Hippocampal expression of murine IL-4 results in exacerbation of amyloid deposition
© Chakrabarty et al.; licensee BioMed Central Ltd. 2012
Received: 15 February 2012
Accepted: 22 July 2012
Published: 29 July 2012
Skip to main content
© Chakrabarty et al.; licensee BioMed Central Ltd. 2012
Received: 15 February 2012
Accepted: 22 July 2012
Published: 29 July 2012
Pro-inflammatory stimuli, including cytokines like Interleukin-1β, Interleukin-6 and Interferon-γ, in the brain have been proposed to exacerbate existing Alzheimer’s disease (AD) neuropathology by increasing amyloidogenic processing of APP and promoting further Aβ accumulation in AD. On the other hand, anti-inflammatory cytokines have been suggested to be neuroprotective by reducing neuroinflammation and clearing Aβ. To test this hypothesis, we used adeno-associated virus serotype 1 (AAV2/1) to express an anti-inflammatory cytokine, murine Interleukin-4 (mIL-4), in the hippocampus of APP transgenic TgCRND8 mice with pre-existing plaques.
mIL-4 expression resulted in establishment of an “M2-like” phenotype in the brain and was accompanied by exacerbated Aβ deposition in TgCRND8 mice brains. No change in holo APP or APP C terminal fragment or phosphorylated tau levels were detected in mIL-4 expressing CRND8 cohorts. Biochemical analysis shows increases in both SDS soluble and insoluble Aβ. mIL-4 treatment attenuates soluble Aβ40 uptake by microglia but does not affect aggregated Aβ42 internalization by microglia or soluble Aβ40 internalization by astrocytes.
Short term focal mIL-4 expression in the hippocampus leads to exacerbation of amyloid deposition in vivo, possibly mediated by acute suppression of glial clearance mechanisms. Given that recent preclinical data from independent groups indicate engagement of the innate immune system early on during disease pathogenesis may be beneficial, our present study strongly argues for a cautious re-examination of unwarranted side–effects of anti-inflammatory therapies for neurodegenerative diseases, including AD.
Amyloid β (Aβ) plaques constitute a hallmark pathological feature of Alzheimer’s disease (AD), the most prevalent neurodegenerative disorder. Neuroinflammation has been hypothesized to play a pathogenic role in the development of sporadic AD, particularly because pro-inflammatory cytokines and chemokines colocalize with neurodegenerative pathology in both AD patient brains as well as in transgenic mouse models of AD type pathology (reviewed in ). However, the placement of the inflammatory response in AD neurodegenerative cascade is still debated, with conflicting ideas of innate immune activation either being the trigger, a homeostatic response mechanism or a bystander phenomenon associated with the disease pathology .
Anti-inflammatory cytokines, such as Interleukin (IL) -4, lead to the suppression of pro-inflammatory responses in macrophages, microglia, T cells, and astrocytes . Anti-inflammatory cytokines are thought to enhance Aβ degradation through phagocytosis and receptor-mediated uptake leading to the abrogation of Aβ induced cell death in APP transgenic mice [4, 5] and primary glia . Down regulation of IL-4 receptors contributes to aging related cognitive impairment . Consequently, IL-4 or minocycline treatment leads to restored synaptic activity in rats following intracerebroventricular infusion of Aβ . However, emerging evidence has also shown that activation of the innate immune system may constitute a beneficial defense mechanism to clear Aβ from the CNS . Data from our lab [10–12] and others [13–17] have demonstrated that glial activation can effectively clear Aβ plaques. It is generally thought that although a sustained inflammatory response is neurotoxic, activation of the innate immune system can indeed have a beneficial function by clearing debris and possibly promoting repair.
In an effort to further understand the role of neuroinflammation on Aβ plaque pathology and specifically to test the role of anti-inflammatory cytokines on Aβ pathology, we utilized recombinant adeno-associated virus serotype 1 (rAAV2/1) to overexpress murine IL-4 (mIL-4) in the hippocampus of Amyloid precursor protein (APP) transgenic mice with pre-existing amyloid plaques. Our results show that mIL-4 expression resulted in exacerbated Aβ deposition in APP transgenic mice brains after 6 weeks of expression. Biochemical analysis of mIL-4 overexpressing mice brains and phagocytosis assays in primary murine glia suggests that mIL-4 expression leads to increased Aβ levels possibly as a result of reduced glia phagocytosis.
A basic tenet of the amyloid cascade hypothesis is that Aβ accumulation triggers the onset and severity of neurodegenerative pathology, including tau hyperphosphorylation. Though occasional CP13 immunoreactive glial cells were visible in both mIL-4 and control mice, mIL-4 mice did not lead to appearance of phosphorylated CP13 immunoreactive tau (phospho Ser202/Thr205) in the hippocampal neurons in TgCRND8 mice, (Additional file 4: Figure S4, A-B). Immunoblotting also failed to show any detectable levels of CP13 or MC6 (phospho Ser235) or PHF1 epitope (phospho Ser396/Ser404) in both cohorts (data not shown).
We have found that overexpression of mIL-4 in the hippocampus of plaque-depositing APP CRND8 mice increased Aβ plaque pathology. We extensively investigated the likely factors that could be responsible for the effects of mIL-4 expression on increased Aβ burden in vivo. APP, APP CTF, BACE, ApoE and Aβ degrading enzyme levels did not appear to be altered. As IL-4 has been reported to inhibit autophagy , we investigated LC3I/II levels, but this was not informative, as we could not reliably detect LC3II. Levels of another key regulator of autophagy, p62, were decreased by mIL-4. As p62 has pleiotropic functions in addition to its role in autophagy  the significance of this finding is unclear. Furthermore if autophagic pathways were being inhibited, p62 would be expected to increase, not decrease. Given i) that these data indicating that mIL-4 does not appear to be affecting APP processing and is not altering major Aβ chaperones or degrading enzymes and ii) that we and others have previously linked proinflammatory activation of microglia to reduced plaque burdens and enhanced microglial scavenging of Aβ [10, 11, 13, 14, 17, 29, 30], we hypothesize that mIL-4 increases Aβ burden via reductions in glial scavenging of Aβ. We explored this hypothesis through Aβ internalization studies in primary microglia and astrocytes cultures. These studies do show that mIL-4 abrogates soluble Aβ40 uptake by microglia without affecting aggregated fAβ42 phagocytosis. Additionally, mIL-4 does not affect the uptake of soluble or aggregated Aβ by astrocytes, though on longer incubation, there was an increasing trend in Aβ ingestion by the astrocytes. Given the low number of astrocytes internalizing Aβ and the fact that astrocyte cultures may contain ~10-20% microglia, it is difficult to conclude whether astrocytes by themselves have a critical role in Aβ phagocytosis.
Microglia can scavenge Aβ via receptor-mediated phagocytosis as well as macropinocytosis . In vivo imaging techniques have demonstrated that microglia can home onto newly appearing plaques and internalize Aβ . Microglial interaction with Aβ occurs through several cell surface receptors, for example, SR-A, SR-B, integrins and Toll-like receptors [15, 32–34]. Previous studies have noted that inflammatory cytokines decrease scavenger receptors levels in microglial cell lines and inhibit Aβ uptake . We find that mIL-4 expression leads to increased SRs in mixed neuroglial cultures, but this appears to be linked to modest inhibitory effects on soluble, but not aggregated, Aβ uptake by microglia. Therefore, it is possible that in the absence of inflammatory mediators, these cells are not competent for Aβ internalization and degradation, even if they express the receptors [36, 37]. Indeed, microglia were found to require stimulation with cytokines or opsonins for internalizing and degrading Aβ efficiently in amyloid vaccination paradigms . Microglial populations from human AD brains display an IL-4/IL-13 induced alternate activation profile  as well as increased IL-1β, suggesting that an imbalance between the inflammatory “classical” and anti-inflammatory alternate activation profiles may result in a dysfunctional glial response and “failed” phagocytosis. Although speculative, these studies may have relevance to human therapy. Many studies show an association of nonsteroidal anti-inflammatory drugs (NSAIDS) use with reduction of dementia risk . However, in the Adult Changes in Thought (ACT) study, there was an increased incidence of all-cause dementia and AD dementia in individuals with heavy nonselective NSAID use, and this was associated with an increase in neuritic plaques at autopsy . Our data could provide an explanation for this association.
It is generally accepted that an altered innate immune response is an inherent feature of AD and most, if not all, degenerative, CNS proteinopathies . Though a large body of literature on this subject refers to this altered immune response as “pathogenic” or “harmful”, in many cases this relationship is assumed and has not been formally proven. Increased pro-inflammatory cytokines levels (TNFα, Interleukin-6, Interleukin-1β), combined with decreased levels of anti-inflammatory cytokines (Interleukin-10), have been correlated with cognitive deficits suggesting that early inflammatory changes may be detrimental . Furthermore proinflammatory stimuli have been linked to overt neuronal degeneration . Although proinflammatory factors may have a detrimental role on neuronal function, priming of CNS microglia and possibly astrocytes by accumulating Aβ, may be an essential innate host response that triggers glial phagocytosis activity and clears Aβ deposits. Indeed, we and others have previously shown that activation of the innate immune signaling pathways can attenuate Aβ accumulation in APP transgenic mice and alter the disease process [10–14, 17, 29, 30].
mIL-4 has been shown to enhance Aβ phagocytosis in primary rodent glial cultures in vitro [45, 46]. Based largely on such evidence, promotion of a glial M2 phenotype has been proposed to be beneficial in preclinical AD models and this concept was directly supported by one recent study, using a similar AAV mediated IL-4 expression paradigm as used in this present study . In this study IL-4 induced a Th2 like glial phenotype, but in contrast to our results, Kiyota et al, showed that IL-4 expression in the hippocampus led to mitigation of Aβ pathology and improved behavior in APP/PS1 mice . Inherent differences in transgenic mice as well as experimental paradigms may underlie the disparate observations. It is possible that presenilins may regulate cytokine secretion , and mutant PS1 especially may confer altered sensitivity to immune challenges in the resident glial cells . Additionally, the two studies were distinct with respect to initiation of treatment. Kiyota and colleagues performed a primary prevention study in which IL-4 expression was initiated in the pre-deposition phase, whereas IL-4 expression in our studies was initiated in mice with modest levels pre-existing plaques. In any case, additional studies of both IL-4 and other anti-inflammatory factors will be required to address this discrepancy and the generalizability of this finding.
In summary, we demonstrate that focal overexpression of mIL-4 in APP transgenic mouse brains leads to exacerbated Aβ plaque pathology. As these results are opposite of effects observed with proinflammatory cytokines, we would suggest that the underlying mechanism appears to be at least in part a failure of glia to successfully clear Aβ. Our data points to the complex relationship between microglial phenotype and the final functional outcome, necessitating a more cautious and thorough examination of potential anti-inflammatory therapies for AD.
Mice. All animal husbandry procedures performed were approved by the Institutional Animal Care and Use Committee.
AAV1 preparation and injection. AAV1 viruses expressing mIL-4 or EGFP, under the control of the cytomegalovirus enhancer/chicken β actin promoter were generated as described previously . For stereotaxic injections, TgCRND8 mice (n = 6/group) were anesthetized with 1.5% isoflurane in 1% oxygen and secured into a Kopf apparatus (Model 900 Small Animal Stereotaxic Instrument, David Kopf Instruments). The coordinates for injection were −1.7 caudal, -1.6 lateral and −1.2 ventral from the bregma. A UMP2 Microsyringe Injector and Micro4 Controller (World Precision Instruments, USA) was used to inject 2 μl (1010 viral genomes) of virus at a constant rate over a 10 minute period. After allowing an additional 10 minutes, the needle was raised and the scalp incision was closed aseptically. Nontransgenic mice used in the study were injected with rAAV2/1-mIL-4 on day P2 into the cerebral ventricles and aged till 5 months of age.
Quantitative real-time PCR. Total RNA from mice hippocampus or primary wild type mouse neuroglial cultures was isolated using the RNaqueous kit (Ambion) and reverse transcribed using Superscript III (Invitrogen). The Q-PCR (initial denaturation cycle of 95 °C /10 min, followed by 40 amplification cycles of 95 °C /15 s and 60 °C/ 1 min) was performed with ABI Prism 7900 Real Time PCR System (Applied Biosystems) using SYBR Green to detect the amplification products. Relative quantification of mRNA expression was calculated by the ΔCT method described by the manufacturer (ABI Prism 7700 Sequence Detection System, User Bulletin #2) after adjusting the levels to the corresponding internal actin control for each sample. Primers and probes were designed following Roche Universal Probe Library sequences (Hoffmann-La Roche, Germany).
Preparation of brain homogenate for immunoblotting and Aβ ELISA assay. Brain were coronally dissected 1 mm anterior and posterior to the point of injection and used for subsequent analysis. Thus, the samples for immunoblotting were obtained by dissecting the hippocampus and overlying cortex and thalamus of injected brains. Protein samples (RIPA soluble or 2% SDS soluble) separated on Bis-Tris 12% XT gels (Bio-Rad, USA) were probed with the antibody CT20 (anti-APP C-terminal 20 amino acid; T. E. Golde; 1:1000); 82E1 (IBL, 1:500), Prion (Abcam; 1:1000); LC3 (MBL; 1:500; Cell Signaling, 1:250; Novus, 1:500); p62 (Cell Signaling, 1:250), CP13 (P. Davies; 1:500) and anti β-actin (Sigma,1:1000). Relative band intensity was quantified using ImageJ software (NIH).
Aβ levels were determined biochemically using human Aβ end-specific sandwich ELISA as previously described  on samples sequentially extracted with RIPA buffer, 2% SDS and 70% Formic Acid. All ELISA results were analyzed using SoftMax Pro software (Molecular Device).
Immunohistochemical imaging and image processing. Brains were coronally dissected at the point of injection for analysis. Immunohistochemical staining was done using pan Aβ antibody 33.1.1 (1:1500, T. Golde), 82E1 (1:500, IBL), Iba-1 (1:1000; Wako), GFAP (1:500; Chemicon), pSTAT6 (1:500, Cell Signaling) and MAP2 (1:1000; Chemicon). 1% Thioflavin S (Sigma) staining was done on paraffin embedded brain sections using established protocols.
Immunohistochemically and fluorescent stained sections were captured using the Aperio Scanscope XT or FL image scanner and analyzed using either Aperio positive pixel count or ImageJ program. Brightness and contrast alterations were applied identically on captured images using Adobe Photoshop CS3.
Quantification of Aβ deposition and gliosis. Paraformaldehyde fixed paraffin embedded brain tissue sections were immunostained with 33.1.1 antibody. Aβ plaque burden and intensity of astrogliosis staining was calculated using the Positive Pixel Count program (Aperio). At least three sections per sample, 30 μm apart, were averaged by a blinded observer to calculate plaque burden. For Thioflavin S quantitation, one section per sample was used by a blinded observer to manually count the plaques using Adobe Photoshop CS5.
Primary murine culture and microglia/astrocyte phagocytosis assay. Primary microglia, astrocytes or neuronal cultures (mixed with astrocytes) were obtained from cerebral cortices of wild type neonate mice as described previously . Glial and astrocytic cultures typically were >90% cd11b (cd11b-APC; 1:200, BD Biosciences) or GFAP (Sigma, 1:500) immunopositive respectively. For phagocytosis assays, microglia or astrocytes, pre-treated with mIL-4 (R&D Systems, USA; 5 ng/ml for 10 hours) or mIL-6 (R&D Systems, USA; 10 ng/ml for 10 hours) were incubated with 0.5 μM Hilyte555-Aβ40 (Anaspec, USA) or fibrillar Hilyte555-Aβ42 (Anaspec, USA). Recombinant fluorescent Aβ was resuspended in DMSO to 1 mg/ml and diluted in DMEM medium to 0.5 μM before addition to cells. Hilyte555-Aβ42 was fibrillized at 370 C for 6 hours in PBS. Cells were analyzed at three timepoints following addition of Aβ to the culture at 370 C: 15 min, 30 min and 1 hr. Cells were washed, gently trypsinised to remove cell surface associated fluorescent entities, fixed in paraformaldehyde and mounted in DAPI containing medium for visualization. Fluorescence intensity of individual cells from at least 5 or more fields of views per sample were calculated from Image J and averaged. For FACS analysis, following washes, cells were collected by trypsinization and resuspended in FACS buffer containing BSA and scanned. Scans were collected using Accuri6 and analyzed with FCS Express 4 Flow Research (BD Biosciences). Neuronal cultures were transduced with AAV (mIL-4 or mIFN-γ; 108 viral genomes) on day 7 for 60 hours and fixed with paraformaldehyde for subsequent immunocytometry. Cytokine ELISA was performed using BD Bioscience OptiEIA reagents.
One-way Anova (with Tukey’s post-hoc test) or two-tailed Student's t test was used for statistical comparison (SigmaStat 3.0 version). Graphical analyses were done using Prism 4 (GraphPad Software) and final images created using Photoshop CS2 (Adobe).
Adult Changes in Thought
Amyloid β precursor protein
β-site APP cleaving enzyme
Enhanced green fluorescent protein
Extracellular signal-regulated kinases
Fluorescence activated cell sorting
Glial fibrillary acidic protein
Institutional animal care and use committee
Ionized calcium binding adaptor protein 1
Microtubule-associated protein light chain 3
Nonsteroidal anti-inflammatory drugs
Signal Transducer and Activator of Transcription
Tumor necrosis factor α.
This work was supported by Mayo Clinic (TEG), National Institutes of Health/National Institute on Aging Grants (RO1AG18454, RO1AG29886, P01AG25531; TEG) and American Health Assistance Foundation Grant A2009061 (PD).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.