As amyloid precursor protein (APP) gets sliced and diced, the fragments coming off the β-secretase tend to steal the show—in particular, the C-terminal piece that contains the amyloid-β peptide found in the hallmark plaques of Alzheimer disease brains. But what about the other half? In this week’s PNAS Early Edition, scientists propose an intriguing new role for the soluble, secreted N-terminal part, otherwise known as sAPPβ. Hui Zheng, Baylor College of Medicine, Houston, Texas, and colleagues report that sAPPβ acts as a transcriptional enhancer of sorts, with transthyretin (TTR) and Klotho as downstream targets. As these genes help cells withstand AD pathology and aging, respectively, the data suggest a potential self-preservation mechanism where each cut of APP by β-secretase not only helps unleash Aβ, but also a soluble APP fragment to upregulate factors counteracting its toxicity. However, while sAPPβ alone was able to recapitulate APP-mediated transcriptional regulation of TTR and Klotho, the fragment was not totally heroic—it could not stand in for its full-length precursor in critical developmental functions. “Soluble APPβ performs, on its own, an important biological function, but this accounts for only part of the overall activity of APP,” said Thomas Südhof, Stanford University, Palo Alto, California, who co-led the research with Zheng. The data could fuel efforts to identify receptors and other signaling molecules that mediate sAPPβ’s protective effects, which could expand possibilities for AD therapeutic approaches.
Zheng and other investigators have established that APP-deficient mice show cognitive and motor weaknesses, but otherwise live full, fertile lives (Zheng et al., 1995; Dawson et al., 1999). On the other hand, double-knockout mice that lack both APP and its family member APLP2 die soon after birth (see Li et al., 1996 and ARF related news story, and von Koch et al., 1997) with severe defects in their neuromuscular synapses (Wang et al., 2005). Similarly, worms deficient in APL-1, the only APP-related gene in C. elegans, do not survive. However, those worms live if just the APL-1 ectodomain is expressed in their neurons (Hornsten et al., 2007), suggesting that this soluble fragment can mediate the essential functions of full-length APL-1.
First authors Hongmei Li and Baiping Wang did the analogous experiment in mice, asking whether sAPPβ is itself functional—in other words, could it rescue the double knockout of APP and APLP2? To address the issue, the researchers generated mice with an APP knock-in (sAPPβ ki) allele that exclusively expresses FLAG-tagged sAPPβ protein, but no full-length APP. They bred those animals onto an APLP2-null background, and crossed double heterozygotes (sAPPβ ki/- APLP2+/-) with each other, in hopes of generating APP/APLP2-null mice with varying sAPPβ gene dosage. Though pups genotyped at birth showed the expected Mendelian gene distribution, exceedingly few survived to weaning age, indicating that sAPPβ fails to rescue the postnatal lethality of APP/APLP2 double-knockout mice.
What about other possible roles for sAPPβ? These emerged after the authors looked more closely at APP’s role in transcription. Prior studies suggest that APP acts through yet another fragment—its intracellular domain also known as AICD—to regulate gene transcription (Cao and Südhof, 2001 and ARF related news story; Cao and Südhof, 2004), and have proposed a good number of downstream targets, many of them “questioned or disputed,” the authors write.
To explore whether the sAPPβ fragment might also play a role in gene transcription, Li, Wang, and colleagues created neuronal-specific APP/APLP2 double-conditional knockout (dcKO) mice. They extracted hippocampal tissue for microarray analysis, looking for genes expressed at different levels in these animals compared with APLP2-null controls. The analysis identified some 30 genes, most of which were downregulated in the APP/APLP2 dcKO samples. The scientists focused on two, TTR and Klotho, that were reported to be upregulated in the brains of APP-overexpressing animals (Stein et al., 2002; Wu et al., 2006). Using quantitative RT-PCR, the authors confirmed that TTR and Klotho expression were elevated in APP-overexpressing hippocampal tissue from Tg2576 mice, and decreased in APP/APLP2 dcKO samples, relative to APLP2-null controls. Curiously, however, they found no downregulation of TTR or Klotho in sAPPβ ki+/- APLP2-null mice. This indicates that sAPPβ was sufficient to support expression of these genes.
“Our findings raise the possibility that impaired APP extracellular processing could contribute to AD pathogenesis during aging through misregulation of TTR and Klotho,” Zheng told ARF via e-mail. TTR has been shown to sequester and degrade Aβ (Costa et al., 2008), as well as protect against AD pathology in mice (Buxbaum et al., 2008 and ARF related news story). Klotho increases resistance to oxidative stress (Yamamoto et al., 2005) and seems to act as an anti-aging factor, based on studies showing that mice lacking Klotho age prematurely and develop cognitive impairment (Kuro-o et al., 1997). Moreover, Klotho expression drops off with age in rhesus monkeys (Duce et al., 2008).
“It would be interesting to see to what extent the upregulation of Klotho and TTR by sAPPβ persists with aging,” said Suzanne Guenette, Massachusetts General Hospital, Charlestown. “Furthermore, identification of the receptors for sAPPβ involved in this protective effect might be useful as therapeutic targets for AD, particularly since TTR is known to be downregulated in AD CSF” (Serot et al., 1997; Merched et al., 1998).
In Südhof’s view, “what’s important here is not which genes are regulated, but that sAPPβ regulates them. This means sAPPβ must bind to something, either a receptor or other signaling molecules. There must be a biological function of sAPPβ that is independent of the function of APP that requires its C-terminus.”
Moreover, the current study reports that the sAPPβ produced by sAPPβ ki/ki mice was highly stable, as judged by the absence of cleavage products in cell lysates and conditioned media probed with N-terminal anti-APP and C-terminal anti-FLAG antibodies. This appeared at odds with a report showing that sAPPβ gets further cleaved to make an N-terminal fragment that binds death receptor 6 (DR6) and mediates axon pruning and neuronal cell death during development (Nikolaev et al., 2009 and ARF related news story). Whether the FLAG tag interferes with degradation is not clear. Also, “it’s possible that the cleavage occurs and is quickly degraded, or that it happens under pathological conditions,” Südhof said. “But in the experiments we performed in the sAPPβ ki/ki mice, this was not a major pathway for sAPPβ clearance.”—Esther Landhuis