Table 1.
Significance of factors in presurgical tests.
Fig 1.
Three weekly sets of behavioral experiments were performed prior to surgery to establish baseline performance for each mouse. After this, mice were randomly assigned to one of three treatment groups: GRIN Lens implant, cranial window (No Lens), or sham surgery (Anesthesia Only). No behavioral experiments were performed on the day of surgery. Post implantation experiments were run 3 days after surgery to implant a GRIN lens, install a cranial window, or conduct a sham surgery to determine the acute effects of lens implantation, craniotomy surgery, and anesthesia, respectively. Tests were conducted weekly at 7, 14, 21, and 28 days after surgery to determine the time course of effects due to chronic implantation, if any. Each mouse was tested at each of the 8 time points for all three behavioral tests (n = 5 mice for each treatment group).
Fig 2.
Comparison of performance on rotarod test.
Mean latency to fall times for each mouse were normalized to the first week of testing (Day -21). Bars indicate the standard error of the mean. Three tests were performed prior to surgery at -21, -14, and -7 days prior to surgery and then 3, 7, 14, 21 and 28 days after surgery. No significant differences were observed between treatment groups (p = 0.26), which include GRIN lens implant (GRIN Lens), cranial window (No Lens), and sham surgery (Anesthesia Only) groups (n = 5 mice per treatment group). Only the effect of time was significant (p = 0.0002). The upward trend in latency times suggests a learning effect.
Fig 3.
Comparison of performance in foot fault test.
The performance of each mouse was normalized to its Day -21 performance. Bars are the standard error of the mean for each treatment group (n = 5 mice per treatment group). Tests were conducted -21, -14, and -7 days prior to surgery and then 3, 7, 14, 21, and 28 days after surgery. Time was the only factor with a significant difference (p = 1 x 10−5). The downward trend in the number of faults suggests a learning effect. No significant difference was found between treatment groups (p = 0.78).
Fig 4.
Comparison of time to find the platform for Morris water maze test.
The performance of each mouse was normalized to its Day -21 performance. Bars denote the standard error of the mean for each treatment group (n = 5 mice per treatment group). Tests were conducted -21, -14, and -7 days prior to surgery and then 3, 7, 14, 21, and 28 days after surgery. There was no significant difference between treatment groups (p = 0.23).
Fig 5.
Comparison of swim speed in the Morris water maze test.
The performance of each mouse was normalized to its Day -21 performance. Bars are the standard error of the mean for each treatment group (n = 5 mice per treatment group). Tests were conducted -21, -14, and -7 prior to surgery and then 3, 7, 14, 21, and 28 days after surgery. There were no significant differences over time (p = 0.44) or between treatment groups (p = 0.51).
Fig 6.
In vivo images of glial cells in cortical layer 5.
Glial cells in layer 5 of the prefrontal cortex of GFAP-GFP mice imaged using implanted ultrathin GRIN lenses and multiphoton microscopy. Digital zoom images were acquired to show details of cells; the full field of view was 180 μm. A. Glial cells expressing GFP under the GFAP promoter are wrapped around a blood vessel counterstained with dextran-Texas red dye. This image is a z-projection of 3 sequential image planes with a 1-μm step size. B. This image is a z-projection of 2 image planes, which was done to average out noise and to better visualize cellular processes. C. Single plane image of glial cells in a different mouse. Scale bars denote 5 μm.
Fig 7.
Placement of lens and measurement of glial scar.
A. Coronal slice containing the track of an implanted GRIN lens is outlined with a dotted yellow line. The implant depth was 935 μm below the pial surface. B. Coronal slice stained with anti-GFAP antibody. The white dashed lines denote the mean thickness of the glial scar under the lens, which was 45.9 ± 4.2 μm (mean ± SEM, 40 areas, 3 mice). The working distance of the GRIN lens, which was 125 μm, is much longer than the thickness of the scar. Therefore, the cells within 125 μm of the lens are too close for inclusion in any in vivo images. A solid light yellow line denotes the region that can be visualized with the GRIN lens and multiphoton microscopy (imaging region). The GFAP+ cells in the imaging region were relatively few and exhibited only weak immunoreactivity compared with the scar. The inset is an enlargement of this area (inset was brightened and contrast was increased to show relatively faint staining). C. An area in the ventrolateral cortex of the contralateral hemisphere was used as a negative control. A solid light yellow line denotes the region which is at the same depth and size as the imaging region and in the same orientation with respect to the pial surface. The inset is an enlargement of this area (inset was brightened and contrast was increased to show relatively faint staining). The number of GFAP+ cells were not significantly different for the imaged region and the negative control region (p = 0.36). Scale bars are 200 μm for A and 100 μm for B and C.
Fig 8.
Assessment of microglia morphology.
A. Several microglial cells had de-ramified or amoeboid morphology, typical of activated microglia, in the non-imaging area up to 125 μm from the end of the lens (cell identified by upper arrow is enlarged in C) with 85.7% of these occurring in the same area as the glial scar (n = 3 mice). The dark area along the upper edge of the image is the lens track. A blue dotted line is placed 46 μm from the bottom of the lens track which represents the mean thickness of the glial scar. Only a few activated microglia (cell identified by lower arrow is enlarged in D) were observed in the imaging areas, which is illustrated using a yellow outline above. Most microglia had a ramified morphology which is indicative of the resting state (arrow in E points to ramified process). B. Similar to the region under the lens, the control areas had very few microglia with activated morphology (cells identified by arrows are enlarged in F and G). Scale bars are 50 μm.