Skin microbiome dysbiosis in Prototheca wickerhamii infection

A 65-year-old male patient presented with a progressively worsening rash on the right hand. As the disease progressed, similar symptoms developed on the patient's scalp. After more than a month of standardized antifungal treatment (an alternating regimen of itraconazole and voriconazole, with dynamic adjustments in administration route and dose over time), the skin lesions showed significant improvement, with complete healing of ulcerations and no recurrence of rashes. Clinical examination, histopathology, microbiology identification, and antifungal susceptibility tests were conducted for the patient. Histopathologic examination and microbial culture confirmed the presence of a high burden of P. wickerhamii at the infection site (Figure 1A).

Swab samples from the dorsum of the hands and scalp were subjected to microbiome sequencing during the course of treatment. Sequencing was performed using the NovaSeq 6000 platform, with taxonomic classification based on the Genome Taxonomy Database (GTDB) and sequencing conducted using the 2bRAD-M technology. For detailed methods, please refer to the Supplementary Materials. These analyses provided a comprehensive overview of the microbial communities at the infection site. The Chao1 index indicated a significant difference in bacterial richness between the infection (IF) and control (C) groups (Figure 1B). The relative abundance of bacterial taxa varied across different time points, primarily driven by changes in Acinetobacter parvus, which decreased over the treatment period. However, this trend was consistent between the IF and the C groups. Notably, Acinetobacter parvus was significantly more abundant in the IF group compared to the C group. Additionally, differences were observed in the relative abundance of certain bacterial genera, including Moraxella and Acinetobacter, between the two groups (Figure 1D). Despite these variations, the overall bacterial microbiome composition remained relatively similar between the IF and C groups.

For the fungal microbiome, the Chao 1 and Shannon indexes of a-diversity showed significant differences between the IF group and the C group (Figure 1E). In the IF group, the relative abundance of P. wickerhamii exceeded 60%. In the C group, Malassezia globosa and Candida albicans exhibited similar relative abundances, while P. wickerhamii was present at much lower levels. Clustering analysis of all samples demonstrated distinct fungal microbiome compositions between the IF and C groups. Comparative analysis of fungal community composition between the Journal two groups (Figure 1G) revealed a significantly higher proportion of P. wickerhamii at the lesion site on the right hand compared to other anatomical sites (Figure 1F). Furthermore, fungal diversity and relative abundance varied across different sampling sites (ulcerated, crusted, and intact skin). The primary fluctuations in relative abundance were observed in M. globosa and P. wickerhamii (Figure 1G). Within the IF group, P. wickerhamii was the dominant species at the infected right-hand lesion, whereas M. globosa was the predominant species in the C group.

Throughout treatment, the relative abundance of P. wickerhamii in the IF group showed a decreasing trend, while the relative abundance of M. globosa increased (Figures 1E, 1G). Comparative β-diversity analysis using Bray-Curtis distance and Adonis analysis revealed a significant difference between the two groups (Figure 1C). Notably, sample P414-2 exhibited the greatest variability compared to other samples.

Finally, spatial analysis of fungal distribution demonstrated that P. wickerhamii was predominantly localized on the dorsum of the right hand, both at the lesion site and within scabs (P47-2, P412-3, P414-2, P417-2). In contrast, M. globosa was the dominant species on intact skin from the scalp and left hand, exhibiting greater fungal diversity.

Based on these findings, we can draw several important conclusions and implications for clinical practice and future research. These findings confirm that histopathological examination is a reliable diagnostic tool for P. wickerhamii infections, while microbiome sequencing offers critical insights into fungal dysbiosis associated with the infection. Alternating treatment of itraconazole and voriconazole may be a promising approach for the treatment of P. wickerhamii infection. Collectively, this study provides valuable insights for optimizing diagnostic approaches and antifungal management in clinical practice.

The skin microbiome plays an important role in maintaining the protective skin barrier, and a diverse microbial diversity may contribute to the maintenance of the skin barrier. Despite the exposure of the skin to the external environment, the fungal community is essentially stable over time. However, microbial diversity cannot be effectively observed in traditional culture techniques limited by the conditions of artificial cultures, and sequencing technology can precisely compensate for this shortcoming. The fungal microbiome was markedly altered in the infection group, with P. wickerhamii dominating the lesion sites, while M. globosa and Candida albicans were predominant in the control group. The significantly higher relative abundance of P. wickerhamii (> 60%) in the IF group indicates severe fungal dysbiosis at the infection site. Given that Malassezia species are known to play an essential role in maintaining skin health and barrier function, the inverse relationship between P. wickerhamii and Malassezia colonization suggests that P. wickerhamii may disrupt the protective skin microbiome, contributing to skin barrier dysfunction. These findings support the hypothesis that alterations in fungal diversity may be linked to impaired skin barrier integrity.

Recent studies have demonstrated that itraconazole and voriconazole are more effective and well-tolerated alternatives. In our study, a 47-day alternating regimen of itraconazole and voriconazole (oral and intravenous) successfully eradicated the infection without recurrence.

However, some limitations should be noted. Due to ethical constraints, we were unable to obtain deep-tissue biopsy samples for sequencing. Instead, swab sampling was used, which may have limited the detection of deeper fungal colonization. Some samples (P47-1, P414-1) likely underwent insufficient fungal collection, leading to failed sequencing quality control and an inability to construct a risk database for further analysis. Furthermore, samples P412-1 and P417-1 exhibited the presence of Aspergillus species, which may have resulted from environmental contamination or sample size limitations. These factors may have introduced variability in sequencing results.

In summary, our findings suggest that P. wickerhamii infection is associated with fungal microbiome dysbiosis, particularly a reduction in fungal biodiversity and a decline in Malassezia abundance. We hypothesize that P. wickerhamii infections may contribute to fungal biodiversity disruption and skin barrier impairment. Future research may aim to elucidate the mechanisms underlying P. wickerhamii induced dysbiosis and investigate potential strategies to restore microbial homeostasis in infected patients.