It was once All About Eve (check out the 1950s movie trailer). But what about Adam, or ADAM10, to be more precise? Unlike the glamorous but snarky Eve, this protease, thought to be the good α-secretase that promotes non-amyloidogenic processing of amyloid-β precursor protein (APP), has stood demurely in the wings as the bad β- and γ-secretases soaked up the limelight. But new insights place this shy sheddase center stage. In the April 7 Journal of Neuroscience, researchers led by Bart De Strooper, Katholieke Universiteit Leuven, Belgium, and Paul Saftig, Christian-Albrechts-Universität zu Kiel, Germany, report that ADAM10 is essential for proliferation of neuronal precursors and for proper development of the mouse brain. They also dispel lingering doubt over the role of ADAM10 in APP processing. “We know from overexpression experiments that ADAM10 can cleave APP, but that does not necessarily imply that ADAM10 is the responsible protease in physiological conditions,” noted De Strooper. “I think this paper provides the answer.” The researchers found that conditionally knocking out ADAM10 in mouse neurons almost completely blocks α-secretase processing of APP.
In related news, researchers led by Sven Lammich at Ludwig-Maximilians-University, Munich, Germany, found that synthesis of ADAM10 is regulated at the level of messenger RNA. In a March 26 paper in the Journal of Biological Chemistry, Lammich and colleagues reported that regulatory elements in the mRNA help suppress production of the protein. The paper raises the intriguing possibility that there may be repressors of ADAM10 that could be blocked to boost α-secretase activity and hence lower production of Aβ. One reason α has been the wallflower among the secretases is that it is much more difficult to boost an enzyme activity for therapeutic purposes than block it.
A disintegrin and metalloproteinase (ADAM), ADAM10 is believed to process transmembrane proteins such as Notch and APP, and in the process releasing several cell-surface proteins into the extracellular milieu in the brain. Saftig and colleagues made full ADAM10 knockout mice before, but those animals did not survive embryogenesis (see Hartmann et al., 2002). This time, joint first authors Ellen Jorissen, Johannes Prox, and colleagues used the Cre/loxP system to selectively ablate the ADAM10 gene from neurons, hoping for better survival and a chance to probe the role of the sheddase in the adult brain. They crossed mice expressing floxed ADAM10 with animals that expressed the Cre recombinase under control of the promoter for the neuron precursor-specific protein nestin. Again, the conditional knockouts (cKOs) were prone to die early, but enough of them survived just long enough for the scientists to tease out the role of ADAM10 in brain development and in Notch and APP processing.
All of the ADAM10-deficient pups died before or within one day of birth. Postmortem examination showed dramatic disruption of brain morphology. The authors attributed this to poor neuronal precursor proliferation and premature differentiation into neurons, which decimated the neuron population. At embryonic day (E) 12.5, the brains of the knockouts looked normal, but by E15.5, precursor proliferation was down dramatically compared to normal brain (as judged by loss of the cell cycle marker K167). This was accompanied by an increase in cells expressing the neuronal marker NeuN. By E17.5, several brain areas in ADAM10 cKOs had severely reduced cell numbers compared to normal brains. The apoptosis marker caspase 3 was barely activated, suggesting that neuron numbers are low because the cells were never produced to begin with, not because they degenerated. Migratory defects seem to contribute to the morphological deficits. The authors found four-day-old neurons uncharacteristically stuck in the intermediate zone between the ventricles and the cortical plate at E17.5; in the normal brain, neurons that age have already made their way up to the cortex.
Since Notch is an important player in neurodevelopment, Jorissen and colleagues looked for effects on its signaling. The cKO animals processed more of the protein via non-α-secretase than α-secretase pathways, and they were depleted of the Notch intracellular domain (NICD), which is derived from sequential α- and γ-secretase cleavage. The latter pathway activates a variety of transcription factors, including Hes1, Hey1, and Hey2, all of which were suppressed by about half. These findings suggest that ADAM10, rather than other candidates such as ADAM17, plays a major role in Notch signaling, the authors conclude.
The authors used primary neuronal cultures from ADAM10 cKO embryos to determine effects on APP processing. Compared to cells from normal embryos, α-secretase products sAPPα and CTF-α were almost non-existent, though APP expression in both neuronal cultures was about the same. The findings mark ADAM10 as the major α-secretase. Curiously, though, β-secretase processing of APP was down by half as well, and cultures from cKO animals only produced about half as much Aβ (38, 40, or 42) as normal cell cultures. Normally, when α-secretase activity is lowered, β-secretase activity increases, and vice versa, noted De Strooper. He is not sure why activity of the β-secretase pathway is lower in the ADAM10 cKOs. “It could be that there is an interaction between these two pathways that we don’t yet fully understand,” he suggested.
That there is still much to learn about the regulation of ADAM10 is evident from the paper by Sven Lammich and colleagues. They noted that the 5’ untranslated region (5’UTR) of ADAM10 is rich in guanine and cytosine bases and contains two small open reading frames (ORFs) upstream of the start site proper. Such features are typical of genes that are translationally regulated. Removing the upstream ORFs by mutating their start codons did not influence ADAM10 translation, but when Lammich and colleagues removed a large swath of the 5’UTR, production of ADAM10 jumped over 100-fold. The authors took this to mean that some secondary structure in the UTR acts as a translational suppressor. Those structures can regulate translation by binding to suppressor proteins. Lammich and colleagues suggest that ADAM10 might be controlled in a similar fashion, raising the prospect that relieving such suppression could become a therapeutic target. De Strooper noted that that could be a specific way of increasing α-secretase activity, as opposed to cholinergic, serotonergic, and other agents that boost α-secretase activity through uncertain mechanisms (see ARF related news story).
In the meantime, a question of interest to AD researchers is what role ADAM10 plays in the adult brain. De Strooper said he had not expected to see such spectacularly lethal effects with the conditional knockout approach and plans next to limit ADAM10 deletion to adult neurons.—Tom Fagan
- Hartmann D, De Strooper B, Serneels L, Craessaerts K, Herreman A, Annaert W, Umans L, Lübke T, Lena Illert A, Von Figura K, Saftig P. The disintegrin/metalloprotease ADAM 10 is essential for Notch signalling but not for alpha-secretase activity in fibroblasts. Hum Mol Genet. 2002 Oct 1;11(21):2615-24. PubMed.
- Lammich S, Buell D, Zilow S, Ludwig AK, Nuscher B, Lichtenthaler SF, Prinzen C, Fahrenholz F, Haass C. Expression of the anti-amyloidogenic secretase ADAM10 is suppressed by its 5'-untranslated region. J Biol Chem. 2010 May 21;285(21):15753-60. PubMed.
- Jorissen E, Prox J, Bernreuther C, Weber S, Schwanbeck R, Serneels L, Snellinx A, Craessaerts K, Thathiah A, Tesseur I, Bartsch U, Weskamp G, Blobel CP, Glatzel M, De Strooper B, Saftig P. The disintegrin/metalloproteinase ADAM10 is essential for the establishment of the brain cortex. J Neurosci. 2010 Apr 7;30(14):4833-44. PubMed.