Gut bacteria influence Blastocystis sp. phenotypes and may trigger pathogenicity

Whilst the influence of intestinal microbiota has been shown in many diseases such as irritable bowel syndrome, colorectal cancer, and aging, investigations are still scarce on its role in altering the nature of other infective organisms. Here we studied the association and interaction of Blastocystis sp. and human intestinal microbiota. In this study, we investigated the gut microbiome of Blastocystis sp.-free and Blastocystis sp. ST3-infected individuals who are symptomatic and asymptomatic. We tested if the expression of phenotype and pathogenic characteristics of Blastocystis sp. ST3 was influenced by the alteration of its accompanying microbiota. Blastocystis sp. ST3 infection alters bacterial composition. Its presence in asymptomatic individuals showed a significant effect on microbial richness compared to symptomatic ones. Inferred metagenomic findings suggest that colonization of Blastocystis sp. ST3 could contribute to the alteration of microbial functions. For the first time, we demonstrate the influence of bacteria on Blastocystis sp. pathogenicity. When Blastocystis sp. isolated from a symptomatic individual was co-cultured with bacterial suspension of Blastocystis sp. from an asymptomatic individual, the parasite demonstrated increased growth and reduced potential pathogenic expressions. This study also reveals that Blastocystis sp. infection could influence microbial functions without much effect on the microbiota diversity itself. Our results also demonstrate evidence on the influential role of gut microbiota in altering the characteristics of the parasite, which becomes the basis for the contradictory findings on the parasite’s pathogenic role seen across different studies. Our study provides evidence that asymptomatic Blastocystis sp. in a human gut can be triggered to show pathogenic characteristics when influenced by the intestinal microbiota.


Introduction
The gut microbiota is composed of not only prokaryotes; but also, certain eukaryotes, most notably, the intestinal protozoans that pose a serious health burden in developing countries. Recent studies suggest that although the gut microbiota is diverse in species, the temporal fluctuation of certain microbial species (denotes instability) commonly occurs [1,2]. Whether this fluctuation in gut microbiota exerts an influence on other eukaryotic inhabitants of the gut is not much explored.
Blastocystis sp. is an intestinal protozoan parasite that has been frequently associated with general gastrointestinal symptoms, colorectal cancer (CRC), and irritable bowel disease (IBS) [3]. Despite its high prevalence, Blastocystis sp. has unresolved controversies regarding its pathogenicity. It was considered as a harmless commensal with recent findings implying that there is a possibility of the existence of two variant forms i.e. disease (pathogenic) and non-disease (non-pathogenic) causing types in a single subtype. Blastocystis sp. has been reported to exhibit strong interaction with its accompanying microbiota [4]. Studies have reported on increased diversity of bacteria in Blastocystis sp. colonized gut [5,6] and strong association of specific subtype with gut microbial composition [7]. However, there is a paucity of studies on the influence of gut microbiota by a single and most prevalent subtype (ST3) of Blastocystis sp. isolated from symptomatic and asymptomatic individuals as well as the influence of varying accompanying microbiota on this intestinal protozoan cells.
Some of the common intestinal protozoan parasites that are pathogenic to human gut include Entamoeba histolytica and Giardia duodenalis. According to epidemiological data, colonization of these eukaryotic pathogens may not necessarily result in the manifestation of symptoms [8]. Some studies have posited that the association of certain bacteria resulted in increased pathogenicity and protective effect in the infection of Giardia sp. [9][10][11] and increased virulence in Entamoeba histolytica [12,13]. It is unknown if a similar mechanism occurs with Blastocystis sp. infection. Whether the microbial environment could influence shaping the parasite's characteristics, thus altering the severity of infection is still a question.
In this study, we investigated the gut microbiota profiles in symptomatic and asymptomatic individuals with or without Blastocystis sp. ST3 infection. Subsequently, we altered the microbiome of Blastocystis sp. in in vitro culture to understand the response of the parasitic cell towards different microbiota.

Ethics statement
A verbal and written consent was obtained from all participants recruited. The study procedure was approved by University Malaya Medical Centre (UMMC) Medical Research Ethics Committee (MRECID: 201914-6975).

Stool sample collection and Blastocystis sp. carriage assignment
A total of 50 fecal samples were studied, with one fecal sample collected each from 50 individuals who participated. Twenty-eight individuals who did not experience any gastrointestinal illness were grouped as asymptomatic. These individuals were recruited from a voluntary stool survey. The balance of 22 was patients visiting the Gastroenterology Unit of University Malaya Medical Centre (UMMC) and Gastroenterology and Hepatology Specialist Clinic of Pantai Medical Centre, Kuala Lumpur, Malaysia. These patients, who experienced frequent non-specific gastrointestinal symptoms such as bloating, abdominal cramps, loose stool, and diarrhea at the time of recruitment were grouped as symptomatic. Both the symptomatic and asymptomatic individuals were sub-grouped into Blasotcystis sp.-infected and Blastocystis sp.-free groups. Only participants with Blastocystis sp. as the sole infective agent were recruited (for all Blastocystis sp.-infected individuals). For all the Blastocystis sp.-free individuals, the fecal specimens were screened to ensure no other parasitic infection was detected. Screening for other parasites was done using the formal-ether concentration technique as described previously [14]. Patients diagnosed with colorectal cancer (CRC), inflammatory bowel disease (IBD), or irritable bowel syndrome (IBS) and those who have consumed antibiotics within the last 30 days were excluded from participating. The clinicians obtained prior written consent before the recruitment of participants. The clinicians also confirmed the diagnosis of these individuals after a thorough examination. To maintain homogeneity in population and environment, only individuals from the most developed part of Malaysia (Kuala Lumpur) and who belonged to a high socio-economic group were included. This study was approved by the Medical Research Ethics Committee of UMMC (201914-6975). Stool samples collected in screwcapped containers were processed and stored at -20˚C within 6 hours of collection.

Fecal DNA extraction
DNA extraction from fecal materials was carried out using MACHEREY-NAGEL NucleoSpin Soil kit (MACHEREY-NAGEL GmbH & Co. KG, Dü ren, Germany). The SL2 buffer was used with Enhancer SX in the lysis step of extraction. DNA was eluted in a final volume of 50 μl and stored at -80˚C.

Patient and public involvement
Patients and the public were first involved during fecal sample collection and questionnaire administration. Recruited individuals were either identified by the healthcare professional or upon voluntary admission. All research questions and outcome measures were approved by the Medical Research Ethics Committee of UMMC and were explained to each participant in detail by the enumerators and the clinicians. There was no involvement of the patient or the public in the design of this study. The participants agreed verbally to have their results published.

Amplification of variable 3 (V3) and variable 4 (V4) region of 16S ribosomal RNA (rRNA) genes
The primer pair sequences that produced a single amplicon for V3 and V4 region of approximately 460 bp were used as described [15]. The library preparation was done according to the study. The amplification process was carried out using the 2X KAPA HiFi HotStart Ready Mix with microbial genomic DNA at a concentration of 5ng/μl in 10mM. The amplification was carried out with thermal cycling consisting of 95˚C for 3 minutes, followed by 25 cycles of 95˚C (30 seconds), 55˚C (30 seconds), and 72˚C (30 seconds) with a final extension of 72˚C for 5 minutes. Subsequently, a dual index barcode Illumina sequencing adaptor was attached to the amplicon using Nextera XT Index Kit (Illumina). Prepared amplicons were cleaned up again using AMPure beads. The V3 and V4 regions were then sequenced on an Illumina MiSeq platform (Illumina, San Diego CA, USA) at the Texas Children's Microbiome Center.

Quality filtering and analysis of filtered reads
The raw sequences were joined at the paired-end by trimming the low-quality bases. All sequences were imported and analyzed through Quantitative Insights Into Microbial Ecology 2 (QIIME 2 Version 2020.6) platform [16]. The paired-end sequences were joined, chimeric sequences filtered and low-quality reads were removed using DADA2 plug-in [17].

Analysis of filtered reads sequence diversity
Quality filtered reads were used as the sequence data. The Operational Taxonomic Units (OTU) abundance was identified and quantified with reference to the Greengenes reference database version 13_5 with a 97% homology cut-off (https://greengenes.secondgenome.com/). We generated a phylogenetic tree by first performing multiple sequence alignment of the sequence using the mafft [18]. Next, highly variable regions that add noise to the tree were removed and the phylogenetic tree was built using FastTree [19]. This pipeline was carried out in QIIME2.
The alpha diversity was measured using the Shannon index metrics, richness and observed OTUs. Alpha rarefaction curve was plotted after samples were rarefied to 4000 sequences per sample. Beta diversity was computed using Bray-Curtis distance matrix ordinated using nonmetric multidimensional scaling (NMDS) and Principal Coordinate Analysis (PCoA).

Linear Discriminant analysis effect size (LEfSe) and predicting the functional composition of a metagenome
LEfSe algorithm from the Galaxy web application (https://huttenhower.sph.harvard.edu/galaxy/) was used to identify taxa/gene/pathways with differential abundance from different experimental classes. In this study, we have used the symptomatic and asymptomatic groups as main classes and Blastocystis sp. infection status as subclasses. LEfSe lists the taxa that are differential among the classes with statistical and biological significance and ranks them according to effect size [20].
Predicted functional metagenome was developed using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt). An OTU table was built using closed-reference clustering method by comparing each OTU representative to the Greengenes database version 13_5 at 97% cutoff. The resulting OTU table was used for metagenome prediction in Galaxy web application using KEGG orthology classification schemes [21]. Subsequently LEfSe was used to compare differential predicted metabolic functions between the classes.

In vitro cultivation and genotyping of Blastocystis sp.
About 50 mg of fecal samples were inoculated into 3 ml Jones medium supplemented only with 10% horse serum as reported previously [22]. There were no antibiotics added to the medium to maintain the original bacterial composition. The cultures were incubated at 37˚C and screened daily for 5 to 7 days. The presence of Blastocystis sp. vacuolar forms was regarded as a positive sample. The xenic parasite cell culture was maintained in vitro and passaged every 3 to 4 days. Basic aseptic techniques were maintained throughout the culture process.
DNA was extracted from Blastocystis sp. in vitro cultures using Macherey Nagel Soil DNA extraction kit. Extracted DNA was used as a template to amplify and sequence the 18S small subunit ribosomal RNA gene (18S SSU-rDNA) at the length of 600bp using the protocols and primers described previously [23]. Five isolates of Blastocystis sp. were selected randomly to represent the symptomatic and asymptomatic groups for subsequent analysis.

Alteration of microbiota surrounding Blastocystis sp.
Cells from three-day-old Blastocystis sp. (isolated from asymptomatic individuals) culture grown in 3 ml of Jones' medium supplemented with only 10% horse serum were spun at 1000 rpm to sediment Blastocystis sp. cells. No antibiotics were added to the Jones' medium. The supernatant containing mostly bacteria was isolated and washed with distilled water three times to lyse any remaining Blastocystis sp. cells. This bacterial suspension was then centrifuged at high speed and the pellet was re-suspended in 100 μl of Jones' medium with 10% horse serum.
Three-day-old Blastocystis sp. cells from symptomatic individuals were washed three times in PBS and counted to a concentration of 1 x 10 5 cells/ml in a final volume of 900 μl of Jone's medium. The bacterial suspension extracted earlier from asymptomatic Blastocystis sp. cultures were added and allowed to incubate at 37˚C. These steps were repeated every 3-4 days for 5 times before assessing the growth characteristics. The cells were then harvested and subjected to downstream processing. The experiment was repeated with asymptomatic Blastocystis sp. cell co-cultured with bacteria from symptomatic cultures. Four replicates were used in this experiment. Throughout this experiment, the xenic cultures that received bacterial suspension were regarded as co-cultured. Control experiments were those xenic cultures that were added with only sterilized Jones medium replacing bacterial suspension.

Phenotypic characterizations
Growth characterization. Three-days old cells from co-cultured and control experiments were counted using heamocytometer chamber and inoculated into 1 ml medium with a final concentration of 1 x 10 5 cells/ml. The cells were counted every day for 10 days using trypan blue exclusion test to determine number of viable cells. The number of granular formation and amoebic formation per ml were counted.
Colorimetric protease quantification assay. Parasite isolates from co-cultured and control experiments were subjected to purification in order to recover parasites cells with minimal bacterial contamination. Purification was done through density-gradient centrifugation as described previously but with slight modification on the centrifugation speed [24]. The protease activity of the minimalized bacterial contamination was ensured to be at a negligible level as noted in the previous study [24]. The solubilized antigen of the purified Blastocystis sp. was extracted using the freeze-thaw technique. The purity and concentration were determined using Bradford Protein Assay (BioRad). The concentration of each sample was standardized to 0.1 mg/ml using a filter sterilized Jones medium before the assay. The specific protease activity was determined using azocasein colorimetric assay as reported previously [25,26].
Colon cell proliferation analysis. The antigen extracted previously were used to study colon cells proliferation. HCT 116 colonic cancer cells were obtained from American Type Cell Culture (ATCC) and maintained in RPMI medium supplemented with L-Glutamine, antibiotics and 10% fetal bovine serum (FBS). Colon cancer cells were maintained in T-25 vented culture flask at 37°C, 5% CO 2 and passaged every 4 to 5 days. To assess the proliferation, cells were standardized and seeded to 1000 cells/well in 100 μl using a 96-well plate as described by previous studies [27,28]. The cells were allowed to incubate at 37°C, 5% CO 2 for 24 hours. The antigens extracted earlier were added to the cells and incubated for another 48 hours. Cell proliferation was then determined using the MTT assay as described by the study cited above.

Statistical analysis
The difference between groups was evaluated by comparing the mean using statistical tests. A student's t-test was conducted to compare the difference in alpha diversity and richness between groups using values derived from the diversity index. The student's t-test was also used to compare means of protease levels and cell proliferation. A p-value less than 0.05 is considered significant.

Microbial profile analysis
A mean of 57828 sequences per sample from 50 individuals who were symptomatic (22/50) and asymptomatic(28/50) was obtained. The sequences were deposited in the National Library of Medicine (NCBI) as a BioProject with accession number PRJNA881789. A total of 1725 unique features were identified. The core phyla across the samples were Firmicutes and Bacteroidetes with more than 50% of relative abundance followed by Actinobacteria and Proteobacteria. At the genus level, the core genera observed across all samples were Prevotella, Feacalibacterium, Bifidobacterium, Bacteroides, and Dialister (S1 and S2 Figs).

Variation of microbiota in symptomatic and asymptomatic individuals and effect of Blastocystis sp. colonization
Using the Shannon diversity index, significantly higher species diversity was observed in asymptomatic individuals (P = 0.028). The alpha rarefaction curve (at 5000 reads/sample) indicated that greater number of features seen in samples of asymptomatic individuals. Coverage index indicates the number of most abundant species occupying 50% of the community ecosystem. In this study, asymptomatic individuals showed significantly greater coverage compared to symptomatic individuals (P = 0.038). The evenness of the samples was measured using Pielou's evenness index. There was lower evenness seen in symptomatic individuals (P = 0.017). Beta diversity analysis through ordination by Non-metric multidimensional scaling (NMDS) using Canberra distance matrix showed clustering of symptomatic and asymptomatic samples (Stress = 0.17). The difference in diversity between samples was significant using PERMANOVA (P = 0.026) (Fig 1).
When the samples were generally classified into Blastocystis sp. infection status, we saw a significantly lower abundance of species in Blastocystis sp.-infected individuals using Chao1 richness estimator (P = 0.032) (Fig 2A). When the samples were grouped according to symptom, a significant difference in the abundance of species was only seen in asymptomatic individuals (P = 0.00026) while in the symptomatic group, there were no significant alterations to the species richness due to Blastocystis sp. infection (Fig 2B). Plotting of

PLOS NEGLECTED TROPICAL DISEASES
Influence of Blastocystis sp. ST3 on microbiome and vice versa distance matrix using Canberra distance with the incorporation of abundance value revealed that infection of Blastocystis sp. regardless of symptoms has significant alterations in the gut microbiota (Fig 2C).

Differential association of bacterial taxa to Blastocystis sp. infection in symptomatic and asymptomatic individual
LEfSe was deployed to determine the bacterial taxa that was differentially present in symptomatic and asymptomatic conditions and in Blastocystis sp.-positive and Blastocystis sp.-negative subjects. In this study, logarithmic LDA score of 3.0 was used as cut-off for the detection of important taxonomic differences. We found that Prevotella sp. was differentially abundant in symptomatic subjects while in asymptomatic group, bacterial phyla belonging to Firmicutes, Bacteroidotes, Verrucomicrobiota and Desulfobacteriota was abundant with LDA score beyond the fixed cut-off value (Fig 3A).
Within the symptomatic group, subjects that were Blastocystis sp.-positive showed that the family Prevotellaceae and Ruminococcceae were abundant whereas Blastocystis sp.-negative subjects showed an abundance of Akkermansia sp. and Bacteroides sp. (Figs 3B and S3). However, in the asymptomatic group, taxa from the phylum Firmicutes, specifically Megasphaera sp. and Butyricicoccaceae were differentially abundant in Blastocystis sp.-positive subjects while in Blastocystis sp. negative subjects, the increased abundance was in taxa belonging to the phyla Verrucomicrobiota, Firmicutes, and Bacteroidota ( Fig 3C). The findings implicate that Blastocystis sp. in symptomatic and asymptomatic infection could be associated with different bacterial taxa.

Blastocystis sp. colonization, and alteration in gut microbial functions
Microbial functions of the microbiota in the subjects were determined by using the inferred metagenomics obtained from PICRUSt. LEfSe was used to identify a differently abundant pathway in the samples. Microbial function in symptomatic individuals with Blastocystis sp. colonization generally showed an abundance of pathways involved in translation, nucleotide metabolism, metabolism of cofactors and vitamins, digestive systems, and also pathways involved in metabolic diseases compared to the microbiota without Blastocystis sp. which had pathways involved in transcription, signal transduction and lipid metabolism (Fig 4A). However, microbiota colonized by Blastocystis sp. in asymptomatic individuals had functional pathways abundant in metabolism of cofactors, vitamins and amino acids (Fig 4B). In general, pathways involved in replication and repair, nucleotide metabolism, translation, metabolic diseases and digestive system are found in Blastocystis sp.-positive subjects. The findings demonstrate that, it was not just the microbes that were differently abundant but also the metabolic functions that seemed to be different in Blastocystis sp.-colonized microbiota isolated from symptomatic and asymptomatic individuals.

Genotyping and phenotypic characteristics of Blastocystis sp.
As reported in past studies [29], here we observed that Blastocystis sp. was found colonizing both symptomatic and asymptomatic individuals. All Blastocystis sp. isolated from this study belong to ST 3. The analysis of 18S partial length rDNA of Blastocystis sp. sequence analysis suggests that the ST 3 isolates belonged to allele 34. The 18S partial rRNA sequences of Blastocystis sp. ST3 isolated from symptomatic and asymptomatic individuals suggest close genotypic similarity and this means that any difference seen phenotypically would be solely due to external pressures. In this study, we assessed phenotypic expressions of the parasite from symptomatic and asymptomatic conditions. The phenotype was studied in terms of in vitro growth profile, specific protease activity, and ability to proliferate cancer cells. High peak cell count was observed specifically in Blastocystis sp. isolated from asymptomatic individuals than parasites isolated from the symptomatic individual (Fig 5A and 5C). Significantly greater total protease activity was seen in Blastocystis sp. isolated from symptomatic individuals. This was noticed in the control experiments in Fig 6. We also observed that the Blastocystis sp. obtained from symptomatic individuals had a predominance of cysteine protease whereas the parasite cells isolated from asymptomatic individuals possessed serine protease predominantly (Fig 6A).

Influence of bacteria on Blastocystis sp.
Blastocystis sp. from symptomatic individuals co-cultured with bacterial suspension from asymptomatic individuals showed increased growth of parasite numbers compared to the parasite isolates without introducing bacterial suspension from asymptomatic isolates. The average peak cell count of 2.46 x 10 6 cells/ml increased about 3-fold to 6.54 x 10 6 cells/ml. Whereas Blastocystis sp.

PLOS NEGLECTED TROPICAL DISEASES
Influence of Blastocystis sp. ST3 on microbiome and vice versa obtained from asymptomatic individuals, which had high cell numbers showed a reduction in cell count upon introducing bacterial suspension from symptomatic individual. The average peak cell count decreased 4-folds from 6.17 x 10 6 cells/ml to 1.45 x 10 6 cells/ml (Fig 5).
Protease activity in Blastocystis sp. isolated from symptomatic individuals co-cultured with bacterial suspension extracted from asymptomatic individual showed only a slight increase, which was insignificant. However, isolate obtained from asymptomatic individuals when cultured with bacterial suspension isolated from symptomatic individuals showed significant increase in the protease activity (from 0.085 to 0.2789). The increase was statistically significant using Student's t-test when compared to the control (P = 0.029). Blastocystis sp. isolated from asymptomatic individuals initially possessed significant predominance of serine protease. When bacterial suspension from symptomatic individuals was introduced to Blastocystis sp. obtained from asymptomatic individual there was an increase in the cysteine protease. This increase was found to be significant using the Student's t-test (P<0.05) (Fig 6A).
Regarding the ability of Blastocystis sp. antigens to promote colonic cell proliferation, antigens isolated from symptomatic individuals that were cultured with bacteria from asymptomatic Blastocystis sp. culture showed insignificant proliferation compared to the control. Nonetheless, antigens from Blastocystis sp. isolated from asymptomatic individuals that were co-cultured with bacteria from symptomatic individuals produced significantly greater colonic cell proliferation than antigens from Blastocysits sp. isolated from asymptomatic individuals in autochthonous culture. There was about 3-fold increase from 19.8% proliferation to 65.8% (Fig 6B).

Discussion
Blastocystis sp. has been reported to have an intricate relationship with its surrounding bacteria [30]. A previous study had orally inoculated axenic, monoxenic and xenic Blastocystis sp. from symptomatic individuals into germ-free guinea pigs. It was found that about half of the rats inoculated with xenic parasite developed infections with watery diarrhea for more than a week duration and increased cellularity at the lamina propria region. In contrast, the rats inoculated with monoxenic had an infection and none was infected in rats inoculated with axenic Blastocystis sp. [31]. This finding was one of the earliest to highlight the importance of accompanying gut bacteria in Blastocystis sp. infection.
In this study, the phyla Firmicutes and Bacteroidetes were most predominant among all the subjects. This is consistent with a previous study on a similar Malaysian population [32,33]. In general, regardless of Blastocystis sp. infection, we found a significant difference in alpha and beta diversity, confirming that bacterial composition in symptomatic and asymptomatic samples is distinct. This confirms that gastrointestinal symptoms are associated with low species richness. LEfSe and relative abundance analysis further confirms alteration in Firmicutes/Bacteroidetes (F/B) ratio. Decreased F/B ratio seen in symptomatic individuals suggests dysbiosis, commonly also seen in inflammatory bowel disease (IBD) patients [34].
Our findings demonstrated that Blastocystis sp.-infected individuals, regardless of symptoms, had decreased alpha diversity and Pielou's evenness. As observed in our recent study, greater amoebic forms and surface fuzzy coat commonly seen in symptomatic isolates [35] suggest a greater interaction with bacteria in these isolates which could contribute to the alteration of microbiota. On the other hand, lower peak cell numbers in the growth profile of symptomatic isolates implicate potential inhibition from accompanying microbiota. These observations suggest a bidirectional interaction between Blastocystis sp. and gut microbiota. However, more data on Blastocystis sp.-gut microbiota across multiple populations is required to corroborate this interaction. A similar study done on pooled symptomatic and asymptomatic populations showed contrasting results where no difference in alpha diversity was detected [36]. This discrepancy could have been potentially contributed by differences in environment and the studied population. Since subtype-influenced associations to gut microbiota have been demonstrated by Tito et al. [7] it is highly likely that the discrepancy is due to analyses being carried out on multiple Blastocystis sp. subtypes (ST 1-7). In this study, the association seen is unique and specific to only Blastocystis sp. ST3. While most other studies have compared Blastocystis sp.-gut microbiota association in diseased or healthy group [5,6,37,38], our study for the first-time reported association of a single subtype (ST 3) of Blastocystis sp. to symptomatic and healthy individuals.
A study by Nagel et al. on Blastocystis sp. from irritable bowel syndrome patients revealed insignificant influence on the gut microbiota [38]. Similarly, in this study, the presence or absence of Blastocystis sp. in symptomatic group did not significantly influence bacterial diversity but changed the abundance of certain bacterial taxa suggesting alterations in bacterial composition. However, in asymptomatic individuals, we saw a significant alteration in gut microbial diversity and composition in Blastocystis sp. infection. Interestingly, different composition of bacteria was seen to be associated with Blastocystis sp. in symptomatic and asymptomatic infections. In symptomatic individuals, bacteria from the family of Prevotellaceae and Rumunicoccaceae were predominant in Blastocystis sp. colonization. Our study is similar to previous reports where Prevotellaceae were positively associated with Blastocystis sp. colonization [5]. Studies have associated bacteria from Prevotellaceae with inflammatory disorders [39] and symptomatic Entamoeba histolytica infection [40]. However, its role especially in symptomatic Blastocystis sp. infection needs further exploration as the parasite often presents features such as amoebic forms and a sticky surface coat [41,42] implicating enhanced interaction with bacteria.
Asymptomatic individuals with Blastocystis sp. colonization were associated with a predominance of bacteria belonging to mainly Firmicutes with reduced diversity. A recent study on the Iranian population has reported similar findings where harmful bacteria were elevated in asymptomatic Blastocystis sp. infection [43]. The findings by Nieves-Ramirez et al [6] showed increased diversity in asymptomatic Blastocystis sp. infection, although there was a similar increase in Firmicutes. Population heterogeneity could be a reason as the study was conducted in the Mexican rural population while the current study was done on the urban population in Malaysia. A significant difference in bacterial composition between Malaysian and western populations [33] suggests the contrasting findings between this study and other studies on gut microbiota-Blastocystis sp. association [5,37]. However, the reduction in richness in asymptomatic Blastocystis sp. infection could be best explained ecologically by an alternative stable state [44], whereby perturbation in gut microflora results in the establishment of a different stable state with associated dynamics such as population fluctuations. This stable state may contribute to specific immunological adjustments as previous reports have noticed reduced fecal calprotectin, IgA level [6], and neutrophil levels [45] in asymptomatic Blastocystis sp. infection. Whether Blastocystis sp. instigates an anti-inflammatory environment for persistent asymptomatic colonization by altering gut bacterial composition warrants more study.
PICRUSt algorithm is commonly used for the functional prediction of the intestinal microbiota [33,46]. In adjunct to microbial diversity and composition, we used the PICRUSt algorithm to further add dimension to metabolic functions in gut microbiota and its alteration after Blastocystis sp. infection in studied subjects. The outcome, for the first time, implies that Blastocystis sp. could be related to modifications of resulting microbial functional pathways. This is likely due to the alteration of microbial composition in Blastocystis sp. infection. We postulated that asymptomatic Blastocystis sp. infection could be associated with an alternative stable state. In symptomatic individuals, Blastocystis sp. altered microbial composition despite the diversity not being significantly affected. Studies suggested that altered bacterial composition can influence how metabolites are processed, resulting in the metabolic pathway and profile changes [2]. Evidence from this study suggests the same, as significant change in metabolic processes are observed in Blastocystis sp. infection. However, the predicted microbial functional pathway only offers preliminary access to understanding microbiota function. Greater depth and details provided by the metabolomic approach and whole genome sequencing would be essential in identifying genes involved in the specific metabolic pathway and metabolite interactions during Blastocystis sp. infection.
To date, no studies have reported the influence of accompanying bacteria on Blastocystis sp. cells. Several studies have suggested important roles of accompanying bacteria in the pathogenesis of intestinal protozoan parasite [9,12]. These studies, however, do not mimic the natural condition of the gut as the parasite cells used were axenic. Hence, the changes seen in the parasites may not translate to the real-time scenario. Here we report the effects of altering the autochthonous bacterial composition in Blastocystis sp. culture. We found that parasite cells isolated from asymptomatic individuals resembled the cells from symptomatic ones upon consistent introduction of bacterial suspension from symptomatic individual and vice versa. These findings demonstrate the role of bacteria in influencing Blastocystis sp. up to protein expression levels where the solubilized protease levels and ability of antigens to proliferate colon cancer cells in vitro were also altered. Although the role of proteases is inconclusive in Blastocystis sp. infection, evidence of degradation of secretory immunoglobulin A [47] and activation of IL-8 gene expression [48] and well-studied pathogenic roles in other intestinal parasites [49,50] potentially implicate it as a virulent factor. Evidence on bacteria engulfing amoebic forms and increased protease activity [24] in Blastocystis sp. as well as lipopolysaccharide (LPS)-induced toll-like receptor activation [51] supports the obligatory role of bacteria in pathogenic characteristics in Blastocystis sp.
Earlier studies demonstrated the presence of physical features such as sticky surfaces, fuzzy coats, and amoebic morphologies indicating interactions with bacteria [52]. With current findings, we are certain that this interaction contributes to shaping the phenotypic feature of Blastocystis sp. cells. Several studies have used phenotypic features to ascribe pathogenic potentials in Blastocystis sp. ST3 when isolated from diseased and healthy groups [41,53,54]. The resulting phenotype reflects a specific microbiota composition, either due to various host-related factors or possibly influenced by Blastocystis sp. itself. The latter is highly likely, as we have seen here and in other recent findings, that Blastocystis sp. modifies the bacterial composition [6,55], although its influence on diversity is contradictory. Even so, nothing is conclusive until the mechanism of specific phyla in cross-talk with Blastocystis sp. is elucidated. Studies on the pathogenicity of Blastocystis sp. have been contradicting for a long time. However, case studies [56] reporting improvement of symptoms upon the extermination of this organism via drug treatment suggest a pathogenic role that could be restricted to only some individuals or certain gut microbial environments. Studies thus far have reported that pathogenic potentials are being assessed in terms of variations in the growth profile, cysteine protease activity, and ability to proliferate cancer cells [57]. Our finding showed that these factors, disparate in Blastocystis sp. isolated from symptomatic and asymptomatic individuals, are largely dependent and altered by the parasite's microbial surroundings. We propose that the pathogenic characteristics may not be wholly exerted by the parasite itself but influenced by factors such as the gut microbiota as well.
Increasing number of studies is beginning to show differences in gut microbiota due to various factors [58]. An individual may undergo alterations in gut microbial environment as a consequence of changing dietary intakes, life stresses, medications, travel and migrations, which are rampant in recent years. A diverse and balanced microbiota profile provide protection to the mucosa [59] and secrete metabolic products such as short chain fatty acids (SCFA) that promote health [60]. Parasitic cells colonizing in such environment may remain asymptomatic [59]. When there is a change in the environment, especially when triggered by certain diet, antibiotic consumption or stress, the microbiota may be altered [1]. This influences the colonizing organism initially harmless to be pathogenic (Fig 7). Although this postulation was derived from the correlation of data without a causal relationship, we believe this is the way forward in understanding the role of Blastocystis sp. in disease and health.
Our study is limited in terms of sample size. This is due to difficulty in obtaining a single subtype of Blastocystis sp. and maintaining the parasite cells in vitro. A similar experimental design, applied to larger sample size, yields more conclusive evidence. This study is also limited in terms of the use of molecular diagnostics for the detections of Blastocystis sp. colonization. Therefore, it is possible that Blastocystis sp.-free group may have individuals who were infected with this organism but was not captured by in-vitro cultivation technique. Also, the quantitative burden of Blastocystis sp. could not be compared between the groups. However, this study, for the first time, has demonstrated gut variation associated with a single Blastocystis sp. subtype. Our study has also shown for the first time the influence of autochthonous bacterial alteration on the phenotype of Blastocystis sp. ST3 cells. In the future, it is essential to characterize the bacterial taxa in close interaction with Blastocystis sp. ST3 and its role in symptomatic and asymptomatic infections.

Conclusion
Many recent studies focused on the effect of Blastocystis sp. in altering the gut microbiota [5,61], however, this is the first study to demonstrate the influence of microbial environment on this prevalent intestinal protozoon. The findings open new vistas in understanding parasite-bacteria interaction, which could help us understand better the pathogenicity of Blastocystis sp. We postulate that the interactions seen between specific intestinal microbiota and Blastocystis sp. influence whether the protozoa will function in a commensal or parasitic role. This study also provides preliminary evidence of a typical intestinal protozoan reverting from a harmless organism to a harmful one.