Symbiodinium is a genus of dinoflagellate algae (in the family of Symbiodiniaceae). These are unicellular microalgae, commonly found in endosymbiotic relationship with cnidarians such as corals, sea anemones, jellyfish. They perform the essential photosynthetic energy generation process and thus play a crucial role in supporting the whole symbiosis (Roth, 2014; Davy et al., 2012). Certain environmental stress conditions (e.g., heat and light) cause the expulsion of these zooxanthellae from the host, which is manifested in coral bleaching (Weis, 2008). Symbiodiniaceae has two main cell types: the motile flagellated cells (referred to as mastigotes or gymnodinioid forms), and the non-motile coccoid form. The motile cells have longitudinal and traversal flagella, which move the cells in helical or cyclic patterns (Freduenthal, 1962; Lee et al., 2015). It has been shown that the motility of Symbiodiniaceae cells exhibits a diel rhythm, and the motility typically starts before or at the beginning of the light period, and peaks within a couple of hours from the start of the light period (Fitt et al., 1981; Yacobovitch et al., 2004), although the diel pattern of motility varies in Symbiodiniaceae species (Yamashita and Koike, 2016). The appearance of motile cells is preceded by cell division, which thus typically achieves its maximum rate before the onset of light, and therefore, the cell division and increase in motile cell forms are correlated (Yamashita and Koike, 2016). The motility of Symbiodiniaceae is controlled by an endogenous rhythm, which determines the alternation between motile and non-motile states (Fitt et al., 1981). These different cell forms play an important role in making contact with the host and can be significantly modified in free-living and symbiotic lifestyles (Fitt et al., 1981). The host cells impose significant regulations on the growth, metabolism and cell cycle of the endosymbiont algae (Koike et al., 2004; Stat et al., 2006), and the algal symbiont exerts a diverse and complex diurnal regulation of their cellular environment, indicating that the symbiosis between the host and algae requires synchronization and coordination of several parallel processes, which are not completely understood (reviewed in Sorek et al., 2014).

The cell cycle dynamics of Symbiodiniaceae has been intensively studied in the past decades using various microscopic tools (Fitt et al., 1981; Fitt and Trench, 1983; Cui et al., 2022; Yamashita and Koike, 2016; Nitschke et al., 2015). However, in these studies repetitive manual sampling was needed, which may be suboptimal to capture fine changes in cell cycle, and it was not possible to modify or manipulate the growth medium or environment of the cells in a non-disruptive manner. High-throughput precisely controlled environment is therefore imperative to monitor cell morphology without potential disturbances or alteration of the ambient conditions. These can be achieved using microfluidic methods, in which a precisely controlled environment with real-time monitoring of physiological events (with a couple of seconds time resolution) is possible, even for a couple of days (Bashir et al., 2022; Patil et al., 2023). Microfluidics has gained increasing interest due to its advantages in controlled microenvironments. This technology offers custom design and solutions to manipulate fluids in microscopic dimensions, particularly tailored for single cells or cell colonies, with widespread applications for microbiology and microbial physiology (Nagy et al., 2018; Weibel et al., 2007). The greatest advantage of microfluidics is the capability of high precision geometric control along with liquid manipulation, which enables laminar liquid flow and thereby precise regulation of flow with minimal shear stress to the cells (Nagy et al., 2018). By this means, several ecologically relevant factors (nutrients, chemical signals, pH, etc.) can be regulated or valuable interactions between organisms can be revealed (Ábrahám et al., 2024; Greipel et al., 2024). Application of microfluidics has attracted considerable interest to study physiological processes of photosynthetic microalgae by the combination of single cell morphology with microscopic scale chlorophyll (Chl) fluorescence imaging (Széles et al., 2022; Széles et al., 2023; Patil et al., 2023). Single-cell Chl fluorescence imaging is a particularly useful tool to reveal photosynthetic light energy utilization and excess energy dissipation, providing therefore valuable insights about the phenotypic responses and heterogeneity of photosynthetic activity in a range of single-celled algae (Oxborough et al., 2000; Oxborough and Baker, 1997; Komárek et al., 2010), reviewed in (Schreiber, 2004; Baker, 2008). Investigation of the photobiology of Symbiodiniaceae has progressed significantly in the past years, and precisely controlled microfluidics-based experimental setups combined with single-cell chlorophyll fluorometry allowed the multiparametric monitoring of thermal responses of Symbiodiniaceae to reveal species specific phenotypic differences (Behrendt et al., 2020; Andersson et al., 2021; Xiao et al., 2022) or monitoring the physiological changes of Symbiodiniaceae cells during protoplast formation and regeneration (Bashir et al., 2022).

Establishing a controlled microenvironment that allows the observation of individual cells in different life stages or forms of Symbiodiniaceae, which also allows the precise modulation of physico-chemical factors is essential. Precise adjustment of microenvironment and concomitant observation of single cells can be achieved by trapping the cells or increasing the viscosity of the medium to slow down rapidly motile cells without affecting the physiological state (i.e., photosynthetic capacity) of the cell population. In the current work, we applied an experimental strategy in which the Symbiodinium cells are not trapped or immobilized, but temporarily slowed down using the physiologically ‘inert’, non-toxic polysucrose Ficoll. Ficoll has been shown to be a useful tool to harvest phytoplankton cells or to improve the stability of photosynthetic activity measurements in microalgae, due to its non-osmotic features which largely preserve cell morphology and physiological parameters, although Ficoll certainly has an impact on viscosity (Joliot and Delosme, 1974; Price et al., 1974). Changing the viscosity using Ficoll has also been found to impact the motility of dinoflagellate Heterocapsa triquetra; increasing viscosity caused slowing down of swimming cells, although the response of the behavior of this alga to changes in viscosity has proven to be complex (Orchard et al., 2016).

The photosynthetic heterogeneity of the different cell forms of Symbiodiniaceae remained largely uncharacterized. This is to a large extent due to the motility of flagellated cells, which prevents accurate estimation of photosynthetic activity at the single-cell level. It was the aim of the current work to monitor the diel changes in motile and non-motile cells using combined morphological analysis and single-cell Chl fluorometry, under varying light-dark periods (normal 12 h: 12 h light: dark and continuous light), using microfluidic cultivating chambers. We aimed to establish conditions when fast-moving cells are transiently slowed down (but not completely immobilized) to mimic symbiotic conditions as opposed to non-symbiotic conditions and to reveal the ensuing changes in photosynthetic activity.

Symbiodinium cells exhibited characteristic motile and non-motile features in the microfluidic chambers (Supplementary Table S1). Based on image segmentation, the motile cells were identified as circular objects due to their fast rotation speed (red objects in Figure 2).

The non-motile cells are separated from motile cells because of the lack of any motion or dislocation of these cells (blue objects in Figure 2). These results indicate that incubating Symbiodiniaceae cells in the cultivating microfluidic chamber maintained the general features of the different cell types observed previously in this alga and the different cell types could be identified and investigated in the microfluidic cultivation chambers.

To resolve the time-dependent changes in cell population and in the fractions of the different cell types, the image analysis was performed at each time point for the entire experimental period (Figure 3).

In normal F/2 growth medium, under a 12 h:12 h light: dark cycle, the cells showed a sustained daily oscillation between the motile and non-motile cell forms, in agreement with previous studies (Yamashita and Koike, 2016). In the dark phase, the motility was minimal as the non-motile coccoid form dominated the cell culture. Towards the end of the night cycle, cells started to divide typically into two moving forms (more detailed presentation of cell division in different conditions are also shown below). As the cell division intensified the number of motile cells started to increase, and the relative number of motile cells peaked 2–3 h after the onset of light. Afterwards, the number of motile cells started to decrease and at the onset of dark phase, the motility ceased nearly completely. This cycle could be repeatedly observed for several days, although it has to be noted that the peak amplitude (i.e., the maximal proportion of moving cells relative to total cell number) varied considerably at the different days and among the different replicates, in the range of moving cells/total cell ratio of 0.2–0.4. In the dark phase, this ratio decreased to 0–0.05 (Figure 3). These results indicate that the assessment of time-resolved changes of the cell types and the relative changes in motile cells of the Symbiodinium population during a diurnal cycle could be studied for extended time periods in the microfluidic chamber. It has to be noted that under constant light, the motile and non-motile cells also displayed a diurnal cycle, however the amplitude of the increase in motile cells decreased over time, as the cells tended to lose the diurnal rhythm due to the presence of constant light (Supplementary Figure S4), in agreement with previous observations in bulk cultures (Fitt and Trench, 1983).

In order to slow down the motile cells and create an environment with higher viscosity that mimics coral symbiotic conditions, the polysucrose Ficoll was applied. It was found based on the preliminary microplate experiment that the motility of the cells changed significantly in the Ficoll-containing solutions. In increasing Ficoll concentrations, the motility of the cells slowed down, cells showed a slow swimming or swinging pattern, or the circular motions were visually much slower at 7% Ficoll as compared to control condition (F/2 medium only), in which the cells displayed regular rapid circular movement (Supplementary Figure S5). Therefore, Ficoll provided an environment in which motile cells are remarkably slowed down but the complete immobilization of cells was avoided. In this manner the slowed flagellated cells could be observed and snap photographed in their characteristic spindle shape.

Ficoll (7%) caused a significant alteration in the cell motility in the cultures maintained in the microfluidic chambers (Figure 4A). In agreement with the microplate incubation experiment, the rotational movement of the flagellated cells slowed down as the circular movement pattern visually ceased or significantly decreased (Figures 4C,D). The diurnal cycle also displayed profound changes. At the onset of light, the number of moving forms increased, similarly to the normal F/2 conditions, the relative number of motile cells peaked 2–3 h after the onset of light. However, after the peak phase, the number of motile cells rapidly declined, and by the 6^th h of the light phase, the motile cells disappeared nearly completely (Figure 4A). Therefore, the phase of the moving forms was much narrower as compared to the non-treated (Ficoll-free) cultures (c.f. Figures 4A,B). It seems that in the presence of Ficoll, the motile cells appear only temporarily as a spike after the light phase of the diurnal cycle sets on, therefore, the cultures in Ficoll conditions largely remained in non-motile form. However, this does not mean that Ficoll prevented the appearance of the moving forms, instead, the phase of the moving forms was restricted to a narrower time window. The motile cells can be distinguished by their characteristic spindle shape, which was much more apparent due to the slowing down of their rotation in high-viscosity medium (Figures 4C,D). These characteristic spindle-shaped cells could not be visually observed in normal F/2 medium, due to their fast rotation (Figures 4E,F).

To investigate whether the normal cell cycle could be regained from Ficoll treatment, the Ficoll solution was washed out and replaced with normal F/2 medium. After washing out Ficoll, the daily cell cycle regained its characteristics similar to the control F/2 condition within 30 h, i.e., the wider phase of motile cells during the light period (Figure 4B). The cells regained their fast rotation and motility and so they appeared as circular objects in the observing area (Figure 4F).

It also has to be noted that Ficoll did not impair the normal cell division capacity of the cultures. It was observed, both in the presence and absence of Ficoll, that in the dark phase largely coccoid cells existed, and towards the end of the dark phase, cell division increased. Under all conditions, the cultures maintained the typical behavior of one non-motile cell dividing into two motile daughter cells (Figure 5). Therefore remarkably, the Ficoll treatment did not impact the cell division capacity of Symbiodinium.

The activity of PSII was measured under control and Ficoll-treated conditions (Figure 6). Non-slowed motile cells displayed incorrect F_v/F_m values, because cells visibly moved between the F₀ (Figure 6A) and F_m (Figure 6B) images, thereby creating artificially high, unrealistic F_v/F_m of 0.7–0.9 (Figures 6C,D). Due to the rapid rotation of the cells in normal F/2 medium, it was not possible to account for the displacement of the cells during F_v/F_m measurement. Ficoll on the other hand, largely prevented the movement of cells, therefore, F_v/F_m could be measured more reliably (Figures 6G,H). However, it has to be noted that even in the case of Ficoll-slowed motile cells, care has to be taken to select the appropriate AOI separately for F₀ and F_m (Figures 6E,F), otherwise artificially high F_v/F_m is calculated (blue-purple colored cells in the Ficoll treatment, showing unrealistic F_v/F_m = 0.8 or higher for Symbiodinium, Figures 6G,H).

In the case of slowed motile cells, the F₀ and F_m were analyzed with different AOIs, to correct for the small movement of cells (decreased motility due to Ficoll environment), and the F_v/F_m was calculated with from the respective F_m and F₀ AOI (similarly to the ‘nudging’ method, which corrects for small cell movements between F′ and F_m’ images, (Oxborough et al., 2000)). In the case of non-motile cells, the same AOI was taken for F₀ and F_m, since the non-motile cells were not displaced. By comparing approx. 20 cells, the F_v/F_m values of the non-motile and motile cells were not significantly different (Figure 7).

An important outcome of the current work is that microfluidic chambers are suitable platforms for the cultivation of the coral endosymbiont algae, Symbiodiniaceae. The microfluidic method applied in the current study represents a large cultivation chamber (in microscopic terms, which is about 1000 × 700 μm) without cell traps. This allowed loading the initial cell culture, observing their morphology, growth, cell division as well as their single-cell photosynthetic activity, with a fixed cultivation setup and geometry for the entire period of the experiment, without the need of repetitive sampling from batch cultures. The cells maintained their motility and division capability and exhibited a normal diurnal cycle typical for Symbiodiniaceae (Figure 3), in agreement with previous findings, which were performed in bulk cultures (Fitt et al., 1981; Yacobovitch et al., 2004; Yamashita and Koike, 2016; Fitt and Trench, 1983).

It has been shown earlier that elevated viscosity significantly alters the motility, swimming patterns, and swimming speed of dinoflagellates (Orchard et al., 2016; Sohn et al., 2013). Freduenthal (1962) showed that increasing the viscosity of the medium using methyl cellulose caused a significant hindrance of the movement of the motile cells of Symbiodiniaceae. Mucus excreted by the coral tissues is accompanied by changes in viscosity, which can therefore fulfil an important physiological role of adjusting the physicochemical and physiological barriers of corals, among other myriad of important functions (Bhagwat et al., 2024). Symbiodinium motility and division, on the other hand, are affected by many factors, including host factor-derived additives and organic compounds (Krueger and Gates, 2012). In the current study, only the viscosity of the medium was changed by the application of Ficoll, which could be precisely controlled by microfluidics. Increasing the viscosity of the medium using Ficoll allowed to reveal the changes in the cell cycle phases in Symbiodinium in a more viscous environment, such as in coral tissues.

The rationale of the application of a medium with elevated viscosity was that the motile cells can be slowed down transiently, and thereby their single-cell characteristics and physiology can be observed without the need for complete immobilization or forcing the cells permanently into non-motile forms. In other words, the establishment of an environment in which slowed motile cells could be observed is imperative.

The typical diurnal cell cycle was maintained in the presence of Ficoll, but with an altered time frame; from the onset to the peak of motile cell number, relative to total cell number, the timeframe was much shorter than in the normal F/2 medium. This could be due to the enhanced viscosity of the medium, which caused the cells to mainly remain in a microenvironment that simulates coral tissue conditions. Nonetheless, the cells retained their active capability for division and did not lose the ability for dividing into motile cells, and typically exhibited the spindle shaped form of mastigotes (Figure 5). However, the motile cells were slowed down, as the circular rotation, typical for motile cells (Nitschke et al., 2015) disappeared, and rather a sliding one-directional motion was observed (Figure 4). Therefore, it appears that a more viscous environment might mimic the coral tissue environment at least in a physico-chemical manner. Importantly, the diurnal alteration of motile and non-motile phases is not lost under these circumstances, only the period length changes. Furthermore, Ficoll did not cause a detrimental effect on cell physiology, as indicated by the fact that the cell cycle regained its properties comparable to control (F/2 only) conditions after washing out Ficoll.

Determination of photophysiological parameters of the motile Symbiodinium cells represent a significant challenge due to their very rapid circular motion in F/2 medium. The basic problem is that the rapid rotation resulted in the overestimation of F_v/F_m since the AOI selected for F_v/F_m calculation was not the same for F₀ and F_m images (Figures 6C,D). Recently, software analysis developments allowed a more high-throughput semi-automated analysis of the PSII activity of Symbiodiniaceae cells, immobilized in microfluidic chambers (Pontén et al., 2024), however, these conditions are far from the physiological ones, as these do not allow the analysis of all cell forms of Symbiodiniaceae (i.e., the motile and non-motile forms). With the presented microfluidic setup, the determination of single-cell PSII activity of the motile form could be significantly improved in the presence of Ficoll. Although, the minor movement of the cells still resulted in somewhat altered AOI for F_v/F_m calculation, this slow displacement could be tracked and the AOI could be selected specifically for the F₀ and F_m image, based on which the F_v/F_m could be accurately determined. By performing this analysis, the F_v/F_m of non-motile and motile cells, slowed down by Ficoll treatment, exhibited similar PSII activity under the investigated conditions. Although it has to be stressed that the F_v/F_m of motile cells cannot be measured in normal F/2 medium, with the method described above the F_v/F_m of slowed cells could be determined with higher precision, and the observed PSII activity was in the range of F_v/F_m = 0.45–0.5, in agreement with (Behrendt et al., 2020; Pontén et al., 2024).

By this means, employing transient slowing of motile cells, the photosynthetic activity (PSII activity) could be monitored more precisely at the single-cell level, along with the physiology of non-motile forms. It has to be noted that the photosynthetic activity measurements are restricted to F_v/F_m determination in the current study. Determination of F_v/F_m solely does not reveal the rate or the regulation of photosynthetic electron transport in the different cell types. For a detailed understanding of the photosynthetic performance of the different cell types, it would be necessary to perform single-cell fluorescence kinetic or light response curve analyses (e.g., Patil et al., 2023), however, this might be problematic even in the case of slowed-down motile cells, due to the long measurement time (at least several minutes). Due to its precisely controlled nature, microfluidics could be used to monitor important single-cell physiological features of coral host-symbiont interactions, bleaching responses under climate stress scenarios.

In conclusion, our results indicate that the application of Ficoll as an inert polysaccharide caused a slowdown of motile cells, which allowed the monitoring of photosynthetic activity of all cell types of Symbiodinium. Although with the presented microfluidic setup, the single-cell PSII activity can be measured, it is important to note that the PSII activity could be precisely determined only in non-motile or immobilized cells. Furthermore, as the viscosity of the medium increased with the application of Ficoll, this method might represent a physico-chemical environment that mimics coral tissues. However, it has to be noted that a microfluidic analysis of Symbiodinium cell cycle under free-living and in hospite conditions requires a thorough investigation taking into account several host factors, which was beyond the scope of the current study. The microfluidics method, as presented here, serves a high-precision-controlled environment, which might allow the investigation of host conditions and other stress factors. For example, co-culturing Symbiodiniaceae with coral host cells, or integrating thermal stress experimental scenarios are promising future applications of the microfluidics setup presented here.