Transition from immotile to motile flagella may involve a series of states, in which some of regulatory mechanisms underlying normal flagellar movement are working with others being still suppressed. To address ourselves to the study of starting transients of flagella, we analyzed flagellar movement of sea urchin sperm whose motility initiation had been retarded in an experimental solution, so that we could capture the instance at which individual spermatozoa began their flagellar beating. Initially straight and immotile flagella began to shiver at low amplitude, then propagated exclusively the principal bend (P bend), and finally started stable flagellar beating. The site of generation of the P bend in the P-bend propagating stage varied in position in the basal region up to 10 μm from the base, indicating that the ability of autonomous bend generation is not exclusively possessed by the very basal region but can be unmasked throughout a wider region when the reverse bend (R bend) is suppressed. The rate of change in the shear angle, the curvature of the R bend and the frequency and regularity of beating substantially increased upon transition from P-bend propagating to full-beating, while the propagation velocity of bends remained unchanged. These findings indicate that artificially delayed motility initiation may accompany sequential modification of the motile system and that mechanisms underlying flagellar motility can be analyzed separately under experimentally retarded conditions.
Ciliary and flagellar motility depends on a series of cooperative interactions between the various components within the axoneme. Starting and stopping transients of motility and the transition between different states of activity have been investigated to functionally dissect these interactions (Miller, and Brokaw, 1970; Rikmenspoel, 1978; Goldstein, 1979; Gibbons and Gibbons, 1980b; Baba and Mogami, 1987a). Sea-urchin sperm cells are stored in a quiescent state in the gonads, but upon release into seawater, flagellar motility is initiated, and a 50-fold activation of respiration occurs (Christen et al., 1982, 1983). Motility initiation involves the switching on of the sperm flagellar apparatus, which occurs simultaneously or successively, probably creating transient states of the axoneme with some mechanisms working but others being suppressed. Motility initiation, however, occurs instantaneously or so rapid that it is hard to capture such transients for analysis, when sperm are diluted in seawater (Clapper et al., 1985; Dorsten et al., 1997).
We developed a solution, which contains 100 mM K+, 2 mM EDTA and 900 mM glycine at pH 8.2, for use in the study of regulatory mechanisms of flagellar movement, which might be revealed during artificially delayed motility initiation. This solution, which will be referred to as KEG, substantially retards motility initiation in sea-urchin sperm cells as described later on. The idea of retardation by KEG is based on the facts that sea-urchin sperm remain quiescent in Na+-free seawater (Lee et al., 1983; Bibring et al., 1984; Schackmann et al., 1984; Gatti and Christen, 1985; Christen et al., 1986), and also immotile at elevated K+ concentrations (Christen et al., 1986; Bracho et al., 1997) and become active when the intracellular pH is elevated, e.g., by addition of Na+ or amine (Christen et al., 1982; Lee et al., 1983; Bibring et al., 1984; Lee, 1984a, b; Bracho et al., 1997), and that activation is, however, delayed in the presence of a variety of metal chelators including EDTA (Clapper et al., 1985; Christen et al., 1986). The KEG retardation of motility initiation first revealed distinct transient patterns of flagellar bending described in the present study. Transition events between these patterns were recorded on a high-speed video camera and were analyzed, although they occurred sporadically and within some ten milliseconds in individual cells.
MATERIALS AND METHODS
Dry sperm were obtained from the sea urchin, Hemicentrotus pulcherrimus, by injecting 0.55 M KCl into the body cavity after removal of the Aristotle's lantern. To get completely inactive sperm, the body fluid was discarded through the oral opening and then the remaining fluid, which otherwise oozed upon spawning, was removed by absorbent tissue. The dry sperm were stored in a refrigerator and used within 6 hours.
Artificial seawater (ASW), 450 mM NaCl, 10 mM KCl, 10 mM CaCl2, 25 mM MgCl2, 28 mM MgSO4, 10 mM Tris-HCl (pH 8.2), and KEG, 900 mM glycine, 60 mM KCl, 40 mM KOH and 2 mM EDTA (pH 8.2), were used. ASW and KEG contained 0.01% bovine serum albumin (Sigma) to prevent sperm flagella from attaching to the surface of cover glass.
Recording and analysis
Sperm diluted and mixed well in KEG or ASW in a test tube were immediately put in an observation chamber made of slide and cover glasses separated with 0.5-mm thick silicone rubber cut into a rectangular frame. Pictures just under the cover glass were recorded in a temperature-controlled box of 20°C, through an Olympus phase contrast microscope with a 20 × objective (BH2 with NH20, Olympus Co., Tokyo) using an NTSC-format video tape recorder (VTR, Victor SVHS HR-VX1, Victor Co., Tokyo) under illumination with xenon flash tubes driven by laboratory-made power controllers (Baba and Mogami, 1987b). Motile spermatozoa were discriminated from non-motile by comparing trajectories, which were made by means of image processing on a DIG98 image processor (Ditect Co., Tokyo) interfaced with a computer (PC). The average number of motile cells of four successive measurements at 5-s intervals was plotted as a percent motility relative to the maximum reached during individual episodes after dilution. Spermatozoa which propagate the principal bend (P bend) and the reverse bend (R bend) alternately in a manner usually found in those diluted in seawater were designated as “full-beating” ones, while spermatozoa propagating exclusively the P bend were referred to as “P-bend propagating” ones (for definition of bends, see Gibbons and Gibbons, 1972). The percentage of full-beating sperm and that of P-bend propagating relative to the motile were also measured respectively as above.
To evaluate bending forms quantitatively, sperm diluted in KEG or in ASW in a test tube were immediately put on a Millipore filter of 13 mm in diameter (HA, Nihon Millipore Ltd., Tokyo) with a hole of 5.5 mm punched through at the center. The layer of solution trapped in the central hole was made as thin as possible to enable us to bring the transient and hence otherwise elusive events into focus. This ring-shaped disk was placed at the center of a cover glass, which was then inverted and supported on a slide glass by a frame of 0.5-mm thick silicone rubber, which in turn formed the walls of a humidity chamber. Pictures of sperm just starting movements in the hanging drop flattened by the surface tension were recorded at room temperatures, through an Olympus phase contrast microscope with a 40 × objective (BH2 with NH40) using VTR or a high-speed video recorder (HSVTR, EktaPro HR1000, Kodak, Tokyo) under illumination with a power xenon flash tube (Miyake et al., 1998). Individual flagellar images captured through the DIG98 image processor from VTR or HSVTR were tracked manually or automatically using flagellar image analysis software, Bohboh (BohbohSoft, Tokyo), on a PC. The angle of the tangent to the flagellar shaft measured with reference to the direction of the axis of sperm head, which will be referred to as shear angle, and the curvature of the shaft were computed as functions of time t and distance s along the flagellum measured from the base (Baba and Mogami, 1985). Based on the description of flagellar bending waves as “sine-generated” by Hiramoto and Baba (Hiramoto and Baba, 1978), the shear angle φ (s, t) was fitted using equations of the form:Bevington and Robinson, 1992). The phase Sφ (s) was further fitted using equations of the form:
To illustrate characteristic features of motility initiation, images were captured, processed to enhance contrast and superimposed by choosing the lightest one for individual pixels using a NEXUS 9000 image processor (Nexus Co., Tokyo).
Retardation of motility initiation
An increasing number of spermatozoa became motile after dilution, but substantially slower in KEG than in ASW (Fig. 1). The retardation effect of KEG was about 6 min when measured as the shift in the time after dilution TM50, at which 50% of spermatozoa became motile (Fig. 1A). The retardation to this degree enabled us to observe the transients of initiation and to discriminate stages of different bending patterns (Fig. 2). Figure 1B shows that the emergence of spermatozoon at the final stage full-beating in KEG delays about 10 min later than in ASW, when measured as the shift in the TF50, at which 50% of spermatozoa are in the full-beating stage.
Progressive stages of initiation
At the very beginning of initiation transients, a number of spermatozoa in KEG were immotile with a straight or slightly bent form (immotile stage). An increasing number of them soon shivered at low amplitude (shivering stage, Fig. 2A, B). The spermatozoon once in the shivering stage was usually found to enter into the next stage, in which its flagellum exclusively propagated a highly asymmetrical bend (Fig. 2C), i.e. convex to one side, instead of alternating bends to either side as found in swimming spermatozoa in seawater. This one-side bend will be called the P bend (principal bend), because it had the same polarity as the P bend of full-beating spermatozoa as it turned out by follow-up VTR. Thus, the third stage is designated the P-bend propagating stage as stated earlier. The spermatozoon then entered into the full-beating stage (Fig. 2D), in which its flagellar beating has the same characteristic features as ordinary ones found in ASW. Individual spermatozoa experience these stages with a sporadic and occasionally reversible transition, making a good number of P-bend propagating sperm and an increasing number of full-beating sperm throughout initiation transients in KEG. In ASW the initiation transients are so rapid that most of beating sperm are identified as full-beating sperm although a small number of irregular sperm remained throughout.
Most of immotile spermatozoa were floating some distance apart from and at an angle to the surface of either the slide or cover so that it was more difficult to get clear images of the initial movement in the shivering stage than those in other later stages, in which they were progressively trapped or guided by the surface with their bending planes being parallel to it. The low amplitude and irregularity in the movement also made it difficult to analyze the flagellar movement in this stage.
A series of events, in which one spermatozoon experienced the C-shaped bending in the shivering stage and then transition to the P-bend propagating stage, were recorded in good focus. This valuable record showed that either bending or unbending occurred simultaneously along the entire length of the flagellum, i.e., the bending wave was “standing” (Fig. 2A, B). The length of period of one bending cycle varied with an occasional long pause at either a bent or straight position. The flagellum bent only in one direction, mostly uniformly but a sharper bend at the base. The direction of bending was opposite to that of the P bend in the following P-bend propagating stage and hence the bend was identified as R bend.
P-bend propagating stage
The curvature of the P bend of the sperm flagella in the P-bend propagating stage was only slightly smaller than that in the full-beating stage in KEG, which was yet small when compared with that in ASW, while the curvature of the R bend in the P-bend propagating stage was substantially smaller than that of the full-beating stage (Fig. 3, Table 1). The P bend was generated at the base and propagated along the flagellum at a speed, not much different from that in the full-beating stage (Figs. 4, 5, Table 1). The generation of the P bend fluctuated in time and site as compared with that of bends in the full-beating stage either in KEG or in ASW (Fig. 5); the time and site of bend generation can clearly be expressed by islands and horizontal bands of warmer-color in pseudocolor maps as shown in Fig. 6. The site of bend generation varies along the basal region up to about 10 μm apart from the base (Fig. 6C) and the beginning of propagation is often delayed after the bend is generated (Fig. 6D), whereas both the site and the time are relatively regular in the full-beating stage (Fig. 7E, F). The P bend occasionally grew at lower speeds, making the start of propagation rather sporadic in time in the P-bend propagating stage than in the full-beating stage (Figs. 5, 6). However, once the propagation of the P bend occurred, it was as smooth as that in the full-beating stage of either KEG or ASW (compare Fig. 6A, B with Fig. 6E, F). The propagation velocity of this bend was also as large as that in the full-beating stage (Table 1). The length of period of one beating cycle varied so much that the beat frequency could not be determined stroboscopically. The fluctuation of beating cycles was also clearly demonstrated by a frame-by-frame analysis (Figs. 5, 6). The swimming velocity in this stage was about 75% of that in the full-beating stage (Table 1). The bending of the flagellum of this stage was still sinusoidal as in the full-beating stage, in terms of shear angle as a function of time (Fig. 7), i.e. the wave is approximately “sine-generated” (Hiramoto and Baba, 1978). Thus, the oscillation was considerably symmetric in terms of shear angle when the bias or offset in angle was ignored, and except for irregular pauses at the end of P-bend generating angular change (rising phase in Fig. 7). The rate of angular change in either the P-bend generating direction (rising) or the R-bend generating (falling) was significantly smaller than that of the full-beating stage (Fig. 7, Table 1). The bending wave damped along the flagellum when measured by the amplitude of shear angle (Fig. 7), while there was no apparent damping in terms of curvature (Figs. 3, 5).
Characteristics of motile spermatozoa after dilution in KEG or ASW.
The spermatozoa diluted in KEG swam regularly in the full-beating stage with an appearance similar to those activated by dilution in ASW (Figs. 2D, 4C, 4D, 5, 6), although with a little lower beat frequency and lower swimming velocity (Table 1). The bending waves in the full-beating stage of spermatozoa in KEG as well as in ASW were more exactly “sine-generated” than those in the P-bend propagating stage in terms of either curvature (Fig. 5) or shear angle (Fig. 7). The P and R bends were generated alternately at a regular position very close to the base at regular time intervals (Fig. 6E, F) as compared with the P bend in the P-bend propagating stage, which was often generated at positions substantially distant from the base at rather irregular intervals (Fig. 6A–D). The waves propagated without any discernible damping as of the P bend in the P-bend propagating stage in terms of shear angle (Fig. 7).
Transition from P-bend propagating to full-beating
The spermatozoon in the late P-bend propagating stage gradually developed a larger R bend at the base, and eventually entered into the next stage. The images showing the intermediate waveforms between the P-bend propagating and full-beating stages were seen only in a couple of fields of VTR, indicating that the transition is very momentary. The transition occurred in a short time, within one or two intermediate beating cycles (Fig. 6B). The parameters of bending changed from characteristic values of the P-bend propagating stage to those of the full-beating stage, which were described earlier under the headings “P-bend propagating stage” and “full-beating stage” and summarized in Table 1.
Diluting dry sperm of the sea urchin, Hemicentrotus pulcherrimus, with KEG instead of ASW delayed the acquisition of motility. This retardation enabled us to observe how individual spermatozoa began their flagellar beating under artificial initiation conditions. The initiation could be divided into four stages: immotile, shivering, P-bend propagating and full-beating. In the course of natural initiation in ASW, these transitional stages have not yet been confirmed most likely because the initiation can proceed to the final full-beating stage before sperm reach the field of observation or because it proceeds through a completely different course of motility acquisition.
Sperm cells are stored in a quiescent state in the gonads, but upon release motility is activated. Activation is triggered by environmental cues through transduction events involving sperm ion channels (Morisawa, 1994; Darszon et al., 1999, 2001). In sea urchins, due to Na+/H+ exchange across the sperm plasma membrane, release into seawater is associated with an increase in intracellular pH (pHi), which, in turn, activates flagellar dynein ATPase (Christen et al., 1982; Schackmann et al., 1984; Lee, 1984a, b). The pHi is about 6.6 in the quiescent state (Lee, 1984a; Dorsten et al., 1997) and shifts to 7.6 by H+ extrusion through the Na+/H+ exchanger upon activation (Dorsten et al., 1997). The removal of Na+ in ASW inhibits the sperm motility even if the pH of ASW is as high as the normal level, while the addition of amine can reactivate the motility (Bibring et al., 1984). Since Na+ is not included in KEG, an increase of pHi by the Na+/H+ exchange might not occur. It is therefore likely that glycine in KEG enters the cell through anion channels (Darszon et al., 2001) rather than through the Na+/H+, elevates the pHi by neutralization of H+ in the cell, and triggers the motility as amine does so (Bibring et al., 1984), which is slower than the process by the Na+/H+ exchange in ASW under normal conditions.
The dynein ATPase utilizes most of the ATP, which in sea-urchin spermatozoa is delivered exclusively by oxidative phosphorylation in a single mitochondrion located in the midpiece of the sperm cell and is transported most efficiently through the sperm creatine kinase and phosphocreatine system known as the CK/PCr shuttle (Christen et al., 1983). When the CK/PCr shuttle is blocked by specific inhibitors, the flagellar bending waves are rapidly damped as they propagate along the flagellum (Tombes and Shapiro, 1985; Tombes et al., 1987). The fact that there are no significant attenuations in bending waves along the flagellum in KEG indicates that the CK/PCr shuttle is functional during the retarded initiation. The fact that the bending waves in KEG, of either P-bend propagating or full-beating sperm, have no resemblance with those reported for demembranated sea urchin sperm reactivated at low ATP concentrations, i.e. lower beat frequency and incomplete bend propagation (Gibbons and Gibbons, 1972), also implies an adequate supply of ATP.
It has been shown that ciliary and flagellar motility are regulated by phosphorylation of the axonemal components. With the known exception of carp sperm (Krasznai et al., 2000), phosphorylation on some axonemal proteins by cAMP-dependent kinase is required for initiation of sperm motility including that of sea urchins (Ishiguro et al., 1982; Brokaw, 1984, 1987a; Takahashi et al., 1985; Murofushi et al., 1986; Bracho et al., 1998). It should be noted that cAMP is elevated by 1.9 fold upon dilution from low-pH ASW to Na+-free, Ca2+-free ASW added with amine (Beltrán et al., 1996), conditions similar to the present. However, the reactivation conditions allow variable levels of motility to be expressed without cAMP incubation, with some species, including Hemicentrotus, being already at a high percent motility (Murofushi et al., 1986; Brokaw, 1987a; Bracho et al., 1997).
Demembranated spermatozoa from the tunicate, Ciona intestinalis, are immotile in an ordinary “reactivation medium”, which has been normally used for sea urchin sperm, and become activated only after incubation with cAMP (Brokaw, 1982, 1987b). Four bending patterns, Patterns C, Q, D and P, appear progressively in the course of cAMP-dependent phosphorylation or reversible dephosphorylation of the axoneme of Ciona, and are also found under modified conditions in the sea urchin, Lytechinus pictus (Brokaw, 1987b). These patterns have counterparts in the present study. Pattern C appears similar to the C-shaped bending and unbending found in the shivering stage. Pattern D is an asymmetric bending pattern and thus appears similar to the bending pattern found in the P-bend propagating stage. There are, however, some significant differences: in Pattern D the bends are damped along the flagellum while they maintain the amplitude unattenuated in the P-bend propagating stage; Pattern D accompanies a significant reduction in both wavelength and bend angle, with active bends in both directions at least in the proximal region, while in the P-bend propagating stage the P bend has the same characteristic features as the normal beating but with the R bend being almost totally collapsed (Table 1). The limited motility in the P-bend propagating stage therefore cannot be due to an incomplete phosphorylation of axonemal components. Pattern P is essentially the same as that of the full-beating stage.
It has been shown that an elevated concentration of intracellular Ca2+ induces asymmetric beating and quiescence in sea-urchin sperm (Brokaw, 1979; Gibbons and Gibbons, 1980a; Gibbons, 1986). The asymmetrical beating in the P-bend propagating stage resembles to this Ca2+-induced asymmetry in that the R bend is reduced in amplitude (Brokaw, 1979; Gibbons and Gibbons, 1980a). There are however substantial differences between these asymmetrical beatings from the following viewpoints: the Ca2+-induced asymmetry does not occur in the Ca2+-free medium as used in the present study (Gibbons and Gibbons, 1980b; Shingyoji and Takahashi, 1995) nor does the intracellular Ca2+ concentration elevate when sperm are diluted in Na+-free, Ca2+-free ASW (Beltrán et al., 1996); the asymmetry in the P-bend propagating stage is due to the selective elimination of the R bend as most clearly shown in Fig. 3 while the Ca2+-induced asymmetry is due to an increase in the P bend compensated by a decrease in the R bend, so that there is no change in mean bend angle, as most nicely explained by “biased baseline” mechanism (Brokaw, 1979; Eshel and Brokaw, 1987).
It has been demonstrated that both live and reactivated sea urchin sperm are straightened at low pH and begin to beat asymmetrically, as the pH of medium is increased (Goldstein, 1979). The bending pattern of this asymmetric motion during pH rise is very similar to that in the P-bend propagating stage during motility initiation. It has been reported that the P bend of the asymmetric beating of live sperm at intermediate pH attains a speed of about one-third of that attained during steady state beating, while the initial bend, which could not be identified as P or R, of reactivated sperm often attains a speed of propagation similar to those of steady-state bends at pH 8.0; the beat frequency of reactivated sperm at intermediate pH is about one-third of that at pH 8.0, while it has not been measured with live sperm (Goldstein, 1979). Thus, our data that there is no significant difference in speed of the P bend between the P-bend propagating and full-beating stage irrespective of irregularity of and reduction in beat frequency (Table 1) may represent more accurately the nature of bending, probably at different pH. The R bend begins to develop later as the pH is increased further (Goldstein, 1979), which may corresponds to the transition from the P-bend propagating to full-beating stage. The asymmetrical beating of this type has also been found during the stopping transients upon a rapid drop in the pH of reactivation medium (Sato et al., 1988). It is therefore likely that P-bend propagating bends are generated under intermediate conditions made in the course of gradual increase in the intracellular pH. It has been demonstrated that one of potentially alternating bends propagates with the other being collapsed during the trypsin digestion in the presence of CO2 (Brokaw and Simonick, 1977), although the propagating bend has not been identified as P or R. These findings may indicate that the axonemal structures including dynein ATPase have subtle differences in sensitivity to their environmental pH or CO2, resulting in selective inhibition of specific sliding between adjacent doublets required for bend generation or propagation, since the P and R bends are topographically related to the axonemal structure (Mohri et al., 1987).
The site of generation of the P bend in the P-bend propagating stage varies in position in the basal region up to 10 μm, making islands or hot spots clearly some distance apart from the base (Fig. 6), which has not been demonstrated with sperm flagella in the full-beating stage nor under other experimental conditions. Goldstein has reported that the development of the first bend near the base (P bend) upon pH rise is accompanied by the formation of a broad, more distal, curve in the opposite direction (R bend) and that this curve occasionally develops into a traveling bend (Fig. 4 of Goldstein, 1979); similar starting transients have been described by tracings of one spermatozoon at normal pH (Fig. 11 of Rikmenspoel, 1978). Thus, our findings indicate more evidently that the ability of autonomous P bend generation at and propagation from a portion of the axoneme is not exclusively possessed by the very basal region but can be unmasked throughout a wider region when the R bend is suppressed.
In conclusion, sea-urchin sperm can selectively propagate the P bend under certain conditions, most likely at low pH. Analyses for components responding to these modified beating remained to be studied.