9 December 2009. Exactly which form the amino end of Aβ takes in the brains of people with Alzheimer disease is a question that increasingly crops up in talks about immunotherapy and even other dementias, as well. For example, last October at the Society for Neuroscience meeting in Chicago, David Cribbs of the University of California at Irvine presented new data from his lab in collaboration with that of Goar Gevorkian at Universidad Nacional Autonoma de Mexico in Mexico City. Cribbs is thinking about how to exploit pyroGluAβ aggregates as a target for immunotherapy. Probiodrug researchers have begun probing this potential new target by way of small-molecule inhibitors of glutaminyl cyclase. For his part, Cribbs noted that immunotherapy research, too, stands to gain by directing new therapeutic antibodies specifically toward the modified N-terminus of Aβ.

Many antibodies in preclinical and clinical research recognize the EFRH sequence at residues 3-6 of full-length Aβ’s N-terminus, and many immunogens studied for use as active vaccines tend to induce antibodies against this epitope. The problem with that, Cribbs said in Chicago, is that this epitope may be absent from plaques in AD brain because it gets lost during in-situ modification of full-length Aβ to the pyroGlu3 form. Furthermore, the EFRH epitope is immuno-dominant in mice but not necessarily in rabbits or humans. “If you produce anti-Aβ antibodies in mice, then humanize them and give them to people, you may be targeting a part of the Aβ peptide that is actually gone from many plaques in human brain,” Cribbs said. To circumvent this problem, the Californian-Mexican colleagues generated rabbit polyclonal antibodies specifically against Aβ pyroglutamated on the 3 position, and used them to characterize the requisite immuno-dominant epitope. “This approach may offer alternatives in AD immunotherapy,” Cribbs concluded. The work appeared online last summer (Acero et al., 2009).

Chris Dealwis at Case Western Reserve University in Cleveland, Ohio, and collaborators elsewhere, are pursuing a similar vein of investigation by way of structural biology. In 2007, these scientists solved the crystal structure of their anti-Aβ antibody PF1 complexed with its quarry, treating the field at large to its first glimpse of an Aβ-antibody interaction in atom-by-atom resolution (see ARF related news story and Q&A). Earlier this year, they followed up with a second paper noting that the PF1 antibody binds pyroGlu3Aβ with 77-fold loss in affinity compared to full-length Aβ. PF1 recognizes the EFRH motif. In their latest study, Dealwis and colleagues dissect exactly why the antibody binds pyroGlu3Aβ so weakly. The scientists present new crystal structures of the antibody complexed with pyroGlu3Aβ, describing how the pyro-glutaminyl ring at residue 3 alters the hydrogen bonds that can form between the antibody and its prey. The goal of this work, as well, is to pave the way for high-affinity antibodies specifically directed against the pyroGlu3 form of Aβ for use in future immunotherapies (Gardberg et al., 2009).

Biomarker research, too, has taken note of pyroGluAβ. Last July, researchers led by Tony Wyss-Coray at Stanford University noted in a broad-based study of circulating natural antibodies that pyroGluAβ forms (there are several different kinds) were the target of surprisingly many of those presumably protective antibodies, and that their concentration in plasma tends to go down with age (Britschgi et al., 2009; ARF related news story). At ICAD in Vienna, and soon after in Neurobiology of Aging, Andrea Marcello and colleagues working with Thomas Bayer at the University of Goettingen, Germany, and Lars Lannfelt at Uppsala University in Sweden, followed on with their own, smaller study of 75 people. Similarly, these investigators found that people with AD tended to have fewer anti-pyroGluAβ IgM autoantibodies in their blood than did people with MCI who, in turn, had lower levels than normal controls. These scientists reported a correlation of plasma anti-pyroGluAβ IgMs and the MMSE in their MCI group of 15 people, though not in the AD group (Marcello et al., 2009).

Finally, a last hint toward therapy development came from Yasushi Tomidokoro at University of Tsukuba in Japan, working with Jorge Ghiso’s group at New York University School of Medicine in New York City. Four years ago, while studying the rare disease Familial Danish Dementia, these authors had noted that a pyroGlu-modified form of the pathogenic protein at play in this illness was a major constituent of its amyloid deposits but was absent from the soluble protein that floats in plasma of patients with this disease (Tomidokoro et al., 2005). At SfN in Chicago, Tomidokoro and colleagues examined the pyroGlu aspect of a similar disease, called Familial British Dementia. FBD is an autosomal-dominant dementia that comes with both central and peripheral amyloidoses. Importantly, the disease serves as a model for AD featuring even neurofibrillary tangles, and yet the original offending peptide has nothing to do with Aβ. The protein comes into being thanks to a stop-to-arginine mutation in the BRI2 gene, which generates a de-novo amyloidogenic peptide that does not exist in normal people.

FBD is a central and peripheral disease, making it more accessible for direct study than AD. In Chicago, the scientists presented a poster showing that soluble ABri peptides circulating in the blood of mutation carriers all are 34 amino acids long, with a glutamine at the N-terminus. In contrast, ABri extracted from FBD brains was in large part pyroglutamated, suggesting that this modification occurs locally where the protein deposits, not throughout the body. To examine this more closely, the scientists extracted ABri from muscle, pancreas, and myocardium and subjected the samples to biochemical analysis to see where in the process of amyloid deposition pyroglutamate formation occurs. More broadly, however, the point was to show that the ability to take amyloid tissue biopsies in FBD might open up opportunities for testing therapeutic strategies directed against pyroGlu formation in FBD, with likely significance for AD.

All told, scientists across the field are taking pyroGluAβ increasingly seriously, Christian Czech of F. Hoffmann-La Roche in Basel, Switzerland, told this reporter. Beyond that, they are hoping to see validation of its clinical potential soon, said Barry Greenberg at University Health Network, Toronto.—Gabrielle Strobel.

This is Part 2 of a two-part series. See also Part 1.