High-density extracellular probes reveal dendritic backpropagation and facilitate neuron classification.

Different neuron types serve distinct roles in neural processing. Extracellular electrical recordings are extensively used to study brain function but are typically blind to cell identity. Morphoelectrical properties of neurons measured on spatially dense electrode arrays have the potential to distinguish neuron types. We used high-density silicon probes to record from cortical and subcortical regions of the mouse brain. Extracellular waveforms of each neuron were detected across many channels and showed distinct spatiotemporal profiles among brain regions. Classification of neurons by brain region was improved with multichannel compared with single-channel waveforms. In visual cortex, unsupervised clustering identified the canonical regular-spiking (RS) and fast-spiking (FS) classes but also indicated a subclass of RS units with unidirectional backpropagating action potentials (BAPs). Moreover, BAPs were observed in many hippocampal RS cells. Overall, waveform analysis of spikes from high-density probes aids neuron identification and can reveal dendritic backpropagation. NEW & NOTEWORTHY It is challenging to identify neuron types with extracellular electrophysiology in vivo. We show that spatiotemporal action potentials measured on high-density electrode arrays can capture cell type-specific morphoelectrical properties, allowing classification of neurons across brain structures and within the cortex. Moreover, backpropagating action potentials are reliably detected in vivo from subpopulations of cortical and hippocampal neurons. Together, these results enhance the utility of dense extracellular electrophysiology for cell-type interrogation of brain network function.

Temporal identity establishes columnar neuron morphology, connectivity, and function in a Drosophila navigation circuit.

The insect central complex (CX) is a conserved brain region containing 60 + neuronal subtypes, several of which contribute to navigation. It is not known how CX neuronal diversity is generated or how developmental origin of subtypes relates to function. We mapped the developmental origin of four key CX subtypes and found that neurons with similar origin have similar axon/dendrite targeting. Moreover, we found that the temporal transcription factor (TTF) Eyeless/Pax6 regulates the development of two recurrently-connected CX subtypes: Eyeless loss simultaneously produces ectopic P-EN neurons with normal axon/dendrite projections, and reduces the number of E-PG neurons. Furthermore, transient loss of Eyeless during development impairs adult flies' capacity to perform celestial navigation. We conclude that neurons with similar developmental origin have similar connectivity, that Eyeless maintains equal E-PG and P-EN neuron number, and that Eyeless is required for the development of circuits that control adult navigation.

Classifying GABAergic interneurons with semi-supervised projected model-based clustering.

OBJECTIVES: A recently introduced pragmatic scheme promises to be a useful catalog of interneuron names. We sought to automatically classify digitally reconstructed interneuronal morphologies according to this scheme. Simultaneously, we sought to discover possible subtypes of these types that might emerge during automatic classification (clustering). We also investigated which morphometric properties were most relevant for this classification. MATERIALS AND METHODS: A set of 118 digitally reconstructed interneuronal morphologies classified into the common basket (CB), horse-tail (HT), large basket (LB), and Martinotti (MA) interneuron types by 42 of the world's leading neuroscientists, quantified by five simple morphometric properties of the axon and four of the dendrites. We labeled each neuron with the type most commonly assigned to it by the experts. We then removed this class information for each type separately, and applied semi-supervised clustering to those cells (keeping the others' cluster membership fixed), to assess separation from other types and look for the formation of new groups (subtypes). We performed this same experiment unlabeling the cells of two types at a time, and of half the cells of a single type at a time. The clustering model is a finite mixture of Gaussians which we adapted for the estimation of local (per-cluster) feature relevance. We performed the described experiments on three different subsets of the data, formed according to how many experts agreed on type membership: at least 18 experts (the full data set), at least 21 (73 neurons), and at least 26 (47 neurons). RESULTS: Interneurons with more reliable type labels were classified more accurately. We classified HT cells with 100% accuracy, MA cells with 73% accuracy, and CB and LB cells with 56% and 58% accuracy, respectively. We identified three subtypes of the MA type, one subtype of CB and LB types each, and no subtypes of HT (it was a single, homogeneous type). We got maximum (adapted) Silhouette width and ARI values of 1, 0.83, 0.79, and 0.42, when unlabeling the HT, CB, LB, and MA types, respectively, confirming the quality of the formed cluster solutions. The subtypes identified when unlabeling a single type also emerged when unlabeling two types at a time, confirming their validity. Axonal morphometric properties were more relevant that dendritic ones, with the axonal polar histogram length in the [π, 2π) angle interval being particularly useful. CONCLUSIONS: The applied semi-supervised clustering method can accurately discriminate among CB, HT, LB, and MA interneuron types while discovering potential subtypes, and is therefore useful for neuronal classification. The discovery of potential subtypes suggests that some of these types are more heterogeneous that previously thought. Finally, axonal variables seem to be more relevant than dendritic ones for distinguishing among the CB, HT, LB, and MA interneuron types.

The morphology and classification of α ganglion cells in the rat retinae: a fractal analysis study.

Rat retinal ganglion cells have been proposed to consist of a varying number of subtypes. Dendritic morphology is an essential aspect of classification and a necessary step toward understanding structure-function relationships of retinal ganglion cells. This study aimed at using a heuristic classification procedure in combination with the box-counting analysis to classify the alpha ganglion cells in the rat retinae based on the dendritic branching pattern and to investigate morphological changes with retinal eccentricity. The cells could be divided into two groups: cells with simple dendritic pattern (box dimension lower than 1.390) and cells with complex dendritic pattern (box dimension higher than 1.390) according to their dendritic branching pattern complexity. Both were further divided into two subtypes due to the stratification within the inner plexiform layer. In the present study we have shown that the alpha rat RCGs can be classified further by their dendritic branching complexity and thus extend those of previous reports that fractal analysis can be successfully used in neuronal classification, particularly that the fractal dimension represents a robust and sensitive tool for the classification of retinal ganglion cells. A hypothesis of possible functional significance of our classification scheme is also discussed.

A universal property of axonal and dendritic arbors.

Axonal and dendritic arbors can be characterized statistically by their spatial density function, a function that specifies the probability of finding a branch of a particular arbor at each point in a neural circuit. Based on an analysis of over a thousand arbors from many neuron types in various species, we have discovered an unexpected simplicity in arbor structure: all of the arbors we have examined, both axonal and dendritic, can be described by a Gaussian density function truncated at about two standard deviations. Because all arbors are characterized by density functions with this single functional form, only four parameters are required to specify an arbor's size and shape: the total length of its branches and the standard deviations of the Gaussian in three orthogonal directions. This simplicity in arbor structure can have implications for the developmental wiring of neural circuits.

Increased proximal bifurcation of CA1 pyramidal apical dendrites in sema3A mutant mice.

Semaphorin-3A (Sema3A) is an attractive guidance molecule for cortical apical dendrites. To elucidate the role of Sema3A in hippocampal dendritic formation, we examined the Sema3A expression pattern in the perinatal hippocampal formation and analyzed hippocampal dendrites of the brains from young adult sema3A mutant mice. Sema3A protein was predominantly expressed in the hippocampal plate and the inner marginal zone at the initial period of apical dendritic growth. Neuropilin-1 and plexin-A, the receptor components for Sema3A, were also localized in the same regions. The Golgi impregnation method revealed that in wildtype mice more than 90% of hippocampal CA1 pyramidal neurons extended a single trunk or apical trunks bifurcated in stratum radiatum. Seven percent of the pyramidal neurons showed proximal bifurcation of apical trunks in stratum pyramidale or at the border of the stratum pyramidale and stratum radiatum. In sema3A mutant mice, proximally bifurcated apical dendrites were increased to 32%, while the single apical dendritic pyramidal neurons were decreased. We designate this phenotype in sema3A mutant mice as "proximal bifurcation." In the dissociated culture system, approximately half of the hippocampal neurons from wildtype mice resembled pyramidal neurons, which possess a long, thick, and tapered dendrite. In contrast, only 30% of the neurons from sema3A mutants exhibited pyramidal-like morphology. Proximal bifurcation of CA1 pyramidal neurons was also increased in the mutant mice of p35, an activator of cyclin-dependent kinase 5 (Cdk5). Thus, Sema3A may facilitate the initial growth of CA1 apical dendrites via the activation of p35/Cdk5, which may in turn signal hippocampal development.

The morphological identity of insect dendrites.

Dendrite morphology, a neuron's anatomical fingerprint, is a neuroscientist's asset in unveiling organizational principles in the brain. However, the genetic program encoding the morphological identity of a single dendrite remains a mystery. In order to obtain a formal understanding of dendritic branching, we studied distributions of morphological parameters in a group of four individually identifiable neurons of the fly visual system. We found that parameters relating to the branching topology were similar throughout all cells. Only parameters relating to the area covered by the dendrite were cell type specific. With these areas, artificial dendrites were grown based on optimization principles minimizing the amount of wiring and maximizing synaptic democracy. Although the same branching rule was used for all cells, this yielded dendritic structures virtually indistinguishable from their real counterparts. From these principles we derived a fully-automated model-based neuron reconstruction procedure validating the artificial branching rule. In conclusion, we suggest that the genetic program implementing neuronal branching could be constant in all cells whereas the one responsible for the dendrite spanning field should be cell specific.

Plasma hormonal profiles and dendritic spine density and morphology in the hippocampal CA1 stratum radiatum, evidenced by light microscopy, of virgin and postpartum female rats.

Successful reproduction requires that changes in plasma follicle-stimulating hormone (FSH), luteinizing hormone (LH), prolactin (PRL), oxytocin (OT), estrogen (E(2)) and progesterone (P(4)) occur together with the display of maternal behaviors. Ovarian steroids and environmental stimuli can affect the dendritic spines in the rat hippocampus. Here, studying Wistar rats, it is described: (a) the sequential and concomitant changes in the hormonal profile of females at postpartum days (PP) 4, 8, 12, 16, 20 and 24, comparing to estrous cycle referential values; (b) the dendritic spine density in the stratum radiatum of CA1 (CA1-SR) Golgi-impregnated neurons in virgin females across the estrous cycle and in multiparous age-matched ones; and (c) the proportion of different types of spines in the CA1-SR of virgin and postpartum females, both in diestrus. Plasma levels of gonadotrophins and ovarian hormones remained low along PP while LH increased and PRL decreased near the end of the lactating period. The lowest dendritic spine density was found in virgin females in estrus when compared to diestrus and proestrus phases or to postpartum females in diestrus (p<0.03). Other comparisons among groups were not statistically significant (p>0.4). There were no differences in the proportions of the different spine types in nulliparous and postpartum females (p>0.2). Results suggest that medium layer CA1-SR spines undergo rapid modifications in Wistar females across the estrous cycle (not quite comparable to Sprague-Dawley data or to hormonal substitutive therapy following ovariectomy), but persistent effects of motherhood on dendritic spine density and morphology were not found in this area.

Dendritic thickness: a morphometric parameter to classify mouse retinal ganglion cells

To study the dendritic morphology of retinal ganglion cells in wild-type mice we intracellularly injected these cells with Lucifer yellow in an in vitro preparation of the retina. Subsequently, quantified values of dendritic thickness, number of branching points and level of stratification of 73 Lucifer yellow-filled ganglion cells were analyzed by statistical methods, resulting in a classification into 9 groups. The variables dendritic thickness, number of branching points per cell and level of stratification were independent of each other. Number of branching points and level of stratification were independent of eccentricity, whereas dendritic thickness was positively dependent (r = 0.37) on it. The frequency distribution of dendritic thickness tended to be multimodal, indicating the presence of at least two cell populations composed of neurons with dendritic diameters either smaller or larger than 1.8 æm ("thin" or "thick" dendrites, respectively). Three cells (4.5 percent) were bistratified, having thick dendrites, and the others (95.5 percent) were monostratified. Using k-means cluster analysis, monostratified cells with either thin or thick dendrites were further subdivided according to level of stratification and number of branching points: cells with thin dendrites were divided into 2 groups with outer stratification (0-40 percent) and 2 groups with inner (50-100 percent) stratification, whereas cells with thick dendrites were divided into one group with outer and 3 groups with inner stratification. We postulate, that one group of cells with thin dendrites resembles cat ß-cells, whereas one group of cells with thick dendrites includes cells that resemble cat a-cells.

Dendritic thickness: a morphometric parameter to classify mouse retinal ganglion cells.

To study the dendritic morphology of retinal ganglion cells in wild-type mice we intracellularly injected these cells with Lucifer yellow in an in vitro preparation of the retina. Subsequently, quantified values of dendritic thickness, number of branching points and level of stratification of 73 Lucifer yellow-filled ganglion cells were analyzed by statistical methods, resulting in a classification into 9 groups. The variables dendritic thickness, number of branching points per cell and level of stratification were independent of each other. Number of branching points and level of stratification were independent of eccentricity, whereas dendritic thickness was positively dependent (r = 0.37) on it. The frequency distribution of dendritic thickness tended to be multimodal, indicating the presence of at least two cell populations composed of neurons with dendritic diameters either smaller or larger than 1.8 microm ("thin" or "thick" dendrites, respectively). Three cells (4.5%) were bistratified, having thick dendrites, and the others (95.5%) were monostratified. Using k-means cluster analysis, monostratified cells with either thin or thick dendrites were further subdivided according to level of stratification and number of branching points: cells with thin dendrites were divided into 2 groups with outer stratification (0-40%) and 2 groups with inner (50-100%) stratification, whereas cells with thick dendrites were divided into one group with outer and 3 groups with inner stratification. We postulate, that one group of cells with thin dendrites resembles cat beta-cells, whereas one group of cells with thick dendrites includes cells that resemble cat alpha-cells.

Effect of hypoxia on the morphology of mouse striatal neurons.

Symptoms of high altitude sickness including headache and neuropsychological dysfunction are thought to result from prolonged exposure to hypoxia. In order to explain how the brain adapts to lower oxygen pressure at high altitude, CD1 mice were exposed to 3 weeks of hypobaric hypoxic conditions. Analyses of the neuronal morphology of striatal medium spiny neurons (MSNs) revealed a significant decrease in dendritic length, yet no change in dendritic volume, in hypoxic mice relative to normoxic mice. Vascular data indicated an increase in blood vessel area in the striatum of mice exposed to prolonged hypoxia. A mouse model of high altitude exposure may assist in elucidating the mechanisms of cerebral adaptation to high altitudes in humans, and therefore aid in developing successful prevention techniques and treatment of problems associated with high altitude disease.

Connections of diffuse bipolar cells in primate retina are biased against S-cones.

In mammalian retina, each diffuse bipolar type stratifies in a distinct layer of the inner plexiform layer. Thus, different types of bipolar cells provide output to distinct visual pathways. Here, the question of whether diffuse bipolar cell types differ with respect to their contacts with short wavelength-sensitive (S-) cones was investigated in the retinas of a New World monkey, Callithrix jacchus, and an Old World monkey, Macaca fascicularis. Subpopulations of OFF bipolar cells were labeled with antibodies to the glutamate transporter Glt-1 and ON bipolar cells were labeled with antibodies to the alpha subunit of the Go protein (Goalpha). Two types of diffuse ON bipolar cells, DB4 and DB6, were identified with antibodies to protein kinase Calpha and CD15, respectively. Cone pedicles were labeled either with peanut agglutinin coupled to fluorescein or with antibodies to the ribbon protein, C-terminus binding protein 2. We found that immunoreactivity for Glt-1 (OFF bipolar cells) is reduced at S-cones in comparison to medium/long wavelength-sensitive (M/L-) cones. Immunoreactivity for Goalpha (ON bipolar cells) is comparable at all cone types. Nearly all M/L-cone pedicles contact the diffuse ON bipolar types DB4 and DB6, but only between 60% and 75% of the S-cone pedicles make contact. Furthermore, the number of dendritic tips of DB4 and DB6 cells at S-cone pedicles is lower than that at M/L-cone pedicles. These results suggest that there is a bias in the S-cone connectivity of diffuse bipolar cells.

Age-dependent glutamate induction of synaptic plasticity in cultured hippocampal neurons.

A common denominator for the induction of morphological and functional plasticity in cultured hippocampal neurons involves the activation of excitatory synapses. We now demonstrate massive morphological plasticity in mature cultured hippocampal neurons caused by a brief exposure to glutamate. This plasticity involves a slow, 70%-80% increase in spine cross-section area associated with a significant reduction in the width of dendrites. These changes are age dependent and expressed only in cells >18 d in vitro (DIV). Activation of both NMDARs and AMPARs as well as a sustained rise of internal calcium levels are necessary for induction of this plasticity. On the other hand, blockade of network activity or mGluRs does not abolish the observed morphological plasticity. Electrophysiologically, a brief exposure to glutamate induces an increase in the magnitude of EPSCs evoked between pairs of neurons, as well as in mEPSC frequency and amplitude, in mature but not young cultures. These results demonstrate an age-dependent, rapid and robust morphological and functional change in cultured central neurons that may contribute to their ability to express long term synaptic plasticity.

Different dendrite and dendritic spine alterations in basal and apical arbors in mutant human amyloid precursor protein transgenic mice.

The extracellular deposition of amyloid-beta peptide (Abeta) in brain parenchyma is one of the characteristic features of Alzheimer's disease and is suggested to induce reactive and degenerative changes in neuronal cell bodies, axons and dendritic processes. In particular, within and in close proximity to amyloid plaques, distinctive morphological alterations have been observed, including changes in neurite trajectory and decreases in dendritic diameter and in spine density. Apart from these plaque-associated focal aberrations, little is known regarding modifications of the global dendritic morphology including the detailed and comparative quantitative analysis of apical and basal arbors. The objective of the present study was to investigate the effects of amyloid plaque deposition and elevated soluble Abeta on neuronal morphology in mutant human amyloid precursor protein (hAPP) transgenic mice (line Tg2576; [K. Hsiao, P. Chapman, S. Nilsen, C. Eckman, Y. Harigaya, S. Younkin, F. Yang, G. Cole, Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice, Science 274 (1996) 99-102]). Retrogradelly labeled callosal-projecting pyramidal cells in the primary somatosensory cortex were three-dimensionally analyzed. Although basal dendrites remained unaffected, analysis of apical trees revealed a number of unambiguous morphological changes. Thus, in TG2576 mice, the apical arbors were shortened in total length and less branched. Furthermore, the diameter of proximal dendritic segments was increased whereas that of distal segments was reduced. Analysis of spine numbers and distribution on basal and apical trees demonstrated a significant reduction in spine densities along the whole course of dendrites. The findings suggest that Abeta-related pathology induces morphological aberrations in basal and apical arbors to different degrees which are unrelated to direct plaque-associated changes.

Development of layer-specific axonal arborizations in mouse primary somatosensory cortex.

In the developing neocortex, pyramidal neurons use molecular cues to form axonal arbors selectively in the correct layers. Despite the utility of mice for molecular and genetic studies, little work has been done on the development of layer-specific axonal arborizations of pyramidal neurons in mice. We intracellularly labeled and reconstructed the axons of layer 2/3 and layer 5 pyramidal neurons in slices of primary somatosensory cortex from C57Bl6 mice on postnatal days 7-21. For all neurons studied, the development of the axonal arborizations in mice follows a pattern similar to that seen in other species; laminar specificity of the earliest axonal branches is similar to that of mature animals. At P7, pyramidal neurons are very simple, having only a main descending axon and few primary branches. Between P7 and P10, there is a large increase in the total number of axonal branches, and axons continue to increase in complexity and total length from P10 to P21. Unlike observations in ferrets, cats, and monkeys, two types of layer 2/3 pyramidal neurons are present in both mature and developing mice; cells in superficial layer 2/3 lack axonal arbors in layer 4, and cells close to the layer 4 border have substantial axonal arbors within layer 4. We also describe axonal and dendritic arborization patterns of three pyramidal cell types in layer 5. The axons of tall-tufted layer 5 pyramidal neurons arborize almost exclusively within deep layers while tall-simple, and short layer 5 pyramidal neurons also project axons to superficial layers.

Control of dendritic diversity.

The dendritic trees of different neuronal types display an astonishing diversity in structure and function. How this diversity is generated remains incompletely understood. However, recent studies have revealed some of the underlying mechanisms by which intrinsic programs of cell-type specification and extrinsic factors exert their effects on the dendritic cytoskeleton to regulate patterns of growth and branching.

Mitochondria form a filamentous reticular network in hippocampal dendrites but are present as discrete bodies in axons: a three-dimensional ultrastructural study.

The fine structure of mitochondria and smooth endoplasmic reticulum (SER) was studied via electron microscopy in dendritic and axonal neuronal segments of hippocampal areas CA1, CA3, and dentate gyrus (DG) of both ground squirrels in normothermic and hibernating conditions, and rats. Ultrathin serial sections of approximately 60 nm (up to 150 per series) were taken and three-dimensional (3D) reconstructions made of dendritic segments, up to 36 microm in length. Mitochondria were demonstrated to be present in filamentous form in every dendrite examined, in each of the hippocampal regions studied, whether in rat or ground squirrel. In addition, apparent continuity between the outer mitochondrial membrane and that of SER was observed by 3D reconstructions of very ultrathin (20 nm) serial sections prepared from dendritic segments. It is believed that SER penetrate into the heads of thin and mushroom spines but mitochondria do not enter the heads of these types of spines in dentate gyrus or CA1 of either rat or ground squirrel. However, in CA3 we have shown here that mitochondria penetrate into the base of the large thorny excrescences. Mushroom dendritic spines (but not thin spines) contained puncta adherentia, formed between pre- and postsynaptic membranes. In contrast to dendrites, the mitochondrial population of axonal processes in the same hippocampal regions were found only in the form of discrete bodies no more than 3 microm in length. The issue of the likely function of this network in dendrites and its potential role in calcium movement is discussed.

Local diameter fully constrains dendritic size in basal but not apical trees of CA1 pyramidal neurons.

Computational modeling of dendritic morphology is a powerful tool for quantitatively describing complex geometrical relationships, uncovering principles of dendritic development, and synthesizing virtual neurons to systematically investigate cellular biophysics and network dynamics. A feature common to many morphological models is a dependence of the branching probability on local diameter. Previous models of this type have been able to recreate a wide variety of dendritic morphologies. However, these diameter-dependent models have so far failed to properly constrain branching when applied to hippocampal CA1 pyramidal cells, leading to explosive growth. Here we present a simple modification of this basic approach, in which all parameter sampling, not just bifurcation probability, depends on branch diameter. This added constraint prevents explosive growth in both apical and basal trees of simulated CA1 neurons, yielding arborizations with average numbers and patterns of bifurcations extremely close to those observed in real cells. However, simulated apical trees are much more varied in size than the corresponding real dendrites. We show that, in this model, the excessive variability of simulated trees is a direct consequence of the natural variability of diameter changes at and between bifurcations observed in apical, but not basal, dendrites. Conversely, some aspects of branch distribution were better matched by virtual apical trees than by virtual basal trees. Dendritic morphometrics related to spatial position, such as path distance from the soma or branch order, may be necessary to fully constrain CA1 apical tree size and basal branching pattern.

Classification of HIV-1-mediated neuronal dendritic and synaptic damage using multiple criteria linear programming.

The ability to identify neuronal damage in the dendritic arbor during HIV-1-associated dementia (HAD) is crucial for designing specific therapies for the treatment of HAD. To study this process, we utilized a computer-based image analysis method to quantitatively assess HIV-1 viral protein gp120 and glutamate-mediated individual neuronal damage in cultured cortical neurons. Changes in the number of neurites, arbors, branch nodes, cell body area, and average arbor lengths were determined and a database was formed (http://dm.ist.unomaha. edu/database.htm). We further proposed a two-class model of multiple criteria linear programming (MCLP) to classify such HIV-1-mediated neuronal dendritic and synaptic damages. Given certain classes, including treatments with brain-derived neurotrophic factor (BDNF), glutamate, gp120 or non-treatment controls from our in vitro experimental systems, we used the two-class MCLP model to determine the data patterns between classes in order to gain insight about neuronal dendritic damages. This knowledge can be applied in principle to the design and study of specific therapies for the prevention or reversal of neuronal damage associated with HAD. Finally, the MCLP method was compared with a well-known artificial neural network algorithm to test for the relative potential of different data mining applications in HAD research.

Class I and class II ganglion cells of rabbit retina: quantitative analysis of dendritic branching patterns.

Class I and class II ganglion cells, distinguished from one another in a companion paper, were analyzed in regard to their dendritic branching patterns by determination of: 1) mean "branching density" (BD), 2) "radial branching frequency" (RBF), and 3) "branch length distributions" (BLDs; Famiglietti [ 1992a] J Comp Neurol 324:295-321). Branching density of class II cells exceeded that of class I cells by a factor of two, when compared at the same retinal location, but declined with increasing distance from the visual streak (dvs). A one-bin difference in RBF between class I and class II cells was not statistically significant. BLDs are scatter-plots of individual preterminal and terminal branch lengths versus the distances of their origins from the soma. The parameters mp and mt, the slopes of regression lines fitted to the preterminal and terminal BLDs, respectively, were determined; mp, but not mt, was relatively independent of dvs, and was used empirically to determine a boundary value, mp = +5.0, separating "radiate" from "tufted" dendritic branching. Similarity of class I (mp = +8.6 +/- 4.6) and class II (mp = +1.4 +/- 5.2) cells did not allow a statistically significant separation of the two classes, based upon branching pattern alone; however, mp together with mt easily separated class I cells (mt = -17.8 +/- 10.0) and particularly "tufted" class II cells (mt = -16 +/- 9.3) from "tufted" class III.1 ganglion cells (Famiglietti, 1992a), with their qualitatively different, more regular branching (mp = +2.1 +/- 0.85; mt = +0.65 +/- 4.9).