The photophobic response in the apusomonad Podomonas kaiyoae is mediated by coordination of cilia and actin filaments

Abstract

Light responsiveness is a fundamental characteristic of eukaryotic organisms, both unicellular and multicellular. However, no photoresponsive behavior has previously been reported in Apusomonadida—a group of small, free-living biflagellates phylogenetically positioned as the sister group to Opisthokonta (comprising animals, fungi, and their unicellular relatives). As such, apusomonads are critical for understanding the evolutionary origins of opisthokonts. Here, we report an avoidance response to blue light in the apusomonad Podomonas kaiyoae. This response is characterized by an increase in gliding velocity, transient changes in flagellar waveforms, and alterations in cell shape. Dynamic cell contraction is triggered by an increase in intracellular free Ca²⁺ and is inhibited by either a dynein inhibitor or an actin-disrupting agent. These findings suggest that the photophobic behavior of Podomonas kaiyoae depends on cytoskeletal responses mediated by both the dynein/tubulin and myosin/actin systems, which likely originated early in eukaryotic evolution. The functional dominance of the posterior flagellum in cilia-driven directional changes further supports the phylogenetic placement of apusomonads as the sister group to opisthokonts, rather than to other eukaryotic lineages.

A light-avoidance response is described in apusomonads, key organisms for understanding the evolutionary origins of opisthokonts. This response involves Ca²⁺-dependent dynamic changes in both cell contraction and flagellar beating.

Introduction

In both aquatic and terrestrial environments, light is an important factor for many organisms in several physiological processes, including energy production, growth regulation, hormonal regulation, circadian system, and reproduction^1–3. Archaeplastida, including green algae, is one of the major eukaryotic supergroups Diaphoretickes (Fig. 1a). In the model organism for the study of photoresponses, the unicellular green algae Chlamydomonas, four patterns of motility change are observed: photophobic response, phototaxis, photokinesis, and surface adhesion/ cell gliding⁴. High light induces a photophobic response caused by a depolarizing signaling cascade. This cascade alters the waveforms of its two morphologically identical flagella, whereas the phototactic response involves changes in their beating balance. Recent studies suggest that the low-light induced positive phototaxis accompanies only a marginal depolarization via the channelrhodopsins⁴.

Fig. 1. Light-induced avoidance response of the apusomonad Podomonas kaiyoae.

a Schematic representation of the phylogenetic position of apusomonads based on published papers^7,20. b A schematic drawing of the cell structure of Podomonas kaiyoae. c Sequential images show P. kaiyoae cells avoiding an illuminated field. The images capture the moments when the blue or green light is turned on (ON) and off (OFF). The scale bar represents 100 µm. The photon flux density of the blue light used was 48.04 µmol/m²/s. d Effect of various light colors on the decrease in intensity within the illuminated field. ‘Control’ represents the cell density under red LED. The alphabets and numbers represent the color and the output % for illumination. Corresponding photon flux densities and irradiance are shown in the Supplementary Table 1. Error bars represent standard deviation. Data was collected from 7 (blue) or 3 (other colors) separate experiments. e Effect of various intensities of green and blue lights on the decrease in intensity within the illuminated field. The alphabets and numbers represent the color and the output % for illumination. Corresponding photon flux densities and irradiance are shown in the Supplementary Table 2. Error bars represent standard deviation from 10 separate experiments. f Behavior of a cell gliding showing a clear step-up response. Top 5 or bottom 5 panels show the trajectories (blue lines) of the cells gliding from a bright to a dark field or a dark to a bright area across the blue light border, respectively. The images were recorded from the initial positions (shown with red spots) for 34–100 s. The long or short side of the rectangle in the image measures 640 or 480 µm, respectively. The photon flux density of the blue light used was 48.04 µmol/m²/s.

Amorphea is a major eukaryotic supergroup comprising Amoebozoa and Obazoa (Fig. 1a)^5–7. Within Obazoa, the group Opisthokonta includes fungi and animals. Animals have evolved complex multicellular visual organs that are integrated with the central nervous system. These visual systems detect environmental light through specialized sensory structures enriched in type-II rhodopsin photoreceptors, which convert light signals into motor responses via the neural circuitry of the motor system⁸. Early diverging fungal representatives, such as chytrids, consist of several phyla of unicellular fungi that produce posteriorly flagellated motile spores. Some chytrid species show clear positive phototaxis^9–11. Light reception in these organisms is mediated by an eyespot-like, lipid-rich structure known as the side-body complex or the microbody-lipid globule complex, which regulates flagellar beating^10,12. Phototaxis in amoebozoans has been observed in the ameboid cells and plasmodia of Dictyostelium and Physarum^13,14, but has not yet been demonstrated in the flagellated cells of Physarum. These observations suggest that light-induced changes in cell motility in early eukaryotes likely involved coordination between ciliary or flagellar movement and dynamic changes in cell shape mediated by microtubule- or actin-filament-based motility^8,15.

Intracellular calcium ion (Ca²⁺) is a critical regulator of cytoskeletal dynamics and motility in eukaryotes, playing essential roles in both ciliary and flagellar beating^16,17, as well as in microfilament contraction^18,19. Photoreception mechanisms are believed to have originated early in eukaryotic evolution. Excavata is a diverse assemblage of early-branching unicellular eukaryotes that includes lineages such as Discoba and Metamonada (Fig. 1a)^7,20. Euglena, a member of Discoba, exhibits a distinctive mode of locomotion known as euglenoid movement^21,22. While this movement typically occurs in the absence of light, it is stimulated by light exposure. In Euglena gracilis, both Ca²⁺-dependent changes in flagellar beat direction and pronounced cellular deformation appear to be associated with light response²³. A recent study employing high-speed imaging and biophysical modeling demonstrated that phototaxis in E. gracilis can be explained by coordinated switching between two distinct flagellar beat patterns²⁴. In contrast, E. mutabilis, which lacks an emergent flagellum, primarily exhibits gliding motility during phototaxis²⁵.

Apusomonads are heterotrophic flagellates found in both marine and terrestrial environments²⁶. Recent phylogenomic analyses suggest that apusomonads form the sister group to opisthokonts, and together with Breviata, they comprise the clade Obazoa (Fig. 1a)^5,7. Apusomonad cells exhibit a distinctive folded membrane structure known as the ‘skirt’, which extends along both the dorsal and ventral surfaces (Fig. 1b). The anterior region of the skirt forms a sleeve-like extension called the ‘proboscis’, which surrounds the anterior flagellum. This structure differs markedly from the flagellar pocket of Trypanosoma, a small invaginated region of the plasma membrane at the flagellar base²⁶. An electron-dense submembranous structure, termed the ‘pellicle’, lies beneath the dorsal cell membrane. The posterior flagellum is situated within a folded portion of the skirt known as the ‘lips’, located on the left ventral side of the cell. Both the anterior and posterior flagella possess slender terminal extensions called ‘acronemes’ (Fig. 1b).

Apusomonads locomote mainly by gliding on surfaces which mostly depends on movement by the posterior flagellum together with the filopodia. The beating of its anterior flagellum does not appear to be involved in the movement^27–29. Although the phylogenetic position and motility characteristics of apusomonads suggest ancestral eukaryotic features^29,30, no phototactic behavior or photoreceptive structures have been described in these organisms. This lack of information represents a significant gap in our understanding of the cytoskeletal roles in photoresponse during eukaryotic evolution, particularly in a group that links Amoebozoa and Obazoa⁸. In this study, we report, to the best of our knowledge, the first observation of a photoresponse in the apusomonad Podomonas kaiyoae. We investigate how two major motility systems, i.e. actin-myosin and tubulin-dynein, cooperate to mediate this response in a representative of one of the earliest-branching eukaryotic lineages.

Results

Photophobic response of the apusomonad Podomonas kaiyoae

To determine if apusomonads display any light-responsive behavior, Podomonas kaiyoae cells were observed under red LED light, followed by exposure to blue light at the center of the field. Stroboscopic illumination with a short pulse red light was used to clearly visualize the flagellar and ciliary waveforms and their propagations³¹. The behavior of the cells was recorded under a phase contrast microscope. Upon blue-light illumination, the distribution of P. kaiyoae cells decreased gradually within the illuminated area, creating a distinct empty circle in the center after 100 s (Fig. 1c, Supplementary Video 1). Most cells in the illuminated area exhibited a photophobic response characterized by cell contraction and extension. When the blue light was turned off, the cells gradually reoccupied the previously empty area, indicating that the reduction in cell numbers within the center field was due to an avoidance response. The avoidance response was most effectively triggered by blue light (440 nm) and was also induced, to a lesser extent, by cyan light (470 nm) (Fig. 1d). In contrast, other wavelengths, including green light and additional illumination with red light in the same area, did not induce a photoavoidance response (Fig. 1d, Supplementary Video 2, Supplementary Table 1).

To rule out potential side effects associated with short-wavelength light, such as heat generation or the production of reactive oxygen species³², we examined the behavior of cells exposed to blue light at varying irradiance levels³³. Blue light at just 8% of the maximum power output (irradiance: 3.0 J /m²/s) caused a substantial reduction in cell density within the illuminated area (Fig. 1e, Supplementary Table 2). In contrast, green light at 40% of the maximum power output (11.1 J/m²/s) had no effect on cell density (Fig. 1d, e, Supplementary Table 2). These results indicate that the avoidance response is specific to blue light and is not attributable to nonspecific energy-dependent effects.

Cells passed through the boundary between bright and dark areas as they escaped from the illuminated region (Fig. 1f, top; Supplementary Video 3). However, when re-entering the illuminated area, cells often stopped gliding at the edge of the light and changed direction away from the illuminated region (Fig. 1f, bottom; Supplementary Video 4). These findings suggest that the response to light is a step-up photophobic response. A typical avoidance response pattern under bright field microscopy included a sequential process: cessation of gliding, a~90-degree turn, and resumption of gliding until the cell crossed the bright-dark boundary.

In Chlamydomonas, the direction of phototaxis is known to switch in response to various factors, including light intensity⁴. In apusomonads, it remained unclear whether cells were initially attracted to the bright–dark boundary—where light leakage may occur—prior to exhibiting an avoidance response under strong blue light (Fig. 1f, bottom). To assess the possibility of positive phototaxis in P. kaiyoae, we illuminated the cells with faint blue light (0.78 µmol photons/m²/s) from one side of a chamber mounted on a glass slide and recorded their behavior. Despite this low-intensity illumination, no clear movement toward the light source was observed, suggesting an absence of positive phototaxis (Supplementary Fig. 1; Supplementary Video 5).

Cell contraction and changes in gliding direction are necessary for photophobic response

Two types of P. kaiyoae cells were observed in terms of motility. Before illumination under red light, many cells were settled on the glass surface by anchoring their posterior filopodia. They showed no progressive gliding but occasionally exhibited contraction-extension movements (termed ‘settled’). Only a small population of cells displayed progressive gliding on the glass surface (termed ‘gliding’) with occasional changes in direction due to cell contraction-extension, as previously reported in P. kaiyoae and other apusomonads^27,28,34. Upon blue light illumination, cells underwent a drastic transition from the settled to the gliding state, with many cells detaching from their anchorage via posterior filopodia and initiating forward gliding motility (Fig. 2a, top; Supplementary Video 6). During blue light illumination, gliding cells frequently changed direction with cell contraction (Fig. 2a, bottom; Supplementary Video 7).

Fig. 2. Light-induced activation of gliding and cell contraction in apusomonads.

a Both settled cells (top) and gliding cells (bottom) on a glass surface contract in response to blue light illumination. Arrows indicate the direction of gliding. The scale bar represents 10 µm. The photon flux density of the blue light used was 48.04 µmol/m²/s. b Rose plots depict the angle displacement in gliding direction post-contraction under blue light illumination. The left plot represents cells that settled and started gliding after contraction, with the angle displacement measured between the anterior-posterior axis before contraction and the gliding direction post-contraction (N = 37 from 7 separate experiments). The right plot represents cells that were already gliding and altered their direction post-contraction (N = 25 from 7 separate experiments). Circular standard deviation (CSD) is noted. Concentric circles represent the scale for percentage (20 and 25% maximum for settled and gliding, respectively. The photon flux density of the blue light used was 48.04 µmol/m²/s. Box plots illustrate gliding velocity (c) and contraction rate (d) under red (8 separate experiments) or blue (7 separate experiments) light. N = 52 (red) and 41 (blue) in (c); 8 (red) and 7 (blue) in (d). ** Significant at p < 0.01, * p < 0.05 (Student’s t test) when compared to red light. The photon flux density of the blue light used was 18.5 µmol/m²/s.

To test if blue-light-dependent cell contraction and the initiation of gliding are correlated with directional blue-light avoidance movement, changes in cell orientation before and after blue light illumination were measured. The rose plots show that settled cells began gliding in various directions relative to their orientation before illumination (Fig. 2b, left). However, in gliding cells, the direction became limited to 60-90° relative to their pre-illumination orientation (Fig. 2b, right). The circular standard deviation demonstrated that the gliding direction was random in settled cells but significantly oriented in gliding cells (Fig. 2b). Blue light illumination activated the gliding movement of P. kaiyoae cells, increasing their gliding velocity to approximately 1.2 times the pre-illumination speed (Fig. 2c). The frequency of contraction also increased to about 1.7 times the total cell population after blue light illumination (Fig. 2d).

The posterior flagellum is linked to cell contraction

The photophobic response in P. kaiyoae involves changes in cell shape, including contraction and subsequent extension for re-gliding. To understand the cellular mechanism underlying blue-light-induced contraction and gliding, the distributions of microtubules and actin filaments were examined. The anti-acetylated α-tubulin antibody is known to label long-lived, stable microtubules in protists³⁵. Immunofluorescence using this antibody revealed that stable microtubules are localized in both the anterior and posterior flagella, as well as in the dorsal and ventral root structures (Fig. 3a, b; Supplementary Fig. 2).

Fig. 3. Roles of cytoskeletal structures in the avoidance response of apusomonads.

a Immunofluorescent microscopy showing stable microtubule and microfilament structures in Podomonas kaiyoae. Cytoskeletal images stained with anti-acetylated tubulin (AcTu, green; left); stained with phalloidin (magenta; middle); merged image (right). The scale bar represents 5 µm. b A schematic drawing of the cytoskeletal structures in P. kaiyoae. Key structures include the anterior flagellum, posterior flagellum, dorsal and ventral root structures, anterior microfilament, trailing pseudopodium, acroneme, and posterior filopodium. c Thin-section electron microscopy of P. kaiyoae, showing the posterior flagellum ventrally positioned along the cell body within a fold called the skirt, formed by two lips. Anterior flagellum (AF); posterior flagellum (PF); dorsal root (dr); ventral root (vr); nucleus (n); pellicle (pe); plasma membrane (pm). The scale bar represents 1 µm. The areas shown in light blue represents root microtubular structures. d Two morphological states of P. kaiyoae are depicted: extension (top) and contraction (bottom). The anterior and posterior flagella are marked with dashed lines in green and magenta, respectively. The scale bar represents 10 µm. e Flagellar curvature of the anterior (AF) and posterior (PF) flagella in the extension (Ext) and contracted (Con) states. N = 7 from separate 3 experiments. Statistical analysis was performed using Dunnett’s test, with AF-Ext and PF-Ext as control groups. A significant difference was observed in PF-Con compared to AF-Ext and to PF-Ext (*** p < 0.001). f Changes in the maximum curvature of the anterior and posterior flagella, measured 2.5 µm from the flagellar base during cell contraction. Cell size is expressed as the distance between the anterior and posterior ends of the cell body. The photon flux density of the blue light used was 48.04 µmol/m²/s. g–j Effects of cytoskeletal or motor drugs on the photoresponse-associated motility of P. kaiyoae. Data were collected under blue light from separate 4 experiments. Cell batches used for experiments with ciliobrevin, cytochalasin, and nocodazole were different from those used for para-aminoblebbistatin. g Box plots show the relative cell density. N = 3 for ciliobrevin (cilio), cytochalasin (cyto), nocodazole (noco), and their control (DMSO); N = 5 or 10 for para-aminoblebbistatin (PAB) or its control (DMSO), respectively. The photon flux density of the blue light used was 48.04 µmol/m²/s. h Box plots show the contraction rate per cell. N = 4 for ciliobrevin, cytochalasin, nocodazole, and their control (DMSO); N = 5 or 9 for para-aminoblebbistatin or its control (DMSO), respectively. The photon flux density of the blue light used was 48.04 µmol/m²/s. i Box plots show the gliding velocity 5 s after blue light illumination. N = 12 (DMSO),15 (ciliobrevin),18 (cytochalasin), and 26 (nocodazole). For para-aminoblebbistatin treatment, N = 5 (para-aminoblebbistatin) and N = 9 (DMSO). The photon flux density of the blue light used was 18.5 µmol/m²/s. j Box plots show the beat frequency of the anterior flagellum. N = 12 (DMSO),15 (ciliobrevin),18 (cytochalasin), and 26 (nocodazole). For para-aminoblebbistatin treatment, N = 5 (para-aminoblebbistatin) and N = 9 (DMSO). The photon flux density of the blue light used was 18.5 µmol/m²/s. *** Significant at p < 0.001, ** p < 0.01, * p < 0.05 (Dunnett’s test) when compared with DMSO.

The anterior flagellum exhibited continuous three-dimensional beating (Supplementary Videos 6 and 7). Staining with Alexa Fluor–labeled phalloidin revealed actin filaments in both the anterior cytoplasm and the trailing pseudopodia, consistent with previous observations^28,29. The trailing pseudopodia extended several filamentous projections that adhered to the glass surface. Based on their strong actin staining and structural characteristics, we refer to these projections as ‘filopodia’ in this study. While the number and length of filopodia varied between cells, they were consistently extended when the cells settled in the dark prior to blue-light exposure (Supplementary Fig. 2).

Thin-section electron microscopy revealed that the anterior flagellum protrudes from the anterior end of the cell, whereas the posterior flagellum is positioned ventrally along the cell body and enclosed by the lips of the skirt (Fig. 3c; Supplementary Fig. 3a–c). Both the dorsal and ventral surfaces feature rows of cortical singlet microtubules—referred to as ‘roots’—which extend from the basal bodies located at the anterior end (Fig. 3c, left; Supplementary Fig. 3d). Notably, no distinct projections such as dynein arms were observed on these singlet microtubules (Supplementary Fig. 3e). P. kaiyoae displays several morphological differences from Thecamonas trahens, including the presence of a line of refractile granules, a reduced sleeve, and an acroneme at the tip of the anterior flagellum^27,28,34. Additionally, in P. kaiyoae, the ventral root microtubules extend across the surface of the lateral pseudopodium— a feature absent in T. trahens (Fig. 3c)²⁷. While the ventral roots are divided into left and right branches in other Podomonas species²⁷, they form a continuous, sheet-like structure within the lateral pseudopodium in P. kaiyoae (Fig. 3c, Supplementary Fig. 3d, e).

The posterior flagellum lay through the center of the cell in the extension state, with the portion protruding from the cell body slightly bent in both contraction and extension states (Fig. 3d, top; Supplementary Video 8). During the contraction state, the flagellum bent significantly towards the ventral side of the cell (Fig. 3d, bottom; Supplementary Video 8). In some cases, the posterior flagellum bent significantly and extended out from the cell body, indicating that the bending of the posterior flagellum was linked to contraction. To investigate the relationship between flagellar motility and cell contraction, we quantified the curvature of the anterior and posterior flagella during contraction events. A significant increase in curvature was observed in the posterior flagellum during contraction, whereas the anterior flagellum showed no such massive increase (Fig. 3e). During the contraction-extension cycle, the anterior flagellum continued to beat rhythmically, displaying oscillatory curvature. In contrast, the posterior flagellum remained relatively straight during the extension phase, with curvature increasing transiently and in parallel with cell contraction (Fig. 3f).

Cooperation of two cytoskeletal systems in contraction-extension transition

To investigate the roles of cytoskeleton-mediated motility in the contraction–extension response of P. kaiyoae, we examined the effects of pharmacological agents targeting cytoskeletal components and their associated motor proteins (Fig. 3g–j). Treatment with the microtubule polymerization inhibitor nocodazole (12.5 µM for 15 min) slightly enhanced the photoavoidance response (Fig. 3g, Supplementary Video 9). However, it did not significantly affect the contraction rate (Fig. 3h, Supplementary Video 9), gliding velocity (Fig. 3i), or anterior flagellar beat frequency (Fig. 3j).

In contrast, treatment with cytochalasin B, an inhibitor of actin polymerization (200 µM for 15 min), markedly suppressed the photoavoidance response (Fig. 3g) and significantly reduced the anterior flagellar beat frequency (Fig. 3j; Supplementary Video 10). Cytochalasin B also led to a significant decrease in both the contraction rate and gliding velocity (Fig. 3h, i).

Inhibitors targeting molecular motors were also tested. Ciliobrevin A, a dynein inhibitor³⁶, is known to inhibit both axonemal^37,38 and cytoplasmic dyneins^39,40. Treatment with ciliobrevin A at 40 µM significantly inhibited both the photoavoidance response and the contraction rate (Fig. 3g, h; Supplementary Video 11), resembling the effects observed with cytochalasin B. As expected, ciliobrevin A treatment also caused a substantial reduction in flagellar beat frequency (Fig. 3j) and significantly decreased gliding velocity (Fig. 3i).

In contrast, treatment with para-aminoblebbistatin (10 µM for 15 min), a non-phototoxic, selective inhibitor of myosin II⁴¹, resulted in a significant or detectable increase in the photoavoidance response or the rate of cell contraction, respectively, compared to the control (Fig. 3g, h, Supplementary Video 12).

Under red light, many cells settled on the glass surface using their posterior filopodia. This settling behavior persisted in some cells even immediately after blue light illumination (Fig. 2a, top; Supplementary Video 6). Upon blue light exposure, the posterior filopodia of the cells detached and were drawn along by the cell body during gliding (Supplementary Video 6 and Video 7). The filopodia were observed in cells treated with nocodazole. However, in cells treated with cytochalasin or ciliobrevin, the filopodia retracted into the cell body. Furthermore, these cells exhibited reduced adhesion to the glass surface, with some cells flipping over (Supplementary Video 13 and Video 14). Cell flipping occurred more frequently in ciliobrevin-treated cells (35.7 ± 20.9%, N = 232 cells, 5 experiments) than those in cytochalasin-treated (12.6 ± 1.8%, N = 190 cells, 4 experiments). In the latter case, the posterior flagellum often remained attached to the glass slide, preventing flipping (Supplementary Video 13).

Ca²⁺ is essential for both progressive gliding and cell contraction

Calcium ions are crucial signaling molecules for regulating cellular phototaxis and chemotaxis^16,42,43. To investigate the regulatory mechanisms underlying the photophobic response in P. kaiyoae, we examined the effects of Ca²⁺ on forward gliding movement and cell contraction. Cells were transferred into Ca²⁺-free (EGTA-chelated) artificial seawater (ASW) medium. Under Ca²⁺-free condition, cells exhibited a dramatic loss of forward motility, cell contraction, and gliding (Fig. 4a–d). Instead, P. kaiyoae cells under Ca²⁺-free condition displayed a pivot-like movement, similar to the motility patterns observed in ciliobrevin- or cytochalasin-treated cells (Supplementary Video 15). Both anterior and posterior flagella appeared to be beating, but the frequency was greatly reduced (Supplementary Video 16). Forward gliding motility and cell contraction were restored when the cells were transferred back to ASW containing Ca²⁺ (Fig. 4a–d).

Fig. 4. Roles of Ca²⁺ in the regulation of the avoidance response in apusomonads.

a Cells were first observed in Ca²⁺-free SW (CFSW). Then the solution was substituted back to ASW. Cell trajectories recorded for 100 s. The long or short side of the rectangle in the image measures 640 or 480 µm, respectively. The red spot indicates the initial position. The photon flux density of the blue light used was 48.04 µmol/m²/s. b–d Ca²⁺ is crucial for cell contraction and gliding. Relative cell density (b), contraction rate per cell (c) and gliding velocity 5 s post-illumination (d) are shown. These variables are plotted as box plots, comparing conditions in CFSW versus standard ASW. ***Significant at p < 0.001 (Student’s t test) when compared with the value in ASW. b, c N = 5 from separate 3 experiments. The photon flux density of the blue light used was 48.04 µmol/m²/s. d N = 11 from separate 3 experiments. The photon flux density of the blue light used was 18.5 µmol/m²/s. e–g Confocal images showing changes in intracellular Ca²⁺ concentration during cell contraction. e Fluo-8H fluorescence is displayed using a pseudocolor lookup table (LUT). Arrows and an arrowhead indicate the direction of gliding and a site of cell contractions, respectively. The anterior (A) and posterior (P) ends of the cell are indicated. Scale bar, 5 µm. f Changes in maximum fluorescence intensity during cell contraction. An increase in [Ca²⁺]_i (red bar) is observed during contraction (green line). g Ratio of [Ca²⁺]_i increases during extension (Ext) and cell contraction (Cont) in Fluo-8H loaded cells. The ratio (F/F_min) represents the relative fluorescence intensity normalized to the minimum intensity within a region of interest (ROI; 29 × 29 pix). * Significant at p < 0.05 (Student’s t test) when compared with the Control. N = 6 from separate 3 experiments.

To visualize Ca²⁺ dynamics during the photophobic response, we introduced a Ca²⁺-selective fluorescent indicator, Fluo-8H, into P. kaiyoae cells. The membrane-permeable acetoxymethyl ester form of Fluo-8H was loaded into the cells in the presence of the non-ionic detergent Cremophor. Fluorescence emission was measured under cyan light (470 nm) excitation, which simultaneously induced the photoresponse. Although the cells exhibited autofluorescence, primarily derived from phagocytosed bacteria within the cytoplasm, no clear increase in the fluorescence was observed during cell contraction (Supplementary Fig. 4a). In Fluo-8H loaded cells, we observed that the fluorescence intensity increased a few seconds after cyan light irradiation (Supplementary Fig. 4b, c), coinciding with cell contraction (Supplementary Fig. 4d). The fluorescence increase was spatially localized rather than uniform with a specific middle region displaying enhanced intensity. Upon resumption of forward gliding, the fluorescence returned to baseline levels.

To rule out the possibility that changes in cell thickness influenced the fluorescence signal, we carried out the imaging of fluorescence change using confocal fluorescence microscopy on optical sections (Fig. 4e, Supplementary Video 17). A large increase in fluorescence was observed with cell contraction around an anterior ventral area (Fig. 4f). In control cells without the fluorescent dye, the ratio of maximum fluorescence intensity between extension and contraction states was 1.18 ± 0.16 (n = 6), whereas in Fluo-8H-loaded cells, the ratio significantly increased to 1.66 ± 0.52 (n = 6) (Fig. 4g).

Discussion

We describe for the first time a clear photoresponse in an apusomonad species. P. kaiyoae was originally isolated from a deep-sea sediment³⁴. It is known that bioluminescence plays a variety of biological roles in the communication among deep-sea organisms⁴⁴. P. kaiyoae, like other apusomonads, is a heterotrophic protist that shows no photosynthesis but feed on bacteria and other small protists³⁴. The light response of P. kaiyoae is unlikely to serve a role in feeding on luminescent bacteria, as our data show that P. kaiyoae does not exhibit positive phototaxis in response to very low-intensity blue light (Supplementary Fig. 1d). Alternatively, the response may function to help the organism evade faint light that reaches the surface of deep-sea sediments, thereby enhancing concealment. It is also conceivable that P. kaiyoae inhabits not only deep-sea sediments but also shallow waters. In such environments, the photoavoidance response could aid in orientation within the sediment or serve as protection against near-UV or blue light exposure.

While the activation of forward gliding by blue light may induce a decrease in the population of P. kaiyoae cells in the illuminated area, suggesting photokinesis, analysis of gliding movement across the bright-dark boundary clearly indicates a photoavoidance behavior (Fig. 1f, bottom). Moreover, the directional change in gliding was more pronounced when cells moved from dark to bright areas, but less evident when moving from bright to dark areas, suggesting a step-up response in the photophobic behavior (Fig. 1f).

Photoreception is a fundamental characteristic in both prokaryotes and eukaryotes, with photoreceptor proteins being the primary molecular players involved⁴⁵. Photoreceptor proteins, including proteorhodopsin (type I) and various vertebrate (type II) opsins, are widely distributed across organisms. Unicellular eukaryotes are also known to possess several types of photoreceptor proteins and associated cellular structures⁸. For instance, zoospores of the chytrid fungus Blastocladiella emersonii, a close relative of apusomonads within the Opisthokonta, exhibit phototactic swimming behavior in response to green light, mediated by a unique fusion protein containing a type I rhodopsin^12,46,47. In B. emersonii, a cyclic nucleotide-gated channel, activated by rhodopsin-mediated cGMP production, regulates flagellar beating¹². In contrast, our findings show that blue light is the most effective wavelength for inducing photoavoidance in P. kaiyoae, suggesting the involvement of blue-light photoreceptors. However, we did not identify any specific photoreceptors or light-sensing subcellular structures in the present study. Future research is essential to identify the photoreceptor(s) in P. kaiyoae and to elucidate the molecular pathway linking blue-light perception to downstream cytoskeletal signaling.

Early branching eukaryotic lineages underwent significant evolutionary events, including the acquisition of novel cytoskeletal motility systems^5,48–50. The photoresponse in P. kaiyoae consists of two steps: activation of forward gliding and cell contraction, which induces a change in gliding direction. The microtubule-polymerizing inhibitor nocodazole did not suppress gliding or cell contraction, whereas the dynein inhibitor ciliobrevin blocked both types of motility. A similar effect of a dynein inhibitor has been observed in trypanosomes^51–53. These findings suggest that the photophobic response is mediated by dynein-driven motility rather than by the dynamics of singlet microtubule structures, such as root microtubules.

Dyneins are classified into axonemal and cytoplasmic types, both of which are inhibited by ciliobrevin⁵⁴. The conversion of the posterior flagellum beat into large asymmetric waveforms suggests that axonemal dyneins play a central role in mediating directional changes in motility. Indeed, the asymmetric waveform of the posterior flagellum is closely coupled with cell contraction and photophobic behavior (Fig. 3d–f). However, it remains possible that cytoplasmic dyneins, such as intraflagellar transport (IFT) dyneins, also contribute to the response. In Chlamydomonas, for example, IFT drives flagellar surface-based gliding^39,55, a mechanism similarly observed in sea urchin larval cilia⁵⁶ and other ciliated eukaryotes⁵⁷. Therefore, the posterior flagellum in P. kaiyoae may contribute not only to cell contraction via large asymmetric bending, but also to gliding along the glass surface through IFT-based motility. In Chlamydomonas, elevated intracellular Ca²⁺ disrupts IFT-dependent cargo transport, which is responsible for generating the force that drives surface gliding^58,59. Two additional points are noteworthy: First, gliding in Chlamydomonas primarily relies on retrograde IFT driven by dynein, whereas in P. kaiyoae and many other protists, gliding may occur via anterograde transport mediated by kinesin-2 rather than dynein⁶⁰. Second, flagellar adhesion in Chlamydomonas is light-inducible⁶¹; while blue light enhances adhesion in Chlamydomonas, P. kaiyoae exhibits the opposite response. Given the low-light conditions characteristic of benthic environments, light-modulated flagellar adhesion may influence the habitat preference of P. kaiyoae, potentially contributing to biofilm formation.

Actin dynamics also appear to be essential for both cell gliding and contraction, as treatment with cytochalasin effectively inhibited both types of motility. Retraction of posterior filopodia was observed in cells treated with both cytochalasin and ciliobrevin, suggesting that these filopodia may function as anchoring structures during cell contraction and directional turning. The anterior actin-rich pseudopods function as the leading edge in the ameboid excavate Naegleria⁶², unicellular slime mold^63,64, and the zoospore of chytrid fungi⁶⁵. Thus, it is possible that the dynamics of the anterior actin network contribute to gliding. Myosin may be involved not in generating the gliding force but rather in anchoring the posterior filopodia. Given that para-aminoblebbistatin significantly increases both the avoidance response and the cell contraction (Fig. 3g, h), myosin likely contributes to maintain the extension state for gliding. Taken together, these observations suggest that P. kaiyoae utilizes a cooperative motility system involving two ancestral mechanisms: ciliary surface motility and actin-based migration³⁰. This dual system may reflect the motility strategies employed by early opisthokont lineages, as supported by evidence of cytoskeleton-based motility in chytrid fungi^65,66 and choanoflagellates^67,68.

In the present study, we were unable to directly demonstrate Ca²⁺-dependence in the conversion of posterior flagellar waveforms to an asymmetric pattern. However, the posterior flagellar waveform in P. kaiyoae becomes highly asymmetric during the contraction state (Fig. 3d bottom; Fig. 3e), a condition in which the intracellular Ca²⁺ levels are presumed to be elevated (Fig. 4e–g). In contrast, the posterior flagellum remains relatively straight during the extension state (Fig. 3d, top; Fig. 3e), which corresponds to a basal level of intracellular Ca²⁺ (Fig. 4f). These changes in flagellar waveform symmetry are consistent with Ca²⁺-dependent responses observed in opisthokonts, including metazoans and fungi¹⁶. P. kaiyoae exhibits a large bend in the posterior flagellum at high Ca²⁺ concentrations, whereas Trypanosoma reverses the direction of wave propagation in the anterior flagellum to a tip-to-base pattern under high Ca²⁺ concentrations, with an asymmetric form⁶⁹. Animal sperm propagate asymmetric waves in the posterior flagellum at high intracellular Ca²⁺ concentrations¹⁶. A similar Ca²⁺-dependent asymmetric movement of the posterior flagellum was observed in the zoospore of the fungus Blastocladiella emersonii⁷⁰, although the conversion of waveform symmetry was extremely slow. Our findings suggest that a large asymmetric bending of the posterior flagellum is induced by an increase in intracellular Ca²⁺ concentration. Genomic surveys of eukaryotes suggest that apusomonads have already acquired a set of molecules necessary for Ca²⁺ response, such as the sperm-specific Ca²⁺ channel CatSper^43,45. Taken together, these findings suggest that the acquisition of Ca²⁺-regulated asymmetric wave propagation in the posterior flagellum was likely a key evolutionary innovation in the common ancestor of opisthokonts and apusomonads.

Methods

Apusomonad culture

The apusomonad P. kaiyoae n. sp., originally isolated from a deep-sea sediment sample³⁴, was cultured in Hemi LB medium, supplemented with 1% v/v horse serum and 0.1% v/v LB medium in artificial seawater (Daigo’s ASW), and maintained at 20 °C. Subculturing occurred weekly in a 25 ml petri flask under a 12-h:12-h light:dark cycle. Cells used for experiments were ten days post-inoculation.

Recording light-induced response of P. kaiyoae

The Hemi LB medium was replaced with Herbst modified Artificial Seawater (ASW) in the cell suspension. The cells were then transferred to a microchamber on a glass slide (~1 mm depth) and kept in darkness for 1 h to allow them to settle. ASW was composed of 462.01 mM NaCl, 9.39 mM KCl, 10.81 mM CaCl₂, 48.27 mM MgCl₂ and 10 mM Hepes-NaOH (pH 8.0). Cell movement was observed using an inverted phase contrast microscope (IX71, Olympus, Japan) equipped with 10× or 40× objectives. For illumination, the entire field was exposed to a laboratory-made red stroboscopic LED lamp (620-630 nm, Power LED, Edison, Taiwan), while the center area was illuminated with blue light (440 nm; Spectra X, Lumencor Inc., OR, USA) using an iris diaphragm through the objective. In experiments involving different colors, cells were illuminated with cyan (470 nm), green (550 nm), and yellow (565 nm) light. Different intensity of blue light was used to test cell behavior (Supplementary Tables 1 and 2). High-speed cameras (HAS-U2, DITECT, Japan) were used to record images at 1 frame per second (fps) for trajectory analysis and 100 or 150 fps for waveform analysis. Photoavoidance was assessed by measuring the relative image intensity in the blue-light-illuminated center at 100 or 150 s after light-on using Bohboh software (Bohboh Soft, Tokyo, Japan). Cell density, velocity, and flagellar waveforms were analyzed using Bohboh software. The rate of cell contraction was calculated based on area measurement by ImageJ. Circular data analysis was performed with Oriana software (Kovach Computing Services, UK).

Effects of drug treatment

Nocodazole (Sigma-Aldrich), ciliobrevin A (Selleck), cytochalasin B (Sigma-Aldrich), and para-aminoblebbistatin (Sigma-Aldrich) were prepared by dissolving in dimethyl sulfoxide to concentrations of 12.5 mM, 40 mM, 200 mM, and 10 mM, respectively. Each drug was added to artificial seawater at a 1:1000 dilution. After the cells in Herbst modified ASW settled in the microchamber, the drug-containing artificial seawater was perfused into the chamber, and cell motility was recorded 15 min post-perfusion.

Effects of extracellular calcium ion concentration

Cells were loaded into a microchamber and allowed to settle. Subsequently, Ca²⁺-free ASW containing no CaCl₂ and chelated with 2.5 mM EGTA (CFSW), was perfused into the chamber, and cell motility was recorded 15 min later. The chamber was then perfused with Herbst modified ASW, and motility was recorded again 25 min later to confirm cell viability and contractility.

Antibodies and immunofluorescent microscopy

Cells were attached to glass slides and fixed with 2% paraformaldehyde for 30 min, followed by 4% paraformaldehyde for another 30 min at room temperature. After washing twice with PBS and once with PBS containing 0.05% Triton X-100 (TPBS), samples were incubated in blocking buffer (10% goat serum in TPBS) for 1 hour at room temperature. Subsequently, they were incubated with an anti-acetylated alpha-tubulin mouse antibody (T6793, Sigma-Aldrich, St. Louis, MO, USA; 1:1000 dilution) in blocking buffer for 1 hour at room temperature. Following four washes with PBS, samples were stained with a mixture of Alexa Fluor 488-conjugated goat anti-mouse antibody (Molecular Probes, Eugene, OR, USA; 1:1000 dilution) and Alexa Fluor 546 Phalloidin (A22283, Invitrogen; 1:1000 dilution) in blocking buffer for 1 hour at room temperature. In some experiments, anti-actin rabbit antibody (1854-1, Epitomics; 1:200 dilution) was used. After three 5-min PBS washes, samples were mounted in SlowFade Gold Antifade Mountant (Thermo Fisher Scientific). Super-resolution images were obtained using structured illumination microscopy (lattice SIM) with Elyra 7 (ZEISS, Germany).

Ca²⁺ imaging

Cells were transferred to a microchamber on a glass slide (~1 mm depth) and kept in darkness for 2 hours. The solution was replaced with Hemi LB medium containing 0.04% Cremophor, 5 μM Fluo-8H AM, and 0.1% DMSO, followed by 30 min in darkness at 20 °C. After four washes with ASW, the cells were further incubated in darkness for 30 min. Photophobic response induction and Fluo-8H excitation were performed simultaneously using cyan light (470 nm, intensity 30 μmol photons/m²/s) using a conventional fluorescent microscope (IX71, Olympus) at 10 fps, using the High-Speed Recording Software (Hamamatsu Photonics). Alternatively, the fluorescent change of cells induced by laser (488 nm, intensity 32 μmol photons/m²/s) was recorded at 6.3 fps using an Olympus IX73 (Olympus) microscope with a MEMS confocal unit (MAICO, Hamamatsu Photonics). Fluorescent intensity was analyzed using Bohboh software, and pseudo-color videos were generated using ImageJ. Control experiments were conducted under the same conditions without 5 μM Fluo-8H AM.

Electron microscopy

Thin-section electron microscopy was conducted following previously described methods^71,72, with some modifications. Cells were harvested, centrifuged at 700 × g for 5 min at 20 °C, and fixed with 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4) for 1 h at room temperature. After washing with 0.1 M cacodylate buffer, cells were post-fixed with 1% OsO₄ in 0.1 M cacodylate buffer at 4 °C for 2 h. Cells were washed five times with ddH₂O and then stained en bloc with 2% aqueous uranyl acetate overnight at 4 °C. Samples were then dehydrated in a 30–100% EtOH series, substituted with propylene oxide, and embedded in Agar Low Viscosity Resin (Agar Scientific, Stansted, UK). Sections (70 nm) were cut using an ultramicrotome (LEICA, Wetzlar, Germany) and mounted on a formvar-coated grid. Samples were stained with Reynolds’ lead citrate for 5 min and observed using a JEM-1200EX transmission electron microscope (JEOL, Tokyo, Japan).

Statistics and reproducibility

Data were collected from more than three independent experiments, each using different cell batches as biological replicates. Except for the experiment measuring cell density, between five to 26 cells were analyzed in each experiment; the number of cells is indicated in each result. Data are expressed as means ± standard deviation in box plots in R. Statistical analyses were performed using Dunnett’s multiple comparison test or Student’s T test. Statistical significance was defined as ^* = P < 0.05, ^** = P < 0.01 or ^*** = P < 0.001.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Author contributions

K.I. designed the outline of the study. A.S, C.Y., and K.I. conducted the main experiments. Y.S. and A.Y. handled cell culture. R.Y. performed electron microscopy; C.Y., A.S., Y.T., K.S., and K.I. conducted analysis. K.I. wrote the manuscript. All authors revised and approved the final version of the manuscript.

Peer review

Peer review information

Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editor: David Favero. A peer review file is available.

Data availability

All data in this paper are available in Supplementary Data 1 (source data for Fig. 2), Supplementary Data 2 (source data for Fig. 3), and Supplementary Data 3 (source data for Fig. 4) (Excel files). Additional data may be requested from the authors.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-025-09209-y.