A monoclonal antibody (RT97) against neurofilament protein specifically and exclusively labelled a subpopulation of rat dorsal root ganglion (DRG) neurones. For seven ganglia (L4 and T13) studied quantitatively the frequency distribution histograms of the size of labelled cells could be fitted by a single normal distribution whose parameters were extremely close to those of the normally distributed large light cell population in that ganglion. On this basis and on the basis of a statistical analysis of the results it was suggested that this antibody can be used as a much needed specific label for the large light population of neurones in rat DRGs. The small dark neurone population was not labelled by this antibody. In one ganglion the subjective analysis of whether each neurone was labelled or not was directly compared with microdensitometric measurements of reaction product intensity. This analysis supported the above conclusion, and furthermore no subdivisions of the labelled population were apparent on the basis of neuronal size plotted against intensity of the reaction product. Other neuronal cell bodies strongly labelled by this antibody were found in association with small unlabelled neurones not only in DRGs, but also in the trigeminal ganglion, the vagal ganglia, and the mesencephalic V nucleus, all of which are made up of primary afferent neurones and all of which are completely or partially derived from the neural crest. Sympathetic and central nervous system neuronal cell bodies were unlabelled or occasionally very lightly labelled although immunoreactive fibres abound in the central nervous system.

The monoclonal antibody UJ13A was raised following immunization of mice with human foetal brain and subsequent somatic cell hyridization of spleen cells with the mouse myeloma cell line P3-X63-AG8-653. The antibody is of the IgG1 subclass and has been shown by indirect immunofluorescence studies on normal foetal, paediatric and adult tissues to selectively bind to most tissues of neuroectodermal origin. Many tumours of neural origin also express the UJ13A antigen and the reagent can be used to distinguish primary intracranial neural tumours from secondary carcinomas and lymphomas. UJ13A is also useful as one of a panel of reagents employed for the identification of metastatic spread of neuroblastoma cells to bone marrow and cerebrospinal fluid. Knowledge of the full spectrum of normal and malignant tissues binding UJ13A suggests that the antibody may have a role in the radioimmunolocalization of neuronal tumours such as neuroblastoma.

Conventional antibodies have long been used in an attempt to produce specific neural markers. Such markers would be invaluable for studying the structural organization and development of the nervous system. Unfortunately, they have not been found to discriminate satisfactorily between neuronal subpopulations. Recent developments of the hybridoma technique, however, promise to provide monoclonal antibodies of adequate specificity. Such antibodies have already generated and shown to be capable of distinguishing between individual neurones of the leech nervous system. We report here two monoclonal antibodies which, although generated against human T cells, react exclusively with Purkinje cells in the vertebrate central nervous system (CNS). This new specificity arose out of a fortuitous observation made during examination of the lymphocyte infiltration of human cerebellar tumours with the monoclonal antibody, UCHT1. Although widely used as a T-cell marker, its reaction with neural tissue has not hitherto been described. To our knowledge, this is the first description of a monoclonal antibody which recognises discrete neuronal population in the vertebrate brain.