Though it has been decades since amyloid-β (Aβ) was implicated in the pathology of Alzheimer disease (AD), it is still not exactly clear how the peptide is toxic. Its modus operandi may involve triggering intracellular signaling cascades—including those activated by protein kinases such as JNK and GSK3—that cause cell death. Another cell-death pathway is outed in the May 19 Journal of Neuroscience. Researchers led by Nancy Ip at the Hong Kong University of Science and Technology, China, report that signal transducer and activator of transcription 3 (STAT3), a transcription factor better known for its role in cytokine signaling, mediates Aβ-induced death of neuronal cells. STAT3 is also robustly activated in the brains of AD mouse models and in hippocampal slices taken postmortem from AD patients, report Ip and colleagues. “We believe this is a good pathway to explore further to see if blocking it would be beneficial,” Ip told ARF. But she cautioned that there is a lot yet to work out.
The transcription factor has been implicated in AD before. STAT3 binds to the promoter region of the β-secretase gene and may boost expression of the protease, leading to increased production of Aβ, according to research carried out at Karen Duff’s lab at Columbia University in New York (see ARF related news story on Wen et al., 2008). The new work from Ip’s lab raises the spectre of a vicious cycle in which STAT3 activation leads to increased Aβ, resulting in more STAT3 activity. “That is totally plausible,” said Duff in an interview with ARF. Duff also has unpublished human autopsy data that fit reasonably well with Ip’s findings (see below).
Ip and colleagues came upon the STAT3 pathway when they screened brain tissue from nine-month-old APP/PS1 double-transgenic mice (APP751SL/PS1M141L) for aberrantly activated signaling molecules. In keeping with previous findings, joint first authors Jun Wan, Amy Fu, and colleagues saw increased phosphorylation of JNK and GSK3 in APP/PS1 tissue compared to control, but also a consistent increase in phosphorylation of STAT3 at a specific amino acid, tyrosine 705 (Tyr705). The latter modification occurred in the hippocampus and cortex, sites of AD pathology, but not in the relatively unscathed cerebellum. STAT3 phosphorylation first appeared in four-month-old mice, two months before amyloid plaques generally emerge in this model, and it co-localized with the neuronal marker NeuN, indicating the transcription factor was activated in neurons. Phosphorylated STAT3 was also evident in a second double-transgenic model (APP695Sw/PS1ΔE9).
Wan and colleagues used various strategies to test if STAT3 activation is linked to Aβ toxicity. The peptide induced STAT3 phosphorylation when added to cultures of rat cortical neurons. An inhibitor that blocks STAT3 dimerization, and therefore activation, protected the neurons against Aβ-induced apoptosis, as judged by reduced caspase-3 activation and improved cell viability. Knocking down STAT3 expression by siRNA also protected PC12 cells against Aβ, but not if an siRNA-resistant STAT3 construct was expressed in the cells as well. And, stereotactically implanting Aβ1-42 in the hippocampus of normal mice caused STAT3 activation in the hippocampus and cortex 20 days later. Aβ42-1 had no effect in this experiment.
These experiments indicated that Aβ toxicity manifests itself, at least in part, through inducing STAT3 activation. But what happens up- and downstream of the transcription factor? Various kinases can phosphorylate STAT3, including Janus kinase family members JAK1, 2, 3, and Tyk2, which can all be inactivated by JAK inhibitor 1. The last completely blocked Aβ-induced phosphorylation of STAT3 in primary cortical neurons. A JAK2-specific inhibitor had no such effect, however, and only neuronal Tyk2 seemed to be activated by Aβ, suggesting it is the upstream kinase that sets the STAT3 pathway in motion. In support of this idea, Aβ failed to activate STAT3 in neurons cultured from Tyk2-negative mice. Downstream, Wan and colleagues found evidence that Aβ leads to activation of STAT3-controlled genes. In PC12 cells, the peptide turned on a luciferase gene driven by a STAT3 response element, while expression of iNOS and TRAIL, two STAT3 responsive genes, was higher in APP/PS1 mouse brain compared to controls. Ip believes that STAT3 probably upregulates other target genes in AD mice as well. For now, she plans to focus on understanding the trigger for STAT3 activation rather than on downstream effects. Since STAT3 knockouts are embryonically lethal but Tyk2-negative mice seem fine, she plans to focus on Tyk2 as a potential therapeutic target. Ip stressed that nothing has been validated yet. Duff noted that their pleiotropic effects generally make kinases make poor therapeutic targets.
To relate these in vitro and mouse experiments to Alzheimer’s, Ip and colleagues examined human postmortem tissue samples. Comparing a small number of control (five) and AD (seven) hippocampal slices, they found a threefold increase in phospho-tyrosine-705 STAT3 in the latter. Ip told ARF that she plans to extend this study to include other tissue samples, for example, from Parkinson disease patients.
Duff’s coworkers also looked at STAT3 in human tissue and have shared their unpublished data with ARF. They analyzed hippocampal tissue taken after a short postmortem interval from six AD patients and six controls. On Western blots, total STAT3 was at least twofold higher in the AD samples, and the data had high statistical significance, Duff told ARF. Her group did not look for changes to the phosphorylated form of STAT3. “If the increase in total STAT3 levels correlates with the increase in phospho-STAT3 level, then our data are complementary with theirs,” suggested Duff. “The bottom line is something is going on with STAT in the disease. Though it is not 100 percent clear what is cause and effect, it is a very interesting molecule in AD pathogenesis,” she said.
Ip thinks that the STAT3 story involves more than neurons. Though Aβ can lead to increased STAT3 activation when added directly to PC12 cells and rat cortical cultures, she believes that in vivo glia play a role. “Our working hypothesis is that increased cytokine release from non-neuronal cells somehow activates Tyk2/STAT3 signaling pathways in neurons,” she told ARF.—Tom Fagan
- Wen Y, Yu WH, Maloney B, Bailey J, Ma J, Marié I, Maurin T, Wang L, Figueroa H, Herman M, Krishnamurthy P, Liu L, Planel E, Lau LF, Lahiri DK, Duff K. Transcriptional regulation of beta-secretase by p25/cdk5 leads to enhanced amyloidogenic processing. Neuron. 2008 Mar 13;57(5):680-90. PubMed.
- Wan J, Fu AK, Ip FC, Ng HK, Hugon J, Page G, Wang JH, Lai KO, Wu Z, Ip NY. Tyk2/STAT3 signaling mediates beta-amyloid-induced neuronal cell death: implications in Alzheimer's disease. J Neurosci. 2010 May 19;30(20):6873-81. PubMed.