Dopamine stimulates salivary duct cells in the cockroach Periplaneta americana

Insects have either morphologically simple tubular or very complex innervated acinar salivary glands. Among the acinar glands, the structure, innervation and several aspects of the aminergic control of salivation have been extensively studied in cockroaches and locusts (for reviews, see Ali, 1997; House and Ginsborg, 1985). Nevertheless, we are far from understanding the complex physiology of salivation with respect to acinar salivary glands. The key cellular mechanisms of saliva production, the mechanisms that contribute to the modification of the primary saliva and the neural control of all cell types engaged in saliva production are not known.

The secretory acini in the salivary glands of the cockroach Periplaneta americana contain two morphologically and functionally distinct cell types: a pair of peripheral cells and eight central cells. The peripheral cells are specialized for electrolyte/water transport, and the central cells for the production and secretion of proteins and mucins (Just and Walz, 1994a–c; Kessel and Beams, 1963). The glands are innervated from the suboesophageal ganglion, the stomatogastric nervous system and the satellite nervous system (Elia et al., 1994; Davis, 1985; Whitehead, 1971), and the biogenic amines dopamine and serotonin have been identified as neurotransmitters stimulating salivation (Evans and Green, 1991; Smith and House, 1979; Whitehead, 1973; for a review, see House and Ginsborg, 1985). We have recently shown that dopamine stimulates the production of a completely protein-free, watery saliva, whereas serotonin is necessary to stimulate secretion of the proteinaceous saliva components (Just and Walz, 1996). Thus, dopamine is not merely more effective in stimulating salivation than serotonin: dopamine and serotonin stimulate different cell types.

Several lines of evidence suggest that the primary saliva is modified by the salivary duct epithelial cells. In situ electron-probe X-ray microanalysis of cryosections through P. americana salivary glands suggests (1) that the acini produce a NaCl-rich primary saliva and (2) that the primary saliva is modified within the ducts by Na⁺ reabsorption and K⁺ secretion (Gupta and Hall, 1983). The distal duct cells have all the ultrastructural characteristics of epithelial cells engaged in transport functions. Both their apical and basolateral plasma membrane domains are enlarged by deep infoldings, and they contain many mitochondria (Just and Walz, 1994a). Immunocytochemical investigations have shown that the basolateral plasma membrane domain is rich in Na⁺/K⁺-ATPase, whereas the apical plasma membrane domain is equipped with a vacuolar-type proton pump (V-H⁺-ATPase; Just and Walz, 1994b). In addition, the salivary duct epithelial cells contain carbonic anhydrase, the enzyme responsible for the reversible hydration of CO₂ (Just and Walz, 1994c). In spite of these findings, the mechanisms by which the salivary ducts modify the primary saliva are unknown. It is also unclear whether the activity of the salivary duct epithelial cells is influenced by the closely juxtaposed salivary duct nerve or neurotransmitter(s) in the haemolymph. Indeed, Whitehead (1971) has demonstrated that branches of the salivary duct nerve form a plexus on the ducts in P. americana and, more recently, Bräunig (1997) has shown that the octopaminergic DUM1B neurones form dense neurohaemal networks on salivary gland nerves in Locusta migratoria.

The immediate aim of the present work was to investigate whether the salivary duct epithelial cells in P. americana are stimulated by the neurotransmitters dopamine and/or serotonin. Conventional intracellular recordings and Ca²⁺-imaging experiments have therefore been used to study the effects of dopamine and serotonin on salivary duct cells.

A colony of Periplaneta americana (Blattodea, Blattidae) was reared at 27 °C under a 12 h:12 h L:D regime. The animals had free access to food and water. Imagines of both sexes were used.

The salivary glands were dissected from the animals in oxygenated cockroach physiological saline as described previously (Just and Walz, 1994a). Small pieces of the salivary glands, consisting of one lobe with its acini and duct system, were examined.

The cockroach physiological saline had the following composition (in mmol l⁻¹): 160 NaCl, 10 KCl, 2 CaCl₂, 2 MgCl₂, 10 glucose, 10 Tris. The pH was adjusted with HCl to 7.4. The nominally Ca²⁺-free solution contained no added CaCl₂ and 1 mmol l⁻¹ EGTA. All solutions were continuously gassed with oxygen. Osmolarity was checked using an automatic osmometer (Knauer, Berlin, Germany) and was 355 mosmol l⁻¹. The Ca²⁺-free solution had a slightly lower osmolarity since it contained EGTA, and less HCl was required to adjust the pH. To avoid osmotic gradients and cell swelling, this difference was corrected using mannitol.

A 10⁻² mol l⁻¹ dopamine stock solution was stored in aliquots of 200 μl at −20 °C, which were diluted in physiological saline immediately before an experiment. Oxygen was passed through the dopamine-containing solutions only while the solution was actually being used. Dopamine-containing solutions were replaced by fresh solution after 10 min of oxygenation.

Fura-2 AM was obtained from Molecular Probes (Eugene, OR, USA), and dopamine and serotonin were obtained from Sigma (Deisenhofen, Germany). All other chemicals were of analytical grade and were obtained from various suppliers.

Ca²⁺-imaging experiments were performed as described previously (Zimmermann and Walz, 1997). Isolated lobes of the salivary glands were loaded with fura-2 at room temperature (20–22 °C) by a 15 min incubation in 5 μmol l⁻¹ fura-2 AM in physiological saline. The lobes were then mounted in a rhomboid recording chamber (Science Products, Hofheim, Germany) and continuously superfused with oxygenated physiological saline. The coverslip-bottom of the chamber was coated with Cell-Tak tissue adhesive (Collaborative Biomedical Products, Bedford, USA) and changed after every experiment.

The chamber was mounted on a Zeiss Axiovert 135TV inverted microscope equipped with epifluorescence optics and a Zeiss Fluar 20 (numerical aperture 0.75) objective. Epifluorescent excitation was produced by a 75 W xenon arc lamp monochromator unit connected to the microscope by a quartz fibre-optic light guide. The epifluorescence filter-block in the microscope contained a 450 nm dichroic mirror and a 515–565 nm bandpass emission filter. Pairs of fluorescence images were excited at 340 and 380 nm and captured and digitized with a cooled image transfer CCD camera (TE/CCD-512EFT, Princeton Instruments Corp., Trenton, NJ, USA) at a rate of 0.5–0.2 s⁻¹ at 12-bit resolution. A 5×5 binning application to individual pixels resulted in a spatial resolution of 3.8 μm. Monochromator control, image acquisition and processing were carried out using the imaging software Metafluor 2.75 (Universal Imaging Corp., PA, USA) on a personal computer.

Lobes of the salivary glands were mounted in the recording chamber as described above and continuously superfused with oxygenated physiological saline at a flow rate of 1.2 ml min⁻¹. Conventional microelectrodes were filled with 3 mol l⁻¹ KCl and had resistances of approximately 80 MΩ. Salivary duct cells were impaled via the basolateral membrane under optical control (Leica DM IBB inverted microscope). Potential differences were measured using an L/M-PC amplifier (List-Medical, Darmstadt, Germany) in current-clamp mode, monitored on an oscilloscope and stored on a personal computer using the software Chart 8.11 (HEKA, Lambrecht/Pfalz, Germany).

In some experiments, intracellular recordings and Ca²⁺-imaging were performed simultaneously on the imaging setup using a Bramp 01 amplifier (npi, Tamm, Germany) for intracellular recordings.

Values are expressed as arithmetic means ± S.E.M. Results were compared statistically using a paired Student’s t-test. P values of less than 0.05 were considered significant. Graphs were plotted using SigmaPlot 4.0 (Jandel Scientific, San Rafael, CA, USA). Figures that show values from a single experiment are representative of at least three independent experiments. Only one region per imaging experiment was used for quantitative analysis.

In the unstimulated salivary duct cells, the intracellular free Ca²⁺ concentration ([Ca²⁺]_i) was 48±4 nmol l⁻¹ (N=18). Serotonin concentrations up to 10 μmol l⁻¹ had no effect on [Ca²⁺]_i, whereas stimulation with 1 μmol l⁻¹ dopamine induced a reversible, slow rise in [Ca²⁺]_i. The elevation in [Ca²⁺]_i began 20–140 s after the addition of dopamine and reached 311±43 nmol l⁻¹ (N=18) (Fig. 1). In some preparations, the dopamine-induced increase in [Ca²⁺]_i was preceded by an initial small brief fall in [Ca²⁺]_i (see Figs 3, 5, 6, 7A). The dopamine-induced elevation in [Ca²⁺]_i was almost tonic (see Fig. 4) or declined slowly after reaching its peak value in the continuous presence of dopamine (see Figs 5, 11A). The spatiotemporal pattern of the dopamine-induced elevation in [Ca²⁺]_i is illustrated in the series of pseudocolour images in Fig. 2A. To evaluate the spatiotemporal [Ca²⁺]_i dynamics, we defined a line (1 pixel wide, 102 pixels long) in the sequence of background-subtracted images, calculated [Ca²⁺]_i in each pixel along the line, and translated [Ca²⁺]_i into pseudocolour values. These were plotted as a space–time plot, where distance was the ordinate and time the abscissa (Fig. 2B). This space–time plot shows that [Ca²⁺]_i starts increasing at several points in the duct epithelium and, from there, the increase in [Ca²⁺]_i seems to spread slowly over the duct. Not all regions of the duct seem to reach the maximum [Ca²⁺]_i. It is interesting to note that, in contrast to the rise in [Ca²⁺]_i, the initial brief fall in [Ca²⁺]_i described above occurs at the same time in all parts of the duct epithelium (Fig. 3). The velocity at which the elevation in [Ca²⁺]_i spreads over the ducts was estimated by fitting a straight line along regions with similar colour, as illustrated in Fig. 3. The velocity of this Ca²⁺ ‘tide’, which is represented by the slope of the line, is 3.7±0.6 μm s⁻¹ (N=4).

The dopamine-induced elevation in [Ca²⁺]_i is dose-dependent and, particularly at lower concentrations, almost tonic (Fig. 4A,B). Another characteristic feature of the dopamine-induced elevation in [Ca²⁺]_i is that its size depends on the history of stimulation. When the duct cells are stimulated twice with 1 μmol l⁻¹ dopamine, the second dopamine application causes an elevation in [Ca²⁺]_i to a concentration 4.2±1.4 times larger than that for the first stimulation (N=4) (Fig. 5).

To test whether the dopamine-induced increase in [Ca²⁺]_i is attributable to Ca²⁺ release from intracellular stores or to Ca²⁺ influx from the extracellular space, we applied 1 μmol l⁻¹ dopamine in a nominally Ca²⁺-free physiological saline. Superfusion of the preparation with Ca²⁺-free saline did not affect the baseline cytoplasmic Ca²⁺ concentration (Fig. 6A). However, under Ca²⁺-free conditions, no elevation in [Ca²⁺]_i could be observed upon stimulation with dopamine. Subsequent superfusion with normal Ca²⁺-containing physiological saline restored the ability of the duct cells to produce an increase in [Ca²⁺]_i upon stimulation with dopamine. The results of six independent experiments, such as that illustrated in Fig. 6A, are summarized in Fig. 6B. The slightly higher [Ca²⁺]_i under Ca²⁺-free conditions in the presence of 1 μmol l⁻¹ dopamine is attributable to the finding that, after the first control stimulation with dopamine, [Ca²⁺]_i returned only very slowly to its pre-stimulus baseline concentration. These experiments show that the dopamine-induced elevation in [Ca²⁺]_i is completely attributable to Ca²⁺ influx across the basolateral plasma membrane of the duct cells.

In many systems, stimulus-induced Ca²⁺ entry into cells is blocked by La³⁺. The experiment illustrated in Fig. 7A shows that, in the presence of 100 μmol l⁻¹ La³⁺, dopamine fails to stimulate an elevation in [Ca²⁺]_i. The La³⁺ block is almost irreversible, and no second stimulation could be obtained with dopamine after washing out the La³⁺. The results from five La³⁺ experiments are summarized in Fig. 7B.

To characterize further the dopamine-induced increase in [Ca²⁺]_i, we tested whether it was mediated by a voltage-sensitive Ca²⁺ entry pathway. For this purpose, we superfused the preparation with a physiological saline in which [K⁺] was elevated to 100 mmol l⁻¹ ([Na⁺] was reduced to 70 mmol l⁻¹). This high-K⁺ medium depolarized the cells by 49±1.0 mV (N=5); [Ca²⁺]_i, however, was unaffected by this depolarization (results not shown).

Unstimulated salivary duct cells had basolateral membrane potentials, PD_b, of −67±0.9 mV (N=10). Serotonin had no effect on PD_b at concentrations up to 10 μmol l⁻¹ (results not shown), while 1 μmol l⁻¹ dopamine produced a slowly rising and long-lasting depolarization of PD_b to 41±2.3 mV (N=10) (Fig. 8).

We tested whether the dopamine-induced depolarization of the duct cells was Na⁺-dependent. For this purpose, we superfused the preparation with a physiological solution in which [Na⁺] was reduced to 10 mmol l⁻¹ by substituting 150 mmol l⁻¹N-methyl-D-glucamine (NMDG) for Na⁺. Superfusion of the preparation with this low-Na⁺ medium produced a small liquid junction potential of several millivolts (Fig. 9). Addition of 1 μmol l⁻¹ dopamine to the low-Na⁺ saline produced no depolarization. The dopamine-induced depolarization developed reversibly, however, when the normal extracellular Na⁺ concentration was re-established in the continuous presence of dopamine (Fig. 9). Thus, the dopamine-induced depolarization requires a high Na⁺ concentration on the basolateral side of the salivary ducts. It should be mentioned that dopamine, although it did not depolarize the duct cells in low-Na⁺ saline, did elevate [Ca²⁺]_i under these conditions (results not shown).

We next tested whether 100 μmol l⁻¹ La³⁺, which blocks the dopamine-induced elevation in [Ca²⁺]_i, also affects the depolarization induced by dopamine. La³⁺ blocked the depolarization almost completely. The remaining depolarization was only 14±1.8 % (N=3) of the control depolarization and was transient (results not shown).

We attempted to correlate the kinetics of the dopamine-induced depolarizations and changes in [Ca²⁺]_i by simultaneous intracellular electrical recordings and Ca²⁺-imaging. In these experiments, the microelectrode tip was localized using differential interference contrast optics, and the region of interest for localized Ca²⁺ measurements was positioned in the fluorescent Ca²⁺ images over the tip of the microelectrode (Fig. 10). The result of such a simultaneous [Ca²⁺]_i and electrical recording is shown in Fig. 11A, together with the raw fluorescence at 340 and 380 nm excitation in Fig. 11B. The first 300 s of this experiment at a higher time resolution are demonstrated in Fig. 11C, during which the time points A–F mark the times when the Ca²⁺ images in Fig. 10A–F were recorded. A comparison of the [Ca²⁺]_i trace and the PD_b trace in Fig. 11A and Fig. 11C shows (1) that the onset of the elevation in [Ca²⁺]_i seems to lag behind the onset of the depolarization by approximately 90 s, (2) that, after this delay, [Ca²⁺]_i rises faster than the depolarization, (3) that [Ca²⁺]_i starts to decrease in the presence of dopamine, whereas the cell is still depolarizing slowly, (4) that, after the dopamine has been washed out, [Ca²⁺]_i recovers monotonically to baseline values, whereas PD_b starts to recover only after a delay of approximately 100 s but then decreases faster than [Ca²⁺]_i. In three further identical experiments, the time differences between the onset of the dopamine-induced depolarization and the increase in [Ca²⁺]_i were 20, 60 and 90 s. The membrane potential always rose earlier than [Ca²⁺]_i. In some experiments (Fig. 12A,B), the often observed initial decrease in [Ca²⁺]_i after dopamine stimulation started simultaneously with the depolarization.

To understand the time delay between the electrical response and the change in [Ca²⁺]_i, it is necessary to inspect the raw traces of the fluorescence measured at 340 and 380 nm excitation (Fig. 11B) and to keep in mind that an increase in [Ca²⁺]_i causes antiparallel changes in the 340 and 380 nm signals: the fluorescence excited at 340 nm rises, whereas the fluorescence excited at 380 nm falls. A parallel rise or fall of both signals occurs when the dye concentration changes, e.g. through changes in cellular volume. At the onset of the depolarization, the fluorescent intensities excited at 340 and 380 nm fall faster than in the unstimulated duct (bleaching), whereas [Ca²⁺]_i appears to remain unchanged (Fig. 11A,B). Only when [Ca²⁺]_i starts rising, do the 340 and 380 nm signals change in an antiparallel way. The most likely explanation for the apparent delay in the onset of the elevation in [Ca²⁺]_i with respect to the depolarization is a dopamine-induced cell swelling, resulting in a decrease in the intracellular dye concentration. An increase in cell volume, not apparent in electrical recordings, could cancel a small initial Ca²⁺-influx-induced elevation in [Ca²⁺] and delay the measurable increase in[Ca²⁺]_i

The main finding of the present study is that the neurotransmitter dopamine, which has been shown to stimulate fluid secretion in the salivary glands of the cockroaches Periplaneta americana and Nauphoeta cinerea (Just and Walz, 1996; for a review, see House and Ginsborg, 1985), also stimulates the salivary gland duct cells downstream of the acini. This result suggests that the most likely function of the salivary gland duct cells, i.e. the modification of the primary saliva, is under the control of the same neurotransmitter that stimulates fluid secretion by the peripheral cells in the acini.

The most likely sources of the dopamine in intact animals are the axons of the paired SN1 neurones. These two neurones are located in the suboesophageal ganglion and innervate the salivary glands via nerve 7b, i.e. the salivary nerve that projects to the glands by following the salivary ducts and is spatially closely juxtaposed to the ducts. The SN1 neurones are most probably dopaminergic, because their axons have been shown to have tyrosine-hydroxylase-like immunoreactivity (Elia et al., 1994; for a review and critical discussion, see Ali, 1997). With respect to a possible dopaminergic input from the salivary duct nerve to the duct cells, it is important to note that small tyroxine-hydroxylase-like immunoreactive neurohaemal processes are associated with the salivary ducts in Carausius morosus (Ali and Orchard, 1996). It has been proposed that these processes, which have not yet been described for either P. americana or L. migratoria, activate the salivary duct cells to induce transport processes for the modification of the primary saliva. In this study, we have now been able to show directly that dopamine has at least two effects on the duct cells: a Na⁺-dependent depolarization of the basolateral membrane that can be blocked by La³⁺, and an increase in [Ca²⁺]_i, which is absent in Ca²⁺-free physiological saline and is also blocked by La³⁺. The latter results indicate that the dopamine-induced elevation in [Ca²⁺]_i is due entirely to an influx of Ca²⁺ across the basolateral membrane.

The temporal pattern of the dopamine-induced elevation in [Ca²⁺]_i requires some attention. Agonist-induced Ca²⁺ signals in a large variety of cells have been shown to exhibit complex patterns: low agonist concentrations evoke periodic oscillations in [Ca²⁺]_i arising from cyclical Ca²⁺ release from, and reuptake into, the endoplasmic reticulum (and/or from Ca²⁺ movements across the plasma membrane), whereas high agonist concentrations cause biphasic elevations in [Ca²⁺]_i consisting of an initial transient followed by a sustained plateau (Berridge, 1993; Fewtrell, 1993). In P. americana salivary duct cells, dopamine causes slowly rising sustained elevations in [Ca²⁺]_i. Thus, information about the agonist concentration is coded in the amplitude rather than the frequency of the Ca²⁺ signal. Such purely amplitude-modulated signalling is rare (Berridge, 1997). Among the few examples are Ca²⁺ signals in B lymphocytes (Dolmetsch et al., 1997) and invertebrate microvillar photoreceptors (Levy and Fein, 1985; Walz et al., 1994). In the tubular salivary glands of the blowfly Calliphora erythrocephala, the neurohormone serotonin evokes intercellular [Ca²⁺] waves and intracellular [Ca²⁺] oscillations. The [Ca²⁺] waves propagate at rates of 12–17 μm s⁻¹ (Zimmermann and Walz, 1997), which is approximately four times faster than the spreading Ca²⁺ ‘tide’ in the cockroach salivary ducts measured in the present study.

The depolarizing effect of dopamine on the salivary duct cells is diametrically opposite to the effects that dopamine exerts on cockroach salivary gland acinar cells. In the latter, dopamine has been shown to cause a Ca²⁺-dependent hyperpolarization, possibly mediated by putative D₂-like receptors (Evans and Green, 1990, 1991; Gray et al., 1984; Ginsborg et al., 1980). This indicates that dopamine activates other, as yet unknown, transduction mechanisms in the duct cells.

We have tried to obtain information on whether both signals, i.e. the depolarization and the elevation in [Ca²⁺]_i, follow the same time course and, thus, might be mediated by the same mechanisms or by functionally closely linked mechanisms. At first sight, the elevation in [Ca²⁺]_i seems to lag behind the onset of the depolarizing response by 20–90 s. A comparison of the two signals is complicated by the observation that, in some preparations, the elevation in [Ca²⁺]_i is preceded by a brief drop in [Ca²⁺]_i and that the latter starts simultaneously with the depolarization. Close inspection of the raw fluorescence signals, however, indicates that dopamine most probably induces an increase in volume of the duct cells; this might mask a small initial Ca²⁺-influx-induced increase in [Ca²⁺]_i. Activation of voltage-gated Ca²⁺ channels by the depolarization seems unlikely, because an elevation of the K⁺ concentration in the superfusate to 100 mmol l⁻¹ depolarizes the duct cells but produces no increase in [Ca²⁺]_i (results not shown). The definitive explanation of the mechanistic basis of the depolarization and the elevation in [Ca²⁺]_i warrants further experiments to address these questions specifically. The observation, however, that dopamine does not depolarize the duct cells but elevates [Ca²⁺]_i in low-Na⁺ saline together with the gross differences in the time courses of the two signals suggest that the depolarization and the Ca²⁺ influx are mediated by different mechanisms.

The luminal plasma membrane of the salivary duct cells in P. americana is equipped with a V-H⁺-ATPase (Just and Walz, 1994b). In several insect ion-transporting epithelia, this proton pump is used to energize plasma membranes in order to drive secondary transport processes (Wieczorek et al., 1991; Klein, 1992). Little is known about the mechanisms that regulate apical proton pump molecules. An additional important subject for future studies will be whether V-H⁺-ATPases can be regulated directly or indirectly by changes in [Ca²⁺]_i.

We wish to thank Drs Bernhard Zimmermann and Otto Baumann for methodological help and critical discussions and Dr R. Theresa Jones for linguistic corrections.