We isolated a strain of P. insidiosum (GZ2020) from a patient with skin ulcers. This pathogen was identified as P. insidiosum based on mycelial morphology on different media and molecular biology. We constructed a phylogenetic tree based on the genes to determine their evolutionary relationship. We observed its biological processes using scanning electron microscopy (SEM) and its inner structure using transmission electron microscopy (TEM). Additionally, we mapped the distribution of human pythiosis using the data from PubMed and Google Scholar and visualized it with R (4.4.1). We improved the sporulation induction method for P. insidiosum. We performed drug susceptibility tests with antibacterial and antifungal agents in vitro and drug susceptibility tests in vivo using a mouse model. The study design is shown in Supplementary Figure 1.

We searched the keywords “P. insidiosum” and “pythiosis” in PubMed¹ (accessed date: 8 March 2025) and Google Scholar² (accessed date: 11 March 2025). The clinical data of human pythiosis cases, including publication year, number of cases, and country of origin, were extracted and collected from the obtained literature. All collected cases met at least one of the following diagnostic criteria: (i) microbiological methods (such as culture identification, zoospore induction, and microscopic examination), (ii) histological assessment (histological examination and staining), and (iii) serological tests. (iv) Molecular assays. A total of 120 articles and 1314 cases were included. Potentially duplicated pythiosis patients were excluded from the study to prevent overestimation of the overall case number. The main criteria for considering a case as a duplicate were an identical institution, overlapping study period, and identical disease manifestation. 86 studies, including 1084 cases (Supplementary Data Sheet 1), were enrolled in the epidemiological distribution map. One case potentially infected in Brazil or Colombia was attributed to Colombia in this analysis.

The P. insidiosum (GZ2020) was isolated from a skin infection patient from the Third Affiliated Hospital of Sun Yat-sen University. We placed infected tissues on blood agar plates or 2% Sabouraud Dextrose Agar (Guangdong Huankai Microbial Technology Co., Ltd. Guangzhou, China) and incubated at 37 °C for 2 days. After hyphae growth, the culture was purified by passaging using dot-planting. According to previous studies, hyphae were picked and stained with lactophenol cotton blue (21, 22).

Genomic DNA from mycelia was extracted with a DNA Extraction Kit (Shanghai Shenggong Biological Engineering Co., Ltd. Shanghai, China), and gDNA amplification was performed using primers for different genes (15). We amplified gene LSU and COI as described by Robideau et al. (12) and Bala et al. (23) with only minor modifications: the reaction volume increased to 50 μL, the denaturation shortened to 2 min and the extension time lengthened to 80s for LSU and the extension time lengthened to 70s for COI. The polymerase chain reaction (PCR) products were separated by 1% agarose gel electrophoresis and stained for observation.

We spliced the sequences into complete sequences using DNAMAN (version 9.0, United States) and compared the amplified primer sequences with BLAST software on the NCBI website. Homology searches were performed in the GenBank database and compared with the identified P. insidiosum sequences.³ We utilized MAFTT (Version 7.0, JPN) to download sequences with high similarity and similar genera. We then used MEGA (Version 12.0, United States) with the maximum likelihood method to construct a phylogenetic tree to determine whether they were in the same branch as P. insidiosum.

This study used a modified method to induce sporulation (Figure 1A) in P. insidiosum, which requires an induction solution combined with plant sap. Liquid A included K₂HPO₄ ⋅ 3H₂O (Tianjin Yongda Chemical Reagent Co., Ltd. Tianjin, China)11.4 g, KH₂PO₄ (Tianjin Yongda Chemical Reagent Co., Tianjin, China) 6.8 g, (NH₄)₂HPO₄ (Tianjin Yongda Chemical Reagent Co Tianjin, China) 6.6 g, distilled water 50 mL, Liquid B includes MgCI₂ ⋅ 6H₂O (Tianjin Yongda Chemical Reagent Co., Ltd. Tianjin, China) 2.54 g, CaCl₂ ⋅ 2H₂O (Tianjin Yongda Chemical Reagent Co., Ltd. Tianjin, China)1.84 g, and distilled water 25 mL. Induction solution: Mix 0.5 mL of Liquid A and 0. 1 mL of Liquid B in 1,000 mL of distilled water, adjusted pH range 7.0 ∼ 7.1, stored at 4 °C (24). Extraction of plant sap: Annona squamosa L leaves (100 g) were weighed, distilled water (500 mL) was added, and sap was extracted, sterilized, and stored at 4 °C. Agar plugs with hyphae on the edge of fresh blood agar plates were inserted into 2% Sabouraud Dextrose Agar plates and incubated at 37 °C for 1 week. The agar plugs with mycelium were placed in centrifuge tubes with 2% Sabouraud Dextrose Broth (Guangdong Huankai Microbial Technology Co., Ltd. Guangzhou, China) culture solution and incubated at 37 °C and 180 rpm for 4 h. The broth was removed from the centrifuge tubes, and an equal amount of distilled water was added and incubated at 37 °C and 180 rpm for 1 h. The centrifuge tubes were incubated with an induction solution and plant sap at a volume ratio of 1:1 at 37 °C and 180 rpm for 10–24 h. We observed the production of zoospores and the release of zoospores from sporangia under a light microscope (MI52-N, Guangzhou Mingmei Optoelectronics Technology Co., Ltd., Guangzhou, China). We also compared the zoospore yield between our modified method and the traditional method (24). We used a Neubauer chamber to count the zoospores after zoosporogenesis (25, 26). We counted the zoospores on the leaves using a light microscope according to the traditional method.

Oomycete preparations were used to remove P. insidiosum from the induction solution. We soaked the samples in 1 mL of 2.5% glutaraldehyde fixative and rinsed them with 0. 1 mL phosphate buffer, washed them in 0. 1 mL phosphate buffer three times and fixed them in 1% osmium solution. After drying the samples with ethanol and dehydrating them (30,50,70,80, and 90% once for 10 min per concentration, and 100% 2 times for 10 min), the developmental cycle of P. insidiosum was observed using SEM (Verios 460, Field Electron and Ion Company (FEI), Brno, Czech Republic).

To evaluate the ultrastructural features of the hyphae, we applied an oomycete preparation to remove the hyphae from the induction solution. Samples were placed in 1 mL of a fixative solution containing 2.5% glutaraldehyde and 2% paraformaldehyde (2.5% GA + 2% PFA). The samples were then rinsed with 0.1 mL of phosphate buffer three times and fixed with a solution of osmium (1% osmium + 1% potassium ferricyanide, prepared at 0.2 M Morphotropic Phase Boundary). Subsequently, the samples were stained with uranyl acetate, dehydrated with ethanol (30,50,70,80, and 90% once for 15 min per concentration, and 100% 2 times for 20 min), washed with 100% acetone 2 times for 10 min, and infiltrated with a mixture of acetone and embedding agent (resin) (volume ration were 3:1,1:1 and1:3). Finally, the samples were polymerized for 24 h at 70 °C (27). 70–100 nm thick were obtained using an ultrathin sectioning machine. The sections were then placed on a nickel grid and stained with uranyl acetate and lead citrate. The internal morphology of the hyphae was examined using TEM (Talos F200S, FEI, Brno, Czech Republic) on a nickel grid.

The susceptibility of P. insidiosum was tested using the broth microdilution method, following the modified Clinical and Laboratory Standards Institute (CLSI) M38-A2 protocol (28). In our study, zoospores were used for the drug susceptibility assays. The zoospores were resuspended in RPMI-1640 (Thermo Fisher Scientific, Waltham, United States) broth with a pH of 6.9–7.1, supplemented with 0.164 mL of 3-[N-morpholino] propane sulfonic acid (Guangzhou Xinglang Biotechnology Co., Ltd., Guangzhou, China), to achieve a final concentration of 2–3 × 10^∧3 cells/mL (29). Microdilution 96-well plates were utilized to test the zoospore suspension (inoculum) against either RPMI-1640 alone (as a no-drug control) or various drug concentrations. After 24 h of incubation at 37 °C, the minimum inhibitory concentration (MIC) was determined according to other studies (30). We conducted three biological replicates to assess the sensitivity of P. insidiosum to several antibiotics (doxycycline (DOX), tetracycline (TET), erythromycin (ERY), and azithromycin (AZM]) (Shanghai Yuanye Biotech Co., Ltd., Shanghai, China), tigecycline (TIG) (Guangzhou Zhuorui Biotech Co., Ltd., Guangzhou, China), chloramphenicol (CHL), and linezolid (LZD) (Shanghai Maclin Biochemical Technology Co., Ltd., Shanghai, China), gentamicin (GCN) (Beijing Pubuxin Biotech Co. Ltd., Beijing, China), and florfenicol (FFC)(Guangzhou Weigu Technology Co., Ltd., Guangzhou, China), compound sulfamethoxazole (SMZ/TMP) (Shandong Fangming Pharmaceutical Group Co. Ltd., Shangdong, China) and antifungals (fluorocytosine and itraconazole [(Shanghai Yuanye Biotech Co., Ltd., Shanghai, China), amphotericin B (Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), and fluconazole (Guangzhou Zhuorui Biotech Co., Guangzhou, China)].

This study utilized 7-week-old female BALB/c mice (18–22 g) that were pathogen-free and obtained from the Guangdong Medical Lab Animal Center in Guangzhou, China. The animal experimental procedures were conducted following the guidelines of the South China Agricultural University (SCAU) Institutional Laboratory Animal Care and Use and were approved by the SCAU Institutional Ethics Committee (2020C407). To induce neutropenia, two intraperitoneal injections of cyclophosphamide were administered at doses of 150 mg/kg on 4 days before infection and 100 mg/kg on 1 day before infection. On day 0, the mice were anesthetized using an intraperitoneal injection of ketamine-xylazine (Syntec Brasil Ltd., São Paulo, Brazil) at 50 and 10 mg/kg, respectively, respectively. Uninfected, just administered cyclophosphamide and completely untreated cohorts (n = 10) served as the baseline controls [immunosuppressed only, uninfected group (control group)]. Subsequently, subcutaneous injections infected mice with 2 × 10∧4 zoospores/mouse (31). The infected mice were randomly divided into three groups, each comprising 10 animals. From day 0 (3 h after zoospore inoculation) to day 14 after infection, the mice received one of the following treatments: (i) oral Saline (PI group), (ii) oral AZM (50 mg/kg of body weight every 12 h; AZM group), (iii) oral DOX (40 mg/kg of body weight every 12 h; DOX group). The experimental process is shown in Figure 2A. Throughout the study, the animals were monitored 3–4 times daily for clinical signs such as changes in breathing, weight, mobility disorders, etc. Criteria for euthanizing the animals during the study included physical and mental alertness, chronic diarrhea, and bleeding. At the end of the experiment, all surviving animals were euthanized by deepening the anesthesia using thiopental (60 mg/kg). Histopathological analyses were performed. The number of hyphae in the tissue slices was measured using Gomori’s methenamine silver (GMS) staining, which enabled the observation of P. insidiosum in the organs. Hematoxylin and eosin (H&E) staining was used to identify inflammatory responses.

Results were presented as means ± standard deviation (SD). The statistical analysis was performed with GraphPad Prism (Version 8, United States). Unless otherwise specified, we used the unpaired t-test or one-way ANOVA to assess the statistical significance of comparisons (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns: not significant).

Seventeen countries reported human pythiosis cases. The epidemiological distribution map revealed that human pythiosis cases maintained their predominant coastal distribution pattern across Asia and the Americas, with India and Thailand continuing to demonstrate the highest endemic burden (94.83% of total reported cases). The secondary prevalence clusters were China (n = 19) and the United States (n = 17), representing 1.75 and 1.57% of the global case load, respectively (Figure 3).

Based on its microscopic features and phylogenetic analysis, the isolate was identified as a Pythium species. At 37 °C, P. insidiosum was characterized by white, sunflower-shaped, radiating colonies on brain heart infusion agar (Figure 4A) and submerged, white to colorless colonies with an irregular radiate pattern on Sabouraud agar (Figure 4B). The hyphae of P. insidiosum had lateral branches with diameters ranging from 4 to 10 μm. Older hyphae usually showed transverse septa, with protoplasmic nutrients flowing inside the hyphae (Figure 4C). Next, we conducted zoospore reproduction and SEM and TEM analyses of P. insidiosum. Zoospore release can be observed through microscopy (Figures 5, 6; Supplementary Video 1), and the process took about 35 min. The zoospores broke the vesicle wall and swam for approximately 20 min, eventually losing their flagella when they became resting spores (Figures 6D–F). When external conditions are suitable, the zoospores re-sprout to form hyphae (Figures 6G–I). In addition, the flow of protoplasm within the hypha was found (Supplementary Video 2). Traditional sporulation methods yield a small number of spores that are difficult to recover. In contrast, the improved method produced more spores that could be easily recovered and utilized. The spore yield of our improved method was 250 times that of the traditional method in shorter periods (Figures 1B–D; Supplementary Data Sheet 2).

The mitochondrial gene COI and the nuclear ribosomal gene LSU are frequently applied in phylogenetic analysis, species identification, and the exploration of evolutionary history. According to the BLAST results from the NCBI database, the COI and LSU gene amplification primer sequences shared a 100.00% sequence identity with P. insidiosum CBS 574.85 (Figure 4). The sequences of P. insidiosum were obtained through homology searches in the GenBank database. Phylogenetic analysis indicated that the P. insidiosum isolate in this study was genetically closely related to the P. insidiosum isolate from Costa Rica.

P. insidiosum has a distinct cell wall consisting of several inner layers and typical internal organelles. The cell wall has a smooth plasma membrane (cytoplasmic membrane) associated with interlaced and meshed inner layers that form coarse and thick inner layers (Figure 7). Cytoplasmic vesicles are transported from the cytosolic compartment to the plasma membrane and integrated into the inner layer. Once integrated, the cytoplasmic bilayer membrane maintains its structural integrity. Large vacuoles containing black electron-dense bodies (EDBs) were observed within the cytosolic compartment. Some fields of the microscope of our sample showed nuclei, each surrounded by a bilayer membrane and containing granular chromatin (Figures 7A,D). Golgi structures with numerous Golgi vesicles pinching off from the Golgi apparatus were also detected in the hyphae. Golgi vesicles were found to integrate into the plasmalemma surrounding the cell wall. There were relatively large microtubules dispersed throughout the hyphae. Endoplasmic reticula (ER) were sometimes detected near the mitochondria, with multiple ribosomes around the ER. Numerous small vacuoles with fine granular contents, which varied in size and shape, were dispersed throughout the hyphae. Some stained samples contained numerous mitochondria with tubular cristae.

To evaluate the effectiveness of existing antibiotics and antifungals against P. insidiosum, we conducted drug sensitivity tests in vitro. Our results showed that P. insidiosum was more sensitive to antibiotics than to antifungal agents. The MIC values of antibiotics against P. insidiosum were significantly lower than those of the antifungals (Table 1). Notably, DOX and AZM exhibited effective results against P. insidiosum. The MIC values for DOX and AZM were 4 μg/mL and 8 μg/mL, respectively. Other antibiotics, such as TIG, TET, CHL, and LZD, also showed varying effectiveness. In contrast, antifungal drugs such as fluorocytosine, amphotericin B, itraconazole, and fluconazole had much higher MIC values (higher than 128 ug/mL, indicating lower effectiveness. However, SMZ/TMP had a MIC value of 375 μg/ml. These results highlight the potential of using antibiotics, particularly DOX and AZM, to treat the infections caused by P. insidiosum.

To certify the drug sensitivity results in vitro, we conducted the drug sensitivity in vivo. The mice in the PI group appeared to have coarse and disheveled fur, were listless, and had a dull response to stimulation after inoculation with zoospores 48 h. In contrast, those in the AZM and DOX groups appeared to have a more lustrous coat, were more vigorous, and were more alert to external stimuli during the entire observation period. One week after inoculation with zoospores, surviving mice in all groups exhibited hair loss and skin ulceration at the injection sites. Ulcerated lesions were observed at the inoculation site in mice infected with P. insidiosum (Figure 2B). Forty eight hours after inoculation with zoospores, the PI group showed mortality, while no deaths were observed in the control, AZM, and DOX groups. All deaths occurred within the first 5 days in the PI group, whereas the fewer deaths in the DOX and AZM groups were delayed 3–9 days. AZM and DOX treatments were more effective in promoting wound healing than the PI group (Figure 2B). The control and PI groups exhibited survival rates of 100 and 20%, respectively. In contrast, the AZM and DOX treatment groups had 80 and 90% survival rates, respectively. AZM and DOX significantly reduced mortality compared to the saline group, the AZM group showing a 3-fold reduction and the DOX group showing a 3.5-fold reduction in mortality (P = 0.0056 for AZM, P = 0.0008 for DOX) (Figure 2C; Supplementary Data Sheet 3). Next, to further confirm the drug effects in vivo, we conducted a detection of P. insidiosum load in the kidneys, spleen, liver, and lungs of BALB/c mice (AZM, DOX, PI groups) that were subcutaneously infected with P. insidiosum (Figure 2E). Microbiological organ cultures revealed that 80% of the infected mice in the PI-treated group had established P. insidiosum infections in the liver. In the PI group, Histological H&E staining revealed prominent neutrophilic periarteritis with thrombosis and perivascular inflammatory infiltration in the liver, lung, spleen, and kidney of mice (Figure 2D). In contrast, the histological features of the mice in the AZM and DOX treatment groups were improved compared to those in the PI group, indicating the therapeutic effects of AZM and DOX on pythiosis. GMS staining revealed fewer positively stained P. insidiosum cells in the tissues off mice in the AZM and DOX treatment groups. GMS staining of the PI group demonstrated a higher presence of black-dyed hyphae in the liver, while no hyphae were observed in other organs (Figure 2E). In contrast, GMS staining of tissues from mice treated with AZM and DOX showed no black-dyed hyphae, suggesting a significant reduction in P. insidiosum burden. Thus, treatment with the antibiotics AZM and DOX resulted in marked improvement in histological lesions and a substantial decrease in P. insidiosum load in vivo.