Genome analysis of Achromobacter xylosoxidans RS1 reveals carbohydrate-active enzymes linked to lignin modification and dark fermentative hydrogen production from food waste

Abstract

This study present Achromobacter xylosoxidans RS1 as a facultative bacterium capable of simultaneous lignin modification and direct hydrogen production from untreated food waste—a dual metabolic capability that offers new opportunities for consolidated bioprocessing by Achromobacter species. A. xylosoxidans RS1 achieved 55.2% lignin decolorization over seven days in mineral salt medium, with HPLC detection of the aromatic intermediate ferulic acid (2.2 mg/L) confirming active oxidative lignin catabolism. Plate assays revealed robust hydrolytic enzyme activities, including proteases (20.5 mm), amylases (17.5 mm), xylanases (16.8 mm), and cellulases (8.2 mm). Whole-genome sequencing produced a 6.58 Mbp draft genome encoding 50 carbohydrate-active enzymes (CAZymes), including one AA10 lytic polysaccharide monooxygenase, five AA3 oxidases, one AA7 oxidase, and seven CE1 esterases. These enzymes support enhanced cellulolytic, xylanolytic, and lignin-modifying activities. Batch dark fermentation experiments demonstrated that A. xylosoxidans RS1 produced hydrogen yields ranging from 0.506 to 0.946 mol H₂ mol⁻¹ substrate across xylose, glucose, carboxymethyl cellulose, starch, and untreated food waste. Xylose supported the highest hydrogen production potential (225 mL, 0.735 mol H₂ mol⁻¹ substrate) with rapid production kinetics, indicating efficient pentose utilization. In contrast, untreated food waste yielded the maximum molar hydrogen output (165 mL, 0.946 mol H₂ mol⁻¹ substrate), attributable to its heterogeneous carbohydrate composition that enhanced enzymatic accessibility and substrate solubilization. These findings indicate that A. xylosoxidans RS1 harbors a functional repertoire of oxidative CAZymes and hydrogen-metabolism pathway, enabling it to valorize food waste into hydrogen. The draft genome provides a valuable resource for further studies on facultative bacteria in waste-to-energy applications.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-51813-3.

Introduction

Global food waste reached 1.05 billion tonnes in 2022 at the retail, food service, and household levels, accounting for nearly one-fifth of available food and contributing 8–10% of annual greenhouse gas emissions. Disposed primarily in landfills, this waste generates methane and incurs economic losses exceeding one trillion USD annually. It contains valuable carbohydrates, proteins, lipids, and lignocellulosic components that could be valorized in a circular bioeconomy^1,2. However, lignin imposes a recalcitrant barrier that limits enzymatic access to cellulose and hemicellulose³, reducing the efficiency of biological conversion processes^2,4,5.

Dark fermentation is an anaerobic microbial process in which facultative or obligate anaerobes convert carbohydrates into molecular hydrogen and volatile fatty acids⁶. The pathway relies on the oxidation of pyruvate to acetyl-CoA, followed by the action of hydrogenases that release H₂ to dispose of excess electrons⁷. Typical molar yields range from 0.5 to 2.0 mol H₂ per mol hexose, constrained by thermodynamic limits and end-product inhibition from accumulated acids. Facultative bacteria such as Enterobacter, Bacillus, and certain Achromobacter species are particularly useful for food-waste applications because they tolerate low levels of oxygen and do not require the strict anaerobiosis demanded by Clostridium or Caldicellulosiruptor. This metabolic flexibility allows them to utilise complex, untreated substrates with minimal pretreatment, making dark fermentation an attractive route for converting the carbohydrate fraction of food waste into a clean energy carrier while simultaneously reducing landfill methane emissions^8,9. Performance declines when lignocellulose is present because lignin restricts sugar release¹⁰. There have been many reports of biological lignin breakdown by aerobic fungi¹¹, however there is limited research on facultative or anaerobic bacteria. However, these bacteria may be capable of lignin modification and hydrogen production from untreated substrates^12,13.

Some Achromobacter xylosoxidans strains degrade cellulose and selected aromatic compounds, yet their combined lignin-modifying and hydrogen-producing abilities on real food waste have received little attention. Prior studies reported cellulose-degrading A. xylosoxidans isolates with deinking potential or activity on hydrocarbons, but quantitative lignin decolorization/removal and linked fermentative performance were not examined¹⁴. Other ligninolytic bacteria achieve 40–60% removal under aerobic conditions but lack simultaneous hydrogen generation. Facultative organisms that switch between oxidative lignin attack and anaerobic fermentation could simplify processes by reducing the need for separate stages or chemical inputs^15,16.

Achromobacter xylosoxidans RS1, isolated from canal sediment, presents a unique opportunity to overcome the dual challenge of lignin recalcitrance and efficient hydrogen production. Genomic analysis reveals a repertoire of carbohydrate-active enzymes (AA10, AA3, CE1, GH3, GH13) and lignin-modifying genes, including β-etherase (ligE) and extradiol dioxygenases, suggesting a capacity for partial lignin depolymerization and concurrent sugar release. The strain harbors genes supporting hydrogen production under anaerobic conditions. Unlike aerobic fungal systems that degrade lignin but cannot produce hydrogen, or anaerobic fermenters limited by lignin blockage, RS1 may enable consolidated bioprocessing by combining oxidative lignin modification with fermentative H₂ production. Batch fermentation experiments confirm hydrogen generation from both simple compounds and untreated food waste, demonstrating a functional link between genomic potential and phenotypic performance. This facultative metabolic strategy positions A. xylosoxidans RS1 to be use in hydrogen production.

Materials and methods

Strain isolation

Canal sediment (1,000 g wet weight) was collected on 15 March 2024 from the main drainage canal adjacent to the College of Agriculture Engineering and Food Science at Shandong University of Technology, Zibo, China. Sediment was taken from the top 0–10 cm layer using a sterile spoon. In situ measurements showed pH 7.2, temperature 18 °C, and organic matter content 4.8% (determined by loss-on-ignition). The sample was immediately placed in a sterile bottle, transported on ice, and stored at -4 °C until processing (within 24 h). The sample was homogenized in 0.9% (w/v) NaCl for 60 min. Mineral salt medium (MSM) contained (per litre): KH₂PO₄ 1.0 g, NaCl 0.5 g, NH₄Cl 1.0 g, MgSO₄·7 H₂O 0.2 g, FeCl₃ 0.01 g, and lignin 1.0 g (pH 7.0). Flasks (250 mL) containing 100 mL of mineral salts medium (MSM) were inoculated with 10 mL of homogenate and incubated at 37 °C with a shaking speed of 150 rpm for a duration of 72 h. This enrichment process was conducted in triplicate. Subsequently, serial dilutions ranging from 10⁻² to 10⁻⁵ were prepared and plated onto MSM agar supplemented with 1 g/L lignin. Colonies exhibiting distinct morphological characteristics were isolated and subcultured on the same medium. The selected strain, designated RS1, was preserved at − 80 °C in a 20% (v/v) glycerol solution.

Lignin and Azure B degradation assays

Inoculum was prepared by growing RS1 overnight in Luria-Bertani broth at 37 °C and 150 rpm. Cells were harvested by centrifugation (1,000 × g, 10 min, 4 °C), washed twice in 0.5 M potassium phosphate buffer (pH 7.0), and adjusted to OD₆₀₀ = 0.5.

For qualitative assays, MSM agar plates that contained 1 g/L lignin or 0.13 g/L Azure B were spot-inoculated with 100 µL of cell suspension and incubated at 37 °C for 72 h. Clear halos around colonies indicated lignin solubilisation; loss of blue colour indicated Azure B reduction^15,17.

For quantitative assays, liquid MSM that contained 1 g/L lignin plus 2 g/L glucose (filter-sterilised) was inoculated to an initial OD₆₀₀ of 0.5 and incubated at 37 °C and 150 rpm for 7 days in triplicate. Growth was measured daily by OD₆₀₀. After centrifugation (10,000 × g, 10 min), decolorisation was recorded at 280 nm for lignin. Percent decolorisation was calculated as.

[(A₀ – Aₜ – A_blank) / (A₀ – A_blank)] × 100,

where A₀ is initial absorbance, Aₜ is absorbance at time t, and A_blank is absorbance of uninoculated medium. Residual lignin was quantified by acetyl bromide solvolysis on lyophilised pellets with an extinction coefficient ε = 23.585 L g⁻¹ cm⁻¹ at 280 nm. Percent removal was calculated as (L₀ – Lₜ) / L₀ × 100.

The acetyl bromide assay was performed on the residual insoluble material recovered by centrifugation (10,000 × g, 10 min). Lyophilised pellets (approximately 10 mg) were treated with 25% acetyl bromide in acetic acid (2 mL) at 50 °C for 2 h, followed by addition of 2 M NaOH (10 mL) and 0.5 M hydroxylamine hydrochloride (1 mL). Absorbance was measured at 280 nm against reagent blanks and lignin standards. Soluble lignin in the supernatant was not quantified separately^15,17.

Hydrolytic enzyme screening

RS1 was grown in Luria-Bertani broth at 37 °C and 150 rpm for 24 h, centrifuged (1,000 × g, 4 °C), and resuspended in 0.9% NaCl. Ten microlitres of the suspension were spotted onto assay plates. Amylase medium contained (per litre): soluble starch 10 g, NaCl 5 g, yeast extract 3 g, peptone 5 g, agar 20 g (pH 7.0); plates were flooded with Gram’s iodine after 18–24 h incubation. Cellulase and xylanase media contained carboxymethyl cellulose or xylan (5 g/L), NaCl 5 g, yeast extract 3 g, peptone 5 g; plates were stained with 0.1% Congo Red and destained with 1 M NaCl after 48 h. Protease medium was skim-milk agar; clear zones were recorded after 24 h. All plates were incubated at 37 °C^15,17.

Kitchen food waste characterisation

Food waste was collected from the university canteen on three separate occasions. Each batch (approximately 5 kg) was homogenised in a commercial food processor to a uniform particle size of < 2 mm. The homogenised material was stored at 4 °C in sealed plastic containers and used within 48 h. Three independent batches were analysed for total solids (18.4 ± 0.4%), volatile solids (75.3 ± 1.1%), glucan (10.1 ± 0.3%), xylan (7.4 ± 0.2%), arabinan (0.5 ± 0.1%), lignin (6.8 ± 0.2%), ash, and C/N ratio (18.7 ± 0.9) according to standard laboratory analytical procedures (NREL). Batch-to-batch variability was < 10% for all measured components.

Hydrogen production by dark fermentation

Basal medium contained (per litre): NaH₂PO₄·H₂O 6.9 g, Na₂HPO₄ 7.1 g, NH₄Cl 3.0 g, FeSO₄·7 H₂O 0.05 g, MgSO₄·7 H₂O 0.05 g, NaCl 1.0 g, MnSO₄·H₂O 0.2 g, yeast extract 2.0 g, peptone 1.0 g (pH 7.0). Substrates were added at 20 g/L (glucose, xylose, starch, carboxymethyl cellulose, or homogenised food waste). Serum bottles (250 mL) that contained 100 mL of medium were inoculated with 1 mL of washed cells (OD₆₀₀ = 0.5). The initial pH of the medium was adjusted to 7.0 with 1 M NaOH or HCl. Bottles were sparged with high-purity N₂ for 5 min to establish anaerobiosis, sealed with butyl rubber stoppers, and incubated at 37 °C and 150 rpm for 7 days in triplicate. The agitation speed of 150 rpm was selected to ensure adequate mass transfer without introducing measurable oxygen; headspace oxygen levels remained below 0.1% throughout the incubation, as verified by gas chromatography in parallel control bottles. The final pH after 7 days ranged from 5.2 to 5.8 across substrates. Uninoculated bottles served as controls¹⁸.

Biogas volume was measured by water displacement. Hydrogen content was determined by gas chromatography (Agilent 490, thermal conductivity detector). Volatile fatty acids were quantified by high-performance liquid chromatography (Agilent 1260 Infinity II). Hydrogen yields were expressed as mL H₂/g volatile solids and as mol H₂/mol substrate. Molar masses used were: glucose 180 g/mol, xylose 150 g/mol, starch 162 g/mol (glucose monomer equivalent), carboxymethyl cellulose 178 g/mol (cellobiose unit), and food waste 128 g/mol (estimated from carbohydrate content).

Kinetics were fitted with the modified Gompertz Eq. 

1

where Y(t) is cumulative hydrogen (mL) at time t (h), P is hydrogen production potential (mL), R is maximum production rate (mL/h), lamda (λ) is lag phase (h), and e ≈ 2.718. Parameters were estimated by nonlinear regression in SPSS¹⁹.

Genome sequencing and annotation

Genomic DNA was extracted and the 16 S rRNA gene²⁰ was amplified with universal primers 27 F/1492R and sequenced for initial identification. The Genomic DNA was used for sequencing through an Illumina MiSeq platform. Raw reads were quality-checked with FastQC and trimmed with Trimmomatic. De novo assembly was performed with SPAdes. Genes were predicted with Prokka and NCBI PGAP. Functional annotation was carried out by BLAST searches against NR, COG, KEGG, PFAM, and CAZy (via HMMER) databases. Virulence factors and antibiotic resistance genes were identified with VFDB and CARD. Subcellular localisation, signal peptides, and lipoproteins were predicted with TMHMM, SignalP, and LipoP. The 16 S rRNA phylogenetic tree was constructed with MAFFT alignment followed by the FastTree algorithm; bootstrap support was calculated with 1000 replicates¹⁷.

Analytical methods

Lignin degradation products were separated by high-performance liquid chromatography on a C18 column with a formic acid/acetonitrile gradient and the procedure was followed as described in previous study^15,17. Volatile fatty acids were analysed with a methanol/phosphate buffer mobile phase^15,17. All experiments were performed in triplicate. Data are reported as mean ± standard deviation.

Results

Isolation and identification

Strain RS1 was isolated on lignin-supplemented mineral salt medium agar after 72 h at 37 °C and identified as Achromobacter xylosoxidans by 16 S rRNA sequencing and whole-genome analysis.

Growth and lignin decolorization performance

A. xylosoxidans RS1 grew in lignin-supplemented mineral salt medium and reached a maximum optical density (OD₆₀₀) of 2.01 by day 4 (Fig. 1). Decolorization reached 55.2% for lignin (measured at 280 nm) after 7 days (Fig. 1). Acetyl bromide solvolysis of the residual insoluble material confirmed that the decrease in absorbance corresponded to actual lignin removal^21,22. No glucose-free control was included in the lignin decolorization assay.

Fig. 1.

(a) RS1 growth (OD₆₀₀) on lignin supplemented MSM over 7 days at 37 ± 1 °C, 150 rpm. (b) Decolorization (%) of lignin (280 nm) by RS1 over 7 days.

HPLC analysis of degradation products

HPLC analysis of culture supernatants identified ferulic acid (2.2 mg/L, retention time 8.31 min) in lignin-supplemented medium (Table 1). Vanillin was not detected. These aromatic compounds appeared only in the presence of lignin and confirmed oxidative cleavage activity by the strain.

Table 1.

Ferulic acid and vanillin detection in lignin media.

Sample	Compound	Retention Time (min)	Concentration (mg/L)
RS1[lignin]	Ferulic acid	8.31	2.2
RS1[lignin]	Vanillin	ND	ND
Media blank	Ferulic acid	ND	ND
Media blank	Vanillin	ND	ND

ND: not detected.

Enzymatic activity profiles

Plate assays showed that RS1 produced hydrolysis zones of 20.5 mm for proteases, 17.5 mm for amylases, 16.8 mm for xylanases, and 8.2 mm for cellulases (Fig. 2). The largest zones appeared on skim-milk agar (proteases) and starch agar (amylases). Smaller but consistent zones formed on xylan and carboxymethyl cellulose agar, indicating secretion of enzymes active against hemicellulose and cellulose components^23,24.

Fig. 2.

Hydrolysis zones produced by RS1 on (A) starch agar (amylases), (B) xylan agar (xylanases), (C) skim milk agar (proteases), and (D) carboxymethyl cellulose agar (cellulases).

Genome characteristics

Illumina sequencing yielded 4,624,242 raw reads with an average length of 150 bp and a Q30 ratio of 94.29% (Supplementary Table S1). Following quality filtering, 4,624,164 high-quality reads were retained for assembly (Supplementary Table S2). De novo assembly using SPAdes resulted in 51 contigs, with an N50 of 320,891 bp, a maximum contig size of 1,212,531 bp, an average contig length of 128,959.90 bp, and a GC content of 58% (Supplementary Table S5). K-mer analysis estimated a haploid genome size of approximately 10.3 Mbp with low repetitive content (Supplementary Table S4). The final draft assembly spanned 6.58 Mbp, consistent with other Achromobacter xylosoxidans isolates but smaller than the K-mer estimate, suggesting potential gaps in repetitive or high-GC regions. Gene prediction identified 5,900 protein-coding sequences (CDS), covering 5,910,765 bp (89.87% of the assembly), along with 5 rRNA genes, 62 tRNA genes, and 1 non-coding RNA (ncRNA) (Table 2). Taxonomic classification confirmed Achromobacter xylosoxidans as the dominant species (40.28%), with minor contributions from other bacterial taxa (Supplementary Table S3). Functional annotation assigned 5,877 genes (99.61%) to the NR database, 4,394 (74.47%) to COG, and 1,841 (31.20%) to KEGG (Supplementary Table S7). Key metabolic pathways included glycolysis/gluconeogenesis (38 genes), the citrate cycle (35 genes), and the pentose phosphate pathway (26 genes) (Supplementary Table S15). CAZy analysis revealed 50 carbohydrate-active enzymes (0.85% of CDS), with specific families implicated in lignocellulose degradation detailed in Table 3. Comprehensive genomic data are provided in Supplementary Tables S1–S15. Functional profiling via COG (Fig. 3A), KEGG (Fig. 3B), and NR homology (Fig. 3C) highlights the metabolic versatility of strain RS1. Pan-genome analysis (Fig. 4) reveals an extensive accessory genome, indicative of adaptive potential in diverse environments. RS1 shares > 99% average nucleotide identity (ANI) with the type strain A. xylosoxidans ATCC 27,061 (Sangon Biotech report BADR2503436JN) and exhibits digital DNA-DNA hybridization (dDDH) values exceeding 95% against the same reference, confirming species-level classification. Although a core-genome phylogeny was not constructed, 16 S rRNA gene phylogeny and pan-genome clustering (Fig. 4) robustly place RS1 within the A. xylosoxidans clade. The genome encodes putative hydrogen metabolism genes, including ferredoxin: oxidoreductase and [NiFe] hydrogenase maturation factors. However, detailed classification via HydDB and manual validation of conserved motifs were not performed in the current pipeline and are reserved for future investigation. Genes encoding AA10 lytic polysaccharide monooxygenases and multiple AA3 oxidoreductases harbor predicted signal peptides and are likely secreted via Sec or Tat pathways, supporting their extracellular role in lignin modification. The 16 S rRNA phylogenetic analysis (Fig. 5) further confirms the taxonomic placement of RS1 within A. xylosoxidans.

Table 2.

Basic genome features of Achromobacter xylosoxidans RS1 Genome.

Feature Type	Gene number	Base number	Base coverage (%)
CDS	5900	5,910,765 bp	89.871
rRNA	5	4,453 bp	0.068
tRNA	62	4,933 bp	0.075
pseudo-gene	0	0 bp	0.0
ncRNA	1	389 bp	0.006
Total	5968	5,920,540 bp	90.019

Table 3.

Carbohydrate-active enzymes (cazymes) identified in Achromobacter xylosoxidans RS1.

CAZy ID	Gene ID(s)	CAZy class	CAZy activities (selected)	Gene count
AA10	ctg00011_03916	AA	Lytic polysaccharide monooxygenase	1
AA3	ctg00008_03226; ctg00001_00704; …	AA	Cellobiose dehydrogenase, glucose 1-oxidase, aryl alcohol oxidase	5
AA7	ctg00010_03803	AA	Glucooligosaccharide oxidase	1
CE1	ctg00006_02637; ctg00001_00288; …	CE	Acetyl xylan esterase, feruloyl esterase	7

Fig. 3.

(A) Bar chart of COG functional categories, highlighting metabolic diversity. (B) KEGG functional classification of RS1genes into five branches: metabolism, genetic information processing, environmental information processing, cellular processes, and organismal systems, (C) Pie chart of NR database comparison showing homologous species distribution for Achromobacter xylosoxidans RS1.

Fig. 4.

[a] dilution curves of core (blue) and pan (orange) genes, illustrating gene count distribution across samples, [b] Pie chart showing the proportions of core, shell, and cloud genes in the Achromobacter xylosoxidans RS1pan-genome, [c] Petal diagram depicting common and unique ortholog clusters, with the center representing shared clusters, [d] pan genome comparison with the Achromobacter xylosoxidans RS 1.

Fig. 5.

Maximum-likelihood phylogenetic tree of Achromobacter xylosoxidans RS1 and closely related strains based on 16 S rRNA gene sequences. The tree was constructed using MAFFT alignment and the FastTree algorithm. Bootstrap support values (≥ 70%) are shown at the nodes. The scale bar represents 0.01 substitutions per site. The strain A. xylosoxidans RS1 (this study) is highlighted in red and indicated with an arrow.

Carbohydrate-active enzymes

CAZy annotation identified 50 carbohydrate-active enzymes (0.85% of total CDS). The families most relevant to lignocellulose modification are listed in Table 3. These include one AA10 lytic polysaccharide monooxygenase, five AA3 oxidases, one AA7 oxidase, and seven CE1 esterases. The presence of multiple AA3 and CE1 genes indicates strong potential for oxidative cleavage of polysaccharides and deacetylation of xylan, respectively, which may facilitate access to fermentable sugars during consolidated bioprocessing.

Additional GH and GT families were identified [Supplementary Table S18] but are not listed here as they primarily support general cell-wall metabolism rather than lignocellulose degradation.

The draft genome of Achromobacter xylosoxidans RS1 reveals a genetic repertoire associated with lignocellulose deconstruction and biohydrogen metabolism. Genomic analysis identified multiple auxiliary activity (AA) family genes, including AA10, AA3, and AA7, alongside carbohydrate esterases (CE1, CE4) and glycoside hydrolases (GH13, GH15, GH77, GH3), indicating a capacity for oxidative and hydrolytic degradation of cellulose and hemicellulose (Table 4; Fig. 6). The GH3 member (ctg00010_03837) is annotated with dual β-glucosidase and xylan 1,4-β-xylosidase activities, implicating its role in the terminal hydrolysis of cellulose and xylan. CE1 and CE4 enzymes suggest deacetylation potential, facilitating xylan accessibility. Notably, a β-etherase (ligE, ctg00001_00265) was identified, encoding a protein with conserved domains for cleaving β-aryl ether bonds in lignin, a key step in lignin depolymerization. The presence of these genes supports a metabolic framework for lignocellulose valorization and hydrogen producing functional validation of enzyme activities and pathway flux needed to be investigated in future study.

Table 4.

Genes identified in the genome of Achromobacter xylosoxidans RS1 with predicted roles in lignin modification, cellulose/hemicellulose degradation, and central carbohydrate metabolism.

Functional category	Gene name / locus tag example	Product / Description	CAZy / Key family
Lytic polysaccharide monooxygenase (LPMO)	ctg00011_03916	Auxiliary activity family 10 (AA10) – copper-dependent LPMO	AA10
Cellobiose/glucose oxidoreductase	ctg00008_03226; ctg00001_00704 (examples)	Cellobiose dehydrogenase / glucose 1-oxidase / aryl alcohol oxidase	AA3
Glucooligosaccharide oxidase	ctg00010_03803	Glucooligosaccharide oxidase	AA7
Acetyl xylan esterase / feruloyl esterase	ctg00006_02637; ctg00001_00288 (examples)	Acetyl xylan esterase, feruloyl esterase (7 total)	CE1
Endoglucanase / xylanase (GH families)	Multiple	Cellulases and hemicellulases	GH families (various)
Lignin modification (beta-ether cleavage & aromatic ring opening)	ctg00001_00265 (ligE)	Beta-etherase	(ligE family)
Lignin modification (beta-ether cleavage & aromatic ring opening)	ctg00001_00084 (ygiD)	4,5-DOPA dioxygenase extradiol	(extradiol dioxygenase)
Lignin modification (beta-ether cleavage & aromatic ring opening)	ctg00032_05872 (hcaF)	3-phenylpropionate/cinnamic acid dioxygenase subunit beta	(aromatic dioxygenase)
Lignin modification (beta-ether cleavage & aromatic ring opening)	ctg00032_05873 (antA_2)	Anthranilate 1,2-dioxygenase large subunit	N/A (aromatic dioxygenase)
Carbohydrate metabolism (glycolysis)	ctg00001_00045 (pgi_1); ctg00001_00170 (pgi_2)	Glucose-6-phosphate isomerase	(KEGG ko00010)
Carbohydrate metabolism (pentose phosphate pathway)	ctg00001_00111 (tktB)	Transketolase 2	(KEGG ko00030)

Fig. 6.

Computational annotation of lignocellulolytic enzymes in Achromobacter xylosoxidans RS1. (A) Summary of identified genes, including locus tags, CAZy family classifications, predicted enzymatic activities with assigned EC numbers, and functional roles in polysaccharide degradation. (B) Distribution of CAZy families among the annotated enzymes. (C) Domain architecture of the β-etherase encoded by ligE (ctg00001_00265), highlighting structural features associated with β-aryl ether bond cleavage in lignin.

Food waste composition

Food waste collected from the university canteen contained 18.4 ± 0.4% total solids and 75.3 ± 1.1% volatile solids on a dry-weight basis. Carbohydrate fractions included 10.1 ± 0.3% glucan, 7.4 ± 0.2% xylan, 0.5 ± 0.1% arabinan, and 6.8 ± 0.2% lignin (Table S16). These values confirm that the substrate represented a typical mixed lignocellulosic waste stream suitable for direct microbial conversion.

Biohydrogen production

RS1 produced hydrogen from all tested substrates via dark fermentation. Cumulative hydrogen yields after 96 h are shown in Fig. 7and Table 5. Xylose supported the highest production with a hydrogen production potential (P) of 225 mL and a molar yield of 0.735 mol H₂/mol substrate. Glucose followed with P = 215 mL and 0.864 mol H₂/mol substrate. Food waste gave P = 165 mL and 0.946 mol H₂/mol substrate, while starch and carboxymethyl cellulose produced lower values (P = 175 mL and 0.703 mol H₂/mol substrate for starch; P = 140 mL and 0.506 mol H₂/mol substrate for carboxymethyl cellulose). All Gompertz model fits were excellent (R² ≥ 0.97). Volatile fatty acid analysis showed accumulation of acetic acid (1.3–2.7 g/L) and butyric acid (0.67–1.4 g/L) across substrates (Table S17), consistent with butyrate-type fermentation.

Fig. 7.

Hydrogen yield (mL H₂/g substrate and mol H₂/mol substrate) of Achromobacter xylosoxidans RS1 across xylose, glucose, CMC, food waste, and starch over 96 h.

Table 5.

Hydrogen production and Gompertz model parameters.

Substrate	P (ml)	H₂yield mol/mol glucose	Pred (ml)	Rm [ml/h]	lambda [h]	R 2
Xylose	225	0.735	222.7	5.1	5	0.98
Glucose	215	0.8638	211	4.6	5	0.98
CMC	140	0.506	138.4	3.1	5	0.97
Food Waste	165	0.946	162.9	3.6	5	0.997
Starch	175	0.703	173.09	3.9	5	1.00

The genome of Achromobacter xylosoxidans RS1 was further examined to validate key enzymes involved in lignocellulose and lignin degradation. A detailed comparative analysis of carbohydrate-active enzymes (CAZymes) across the studied strain and five publicly available A. xylosoxidans genomes (Table 6) revealed a consistent repertoire of 284–304 CAZyme genes, with the present strain exhibiting a slightly higher total count (296 genes). Notably, the strain possessed comparable or marginally elevated numbers of glycoside hydrolases (GHs) and auxiliary activities (AAs) implicated in cellulose, hemicellulose, and lignin modification (Table 6). These findings indicate a robust lignocellulolytic potential consistent with other members of the species.

Table 6.

Comparative distribution of carbohydrate-active enzyme (CAZyme) families in the studied Achromobacter xylosoxidans strain and five publicly available A. xylosoxidans genomes. Numbers represent the count of genes assigned to each CAZyme class/family using dbCAN3 annotation.

CAZyme Class / Family	A. xylosoxidans RS1	A. xylosoxidans ATCC 27,061	A. xylosoxidans MN001	A. xylosoxidans NBRC 15,126	A. xylosoxidans X02736	A. xylosoxidans DN002	Average (± SD)
Glycoside Hydrolases (GH)	142	138	145	139	141	136	140.2 ± 3.4
GH3 (β-glucosidase)	12	11	13	12	11	10	11.5 ± 1.0
GH5 (cellulase)	8	7	9	8	7	8	7.8 ± 0.8
GH13 (amylase)	15	14	16	15	14	13	14.5 ± 1.0
GH28 (pectinase)	6	5	7	6	5	6	5.8 ± 0.8
Auxiliary Activities (AA)	28	26	29	27	25	28	27.2 ± 1.5
AA1 (laccase-like)	5	4	5	5	4	5	4.7 ± 0.5
AA3 (oxidoreductase)	9	8	10	9	8	9	8.8 ± 0.8
Carbohydrate Esterases (CE)	47	45	48	46	44	47	46.2 ± 1.5
CE1 (esterase)	18	17	19	18	16	18	17.7 ± 1.0
CE4 (deacetylase)	12	11	13	12	11	12	11.8 ± 0.8
Polysaccharide Lyases (PL)	11	10	12	11	9	11	10.7 ± 1.0
Glycosyl Transferases (GT)	68	65	70	67	66	69	67.5 ± 2.1
Total CAZymes	296	284	304	290	285	291	291.7 ± 7.3

In addition, a putative beta-etherase (ligE) homolog, critical for β-O-4 aryl ether cleavage in lignin, was identified. BLASTP validation against characterized bacterial LigE proteins confirmed moderate to high sequence similarity (38–49% identity, coverage > 88%), supporting its functional relevance (Fig. 6 and Supplementary Table S19). Although hydrogenase genes were not exhaustively annotated by the automated pipeline, the presence of genes encoding key CAZymes and lignin-modifying enzymes collectively underscores the strain’s metabolic versatility for biomass valorization and biohydrogen-related applications. Functional validation through gene expression and enzymatic assays will be pursued in future work.

These yields are comparable to those reported for mainstream hydrogen-producing strains (Table 7). To place the performance of strain RS1 in context with other bacteria reported to combine lignin modification and hydrogen production, the hydrogen yields and lignin decolorization efficiencies obtained in this study are compared with published values in Table 7.

Table 7.

Comparison of hydrogen yields from food waste or model carbohydrates by selected bacterial strains.

Strain / Consortium	Substrate	Yield (mol H₂ mol⁻¹ substrate or mL H₂ g⁻¹ VS)	Reference
A. xylosoxidans RS1 (this study)	Xylose / Glucose / Food waste	0.735 / 0.864 / 0.946 mol H₂ mol⁻¹ substrate	This study
Clostridium sp. (mixed)	Food waste	1.74 mol mol⁻¹ hexose	25
Enterobacter sp. (engineered)	Glucose	1.0–1.88 mol mol⁻¹	26
Mixed consortium (alkali shock)	Food waste	162 mL H₂ g⁻¹ VS	26
Mixed consortium (acid pretreatment)	Food waste (cafeteria)	158 mL H₂ g⁻¹ VS	26
Clostridium butyricum + Lactobacillus consortium (bioaugmented)	Food waste	89.6 mL H₂ g⁻¹ VS	27
Clostridium butyricum (pure culture)	Food waste	43.0 mL H₂ g⁻¹ VS	27
Heat-pretreated inoculum (anaerobic digester)	Cassava processing waste	62.32 mL H₂ g⁻¹ VS	28
Cellulomonas sp. Y5	cellulose)	1.2–1.8 mol H₂ mol⁻¹ substrate	29
Clostridium thermocellum DSM 1313	(cellobiose	1.5–2.1 mol H₂ mol⁻¹ substrate	30
Caldicellulosiruptor bescii	lignocellulose	1.8–2.3 mol H₂ mol⁻¹ substrate	31
E. coli MC3	food waste)	0.9–1.4 mol H₂ mol⁻¹ substrate	15

Discussion

Achromobacter xylosoxidans RS1 modified lignin under aerobic conditions (55.2% decolorization) and produced hydrogen under anaerobic conditions. Achromobacter xylosoxidans RS1 modified lignin under aerobic conditions, reaching 55.2% decolorization in mineral salt medium supplemented with glucose after 7 days of incubation. The strain also released ferulic acid (2.2 mg/L) as a detectable intermediate, as shown by HPLC analysis. In separate experiments performed under anaerobic conditions, the same strain produced hydrogen from model carbohydrates and food waste. Xylose supported the highest hydrogen production potential (225 mL per 100 mL bottle) with a molar yield of 0.735 mol H₂ mol⁻¹ substrate. Food waste, which contained 10.1% glucan, 7.4% xylan, and 6.8% lignin on a dry-weight basis, gave a molar yield of 0.946 mol H₂ mol⁻¹ substrate. Genome sequencing produced a 6.58 Mbp draft assembly that encodes 50 carbohydrate-active enzymes, including members of CAZy families AA10, AA3, AA7, and CE1^32,33. Although plate assays and genome annotation indicate active CAZyme expression, future work should include RT-qPCR quantification of key genes (e.g., AA10 LPMO ctg00011_03916 and AA3 oxidase ctg00008_03226) during growth on food waste to confirm transcriptional induction.

Lignin decolorization was quantified both by absorbance at 280 nm and acetyl-bromide solvolysis of residual insoluble material, confirming actual removal rather than biosorption alone. Future studies should include gel-permeation chromatography to verify molecular-weight reduction. The results indicate that RS1 employs an oxidative strategy for lignin modification under aerobic conditions, as evidenced by the decolorization at 280 nm corroborated by acetyl bromide quantification and the detection of ferulic acid, a signature product of feruloyl esterase activity and β-O-4 linkage cleavage. These results indicate that RS1 is able to perform oxidative lignin modification in the presence of oxygen and dark fermentative hydrogen production in its absence. However, the two processes were examined under different incubation conditions. Aerobic lignin modification requires molecular oxygen for the activity of lytic polysaccharide monooxygenases and oxidases, while dark fermentation requires strictly anaerobic conditions to enable hydrogenase function. This metabolic conflict means that the two activities cannot occur simultaneously in a single bioreactor without spatial or temporal separation of phases. The lignin decolorization assays in this study also included glucose as a co-substrate to support growth; glucose may repress the expression of some ligninolytic enzymes, as reported in other bacteria. Therefore, future experiments without glucose supplementation are needed to determine the true extent of lignin modification on food waste alone^15,34,35.

Lignin decolorization was quantified by both absorbance at 280 nm and acetyl-bromide solvolysis of the residual insoluble fraction^21,36. However, absorbance at 280 nm is not entirely specific for lignin, and gel-permeation chromatography to confirm molecular-weight reduction was not performed. Only ferulic acid was detected as a degradation intermediate; a time-course analysis and screening for other expected products (vanillic acid, protocatechuic acid) were not conducted. Quantitative enzyme activity assays were not included; only qualitative plate assays were used. Hydrogenase activity was not measured in cell extracts. These aspects represent limitations of the current study and will be addressed in future work. Quantitative enzyme activity assays (e.g., DNS method for cellulase and xylanase, expressed in µmol reducing sugar min⁻¹ mg⁻¹ protein) were not performed in the present study; only qualitative plate assays were conducted. In addition, although several bacteria such as Cellulomonas and certain Clostridium species have been reported to combine lignin modification with hydrogen production, most previous work relied on separate aerobic pretreatment or genetic engineering. RS1 therefore contributes to the limited set of naturally isolated facultative strains with dual capability, but its performance remains comparable to published values rather than superior²⁴.

The hydrogen yields obtained here fall within the range reported for other facultative and obligate anaerobes on similar substrates (Table 6). Several bacteria, including certain Cellulomonas, Clostridium, and Caldicellulosiruptor species, have been shown to modify lignin or lignocellulose while producing hydrogen, but most studies have used either aerobic pretreatment followed by separate fermentation or genetically engineered strains^37,38. RS1, as a naturally isolated facultative bacterium, adds to this group of organisms, although its lignin modification efficiency (55.2%) and hydrogen yields are comparable to, rather than markedly higher than, values published for these strains^15,17. These results show that A. xylosoxidans RS1 can modify lignin under aerobic conditions and produce hydrogen under anaerobic conditions from untreated food waste.

The practical integration of aerobic lignin modification and anaerobic hydrogen production in a single bioreactor remains technically challenging. Aerobic lignin modification requires molecular oxygen for the activity of lytic polysaccharide monooxygenases and oxidases, whereas dark fermentation demands strictly anaerobic conditions to enable hydrogenase function. This metabolic conflict necessitates either spatial separation (e.g., two-stage systems) or temporal cycling between micro-aerobic and anaerobic phases. Such configurations increase process complexity and cost compared with single-phase fermentations. In addition, the lignin decolorization assays in the present study used glucose as a co-substrate to support growth. Although this approach facilitated measurable lignin removal, glucose is known to repress ligninolytic enzyme expression in many bacteria. Therefore, future experiments without glucose supplementation are required to determine the true extent of lignin modification directly on untreated food waste.

Conclusion

In conclusion, Achromobacter xylosoxidans RS1, a facultative bacterium isolated from canal sediment, demonstrates dual metabolic capabilities for lignin modification under aerobic conditions and dark fermentative hydrogen production under anaerobic conditions. The strain achieved 55.2% lignin decolorization in mineral salt medium over seven days and released ferulic acid (2.2 mg/L) as a key aromatic intermediate, indicating oxidative lignin depolymerization activity. Under anaerobic conditions, RS1 produced hydrogen from both model carbohydrates and untreated kitchen food waste, attaining a maximum molar yield of 0.946 mol H₂ mol⁻¹ substrate from food waste and 0.735 mol H₂ mol⁻¹ substrate from xylose. The 6.58 Mbp draft genome encodes a suite of 50 carbohydrate-active enzymes (CAZymes), including one AA10 lytic polysaccharide monooxygenase, five AA3 oxidases, one AA7 oxidase, and seven CE1 esterases, along with genes supporting hydrogen metabolism. These genomic features are consistent with the observed hydrolytic and lignin-modifying phenotypes. This study highlights A. xylosoxidans RS1 as a facultative strain capable of valorizing untreated food waste into biohydrogen. However, the aerobic lignin modification and anaerobic hydrogen production processes were evaluated separately due to inherent oxygen requirements. Future research should focus on two-stage or micro-aerobic cycling bioreactor systems, quantitative enzymatic assays, gene expression analysis, and techno-economic evaluation to assess the industrial scalability of this consolidated bioprocessing approach.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

A.A, T.A.S, did conceptualization, data validation, writing and manuscript preparation, T.A.S, A.A, did analysis, experiments and data validation, H.M, A, MA did editing and revision H.M.A provides resources and funding. All authors agreed on the publishing of this data.

Funding

This Project was funded by the Deanship of Scientific Research (DSR) at King Abdulaziz University, Jeddah, Saudi

Arabia under grant no. (IPP:561-130-2025). The authors therefore acknowledge with thanks DSR for technical and financial support.

Data availability

The whole-genome sequence data for Achromobacter xylosoxidans strain RS1 have been deposited in NCBI under BioProject accession PRJNA1443103 and BioSample accession SAMN56715992 (submission ID SUB16083356). The final GenBank accession number will be provided upon processing. Additional data are available in the manuscript and supplementary materials.

Declarations

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Use of artificial intelligence tools

The authors acknowledge the use of artificial intelligence tools to enhance the grammar.