Treatment of γFL Had no Effects on the Morphologies of Astrocytes and Microglia in the Whole Neural Retinas of Different-Age Mice
It has been reported that γFL functions significantly in the biological processes that are related to immunity in the mouse brain [14], and astrocytes and microglia are susceptible to changes of the immune state [20, 21], so we first explored the morphological changes of these immune cells in the neural retina after γFL treatment. The retinal flat-mounts from different-age mice were photographed at different layers of the neural retina. Anti- glial fibrillary acidic protein (GFAP) and anti-ionized calcium binding adaptor molecule 1 (IBA-1) antibodies were used to stain retinal astrocytes and microglia respectively, while isolectin B4 (IB4) was used to outline the retinal blood vessels [6].
As shown in Fig. 1b-c, the retinal astrocyte density decreased significantly in all 9-month-old mice versus all 8-week-old mice (P < 0.001), and further decreased in all 18-month-old mice (P < 0.05, compared with all 9-month-old mice; P < 0.001, compared with all 8-week-old mice). Nevertheless, retinal astrocyte densities of mice aged 8 w, 9 m, and 18 m after 6 d of γFL treatment showed no significant difference compared with those of mice without γFL treatment.
Normally, microglia are stratified in retinal tissue in the same way as vascular stratification, thus distributed in superficial, middle, and deep layers (corresponding to the nerve fiber layer, inner plexiform layer, and outer plexiform layer respectively) [22]. Since the boundary between superficial and middle microglia is not clear, we observed and analyzed these two microglia layers simultaneously (Fig. S1a-d). As shown in Fig. S1b-d, except for the microglia cell body diameter, the retinal microglia count and average process length changed with age. The microglia count in all 18-month-old mice was highest among all age groups (versus all 9-month-old mice, P < 0.001; versus all 8-week-old mice, P < 0.001), while the average process length was shortest (versus all 9-month-old mice, P < 0.001; versus all 8-week-old mice, P < 0.001) (Fig. S1b-c). However, the retinal microglia count, average process length, and cell body diameter were showed no significant differences between the γFL-treated group and control group aged 8 w, 9 m, or 18 m (Fig. S1b-d).
Similarly, for the microglia in the deep layer, the cell count and average process length changed with age, except for the cell body diameter (Fig. 2a-d). The microglia count in all 8-week-old mice was lowest among all age groups (versus all 9-month-old mice, P < 0.01; versus all 18-month-old mice, P < 0.001), while the average process length was longest (versus all 9-month-old mice, P < 0.05; versus all 18-month-old mice, P < 0.05) (Fig. 2b-c). Meanwhile, γFL treatment did not significantly impact the retinal microglia count, average process length, and cell body diameter for mice aged 8 w, 9 m, or 18 m (Fig. 2a-d).
Collectively, γFL seemed to have no significant effects on astrocyte density, microglia density, microglia average process length, and microglia cell body diameter in the retinas of mice at 8 w, 9 m, and 18 m of age. With the increase of age, the density of retinal astrocytes decreased, while the density of microglia increased, the average process length of microglia decreased, and the cell body diameter of microglia did not change.
Retinal Immune-Related Transcriptome Analysis Highlighted the MHC-II Mediated Antigen Processing and Presentation in Aged Mice Treated with γFL
To find the potential changed retinal immune factors induced by treating γFL, we then analyzed the mRNA sequencing data of the retina of 20-month-old mice treated with or without γFL (sequencing data were deposited in the database NCBI Sequence Read Archive, No. PRJNA748184). As shown in Fig. 3a, there were 304 differentially expressed genes in the retinas of aged mice treated with γFL, of which 145 were up-regulated and 159 were down-regulated, compared with aged control mice. Following gene enrichment analysis via clusterProfiler in R and ranking with P values, we obtained the top 6 immune-related biological processes (Fig. 3b). These biological processes include the antigen processing and presentation via MHC class II, innate immune response, biological adhesion, leukocyte migration, regulation of cell proliferation, and positive regulation of cytokine production. We noticed that the biological process with the highest enrichment was antigen processing and presentation via MHC class II, in which all significant 5 genes were up-regulated with high -log10 (P value). Such 5 genes include Cd74, H2-Aa, H2-Eb1, H2-Ab1, and H2-DMb1, among which Cd74, H2-Aa, and H2-Ab1 were also enriched in the other 5 processes (Fig.3a-b).
Noting that MHC-II mediated antigen processing and presentation may play an important role in the effects of γFL on the aged retina, we next focused our attention on this biological process. Further RT-qPCR assays were performed for the 5 genes involved in this process. Similar to the RNA sequencing results, the expression of these 5 genes were significantly increased in the γFL-treated group (Cd74, P < 0.01; H2-Aa, P < 0.01; H2-Ab1, P < 0.001; H2-DMb1, P < 0.05; and H2-Eb1, P < 0.01) (Fig. 3c).
Specifically, the progress MHC-II mediated antigen processing and presentation may play a pivotal role in the effects of γFL on the retinal aging of mice.
Treatment of γFL Increased Para-Venous MHC-II Positive Microglia in the Retina of Aged Mice
We next directly stained anti-MHC-II antibody for the retinal flat-mounts of 18-month-old mice. As shown in Fig. 4a and S2c, several MHC-II positive cells were seen in the mouse retina without γFL treatment, overlapped with IBA-1 positive cells around the retinal veins. After γFL treatment, the MHC-II positive cells were increased, co-located with IBA-1 positive cells, and arranged in a linear style with a radial-axis distribution (Fig. 4b-c, S2c). Compared with the distribution of blood vessels, we found that the distribution of these MHC-II positive cells coincided with the direction of the veins (Fig. 4b, S2c).
In the same batch of experiments, we also performed the same intervention and retinal flat-mount staining for the mice aged 8 w (Fig. S2a, S3) and 9 m (Fig. S2b, S4) to investigate whether the effect of γFL on retinal para-vascular MHC-II positive microglia is suitable for other age mice. As shown in Fig. S2-S4, the retinas of mice at these two ages had a very small number of MHC-II positive spots, which co-located with para-venous microglia. Due to the small number, the radio-linear appearance could not be formed (Fig. S3, S4).
We counted these para-venous MHC-II positive microglia in the retinal flat mounts and found that γFL treatment had no significant influences on the numbers of para-venous MHC-II positive microglia in the retinas of the 8-week- and 9-month-old mice (P > 0.05), however enhanced that of the 18-month-old mice (P < 0.001) (Fig. 4c). Additionally, we analyzed this index among the control groups. We found that para-venous MHC-II+ microglia count displayed an increase tendency from the 8-week-old to 9-month-old, and to 18-month-old groups, and that there was a significant difference between the 8-week- and 18-month-old groups (P < 0.01) (Fig. 4c).
Together, γFL significantly increased the number of MHC-II positive microglia around the retinal veins of mice at 18 m of age, but not that at 8 w or 9 m of age. In the absence of γFL intervention, the MHC-II positive microglia around the retinal veins of mice increased with age, and the difference was significant between the 18-month- and 8-week-old groups.
Treatment of γFL Increased MHC-II Positive Microglia in the Sub-Retina of Aged Mice
We then performed immunostaining for retinal sections from mice aged 8 w, 9 m, and 18 m. As shown in Fig. 5, there were no IBA-1 positive cells, namely, microglia, in the retinas of mice aged 8 w and 9 m with control light treatment. However, several IBA-1 positive cells were seen in the sub-retinal space of 18-month-old mice with control light treatment, and these cells were MHC-II low positive or even negative (Fig. 5). After γFL treatment, the IBA-1 cells in the sub-retina of 18-month-old mice exhibited obviously positive MHC-II staining (Fig. 5).
In the section staining results above (Fig. 5), we noted that subretinal microglia of all the 18-month-old mice were embedded in the side of the photoreceptor outer segment layer rather than the RPE layer. Considering this finding, we stained the mouse retinal flat-mounts and observed the photoreceptor side of the neural retinas to evaluate the subretinal microglia state. We saw IBA-1 positive cells in the sub-retinal space of the 18-month-old control mice, the cells being IB4 positive and MHC-II weakly positive or negative (Fig. 6a). After γFL treatment, subretinal IBA-1+ cell count was significantly increased (P < 0.01) (Fig. 6a, c). These microglia showed IB4 and MHC-II positive, and the number of MHC-II positive microglia were also increased significantly (P < 0.01) (Fig. 6a, d). In the aspect of cell body diameter, the subretinal microglia in γFL treatment group were larger than those in control light group (P < 0.01) (Fig. 6a-b). In the same way, sub-retinal microglia of 8-week- and 9-month-old mice were observed, with or without γFL treatment (Fig. S5). We found a few IBA-1 positive cells in the sub-retina of mice at both ages with or without γFL treatment, and these cells were IB4 positive but MHC-II nearly negative (Fig. S5).
Furthermore, we performed western blot for the whole neural retina of mice aged 18 m and assessed the expression level of CD74, which is also known as the γ chain of the MHC-II complex [23]. We found a significantly higher CD74 expression level in the neural retina of γFL-treated mice (P < 0.01) (Fig. 7).
Taken together, γFL treatment significantly increased the number of subretinal microglia, especially MHC-II positive microglia, in 18-month-old mice, and the microglia cell bodies were significantly larger in diameter than those of control-light-treated mice of the same age. In addition, there were only a few IB4 positive and MHC-II negative subretinal microglia in mice aged 8 w and 9 m, and no significant changes were observed after γFL treatment. For the whole layers neural retina tissue, γFL treatment significantly increased the expression level of CD74 of 18-month-old mice.
Treatment of γFL Increased Retinal Para-Venous MHC-II+ Microglia and Aβ Phagocytosis of Microglia in Mice Intravitreally Injected of Aβ1-40 Oligomers
It has been determined that Aβ is an important antigen in the central nervous system during aging, and that intravitreal injection of Aβ oligomers can induce retinal aging changes [24-30]. Next, we attempted to promote retinal aging in mice aged 8 w and 9 m by intravitreal injection of Aβ oligomers and then treated with γFL to observe MHC-II-associated changes in the retina (Fig. 8a).
As shown in Fig. 8b-c, the expression levels of CD74 in mice aged 8 w and 9 m after Aβ injection showed increased trends, but no statistical significances (P > 0.05). However, γFL treatment significantly promoted the expression of CD74 (versus vehicle injection/control light group, 8w, P < 0.05, and 9 m, P < 0.01; versus Aβ injection/control light group, 8w, P < 0.05, and 9 m, P < 0.05). We also evaluated the expression of CD68, a microglia marker reflecting the activation state. After Aβ injection, the levels of CD68 in the retina of mice at 8 w and 9 m of ages showed increased trends, the former had no statistical significance (P > 0.05), while the latter displayed a statistical significance (P < 0.05) (Fig. 8b-c). Similarly, γFL treatment significantly enhanced the expression of CD68 (versus vehicle injection/control light group, 8w, P < 0.01, and 9 m, P < 0.001; versus Aβ injection/control light group, 8w, P < 0.05, and 9 m, P < 0.001) (Fig. 8b-c).
We then observed the activation of microglia and expression of MHC-II in the retina of 8-week-old mice after Aβ injection by the method of retinal flat-mount immunostaining. We found that there was almost no overlap of CD68, MHC-II, and Aβ1-40 positive staining in the retinas of the vehicle and Aβ oligomers injection groups following control light treatment (Fig. 9a-b). When combined with γFL treatment, the Aβ-injected mouse retina displayed CD68 positive cells, a part of which exhibited MHC-II positive co-location and obviously distributed in a radial axis taking the optic disc as the center, while another part of which displayed Aβ1-40 positive co-staining and were located outside the above radial axis regions (Fig. 10a). For the linear and radial-axis-like distribution of retinal microglia, in view of the findings in γFL-treated 18-month-old mcie, we further stained the flat-mounts with a vascular marker, IB4. We found that the vein area corresponded to the radial-axis-distribution region shown in Fig. 10a, whereas there was no similar change in the artery area (Fig. 10b). Additionally, we also checked the subretinal microglia state and found that there was not recruited microglia in the subretinal space following vitreously injecting Aβ1-40 oligomers with or without treating γFL (data not shown).
These results indicated that under retinal aging situation induced by Aβ1-40 injection, γFL treatment promotes the activation of retinal microglia, the phagocytosis of Aβ by microglia, and the MHC-II-mediated antigen presentation in para-venous microglia.