The triple complex Bredesen and colleagues described seems to parallel another triumvirate, that of APP, the adaptor protein Fe65, and Tip60, which also influences transcription (see ARF related news story on Cao and Südhof, 2001 and ARF related news story on Stante et al., 2009). Both complexes appear to rely on the cleavage of APP by γ-secretase, which frees the APP intracellular domain (AICD) to join other cytoplasmic proteins in entering the nucleus (reviewed in Slomnicki and Lesniak, 2008). But the two complexes are not entirely similar, notes Suzanne Guenette of the MassGeneral Institute for Neurodegenerative Disease in Charlestown, Massachusetts, who was not involved with the current research. Tip60 is a histone acetyltransferase linked to DNA repair, whereas TAZ and YAP are transcriptional coactivators. “The two mechanisms might be completely different,” she said.

The Bredesen group was interested in the Mint family, Mint1, Mint2, and Mint3 (also known as X11 proteins), because they interact with APP (see ARF related news story on Ho et al., 2008). In mouse studies, Mint proteins seem to be involved in regulating neurotransmitter release (Ho et al., 2006). Mint1 and Mint2 are expressed mainly in neurons, and Mint3 is ubiquitous. They also appear to compete with Fe65 to bind APP (Lau et al., 2000).

To look for new proteins that might interact with the Mints, the researchers applied a phage-based screen developed in their lab (Kurakin et al., 2004). The goal of the screen was to identify, from a library of short peptides, amino acid sequences likely to interact with Mint1. They used beads bound with Mint1’s PDZ domain, a domain commonly involved in protein-protein interactions, as a target and exposed the beads to a collection of phage expressing a library of 16-amino-acid peptides on their surfaces. “You literally pour these [phage] over your target,” Bredesen said. The researchers then collected the beads to identify the peptides that interacted with the PDZ domain.

Any protein containing those amino acid sequences, then, would be a prime candidate for a Mint1 interactor. The researchers searched computer databases for proteins that fit the bill. Ultimately, they identified 46 novel proteins that may interact with Mint1. Among those were 10 transcriptional regulators. Membrane proteins, receptors, and enzymes were also well represented.

The researchers selected two transcriptional regulators for further study. They were interested in TAZ because it appears to repress the effects of PPARγ (Hong et al., 2005), a member of a protein family that appears to be protective in Alzheimer’s (reviewed in Kummer and Heneka, 2008). YAP was also intriguing because it appears to help control organ size (Dong et al., 2007) and, Alzheimer’s, of course, involves a shrinking brain.

The researchers showed that APP, Mint proteins, and TAZ and YAP interact by overexpressing the genes into 293T kidney cells and B103 neuroblastoma cells. They used co-immunoprecipitation to define a physical interaction among APP, Mint proteins, and TAZ in 293T cells. When they tagged TAZ with FLAG and used FLAG antibodies; APP came along for the ride—but only at very low levels unless Mint1 or Mint3 was overexpressed as well. (The researchers did not observe APP-Mint-YAP co-immunoprecipitation in a similar assay, although further experiments, described below, indicated that the three interact.)

Being transcriptional coactivators, TAZ and YAP must reach the DNA in the nucleus to carry out their function. Accordingly, the researchers used microscopy to look for an influence of Mint proteins over TAZ and YAP localization in 293T cells.

Mint1 and Mint2 tend to hang out in the cytoplasm, but Mint3 shuttles back and forth between the cytoplasm and the nucleus. When co-expressed with Mint1, both TAZ and YAP stayed in the cytoplasm. With Mint3, they were found in both compartments. The results suggest that the Mint proteins can dictate the location of the transcriptional co-activators.

Next, the researchers looked for a functional role for the APP-Mint-TAZ and APP-Mint-YAP complexes. They transfected B103 cells with an APP-GAL4 hybrid construct, as well as a GAL4-controlled luciferase reporter. If APP, Mint, and TAZ or YAP (and thus GAL4) reached the nucleus, the reporter should produce luciferase.

When the scientists expressed the reporter, APP-GAL4 and TAZ or YAP along with Mint3, they did see luciferase activity, in keeping with the localization results since Mint3 might be expected to carry the APP-GAL4 to the nucleus. But with Mint1 or Mint2, there was no reporter activity, presumably because the triple complex remained cytoplasmic.

The results suggest that Mint proteins act in concert with the APP cytoplasmic fragment to regulate transport of transcription factors to the nucleus. Bredesen was especially intrigued by the YAP connection, because in a mouse model of AD, brain size appears to decrease before any loss of neurons or synapses (Galvan et al., 2006). “It’s backwards,” Bredesen said. “Why would you have atrophy early, with no cell loss?” YAP’s role in organ size, influenced by APP cleavage, makes for an appealing but unproven explanation.

The work opens up many new questions. For one, Bredesen would like to explore more of the 46 novel Mint1 interactors identified in the phage assay. In addition, Guenette suggested that it would be important to find evidence for the APP-Mint-YAP and TAZ interactions in cells or animals expressing normal levels of the proteins. At this point, the results say that the three can interact—but not that they actually do come together in any physiologically relevant system. Future work could also address the potential gene targets of YAP and TAZ, which may or may not be involved in AD pathology.—Amber Dance.

Reference:
Swistowski A, Zhang Q, Orcholski ME, Crippen D, Vitelli C, Kurakin A, Bredesen DE. Novel mediators of amyloid precursor protein signaling. J Neurosci. 2009 Dec 16;29(50):15703-15712. Abstract