigal Son (prokaryotic dynamic programming gene discovery algorithm). GeneMarkS was used to predict the plasmid genome. tRNAscan-SE v2.0 software was used to predict the tRNA contained in the genome, and Barrnap software was used to predict the rRNA contained in the genome. Functional annotation of the predicted coding gene was carried out by comparison with 6 major databases (NR, Swiss-Prot, Pfam, EggNOG, GO and KEGG) [79,80,81,82].

Alkali lignin degradation products determined by GC–MS

DF3-3 was inoculated in 100 ml of medium with AL as the carbon source and cultured on a shaker for 7 days. Samples were collected every 24 h, and a number of control groups was set up. The sample was centrifuged (10,000 rpm, 15 min) to remove the bacteria, the supernatant was acidified with HCl to pH 2–3, and it was thoroughly extracted with a threefold volume of ethyl acetate. The extract was rotary evaporated to 10 ml at 37 °C and dried with anhydrous Na2SO4. After evaporating the solvent in a nitrogen stream, 100 µl of the organic layer was derivatized. Then, 100 μl of dioxane and 10 μl of pyridine were added to the sample and vortexed, and 50 μl of bis(trimethylsilyl)trifluoroacetamide (BSTFA) was added. The mixed solution was placed in a water bath at 80 °C for 45 min and shaken regularly. The silanized sample was tested by GC–MS [83].

The analytical column was a DB-5 capillary column (30 m length, 0.25 mm inner diameter, 0.25 mm film thickness). The carrier gas was helium. The column temperature was initially 50 °C (5 min), then it was raised to 280 °C (10 °C/min, holding time of 5 min). The transmission line and ion source temperatures were 200 and 250 °C, respectively. The solvent delay time was 4.0 min. The injection volume was 1 μl. Electron ionization mass spectra were recorded in the range 30–550 (m/z) in full scan mode.