The distribution of tau in the mammalian central nervous system.

We have determined the biochemical and immunocytochemical localization of the heterogeneous microtubule-associated protein tau using a monoclonal antibody that binds to all of the tau polypeptides in both bovine and rat brain. Using immunoblot assays and competitive enzyme-linked immunosorbent assays, we have shown tau to be more abundant in bovine white matter extracts and microtubules than in extracts and microtubules from an enriched gray matter region of the brain. On a per mole basis, twice-cycled microtubules from white matter contained three times more tau than did twice-cycled microtubules from gray matter. Immunohistochemical studies that compared the localization of tau with that of MAP2 and tubulin demonstrated that tau was restricted to axons, extending the results of the biochemical studies. Tau localization was not observed in glia, which indicated that, at least in brain, tau is neuron specific. These observations indicate that tau may help define a subpopulation of microtubules that is restricted to axons. Furthermore, the monoclonal antibody described in this report should prove very useful to investigators studying axonal sprouting and growth because it is an exclusive axonal marker.

Differential localization of class III, beta-tubulin isotype and calbindin-D28k defines distinct neuronal types in the developing human cerebellar cortex.

This immunohistochemical study compares the localization of the neuronal class III beta-tubulin isotype (beta III) to that of calbindin-D28k in 40 human fetal and postnatal cerebella ranging from 12 weeks gestation to adulthood. In the external granule layer of the developing cerebellar cortex, beta III staining was present in the premigratory (postmitotic) zone of horizontal neurons but was absent in "epithelioid" cells of the subpial proliferative mitotic zone. In the molecular layer, intense beta III staining was associated with parallel fibers, stellate/basket neurons and migrating fusiform granule neurons. beta III staining was also present in internal granule neurons. In contrast, beta III was not detectable in fetal and neonatal Purkinje neurons and Golgi II neurons, but was evident in these neurons from juvenile and adult cerebella. Calbindin-D28k staining was present in Purkinje neurons also delineating their somatic spines ("pseudopodia"), lateralizing and apical dendrites (including dendritic spines), subpopulations of small to intermediate-sized Golgi II neurons in the internal granule layer ("synarmotic cells" of Landau), large to medium-sized subcortical Golgi II neurons and neurons of cerebellar roof nuclei, at various gestational stages and postnatally. It was absent in the external granule layer, parallel fibers, stellate/basket and internal granule neurons. Variable degrees of beta III and calbindin-D28k staining were detected in subpopulations of immature neuroepithelial cells of the ventricular matrix at the roof of the fourth ventricle. Glial (including Bergmann glia) and mesenchymal cells were not stained for either antigenic determinants. The differential expression of calbindin-D28k and beta III defines distinct populations of neurons in the developing human cerebellar cortex and supports the ontogenetic concept of Ramon y Cajal.

Glomus cell differentiation in the carotid body region of chick embryos studied by neuron-specific class III beta-tubulin isotype and Leu-7 monoclonal antibodies.

Development of the carotid body and the glomus cell groups in the wall of the common carotid artery and its branches was examined in chickens at various developmental stages by immunohistochemistry using three different monoclonal antibodies, i.e., anti-neuron-specific class III beta-tubulin isotype (TuJ1), anti-rat brain beta-tubulin, and anti-Leu-7 (HNK-1) antibodies. All the antibodies reacted with neurons. The carotid body anlage was first discerned at 6 days of incubation at the lateral portion of the third branchial artery. The cells and nerve fibers immunoreactive for TuJ1, brain beta-tubulin and Leu-7, which were connected with the distal ganglion of the vagus nerve, were found around the carotid body anlage at this stage. Within the carotid body anlage, no immunoreactivity yet appeared. The immunoreactive cells were accumulated around the carotid body anlage until 8 days of incubation. From 9 days of incubation, the immunoreactive cells continuing with the distal vagal ganglion began to enter into the carotid body anlage and also dispersed widely along the common carotid artery and its branches, giving rise to the glomus cells. At 12 days of incubation, a large portion of the carotid body was occupied by the immunoreactive cells. Thus, the present study evidences that the glomus cells in the carotid body and around the arteries are emigrés that arrive in each residential place from the distal vagal ganglion. Immunoreactivity for TuJ1, brain beta-tubulin, and Leu-7 in the glomus cells started to decrease at late stages of embryonic development. After hatching, no TuJ1-immunoreactive cells were detected in the carotid body region.

Studies of the principal sensory and spinal trigeminal nuclei of the rat: projections to the superior colliculus, inferior olive, and cerebellum.

We have analyzed the connections between the sensory trigeminal nuclei and two major sensorimotor areas (i.e., the superior colliculus and crura I and II of the cerebellar cortex) in which tactile input from peri-oral and other facial regions is a prominent feature. Following injections of horseradish peroxidase into the superior colliculus, retrogradely labeled cells occupy the ventral one-third of the contralateral principal sensory and spinal trigeminal nucleus; trigeminocollicular neurons are especially numerous within the subnucleus interpolaris (Svi). Injections of either 3H-proline or horseradish peroxidase (HRP) into the Svi reveal that trigeminocollicular axons reach the rostral two-thirds to three-quarters of the contralateral superior colliculus, where they distribute in a nonuniform, patchy manner within layers IV-VI. In addition to demonstrating the trigeminocollicular projection, anterograde and retrograde transport studies of the Svi also reveal a trigeminoolivary projection which terminates primarily within the contralateral rostral dorsal accessory (DAO) and adjacent principal (PO) olives; some of the Svi neurons innervate both the superior colliculus and the DAO-PO via axon collaterals. Data from a final set of retrograde tracing experiments show that the trigeminorecipient zone of the DAO-PO contains neurons which project to crura I and/or II of the cerebellar cortex. Of the various submodalities conveyed by the trigeminal system, it is likely that the trigeminal connections we have demonstrated are carrying tactile information. This is indicated by the fact that responses to tactile stimulation of the face have been reported for cells in (1) the deeper collicular layers, (2) the trigeminorecipient zone of the DAO-PO, and (3) cerebellar targets of this zone, crura I and II. All data are discussed in the context of the anatomical and physiological literature.

Development of the peripheral trigeminal system in the chick revealed by an isotype-specific anti-beta-tubulin monoclonal antibody.

Using a monoclonal antibody directed against a class III beta-tubulin isotype, c beta 4, we studied the time course of the expression of this protein, the morphological differentiation of the immunoreactive cells, and the time course of peripheral axon outgrowth in the chick trigeminal (V) system. The neural crest precursors of the V ganglion neurons do not express the antigen until they begin to differentiate as neurons. The placodal precursors of the V ganglion neurons express the antigen while they still are cuboidal epithelial cells. They continue to be immunoreactive as they migrate from the placode and settle in the ganglion, prior to sprouting axons. The V motoneurons express the antigen near the time of their terminal mitotic division. Using this antibody to visualize axons, we demonstrate that both ganglionic and motoneuron axons grow out as individual fibers, much like pioneer axons. Both halt their extension for several hours once they attain the vicinity of their targets. During this pause many other axons join the nerve bundles. Finally, single pioneer axons split from the main trunk to begin local target innervation.

Developmental expression of a neuron-specific beta-tubulin in frog (Xenopus laevis): a marker for growing axons during the embryonic period.

In mammals, there are seven classes of beta-tubulin genes, one of which, class III, is neuron specific. Using class-specific monoclonal antibodies, class III beta-tubulin protein could not be detected in frog embryos or in adults with either Western blotting or immunohistochemical techniques. In contrast, the class II beta-tubulin protein, which is predominant in mammalian brain but is also expressed in other tissues, is expressed only in neurons in frog embryos. Protein was detected only in neurons from late stages of neural tube closure through premetamorphic stages. At stages 21-28, the pioneering axons of Rohon-Beard, commissural, primary motor, and trigeminal ganglion neurons were distinctly stained in the axon scaffolds that they formed in the embryonic brain and the peripheral mesenchyme. Nonneuronal cells, both outside the nervous system and within it (e.g., radial glia, Müller glia, roof plate, and floor plate cells) were not immunoreactive. Throughout swimming and premetamorphic stages, neuronal cells in all brain regions became immunoreactive as they differentiated and extended axons. Whereas many embryonic neurons became postmitotic during gastrulation stages, neurons expressed detectable levels of class II beta-tubulin protein only beginning at the onset of overt axon outgrowth. These observations demonstrate that the neuron-specific beta-tubulin in frog is a different gene from that in mammals, and its protein product is detectable at the time of axonogenesis rather than neurogenesis.

Ascending pathways from the monkey superior colliculus: an autoradiographic analysis.

The autoradiographic tracing method has been used to analyze the distribution of ascending tectofugal pathways in the rhesus monkey. Our findings show that axons which arise from deep collicular neurons terminate within several dorsal thalamic nuclei which in turn project upon the frontal eye fields (area 8) and the inferior parietal lobule (area 7). Both of these cortical areas are functionally quite similar to the deep colliculus, and we suggest that ascending channels from the deep tectum must account, at least in part, for these functional similarities. The present autoradiographs reveal projections to several nuclear zones previously not identified as deep collicular targets in the monkey. Such targets include the visceral cell columns of the oculomotor complex, the rostral interstitial nucleus of the medial longitudinal fasciculus, and the magnocellular division of the ventral anterior nucleus. Deep tectal input also has been shown to terminate quite extensively within the paralamellar region of the mediodorsal nucleus and in the parafascicular nucleus; very little input to the central lateral and centromedian nuclei was observed. Radioisotope injections restricted to the superficial layers reveal dense projections to the parabigeminal nucleus, the pretectum, the inferior and lateral pulvinar nuclei, and to the ventral and dorsal lateral geniculate nuclei. Transported protein within the dorsal lateral geniculate nucleus occupied the "S" layers and the interlaminar zones.

Cytoskeletal immunohistochemistry of central neurocytomas.

Central neurocytomas are rare intraventricular tumors. Patients with such tumors have a favorable prognosis after surgical removal. These tumors may be misdiagnosed as neuroblastomas or gliomas, risking the complications of adjuvant therapy. Diagnosis of central neurocytoma requires that the tumor shows the ultrastructural features of mature neuronal differentiation, including the presence of synapses and dense-core and clear vesicles in addition to profiles of neuritic processes with microtubules. The cytoskeletal phenotype of central neurocytomas has not been previously characterized, but it may facilitate their definitive recognition when ultrastructural examination is not possible. Ten central neurocytomas were examined by immunohistochemistry for phosphorylation-dependent/independent neurofilament epitopes, neuron-associated class III beta-tubulin, microtubule-associated proteins (MAP2, tau), and glial fibrillary acidic protein (GFAP). The neuronal nature of all neoplasms was documented by immunoreactivity for synaptophysin in nine tumors and for phosphorylation-independent neurofilament-H/M in the remaining case. Electron microscopy in four cases showed synapses and dense core vesicles. All tumors were immunoreactive for class III beta-tubulin and MAP2, which were seen in cytoskeletal structures by immunoelectron microscopy. Two thirds of the cases were immunohistochemically positive for neurofilament epitopes. None of the tumor cells displayed GFAP immunoreactivity, although reactive astrocytes were present. These data suggest that central neurocytomas may be recognized by synaptophysin immunoreactivity and that the expression of cytoskeletal epitopes indicates that these tumors are well-differentiated neuronal neoplasms.

The distribution and some morphological features of substantia nigra neurons that project to the thalamus, superior colliculus and pedunculopontine nucleus in the monkey.

Neurons of the substantia nigra's pars reticulata that send axons to the thalamus, superior colliculus and midbrain reticular formation (including the pedunculopontine nucleus) have been revealed in monkeys by the technique of retrograde transport of horseradish peroxidase. The populations of nigrothalamic, nigrotectal and nigroreticular neurons differ from one another in their number, intranigral distribution and somatodendritic size and shape. Nigrothalamic cells are the most abundant and, although scattered throughout the mediolateral expanse of the pars reticulata, their numbers progressively diminish from rostral to caudal levels. Nigrotectal cells are least numerous and are restricted almost exclusively to the lateral margin of the rostral one-half of the pars reticulata. Nigroreticular cells, like nigrothalamic, are scattered throughout the mediolateral dimension of the nucleus, but are more commonly located at middle to caudal levels. In addition to their restricted intranigral location, the nigrotectal cells are larger, polygonal and have more major dendritic processes than the smaller nigrothalamic and nigroreticular cells which are usually triangular or fusiform. A small proportion of cells of all three types appears to project contralaterally. These findings indicate that the efferent organization of the primate pars reticulata differs markedly from that of the rodent and the monkey's nigrotectal cells constitute a spatially and morphologically distinct subpopulation within the pars reticulata. These data should be useful in understanding the functional organization of topographic inputs to the pars reticulata such as that from the neostriatum.

Olfactory neuroblastoma. Additional immunohistochemical characterization.

A panel of 12 antibodies was used to further characterize the immunohistochemical staining profile of olfactory neuroblastoma. The following results were obtained for the 11 neoplasms that were immunostained: neuron-specific enolase 11/11(+), S-100 protein 8/11(+), microtubule-associated protein-2 8/11(+), class III beta-tubulin isotype 9/11(+), neurofilament 200 kD 8/11(+), synaptophysin 7/11(+), glial fibrillary acidic protein 1/11(+), chromogranin A 1/11(+), vimentin 1/11(+), keratin (CAM 5.2) 4/11(+), keratin (AEI/AE3) 0/11(+), and epithelial membrane antigen 0/11(+). Expression of two intermediate filaments was found in 4 of the 11 tumors. The authors' data showing that 72% of olfactory neuroblastomas were S-100 protein positive and only one was immunoreactive for glial fibrillary acidic protein agree with other published immunohistochemical studies. With only a single exception, each of the 11 neoplasms was labeled with one or more antibodies that detect neuronal cytoskeletal proteins (class III beta-tubulin isotype, microtubule-associated protein-2, neurofilament 200 kD). These immunohistochemical results are complementary to the reported electron microscopic findings of intermediate filaments and microtubules in olfactory neuroblastomas.

Immunohistochemical localization of a neuron-specific beta-tubulin isotype in the developing chicken ultimobranchial glands.

The localization of a neuron-specific beta-tubulin isotype in the ultimobranchial glands from chickens at various stages of development was studied by means of light- and electron microscopic immunohistochemistry with a monoclonal antibody (TuJ1) against the beta-tubulin isotype, c beta 4. At 8 days of incubation, many C cells in the ultimobranchial glands showed immunoreactivity for TuJ1 invariable degrees, weak to intense. At 12 days of incubation, a vast majority of C cells were intensely immunoreactive for TuJ1, and further TuJ1-immunoreactive nerve fibers were distributed in the ultimobranchial glands. At 14 and 16 days of incubation, intense immunoreactivity for TuJ1 was sustained in the C cells. Electron microscopic analyses revealed that TuJ1 immunoreactivity was diffusely distributed throughout the cytoplasm and also localized on the secretory granules of C cells at these stages. TuJ1 immunoreactivity in the C cells started to decrease at late stages of embryonic development. At the hatching period, dense distributions of TuJ1-immunoreactive nerve fibers were observed in the ultimobranchial glands, whereas TuJ1 immunoreactivity of the C cells became very weak. In 10-day-old chickens, TuJ1 immunoreactivity was restricted to the nerve fibers.

The trigeminocollicular projection in the cat: patch-like endings within the intermediate gray.

We have used two neuroanatomical tracing techniques to study the trigeminocollicular projection in the cat. In one series of experiments we injected [3H]proline into the alaminar division of the spinal trigeminal nucleus and analyzed the distribution and pattern of anterogradely transported label within the superior colliculus. These autoradiographic data reveal that the trigeminocollicular projection: (1) is primarily contralateral; (2) reaches only the rostral 60-70% of the colliculus; and (3) terminates in a discontinuous, patch-like tier within the middle of the dorsal-ventral axis of the stratum griseum intermediale (SGI). The patches of label measure approximately 330 micrometer in the medial-lateral dimension and 250 micrometer in the dorsal-ventral extent. There seems to be an alignment of the patches in the rostral-caudal direction, suggesting that the trigeminal input forms longitudinal columns in the colliculus. Anterogradely transported protein is also present within the stratum griseum profundum (SGP). In contrast to the patch-like pattern present in the SGI, label in the SGP is more diffusely distributed. In a second series of experiments we injected horseradish peroxidase-wheat germ agglutinin (HRP-WGA) into the spinal trigeminal nucleus. While the distribution and pattern of the contralateral trigeminocollicular axons is similar to that in the autoradiographic experiments, patches of anterogradely transported HRP-WGA are also present within the ipsilateral SGI and SGP. Furthermore, the HRP-WGA data reveal groups of retrogradely labeled collicular neurons which lie in intimate association with the patches of anterogradely transported tracer. Our findings are discussed in relation to other collicular afferents which terminate in a patch-like manner within the SGI. We hypothesize that the colliculus contains many vertically oriented modules. Each module consists of tectopetal axon terminals -- arranged in sandwich fashion -- and functionally related collicular (tectofugal) neurons.

An autoradiographic analysis of the tecto-olivary projection in primates.

Autoradiographic tracing methods were used to demonstrate a well-defined projection from the superior colliculus to the inferior olivary complex in the monkey. This projection originates within the deep layers of the superior colliculus, descends within the contralateral tecto-spinal tract, and terminates within the caudal 1/3 of the medial accessory nucleus. The terminal field is restricted to a densely packed, darkly stained group of cells located in the most dorsal segment of subnucleus b. In one animal, another group of olivary afferents was identified. These fibers also descend within the contralateral tecto-spinal tract, and terminate within the dorsal cap of Kooy. While it was not possible to determine the origin of this projection, our data suggest that it arises within a region adjacent to the rostral pole of the superior colliculus. The present study further indicates that in the monkey relatively few axons which course within the classical tecto-spinal tract pass caudal to the medulla.

The neuronal response to injury as visualized by immunostaining of class III beta-tubulin in the rat.

The neuronal response to trauma of the brain and spinal cord was examined by staining sections of injured central nervous system (CNS) with a monoclonal antibody (TuJ1) that recognizes class III beta-tubulin exclusively. Because class III beta-tubulin is expressed by neurons and not by glia, this monoclonal antibody stains neuronal cell bodies, dendrites, axons and axonal terminations darkly with a pale staining background. Thus, the TuJ1 antibody is extremely useful, revealing the fine details of axons and their terminations, as well as significant injury-related alterations in the composition of the somatic cytoskeleton.

A cytomorphological scheme of differentiating neuronal phenotypes in cerebellar medulloblastomas based on immunolocalization of class III beta-tubulin isotype (beta III) and proliferating cell nuclear antigen (PCNA)/cyclin.

This immunohistochemical study compares the localization of the neuronal class III beta-tubulin isotype (beta III) to that of the proliferating cell nuclear antigen (PCNA)/cyclin in 46 cerebellar neuroblastic tumors (medulloblastomas). Both class III beta-tubulin (beta III) and PCNA/cyclin reactivities were present in all tumors, but the topographic distribution and cytomorphologic features of stained cells varied considerably between classic and desmoplastic medulloblastomas. Four neoplastic phenotypes, representing gradations of neuronal differentiation, were identified: [Allegranza 1991] apolar, blast-like PCNA/cyclin(+) cells devoid of beta III reactivity (Nb1); [Bravo et al. 1987] apolar, often binucleated and/or fusiform, PCNA/cyclin (+) cells with pronounced beta III staining in their protoperikarya and their growth cones (Nb2); [Burger et al. 1987] beta III-immunoreactive immature polar neurons with varying degrees of neuritic development, reading to significant neuritogenesis in the "pale islands" of desmoplastic medulloblastomas (Nb3). The majority of Nb3 phenotypes were PCNA/cyclin (-), although subpopulations of such polar tumor cells exhibiting PCNA staining were also identified; and [Burger et al. 1991] beta III-immunoreactive, PCNA/cyclin (-) mature ganglion-like cells (Nb4). A high PCNA/cyclin labeling index (> 80%) was obtained in 20 poorly differentiated classic medulloblastomas while, significant intratumoral staining heterogeneity was observed in 23 cases of desmoplastic medulloblastomas and 3 cases of "medulloblastomas with ganglion cells": A high labeling index (LI)(> 80%) in the reticulin-impregnated poorly differentiated areas of tumor contrasted with sharp decline of PCNA staining and a very low LI (< 10%) in areas of overt neoplastic neuritogenesis ("pale islands") displaying strong beta III reactivity. Neoplastic ganglion cells were beta III (+)/PCNA (-). Our findings indicate that the majority of differentiating neuronal phenotypes undergoing cytomorphological changes of neuritic development (Nb3), and all neoplastic ganglion cells (Nb4 phenotypes) are PCNA (-), in contrast to actively proliferating, poorly differentiated, tumor cells that are PCNA (+). Although PCNA staining corresponded in part, to beta III (-) blast-like elements (Nb1), a co-expressive pattern of staining for beta III and PCNA/cyclin also was observed in subpopulations of poorly differentiated tumor cells (Nb2), indicating that transformed neuroblasts are capable of expressing differentiation-associated neuronal cytoskeletal proteins while still remaining in the proliferative compartment of the cell cycle. Our observations suggest that only neuritogenesis and acquisition of ganglionic phenotype are significant maturational events in medulloblastomas (indicating entry into the quiescent phase of the cell cycle) and provide further support for the neuronal lineage and differentiation potential of these cerebellar embryonal tumors.

The stromal Schwann cell during maturation of peripheral neuroblastomas. Immunohistochemical observations with antibodies to the neuronal class III beta-tubulin isotype (beta III) and S-100 protein.

This immunohistochemical study compares the localization of the neuronal class III beta-tubulin isotype (beta III; analogous to the beta' 1-/beta 2-tubulin isoform) to the Schwann cell-associated S-100 protein focusing on topographic relationships of Schwann-like cells to differentiating neuronal phenotypes during stromal development in human peripheral neuroblastomas. The earliest appearance of Schwann cells in poorly differentiated (classical) neuroblastomas is heralded by S-100 protein-immunoreactive cells in close association with tumor blood vessels. In subsequent stages of maturation, i.e. maturing neuroblastoma (ganglioneuroblastoma and gangliocytoma), S-100 protein-positive cells are mostly confined to the connective tissue septa dividing tumor into lobules, and are not freely interspersed with beta III-immunoreactive neoplastic neurons. Significant ensheathment of individual axon-like processes by Schwann cells occurs only in mature ganglioneuromas. beta III is localized in a full spectrum of neoplastic neuronal phenotypes, ranging from poorly-differentiated apolar neuroblasts (often signaling ensuing neuritogenesis) to mature ganglion cells, but not in Schwann cells, or other cell types of the stroma. Our observations suggest that Schwann cells in peripheral neuroblastomas are stroma-derived cells and not an expression of divergent neoplastic differentiation.

Localization of the neuronal class III beta-tubulin isotype in foci of early neuritogenesis supports divergent neuroblastic differentiation potential in Wilms' tumors.

Wilms' tumors are embryonic neoplasms that have been proposed to originate from the metanephric blastema and are capable of divergent epithelial and mesenchymal differentiation. Neuroepithelial differentiation in these tumors remains controversial. The aim of this study was to examine the phenotypic profile of certain neuronal and glial antigenic determinants in a series of 21 Wilms' tumors. Immunohistochemical studies were performed by using monoclonal antibodies against the neuronal class III beta-tubulin isotype (beta III), the phosphorylated and phosphorylation-independent epitopes of neurofilament protein, and synaptophysin; antisera to gamma-enolase (neuron-specific enolase) glial fibrillary acidic protein, and S100 protein were also used. Foci of neoplastic cells with neurite-like processes that exhibited intense beta III staining were demonstrated in blastemalike areas of three of 21 tumors. In one case, Homer Wright rosettes (stained for beta III) were identified. Areas of abortive neuritic development were also labeled with antibodies to gamma-enolase. No reactivity was obtained in these foci for phosphorylated and phosphorylation-independent epitopes of neurofilament protein, synaptophysin, glial fibrillary acidic protein, and S100 protein. The remainder of the tumors (18 of 21) were negative with the panel of neural markers. Our results indicate that divergent neuroblastic differentiation, evidenced as early neoplastic neuritogenesis, may be present in the blastematous component of Wilms' tumor subsets.

Cerebellar desmoplastic medulloblastomas. A further immunohistochemical characterization of the reticulin-free pale islands.

We studied by immunohistochemistry the features of differentiation in 24 desmoplastic and 16 classic medulloblastomas (median patient ages, 18 and 6.5 years, respectively) with the use of a panel of cytoskeletal and synaptosomal markers. A distinctive pattern of immunoreactivity with a series of monoclonal antibodies (Mabs) was documented in the polar tumor cells forming the reticulin-free pale islands of the desmoplastic variant, denoting overt neuritogenesis. These comprised the following: (1) Mab Tp-NFP1A3 recognizing an epitope in the high-molecular-weight (Mr) isoform of neurofilament protein; (2) Mab AP18 to the high-Mr microtubule-associated protein 2; (3) Mab TUJ1 recognizing the class III beta-tubulin isotype (human h beta 4); and (4) Mab SY38 to synaptophysin. Immunoblot analysis confirmed the expression of h beta 4 in three medulloblastomas, yielding strong single bands in two desmoplastic medulloblastomas and a considerably weaker band in one classic medulloblastoma. Glial fibrillary acidic protein-positive tumor cells frequently formed an integral component of the pale islands. Oligodendrogliallike areas in one classic and in three desmoplastic medulloblastomas were immunopositive for the Mabs to synaptophysin, microtubule-associated protein 2, and h beta 4, indicating a neuroblastic nature. We propose that the reticulin-free structures of desmoplastic medulloblastomas constitute neoplastic foci with features of predominantly neuronal and, to a lesser degree, astroglial differentiation.

Differential distribution of the neuron-associated class III beta-tubulin in neuroendocrine lung tumors.

OBJECTIVE: To study the immunoreactivity profile of the neuron-associated class III beta-tubulin isotype (beta III) in epithelial lung tumors. DESIGN: One hundred four formalin-fixed, paraffin-embedded primary and metastatic lung cancer specimens were immunostained with an anti-beta III mouse monoclonal antibody (TuJ1) and an anti-beta III affinity-purified rabbit antiserum. Paraffin sections from fetal, infantile, and adult nonneoplastic lung tissues were also examined. RESULTS: In the fetal airway epithelium, beta III staining is detected transiently in rare Kulchitsky-like cells from lung tissues corresponding to the pseudoglandular and canalicular but not the saccular or alveolar stages of development. beta III is absent in healthy, hyperplastic, metaplastic, and dysplastic airway epithelium of the adult lung. In contrast, beta III is highly expressed in small cell lung cancer, large cell neuroendocrine carcinoma, and in some non-small cell lung cancers, particularly adenocarcinomas. There is no correlation between expression of beta III and generic neuroendocrine markers, such as chromogranin A and/or synaptophysin, in pulmonary adenocarcinomas. Also, focal beta III staining is present in primary and metastatic adenocarcinomas (to the lung) originating in the colon, prostate, and ovary. beta III is expressed to a much lesser extent in atypical carcinoids and is rarely detectable in typical carcinoids and squamous cell carcinomas of the lung. The distribution of beta III in small cell lung cancer and adenocarcinoma metastases to regional lymph nodes and brain approaches 100% of tumor cells, which is substantially greater than in the primary tumors. CONCLUSIONS: In the context of neuroendocrine lung tumors, beta III immunoreactivity is a molecular signature of high-grade malignant neoplasms (small cell lung cancer and large cell neuroendocrine carcinoma) although its importance in atypical carcinoids must be evaluated further. In addition, beta III may be a useful diagnostic marker in distinguishing between small cell lung cancers and certain non-small cell lung cancers (poorly differentiated squamous cell carcinomas), especially in small biopsy specimens. To our knowledge, beta III is the only tumor biomarker that exhibits a substantially more widespread distribution in poorly differentiated than in better differentiated pulmonary neuroendocrine tumors. However, the significance of beta III phenotypes in non-small cell lung cancer, particularly adenocarcinoma, with respect to neuroendocrine differentiation and prognostic value, requires further evaluation.

Aberrant localization of the neuronal class III beta-tubulin in astrocytomas.

BACKGROUND: The class III beta-tubulin isotype (betaIII) is widely regarded as a neuronal marker in development and neoplasia. In previous work, we have shown that the expression of betaIII in neuronal/neuroblastic tumors is differentiation dependent. In contrast, the aberrant localization of this isotype in certain nonneuronal neoplasms, such as epithelial neuroendocrine lung tumors, is associated with anaplastic potential. OBJECTIVE: To test the generality of this observation, we investigated the immunoreactivity profile of betaIII in astrocytomas. DESIGN: Sixty archival, surgically excised astrocytomas (8 pilocytic astrocytomas, WHO grade 1; 18 diffuse fibrillary astrocytomas, WHO grade 2; 4 anaplastic astrocytomas, WHO grade 3; and 30 glioblastomas, WHO grade 4), were studied by immunohistochemistry using anti-betaIII monoclonal (TuJ1) and polyclonal antibodies. A monoclonal antibody to Ki-67 nuclear antigen (NC-MM1) was used as a marker for cell proliferation. Antibodies to glial fibrillary acidic protein (GFAP) and BM89 synaptic vesicle antigen/synaptophysin were used as glial and neuronal markers, respectively. RESULTS: The betaIII immunoreactivity was significantly greater in high-grade astrocytomas (anaplastic astrocytomas and glioblastomas; median labeling index [MLI], 35%; interquartile range [IQR], 20%-47%) as compared with diffuse fibrillary astrocytomas (MLI, 4%; IQR, 0.2%-21%) (P <.0001) and was rarely detectable in pilocytic astrocytomas (MLI, 0%; IQR, 0%-0.5%) (P <.0001 vs high-grade astrocytomas; P <.01 vs diffuse fibrillary astrocytomas). A highly significant, grade-dependent relationship was observed between betaIII and Ki-67 labeling and malignancy, but this association was stronger for Ki-67 than for betaIII (betaIII, P <.006; Ki-67, P <.0001). There was co-localization of betaIII and GFAP in neoplastic astrocytes, but no BM89 synaptic vesicle antigen/synaptophysin staining was detected. CONCLUSIONS: In the context of astrocytic gliomas, betaIII immunoreactivity is associated with an ascending gradient of malignancy and thus may be a useful ancillary diagnostic marker. However, the significance of betaIII-positive phenotypes in diffuse fibrillary astrocytomas with respect to prognostic and predictive value requires further evaluation. Under certain neoplastic conditions, betaIII expression is not neuron specific, calling for a cautious interpretation of betaIII-positive phenotypes in brain tumors.