*2.9. In Vitro Analysis of P2RY12 Expression by Human Microglia*

Our brain tissue observations of P2RY12-expressing microglia have shown multiple features of these cells in relation to their activation states. To further investigate if increased P2RY12 expression represents a marker of alternative activation, cultured human brain-derived microglia were treated with interleukin-4 (IL-4), Aβ peptide and other proinflammatory agents to determine how activation affects P2RY12 mRNA expression. Samples analyzed by qPCR for P2RY12 mRNA expression showed strong induction of expression by IL-4 treatment, and downregulation to different extents with Aβ and other proinflammatory agents (Figure 11A). Induction of P2RY12 protein was also observed by western blotting in IL-4-treated microglia (Figure 11B).

**Figure 11.** Expression of P2RY12 mRNA and protein by in vitro cultures of human microglia. (**A**). Interleukin-4 stimulates P2RY12 mRNA expression. Bar chart showing results real time PCR analysis for P2RY12 mRNA of human microglia stimulated with indicated agents. Results of analysis of single human microglia case (each in triplicate) and representative of other analyses. Abbreviations: Con, control unstimulated: IL-4, interleukin-4 (40 ng/mL): Aβ2 and Aββ5 (Aβ (1–42) 2 μM and 5 μM): IFNγ, interferon-γ (100 ng/mL): LPS, lipopolysaccharide (100 ng/mL): LPS/IFNγ (doses combined): IL-6, interleukin-6 (40 ng/mL). \*\*\*\* *p* < 0.0001. \* *p* < 0.05. (**B**). Western blot of human microglia protein samples probed with antibody to P2RY12. Increased amounts of P2RY12 (58 kDa) in IL-4 treated samples.

#### **3. Discussion**

The aim of this work was to examine phenotypes of P2RY12-immunopositive microglia in aging and AD brains in relation with AD-associated pathological structures. P2RY12 has been defined as a specific marker to discriminate between microglia, with high levels of expression, and macrophages, with low levels of expression [16,30]. In addition, based on experimental findings, continued expression by microglia of P2RY12 in brain should define them as non-activated, namely those not producing proinflammatory cytokines associated with enhanced inflammation. As experimental studies have shown that proinflammatory activation of microglia resulted in significant reduction in P2RY12 expression [16,26], it was hypothesized that identifying microglia with high expression of P2RY12 compared to microglia positive for classical activation markers but with low to negative expression of P2RY12 would provide a means of identifying areas of active inflammation in brain tissue [30]. The findings of this work showed that microglial expression of P2RY12 was downregulated in AD tissue samples, but immunohistochemistry identified more complex patterns of increased P2RY12 expression associated with pathological structures than previously identified [30].

Recent gene expression profiling of single-cell microglia from rodent and human sources had confirmed that P2RY12 mRNA expression was associated with a non-activated phenotype and expression was downregulated with progression of disease [10,14,15]. However, our initial observations in human brains suggested that this might not cover all features of P2RY12 expression by microglia; for this reason, we sought to provide detailed characterization of P2RY12 microglia in human aged and AD brains. The importance of microglial phenotyping to identifying functional markers is now appreciated along with the need for greater numbers of markers [39]. For a number of years, the classification of microglia (and macrophages) into functional M1 and M2 groups was applied but it is now appreciated that this system does not account for the complexities of microglial phenotypes in diseased brains [40]. Recent findings have defined a phenotype of microglia designated "disease-associated microglia" (DAM), which describes a transcriptional signature first associated with response to neurodegeneration-associated molecular patterns (DAM stage 1) that progresses to a signature associated with a protective role to limit inflammation (DAM stage 2) that is coordinated by TREM-2 signaling. Downregulation of P2RY12 from homeostatic to stage 1 DAM confirms earlier findings but our cellular localization findings suggest that upregulation of P2RY12 may also be a feature associated with later stages of AD.

We have made some new findings on P2RY12 expression by microglia in human brains as part of this study that extend previous findings [30]. There were decreased levels of total P2RY12 protein of 58 kDa in brain extracts from AD cases compared to non-demented low and high pathology cases, as could be expected, but we also identified increased levels of 30 kDa P2RY12 polypeptide in AD cases. This indicates that downregulation of P2RY12 protein levels might be due to enhanced cleavage of this plasma membrane protein. The antibody we used for our study was prepared against a 40-amino acid recombinant peptide corresponding to the C-terminal cytoplasmic domain of P2RY12 (amino acids 303–342). P2RY12 is a G-protein-coupled receptor for ADP containing 7-transmembrane domains. Based on this structure, the 30 kDa polypeptide would represent a remaining cell-associated peptide that does not contain the N-terminal sequences. As P2RY12 has been reported to have multiple ADP-binding domains, it is unclear if this fragment will be biologically active for ADP binding and signal transduction. Downregulation of P2RY12 expression as a result of proinflammatory activation and upregulation as a response to IL-4 were regarded as features of alternatively activated microglia though the mechanisms for this to occur in brains is unclear as IL-4 expression has not been consistently detected in brain tissue [41]. We showed that downregulation of P2RY12 expression following inflammatory stimulation also occurred in cultured human microglia as did others [42]. The involvement of progranulin, a neuroprotective and anti-inflammatory molecule, in microglial function is still unclear. Progranulin positive microglia are found throughout brain but our findings showed P2RY12-positive microglia in all disease groups were positive for intracellular progranulin. It has been shown that IL-4 upregulates progranulin expression by cultured human microglia [43]. Based on current findings of gene regulation, one can speculate that P2RY12/progranulin positive microglia are protective rather than reactive, but further studies are required.

A previous study of P2RY12 microglia across different human brain regions and ages made similar observations as this study that most P2RY12-positive microglia were also CD68 positive [30]. CD68, a myeloid specific lysosomal-associated membrane protein associated with phagocytosis, has been considered as a microglial activation marker in a number of studies [23,44], but colocalization of CD68 and P2RY12 would suggest its involvement in normal microglial function. Two earlier studies have observed that P2RY12 was not expressed by microglia accumulating around plaques in AD brains. We also observed this for many plaques, but there were noticeable exceptions as many diffuse-like Aβ(6E10-immunoreactive) plaques had P2RY12-positive microglia interacting with them. We also observed varied morphologies for P2RY12-microglia interacting with plaques, including fragmented, tufted and rod-shaped. The study of Mildner et al. ([30]) employed a different antibody (Sigma-Aldrich HPA014518) than we used (Novus NBP2-33870), however both antibodies were produced against the same C-terminal 40-amino acid peptide sequence so should have the same properties. Detection differences in these studies could be due to tissue fixation conditions. Our study employed lightly-fixed free-floating brain sections for immunohistochemistry, while Mildner et al. employed harder-fixed paraffin-embedded sections [30]. For their study, antigen retrieval was required for all antibodies, while we found this not necessary for free-floating sections when using the Novus antibody. However, sensitivity of P2RY12 detection to fixation was observed as the alternate antibody (Alomone-APR-012) we used only worked when free-floating sections underwent antigen retrieval processing.

Although P2RY12 appears to be an excellent marker for microglia in brain, it is unclear whether expressing microglia can be classified as protective or proinflammatory. How P2RY12-mediated responses by microglia are involved in AD pathogenesis is unresolved. P2RY12-mediated chemotactic responses to ATP and ADP, which are released by damaged or dying cells, appears to be an early inflammatory response. The rapid downregulation of P2RY12 expression with proinflammatory activation would appear to function to anchor microglia at sites of inflammation. It has been proposed that downregulation of P2RY12 is accompanied by increased expression of adenosine A2 receptor, the breakdown products of P2RY12 ligands ADP/ATP [45]. A recent in vitro study of microglia demonstrated the proinflammatory consequences of inflammasome and NFκB activation by extracellular ADP activation of P2RY12 [46]. In a rodent ischemia model, blockade of microglial P2RY12

with ticagrelor, an antagonist, reduced ischemic damage by microglia by reducing their migration to sites of injury [47]. In another ischemia animal model, and with an in vitro model of oxygen–glucose deprivation (OGD), inhibition of microglia P2YR12 with clopidogrel, another antagonist, or by P2RY12 gene expression knock-down significantly reduced microglial migration and neurotoxicity [48].

The potential for using P2RY12-positive microglia distribution to define distribution of neuroinflammation in AD or other neurodegenerative diseases is still valid, though not as clear as originally hypothesized. There has been a widely-held concept that AD neuroinflammation is widespread through affected areas, but these findings suggest it is highly localized. If one considers the scheme outlined in Figure 12, which is a representation of Figure 1E,F,G and Figure 9D,F, that activated microglia (low-negative P2RY12 microglia) associated with mature amyloid plaques are secreting cytokines and other molecules that downregulate P2RY12 expression by microglia, these will be present in the zone around the plaques (Zones 1 and 2), but not in Zone 3 where the P2RY12-positive microglia are located. With the use of laser capture microscopy and proteomics techniques, analyses of micro-dissected regions are feasible and could identify key neuroinflammatory factors associated with AD that are not represented when larger dissected pieces of brain tissue are analyzed. The feasibility of this approach has been demonstrated in proteomic analyses of micro-dissected plaques and tangles from AD brains [49,50].

**Figure 12.** Proposed scheme of arrangement of different P2RY12-expressing microglia around Aβ plaques. Suggested scheme to describe localized areas of microglial inflammation around plaques. Zone 1: microglia interacting with mature plaques (HLA-DR high, P2RY12 negative) producing proinflammatory cytokines. Zone 2: Area adjacent to plaque with low or negative P2RY12 positive microglia. Zone 3. P2RY12 high expression in surrounding area defining the boundary between proinflammatory area (Zone 1 and 2) and non-affected area (Zone 3 and beyond). As described in this report, exceptions to this scheme were observed.

#### **4. Materials and Methods**

#### *4.1. Human Brain Tissue Samples*

Human brain tissue samples used in this study were obtained from the Banner Sun Health Research Institute Brain and Body Donation Program, Sun City, Arizona, U.S.A. [51,52]. The operations of the Brain and Body Donation Program (BBDP) have received continuous approval of Institutional Review Boards (IRB). Current operations have been reviewed by Western IRB (Puyallup, WA, USA). Written informed consent for collection and use of brain and other tissues for research purposes were obtained from donors or next-of-kin. Tissue studies carried out in the U.S.A. were considered non-human subject research under federal regulations. Tissue studies carried out in Japan were approved by Shiga University of Medical Science Ethical Committee (Project Certificate no. 29-114). Demographic details of cases used in this study are summarized in Table 1.
