Comparison of Monoamine Oxidase-A, Aβ Plaques, Tau, and Translocator Protein Levels in Postmortem Human Alzheimer’s Disease Brain

Increased monoamine oxidase-A (MAO-A) activity in Alzheimer’s disease (AD) may be detrimental to the point of neurodegeneration. To assess MAO-A activity in AD, we compared four biomarkers, Aβ plaques, tau, translocator protein (TSPO), and MAO-A in postmortem AD. Radiotracers were [18F]FAZIN3 for MAO-A, [18F]flotaza and [125I]IBETA for Aβ plaques, [124/125I]IPPI for tau, and [18F]FEPPA for TSPO imaging. Brain sections of the anterior cingulate (AC; gray matter GM) and corpus callosum (CC; white matter WM) from cognitively normal control (CN, n = 6) and AD (n = 6) subjects were imaged using autoradiography and immunostaining. Using competition with clorgyline and (R)-deprenyl, the binding of [18F]FAZIN3 was confirmed to be selective to MAO-A levels in the AD brain sections. Increases in MAO-A, Aβ plaque, tau, and TSPO activity were found in the AD brains compared to the control brains. The [18F]FAZIN3 ratio in AD GM versus CN GM was 2.80, suggesting a 180% increase in MAO-A activity. Using GM-to-WM ratios of AD versus CN, a >50% increase in MAO-A activity was observed (AD/CN = 1.58). Linear positive correlations of [18F]FAZIN3 with [18F]flotaza, [125I]IBETA, and [125I]IPPI were measured and suggested an increase in MAO-A activity with increases in Aβ plaques and tau activity. Our results support the finding that MAO-A activity is elevated in the anterior cingulate cortex in AD and thus may provide a new biomarker for AD in this brain region.


Introduction
Early diagnosis and disease monitoring strategies for Alzheimer's disease (AD), the most common type of dementia, are critically important due to its high prevalence in an aging worldwide population. Characterized by the accumulation of amyloid β (Aβ) plaques and neurofibrillary tangles (NFT) in the brain [1,2], molecular biomarkers for AD are now indispensable for the clinical definition of the process and stage of the disease [3]. PET imaging of Aβ plaques is currently being used to monitor drug treatments [4]. There is now an increased focus on tau as a more accurate early predictive marker for AD diagnosis. Phosphorylated p-tau immunoassays p-tau181 [5] and p-tau217 [6] confirmed that plasma p-tau has high sensitivity and specificity to detect AD neuropathology, which was ascertained with [ 18 F]flortaucipir PET [7,8]. Thus, both Aβ plaque and NFT PET imaging are now playing a major role in the diagnosis, staging, and treatment evaluations of AD (Figure 1). MAO-A may play a role in the regulation of neuronal survival in neurodegenerative disorders [14] as well as monoaminergic dysfunction and mitochondrial dysfunction in AD [15]. MAO-A gene polymorphisms have been shown to be associated with neurological changes in AD pathology [16]. An enzymatic assay of MAO-A activity in AD brain tissue showed an increase or no change in several brain regions [17], while another report using similar assay methods indicated a decrease in MAO-A activity in the prefrontal cortex [18]. MAO-A hyperactivity has been shown to be associated with depression, suggesting MAO-A inhibitors as effective therapeutics against clinical depression and anxiety [19]. Upregulation of MAO-A has been reported with chronic intermittent hypoxia and oxidative stress leading to neurodegeneration [20]. To the best of our knowledge, no imaging studies (autoradiographic or PET) of MAO-A in AD have been reported.
Imaging studies using PET on the isozyme MAO-B have been reported in AD. Initial studies on MAO-B as an off-target of tau imaging agents and subsequent studies have shown an increase in MAO-B activity in AD [21,22]. Recent PET imaging findings using Neuroinflammation in AD is now regarded as an early indicator of disease [9]. Translocator protein-18 kDa (TSPO) is currently being used as one such biomarker for PET imaging of neuroinflammation [10]. Alternate molecular targets of inflammation other than TSPO are now receiving attention for PET radiotracer development [11]. Monoamine oxidase-A (MAO-A), a pro-oxidative enzyme encoded by the X chromosome and located in the outer mitochondrial membrane and cytosol, has been suggested as a biomarker for activated monocytes and macrophages [12]. MAO-A plays a major role in neurotransmitter degradation in the human brain [13].
MAO-A may play a role in the regulation of neuronal survival in neurodegenerative disorders [14] as well as monoaminergic dysfunction and mitochondrial dysfunction in AD [15]. MAO-A gene polymorphisms have been shown to be associated with neurological changes in AD pathology [16]. An enzymatic assay of MAO-A activity in AD brain tissue showed an increase or no change in several brain regions [17], while another report using similar assay methods indicated a decrease in MAO-A activity in the prefrontal cortex [18]. MAO-A hyperactivity has been shown to be associated with depression, suggesting MAO-A inhibitors as effective therapeutics against clinical depression and anxiety [19]. Upregulation of MAO-A has been reported with chronic intermittent hypoxia and oxidative stress leading to neurodegeneration [20]. To the best of our knowledge, no imaging studies (autoradiographic or PET) of MAO-A in AD have been reported.
Imaging studies using PET on the isozyme MAO-B have been reported in AD. Initial studies on MAO-B as an off-target of tau imaging agents and subsequent studies have shown an increase in MAO-B activity in AD [21,22]. Recent PET imaging findings using the new reversible MAO-B [ 18 F]SMBT-1 suggest an increase in MAO-B activity in the human AD temporal cortex that is concordant with the presence of tau and may suggest inflammatory changes [23]. This increased MAO-B in AD patients using PET imaging is consistent with previous studies which found increased MAO-B activity in AD [17,18].
Using autoradiographic imaging methods, we have evaluated MAO-A activity in postmortem human Parkinson's disease (PD) brain sections of the anterior cingulate using [ 18 F]FAZIN3, a new reversible MAO-A inhibitor [24]. Our ongoing multitarget approach involves comparing Aβ plaque imaging using [ 18 F]flotaza [25], tau imaging using [ 125 I]IPPI [26], and TSPO imaging for microglia using [ 18 [28,29]) for MAO-A imaging. The human anterior cingulate (AC) cortex has been shown to be enriched with MAO-A in PET studies [30], and this region has abundant accumulation of Aβ plaques [31] and tau [32] in AD patients. Microglial activity in the anterior cingulate cortex is being investigated in brain disorders and is therefore a good biomarker to be explored [33]. Because of the presence of these biomarkers in AC and the important role AC plays in cognitive function [34], this brain region was chosen for the present study. The four biomarkers were evaluated in cognitively normal (CN) and AD subjects using autoradiography to assess potential relationships between the biomarkers in AD ( Figure 1).

MAO-A Imaging
is a polyethyleneglycol (PEG) fluoroalkyl azaindole derivative and binds reversibly and selectively to MAO-A [24]. Figure 2 shows images of two AD subjects and two CN subjects from the list in Table 1. Anti-Aβ IHC of the two AD subjects are shown in Figure 2A,C, confirming the presence of abundant Aβ plaques in the gray matter regions of the anterior cingulate. Greater binding of [ 18 F]FAZIN3 was distinctly seen in the anterior cingulate in both AD subjects ( Figure 2B,D), with lower levels in the WM regions.
The average ratio of GM/WM in the six AD subjects was 3.43. Two CN subjects shown in Figure 2E,G also showed more [ 18 F]FAZIN3 binding in the GM compared to that in the WM ( Figure 2F,H). This is consistent with the presence of MAO-A in the human anterior cingulate [30]. The ratio of GM/WM in the six CN subjects averaged 2.17. None of the CN subjects except for CN12-21 ( Figure 2E,F) exhibited a presence of Aβ plaques (Table 1), and this was confirmed using [ 18 F]flotaza and [ 125 I]IBETA [35]. The binding of [ 18 F]FAZIN3 in the GM and WM of all the AD and CN subjects is shown in Figure 2I, and the average of the GM and WM in the two groups of subjects is also shown in Figure 2J. The average [ 18 F]FAZIN3 AD GM-to-CN GM ratio = 2.80, suggesting a significant increase in [ 18 F]FAZIN3 in AD subjects (p < 0.001). When comparing the GM/WM ratios of AD (3.43) to those of CN (1.58), a 58% increase in [ 18 F]FAZIN3 binding was observed, suggesting an increase in MAO-A activity.
It should be noted that subject CN 12-21 was found to exhibit more [ 18 F]FAZIN3 binding compared to other CN subjects, which may be due to the presence of abundant Aβ plaques (Table 1). This subject was not confirmed to have AD. The exclusion of CN 12-21 from the CN group would further increase the percent difference between the AD and CN groups.
The average ratio of GM/WM in the six AD subjects was 3.43. Two CN subjects shown in Figure 2E,G also showed more [ 18 F]FAZIN3 binding in the GM compared to that in the WM ( Figure 2F,H). This is consistent with the presence of MAO-A in the human anterior cingulate [30]. The ratio of GM/WM in the six CN subjects averaged 2.17. None of the CN subjects except for CN12-21 ( Figure 2E,F) exhibited a presence of Aβ plaques (Table 1), and this was confirmed using [ 18 F]flotaza and [ 125 I]IBETA [35]. The binding of [ 18 F]FAZIN3 in the GM and WM of all the AD and CN subjects is shown in Figure 2I, and the average of the GM and WM in the two groups of subjects is also shown in Figure 2J   In order to further ascertain this increase in MAO-A activity in the six AD subjects, we prepared [ 18 F]FEH, a known fluorine-18 analog of [ 11 C]harmine, a MAO-A radiotracer [28,29]. All AD and CN subjects exhibited preferential binding in the anterior cingulate (GM) regions. Increased binding of [ 18 Figure 3I shows the extent of the MAO drug effects on [ 18 F]FAZIN3 binding. All CN subjects exhibited similar drug effects, further confirming that the observed binding of [ 18 F]FAZIN3 is to MAO-A. This MAO-A selectivity of [ 18 F]FAZIN3 was also reported previously [24]. total binding, in the presence of (R)-deprenyl and clorgyline in AD subjects showing no effect of (R)deprenyl while clorgyline displaced >90% of [ 18 F]FAZIN3 ("*** = p < 0.001" for total versus clorgyline; total versus (R)deprenyl was not significant. Autoradiography scale bar: 0 to 700 DLU/mm 2 .

Aβ Plaque Imaging
[ 18 F]Flotaza is a new fluorinated PEG derivative for Aβ plaque PET imaging and provides high-contrast imaging of postmortem AD brains [25]. All AD subjects showed high levels of [ 18 F]flotaza binding in the anterior cingulate regions. Figure 4A shows AD subject 11-38, confirming the presence of extensive Aβ plaques in the anterior cingulate GM regions (shown in the inset). The adjacent brain slice for subject AD 11-38 showed consistent cortical binding of [ 18 F]flotaza corresponding to anti-Aβ ( Figure 4B). Further confirmation of Aβ plaques was achieved by [ 125 I]IBETA, a new Aβ plaque imaging agent shown in Figure 4C   ]FAZIN3 total binding, in the presence of (R)-deprenyl and clorgyline in AD subjects showing no effect of (R)deprenyl while clorgyline displaced >90% of [ 18 F]FAZIN3 ("*** = p < 0.001" for total versus clorgyline; total versus (R)deprenyl was not significant. Autoradiography scale bar: 0 to 700 DLU/mm 2 .

Aβ Plaque Imaging
[ 18 F]Flotaza is a new fluorinated PEG derivative for Aβ plaque PET imaging and provides high-contrast imaging of postmortem AD brains [25]. All AD subjects showed high levels of [ 18 F]flotaza binding in the anterior cingulate regions. Figure 4A shows AD subject 11-38, confirming the presence of extensive Aβ plaques in the anterior cingulate GM regions (shown in the inset). The adjacent brain slice for subject AD 11-38 showed consistent cortical binding of [ 18 F]flotaza corresponding to anti-Aβ ( Figure 4B). Further confirmation of Aβ plaques was achieved by [ 125 I]IBETA, a new Aβ plaque imaging agent shown in Figure 4C

Tau Imaging
We developed [ 125 I]IPPI [26], an analog of [ 18 F]MK-6240, for selective binding to tau, which is useful in autoradiographic studies. More recently, [ 124 I]IPPI was developed as a potential in vivo PET imaging agent for tau [37]. Figure 5A shows anti-tau in AD subject 13-10, confirming the presence of NFT in the anterior cingulate GM regions (shown in the inset and arrow). Figure 5B shows the adjacent brain slice for subject AD 13-10 with binding of [ 124 I]IPPI in the anterior cingulate regions rich in NFT. The GM regions showed significantly higher binding compared to that in the WM ( Figure 5B). All AD subjects showed high levels of [ 125 I]IPPI binding in the anterior cingulate regions, although there was greater variability in the level of binding compared to that in [ 18

Tau Imaging
We developed [ 125 I]IPPI [26], an analog of [ 18 F]MK-6240, for selective binding to tau, which is useful in autoradiographic studies. More recently, [ 124 I]IPPI was developed as a potential in vivo PET imaging agent for tau [37]. Figure 5A shows anti-tau in AD subject 13-10, confirming the presence of NFT in the anterior cingulate GM regions (shown in the inset and arrow). Figure 5B shows the adjacent brain slice for subject AD 13-10 with binding of [ 124 I]IPPI in the anterior cingulate regions rich in NFT. The GM regions showed significantly higher binding compared to that in the WM ( Figure 5B Figure 5D). Spearman's correlation coefficient was also not found to be significant. in all AD subjects but was not significant ( Figure 5D). Spearman's correlation coefficient was also not found to be significant.

TSPO Imaging
Shown in Figure 5E is the brain slice of AD subject 13 Figure 5F). The binding of [ 18 F]FEPPA from the anterior cingulate was displaced by using PK 11195 (10 µM; [23]). Because of the high levels of specific binding of [ 18 F]FEPPA in all the subjects, it may be assumed that the subjects are high-affinity binders. All CN subjects exhibited lower [ 18 F]FEPPA binding in the GM ( Figure 5G) compared to the AD subjects. The increased binding of [ 18 F]FEPPA in the GM of the AD subjects versus that in the CN subjects was significant (p < 0.01; Figure 5G). The average GM/WM ratio was 1.85 for the AD subjects and 1.72 for the CN subjects, suggesting an approximate 11% increase in AD subjects. The ratio of [ 18 F]FEPPA binding in GM alone of the 6 AD subjects and 6 CN subjects was found to be 1.65. The specific binding of [ 18 F]FEPPA and [ 18 F]FAZIN3 exhibited a significant, low positive correlation in the AD subjects, as well as in CN 12-21 ( Figure 5H).
The correlation of the GM/WM binding ratios of the four individual biomarkers in the 6 AD subjects and one CN subject is shown in Figure 6. A positive correlation of MAO-A with Aβ plaque levels was observed with both [ 18 F]flotaza ( Figure 6A) and [ 125 I]IBETA ( Figure 6B) and was found to be significant (p = 0.02 and p = 0.005, respectively). A stronger MAO-A-to-tau linear regression was seen with the GM/WM ratios of [ 125 I]IPPI ( Figure 6C), compared to when only GM binding was used ( Figure 5D). However, in both cases, the correlation between [ 125 I]IPPI and [ 18 F]FAZIN3 was not found to be significant. In the case of TSPO, labeled with [ 18 F]FEPPA, there was a lower but significant correlation that appeared negative ( Figure 6D). When Aβ plaque and tau were correlated together with [ 18 F]FAZIN3, a significant stronger correlation was observed (p = 0.02; Figure 6E). This positive correlation of MAO-A to both Aβ plaque and tau suggests that at least in the anterior cingulate cortex, in this limited number of subjects, the three biomarkers may be positively correlated. However, using one-way ANOVA multiple comparisons, the correlation of MAO-A levels to Aβ plaque levels was highly significant ([ 18 Figure 7 is the comparison of the binding (GM/WM) of the five radiotracers with respect to the Braak stages. As expected, the Aβ and tau radiotracers did not exhibit any binding in Braak stages I and II, except for subject CN12-21, who exhibited significant presence of Aβ plaques but no tau (Supplementary Figure S2). An increase in Aβ is seen in Braak Figure 7 is the comparison of the binding (GM/WM) of the five radiotracers with respect to the Braak stages. As expected, the Aβ and tau radiotracers did not exhibit any binding in Braak stages I and II, except for subject CN12-21, who exhibited significant presence of Aβ plaques but no tau (Supplementary Figure S2). An increase in Aβ is seen in Braak   [ 125 I]IPPI GM/WM ratios >5 in Braak stage VI. There was greater variability in the presence of tau compared to that of the presence Aβ within each of these advanced stages. In the case of [ 18 F]FAZIN3, there was a gradual increase in the GM/WM ratios, with the highest being in Braak stage VI. [ 18 F]FEPPA exhibited the least change in its GM/WM ratio at the different stages.

Discussion
Studies in patients suffering from depression found increased levels of MAO-A radiotracer [ 11 C]harmine in the anterior cingulate and temporal cortex [38]. Our recent findings in the postmortem PD anterior cingulate using [ 18 F]FAZIN3 found significant increases in MAO-A binding [24]. No imaging studies on the status of MAO-A in AD have been reported. Since the cingulate cortex is amongst one of the brain regions affected by both Aβ plaques and NFT [31,32], this study evaluated the anterior cingulate in postmortem AD subjects. Additionally, previous PET imaging reports using the MAO-A radiotracer [ 11 C]harmine confirmed significant levels of MAO-A in the anterior cingulate of healthy human subjects [30], and elevated levels of [ 11 C]harmine binding to MAO-A were found in the anterior cingulate in major depression [38].
At least a 59% increase in [ 18 F]FAZIN3 in AD subjects was observed when comparing ratios of GM/WM in AD and CN subjects. This increase in [ 18 F]FAZIN3 binding was very significant and greater than that reported in depressed patients [38]. This is indicative of an increase in MAO-A in the anterior cingulate cortex in AD (either more MAO-A per mitochondrion or more mitochondria or more dysphoric mitochondria in AD). It should be noted that depression was not a comorbidity in the AD subjects. Based on the selectivity of [ 18 F]FAZIN3 in AD subjects ( Figure 3) and our previous finding in PD subjects [24], this study confirms that the increases reported here are MAO-A and not

Discussion
Studies in patients suffering from depression found increased levels of MAO-A radiotracer [ 11 C]harmine in the anterior cingulate and temporal cortex [38]. Our recent findings in the postmortem PD anterior cingulate using [ 18 F]FAZIN3 found significant increases in MAO-A binding [24]. No imaging studies on the status of MAO-A in AD have been reported. Since the cingulate cortex is amongst one of the brain regions affected by both Aβ plaques and NFT [31,32], this study evaluated the anterior cingulate in postmortem AD subjects. Additionally, previous PET imaging reports using the MAO-A radiotracer [ 11 C]harmine confirmed significant levels of MAO-A in the anterior cingulate of healthy human subjects [30], and elevated levels of [ 11 C]harmine binding to MAO-A were found in the anterior cingulate in major depression [38].
At least a 59% increase in [ 18 F]FAZIN3 in AD subjects was observed when comparing ratios of GM/WM in AD and CN subjects. This increase in [ 18 F]FAZIN3 binding was very significant and greater than that reported in depressed patients [38]. This is indicative of an increase in MAO-A in the anterior cingulate cortex in AD (either more MAO-A per mitochondrion or more mitochondria or more dysphoric mitochondria in AD). It should be noted that depression was not a comorbidity in the AD subjects. Based on the selectivity of [ 18 F]FAZIN3 in AD subjects ( Figure 3) and our previous finding in PD subjects [24], this study confirms that the increases reported here are MAO-A and not MAO-B. Although this is a small study and will need a larger patient sample, our preliminary findings suggest that a 59% increase in [ 18 F]FAZIN3 binding compared to that of the control subjects in a postmortem study may be a sufficient increase to detect changes in MAO-A levels in AD. This will depend, however, on the need to successfully translate the use of [ 18 F]FAZIN3 to in vivo human PET studies for the evaluation of MAO-A in AD.
Increases in [ 18 F]FAZIN3 binding to MAO-A in AD subjects positively correlated with Aβ plaques labeled with [ 18 F]flotaza [25] and [ 125 I]IBETA [35]. It is noteworthy that one CN subject (CN 12-21) with significant Aβ plaques ( Figure 4E) also had higher levels of [ 18 F]FAZIN3 binding compared to the rest of the CN subjects without Aβ plaques ( Figure 2I and Supplementary Figure S2). Since the formation and accumulation of Aβ plaques causes an inflammatory response [39], it is likely that cellular pathways trigger an increase in MAO-A levels [40,41]. This increase in MAO-A levels potentially may influence several detrimental effects, including depleting neurotransmitter levels and inducing further neurodegeneration from oxidative processes. It may be noted that the selective serotonin reuptake inhibitor fluoxetine, a common antidepressant, has been shown to accumulate in the mitochondria in micromolar concentrations and may be contributing to secondary mechanisms for its antidepressant effects [42][43][44].
The presence and spread of NFT in AD is currently being actively pursued as a more accurate biomarker for the staging of the disease [45]. All AD subjects in this study had significant levels of  [35,38,46,47].
The inflammatory response in AD may include changes in the microglia morphology-from ramified (resting) to amoeboid (active)-and astrogliosis (manifested by an increase in the number, size, and motility of astrocytes) surrounding the senile plaques [48]. Microglia activation, a biomarker for inflammation, is characterized by an increased expression of TSPO. Postmortem studies and PET studies of AD have shown significantly elevated TSPO expression in several brain regions [49]. [ 18 F]FEPPA, which has a high affinity for TSPO, has shown promise in PET studies of AD [50]. In the present study, although the AD subjects had a marginal increase in [ 18 F]FEPPA binding compared to the CN subjects, there was a low correlation with [ 18 F]FAZIN3 binding. In contrast, for MAO-B, there appeared to be a better correlation between deprenyl and PK11195 [23]. Since both MAO-B and TSPO are located in glial cells, a stronger correlation of the two biomarkers may be expected. On the other hand, the distribution of MAO-A is different and may be predominantly presynaptic, which could account for the poor correlation of [ 18 F]FAZIN3 binding to MAO-A with [ 18 F]FEPPA binding to TSPO. However, these findings are preliminary, and a larger study will be needed to ascertain this. Additional biomarkers for neuroinflammation will have to be explored for potential correlations with changes in MAO-A in AD [40,41].
Our results suggest that increases in MAO-A levels appear to strongly correlate with Aβ plaques and may serve as a complimentary biomarker for AD. The presence of NFT with SP may further increase the levels of MAO-A. Using measurements of the enzymatic activity of MAO-A in the prefrontal cortex, it has been suggested that changes in MAO-A and B levels occur early in AD and remain at the same levels with increasing duration of the disease [18]. Mitochondrial dysfunction appears to affect both MAO-B and MAO-A [51]. Their potential involvement in the continued accumulation and increase in both SP and NFT loads as well as in the depletion of neurotransmitter function in the AD brain needs further studies. Our assessment in this preliminary study suggests that significant increases in MAO-A levels occur alongside the formation of Aβ plaques and may continue with the formation of NFT. A direct role of TSPO in increasing MAO-A levels was not found in this study.
This is an initial imaging study demonstrating that levels of MAO-A in the AD brain are altered. Limitations of the study include the small number of subjects in advanced stages of AD. A larger study with more subjects at different disease stages is needed to ascertain the correlation of MAO-A changes with the progression of disease. Our study here reports only one brain region; other brain regions, such as the temporal cortex and hippocampus, need to be studied to assess MAO-A changes. It must also be noted that depressive behaviors are often present in AD [52], although our cohort of subjects in this study were not diagnosed with depression. Since MAO-A appears to be upregulated in depression [38], our future studies will also have to evaluate earlier, clinically asymptomatic cases of AD. Finally, attempts are also underway to optimize immunohistochemical staining methods for MAO-A in adjacent brain slices so that the binding of [ 18 F]FAZIN3 may be compared.

Postmortem Human Brain
Human postmortem brain tissue samples were obtained from Banner Sun Health Research Institute (BHRI), Sun City, AZ, USA, brain tissue repository for in vitro experiments. Well-characterized frozen brain samples were obtained from BHRI, Sun City Arizona (Table 1; [53]). Brain tissue samples from AD and cognitively normal (CN) subjects were selected by observing the presence and absence of end-stage pathology. The brain slices contained the anterior cingulate and corpus callosum regions (CN, n = 6; ages 81-90 and AD, n = 6, ages 64-89; Table 1). Brain sections were stored at −80 • C. All postmortem human brain studies were approved by the Institutional Biosafety Committee of University of California, Irvine.

Autoradiography
Brain slices were placed in separate incubation chambers and were allowed to thaw from −80 • C to ambient temperature for 10-15 min. Subsequently, they were preincubated in PBS (pH 7.4) or Tris buffer (pH 7.4) at ambient temperature for 10 min. Fresh PBS buffer (pH 7.4) or Tris buffer (pH 7.4) containing the respective radiotracer was added to all the chambers and incubated for 60-90 min. The brain sections were air-dried, exposed (24 h to 7 days, depending on radioisotope) on phosphor screens, and then placed on the Phosphor Autoradiographic Imaging System (Packard Instruments Co., Boston, MA, USA). Using the Optiquant acquisition and analysis program (Packard Instruments Co., Boston, MA, USA), regions of interest were drawn in the gray matter regions of the anterior cingulate and white matter regions of the corpus callosum. Digital light units/mm 2 (DLU/mm 2 ) were used to quantify the extent of binding. Slides were scanned on a Hewlett Packard scanner to help delineate regions of the brain sections more clearly. At least n = 3 to 6 tissue sections per subject, per radiotracer were used for the study.

MAO-A Imaging
The azaindole derivative [18F]FAZIN3 was prepared in-house, as described previously [24]. Human brain slices containing the anterior cingulate and corpus callosum (10 µm thick) were placed in glass chambers and preincubated in PBS buffer for 10 min. The brain sections (CN and AD) were then incubated with [18F]FAZIN3 (approximately 120 kBq/mL; 1 nM; specific activity >35 GBq/µmol) in PBS at 25 • C for 1 h. The slices were then washed with cold PBS buffer (2 × 5 min) and rinsed with cold deionized water for 2 min. Drug competition studies using 1 µM clorgyline for MAO-A and 1 µM (R)-deprenyl for MAO-B were carried out on all CN and AD brain sections [26]. The brain sections were air-dried and exposed overnight on phosphor screens.
[18F]FEH was in 10% ethanol in sterile saline for in vitro studies. Human brain slices containing the anterior cingulate and corpus callosum (10 µm thick) were placed in a glass chamber and preincubated in PBS buffer for 10 min. The brain sections were placed in a glass chamber and incubated with [18F]FEH (approximately 37 kBq/mL; 1 nM; specific activity >35 GBq/µmol) in PBS at 25 • C for 1 h. The slices were then washed with cold buffer (2 × 5 min) and rinsed with cold deionized water for 2 min. The binding of [18F]FEH to MAO-A was confirmed by the blocking effects of 10 µM clorgyline. The brain sections were air dried and then exposed overnight on a phosphor film.

Tau Imaging
For tau imaging, [125I]IPPI was used for autoradiographic studies [26]. Human brain tissues from the 6 AD and 6 CN subjects were preincubated in PBS buffer for 15 min. After the preincubation buffer was discarded, [125I]IPPI in 10% ethanol PBS buffer with pH 7.4 (60 mL; 3.7 kBq/mL; 0.2-0.5 nM; specific activity >90 GBq/µmol) or [124I]IPPI (6 kBq/mL; 0.2-0.5 nM; specific activity >200 GBq/µmol)) [37] were added to the chambers and incubated at 25 • C for 1.25 h. Nonspecific binding was measured in separate chambers in the presence of 1 µM MK-6240. The slices were then washed with cold PBS buffer for 2 min, 50% ethanolic PBS buffer twice for 2 min each, PBS buffer for 2 min, and cold water for 1 min, respectively. The brain sections were air-dried and then exposed for a week on a phosphor film.

TSPO Imaging
The TSPO PET probe [18F]FEPPA, with >95% radiochemical purity and a specific activity of >70 GBq/µmol (>2 Ci/µmol), was used for autoradiographic studies [54]. Human brain slices from all 6 CN and 6 AD subjects were placed in glass chambers and preincubated in 0.1 M Tris buffer (pH 7.4) for 10 min. Following preincubation, fresh buffer was added to the chambers along with [18F]FEPPA (approximately 37-50 kBq/mL; 0.5-0.75 nM; specific activity >15 GBq/µmol) in 0.1 M Tris buffer, pH 7.4 at 25 • C, and incubated for 1 h. The slices were then washed with cold Tris buffer (2 × 5 min) and rinsed with cold deionized water for 2 min. Nonspecific binding was measured in the presence of 10 µM PK 11195. The brain sections were air-dried and then exposed overnight on phosphor screens.

Immunohistochemistry
Immunostaining of all brain sections was carried out by University of California-Irvine, Pathology services using Ventana BenchMark Ultra protocols. Neighboring slices were immunostained with DAKO polyclonal total tau antibody which detects all 6 six isoforms of tau, dilution 1: 3000, A0024 (Agilent, CA, USA) using reported protocols [55]. For Aβ plaques, slices from all subjects were immunostained with anti-Aβ Biolegend 803015 (Biolegend, CA, USA), which is reactive to amino acid residue 1-16 of β-amyloid [56]. All IHC-stained slides were scanned using the Ventana Roche instrumentation and analyzed using QuPath [37,57].

Image Analysis
All regions of interest (ROI) in the anterior cingulate (GM) and corpus callosum (WM) autoradiographic images of [ 18  F]FEH were quantified using measurements (DLU/mm 2 ). Immunostained sections were analyzed using QuPath. Gray matter (GM) and WM binding of each radiotracer in the AD and CN subjects were measured, and the GM/WM ratios of the AD and CN subjects were compared for each radiotracer. Using the ratio method is akin to in vivo PET methods, where the standard uptake value (SUV) between the target region is compared to a nonspecific binding reference region as a ratio and expressed as SUVR [58,59]. Group differences of GM/WM ratios between AD and CN subjects were evaluated using t-tests. For each radiotracer, the specific binding was calculated by subtracting the WM from the GM of each individual subject. The specific binding of [ 18 F]FAZIN3 was correlated to each specific binding of [ 18 F]flotaza, [ 125 I]IPPI, and [ 18 F]FEPPA in order to assess any relationship between the different biomarkers.

Statistical Analysis
Group differences between AD and CN subjects were assessed using the average GM/WM ratios and were determined using Microsoft Excel 16 and GraphPad Prism 9. The statistical power was determined with Student's t-test, and p values of <0.05 were considered to indicate statistical significance. Spearman's correlation was carried out in certain cases. The linear correlations and ANOVA analysis of the bindings between the different radiotracers were used to evaluate potential relationships between the different biomarkers.

Conclusions
Four different biomarkers, which included Aβ plaques, tau, the translocator protein for microglia, and MAO-A, were used to study the anterior cingulate cortex of wellcharacterized AD subjects. Our results showed that significantly greater MAO-A activity was present in the anterior cingulate of AD subjects compared to the cognitively normal controls. This increased MAO-A in AD subjects was positively correlated to Aβ plaque and tau levels. Early assessment of increased MAO-A levels using PET imaging in AD patients may provide opportunities for therapeutic interventions by MAO drugs [60]. This may provide neuroprotection and slow down AD progression. Increased MAO-A levels did not correlate with the expression of the translocator protein. A larger study with more subjects at different disease stages is needed in order to ascertain the correlation of MAO-A imaging with Aβ plaque and tau levels. Other brain regions such as the temporal cortex and hippocampus need to be studied to assess MAO-A increases.