Characterizing the Proliferation Patterns of Representative Microsporidian Species Enlightens Future Studies of Infection Mechanisms

Background: Microsporidia are a group of pathogens that infect all kinds of animals, such as humans, silkworms, honeybees, and shrimp; they, therefore, pose a severe threat to public health and the economy. There are over 1500 species of microsporidia that have been reported, among which Encephalitozoon hellem and Nosema bombycis are the representative zoonotic and insect-infecting species, respectively. Investigating their cell infection patterns is of great significance for understanding their infection mechanisms. Methods: Specific probes were designed for the ribosomal RNA sequences of microsporidia. Fluorescence in situ hybridization (FISH) was used to trace the proliferation cycle of the pathogens in different cells. Results: Here, two rRNA large subunit gene (LSUrRNA) probes specifically labeling N. bombycis were obtained. The life cycle of N. bombycis in silkworm cells and E. hellem in three kinds of host cells was graphically drawn. N. bombycis meronts were first observed at 30 hours post-infection (hpi), and they began merogony. Sporonts were observed at 42 hpi, and the first entire proliferation cycle was completed at 48 hpi. The proliferation cycle of E. hellem in RK13 and HEK293 epithelial cells was almost the same, completing the first life cycle after 24 hpi, but it was significantly delayed to 32 hpi in RAW264.7. Conclusions: Specific FISH probes were established for labeling microsporidia in multiple host cells. The proliferation characteristics of representative zoonotic and insect-infecting microsporidian species were clarified. This study provides an experimental pattern for future analyses of microsporidian infection mechanisms.


Introduction
Microsporidia are obligate intracellular pathogens that infect a wide variety of animal hosts. Nosema bombycis infects silkworms via both horizontal and vertical transmissions, making it the causative agent of pébrine disease and a significant threat to sericulture worldwide [1][2][3]. Encephalitozoon hellem is a zoonotic microsporidian species that is frequently identified in avian hosts, livestock, and even humans. E. hellem infections cause keratoconjunctivitis, nephritis, and diarrhea in humans and could be fatal in immunocompromised individuals [4][5][6][7].
All microsporidian species share a general life cycle. It could be divided into the following three phases: the extracellular invasive phase, the proliferative (vegetative) phase, and the sporogonial (generative) phase [8,9]. During the infection phase, the microsporidian spores germinate and extrude a polar tube, which can penetrate host cell membranes and enable the transport of infective sporoplasms into the host cells [10]. The intracellular meront has a single-layer membrane structure, but there is no clear outline under white light, making it difficult to observe. The generative stage includes sporonts (cells that The microsporidian Nosema bombycis CQ 1 was isolated from the silkworm Bombyx mori in Chongqing, China (102059, CVCC). Encephalitozoon hellem was gifted by Dr. Han Bing from the Shandong University, China. Heavily infected silkworm pupae were removed from the skin and degenerated midgut. The pupa tissues were gently homogenized in physiological saline and filtered using a 10 mL centrifuge tube containing thick cotton. The spores were purified by centrifugation at 10,000× g for 20 min in 90% Percoll (17089102, Cytiva). The purified spores were washed with sterile water three times, suspended in 0.5 mL of water, and, finally, stored at 4 • C for one month [28]. E. hellem spores were collected from infected rabbit kidney cells (RK13) (CCL37, ATCC) and purified by centrifugation at 10,000× g for 15 min in 75% Percoll to collect the lowest pathogen [29]. In this research, we used fresh spores for each infection.

FISH Probes for Labeling the Microsporidia
The complete sequence of the ribosomal RNA (rRNA) and the secondary structure model of the rRNA large subunit gene (LSUrRNA) of N. bombycis were obtained [35]. According to the N. bombycis rRNA complete sequence, we designed nine rRNA probes and synthesized them with a Quasar 570 (Cy3) 5 modification with high-performance liquid chromatography purification by Sangon Biotech Co., Ltd. (Shanghai, China) ( Table 1). In addition, a Cy3-ACTCTCACACTCACTTCAG probe (HEL878F probe) had been reported for the specific labeling of E. hellem [20,36].

FISH Staining
The infected cells were fixed with 4% paraformaldehyde for 20 min and washed with phosphate-buffered saline (PBS) three times, followed by incubation overnight at 46 • C with hybridization buffer (900 mM NaCl, 20 mM Tris [pH 7.5], and 0.01% sodium dodecyl sulfate [SDS]) containing 5 ng/µL of the probes. The infected cells were then washed at 48 • C for 1 h in wash buffer (900 mM NaCl, 20 mM Tris [pH 7.5], 0.01% SDS, and 5 mM ethylenediaminetetraacetic acid) [24], after which they were washed in PBS three times and then stained in DAPI (4 6-diamidino-2-phenylindole, Sigma) for 30 min. After three additional washes with PBS and the addition of ProLong1 Gold anti-fade reagents, the infected cells were observed using Olympus FV1200 laser scanning confocal microscopy. With the photos, we randomly selected 10 visual fields, and the ratio of infected cells in each field was calculated by manual counting. The lengths and widths of the parasites were determined using ImageJ 1.52v (https://imagej.nih.gov/ij/index.html, accessed on 13 April 2020), and they were analyzed for a t-test with GraphPad Prism v6.01 (https://www.graphpad.com/scientific-software/prism/, accessed on 21 September 2012). The experiments were conducted at least three times.

Pathogen Infections Demonstrated by FISH
We designed and tested nine probes. The results showed that there are two rRNAspecific fluorescence in situ hybridization (FISH) probes NbLSU-V1 and NbLSU-1 that could specifically label N. bombycis meronts inside BmEs but not Vairimorpha necatrix ( Figure 1A,B). The NbLSU-V1 probe was chosen for the subsequent experiments. On the other hand, E. hellem meronts inside host cells could be labeled by the rRNA-specific HEL878F probe ( Figure 1C). The FISH probes can label meronts, sporonts, and sporoblasts, but not mature spores. Thus, the combined application of DAPI and FISH can distinguish pathogens in developmental stages from dormant spores. could specifically label N. bombycis meronts inside BmEs but not Vairimorpha necatrix ( Figure 1A,B). The NbLSU-V1 probe was chosen for the subsequent experiments. On the other hand, E. hellem meronts inside host cells could be labeled by the rRNA-specific HEL878F probe ( Figure 1C). The FISH probes can label meronts, sporonts, and sporoblasts, but not mature spores. Thus, the combined application of DAPI and FISH can distinguish pathogens in developmental stages from dormant spores.  The specificity of the NbLSU-1 (A) and NbLSU-V1 (B) probes was detected in N. bombycisand V. necatrix-infected BmE cells, respectively. The specificity of the HEL878F probes (C) was detected in E. hellem-infected RK13 cells. Laser confocal microscopy showed a nucleus labeled with DAPI (blue). The microsporidia were labeled with the rRNA FISH probe (red). The white arrow indicates FISH staining of multinucleate microbes (proliferative or sporogonial phases); the yellow arrow indicates V. necatrix xenoma; the arrowhead indicates the mature spores. Scale bar, 5 µm.

The Proliferation and Life Cycle of N. bombycis in BmE Cells
A typical time course of an N. bombycis infection and proliferation within host cells is illustrated in Figure 2. As shown, no pathogens were detected by 24 hpi. By 30 hpi, the parasites were detected by FISH staining for the first time, indicating that they were in merogony ( Figure 2A). The meronts gradually grew larger and became irregular in shape until they occupied about a quarter of the host cytoplasm. The nuclei were irregularly shaped and less compact, possibly suggesting that the chromosomes were replicating. By 42 hpi, sporogony was observed. The sporonts were bounded by a typical plasma membrane and were fairly round in shape, with 1.7-2.8 µm in width and 3 µm in length on average ( Figure 2B). The two sets of paired nuclei of sporonts were clearly visible. By 45 hpi, the parasites were in the sporoblast stage and elongated up to 7 µm in length, each containing two nuclei (Figures 2 and S1). By 48 hpi, some pathogens had completed sporulation and produced mature spores, which could not be labeled by the probe. Finally, a massive production of mature spores was observed after 54 hpi ( Figure 2). until they occupied about a quarter of the host cytoplasm. The nuclei were irregularly shaped and less compact, possibly suggesting that the chromosomes were replicating. By 42 hpi, sporogony was observed. The sporonts were bounded by a typical plasma membrane and were fairly round in shape, with 1.7-2.8 µm in width and 3 µm in length on average ( Figure 2B). The two sets of paired nuclei of sporonts were clearly visible. By 45 hpi, the parasites were in the sporoblast stage and elongated up to 7 µm in length, each containing two nuclei (Figures 2 and S1). By 48 hpi, some pathogens had completed sporulation and produced mature spores, which could not be labeled by the probe. Finally, a massive production of mature spores was observed after 54 hpi ( Figure 2).  Additionally, some adjacent host cells were found to fuse with each other after 48 hpi ( Figure 3). Host cell fusion was also observed in Nematocida parisii-infected Caenorhabditis elegans [23]. Therefore, in BmEs, N. bombycis begins to proliferate at 30 hpi and completes the first round of life cycles at 48 hpi, after which the pathogen probably starts the secondary infections.
Additionally, some adjacent host cells were found to fuse with each other after 48 hpi (Figure 3). Host cell fusion was also observed in Nematocida parisii-infected Caenorhabditis elegans [23]. Therefore, in BmEs, N. bombycis begins to proliferate at 30 hpi and completes the first round of life cycles at 48 hpi, after which the pathogen probably starts the secondary infections.

The Proliferation and Life Cycle of E. hellem in RK13, HEK293, and RAW264.7 Cells
It was found that the proliferation cycles of E. hellem in RK13 and HEK293 were identical. At 8 hpi, the pathogen was labeled with the FISH probe for the first time in both cells ( Figure 4A,B). Multinucleated meronts were observed at 12 hpi, which were thin or irregular and located at the edge of the plasma membrane. At 16 hpi, the volume of the pathogen was further expanded, the shape was more irregular, and the number of nuclei significantly increased, but the pathogen outline was still not visible at this time, indicating they were also in the proliferative phase. At 20 hpi, sporogony was observed, and the parasitophorous vacuole (PV) was visible in the DIC. The mature spores were visible after 24 hpi, and they were detected with UV light but not with the FISH probe ( Figure 4A,B). In the case of the RK13 and HEK293 cell infections, E. hellem invaded the host cells after 8 hpi and began to proliferate, followed by the proliferative phase at 8-16 hpi and sporogonial stage at 16-20 hpi, and finally completed a life cycle round at 24 hpi.
In comparison, the E. hellem meronts in the RAW264.7 cells appeared after 12 hpi. At 24 hpi, the spores gradually formed the spore wall in the RAW264.7 cells. After 32 hpi, a round of proliferation in RAW264.7 was completed. In general, the proliferation cycle of E. hellem in RAW264.7 was delayed by about 8 h compared with that in RK13 and HEK293 epithelial cells ( Figure 4C).
Our results also demonstrated that an E. hellem infection was not completely synchronous. At 36 hpi, a larger PV was observed, and the mature spores in the inner layer, while in the outer layer, were still in the stage of meronts ( Figure 4A), which further confirms that the outer edge meronts of PV were more favorable to exchange substances. At 24 and 48 hpi, the developmental stages were uneven, and we could observe the co-existence of

The Proliferation and Life Cycle of E. hellem in RK13, HEK293, and RAW264.7 Cells
It was found that the proliferation cycles of E. hellem in RK13 and HEK293 were identical. At 8 hpi, the pathogen was labeled with the FISH probe for the first time in both cells ( Figure 4A,B). Multinucleated meronts were observed at 12 hpi, which were thin or irregular and located at the edge of the plasma membrane. At 16 hpi, the volume of the pathogen was further expanded, the shape was more irregular, and the number of nuclei significantly increased, but the pathogen outline was still not visible at this time, indicating they were also in the proliferative phase. At 20 hpi, sporogony was observed, and the parasitophorous vacuole (PV) was visible in the DIC. The mature spores were visible after 24 hpi, and they were detected with UV light but not with the FISH probe ( Figure 4A,B). In the case of the RK13 and HEK293 cell infections, E. hellem invaded the host cells after 8 hpi and began to proliferate, followed by the proliferative phase at 8-16 hpi and sporogonial stage at 16-20 hpi, and finally completed a life cycle round at 24 hpi.
In comparison, the E. hellem meronts in the RAW264.7 cells appeared after 12 hpi. At 24 hpi, the spores gradually formed the spore wall in the RAW264.7 cells. After 32 hpi, a round of proliferation in RAW264.7 was completed. In general, the proliferation cycle of E. hellem in RAW264.7 was delayed by about 8 h compared with that in RK13 and HEK293 epithelial cells ( Figure 4C).
Our results also demonstrated that an E. hellem infection was not completely synchronous. At 36 hpi, a larger PV was observed, and the mature spores in the inner layer, while in the outer layer, were still in the stage of meronts ( Figure 4A), which further confirms that the outer edge meronts of PV were more favorable to exchange substances. At 24 and 48 hpi, the developmental stages were uneven, and we could observe the coexistence of many mature pathogens and meronts in different cells, possibly related to the secondary infection of the spores ( Figure 4A).

The Infection Patterns of N. bombycis and E. hellem
Next, we calculated the percentage of infected cells at each sampling time. The infection rate increased in RK13 and RAW264.7 and synchronously reached 23% at 40 hpi. This implied that most pathogens had completed their first life cycle at this time ( Figure 5A). Furthermore, the infection rate increased sharply after 40 hpi in RK13, but the infection

The Infection Patterns of N. bombycis and E. hellem
Next, we calculated the percentage of infected cells at each sampling time. The infection rate increased in RK13 and RAW264.7 and synchronously reached 23% at 40 hpi. This implied that most pathogens had completed their first life cycle at this time ( Figure 5A). Furthermore, the infection rate increased sharply after 40 hpi in RK13, but the infection was always relatively slower in RAW264.7, suggesting that the infection rate of E. hellem was significantly lower in RAW264.7 than that in RK13. This discrepancy may be caused by the different mechanisms of pathogen infection and immune pressure. after 42-48 hpi, implying that most pathogens had completed the initial infection phase and began the proliferate phase at 42 hpi. Subsequently, the infection rate stayed at a certain higher level from 48 to 60 hpi, suggesting that the rates of parasite infection and host cell proliferation had achieved an equilibrium. After 60 hpi, however, the infection rate sharply increased to about 40%, indicating that the pathogens quickly multiplied by the second or third round of infection ( Figure 5B).

Discussion
Studies on microsporidia have been carried out for more than 150 years. To observe their infections in situ, different staining methods were developed [8,19,24,37]. Among which, fluorescence in situ hybridization (FISH) shows comparative advantages in resolution and usability. As shown in this study, nearly all of the stages of the life cycles of N. bombycis and E. hellem were labeled, except for the mature spores.
A total of nine probes were designed, which were derived from the small subunit (SSU) rRNA and internal transcribed spacer (ITS) region, as well as the linear and hypervariable areas of the secondary structure LSU rRNA of N. bombycis [35]. However, only the probes from the linear (NbLSU-1) and hypervariable regions (NbLSU-V1) were well labeled. This phenomenon was also found in the relative fluorescence intensity analysis of the oligonucleotide probes for the model 16S rRNA secondary structure in Escherichia coli [38]. Although the E. hellem HEL878F probe had been reported, the FISH protocol was completely different from the previous one. In summary, the current study provides an effective and useful method for precisely tagging the microsporidian life cycle. Although the life cycle of E. hellem has been explored in a single host cell [25,26,36,39,40], our study compared E. hellem proliferation in multiple cells for the first time. The proliferation of E. hellem in the macrophage cell line showed up 8 h later than that in the epithelial cell line, which may be caused by the immune pressure of macrophages. It is known that macrophages are the intermediate link between innate immunity and adaptive immune response and are essential for the early invasion of pathogens [41]. In addition, macrophages are reported to recognize microsporidian infections via toll-like receptors (TLR2) and the release a large number of defense mediators, such as IFN-γ and IL-12, against other Similarly, we analyzed the pattern of N. bombycis in BmE cells. We found that the infection rate did not significantly change until 42 hpi, but a rapid increase was detected after 42-48 hpi, implying that most pathogens had completed the initial infection phase and began the proliferate phase at 42 hpi. Subsequently, the infection rate stayed at a certain higher level from 48 to 60 hpi, suggesting that the rates of parasite infection and host cell proliferation had achieved an equilibrium. After 60 hpi, however, the infection rate sharply increased to about 40%, indicating that the pathogens quickly multiplied by the second or third round of infection ( Figure 5B).

Discussion
Studies on microsporidia have been carried out for more than 150 years. To observe their infections in situ, different staining methods were developed [8,19,24,37]. Among which, fluorescence in situ hybridization (FISH) shows comparative advantages in resolution and usability. As shown in this study, nearly all of the stages of the life cycles of N. bombycis and E. hellem were labeled, except for the mature spores.
A total of nine probes were designed, which were derived from the small subunit (SSU) rRNA and internal transcribed spacer (ITS) region, as well as the linear and hypervariable areas of the secondary structure LSU rRNA of N. bombycis [35]. However, only the probes from the linear (NbLSU-1) and hypervariable regions (NbLSU-V1) were well labeled. This phenomenon was also found in the relative fluorescence intensity analysis of the oligonucleotide probes for the model 16S rRNA secondary structure in Escherichia coli [38]. Although the E. hellem HEL878F probe had been reported, the FISH protocol was completely different from the previous one. In summary, the current study provides an effective and useful method for precisely tagging the microsporidian life cycle. Although the life cycle of E. hellem has been explored in a single host cell [25,26,36,39,40], our study compared E. hellem proliferation in multiple cells for the first time. The proliferation of E. hellem in the macrophage cell line showed up 8 h later than that in the epithelial cell line, which may be caused by the immune pressure of macrophages. It is known that macrophages are the intermediate link between innate immunity and adaptive immune response and are essential for the early invasion of pathogens [41]. In addition, macrophages are reported to recognize microsporidian infections via toll-like receptors (TLR2) and the release a large number of defense mediators, such as IFN-γ and IL-12, against other intracellular viruses, bacteria, and parasites [41,42]. The discrepancy in the proliferation patterns is also reported in the infection cycle of Toxoplasma gondii, which showed a 2.6 h delay in HeLa cells compared with bovine embryonic skeletal muscle cells [43,44]. The proliferation pattern of N. bombycis used to be studied by Giemsa staining and TEM and was found to be variable between Bombyx mori ovarian (BmN) and Antheraea eucalypti cells [8,37,[45][46][47][48]. In A. eucalypti cells, N. bombycis sporoplasms were initially observed after 6 hpi and developed to meronts by 18 hpi, and the mature spores were present after 42 hpi [49]. N. bombycis were found to develop slower in BmN cells, in which the sporoblasts appeared after 54 hpi, the mature spores were observed after 60 hpi, and the massive mature spores were formed by 96 hpi [27]. In our study, the N. bombycis life cycle in BmE cells is shorter than that of BmN but longer than that of A. eucalypti cells. None of these cells are immune cells; the reason for the discrepancy is unclear. However, pathogen adaptation, dosage of infection, culture environment, and different staining methods are probably important factors.
Moreover, it is worth mentioning that the FISH probes could specifically identify meronts, sporonts, and sporoblasts. Our study provides a reference for the selection of appropriate cell lines for future E. hellem infection studies, as well as for the determination of the infection time in different cell lines. Additionally, the current study is based on in vitro cell cultures and does not implicate the adaptive immunity of the host. For future studies, we will expand our investigations to different cell lines and tissues to provide a better understanding of microsporidian proliferation patterns within hosts.

Conclusions
In summary, our study verified two FISH probes for specific microsporidian labeling, characterized the proliferation patterns of the representative microsporidian species N. bombycis and E. hellem for the first time, and enlightened future investigations of microsporidian infection mechanisms.
Author Contributions: T.L. and Z.Z. contributed to the conception and design of the study; J.L., H.G., C.X. and J.X. contributed to the experimental analysis; J.L. and Z.Z. contributed to the data analysis; J.L. and T.L. wrote the first draft of the manuscript. All authors have read and agreed to the published version of the manuscript.