Graviresponses in Paramecium Biaurelia Under Different Accelerations: Studies on the Ground and In Space

As long ago as 1906, Jennings noted that at least some and probably many protozoa use gravity as an environmental stimulus for orientation. Some, such as Paramecium, are even able to control both their swimming direction (gravitaxis) and their swimming velocity in response to gravity. This gravikinesis enables the protozoa to compensate, at least in part, for sedimentation by increasing their upward swimming rate and by reducing their downward swimming rate (Machemer et al. 1991; Ooya et al. 1992). The mechanisms controlling gravitaxis and gravikinesis are still under discussion. Hypotheses based on purely physical or purely physiological reactions have been put forward to explain the graviresponses of Paramecium. There are many indications that a passive mechanism, such as the principle of a buoy, is insufficient to explain the variety of behavioural responses that occur (for a review, see Bean, 1984; Machemer and Bräucker, 1992). In consequence, gravitaxis and gravikinesis are thought to involve a physiological signal transduction chain similar to that involved in the perception of light (Colombetti, 1990) and chemical (Van Houten and Preston, 1988; Van Houten, 1992) stimuli. Superposition of different stimuli (light and gravity) has been demonstrated, e.g. for the flagellate Euglena gracilis (Häder, 1987), where such behaviour is ecologically essential for the survival of the organisms. Such superposition may involve common pathways for different signal transduction chains.

A greater knowledge of the mechanism of graviperception can be gained by studying the effects of varying the stimulus (gravity). In previous studies, it was shown that an absence of gravity resulted in random swimming of Paramecium aurelia, Paramecium biaurelia (Hemmersbach-Krause et al. 1993a,b) and Euglena gracilis (Häder et al. 1990) accompanied by equal swimming velocities in all directions, indicating that gravity was the sole reason for the behaviour that occurs under normal conditions of gravity (1 g). In the present study, the behaviour of Paramecium biaurelia in response to gravity changes was investigated using a slow-rotating centrifuge microscope (NIZEMI, Dornier, Friedrichshafen, Germany) on the ground and in space. This enabled reactions to increasing and decreasing gravitational stimulation to be examined. Experiments under decreasing accelerations below 1 g were performed during the Spacelab IML-2 (International Microgravity Laboratory) mission to determine the minimum gravity level at which the transition from oriented movement (negative gravitaxis) to random swimming occurred.

Paramecium biaurelia Sonneborn (obtained from the Culture Collection of Algae and Protozoa, Ambleside, UK) was cultivated in straw medium (pH 7.2, 22 °C). The cultures had to be enriched without centrifugation to exclude any effects of accelerations before experiments were begun. To achieve this, cell suspensions were placed in volumetric flasks, in which the protozoa assembled in the narrow neck owing to their negative gravitaxis and positive aerotaxis. Experimental observations were performed in custom-made observation chambers, which in the case of the 15 day space experiments had to be developed to allow cultivation as well as observation of the cells. In over 20 tests, Plexiglas chambers proved well-suited for the survival of the cells for up to 15 days, but the precision of gravitaxis of Paramecium biaurelia was rather low in these chambers. In contrast, within 6 h of transfer into a glass chamber, Paramecium biaurelia showed a precise gravitaxis, but the cells died after 1 day. From previous studies, we know there is a correlation between oxygen concentration and degree of orientation, so that under hypoxic conditions the precision of orientation is high (Hemmersbach-Krause et al. 1991). The preliminary observations indicated that the permeability for oxygen is high through Plexiglas and low through glass, and these results were taken into account in the construction of the chambers. Experiments on the ground were normally performed in simple cylindrical Plexiglas chambers (diameter 23 mm, 0.5 or 1.0 mm depth) covered by a glass plate sealed with silicone paste. In addition, control experiments were performed in the more complex chambers designed for the experiments carried out in space. These controls showed that there was no difference in the behaviour of the cells in the different containers. The chambers used in space were divided into two parts, one for cultivation (Plexiglas, 27 mm×35 mm, 3 mm depth) and one for observation (glass, 27 mm×27 mm, 1 mm depth). The two parts of the chamber were separated by a plate with two holes which were initially sealed by silicone. On the ground, before lift-off of the space shuttle, the observation part was filled with sterile culture medium, the cultivation part with cells. Six hours before the experiments were begun, an astronaut transferred the cells into the observation part by moving a piston.

The cuvettes were stored on board the shuttle Columbia 19 h before launching. The samples were kept at 10 °C to reduce the effects of acceleration on the cells during the ascent of the rocket. In space, the samples were incubated at 22 °C on a 1 g reference centrifuge. In preparation for the experimental run on mission day 3, one sample was transferred from the 1 g reference centrifuge to the slow-rotating centrifuge microscope (NIZEMI), accelerated to 1 g and subsequently slowed down in 10 steps to the conditions of weightlessness ‘0 g’ (NIZEMI halted) and was then taken back in a single step to 1 g. Each acceleration or deceleration step lasted for 5 min. Conditions within the spacecraft are such that only near-weightlessness (microgravity) is achieved (indicated by ‘0 g’ in Fig. 6). The residual acceleration during the ‘threshold experiments’ did not exceed 10⁻³g as measured by accelerometers.

The experiments were performed on a slow-rotating centrifuge microscope (NIZEMI) (Fig. 1) constructed by Dornier (Friedrichshafen, Germany) and sponsored by the German space agency DARA. Two models of the NIZEMI were used for the experiments, a ground model and a space model, differing in mass and dimensions. The NIZEMI is a horizontal microscope (Axioplan, Zeiss, Jena, Germany) mounted tangentially on a circular platform driven by an electric motor. In order to achieve good contrast for computer-controlled image analysis, the cells were observed using dark-field illumination and red light (645 nm) at an objective magnification of 1.25× or 2.5×. During rotation, the optical axis remains perpendicular to the resultant vector of gravity and the centrifugal acceleration. In the NIZEMI ground model, the effective radius is 170 mm. The maximal acceleration of 5 g was achieved at a rotational speed of 163 revs min⁻¹. In the space-adapted NIZEMI, the effective radius is 110 mm and the maximum acceleration of 1.5 g was achieved at 111 revs min⁻¹. In the space experiment, the temperature of the cuvettes was maintained at 22 °C by eight Peltier elements located around the cuvette holder. In the ground experiments, no temperature control of the cuvettes was implemented and these experiments were performed at a room temperature of 22 °C. During rotation, the microscope image was recorded by a black and white CCD camera mounted on the microscope. The video sequences were either stored on an NTSC video recorder (V-250 AB-R, TEAC, USA) in the space experiment or on an SVHS recorder for the ground experiments. The shuttle was launched from Kennedy Space Center on 8 July 1994 for a 15 day mission. During the experimental run, the video image was transmitted to the mission control center in Huntsville, Alabama, to allow control of the viewing area and focus by the experimenter.

At 1 g, the precision of gravitactic orientation of Paramecium biaurelia differs from culture to culture. Under controlled conditions (closed observation chamber), negative gravitaxis is usually observed several hours after placing the cells in the observation chamber (Hemmersbach-Krause et al. 1991, 1993a). When a cell culture with a clear negative gravitaxis was accelerated (tested up to 5 g), the cells did not sediment and negative gravitaxis persisted. Fig. 2 shows a typical example in which the r-value rose from 0.37 at 1 g (preferred direction θ=16.5 °, N=521) to 0.73 at 5 g (θ=358 °, N=509) (Fig. 2A,B). In cultures with a lower degree of orientation under 1 g conditions (r-value of 0.22, θ=20 °, N=585), hypergravity of 5 g also induced a clear gravitactic response, demonstrated by a significant rise in the r-value to 0.72 (θ=1 °, N=750) (Fig. 2C,D) accompanied by a closer alignment of the cells towards the g-vector, as indicated by the decrease in the angle θ. The reaction occurred within seconds; the circular histograms represent the data from the first minute under hypergravity.

Repetitive stimulation between 1 g and hypergravity (for an example of changes between 1 g and 5 g, see Fig. 3, each acceleration step lasting 2 min) revealed that the reactivity of the cells was maintained. They always showed clear responses to hypergravity (increase in r-value) accompanied by acceleration-dependent changes in the upward swimming velocities. In comparison with the mean 1 g values (599± 24 μm s⁻¹, N=7624), the mean upward swimming velocities decreased by approximately 200 μm s⁻¹ at 5 g. This value corresponds to about 2.5 times the sedimentation velocity at 1 g (82±22 μm s⁻¹; Hemmersbach-Krause et al. 1993b) (see Discussion).

To determine the length of time needed for the cells to reorient with respect to a new acceleration vector and to adjust their swimming velocities, the measurements were divided into 20 s intervals, the minimum period needed for the registration of a sufficient number of cells for statistical analysis. Five experiments were carried out and a representative example is shown in Fig. 4. Immediately upon acceleration, the cells showed a reaction (Fig. 4); within 20 s, the orientation histogram revealed a clear negative gravitaxis towards the new acceleration vector, accompanied by an increased precision of orientation (r-value=0.96, θ=355 °, N=289), which subsequently decreased slightly to 0.9. Since 97 % of the cells were swimming upwards (negative gravitaxis) at 5 g, only the mean upward (0±30 °) swimming velocities (with standard error of the means) were considered. Within the first 20 s at 5 g, the upward swimming velocity decreased significantly from 642±12 μm s⁻¹ (N=232) (1 g value) to 527±10 μm s⁻¹ (N=289) and decreased further within the next 20 s to 465±10 μm s⁻¹ (N=275). The decrease corresponds to twice the sedimentation velocity at 1 g. When the centrifuge was stopped, the upward swimming velocity increased to 645±12 μm s⁻¹ (N=226) within the first 20 s, a value not significantly different from the initial 1 g value (Fig. 4). Fig. 4 also shows that the degree of orientation was higher than the initial 1 g value (r-value 0.84, θ=26.5 °). After 60 s, the precision of orientation had declined to that of the starting conditions during the last 20 s at 1 g (data not shown). These results indicate that both responses (gravitaxis and gravikinesis) occurred within a few seconds.

In order to test whether cells adapted to hypergravity, they were exposed to 5 g for 10 min (Fig. 5). Hypergravity induced negative gravitaxis and a significant decrease in the mean upward swimming velocity. Both responses were maintained throughout the exposure to increased gravity.

A cell culture cultivated in a 1 g reference centrifuge in space was transferred to the centrifuge microscope and accelerated to 1 g (Fig. 6) which induced negative gravitaxis, indicated by an r-value of 0.26 (θ=29 °, N=693), comparable to the values obtained on Earth (Fig. 2). During stepwise deceleration, negative gravitaxis became less precise and could no longer be measured at 0.16 g, when the r-value had decreased to 0.02 (N=1409) (Figs 6, 7C). Randomness was tested by calculating of the statistical value z. In microgravity (0 g), after stopping the centrifuge, an apparent orientation was measured. However, since the preferred direction of the culture was 177 °, this value does not provide information on the threshold for gravitaxis (Fig. 7D). The orientation might be induced by a temperature gradient (see below). Acceleration to 1 g in the last step re-established negative gravitaxis, indicated by an r-value of 0.33 (θ=26 °). The results indicate that the minimum acceleration required to induce negative gravitaxis was below 0.3 g and above 0.16 g. The occurrence of thermoconvection in the cuvettes during acceleration means that the data for the swimming velocities must be considered with care. The significant difference between upward and downward velocities, which was registered at 1 g, was no longer visible at 0.6 g (data not shown).

The graviresponses of Paramecium biaurelia consist of two components: an oriented turn (gravitaxis) and regulation of the swimming velocity (gravikinesis). Measurements at 1 g revealed that the sedimentation velocity (82 μm s⁻¹ for Paramecium biaurelia; Hemmersbach-Krause et al. 1993b) is partially compensated for in negative gravitactic cells by an increase in their active upward swimming velocity (Machemer et al. 1991). Bräucker et al. (1994) demonstrated a threefold ‘quasi-linear’ increase in sedimentation rate between 1 g and 5 g, assuming that Stokes law does not apply completely for these large organisms since otherwise a fourfold increase would be expected. Nevertheless, our data show that the net upward swimming velocity at 5 g only decreased by a maximum of 2.5 times the sedimentation velocity at 1 g, indicating that the cells had increased their active upward swimming velocity and thus their gravikinetic response.

Application of hypergravity showed that gravitaxis and gravikinesis increased, but with obviously different time courses, supporting data from Machemer et al. (1993), who stated that there was no ‘direct relationship between graviorientation and gravikinesis’. The question of adaptation to the stimulus (gravity) is difficult to resolve and results given in the literature are controversial. The parameters investigated by our experimental design did not adapt to changes in gravity, supporting data for Paramecium caudatum (Ooya et al. 1992), for Loxodes striatus (Machemer-Röhnisch et al. 1993), for Euglena gracilis (Häder et al. 1995) and even for vertebrates in the case of Xenopus laevis (Neubert et al. 1994). Machemer et al. (1993), Machemer-Röhnisch et al. (1993) and Bräucker et al. (1994) stated that gravikinesis did not adapt, but that gravitaxis did adapt, on the basis of the observation that gravitaxis declined over several hours at 1 g. These results are in contrast to our observations and to those reported for Euglena gracilis (Lebert and Häder, 1996), because the precision of gravitaxis increased with time after filling the cells in the observation chambers. Further experiments will be necessary to show whether these differences are a result of experimental design, e.g. the materials used to construct the observation chamber and differences in oxygen permeability.

By using a centrifuge microscope in space, it was possible to observe swimming at accelerations below 1 g and to determine for the first time the lowest acceleration necessary to maintain the gravitactic response. A random distribution of Paramecium biaurelia occurred at and below 0.16 g. This threshold value is within the range of values obtained for other species, e.g. Euglena gracilis, with the same experimental set-up (Häder et al. 1995). Use of the term ‘threshold’ implies that gravitaxis is not a result of buoyancy, but rather the result of a physiological signal transduction chain as indicated by previous results (see below); however, this has not been unequivocally demonstrated in Paramecium. A rigorous discrimination between the first sign of directivity due to the rising acceleration and the first sign of gravisensory transduction is difficult and needs further experimental or theoretical approaches.

The mechanism underlying the gravity-dependent behaviour of Paramecium has been under discussion for many years, with either a pure passive physical mechanism being proposed or a physiological mechanism involving a sensory transduction pathway (for a review, see Bean, 1984; Machemer and Bräucker, 1992). Electrophysiological studies revealed a characteristic distribution of specific ion channels in the membrane of Paramecium, with depolarizing Ca²⁺ channels in the anterior region and hyperpolarizing K⁺ channels in the posterior one (Naitoh, 1984; Machemer, 1988). Mechanical stimulation of these channels induces either a decrease in the frequency of the ciliary beat (depolarization) or an increase in the frequency (hyperpolarization). This lead to the hypothesis for gravikinesis that the faster active forward swimming rate in upward-swimming cells and the decreased forward swimming rate in downward-swimming cells are the result of differential ion stimulation. It is proposed that the whole cytoplasm acts as a statolith by exerting a distinct pressure of approximately 0.1 Pa on the lower cell membrane (Machemer et al. 1991), thus initiating a signal transduction cascade by stimulation of specific ion channels. If stimulus saturation and, thus, maximum activation of the channels is not achieved at 1 g, the increase in pressure under hypergravity should increase gravikinesis, as shown by the present data, which support the results of Bräucker et al. (1994). It is not clear whether gravitaxis is controlled by the same mechanism. The precision of orientation under hypergravity and the maintenance of reactivity during repetitive stimulation provide no information on the physiological mechanism involved and could be explained by the buoyancy principle. However, the variability in the degree of gravitaxis shown by different cultures, the variable orientations of sedimentating Paramecium cells (Kuznicki, 1968; R. Hemmersbach, personal observations), the changes in the directionality of gravitaxis in Euglena gracilis induced by heavy metals (Stallwitz and Häder, 1994) and the inhibition of gravitaxis by the stretch-activated channel blocker gadolinium in Euglena gracilis (Lebert and Häder, 1996) are strong indications that purely physical mechanisms are not sufficient to explain the complex responses to gravity. Although Paramecium and Euglena have no specific statocyst organelle, such as the Müller organelle in Loxodes, common cellular structures are able to take over this function as has been shown in plant systems (Björkman, 1988; Buchen et al. 1993; Sack, 1991; Sievers and Volkmann, 1979; Sievers et al. 1991). Further experiments will be necessary to identify the perception/transduction mechanism. In this context, hypergravity experiments which increase weak graviresponses are a promising experimental approach and could be used, for example, to investigate the behaviour of genetic mutants, the effects of density-adjusted medium and the effects of channel blockers.

We are indebted to W. Briegleb both for fruitful discussions and for technical assistance. The authors thank B. Bromeis, I. Block and A. Wolke for skilful technical support. Our special thanks are due to the astronauts of the IML-2 mission for their devoted work during training and in the performance of the experiment. We would also like to thank the teams of DORNIER, MUSC and DARA for support. We thank NASA for the opportunity of performing this experiment in space.