Parvalbumin (PV) is a member of the family of calcium-binding proteins. The three CBPs, PV, calbindin D28k (CB) and calretinin (CR), are important modulators of intracellular calcium dynamics in neurons. PV is a slow-onset buffer, whereas CB and CR are considered as fast buffers. In addition, PV potently modulates short-term synaptic plasticity. There are data suggesting that the CBPs play a neuroprotective role in neurodegenerative disorders.

Immunocytochemical staining of PV has been repeatedly used as marker of specific populations of neurons of the central nervous system. Immunocytochemical staining of PV also distinguishes separate subpopulations of GABAergic neurons and demonstrate morphological characteristics. In a series of previous studies, distribution patterns of PV have been characterized in nuclei and pathways of the auditory system of species such as birds and Braun et al (1991), bats, gerbils, mice and Idrizbegovic et al (2003), rats, rabbits, pigs, dolphins, monkeys, and humans. However, detailed studies describing the ultrastructure of PV-immunopositive neurons and fibers within the cat’s inferior colliculus (IC) are lacking. Here, we describe ultrastructural characteristics of PV-immunopositive neuronal perikarya, dendrites, dendritic spines, axons, and terminal boutons in the IC of adult cats.

Seven adult, untreated cats of both sexes were used in the present study. After anaesthesia with sodium pentobarbital (40 mg/kg) by intraperitoneal injection, animals were perfused transcardially with 500 ml saline, followed by 3000 ml phosphate-buffered saline (PBS; pH 7.4) containing 2.5% glutaraldehyde and 4% paraformaldehyde. Two hours after the end of the perfusion, brains were removed and stored in the same fixative for 2 h. The mesencephalon was removed. The blocks were sectioned with a thickness of 40 ?m on a Vibratome. In five cases, blocks were cut in the coronal plane, and in two cases, in the sagittal plane. In order to quench aldehyde groups, sections were treated with 1% sodium borohydride for 45 min, followed by extensive rinsing in 0.01 M PBS. Incubation for 30 min in a solution of 1% bovine serum albumin (BSA) was followed by incubation overnight in a solution of a monoclonal anti-PV antibody (Sigma, St. Louis, MO, USA), in a dilution of 1:1000. After rinsing in PBS, sections were incubated for 20 min in 1% BSA in PBS, followed by incubation for 2 h in biotinylated anti-mouse IgG (Vector, Burlingame, CA, USA) in a dilution of 1:500. After rinsing in PBS, sections were incubated in a solution of avidin–biotin–peroxidase complex (Vector) for 1 h. All incubations were carried out on a shaker at room temperature. A new series of rinses, first in PBS and then in Tris-buffer, pH 7.6, preceded visualization of peroxidase activity with H₂O₂ and 3,3?-diaminobenzidine as substrates. After rinsing in Tris buffer and phosphate buffer, part of the sections were processed for electron microscopy. Control sections were incubated in the same medium lacking the primary antibody and the results of control incubations were negative. Sections were postfixed with 1% OsO₄ in phosphate buffer for 1 h, dehydrated in graded series of ethanols and flat embedded in Durcupan (Fluka, Buchs, Switzerland) between acetate sheets. Blocks were trimmed out under a dissecting microscope and glued to epoxy blanks. Thin sections were cut with an ultramicrotome (LKB, Stockholm-Bromma, Sweden). Sections were counterstained with uranyl acetate and lead citrate and examined with a 500 electron microscope (Hitachi, Tokyo, Japan).

Sixty-five PV-immunostained neuronal perikarya were measured on the electron micrographs. The means of longest and shortest diameters were calculated. The ratio of the mean nuclear diameter and the mean diameter of the perikaryon was calculated as nucleocytoplasmic ratio. Calculation was made with an ultrastructural size calculator (Ted Pella, Tustin, CA, USA).

We discriminated three major parts of the IC as defined by Berman (1968): central nucleus, pericentral nucleus, and external nucleus. Immunostaining of PV was detected in all parts of the IC. Especially towards the rostral pole of the IC, prominent clusters of PV-immunopositive perikarya were observed, surrounded by a dense network of immunostained processes. Most of the PV-positive cells in this region were small to medium sized, and were found predominantly in the ventromedial part of the central nucleus. Large PV-positive neurons were scattered throughout the IC and were more frequently found in the dorsomedial part of the central nucleus and in the deep part of the external nucleus. The immunopositive perikarya exhibited various shapes: oval, multipolar to strongly elongated, fusiform. The intensity of immunostaining also varied: from strongly-stained cells that almost looked like impregnated perikarya to weakly-stained cells. Cell nuclei were usually unstained but some nuclei also contained a large amount of reaction product. As a rule, dendrites were only faintly stained, whereas proximal dendritic trunks of larger neurons exhibited substantial immunostaining only. In contrast, heavily immunostained axons and “puncta” (presumably synaptic boutons–SB) were found in all territories of the IC. Their density was largest in the central nucleus, and somewhat less in the pericentral nucleus, whereas the number of immunostained axons was lower in the external nucleus.

Figure 1. Coronal section through the rostral half of the IC. NC, nucleus centralis; NP, nucleus pericentralis; NE, nucleus externus; SGC, substantia grisea centralis. The boxed region indicates the ventromedial part of the central nucleus in which the number of PV-immunopositive neurons is high and is shown at a higher magnification in Fig. 2.

Ultrastructural analysis demonstrated PV immunostaining in neuronal perikarya, dendrites, dendritic spines, myelinated and unmyelinated axons, and terminal SBs.

Neurons of various shapes and sizes were stained for PV. On the basis of the size of perikarya, the ultrastructural characteristics of organelles, and occurrence and number of axosomatic synaptic contacts, PV-stained neurons were classified in three types: large, medium-sized, and small.

Approx. 15% of the immunolabeled perikarya in the IC belong to the large-sized group of PV-positive neurons. The dimensions of perikarya of large-sized neurons were >25 ?m in diameter whereas some elongated cells were up to 40 ?m in length. The labeled perikarya had an irregular appearance, being either slightly elongated, oval, triangular, or fusiform. Nuclei of large neurons were located centrally and surrounded by abundant cytoplasm. The shape of the nucleus usually reflected the shape of the perikaryon. The chromatin was finely dispersed as euchromatin throughout the nucleus and little heterochromatin was present. Nucleoli were prominent and centrally located. Slightly more than half of the large neurons were immunostained for PV in the nucleus. The reaction product was homogeneously distributed throughout the cytoplasmic compartments, occasionally extending into somatic spines. The number of organelles was high. In some sections of perikarya, more than 150 mitochondrial profiles could be counted. The Golgi apparatus was well developed; numerous relatively small dictiosomes were mainly found in perinuclear position. Granular endoplasmic reticulum was prominent. It was organized in Nissl bodies, but there were also numerous scattered cisternae, intermingled with free ribosomes. Cisternae located underneath the plasma membrane were not a rare finding. The perikaryal surface was contacted by numerous terminal boutons, forming symmetrical and, more rarely, asymmetrical axosomatic synapses. Some of the axosomatic boutons were stained for PV and formed symmetrical synaptic contacts.