Carotid chemoreceptor function and structure in the atherosclerotic rabbit: respiratory and cardiovascular responses to hyperoxia, hypoxia and hypercapnia.

We tested the following hypothesis: if carotid body blood flow, and hence the relationship of the frequency of discharge in chemoreceptor afferent fibres to arterial PO2, were affected by atherosclerotic change, then a modification of the control of the respiratory and cardiovascular systems might result. Carotid body reflexes were therefore studied in conscious atherosclerotic rabbits and a control group of normal animals breathing 100% O2, three hypoxic gas mixtures to which was added sufficient CO2 to maintain the arterial PCO2 constant, and 2% and 4% CO2 in 21% O2 and N2. When breathing room air, the atherosclerotic rabbits breathed at a higher respiratory frequency and lower tidal volume than the normal animals, although there was no difference in the respiratory minute volume. The respiratory and cardiovascular responses to hyperoxia, isocapnic hypoxia and hypercapnia were essentially the same in both groups of animals. Serial sections of the carotid bodies showed pathological changes including interstitial fibrosis in the caudal part with interstitial haemorrhages. The proximal part of the ascending pharyngeal artery, the vessel supplying the organ, and its origin from the external carotid, and the arterioles in the caudal part of the carotid body were nearly always occluded to a varying extent by atheromatous plaques. The capillaries appeared normal under light microscopy. The rostral-caudal lengths of the carotid bodies were similar in the two groups. We conclude that the peripheral arterial chemoreceptor responses in atherosclerotic rabbits are relatively normal even though the arteries to, and arterioles within, the carotid body are partly occluded.

The carotid body of the harbour seal (Phoca vitulina richardsi).

The bilateral distribution of carotid body type 1 and 11 cells was investigated in five harbour seals (Phoca vitulina richardsi), by serially sectioning the carotid bifurcation regions. The cells occurred bilaterally in the animals and were also present in one specimen from a sixth animal available for study. The type 1 and 11 cells were located in the space between the internal and external carotid arteries and had a varied relationship to the occipital and condyloid arteries. They lay within a division of connective tissue with irregular but defineable borders and this combination of connective tissue and type 1 and 11 cells constituted the principal mass of the carotid body. The carotid body occurred in a variety of forms: wedge-shaped, crescentic or horse-shoe shaped, or as a discrete oval structure. In some specimens the carotid body had a central 'neurovascular' core of small blood vessels and nerves. The artery to the organ originated from either the external carotid, internal carotid or common carotid arteries. Using an interactive image analysis system in eight specimens, which had been perfusion-fixed at a normal arterial pressure, the mean volume of the carotid body was 1.666 +/- 0.45 (SD) mm3. Caudally and separate from the principal mass of the carotid body periadventitial type 1 and 11 cells were noted in 4 out of 11 specimens in the connective tissues adjacent to the external carotid artery, origin of the occipital, and the rostral part of the common carotid artery and its bifurcation.

Chemoreceptor reflexes and cardiovascular control.

Evidence is presented which indicates that in the absence of other known inputs to the nervous system and during controlled pulmonary ventilation, stimulation of the carotid body chemoreceptors causes bradycardia and selective peripheral vasoconstriction. These responses may be attenuated, however, by concomitant changes in respiration and arterial blood pressure, and by activity of higher parts of the brain stem. Stimulation of the aortic bodies in mammals in which they are functionally active, causes bradycardia or tachycardia and selective peripheral vasoconstriction. The reflex vascular effects from the peripheral arterial chemoreceptors are mediated by alpha-adrenergic sympathetic fibres. A potential mechanism exists therefore whereby the peripheral arterial chemoreceptors could contribute to the neurogenic component of hypertension.

Cardiovascular responses to stimulation of cardiac receptors in the cat and their modification by changes in respiration.

1. In cats anaesthetized with a mixture of chloralose and urethane, stimulation of cardiac receptors by left atrial injections of veratridine had little or no effect on pulmonary ventilation but caused bradycardia, systemic hypotension and hindlimb vasodilation with a latency of 3.3 s. 2. The hindlimb vasodilatation was due largely, if not entirely, to a reduction in sympathetic vasoconstrictor activity. 3. Similar cardiovascular responses occurred when the arterial blood pressure was maintained constant and also in artificially ventilated animals. 4. When the cardiac receptors were excited during a period of apnoea which was induced reflexly by electrical stimulation of the central cut end of a superior laryngeal nerve, the cardio-inhibitory response to left atrial injections of veratridine was enhanced but the size of the vasodilator response was unaffected. 5. In contrast, the cardiovascular effects of stimulation of the carotid body chemoreceptors, bradycardia and hindlimb vasoconstriction were enhanced by the laryngeal input. 6. The possible central mechanism responsible for the differential modulation of cardiac receptor and carotid chemoreceptor reflexes by respiration are discussed.

The volume of the carotid body and periadventitial type 1 and type 2 cells in the carotid bifurcation regions of the pregnant and lactating cat.

The bilateral distribution of carotid body type 1 cells was investigated in 2 pregnant cats at 95% full term and two lactating cats (2 and 4 d after parturition). Carotid body type 1 and 2 cells occurred bilaterally in close proximity to the occipital artery or one of its branches in a division of connective tissue with defineable but irregular borders. This combination of connective tissue and type 1 and 2 cells constituted the principal mass of the carotid body, which received its blood from the occipital and/or the ascending pharyngeal arteries. Using an interactive image analysis system, the area of the carotid body in each serial section was measured by accurately contouring its perimeter. The volume of the carotid body was calculated by multiplying the sum of the contoured areas of the serial sections by the thickness of the section. The volumes for the carotid bodies in pregnant cats ranged between 0.186 and 0.278 mm3 while the values in lactating animals lay between 0.287 and 0.356 mm3. Caudally and separate from the carotid body, isolated groups of periadventitial type 1 and 2 cells were found in 5 out of 8 specimens around the occipitoascending pharyngeal trunk, origin of the occipital artery, external carotid artery and rostral part of the common carotid artery. The volumes of the periadventitial type 1 and 2 cells were variable in pregnant and lactating cats.

The carotid body of the mini-pig.

The bilateral distribution of carotid body type 1 and 11 cells was investigated in 3 mini-pigs by serially sectioning the carotid bifurcation regions. The majority of type 1 and 11 cells occurred bilaterally in close proximity to the wall of the occipital artery or one of its small proximal branches, the internal carotid artery and the common arterial trunk. A division of connective tissue surrounded the type 1 and 11 cells with defineable but very irregular borders and this combination of connective tissue and cells constituted the principal mass of the carotid body. In the majority of specimens the cells were diffusely arranged giving a fragmented appearance to the organ but in 2 specimens part of the carotid body was discrete, adopting an ovoid or crescent-shaped appearance over a limited rostral-caudal part of its extent. Arteries to the carotid body originated from the occipital arterial tree in the vicinity of the division of the common arterial trunk and less commonly from the common arterial trunk itself. A supplementary blood supply came from unidentified connective tissue arterioles. In 2 out of 6 specimens type 1 and 11 cells were associated with cervical nerve trunks. Periadvential type 1 and 11 cells were observed in one out of 6 specimens lying dorsal to the common arterial trunk. From our data on 6 specimens, three dimensional reconstructions were made of ventral views of the distribution of carotid body type 1 and 11 cells. On account of the diffuse arrangement of the mini-pig carotid body, morphometric analysis of the organ volume was not feasible.

Cardiovascular responses to carotid chemoreceptor stimulation in the dog: their modulation by urinary bladder distension.

Respiratory, heart rate and hindlimb vascular responses were studied in response to increasing levels of stimulation of the carotid body chemoreceptors, together with an examination of the modulation of their effects by distension of the urinary bladder in the dog anaesthetized with a mixture of chloralose and urethane. The vascularly isolated carotid bifurcation regions were perfused with blood, stimulation of the carotid bodies being carried out by three different levels of hypoxic isocapnic blood (PO2 approximately 58, 40 and 22 mmHg) obtained from a donor animal. A vascularly isolated hindlimb was autoperfused at constant blood flow through its femoral artery. In spontaneously breathing animals, increasingly intense hypoxic stimulation of the carotid bodies caused a progressive augmentation of respiratory minute volume. Superimposition of distension of the bladder increased ventilation further, by the same amount during hypoxic as during normoxic blood perfusion of the chemoreceptors. Prevention of the effects of lung stretch afferent stimulation by artificial ventilation modified the heart rate and hindlimb vascular responses to excitation of the carotid bodies by revealing or accentuating the primary cardiovascular responses, bradycardia and vasoconstriction. In contrast, no such respiratory modulation was apparent in the cardiovascular responses to bladder distension. When, under conditions of artificial ventilation and in the absence of changes in the arterial baroreceptor input, the primary cardio-inhibitory and vasoconstrictor responses to carotid chemoreceptor stimulation predominated, the heart slowed progressively as the stimulus was increased. At the same time the cardio-accelerator effects of bladder distension progressively diminished, indicating an interaction between the cardiac reflex responses evoked by the two inputs. In contrast, the reflex vascular responses resulting from stimulation of the two inputs were additive, at least for PO2 levels of carotid body perfusate down to approximately 40 mmHg. In conclusion these experiments demonstrate the differential nature of the integration of respiratory and cardiovascular responses evoked by stimulation of the carotid chemoreceptors and bladder distension.

Reflex cardiac dromotropic responses to stimulation of the carotid and aortic chemoreceptors in the anaesthetized cat.

1. The reflex changes in the dromotropic state of the heart (P-R interval or atrioventricular conduction time) in response to selective stimulation of the carotid and aortic bodies by sodium cyanide were studied in the anaesthetized cat. The heart was paced and the arterial blood pressure was kept constant to minimize secondary effects of changes in arterial baroreceptor activity. 2. Stimulation of the carotid and aortic bodies caused an increase in the atrioventricular conduction time. 3. Evidence is presented to suggest that this negative dromotropic response was due predominantly to a vagal cholinergic mechanism. There is a small sympathetic component but only in so far as the carotid body reflex is concerned. 4. The negative dromotropic responses were enhanced during reflex suppression of the central inspiratory neuronal drive combined with minimal activity of the slowly adapting pulmonary stretch afferents indicating that they are respiratory modulated. 5. The clinical implications of these results are discussed.

The reflex effects of alterations in lung volume on systemic vascular resistance in the dog.

1. The reflex effects of alterations in lung volume on systemic vascular resistance have been studied in anaesthetized dogs under conditions in which the systemic circulation was perfused at constant blood flow. The pressures in the isolated perfused carotid sinuses and aortic arch, and the arterial blood P(O2) and P(CO2) were maintained constant.2. A maintained inflation of the lungs produced by injection of air into the trachea caused a fall in systemic arterial perfusion pressure, indicating vasodilatation. The size of the systemic vasodilator response varied directly with the pressure and volume of gas used to inflate the lungs. A similar effect was observed when the tidal volume of lungs ventilated by an intermittent positive pressure was increased.3. Collapse of the lungs by creating a pneumothorax in closed-chest spontaneously breathing animals evoked a systemic vasoconstrictor response which was reversed when the lungs were re-expanded.4. These vasodilator responses were abolished by dividing the pulmonary branches of the thoracic vagosympathetic nerves. Evidence is presented that the afferent fibres run in the cervical vagosympathetic nerves and through the stellate ganglia.5. The responses were unaffected by atropine, but were abolished by hexamethonium, guanethidine and by bretylium tosylate, indicating that they are mediated via the sympathetic nervous system.6. Evidence is presented that the lungs are a constant course of afferent impulses inhibiting the ;vasomotor centre', and that the lung inflation-systemic vasodilator reflex is a potential mechanism operating in eupnoeic breathing.

Reflex respiratory and cardiovascular effects of stimulation of receptors in the nose of the dog.

1. In forty-one out of forty-seven dogs under chloralose-urethane or Nembutal anaesthesia, mechanical stimulation of the nasal mucous membrane caused a reduction or inhibition of respiration, bradycardia, variable changes of arterial blood pressure, and a small rise in venous pressure.2. Simultaneous measurements of arterial and venous pressures, and also of blood flow in various arteries by means of an electromagnetic flowmeter indicate that the calculated vascular resistance increases in the intact limb, muscle, and skin, and the vascular beds of the vertebral, superior mesenteric, renal and splenic arteries. No changes in vascular resistance occur in the common carotid circulation.3. Evidence is presented that the increase in vascular resistance is due to vasoconstriction, and occurs in the absence of changes in pulmonary ventilation.4. Stimulation of the nasal mucous membrane causes a reduction in volume of the spleen.5. The respiratory and cardiovascular responses are reflex in nature, being abolished by the application of a local anaesthetic to the nose or by combined division of the maxillary and ethmoidal branches of the trigeminal nerves. The cardiac response is mediated largely by the vagus nerves, and the vascular responses by sympathetic adrenergic fibres.6. Cessation of the stimulus to the nose not infrequently results in the following temporary after-effects: hyperventilation, tachycardia, hypertension, and vasodilatation in the intact limb and in muscle.

Distribution of carotid body type I cells and other periadventitial type I cells in the carotid bifurcation regions of the rabbit.

The distribution of carotid body type I and periadventitial type I cells in the carotid bifurcation regions was investigated unilaterally in seven and bilaterally in two New Zealand White rabbits. Carotid body type I cells occurred in close proximity to the wall of the internal carotid artery immediately rostral to the carotid bifurcation, within a division of connective tissue with definable but irregular borders. Caudally, and separate from the main mass of carotid body type I cells, isolated groups of periadventitial type I cells lay freely in the connective tissue around the internal carotid artery and alongside the carotid bifurcation and common carotid artery. A overall picture of the carotid body in the rabbit was reconstructed and the occurrence and significance of periadventitial type I cells discussed.

A comparative study of the distribution of carotid body type-I cells and periadventitial type-I cells in the carotid bifurcation regions of the rabbit, rat, guinea-pig and mouse.

The bilateral distribution of carotid body type-I cells was investigated in five rabbits, rats, guinea-pigs and mice by serially sectioning the carotid bifurcation regions. Carotid body type-I cells occurred bilaterally in close proximity to the wall of the internal carotid artery in the rabbit, rat and mouse and to the wall of the ascending pharyngeal artery in the guinea-pig. The rat carotid body was sometimes recessed into the lateral aspect of the superior cervical ganglion and was the most easily defined organ in the four animals studied. Caudally, and separate from the principal mass of carotid body type I cells, isolated groups of periadventitial type-I cells were observed in the connective tissues around the internal carotid artery and adjacent to the carotid bifurcation and common carotid artery in the rabbits only. An overall picture of the carotid body in the four animals was constructed. In all specimens rostral-caudal dimensions were recorded and compared bilaterally.

Distribution of carotid body type I cells and periadventitial type I cells in the carotid bifurcation regions of the dog.

The bilateral distribution of carotid body type I cells was investigated in 5 mongrel dogs and compared with the arrangement of type I cells in beagles by serially sectioning the carotid bifurcation regions. The distribution of type I cells was not affected by the pedigree of the dogs. In both mongrel dogs and beagles type I cells were arranged in close proximity to the wall of the ascending pharyngeal artery within a division of connective tissue with defineable but irregular borders. Occasionally, type I cells were observed in relation to the occipital and external carotid arteries. This association of type I cells and connective tissue formed the principal mass of the carotid body. Caudally, and separate from the principal mass, isolated groups of periadventitial type I cells lay freely in the connective tissue adjacent to the internal and external carotid arteries in both mongrel dogs and beagles. Less commonly, and in mongrel dogs only periadventitial type I cells were noted alongside the carotid bifurcation and the rostral end of the common carotid artery. Three-dimensional reconstructions of the distribution of carotid body type I cells and periadventitial type I cells from the left and right carotid bifurcation regions were made. In all specimens rostral-caudal dimensions of the distribution of carotid body type I cells and periadventitial type I cells were recorded and compared bilaterally.

Effects of distension of the urinary bladder on the cardiovascular reflexes from the carotid baroreceptors in the dog.

1. The hindlimb vasoconstrictor effects of distension of the urinary bladder were studied at different levels of input from the carotid sinus baroreceptors in the dog anaesthetized with a mixture of chloralose and urethane. 2. The vascularly isolated hindlimb was perfused at constant blood flow through its femoral artery, so that a change in pressure gradient (mean femoral arterial perfusion pressure minus mean inferior vena caval pressure) indicated a similar directional change in vascular resistance. The vascularly isolated carotid sinus regions were perfused with blood at a constant pulsatile flow. 3. Raising the carotid sinus mean perfusion pressure in randomly selected steps of 30 mmHg from 60 to 210 mmHg had little effect on heart rate unless the blood pressure was controlled, when a progressive bradycardia occurred, but caused a progressive reduction in arterial blood pressure and vasodilatation in the perfused hindlimb. Distension of the bladder at each level of carotid sinus pressure resulted in tachycardia, hypertension and hindlimb vasoconstriction. 4. The cardiac responses to bladder distension were the same at all carotid sinus pressures. When the blood pressure was controlled, however, the response was reduced at high and low sinus pressures. 5. The relationship between the carotid sinus perfusion pressure and hindlimb perfusion pressure (i.e. vascular resistance) was affected by distension of the bladder in two ways. In the one, hindlimb perfusion pressure increased by approximately the same amount at all levels of carotid sinus pressure indicating resetting of the carotid sinus baroreceptor reflex control of hindlimb vascular resistance towards vasoconstriction without change in gain of the reflex. In the other, the pressure increases were diminished at the higher levels of carotid sinus pressure indicating both resetting and an increase in gain of the reflex. 6. Both types of response occurred in the spontaneously breathing animal, in animals artificially ventilated, while pacing the heart, with the arterial blood pressure maintained constant at about 100 mmHg, and after division of the cervical vagosympathetic nerves. The frequency of occurrence of each type of response, however, varied under the different conditions. 7. The possible reasons for the two types of vascular response are discussed.

Trigeminal and carotid body inputs controlling vascular resistance in muscle during post-contraction hyperaemia in cats.

1. In anaesthetized cats, the effects of stimulation of the receptors in the nasal mucosa and carotid body chemoreceptors on vascular resistance in hindlimb skeletal muscle were studied to see whether the responses were the same in active as in resting muscle. The measurements of vascular resistance were taken, first, in resting muscle, and second, in the immediate post-contraction hyperaemic phase that followed a 30 s period of isometric contractions. 2. Stimulation of the receptors in the nasal mucosa caused reflex apnoea and vasoconstriction in muscle. The latter response was attenuated when the test was repeated during post-contraction hyperaemia. 3. Stimulations of the carotid bodies were made during a period of apnoea evoked reflexly by electrical stimulation of both superior laryngeal nerves. This apnoea prevented any effects of changes in respiration on the carotid body reflex vascular responses. Stimulation of the carotid bodies evoked hindlimb muscle vasoconstriction. In the post-contraction hyperaemic period, the response was reduced or abolished. A similar attenuation of the reflex vasoconstrictor responses occurred in decentralized muscles stimulated through their motor roots in the cauda equina. 4. Evidence is presented that the attenuation of the vasoconstrictor responses evoked by the two reflexes is a phenomenon localized to the contracting muscles themselves resulting from an interaction between sympathetic neuronal activity and the local production of metabolites. 5. The results are discussed in relation to the metabolic needs of tissues in relation to asphyxial defence mechanisms such as occur in the diving response.

Comparison of the size of the vascular compartment of the carotid body of the fetal, neonatal and adult cat.

The carotid bodies from full-term fetal cats, 3- to 4-day-old neonates and adult cats were perfusion-fixed at normal arterial blood pressure with 3% phosphate-buffered glutaraldehyde. Serial 5-microns sections were cut and stained by the MSB method. Using an interactive image analysis system, determinations were made of the volumes of the carotid body and of its vascular and extravascular compartments. Compared to the fetus, the carotid body of the neonate increased in volume by 51% and by 286% in the adult cat. There was a proportional increase in the volumes of the vascular compartment and of the small vessels (5-12 microns diameter) in that compartment. The volume of the small vessels, expressed as a ratio of the total volume of the organ, remained constant in the three animal groups at 5-7%. The small vessel endothelial surface area, expressed as a ratio of the extravascular volume (which was assumed to consist largely of type 1 and type 2 cells), was the same in the neonate as in the full-term fetus. Thus, there were no apparent quantifiable morphological features of the carotid body and its vasculature which would account for the resetting of the hypoxic sensitivity of the organ from the fetal to the adult range within a few days of birth.

Cardiovascular and respiratory effects of carotid body stimulation in the monkey.

1. The carotid bodies were stimulated in the anaesthetized pig-tailed macaque monkey (M. nemestrina) using (i) brief injections of cyanide or CO2-equilibrated bicarbonate solution into a common carotid artery, and (ii) longer perfusion with hypoxic hypercapnic blood in vascularly isolated chemoreceptor preparations. 2. In spontaneously breathing animals brief stimuli (thirty-one tests, seven monkeys) consistently increased pulmonary ventilation (by 97 +/- 10% of control), slowed the heart rate (the pulse interval increasing by 36 +/- 7.5%), and increased femoral vascular resistance (by 44 +/- 7%). 3. More sustained chemoreceptor stimulation with asphyxial blood (nineteen tests, five monkeys) increased ventilation by 187 +/- 23%, but transient bradycardia occurred in only eight of nineteen tests and was followed by tachycardia; in the remaining tests, only tachycardia occurred. After 20--40s, the pulse interval was 5.8 +/- 0.9% below the control level. Femoral vascular resistance either increased (five tests, two animals) or decreased (six tests, two animals). 4. Evidence is presented that in the monkey the autonomic effects of chemoreceptor stimulation are influenced by the level of respiratory activity with bradycardia and vasoconstriction occurring when the level is low, and tachycardia and vasodilatation when it is high. 5. The interaction of autonomic responses resulting from carotid body stimulation and from mechanisms initiated by the concomitant hyperventilation are qualitatively similar in the monkey and in subprimate species, although there may be quantitative differences such as would account for the species differences to distrubances produced, for instance, by arterial hypoxia.

Carotid body chemoreceptor reflexes and their interactions in the seal.

In the anesthetized spontaneously breathing harbor seal Phoca vitulina stimulation of the carotid body chemoreceptors by intracarotid injections of sodium cyanide or by hypoxic hypercapnic blood causes an increase in tidal volume, respiratory frequency, and respiratory minute volume. The heart rate invariably decreased. Experimental dives caused apnea and bradycardia. When the carotid bodies are stimulated within 10 s of the commencement of a dive, the chemoreceptor-respiratory response is abolished, but the chemoreceptor-cardioinhibitory response is considerably enhanced. Electrical stimulation of the central cut end of a superior laryngeal nerve also causes apnea and bradycardia; stimulation of the carotid body now fails to produce a respiratory response but the cardioinhibitory effect is enhanced. These results indicate that the carotid bodies cause reflexly hyperventilation and bradycardia, and that these responses are considerably modified by other inputs to the central nervous system.

Cardiorespiratory control by carotid chemoreceptors during experimental dives in the seal.

The diving responses of apnea and bradycardia, produced experimentally by immersing the face in water, were successfully elicited in the harbor seal Phoca vitulina anesthetized with urethan. The role of the carotid body chemoreceptors in the production of the diving bradycardia was studied in isolated carotid sinus-body preparations autoperfused with blood from the arterial circulation. When asphyxia was well developed during a dive the chemoreceptor drive was withdrawn by temporarily perfusing the chemoreceptors with blood of high PO2 (greater than 400 mmHg) and normal PCO2 from a disk oxygenator. The heart rate immediately rose to its predive value. Reestablishing hypoxic hypercapnic blood perfusion of the chemoreceptors from the animal's own circulation caused bradycardia with persistence of the apnea. Breathing restarted only on emersion. Substitution of normal arterialized blood from the oxygenator before or at the onset of a dive had no effect on the existing heart rate. It is concluded that the carotid bodies play an important part in maintaining the diving bradycardia during developing asphyxia without affecting respiration.

Lung inflation: effects on heart rate, respiration, and vagal afferent activity in seals.

In the anesthetized harbor seal, Phoca vitulina, the Hering-Breuer inflation reflex was weak and comparable to that in humans. Single inflations of the lungs from a syringe during the expiratory phase of normal breathing caused temporary inhibition of breathing and an immediate tachycardia dependent on the integrity of the cervical vagosympathetic nerves. A similar cardiac response occurred when the lungs were artificially inflated during an experimental dive and under conditions in which apnea and bradycardia were reflexly induced by a combination of stimulation of the carotid body chemoreceptors and of the trigeminal or laryngeal input. Recordings from single vagal afferent nerve fibers innervating presumptive pulmonary stretch receptors showed a close relationship between the increase in impulse frequency and increase in lung volume or transpulmonary pressure. It appears that in diving the decrease in pulmonary stretch receptor activity during apnea, combined with cessation of central inspiratory neuronal drive, is an important integrative mechanism that helps development of the reflex bradycardia of trigeminal, carotid, chemoreceptor, and baroreceptor origin.