Longitudinal in vivo imaging reveals balanced and branch-specific remodeling of mature cortical pyramidal dendritic arbors after stroke

Abstract

The manner in which fully mature peri-infarct cortical dendritic arbors remodel after stroke, and thus may possibly contribute to stroke-induced changes in cortical receptive fields, is unknown. In this study, we used longitudinal in vivo two-photon imaging to investigate the extent to which brain ischemia can trigger dendritic remodeling of pyramidal neurons in the adult mouse somatosensory cortex, and to determine the nature by which remodeling proceeds over time and space. Before the induction of stroke, dendritic arbors were relatively stable over several weeks. However, after stroke, apical dendritic arbor remodeling increased significantly (dendritic tip growth and retraction), particularly within the first 2 weeks after stroke. Despite a threefold increase in structural remodeling, the net length of arbors did not change significantly over time because dendrite extensions away from the stroke were balanced by the shortening of tips near the infarct. Therefore, fully mature cortical pyramidal neurons retain the capacity for extensive structural plasticity and remodel in a balanced and branch-specific manner.

Figure 1

Summary of imaging sessions and laminar position of each neuron. Each mouse was assigned an arbitrary name (e.g., C39) and imaged every other week. Only one neuron was imaged in each mouse, except for mouse C13. Neurons were imaged for 2 to 4 weeks before, and up to 6 weeks after stroke. As one cannot control or predetermine the exact location of the infarct border, some imaged neurons were destroyed by the stroke (C20, 33, 48; resided within the infarct core) and therefore could not be imaged after stroke induction.

Figure 2

Arrangement of cortical dendrites before and after focal stroke. (A) Surface view of the control brain and two-photon images (maximal intensity z-projection in the x–y plane) of a GFP-labeled apical dendrite named ‘C13-N2' located in the boxed area, imaged every 2 weeks. The 0-week time point corresponds to images taken 1 h before the induction of stroke. (B) Photomicrograph of the brain after stroke and time-lapse images of the same dendritic arbor is shown in panel A. (C) Side-view projection in the y–z plane showing progressive structural changes. It must be noted that tips 5 and 6, which were not well differentiated in z-projections, were indeed resolved in side-view projections at different depths. (D) Plots showing changes in dendritic tip length over time. wk, week.

Figure 3

Remodeling of dendritic arbors after stroke. (A) Surface of the brain 2 weeks after stroke e.g., as shown in B. (Panel B) Time-lapse two-photon images of dendritic arbor (maximal intensity projection in the x–y plane) situated close to the infarct border (∼145 μm), identified as ‘C39'. The growth of dendrites away from the site of infarction must be noted. (C) Postmortem identification of GFP-labeled neuron imaged in vivo. (D) Reconstructed drawings of the imaged dendrite shown from the side or the y–z axis. (E) Stroke significantly increases the length of dendrite remodeled (i.e., absolute value of extensions and retractions, average of 6 neurons with a total of 82 tips analyzed), especially in the first 2 weeks of recovery. Dashed lines show estimates of measurement error. (F) Color-coded correlation for total amount of remodeling observed for each neuron over 6 weeks recovery and their relative distance (measured from the primary apical dendrite) to the infarct border. (G) Histogram showing average length of dendrite gained or lost for each time point after stroke. (H) Total amount of dendrite gained and lost for all imaged neurons over 4 to 6 weeks of recovery is almost identical. ^*P<0.01. wk, week.

Figure 4

Examples of dendrite tip remodeling after stroke. (A) Shows a maximal intensity projection of dendrite tip that grew after stroke. (B) Maximal intensity projection showing dendritic tip retraction after stroke. wk, week.

Figure 5

Region-specific remodeling of dendrites. (A and C) Diagrams summarizing how dendrites appear before stroke and the orientation-specific remodeling that occurs after stroke. (B) Average change in dendrite length over recovery period. It must be noted that dendrites oriented toward the infarct (proximal) tend to retract, whereas those oriented away from the stroke (distal) preferentially grew. ^*P<0.05.