Call it a case of lost inhibitions. Increased β-site cleavage enzyme 1 (BACE1) is one plausible risk for sporadic Alzheimer disease, but what causes this increase? A study in the April 23 PNAS online suggests that certain microRNAs are to blame. Researchers led by Bart De Strooper at KU Leuven, Belgium, report that several of these pint-sized nucleic acids normally prevent translation of BACE1 into protein, and that some of them are downregulated in sporadic AD brain tissue. The findings help explain why β-secretase is elevated in AD despite normal levels of BACE1 mRNA, and may even offer a new therapeutic target for treating sporadic forms of the disease.
MicroRNAs may be small, but they can certainly pack a punch. By binding to the 5’ or 3’ untranslated regions of messenger RNA, these regulatory molecules can profoundly affect translation. MicroRNAs have recently been linked to Parkinson disease (see ARF related news story). At least one microRNA, miR-107, seems to herald destruction of BACE1 mRNA and may be linked to increased BACE1 transcripts in the AD brain (see ARF related news story). But this latest finding from De Strooper’s lab suggests that other microRNAs act in a different manner.
First author Sebastien Hébert and colleagues looked in the brain of AD patients to see if any other microRNAs are up- or downregulated. In a pilot test of five patients and five age-matched controls, they found that levels of 13 out of 328 human microRNAs were significantly lower in samples of AD cortex. Computer-based structural predictions suggested that seven of those microRNAs potentially bind to the 3’ UTR of BACE1, APP, or PSEN1 genes. To see if those in silico predictions hold water, the researchers used an in-vitro reporter system to test if any of the microRNAs alter translation of a luciferase gene coupled to the BACE1 3’UTR. They found that two of the microRNAs, miR-9 and -29b-1, significantly suppressed translation of the reporter, as did miR-29a. The latter micro RNA is co-transcribed with miR-29b-1 and was also suppressed in AD tissue, though the drop did not reach statistical significance, possibly because of cross-reactivity of microRNA probes, suggest the authors.
Given that these microRNAs suppress BACE1 translation, could their downregulation in AD explain increased levels of BACE1? Hébert and colleagues found that loss of miR-29a and miR-29b-1 correlated with increased BACE1 in a subgroup of 11 patients with high BACE1. In contrast, patients with normal BACE (n = 23) had normal levels of the microRNAs. The scientists also found normal miR-29a/b-1 levels in a group of nine patients with non-AD dementia. “Thus, overall, we find that in the subgroup of AD patients with increased BACE1 expression, miR-29a and -29b-1 expression is significantly and specifically decreased,” write the authors.
There are, of course, various other ways of controlling BACE1 expression. Protein trafficking (see ARF related news story), transcriptional (see ARF related news story), other forms of translational (see ARF related news story), and post-translational regulation (see, e.g., Marambaud et al., 1998) have all been implicated. Where in the pecking order do these microRNAs stand? That remains to be determined, but the authors do provide evidence that these microRNAs have functional significance. They showed that there is co-regulation of miR-29a/b-1 and BACE1 in the brain during development and that expressing the microRNAs in HEK293 cells shuts off BACE1 expression and reduces APP processing and Aβ production. Unfortunately, the researchers found it very difficult to stably express these microRNAs in primary neurons.
Interestingly, microRNAs have been linked to the control of lifespan and aging in worms, and the authors speculate that microRNAs may become compromised during human aging. “Loss of specific microRNAs sets thus the stage for considering a ‘multiple hit hypothesis’ for sporadic AD,” they write.—Tom Fagan