|Home | About | Journals | Submit | Contact Us | Français|
Activated (phosphorylated) mitogen-activated protein kinase p38 (MAPK-p38) and interleukin-1 (IL-1) have both been implicated in the hyperphosphorylation of tau, a major component of the neurofibrillary tangles in Alzheimer’s disease. This, together with findings showing that IL-1 activates MAPK-p38 in vitro and is markedly overexpressed in Alzheimer brain, suggest a role for IL-l-induced MAPK-p38 activation in the genesis of neurofibrillary pathology in Alzheimer’s disease. We found frequent colocalization of hyperphosphorylated tau protein (AT8 antibody) and activated MAPK-p38 in neurons and in dystrophic neurites in Alzheimer brain, and frequent association of these structures with activated microglia overexpressing IL-1. Tissue levels of IL-1 mRNA as well as of both phosphorylated and non-phosphorylated isoforms of tau were elevated in these brains. Significant correlations were found between the numbers of AT8- and MAPK-p38-immunoreactive neurons, and between the numbers of activated microglia overexpressing IL-1 and the numbers of both AT8- and MAPK-p38-immunoreactive neurons. Furthermore, rats bearing IL-1-impregnated pellets showed a six- to seven-fold increase in the levels of MAPK-p38 mRNA, compared with rats with vehicle-only pellets (P < 0.0001). These results suggest that microglial activation and IL-1 overexpression are part of a feedback cascade in which MAPK-p38 overexpression and activation leads to tau hyperphosphorylation and neurofibrillary pathology in Alzheimer’s disease.
Neurofibrillary pathology, in the form of intraneuronal tangles and dystrophic neurites within neuritic β-amyloid plaques, is a histological hallmark of Alzheimer’s disease. These abnormal structures consist primarily of aggregated, paired helical filaments composed of hyperphosphorylated tau protein. Tau is a microtubule-associated protein that normally promotes microtubule assembly, stabilization, nucleation, and elongation. Hyperphosphorylation of tau is thought to be an important step in the formation of paired helical filaments, dystrophic neurites, and neurofibrillary tangles in Alzheimer’s disease (Braak et al., 1994).
Phosphorylation of tau protein may result from the actions of a number of protein kinases (Drewes et al., 1992). The mitogen-activated protein kinase p38 (MAPK-p38), in particular, phosphorylates tau protein at precisely those sites that are hyperphosphorylated in Alzheimer’s disease (Reynolds et al., 1997). MAPK-p38, a 38-kDa polypeptide, is a member of a group of protein serine/threonine kinases that are activated in response to extracellular stimuli through dual phosphorylation at conserved threonine and tyrosine residues. Immunoreactivity for phosphorylated (activated) MAPK-p38 has been shown to be associated with structures bearing neurofibrillary changes in Alzheimer’s disease (Hensley et al., 1999; Knowles et al., 1999).
Activation of MAPK-p38 occurs in response to treatment of cultured cells with pro-inflammatory cytokines and, in particular, in response to interleukin-1 (IL-1) (Raingeaud et al., 1995; Guan et al., 1997). Such activation may be a component of at least some IL-1 signal transduction mechanisms and consequent IL-1-driven cascades. IL-l-driven cascades have been implicated in the formation and evolution of neuritic plaques and neurofibrillary tangles in Alzheimer’s disease (Griffin et al., 1996; Sheng et al., 1997), and overexpression of IL-1 by activated microglia is a prominent feature of Alzheimer’s disease (Griffin et al., 1989). Indeed, certain polymorphisms in the two IL-1 genes have been associated with increased risk for Alzheimer’s disease (Du et al., 2000; Grimaldi et al., 2000; Nicoll et al., 2000; Rebeck, 2000).
We have recently shown that implantation of IL-1-containing, slow-release pellets into rat brain increases expression of hyperphosphorylated tau protein (Sheng et al., 2000), supporting our hypothesis that IL-1 acts as a driving force in the development of neurofibrillary pathology in Alzheimer’s disease (Mrak et al., 1995; Griffin et al., 1998,2000). Using this pellet paradigm, we find here that in addition to this increase in hyperphosphorylated tau, IL-1 also increases MAPK-p38 expression. Moreover, we demonstrate significant correlations between the numbers of hyperphosphorylated tau-immunoreactive neurons and the numbers of MAPK-p38-immunoreactive neurons, and between the numbers of activated microglia overexpressing IL-1 and the numbers of both hyperphosphorylated tau- and MAPK-p38-immunoreactive neurons in Alzheimer brain. Eighty-two percent of neurons immunoreactive for MAPK-p38 were also immunoreactive for hyperphosphorylated tau. These results are consistent with the idea that the overexpression of IL-1 in Alzheimer brain contributes to tau protein hyperphosphorylation and neurofibrillary pathology, through promotion of MAPK-p38 activation.
Pellets impregnated with IL-1 β (100 ng of recombinant mouse IL-1 β (Sigma Chemical Co., St. Louis, MO)) and control pellets (without IL-1β impregnation) were obtained from Innovative Research of America (Sarasota, FL). As described earlier (Sheng et al., 2000), these pellets were 1.5 mm in diameter and designed for controlled, slow release of IL-1 over a 21-day period. Forty male Sprague–Dawley rats, weighing 264 ± 6 g, were randomly assigned to three groups. Sixteen rats received implants of IL-1β-containing pellets (IL-1), 14 rats received pellets without IL-1 impregnation (sham), and 10 rats served as unoperated (normal) rats. The pellets were implanted 2.8 mm caudal to bregma, 4.5 mm right of the midline, and 2.5 mm deep to the pial surface. Twenty-one days following pellet implantation, rat brains were processed either for immunohistochemistry or for mRNA analysis.
Tissues from 12 demented patients, with postmortem neuropathological confirmation of Alzheimer’s disease according to CERAD criteria (Mirra et al., 1991), were used in this study (four males, eight females; age 63–92 years; postmortem interval 2–13 h). Tissues from nine patients with no clinical or pathological evidence of neurological or psychiatric disease were used as controls (seven males, two females; age 50–93 years, postmortem interval 1–15 h). Parahippocampal tissue samples were obtained fresh at the time of autopsy from the left cerebrum. These were frozen in liquid nitrogen, and stored at −80°C until used for Western immunoblot analysis or for analysis of IL-1 mRNA. The right half of autopsied brains was fixed in 20% formalin for 7–10 days. Tissue blocks of hippocampus and adjacent mesial temporal cortex at the level of lateral geniculate nucleus were then obtained and embedded in paraffin for sectioning for immunohistochemical studies.
The polyclonal anti-human IL-1 antibody used here is specific for the IL-1α isoform (Cistron Biotechnology, Pine Brook, NJ). This antibody recognizes the 33-kDa (uncleaved) form of IL-1α, and reliably labels IL-1-expressing cells in tissues. The polyclonal anti-phospho-p38 MAP kinase antibody used here detects p38 MAP kinase only when activated by dual phosphorylation at threonine 180 and tyrosine 182 (New England Biolabs Inc, Beverly, MA). AT8 is a monoclonal antibody that recognizes tau protein only when phosphorylated at both serine 202 and threonine 205 (Goedert et al., 1995). Monoclonal anti-Tau antibody (Zymed Laboratories Inc., South San Francisco) recognizes tau protein independent of phosphorylation state, at an epitope located within amino acids 404–441 of the rat sequence.
Single-label immunohistochemistry was performed on 10-μm sections cut from paraffin-embedded blocks of hippocampus and adjacent mesial temporal cortex. Paraffin sections were deparaffinized, rehydrated, permeabilized, and peroxidase blocked as described earlier (Sheng et al., 1997). The primary antibody, either anti-IL-1 (diluted 1:20), anti-MAPK-p38 (1:300), or AT8 (1:300), was diluted in 2% normal goat serum in tris-buffered saline (TBS) and was incubated on the sections overnight at room temperature. The link antibody, either anti-mouse IgG or anti-rabbit IgG (Cappel), was diluted 1:50 in 2% normal goat serum in TBS and incubated on the sections for 30 min. The secondary antibody, either goat anti-rabbit or anti-mouse peroxidase-anti-peroxidase (diluted 1:300, DAKO, Carpinteria, CA), was incubated for 30 min. Immunoreactive structures were developed as a brown color using a diaminobenzidine kit (Zymed, South San Francisco).
Dual-label immunohistochemistry was performed (10 Alzheimer and seven control patients) using a commercially available kit (K1395, DAKO). Tissue sections were deparaffinized, rehydrated, and permeabilized as described above, and primary antibodies were diluted as described earlier (Sheng et al., 1997). For IL-1/MAPK-p38 double labeling, anti-IL-1 antibody was applied directly to the tissue sections for overnight incubation. The sections were then washed in TBS (3 × 5 min), incubated in the kit peroxidase-labeled polymer (HRP from bottle 2) for 15 min, and covered with substrate-chromagen solution to produce a brown color reaction. The reaction was stopped with double stain block (bottle 4) for 3 min. The tissue sections were then incubated for 30 min with the second primary antibody, anti-MAPK-p38, followed by alkaline phosphatase-labeled polymer (bottle 5) for 30 min. Sufficient substrate-chromagen solution was applied to produce a red color reaction. MAPK-p38/AT8 double labeling was performed using a similar technique.
The numbers of cells immunoreactive for IL-1, for MAPK-p38, or for AT8 were counted using a CCD video camera attached to a Macintosh computer. Five 25 × microscopic fields (0.1 mm2 each) were analyzed for each patient.
For analysis of tissue levels of total tau protein and AT8-immunoreactive (hyperphosphorylated) tau protein, aliquots of tissue homogenate (containing 20 μg of sample protein) from 12 Alzheimer and nine control patients were subjected to 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis. The gels were transferred to nitrocellulose membranes (0.45 μm; Schleicher and Schuell, Keene, NH) using 100 V × 1 h. Subsequently, the membranes were stained with Ponceau S (Sigma), then incubated overnight in 10 ml blotto (10% non-fat dry milk, 0.01% sodium azide) containing either anti-Tau (1:500) or AT8 (1:1000) antibodies, and washed for 1 h in blotto and phosphate-buffered saline (PBS) before incubation with 0.5 μCi of I125-labelled IgG. The membranes were then washed in PBS and exposed to Kodak-XAR film for either 42 h (for AT8) or 32 h (for anti-Tau). The autoradiographic images were analyzed using a MIRROR scanner and an Apple computer.
Reverse transcription-polymerase chain reaction (RT-PCR) was carried out using an Advantage™ RT-for-PCR Kit (Clontech Laboratories, Inc., Palo Alto, CA). One microgram of total RNA was reverse transcribed in a 20 μl volume containing 20 pM oligo (dT)18 primer, 0.5 mM each dNTP, 200 U Moloney-Murine Leukemia Virus reverse transcriptase, 75 mM KC1, 3 mM MgCl2, and 50mM Tris–HCl, pH 8.3. After annealing the oligo (dT)18 primer at 70°C for 15 min, the samples were placed on ice and the remaining reaction mix was added. Following incubation at 42°C for 60 min, the reverse transcriptase was inactivated at 94°C for 5 min.
PCR was performed using an Advantage™ cDNA PCR Kit (Clontech). For the analysis of human tissue, a 30 μl reaction volume contained 5 μl template cDNA, 60 nM GAPDH (Clontech) and 0.2 μM IL-lα in 15 mM KOAc, 3.5 mM Mg(OAc)2, 75 μg/ml bovine serum albumin (BSA), and 40 mM Tricine–KOH, pH 9.2, supplemented with 0.2 mM dNTP, and 1 μl of Advantage™ cDNA Polymerase Mix (Klentaq-1 DNA polymerase and TaqStart Antibody). For the analysis of rat tissue, a similar amplification mix was used except the 30 μl reaction volume containing 2 μ (il template cDNA, 60 nM rat GAPDH and 0.2 μM MAPK-p38 primers (Ransom Hill Bioscience, Inc., Ramona, CA). Following addition of the Taq mix, cDNA was amplified using a Perkin-Elmer 2400 GeneAmp PCR System (Perkin-Elmer, Norwalk, CT). Linearity was established by generating a standard curve over a range of cycles for each primer and for pellet, sham, and control animals as well as for Alzheimer and control patients. Differences reported were evident over a range of cycles, demonstrating linearity. The mixture was subjected to 30 cycles of amplification for IL-1 and 29 cycles of amplification for MAPK-p38, with denaturation at 94°C for 1 min; annealing at 55°C for 1 min for IL-1 and at 60°C for 1 min for MAPK-p38; extension at 72°C for 1 min; and a final extension step at 72°C for 7 min.
Ten microlitres of amplified cDNA products (from 11 Alzheimer and seven control patients) were loaded onto each lane of a 1.5% agarose gel. The samples and a 1Kb ladder (Life Technologies, Gaithersburg, MD) were fractionated by size. The gel image was produced on a Stratagene Eagle Eye Digital Imaging system (Stratagene, La Jolla, CA) and saved for further analysis.
The statistical significance of differences in MAPK-p38 mRNA expression in experimental animals was assessed using ANOVA followed by Fisher’s test. The statistical significance of differences in numbers of IL-1-immunoreactive microglia, MAPK38-, and AT8-immunoreactive neurons between Alzheimer’s and control tissues was assessed using Student’s t-test. The significance of correlations between tissue IL-1 mRNA levels and numbers of IL-1-immunoreactive microglia, MAPK38-, and AT8-immunoreactive neurons were assessed using regression analysis.
Alzheimer tissue showed increased numbers of activated microglia, overexpressing IL-1, accompanying increased numbers of neurons immunoreactive for phospho-MAPK-p38, and for hyperphosphorylated tau (AT8 antibody), relative to control tissue, in parahippocampal cortex. The IL-1-immunoreactive microglia in Alzheimer’s disease were enlarged, with prominent processes, and were more intensely immunoreactive than those found in control brain. Quantitative analysis showed that the numbers of immunoreactive cells in Alzheimer tissue were 2.5-fold (for IL-1-immunoreactive microglia, P = 0.0002), nine-fold (for phospho-MAPK-p38-immunoreactive neurons, P = 0.0007), and 14-fold (for AT8-immunoreactive neurons, P = 0.0001) those of control tissue (Fig. 1A).
Dual-label preparations showed that most (82%) neurons immunoreactive for active (phospho) MAPK-p38 were also immunoreactive for AT8 (Figs. 1B, ,2d),2d), as were dystrophic neurites associated with neuritic plaques (Fig. 2e and f). These neurons and neurites were closely associated with activated microglia overexpressing IL-1α (Fig. 2b).
Western immunoblot analysis of tissue tau and AT8 protein levels showed increases of 60% (P = 0.01) and 1000% (P < 0.001), respectively, in Alzheimer brain tissue homogenates relative to controls (Fig. 3).
The human IL-1α and GAPDH sequences used for RT-PCR analysis of these mRNAs and an example of their tissue levels in Alzheimer and control brains are shown in Fig. 4a and b. In Alzheimer brain, IL-lα levels were 1.6 times those in control (0.8 ± 0.05 vs. 0.5 ± 0.07, P < 0.01, Fig. 4c).
The numbers of IL-1-immunoreactive microglia correlated with the numbers of phospho-MAPK-p38-immunoreactive neurons (R = 0.81, P < 0.0001) and with the numbers of AT8-immunoreactive neurons (R = 0.86, P < 0.0001), as well as with the relative levels of IL-1 mRNA (R = 0.82, P < 0.001, Fig. 5) in Alzheimer patients. The numbers of phospho-MAPK-immunoreactive neurons also correlated with the numbers of AT8-immunoreactive neurons (R = 0.95, P < 0.0001, Fig. 5).
Implantation of slow-release, IL-1-impregnated pellets into rat brain resulted in significant elevation of cerebral cortical tissue MAPK-p38 mRNA levels. Twenty-one days following implantation, tissue MAPK-p38 mRNA levels relative to GAPDH mRNA in contralateral cerebral cortex were elevated six to seven fold above levels in cortex of unoperated control rats or in cortex of rats receiving pellets containing vehicle only (P < 0.0001, Fig. 6).
We show cerebral cortical overexpression of MAPK-p38 mRNA in response to chronic elevation of cerebral IL-1 levels in vivo, confirming earlier in vitro studies (Raingeaud et al., 1995; Guan et al., 1997) and suggesting that the overexpression of IL-1 directly contributes to overexpression of MAPK-p38 in Alzheimer brain. This suggestion is supported by our findings of morphological associations between activated microglia overexpressing IL-1 and both neurons and neurites immunoreactive for active MAPK-p38 in Alzheimer brain, as well as our earlier findings that IL-1 is markedly overexpressed by activated microglia in Alzheimer’s disease (Griffin et al., 1989). Such activated microglia have been shown to be associated both with neurons bearing neurofibrillary tangles (Sheng et al., 1997) and with dystrophic neurites in β-amyloid plaques (Griffin et al., 1996) in Alzheimer’s disease.
MAPK-p38, like several other members of the MAPK group, is activated by dual phosphorylation (Su and Karin, 1996), and cytokines, including IL-1, can mediate this activation. Although the mechanism underlying IL-1 promotion of MAPK-p38 phosphorylation and activation is not entirely clear, protein phosphorylation changes induced in fibroblasts by IL-1 are similar to those seen after treatment with the phosphatase inhibitor okadaic acid (O’Neill, 1995), suggesting that IL-1 acts through inhibition of a phosphatase, possibly protein phosphatase-2A. The substrate for this phosphatase may be a kinase suppressor that is active in the dephosphorylated form (Guan et al., 1997). Hyperphosphorylated tau, as detected by the AT8 antibody used in this study (Goedert et al., 1995), is a key molecular component of neurofibrillary changes in neurons and in dystrophic neurites in Alzheimer’s disease. Our finding that neurons and neurites immunoreactive for AT8 showed frequent concurrent immunoreactivity for activated MAPK-p38 protein confirms the findings of earlier studies (Hensley et al., 1999; Knowles et al., 1999). These results are also consistent with in vitro studies showing that activated MAPK-p38 phosphorylates tau protein at precisely those sites where hyperphosphorylation is seen in Alzheimer’s disease (Reynolds et al., 1997).
Our finding of direct upregulation of MAPK-p38 mRNA by exogenously applied IL-1 suggests that the elevated tissue levels of IL-1 in Alzheimer’s brain contribute to the observably increased levels of activated MAPK-p38 and, through the phosphorylation of tau by MAPK-p38, to neurofibrillary pathology. This suggests the presence of a feedback loop in Alzheimer’s disease that involves IL-1 and MAPK-p38, with the collateral effect of tau hyperphosphorylation (Fig. 7).
The authors thank S. Woodward for technical assistance, P. Free for secretarial assistance, and P. Green for reviewing the manuscript. This research was supported in part by NIH AG 13939, NIH AG15501, NIH AG12411, and the Donald W. Reynolds Foundation.