Data Information of Samples

Microbial samples were collected from five parts including sealing strip, detergent drawer, inner drum, water filter and waste water (Fig. 1a). The samples were collected from 10 domestic washing machines before and after a thorough machine cleaning by the disassembly of inner parts (Table 1), respectively. The sample ID, using years, the last machine cleaning time, number of family members, laundry frequency and detergent type were listed in Table 2.

Fig. 1

The sampling parts of washing machine and the representative images of microbial cultivation. a The samples from sealing strip, detergent drawer, inner drum and water filter were collected by soaked cotton swabs. Waste water samples were collected from drain hose. b Representative cultivation images of bacteria, fungi and coliform

Table 1 Information on the number of samples and the detection indexesTable 2 Sample ID and the information of washing machinesMicrobial Count in the Different Parts of Washing Machines

Representative of microculture images were showed in Fig. 1b. The average numbers of total bacterial count were 4976.0, 5560.4, 8166.4, 10617.8 CFU/cm2, and 9765.0 CFU/mL in sealing strip, detergent drawer, inner drum, water filter and waste water before the washing machine cleaning, respectively (Tables 3, S1). For total fungal count, the average numbers were 738.2, 780.4, 1368.0, 1622.0 CFU/cm2,and 2560.0 CFU/mL in sealing strip, detergent drawer, inner drum, water filter and waste water before the washing machine cleaning, respectively (Tables 3, S2). The average numbers of coliform count were 353.4, 1037.0, 916.2, 1563.0, and 4276.0 MPN/mL in sealing strip, detergent drawer, inner drum, water filter, and waste water before the washing machine cleaning, respectively (Tables 3, S3). Notably, washing machine cleaning by the disassembly of inner parts remarkably reduced bacterial, fungal, and coliform counts in all samples (Table 3), so the microbial counts after the machine cleaning were used as a control to reflect the severe microbial contamination of washing machines before the cleaning.

Table 3 The data of microbial culture results

We further analyzed the difference of microbial count between the four sampling parts. The number of total bacterial count was the highest in water filter, followed by inner drum (Fig. 2a). The numbers of total fungal count in water filter and inner drum were significantly higher than those in sealing strip and detergent drawer (Fig. 2b). For coliform, the count in water filter was significantly higher than that in sealing strip (Fig. 2c). Moreover, although the number of total bacterial count was increased in the group of the last machine cleaning time > 3 years, the differences were not statistically significant in the four sampling parts (Fig. 2d). For fungi, the counts in sealing strip and detergent drawer of the last machine cleaning time > 3 years were significantly increased than the last machine cleaning time of 0–3 years (Fig. 2e). And the coliform counts were significantly increased in sealing strip, inner drum and water filter of the last machine cleaning time > 3 years than 0–3 years (Fig. 2f). For waste water samples, the microbial counts of the last machine cleaning time > 3 years had an increasing trend (Fig. 2g–i). Taken together, bacterial, fungal and coliform counts were expressed differently in the different parts of washing machine, and were increased with the increase of uncleaning years.

Fig. 2

The microbial culture results of different sampling parts in washing machine. a Total bacterial count in sealing strip, detergent drawer, inner drum and water filter. b Total fungal count in sealing strip, detergent drawer, inner drum and water filter. c Coliform count in sealing strip, detergent drawer, inner drum and water filter. d Total bacterial count of the last machine cleaning time > 3 years and 0–3 years in the different sampling parts. e Total fungal count of the last machine cleaning time > 3 years and 0–3 years in the different sampling parts. f Coliform count of the last machine cleaning time > 3 years and 0–3 years in the different sampling parts. g Total bacterial count of the last machine cleaning time > 3 years and 0–3 years in the waste water samples. h Total fungal count of the last machine cleaning time > 3 years and 0–3 years in the waste water samples. i Coliform count of the last machine cleaning time >3 years and 0-3 years in the waste water samples. *P < 0.05, **P < 0.01, ***P < 0.001, compared to the corresponding group indicated by the horizontal line

Samples from Different Parts Showed Characteristic Microbial Diversity

We further analyzed the 50 samples prior to the machine cleaning using 16S rRNA gene sequencing to detect the bacterial diversity in the different parts. The sample ID of 16S ribosomal RNA (rRNA) gene sequencing was list at Table S4. Raw sequence data were deposited in the NCBI Sequence Read Archive database and the BioProject accession number was PRJNA971391 (https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA971391). Alpha-diversity analysis was performed to describe bacterial diversity by Coverage, Chao and Shannon index. Coverage index reflects the real situation of the microorganisms detected in the samples. The values of Coverage index were higher than 99.0% in the five groups (Fig. 3a), indicating that the measured sequence could represent the real microbial diversity. Chao index reflects species richness based on the operational taxonomic unit (OTU), and the value of Chao index in waste water (744.9) was significantly higher (P < 0.001) than samples from sealing strip (499.2), detergent drawer (496.1), inner drum (513.7), water filter (514.2) (Fig. 3b). Shannon index reflects species richness and evenness, and the value of Shannon index in waste water (4.4) was significantly higher (P < 0.001) than samples from sealing strip (2.8), detergent drawer (2.8), inner drum (3.0), water filter (2.9) (Fig. 3c). These data indicated that the microbial diversity of waste water was greater than samples from sealing strip, detergent drawer, inner drum, water filter.

Fig. 3

Comparisons on the diversity among different sample groups. a Coverage index to show alpha-diversity. b Chao index to show alpha-diversity. c Shannon index to show alpha-diversity. d Analysis of similarities (ANOSIM). e Principal component analysis (PCA) to show beta-diversity. f Nonmetric multidimensional scaling (NMDS) to show beta-diversity. g Principal coordinate analysis (PCoA) to show beta-diversity. h Partialleast squares discriminant analysis (PLS-DA). Group A: sealing strip; Group B: detergent drawer; Group C: inner drum. Group D: water filter. Group E: Waste water. Sample ID: WM (washing machine) + The household number + A (prior to the machine cleaning) + Sampling parts (1 represented sealing strip, 2 represented detergent drawer, 3 represented inner drum, 4 represented water filter, 5 represented waste water). ***P < 0.001, compared to the corresponding group indicated by the horizontal line

Analysis of similarities (ANOSIM) showed that the difference between five groups was significantly greater (P < 0.01) than the difference within each group (Fig. 3d). We further explored the difference of community compositions between all the samples by beta-diversity analysis using principal component analysis (PCA), nonmetric multidimensional scaling (NMDS) and principal coordinate analysis (PCoA). PCA results showed that the samples had a similarity in community composition (Fig. 3e). However, samples from waste water and sealing strip showed a separately characteristic of community composition respectively in NMDS and PCoA results (Fig. 3f–g). Partialleast squares discriminant analysis (PLS-DA) further showed the distinct aggregation of samples of sealing strip, detergent drawer and waste water (Fig. 3h). These results suggested that different parts had distinct microbial diversity.

The Dominant Bacteria Community in Different Parts of Washing Machine

Overall, 45 phyla, 129 classes, 316 orders, 567 families, 1261 genera and 2365 species were detected in this study. The five groups showed an overlap of 29 phyla (Fig. 4a). The relative abundance of Proteobacteria showed a trend of increase in sealing strip, detergent drawer, inner drum, water filter and waste water groups (Fig. 4b). Actinobacteriota was dominant in the five groups and Firmicutes was exclusively dominant in the group of sealing strip (Fig. 4b). Surprisingly, the relative abundance of Bacteroidota was low in the five groups (Fig. 4b). Community heatmap showed the detailed information of relative abundance for each sample at the phylum and illustrated that Proteobacteria and Actinobacteriota (dark red in the grids) expressed dominantly in all samples (Fig. 4c). Hierarchical clustering tree further showed the relative abundance of Proteobacteria, Actinobacteriota, Bacteroidota, and Firmicutes in each sample and indicated that the samples in each group had similarity in community composition at the phylum level (Fig. 4d).

Fig. 4

Analysis on the microbial composition among different sample groups. a The Venn diagram at the phylum level. b Community barplot analysis of groups at the phylum level. c Community heatmap at the phylum level. d Community barplot analysis of samples at the phylum level. e Linear discriminant analysis Effect Size (LEfSe) barplot at various taxonomic levels. f Circular phylogenetic tree at the genus level. g Phylogenetic tree analysis based on the sequencing reads of different groups. Group A: sealing strip; Group B: detergent drawer; Group C: inner drum. Group D: water filter. Group E: Waste water. f-, o-, c-, p- represented the level of family, order, class and phylum, respectively. Sample ID: WM (washing machine) + The household number + A (prior to the machine cleaning) + Sampling parts (1 represented sealing strip, 2 represented detergent drawer, 3 represented inner drum, 4 represented water filter, 5 represented waste water)

The dominant bacteria communities from the phylum to genus level in the five groups were further analyzed to identify the important microbiomes as the bioindicators. LEfSe barplot showed that the genus of Caproiciproducens (belonging to the family Ruminococcaceae, class Clostridia, phylum Firmicutes) and the genus of Acidipropionibacterium (belonging to the family Propionibacteriaceae, class Actinobacteria, phylum Actinobacteriota) were the predominant bacteria in the group of sealing strip (Fig. 4e). The genus of Barnesiella (belonging to the phylum Bacteroidota), the genus of Shinella (belonging to the phylum Proteobacteria) and the genus of Sellimonas (belonging to the phylum Firmicutes) had relatively high abundance in detergent drawer (Fig. 4e). For inner drum and water filter, the genus of Luteibacter (belonging to the phylum Proteobacteria) and the genus of Corynebacterium (belonging to the phylum Actinobacteriota) were the dominant bacteria, respectively (Fig. 4e). And the genus of Azospira (belonging to the phylum Proteobacteria), the genus of Roseococcus (belonging to the phylum Proteobacteria), the genus of Elstera (belonging to the phylum Proteobacteria) and the genus of Aquicella (belonging to the phylum Proteobacteria) expressed abundantly in waste water samples (Fig. 4e). Circular phylogenetic tree showed the phylogenetic diversity of bacterium and the evolutionary difference between the dominant bacteria of all samples at the genus level (Fig. 4f). The detailed evolutionary relationship and expression difference of important bacteria between the five groups were also showed in phylogenetic tree at the genus level (Fig. 4g). Taken together, these results demonstrated that different parts from washing machine had characteristic dominant bacteria community.

The phenotypes and healthy effects of microbial community in washing machine

The functional profiling and phenotypes of microbial community were further analyzed. Clusters of Orthologous Groups of proteins (COG) function classification showed that the bacteria from washing machines were associated with energy production and conversion, amino acid transport and metabolism, carbohydrate transport and metabolism, lipid transport and metabolism, translation, ribosomal structure, and biogenesis (Fig. 5a). Consistently, Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis illustrated that the bacteria from washing machines were related to the pathways of carbohydrate metabolism, amino acid metabolism, energy metabolism, xenobiotics biodegradation and metabolism, bacterial infectious disease, and antimicrobial drug resistance (Fig. 5b).

Fig. 5

Functional profiling of microbial community among the different groups. a Analysis of metabolic pathways by Clusters of Orthologous Groups of proteins (COG). b Functional annotation by Kyoto Encyclopedia of Genes and Genomes (KEGG). c Functional profiling between the different groups. d The comparison of stress tolerant between the different groups. e The comparison of mobile elements between the different groups. f The comparison of anaerobe between the different groups. g The comparison of forming biofilms between the different groups. Group A: sealing strip; Group B: detergent drawer; Group C: inner drum. Group D: water filter. Group E: Waste water. *P < 0.05, **P < 0.01, ***P < 0.001, compared to the corresponding group indicated by the horizontal line

The results of microbial phenotype indicated that the bacteria from the different parts had distinct characteristics on anaerobic, facultatively anaerobic, stress tolerant, mobile elements, gram negative and gram positive (Fig. 5c). Specifically, bacteria from the samples of detergent drawer, inner drum, water filter and waste water showed a notable increase in gram negative phenotype and mobile elements compared to sealing strip (Fig. 5d, e). However, bacteria from sealing strip had an obvious characteristic of anaerobic (Fig. 5f). Furthermore, bacteria from the samples of detergent drawer, inner drum, water filter and waste water showed an increase in the ability of biofilm forming compared to sealing strip (Fig. 5g). These results indicated that the bacteria from washing machines were tightly associated with xenobiotics biodegradation, energy metabolism, bacterial infectious disease and antimicrobial resistance, which might result in an adverse effect on the health of residents.