Previous reports have shown that the repeated restraint procedure (6 hours/day, 21 days) used in the present study activates the stress-responsive HPA axis, and produces modest, but significant increases in adrenal weights compared with unstressed rats (Watanabe et al., 1992
; Magariños and McEwen, 1995
). Although we did not assess either of these indices of stress activation, we did verify that 21 days of repeated daily restraint stress resulted in a15% reduction in body weight compared to unstressed controls (P
< 0.005; see Watanabe et al., 1992
The precise location of Lucifer Yellow-filled pyramidal neurons in layers II/III of the dorsal mPFC was evaluated with reference to standard cytoarchitectonic parcellations of the region (Krettek and Price, 1977
; Vogt and Peters, 1981
). The two fields of primary interest are the dorsal subdivision of the anterior cingulate area (ACd), and prelimbic area (PL). Using the fluorescent nucleic acid counterstain (4,6-diamidino-2-phenylindole), the ACd is characterized by a sparse layer II and a broad layer V, distinguishing it from PL, which contains a more homogeneous layer V comprised of larger nuclei. Layer I is nearly devoid of neurons, in contrast to the cell-dense layers II/III, permitting clear establishment of laminar borders. provides an example of the locations for each of the Lucifer Yellow-filled pyramidal neurons in the dorsal mPFC that were subject to the dendritic tracing and spine morphometric analysis.
Anatomical localization of Lucifer Yellow-filled layers II/III pyramidal neurons in mPFC for subregional analysis
In a previous report repeated stress was shown to induce spine loss in apical dendrites of dorsal mPFC pyramidal neurons (Radley et al., 2006b
). While this earlier study reported spine density estimates from confocal stacks of three-dimensionally reconstructed dendritic segments (25–50 µm in length), spine densities were manually determined by counting the number of spines for each length of dendritic segment and expressing values in terms of spines/µm. Changes in mean spine density as a function of treatment status were readily determined from the Rayburst-based automated system, as an initial step in the algorithm involves the extraction of the medial axis of the dendritic segment and recognition of spines using a dynamically adjusting local segmentation technique. Fifty branches from the data set were randomly subjected to counting using both manual and Rayburst-based automated approaches. The Rayburst-based automated spine analysis system yielded spine densities that were 87% similar to the manual counting approach.
Repeated restraint stress produced an overall 11% decrease in spine density throughout apical and basal dendrites of dorsal mPFC pyramidal neurons in layers II/III (F(1,8) = 8.3; P < 0.05; , upper left) and an overall 12% decrease in apical dendritic spine density (P < 0.05; , middle left). There was also a main effect for dendritic spine density relative to distance from the soma (F(4,8) = 10.1; P < 0.005), and no interaction between experimental treatment and distance from the soma (F(4,8) = 1.6; P = 0.2). By location on the apical dendritic arbor, the stress-induced reduction in spine density was most pronounced at >150 µm relative to the soma (15% decrease, P < 0.05; , middle left). Although stress did not have any significant overall effect on basal dendritic spine density, a 13% decrease in spine density was present at radial distances of <75 µm (P < 0.05; , lower left).
Repeated restraint stress reduces dendritic spine density in layer II/III pyramidal neurons
The dendritic spine morphometric parameters of volume, surface area, and length were assessed for each dendritic spine using the Rayburst-based automated approach. A total of 187 dendritic segments were analyzed (92 control, 97 repeated stress), and 17,078 dendritic spines (8,091 control, 8,987 repeated stress). Based on 5 animals/group, and 5 neurons/animal, a mean of 3.7 and 3.9 dendritic segments/neuron were analyzed for the control and repeated stress groups, respectively. In unstressed controls, an average of 12 apical and 6.4 basal dendritic segments were analyzed per animal (totaling 5,383 and 2,708 spines, respectively), and in the repeatedly stressed group, an average of 12.4 apical and 7 basal dendritic segments were analyzed per animal (totaling 5,587 and 3,400 spines, respectively).
Repeated restraint stress resulted in an overall 10% decrease in mean dendritic spine volume (F(1,8) = 7.3; P < 0.05; ) and surface area (F(1,8) = 15.3; P < 0.005; ), and no effect on spine length (F(4,8) = 2.2; P = 0.2; ), in pyramidal neurons in layers II/III of the dorsal mPFC. Regarding mean spine volume, there was also a main effect for distance from the soma (F(4,8) = 3.9; P < 0.05), and no interaction (F(4,8) = 0.6; P = 0.6), whereas no effects for distance (F(4,8) = 1.3; P = 0.3) or interaction (F(4,8) = 0.6; P = 0.7) were evident for mean spine surface area.
Effects of repeated restraint stress on Rayburst volumes of dendritic spines in layers II/III pyramidal neurons in the dorsal mPFC
Effects of repeated restraint stress on Rayburst surface areas of dendritic spines in layers II/III pyramidal neurons in the dorsal mPFC
Effects of repeated restraint stress on dendritic spine lengths in layers II/III pyramidal neurons in the dorsal mPFC
Similar to the effects of stress on dendritic spine density, the reduction in mean spine volume and surface area is most prominent in the more distal aspect (>150 µm) of apical dendrites (10% decrease, P < 0.05; , ). Moreover, basal dendrites show a significant reduction in mean spine volume and surface area, focused in the proximal aspect (<75 µm) following repeated stress (, ). While 21 days of repeated restraint stress did not significantly affect overall mean spine length (F(1,8) = 2.3; P = 0.2), there was a main effect for distance from the soma (F4,8 = 4.9; P < 0.05), and no interaction between treatment and distance (F(4,8) = 1.1; P = 0.4). Repeated stress produced a smaller, albeit significant decrease in spine length (6% decrease), which was once again focused in apical dendrites at radial distances >150 µm ().
To determine whether there was an overall shift in the spine size distribution following repeated stress, or a disproportionate change in the incidence of spines of smaller volume, surface area, and length, the frequency of spines in the top and bottom quartile for the subject population as a whole (i.e., control and repeated stress groups) were analyzed. For volume and surface area, the proportions of spines on apical and basal dendrites following repeated stress were higher in the lowest quartile, and lower in the highest quartile, relative to unstressed controls (P < 0.001; ). While a similar shift in spine length distribution was also present, the effect was limited to apical dendrites (P < 0.001; ). Spines on apical dendritic segments at radial distances >150 µm from soma exhibited the most pronounced shift in frequency distribution in all three dimensions examined (, ).
Dendritic Spine Dimensions in Layers II/III Pyramidal Neurons in Dorsal mPFC
Probability distributions for spine volume and surface area for distal apical dendritic segments