All the patients were enrolled for this analysis from January 2022 to December 2022 and received wound management at Wound Care Center, West China Hospital, Sichuan University, which was a single-center, prospectively designed observational research. The inclusion/exclusion criteria for sample collection were as follows: (1) age from 18 to 75 years old; (2) not in pregnancy or lactation; (3) no use of steroids or immunosuppressants within 2 weeks; (4) exclusion of cancerous or secondary hemorrhagic wounds. Consecutive patients with traumatic or ulcerative wounds had been enrolled, and a total of 263 wound samples were collected, including 106 from traumatic wounds and 157 from ulcerative wounds. Traumatic wounds include acute wounds such as incised wounds from sharp instruments, surgical incisions, animal bites, burns, and scalds (Lazarus et al., 1994). Ulcerative wounds consist of ulcerative lesions, such as chronic wounds, including diabetic foot, venous ulcers, arterial ulcers, and pressure ulcers (Lazarus et al., 1994; Powers et al., 2016). Before sample collection, the wound was rinsed twice with normal saline for cleansing, followed by disinfection of the periwound skin twice with povidone-iodine antiseptic solution to eliminate potential contaminants. Then, expressed the exudate from the wound bed and used three sterile swabs to soak up the exudate and store them in sterile tubes. Following collection, specimens were transferred and stored at -80°C within 15 minutes to maintain microbiological integrity for subsequent analytical procedures. The Ethics Committee of the West China Hospital of Sichuan University approved this study (2020No.239). The basic information of the project and the requirements for sample collection were provided to the patients by professional nurses. After the patients agreed to participate in the project, all the involved patients signed the written informed consent form before sampling.

Total DNA from samples was extracted utilizing the QCTAB/SDS method. DNA concentration and purity were monitored on 1% agarose gels. After the quality inspection, DNA was diluted to 1 ng/μL using sterile water. Amplicon PCR targeted the V3–V4 region of the 16S rRNA gene using the 341F/806R primer (341F: 5'-CCTAYGGGRBGCASCAG-3', 806R: 5'-GGACTACNNGGGTATCTAAT-3') with the barcode. All PCR reactions contained 15 µL of Phusion^® High-Fidelity PCR Master Mix (New England Biolabs), 0.2 µM of each primer and 10ng target DNA, and cycling conditions consisted of a first denaturation step at 98°C for 1 min, followed by 30 cycles at 98°C (10s), 50°C (30s) and 72°C (30s) and a final 5 min extension at 72°C. Mix an equal volume of volume of 1X loading buffer (contained SYB green) with PCR products and perform electrophoresis on 2% agarose gel for detection. PCR products were mixed in equidensity proportions. Then, the mixture PCR products was purified with Qiagen Gel Extraction Kit (Qiagen, Germany). Sequencing libraries were generated by NEBNext^® Ultra™ IIDNA Library Prep Kit (Cat No. E7645) following the manufacturer’s recommendations, and index codes were added. The library quality was assessed on the Qubit@ 2.0 Fluorometer (Thermo Scientific) and Agilent Bioanalyzer 2100 system. At last, the library was sequenced on an Illumina NovaSeq platform and 250 bp paired-end reads were generated.

Paired-end reads were assigned to samples based on their unique barcode and truncated by cutting off the barcode and primer sequence. Paired-end reads were merged using FLASH (VI.2.7, http://ccb.jhu.edu/software/FLASH/) (Magoc and Salzberg, 2011), which was designed to merge paired-end reads when at least some of the reads overlap the read generated from the opposite end of the same DNA fragment. The splicing sequences were called raw tags. Quality filtering on the raw tags was performed under specific filtering conditions to obtain the high-quality clean tags (Bokulich et al., 2013) according to the FASTP. The UCHIME algorithm (UCHIME Algorithm, http://www.drive5.com/usearch/manual/uchime_algo.html) was employed to compare the raw tags with the reference Silva database (v138.1, https://www.arbsilva.de/) (Robeson et al., 2021), aiming to detect and remove chimera sequences. The QIIME2 (v2020.11.1) (Hall and Beiko, 2018) pipeline with default parameters was used for data analysis. Specifically, DADA2 was applied for filtering noise sequences, correcting edge sequence errors, removing chimeric sequences and accidental sequences, thereby obtaining a high-resolution analogue of the operational taxonomic unit (OTU) table. Subsequently, the scikit-learn method was adopted to assign and classify taxonomy against the Silva database with a 99% similarity threshold.

The variations in alpha diversity between groups were compared using the t-test. Linear discriminant analysis (LDA) Effect Size (LEfSe) was used to reveal intergroup differences in microbial communities, with the screening criteria LDA > 3, P < 0.05 (Chang et al., 2022). The microbial functional differences between groups were performed and visualized by STAMP v2.1.3 (Parks et al., 2014). The Principal Co-ordinates Analysis (PCoA) was tested by PERMANOVA (permutations=999) using the adonis2 function in vegan (v2.6.4) (Dixon, 2003). Graphical visualization used R packages including ggplot2 v3.4.1 (Wickham, 2016), ggpubr v0.6.0 (Kassambara, 2023), ggprism v1.0.4 (Dawson, 2024), permute v0.9.7 (Simpson, 2022), dplyr v1.1.0, phyloseq v1.40.0 (McMurdie and Holmes, 2013).

To elucidate the composition and structure of the microbial community in wounds and its influence on wound healing, we collected microbial samples from various types of wounds and conducted 16S rRNA gene sequencing. A total of 263 wound samples were obtained, including 106 traumatic and 157 ulcerative wound samples. Traumatic wounds include home pet bite injuries, surgical incisions, burns, and scalds, while ulcerative wounds primarily encompass venous ulcers, scar ulcers, immune-related ulcers, diabetic foot ulcers, and radiation-induced ulcers.

We first revealed the composition of the microbial community and its differences between traumatic and ulcerative wounds. Initially, we characterized the microbial OTUs present in the wound samples. The traumatic group exhibited a total of 3,072 OTUs, while the ulcerative group revealed 3,932 OTUs. Notably, 2,392 OTUs were shared between the two groups, however, 680 OTUs were exclusive to the traumatic group and 1,540 OTUs were exclusive to the ulcerative group ( Figure 1A ). The exclusive microbiota in the traumatic group included Truepera radiovictrix, Cnuella takakiae. In contrast, the exclusive microbiotas of the ulcerative group comprised Prevotella denticola, Bulleidia extructa, which have been previously identified at the abscess site, and Clostridium cadaveris.

At the phylum level, the most abundant phyla in the traumatic group were Firmicutes_A, Bacteroidota, and Actinobacteriota. In contrast, the ulcerative group displayed a distinct composition with Firmicutes_A, Actinobacteriota, and Bacteroidota as its top three predominant phyla. Notably, the relative abundance of Actinobacteriota was increased in the ulcerative group compared to that in the traumatic group ( Figure 1B ). At the genus level, Anaerococcus, Corynebacterium, and Fusobacterium_C ranked as the top three genera within the traumatic group. Conversely, Corynebacterium, Anaerococcus, and an unannotated genus dominated in the ulcerative group. Importantly, an increase in Corynebacterium relative abundance was observed in the ulcerative group relative to that of the traumatic group ( Figure 1C ).

Firstly, diversity analysis revealed that the alpha diversity (Shannon and Simpson indices) of bacterial communities on traumatic and ulcerative wounds was not significantly different ( Figures 1D, E ). Furthermore, the beta diversity (PCoA) of bacterial communities showed significant yet slight differences between traumatic wounds and ulcerative wounds ( Figure 1F ).

Then, we performed Linear discriminant analysis Effect Size (LEfSe) to identify the differential microbiota between traumatic and ulcerative wounds ( Figure 1G ). As a result, we identified seven microbial species with significantly higher relative abundance in ulcerative wounds, and five species with significantly higher in traumatic wounds. In the ulcerative wound, the differential species were found associated with infection and pro-inflammatory reactions, such as Corynebacterium auriscanis, Finegoldia magna, and Corynebacterium striatum. Conversely, the differential species in traumatic wounds included Brevundimonas diminuta, a multidrug-resistant opportunistic pathogen, and Bifidobacterium longum, which has been found that can enhance wound healing by regulating inflammatory response, promoting angiogenesis and inducing tissue repair. The significantly higher relative abundance of Bifidobacterium longum in traumatic wounds implies that traumatic wounds might experience improved recovery.

Furthermore, we identified functional differences in microbiota between traumatic and ulcerative wounds ( Figure 1H ; Supplementary Table S1 ). We found that the relative abundance of lipopolysaccharide (LPS) biosynthesis, fatty acid biosynthesis, metabolism of xenobiotics by cytochrome P450, flagellar assembly, bacterial chemotaxis, and tetracycline biosynthesis were significantly higher in the traumatic wound group. In contrast, the relative abundance of drug metabolism-other enzymes, folate biosynthesis, cysteine and methionine metabolism, DNA replication, glycine, serine and threonine metabolism were significantly higher in the ulcerative wound group.

Considering the different types of traumatic wounds, we compared the differences in microbiome composition and function between postoperative and non-postoperative wounds. Postoperative wounds refer to the surgical incision, while Non-postoperative wounds refer to any other traumatic wound except the surgical incision, e.g., trauma, burns, scalds, animal bites. Firstly, diversity analysis showed that there was no significant difference in α-diversity (Shannon and Simpson index) between postoperative wounds and non-postoperative wounds ( Figures 2A, B ). In addition, there was no significant difference in microbial β diversity (PCoA) between postoperative and non-postoperative wounds ( Figure 2C ).

Subsequently, we assessed the microbiota composition of postoperative and non-postoperative wounds. At the phylum level, Firmicutes_A, Bacteroidota, and Actinobacteriota were identified as the most abundant phyla in the postoperative wound group ( Figure 2D ). In contrast, the three most abundant phyla in the non-postoperative wound group were Bacteroidetes, Proteobacteria, and Firmicutes_A. Compared to the non-postoperative wound group, the relative abundance of Proteobacteria decreased in the postoperative wound group; conversely, the relative abundance of Firmicutes_A increased. At the genus level, Anaerococcus, Corynebacterium, and Fusobacterium_C were found to be the most abundant genera in the postoperative wound group, while Fusobacterium, Bacteroides_H, and Corynebacterium dominated in the non-postoperative group ( Figure 2E ). Notably, compared to the non-postoperative wound group, the relative abundance of Fusobacterium_C was reduced in the postoperative wound group, while the abundance of Anaerococcus increased significantly in the postoperative wound group.

The LEfSe analysis identified six and 19 microbiota species with significantly higher relative abundance in the postoperative and non-postoperative wound groups, respectively ( Figure 2F ). In the non-postoperative wound group, we noticed that many species with significantly higher relative abundance were pathogenic bacteria, such as Brevundimonas diminuta, Cutibacterium acnes, Bacteroides Hpyogenes, indicating a potential poor wound healing ability. However, Bifidobacterium longum also showed significantly higher relative abundance in the non-postoperative wound group. In the postoperative wound group, we also identified one abscesses-related species with significantly higher relative abundance, Parvimonas micra.

Furthermore, the functional difference analysis showed that various amino acid metabolism (D-arginine, D-ornithine, phenylalanine), and metabolism of xenobiotics by cytochrome P450, fatty acid biosynthesis, polyketide sugar unit biosynthesis, cyanoamino acid metabolism, etc. displayed significantly higher relative abundance in the non-postoperative wound group ( Figure 2G ; Supplementary Table S2 ). Some bacterial infection pathways (Staphylococcus aureus, Vibrio cholera, and Escherichia coli), and saccharometabolism pathways (e.g. glycolysis/gluconeogenesis, starch and sucrose metabolism, and fructose and mannose metabolism), and terpenoid backbone biosynthesis, riboflavin metabolism, and other pathways showed significant higher relative abundance in the postoperative wound group.

Meanwhile, we performed a comparative analysis between different types of ulcerative wounds, specifically venous and non-venous. The non-venous ulcerative wound contained cicatricial ulcer, radiation ulcer, immune ulcer, diabetic foot, gouty ulcer and others. Similarly, we observed no significant microbial diversity differences between venous and non-venous ulcerative wounds regarding both α diversity (Shannon and Simpson index, Figures 3A, B ) and β diversity (PCoA, Figure 3C ).

Then, we displayed the microbial composition of the venous and non-venous groups. At the phylum level, the three most abundant microbiota in the venous group were Actinobacteriota, Firmicutes_A, and Bacteroidota ( Figure 3D ). In contrast, the top three abundant microbiotas in the non-venous group were Firmicutes_A, Actinobacteriota, and Bacteroidota. The relative abundance of Firmicutes_A decreased in the venous group compared to that in the non-venous group, whereas Actinobacteriota increased. At the genus level, Corynebacterium, Bacteroides_H, and Dolosigranulum were identified as the most abundant genera in the venous group ( Figure 3E ). The top three abundant genera in the non-venous group were Corynebacterium, Anaerococcus, and Peptoniphilus_A. Notably, we noticed that the relative abundance of Corynebacterium increased in the venous group compared to that in the non-venous group.

Additionally, we conducted LEfSe to elucidate whether the two groups exhibit differential microbiota. In results, we identified seven differential microbial species in each group (venous and non-venous groups) with significantly higher relative abundance ( Figure 3F ). In the venous group, the microbiotas with significantly higher abundance included Corynebacterium striatum, Helcococcus kunzii, Faecalibacterium prausnitzii C 71358, Corynebacterium jeikeium, Bifidobacterium pseudocatenulatum, Frankia sp., Agathobacter rectalis. The non-venous group identified Porphyromonas endodontalis A 859424, Prevotella oris, Pyramidobacter piscolens, JC017 sp004296775, Prevotella nigrescens, Parvimonas micra, and Alloprevotella tannerae as differential species. We observed that several microbes in the venous group were associated with anti-inflammatory regulation, including Faecalibacterium prausnitzii and Bifidobacterium pseudocatenulatum, suggesting that venous ulcer wounds heal more readily than non-venous ulcer wounds.

Finally, we compared the differences in microbial function between venous and non-venous groups ( Figure 3G ; Supplementary Table S3 ). In the venous group, we found that pathways like streptomycin biosynthesis, lysine biosynthesis, photosynthesism, sulfur metabolism, amino sugar and nucleotide sugar metabolism, secondary bile acid biosynthesis, were significantly more abundant. In the non-venous group, pathways like ABC transporters, taurine and hypotaurine metabolism, linoleic acid metabolism, styrene degradation, riboflavin metabolism, beta-Alanine metabolism, were significantly more abundant.