INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The eyes of unicellular flagellate algae contain one or more rhodopsin-type photoreceptors, which, upon light stimulation, evoke electrical inward currents within the eyespot area (photoreceptor currents, I_P) (Sineshchekov and Govorunova, 1999). These inward currents were recorded from Hematococcus pluvialis wild-type cells (Litvin et al. 1978) or from a cell wall-deficient Chlamydomonas reinhardtii mutant (Harz and Hegemann, 1991). In C. reinhardtii the small photoreceptor current I_P, saturated at low light, induces an asymmetrical change of the flagellar beating (Holland and Hegemann, unpublished). Because of the directional properties of the eye, the transient beat asymmetry occurs in such a way that the cells preferentially orient toward or away from the light source (Foster and Smyth, 1980; Rüffer and Nultsch, 1990). When the flash photon exposure exceeds a certain threshold, I_P is superimposed by a fast all-or-none flagellar current, I_FF, followed by a slow and small current-tail <1 pA, I_FS. It is clear that both flagellar current components are carried by Ca²⁺, and that Ca²⁺ down-regulates I_FS from inside (Holland et al., 1996). I_FF is the current that causes the switch of the flagellar waveform from forward swimming to undulation (stop response or photophobic response), as shown by simultaneous recording of photocurrents and flagellar beating (Holland et al., 1997). The ion specificity of the photoreceptor current I_P and its signaling properties are less clear by far. I_P depends on extracellular Ca²⁺, and in the early experiments on C. reinhardtii, I_P had disappeared completely at <0.1 µM Ca²⁺ (Harz and Hegemann, 1991). In other experiments on H. pluvialis and C. reinhardtii, 15 to 30% of the I_P persisted at Ca²⁺ levels below 1 nM (Sineshchekov, 1991; Holland et al., 1996). It could not be unequivocally clarified which ion carries the residual current. A supposedly Ca²⁺-independent portion, I_P1a, and a Ca²⁺-dependent component, I_P1b, have been discriminated since then (I_P = I_P1a + I_P1b) (Sineshchekov, 1991; Holland et al., 1996).

Because the more Ca²⁺-sensitive component I_P1b dominates the photocurrent I_P1 at high flash energies under otherwise physiological conditions, I_P1b was studied in more detail than I_P1a (Sineshchekov and Govorunova, 1999). I_P1b saturates only at very high flash intensities, it rises with a delay of <30 µs, and it peaks at ~1 ms after the flash. The short delay and the high saturation level originally lead to the suggestion that I_P1b is indicative of a charge redistribution within the rhodopsin photoreceptor and does not reflect translocation of ions, as in the early receptor potential of animal vison (Sineshchekov et al., 1990). This hypothesis, however, did not hold true for C. reinhardtii where I_P1b was assigned to a real inward current, for the following reasons. First, estimating 10 000 rhodopsin molecules per cell, ~100 charges will be displaced across the membrane per rhodopsin molecule (Harz et al., 1992). Second, in experiments with double flashes of moderate energy, the second flash could not evoke a photoreceptor current (Holland et al., 1996). To explain the short delay between flash and photocurrent rise we have proposed that the light-induced conductance for I_P1b is either closely linked to rhodopsin or is a light-activated form of rhodopsin itself (Harz et al., 1992; Holland et al., 1996).

Recording photoelectric responses in cell suspension and suction experiments with improved time resolution allowed the analysis of I_P1 in response to dim flashes in C. reinhardtii and Volvox carteri (Sineshchekov et al., 1992; Holland et al., 1996; Braun and Hegemann, 1999a). At these low-light levels, I_P1 is dominated by the less Ca²⁺-sensitive component I_P1a (formerly named P_S). The delay time between flash and beginning of the photocurrent increased to several milliseconds and the flash-to-peak time was accordingly extended up to 10 ms (Braun and Hegemann, 1999a). Extended delays and flash to peak times are especially evident in V. carteri, because all electrical responses are generally slower and I_P1a is more prominent in this colonial species as compared with its unicellular relatives. The observations were interpreted in terms of a second ionic process involving biochemical signal transduction processes.

Here we extend the characterization of the photoreceptor currents I_P1a and I_P1b in C. reinhardtii. In addition, we report on a new acid-induced photoreceptor current, I_P2. This current is slowly activated and appears with a large amplitude at acidic pH in response to brief flashes or, as a large stationary current, upon step-up stimulation.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

All experiments were carried out with the C. reinhardtii strain CW2 cells. Cells were converted into gametes in N-methyl-D-glucamine (NMG)⁺/K⁺-buffer (5 mM Hepes adjusted to pH 6.8 with NMG, 200 µM Ca²⁺, 100 µM K⁺, and 10 µM BAPTA) overnight.

Electrical measurements

CW2 cells were dark-adapted for more than 1 h before they were used for experiments. The NMG⁺/K⁺-buffer was used as electrode and bath solution. Ca²⁺ was added as CaCl₂. The total amount of Ca²⁺ required for a defined concentration of free Ca²⁺ was calculated according to Holland et al. (1996).

Ca²⁺-free conditions were accomplished by using 100 µM KPi-buffer. The buffer was prepared from bi-deionized H₂O finally poured through a strong acid cation exchange resin (Dowex-50 W, Sigma, St. Louis, MO). The buffer was adjusted to pH 6.8, the Ca²⁺ concentration was measured using Fura-2 (Molecular Probes, Eugene, OR), and the buffer was finally adjusted to the desired pH by titration with ultrapure HCl (Merck, Darmstadt, Germany). The gametes were washed three times in KPi-buffer and stored in the dark. Because of the low ionic strength of the KPi-buffer, electrodes were kept in 9 mM KCl and separated by salt bridges as illustrated in Fig. 1. The fused silica capillary used for the pipette electrode had a diameter of 300 µm (o.d., TSP 180350, Composite Metal Services, Worcester, UK).

View larger version (25K): [in this window] [in a new window]   FIGURE 1   Experimental setup for current recordings in suction-pipette configuration with low ( 15 nM), external [Ca2+].

Suction pipette measurements

Patch pipette measurements

2

2

1

Action spectra

Action spectra for I_P1b and I_P2b were obtained by evaluating data from at least three different cells (Harz and Hegemann, 1991). The linear part of the stimulus-response curves (peak current vs. log Q) was extrapolated to zero. The absolute light sensitivities (1/threshold photon exposure) as determined for six different wavelengths were normalized to the 510-nm value and plotted versus wavelengths.

Reaction scheme

The numerical values given here refer to Rh_b There are 10⁴ Rh molecules in the eye (Beckmann and Hegemann, 1991). In the unilluminated eye all Rhs are in the inactive ground state I. Upon illumination with L photons per Rh in a 10-µs flash, part of I (L times  ·  = 10²⁰ · 0.7) will be converted to the first non-excited state K with the apparent rate constant L · k_IK (k_IK = 1000 s¹/L); K will fall back to the dark state I through a cascade of intermediate states (in further alphabetical order), L, M, N, O, and P, with the corresponding rate constants k_KL 200, k_LM 3000, k_MN 500, k_NO 100, k_OP 20, and k_PI 10 s¹. The state M permits 0.01 pA Ca²⁺ inward current per Rh/G_1b unit, with 0.01 mM half-saturating [Ca²⁺]_o, and the state O a H⁺ entry by a constant field H⁺ conductance G_H of 2 pSmM¹.

After illumination, the currents through M and O will change the voltage in the area of the membrane enclosing the eye and will thus alter the driving forces in the equivalent circuit. These changes can be calculated by numerical integration.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Classification of flash-induced transient inward currents

For an overview and for reference purposes, Fig. 2 shows typical current tracings as recorded from C. reinhardtii upon stimulation with short light flashes. In the experiment of Fig. 2 a the cell was sucked into a suction pipette with both eye and flagella exposed to the bath medium, whereas in Fig. 2 b the eye was sucked into a patch pipette with small tip diameter and steep cone angle. The suction mode (a) allows the recording of the photoreceptor currents I_P and flagellar currents, I_F, (Litvin et al., 1978; Harz and Hegemann, 1991). Photoreceptor currents have been defined in the past as eyespot-localized currents, whereas flagellar currents are distributed along the whole flagella (Beck and Uhl, 1994). Under the configuration with the eyespot facing the bath compartment, all currents are only ~25% of the total (Holland et al., 1996). In contrast, patch pipette experiments with the eyespot in the pipette allow the recording of 80% of I_P, whereas the flagellar currents are generally much smaller (Fig. 2 b). The I_P sign inversion is attributable to the current vectors, i.e., either from the bath into the eye (a) or from the pipette into the eye (b), respectively.

View larger version (16K): [in this window] [in a new window]   FIGURE 2   Two representative current records in response to a saturating light flash under different measuring configurations. (a) Suction pipette measurement with eyespot and flagella exposed to the bath medium. (b) Patch pipette measurement with eyespot in the pipette.

In this report we will present data about a second photoreceptor current I_P2, which follows I_P1 when [H⁺]_o at the eye is high enough. Before presenting details of the novel event I_P2, we examine the hypothesis that the well-known current I_P1 consists of two distinct components. Photoreceptor currents recorded at a few selected photon exposures are shown in Fig. 3 a. In Fig. 3 b, the peak current of I_P1 is plotted versus the photon exposure of the actinic flash. This experimental relationship is well fitted by the sum of two hyperbolic Michaelis-Menten-type kinetic components (relative deviation from solid line: 0.6 pA):

(1)

with the saturating amplitudes I_Sa = 4.85 pA and I_Sb = 38.36 pA, and the half-saturating photon exposures Q_{1/2 a} = 0.14 and Q_{1/2 b} = 57.72 µE m² of the two respective components. The line in Fig. 3 c represents an alternative fit with the sum of two exponentials,

(2)

which resulted in an inferior approximation, especially for low photon exposures attributable to the stronger curvatures in exponentials compared with hyperbolas (relative deviation from the solid line: 1.3 pA).

View larger version (14K): [in this window] [in a new window]   FIGURE 3   Light dependence of the peak amplitude of IP1a+ IP1b. (a) Four representative traces recorded at 1:690; 2:54; 3:6.9; and 4:0.55 µE m2. (b and c) The peak amplitude is plotted vs. the photon exposure of the actinic light flash; lg/lg plot; points, data; lines, theoretical approximations (fits). (b) Solid line represents the sum of two Michaelis-Menten components: IP1 (Q) = ISa/(1 + Q/Q1/2a) + ISb/(1 + Q/Q1/2b  ) with fitted values of ISa = 4.85 pA Q1/2a = 0.14 µE m2, ISb = 38.36 pA, and Q1/2b = 57.72 µE m2. (c) Sum of two exponentials: IP1(Q) = Ia(1  exp(Q/Qa)) + Ib(1 exp(Q/Qb) with fitted values of Ia = 4.01 pA, Qa = 0.27 µE m2, Ib = 31.72 pA, and Qb = 80.93 µE m2

I_P2: Light-induced H⁺ entry

The novel phenomenon reported here is I_P2, a flash-induced, slow, transient, inward current through the eye region. I_P2 is virtually absent at pH 6.8 (Fig. 2) but is clearly visible at pH 4.0 (Fig. 4, a and b). The magnitude of I_P2 has been titrated with respect to the pH in the eye region. For this purpose we did not simply pick the peak of I_P2 which may be distorted by the flagellar currents I_FF between I_P1 and I_P2. Rather, we picked I_P2 100 ms after the flash, when I_FF has decayed. These values of I_P2,100ms have been related to the peak of I_P1, the latter being independent of pH. Fig. 4 c shows an example of a resulting Lineweaver-Burk plot of I_{P1, peak}/I_P2,100ms versus 1/[H⁺]. These straightforward Michaelis-Menten kinetics indicate that I_P2 is carried by H⁺. However, strictly speaking, such kinetics might also be brought about by H⁺ gating of a non-H⁺ conductance. Means from n = 7 independent cells are 1/K_m = 4.06 ± 0.26 mM¹ and (I_{P2,100 ms}/I_P1,peak)_max = 0.93 ± 0.22. Corresponding kinetic evidence for I_P1 being carried by Ca²⁺ is not available. Nevertheless, this strong kinetic evidence in Fig. 4 c is not absolutely compelling. There is still a possibility that H⁺ has only a catalytic effect on I_P2. Usually, this discrimination can be made in thermodynamic terms through the Nernst equation and reversal voltage. Unfortunately, this approach is not technically feasible in our system because the membrane voltage is not accessible.

View larger version (11K): [in this window] [in a new window]   FIGURE 4   (a and b) Transient inward currents through the eye of C. reinhardtii upon short light flashes, recorded under different pipette configurations with pHo 4 in the compartment of the eye and 200 µM Ca2+ in both compartments. (c) The influence of [H+]o on the second photocurrent, IP2, of a typical cell is shown in a Lineweaver-Burk plot; data were sampled 100 ms after the flash (not peak) and related to peak of IP1. Mean values of seven cells are discussed in the text.

Low [K⁺]_o promotes I_P2

Taking I_P2 for a proton current, I_H, led to the expectation that its size does not depend on the osmotic component, pH, of the pmf alone. The transmembrane voltage, V_m, as the electrical component of the pmf should have an equivalent influence. Unfortunately, corresponding V-clamp experiments can not be performed in C. reinhardtii at the moment. However, predominant K⁺ diffusion is common in biological membranes, including the plasma membrane of C. reinhardtii (Malhotra and Glass, 1995). Thus, we expected that [K⁺]_o influences V_m according to the Nernst equilibrium voltage for K⁺, E_K  (59 mV)lg([K⁺]_i/[K⁺]_o), and the H⁺-inward current I_P2.

These relationships were investigated by the experiments presented in Fig. 5. In the given configuration, the large membrane portion in the bath will guarantee a high impact of [K⁺]_o on the uniform, internally short-circuited V_m. This means that in the experiment depicted in Fig. 5 a, V_m was presumably rather negative because of the low [K⁺]_o in both compartments. Upon exposure of the predominant membrane portion in the bath to high [K⁺]_o, V_m was presumably also less negative in the membrane region within the pipette not directly exposed to high [K⁺]_o. This intended manipulation of V_m resulted in the expected reduction of I_P2  I_H. As seen from the graph in Fig. 5 b, 10 mM [K⁺]_o inhibits I_P2 by 50%, and >20 mM [K⁺]_o in the bath causes K⁺ to move from the bath compartment through the cell into the pipette (I_K in Fig. 5 a). The large I_P2 and the dependence on [K⁺]_o are consistent with the gating properties of the K⁺ outward-rectifying channels (Blatt and Gradmann, 1997) that seem to be ubiquitous in plant cells. A transient depolarization from a pump-dominated, negative resting voltage V_r (here  150 mV) will cause a transient activation of these channels which results, at 1 mM [K⁺]_o, in an accelerated repolarization toward E_K (Govorunova et al., 1997). At [K⁺]_o >10 mM, the repolarization is inhibited and the driving force for H⁺ flux at acidic pH is suppressed; the small outward current observed at pH 6.8, 50 ms after flash (Fig. 5 b) is not affected by high [K⁺] in the bath.

View larger version (18K): [in this window] [in a new window]   FIGURE 5   Influence of external K+ on early and late flash-induced inward currents, IP1 and IP2, through the eye. (a) Three representative traces at three different K+-concentrations in the bath medium. Note that 50 mM K+ leads to an inversion of the IP2 signal, whereas IP1 remains almost unchanged. (b) Plot of the current amplitudes 50 ms after the flash vs. the K+ concentration in the bath compartment (the lines are drawn to guide the eye).

It is noted that this effect of V_m on I_H is not specific; rather, it applies for the inward currents of any ion at physiological driving forces (V  E_X) < 0. Transcellular currents at asymmetric [K⁺]_o also accompany the large flagellar currents observed at high Ba²⁺ concentration (Nonnengäer et al., 1995).

I_P1 and I_P2 depend on [Ca²⁺]_o

The high light-saturating component of I_P1, I_P1b, has been known to be carried by Ca²⁺ (Sineshchekov, 1991; Holland et al., 1996), whereas the ionic nature of the smaller, low light-saturating component I_P1a was not yet known. To isolate the inward currents I_P2 from I_P1, we intended to record I_P2 at low [Ca²⁺]_o. At acidic pH, a defined [Ca²⁺]_o can not be adjusted in the usual way by buffering Ca²⁺ with common chelating agents because the carboxyl groups of EDTA or BAPTA are protonated at acidic pH. Therefore, the cells were repeatedly washed with ultrapure 100 µM KPi-buffer that contained ~1 nM Ca²⁺. Measurements were carried out in this low ionic-strength buffer, and the cell was decoupled from the Ag/AgCl electrodes by salt bridges as explained in Fig. 1. Under reference conditions (Fig. 6 a) with 200 µM [Ca²⁺]_o in phosphate buffer of pH_o = 6.8, I_P1 had its usual size. When cells of the same batch were tested in Ca²⁺-depleted phosphate buffer (15 nM final concentration), I_P1 was almost absent, indicating that not only I_P1b but also I_P1a is suppressed. Moreover, after acidification of the bath compartment toward pH 4.0, no I_P2 could be evoked either, which is a strong indication that all three photoreceptor current components are in some way Ca²⁺-sensitive. It should be pointed out that, compared with all former experiments, the bath medium in the experiment to Fig. 6, b and c was also Cl-depleted. But, addition of Cl did not restore I_P1 or I_P2.

View larger version (16K): [in this window] [in a new window]   FIGURE 6   Low extracellular Ca2+ suppresses IP1 and IP2. Measuring KPi-puffer including 200 µM Ca2+ (a) was replaced by deionized KPi with 15 nM Ca2+ and pH 6.8 (b) or pH 4.0 (c).

Light dependence of I_P2

The similar response of I_P1 and I_P2 to depletion of external Ca²⁺ (Fig. 6) may suggest coupling between these two events. An equivalently close relationship between these two currents exists with respect to their light dependence. The dependence of I_P1 on the photon exposure provided by individual flashes has already been presented in detail in Fig. 3. Fig. 7 shows this dependence for eyespot in conditions and pH 4.0 in the pipette. The peak amplitude of I_P2 throughout is ~20% of that of I_P1 at all flash-conditions tested. I_P1 and I_P2 increase in parallel with the flash photon exposure between 6.2 and 1250 µE m². Again, like the saturation curve for I_P1 at pH 6.8, both curves for I_P1 and I_P2 are too flat to be described by a single Michaelis-Menten or a single exponential saturation curve. Two Michaelis-Menten curves shown in Fig. 7 c describe the data for acidic conditions adequately.

View larger version (16K): [in this window] [in a new window]   FIGURE 7   Dependencies of amplitudes and temporal responses of IP1 and IP2 on photon exposure, Q, of the light flash. (a) Selected current recordings at three selected photon exposures (10% = 156 µE m2). (b) Action spectra for IP1b and IP2b constructed from stimulus response curves as described in Material and methods. The solid line shows a standard rhodopsin spectrum taken from Knowles and Dartnall (1977). (c) Peak amplitudes of IP1 and IP2 plotted vs. the flash photon exposure Q. The data for the grouped fit for photocurrents at acidic conditions are: Q1/2 a = 1.52 and Q1/2 b = 136.24 µE m2. The resulting amplitude values are: IP1a = 3.03 pA; IP1b = 31.32 pA; IP2a = 0.31 pA; and IP1b = 7.58 pA. (d) Plot of the reciprocal flash-to-peak-times (tftp1) vs. the flash photon exposure Q.

Using common Q_1/2a and Q_1/2b for grouped fits of I_P1(Q) and I_P2(Q) with Eq.(1) resulted in Q_1/2a = 1.52 and Q_1/2b = 136.24 µE m² for photocurrents at acidic conditions. This supports the close relationship between I_P1 and I_P2. The resulting amplitudes are: I_P1a, 3.03 pA; I_P1b, 31.32 pA; I_P2a, 0.31 pA; and I_P2b, 7.58 pA.

In addition, in both time courses, I_P1(t) and I_P2(t), the times t_P1 and t_P2 between light flash and current peak (flash-to-peak time, t_ftp) contract with increasing light intensity toward a discrete minimum (t_P1,min  1 ms, t_P2,min  3 ms) (Fig. 7 d). However, the I_P1 peak times of the a and b components overlap over wide range of photon exposure, whereas those of a and b of I_P2 are more clearly separated with a clear switch near 25 µE m² s¹. The change of the flash-to-peak-time is not trivial because in a simplistic model with defined signal transduction mechanism light affects only the prime reaction step of photoconversion and, in the following reactions, only the amplitudes, but not the time courses, are affected by the light intensity. Because the light sensitivity of I_P1(t) and I_P2(t) differs from this simplistic pattern in the same fashion, a close mechanistic relationship between I_P1 and I_P2 can be inferred (further discussion below).

Stimulus-response curves for the range between 10 and 90% saturation where I_P1b and I_P2b dominate the photocurrents were recorded for different wavelengths. The amplitudes were plotted versus log Q and extrapolated to zero. The sensitivity defined as 1/threshold was plotted versus the wavelength and the resulting action spectra for I_P1b and I_P2b at pH 4 are seen in Fig. 7 b. Both spectra are compared with a standard rhodopsin nomogram (Knowles and Dartnall, 1977). The shape and the peak at 500 nm are nearly identical to the earlier presented action spectrum for I_P1b recorded at neutral pH (Harz and Hegemann, 1991). The close agreement between the data and the nomogram leaves little doubt that the photoreceptor for I_P2b is also a rhodopsin.

Step responses

To inquire to what extent the photocurrents in continuous light are carried by H⁺, responses to step-up stimulation were analyzed at different pH. From responses to on- and off-steps of light we expected a distinction between transient and permanent components. For this purpose, light pulses of 300 ms were applied in a separate set of experiments. At pH_o 6.8 at the eye region, transient current peaks provisionally assigned as I_P1 and I_FF are followed by a small stationary current of 1 to 2 pA (Fig. 8 a). I_FF is an all-or-none response and it disappears when the flagella are chopped off (data not shown), which justifies its assignment as a fast flagellar current corresponding to I_FF in flash experiments. Stationary photocurrents were observed before in H. pluvialis (Sineshchekov, 1991), but because of the small amplitude they have never been investigated systematically.

View larger version (13K): [in this window] [in a new window]   FIGURE 8   pH- and intensity-dependence of temporal responses of IP upon light-on and -off. Currents traces were recorded in response to step-up and down stimulation at pH 6.8 (a) and pH 4.0 at the eye (b); 100% = 40 µE m2 s1.

Upon acidification of the eye region we expected additional current responses related to I_P2. Fig. 8 b shows such additional responses, which consist of a second transient, peaking at ~9 pA some 50 ms after light-on before it decays with   100 ms to a stationary level, I_SS, of ~7 pA under maximum (100%) step-illumination. Upon turning the light off, the current decays with only one predominant exponential (  200 ms). At lower intensities (<15% light) we noticed that the slow and transient peak I_P2 vanishes and the stationary level is apparently reached with one exponential (  120 ms). However, the light-off response now shows two exponential components, e.g., the 26% response in Fig. 8 b, a fast one (1/  25 s¹) with an amplitude A₁  1 pA and a slow one (1/  2.5 s¹) of a similar amplitude A₂  1 pA. Comparing the tracings in Fig. 8 b shows that the fast off-component (₁) becomes smaller with decreasing intensity of the preceding continuous illumination. Thus, we may conclude that at 100% light in Fig. 8 b the stationary current is mainly attributable to the high saturating component G_2b, whereas G_1a, G_2a, and G_1b significantly contribute at lower but more physiological intensities. Unfortunately, the rise of I_P2 is disturbed by the positive I_FF, precluding a more detailed analysis.

A systematic evaluation of these features from three experiments over an intensity range of more than two orders of magnitude is illustrated by Fig. 9. The two current transients I_P1 and I_P2 and the stationary current I_SS increase more or less in parallel with the intensity of the light (Fig. 9 b). Above 10 mE m² s¹ the new transient I_P2 is distinct from I_SS. Because I_P2 and I_SS are only visible or largely enhanced at low pH, we anticipate that both mainly reflect H⁺-carried I_P2 (a and b), discussed extensively above. The step-up-to-peak time, t₁, is shortened at higher light intensities. The peak I_P2 appears only at high irradiance, whereas two exponential components A₁exp(t/₁) and A₂exp(t/₂) in the relaxation of I_P(t) from I_SS to zero after light-off become most clearly visible after moderate irradiance  10 mE m² s¹ (26% in Fig. 8 c). Owing to the fact that I_P1 is quite independent of the extracellular pH and rapidly decays after a flash, we assume that the conductances G_1a and G_1b define the small stationary current seen at pH 6.8, but hardly contribute to the stationary current observed at acidic pH. In line with this conclusion we assigned the components contributing to I_SS to a slow, high-sensitivity component, I_P2a, and to a fast, low-sensitivity component, I_P2b, of H⁺ current with the relaxation constants 1/_P2a = ₂¹ and 1/_P2b = ₁¹, respectively (Fig. 9, b and c).

View larger version (14K): [in this window] [in a new window]   FIGURE 9   Synopsis of changes in characteristic parameters in step response of light-induced inward currents across the eyespot overlaying part of the plasmalemma as functions of the intensity of the 300-ms pulse of light. (a) Typical current trace and the parameters that were evaluated. (b) Double logarithmic plot of the three current amplitudes IP1, IP2, and ISS. (c) Double logarithmic plot of the step-up to peak time t11, and the two decay times after light off, 11 and 21.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The four eyespot photoreceptor currents I_P1a, I_P1b, I_P2a, I_P2b

At neutral pH, all eyespot-restricted photocurrents appearing in C. reinhardtii after a flash or upon step-up stimulation are sufficiently explained by two light-induced conductances, G_1a and G_1b. The few studies carried out previously on I_P1 in the green algae C. reinhardtii and H. pluvialis have generally focused on the fast high-light saturating component I_P1b. This major photoreceptor current was proposed to result from a conductance (G_1b)_, that is closely coupled to rhodopsin (Harz et al., 1992). A large body of evidence has accumulated showing that the I_P1b-induced depolarization triggers the flagellar currents, I_FF, that, in turn, cause a switch from forward to backward swimming (Holland et al., 1997). The other component of I_P1, I_P1a, saturating at low light has not been studied in detail because of its small amplitude. It was not clear which ion was the carrier of I_P1a because adjustment of Ca²⁺ in the medium by EDTA or BAPTA did not reduce I_P1 below 15% of its full size (Sineshchekov, 1991; Holland et al., 1996). We have shown above that a careful removal of Ca²⁺ by an ion exchanger also suppresses I_P1a. In earlier experiments, Ca²⁺ might not have been removed completely from the inner cell wall layer that is still present in the cell wall-deficient CW2 mutant. The possibility has been under discussion for years that a low-saturating photoreceptor current, here named I_P1a, triggers distinct direction changes in response to flashes or phototaxis in continuous light. This current has not yet been characterized, and the mechanism through which it might induce direction changes is still obscure. The new finding at least is consistent with the decade-old knowledge that phototaxis is a strongly Ca²⁺-dependent process (Stavis and Hirschberg, 1973), supporting the hypothesis that I_P1a serves as a trigger for direction changes and phototaxis. It should also be clear now that I_P1a involves a multicomponent signaling system, the saturation of which is described by a low light-saturating Michaelis-Menten kinetics.

The stimulus/response relationship of a nth order reaction has the linear slope m = n in the low-dose range of a lg/lg plot. This means that in this range, far from saturation, the observed linear slopes with 0< m <1 can not be explained without additional assumptions. However, apparent slopes 0< m <1 are frequently observed in lg/lg plots of biological dose-effect relationships, such as Figs. 3 b and 7 c. Stimulus-response curves with such small slopes provide the physiological benefit of covering a wide sensitivity range with a finite modulation depth of the sensor. This can not be realized by any signaling system that is simply proportional to the fraction of photoreceptor bleaching. The fact that the data in Figs. 3 b and 7 c are fitted well by two kinetic components suggests the operation of two kinetic systems with sensitivity ranges that differ by approximately a factor of 100: a high-sensitivity system (a) and a low-sensitivity system (b).

Owing to the fact that I_P1a and I_P2a saturate at low photon exposures and occur with millisecond delays after flash (Braun and Hegemann, 1999a), we suggest that the responsible rhodopsin evokes the two conductances G_1a and G_2a via a transmitter (Fig. 10 a). In contrast, the simplest explanation for the low sensitivity system is a single-reaction cycle of a protein complex including rhodopsin and two conductances, G_1b and G_2b, as shown in Fig. 10 b. In this reaction scheme, a defined 1:1 stochiometry between rhodopsin and each conductance is anticipated without specifying how many proteins are involved. The cyclic reaction sequence shown in Fig. 10 b is intended to explain qualitatively the kinetics of the two photocurrents I_P1b and I_P2b at a given high photon exposure. The reaction cycle comprises seven states and transitions of the rhodopsin/ion channel complex. The numbering is in alphabetical order starting with the inactive, resting state I. In the scheme the states M and O represent the conducting states G_1b(Ca²⁺) and G_2b(H⁺), respectively. K does not directly convert into M because the rise of M occurs with a delay of several microseconds (Holland et al., 1996; Braun and Hegemann, 1999a), whereas the lifetime of K should be in the nanosecond range. Similarly, M does not directly convert into O because the decay of I_P1 is faster than the rise of I_P2. This purely sequential model is consistent with the observed sequential appearance of I_P1b and I_P2b and with their light saturation in parallel with rhodopsin bleaching.

View larger version (10K): [in this window] [in a new window]   FIGURE 10   Schematic representation for the activation of light-induced conductances in the C. reinhardtii eye by two rhodopsins, Rha and Rhb. (a) Rha activates G1a within 1 ms and G2a within >100 ms via a side reaction starting from the intermediate M. The activation of G1a and G2a might occur in sequence or via different pathways (as shown here). (b) Simplified cyclic reaction cycle used for the calculation of the high-light saturating photocurrents mediated by Rhb and two conductances, G1b and G2b. The delay for the activation of the Ca2+ conductance G1b only a few microseconds and the Ip1b maximum is reached within 1 ms (flash-to-peak), whereas that of the H+ conductance G2b is reached only after >5 ms.

It had been previously noted on studies on H. pluvialis (Sineshchekov et al., 1990) and C. reinhardtii (Harz et al., 1992) that the flash-to-peak time, t_ftp, depends on Q. This finding is surprising in terms of straightforward photochemistry, because the velocity of the reaction cascade after the initial and only photochemical excitation is expected to be independent to the light intensity. If we assign different velocities of the cascades in systems a and b, superpositions of the current transients from the two systems, a and b of Fig. 10, will cause apparent dependencies of t_ftp on Q. Systematic investigation of this relationship on C. reinhardtii and V. carteri shows that this dependency seems to be steeper in the low-intensity range than at high intensities as seen in Fig. 7 d (above) and in Fig. 4 of Braun and Hegemann (1999a). Moreover, the delay between flash and beginning of the I_P1 also increases considerably at low light levels when the amplitude is relatively large compared with the fraction of rhodopsin bleached (Braun and Hegemann, 1999a). At high photon exposures, the behavior of the total system becomes simpler as the properties of the low-sensitivity system b predominate. Correspondingly, the gentle slopes of the t_ftp(Q) relationships in the right part of Fig. 7 d are fairly consistent with the above notion that in a straightforward photoreceptor t_ftp is quite independent of Q.

In conclusion, at acidic pH, four distinct photoreceptor currents are identified in the eye of C. reinhardtii: the two early and probably Ca²⁺-carried currents I_P1a and I_P1b, and the two late H⁺ currents I_P2a and I_P2b. The same action spectra of I_P1b and I_P2b (Fig. 7 b) and the similar light dependence of the amplitudes (Fig. 7 c) and t_ftp (Fig. 7 d) suggest that the underlying conductances G_1b and G_2b are part of one reaction cascade b. The finding that G_1a saturates at ~100 times smaller light intensities than rhodopsin itself (e.g., Fig. 7 c) may be accounted for by the signaling cascade a (Fig. 10 a). The conductances G_1a and G_2a are activated when only 1% of the photoreceptor is excited. The number of channels might be smaller than that of the photoreceptor to explain the small maximal current I_P1a.

The responses of I_P on up- and down-steps of the quantum flux density E of various intensities, as summarized in Fig. 9, confirm the results of Fig. 7. The two time constants of the exponential current relaxation upon light off (₁ and ₂ in Fig. 9, a and c) turned out to be insensitive to the intensity of the preceeding light (within the accuracy of the measurements, of course). They compare well with the t_ftp levels of I_P2b and I_P2a in Fig. 7 c, respectively. A quantitative analysis of the results of Figs. 7-9 by mechanistic models will be the subject of a separate study.

Geometric aspects

During the investigation of I_P2 it was noticed that the recorded I_P1 and I_P2 were larger when the eye was in the pipette and the membrane portion in the pipette was small, compared with the configuration with the eye exposed to bath and a larger membrane portion outside the pipette (Figs. 2 and 4). This is attributable to the fact that in the configuration of Fig. 4 a only a fraction of the total current through the eye is recorded (Braun and Hegemann, 1999a). Correspondingly, the de- or hyperpolarizing effects of external anions (such as H⁺ or K⁺) on the resting voltage are expected to increase as the surface portion (bath/pipette) is exposed to changes in external ionic composition.

The number of photoreceptor species involved

As shown in the past by electrical measurements on single cells, the photocurrent I_P1b has a rhodopsin action spectrum peaking at 495 nm (Harz and Hegemann, 1991). Later, in populations of retinal-deficient cells that do not express any or extremely small photocurrents in response to light flashes, low- and high-intensity photoreceptor currents (I_P1a and I_P1b) were reconstituted after addition of exogenous retinal (Sineshchekov et al., 1994; Govorunova et al., 2001). Both findings taken together strongly suggest that I_P1a and I_P1b are both rhodopsin-mediated processes, although their Q_1/2 differ by two orders of magnitude of light intensity. Because of the extremely small amplitudes, reliable action spectra of the I_P1a could not be generated in earlier studies nor in this one. Motion analysis and light scattering experiments with white cells that were reconstituted with retinoids and analog compounds have shown a close similarity in the chromophore structural requirements of the two behavioral responses, i.e., phototaxis and photophobic responses (Hegemann et al., 1991; Lawson et al., 1991; Takahashi et al., 1991). However, when the dependence of the direction changes and the phobic responses were analyzed with the help of Poisson statistics, the conclusion was drawn that both are independent processes and may be triggered by individual photon absorptions (Marwan and Hegemann, 1988). Moreover, competition experiments with a 13-trans-locked retinal analog supported this notion by suggesting that slight differences exist between the two responses at the level of their photoreceptor proteins (Zacks et al., 1993). Despite the similarity of the action spectra for all flash-induced behavioral responses respective shape and peak maximum, it is conceivable that I_P1a and I_P1b are controlled by two different rhodopsin species as expressed by the models of Fig. 10.

The next question to be discussed is whether G_2a and G_2b are also rhodopsin-activated conductances. The retinal dependence of I_P2 could not be demonstrated because of the instability of white cells at low pH (data not shown). We presented evidence that G_2a and G_2b conduct H⁺ and are distinct from the two Ca²⁺-conductances G_1a and G_1b. As stated before, I_P2b shares important features with I_P1b, most prominently the high light saturation with Q_1/2 of 58 µE·m². Assuming common values of an absorption cross-section of 1.9·10²⁰ m² and a quantum efficiency of 0.7 for all known rhodopsins ( = 50 000 M¹ cm¹), both I_P1b and I_P2b saturate almost with the rhodopsin bleaching. We argued before that this implies a 1:1 ratio between rhodopsin and the photoreceptor channels or an ion-conducting photocycle intermediate (Harz et al. 1992) (Fig. 10 b). This may not be completely true, because with a fixed rhodopsin/channel stochiometry, light saturation of I_P1b should follow an exponential saturation curve which it does not do perfectly (Fig. 3 c). A multimeric photoreceptor complex (Melkonian and Robenek, 1984) with negatively cooperative ion-conducting properties, however, may explain the slightly better fit of the I_P1b data to the Michaelis-Menten equation of Fig. 3 b. In addition, to cover a dynamic light intensity range of three log units, the unitary conductance of the I_P1b and I_P2b must be small. Finally, taking the similar action spectra of Fig. 7 b into account, we propose that I_P1b and I_P2b are components of the same rhodopsin that has two conductances, G_1b and G_2b (Fig. 10 b); these might be two conducting intermediates of one rhodopsin photocycle. A fast cycling of the rhodopsin complex within 100-300 ms elegantly explains the observed large stationary currents, as only a small fraction of rhodopsin is in an inactive state.

The characterization of I_P1a and I_P2a remains less complete. The major drawback is that high resolution action spectra are missing. The earlier shown retinal-dependence of I_P1a (Sineshchekov et al., 1994) and the rhodopsin action spectra for phototaxis and all flash-induced behavioral responses (Uhl and Hegemann, 1990) lead us to the conclusion that I_P1a and probably I_P2a are activated by the same photoreceptor as shown schematically in Fig. 10 a. This conclusion needs further verification.

Contribution of [K⁺]_o

Finally, one could ask the question why I_P2 is not seen at neutral pH. In earlier experiments we demonstrated that the change of the extracellular pH from 6.8 to 4.0 only altered the internal pH by 0.25 units from pH 7.48 to 7.23 (Braun and Hegemann, 1999b). Consequently, the pH-gradient increased from pH = 0.68 to pH = 3.23, which would explain the observed I_P2 increase. Concomitantly, the membrane potential in C. reinhardtii should shift from 115 to 75 mV, reducing the driving force for H⁺ accordingly (Malhotra and Glass, 1995). But, we argue, under conditions when only the eyespot region was exposed to high proton concentrations, the eyespot is exposed to a high H⁺ driving force without strong depolarization. In contrast to this view, I_P2 increased less than I_P1 in the eyespot-in configuration compared to eyespot-out (Fig. 4) and completely disappeared at high K⁺. To understand this finding, the counterbalancing K⁺-efflux must be considered. The nonlocalized K⁺ efflux is facilitated by a less negative membrane potential as it should be realized under eyespot-out conditions, where the majority of the cell surface is exposed to low pH. Thus, we conclude that at acidic pH I_P2 is determined by the H⁺ influx into the eyespot and by a counterbalancing K⁺ efflux.