Local microglial activation induced and labeled in the retina in a novel subretinal hemorrhage mouse model

Abstract

Subretinal hemorrhage (SRH) is caused by the accumulation of blood between the neurosensory retina and the retinal pigment epithelium or between the retinal pigment epithelium and the choroid. SRH often arises from age-related macular degeneration, traumas, and may occur spontaneously caused by other diseases like hypertension and diabetes. Here, we developed a novel technique - co-injection of blood and a dye-coupled tracer protein, Cholera toxin subunit B (CtB) - to better localize and understand the disease and how it can cause microglial activation, inflammation, and partial vision loss. Our results show that microglia are activated in the inner retinal layers in zones adjacent to blood injection. In contrast, the non-affected zone of the injected eye showed no microglial activation. For the first time, we used phosphate-buffered saline (PBS) injections as a control to assess the specific effects of injected blood. The results demonstrated that blood induced a markedly stronger activation response in the surrounding tissue, whereas PBS elicited a comparatively milder effect. PBS did cause microglial activation, but it was largely confined to the injection site and adjacent regions, and to a lesser extent than that observed with blood. We also observed microglial activation in the inner retina, along with the emergence of microglia and macrophages in the retinal pigment epithelium. Using advanced imaging techniques, we were able to better localize the affected area which comprises not only the immediate retinal area over the blood clot but the neighboring regions as well. These findings will provide the basis for novel therapeutic interventions targeting neuroinflammation in the retina after subretinal hemorrhage and other diseases affecting the eye.

Keywords: Bleeding; Blood; Cholera toxin; Experimental ophthalmology; Eye; Inflammation; Retinal pigment epithelium.

Fig. 1

Subretinal hemorrhage. (A) Schematic diagram represents the human eye and (B) the retinal layers with the respective cell types. Dashed red boxes show microglia layers in the healthy retina. Abbr. Astrocyte (AS), Ganglion Cell (GC), Microglia (MG), Amacrine Cell (Am), Bipolar Cell (BC), Horizontal Cell (HC), Cone (C), Rod (R). Neurofilament Layer (NFL), Ganglion Cell Layer (GCL), Inner Plexiform Layer (IPL), Inner Nuclear Layer (INL), Outer Plexiform Layer (OPL), Outer Nuclear Layer (ONL), Outer Segment Layer (OSL), Retinal Pigment Epithelium (RPE), Subretinal Hemorrhage (SRH). Dashed red lines indicate the Superficial Layer (SL) and Deep Layer (DL) of microglia.

Fig. 2

Subretinal injection in the mouse. (A) SRH injections. (B) Stereo-microscopic images during the eye dissection validating the injection site and the presence of subretinal blood (*). Dashed red areas show the SRH sites, demonstrating the healing of the injury at the ora serrata.

Fig. 3

Control, SHAM (PBS), SRH - blood, SRH - CtB injection sites. Whole mount control (A), SHAM (PBS) injected (B), blood injected (C) and blood/CtB-A555 co-injected (D) retina. The dashed white square (top row) shows the injection site, while the image below (bottom row) focuses on the injection site in 20x magnification of the same sample.

Fig. 4

Example microscopic images of the SRH site and the surrounding areas. (A) SRH site (Z1) at the level of the superficial layer (left) and the deep layer (right). (B) Neighboring zone (Z2) superficial layer (left) and the deep layer (right). (C) Z3, the farthest zone from the SRH site at the level of the superficial layer (left) and the deep layer (right). (D) Control retina (Z4; contralateral eye) at the level of the superficial layer (left) and the deep layer (right). All images are 20x magnification confocal images.

Fig. 5

Change in the number of activated microglia due to SRH. (A) Typical morphological changes from our dataset, ranging from non-activated (top row), ramified (middle row) to activated, amoeboid (bottom) scale bar: 10 μm. (B) The SRH (BZ1) site showed a pronounced increase in the activated microglia count in comparison to all other SRH zones, including neighboring (BZ2), off-site (BZ3), and contralateral retinal (Z4). PBS injected and neighboring retinal zones (PZ1-2) showed a significant, but mild increase in activated numbers versus Z4. (C) After 24 h, we detected a small decrease in the total number of microglia in BZ1-4. Interestingly, a small, non-significant decrease was detectable in the PBS injected eye (PZ1-3) vs. the contralateral eye (Z4) in the total number of microglia. Kruskal-Wallis, Dunn’s post hoc **p < 0.01 o = outlier x = mean. Original dataset available in the Supplement.

Fig. 6

Change of activated microglia counts in the superficial (SL) and deep layers (DL) of the SRH inner retina. A-B) SL microglia (MG) activation states in different zones. C-D) DL microglia distribution in the different zones and activation states. BZ indicates SRH-blood injection, PZ indicates PBS SHAM-injection. Z4 = non-treated contralateral retina. For zone details see Figs. 4 and 5. Kruskal-Wallis, Dunn’s post hoc **p < 0.01 *p < 0.05 o = outliers x = mean. Original data available in the Supplementary Dataset.

Fig. 7

Microglial activation based on automated cell morphologies with MotiQ based on a sample set of pooled cell images (15–15 cells, random) from all zones (BZ1: SRH, BZ2: neighboring, BZ3: contralateral side from the same retina, PZ1: PBS subretinal injected, PZ2: PBS neighboring, PZ3: PBS injected-contralateral side, Z4: healthy contralateral eye). (A) Total occupied area of cells decreasing when activated together with cell outlines (B), spanned areas (C), and Spanned outlines (D). The Ramification index shows the complexity of MG end-feet. ANOVA, Tukey’s post hoc **p < 0.01 *p < 0.05 o = outliers x = mean. Original data available in the Supplementary Dataset.

Fig. 8

Iba1⁺ cells infiltrated into the subretinal space showed activated morphology in apposition to Z1. (A) RPE from SRH-treated animals. B-C) High-magnification of the CtB⁺ Iba1⁺ cells. The cells on (B) show ellipsoid macrophage morphology, possibly infiltrated macrophages or activated microglia, while most of the Iba1⁺ cells were classified as activated microglia as on panel C). D) The cells marked in panel C are shown Y-angle rotated in a position to the RPE.

Fig. 9

High-resolution images from microglia in the inner retina (A; GCL) and RPE (B). CtB is only internalized by RPE microglia (B) but not inner retinal ramified microglia contrary to the direct contact with the CtB-A555 particles of these cells (A). In the RPE, Iba1⁺ cells internalized the dye and clearly expressed JamB tight junction protein (blue arrowheads) (Suppl. Video 1 for rotation of B); JamB is normally expressed only by RPE cells (blue, highlighted with light blue dots).

Fig. 10

SRH-induced microglial motility in the live retina. (A) Ex vivo confocal images from the live imaging sample area (the dashed area outlines the visible area in B). (B) Image created by collapsing 81 frames of a 20-minute-long time-lapse recording along the time domain, where colors represent the frames (see the color scale on top). This image sequence was recorded at the SRH injection site, 24 h post-injection with CtB-A555. The presence of rainbow-colored structures reflects the time-dependent change in the location of structures; when the location of a structure is unchanged, the corresponding pixels contain all colors thus they appear white. See also the time-lapse Suppl. Video 2. (C) Motion detected in the inset during the 20-minutes timelapse video (15 s/frame time-lapse, 4xZ-merged). Each image represents the accumulated movement in 2 min. Color code shows movement activity across frames.