Retinal Microglial Cell Analysis in Rat Models

Morphology, spatial distribution and mean number of microglial cells were analyzed in the inner and outer plexiform layers (IPL, OPL), the ganglion cell layer (GCL) and the subretinal space (SS) of whole-mount retinas. Macrophages detected in the GCL during the analysis were also accounted. Four to six retinas from each animal group (TUDCA-treated, untreated and SD animals) were examined. In each retina, 12 representative regions of 0.227 mm² each were analyzed, 6 regions equidistantly arranged on the superior-inferior axis of the retina and 6 fields disposed on the temporal-nasal axis; thus sampling representative peripheral, medial and central areas of the superior, inferior, temporal and nasal quadrants of each retina. In each of the 12 regions analyzed in each retina, every one of the cell bodies labeled with immunoperoxidase in each of the 3 layers analyzed was manually traced using a camera lucida attached to the Leica DMR microscope (Leica Microsystems, Wetzlar, Germany). Each retinal layer was determined according to the vascular stratification of the retinal tissue. The images created were subsequently digitized using the image-editing software Photoshop (Adobe Systems Inc., San Jose, CA, USA). The distribution pattern of microglial cells in each retinal layer was assessed by measuring the distances to the nearest neighbors of each microglial cell using ImageJ software [39 ]. The distances to the nearest neighbors were classified in histograms, statistically analyzed and compared with a nearest neighbor analysis of a random pattern of the same density and standard deviation [39 ,40 (link)]. For nearest neighbor distance analysis we used images collected from the medial area of the retina (superior quadrant).Figure 1

Distribution and morphology of microglial cells in Sprague-Dawley (SD) (A) and P23H (B) rats. Retinal vertical sections were immunolabeled with CD11b (OX-42). Note the presence of amoeboid CD11b-positive cells in different layers of the P23H rat retina, including the subretinal space. GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer. Scale bar: 10 μm.

Figure 2

Drawing of the most representative morphology and location of microglial cells in Sprague-Dawley (SD) (A), untreated P23H (B) and tauroursodeoxycholic acid (TUDCA)-treated P23H (C) rats. Note the absence of microglial cells into the subretinal space of P23H rats treated with TUDCA. GCL, ganglion cell layer; IPL, inner plexiform layer; INL, inner nuclear layer; OPL, outer plexiform layer; ONL, outer nuclear layer; SS, subretinal space.

Figure 3

Migration of microglia into the subretinal space of untreated P23H rats. Whole-mount retina of a SD (A), untreated P23H (B) and tauroursodeoxycholic acid (TUDCA)-treated P23H (C) rat immunolabeled with CD11b (OX-42), showing the presence of amoeboid microglial cells in the subretinal space of untreated P23H rats. Scale bar: 40 μm.

Figure 4

Quantification of retinal microglial cells. (A) Average number of positively stained microglial cells quantified in whole-mount retinas from SD (n = 4; green), untreated P23H (n = 6; red) and tauroursodeoxycholic acid (TUDCA)-treated P23H rats (n = 6; grey). (B) Average number of microglial cells in the GCL, IPL, OPL and SS of the retinas analyzed in (A). *P <0.05, **P <0.01, ***P <0.001; ANOVA, Bonferroni’s test.

To analyze microglial activation in each animal group, we used cryostat vertical retinal sections immunostained with Iba1 and MHC-II. We examined three vertical sections non-consecutive per animal and we studied six animals per experimental group. All the images analyzed were collected from the central area of the retina, close to the optic nerve. In every retinal section, the total number of microglial cells expressing one or both markers (Iba1⁺/MHC-II⁻, Iba1⁻/MHC-II⁺ and Iba1⁺/MHC-II⁺) was counted using light microscopy at x63 magnification. The obtained data were referred to the length of each retinal section, measured using the ImageJ software.