University of California, Berkeley | Department of Molecular & Cell Biology

Winer Laboratory








The vascular organization of the cat inferior colliculus1


Brain neural activity depends critically on the blood supply to a given structure, and this may have functional significance and can differ between and within subdivisions.  We analyzed the vascular organization of the cat inferior colliculus (IC) to determine whether its divisions share a common pattern.  The IC consists of the central nucleus (CN), the dorsal cortex (DC), and the lateral cortex (LC), each with different roles in auditory behavior and perception.  We compared IC capillary distribution, density, and size to quantify differences between subdivisions.  Plastic-embedded material was studied from two adult cats in 1 Ám-thick semithin sections stained with toluidine blue; tissue was sampled from IC in a rostral-caudal series of sections.  The architectonic subdivisions were drawn independently for unrelated studies.


We randomly selected ten non-overlapping 200 x 200 Ám2 samples in each CN, DC, and LC.  Samples near borders were excluded as were vascular profiles with a minor–axis diameter >8 Ám.  We found significant differences in the number of capillaries between the CN and DC (p< 0.05; t-test, df: 2; n=30).  There were also significant differences in capillary number between the CN and LC (p< 0.05; t-test, df: 2; n=30).  The differences in capillary density followed similar patterns: CN and DC differed, as did CN and LC.  The average capillary lumen area was 40–60 Ám2 in each subdivision and not significantly different.


The CN has 50% more capillaries than other IC subdivisions and, presumably, a proportionally higher level of basal metabolic activity. The IC vascular organization confirms the CN border based on neuronal architecture, efferent and afferent connections, and physiological properties.  The DC and LC share a similar vascular architecture despite their functional differences.  The significantly greater CN blood supply suggests that the lemniscal auditory pathway has higher levels of metabolic activity than non-lemniscal parts of the IC.


1 Song, Y., and Winer J.A.  Vascular architecture of the cat inferior colliculus. Society for Neuroscience Abstracts, 2008, 34, in press.




Neocortical horizontal cells: distribution and morphology2


Interneurons play key roles in neocortical function.  Most interneurons are GABAergic and have dendrites with a vertical or stellate orientation.  Horizontal cells (HCs) have been identified in layers I and VI and in the white matter (WM) and have dendrites parallel to the layers.  However, they are less well documented in layers II-V.  We define neocortical and white matter HCs as having a major dendrite at each somatic lateral pole oriented within 0-10╝ parallel to the pia or layer VI-WM border, a spindle shaped cell body, and a minimum somatic length-width ratio of 3:2.  HCs (n=200) in layers I-VI have an average length-to-width ratio of 2.2:1 and somata ~27 mm long and ~13 mm wide.  In the rat, NADPH-d-positive HCs are found in every layer and each area studied.  Transverse and horizontal sections show a non-uniform HC neocortical distribution, with the highest concentration in the sensory areas: 27% of the total sample in somatic sensory cortex, 21% in auditory cortex, 15% in visual cortex, 21% in insular cortex, and <15% total in the entorhinal, cingulated, and motor cortices.  The laminar distribution of NADPH-d-positive HCs is 2% in layer I; 11% in layer II; 21% in layer III; 4% in layer IV; 24% in layer V; and 38% in layer VI.  Sections were not normalized for laminar thickness.  We confirm prior NADPHD-d studies which found 70% colocalization with somatostatin and 25% colocalization with GABA.


Our findings suggest a species difference in the distribution and morphology of NADPH-d-positive HCs in the rat and cat.  In the rat, ~15% of DNADPH-d-positive neurons are HCs and only 20% are in the Wm.  In contrast, almost all cat NADPH-d-positive neurons are HCs in layer VI or the WM, with most at the layer VI-WM order.  These WM HCs decrease with distance from the gray matter and are concentrated near sulci and gyri.  HCs in the cat have thin lateral dendrites of uniform thickness, whereas rat HCs have thicker lateral dendrites that taper at their poles.


Cajal-Retzius layer I cells and WM HCs have roles in neocortical development, and the HCs in layers II-VI could provide a lateral influence on pyramidal cells related to columnar organization.  HCs at the layer VI-WM border and WM HCs near sulci and gyri may establish cortical-WM boundaries and influence cortical developmental folding, respectively.  Our findings suggest multiple and diverse functional roles for HCs.


2 Shen. A., and Winer, J.A.  Neocortical horizontal cells: distribution and morphology.  Society for Neuroscience Abstracts, 2008, 34, in press.



Auditory corticocollicular horizontal cells in layer VI3


We studied layer VIb horizontal cells projecting to the rat inferior colliculus (IC). Deposits of retrograde tracers in the pericentral nuclei of the IC label small pyramidal and horizontal cells in layer VIb in auditory and adjacent cortical areas. This is in addition to the expected and much larger contribution from layer V. The horizontal cells lie at the base of layer VI, just above the white matter. They are fewer and typically smaller than their layer V counterparts but are present consistently and are labeled by with WGA-HRP, CT?, CT?G, HRP, and BDA. Deposits that involve the IC lateral or caudal cortices label more layer VI cells than injections largely restricted to the central nucleus.

The layer VI labeling begins as far caudally as IA-zero (cells in the subjacent white matter) and continues through IA+4.2 mm. Labeled cells are found in areas Ect, TeA, V2L, AuV and AuI.


Some layer VIb cells were labeled sufficiently with CT? to reveal lateral dendrites projecting laterally for more than 50 Ám. This population of horizontal cells represents a non-pyramidal corticofugal projection. Labeled somatic profiles were small (approximately 10 Ám wide by 4 Ám high), fusiform in shape, and oriented parallel to the white matter. Their average length/width ratio was 2.48:1.


In most species studied, only layer V pyramidal cells contribute to the corticotectal pathway. There are few reports of layer VI neurons projecting to the IC in rodents (rat and guinea pig) but these studies did not specify the cell types involved. In the cat, this layer VI contribution is absent. It also is absent in the squirrel monkey and the mustached bat but it has been reported in the lesser hedgehog. This suggests that the layer VI corticofugal projection is unique to rodents and basal insectivores. Moreover, these layer VI horizontal cells may be the first evidence that non-pyramidal cells participate in this projection.


3 Larue, D.T., and Winer, J.A.  Auditory corticocollicular horizontal cells in layer VI.  Society for Neuroscience Abstracts, 2008, 34, in press.




GABAergic Networks Differ in Auditory Cortex Areas4


GABAergic interneurons constitute ~15-25% of auditory cortex cells and are essential for sensory processing. The correlation between GABAergic connections and neuronal response properties in physiologically defined auditory cortical areas is not understood. We studied the GABAergic connections in the AI (primary auditory cortex) central narrowband subregion (cNB), and in a nontonotopic auditory cortex area (AII) with a retrograde tracer (WAHG) and the Ca2+ binding protein, parvalbumin (Pv), a marker for GABAergic neurons usually described physiologically as fast-spiking neurons, in 2 cats.


After mapped injections, we found: (1) Dorsal-ventral excitatory projections are clustered in AI and AII.  (2) In central AI, 99% of retrogradely labeled Pv+ neurons are in cNB.  The spread of retrogradely labeled excitatory neurons in AI is 3 mm, whereas that of Pv+ cells is only 1.5 mm.  In AII, the spread of Pv+ cells is almost equal to that of excitatory neurons and up to 2 mm.  (3) The percentage of labeled cells identified as inhibitory in AI and AII differs significantly, with 20% in AI and 15% in AII.  (4) There is a significant difference in the dispersion of layer IV labeled inhibitory and excitatory neurons in AI and AII.  In AI layer IV labeled excitatory neurons do not spread as far as those in other layers.  In AII layer IV, labeled excitatory neurons are as dispersed as those in layers II and III and many are in the dorsal or ventral patches within which Pv+ cells are observed.


The findings suggest that the connections of cat GABAergic AI and AII cells follow different anatomical plans, which may shape the different neural response properties.  This implies that different local circuit architectures in AI and AII could embody unique functional roles in sound information processing.


4 Yuan, K., Shih, J.Y., Schreiner, C.E., and Winer J.A.  GABAergic Networks Differ in Auditory Cortex Areas. Society for Neuroscience Abstracts, 2008, 34, in press.