Ion and Water Balance in the Ixodid Tick Dermacentor Andersoni: II. Mechanism and Control of Salivary Secretion

The salivary glands of at least two species of ixodid ticks are the major route for water excretion in the adult female during the feeding period (Tatchell, 1967b; Kaufman & Phillips, 1973). The present paper examines: (1) whether fluid production by these glands is generated by means of a secretory (as opposed to a filtration resorption) mechanism, and (2) some of the pharmacological properties of the salivary gland.

Inulin traverses most filtration membranes recognized to date (capillaries of the vertebrate glomerulus (Pitts, 1963); coelomic sac of the crayfish antennal gland and pericardium of the octopus (Riegel, 1972); filtration membrane of the tick coxal gland (S. Kaufman, 1971), whereas the salivary glands of Dermacentor andersoni are completely impermeable to inulin (Kaufman & Phillips, 1973). This suggests that, in the latter gland, fluid may be produced by means of a secretory mechanism since secretory systems are usually impermeable to molecules as large as inulin. Even when inulin is permitted to enter a secreted fluid, the urine-to-plasma ratio (U/P) is less than unity (e.g. the Malpighian tubules of Carausius morosus (Ramsay & Riegel, 1961)). Much more compelling evidence for the secretory hypothesis would be to demonstrate that tick salivary glands can produce fluid in the absence of an externally applied hydrostatic pressure, for example, by removing the glands from the body and stimulating them to function in vitro. This approach has been previously used for studying fluid secretion by insect Malpighian tubules (Ramsay, 1954, 1955; Maddrell, 1964,a, b, c; Berridge, 1966a, b; Irvine, 1969; Pilcher, 1970a, b) and salivary glands (Berridge & Patel, 1968 ; Berridge, 1970). In view of the successes of the above-mentioned authors, an in vitro preparation of the salivary glands of Dermacentor andersoni was attempted. Furthermore, it was hoped that an examination of those agents which are capable of stimulating salivation in vitro would provide clues as to whether fluid secretion is controlled by hormones or nerves.

The adult female ticks used in this study were taken from a colony which had been cold-diapaused for at least 3 months. The rearing procedures and conditions were previously described (Kaufman & Phillips, 1973).

Salivary glands were excised from partially fed ticks (normally 40–300 mg).

Partially fed ticks were immobilized dorsum-up with modelling clay in a small dish, covered with a simple Ringer solution (dissection medium l, Table 1) and dissected to reveal the full extent of the salivary glands and ducts (figured well by Gregson, 1960). To free the glands from the body, tracheae and other connective tissue were cut away; the portion of the pharyngeal wall just beneath the salivarium at the origin of the ducts was then sliced, thus releasing the region around the ducts’ orifices from the buccal cavity region. The two glands were transferred in a medicine dropper to a Petri dish containing fresh Ringer solution, and separated one from the other by severing the cuticle between the two orifices.

A fine strand of silk thread was tied to the cuticle associated with each duct and thus the gland could be manipulated. The gland was then cleared of extraneous tissue and transferred to a drop of bathing medium (normal Ringer solution, Table 1) sitting under liquid paraffin in a Petri dish lined with Parafilm. A short piece of glass capillary was used to anchor the drop of bathing medium to the Parafilm. The proximal portion of the salivary duct was then drawn out of the drop of medium into the liquid paraffin by means of the silk thread. When the appropriate stimulus was applied, fluid emerged from the orifice of the duct and assumed spherical shape in the liquid paraffin. Although the drop of bathing medium was not continuously oxygenated, pure oxygen was bubbled through the stocks of media from which the drop bathing the gland was frequently renewed.

The temperature of the liquid paraffin was maintained at 25–29 °C by placing the Petri dish on a water-circulation plate (A. H. Thomas Co.) connected to a constant temperature bath. During any single experiment, however, the temperature did not vary by more than 0·5 °C. The volume of the secreted droplet was calculated from its radius which was measured by an ocular micrometer through a dissecting microscope.

Initially, only a few isolated glands showed any sign of secretory activity when exposed to the complete tissue culture medium of Rehacek & Brzostowski (1969). When bathed in the complex medium of Berridge (1966b) about 10% of the excised glands secreted saliva, but of the latter, no gland ever secreted as much as 1 μl of fluid, and the maximum rate of secretion at room temperature was 5 nl/min. In order to improve upon the latter performance, some potential stimuli were tested on glands set up in Berridge’s medium. The following were without effect:

• Augmentation of the bathing medium with fresh haemolymph (up to 30%,v/v) ; pure haemolymph also failed to initiate secretion even though the donors were in the active salivation stage of the feeding cycle (13 preparations).

• Dilution of the medium with an equal volume of distilled water (3 preparations).

• Vasopressin at doses ranging from 0·125 to 13 m.u./ml (4 preparations).

• Since the cholinomimetic drug pilocarpine stimulates salivation in vivo (Howell, 1966; Tatchell, 1967a; Purnell, Branagan & Radley, 1969), pilocarpine (10⁻⁷ to 10⁻²M; 12 preparations) and acetylcholine (10⁻³M; 12 preparations) were tested on excised glands; neither of these was effective.

Finally, a marked stimulation of secretion was consistently induced by adrenaline, with a threshold dosage of approximately 10⁻⁶ M (Fig. 1). The usual rate of secretion at 10⁻⁴ M was 80–130 nl/min compared to a maximum of 5 nl/min in the absence of adrenaline. Notwithstanding these initial successes, the rate of secretion decayed steadily with time. Several modifications of medium and protocol were tested in order to establish the optimal conditions for secretion. Secretory rates were inhibited at pH below 6 and above 8 so all media were subsequently adjusted to a pH close to neutrality. Secretory activity improved somewhat when dissection medium 2 (Table 1) replaced dissection medium 1, and when the NaCl concentration of Berridge’s medium (1966b) was tripled (see normal Ringer solution, Table 1).

Although the in vitro preparation gradually improved with the above modifications, the rate of secretion declined considerably during the first 2 h after dissection. This decay in rate was possibly due to the continual exposure of the glands to adrenaline. To test this hypothesis, the following experimental protocol was introduced.

Instead of continually exposing glands to adrenaline throughout the experimental period, glands were left in the adrenaline-containing Ringer solution only until a maximum rate of secretion was reached. At that point the Ringer solution was replaced with a medium lacking adrenaline. After several washings with this medium the secretory rate fell to zero. This procedure (a ‘pulse’ of adrenaline with rest periods between) was repeated two or three times throughout the day. There was a noticeable increase in the duration of secretory activity of the ‘pulsed’ glands as compared to glands which were continually bathed in adrenaline; furthermore, the decay in rate (which was still significant), became a linear function of time (Fig. 2). Finally, a number of individual preparations occasionally showed no appreciable decay over the 6 h experimental period. The latter was never observed when glands were subjected to a maintained dosage of adrenaline. In the above experiments the albumin and T. C. yeastolate had been omitted from Berridges’ original complex medium (1966 b). Berridge (1966b) found that the Malpighian tubules of Calliphora secreted urine as rapidly and for similar durations in media lacking albumin as in media containing it. However, it is well established that tissue culture media must be supplemented with protein if most cells are to survive in vitro for extended periods (or even for shorter periods if the cells are very active (Waymouth, 1965)). Consequently, it seemed reasonable to test whether inclusion of albumin in the medium might prolong secretory activity of the tick salivary gland. Fig. 2 shows that although the activity of the glands on the average did not decay beyond the second pulse, the overall decay was undiminished.

The maximum response to a given dose of adrenaline did not occur quickly (Fig. 3). Although secretory activity usually commenced within 5 min of adding adrenaline, the rate of secretion rose steadily, so that the maximum rate eventually achieved occurred approximately 25 min after adding adrenaline. This suggests that it took some time for the concentration of adrenaline at its site of action to come to equilibrium with that in the medium. On removal of adrenaline the response of the gland was more rapid. Whereas the time for attainment of the half-maximal rate after adding adrenaline was 11–12 min, the half time for activity decay after removal of adrenaline was only 2 min (Fig. 3).

In vitro salivary glands were most sensitive to the alpha-adrenergic drugs adrenaline, noradrenaline and dopamine (Fig. 4A-C), but could be stimulated by the beta adrenergic drug isoproterenol, at higher concentrations (Fig. 4D-E). DL-Dopa at 10⁻⁵ M (Fig. 4B) and phenylalanine (2·6 × 10⁻⁴ M in normal Ringer solution), which are intermediates in the biosynthesis of adrenaline (Malmejac, 1964) were inactive. The medium of Rehacek & Brzostowski (1969) and that of Berridge (1966b) each contained a potential neuroactive substance (malate and glutamate respectively), yet as mentioned earlier neither of these media alone stimulated salivation in a consistent manner. 5-hydroxytryptamine (5-HT) was active only at 10⁻³ M (Fig. 4A) but 10⁻² M cyclic AMP (Fig. 4B) was unable to stimulate secretion.

To test whether adrenaline could also stimulate salivation in vivo, a 10⁻⁴ M solution in 1·2% NaCl (approximately iso-osmotic with the haemolymph) was injected into ticks according to the following procedures:

(A) Without injection, most feeding ticks (over 60 mg) can be coaxed to secrete a limited volume of saliva into glass capillaries (‘spontaneous secretion’). Adrenaline was injected into such ticks to see whether they would resume salivating. Control ticks were injected with saline alone.

(B) Adrenaline was injected immediately after removing the ticks from the host in order to see whether the rate and volume of spontaneous secretion could be increased above the values characteristic of non-injected ticks. Control ticks were injected with saline alone.

(C) After some of the experimental ticks in ‘B’ above had terminated salivation, they were re-injected with a volume of the adrenaline solution equal to that of the secreted saliva. If salivation did not resume within 10 min, the injection was considered to be ineffective.

The yield of saliva from ticks injected with adrenaline was increased significantly over that from control ticks provided that injection was accomplished before the ticks had salivated spontaneously (Table 2). Thus adrenaline elicits a profound response from the salivary gland both in vivo and in vitro, but pilocarpine appears to exert an effect only in vivo. Furthermore, adrenaline is more potent than pilocarpine as an in vivo stimulant, since we estimate that the final concentration of the former in the haemocoele of Dermacentor (2 ×10⁻⁵ M) was between th and th the concentration of pilocarpine injected into other ticks by previous workers (see Discussion).

The salivary gland of Dermacentor andersoni will produce fluid in the absence of an externally applied hydrostatic pressure gradient; according to the criteria proposed by Kirschner (1967), this constitutes firm evidence that salivation occurs by means of a secretory rather than a filtration-resorption process. However, the in vitro salivary glands will not secrete unless one of a variety of catecholamines (e.g. adrenaline, noradrenaline, dopamine) is included in the bathing medium. The latter suggests (coupled with the fact that freshly collected haemolymph from salivating ticks does not induce secretion in vitro) that salivation might normally be under nervous (cate-cholaminergic) rather than hormonal control. We have other information that is consistent with this hypothesis. (1) Nerves are seen in close proximity to the tick salivary gland (Meredith & W. R. Kaufman, in preparation). (2) Adrenaline stimulates the in vitro gland to half the maximal secretory rate in 11 min. By way of contrast, the salivary gland of Calliphora (which is not innervated; Oschman & Berridge, 1970) responds maximally to exogenous 5-HT within 60 sec (Berridge, 1970). Perhaps the relatively long time required for stimulation in Dermacentor reflects the existence of a‘synaptic barrier’ (see McLennan, 1963) which shields the receptor sites on the secretory cell from the bathing medium. Although the latter is one way of explaining the relatively slow response of Dermacentor salivary glands, other explanations based on differing drug-receptor affinities cannot be ruled out.

The suggestion that salivation in Dermacentor is mediated by nerves rather than by a‘salivation hormone’ must still remain tentative. Perhaps a hormone does exist, but by the time fresh haemolymph has been collected and applied to the gland (several minutes), the hormone may have been inactivated by an enzyme in the haemolymph. M. J. Berridge (personal communication) found that the diuretic hormone of tse-tse fly haemolymph is short-lived unless suitable precautions are taken to inhibit its degradation. Also, the demonstration that a widely occurring neurotransmitter substance (such as noradrenaline) can stimulate salivation is not by itself sufficient evidence that the normal activation of the gland is mediated by noradrenaline issuing from nerve terminals. For instance, the Malpighian tubules of Rhodnius and Carausius are not innervated, and are normally activated by specific diuretic hormones, yet the tubules of both insects are also highly responsive to 5-HT ; furthermore, 5-HT and the two diuretic hormones in the above insects are three distinct substances (Maddrell, Pilcher & Gardiner, 1971).

Although 10⁻⁵ M adrenaline causes the salivary gland to attain half the maximal secretory rate within 11 min, after removal of adrenaline from the bathing medium the half-maximal rate is reached in only 2 min (Fig. 3). The more rapid decay of secretory rate after removal of adrenaline from the bathing medium suggests the presence of an inactivating system for a transmitter substance near the receptor sites. The demonstration of a monoamine oxidase or a catechol-O-methyl transferase in the tick salivary gland would fulfil one criterion in establishing that a catecholamine-like substance is the natural transmitter substance. ^*

The inability of cholinergic drugs to stimulate secretion of the tick salivary gland in vitro appears to be at variance with the findings of other authors (see below) who use pilocarpine routinely to increase the yield of saliva in vivo. However, these authors injected excessive amounts of pilocarpine into the experimental animals. On the assumption that the haemolymph volume represents approximately 20% of the body weight (as in the case of Dermacentor ; Kaufman & Phillips, 1973), we estimate that the final concentration of pilocarpine was 4 to 6 × 10⁻² M in the haemolymph of Omitho-dorus savignyi (Howell, 1966), 8 × 10⁻² M in the haemolymph of Boophilus microplus (Tatchell, 1967a), and 4 to 12 × 10⁻² M in the haemolymph of Rhipicephalus appen-diculatus (Purnell et al. 1969). On the basis of the present in vitro observations, it is unlikely that pilocarpine exerts its in vivo effect at the secretory membrane itself. Possibly pilocarpine activates neurones in the central nervous system which then stimulate efferent axons innervating the salivary gland (e.g. see Burn & Rand, 1965). Alternatively, this drug may stimulate the absorption of fluid from the gut and that in turn might indirectly lead to events which trigger salivation.

Cyclic AMP has been implicated in many cells as a link between the primary (hormonal) message and the effector system itself (reviewed by Robison, Butcher & Sutherland, 1971). Recently, cyclic AMP has been similarly implicated as a‘second messenger’ in the salivary gland of Calliphora (Berridge, 1970) and the Malpighian tubules of Rhodnius and Carausius (Maddrell et al. 1971). Interestingly, exogenous-cyclic AMP at concentrations as high as 10⁻² M had no stimulatory effect on the salivary gland of Dermacentor. The ineffectiveness of exogenous cyclic AMP in the present case may be explained in one of three ways: (1) cyclic AMP plays no role in salivary fluid secretion; (2) the basal membrane of the secretory cell is only very slightly permeable to cyclic AMP, and (3) in addition to the latter, the cell may possess an extremely active phosphodiesterase which is capable of maintaining the intracellular concentration of cyclic AMP below the threshold for activating fluid secretion in spite of the large external concentration of cyclic AMP. Treatment of glands with methyl xanthines (such as theophylline), which interfere with the action of phosphodiesterase on cyclic AMP (Robison et al. 1971), might clarify this question.

We are grateful to Dr M. J. Berridge for his most helpful comments on a draft of the manuscript. We would also like to extend our thanks to the National Research Council of Canada who generously provided research funds and a scholarship to W.R.K.