**1. Introduction**

Alzheimer's disease (AD) is the leading cause of dementia, currently affecting an estimated 47 million people worldwide, but this number will increase unless effective treatments are discovered [1]. Since the identification of strongly immunoreactive major histocompatibility class II HLA-DR (MHC-II)-positive microglia associated with AD pathological structures [2,3], neuroinflammation is considered a prominent feature of AD pathology [4,5]. These early studies established the hypothesis that inflammatory responses to extracellular Aβ plaques and neurofibrillary tangles

might be accelerating neurodegeneration through the production of toxic inflammatory cytokines, reactive oxygen species and enzymes [6,7]. Microglia, the brain-resident macrophages, are considered the main source of these molecules. These studies suggested that anti-inflammatory agents might be effective in slowing disease progression [8], but clinical trials of anti-inflammatories have generally shown no protective effect for AD subjects [9].

There is now greater appreciation of the complexity of microglia and their many specialized functions, both pathological and reparative. Recent gene expression profiling studies of microglia isolated from human AD tissue or AD animal models have provided large amounts of data on microglial properties and identified potentially new phenotypic markers for studying microglia in disease [10–13]. These and other studies have consistently identified the purinergic adenosine diphosphate/triphosphate (ADP/ATP)) receptor P2RY12 as a significant marker for non-activated/homeostatic microglia (examples: [10,12,14–16]). Increased understanding of neuroinflammation will come from further classification of microglia for expression of these newly identified functional markers in human AD brain tissues. Microglial markers studied in human brains only represent a small number of potential targets. The most widely-used markers in neuropathology studies of human brains have been MHCII protein HLA-DR and ionized calcium-binding adaptor molecule (IBA-1). Increased expression of HLA-DR by microglia in AD gray matter has been consistently observed, but the specificity, significance or mechanism for this is unclear, while IBA-1 identifies all microglia and does not discriminate between phenotypes of microglia in human brains. Other microglial markers characterized in AD brains include CD68, a lysosomal-associated membrane protein associated with phagocytosis, CD32 and CD64, immunoglobulin Fc receptors, CD11b, colony stimulating factor-1 receptor (CSF-1R), Toll-like receptors (TLR)-2, 3 and 4, ferritin, CD163, Transmembrane Protein (TMEM)-119 [17–23] as well as Triggering receptor expressed on myeloid cells-2 (TREM-2) and CD33, microglial genes with genetic associations to AD [24,25].

P2RY12 is a member of the P2 purinergic family of receptors, a seven transmembrane-spanning G protein-coupled receptor that responds to ADP/ATP by increasing cell migration [26]. P2RY12 is mainly expressed by platelets and microglia [27]. Its function has been widely studied in relation to platelet activation and blood clotting, but its role in neuroinflammation requires further investigation. Microglia expressed significantly higher levels of P2RY12 than macrophages, in culture and in tissue, allowing discrimination between microglia and blood macrophages [11,28]. Activation of microglial P2RY12 by ADP/ATP promotes microglial chemotaxis towards sites of release [29]; these molecules are released in increased amounts by necrotic and apoptotic cells There was significantly reduced microglial chemotaxis and process formation in response to injury in P2RY12 gene-deficient mice [26]. Expression of P2RY12 by microglia (rodent and human) is downregulated after inflammatory stimulation. Injection of lipopolysaccharide (LPS) into rat brains resulted in rapid loss of P2RY12 immunoreactivity [16,22,26].

There have been limited numbers of studies of P2RY12 expression in microglia in human brains by immunohistochemistry. The most detailed previous study of P2RY12 microglia in human brains across different ages, brain regions and diseases showed expression early in brain development in all regions with limited decline with aging, while characterization of P2RY12-immunopositive microglia in 3 AD cases identified absence of P2RY12-positive microglia around Aβ plaques [30]. P2RY12-expressing microglia in astrocytomas were increased in low-grade but reduced in high-grade tumors [31]. Immunohistochemistry of human brain sections from multiple sclerosis cases confirmed loss of P2RY12 microglial immunoreactivity in areas associated with enhanced inflammation [22,30,32,33]. Another study in 2 AD cases showed that microglia positive for P2RY12 did not express TREM-2 [34].

In this report, we sought to determine if P2RY12 could be used for phenotyping the progression of human brain microglial changes in response to AD pathology by extending previous studies [30] by characterizing P2RY12 expression by microglia in detail in a staged series of AD and non-demented aged cases. The major findings showed that P2RY12 identified populations of microglia with features of resting microglia but also other populations of microglia. There was significant reduction in P2RY12 total protein levels in AD compared to ND cases, but significant amount of P2RY12 expression was present

even in severe AD cases. We identified CD68 and progranulin expression in most P2RY12-positive microglia. In pathologically-involved brains, as P2RY12 expression identified microglia with many of the different morphologies associated with inflammatory activation, classifying P2RY12 expression as a marker of homeostatic (non-activated) microglia needs to be reconsidered.

#### **2. Results**

#### *2.1. Patterns of Expression of P2RY12 by Microglia*

The aim of this study was to determine if P2RY12 expression delineates populations of resting, non-activated or reparative microglia in human brains affected by different amounts of AD plaque and tangle pathology from microglia considered as pro-inflammatory and activated. If correct, areas of inflammation around pathological structures could be defined by the presence of P2RY12-positive microglia surrounding areas with P2RY12-negative (activated) microglia.

Figure 1 illustrates initial observations of P2RY12-immunopositive microglia in middle temporal gyrus (MTG) and selected hippocampal sections. Figure 1A,B illustrates MTG sections stained with antibody to P2RY12 (Novus, rabbit polyclonal) from a non-demented (ND) (A) and AD (B) case. P2RY12-immunoreactive microglia with similar morphology were in sections of hippocampus from ND (Figure 1C) and AD cases (Figure 1D) (CA 2–3 region). A feature to note in Figure 1D is that although this region of hippocampus had significant neurodegenerative pathology as tangles, there was little difference in numbers and abundance of P2RY12 microglia compared to the ND case shown (Figure 1C) The distribution of P2RY12-positive microglia was noticeable with areas showing minimal staining in high pathology and AD cases (AD case shown in Figure 1B—red arrows). Areas with reduced staining were identified by accumulations of HLA-DR-positive microglia (Figure 1E,F). However, within these clusters of HLA-DR-positive microglia were isolated P2RY12-positive microglia (Figure 1E,F, blue arrows). Absence of P2RY12-positive microglia in the area around a cored Aβ immunoreactive plaque is shown (Figure 1G). However, the observation of P2RY12-expressing microglia with activated morphologies in close association with diffuse Aβ-positive plaques (high plaque non-demented (HPND) case—Figure 1H) suggested that P2RY12 expression by microglia was not restricted to non-activated microglia. These observations were the basis for further examination of P2RY12 expression by different types of microglia in this report.

Antibody validation was carried out to confirm that the observed immunostaining represented P2RY12 expression by microglia. Firstly, absorption of the Novus antibody with its immunizing peptide was carried out. Preincubation of diluted antibody with this 40 amino acid-recombinant peptide resulted in absence of microglial staining (Figure 1I—absorbed + PEP) compared to staining with non-absorbed antibody (Figure 1J, -PEP). The second validation step was to demonstrate the same microglial staining pattern using an independent P2RY12 antibody (Alomone Labs) produced against a different immunizing sequence. Both antibodies produced similar immunoreactivity patterns in sections from a low plaque non-demented (LPND) case (Figure 1K: Novus, Figure 1L: Alomone). To demonstrate immunoreactivity with the Alomone antibody, sections required antigen retrieval (80 ◦C, 30 min, 1 mM EDTA, pH 8.0), which was not required for the Novus antibody. The third stage was western blot analyses using brain protein extracts to demonstrate that the Novus antibody could detect a P2RY12 polypeptide of approximately 58 kDa, (representative western blots are shown in Figure 2A). This antibody also identified a polypeptide of approximately 30 kDa, a presumptive cleavage fragment of the full-length P2RY12 polypeptide. Although a number of different molecular weights for P2RY12 have been shown depending on cell source and antibody, a 58 kDa polypeptide is consistent with previous observations.

**Figure 1.** Features of P2RY12-immunoreactive microglia. (**A**,**B**). Morphology of P2RY12-immunoreactive microglia (purple) in a low plaque non-demented (LPND) case (**A**) and AD case (**B**). Sections of middle temporal gyrus (MTG) were single-stained with antibody to P2RY12. Red arrows in panel B illustrate the lack of P2RY12 immunoreactive cells in an area occupied by plaque. (**C**,**D**). P2RY12-immunoreactive microglia (purple) are a feature in hippocampus sections from non-demented (ND) case (**H**) and Alzheimer's disease (AD) case. Section shows staining in CA2 region of hippocampus. Continued presence of P2RY12-positive microglia in AD hippocampus (**D**) was noticeable. (**E**,**F**). Double-staining of section of ND and AD case with P2RY12 (purple) and HLA-DR (brown) showed limited overlap. HLA-DR-positive microglial clusters over plaques were P2RY12-negative except for single cells observed within the cluster (arrows). (**G**,**H**). Interaction of P2RY12-immunoreactive microglia (purple) and Aβ plaques (brown). The panels show two types of interactions of P2RY12-positive microglia with plaques. Positive microglia are not present in close association with mature cored plaque (**G**), while they are present in close association with diffuse type of plaques (**H**). Specificity controls for P2RY12 staining of microglia. (**I**,**J**). Staining of representative sections with P2RY12 (Novus) antibody preabsorbed with immunizing peptide (I, +Pep) compared to staining of matched section with P2RY12 antibody non-absorbed (**J**, -Pep). (**K**,**L**). Staining of matched sections with alternative P2RY12 antibody. Same staining pattern of microglia revealed with P2RY12 (Novus) antibody (**C**) as with P2RY12 (Alomone Labs) antibody **(D**). Sections reacted with Alomone Lab P2RY12 required antigen retrieval to obtain positive staining pattern. All sections shown had been counterstained with neutral red to identify nuclei (red color). Abbreviations: ND: non-demented. AD: Alzheimer's disease. MTG: middle temporal gyrus.—Pep: antibody without immunizing peptide. + Pep: antibody with immunizing peptide. Scale bars represent 50 μm.

### *2.2. Continued Expression of P2RY12 in MTG Brain Samples with Increasing Pathology and AD: Biochemical Measurements*

Increased microglial activation and pro-inflammatory cytokines have long been considered a feature of AD. If widespread throughout the AD brain, one would expect levels of P2RY12 to be very low in severe AD cases if expression was restricted to non-activated microglia. We measured the levels of P2RY12 expression in MTG sections from brains with increasing amounts of plaque pathology (Table 1, set 2). Western blot measurements of levels of P2RY12 polypeptides were made in protein samples from LPND (*n* = 10), HPND (*n* = 9) and AD cases (*n* = 9) (representative western blot–Figure 2A). Protein extracts were not available from 6 cases used in immunohistochemistry. The Novus P2RY12 antibody detected two polypeptide bands in brain samples, one of approximately 58 kDa (considered to represent full-length P2RY12) and one of approximately 30 kDa. Semi-quantitative measurements of band intensities, normalized for β-actin levels, showed significant decreased levels of 58 kDa polypeptide in AD cases (Figure 2B) but increased levels of the 30 kDa band (Figure 2C). Spearman non-parametric correlation analysis between levels of 58 kDa polypeptide and plaque and tangle scores showed significant negative correlation (P2RY12 (58 kDa)/β-actin levels compared to plaque scores; r = −0.503, *p* = 0.0039 (F1,29 = 7.618, *p* < 0.001): compared to tangle scores; r = −0.612, *p* = 0.0002 (F1,29= 24.06, *p* < 0.0001). As the Novus antibody was raised against an intracellular C-terminal peptide sequence of P2RY12, the 30 kDa bands could represent accumulations of P2RY12 after proteolytic cleavage. Its functional significance is unclear. Measurement of P2RY12 mRNA expression in cDNA derived from a similar but separate group of MTG samples (Table 1, set 3) showed non-statistically significant decrease in expression in the AD cases (Figure 2D). These results showed that significant amounts of microglial expression of P2RY12 mRNA and protein were still occurring in AD brains.


**Table 1.** Demographic details of human brain cases used.

**Abbreviations**: ApoE4: % ApoE4 alleles; Plaques: mean plaque score + SEM (scale 0–15); tangles: mean tangle score + SEM (scale demented. AD: Alzheimer's disease 0–15); LPND: low plaque non-demented; HPND: high plaque non- demented.

**Figure 2.** Quantitative biochemical measurements of P2RY12 protein and mRNA in human brains. (**A–C**). Western blot measurements of P2RY12 levels in MTG samples from LP, HP and AD brains. (**A**). Representative western blot image of P2RY12 polypeptide of MTG protein extracts identified with Novus antibody. Blots were normalized for levels of β actin. (**B**). Scatter plot showing individual P2RY12 expression levels. Significant decrease in protein levels of 58 kDa full-length P2RY12 band in AD (green shapes) compared to LPND (black) and HPND (red) cases. Chart indicates mean + Standard error of mean (SEM). Statistical analysis by one-way ANOVA with Tukey post-hoc test (F2,26 = 11.54, *p* < 0.001). (**C**). Scatter plot showing significant increase in protein levels of 30 kDa cleaved P2RY12 band in AD cases compared to LPND cases. Statistical analysis by one-way ANOVA with Tukey post-hoc test (F2,26 = 5.649, *p* < 0.01). Bar chart indicates mean + SEM. (**D**). Scatter plot showing expression levels of P2RY12 mRNA (normalized for β actin mRNA) in MTG samples. Lack of significant difference in expression of P2RY12 mRNA between LPND, HPND, and AD cases (MTG set 2). Samples measured by real time polymerase chain reaction. Statistical analysis by one-way ANOVA with Tukey post-hoc test (F2,32 = 1.031, *p* >0.05). Abbreviations: LPND: low plaque non-demented. HPND: high plaque non-demented. AD: Alzheimer's disease. NS: non-significant. \* *p* < 0.05, \*\* *p* < 0.01,\*\*\* *p* < 0.001.

#### *2.3. Patterns of Expression of P2RY12-Positive Microglia and Amyloid Beta Plaques*

When considering features of microglial activation in AD, one feature not generally considered is the anatomical distribution of microglia within brain structure. The features of neuronal architecture throughout the cortex results in different patterns of pathological development and different phenotypes of microglia. Figure 3A–C illustrate the distributions and morphology of P2RY12-positive microglia between disease groups, with boxed areas of layer I and II for each of these sections shown at higher magnification (Figure 3D–F). These areas appeared to have high expression of P2RY12 in AD cases. Sections were double-stained for P2RY12 (purple) and Aβ brown). In the HPND case illustrated (Figure 3B,E), diffuse-type Aβ plaques are observable, particularly in layers I and II (panel E), while in the AD case, the plaques are more consolidated. The morphologies of P2RY12-positive microglia varied, particularly in HPND and AD cases, and many of the plaques had associated P2RY12-positive microglia with activated morphologies. Due to the shrinkage of cortical layers in AD, the layers in LPND and HPND sections are slightly wider; white matter regions in Figure 3A (LPND) and Figure 3B (HPND) are not visible.

**Figure 3.** Distribution of P2RY12 microglia within cortical layers in relation to amyloid beta plaques in pathologically staged samples (**A**–**C**). Lower magnification photomicrographs showing the changes in P2RY12 microglia distribution compared to Aβ plaques within cortical layers of MTG. Sections from low plaque, high plaque and AD cases stained for P2RY12 (purple) and Aβ (brown). Sections were counterstained with neutral red to identity cellular morphology. Scale bars represent 200 μm. (**D**–**F**). Higher magnification photomicrographs of the areas in panels (**A**–**C**) indicated by frames. Scale bars represent 50 μm.

#### *2.4. Immunohistochemical Measurements of P2RY12 Expression*

Measurement of the occupied areas of P2RY12 immunoreactivity in a complete series of MTG sections was carried out. These sections were imaged at low magnification to include all cortical layers and analyzed using ImageJ software. There was a small but significant decrease in mean occupied area of P2RY12 immunoreactivity in AD cases (Figure 4A). Representative low-magnification images of sections used for measurements are shown (Figure 4B). Determining the numbers of P2RY12 immunoreactive microglia in stained sections in a defined area using a microscope eyepiece reticule was also carried out. These were counted in five individual fields through the cortical laminar columns (layer I and II, layer III, layer IV, layer V and layer VI), and three separate columns from each section. Mean total values for all laminar showed no significant difference between disease groups (Figure 4C),

but we observed a redistribution of P2RY12-positive microglia into layers I and II in AD cases. As shown in Figure 3F, this area contained significant amount of Aβ. When cell counts for layer I and II were excluded from the analyses, the decreased number of P2YR12-positive microglia in the other layers in the AD cases was statistically significant (*p* < 0.05) (Figure 4D).

**Figure 4.** Quantitative measurements of P2RY12 immunoreactive structures in human brain tissue sections. (**A**,**B**). Measurements of area occupied of P2RY12 immunoreactivity in LPND (*n* = 11) (black), HPND (*n* = 11) (red) and AD (*n* = 12) (green) cases. (**A**). Sections were imaged at 4x magnification (three random areas/cases) and area occupied in thresholded images measured using Image J software. Results show significant decrease between HPND and AD cases. Statistical analysis by one-way ANOVA with Tukey post-hoc test (F2,29 = 3.903, *p* < 0.05). (**B**). Representative images of low magnification images of LPND, HPND and AD used for measurements show distribution of P2RY12 immunoreactive microglia. Scale bars represent 400 μm. (**C**). Numbers of P2RY12 immunoreactive microglia in all cortical layers. Scatter plot showing the estimated total mean number of microglia in cortical layers I–VI. Individual points represent the mean of total numbers from three separate measures for each slide. The numbers of microglia/2 mm<sup>2</sup> field were counted. Results show insignificant decline in mean number of P2RY12-positive microglia in AD cases. Statistical analysis by one-way ANOVA with Tukey post-hoc test (F2,29 = 1.704, *p* = 0.14). (**D**). Numbers of P2RY12 immunoreactive microglia in cortical layers III-VI. Scatter plot showing the estimated total mean number of microglia in cortical layers III-VI. The numbers of microglia counted in Layers I and II were subtracted from the total. Individual points represent the mean of total numbers (except Layers I and II) from three separate measures for each slide. The numbers of microglia/2 mm<sup>2</sup> fields were counted. Results show significant decline in mean number of P2RY12 positive microglia in AD cases. Statistical analysis by one-way ANOVA with Tukey post-hoc test (F2,29 = 3.201, *p* < 0.05). **Abbreviations**: LPND: low plaque non-demented. HP: high plaque non-demented. AD: Alzheimer's disease. ND: non-demented. NS: non-significant.

#### *2.5. P2RY12 Expression and Microglial Morphology*

The morphology of microglia has been considered to reflect their activation states, but these types of classifications have limitations especially in aged human brains [35,36]. The typical ramified appearance of resting microglia positive for P2RY12 in layer III of LPND case is shown (Figure 5A). Other morphologies can be observed, including dystrophic (fragmented) (Figure 5B,C) and "tufted" microglia (Figure 5D). It was observed that all P2RY12-positive microglia showed immunoreactivity for IBA-1 (purple/brown colocalization), but there were IBA-1-positive/P2RY12-negative (arrowhead) or IBA-1-positive/P2RY12-weakly positive (arrow) (Figure 5E—LPND case). Figure 5F shows the appearance of a cluster of IBA-1-positive/P2RY12-negative activated microglia (arrowhead) surrounded by strongly stained P2RY12-positive microglia. Considering the interaction of P2RY12-positive

microglia with Aβ plaques, variable responses were seen. Figure 5G shows that most of the diffuse plaques (brown) in an HPND case had associated P2RY12-positive microglia (purple) with extended processes. Figure 5H,I from AD cases show microglia with more activated morphologies (shorter processes and enlarged cell bodies) interacting with more mature plaques (arrowheads). Another frequent observation was the presence of large P2RY12-positive rod-shaped microglia (Figure 5J,K), with some directly interacting with plaques (Figure 5K). One other noticeable feature seen in most cases was that many P2RY12-positive microglia, especially those with ramified processes, showed close interactions of processes with neurons (Figure 5L).

**Figure 5.** Different microglial morphologies associated with P2RY12 expression. Representative immunohistochemistry results of tissue sections stained to identify P2RY12 (purple) alone and Aβ (brown), IBA-1 or phosphorylated tau. (**A**). Ramified microglia in LPND case. (**B**,**C**). Microglia with fragmented morphology in HPND cases. (**C**). Fragmented microglia associated with diffuse Aβ plaques. (**D**). P2RY12 microglia with tufted morphology in AD case. (**E**,**F**). Colocalization of P2RY12 and IBA-1.

(**E**). Rod shaped IBA-1-positive microglia (brown arrow) with minimal P2RY12 immunoreactivity. P2RY12-positive, IBA-1-positive microglia (arrowheads (**E**,**F**)). All P2RY12 immunoreactive microglia showed some IBA-1 immunoreactivity. (**F**). IBA-1 positive cluster surrounded by P2RY12 microglia. (**G**–**I**). Different morphologies of P2RY12-positive microglia interacting with Aβ plaques. (**G**). P2RY12-positive microglia with long processes interacting with diffuse plaques in HPND case. (**H**,**I**). P2RY12-positve microglia with activated morphologies (large cell bodies, short processes) interacting with dense Aβ plaques. (**J**,**K**). P2RY12-positive rod-shaped microglia in LPND (**J**) and HPND (**K**) sections. (**L**) P2RY12-positive microglial processes show interactions with neurons. Scale bars represent 50 μm.

#### *2.6. Confocal Microscopy Localization of P2RY12 with HLA-DR, CD68 and Progranulin in Brain Microglia*

To follow up observations, further investigations of the phenotypes of P2RY12-positive microglia using antibodies to activation markers HLA-DR, CD68 and progranulin were carried out using multicolor laser confocal microscopy to demonstrate cellular colocalization.

In LPND cases with fewer HLA-DR positive microglia, those present showed colocalization with P2RY12 immunoreactivity (example: Figure 6A–C). In HPND cases, with greater numbers of HLA-DR positive microglia cases, there was less colocalization between P2RY12 (green) and HLA-DR (red) immunoreactivity (Figure 6D–F). Similarly, in AD sections (Figure 6G–I), separation of P2RY12 and HLA-DR immunoreactivity can be observed though small amounts of P2RY12 immunoreactivity with HLA-DR (Figure 6I—yellow arrow) were present in some cells.

**Figure 6.** Confocal microscopy of P2RY12 and HLA-DR positive microglia in pathologically staged cases. (**A**–**I**) Images of P2RY12 (green), HLA-DR (red) and merged (yellow) with DAPI (blue) in MTG of LPND (**A**–**C**), HPND (**D**–**F**) and AD case (**G**–**I**) to show the distribution of P2RY12 and HLA-DR immunoreactivity. Examples of colocalization (yellow arrows) are shown in Merge image. Scale bar represents 50 μm.

Similar analyses were carried out comparing co-expression by microglia in LPND, HPND and AD MTG sections for P2YR12 and CD68, the monocyte-specific phagocytic lysosomal marker (Figure 7), and with progranulin, another lysosomal-associated marker, (Figure 8). Both proteins have been considered as markers of activated microglia. We had hypothesized that there would be a clear discrimination between P2RY12 and CD68 staining to distinguish between resting and phagocytic microglia, but this was incorrect. We observed that most P2RY12 positive microglia were also positive for CD68. The distribution of P2RY12 and CD68 in top cortical layers are shown in Figure 7A–C at low-magnification. Higher magnification images of these cases show that there was strong CD68 staining in low pathology cases (Figure 7D—red CD68 alone and Figure 7G—merged images of P2RY12 and CD68). There was no noticeable increase in CD68 intensity in HPND and AD cases (Figure 7E,F), but there were more CD68 positive cells with limited P2RY12 staining (Figure 7I) in AD cases.

**Figure 7.** Confocal microscopy of P2RY12 and CD68-positive microglia in pathologically staged cases. **A**–**C).** Low magnification merged images of P2RY12 (green), CD68 (red) and DAPI (blue) in MTG of LPND (**A**), HPND (**B**) and AD cases (**C**) to show the distribution of P2RY12 and CD68 immunoreactivity through cortical layer. Scale bars represent 50 μm. (**D**–**F**). Higher magnification merged images of CD68 (red) in MTG of LPND (**D**), HPND (**E**) and AD cases (**F**) to show the distribution of CD68 immunoreactivity. Similar amounts of CD68 immunoreactivity was present in each disease group. Scale bars represent 50 μm. (**G**–**I**). Merged images of P2RY12 (green) with the CD68 (red) images shown in (**D**,**E**)) with DAPI (blue). Scale bars represent 50 μm.

Analyses of expression of progranulin in P2RY12-positive microglia in staged samples were also carried out. Progranulin is a multi-functional protein with anti-inflammatory, growth factor and lysosomal regulatory properties. Similar to CD68, microglial expression of progranulin has been associated with activated microglia [37]. The representative images shown in Figure 8 show that most P2RY12-positive microglia, whether in LPND, HPND or AD MTG sections, are also

positive for progranulin. The lower magnification images (Figure 8A–C) show the distribution of P2RY12 and progranulin-positive microglia through sections of Layer II and III. All progranulin immunoreactivity was associated with cells showing different amounts of P2RY12 immunoreactivity. Although progranulin immunoreactivity can be detected in neurons with the antibody used, this was mainly detectable only in large pyramidal neurons in Layer V and not in the layers shown in this figure [38]. Higher magnifications images show progranulin immunoreactivity alone (Figure 8D–F), and merged images combined with P2RY12 show that all P2RY12 microglia appeared to be progranulin positive (Figure 8G–I).

**Figure 8.** Confocal microscopy of P2RY12 and progranulin-positive microglia in pathologically staged cases. (**A**–**C**). Low magnification merged images of P2RY12 (green), progranulin (PGRN) (red) and DAPI (blue) in MTG of LPND (**A**), HPND (**B**) and AD cases (**C**) to show the distribution of P2RY12 and PGRN immunoreactivity through cortical layers. Scale bar represents 50 μm. (**D**–**F**). Higher magnification images of PGRN (red) in MTG of LPND (**D**), HPND (**E**) and AD cases (**F**) to show the distribution of immunoreactivity. Similar amounts of PGRN immunoreactivity was present in each disease group. Scale bar represents 50 μm. (**G**–**I**). Merged images of P2RY12with PGRNimages shown in (**D**–**F**) with DAPI (blue) showing expression in same cells (yellow arrows). Scale bar represents 25 μm.

### *2.7. Patterns of Expression of P2RY12-Microglia with Di*ff*erent Types of Plaques in Pathologically Staged Cases*

Confocal microscopy with multi-layered images of P2RY12-immunoreactive microglia with Aβ provide additional information on their interactions. We had earlier observed that there were two types of P2RY12-positive microglia having interactions with Aβ plaques. Those surrounding plaques and those interacting with plaques. It might be assumed that all Aβ plaques would activate the microglia in a proinflammatory manner resulting in downregulation of P2RY12 expression. Figure 9 illustrates that P2RY12-positive microglia interact with diffuse-like, non-cored amyloid (earlier plaques) (Figure 9A—LPND case: Figure 9C—HPND case: Figure 9E—AD case), while the cored plaques had zones without P2RY12-expressing microglia (Figure 9B—LPND: Figure 9D—HPND: Figure 9F—AD).

**Figure 9.** Interaction of P2RY12-positive microglia with different types of Aβ-positive plaques in pathologically-staged cases. (**A**–**F**). P2RY12-positive microglia (green) interacting with Aβ-positive diffuse-type plaques (red) (**A**,**C**,**E**), in LPND (**A**), HPND (**C**), and AD (**E**), but not with mature-type cored plaques (**B**-LPND) (**D**-HPND) (**F**-AD). Scale bars represent 50 μm.
