Once in a great many experiments, a negative control can turn up surprisingly positive results. That happened to Todd Golde, Jungsu Kim, Terrone Rosenberry, and colleagues at the Mayo Clinic in Jacksonville, Florida, recently, when they identified an unexpected anti-amyloid activity in a peptide fragment of the Bri2 protein. The result is even more intriguing, given that mutated forms of the Bri2 peptide are themselves amyloidogenic and cause familial British and Danish dementias.
The work, published in the June 4 issue of the Journal of Neuroscience, puts the Bri2-23 peptide, comprising 23 C-terminal amino acids liberated from the parent protein by furin and related proteases, in a group with several other recently identified anti-amyloid peptides, including transthyretin (see ARF related news story) and cystatin C (see ARF related news story).
In a previous study, first author Kim used a Bri2-Aβ40 fusion protein to make transgenic mice that accumulated Aβ40 after furin cleavage of the fusion protein. Crossing those mice with Tg2576 AD mice revealed that elevating Aβ40 reduced amyloid buildup (see ARF related news story). That data led to the idea that Aβ40 was “anti-amyloidogenic” and protected against the aggregation of Aβ42. The new study replicates that finding in another AD mouse model (TgCRND8, expressing human APP with Swedish and Indiana mutations) using an adeno-associated virus (AAV) technology (“somatic brain transgenesis”) to introduce the fusion protein. Injection of the adeno-associated virus (AAV) vector into the cerebral ventricles of newborn mice results in robust expression of the fusion protein in the brain, and a reduction in total amyloid load and Aβ plaque load.
The surprise result came when Kim transduced mice with the full-length Bri2 protein as a control, and found that the Bri construct was as effective at reducing total amyloid load as Aβ40.
There have been previous reports that Bri2 interacts with APP and inhibits its processing in vitro (Matsuda et al., 2005; Fotinopoulou et al., 2005), but Kim and colleagues found no evidence for changes in APP processing in vivo, or for induction of a plaque-clearing immune response in the transduced mice.
Instead, they found that the Bri2-23 peptide could directly inhibit Aβ aggregation in vitro. In vivo, a mutant Bri2 protein lacking the C-terminal 23 amino acids had no effect on Aβ deposition. “Our data support anti-aggregation effects of peptides released from Bri2 as the mediator of anti-amyloid effects,” Golde told ARF. The results don’t rule out the possibility that Bri might under some circumstances affect APP processing, but either way, he said, “Bri2 looks like it is going to be a modulator of Aβ deposition.” If someone can find a way to deliver the peptide to the brain, Golde said, “We may have something that can be effective in slowing the course of amyloid deposition.”
Like transthyretin and cystatin C, mutant Bri2 peptides have the potential to self-aggregate and also interact with Aβ. “There may be a general theme emerging,” Golde said, “that if you allow two proteins that are potentially amyloidogenic or just in some way ‘unfolded’ to interact, maybe under some circumstances they may inhibit each others’ deposition.”
Could Bri2-23 be an endogenous binding partner for Aβ? In support of this idea, Kim and colleagues detected Bri2-23 in cerebral spinal fluid samples from three out of three people tested. There is also some unpublished genetic evidence that may point to a role for Bri2 in AD, says Golde.
Beyond the Bri2 results, the new study highlights the value of “somatic brain transgenics,” an approach Golde and colleagues have championed (see ARF related news story). Their technique of AAV-mediated gene transduction into newborn mice allows the rapid evaluation of transgenes in AD models at a relatively low cost. “A complete study with one construct in one or even two APP transgenic mouse models takes three to five months and costs $5,000 to $10,000,” Golde said. “That’s a fraction of the time and expense of making transgenics.”
Although Bri2 is highly abundant in the brain, the function of the normal protein is unknown, as is the basis for the neurotoxicity of mutated Bri2. A new transgenic mouse could be a start to understanding the pathology of the mutation. Ruben Vidal, Bernardino Ghetti, and colleagues at the Indiana University School of Medicine in Indianapolis have created a mouse expressing the Danish mutant form of Bri2 under the control of the mouse prion promoter, which they describe in a publication in Brain Pathology that appeared online in April (see also comment below). In the new mice, just as in human familial Danish dementia, Bri2 gets cleaved to yield the 34 amino acid, amyloidogenic ADan peptide, and similarly to the human disease, the animals show cerebral amyloid angiopathy, parenchymal ADan deposits, and neuroinflammation. Despite a large amount of CAA, there is no evidence of hemorrhages. The mice have abnormal grooming behavior, and gait changes. Oligomeric ADan (detected with the A11 antibody, which also recognizes oligomeric Aβ) is apparent in both intracellular and extracellular locations, along with tau immunoreactive deposits. Further studies on the mice may shed light on the question of whether amyloid peptides of different primary sequence share common pathogenic mechanisms.—Pat McCaffrey