[The salivary glands within the history of physiology]. / Salivkörtlarna i fysiologins historia.

In the rapid development of physiology during the second half of the 19th century the salivary glands played a remarkable part. Numerous of the famous physiologists of that age used these glands in experiments, which yielded new knowledge of far-reaching importance, particularly to neurophysiology. Nerves with hitherto unknown functions were found when Carl Ludwig discovered the secretory nerves and Claude Bernard the vasodilator nerves in experiments on the submaxillary gland. Motor nerves to myoepithelial cells were first observed in parotid glands. The conception of an "autonomic nervous system", composed of a parasympathetic and a sympathetic part, was created by Langley after observations, inspired by decades of work on salivary glands. Pavlov based the idea of the "conditioned reflexes", through which he studied the function of the cerebral cortex, mainly on experiments on salivary glands. Present research on nonadrenergic non-cholinergic transmission of nerve impulses has its origin in Heidenhain's observations on the action of atropine on salivary glands. The understanding of the secretory process in general was also promoted in these early experiments on salivation. Results from simultaneous measurements of the pressure in the sallivary duct and the arterial system, carried out by Ludwig, excluded the then current hypothesis that saliva is formed by filtration of the blood. Barcroft demonstrated that activity in salivary glands is accompanied by increased consumption of oxygen and formation of carbon dioxide. The salivary glands have also attracted physiologists in more recent years. For instance, the first observations on the electrical events on the cellular level in glands were made in submaxillaries; and the first normal tissue in which the presence of the nerve growth factor was found, was the submaxillary gland of the mouse.

Nerve-induced secretion of parotid acinar granules in cats.

Electron microscopy of cat parotid glands revealed great heterogeneity in the secretory granules of normal unstimulated acinar cells. Electrical stimulation of the parasympathetic nerve to the gland evoked a copious flow of parotid saliva which was accompanied by an extensive depletion of the secretory granules from the acinar cells. Exocytosis was captured as it was occurring by means of perfusion-fixation, and showed that the events occur in a conventional manner. Stimulation of the sympathetic nerve caused only a very small flow of saliva, and no acinar degranulation was detected. It can be concluded that the parasympathetic secretomotor axons provide the main drive for parotid acinar degranulation in the cat. This contrasts with the rat in which sympathetic impulses provide the main stimulus for parotid acinar degranulation. These dissimilarities serve to emphasise how extensively species differences may influence autonomic responses in salivary glands.

Protein secretion from the cat parotid gland on nerve stimulation.

1. The output and composition of proteins in nerve stimulated saliva samples were compared. 2. Protein output upon parasympathetic stimulation was higher than following sympathetic stimulation and was accompanied by an obvious degranulation of acini with the former but not the latter. These events are the converse of those in the rat parotid gland. 3. Superimposition of sympathetic upon parasympathetic stimulation caused an augmented output of salivary protein. 4. Electrophoresis of salivas revealed differences between individual cats in protein composition but not between differently stimulated salivas.

Secretion of amylase from the rat parotid salivary gland after degeneration of the auriculotemporal nerve.

The output of amylase into saliva secreted after injection of methacholine or substance P was increased after parasympathetic denervation, but the salivary concentration of amylase was unchanged. The increased output corresponded to the increased flow. Isoprenaline injected during the methacholine-induced secretion raised the output, more being secreted from the denervated than from the contralateral gland. Vasoactive intestinal peptide, given while substance P caused salivation, also increased the amylase output, but equally from the two glands.

Nerve interactions in salivary glands.

In the salivary reflex, not only secretory cells are activated, but also myo-epithelial cells are contracted to support these cells and promote the flow of saliva, and blood vessels dilate to meet the increased demands of the tissues. The various effector cells often receive nerves from both parts of the autonomic system, and interactions may occur when the nerves act on the same type of effector, or on different types of effectors. While in an experiment electrical stimulation of the sympathetic trunk may decrease a parasympathetic salivary flow by causing marked vasoconstriction, this does not occur in the salivary reflex, since the vasoconstrictors do not take part. On the contrary, the normal sympathetic vasoconstrictor tone of the resting gland is easily overcome by activity in parasympathetic vasodilator nerves when secretion starts. Pronounced synergism can be demonstrated between sympathetic and parasympathetic secretory nerves. In dogs, for instance, in which sympathetic secretion is beta-adrenoceptor-mediated, this is marked in the case of fluid secretion. In rats and rabbits, in which beta-receptors elicit secretion of amylase, the potentiating interaction among the nerves is striking when amylase secretion is considered. Even the random release of acetylcholine from the post-ganglionic parasympathetic axons, by itself insufficient to evoke secretion, can increase the sympathetic effects. Motor nerves interact with secretory nerves by causing myo-epithelial contraction, mechanically promoting secretion. Interactions between the nerves in their long-term regulatory function on the sensitivity of the acinar secretory and myo-epithelial cells can also be demonstrated.

Amylase-secretory responses to sympathetic nerve impulses in the rat parotid gland following partial sympathetic ganglionectomy.

Some sympathetic axons reach the parotid gland from the contralateral sympathetic chain. Such contralateral nerves were stimulated and saliva secreted after intravenous injections of methacholine was collected from the parotid duct at various times after removal of the ipsilateral superior cervical ganglion. The secretion of amylase caused by sympathetic nerve stimulation was greatly increased three days after the sympathectomy and even more after ten weeks. This effect is attributed to denervation supersensitivity, prejunctional after three days and, in addition, postjunctional later.

Amylase in parotid saliva of rats after sympathetic nervous decentralization.

The marked increase in the amylase secretion that occurs in parasympathetic saliva some days after sympathetic ganglionectomy occurred after decentralization also; hence, it results from lack of impulses from the central nervous system. It occurred whether salivation was evoked by activating muscarinic receptors with methacholine, peptidergic receptors with substance P or physalaemin, or alpha-adrenoceptors with phenylephrine. Unilateral operation affected to some degree the contralateral gland also.

Adrenergic nerves to the rat parotid gland originating in the contralateral sympathetic chain.

The parotid gland of the rat seems to receive some adrenergic nerves from the sympathetic chain of the opposite side. This is suggested by the following evidence: after unilateral removal of the superior cervical ganglion, parotid tissue from the contralateral gland shows degeneration secretion of amylase in vitro similar to, but much smaller than that known to occur ipsilaterally. When parotid secretion is evoked parasympathetically in the anesthetized rat, superimposed stimulation of the contralateral cervical sympathetic trunk can be shown to increase the secretion of amylase into this parasympathetic saliva; as it does, much more, ipsilaterally. It may also cause an evanescent decrease of the salivary flow, suggesting that not only secretory, but also vasoconstrictor nerves had been activated. After removal of one sympathetic ganglion, some undergenerated adrenergic nerves remain ipsilaterally, as earlier demonstrated; but no such fibers can be detected when the ganglion has been removed on both sides.

Degeneration secretion of amylase from the parotid gland of the rat.

In anaesthetized rats the parotid saliva secreted during degeneration of the auriculo-temporal nerve contains amylase in fairly constant concentration, the amylase output varying with the secretory rate. The amount of amylase was not reduced by adrenoceptor blocking agents; it was increased by sympathectomy carried out one week in advance. A sympathetic degeneration secretion of amylase could also be demonstrated while the post-ganglionic sympathetic nerves were degenerating; a simultaneous parasympathetic degeneration secretion provided the salivary flow needed for the transport of the amylase.

Parotid degeneration secretion of amylase in vitro in the rat following sympathectomy.

After extirpation of the superior cervical ganglion in the rat a degeneration secretion of amylase occurs in an in vitro preparation of the parotid gland. It can be detected about 14 h after the sympathectomy, reaches a maximum after about 17 h and then slowly subsides. It is abolished by atenolol, but not by dihydroergotamine or atropine, nor by tetrodotoxin. From this it is concluded that the phenomenon is due to an action mainly on beta 1-adrenoceptors exerted by noradrenaline, which is released from the degenerating sympathetic nerves at the neuro-glandular junctions, independently of propagated nerve impulses.

Nervous control of mammalian salivary glands.

In the production and flow of saliva, sympathetic and parasympathetic nerves generally cooperate, although variations between the different salivary glands are considerable, particularly in the sympathetic innervation. In the submandibular gland of the dog, sympathetic impulses cause secretion via beta-adrenoceptors, and since sympathetic motor effects are elicited via alpha-adrenoceptors it is possible to study separately motor and secretory effects in this gland. Such experiments indicate that myoepithelial contractions serve to accelerate the salivary flow and to support the secreting acinar cells and prevent back-flow of fluid from the luminal system into the glandular tissues. The contractions are elicited reflexly from the oral mucosa together with secretion. A potentiation interaction between sympathetic and parasympathetic nerves occurs in the formation of the primary saliva. In parotid glands of rabbits and rats such an interaction has been demonstrated in the secretion of amylase.

Functional and structural studies concerning the control of activity in zygomatic glands of cats.

The zygomatic gland of the cat consists of at least two parts, separated by a septum, and each drains by a separate duct opening on a mucosal ridge postero-medial to the parotid orifice. It is composed on thin-walled ducts and tubulo-acini, containing mainly mucous cells, but scattered cap cells are also present. In cats under chloralose anaesthesia there is a spontaneous flow of extremely viscous saliva and often, in addition, there is a reflexly elicited component to the secretion. In fact, in contrast to the other major salivary glands, the zygomatic gland is easily made to secrete reflexly even in deep] anaesthesia, e.g. by pinching ipsilateral parts of the tongue, by stimulation of the oesophagus mimicking swallowing, or by afferent excitation of ipsilateral lingual, glossopharyngeal or vagal nerves. The efferent link of the reflex arc is contained in the buccal branch of the mandibular nerve. Section of this nerve abolishes these reflexes, and two weeks later a great loss of acetylcholinesterase positive nerves can be demonstrated in the gland. Efferent stimulation of the buccal nerve evokes a lively secretion that is not affected by hexamethonium but is abolished by atropine. Histochemically, adrenergic nerves are also found surrounding the acini and these nerves disappear after excision of the superior cervical ganglion. Electrical stimulation of the cervical sympathetic trunk causes some secretion, mainly by way of beta 1-adrenoreceptors. Myoepithelial cells are present around the tubulo-acini, and indications of the effect caused by their contraction on the flow from the glands have been observed. Such activity can be induced reflexly and it is then abolished by cutting the buccal nerve or by injecting atropine.

Sympathetically evoked secretory potentials in the parotid gland of the cat.

1. In cats under chloralose anaesthesia micro-electrodes were inserted into parotid gland cells. 2. The average resting potential was found to be -35 . 6 +/- 4 . 7 (S.D.) mV. 3. Stimulation of the auriculo-temporal nerve caused hyperpolarizing, or occasionally depolarizing, secretory potentials of 5--10 mV, which were abolishable with atropine. 4. Stimulation of the cervical sympathetic trunk regularly caused, after long latency (several seconds), slow depolarizations of 15--20 mV, accompanied by a decrease in input resistance. They were antagonized by practolol and therefore assumed to be mediated by beta 1-adrenoceptors. Occasionally hyperpolarization, ascribed to an effect on alpha-adrenoceptors, was observed. 5. In many units the slow depolarization on sympathetic stimulation was preceded by short-lasting hyperpolarizing (sometimes depolarizing) transients. They resembled the evoked cholinergic responses but could be abolished not only by atropine but also by guanethidine.

Amylase secretion from rat parotid glands as dependent on co-operation between sympathetic and parasympathetic nerves.

A slow, long-lasting 'degeneration secretion' from the parotid gland was brought about in anaesthetized rats by section of the auriculo-temporal nerve 16--19 h in advance. This parasympathetic background activity greatly increased the secretion of amylase elicited by sympathetic nerve stimulation.

Parasympathetic degeneration secretion of saliva in rats.

In rats under chloralose anaethesia saliva was found to flow from the submandibular and parotid glands previously subjected to (partial) postganglionic parasympathetic denervation. Secretion started in the submandibular glands 8.8-11.8 hours, and in the parotid glands 14.0-12.6 hours after the denervation and lasted about 7 hours in both glands. It was not abolished by sympatholytic drugs but by atropine. It is regarded as an example of the "degeneration activity" described in many organs and species and provides a method for prolonged stimulation of salivary glands in rats.

Activities of salivary myoepithelial cells: a review.

Historical developments concerning salivary myoepithelial cells have been outlined and structural features of the cells have been described. Evidence from structural and functional studies supports the belief that myoepithelial cells usually have a dual innervation by parasympathetic as well as sympathetic nerves, and impulses from both types of nerve cause the cells to contract. Functional assessment using salivary flow phenomena and intraluminal pressure changes have been used to determine, so far as is possible, the effects of myoepithelial contractions. In some instances tissues have been examined structurally after such experimental procedures. These investigations indicate that salivary myoepithelial activity 1. Speeds up the outflow of saliva 2. Reduces luminal volume 3. Contributes to the secretory pressure 4. Supports the underlying parenchyma 5. Helps salivary flow to overcome increases in peripheral resistance. However, beyond a certain point, this may lead to sialectatic damage of striated ducts with increase in glandular permeability. Myoepithelial activity may also help to expel parenchymal cell contents in certain instances.

Supporting effects of myoepithelial cells in submandibular glands of dogs when acting against increased intraluminal pressure.

1. In dogs under chloralose-urethane anaesthesia secretion from the submandibular gland was recorded with the outflow at gland level and at heights of up to 50 cm above the gland.2. With the outflow level increased, secretion elicited by sympathetic nerve stimulation was far better maintained before than after injection of the alpha-adrenoceptor blocking agent dihydroergotamine.3. When the outflow level was raised while no secretion occurred, fluid flowed into the gland. Part of this amount was returned on lowering the outflow to gland level. This volume was assumed to have been accommodated in the distended luminal system, whereas some fluid was obviously lost into the gland tissues.4. Both these fractions of the fluid flowing into the gland when the outflow level was high could be reduced by injecting the alpha-receptor agonist phenylephrine. Bradykinin, which like phenylephrine is known to contract salivary myoepithelial cells, had the same effect on the two inflow volumes.5. It is concluded that myoepithelial contraction reduced the distensibility of the luminal system and in addition supported the acini, thereby diminishing backflow into the glandular tissues and enabling the gland to discharge saliva against a high pressure.6. Morphologically it was found that in resting glands PAS-positive saliva was displaced from the ductal system when the outflow cannula was raised, but it was preserved in the lumina under similar conditions when the myoepithelial cells were being stimulated by phenylephrine or bradykinin.7. Although sympathetic secretion could be maintained against a head of pressure, so long as it was accompanied by myoepithelial contraction, the increased force caused by the secretion led to disruptive damage of striated ducts which are the first part of the luminal system not supported by myoepithelial cells.8. The morphological findings reinforce the belief that contraction of myoepithelial cells gives active support to the underlying parenchyma.

On the existence of parasympathetic motor nerves to the submaxillary gland of the dog.

Duct pressure and salivary flow were recorded in submaxillary glands of anaesthetized dogs, to study whether parasympathetic stimulation caused effects referable to activity in myoepithelial cells. At fairly low frequency of stimulation, e.g. 3 Hz, the pressure curve had two distinct components, with initial steep and a secondary gradual rise. It resembled that obtained on sympathetic stimulation, where the first phase is ascribed to myoepithelial contraction, the second phase to secretion. When parasympathetic stimulation ceased, there was first a steep fall, then a more gradual decline of the pressure. The steep fall was of the same magnitude as the steep rise; both increased with the frequency of stimulation. The size of the initial fall was fairly independent of the pressure level from which it started. Such a steep fall did not occur subsequent to parasympathetic stimulation if the myoepithelial cells were already in a state of strong contraction caused by sympathetic impulses or bradykinin. The phase of steep fall was inferred to be due mainly to relaxation of contracted myoepithelial cells, the following decline to back-flow of fluid into the gland. The salivary flow rate was highest at the beginning of a period of parasympathetic stimulation, particularly if the duct system was well filled and the saliva thin. It was concluded that myoepithelial contraction had initially expelled saliva. A brief period of parasympathetic stimulation while a slow basal secretion at constant rate was going on was found to accelerate this flow, and afterwards there was a transient deceleration of the flow. Acceleration was attributed partly to myoepithelial contraction, mainly to superimposed secretion; retardation to myoepithelial relaxation. The effect appeared independently of the way in which the basal flow was evoked, and the retardation resembled that seen after sympathetic stimulation or bradykinin.