In October 2022 and March 2023, a microbiological survey was conducted on shadow puppet artifacts stored in the repository of the National Shadow Puppetry Museum in Chengdu. Crafted entirely from cowhide, these shadow puppet samples were sourced from various rural areas, primarily including Shaanxi, Sichuan, and Hebei. The majority of the shadow puppets date back to the period spanning from the Republic of China era to modern and contemporary times. The artifacts were housed in a climate-controlled warehouse. The temperature was maintained at 20 ± 2°C, with a daily variation of less than 2°C, while the humidity was kept at 50 ± 5%, with a daily fluctuation of less than 5%. The study detected microbial colonization on the surfaces of certain cultural relics (Figure 1). To evaluate the severity of microbial degradation, a grading system was established based on the colony coverage area. Samples S1–S5 were classified as having mild contamination, S6–S10 as moderate, and S11–S16 as severe. Microbial plaques were collected from the shadow puppet surface using sterile cotton swabs and then streaked onto Potato Dextrose Agar (PDA), Luria-Bertani (LB), and Gauze’s No. 1 media plates. These plates were transported to the laboratory for subsequent microbial isolation and purification. Meanwhile, surface plaques were carefully scraped with a sterile scalpel and placed into sterile EP tubes. These samples were transported on dry ice to the laboratory and stored at −80°C for subsequent amplicon sequencing and metagenomic analysis. During the survey, the ambient temperature in the repository was 18°C with a relative humidity of 56%.

Microbial samples were collected from the surfaces of shadow puppet artifacts and adhered to a substrate using carbon conductive adhesive. The samples were then dried in a desiccator. Once dried, each sample was mounted onto an SEM specimen stage. Gold sputtering was performed at a current of 24 mA for 300 s. Subsequently, the samples were examined using a SEM, and images were captured. The measurement parameters were set as follows: electron high tension (EHT) at 15.0 kV, working distance (WD) ranging from 9.5 to 10.7 mm, and magnification (Mag) between 2,000× and 10,000×.

For the isolation and cultivation of fungi, PDA medium (1.2% potato extract, 2% glucose, 2% agar) was employed. LB medium (1% tryptone, 0.5% yeast extract, 1% NaCl, 2% agar, pH7.2) was utilized for bacterial isolation and cultivation. Gauze’s No.1 Medium (1% soluble starch, 0.2% KNO₃, 0.05% K₂HPO₄·3H₂O, 0.05% MgSO₄·7H₂O, 0.001% FeSO₄·7H₂O, 2% agar, pH7.5) served as the growth medium for actinomycetes. These microorganisms were incubated under specific conditions: fungi were cultured at 28°C for 5 days, bacteria at 37°C for 2 days, and actinomycetes at 28°C for 10 days. After the incubation period, all microbial colonies that developed on the media were systematically isolated and purified. Genomic DNA was then extracted from these pure cultures and amplified using polymerase chain reaction (PCR) to generate PCR products. For fungal amplification, the primers ITS1 (5′-TCCGTAGGTGAACCTGCGG-3′) and ITS4 (5′-CCTCCGCTTATTGATATGC-3′) were used, while the bacterial amplification employed primers 341F (5′-CCTACGGGAGGCAGCAG-3′) and 907R (5′-CCCCGTCAATTCATTTGAGTTT-3′). The PCR reaction conditions remained uniform across all samples, with the following reaction program: an initial denaturation step at 98°C for 3 min, followed by 33 cycles of 98°C for 10 s, 57°C for 10 s, and 72°C for 50 s. A final extension step was conducted at 72°C for 5 min, and the samples were held at 4°C. The resulting PCR products were sent to GENEWIZ (Beijing, China) for sequencing. Subsequently, the sequence homology of the amplified fragments was analyzed using tools available on the National Center for Biotechnology Information (NCBI) platform. The obtained sequences have been deposited in GenBank for public access and further research.

Samples S1–S10 exhibited relatively mild microbial contamination, and their microbial compositions were analyzed using amplicon sequencing. Conversely, samples S11–S16 presented more severe microbial challenges. To comprehensively examine both their microbial communities and the leather degradation processes mediated by these microorganisms, metagenomic sequencing was employed. Total genomic DNA was extracted using the DNeasy PowerSoil Pro Kit (QIAGEN, Germany; Cat. No. 47014), strictly following the manufacturer’s protocols. The extracted DNA samples were then submitted to NovoMagic Technology Co., Ltd. (Beijing, China) for sequencing. The raw sequencing data are publicly available in the NCBI Sequence Read Archive (SRA) under the study accession number PRJNA1231699 and PRJNA1230850.

When microorganisms were inoculated into a medium where gelatin served as the sole nitrogen source (comprising 1% gelatin, 1% glucose, 0.1% KH₂PO₄, 0.05% MgSO₄·7H₂O, 0.05% NaCl, 2% agar), their growth on this medium indicated their capacity to degrade collagen (Osathanunkul et al., 2013; Fuente et al., 2024).

Building on previous experimental findings and relevant literature (Wang et al., 2024; Li et al., 2025), the antimicrobial agents selected for this study were the isothiazolinone-based agent BC01 and carvacrol, a principal component of a specific plant essential oil (Table 1). The experimental procedure was as follows: First, microorganisms were uniformly spread across the appropriate culture medium. Subsequently, four sterile filter paper disks, each with a diameter of 7 mm, were placed on the medium’s surface. Then, 5 μL of each antimicrobial agent, along with the negative control solution, were separately dispensed onto the filter paper disks. Finally, the inoculated culture medium was transferred to the corresponding incubator for cultivation. The presence and size of inhibition zones were observed to assess the antimicrobial efficacy of the tested agents.

SEM analysis provided conclusive evidence for the presence of microbial hyphae and spores (Figure 2). Significantly, both fungal and actinomycete species were simultaneously detected in multiple samples, including S1, S2, S7, and S14. This finding indicated that diverse microbial community inhabited the surface of shadow puppet cultural relics. These results strongly suggested that the visible microbial colonies on the artifacts are the result of complex synergistic interactions among multiple microbial species, rather than the growth of a single organism. In addition to microbial structures, SEM imaging also revealed the presence of insects on the surfaces of the shadow puppets (Supplementary Figure S1). These insects, when crawling on the puppets, can inadvertently transport microbial hyphae and spores, thereby facilitating the spread and exacerbating the contamination of these cultural artifacts by microorganisms. This biotic interaction represents a significant threat to the long—term preservation of shadow puppet cultural relics.

High-throughput sequencing analysis of samples S1–S16 (Supplementary Figures S2–S4) revealed the presence of diverse microbial communities. Fungi predominantly belonging to the phylum Ascomycota and bacteria primarily classified under Actinobacteria were the dominant taxonomic groups on the surfaces of these cultural relics (Supplementary Figures S2–S4). Specifically, Firmicutes accounted for 14.22% of the prokaryotic community in samples S1–S5 (Supplementary Figure S2B), Proteobacteria constituted 30.68% of prokaryotes in samples S6–S10 (Supplementary Figure S3B), and Pseudomonadota represented 5.05% of the total microbial population in samples S11–S16 (Supplementary Figure S4). In-depth taxonomic profiling demonstrated significant variations in the dominant microbial communities across different shadow puppet artifacts. Samples S1–S5 shared similar microbial profiles, with Aspergillus (fungi) and Streptomyces (bacteria) as the predominant genera (Figures 3A,C). In contrast, samples S6–S10 exhibited notable heterogeneity: Aspergillus and Pseudomonas were enriched in S6; Pseudomonas was the dominant genus in S7; S8 was characterized by the co-occurrence of Aspergillus, Microascus, Pseudomonas, and Nocardia, S9 featured elevated abundances of Aspergillus, Streptomyces, and Pseudomonas; and S10 was dominated by Aspergillus, Arachnomyces, and Streptomyces (Figures 3B,D). Metagenomic sequencing of samples S11–S16 further identified prokaryote-dominated communities, with Nocardiopsis, Streptomyces, Saccharopolyspora, Brevibacterium, and Brachybacterium as the key taxonomic representatives (Figure 4). These findings highlight the complex and diverse microbial ecosystems associated with shadow puppet cultural relics, which may have implications for their conservation and preservation.

To elucidate the potential degradation mechanisms of microorganisms on shadow puppets, the sequencing results were annotated across multiple databases. Analysis using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database revealed that metabolic pathways represented a significant proportion in each sample, ranking among the primary functional categories. This finding underscores the critical role of metabolic functions within these microbial communities (Figure 5A). Functional annotation based on the eggNOG database demonstrated that, while there was a degree of similarity in the overall composition of functional categories across samples, notable variations in specific proportions were observed. Transcription, amino acid transport, metabolism, and related categories were consistently highly represented in multiple samples. For example, in sample S11, the category of unknown function was the most abundant, followed by transcription, amino acid transport and metabolism. In contrast, sample S12 showed elevated proportions of carbohydrate transport and metabolism, inorganic ion transport and metabolism, among others (Figure 5B). Examination of the Comprehensive Antibiotic Resistance Database (CARD) revealed significant differences in the relative abundance and composition of antibiotic-resistance genes. Sample S13 exhibited the highest relative abundance, primarily attributable to genes such as vanW gene in vanI cluster (Drug Class: glycopeptide antibiotic), vanX gene in vanO cluster (Drug Class: glycopeptide antibiotic), and vanY gene in vanA cluster (Drug Class: glycopeptide antibiotic). Sample S15 also had a relatively high relative abundance with a complex gene composition, including genes such as aadA5 (Drug Class: aminoglycoside antibiotic). In contrast, samples S11, S12, and S16 showed lower relative abundances, characterized by dispersed gene components and minor proportions. Sample S14 displayed a moderate relative abundance, predominantly composed of genes like the vanY gene in vanG cluster (Drug Class: glycopeptide antibiotic). These results illustrated the distinct distribution patterns of antibiotic-resistance genes across samples, providing essential data for understanding the dissemination and evolution of these genes in different environmental contexts (Figure 5C).

Eleven distinct microbial species were successfully isolated and purified using standard microbiological culture techniques, including four bacterial strains, two actinomycete strains, and five fungal isolates (Table 2). Among these, four strains—Pseudomonas sp. (WH. S-B1), Streptomyces sp. (WH. S-B2), Nocardiopsis sp. (WH. S-B6), and Aspergillus fumigatus (WH. S-F2)—corroborated the findings from high-throughput sequencing data, confirming their status as predominant microorganisms on the shadow puppet artifacts (Figure 6). These four strains were thus selected for further in-depth investigation. Growth assays on media supplemented with gelatin as the sole nitrogen source demonstrated that Pseudomonas sp., Streptomyces sp., Nocardiopsis sp., and Aspergillus fumigatus were capable of utilizing gelatin for growth (Figure 7). This phenotypic characteristic strongly suggests that these four strains possess the metabolic potential to degrade collagen, a major structural protein component of shadow puppets. These results provide crucial insights into the biochemical mechanisms underlying the deterioration of shadow puppet cultural relics.

To identify effective antimicrobial agents with minimal adverse effects on cultural relics, two candidates were selected for laboratory testing: BC01, an isothiazolinone-based antimicrobial agent currently utilized in cultural heritage conservation, and carvacrol, a main component of certain plant essential oils. The antimicrobial assay results indicated that 0.3% BC01 and 50 mg/mL carvacrol exhibited limited inhibitory activity against Pseudomonas sp. (WH. S-B1). Conversely, BC01 demonstrated notable antibacterial and antifungal effects against Streptomyces sp. (WH. S-B2), Nocardiopsis sp. (WH. S-B6), and Aspergillus fumigatus (WH. S-F2), with the most pronounced effect observed against Nocardiopsis sp. Carvacrol showed significant inhibitory efficacy against Streptomyces sp. and Nocardiopsis sp. (Figure 8). These findings suggest that BC01 and carvacrol hold potential for application in the conservation of shadow puppet cultural relics, warranting further investigation into their long-term efficacy and compatibility with artifact materials.