Type I cells of carotid body from rats treated with 5-OH-Dopa and L-Dopa: an electron microscopical study

Summary The ultrastructure of the Type I cells in rat carotid bodies was studied after treatment with 5-OH-Dopa and L-Dopa. Type I cells from the L-Dopa treated rats were also analysed by morphometric methods.After 5-OH-Dopa treatment, the majority of dense-cored vesicles of the Type I cells were almost completely filled withvery electron-dense material. The vesicles were mainly distributed close to the plasma membrane.After L-Dopa treatment, the ultrastructure of the Type I cells resembled that after 5-OH-Dopa treatment, except that the increase in electron density of the vesicle content was less pronounced. The morphometric analysis revealed, as previously described for normal carotid bodies, two subclasses of Type I cells, small vesicle cells (SVC) and large vesicle cells (LVC). Compared to those of normal rats, the mean profile diameter of the vesicles in theL-Dopa treated rat carotid bodies were enlarged (for SVC from 47 to 55 nm and for LVC from 63 to 76–78 nm). Furthermore, the volume density of vesicles in both SVC and LVC was increased. The mean volume of the SVC was markedly increased after L-Dopa treatment for 60 h.These observations indicate that the Type I cells possess mechanisms for uptake of amine precursors and synthesis of monoamines. Though the vesicle may be involved in an endocrine-like function, they may also be related to the chemoreceptor function of the carotid body.     

A quantitative morphological study of the carotid bodies of rats living at a simulated altitude of 4300 metres.

A quantitative histological study was carried out on the carotid bodies of 10 normal rats and 10 rats living in a hypobaric chamber at a pressure of 460 mm Hg from 25 to 96 days. In the chronically hypoxic rats there was a four-fold increase in the mean combined volume of the carotid bodies. Morphometric analysis disclosed a three-fold increase in the mean volume of specialised glomic cells and a ten-fold increase in the mean volume of capillaries, although the proportion of glomic cells was actually significantly decreased. In all our hypoxic rats there was evidence of both right and left ventricular hypertrophy. However, there was no linear relation between total carotid body volume or volume of glomic cells on one hand and the right and left ventricular weight, on the other hand. Although there was no linear relation between combined total carotid body volume and duration of hypoxia, the linear relation between glomic cell volume and duration of hypoxia was significant at the 5 per cent. level. The increase in vascularity of the hypoxic carotid body may be a mechanism to increase blood flow and thus oxygen transport to a hypoxic organ with increased metabolic activity. Small quantities of an amorphous hyaline material of unknown nature were found in relation to capillaries and type I cells in all the hypoxic rats.     

Paraganglia of the rat recurrent laryngeal nerve after long-term hypoxia: A morphometric and biochemical study

Summary Paraganglia of the recurrent laryngeal nerve of rats exposed to hypoxia (10±0.5% O₂) for 3 weeks became enlarged in volume by 2.6-fold. The volume densities of blood vessel lumen and endothelial cells were unchanged. The enlargement of the paraganglia was mainly attributed to hypertrophy of the dense-cored vesicle-containing cell. Concerning the profile diameter of the vesicle, a unimodal distribution of dense-cored vesicle cells was found in both control and hypoxic paraganglia. The vesicle diameter increased by about 19%, whereas the numerical density of dense-cored vesicles and the vesicle volume density were unaltered after hypoxia.Recurrent laryngeal nerves assayed by a sensitive high performance liquid chromatography technique contained an average of 1.9pmol dopamine and 1.3pmol norepinephrine. Exposure to 2 weeks of hypoxia increased the dopamine level to 4.9pmol, whereas the norepinephrine content was unchanged. It is suggested from this study that the catecholamines determined are housed in the paraganglia of the recurrent laryngeal nerve.     

The paraganglia within the carotid bifurcation regions of young and old spontaneously hypertensive rats (SHR) after exposure to chronic hypobaric hypoxia. II. The carotid body-like organs

The following paraganglia in the carotid bifurcations regions of spontaneously hypertensive rats (SHR) were studied: Endoneural paraganglia within the external carotid nerve, the carotid sinus nerve, the glossopharyngeal nerve, and the pharyngeal branch of the vagus nerve, the so-called periadventitial type I cells, and so-called miniglomera. Number and distribution of these paraganglia vary among different individuals. After chronically hypobaric hypoxia the volume of these paraganglia was increased but their number remained unchanged. The increase of volume was dependent on the duration of hypoxia. There were no differences between young and old SHR when the hypoxia-time was the same.     

Morphometric analysis of glomus cells within the recurrent laryngeal nerve of the rat

Summary An endoneurial paraganglion located in the recurrent laryngeal nerve of the rat was found in 17 of 20 nerves studied. The median size of the paraganglia was estimated to be 0.8 × 10⁶ m³. The number of type I like cells within a paraganglion ranged between 8 and 24. Altogether, 120 endoneurial dense-cored vesicle cells from 8 paraganglia were subjected to a morphometric analysis at the ultrastructure level. The mean vesicle profile diameter was estimated to be 101.3 nm and only one type of granulated cell could be distinguished. The volume density of the dense-cored vesicles was estimated to be about 6%, a value amounting to two to three times that in the carotid glomus. Moreover, the mean cell profile area of vesiculated cells of recurrent laryngeal nerve exceeds that of the cells of the carotid glomus.     

Mouse chromaffin cells have two populations of dense core vesicles

The quantal hypothesis states that neurotransmitter is released in discrete packages, quanta, thought to represent the neurotransmitter content of individual vesicles. If true, then vesicle size should influence quantal size. Although chromaffin cells are generally thought to have a single population of secretory vesicles, our electron microscopy analysis suggested two populations as the size distribution was best described as the sum of two Gaussians. The average volume difference was fivefold. To test whether this difference in volume affected quantal size, neurotransmitter release from permeabilized cells exposed to 100 microM Ca2+ was measured with amperometry. Quantal content was bimodally distributed with both large and small events; the distribution of vesicle sizes predicted by amperometry was extremely similar to those measured with electron microscopy. In addition, each population of events exhibited distinct release kinetics. These results suggest that chromaffin cells have two populations of dense core vesicles (DCV) with unique secretory properties and which may represent two distinct synthetic pathways for DCV biogenesis or alternatively they may represent different stages of biosynthesis.     

The ultrastructure of the carotid body in chronically hypoxic rabbits

1. The ultrastructure of the carotid body in the rabbit has been examined by electron microscopy.2. A comparison was made between the ultrastructure of the carotid bodies in sea level rabbits, in rabbits which had been exposed to hypoxia equivalent to an altitude of 6000 m for 7 days and in rabbits which always had lived at an altitude of 4000-4300 m.3. We could not detect any difference in the ultrastructure between the two groups of hypoxic rabbits.4. When the hypoxic rabbits were compared with sea level rabbits there was a marked increase in the number of dense cored vesicles and mitochondria in the type I cells in the hypoxic rabbits. The Golgi region also appeared to be enlarged in the type I cells in the hypoxic rabbits.5. The finding suggests that in the rabbit the production of amines, probably dopamine, within the type I cells is increased during prolonged hypoxia which might explain the lowered ventilatory response to hypoxia observed in human high altitude residents.6. If the carotid bodies are organs of internal secretion the finding is compatible with an increased production of a hormone produced within the type I cells.     

A quantitative study of some ultrastructural features of the type I cells in the carotid bodies of rats living at a simulated altitude of 4300 metres

Summary A quantitative study was carried out on the ultrastructure of the type I cells of the carotid bodies of three normal rats and three rats living in a hypobaric chamber at an atmospheric pressure of 460 mm Hg for 27, 28 and 35 days. Point-counting methods were performed on electron micrographs of randomly selected cross-sections of type I cells. These sections always included the nucleus of the cell. The same electron micrographs were used to determine the number of mitochondrial cross-sections and electron-dense core vesicles appearing in each type 1 cell profile. The morphometric analysis disclosed that the mean cross-sectional area of the type I cells was 41.35 m³ in the untreated rats and 82.03 m³ in the rats exposed to chronic hypoxia. Assuming that this area of cross-section was in direct proportion to the volume of the cell, this result indicates an approximately three-fold increase in volume of the type I cells of the carotid bodies in hypoxaemic rats. There was no change in the volume proportion of the type I cells occupied by the mitochondria. However, as the cells had increased in volume in hypoxaemic rats, it was concluded that the number of mitocondria in each type I cell was increased. The concentration of electron-dense core vesicles was 21.9/m³ cytoplasm in the type I cells of untreated rats and 7.3/m³ in the hypoxaemic rats. The dense-core vesicles were increased in diameter in states of chronic hypoxia. Their mean diameter was 102.3 nm in normal rats and 117.0 nm in hypoxaemic rats. The enlargement of the type I cells and the increase in the number of mitochondria within each cell suggests that this depletion is more likely to be due to an increased rate of release of dense-core vesicles, than to a reduced rate of their synthesis.     

The influence of chronically hypoxemic states on human carotid body structure and cardiac hypertrophy

Summary Quantitative and qualitative changes in the human carotid body morphology, and their relationship to changes in the weight of right and left ventricles were investigated in 10 patients with a history of chronic hypoxemia. 5 patients without a history of cardiac, pulmonary or cerebral respiratory failure served as the control group. In the chronically hypoxemic group, a 2.67-fold increase in the total specific glomus cell volume was found. Up to a critical volume this increase is due to hypertrophy, beyond that it is due to hyperplasia. The course of the morphologic changes under the influence of slowly progressive chronic hypoxemia is discussed in a frame work of three stages (stage I = hypertrophy, stage II = nodular hyperplasia, stage III = atrophy). Plasmacellular infiltrates are constant though sometimes sparse. They are mostly perineural in location, less often intralobular and if so almost exclusively periglomoidal. In one case, we found an increase of Schwann cells in the interstitial and periglomoidal space without demonstrable degeneration of the nerve fibres themselves. Our hypothesis suggests that degeneration of special nerve terminals of the reciprocal type occurs in afferent nerve fibers. The increase of right ventricular weight (by a factor of 2.05) is significant, in contrast to that of the left. A linear correlation between the increase of right ventricular weight and the increased total glomus cell volume was not established. In 4 cases, however, we found pulmonary hypertensive vascular changes, which might be responsible for the disparity in the linear relationship.Dedicated to Prof. Dr. J.R. Rüttner on the occasion of his 60th birthday     

Ultrastructure of the glomus cells in the carotid body of chronically hypoxic rats: With special reference to the similarity of amphibian glomus cells

The ultrastructural characteristics of the glomus cells in the rat carotid body exposed to extremely long-term hypoxia (10–12 weeks) were investigated. The glomus cells could be classified into four distinct types according to the shape of dense-cored vesicles in the glomus cell cytoplasm: (1) small vesicle cells (SVCs, 50 nm in mean diameter), (2) large vesicle cells (LVCs, 80 nm in mean diameter), (3) dilated eccentric vesicle cells (EVCs, 400–800 nm in diameter), and (4) mixed vesicle cells (MVCs, large and eccentric vesicles). Many clusters of glomus cells were found to contain all four categories of cell types. The appearance of EVCs was a unique and common characteristic of glomus cells in this long-term hypoxia model. We also noted other ultrastructural features with chronic hypoxia which are characteristic of the amphibian carotid labyrinth glomus cells: (1) incomplete covering of glomus cells with the supporting cell missing over a wide area, (2) long thin cytoplasmic projections in the intervascular stroma, and (3) intimate apposition of the glomus cells and pericytes (g-p connection), endothelial cells (g-e connection), plasma cells, and fibrocytes. Because arterial PO₂ is generally low in amphibia, these may be general features of hypoxic adaptation and facilitate both uptake of oxygen from blood and release of catecholamine into the blood. The g-p and g-e connections may take part in the regulation of the microcirculation in the enlarged carotid body. © 1993 Wiley-Liss Inc.     

Ultrastructural localization of active zone and synaptic vesicle proteins in a preassembled multi-vesicle transport aggregate

Although it has been suggested that presynaptic active zone (AZ) may be preassembled, it is still unclear which entities carry the various proteins to the AZ during synaptogenesis. Here, I propose that aggregates of dense core vesicles (DCV) and small clear vesicles in the axons of young rat hippocampal cultures are carriers containing preformed AZ and synaptic vesicle (SV) components on their way to developing synapses. The aggregates were positively labeled with antibodies against Bassoon and Piccolo (two AZ cytomatrix proteins), VAMP, SV2, synaptotagmin (three SV membrane proteins), and synapsin I (a SV-associated protein). Bassoon and Piccolo labeling were localized at dense material both in the aggregates and at the AZ. In addition to the SV at the synapses, the SV membrane proteins labeled the clear vesicles in the aggregate as well as many other SV-like and pleiomorphic vesicular structures in the axons, and synapsin I labeling was associated with the vesicles in the aggregates. In single sections, these axonal vesicle aggregates were approximately 0.22 by 0.13 microm in average dimensions and contain one to two DCV and five to six small clear vesicles. Serial sections confirmed that the aggregates were not synaptic junctions sectioned en face. Labeling intensities of Bassoon and Piccolo measured from serially sectioned transport aggregates and AZ were within range of each other, suggesting that one or a few aggregates, but not individual DCV, can carry sufficient Bassoon and Piccolo to form an AZ. The present findings provide the first ultrastructural evidence localizing various AZ and SV proteins in a preassembled multi-vesicle transport aggregate that has the potential to quickly form a functional active zone.     

The effects of nerve stimulation and hemicholinium on synaptic vesicles at the mammalian euromuscular junction

1. Electron micrographs of nerve terminals in rat phrenic nerve-diaphragm preparations have been studied. This has been done before and after prolonged nerve stimulation. The effectiveness of nerve stimulation has been monitored by intracellular micro-electrode recordings from the muscle cells.2. Characteristic changes in the form and distribution of the nerve terminal mitochondria were noted after nerve stimulation.3. Synaptic vesicle numbers in the region of nerve terminal less than 1800 A from the synaptic cleft were significantly greater in tissue taken 2 and 3 min after nerve stimulation, than in unstimulated preparations.4. The long and short diameters of the synaptic vesicle profiles less than 1800 A from the synaptic cleft were measured. Analysis of the distribution of the diameters indicated synaptic vesicles to be basically spherical structures. Estimates of synaptic vesicle volume were made from the measurements. Synaptic vesicle volume was significantly reduced in tissue taken 2 and 4 min following nerve stimulation.5. If hemicholinium, a compound which inhibits acetylcholine synthesis, was present during the period of nerve stimulation, much greater reductions in synaptic vesicle volume occurred. Synaptic vesicle numbers in the region of nerve terminal less than 1800 A from the synaptic cleft were also reduced, compared with unstimulated control preparations.6. These results are regarded as support for the hypothesis that the synaptic vesicles in nerve terminals at the mammalian neuromuscular junction represent stores of the transmitter substance, acetylcholine.     

Postnatal growth of the carotid body

The size of the carotid body (CB) is increased significantly during the postnatal period. Type I cells in the CB are the chemoreceptive element and possess many neuron-like characteristics. In contrast to previous opinions that the number of type I cells is determined before birth, we have found that type I cells continue to proliferate over a period of at least 1 month after birth in rats. The proliferation of type I cells is influenced by oxygen concentration in ambient air. Specifically, hyperoxia inhibits the type I cell proliferation, resulting in small CBs throughout life and the permanent impairment of CB chemoreception. On the other hand, hypoxia enhances the type I cell proliferation. Whether hypoxia causes long-lasting effects on CB morphology and function remains to be determined. Besides type I cell proliferation, other cellular components in the CB undergo proliferation and growth as well. In the nearby petrosal ganglion and superior cervical ganglion, both involved in CB chemoreception, cellular proliferation is limited to glial cells and no proliferation of neurons is observed. Also, expression of neurotrophic factors, particularly, BDNF and GDNF, is observed in type I cells of neonatal rats. Taken together, the CB undergoes significant morphological and functional changes during the postnatal period over at least 1 month. This process can be altered by oxygen concentration in ambient air.     

In vitro recording of chemoreceptor activity in catecholamine-depleted rabbit carotid bodies

Carotid bodies, together with Hering's nerves, were excised from anesthetized rabbits 24, 48 or 72 h after single reserpine injections (5 mg kg-1, i.v. or i.p.) and were superfused in vitro. Some carotid bodies were processed for formaldehyde-induced fluorescence microscopy to assess catecholamine depletion. Twenty-four hours after reserpine treatment, most of the type I cell islets had lost their fluorescence and the number of spontaneously active chemoafferent units was dramatically reduced. Forty-eight hours after reserpine injection, both the fluorescence of type I cells had partially recovered and the number of chemoreceptor units was almost normal. A significant reduction of both the normoxic and hypoxic frequencies of discharge was demonstrated in carotid bodies examined 24 or 48 h after reserpine pretreatment. Superfusions with dopamine (1, 10, 100 microM) transiently restored the response to hypoxia. It is proposed that catecholamines contained in type I cells play a prominent role in the genesis of chemoafferent activity and in the chemoreceptor response to hypoxia.     

Hypoxia decreases cellular ATP demand and inhibits mitochondrial respiration of a549 cells

Hypoxia inhibits activity and expression of transporters involved in alveolar Na reabsorption and fluid clearance. We studied whether this represents a mechanism for reducing energy consumption or whether it is the consequence of metabolic dysfunction. Oxygen consumption (JO2) of A549 cells and primary rat alveolar type II cells was measured by microrespirometry during normoxia, hypoxia (1.5% O2), and reoxygenation. In both cell types, acute and 24-h hypoxia decreased total JO2 significantly and reoxygenation restored JO2 after 5 min but not after 24 h of hypoxia in A549 cells, whereas recovery was complete in type II cells. In A549 cells under normoxia Na/K-ATPase accounted for approximately 15% of JO2, whereas Na/K-ATPase-related JO2 was decreased by approximately 25% in hypoxia. Inhibition of other ion transporters did not affect JO2. Protein synthesis-related JO2 was not affected by acute hypoxia, but decreased by 30% after 24-h hypoxia. Acute and 24-h hypoxia decreased JO2 of A549 cell mitochondrial complexes I, II, and III by 30-40%. Reoxygenation restored complex I activity after acute hypoxia but not after 24-h hypoxia. ATP was decreased 30% after 24-h hypoxia, but lactate production rate was not affected. Reduced nicotinamine adenine dinucleotide was slightly elevated in acute hypoxia. Our findings indicate that inhibition of the Na/K-ATPase by hypoxia contributes little to energy preservation in hypoxia. It remains unclear to what extent hypoxic inhibition of mitochondrial metabolism affects ATP-consuming processes.     

Influence of age on position, shape and size of the carotid bodies in spontaneously hypertensive (SHR) and normotensive (NCR) rats

In normotensive Wistar rats (NCR) and spontaneously hypertensive rats (SHR) aging 3-6 d and 5-6, 15-20, 30-40 and 50-70 weeks respectively, position, shape and size of the carotid bodies were studied using ligh-microscopic methods. The anatomical position of the carotid bodies in the SHR was found to be less variable than in the NCR. In the hypertensive animals the glomera carotici were usually situated near the internal carotid artery. The carotid bodies of the NCR were of a more round-oval shape. In all age-groups the carotid bodies were compact and clearly demarked corpuscules. But also outside the principal mass of the glomera carotici small groups of type I cells were found in the surrounding nervous tissue or they appeared as miniglomera or periadventitial type I cells, respectively. These small groups of type I cells were predominantly provable in the 5-6 weeks old animals and more frequently in the NCR than in the SHR. The absolute carotid body volumes increased from birth up to an age of 30-40 weeks, whereas the relative carotid body volume, i.e. when related to the same body mass, became much smaller from birth up to an age of 15-20 weeks in both the normotensive and hypertensive animals. Except for the 5-6 weeks old group the SHR showed significantly greater carotid bodies when compared with the age-matched normotensive rats. In the course of aging in both strains of animals a stronger and more clearly developed interstitial fibrosis of the specific tissue of the carotid bodies occurred.     

Stimulus-specific signaling pathways in rabbit carotid body chemoreceptors

The carotid body is an arterial chemosensory organ which responds to multiple natural and pharmacological stimuli, including hypoxia and nicotine. Numerous studies have investigated the initial molecular events which activate chemosensory type I cells in the carotid body, but less attention has been focused on later steps in the transduction cascade, which mediate neurotransmitter release from type I cells and excitation of chemoreceptor afferent fibers in the carotid sinus nerve. In the present study, we examined the effects of a highly specific inhibitor of calcium/calmodulin-dependent kinase II, KN-62, and a calmodulin inhibitor, trifluoperazine, on carotid sinus nerve activity and catecholamine release evoked from rabbit carotid bodies superfused in vitro. KN-62 did not alter sinus nerve activity and catecholamine release evoked by hypoxia, but this agent significantly reduced nerve activity and neurotransmitter release evoked by 100 microM nicotine. Trifluoperazine (10 microM), likewise inhibited activity evoked by nicotine, as well as hypoxia. Basal levels of nerve activity and catecholamine release (established in superfusate equilibrated with 100% O2) were unaffected by all drug treatments. Separate biochemical experiments showed that Ca2+/calmodulin-dependent incorporation of 32P into carotid body particulate proteins is significantly reduced following incubation of intact carotid bodies in nicotine, but not following exposure to hypoxia. Our observations suggest that excitation of the carotid body by diverse stimuli may involve the activation of distinct, stimulus-specific transduction pathways. Furthermore, these data correlate with our previous findings which showed that hypoxia, on the one hand, and nicotine on the other, evoke the preferential release of either dopamine or norepinephrine, respectively, from carotid bodies incubated in vitro.     

IgG subclass antibodies to serogroup B meningococcal outer membrane antigens following infection and vaccination

IgG and IgG subclass antibodies to the outer membrane antigens from Neisseria meningitidis (serogroup B, serotype 15:P1.16) were quantitated by an enzyme-linked immunosorbent assay (ELISA) in sera from 40 patients with group B:15:P1.16 meningococcal disease and 24 volunteers immunized with a serotype 15:P1.16 outer membrane vesicle vaccine. A second injection was given 6 weeks after the first immunization. Patient sera obtained two and six weeks after onset of the disease had significantly higher levels of total IgG, IgG1, IgG2, and IgG3 antibodies to the outer membrane antigens than acute sera, convalescent sera from patients with systemic non-meningococcal bacterial infections and sera from healthy controls. The levels of total IgG and IgG1 remained high one and three years later. Sera from the vaccinees showed high levels of total IgG and IgG1 6, 12 and 26 weeks after the first immunization and high levels of IgG3 6 weeks after the second immunization. No increase of IgG2 or IgG4 levels was observed in the postimmunization sera. Immunoblotting of three convalescent sera demonstrated individual patterns of IgG subclass binding to various outer membrane antigens with most distinct binding of IgG1 and IgG3 antibodies to the class I protein, the H.8 lipoprotein and the lipopolysaccharide. Since IgG1 and IgG3 are the most effective antibodies for complement activation and phagocytosis, group B meningococcal disease and immunization with the serotype 15:P1.16 outer membrane vesicle vaccine stimulate production of those IgG subclasses which have the strongest opsonic and bactericidal activity.     

Calcium distribution in aortic smooth muscle cells of deoxycorticosterone-hypertensive rats. A quantitative cytochemical study

The effect of deoxycorticosterone (DOC)-induced hypertension on the calcium content within the aorta was studied before the increase in pressure (one week) and after the pressure had reached hypertensive levels (4 weeks). The volume density of free calcium detected ultrastructurally by pyroantimonate precipitation was quantitated by stereological techniques in aortic smooth muscle cells. An increase in the volume density of electron opaque precipitate was observed in the cytoplasm at one week of DOC treatment when neither the systolic blood pressure, the thickness of the media nor volume fraction of medial smooth muscle as compared to the extracellular space was increased significantly. The total aortic calcium as measured by atomic absorption spectroscopy was not increased at one week. By 4 weeks when the rats were hypertensive, the cytoplasmic free calcium in the smooth muscle cells and the number of peripherally-located cytoplasmic vesicles with precipitate was increased significantly. Total aortic calcium was also increased significantly in the DOC-saline group but not in the DOC group drinking tap water or in the saline drinking controls. An elevation of calcium within the cytoplasm of vascular smooth muscle cells may precede the development of hypertension and play a role in the pathogenesis of the increased blood pressure, increased medial thickness and hypertrophy of the vascular smooth muscle cells.     

Effects on the rabbit carotid body of stimulation by almitrine, natural high altitude, and experimental normobaric hypoxia

Rabbits were given intraperitoneal injections of almitrine in ascending doses for 5 weeks. They were compared with a control group and with a group of rabbits which had been exposed from birth to the natural hypobaric hypoxia found at Cerro de Pasco (433 m) in the Peruvian Andes. A further group of animals was placed in an experimental normobaric chamber for either 3 or 6 months to subject them to the same degree of hypoxia as that occurring in Cerro. The carotid bodies of the rabbits in all these groups were processed for light and electron microscopy, and examined both qualitatively and quantitatively. The carotid bodies in the group given almitrine showed no changes in their size or in the population of their glomic cells when compared with controls. In contrast, the carotid bodies of Peruvian rabbits were greatly enlarged with a disproportionate increase in the population of the light variant of chief cell. Rabbits from the hypoxic chamber also had enlarged carotid bodies but those killed after 3 months showed an increase in the dark variant of chief cell, whereas after 6 months this cell was reduced in number. There was also intense cytoplasmic vacuolation. Election microscopy confirmed these changes and revealed that dark cells had larger, more pleomorphic granules than the light variant. Vacuolation of the granules in light cells was most pronounced in Peruvian rabbits, but was uncommon in animals exposed to hypoxia for 3 months. We suggest that the dark cell responds to the early stages of hypoxia but later matures into the light variant of chief cell.