Phylogenetic analyses of Chilomastix and Retortamonas species using in vitro excysted flagellates

Abstract In vitro excystation of cysts of microscopically identified Chilomastix mesnili and Retortamonas sp. isolated from Japanese macaques and Retortamonas sp. isolated from small Indian mongooses could be induced using an established protocol for Giardia intestinalis and subsequently by culturing with H2S-rich Robinson’s medium supplemented with Desulfovibrio desulfuricans. Excystation usually began 2 h after incubation in Robinson’s medium. DNA was isolated from excysted flagellates after 4 h of incubation or from cultured excysted flagellates. Phylogenetic analysis based on their 18S rRNA genes revealed that two isolates of C. mesnili from Japanese macaques belonged to the same cluster as a C. mesnili isolate from humans, whereas a mammalian Retortamonas sp. isolate from a small Indian mongoose belonged to the same cluster as that of an amphibian Retortamonas spp. isolate from a ‘poison arrow frog’ [sequence identity to AF439347 (94.9%)]. These results suggest that the sequence homology of the 18S rRNA gene of the two C. mesnili isolates from Japanese macaques was similar to that of humans, in addition to the morphological similarity, and Retortamonas sp. infection of the amphibian type in the small Indian mongoose highlighted the possibility of the effect of host feeding habitats.


Introduction
Retortamonadida (retortamonads) are a group of bacterivorous metamonads that produce rigid pyriform cysts and are currently represented by the single family Retortamonadidae, comprising two genera, Chilomastix and Retortamonas (Kulda et al., 2017).These metamonads live predominantly as intestinal endocommensals in both vertebrates and invertebrates (Kulda et al., 2017).However, two species have been reported as potential pathogens that cause diarrhoea in humans (Chilomastix mesnili) or unadapted avian hosts (C.gallinarum) (Boeck et al., 1921;Cepicka et al., 2008;Kulda et al., 2017).Among nonhuman primates, C. mesnili infection cases have also been reported (Levecke et al., 2007); However, genetic information on C. mesnili isolated from nonhuman primates is extremely limited (Jeong et al., 2019).C. mesnili-infected nonhuman primates are often difficult to morphologically distinguish from infected humans.However, Hegner (1924) reported that Chilomastix sp.cysts from Geoffroy's spider monkey (Ateles geoffroyi Kuhl) resembled those of C. mesnili isolated from humans in shape, although they were considerably larger.The results of phylogenetic analyses of Chilomastix spp.and Retortamonas spp., based on 18S rDNA, suggested that Retortamonadida had tendencies of non-monophyly and high genetic diversity of the genus Chilomastix (Cepicka et al., 2008).
In the present study, we attempted to determine the phylogenetic position of morphologically identified C. mesnili and Retortamonas sp.isolated from a Japanese macaque (Macaca fusucata: a nonhuman primate and mainly vegetarian) and determine the effect of feeding habitat on infective Retortamonas sp.species in a small Indian mongoose (Urva auropunctata), which habitually feeds on frogs that may be infected with a different phyletic Retortamonas sp. from that of mammalian hosts.
To obtain valid results from the first genetic analysis of a protozoan parasite and determine the cause of genetic diversity, such as in the genus Chilomastix, establishing a culture system to maintain the protozoan species continuously is necessary.Before establishing culture isolates and performing phylogenetic analyses, an in vitro excystation system for C. mesnili and Retortamonas spp.using cyst-positive samples was established.This is because treatment with the low-pH induction solution in the in vitro excystation protocol using partially purified cysts could be eliminated, often contaminating and easily isolating non-cyst-forming parasitic and coprozoic flagellates into the medium from stool samples of Japanese macaques and small Indian mongooses.Another reason is that, even when the culture isolate could not be established, DNA samples from the excysted flagellates could be obtained directly in a short period without cultivation.In the present study, culture isolates of Retortamonas spp.were established and maintained for over 3 years in Robinson's medium (Robinson, 1968) supplemented with Desulfovibrio desulfuricans (Yoshida et al., 2019).However, a continuous culture system for C. mesnili using motile flagellates excysted from cysts in stool samples of Japanese macaques could not be established using established culture media [Robinson's medium (Robinson, 1968), Balamuth's medium (Balamuth, 1946), TYSGM-9 (Diamond, 1982), and modified TYI-S-33 medium (Keister, 1983)].Through this culture trial, an excystation protocol and a short-term culture system using an egg yolk medium (Balamuth's medium) were established by supplementing with mucin and mucin-degradable Bifidobacterium longum.
Subsequently, we used the established excystation protocol (Boucher & Gillin, 1990) for Giardia lamblia (syn., G. intestinalis) by culturing with Robinson's medium supplemented with D. desulfuricans.In addition, we designed C. mesnili and Retortamonas sp.specific primers for phylogenetic analyses based on the genomic sequences of 18S rRNA genes that were revealed using culture isolates and eukaryote-specific primers (Silberman et al., 2002;Cepicka et al., 2008).

Obtaining cysts
Stool samples from 213 wild Japanese macaques were collected from monkey trails near villages close to mountains at three different places in the Chubu and Kanto regions, Japan, where they were usually populated and obtained food, and from a small wild Indian mongoose obtained from a rural village in West Bengal, India, where it was commonly found.Consequently, 17 Chilomastix mesnili and 6 Retortamobas sp.cyst-positive stool samples of wild Japanese macaques could be obtained and used for establishing those culture isolates.
C. mesnili cysts (MfOkutamaCm and MfChichibuCm) were obtained from C. mesnili cyst-positive stool samples of wild Japanese macaques (inhabiting Okutama, Tokyo, Japan and Chichibu, Saitama, Japan) and identified morphologically by microscopic stool examination.However, a sufficient number of cysts from isolates obtained Braz J Vet Parasitol 2023; 32(4): e011923 3/9 Phylogenetic analyses of Retortamonadidae strains four different samples of Retortamonas spp.(UaBengalR: from a small wild Indian mongoose inhabiting West Bengal, India; MfHotakaR1, R2, and R6: from wild Japanese macaques inhabiting Azumino, Nagano, Japan) for the excystation experiment were difficult to obtain from the cyst-positive stool samples.Nevertheless, every four culture isolates from each cyst-positive stool sample could be established using Robinson's medium supplemented with D. desulfuricans and maintained at 35.5 °C, which exerts an excystation stimulatory effect on the cysts of Entamoeba chattoni (E.polecki subtype 2) isolated from Japanese macaques, E. polecki (subtype 1 and 3), E. suis and Endolimax sp.isolated from swine, and Nyctotherus teleacus isolated from a radiated tortoise.They could be subcultured using the same Robinson's medium for over a year (Robinson, 1968;Yoshida et al., 2019;Suzuki et al., 2020).Under these culture conditions, the four Retortamonas spp.isolates spontaneously and continuously produced a sufficient number of cysts in the medium for over 3 years.The size of the culture flagellates of these isolates was approximately 2-3-fold the size of the cysts produced.The cysts of UaBengalR and MfHotakaR1, R2, and R6 were obtained by culturing these isolates and used as cyst samples for in vitro excystation experiments.

Partial purification of cysts
Cyst-positive stool samples were suspended in 10 times the volume of high-purity water in 50-mL plastic centrifuge tubes and left to stand for a few minutes to allow the precipitation of large debris.Thereafter, the supernatants were transferred into new plastic tubes using a transfer pipette (volume: 3 mL, drops per mL: 22) to further remove debris which could not pass through the aperture of the pipette.Next, the supernatants were stored at 4 °C with suspension in high-purity water after washing via centrifugation (275 × g for 5 min twice) with 10 times the volume of high-purity water and used within a few weeks.
Before excystation, the cysts in these suspensions (2 mL) were partially purified using the Percoll density gradient centrifugal separation method (400 × g, 20 min) with 4 mL of 40 and 100% Percoll PLUS (GE Healthcare Bio-Sciences AB, Little Chalfont, UK) in a 50-mL plastic tube.The cyst-rich samples of C. mesnili and four Retortamonas spp.were collected into 10 times the amount of high-purity water from the border fraction, washed twice with high-purity water via centrifugation (275 × g for 5 min), and stored at 4 °C until use.

Excystation and harvest of excysted flagellates
The established excystation protocol for G. lamblia (syn., G. intestinalis) cysts (Boucher & Gillin, 1990) that employs a low-pH induction solution (pH 4.0) consists of solutions of L-cysteine hydrochloride and reduced glutathione and is incubated with cysts at 38 °C for 30 min (step 1).After washing with Hanks' solution (pH 7.4 ) via centrifugation (275 × g for 5 min) once, a digestive enzyme (chymotrypsin) solution (pH 8.0) was applied and incubated with the cysts at 38 °C for 30 min (step 2).Thereafter, the cyst fraction obtained via centrifugation (275 × g for 5 min) was transferred into H 2 S-rich Robinson's medium supplemented with D. desulfuricans and incubated at 35.5 °C (step 3).
Percoll density gradient centrifugal separation and treatment with a low-pH induction solution in excystation (step 1) could eliminate coexisting non-cyst-forming parasitic flagellates, such as Pentatrichomonas spp., and coprozoic flagellates.However, these steps could not eliminate the cysts of parasitic flagellate species, such as Enteromonas sp., in the case of coexistence in stool samples of Japanese macaques and small Indian mongooses.
Excysted motile flagellates of the two C. mesnili and four Retortamonas spp.isolates usually began to appear 2 h after incubation in H 2 S-rich Robinson's medium (excystation step 3).Excysted motile flagellates of C. mesnili (MfOkutamaCm and MfChichibuCm) in the medium were harvested after 4 h of incubation via centrifugation (275 × g for 5 min).The number of motile C. mesnili flagellates gradually decreased after excystation and disappeared 24-48 h after incubation in Robinson's medium.The incubation time required to obtain the highest number of motile C. mesnili flagellates was 3-5 h (e.g. when the initial concentration of cysts was 5-6 × 10 4 /mL, approximately 2-3 × 10 4 /mL motile flagellates usually appeared after 4 h of incubation).One portion of the excysted flagellates of C. mesnili was inoculated into Balamuth's medium supplemented with mucin and Bifidobacterium longum, and grown flagellates were harvested.Excysted flagellates of the four Retortamonas spp.isolates were harvested after 48-72 h of incubation.This was because the cells were efficiently grown in H 2 S-rich Robinson's medium and could be cultured continuously.It was also confirmed that the size and morphological features of spontaneously produced cysts in the medium before and after the excystation experiment were the same.
Mucin was used because of its growth-promoting effect on some intestinal protozoan parasites, such as Entamoeba histolytica, in TYGM9 culture medium (Diamond, 1982).Balamuth's medium was developed to establish a culture isolate of C. mesnili from Japanese macaques for a more reliable phylogenetic analysis.
PCR products were directly sequenced using an ABI Prism BigDye Terminator v3.1 cycle sequencing ready reaction Kit on an ABI Prism 3500 Genetic Analyser (Applied Biosystems Japan, Ltd., Tokyo, Japan).Multiple alignments and phylogenetic analyses of the 18S rRNA gene sequences of C. mesnili and Retortamonas spp.isolates were performed using ClustalW and the maximum-likelihood (ML) method in MEGA software version 6 (Ronquist et al., 2012;Tamura et al., 2013).The ML tree was derived using the Kimura 2-parameter model, employing estimates of the gamma distribution with five rate categories.Statistical significance was evaluated by bootstrapping with 1,000 replicates.The ML tree data files were visualised using MEGA version 6.
The two PCR primer sets for C. mesnili and Retortamonas spp.also confirmed that the primer sets did not amplify the DNA extracted from the excysted flagellates of the microscopically identified Enteromonas sp.isolated from Japanese macaques.
A flowchart of the phylogenetic analysis approach used in the present study is shown in Figure 1.

Results
The excysted flagellates of the C. mesnili (MfOkutamaCm) and Retortamonas sp.(MfHotakaR1) isolates are shown in Figure 2d and e.The cysts of C. mesnili (MfOkutamaCm) from the Japanese macaque measured 9.4-11.3μm in length and 6.6-7.8 μm in breadth (see Figure 2 legend).C. mesnili infecting Japanese macaques (a nonhuman primate) (Figure 2a, b, and c) is often difficult to distinguish morphologically from C. mesnili infecting humans.However, the size of the cysts of C. mesnili from humans differed from that from Japanese macaques.In our results, the cysts of another C. mesnili-infected Japanese macaque (MfChichibuCm) from a different populating area showed a tendency to be larger, mainly in breadth, than those from humans (7.92-9.9 and 6.93-7.92μm in length and breadth, respectively).Continuous culture of the excysted flagellates of C. mesnili (MfOkutamaCm and MfChichibuCm) confirmed the proliferation of flagellates in Balamuth's medium supplemented with mucin and B. longum.They could survive for up to 13 days, although the culturable period was still short.However, this result demonstrates the possibility of establishing a culture of C. mesnili isolated from Japanese macaques.

6/9
Phylogenetic analyses of Retortamonadidae strains Notably, we found that the size of Retortamonas spp.flagellates (UaBengalR) was notably larger (approximately 1.5 times) than that of MfHotakaR1, although the cysts of both isolates were approximately of the same size (Figure 3a, b, c, d, and e, and their measured values are described in the figure legend).

Discussion
To the best of our knowledge, this is the first study to show the genetic similarity of 18S rRNA between C. mesnili (NOVA: KC960589 and KC960590) isolated from humans and C. mesnili [MfOkutamaCm (LC752183) and MfChichibuCm (LC771049)] isolated from Japanese macaques, in addition to their morphological similarity.However, the cysts of C. mesnili in infected Japanese macaques appear larger than those in infected humans.This result is consistent Hegner (1924), who reported on the size of Chilomastix sp.cysts from Geoffroy's spider monkey (Ateles geoffroyi Kuhl).They found that cysts in humans infected with C. mesnili ranged from 7 to 9 μm in length and from 4 to 6 μm in breadth, while those from Geoffroy's spider monkey measured from 8.47 to 10.26 μm in length and from 5.92 to 7.62 μm in breadth.The reason for the difference in cyst size between humans and nonhuman primates [Japanese macaque and Geoffroy's spider monkey (Hegner, 1924)] is unclear.However, differences in cyst size may be affected by flagellate size.The motile flagellates of C. mesnili in humans display a wide range of sizes, ranging from 3 to 19 μm in length and 2 to 9 μm in width (Boeck, 1921).However, in the case of C. mesnili from Japanese macaques (MfOkutamaCm and MfChichibuCm), no motile flagellates <10 μm in length [e.g., 12.5-17.8 μm in length and 4.4-7.6 μm in width (MfOkutamaCm)] were observed.
In the culture process, we also showed that mucin and B. longum exerted growth-promoting effects on C. mesnili (MfOkutamaCm and MfChichibuCm).Bifidobacterium species can colonise the intestinal mucosa through their adhesion proteins (Nishiyama et al., 2020) and their ability to degrade mucin (Ruas-Madiedo et al., 2008), which is the principal component of mucus in the mucosa.Bifidobacterium species (B.animales, B-12 strain), other than B. longum, exerted the same growth-promoting effect as mucin in C. mesnili (MfOkutamaCm) (data not shown).These bacteria inhabit the gut of primates (Modrackova et al., 2021).These results suggest that C. mesnili (MfOkutamaCm and MfChichibuCm) secretes an adhesin-like substance that adheres to the intestinal mucosa, such as some intestinal protozoan parasites that cause diarrhoea [Tritrichomonas foetus (Tolbert et al., 2013) and E. histolytica (Mann et al., 1991)].
The phylogenetic analysis of Retortamonadida and genus Chilomastix, based on 18S rDNA, first reported by Cepicka et al. (2008), suggested that Retortamonadida had tendencies of non-monophyly and high genetic diversity of the genus Chilomastix.Up to 2023, the 18S rRNA gene sequences of 12 DNA clones obtained from 4 isolates (CAR5A, FAB, FAB2, and NOVA) of C. mesnili have been registered to GenBank by Cepicka et al. (2008) and direct submitted and registered to GenBank by Cepicka in 2013.According to our phylogenetic analysis using Phylogenetic analyses of Retortamonadidae strains the 12 18S rRNA gene sequences of C. mesnili from humans, it branched into 2 clusters: cluster A (CAR5A, FAB, and FAB2) and cluster B (NOVA) of C. mesnili isolates in the phylogenetic tree.In the present study, the 18S rRNA gene sequences of two C. mesnili isolates [MfOkutamaCm (LC752183) and MfChichibuCm (LC771049)] from Japanese macaque belonged to cluster B (NOVA) in the phylogenetic tree.Silberman et al. (2002) reported that the identities among mammalian isolates ranged from 99.5 to 99.8%, whereas those among two amphibian isolates were slightly lower (96.5%),suggesting a tendency for 18S rRNA gene sequence homologies among Retortamonas spp. of mammalian and amphibian animals.Furthermore, they found that the overall identity between the mammalian and amphibian retortamonad 18S rRNA genes was 74.1%.We assumed that the high homology (94.9%) of the 18S rRNA gene sequences of Retortamonas spp.(UaBengalR, LC752184) with that of the isolate from the 'poison arrow frog' (AF439347) might be due to the feeding habits of the small Indian mongoose, which feeds on amphibians, insects, small mammalian animals, and reptiles.These results suggest that Retortamonas sp.infection may be affected by the feeding habitat of its host.
Based on the results, the flagellates excysted from cysts appeared to be reliable specimens, similar to culture isolates for the genetic analysis of cyst-forming parasitic flagellates, such as those belonging to Retortamonadidae and Enteromonadidae.Due to the high genetic diversity of the Chilomastix genus (Cepicka et al., 2008) and low but non-negligible genetic diversity even among C. mesnili species, and the limited knowledge of the genetic information on C. mesnili isolated from nonhuman primates, further genetic analyses using more culture isolates of C. mesnili from humans and nonhuman primates are necessary.

Conclusion
The in vitro excystation of Chilomastix and Retortamonas species was induced.The excysted flagellates were reliable specimens for phylogenetic analysis.Novel insights into C. mesnili isolated from nonhuman primates were gained from genetic analyses.

Figure 1 .
Figure 1.Flowchart of the approach to phylogenetic analysis in the present study.

Figure 4 .
Figure 4. Phylogenetic analysis of Chilomastix mesnili (MfOkutamaCm and MfChichibuCm) isolates based on their 18S rRNA gene sequences.Maximum likelihood (ML) tree of C. mesnili based on 18S rRNA sequences.An ML tree was derived using the Kimura-2 parameter model.Significant bootstrap support (>500) from 1,000 replicates is indicated on the left of the supported nodes.The scale bar represents the evolutionary distance based on the number of changes per site.The numbers in parentheses represent GenBank accession numbers.*C.mesnili isolates from Japanese macaques in the present study.

Figure 5 .
Figure 5. Phylogenetic analysis of Retortamonas spp.(UaBengalR and MfHotakaR1, R2, and R6) isolates based on their 18S rRNA gene sequences.Maximum likelihood (ML) tree of Retortamonas spp.and those of Fornicata parasites based on 18S rRNA sequences.Phylogenetic analysis of Retortamonas spp. was performed as described in Figure 4. *Retortamonas spp.isolated from Japanese macaques and small Indian mongooses.