Conditional Inhibition of Screening-Pigment Aggregation by Lidocaine in Crayfish Photoreceptors and Frog Retinal Pigment Epithelium

In addition to their effect on membrane excitability, local anesthetics are known to impair the transport of intracellular components along nerve cell processes. Procaine, lidocaine, tetracaine and other analogous compounds reversibly inhibit fast axoplasmic transport in mammalian and frog nerve at concentrations generally higher than those required to block the propagation of the electric impulse (Fink et al. 1972; Anderson and Edstrom, 1973; Aasheim et al. 1974; Fink and Kish, 1976; Lavoie, 1982a,b, 1983). These two actions of the local anesthetics are not necessarily correlated as regards their effective concentrations. Whereas dibucaine is slightly more potent than tetracaine on both excitability and axoplasmic transport, etidocaine is as potent as tetracaine on excitability but five times less effective in inhibiting transport (Lavoie, 19826). Therefore, different actions of the drugs may be responsible for the effects observed on the two physiological processes.

In contrast to the vast literature concerned with the mechanisms of action of local anesthetics, comparatively few reports are available on the inhibitory action of these compounds on intracellular transport. Electron microscopy studies of axons treated with local anesthetics have shown reductions in the numbers of microtubules (Fink et al. 1972; Edstrom et al. 1973). Nevertheless, though such alteration of the cytoskeleton would be expected to interfere with its fundamental role in intra-axoplasmic motion, the decrease in microtubule count does not always correlate with inhibition of transport (Byers et al. 1973, 1979; Lavoie et al. 1989). The events leading to these morphological effects of the local anesthetics are far from understood, and it is conceivable that the drugs might have a concomitant, or even earlier, influence on other factors critical for intracellular transport. One of these, depletion of metabolic energy, has been ruled out as the cause of axoplasmic transport inhibition by lidocaine and has been shown to be, at most, only partly responsible for the inhibition observed with other local anesthetics (Lavoie et al. 1989).

This paper reports a first survey of the effects of lidocaine on two experimental models used for the study of intracellular transport: the migration of screening pigment granules along the photoreceptors of the crayfish compound eye and within the cells of the frog retinal pigment epithelium (RPE). Like fast axoplasmic transport, screening-pigment migration in both types of retinal cells is dependent on microtubules (Frixione et al. 1979; Frixione, 1983a,b; Burnside et al. 1983) and on oxidative metabolism (Frixione et al. 1979; Mondragón and Frixione, 1989). Unlike fast axoplasmic transport, which is a rapid, continuous process, screeningpigment migration is a relatively slow response to changes in illumination and can be activated by light or darkness, even under in vitro conditions (Snyder and Zadunaisky, 1976; Frixione et al. 1979; Burnside and Basinger, 1983; Dearry and Burnside, 1984; Mondragón and Frixione, 1989). The movement of pigment in the RPE also differs from both fast axoplasmic transport and pigment migration in crustacean photoreceptors in that the pigment response to light -but not to darkness -is mediated by actin filaments in addition to microtubules (Burnside et al. 1983).

Pigment migration in the crayfish photoreceptor closely resembles fast axoplasmic transport in nerves in that the pigment granules move along the axon, advancing anterogradely -i.e. away from the nucleus and in the same direction that the visual input is conducted -during dark-adaptation and retrogradely as they slip back into the cell body when returning to the light-adapted position. In contrast to typical axoplasmic transport, however, the pigment granules shuttle within a restricted area of the photoreceptor length, so that they never get down to the synaptic terminals. In the RPE cells, the pigment moves within several slender processes that stem from the cell body, invading these projections in the light and retreating towards the cell body in the dark. Although in both systems the pigment movements are bidirectional, the granules migrate in just one direction at any given time, the movement being in this regard analogous to that of chromatophores (Rodionov et al. 1991). As in chromatophores, movement in one direction aggregates the granules whereas the reverse motion causes dispersal, so the pigment redistributes according to the physiological state of the cells; in both the crayfish photoreceptors and the frog RPE the pigment aggregates as it moves to the dark-adapted position and is dispersed by the action of light. Dark-adaptation is in each case more dependent on metabolic energy than is light-adaptation (Frixione et al. 1979; Mondragon and Frixione, 1989).

Our experiments show that lidocaine blocks pigment migration to the dark-adapted position or promotes movement in the opposite direction in the dark in the two types of retinal cells at concentrations equivalent to, or lower than, those found to inhibit fast axoplasmic transport. Tests with ionic substitution in the incubation media indicate that the effect is dependent on the external levels of Na⁺ and Ca²⁺. The results suggest that lidocaine affects pigment migration by shifting the ionic balance of the cytoplasm in each case and argue against the necessary involvement of an efferent nervous influence in the control of pigment position.

The procedures followed for dissection, incubation and fixation of the two types of eyes have been described in detail elsewhere (Frixione et al. 1979; Mondragón and Frixione, 1989). Briefly, eyestalks severed from adult crayfish of the species Procambarus clarkii (Girard) were dissected to remove the optic ganglia, leaving intact the lamina ganglionaris, which is intimately associated with the axons of the photoreceptors. Frog eyecups were prepared by enucleation of eyes excised from Rana pipiens (Schreber). In each case, three or four prepared eyes were then fastened to a rubber base at the bottom of a wide-mouthed 100 ml glass vial containing 6–10 ml of the desired incubation medium (see Tables 1 and 3) under an O₂-rich atmosphere. Lidocaine chlorhydrate (2 % stock solution from Astra Chemicals, S.A., México) was diluted in the test media to give the indicated concentrations. All solutions were buffered at pH7.4 with 10 mmoll⁻¹ imidazole, maintained at 17°C, and changed under strictly light-proof conditions when so required. Incubation with frequent mild agitation was carried out for the specified periods in a dark-room or under approximately 500 l× of fluorescent illumination. Each group of eyes was then fixed (heat/formaldehyde method for crayfish eyes; 2.5 % glutaraldehyde in Ringer for frog retinas; see above references for details) at the appropriate time either in the light or in total darkness.

After fixation, crayfish eyes were sagittally bisected and measurements of the pigment position over the longitudinal axis were taken with a micrometer installed in a stereomicroscope. The frog eyecups were divided into narrow strips, postfixed with 1 % OsO₄, dehydrated through ethanol, and embedded in Spurr’s resin. The pigment position in the RPE was then micrometrically determined in semi-thin sections (150-200 nm) stained with Toluidine Blue and cut parallel to the photoreceptors located around the posterior pole of the eye. Micrographs of crayfish eyes were obtained with a 35 mm camera coupled to a stereomicroscope, and those of the frog RPE were produced with a photomicroscope. At least three eyes were included in each experiment with a given set of conditions, and all experiments were repeated at least twice, so the data shown in Fig. 2 and Tables 2 and 4 represent averages and standard deviations of determinations of pigment position in a minimum of six different eyes.

Fig. 1A,B shows the sagittal plane of two crayfish eyes, fixed and bisected after being incubated in normal saline (solution A, Table 1, slightly modified from van Harreveld, 1936) for full adaptation to either light or darkness. In the light-adapted state (Fig. 1A), the photoreceptor layer of the retina appears to be filled with the screening pigment up to the distal or nuclear ends of the photoreceptors (see also scheme in Fig. 4; for histological details see Krebs, 1972; Nassel, 1976). In the dark-adapted condition (Fig. IB), the pigment is found aggregated below the basal membrane of the retina, where it accumulates within the axons of the photoreceptors. When the pigment is so retracted, the photoreceptor cells become fully exposed, with a consequent increase in visual sensitivity. As the pigment recedes or expands between these extreme positions, its outer margin travels about 180 μm in an adult crayfish. Since the length of the cells themselves does not vary during these movements, a convenient pigment index (PI) can be obtained as the quotient of two measurements (Frixione and Aréchiga, 1981): (a) the distance from the basal membrane to the outer margin of the pigment mass, which varies with the level of adaptation; and (b) the distance from the basal membrane to the distal tip of the photoreceptors, which is constant for a given eye. Then PI=a/b is a normalized figure which, under usual conditions, takes values ranging from 1.0 for maximal pigment dispersion in the light to 0.0 for extreme aggregation in the dark. The data shown in Fig. 2 and Table 2 are expressed in terms of this assessment convention.

Pre-incubation of crayfish eyes for 90 min in the light with saline containing 25 mmol l⁻¹ lidocaine blocked pigment migration to the dark-adapted position (Table 2). A PI of nearly 1.0 was obtained in all treated eyes after subsequent incubation in the same solution for 120 min in darkness, when the controls were typically dark-adapted. The inhibitory effect of the anesthetic on pigment aggregation is largely reversible. Since, even under normal conditions, complete dark-adaptation occurs just once in isolated eyes, reversibility was tested with experiments that involved a single challenge in the dark. Eyes pre-incubated in the light for 90min in the presence of 25 mmol l⁻¹ lidocaine and then thoroughly washed with normal saline, aggregated the bulk of the pigment mass to the dark-adapted position when placed in darkness for 120 min. It should be noted, though, that the ability to recover from the treatment was not as homogeneous as might be inferred from the PI values in Table 2, because some of these eyes showed faint traces of pigment left in the proximal half of the effective migration zone. In a second protocol, the eyes were exposed to lidocaine for 30min in the light and 90 min in the dark, before repeated washing and continued incubation in darkness for an additional 120min. Pigment aggregation in this case was less complete (about 75 % of the maximum in average), but nevertheless characteristic of darkadaptation and clearly different from the PI value of 0.87±0.07 (S.D., N=6) observed in the controls, which were not washed. Also, the region deserted by the pigment in the washed eyes was remarkably clean in comparison with the same region after treatment with the first protocol.

The anesthetic not only inhibits pigment aggregation, but also promotes pigment dispersion in the absence of light. When dark-adapted eyes were exposed to 25 mmol l⁻¹ lidocaine in the dark the pigment dispersed to near the fully light-adapted position over 3h, a period during which the photoreceptors can easily maintain a dark-adapted state (Figs 1C and 2). This light-adapting effect of lidocaine was considerably reduced when the concentration of anesthetic was lowered by just 25 % (i.e. down to 19mmoll⁻¹) and it was virtually undetectable below this level. In contrast, doubling the concentration (to 50 mmol l⁻¹) produced pigment dispersion with the same time course as with 25 mmol l⁻¹ lidocaine (Fig. 2).

Even though lidocaine produces pigment dispersion by itself, the natural response of the screening pigment to light could be detected in photoreceptors treated with the drug. Incubation of dark-adapted eyes with 25mmoll⁻¹ lidocaine for 2h in the dark (i.e. when according to data in Fig. 2 the drug-induced migration has reached a PI value of about 0.75), followed by illumination for 60min, resulted in a uniform PI of 1.0. This value is significantly higher (P<0.05 by the Student’s t-test) than that produced by lidocaine during an equivalent total period in the dark (Fig. 2), indicating that the anesthetic does not interfere with the pigment movement elicited by light. This finding was not surprising because pigment dispersion is a relaxed state to which the cells revert when exposed to light and to many disruptive agents, including metabolic inhibitors (Frixione et al. 1979). Therefore, all the further tests of lidocaine on crayfish photoreceptors were designed to look at its effects on the ability of the cells to maintain the aggregation of pigment in the dark-adapted position, which is the active phase of adaptation. From the above results, a 2h exposure of dark-adapted eyes to 25mmoll⁻¹ anesthetic was selected as the standard treatment.

Studies on the ionic dependence of the screening-pigment responses in arthropod photoreceptors have suggested that a light-elicited rise in intracellular levels of Ca²⁺ and/or Na⁺ is probably involved in the mechanism that drives the pigment to the light-adapted position (Kirschfeld and Vogt, 1980; Frixione and Aréchiga, 1981; Howard, 1984). Local anesthetics have recently been reported to alter the normal fluxes of Na⁺ and Ca²⁺ across plasma membranes in preparations of synaptosomes (García-Martín et al. 1990) and epithelial cells (Shibamoto et al. 1990), so one possible mechanism for the inhibitory effect of lidocaine on pigment aggregation is that it may increase Na⁺ and/or Ca²⁺ levels in the photoreceptor cells. To test this hypothesis, we tested the anesthetic in media lacking these ions (solutions B and C, Table 1). The eyes were first dark-adapted in the modified solutions, which permit or even enhance pigment aggregation in the dark (Frixione and Aréchiga, 1981), and they were then transferred in complete darkness to the same medium containing lidocaine. Substitution of Na⁺ with choline (solution B) eliminated the pigment dispersion induced by the anesthetic in the normal saline (Table 2). A similar protection was observed with an imidazole-buffered 300 mmol l⁻¹ sucrose solution (solution C), that is in the absence of all the physiological ions.

It has been reported that abnormally high concentrations of Na⁺ and Ca²⁺ in the external medium have opposite effects on pigment position (Frixione and Aréchiga, 1981). High Na⁺ concentration inhibits, whereas high Ca²⁺ concentration facilitates, aggregation, and the addition of Co²⁺ to the normal saline also favors the dark-adapted state. Hence, it was of interest to find out whether the divalent cations could protect the photoreceptors against the dispersive action of lidocaine on the pigment. Eyes dark-adapted in high-Ca²⁺ or in Co²⁺-containing media (solutions D and E, Table 1) were able to keep the pigment in the dark-adapted position despite the ensuing addition of the drug (Table 2). Most eyes incubated in the modified solutions presented the maximal pigment aggregation characteristically produced in such media in the absence of anesthetic.

The RPE cells are attached at their basal ends to Bruch’s membrane of the retina and present long apical projections that interdigitate with the outer segments of the photoreceptors (Fig. 3 and scheme in Fig. 4; for a detailed morphological description, see Nguyen-Legros, 1978). Under bright illumination numerous pigment granules disperse along the narrow projections, forming a light absorbing screen around the outer segment of each photoreceptor (Fig. 3A). In the dark the granules migrate towards the basal ends of the epithelial cells (Fig. 3B), so the full length of each photoreceptor becomes exposed and visual sensitivity is increased. As a result of these movements the distance from Bruch’s membrane to the edge of the pigment mass changes by about 70μm. Thus, as in the case of crayfish eyes, the pigment position can be expressed as a normalized index obtained from the quotient of two measurements (Burnside et al. 1983; Mondragón and Frixione, 1989): (a) the distance from Bruch’s membrane to the edge of the pigment, which varies with adaptation; and (h) the distance from Bruch’s membrane to the outer limiting membrane of the retina, which is fixed in each particular eye. Typical extreme values of PI=a/b are 1.0 for full light adaptation and 0.3 for complete dark adaptation, both of which occur and are stable in vitro for up to 6 h when the eyecups are incubated in Ringer (solution A, Table 3) under the present conditions (Mondragón and Frixione, 1989).

Except for its effective concentration, lidocaine affected the RPE and the crayfish photoreceptors in a very similar way. Preliminary tests showed that 5 mmol l⁻¹ anesthetic in the Ringer is a critical level, which has an effect on pigment position and can be tolerated by the tissue. Higher concentrations (⩾10mmoll⁻¹) caused detachment of the retina from the RPE, particularly in dark-adapted eyes, whereas lower levels (⩾ mmol l⁻) failed to produce clear and reproducible effects on pigment position. Therefore 5 mmol l⁻¹ lidocaine was used for the following experiments.

Pigment aggregation to the dark-adapted position was arrested when the Ringer contained lidocaine (Table 4). All the eyecups gave a PI close to 1.0 after 3h of dark incubation in the presence of the drug, in contrast with an average PI value of 0.32 for the controls. Treatment of dark-adapted eyes with the anesthetic in the dark dispersed the pigment up to a PI value of 0.68 within 2h (Fig. 3C). Dark-adapted eyecups transferred in the dark to Ringer with anesthetic, and then incubated for 2h in darkness before 60 min of illumination, gave a uniform PI of 1.0. Since the total incubation time in this case was 6h, i.e. the maximum period during which the eyecups can maintain the dark-adapted state, full pigment dispersion under these conditions may reflect a decreased metabolic capacity of the cells in addition to the combined result of the treatment with the drug and the effect of illumination. However, as with crayfish eyes, it is clear that lidocaine blocks pigment aggregation but does not interfere with the movement to the full light-adapted position in RPE cells.

Lidocaine-treated eyecups showed morphological alterations of the retinal tissue. The photoreceptors appeared crumpled, and wide interstitial spaces separated the epithelial cells (Fig. 3C). These histological changes were not completely reversible, even though the pigment-transporting ability was recovered after removal of the anesthetic. Eyecups incubated with lidocaine for 2h in the dark, and then changed several times to fresh Ringer in darkness, had the pigment aggregated in the typical dark-adapted position after an additional 3h of incubation.

The movement of pigment in the RPE has also been found to be sensitive to the ionic balance in the external and intracellular environments, particularly as regards Ca²⁺ and Na⁺ (Snyder and Zadunaisky, 1976; Dearry and Burnside, 1984; Mondragón and Frixione, 1989). Therefore, for the reasons already mentioned for crayfish eyes, we tested lidocaine in combination with manipulations of the incubation medium (solutions B-D, Table 3). Replacement of Na⁺ with choline in the Ringer (solution B) prevented the disruptive effects of the anesthetic on both dark-adaptation and retinal morphology (Table 4 and Fig. 3D). Equivalent results were obtained when lidocaine was applied in the presence of a fivefold increase in the normal concentration of Ca²⁺ (solution C) or with the addition of 5 mmol l⁻¹ Co²⁺ to the Ringer (solution D).

Fig. 4 summarizes schematically these observations on the effects of lidocaine on screening-pigment behavior in the two types of retinal cells. In each case treatment of the isolated eyes with the anesthetic blocked pigment migration to the aggregated, energy-requiring state characteristic of dark-adaptation. The pigment was thus found in these eyes in a fully light-adapted position, equivalent to that shown by the drawings on the left. Moreover, lidocaine antagonized the aggregated state because it caused pigment reversion towards the dispersed, light-adapted position when administered to dark-adapted eyes subsequently maintained in the dark (drawings on the right). Nevertheless, the inhibitory action of lidocaine on dark-adaptation was found to depend on the composition of the incubation medium; low-Na⁺, high-Ca²⁺ or Co²⁺-containing solutions allowed dark-adapted eyes to hold the pigment in the fully aggregated state despite the presence of the drug. Therefore, the pigment position under these conditions was typical of dark-adaptation (illustrated by the drawings in the middle).

Screening-pigment position is controlled through intrinsic and extrinsic mechanisms in arthropod photoreceptors. The cells are known to react to light and darkness as independent pigmentary effectors (Frixione et al. 1979; Kirschfeld and Vogt, 1980), but they are also subject to an efferent nervous input that influences the adaptation state according to a circadian rhythm (see Barlow et al. 1989). Recent evidence suggests that pigment aggregation in crayfish photoreceptors is facilitated through an efferent serotonergic pathway (Frixione and Hernández, 1989; Aréchiga et al. 1990). Likewise, dopamine, γ-aminobutyric acid and other neurotransmitters have been shown to modulate retinomotor movements in some vertebrate eyes (Dearry and Burnside, 1986a,b). Thus, although isolated eyes are disconnected from the central nervous system, it is possible that the local circuitry may still promote pigment aggregation in the dark, e.g. by spontaneous firing of the severed efferent fibers or activation of certain interneurons. In the presence of an action-potential-blocking and neurotransmission-depressing agent such as lidocaine (Tabatabai and Booth, 1990) most, if not all, of these extrinsic mechanisms may be expected to be inoperative. Therefore, the failure of the anesthetized eyes to aggregate their pigment or to keep it in that position could result from the lack of some facilitatory influence. This interpretation can be virtually excluded by the observation that in both preparations the effect of the drug was dependent upon the ionic composition of the incubation medium. The retinal cells were able to keep the screening pigment in the aggregated state in the presence of lidocaine provided that Na⁺ was excluded from the medium; elevated Ca²⁺ concentration or the addition of Co²⁺ also prevented the inhibition. While a high Ca²⁺ concentration could perhaps promote the release of a facilitatory neurotransmitter and thus bypass the effect of lidocaine, a low Na⁺ concentration and especially the presence of Co²⁺ should antagonize synaptic transmission (Dearry and Burnside, 1986a,b). Although an anesthetic-induced selective release of light-adapting neuromodulators cannot be rigorously ruled out, the present results seem to indicate that the light-adapted and dark-adapted positions of the screening-pigments can occur in both types of eye without the mediation of synaptic inputs. Moreover, the demonstrated ability of local anesthetics to inhibit fast axonal transport suggests that pigment migration may be impaired by a direct effect of lidocaine on the photoreceptors and RPE cells.

In the crayfish photoreceptor -a primary neuron -lidocaine affected the anterograde migration of the screening pigment to the dark-adapted position and its maintenance in the aggregated state at concentrations (19–25 mmol l⁻¹) equivalent to those reported to inhibit fast axoplasmic transport (Fink et al. 1972; Anderson and Edstrom, 1973; Byers et al. 1973; Lavoie, 1982b). However, the level of anesthetic in the interstitial spaces around the photoreceptors is probably lower than this, because even small ions appear to diffuse slowly into perfused nervous tissues of crustaceans (Abbott et al. 1975). Therefore, the effective concentration is likely to be closer to that found to inhibit pigment aggregation in the frog RPE (5mmoll⁻¹), a much thinner and presumably more readily perfusable preparation. Such levels are within the range in which lidocaine abolishes excitability, but not axoplasmic transport, in nerve (Fink et al. 1972; Anderson and Edstrom, 1973; Lavoie, 19826). Accordingly, pigment migration in the two types of retinal cells appears to be more sensitive to the anesthetic than is protein transport along axons. The accumulated evidence from these various systems suggests that lidocaine -and perhaps other local anesthetics -interferes with a basic mechanism common to different types of intracellular transport.

Pigment aggregation in the dark-adapted position is the active, energydemanding phase of the adaptation cycle in both crayfish photoreceptors and RPE cells, whereas movement in the opposite direction seems to be the result of relaxation induced by light or limiting conditions (Snyder and Zadunaisky, 1976; Frixione et al. 1979; Mondragon and Frixione, 1989). Since oxidative metabolism in nervous tissue and other systems has been found to be sensitive to local anesthetics (Geddes and Quastel, 1956; Haschke and Fink, 1975; Tarba and Cracium, 1990), one reasonable assumption is that lidocaine might diminish the energy available in the retinal cells for aggregating the pigment and holding it in that state. However, the occurrence of stable pigment aggregations in low-Na⁺ media, despite the presence of lidocaine, is contrary to this view. The same observation argues against the inhibition being due to a specific effect of the anesthetic on microtubules (Fink et al. 1972; Edström et al. 1973), which are involved in pigment migration to the dark-adapted position in crayfish photoreceptors (Frixione et al. 1979; Frixione, 1983a,b) and RPE cells (Burnside et al. 1983). Hence, as has been concluded in the case of fast axoplasmic transport (Lavoie et al. 1989), the effect of lidocaine on the movement of the screening pigment is unlikely to be the result of an action of the drug on the structure or the fuelling of the translocation apparatus.

An alternative candidate to account for the vulnerability of pigment migration to lidocaine is the intracellular mechanism that controls pigment position in the retinal cells. A rise in cytoplasmic Ca²⁺ concentration has been associated with pigment migration to the light-adapted position in arthropod photoreceptors (Kirschfeld and Vogt, 1980; Frixione and Aréchiga, 1981; Howard, 1984) and RPE cells (Burnside and Basinger, 1983; Dearry and Burnside, 1984; Mondragón and Frixione, 1989). Since local anesthetics are known to inhibit the Ca²⁺-storing ability of muscle sarcoplasmic reticulum (Bianchi, 1968; Johnson and Inesi, 1969), non-mitochondrial compartments in nerve terminals (Blaustein et al. 1978; Lavoie et al. 1986) and smooth endoplasmic reticulum of invertebrate photoreceptors (Walz, 1982), it is conceivable that lidocaine could hinder adequate buffering of internal Ca²⁺ and thus prevent the retinal cells from attaining the low levels required to keep the pigment in the dark-adapted position. This explanation is also unlikely, however, because the drug would be expected to affect the Ca²⁺-sequestering organelles regardless of the ionic composition of the external medium, which is not the case with either preparation used in the present study. Moreover, the levels of lidocaine tested in our experiments are within the range found to stimulate rather than to inhibit Ca²⁺ uptake by lysed brain synaptosomes (Lavoie et al. 1986). Several local anesthetics may cause a net decrease in Ca²⁺ extrusion from nerve terminals as a result of inhibition of the Ca²⁺ pump at the plasma membrane, in spite of some stimulation of the Na⁺/Ca²⁺ exchanger (Garcia-Martin et al. 1990). Lidocaine could produce similar effects in the retinal cells and thus cause the pigment to move to the light-adapted position through Ca²⁺ accumulation in their cytosol. However, Na⁺-deficient and Ca²⁺-rich solutions would then be expected to enhance the light-adapting action of lidocaine, which is the reverse of the results of our experiments.

Conditions designed to increase internal Na⁺ concentration promote pigment dispersion, or inhibit aggregation, in both crayfish photoreceptors and frog RPE cells (Mondragón and Frixione, 1989; Frixione and Pérez-Olvera, 1990). In both instances there is reason to suppose that elevation of Na⁺ concentration might cause Ca²⁺ release from internal stores, which in turn could inactivate the energy-consuming mechanism that tends to aggregate the pigment in the dark (see also Frixione and Ruiz, 1988). The dependence of the lidocaine effect on external Na⁺ concentration could, therefore, indicate that the anesthetic induces a higher than normal Na⁺ leakage into the cells, leading to internal Ca²⁺ release and inhibition of the pigment-aggregating mechanism. A similar mechanism has been considered to explain the Na⁺-dependent inhibition of fast axoplasmic transport and saltatory organelle movement in nerve cells by batrachotoxin, a steroidal compound that opens the voltage-dependent Na⁺ channels (Forman and Shain, 1981; Worth and Ochs, 1982). In this context it also becomes particularly significant that ouabain and veratridine, which have in common the ability to cause an abnormal increase in intracellular Na⁺ concentration, are potent inhibitors of fast axoplasmic transport (Edstrom, 1977). Like lidocaine, ouabain inhibits pigment aggregation in crayfish photoreceptors (Frixione and Aréchiga, 1981; Frixione and Pérez-Olvera, 1990) and RPE cells (Mondragón and Frixione, 1989). A direct action of lidocaine on the Na⁺/K⁺-ATPase is unlikely because the anesthetic seems to be unable to alter the pumping rate of the enzyme, even at concentrations that were found to be effective in the present experiments (den Hertog and Ritchie, 1969).

The possibility that lidocaine might inhibit the pigment-aggregating mechanism by inducing an increase in Na⁺ leakage into the retinal cells is supported by evidence that several anesthetics can enhance permeability to cations in artificial and natural membranes (Barchfeld and Deamer, 1985; Escudero and Gutiérrez-Merino, 1987). High levels of external Ca²⁺, or the addition of Co²⁺, which probably decrease Na⁺ leakage into the cells (Blaustein and Goldman, 1968; Fulpius and Baumann, 1969), would be expected to protect the cells against this lidocaine effect. This is in accordance with our results. Competition between Ca²⁺ and local anesthetics has also been observed for active conductance changes in both frog and crustacean nerves (Aceves and Machne, 1963; Blaustein and Goldman, 1966). If our interpretation is correct, Ca²⁺ release from internal stores in response to a rise in Na⁺ concentration in the cytoplasm should outweigh any stimulation of Ca²⁺ uptake that might be produced at the same time in these compartments by low doses of lidocaine (Lavoie et al. 1986). Direct measurements with ion-sensitive microelectrodes will be necessary to test this hypothesis for the action of lidocaine on pigment transport in the two types of retinal cells.

The micrographs presented in this work were made at the Unidad de Microscopía Electrónica del CINVESTAV (México). The authors are grateful to Ms Leonor C. Fierros for her assistance in typing the manuscript. During the execution of this investigation, R.M. was an undergraduate student of the Facultad de Estudios Superiores Cuautitlán, Universidad Nacional Autónoma de México. Supported in part by the research grant PCCBBNA-020479 from the CONACYT (México).