At the light microscope level, the minute concentrations of substance P (SP) in rat spinal ventral horn can be visualized best by amplification with the double bridge PAP method of Vacca et al. (1975; 1980) in 5 microns paraffin tissue sections. Morphologically, the immunoreactive sites resemble punctate bodies. They occur in close apposition with the large ventral horn cells and their associated neuronal processes. By the Sternberger PAP procedure, we now describe these punctate bodies at the electron microscope level. Ultrastructurally, they appear as tiny boutons (terminal and preterminal) and small unmyelinated processes. The boutons and processes typically contain one to several immunolabeled dense core vesicles among many immunolabeled clear vesicles. They range in size near the limit of resolution of the light microscope (LM), thereby justifying further the use of LM amplification staining by the double bridge method. The immunolabeled boutons often synapse with large smooth dendrites (which may originate from motoneurons) by asymmetrical or symmetrical synaptic densities. Their synaptic densities appear immunostained as well. The data support the view that the electrophysiological action of SP in the ventral horn occurs in part by synaptic action along the processes of the ventral horn cells. Other mechanisms of action are considered for the peptide as well. Additional types of membrane specializations (synaptoid junctions) and SP neural circuits are described below.

To the editor: Sotolongo et al. (Neurourol 8:245-254 [I 9891)Regarding the above article and the published data reported on eight feline spinal cord injury models treated on a dosage of 1.3 mg/kg of guanabenz: Comparing their urodynamic parameters (residual urine, capacity, and detrusor contraction pressure) to a control group of four untreated spinal-cord-injured animals, a significant difference was noted in all parameters between the treated and the untreated groups. In the course of a recent reanalysis of the data, however, it was noted that only three animals actually received the guanabenz at 1.3 mg/kg. In comparing their urodynamic parameters to the recalculated values for the control group, only the bladder capacity showed a trend towards a significant difference between the treatment and the control group.It has become apparent that errors were committed in our laboratory in the identification of the medications used as well as in the recorded dosage administered to the feline models. Two different analogs were used in our laboratory, and therefore the calculation of the published urodynamic parameters was not based on accurate data, since significantly fewer animals than reported actually received the guanabenz at the reported dose.