Parallel changes in ultrastructure and noradrenaline content of nerve terminals in rat vas deferens following transmitter release

Transmural electrical stimulation and exposure to incubation media where some or all of the Na⁺ had been replaced with K⁺ were used to elicit transmitter release. Changes in noradrenaline content and ultrastructure of the nerve terminal varicosities in rat vas deferens were measured. Electrical stimulation in the presence of 4-aminopyridine had little effect, but high [K⁺] solutions caused a parallel reduction in noradrenaline content and the number of small dense-cored vesicles; large densecored vesicles showed no change, and small clear vesicles increased in number. In spite of a reduction in total vesicle number there was no evidence of expansion of the varicosity membrane.The parallel fall in noradrenaline content and in the number of small dense-cored vesicles suggests that the latter are the source of the released noradrenaline under the conditions of high [K⁺] stimulation we have used.     

Biochemical and functional properties of large and small dense-core vesicles in sympathetic nerves of rat and ox vas deferens

A comparative study of the noradrenaline storing vesicles in vas deferons from ox and rat was performed. Microsomal fractions were subjected to density gradient centrifugation. In rat, noradrenaline and dopamine β-hydroxylase were mainly present in the upper fractions of the gradient, which is consistent with the predominance of light (small dense-core) vesicles in this species. In ox, noradrenaline, dopamine and dopamine β-hydroxylase were found in the gradient in a bimodal distribution. This is consistent with the presence of about equal numbers of small and large dense-core vesicles in this species. On the other hand, chromogranin A, immunologically related proteins and enkephalin-like immunoreactivity were only present in the dense (large dense-core) vesicle population. In order to study the capability of light and dense vesicles to synthesize noradrenaline we “pulse-labelled” ox vasa deferentia with [³H]tyrosine. Already 3.5 min after the pulse both types of vesicles contained [³H]noradrenaline and [³H]dopamine. During longer “chase” periods the amount of [³H]dopamine gradually declined.We conclude that dense (large dense-core) vesicles contain chromogranin A, immunologically related proteins and enkephalin-like immunoreactivity whereas light (small dense-core) vesicles are devoid of these components. Both types of vesicles contain dopamine β-hydroxylase and can synthesize noradrenaline from dopamine under in vivo conditions.     

Improved isolation of small noradrenergic vesicles from rat seminal ducts following castration. A density gradient centrifugation and morphological study

Noradrenaline storage vesicles from nerve endings have been prepared by differential and density gradient centrifugation. Seminal ducts from castrated rats have been used as a source since fluorescence micrographs and electron microscopy demonstrated an increased concentration of sympathetic nerve endings in this tissue; probably as a result of muscle atrophy. The fractions obtained were analyzed biochemically and partly by electron microscopy.By centrifugation techniques, a density gradient fraction (at the density of 0.4–0.45m sucrose) was obtained in which noradrenaline was enriched about 10-fold from the homogenate. By using organs from castrated rats a further 4-fold increase was obtained in the noradrenaline/protein ratio of the density gradient fraction. This gradient fraction was studied by electron microscopy, after incubation with 5-hydroxydopamine as an exogenous marker, and was found to contain a high proportion of granular vesicle profiles. More than ? of all vesicles with a diameter below about 1000A?had an electron dense core. The vesicle preparation contained low activities of mitochondrial (cytochrome oxidase) and lysosomal (acid phosphatase) marker enzymes, while microsomes and plasma membrane fragments as indicated by NADH-cytochrome c reductase and adenosine triphosphatase had only been partially removed and probably constituted the main contaminants.The distribution of dopamine β-hydroxylase activity suggested that the small, light dense cored vesicles are deficient of this enzyme compared with the large, heavy dense cored vesicles, assumed to occur at the bottom of the gradient. The noradrenaline/adenosine 5'-triphosphate molar ratio was found to be 5.9 in the main noradrenaline fraction, but this fraction was not sufficiently pure to draw any definite conclusions about whether or not noradrenaline is stored together with adenosine 5'-triphosphate in small noradrenergic vesicles.     

The storage of endogenous noradrenaline in sympathetic nerve terminals

1. The subcellular distribution of noradrenaline in sympathetic nerve terminals of rat vas deferens and cat spleen has been studied by cell fractionation methods combined with fluorescence and electronmicroscopic histochemical methods for noradrenaline.2. Pinched-off axon varicosities (synaptosomes) were isolated and identified by fluorescence and electronmicroscopy in the mitochondrial pellet.3. The proportion of large to small dense-cored vesicles in electronmicrographs of sympathetic nerve terminals varies in different organs. In rat vas deferens 4% and in cat spleen 20% are large vesicles.4. Density gradients of rat vas deferens have a single low density peak of noradrenaline at 0.6 M sucrose, whereas those of cat spleen have an additional peak of noradrenaline at 1.1 M sucrose.5. Small dense-cored vesicles were identified electronmicroscopically in the low density fractions and large dense-cored vesicles in the high density fractions from density gradients.6. We conclude that both small and large dense-cored vesicles store noradrenaline.     

Release of [3H]norepinephrine from synaptic vesicles isolated from rat brain after the intracisternal administration of [3H]norepinephrine: influence of nucleotides, ions and drugs, and destabilization of transmitter storage caused by acute or chronic lithium administration

Following the intracisternal administration of [³H]norepinephrine to rats pretreated with a monoamine oxidase inhibitor, synaptic vesicles containing the radio label could be isolated from isotonically prepared microsomal fractions of rat brain. Incorporation of [³H]norepinephrine into the vesicles was reduced by pretreatment of the rats with desmethylimipramine and was also reduced if the rats had not been pretreated with a monoamine oxidase inhibitor. Incorporation of the label was totally eliminated by pretreatment with reserpine. Release in vitro of [³H]norepinephrine from the labeled vesicles was monophasic with a half-time of about 12 min at 30°C. The release was slowed by addition of adenosine 5′-triphosphate plus Mg²⁺ by a mechanism different from that of the vesicular amine uptake system; this was shown by the failure of inhibitors of adenosine triphosphate-Mg²⁺-stimulated uptake (reserpine,N-ethylmaleimide, lithium) to block the effect of adenosine triphosphate plus Mg²⁺ on release. Several other nucleotides also were able to slow the release of [³H]norepinephrine. Unlike adrenomedullary vesicles, rat brain synaptic vesicles did not show enhancement of amine release by chloride in the presence or absence of adenosine triphosphate plus Mg²⁺. The yield of labeled vesicles was substantially reduced if vesicles were prepared by hypotonic lysis of synaptosomes instead of isotonically from the microsomal fraction; the isotonic preparation appears to be superior for studies of vesicle uptake and storage properties.This preparation is readily utilizable for studies of the effects of in vivo administration of drugs thought to act on vesicular storage of catecholamines, a point illustrated by the destabilization of norepinephrine storage caused by acute or chronic lithium administration.     

Atrial sympathetic nerves after perfusion of the rabbit heart with low sodium solutions containing potassium chloride or urea—a biochemical,histochemical and fine structural study

Rabbit hearts were perfused in vitro with Tyrode's solution or low sodium media (17mm Na⁺) substituted with either KCl (low Na-high K) or urea (low Na-urea). Samples of right atrial tissues were removed for fluorescence and electron microscopy after 5, 15 and 120 min. In separate experiments, the noradrenaline concentrations in the right and left atria and ventricles were estimated at these times, and the amine overflow into the perfusion fluid determined. The cardiac noradrenaline concentrations and ultrastructural appearance of adrenergic axon terminals were not significantly affected by perfusion with Tyrode's solution for 120 min.Low Na-urea caused a progressive loss of noradrenaline from all parts of the heart which resulted in a 86–88% decrease of amine concentrations. Similar changes were observed after low Na-high K and also if the noradrenaline concentrations were expressed in terms of protein content of tissue. The amounts lost from the hearts were almost quantitatively recovered from the perfusates. There was greater tissue edema in hearts perfused with low Na-urea than after perfusion with the other media.After 5 min perfusion with either of the low Na solutions, little change was apparent in the ultrastructure of nerve terminals; there was no obvious alteration in histofluorescence, but after 15min it was reduced and at 120 min nearly abolished. Low Na-high K reduced the size of dense cores within the vesicles at 15 min. At 120 min the vesicles were agranular and their number per unit axon profile had dropped by 50%. In contrast, low Na-urea did not significantly alter the number of small dense-cored vesicles per profile despite causing myocardial cell damage from 5 min onwards which intensified with prolonged perfusion.The concomitant loss of histofluorescence, dense-cored vesicles and noradrenaline after low Na-high K is compatible with the view that the amine was released by exocytosis, as reported by others. However, despite myocardial tissue damage by low Na-urea, the noradrenaline overflow in this case is not due to disruption of vesicles and axon membranes. The preservation of electron-dense material in the vesicles is consistent both with previous reports that macromolecules rather than amines are responsible for the electron opaque cores and that amine release by low Na-urea is probably not by exocytosis.     

Pharmacological and ultrastructural studies on the electron dense cores of the vesicles that accumulate in noradrenergic axons constricted in vitro.

The noradrenaline storage vesicles that accumulate in axons proximal to ligatures applied to cat hypogastric nerves incubatedin vitro have been examined in the electron microscope with different fixation and staining procedures after exposure of the nerves to a variety of drugs which alter their noradrenaline content. After glutaraldehyde/osmium tetroxide fixation and lead citrate staining the granules possess electron-dense cores irrespective of their noradrenaline content. After fixation with Palade's osmium tetroxide or glutaraldehyde/dichromate followed by lead citrate staining, dense cores are not seen after treatment with noradrenaline-depleting agents but reappear after staining additionally with uranyl acetate.It is concluded that the matrix of axonal (large) noradrenaline storage vesicles contains some material in addition to catecholamines and this possibility is discussed in connection with the relationship between large and small dense-cored vesicles seen in sympathetic nerve endings.     

Release of adenosine 5′-triphosphate from synaptosomes from different regions of rat brain

The release of adenosine 5′-triphosphate by elevated extracellular concentrations of KC1 and by veratridine was determined in synaptosomal fractions prepared from different regions of rat brain. Following correction for yields of synaptosomes from the various regions, the relative distribution of K⁺-induced release was corpus striatum > cerebral cortex > medulla > hypothalamus > cerebellum. In contrast, the relative distribution of veratridine-induced release of adenosine 5′-triphosphate was medulla > corpus striatum > hypothalamus > cerebral cortex > cerebellum.From these findings, it was concluded that (1) depolarization-induced release of adenosine 5′-triphosphate was not distributed uniformly throughout the brain but varied from region to region, (2) the K⁺-induced release of adenosine 5′-triphosphate which is Ca²⁺-dependent, had a different regional distribution than the veratridine-induced release, which is greatest in Ca²⁺-free medium, and (3) the distribution of K⁺-induced release of adenosiae 5′-triphosphate did not correlate well with the known distribution of noradrenaline concentrations in rat brain, but did correlate to some extent with the distributions of 5-hydroxytryptamine, dopamine and especially acetylcholine, so that co-release of adenosine 5′-triphosphate with these transmitters may possibly occur.     

Effects of Mg2+ and Ca2+ on Noradrenaline Release and Uptake in Adrenergic Nerve Granules in Different Media

Euler , U. S. V. and F. Lishajko . Effects of Mg²⁺ and Ca²⁺ on noradrenaline release and uptake in adrenergic nerve granules in different media. Acta physiol. scand. 1973. 89. 415–422. Uptake and release of noradrenaline (NA) from a suspension of storage granules from bovine splenic nerves was measured after incubation in various media and effects of added Mg²⁺ and Ca²⁺ observed. No significant differences in NA release rate were observed in different media when NA uptake was prevented by addition of potassium ferricyanide to the medium. Uptake was facilitated by phosphate ions 5–30 mM but counteracted by monovalent cations. In sucrose-phosphate media uptake of NA almost balanced spontaneous release. Mg²⁺ and Ca²⁺ had only a small action on NA release in the absence of NA uptake. In sucrose phosphate media Mg²⁺ facilitated uptake, whereas Ca²⁺ enhanced NA release. This action was associated with the formation of a calcium phosphate precipitate. No release or uptake of NA was observed during incubation at ice-water temperature, nor did Mg²⁺ or Ca²⁺ have any effect. The results tend to show that release and uptake of NA in storage granules are separate, temperature-dependent processes, influenced by mono- and divalent cations and phosphate ions.     

Opioid peptides and noradrenaline co-exist in large dense-cored vesicles from sympathetic nerve

The possibility that opioid peptides and noradrenaline co-exist not only in the desheathed bundle of bovine splenic nerve which contains ~98% sympathetic C-fibers, but also in the population of large dense-cored noradrenergic vesicles from these fibers, has been investigated. The primary fraction of large dense-cored vesicles which can be prepared at about 85% purity has been further subjected to density gradient and fractional centrifugation procedures, including D₂O-loading and unloading on modified second gradients, in an attempt to separate any minor population of particles which potentially could contain opioid peptides and contaminate the large dense-cored vesicle fraction. Measurement of opioid peptides, noradrenaline, dopamine and dopamine β-hydroxylase activity supports the conclusion that opioid peptides are stored in the primary population of large dense-cored vesicles per se, rather than in a minor population of contaminating particles from cells other than sympathetic C-fibers.This conclusion has implications for exocytotic release and the physiological role of the opioid peptides intra- and extra-neuronally. Nerve vesicle opioid peptides have a size less than 5000 daltons, in contrast to the high proportion of large peptides containing enkephalin sequences in the bovine adrenal medulla.     

Isolation and characterization of specialized transducing bacteriophages for the recA gene of Escherichia coli.

Mg²⁺ stabilized potato virus Y helper component during partial purification by precipitation using polyethyleneglycol. In solutions containing 0.02 M Mg²⁺ the helper component retained most of its activity for 2 days at 4° and for at least 8 months at ?15°. Activity was destroyed on incubation with pronase or trypsin, or by heating for 5 min at 55°, but not by incubation with ribonuclease. Incubation with its own antiserum strongly inhibited helper component activity, but antisera to potato virus Y coat protein or inclusion protein had no more effect than a control serum showing that the component was unrelated to these two proteins. Filtration through a Sephadex G-200 column resulted in a broad peak of activity which produced many protein-staining bands when electrophoresed on polyacrylamide gel. Gel filtration and ultrafiltration experiments both indicated a molecular weight of 100,000–200,000. Some helper component activity was retained by aphids allowed to probe into a sucrose solution for 20 min showing that the helper component is more firmly bound to the aphid than is the tobacco mosaic viruspoly-l-ornithine complex.     

Evidence for the coexistence and co-release of [met]enkephalin and noradrenaline from sympathetic nerves of the bovine vas deferens

The localization and neurosecretion of methionine-enkephalin was studied in sympathetic nerves of the bovine vas deferens. Immunostaining showed methionine-enkephalin-like immunoreactivity in a network of varicose nerve fibres in the smooth muscle layers of the vas deferens. When vas deferens homogenates were subjected to differential and sucrose density gradient centrifugation, methionineenkephalin was found to parallel the distribution of noradrenaline in the more dense region of the gradient, where “heavy” or large dense-cored vesicles are present. Electron microscopic immunochemistry confirmed this finding and showed methionine-enkephalin-like immunoreactivity in large dense-cored vesicles. The release of methionine-enkephalin upon electrical stimulation was studied in superfusion experiments. The methionine-enkephalin secretion was shown to be Ca²⁺-dependent and was inhibited by adding the adrenergic neuron blocking drug guanethidine to the superfusion medium.

We conclude that in the bovine vas deferens methionine-enkephalin is only present in large dense-cored vesicles of adrenergic neurons and that the peptide is released from these vesicles together with noradrenaline by a Ca²⁺-dependent mechanism.     

Ionic currents in the somatic membrane of rat dorsal root ganglion neurons-II. Calcium currents

The experiments have shown that the introduction of cyclic adenosine monophosphate, adenosine 5'-triphosphate and Mg²⁺ ions into dialysed isolated dorsal root ganglion neurons of 5–8 day-old rats not only prevents the rapid decline of calcium inward currents during the course of dialysis but restores to a considerable extent the calcium conductance. Introduction of adenosine 5'-triphosphate and Mg²⁺ has a much weaker stabilizing effect. This finding made it possible to separate and to investigate in detail the calcium current (I_Ca) in the somatic membrane of all investigated neurons. The maximal amplitude of I_Ca was proportional to the concentration of Ca²⁺ ions in the extracellular solution between 2 and 14mM; with higher concentration a saturation effect was observed. Replacement of Ca²⁺ by Ba²⁺ caused about a two-fold increase in the maximal amplitude of inward currents. Addition of Co²⁺, Mn²⁺, verapamil and related substances blocked the calcium current. The activation kinetics of I_Ca could be approximated by a modified Hodgkin-Huxley equation using a square power of the m-variable. The activation time constantτ_m changed in the range from 16 to 1.8 ms with testing potential change from ?40 to +20 mV. The inactivation of I_Ca was extremely slow; the half value of steady-state inactivation was observed at holding potential about ?60 mV. The potential-dependent and kinetic characteristics of the calcium currents obtained on several neurons from adult rats were similar to those for neurons of new-born ones.It is concluded that the somatic membrane of the rat neurons has a system of electrically-operated selective calcium channels, the normal functioning of which is dependent on intracellular cyclic nucleotide metabolism.     

Metabolic aspects of the synthesis and intra-axonal transport of noradrenaline storage vesicles

1. Constricted cat hypogastric nerve/inferior mesenteric ganglion preparations have been maintained for 24 hr in vitro, in either test-tubes or a two compartment box, and used to study the roles of energy metabolism, protein synthesis and calcium ions in the synthesis and transport of noradrenaline storage vesicles in sympathetic neurones.2. When preparations were incubated so that the nerve trunks but not the ganglia were made anoxic, dense-cored vesicles lost their store of noradrenaline and did not accumulate above the ligature. This was accompanied by profound ultrastructural damage to the nerves.3. When in contact with the ligated nerve trunks, sodium fluoride inhibited the intra-axonal movement of dense-cord vesicles but did not deplete them of their stored noradrenaline. When sodium fluoride and pyruvate were present in the medium bathing the nerve trunks the movement of noradrenaline storage vesicles was unaffected.4. Whilst a reduction of the calcium ion concentration in the incubation medium and the inhibition of protein synthesis by cycloheximide prevented the synthesis of dense-cored vesicles in the neuronal perikaryon they had no effect on the movement of preformed dense-cored vesicles along non-myelinated axons.     

Selective uptake of 3H tyramine by chromaffin cells in vitro and simultaneous release of noradrenaline

Summary Cultures of adrenal explants derived from rat and mouse embryos were exposed to³H tyramine for intervals of 5 to 25 minutes. Light and electron microscopic autoradiography revealed a selective up-take of³H tyramine in chromaffin cells, and not in adrenocortical cells or fibroblasts. During this same period release of noradrenaline by chromaffin cells was demonstrated. Grains were observed over dense core vesicles at all times, and labelled vesicles were always distributed throughout the cytoplasm with no concentrations close to the plasma membrane. The mechanism of tyramine-induced secretion is discussed in the light of these findings.     

Presynaptic inhibitory actions of adenine nucleotides and adenosine on neurotransmission in the rat vas deferens.

Adenine nucleotides and adenosine inhibited the isometric contractions of the rat vas deferens in vitro in response to field stimulation but had no effect on the responses to exogenous noradrenaline. The inhibitions were potentiated by dipyridamole and compound 555, antagonized by theophylline and unchanged by indomethacin, 2-2′-pyridylistogen, phenoxybenzamine and atropine. Adenosine and adenosine 5′-triphosphate inhibited the release of [³H]noradrenaline produced by field stimulation.These results indicate that adenine nucleotides, probably acting via the common metabolite adenosine, inhibit adrenergic neurotransmission at a presynaptic site. Their antagonism by theophylline suggests that a presynaptic ‘purinergic’ receptor system could be involved.     

Noradrenaline release and uptake in isolated small dense cored vesicles from rat seminal ducts

Noradrenaline (NA) release and uptake was investigated in an improved preparation of isolated small dense cored vesicles from rat seminal ducts. The vesicle preparation exhibited an Mg²⁺-ATP-dependent uptake of NA. Half maximal uptake after 20 min was seen at 22 μM. Exchangeability of NA between medium and vesicles was 100%. The kinetics of exchange suggested that mos NA is stored in a single pool. The t1/2 for release of NA was 43 min in the presence of Mg²⁺-ATP, as compared to 15 min in the absence of Mg²⁺-ATP. Addition of reserpine (20 μM) did not significantly alter the NA release. The kinetic properties of this preparation was compared to those of earlier reported noradrenergic vesicle preparations from different tissues.     

Separation of synaptic vesicles of different functional states from the cholinergic synapses of the Torpedo electric organ

Synaptic vesicles were isolated on sucrose zonal gradients from perfused tissue blocks of Torpedo electric organ. They give rise to a coincident peak in the concentrations of acetylcholine and adenosine 5′-triphosphate. On low-frequency stimulation (0.1 Hz) of the nerve attached to the tissue block a distinct population of synaptic vesicles is found that sediments further into the density gradient forming a second (denser) vesicle peak. When dextran is added to the perfusate, only these denser vesicles contain electron-dense granules. This second (denser) peak contains about 25% of all vesicular acetylcholine and about 30% of the adenosine 5′-triphosphate and most of the newly synthesized acetylcholine as shown by incorporation of radiolabelled acetate. The specific radioactivity of acetylcholine in the denser vesicles after 1800 impulses is on average 16.5 times higher than that of the vesicles sedimenting at the original density and 9.2 times higher than the average of all vesicles isolated. The specific radioactivity of total tissue acetylcholine is lower than the average for all vesicles.It is concluded that stimulation makes apparent metabolic and morphological heterogeneity of synaptic vesicles. The increase in density of the vesicles containing newly synthesized acetylcholine could be due to endocytotic uptake of sucrose contained in the perfusate after the vesicle has undergone exocytosis. The results suggest that synaptic vesicles can be reloaded with transmitter and re-used even after uptake of extracellular marker.     

Modulation of the hepatocyte rough endoplasmic reticulum single chloride channel by nucleotide–Mg2+ interaction

The effect of nucleotides on single chloride channels derived from rat hepatocyte rough endoplasmic reticulum vesicles incorporated into bilayer lipid membrane was investigated. The single chloride channel currents were measured in 200/50?mmol/l KCl cis/trans solutions. Adding 2.5?mM adenosine triphosphate (ATP) and adenosine diphosphate (ADP) did not influence channel activity. However, MgATP addition inhibited the chloride channels by decreasing the channel open probability (Po) and current amplitude, whereas mixture of Mg²⁺ and ADP activated the chloride channel by increasing the Po and unitary current amplitude. According to the results, there is a novel regulation mechanism for rough endoplasmic reticulum (RER) Cl^? channel activity by intracellular MgATP and mixture of Mg²⁺ and ADP that would result in significant inhibition by MgATP and activation by mixture of Mg²⁺ and ADP. These modulatory effects of nucleotide?CMg²⁺ complexes on chloride channels may be dependent on their chemical structure configuration. It seems that Mg?Cnucleotide?Cion channel interactions are involved to produce a regulatory response for RER chloride channels.     

Ultrastructure of the primate carotid body: a morphometric study of the glomus cells and nerve endings in the monkey (Macaca fascicularis)

Summary The carotid body of the monkey (Macaca fascicularis) was studied at both the light and electron microscopic levels in an effort to provide a detailed quantitative characterization of this chemoreceptor organ in the primate. Structurally, the monkey carotid body was organized into lobules of from three to eight glomus cells (in section) and their ensheathing supporting cells. Interspersed among the lobules was abundant connective tissue stroma, fibroblasts and mast cells. Fenestrated capillaries, small arterioles and venules also permeated the organ. Each supporting cell partially ensheathed about three glomus cells and could be easily differentiated from glomus cells by their darker cytoplasmic staining, lack of dense-core vesicles and angular nuclear profile. Glomus cells exhibited an intense catecholamine histofluorescence and contained abundant dense-core vesicles. On the basis of dense-core vesicle size, shape and numerical density, four types of glomus cells were identified. The most common type (62% of all glomus cells) contained vesicles with an average diameter of 219 nm and a density of 8 vesicles per m² of cytoplasm. The second type possessed larger vesicles (264 nm in diameter) and accounted for about 14% of all glomus cells. A third type of glomus cell contained smaller (167 nm) and fewer (5 vesicles per m²) dense-core vesicles. The fourth type of glomus cell contained pleomorphic-shaped vesicles with a maximal diameter of 232 nm. Each of these last two types accounted for about 12% of all glomus cells. All four types of glomus cells were innervated, averaging 1.43 nerve endings per glomus cell (in sections). Nerve endings were primarily of the bouton-like variety averaging 2 m² in sectional area and containing 34.3 clear-core synaptic vesicles (average size 73.5 nm in diameter) per m² of cytoplasm. Of the 57 nerve endings examined in single sections, 16% displayed junctions typical of synaptic specializations and most of these were presynaptic to glomus cells. Glomus cell-glomus cell synapses were not observed. Based on these quantitative observations and on previous studies of carotid body cytoarchitecture in other laboratory species, it appears that the primate organ most closely resembles the cat carotid body, although several differences exist.