3D reconstruction of A. algerae spores

2D transmission electron microscopy (TEM) images of microsporidia have shown that the PT is arranged as a coil within the spore, alongside other organelles[22,33,35,36]. How the coils of the PT are organized in 3D and its spatial relationship to other organelles within the spore is less well understood. We used SBFSEM[37] to generate 3D reconstructions of the PT in intact dormant spores of the microsporidian species Anncaliia algerae, which can infect both invertebrate and vertebrate hosts, including humans[38–41]. SBFSEM allows us to automate the sectioning and imaging of a series of 50 nm slices through a block of spores in a high throughput manner[42]. A typical field of view contained approximately 50 spores in several different orientations, from which we reconstructed 20 spores in 3D (Fig 1A, S1 Table).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Serial block-face scanning electron microscopy imaging of intact A. algerae spores.

(a) Representative data from SBFSEM imaging. Samples were serially sliced at 50 nm thickness (left), and a representative slice is shown (right). (b) Representative SBFSEM slice highlighting segmented organelles. Original micrograph is shown (left), as well as the same image with color overlays indicating segmented organelles (right): exospore (orange), endospore (yellow), PT (blue), vacuole (red), posterior polaroplast (purple), and anchoring disc (green). The anterior polaroplast is not segmented, as we do not clearly observe it. Magenta arrow indicates the thinnest part of the endospore layer where the anchoring disc is localized. (c) Representative 3D reconstruction of an A. algerae spore from SBFSEM data. Each color represents an individual organelle; color code as in (b).

https://doi.org/10.1371/journal.ppat.1008738.g001

We segmented the outermost layer of the spore, the exospore, as well as the underlying chitinous endospore, which is adjacent to the plasma membrane (Fig 1B, S1 Video). In the egg-shaped A. algerae spores, the exospore layer is uniformly 0.16 ± 0.03 μm thick, and the thickness of the endospore varies from 0.14 μm to less than 0.08 μm towards the anterior tip of the spore. We found that the apical part of the endospore is the thinnest, as previously reported[35,43], and this region has been hypothesized to be the site of initiation for PT firing[23,44] (Fig 1B and 1C). The spores are 3.9 ± 0.4 μm in length along the anterior-posterior (A-P) axis (S1A Fig), consistent with previous work[40], and the average spore volume is 8.8 ± 1.4 μm³ (S1B Fig).

Spatial organization of the polar tube in dormant A. algerae spores

We segmented the entire A. algerae PT to obtain spatial information on how it is organized in the dormant spore (Fig 1C, S1 Video). The PT occupies only ~3.5% of the total spore volume (S1B Fig), yet visually is the most striking feature of the spores when imaged by either SBFSEM or 2D TEM. As previously reported[45], the PT can be divided into two main parts: 1) a linear segment that emanates from the anterior tip of the spore and extends towards the posterior end, and 2) a coil of PT around the middle and posterior end of the spore (Fig 2A). At least two models for the connection between these parts have been proposed (S2 Fig). In the first model, the linear segment extends nearly the length of the spore, and connects to the posterior end of the coiled region[46]. In the second model, the linear segment is shorter, and connects with the anterior end of the coiled region[44]. Our data show that the straight segment connects to the anterior end of the coiled region (Fig 2A), consistent with the second model. The coils of the PT are approximately parallel to each other, but the stack of coils is tilted relative to the A-P axis of the spore (Fig 2B, S1D Fig). Here we found that the coiled segment of the PT consists of 7 turns on average (S1C Fig), in contrast to previous data, which suggested 8–11 coils[40]. The differences may be due to differences in the source of spores or propagation of the spores in different hosts.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Configuration of the PT and other organelles in intact A. algerae spores.

(a) Representative 3D reconstruction of an A. algerae spore showing the relative orientations of the PT (blue), anchoring disc (green), posterior polaroplast (purple), and vacuole (red). (b) Measurement of the angle of PT coils relative to the anterior-posterior (A-P) axis of the spore. Schematic showing how the angle was measured between the A-P axis and the PT coil plane (left). Representative spore with the average angle and standard deviation annotated (right, n = 18). See also S1D Fig. (c) Chirality of the A. algerae PT. Red arrows indicate the right-handed helix of the PT. (d) Heterogeneity of PT ends observed in A. algerae PTs (n = 20). Black arrows indicate the location of the PT end. (e) Transmission electron microscopy (TEM) section of an A. algerae spore showing interdigitation between vacuole membrane (red) and the PT, and position of the nuclei (N, cyan) relative to the vacuole. Inset shows the region boxed in orange.

https://doi.org/10.1371/journal.ppat.1008738.g002

Remarkably, in every spore examined the PT takes the form of a right-handed helix (Fig 2C). In the absence of any mechanism to bias PT assembly, the expectation would be a 50:50 mixture of left-handed and right-handed helices. This strong bias towards a right-handed helix implies the presence of an organizing principle. One explanation is that handedness could arise from the physical properties of a polymer, possibly formed by the polar tube proteins (PTPs). Five PTPs have been reported in microsporidia[19,47–49] and three of them (PTP1-3) localize along the entire length of the PT[48], likely acting as key structural elements. The presence of chirality in the PT is consistent with other biopolymers—such as DNA, actin filaments, and microtubules—which adopt superhelical coils with characteristic twist and handedness under strain. Indeed, recent cryo EM images and corresponding Fourier transforms of the PT show clearly repeating patterns along the tube[25]. Alternatively, or in addition, this right-handed bias may arise from asymmetry in the machinery involved in PT assembly.

We observed heterogeneity at the posterior end of the PT among spores, which can be divided into two main classes (Fig 2D). In the majority of spores, the PT terminates abruptly with a clean, blunt posterior end, remaining well-aligned with the preceding coils. However, in a few spores we observed irregular, tangled ends (Fig 2D, S3 Fig). The significance of blunt versus tangled PT ends is unclear, but the tangled ends may, for example, result from different developmental stages of the spores, or from abnormal development of the PT during spore formation[50]. It seems likely that these tangled ends pose a problem for PT extrusion, and if not resolved may result in incomplete germination and failed infection.

Configuration of organelles surrounding the PT in A. algerae

In addition to the PT, we segmented other organelles, including the anchoring disc, vacuole, and posterior polaroplast (Fig 1B and 1C, S1 Video). First, the anchoring disc forms an umbrella-shaped structure at the anterior tip of the spore, underlying the thinnest region of the endospore, as previously described[35,51] (Fig 1B and 1C). The anchoring disc is thought to be the site of spore germination, and thus some rearrangement or disruption of the anchoring disc is required to allow egress of the PT. Second, the vacuole at the posterior end of the spore occupies approximately 6.9% of the spore volume, and is roughly bowl-shaped (Fig 2A, S1B Fig). The vacuole has previously been shown to expand during the germination process, and one possibility is that it plays an active role in facilitating PT extension and the translocation of spore contents through the PT[34,52]. Our 2D TEM sections show that the vacuole is surrounded by a single membrane (Fig 2E). The convex side is in close proximity to the posterior pole, and the concave side faces towards the anterior end of the spore (Fig 2A and 2E). While there is no clear cellular structure interacting with the concave surface of the vacuole in our SBFSEM images, examination of 2D TEM images reveals that the nucleus often rests against this depression in the vacuole (Fig 2E). Interestingly, the vacuolar membrane is tightly interdigitated between PT coils (Fig 2E, S4A Fig). Third, the polaroplast is a multi-layered membranous organelle that is thought to be important during the germination process, perhaps providing membrane to accommodate the increased cell surface area during cargo transport through the PT[52]. Typically, the polaroplast consists of two regions with different morphologies, as observed by 2D TEM[30,33]: anterior polaroplast and posterior polaroplast (S4 Fig). Reconstruction of the posterior polaroplast showed that it snugly surrounds the linear part of the PT (Fig 2A), possibly stabilizing this region and is well-positioned to contribute membrane to the extending tube as it exits the spore.

Comparison of E. hellem and A. algerae 3D reconstructions

To assess whether the PT configuration and the relative orientation of other organelles are conserved in other microsporidian species, we carried out SBFSEM analysis of another human-infecting microsporidian species, Encephalitozoon hellem (Fig 3A, S1 Table). We segmented the spore wall layers, PT, anchoring disc, and anterior polaroplast (Fig 3B, S2 Video). While the overall organization of organelles in E. hellem spores is similar to the organization in A. algerae spores, including a right-handed helical arrangement of the PT (Fig 3C), there are also some notable differences. First, in contrast to the egg-shaped A. algerae spores, the E. hellem spores are more cylindrical (Fig 3C). The length of the E. hellem spores is 2.8 ± 0.3 μm along the A-P axis, and the volume is approximately half that of A. algerae (S1A and S1B Fig). Second, the spacing between PT coils is smaller in E. hellem (an average distance of 0.12 ± 0.03 μm, compared with 0.22 ± 0.02 μm for A. algerae (Fig 3D)), resulting in a more tightly packed coil. Third, in most E. hellem spores, the anchoring disc is located off-center with respect to the apical tip of the spore, rather than at the center of the apical tip as in A. algerae (Fig 3E). The region surrounding the anchoring disc is the thinnest part of the spore wall (Figs 1B and 3B), regardless of whether it is at the apical tip, or off-centered, and this region is where the PT is expected to exit the spore. Overall, the SBFSEM results provide insights into how the PT is packed inside the spore, and these data provide a static snapshot of the PT in its pre-germination state.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Serial block-face scanning electron microscopy imaging of E. hellem spores and comparison with A. algerae.

(a) A representative slice from SBFSEM imaging of E. hellem spores. (b) Representative SBFSEM slice highlighting segmented organelles. Original micrograph is shown (left), as well as the same image with color overlays indicating segmented organelles (right): exospore (orange), endospore (yellow), PT (blue), anterior polaroplast (pink), and anchoring disc (green). Magenta arrow indicates the thinnest part of the endospore layer where the anchoring disc is localized. (c) Representative 3D reconstruction of an E. hellem spore from SBFSEM data. Each color represents an individual organelle; color code as in (b). We did not observe any tangled PT ends in E. hellem spores. The average angle of E. hellem PT coils relative to the A-P axis is annotated with a standard deviation (n = 20, See also S1D Fig). The angle measurement was performed as in Fig 2B. (d) Quantification of the distance between the PT coils in A. algerae and E. hellem. Error bars represent standard deviation (n = 18 for A. algerae and n = 20 for E. hellem), ****p<0.0001 (unpaired Student’s t-test). (e) Position of the anchoring disc (AD, green) relative to the spore coat. Black line indicates the anterior-posterior (A-P) axis of the spore. Coincidence of the AD and the A-P axis was scored as “center”, indicating that the AD is at the apical tip of the spore. Separation of the AD from the A-P axis was scored as “off-center”, indicating that the AD is not centered at the apical tip of the spore (n = 18 for A. algerae and n = 20 for E. hellem).

https://doi.org/10.1371/journal.ppat.1008738.g003

Kinetics of polar tube germination

In order to infect the host, microsporidia must first extend the PT to penetrate a target cell. This process occurs extremely rapidly[24,25,34], making it challenging to image in real time. To better understand the kinetics and mechanics of PT germination, we performed high-speed, live-cell imaging to capture in vitro PT germination events in three microsporidian species that infect humans: A. algerae, E. hellem, and Encephalitozoon intestinalis. Although the in vivo triggers for PT firing are not well understood, in vitro triggers have been reported[53–56]. We used small variations of these conditions to optimize PT firing in vitro for our light microscopy assay (see Methods), and captured the entire germination process, including release of the cargo after transport through the PT (Fig 4A, S3–S5 Videos, S2 Table). In all three species, we observed three distinct phases of the germination process: 1) PT elongation, 2) a static phase, where the PT length does not change, and 3) emergence of cargo at the distal end of the PT. PT elongation occurs rapidly (Fig 4B, S5–S7 Figs); the time to reach 90% PT extension (T_EXT90) is significantly shorter in E. intestinalis (T_EXT90 160 ± 20 ms; Fig 4C) and E. hellem (T_EXT90 290 ± 200 ms) compared with A. algerae (T_EXT90 830 ± 170 ms). The entire germination process, from the start of PT extrusion to cargo ejection, is completed in less than 500 ms in E. hellem and E. intestinalis, and in approximately 1.6 seconds in A. algerae. We observed that the PT emerges from the center of the apical tip in A. algerae and off-center in E. hellem and E. intestinalis (S8 Fig). This is consistent with the preferential positioning of the anchoring disc in our SBFSEM data, at the center of the apical tip in A. algerae and off-center in E. hellem spores, supporting the idea that the anchoring disc position determines the PT exit site (Fig 3E).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Live-cell imaging of PT germination.

(a) Kymographs showing representative spore germination events from three microsporidian species: A. algerae, E. hellem, and E. intestinalis. Scale bars for distance and time intervals are shown on each kymograph. Three phases of the germination process are color-coded. The A. algerae PT appears to have a wavy pattern as it emerges from the spore, possibly forming a spiral, while the E. hellem and E. intestinalis PTs are less wavy in nature. As infectious cargo emerges from the distal end of the PT, the amplitude of the waveform is reduced, as has been previously noted[24]. After the A. algerae PT is fully extended, the distal end of the fired PT is curved and forms a hook-like structure, as previously described[25]. See also S3–S5 Videos. (b) Quantification of the PT length as a function of time for A. algerae (black), E. hellem (blue), and E. intestinalis (orange) (n = 20 for each species). The raw data are shown as faint lines with time points indicated as circles. The overall trends of each species were fitted (see Methods) and are represented as thick lines in the corresponding color. (c) Time taken for PT to extend to ≥ 90% (T_EXT90) of its maximum length (n = 20 for each species). ****p<0.0001, **p = 0.0096 (unpaired Student’s t-test). (d) Average maximum velocity of PT extension calculated as described in Methods (n = 20 for each species). **p = 0.006; ns, not significant (unpaired Student’s t-test). (e) Average maximum acceleration of PT extension (n = 20 for each species). ****p<0.0001; ns, not significant (unpaired Student’s t-test). (f) Shortening of the E. hellem PT after the full extension (maximum length) has been reached and cargo ejected. Representative phase-contrast micrographs are shown, and orange lines represent the length of the PT quantified from the micrographs. (g) Reduction of PT length measured 3 seconds after sporoplasm ejection, as a percentage of the maximum PT length (n = 20 for each species, see other representations of these data in S9A and S9B Fig). The PT shortens by 3%, 24%, and 5% for A. algerae, E. hellem, and E. intestinalis, respectively. ****p<0.0001 (One sample t-test). (h) Comparison of the PT length for E. hellem, in dormant spores, obtained from SBFSEM (labeled “inside”) and the maximum PT length after germination, obtained from live-cell optical microscopy experiments (labeled “outside”) (n = 20, see Methods for how these measurements were made). ****p<0.0001 (unpaired Student’s t-test). All error bars in this figure represent standard deviations.

https://doi.org/10.1371/journal.ppat.1008738.g004

In phase 1, the PT is fired and elongates to its maximum length (Fig 4A, S5–S7 Figs). Our data show that on average, the maximum velocity (V_max) of PT firing is 235 ± 61 μm⋅s^-1 and the maximum acceleration (A_max) is 1,503 ± 1,082 μm⋅s^-2 in A. algerae (Fig 4D and 4E). The V_max we observe is approximately double the previously reported value for this species[24], which may reflect differences in A. algerae spores purified from different sources or more accurate measurements due to higher temporal resolution. In E. hellem and E. intestinalis, V_max and A_max are higher than that measured for A. algerae. The V_max and A_max reaches 336 ± 142 μm⋅s^-1 and 5,219 ± 2,521 μm⋅s^-2 for E. hellem, and 290 ± 38 μm⋅s^-1 and 5,045 ± 2,242 μm⋅s^-2 for E. intestinalis, respectively (Fig 4D and 4E). In contrast, other cellular processes involving movement are typically slower, such as organelle transport in cells by motor proteins (~0.51 μm⋅s^-1 for human kinesin[57], ~1 μm⋅s^-1 for porcine dynein[58]), gliding motility of apicomplexan parasites (1–3 μm⋅s^-1)[59], and mobility of zoospores from chytrid fungi (~104 μm⋅s^-1)[60].

In phase 2, the PT is fully extended and remains static at its maximum length: on average, 101, 53 and 36 μm for A. algerae, E. hellem, and E. intestinalis respectively (S9A Fig). In A. algerae this phase persists for 790 ± 360 ms (S9C Fig), and we hypothesize that during this time, the infectious cargo is traveling through the PT. If cargo transport begins once the tube is fully extended and all the species transport cargo at the same rate, we would expect phase 2 in E. hellem and E. intestinalis to be approximately 400 ms, since their PTs are approximately half the length of the A. algerae PT. However, we observe E. hellem and E. intestinalis spend less time in this phase than predicted (60–140 ms; S9C Fig), suggesting that either cargo transport is faster in these species, or that cargo may begin moving through the PT before extension is completed. Intriguingly, for E. hellem, in which the source of the spores was the same for SBFSEM experiments and light microscopy (LM) experiments, we found that the fully extended PTs observed by LM are more than twice the length of pre-germination PTs packaged inside the spore, as assessed by SBFSEM (Fig 4H, S9D Fig, see Methods). A previous study has also suggested that the PT length outside the spore may be longer than when it is packed inside the spore[22]. The PT is thought to be built of repetitive protein polymers[25], and our observation raises the possibility that a significant conformational change may occur in some or all of these proteins between pre-germination and post-germination states, potentially leading to a change in PT length.

In phase 3, the cargo is expelled at the distal end of the PT, and appears approximately circular in shape (Fig 4A). The cargo remains attached to the tube for the duration of the experiment, and it is unclear what mediates this contact. There may be specific interactions between components of the PT and the cargo, or alternatively, some cargo may remain inside the tube itself, thereby creating a membranous bridge that leads to stable adhesion. When we continued imaging for several seconds after the cargo had been expelled, we observed that the E. hellem PT rapidly shortens (Fig 4F), and a similar observation has been reported for A. algerae[24]. To assess this observation quantitatively, we measured the length of the PT after germination is complete (full PT extension), and also 3 s later. The PT of E. hellem shortens by 24%, while the A. algerae and E. intestinalis PTs shorten by only 3% and 5% respectively (Fig 4G, S9A and S9B Fig, S6 Video). This suggests that there are differences in the mechanics, plasticity and behavior of the PT even between closely related species such as E. hellem and E. intestinalis. Shortening of the E. hellem PT appears to be distributed across the length of the tube, as opposed to shortening due to retraction of the tube back inside the spore coat (S6 Video).

Occasionally, we observed incomplete germination of the PT in all three species, which we define as being stuck in phase 1 or phase 2: the PT may not be fully extended, or it is extended but no cargo is observed at the distal end of the tube (S10A and S10B Fig, S7–S9 videos). In these events, the time to maximum extension of the PT was longer than in productive germination events (S10C Fig). Incomplete PT firing events have been previously described[24], and may represent spores that are not infectious, since cargo does not emerge at the distal end of the tube.

Cargo transport through the tube

An enigma of the microsporidian invasion process is that the cargo to be transported through the tube is much larger than the diameter of the tube itself. While it is unclear precisely what is transferred from the spore to the host cell, at the very least the nucleus containing the parasite genome must be transported through the tube[33,61]. In the case of A. algerae, there are two nuclei[23], and the diameter of each nucleus is about 0.7 μm, while the diameter of the PT is approximately 100 nm, as measured from negative stain images of the PT[25]. The nucleus is at least 7 times larger in diameter than the PT, and this mismatch in scale must be overcome to achieve cargo transport through the tube. As tools for genetic modification and transgenesis are not yet available for microsporidia, labeling specific subcellular structures for live-cell imaging is challenging, and consequently tracking the movement of cargo through the polar tube has not been reported.

To track nuclear movement within the spore body during germination, we used NucBlue to stain the nucleus of intact spores, which yields sufficient signal for live-cell imaging (Fig 5A). We used a modified version of our live-cell imaging assay, in which we pre-incubated dormant A. algerae spores with NucBlue, and tracked the nucleus during PT firing. We used dual detection of fluorescence and transmission to track the nucleus and to visualize the PT, respectively. Prior to nuclear exit from the spore, we clearly observed two distinct fluorescent lobes, corresponding to the two nuclei in A. algerae (Fig 5C, S10 Video). The two nuclei briefly move around and rearrange inside the spore, perhaps in response to the rapid ejection of the PT, then enter the PT together, on a very fast timescale. Interestingly, while T_EXT90 is 830 ms in A. algerae on average, nuclear translocation into the PT begins only ~500 ms after the initiation of PT firing, suggesting that the tube is not fully extended prior to cargo transport (Fig 5C, S11A Fig).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Live-cell imaging of nuclear transport through the PT.

(a) Image of a fixed germinated A. algerae spore stained with a nuclear dye, NucBlue, overlaid with a phase-contrast image of the same spore. (b) Schematic diagrams show two possible hypotheses of how large nuclei may travel through the narrow PT. Nucleus is depicted in blue; in “hypothesis 1” it is drawn as an oval, to scale with the PT (nucleus diameter = ~0.7 μm; PT diameter = ~100 nm). (c) Time-lapse images of the two A. algerae nuclei inside the spore during PT germination, with a time interval of 36 ms. Nuclei are pre-stained with NucBlue, and white light was applied in order to observe the PT firing event simultaneously. Red arrow indicates the frame in which PT firing is initiated; black arrow indicates the frame in which the nuclei begin to leave the spore body. Inset highlights the two nuclei, labeled with numbers and color overlays, from the frame boxed in orange. (d) Kymograph of nuclear translocation through the PT, with a time interval of 50 ms. (e) Quantification of the aspect ratio of the nuclei during transport (inside the tube) and after being expelled (outside the tube), which represents the extent of nuclear deformation during the process. Error bars represent standard deviation (n = 7), ****p<0.0001 (paired Student’s t-test).

https://doi.org/10.1371/journal.ppat.1008738.g005

Next, we shifted our attention to monitoring the nucleus as it moves through the PT. Given the discrepancy in diameter between the nucleus and the PT, we hypothesized that either 1) the PT is flexible enough that it could deform to accommodate the nuclei, or 2) the nuclei must deform to fit the PT diameter (Fig 5B). To assess which hypothesis may be correct, we performed live-cell imaging of A. algerae PT germination in vitro, and monitored the nuclei as they traversed through the PT. Our results show gross deformation of the nuclei on a millisecond timescale as they travel through the PT (Fig 5D, S11B Fig, S11 Video). However, after exiting the tube, the nuclei return to a globular shape (Fig 5D). To better quantify this nuclear deformation, we calculated the aspect ratio (the shortest dimension divided by the longest dimension) of the nuclei inside the PT and after exit from the PT (Fig 5E). The aspect ratio should be close to 1 for an isotropic object (e.g. circle), whereas an aspect ratio less than 1 indicates an elongated object (ellipse). The average aspect ratio of the paired nuclei is 0.11 ± 0.04 inside the tube and 0.87 ± 0.07 after they are expelled, outside the PT (Fig 5E). The nuclei travel through the tube at remarkable velocities, approaching 270 ± 115 μm⋅s^-1 (S11D Fig), which is comparable to PT firing velocities in A. algerae. Of note, the nuclei pause within the PT during translocation in all seven movies we recorded in vitro (Fig 5D, S11B and S11C Fig, S11 Video). This may reflect a change in the forces that are driving cargo transport through the PT, though it remains to be determined whether pausing also occurs during germination in vivo.

Mechanistic insights into polar tube germination

Microsporidia comprise an entire phylum of extremely successful parasites, all of which share the common feature of the polar tube infection apparatus. Our data combine spatial and temporal information to provide new insights about how the PT is packed in dormant spores, the PT germination process, and synchronization of PT firing and cargo transport. Synthesizing these data, we present a model for PT germination (Fig 6), in which we can clearly define some aspects of the germination process, while others still remain ambiguous. We focus on A. algerae, as we have the most information for this species. In the dormant spore, the PT is a right-handed helix packed at an angle relative to the A-P axis of the spore, and interacts closely with the vacuole. The two nuclei in A. algerae are nestled in the bowl-shaped vacuole, and surrounded by the coiled PT (Fig 6A). Previous data suggest that initiation of germination is dependent on osmotic pressure buildup within the spore[62], which may trigger the rupture of the spore wall and PT firing[63]. Once triggered, the PT fires and reaches its full length in under 1 second (T_EXT90 = ~830 ms). By tracking the nuclei during this germination process, we find that as the germination process initiates, the two nuclei begin to rearrange in the spore (Fig 6B, step 3). Approximately 500 ms after PT firing is initiated, the nuclei exit the spore body together, and deform drastically to fit into the PT. The average time to 50% extension (T_EXT50) in A. algerae is approximately 600 ms, suggesting that once the tube has reached ~50% extension, cargo transport is initiated. This observation is consistent with a model in which the PT everts[21,64], which would only allow cargo to enter the tube after 50% extension. The velocity at which the cargo is transported through the tube is comparable to the velocity of PT extension, and the cargo regains a globular shape after exiting the PT. Many open questions remain and further studies will be required to definitively address the mechanistic basis of PT germination.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. Model for PT germination and nuclear transport through the PT.

(a) Model for 3-dimensional organization of a microsporidian spore. Combining our SBFSEM and 2D TEM data for both A. algerae and E. hellem, we generate a composite 3D model that shows spatial organization of organelles inside the microsporidian spore. Elements that were not observed in the A. algerae SBFSEM data (anterior polaroplast and nuclei) have been added manually and are in gray. The anterior part of the PT (blue) is straight and connected to the anchoring disc (green). This part of the PT is surrounded by the anterior (gray) and posterior (purple) polaroplast. The posterior part of the PT is coiled and packed at an angle relative to the A-P axis. The PT resembles a rib cage around the vacuole (red) and nuclei (gray). The vacuole is bowl shaped, and localized beneath the nuclei. (b) Model for PT germination and cargo transport in A. algerae. (1) The spore is triggered in the presence of a stimulus (in vivo triggers remain largely unknown; in vitro triggers are described in Methods). (2) PT firing is initiated at the thinnest part of the spore coat, where the anchoring disc is localized. (3) After initiation of PT firing and during Phase I (PT elongation), the nuclei, and presumably other organelles, are reoriented. (4) At the stage in which the PT is extended to just past 50%, the nuclei deform to fit into the PT and exit the spore. (5) Nuclei (and likely other cargo) are translocated through the PT, at a speed comparable to that of PT extension. (6) The nuclei exit the PT and regain a circular shape at the tip of the PT.

https://doi.org/10.1371/journal.ppat.1008738.g006

Rapid nuclear deformation in microsporidia

In our study, we observed incredibly fast and large deformation of the nuclei as they traveled through the PT. This is reminiscent of other cell translocation and migration processes, such as immune cells squeezing through tight junctions to exit the bloodstream, or tumor cells penetrating tissues[65–67]. Nuclear deformation is necessary to facilitate tumor cell invasion, as the cells must adapt their shape to accommodate available space within the three-dimensional architecture of the tissue. Two key differences between the nuclear deformation we observe in microsporidia and analogous processes in tumor cell invasion is that the timescale is much faster in microsporidian nuclear transport (milliseconds compared with minutes/hours), and the microsporidian nuclear distortion is much more severe, as assessed by the aspect ratio. During tumor cell invasion and nuclear deformation, DNA damage is reported to occur[68,69]. Given the extent of microsporidian nuclear deformation during transport, it is plausible that the microsporidian genome may also be subject to shearing and other stresses. In migrating cancer cells, nuclear envelope breakdown has also been noted[68,69], which likely enhances deformability of the packaged DNA. Some cancer cells express lower levels of lamins[70–73], which are known to be important cytoskeletal proteins that rigidify the nuclear envelope. While cells expressing low levels of lamin A show a high migration rate through tissues[74], overexpression of lamin A has been shown to result in reduced nuclear deformability and impaired cell passage through narrow constrictions[75], consistent with the idea that the presence of lamins contributes substantially to rigidity of the nucleus. In agreement with previous reports that lamins are only present in Metazoans[76,77], we were unable to identify any lamins or homologs in fully sequenced microsporidian genomes[17,78,79]. This suggests that microsporidia may lack lamins entirely, perhaps explaining why nuclear deformation occurs to a much higher degree than in mammalian cells, and on a rapid timescale. Future studies will reveal the nature of cargo transport and address questions of the chromatin state during nuclear transport, whether the nuclear envelope is intact, and how cargo is modulated once delivered to the host.

Propagation of microsporidian parasites

A. algerae was propagated either in Helicoverpa zea larvae or in Vero cells. E. hellem and E. intestinalis were propagated in Vero cells. A. algerae spores propagated in H. zea were purified using a continuous Ludox gradient, as previously described[80], and kindly provided by James J. Becnel and Neil Sanscrainte. To propagate the parasites in Vero cells, microsporidia A. algerae (ATCC PRA-168), E. hellem (ATCC 50504), and E. intestinalis (ATCC 50506) were grown in Vero cells (ATCC CCL-81) using Eagle's Minimum Essential Medium (EMEM) (ATCC 30–2003) with 10% heat-inactivated fetal bovine serum (FBS) at 37°C and with 5% CO₂. At 70%-80% confluence, parasites were added into a 25 cm² tissue culture flask and the media was switched to EMEM supplemented with 3% FBS. Cells were allowed to grow for fourteen days and medium was changed every two days. To purify spores, the infected cells were detached from tissue culture flasks using a cell scraper and placed into a 15 ml conical tube, followed by centrifugation at 1,300 g for 10 min at 25°C. Cells were resuspended in sterile distilled water and mechanically disrupted using a G-27 needle. The released spores were purified using a Percoll gradient[79]. Equal volumes (5 mL) of spore suspension and 100% Percoll were added to a 15 mL conical tube, vortexed and then centrifuged at 1,800 g for 30 min at room temperature. The purified spore pellets were washed three times with sterile distilled water and stored at 4°C in distilled water for further analyses.

Germination conditions for microsporidian spores

Germination conditions, to activate PT firing, are different among microsporidian species[53,54,81]. For A. algerae, the germination buffer was slightly modified from the previously reported condition[53]. Spore germination was triggered by incubating spores at 30°C for 30 min in 10 mM Glycine-NaOH buffer pH 9.0 and 100 mM KCl. These conditions typically yield ~70% germination. For both Encephalitozoon species, spores were incubated at 37°C and spore germination was triggered using germination buffer containing 140 mM NaCl, 5 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, and 5% (v/v) H₂O₂ at pH 9.5, as previously described[54]. Germination buffer for Encephalitozoon species was freshly prepared for each experiment.

Electron microscopy

Purified spores from tissue culture were fixed in 0.1 M sodium cacodylate buffer (pH 7.2) containing 2.5% glutaraldehyde and 2% paraformaldehyde. For transmission electron microscopy, fixed spores were post-fixed with 2% osmium tetroxide (OsO₄), embedded in 2% agar, block stained in 1% aqueous uranyl acetate, and dehydrated using a gradient of cold ethanol. The samples were then transferred in propylene oxide and embedded in EMbed 812 (Electron Microscopy Sciences, catalog #14121). 70 nm thin sections were cut, mounted on copper grids and stained with uranyl acetate and lead citrate. Stained grids were inspected using a Thermo Fisher Talos 120C electron microscope and imaged with a Gatan OneView camera using a nominal magnification of 22,000x corresponding to a pixel size of 0.652 nm/pixel on the specimen.

For serial block-face scanning electron microscopy (SBFSEM), the sample block was mounted on an aluminum 3View pin and electrically grounded using silver conductive epoxy (Ted Pella, catalog #16014). The entire surface of the specimen was then sputter coated with a thin layer of gold/palladium and imaged using the Gatan OnPoint BSE detector in a Zeiss Gemini 300 VP FESEM equipped with a Gatan 3View automatic microtome. The system was set to cut 50 nm slices, imaged with gas injection setting at 40% (2.9 x 10⁻³ mBar) with Focus Charge Compensation to reduce electron accumulation charging artifacts. Images were recorded after each round of sectioning from the blockface using the SEM beam at 1.5 keV with a beam aperture size of 30 μm and a dwell time of 1.0 μs/pixel. Each frame is 22 x 22 μm with a pixel size of 2.2 nm. Data acquisition occurred automatically using Gatan Digital Micrograph (version 3.31) software. A stack of 150 slices was aligned and assembled using Fiji[82]. A total volume of 22 x 22 x 11 μm³ was obtained from the sample block.

Segmentation of organelles of interest, 3D reconstruction, and quantification of the spore size, volumes and PT length in the intact spores were performed using Dragonfly 4.1 software (Object Research Systems, ORS). SBFSEM sections were automatically aligned using SSD (sum of squared differences) method prior to segmentation. Graphic representation of the spores and PT were performed with either Dragonfly software or UCSF Chimera[83].

Measurement of the PT angles and the distance between coils were made in UCSF Chimera[83]. To facilitate this, two atoms were placed as markers on the anterior-posterior ends of the spore to create an A-P axis. 3 more atoms were placed along the PT coil to generate a plane corresponding to the PT. Then, the PT angles were measured between the A-P axis and the PT plane. To measure the distance between the coils, atoms were manually placed at the center of two adjacent coils, and the distance between these was measured.

Optical microscopy

Validation of PT length comparison between pre-germination and post-germination states

Our measurements of PT length inside the spore (pre-germination) are made from analyzing SBFSEM data, while measurements of the PT outside the spore (post-germination) are made using optical microscopy. To assess whether measurements from these methods are comparable, we measured the spore length for E. hellem from both these techniques. The spore length obtained from SBFSEM and LM are similar ( 2.8 ± 0.3 μm). These results validate comparing measurements between the two methods.

Statistical analyses

GraphPad Prism 8 software was used for all statistical analyses. In all analyses, we used a two-tailed unpaired Student’s t-test to compare the difference between two groups, with the exception of two analyses: 1) PT shortening after germination and 2) the nuclear aspect ratio. For PT shortening after germination, both a one sample t-test and a two-tailed paired Student’s t-test were used. For the nuclear aspect ratio, a two-tailed paired Student’s t-test was used. P values are reported in the figure legends.