A Mechanosensory Mechanism for Evoking Cellular Contraction in Vorticella

The peritrich ciliate Vorticella is known to exhibit a cellular contraction, involving shrinkage of the cell body and coiling of the stalk, in response to mechanical agitation (such as water current), vibration of the water and/or of the substratum, or collision with a solid object. Mechanosensitivity is a common characteristic among a variety of living cells, from bacteria to mechanoreceptor cells of vertebrates (Martinac et al. 1987, 1990; Gustin et al. 1988; Naitoh, 1982, 1984; Machemer, 1985; Hudspeth, 1989).

In many ciliate protozoans, the mechanosensitive Ca²⁺ channels responsible for producing an avoiding response are present predominantly in the anterior membrane, while mechanosensitive K⁺ channels responsible for the escape response are present predominantly in the posterior membrane (Naitoh and Eckert, 1969; Naitoh, 1974). Mechanosensitive ion channels are also distributed in localized regions in vertebrate mechanoreceptor cells. They are restricted to the nerve ending of a Pacinian corpuscle (Loewenstein, 1971) and mechanoelectrical transduction channels are present around the tip (Hudspeth, 1979, 1982) or the base (Ohmori, 1988) of stereocilia of the hair cell.

The aspects of a mechanical stimulus effective in evoking a cellular response are different in different types of cells. Mechanoreceptor channels of patch-clamped chick skeletal muscle cells, frog oocytes, yeast cells and bacteria can be gated by hydrostatic pressure (Guharay and Sachs, 1985; Martinac et al. 1987; Taglietti and Toselli, 1988; Gustin et al. 1988). The magnitude of the mechanoelectrical transducer current of a voltage-clamped vertebrate hair cell is dependent on the absolute displacement of its stereocilium (Hudspeth, 1989). Both the rate and the extent of stretching of a tendon influence the activation of the tendon receptor of the cat (Ruffini ending of the knee-joint) (Boyd and Roberts, 1953). The rate of creation of a localized depression in the surface membrane affects the activation of mechanosensitive ion channels in many ciliate protozoans such as Paramecium caudatum, Euplotes sp., Stylonychia lemnae and Tetrahymena pyriformis (Naitoh et al. 1972; De Peyer and Machemer, 1978; Naitoh, 1984; Machemer, 1985). The great variation in the effective factor among different types of mechanoreceptor cells is attributable, at least in part, to differences in their viscoelastic properties.

Vorticella is an excellent model system for quantitative studies of mechanosensory transduction, because of the cell’s high mechanosensitivity, its lack of mechanoreceptor specializations (which results in simpler mechanical properties, which are easier to analyse) and its large size.

The primary objectives of the research described in this paper were (1) to examine any localized difference in the mechanosensitivity for initiating a cellular contraction; (2) to investigate the possible involvement of ion channels in the mechanosensory mechanism; and (3) to determine the factor(s) of a mechanical stimulus that is effective in evoking a cellular contraction in Vorticella. The biological significance of the contraction is discussed in relation to the underlying factor(s). Some of these results have been presented verbally elsewhere (Katoh and Naitoh, 1988, 1991).

Specimens of Vorticella sp. were grown at 20°C on a glass slide in a bacterized saline solution (final concentrations O.lmmoll⁻¹ KC1, 0.09mmoll⁻¹ CaCl₂ and 0.1 mmol 1⁻¹ MgSO₄) made from dehydrated cereal leaves (Sigma) dissolved at 0.1% (w/w) in water. Specimens of medium size (approximately 60 in cell body length) were washed with the standard saline solution, which contained 1 mmol 1⁻¹ KC1, 1 mmol 1⁻¹ CaC1₂ and 10 mmol 1⁻¹ Tris-maleate buffer (pH 7.0), and were kept immersed in the solution for about 10 μm prior to experimentation.

The tip of a microneedle (10 μm in diameter) mounted on a piezoelectric phonocartridge was placed in gentle contact with the cell surface with the aid of a micromanipulator. The tip was driven perpendicular to the cell surface by activating the piezoelectric phonocartridge with an electric pulse.

The magnitude and rate of excursion of the tip were controlled by changing the voltage and the rate of change of the voltage of an electric current pulse applied to the cartridge. The excursion of the tip increased in proportion to the voltage of the pulse as shown in Fig. 1. The rate of excursion was calculated by dividing the degree of excursion by the time taken for the electric pulse to reach its final level. All the experiments were performed at room temperature (20-21°C). Further details of the experimental procedures will be described in the Results section.

Specimens of Vorticella contracted when the tip of the microneedle was pressed against them. Such a contraction is illustrated in Fig. 2, a photograph taken on moving film using a slit camera. Contractions were of an all-or-nothing type. The latency between a stimulus and the contraction was 1.7±0.2ms (mean±s.D., N=10). The maximum rate of contraction (calculated from the slope of the contraction curve) was 6.7±2.1cms⁻¹ (N=10). Contracted specimens slowly extended to their original length in 3.3±0.9s (N =10).

To determine local differences in the mechanosensitivity which initiates contraction, the microneedle was pushed against various portions of a specimen attached to the substratum, and the threshold excursion required to elicit a contraction was determined in each portion. Four groups of specimens with stalks of different lengths (80, 120, 160 and 200 μgm), obtained from the same culture, were examined.

Fig. 3 shows the relationship between the threshold excursion of the microneedle and the site of stimulation, which was expressed as the distance from the junction between the cell body and the stalk to the stimulation site (λ). The threshold was lowest and identical at all sites on the cell body (negative λ region) (the threshold at -20μm was not statistically different from that at -40μm). On the stalk, it was higher when the site of stimulation was farther from the cell body (λ was longer) (the threshold at 160μm was significantly higher than that at 40 μm; P<0.00l). The relationships between threshold and λ were not statistically different among the four different groups of specimens, irrespective of their stalk length.

A specimen was first detached from its substratum (a glass slide) and laid on a second glass slide. The middle portion of the stalk was then mechanically clamped by pressing it against the glass slide with a microneedle. A mechanical stimulus (with a supramaximal excursion) was applied to the clamped stalk to examine whether it evoked a contraction.

When Vorticella was stimulated in a region between the clamped site and the cell body (proximal region), the whole cell contracted (the cell body shrank and the stalk coiled). In contrast, when Vorticella was stimulated in a region between the clamped site and the distal end of the stalk (distal region) there was no resulting contraction. When the clamp was removed, stimulating the same distal region again evoked a contraction of the whole cell.

When the middle portion of the stalk of a specimen (detached from its substratum and laid on a glass slide) was pressed very firmly against the glass slide with a microneedle, the portion was crushed flat and the spasmoneme was cut. A mechanical stimulus was applied to the stalk either proximal or distal to the point of crush after the microneedle had been removed from the crushed portion. The cell body shrank and the proximal region of the stalk coiled, while the distal region of the stalk did not.

The spasmoneme often became detached from the myoneme while the sheath of the stalk remained attached to the cell body and the specimen swam away as a free-swimming form, leaving its stalk behind. Mechanical stimuli applied to the stalk of spasmoneme-detached specimens evoked only shrinkage of the cell body. The stalk did not exhibit coiling. However, the mechanical threshold of the stalk was not statistically different from that of normal specimens (spasmoneme-detached specimens, 0.86±0.16μm, N=5; normal, spasmoneme-attached specimens, 0.96±0.09 μm, N=5; λ=40μm).

It should be noted here that pressing the stalk against the glass slide in order to cut the spasmoneme did not evoke a contraction if care was taken not to agitate the cell body mechanically. Furthermore, drawing a small portion of the stalk into a glass pipette (approximately 7 μm inner diameter) by suction did not evoke a contraction, whereas drawing any portion of the cell body into the capillary in a similar way did.

To examine whether ion channel activity in the surface membrane was involved in Vorticella’s mechanoreception, the relationships between stimulus threshold and A were determined in solutions with different cationic compositions. As shown in Fig. 4, the relationship was not affected by raising the K⁺ concentration to 20 mmol 1”‘or raising the Ca²⁺ concentration to 10 mmol 1⁻¹. The relationship remained the same after Ca²⁺ had been removed from the external solution by addition of EGTA (5 mmol 1⁻¹ final concentration).

Pressing a microneedle against a specimen attached to the substratum by its stalk caused translocation and rotation of the cell as well as depression of the cell surface at the point of contact or a bending of the stalk. To eliminate the translocation and rotation, specimens were detached from their substratum and their cell body was fixed at the tip of a micropipette (3 μm internal diameter) by lowering the hydrostatic pressure inside the pipette. In these specimens, stimulating the cell body mechanically produced only a depression in the surface membrane, while stimulating the free stalk caused it to tilt relative to the cell body. Tilting the stalk evoked a cellular contraction similar to that caused by stimulating the cell body.

The excursion of the microneedle required to tilt the stalk just far enough to evoke a contraction (the threshold excursion) was determined while the position of the tip of the microneedle on the stalk was varied. The rate of excursion of the microneedle was adjusted so that the stalk tilted at a constant angular velocity, ω, irrespective of the position of the tip. The threshold excursion was plotted against the position of the tip on the stalk, which was expressed as the distance from the junction between the cell body and the stalk to the tip. Fig. 5 shows two sets of plots obtained in two series of experiments with different values of ω (17.5 rad ms⁻¹ and 30.7 rad ms⁻¹). The threshold excursion increased in proportion to the distance, and it was significantly higher when ω was lower (P<0.01 for the thresholds at 30 and 40 μm). The slope of the plot corresponds to the threshold degree of tilting, θ_t, which is expressed as the angle between the stalk and the longitudinal axis of the cell body. θ_t was lower when ω was higher (0.036 rad when co was 30.7 rad ms⁻¹ and 0.062 rad when co was 17.5 rad ms⁻¹).

In the next series of experiments, the relationship between θ_t and ω was examined more precisely. Fig. 6 shows a plot of θ_t against ω, both on logarithmic axes. The relationship was linear with a slope of −1. This indicates that θ_t is inversely proportional to co. The proportionality constant was calculated from the intersection of the plot with the threshold axis (ordinate) to be 1.09 rad² ms⁻¹.

In this series of experiments, the microneedle was pushed against the middle portion of the cell body of a detached and fixed specimen, and the degree of depression just sufficient to evoke a contraction (the threshold depression), x_t, was determined while the rate of depression, dx/dt, was varied.

Fig. 7 shows a plot of x_t against dx/dt, both on logarithmic axes. The plot was linear with a slope of −1 for values of dx/dt in the range 8-40 μm ms⁻¹. This indicates that x_t is inversely proportional to dx/dt in this range. The proportionality constant was calculated from the intersection of the plot with the threshold axis (ordinate) to be 11.8_Jum²ms”¹.

The deviation from linearity at the lowest value of dx/dt (4μmms⁻¹) (lower than the value expected from the line) is attributable to the limitation of the excursion of the microneedle. As shown in Fig. 1, the maximum excursion of the microneedle is 2.7 μm (at 50 V). This value is close to the threshold depression at 4μmms⁻¹. Some specimens, therefore, contracted in response to the depression (and some responded to a lesser depression) while others did not respond. Thus, the mean threshold value for the responding specimens was lower than the true value. The deviation from linearity at the highest dx/dt (50 μmms⁻’) (higher than the value expected from the line) might be attributable to the limitation of the rapid movement of the microneedle (due to the overall mass of the phonocartridge and the microneedle).

Pushing of the tip of a microneedle against the stalk of Vorticella evoked a contraction. The threshold excursion of the microneedle required was larger in regions farther from the cell body and smaller in the cell body itself (Figs 3, 4). However, stimulating the stalk did not evoke a contraction when the stalk was kept mechanically clamped in a region between the site of stimulation and the cell body. Moreover, pressing the stalk against a glass slide using a microneedle, or quickly sucking a portion of the stalk into a micropipette, did not evoke a contraction, whereas similar suction of a portion of the cell body always evoked a contraction. These observations imply that the cell body has a mechanoreceptor mechanism for evoking a contraction, while the stalk has not. We therefore presume that stimulating the stalk exerts a mechanical effect on the mechanosensitive cell body, thus evoking a cellular contraction, and that the higher mechanical threshold in regions of the stalk farther from the cell body is attributable to a decrement of the mechanical effect according to the distance from the site of stimulation to the cell body. Machemer-Röhnisch and Machemer (1984) reported a similar decrement in the efficiency with which mechanical stimuli applied to the caudal cilia of Paramecium could activate mechanosensitive ion channels in the posterior membrane.

Stimulating a stalk with an experimentally cut spasmoneme evoked a contraction. This implies that continuity of the spasmoneme in the stalk is not necessary for the transfer of the mechanical effect to the cell body. The sheath of the stalk seems to be responsible for the transfer.

Activation of the myoneme to shrink the cell body and activation of the spasmoneme to coil the stalk are dependent on Ca²⁺ (Hoffmann-Berling, 1958; Amos, 1972a, 1975; Amos et al. 1975, 1976; Ochiai et al. 1983). We found that the mechanical threshold for evoking a contraction was not affected by removal of Ca²⁺ from the external solution (Fig. 4). This is consistent with the finding of Allen (1973) that Vorticella could contract even in a Ca²⁺-deficient solution (see also Shiono and Naitoh, 1978). Ca²⁺, therefore, must be liberated from some intracellular storage sites in response to a mechanical stimulus (Carasso and Favard, 1966; Amos, 1972b). Some previous workers (Moreton and Amos, 1979; Shiono et al. 1980) proposed that an influx of Ca²⁺, accompanied by a Ca²⁺ spike generated in response to a mechanical stimulus, is responsible for evoking contraction in vorticellid protozoans. However, we found that the mechanical threshold for evoking a cellular contraction was largely unaffected by a large increase in external K⁺ concentration (Fig. 4), which brought about a large depolarization of the membrane and thus diminished the Ca²⁺ spike (Shiono and Naitoh, 1978). It is therefore highly probable that activation of Ca²⁺ channels in the surface membrane is not directly involved in the mechanoreception in Vorticella that evokes a contraction.

The stalks of specimens with experimentally cut spasmonemes coiled only in the region between the cell body and the cut portion. Moreover, coiling of the stalk was not observed in specimens in which the spasmoneme had been detached from the myoneme at the junction between the cell body and the stalk. These observations indicate that the connection between the spasmoneme and the myoneme is required for initiation of coiling and that continuity of the spasmoneme in the stalk is needed for propagation of coiling along the stalk.

Sugi (1959, 1960) found that coiling of the stalk of Carchesium sp., a relative of Vorticella, started at the junction between the cell body and the stalk and propagated down the stalk. The coiling propagated beyond a mechanically clamped region (see also Jones et al. 1970). We confirmed these findings. This suggests that an increase in Ca²⁺ concentration around the spasmoneme occurs first at the myoneme-spasmoneme junction, then spreads down the spasmoneme. We found that the stalk itself was mechanically insensitive. It is, therefore, not feasible that the sequential liberation of Ca²⁺ from storage sites, such as membrane-bound tubules along the spasmoneme (Carasso and Favard, 1966; Amos, 1972b; Allen, 1973), is activated by their consecutive mechanical stretching. Instead Ca²⁺-induced Ca²⁺ release from the sites (Endo et al. 1970) is the most probable mechanism by which Ca²⁺ release is propagated along the spasmoneme. It should be noted that the initiation of a contraction of Vorticella by injection of an inward current requires the presence of external Ca²⁺ (Shiono and Naitoh, 1978). Moreover, Jones et al. (1970) reported that coiling initiated in the stalk by injection of electric current propagated towards both the distal and the proximal ends of the stalk. Ca²⁺ driven into the cell body by the current might induce a release of Ca²⁺ from Ca²⁺ storage sites.

Our quantitative examination of detached specimens and those with fixed cell bodies revealed that the threshold degree of tilting of the stalk required to evoke a contraction (θ_t) was inversely proportional to ω (Fig. 6). This means that the product θ_tω is a constant. In other words, a contraction occurs whenever the product exceeds a certain threshold value (the threshold product), irrespective of the value of ω. The threshold product corresponds to the proportionality constant (1.09 rad² ms⁻¹).

It is presumed that tilting of the stalk causes a localized depression and/or elevation of the surface membrane around the junction between the cell body and the stalk, which activates a hypothetical mechanoreceptor mechanism in the membrane and evokes a contraction. As shown schematically in Fig. 8, one edge of the stalk (A) produces a depression in the surface membrane when the stalk pivots around the opposite edge (fi). When the stalk pivots around a point between A and B (such as C), edge A produces a depression, while edge B produces an elevation. When the stalk pivots around A, edge B produces an elevation. For simplicity, we will consider here only the case where the stalk pivots around B. The size of the depression x(θ) caused by a tilt ((θrad) can be expressed as:

We also examined the relationship between the threshold extent of a localized depression of the surface membrane caused by pushing the surface and the rate of depression required to evoke a contraction. As shown in Fig. 7, the threshold extent of depression, x_t, was inversely proportional to the rate of depression, dx/dt. This means that a contraction occurs when the product xdx/dt is above a certain threshold value (the threshold product; x_tdx/dt). This threshold is estimated to be 11.8μm²ms⁻¹, similar to that for x(θ)_tdx(μ)/dt (estimated from the threshold product 0_tm), which is the threshold for evoking a contraction by tilting the stalk (12.7 μm² ms⁻¹). This strongly supports the hypothesis that tilting causes a localized depression of the cell membrane, thus producing a contraction.

It is interesting to note that the threshold product x_tdx/dt [and also x(θ)_tdx(θ)/dt] has the dimensions of L²T⁻¹(L, length; T, time), which corresponds to that for the rate of change in the area. Pushing a microneedle against the cell surface stretches the inner membrane (and thus the endoplasmic reticulum) as well as the surface membrane, since the endoplasmic reticulum is present immediately beneath the surface membrane (Allen, 1973). It is, therefore, highly probable that the rate of expansion of the membrane is a major factor activating the hypothetical mechanosensory system responsible for evoking a contraction in Vorticella. The mechanism by which the rate of membrane expansion is detected remains unclear.

It is presumed that a contraction of Vorticella is a behavioural response which allows it to escape from a dangerous stimulus. Mechanical stimuli which produce more rapid expansion of the membrane may be more dangerous to Vorticella, because they might damage the cell. Indeed, Vorticella always contracted when subjected to a fast water current, whereas it rarely contracted even when its cell body was tilted at a sharp angle against the stalk by the pressure of a slow water current. The fast water current produces more rapid tilting of the cell body against the stalk, and this apparently causes a membrane expansion that is fast enough to activate the contractile mechanism. In contrast, the slow tilting of the cell body produced by a slow water current causes a membrane expansion that is too slow to activate the contractile mechanism.

This work was supported by a grant in Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan and a grant from the Mitsubishi Foundation to Y.N. We are grateful to Professor D. R. J. Macer for critically reading the manuscript and to Dr H. Horigami of Hosei University for giving us specimens of Vorticella.