CD8+ T cells are increased in the subventricular zone with physiological and pathological aging

Abstract Age‐related cognitive decline and neurodegenerative diseases are associated with less functional neurogenic niches. It has been recently shown that aged subventricular zone (SVZ) suffers an infiltration of T cells, which affects neural stem cell activity in mice. Whether this occurs in human neurogenic niches or to which extent T‐cell infiltration is also taking place in neurodegenerative diseases remains unknown. In this work, we studied the presence of T cells in both human neurogenic niches in young and old individuals. There was a significant increase in the number of CD3+ and CD8+ T cells in the SVZ of elderly individuals, which was not detected in the dentate gyrus. Moreover, we also found CD3+ and CD8+ T cells in the SVZ of individuals with neurodegenerative diseases. However, T‐cell count was similar when compared non‐neuropathological elderly with disease diagnosed patients. Our study reveals the infiltration of T cells in old human brains, particularly in the SVZ under non‐pathological conditions and also in neurodegenerative contexts.

seems to be dependent on interferon-γ response (Dulken et al., 2019). Remarkably, CD8 + T-cell infiltration within brain seems to be a stochastic event rather than a passive diffusion due to blood-brain barrier disruption (Dulken et al., 2019;Ritzel et al., 2016). Moreover, once T cells are into the brain neurogenic niches, they may recognize specific antigens and clonally expand to become activated T cells that are different from circulating T cells (Dulken et al., 2019). In this work, we studied the presence of T-cell infiltration in human neurogenic niches of non-neuropathological individuals of different ages and also with neurodegenerative diseases.

| CD3 + and CD8 + T-cell infiltration is increased in the subventricular zone of aged individuals
We investigated the presence of T cells in both neurogenic niches, SVZ and DG, which were obtained in the same section and processed jointly (Figure 1a,b). In the case of SVZ, we defined a surrounding area of 1 mm 2 ( Figure 1c) within which all parenchyma-infiltrating positive cells, but not those located into blood vessels or vascular walls, were considered. Regarding DG, we encompassed the whole area of this structure (see Figure 1d), excluding again those positive cells located into blood vessels or vascular walls.
The SVZ is a complex niche containing multiple cell types. The most abundant ones are glial cells, which are characterized by a medium, vesicular nucleus with an eosinophilic cytoplasm, sometimes visible, from which fine prolongations project giving them a reticular aspect (Figure 1e; red arrows), and express the GFAP marker ( Figure S1). In contrast, microglial cells are characterized by a small, elongated nucleus with a little apparent cytoplasm (Figure 1e; black arrow). The typical morphology of T cells includes a small, round, and dark nucleus with scarce cytoplasm, often very little apparent. Cells with these characteristics were observed in the SVZ of individuals over 65 years old (Figure 1e; green arrows). At the central nervous system (CNS), T cells could be confused with oligodendrocytes, but not in the SVZ, where no myelinated axons are found.
To characterize the populations of T cells in aged patients, we performed immunohistochemistry (IHC) analyses using the pan T-cell marker, CD3 (Figure 2a; red arrows). Other cells consistent with a T-cell morphology not stained using the CD3 antigen were also observed ( Figure 2a; green arrows). Moreover, CD4 staining revealed some T helper cells ( Figure 2b) and several cells were also co-stained for CD3 and CD4 markers ( Figure S2). IHC further showed the presence of cells expressing the specific marker of cytotoxic T cells, CD8 ( Figure 2c). This protein, as well as CD3 and CD4, is membrane protein complexes, so IHC enables to visualize some structures that are not revealed by hematoxylin-eosin, giving cells a larger size than observed after hematoxylin-eosin staining. To further determine the population of cytotoxic T cells, we performed co-immunofluorescence of CD8 with CD3 or CD4. Confocal analysis revealed that the majority of CD8 + cells were also positive for CD3 ( Figure 2d) but negative for CD4 markers (Figure 2e). This staining pattern is indicative F I G U R E 1 Identification of selected brain regions. (a) Coronal section of adult human brain showing the hippocampus and the temporal horn of the lateral ventricles. (b) Microscopic coronal section showing the hippocampus, entorhinal cortex, and lateral ventricle at the lateral geniculate nucleus level (our reference point). (c) Subventricular zone: We selected an area of 1 mm 2 that included the four layers (I-IV); all positive cells with immunohistochemical staining that infiltrate the tissue were quantified (excluding those intravascular and vessels wall located). (d) Dentate gyrus: We selected the whole area of this structure, which is defined by three layers (subgranular cell layer, granular cell layer, and molecular layer); all positive cells with immunohistochemical staining that infiltrate the tissue were quantified (excluding those intravascular and vessels wall located). (e) Hematoxylin-eosin staining of the subventricular zone with the main cell types: glial cells (red arrows), T cells (green arrows), and microglia (black arrow)

IFNγ
of the presence of cytotoxic T-cell infiltration (CD3 + CD8 + cells) in the SVZ of elderly individuals. Moreover, some cells from SVZ of old individuals expressed high levels of interferon-γ ( Figure 2f).
Next, we performed IHC assays to quantify the presence of

| CD3 + and CD8 + T cells do not infiltrate the dentate gyrus in aged individuals
Next, we moved to the other neurogenic niche in adults, that is, the DG, and performed the same analyses of CD3, CD4, and CD8 expression by IHC. In the young group, we observed a count of CD3 + T cells less than 1 in all cases except for one patient with 3 cells per mm 2 who presented severe liver failure due to herpes viremia These results point out that T-cell infiltration in the DG is not increased with aging unlike that observed in the SVZ.

| CD3 + and CD8 + T cells are present in subventricular zone of aged individuals with neurodegenerative diseases
We

| D ISCUSS I ON
In this study, we provide new evidence about the existence of T-cell infiltration in brain with physiological and pathological aging. A recent report by Brunet´s Lab showed that T-cell infiltration occurs in old mice and human SVZ, where CD3 + and CD8 + cells were elevated when compared to young individuals (Dulken et al., 2019).
In our work, we extend these data and show that CD3 + and CD8 + cells are increased in the SVZ of elderly individuals but not in the DG, and they are also present in neurodegenerative diseases such as AD. We found higher SVZ CD8 + total counts than Dulken´s study although the ratio of increase between young and old individuals was the same in both works. This could be due to different inclusion criteria of patients (Table 1), different antibodies employed or the anatomical region considered, since they studied the SVZ at basal ganglia level and we did so at temporal horn. Concerning this, it is important to mention that we selected this region to obtain both neurogenic niches in the same section in order to be processed jointly. Conversely, our results show that T-cell infiltration is negligible in the DG of both elderly individuals and patients CD8 but not the CD3 pan T-cell marker (Campbell, Guy, Cosgrove, Florida-James, & Simpson, 2008) and also subsets of CD8 + cells with markers of NK cells have been identified in human peripheral blood (Ohkawa et al., 2001).
It is also important to mention that we described the same trend of T-cell infiltration in neurodegenerative diseases as we found in aged individuals with no neuropathological injuries (presence of CD3 + and CD8 + T cells in the SVZ but not in the DG). All our patients with a neurodegenerative disease displayed AD. The contribution of brain inflammation and T cells to AD is still controversial, but other groups have also found CD8 + cells in the brain parenchyma from patients with AD (Itagaki, McGeer, & Akiyama, 1988;Liang et al., 2017;Rogers, Luber-Narod, Styren, & Civin, 1988 interferon-γ, and the evidence that CD8 + T cells inhibit NSC proliferation (Dulken et al., 2019), might indicate that CD8 + cells infiltrating the human SVZ trigger a cytotoxic immune response in this neurogenic niche, eventually leading to a reduction in the NSC pool in aged individuals. The increased expression of CD68 and IBA1 microglia markers as well as IL1α in the aged SVZ, which suggest a pro-inflammatory environment into the niche, may support this point.
In conclusion, we report an increase in CD8 + T cells in aged human SVZ, but not in the DG, which is also observed in neurodegenerative diseases at the same level. Our results point out to a specific role of inflammation in human SVZ associated with physiological and pathological aging.

| Samples
Human brain samples were collected from autopsies conducted at Donostia University Hospital (Spain). Postmortem interval (PMI) was limited to 12 hr due to its effects on brain proteins. Brains were kept in a fixative solution (4% paraformaldehyde) for a period of not less than 24 hr. Samples were divided into three groups: "young" (n = 4; individuals ranged from 36 to 58 years old), "old" (n = 10; individuals aged 65-87 years old), and "neurodegenerative" (n = 5; individuals between 60 and 90 years old). Inclusion criteria for selection of "young"

| T-cell quantification
The quantification of positive T cells for the different markers in the SVZ and the entire DG was manually performed in entire coronal sections from previously scanned images (Figure 1a,b). In the case of SVZ, we defined a surrounding area of 1 mm 2 (Figure 1c). All parenchyma-infiltrating positive cells within this area, but not those located into blood vessels or vascular walls, were counted and considered for further analyses. Regarding DG, we encompassed the whole area of this structure (see Figure 1d), counted the total number of positive cells in such area, excluding those located into blood vessels or vascular walls, and normalized this number to the area considered. T cells were identified as CD3 + cells, helper T cells as CD4 + , and cytotoxic T cells as CD8 + .

| Statistical analysis
The number of T cells was expressed as total number of positive cells per unit area (mm 2 ). Two-tailed Mann-Whitney U test was performed to compare CD3-, CD4-, and CD8-positive cell counts between groups. Asterisks (*, **, and ***) indicate statistically significant differences (p < 0.05, p < 0.01, and p < 0.001, respectively).

CO N FLI C T O F I NTE R E S T
None.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.