1
Reduced miR-512 and the elevated expression of its anti-apoptotic targets cFLIP and MCL1 localize to dysfunctional neurons in Alzheimer’s disease
Louisa Mezache1 Madison Mikhail1 Michela Garofalo2 Gerard J. Nuovo 1,2
Phylogeny Inc.1, OSU Comprehensive Cancer Center2, Columbus Ohio
2 Direct all correspondence to: Gerard J Nuovo, MD 1476 Manning Parkway Powell, Ohio 43065 email – nuovo.1@osu.edu; telephone – 614 519 8890; FAX - 877 591-1815 Running title: Anti-apoptosis and Alzheimer's disease
3
ABSTRACT The cause for the neurofibrillary tangles in Alzheimer's disease is assumed to be an abnormal increased rate of production of their key neurotoxic components which include beta amyloid and hyperphosphorylated tau protein. We segregated Alzheimer’s brain sections into those with abundant hyperphosphorylated tau protein and those without and compared each to normal brains for global microRNA patterns. A significant reduced expression of several microRNAs, including miR-512, was evident in the Alzheimer’s brain sections with abundant hyperphosphorylated tau. Immunohistochemistry documented that two known targets of microRNA-512 that can each inhibit apoptosis, cFLIP and MCL1, were significantly over expressed and each co-localized to neurons with the abnormal tau protein. Analysis for apoptosis including activated caspase-3, increased caspase 4 and 8, apoptosis initiating factor, APAF-1 activity, and the TUNEL assay was negative in the areas where neurons showed hyperphosphorylated tau. MCM2 expression, a marker of neuroprogenitor cells, was significantly reduced in the Alzheimer’s sections that contained the hyperphosphorylated tau. These results suggest that a basic defect in Alzheimer’s disease may be the microRNA-driven increased expression of anti-apoptotic proteins that prevent the normal turnover of neurons. The abnormal longevity of these neurons could, in turn, lead to the accumulation of key Alzheimer’s proteins such as hyperphosphorylated tau that ultimately prevent normal neuronal function and lead to disease symptomatology.
Key words: microRNA, tau, hyperphosphorylated, cFLIP, Mcl1
4 INTRODUCTION A primary question in Alzheimer's research is whether the larger pyramidal neurons with the neurofibrillary tangles have an increased rate of accumulation of abnormal proteins such as beta amyloid and hyperphosphorylated tau protein that, in turn, triggers their apoptotic death (1-5)? Various investigators have noted that classic apoptotic proteins such as caspase-3, caspase-8, and apoptosis initiating factor can be found either directly in the neurons with neurofibrillary tangles or in the microglial cells that surround them (6-10). Others have theorized that the increased hyperphosphorylated tau may reflect enhanced activity of many of the tyrosine kinases known to function in the brain, such as p38, pERK, and Jun which, in turn, may explain why neurons in Alzheimer’s disease apparently have increased synthesis of this neurotoxic protein (11-13). MicroRNAs play a key role in the normal physiology and pathophysiology of the brain although there does not appear to be a strong consensus regarding their dysregulation in Alzheimer's disease among different research groups (14). This may be because the brain shows an extreme level of compartmentalization with groups of neurons involved in a given function typically tightly bundled in fascicles. Thus, biochemically abnormal groups of neurons can co-exist with adjacent groups of normal functioning neurons. The central nervous tissue undergoes a normal turnover of cells called programmed neuronal death whereby neurons are replaced on a regular basis by neuron progenitor cells (15). Neuronal turnover by programmed cell death may be important for normal brain function as it may prevent the accumulation of neurotoxic chemicals. The rate of neuronal turnover in the brain is not known but has been shown to be about 3% per month in the olfactory bulb of adult mice (16). Neuronal programmed cell death and generation of new neurons in adults has also been documented in the human olfactory bulb (17). MCM2 is a phenotypic marker of neuronal progenitor cells (18). Immunohistochemistry provides a thorough assessment of the extent of apoptosis in a given tissue as it can readily detect proteins associated with the intrinsic apoptotic pathway such as APAF-1, apoptosis initiating factor, and activated caspase 3 as well as caspase-8 which catalyzes the extrinsic as
5 well as intrinsic pathways via the death initiating signaling complex (19). In either apoptotic pathway the end result is nuclear DNA that is extensively nicked and, thus, readily detected by the TUNEL assay. Since apoptosis is invariably countered by anti-apoptotic proteins such as bcl-2, cFLIP, and MCL1, and since these proteins are also readily detected in situ, one can molecularly dissect whether apoptosis or “anti-apoptosis” is dominating in a given tissue (19). The purpose of this paper is to provide evidence that the basic defect in Alzheimer's disease is the loss of the normal neuronal cell turnover due to the predominance of anti-apoptotic factors which, in turn, are present due to dysregulation of microRNA expression. It is thus theorized that it is the abnormal longevity of the affected neurons that is the primary factor responsible for the accumulation and appearance of the hyperphosphorylated tau, beta amyloid, and the neurofibrillary tangles that allow the pathologic diagnosis of Alzheimer’s disease and ultimately are responsible for the clinical symptomatology.
6
METHODS Patient samples The patient samples (formalin fixed paraffin embedded excisional autopsy material) originated from the files of Folio Biosciences. Clinical information included the diagnosis, age/sex of patient, extent of disease, and the clinical workup to rule out other possible causes of dementia. Thirteen tissues of Alzheimer's disease from ten patients (mean age 83.9 years) were identified. Negative controls included thirteen tissues from ten age matched controls (mean age 85.0) where there was no clinical evidence of Alzheimer's disease. In each case, large (2.0 cm) full thickness sections of the frontal cerebral cortex and hippocampus were available for analysis. We also studied five TMAs that each consisted of four cores of cortical/hippocampal sections from Alzheimer's disease patients and four aged matched controls (thus, twenty normal and twenty Alzheimer tissues). Each case of Alzheimer's disease and each negative control was confirmed by a Bielschowsky stain. Immunohistochemical analysis Our immunohistochemical protocol has been previously described and uses the Leica BOND-MAX (Leica Biosystems, Buffalo Grove, IL) (19). Molecular analysis began with immunohistochemistry for hyperphosphorylated tau using an antibody from ABCAM (phospho S404) that is specific for end-stage hyperphosphorylation (20). The initial analyses were blinded to the clinical diagnosis. After “unblinding� it was clear that each of the controls had from 0 to 0.5 tau hyperphosphorylated positive neurons per 4 mm2. The Alzheimer's disease tissues showed marked variation in hyperphosphorylated tau expression that fell into two disparate categories: rare positive neurons with a mean number of 1.1/4mm2 and clusters of many positive
7 neurons in a given tissue with a mean number of positive cells 36.3/4mm2. We thus segregated the Alzheimer's disease tissues into these two groups which will be referred to as AD-tauP(which consisted of seven tissues) and AD-tauP+ (which consisted of the remaining six tissues), respectively. With regards to the twenty cores from Alzheimer's disease cases, ten showed many neurons with hyperphosphorylated tau and the other ten showed zero to 2 positive cells per 4 mm core. The AD-tauP- and + cases and controls were tested for the following tyrosine protein kinases that can hyperphosphorylate tau: p38, pERK-1 and 2, Akt, Jun, RAS, and cMYC plus TrkC which can regulate the protein kinases. The cases and controls were also tested for the apoptotic proteins caspase 8, activated caspase 3, caspase 4, APAF-1, and apoptosis initiating factor. Similarly, the cases and controls were tested by immunohistochemistry for the mitochondrial protein pyruvate dehydrogenase that is strongly expressed in viable neurons, and the antiapoptotic proteins cFLIP, and MCL1. Optimization of the antibodies which were all obtained from ABCAM except for cFLIP, NSE, and RAS which were obtained from Enzo Life Sciences for immunohistochemistry was as previously described (19). Co-expression analysis Our co-expression analyses protocol has been previously reported (19). The computerbased analysis by the Nuance system (Caliper Life Sciences, Hopkinton, MA) separates each chromogenic spectral signal, converts it to a fluorescent signal, then mixes the two and indicates if cells contain the two targets of interest. Global microRNA analysis Global microRNA analysis was compared for the RNA extracted from the normal, the AD-tau-, and AD-tau+ brain sections which was done as previously described using the Qiagen
8 RNA extraction kit (19). The NanoString nCounter Human miRNA Expression Assay Kit (http://www.nanostring.com/) was used to profile more than 700 human and human-associated viral miRNAs as previously described (19). Hybridized probes were purified using the nCounter Prep Station (NanoString Technologies) to remove excess capture and reporter probes and to immobilize transcript-specific ternary complexes on a streptavidin-coated cartridge. Data collection was carried out on the nCounter Digital Analyzer (NanoString Technologies) following the manufacturer’s instructions to count individual fluorescent barcodes and quantify target RNA molecules present in each sample. In situ hybridization analyses Our microRNA in situ hybridization protocol has been previously described (19). In brief, in situ hybridization was performed with LNA modified and 5’ digoxigenin tagged probes specific for the miRNA-512, and 765. The probe/target complex was visualized after the alkaline phosphatase-linked conjugate reacted with the chromogen, nitroblue tetrazolium and bromochloroindolyl phosphate (NBT/BCIP) with a nuclear fast red counterstain. Negative controls included omission of the probe and the use of a scrambled probe. The TUNEL assay was performed using the Roche in situ cell apoptosis kit as previously described (19).
9
RESULTS In the initial part of the study, the distribution of neurons with neurofibrillary tangles, beta amyloid precursor protein, and hyperphosphorylated tau was determined. The control tissues showed very rare to no neurons with hyperphosphorylated tau, abnormal aggregates of beta amyloid precursor protein, or neurofibrillary tangles using the Bielschowsky stain. The Alzheimer's disease cases showed a strikingly variable distribution of hyperphosphorylated tau. This allowed us to segregate the Alzheimer's disease sections into two groups: AD-tauP- with rare positive neurons (mean 1.1/4mm2 with SEM 0.1 - seven tissues) and AD-tauP+ with many positive neurons (mean 36.3/4mm2 with SEM 9.1 - six tissues, p <0.001) that were primarily in levels III and V of the cortical gray matter (Figures 1 A-C). The neurons expressing abundant hyperphosphorylated tau showed a strong tendency to cluster into a fascicular-type pattern (Figure 1C). Interestingly, some cases had blocks that were scored as AD-tauP- and other blocks scored as ADtauP+. Co-expression analysis indeed confirmed that the neurons positive for hyperphosphorylated tau also expressed aggregates of beta amyloid precursor protein and showed neurofibrillary tangles (Figure 1D-F). We then extracted the total RNA from the three sets of brain sections (controls, AD-tauP, AD-tauP+) and did global microRNA analysis. No microRNA was up-regulated in the ADtauP+ group compared to the other two groups. We noted that microRNAs 512, 765, 1181, and 1292 were significantly decreased in the tissues with AD-tau-P+ when compared to the control and AD-tauP- groups (Supplemental Figure 1). TargetSpan Version 6.2 analysis demonstrated that reduced microRNA 512 has been verified to be able to increase expression of two antiapoptotic proteins (cFLIP and MCL1) and that decreased miR-765 expression can lead to in-
10 creased expression of the tyrosine-protein kinase receptor TrkC (also known as neurotrophic tyrosine kinase, receptor, type 3) (21,22), No validated targets have been reported yet for miRNAs 1181 or 1292. Given the possible roles of anti-apoptotic and tyrosine protein kinase proteins as suggested by the global microRNA analysis, and the putative role of apoptosis and phosphorylation of tau in Alzheimer's disease, we next examined if there was evidence of apoptosis and/or enhanced protein kinase activity in the AD-tauP+ sections by comparing the distribution of each protein to serial sections analyzed for hyperphosphorylated tau. A primary assumption was that any protein directly correlated with the neurons with hyperphosphorylated tau would show minimal expression in the normal brains or AD-tauP- tissue sections and enhanced expression that paralleled the neurons showing hyperphosphorylated tau. Immunohistochemical analysis of pERK 1 and 2, p38, RAS, JUN, and AKT, which each can phosphorylate tau protein (11-13), showed that each was expressed in normal brain where they localized primarily to neurons and endothelial cells with the exception of JUN which was present mostly in cells with the cytologic features of astrocytes and oligodendroglial cells. None of these proteins mirrored the expression pattern of hyperphosphorylated tau and none showed increased number of positive cells in the AD-tauP+ group compared to the other two groups (data not shown). Co-expression analysis for pERK-1 and 2, p38, and RAS with hyperphosphorylated tau confirmed minimal co-expression (data not shown). We next studied the distribution of the two anti-apoptotic proteins cFLIP and MCL1 in the control and Alzheimer's disease brain sections. There was a statistically significant increase in the neuronal expression of each protein in the AD-tauP+ sections compared to both the control and AD-tauP- sections (Table 1). Also note that each protein was also significantly increased
11 when comparing the AD-tauP- and the control tissues. As evident in Figure 2, the pattern of distribution of these two anti-apoptotic proteins closely mirrored hyperphosphorylated tau (panel A, control, MCL1, panel B-D AD-tauP+ serial sections analyzed for MCL1, cFLIP, and hyperphosphorylated tau, respectively). Co-expression analysis confirmed that over 90% of cells with hyperphosphorylated tau were also positive for MCL1 (Figure 2E) and cFLIP (Figure 2F). .
Next activity of key components of the intrinsic apoptotic system were examined includ-
ing caspase 3, 4, APAF-1, and apoptosis initiating factor as well as the key coordinating molecule of both apoptotic pathways, caspase 8. Since each apoptotic pathway would be expected to lead to nuclear fragmentation, the TUNEL assay was also performed. Analysis of the control brain tissues and twoAlzheimer's disease groups showed equivalent distribution of APAF-1 (rare neurons and scattered endothelial cells) and apoptosis initiating factor (cytoplasmic localization in many neurons in each of the three groups with no evidence of nuclear translocation). The control brain tissues each showed expression of caspase-3, 4, and 8. The cells expressing each caspase had the cytologic features of astrocytes and microglial cells with the processes at times directly communicating with adjacent neurons (Figure 3B). The number of positive cells expressing caspase-3, 4, and 8 and the distribution of the cells was equivalent in the brain sections of the control and the AD-tauP- as well as the AD-tauP+ tissues. Figure 3B shows caspase-8 expression in an AD-tauP+ brain. Since cFLIP (also known as CASPASE8 and FADD like apoptosis regulator protein) and MCL1 can each inhibit caspase-8 activation (23-25), we did co-expression of caspase-8 and these proteins. As seen in Figure 3 CE, cFLIP and MCL1 each strongly co-expressed with the caspase-8 in the AD-tauP+ sections. Next the histologic distribution of miR-512 was examined since it has been shown to negatively regulate cFLIP and MCL1 (21,22). miR-512 was robustly expressed in the control brain
12 sections where it was neuron specific and evident in the cytoplasm of over 90% of these cells (Figure 3F NL). The insert shows the loss of the signal with a scrambled probe. In comparison, there was a dramatic decrease in miR-512 expression in the AD-tauP+ sections and it was virtually absent in the areas with strong hyperphosphorylated tau expression (panel F-AZ). The insert shows that miR-125b, which was not altered in the Alzheimerâ&#x20AC;&#x2122;s tissue sections, was strongly expressed in these sections and co-expressed with hyperphosphorylated tau. Interestingly, miR512 expression was also reduced in the AD-tauP- sections compared to the controls although not to the same degree as with the AD-tauP+ sections (data not shown). In comparison, miRNA-765 was also significantly decreased in the AD-tauP+ sections. This microRNA was evident in all cell types in the control brains (supplemental Figure 2A). Its expression was not altered in the AD-tauP- tissues (data not shown) but was much decreased in the AD-tauP+ tissues. However, as seen in supplemental Figure 2B, the loss of signal in the AD-tauP+ tissues was in the white matter alone; miR-765 was robustly expressed in the neurons with hyperphosphorylated tau (supplemental Figure 2C, D compared to miRNA-512 in supplemental Figure 2E,F). Reduced miR-765 expression has been shown to lead to increased TrkC expression. However, we noted that TrkC expression actually was reduced in the neurons in the Alzheimerâ&#x20AC;&#x2122;s tissue sections (Table 1) suggesting that in the brain it was not being regulated by miR-765. Finally, in situ hybridization analysis of miR-1181 and -1292 showed results equivalent to miR-512 with ample expression in neurons in normal brain and the loss of expression in the areas of the brain where there was hyperphosphorylated tau protein (data not shown). The TUNEL assay corroborated the immunohistochemistry based conclusion that apoptosis was not operative in the neurons with hyperphosphorylated tau. Specifically, the multiple myeloma cell line U266 treated with reovirus or reovirus plus ammonia demonstrated the speci-
13 ficity of the assay since this virus can dramatically enhance apoptosis by its ability to directly up regulate caspase 3 (19) (Figure 4 A+virus) which in turn is blocked by ammonia (Figure A NH3+virus). TUNEL analysis of the normal brain sections showed rare positive neurons that numbered less than ten per 2.0 cm section (Figure 4B) which compared to an approximate doubling of neuronal apoptosis in five viral encephalitis cases (Figure 4V insert) and a dramatic increased apoptosis in glioblastoma multiforme (data not shown). The serial sections that showed that neurons with hyperphosphorylated tau were negative for the TUNEL assay (Figure 4C). For further evidence that the neurons with the classic pathologic features of Alzheimer's disease were in an â&#x20AC;&#x153;anti-apoptotic stateâ&#x20AC;?, we did co-expression analysis for the key mitochondrial protein pyruvate dehydrogenase and hyperphosphorylated tau. These data showed strong co-expression of these two proteins (Figure 4F) and equivalent expression of pyruvate dehydrogenase in neurons in the control brain sections (Figure 4F) insert. The data strongly suggested that the basic defect in the dysfunctional Alzheimer's disease was the domination of anti-apoptosis that could lead to increased longevity of the neurons. The putative decreased neuronal cell turnover could in turn reduce the activity of the neural progenitor cells. We thus compared the distribution of the neural progenitor cell marker MCM2 in the three groups. As seen in Table 1, there indeed was a significant decrease in the neural progenitor cells in the AD-tauP+ sections when compared to the other groups (Table 1), consistent with decreased turnover of the neurons (Figure 4D control and 4E, AD-tauP+ tissue).
14
DISCUSSION A central issue in understanding of Alzheimer's disease is the biologic basis of the accumulation of the abnormal proteins, such as beta amyloid and hyperphosphorylated tau, that mark the dysfunctional neurons. It has been theorized that in Alzheimerâ&#x20AC;&#x2122;s disease these proteins accumulate at an increased rate in neurons and lead to their apoptotic death (1-5). The primary finding in this study is to suggest that the pathophysiology of Alzheimer's disease may center on an abnormal longevity of the dysfunctional neurons due to, in part, the dysregulation of microRNA expression that, in turn, induces increased expression of anti-apoptotic proteins such as cFLIP and MCL1. The increased longevity of the neurons so affected would allow the accumulation of proteins such as hyperphosphorylated tau and amyloid beta. The tau protein plays a key role in microtubular stabilization in neurons and, thus, is essential in the trafficking of neurotransmitters (20). The neuronal isoforms of tau protein have about 80 serine and threonine phosphorylation sites. It has been documented that phosphorylation of less than 50% of these sites invariably in present in normal brain and that increased phosphorylation at specific sites, including phosphor S404, is associated with the loss of function of tau (20). Hyperphosphorylated tau was basically absent in normal brain and very rare in some cerebral sections of Alzheimer's disease, even though adjoining areas showed many positive cells that clustered in a fascicular fashion. Two important implications of these observations were: 1) it suggests that hyperphosphorylated tau may be a more accurate immunohistochemical indicator of Alzheimer's disease since beta amyloid protein is normally present in the brain and 2) it underscores the importance when studying Alzheimer's disease to recognize that the molecular changes are not homogenous but can vary markedly between samples from the same person and
15 inside a given tissue block. By segregating the Alzheimerâ&#x20AC;&#x2122;s disease brain tissues into AD-tauPand AD-tauP+, we were able to document down-regulation of certain miRNAs that correlated strongly with hyperphosphorylated tau. This, in turn, allowed the demonstration that several anti-apoptotic proteins were also significantly correlated with hyperphosphorylated tau expression. The death-inducing signaling complex, which is formed via the aggregation of FasR after binding to its ligand FasL, induces apoptosis via the widespread activation of both the intrinsic and extrinsic pathways that is activated by caspase 8 and effected by caspase 3 (23). cFLIP can directly interact with the death inducing signaling complex and inhibit caspase 8 activity; MCl1 also can directly appose caspase-8 activity (24,25). Each protein has been shown to strongly inhibit apoptosis in murine models of neuronal apoptosis (24,25). As shown in this study and corroborated by others MCL1 is not made by neurons in normal brain and cFLIP is minimally expressed (24,25). Each protein showed a marked increase in the Alzheimer's disease brains with hyperphosphorylated tau and each strongly co-expressed with this abnormal protein. It is theorized that the presence of these proteins in the neurons was indicative of increased longevity of these cells since they effectively block the apoptotic pathways that normally would lead to their slow but programmed death and orderly replacement by neural progenitor cells. Although this theory will certainly require more study, the observations in this study that there was a significant decrease in the MCM2 positive neural progenitor cells in the AD-tauP+ tissues, that caspase-8 co-expressed with the anti-apoptotic proteins, and that there was no evidence of increased activity of either the intrinsic or extrinsic apoptotic pathways in the neurons with hyperphosphorylated tau all support this hypothesis. Interestingly, both cFLIP and MCL1 were also significantly increased in the AD-tauP- tissues though not to as strong a degree as with the AD-tauP+ tissues. It is possible that the inhibition of neuronal turnover from the anti-apoptotic state may
16 be occurring early in the Alzheimerâ&#x20AC;&#x2122;s disease process and that the cells with neurofibrillary tangles, beta amyloid, and hyperphosphorylated tau represent the end stage of this process after perhaps years of added longevity of the affected neurons. In summary, the data in this manuscript suggests that the neuronal dysfunction in Alzheimer's disease may be based in the miRNA directed increased expression of anti-apoptotic proteins such as cFLIP and MCL1 that prevent the normal and necessary programmed death of neurons. The so-affected neurons may live for many years past the time they were supposed to undergo programmed cell death. This allows the accumulation of abnormal proteins such as hyperphosphorylated tau and amyloid beta that ultimately prevent the cell from functioning. The slow process of neuronal cell turnover and accumulation of these neurotoxic molecules could explain in part why Alzheimerâ&#x20AC;&#x2122;s disease typically affects the elderly. This hypothesis suggests that treatments directed towards restoring the involved microRNAs and/or inhibiting the antiapoptotic proteins may be able to prevent the extensive neuronal dysfunction that is at the epicenter of Alzheimer's diseases.
17
Disclosure/Conflict of interest. The authors have no conflict of interests to report. Acknowledgements. We appreciate the assistance of Dr. Margaret Nuovo who did the photomicroscopy/digital imaging and the assistance with reagents from Phylogeny.
18
REFERENCES
1. Aguzzi A, Barres BA, Bennett ML. Microglia: scapegoat, saboteur, or something else? Science 2013; 339(6116):156-61. 2. Jucker M, Walker LC. Self-propagation of pathogenic protein aggregates in neurodegenerative diseases. Nature 2013; 501(7465):45-51. 3. Roberson ED, Mucke L. 100 years and counting: prospects for defeating Alzheimer's disease. Science 2006; 314(5800):781-4. 4. Swarup V, Geschwind DH. Alzheimerâ&#x20AC;&#x2122;s diseease: from big data to mechanism. Nature 2013; 500: 34-35. 5. Chu CT, Caruso JL, Cummings TJ, Ervin J, Rosenberg C, Hulette CM. Ubiquitin immunochemistry as a diagnostic aid for community pathologists evaluating patients who have dementia. Mod Pathol 2000; 13(4):420-6. 6. Burguillos MA, Deierborg T, Kavanagh E, Persson A, Hajji N, Garcia-Quintanilla A, et al. Caspase signalling controls microglia activation and neurotoxicity. Nature 2011; 472(7343):319-24. 7. Engidawork E, Gulesserian T, Yoo BC, Cairns N, Lubec G. Alteration of caspases and apoptosis-related proteins in brains of patients with Alzheimer's disease. Biochem Biophys Res Commun 2001; 281(1):84-93. 8. Rohn TT, Head E, Nesse WH, Cotman CW, Cribbs DH Activation of caspase-8 in the Alzheimer's disease brain. Neurobiol Dis 2001; 8(6):1006-16.
19 9. Su JH, Kesslak JP, Head E, Cotman CW. Caspase-cleaved amyloid precursor protein and activated caspase-3 are co-localized in the granules of granulovacuolar degeneration in Alzheimer's disease and Down's syndrome brain. Acta Neuropathol 2002; 104(1):1-6. 10. Vaisid T, Barnoy S, Kosower NS. Calpain activates caspase-8 in neuron-like differentiated PC12 cells via the amyloid-beta-peptide and CD95 pathways. Int J Biochem Cell Biol 2009; 41(12):2450-8. 11. Anderson AJ, Suh JH, Cotman CW. DNA damage and apoptosis in Alzheimer’s disease: Colocalization with c-JUN immunohistochemistry, relationship to brain area, and effect of postmortem delay. J Neuroscience 1996; 16: 1710-9. 12. Hensley K, Floyd RA, Zheng NY. p38 kinase is activated in the Alzheimer’s disease brain. J Neurochem 1999; 155:2053-7. 13. Pei JJ, Braak H, An WL. Up-regulation of mitogen-activated protein kinases ERK1/2 and MEK1/2 is associated with the progression of neurofibrillary degeneration in Alzheimer’s disease. Molec Brain Res 109; 2002: 45-55. 14. Satoh JI. microRNAs and their therapeutic potential for human diseases: Aberrant microRNA expression in Alzheimer’s disease brains. J Pharmacological Sciences 2010, 114:269-74. 15. Manganas LN, Zhang X, Li Y, Hazel RD, Smith SD, Wagshul ME, Henn F, et al. Magnetic resonance spectroscopy identifies neural progenitor cells in the live human brain. Science 2007; 318(5852):980-5. 16. Mizrahi A, Lu J, Irving R, Feng G, Katz LC. In vivo imaging of juxtaglomerular neuron turnover in the mouse olfactory bulb. Proc Natl Acad Sci USA, 2006; 103:1912-7.
20 17. Bedard A, Parent A. Evidence of newly generated neurons in the human olfactory bulb. Develop Brain Res 2004; 151:159-68. 18. Thom M, Martinian L, Williams G, Stoeber K, Sisodiya SM. Cell proliferation and granule cell dispersion in human hippocampal sclerosis. J Neuropathol Exp Neurol 2005; 64:194-201. 19. Nuovo GJ, Garofalo M, Valeri N, Roulstone V, Volinia S, Cohn DE, et al Reovirusassociated reduction of microRNA-let-7d is related to the increased apoptotic death of cancer cells in clinical samples. Mod Pathol (2012 Oct) 25(10):1333-44. 20. Augustinack JC, Schneider EM, Mandelkow B, Hyman T. Specific tau phosphorylation sites correlate with severity of neuronal cytopathology in Alzheimerâ&#x20AC;&#x2122;s disease. Acta Neuropathol 2002; 103: 26-35. 21. Chen F, Zhu HH, Zhou LF, Wu SS, Wang J, Chen Z Inhibition of c-FLIP expression by miR-512-3p contributes to taxol- induced apoptosis in hepatocellular carcinoma cells. Oncol Rep 2010; 23(5):1457-62. 22. Saito Y, Suzuki H, Tsugawa H, Nakagawa I, Matsurzaki J, Kanai Y, Hibi T. Chromatin remodeling at Alu repeats by epigenetic treatment activates silenced microRNA-512-5p with downregulation of Mcl-1 in human gastric cancer cells. Oncogene 2009, 28: 273844. 23. Kishckel FC, Hellbardt S, Behrmann I, Germer M, Pawlita M, Krammer PH, Peter ME. Cytotoxicity-dependent APO-1(Fas/CD95) associated proteins form a death inducing signaling complex (DISC) with the receptor. The EMBO J 1995; 14:5579-88.
21 24. Mori M, Burgess DL, Gefrides LA, Foreman PJ, Opferman JT. Expression of apoptosis inhibitor protein Mcl1 linked to neuroprotection in CNS neurons. Cell Death Differ 2004; 11:1223-3. 25. Matsumori Y, Northington FJ, Hong SM, Kayama T, Sheldon RA, Vexler ZS, Ferriero DM, et al. Reduction of caspase-8 and -9 cleavage is associated with increased c- FLIP and increased binding of Apaf-1 and Hsp70 after neonatal hypoxic/ischemic injury in mice overexpressing Hsp70. Stroke 2006; 37(2):507-12.
22
Table 1. Compilation of expression of various proteins with tau hyperphosphorylation in Alzheimerâ&#x20AC;&#x2122;s disease. Category
Mean*
SEM**
p value***
Control AD-tauPAD-tauP+
0.5 1.1 36.3
0.1 0.3 9.1
<0.001
MCL1 Control AD-tauPAD-tauP+
1.8 28.4 52.1
0.7 3.9 6.9
<0.001 0.024
Control AD-tauPAD-tauP+
17.5 39.6 64.3
2.8 6.3 8.8
0.005 0.023
MCM2 Control AD-tauPAD-tauP+
18.1 8.5 2.7
1.5 1.1 0.7
<0.001 <0.001
TRKC Control AD-tauPAD-tauP+
115.4 49.0 27.5
7.0 3.1 2.2
<0.001 <0.001
HYPERPHOSPHORYATED TAU
CFLIP
* Number of positive cells per 4 mm2 **Standard error of the mean *** p value comparison is between Control and AD-tauP- plus AD-tau-P- and AD-tauP+.
23 Figure legends Figure 1. Histologic distribution of hyperphosphorylated tau protein. Panel A shows absence of hyperphosphorylated tau protein in normal brain. Panels B and C are from the same case of Alzheimer’s disease and demonstrate how large areas of the gray matter can be negative for hyperphosphorylated tau (panel B) whereas adjacent areas can show many neurons positive for the protein in a fascicular type pattern (panel C; signal is brown and counterstain blue). A higher magnification (panel D) shows that hyperphosphorylated tau localizes primarily to pyramidal neurons which also contain beta amyloid precursor protein (panel E, serial section). Coexpression of hyperphosphorylated tau and beta amyloid precursor protein confirms the strong co-localization (seen as fluorescent yellow). The scale bar for each image is 100 microns. Figure 2. The anti-apoptotic proteins MCL1 and cFLIP co-localize with hyperphosphorylated tau. Panel A demonstrates the absence of MCL1 protein in the control brain. Panels B-D shows serial sections from a case of Alzheimer’s where the same groups of neurons express MCL1 (panel B), cFLIP (panel C) and hyperphosphorylated tau (panel D; signal is brown and counterstain blue). anels C and D are serial sections showing cFLIP and hyperphosphorylated tau protein. Co-expression of hyperphosphorylated tau and MCL1 (panel E) and cFLIP (panel F) confirms the strong co-localization (seen as fluorescent yellow). Figure 3. Distribution of caspase8 and its regulators, cFLIP and MCL1, in Alzheimer’s disease. Panels A and B demonstrate that caspase 8 expression in normal brain and in the AD-tauP+ brain tissues shows an equivalent distribution. Note that the signal is primarily in the processes of supporting cells which also can reach neurons. Panel C shows the routine microscopy image after co-expression with caspase-8 and MCL1 whereas panel D is the Nuance computer based im-
24 age that demonstrates the strong co-expression between these two proteins seen as fluorescent yellow. A similar strong co-localization can be seen in the serial section between cFLIP and caspase 8 (panel E). Panel F-NL shows the strong expression of miRNA-512 in the control brain (blue is signal and pink is counterstain). Note that the signal is neuron specific and is lost with a scrambled probe (insert). Panel F-AZ shows the marked reduction of miRNA-512 expression in the AD-tauP+ brain in an area with many neurons positive for hyperphosphorylated tau. Note that these cells show a strong signal for miRNA-125b (blue) which co-localizes with hyperphosphorylated tau (brown, insert). Figure 4. Demonstration of the anti-apoptotic state and loss of neuroprogenitor cells in the ADtauP+ tissues. Panel A- shows the rare apoptosis in U266 myeloma cells without reovirus that is markedly increased after viral infection (panel A+virus) and again lost with pre incubation with ammonia (panel A NH3+ virus; the signal is blue and the counterstain pink using the TUNEL assay). The TUNEL assay does detect rare apoptotic cells in the control brain (panel B) which are more numerous in fatal viral encephalitis (insert, enteroviral encephalitis). Apoptotic cells were not evident in the AD-tauP+ brain in the areas with hyperphosphorylated tau (panel C). MCM2, a marker of neuroprogenitor cells, is positive in scattered cells in the control brain (panel D) whereas the signal is substantially reduced in the AD-tauP+ brain (panel F, signal is brown with blue counterstain). in normal brain with insert showing viral encephalitis. Panel C shows no TUNEL + in Alzheimerâ&#x20AC;&#x2122;s brain. Panel F shows strong co-expression of pyruvate dehydrogenase (fluorescent red) and hyperphosphorylated tau (fluorescent green) in AD-tauP+ brain that co-expresses with hyperphosphorylated tau protein (fluoresecent yellow) indicative of antiapoptic state. The insert shows that the signal for pyruvate dehydrogenase is equivalent in neurons from the control brains.
25
Supplemental Figure 1. Representative heat map for the global microRNA analysis. This heat map compares the expression of a series of microRNAs in Group 1 (control brain sections) and Group 3 (AD-tauP+ brain sections). Note the reduced expression of miR-512-5p in the latter group. Supplemental Figure 2. Correlation of microRNA and hyperphosphorylated tau expression in the AD-tauP+ brain sections. Both miRNA-765 and -512 showed significantly reduced expression in the AD-tauP+ brain sections. Panel A shows the marked miR-765 expression in normal brain sections present in both the gray (right side of image) and white matter (left side of image). Note that the neuron based signal is still strong in the AD-tauP+ brain tissue while the white matter signal is much reduced (panel B). Co-expression of miR-765 (blue) and hyperphosphorylated tau (red) demonstrates that the neurons with the abnormal tau protein also express miR-765 (fluorescent yellow, panels C and D). In comparison, the co-expression analysis of the serial section for the abnormal tau protein and miR-512 shows that the cells with hyperphosphorylated tau do not express miR-512 (panels E and F).
26