Xu and colleagues’ search for APP partners broadened when they identified a tumor suppressor role for the intracellular chunk of APP produced by γ-secretase cleavage. The APP intracellular domain (AICD) bound the promoter of epidermal growth factor receptor (EGFR), curbing its transcription (Zhang et al., 2007). EGFR is upregulated in various cancers, and Xu said the AICD data seem to fit with the discovery of increased skin cancer rates among treated participants in the halted Phase 3 AD trial of Eli Lilly’s γ-secretase inhibitor (see ARF related news story). Xu’s team did a yeast two-hybrid screen for AICD-associated proteins—one of which turned out to be appoptosin. Xu previously reported on this protein, then called 168, at the 2007 Society for Neuroscience meeting in San Diego (see ARF conference story), and has been characterizing the protein ever since.

At the time, Han Zhang, Xiamen University, Fujian, China, first author on the paper, and colleagues could not find any published data on this protein. But they found that it promoted apoptosis: Overexpressing the gene triggered death in human embryonic kidney cells, and knocking it out made neurons more resistant to Aβ toxicity. Further experiments honed the researchers’ focus by showing that the AICD-binding protein induced cell death through mitochondrial caspase-3. Still, Xu said, they wondered how this protein causes apoptosis.

A 2009 paper provided the critical hint. Other scientists linked a rare form of anemia to mutations in a mitochondrial gene (SLC25A38) that overlapped with the p168 sequence from Xu’s yeast two-hybrid screen (Guernsey et al., 2009). Xu then suspected the mutated gene might be important for producing heme.

Consistent with Xu’s hunch, heme levels rose, as did amounts of cleaved caspase-3, in mouse neuroblastoma cells transfected with appoptosin. Treatment with a heme synthesis inhibitor prevented the caspase-3 activation. Furthermore, overexpression of appoptosin caused human neuroblastoma cells to overproduce reactive oxygen species (ROS), and weakened their mitochondrial membrane potential. Addition of Aβ1-42 and glutamate ramped up heme and cleaved caspase-3 levels—but not if appoptosin was downregulated by RNA interference (RNAi). Plus, appoptosin knockdown prevented neuritic damage, mitochondrial fragmentation, and other morphological abnormalities typically induced by Aβ treatment. The findings show that appoptosin promotes cell death by turning up heme synthesis, whereas downregulating appoptosin helps cells resist neurotoxicity.

How does APP fit in? Previous work suggested a connection between AICD and caspase activation (see Park et al., 2009; ARF conference story; Kajiwara et al., 2009). In the present study, Zhang and colleagues looked at caspase-3 activation in HEK293T human embryonic kidney cells transfected with appoptosin and different forms of APP—full length or one of several truncation mutants. Overexpression of APP relieved appoptosin-induced caspase-3 cleavage in all cases but one—the APP mutant lacking the 57-amino-acid intracellular domain that binds appoptosin. Curiously, expression of this intracellular fragment (APP C57)—the only non-membrane-tethered form of APP used in the study—also had little influence on caspase-3 activation. These findings suggested to the authors that membrane-anchored APP binds to appoptosin and keeps it in the cytosol, reducing its transport into mitochondria where it can cause heme overproduction and apoptosis. But “when membrane-anchored APP is cleaved by γ-secretase, APP-bound appoptosin is released together with APP C57/AICD and translocated to the mitochondria to exert its function,” the authors write. In support of this idea, fluorescence microscopy of HEK293T transfectants showed that appoptosin binds full-length APP primarily in the cytosol, whereas association between appoptosin and APP C57 mainly occurs in mitochondria.

“This is an important study,” said Ashley Bush of the Mental Health Research Institute, Victoria, Australia. The findings seem to jibe with recent work suggesting that neurons rack up hemoglobin in AD in response to hypoxia (Chuang et al., 2012), Bush noted in an e-mail to Alzforum. From immunoblot analyses in the present work, the authors detected more appoptosin and cleaved caspase-3 in cortical extracts from 10 AD patients and four stroke patients, relative to brain samples of cognitively healthy controls. There are other potential links to dementia. Heme metabolism falters in AD and other neurodegenerative disorders (Ryter and Tyrrell, 2000; Atamna, 2004), and iron chelators are emerging as potential therapeutic approaches because they can prevent iron-based reactions that lead to reactive oxygen species (see ARF conference story). The new data also appear consistent with research by Bush and colleagues highlighting iron transport roles for APP and tau (ARF related news story on Duce et al., 2010; ARF related news story on Lei et al., 2012). Moreover, appoptosin expression correlates with polymorphisms in or near MOBP (myelin-associated oligodendrocyte basic protein)—a new gene linked to risk for the tauopathy progressive supranuclear palsy (Höglinger et al., 2011).

So far, apoptosis genes have not been tied to increased AD risk, making it hard to connect the disease with programmed cell death. The current work “provides what is perhaps the most direct molecular pathway linking APP metabolism and apoptosis,” noted Sam Gandy of Mount Sinai School of Medicine, New York, in an e-mail to Alzforum (see full comment below). The data suggest that reducing brain appoptosin could be a good therapeutic strategy for AD, he wrote.

To get a better handle on mechanisms underlying appoptosin upregulation during disease or injury, Xu’s team has made appoptosin knockout mice and are crossing appoptosin-overexpressing mice with mice that model AD or other neurodegenerative disorders. They are also trying to develop compounds that inhibit appoptosin transcription or increase its interaction with APP.—Esther Landhuis.

Reference:
Zhang H, Zhang YW, Chen Y, Huang X, Zhou F, Wang W, Xian B, Zhang X, Masliah E, Chen Q, Han JD, Bu G, Reed JC, Liao FF, Chen YG, Xu H. Appoptosin is a Novel Pro-Apoptotic Protein and Mediates Cell Death in Neurodegeneration. J Neurosci. 31 Oct 2012;32(44):15565-76. Abstract