Fig 1.
Scheme for quantifying bias in the pattern of dendritic growth with respect to the location of astroglia.
A) To measure the radial extent of the dendritic arbor, concentric Sholl rings were placed around the neuronal cell body and repeated at 10 μm intervals. To provide an index of the potential growth zone occupied by glia, the 360° Sholl area was divided into a sector occupied by glia and a sector absent of glia indicated by the angle shown in yellow. B) The fraction of the Sholl area occupied by glia was measured by superimposing a 10 μm x 10 μm grid onto the image and counting intersections. Scale bar = 20 μm.
Fig 2.
Plating neurons and astroglia together at low density produced varied degrees of interaction between the two cell types.
Cells were stained to reveal the dendritic arbor of neurons (MAP2, red) and extent of interaction with astroglia. Phalloidin (green) binds filamentous actin, is present throughout astroglia, as well as neurons, and is notably concentrated in growth cones. Combining these two fluorescent channels distinguished the boundaries and overlap between the two cell types (as in A, B, and C). Comparing MAP2 staining (D, E, F) against neuron-specific tubulin (G, H, I) allowed dendrites to be distinguished from axons and further separated neurons from astroglia. Each column shows representative neurons in differing degrees of interaction with astroglia: neurons with dendritic arbors growing without physically contacting astroglia (A, D, G); arbors in partial contact with astroglia (B, E, H); and arbors in full contact with astroglia (C, F, I). Scale bar = 20 μm.
Fig 3.
Contact with astroglia produced asymmetric dendritic arbors through local restriction of growth.
A) The total number of Sholl ring intersections for dendrites growing completely on (On), partially on (Partial), or not in contact (Off), showed that dendritic growth was inversely proportional to the extent of physical contact with astroglia. Bars with different letters were significantly different from each other. B) For neurons in partial contact, the number of dendrites intersecting Sholl rings was significantly reduced for the portion of the arbor in contact with astroglia (On Astroglial Territory) compared with the portion that was not (Off Astroglial Territory). C) Scatter plot showing localized limited dendritic growth based on physical contact with astroglia. If the presence of glia had no effect on growth, the probability of dendrites falling on glia would simply be a function of the proportion of the field occupied by glia (“neutral,” red line). The slope of the line of best fit to data (“observed,” blue line) falls below the neutral line, indicating that dendritic growth was restricted when it entered the territory of an astroglial cell. Neurons were quantified at 4 DIV. Data are reported as mean +/- standard error (SE).
Fig 4.
Dendrites and filopodia orient toward astroglia.
A) Representative image of neuronal dendrites at 3 DIV (green, MAP2 staining) plated with astroglia (red/blue, phalloidin/GFAP), scale bar = 50 μm. B) Quantification. Dendritic arbors of neurons at 3, 5 and 8 DIV were divided into zones based on the relative position of astroglia. The dendrites facing vs. not facing astroglia were counted using a pizza-wedge analysis, as described in Methods (see Fig 1). To correct for varying amounts of territory occupied by astroglia in the field, the number of dendrites per 10° is reported. C) Representative SEM image of a neuron and neighboring astroglia at 3 DIV, false-colored to highlight processes and filopodia of both cell types. Scale bar = 10 μm. D) Mean number of filopodia per micron of neuronal process that oriented toward vs. away from nearby astroglia. For panels B and D, data are reported as means +/- SE; each DIV was analyzed independently using a paired, two tailed test as described in results. ****, p < 0.0001; **, p < 0.01.
Fig 5.
Astroglia and dendrites interact via filopodial contact.
Phalloidin staining of filamentous actin reveals extensive filopodia on both neurons (shown at 5 DIV) and astroglia. A) High magnification image shows numerous zones of astroglial filopodia (arrows), resulting in frequent contact between filopodia of both cell types. The combined image shows filamentous actin (phalloidin staining, green), MAP2 staining (pink), and neuron-specific tubulin (blue), arrowhead points to the axon, scale bar = 20 μm. B, C) Images acquired using lower magnification show patches of astroglia with numerous, long filopodia (arrows, orange) extending toward neurons stained with neuron-specific tubulin (teal), to reveal all processes (including both axons and dendrites). Scale bar for B, C = 50 μm.
Fig 6.
Soluble astrocytic factors modulate dendritic growth as well as synapse formation.
A—D) Neurons at 5 DIV, immunostained for MAP2 (green) and synapsin I (red) have few presynaptic contacts when placed under glial deprivation, but more elaborate dendritic arbors. Scale bar = 15 μm. E) The total number of branches per neuron was counted and averaged per experimental condition. Significant differences in the total number of branches between glial deprived and +TSP conditions suggest additional factors beyond TSP are likely involved. F) The total number of presynaptic contacts per neuron was counted and averaged per experimental condition. For both Panels E and F, the sample size (N) of each condition at 4 DIV was 41, 60, 39, 42, from 2 separate culture preparations; at 5 DIV, 114, 112, 107, 110, from 3 separate culture preparations; at 6 DIV, 58, 50, 49, 43, from 2 separate culture preparations, respectively. Data are reported as mean +/- SE. For each age group, bars with different letters are significantly different from each other at p < 0.05 as determined by a Kruskal-Wallis test combined with Dunn’s Multiple Comparisons Test. See results for additional details of statistics.
Fig 7.
Evidence that paracrine effects act generally throughout the dendritic arbor.
A) Nomenclature for order of branches. Primary branches emerge from the cell body, higher order branches are the sum of all branches above primary. B) Number of both primary, and C) higher order branches were increased significantly in neurons deprived of astroglia, with significant differences detected at 4 DIV (24 hours after removal of astroglia from the culture). Bars with different letters are significantly different from each other at p < 0.05. D, E) Sholl ring analyses corroborated branch analyses, with mean number of intersections for neurons under glial deprivation (glia-dep and glia-dep + TSP, both numerically higher at 4 DIV (D), and significantly higher at 6 DIV (E). When taken with data shown in Fig 3, these data suggest that astroglia produce general as well as contact-mediated effects on dendritic growth, potentially involving different factors. Data are reported as mean +/- SE, N is identical to that reported in Fig 6.
Fig 8.
Local and physical interactions between dendrites and astroglia can alter the shape and growth of the dendritic arbor.
Dendrites not in contact with astroglia are significantly longer than those in contact. Dendrites, and associated filopodia that are in close proximity to astroglia orient toward the glial cell which, upon contact, restricts growth.