| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online February 15, 2006 as doi:10.1096/fj.05-5423fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Pathology, University of Washington, Seattle, Washington, USA; and
* Department of Pathology, University of North Carolina, Chapel Hill, North Carolina, USA
1Correspondence: Department of Pathology, University of Washington, Box 359791, Seattle, WA 98104, USA. E-mail: tmontine{at}u.washington.edu
SPECIFIC AIMS
Glial innate immune response has been associated with several degenerative diseases of brain including Alzheimer’s disease (AD), Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS), traumatic brain injury, and HIV-encephalitis; and numerous genetic studies have associated inheritance of APOE4 with increased risk, earlier onset, or poorer clinical outcome for these same diseases of brain. ApoE has an immune modulatory function, at least in the peripheral adaptive immune response to some bacteria and viruses. Here we tested the hypothesis that inheritance of different APOE alleles significantly modulates neurotoxicity arising from CD14/Toll-like receptor (TLR) 4-activated glial innate immune response by lipopolysaccharide (LPS).
PRINCIPAL FINDINGS
1. Neurotoxicity from glial innate immune response is greatest with TR APOE4 microglia in primary dissociated cultures
We exposed primary dissociated cultures of wild-type (WT) neurons mixed with or without TR APOE microglia or astrocytes to LPS and quantified the number of NeuN immunoreactive cells as an index of neurotoxicity. We confirmed that that dissociated cultures enriched in neurons to the exclusion of glia are insensitive to LPS by showing that the number of NeuN-immunoreactive cells in culture was not significantly different between WT neuron cultures exposed to PBS or LPS (P>0.05). Mixed TR APOE glia-WT neuron cocultures did not show any reduction in NeuN-immunoreactive cells when exposed to PBS vehicle (P>0.05 for microglia or astrocytes); however, they displayed neurotoxicity after LPS exposure that was much greater with microglia than astrocytes and modulated by TR APOE with TR APOE4 being most neurotoxic and TR APO2 being least neurotoxic (Fig. 1
); only TR APOE2 astrocytes did not produce a toxic response. We confirmed our findings with LDH release, a cell nonspecific measure of toxicity. Since the vast majority of paracrine damage to neurons in this model derived from microglia, subsequent experiments focused on this cell.
|
Neither TLR4 expression, apoE secretion, nor LPS binding to conditioned medium was significantly different among cultures of TR APOE microglia. NO secretion, as measured by medium nitrate and nitrite levels, from TR APOE4 microglia was greater than TR APOE3 or TR APOE2; however, this was after 72 h stimulation with LPS and so lies distal to whatever processes underlie TR APOE-dependent neurotoxicity shown in Fig. 1
.
2. Cytokine secretion by TR APOE microglia is TR APOE-dependent
We screened for changes in medium cytokine concentrations using the LiquiChipTM Mouse 10-Cytokine assay and a Luminex 100 xMap reader that simultaneously determines 10 mouse cytokines in medium from TR APOE microglia. The cytokines quantified were GM-CSF, INF-
, IL-1ß, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12, and TNF-
. Only IL-6 and TNF- changed significantly after LPS exposure for 12 h; IL-1ß was near the limit of detection for this assay. Using individual ELISAs, we showed that after 12 h incubation with LPS microglial secretion of IL-6, TNF-, and IL-1ß followed the rank order of TR APOE4 > TR APOE3 > TR APOE2. Astrocytes did not show this pattern of TR APOE-dependent cytokine secretion.
3. p38MAPK activation in TR APOE microglia
We assessed relative p38MAPK activity by quantifying IL-6 and TNF- concentrations after LPS exposure with and without a specific inhibitor of p38MAPK activity, SB203580 (SB). Secretion of both cytokines by TR APOE4 microglia was significantly more sensitive to SB than TR APOE2 microglia, while only secretion of TNF- by TR APOE3 microglia was significantly more sensitive to SB than TR APOE4 microglia. There was no difference in SB-dependent cytokine secretion among astrocyte cultures from the 3 TR APOE mice.
4. Innate immunity-mediated neural damage in hippocampal slice cultures occurs with TR APOE4 and TR APOE3, but not TR APOE2, and is p38MAPK-dependent
Our results from dissociated cultures pointed to p38MAPK activation in microglia, particularly TR APOE4 microglia, as the dominant mechanism for neurotoxicity in our model. However, dissociated cultures lack the physiologically relevant proportion of cells and their contacts as well as the likely functionally antagonistic effects of microglia and astrocytes. Therefore, we examined hippocampal slice cultures from TR APOE mice exposed to LPS. Our results showed that LPS-activated innate immune response in hippocampal slices had the same pattern of TR APOE-dependent cytokine secretion as dissociated microglial cultures with TNF- secretion that was lowest with TR APOE2, increased 3.4-fold in TR APOE3, and increased 6.7-fold in TR APOE4 slices (ANOVA on absolute concentrations had P<0.0001 with all possible Bonferroni-corrected repeated pair comparisons showing P<0.001). As expected, we colocalized phosphorylated- (P- ; activated) p38MAPK with lectin (microglia) reactive cells in TR APOE4 hippocampal slices. Together these quantitative and localizing data indicate that LPS activation of innate immune response in hippocampal slices was greatest with TR APOE4 and associated with microglial p38MAPK activation, similar to microglial dissociated cultures.
We showed earlier that LPS-activated innate immune response in wt mice does not lead to hippocampal neuron death but rather a CD14/TLR4-dependent reversible reduction in dendritic arbor and spine density that peaked at 24 h postexposure in vivo; therefore, we assessed neurotoxicity in slice cultures by determining relative levels of presynaptic and postsynaptic membrane proteins 24 h after LPS exposure. There was no difference in the relative concentration of any of these synaptic proteins among TR APOE slices exposed to PBS vehicle, nor were synaptophysin (presynaptic) levels different among the 3 TR APOE slices exposed to PBS or LPS; the 6 average synaptophysin values (normalized to TR APOE2 slices exposed to PBS) ranged from 96.3 ± 9.7% to 100.5 ± 10.1%. Similar to what others have observed in mouse models of cerebral Aß amyloidogenesis, there was significant reduction in the relative concentration of both postsynaptic proteins (drebrin and PSD-95) after exposure of TR APOE3 or TR APOE4 hippocampal slices to LPS. There was no significant effect of LPS in TR APOE2 hippocampal slices. We investigated this further by immunohistochemistry for MAP2 in WT and the 3 TR APOE slices exposed to PBS or LPS for 24 h. The pattern of MAP2 immunoreactivity was essentially the same among WT and TR APOE slices exposed to PBS as well as WT and TR APOE2 slices exposed to LPS. In contrast, LPS exposure produced disorganized, fractured, and beaded MAP2 immunoreactivity that was similar in TR APOE3 and TR APOE4 slices.
Considering the TR APOE-dependent effect of a p38MAPK inhibitor on dissociated microglia and the apparent microglial dominance in the innate immune response of hippocampal slices, we hypothesized that the TR APOE4-specific neurotoxicity in hippocampal slices may be p38MAPK-dependent (Fig. 2
). TR APOE4 hippocampal slices were incubated with PBS or LPS, with or without SB, and relative levels of presynaptic and postsynaptic proteins determined. LPS plus SB did not produce any change in density of synaptophysin immunoreactivity with values for these 6 groups ranging from 98 ± 4 to 104 ± 3% (ANOVA had P>0.05). LPS-activated neuronal damage in TR APOE3 or TR APOE4 hippocampal slice cultures was largely suppressed by coincubation with SB, so that LPS plus SB was not significantly different from PBS plus SB; similar results were obtained with PSD-95. Thus inhibition of p38MAPK was sufficient to abolish the TR APOE4-specific neurotoxicity from innate immune activation in hippocampal slices.
|
CONCLUSIONS AND SIGNIFICANCE
While there are several apoE transgenic mouse models that have been used to model diseases of brain, an advantage of the TR APOE mice is that express "humanized" apoE isoforms under control of endogenous promoters and repressors and thereby more closely reflect the appropriate expression of apoE in the various cells of brain than earlier transgenic models that used different promoters.
Inheritance of APOE alleles is associated with varying clinical outcomes in several neurodegenerative diseases, including AD, PD, ALS, head trauma, multiple sclerosis, and HIV encephalitis. Since activation of innate immunity is also associated with these same brain diseases, we tested the hypothesis that apoE isoforms may act by modulating glial innate immune response and thereby altering neurotoxicity. Our results show that indirect neurotoxicity from activation of CD14/TLR4-dependent innate immune response is much greater from microglia than astrocytes and follows the rank order of greatest from TR APOE4, intermediate from TR APOE3, and least from TR APOE2 in both microglia-neuron cocultures and hippocampal slice cultures.
Although the precise mechanism by which LPS-activated TR APOE4 microglia were most neurotoxic is not entirely clear from our experiments, whatever exact combination of effectors is, their production was largely if not completely p38MAPK-dependent. Interpretation of specific neurotoxic effects of cytokines is confounded in our system because of the multiple autocrine and paracrine effects of cytokines and other products simultaneously secreted during innate immune response. While we confirmed TR APOE4-dependent differences in NO secretion 3 days after exposure to LPS and extended them to include analysis of TR APOE2 glia, NO secretion did not show TR APOE-dependent differences or achieve neurotoxic levels until after neurotoxicity occurred. For this reason, as well as the limitations mentioned above for cytokines, it seems unlikely that NO, or its products, directly contributed to neurotoxicity in this mouse model.
Acute injury to brain, such as stroke and trauma, and chronic neurodegeneration, especially AD, have been repeatedly associated with p38MAPK-mediated signaling in microglia was greatest in TR APOE4, intermediate in TR APOE3, and least in TR APOE2, and thus positively correlated with both increased innate immune activation and paracrine damage to neurons from microglia. Moreover, the indirect neurotoxic effects of glial innate immune activation in TR APOE4 and TR APOE3 hippocampal slices were suppressed by inhibition of p38MAPK, converting the neurotoxic phenotype of these hippocampal slices to that of TR APOE2. These data suggest a new mechanism by which inheritance of different APOE alleles may influence the outcome of acute and chronic diseases of brain that are associated with innate immune activation.
|
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.05-5423fje;
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
P < 0.05.
4 per group). Two-way ANOVA had P < 0.01 for PBS vs. LPS, and P < 0.05 for SB or no SB, but P > 0.05 for interaction between these 2 terms for both TR APOE3 and TR APOE4. Bonferroni-corrected posttests had P > 0.05 for SB but *P < 0.01 for no SB.