Read a PDF of the entire series.
20 March 2013. Pyroglutamate Aβ is a particularly toxic version of the peptide at the center of Alzheimer’s disease pathogenesis; in fact, some scientists claim it seeds the formation of plaques (for a review, see Gunn et al., 2010). It forms when a few amino acids are lopped off of Aβ’s N-terminus to expose a glutamate residue, which is then cyclized by an enzyme called glutaminyl cyclase (QC). The resulting pyroglutamate Aβ (pGlu-Aβ) is stickier than other forms of the peptide and, according to some estimates, constitutes up to half of the Aβ in AD plaques. “Pyroglutamate Aβ is a very pathogenic form of β amyloid, so it’s something you want to get rid of and prevent from occurring,” said Cynthia Lemere, Brigham and Women’s Hospital, Boston, Massachusetts. Researchers are targeting pGlu-Aβ in two ways: by generating antibodies against it and by inhibiting QC with small-molecule drugs. They presented the latest data on both strategies at the 11th Annual AD/PD 2013 Conference held 6-10 March 2013 in Florence, Italy.
Frank Weber of Probiodrug AG, based in Halle/Saale, Germany, presented initial clinical trial data on the company’s QC inhibitor, PQ912. It is the first one to be used in humans. In this single-site Phase 1 study, researchers tested the compound on 108 healthy volunteers, aged 18-50, at either a single daily oral dose of 10 to 1,200 mg or twice daily at 20 to 500 mg for 11 days. Participants took the drug either in liquid or pill form. Twenty-eight people took placebo.
In terms of pharmacokinetics, PQ912 performed quite well, Weber said. It reached a peak concentration in plasma rapidly—in two hours—and declined gradually over 12 hours. At six hours, the drug’s half-life in cerebrospinal fluid (CSF) was longer than in plasma, and CSF concentrations were about 30 percent of those in blood. In all, the drug appeared safe even at the high end of the dose range tested, and produced no serious adverse events thus far. It gave a dose-proportional response, where more inhibitor led to less QC activity measured in blood and CSF.
A critical parameter in drug development is the drug’s effective concentration that produces 50 percent of the maximum response (EC50), said Weber. In this case, researchers wanted to know how much of the inhibitor blocks half of QC’s activity. In both plasma and CSF, 11.3 nM was sufficient to knock out enzyme action halfway. The twice-daily 400 mg dose reached a 24 nM mean concentration in CSF and blocked an average of 70 percent of the enzyme’s activity. Is that enough to bring down pGlu-Aβ formation? Yes, Weber told Alzforum, because given the specific properties of the substrates Aβ3-42 and Aβ11-42, “even a small inhibition of the enzyme leads to a big increase in preventing cyclization of the Aβ peptide.” The research team tested other QC substrates and found that this inhibitor left them alone, even at the highest doses, said Weber.
Given the level of tolerability, the company is conducting a series of tests in elderly people at even higher doses to find out if they absorb or metabolize PQ912 differently. Next, they intend to test the inhibitor in AD patients and measure what it does to biomarkers of disease and cognition in this group. Weber said Probiodrug has yet to decide at what stage to treat, but that it will likely be in the early stages of AD or MCI. As plaques are known to be abundant even in early AD, will the QC inhibitor help? Yes, Weber claimed, because the compound would quell new pGlu-Aβ that forms from peptides that escape deposits.
“The compound looks exciting,” said Wiep Scheper, Academic Medical Center, University of Amsterdam, the Netherlands. “The safety and toxicology look promising, and I found the extensive dose range they tested to be impressive.” About Phase 2 trials with this drug, Scheper said, “It will be important to see which patient group they are going to test—many trials have failed because they targeted patients too late in the course of the disease.”
In response to an audience question, Weber said that PQ912 inhibits isoQC—a closely related form of the QC enzyme—as well. IsoQC cyclizes CCL2, a chemokine that attracts immune cells to atherosclerotic plaques (see Cynis et al., 2011). In previous studies, QC inhibitors have been shown to reduce such pathology. “This inhibitor could have multiple therapeutic uses,” said Lemere, “not just for AD, but for atherosclerosis and other CCL2-driven conditions.” Lemere collaborates with Probiodrug.
Other researchers are exploring ways to deploy antibodies to prevent or rid the brain of pGlu-Aβ after formation. Jeffrey Frost in Lemere’s lab presented mouse behavioral data on the group’s pGlu-Aβ-specific monoclonal antibody, 07/1. These researchers recently reported that the antibody lowered general Aβ burden in APPswe/PS1 ΔE9 mice in both prevention and therapeutic pilot studies (see ARF related news story and Frost et al., 2012). Now they find that the same antibody benefits cognition.
Starting when the mice were about six months old, just about the time they begin to deposit Aβ plaques, Frost injected groups of 11-12 male transgenic mice with either 150 or 500 micrograms of 07/1 or 200 micrograms of a control Aβ antibody, 3A1, weekly for seven months. Twelve transgenic and 12 wild-type mice injected with saline served as controls. At 12 months, the 07/1 high-dose group outperformed the saline-treated transgenics on the water T maze (see Locchi et al., 2007), nearly matching wild-type. Those receiving the 07/1 low dose showed a trend for improvement in contextual fear conditioning compared to the PBS-treated transgenic mice, and performed almost as well as wild-type mice. In general, 07/1-treated mice performed better than 3A1 treated mice. At 13 months, the mice were sacrificed. Relative to transgenic controls, 07/1-treated mice had 20-25 percent less plaque burden in the frontal cortex and 35-45 percent less in the hippocampus. Microgliosis and astrocytosis had come down as well. Based on these results, Frost told the audience that “monoclonal antibodies may trap this toxic form of Aβ, sparing mice from neurodegeneration and cognitive dysfunction.”
Frost did not find pGlu-Aβ in the periphery with use of the antibody, arguing against a peripheral sink effect. One listener asked what mechanism the researchers proposed instead. The antibody may work through microglial phagocytosis, responded Lemere, who chaired the session. Another said that the antibody’s behavioral effects seemed small. Lemere responded that the changes were statistically significant, and that mutant mice sometimes performed up to wild-type standards. Scheper pointed out that pGlu-Aβ levels are lower in mice than in humans, and considers it notable that antibodies against the rarer peptide can remove a significant portion of Aβ deposits. “That points to an important role for pGlu-Aβ in formation of the pathology,” she told Alzforum.
Exactly how pGlu-Aβ relates to plaques—whether it seeds them or forms as a modification on existing plaques later—remains controversial among scientists.
Scheper presented her own recently published work on a potential trigger and subsequent consequences of rising QC levels in aging or AD brains, which she and other groups have shown. In neuroblastoma cells, falling calcium levels in the endoplasmic reticulum—a cellular change that has been reported in human aging—led to elevated mRNA and activity of QC (see De Kimpe et al., 2012). It also resisted degradation in vitro while disrupting the lysosomal membrane in neuroblastoma cells (see De Kimpe et al., 2012). This likely allows lysosomal enzymes to leak into the cytoplasm, Scheper said. In postmortem brain tissue, she found pGlu-Aβ had built up in lysosomes of neurons and glial cells. Such cell biological havoc reinforces the rationale of targeting pGlu-Aβ therapeutically, she told Alzforum.
In Florence, another group presented human data. Milos Ikonomovic, University of Pittsburgh, Pennsylvania, showed that fibrillar pGlu-Aβ correlates with severity of disease and poorer cognition. Ikonomovic analyzed tissue samples from the posterior cingulate cortex of 64 participants in the Religious Orders Study who had had either early AD, mild cognitive impairment (MCI), or had been cognitively normal when they died. In all samples, levels of soluble pGlu-Aβ were about 100-fold lower than soluble Aβ42, probably because the pyroglutamate form aggregates more rapidly, Ikonomovic said. Soluble pGlu-Aβ levels are likely too low to be clinically relevant, and showed no association with progression of disease or cognitive impairment, he said. By contrast, fibrillar pGlu-Aβ, and both soluble and insoluble Aβ42, were more abundant in early AD than in MCI or cognitively normal samples. Higher levels of all three forms correlated with more advanced Braak stages, CERAD pathology scores, and worse performance on tests of cognition and episodic memory. Ikonomovic concurred that pGlu-Aβ may be a viable target for treatment, and said it may have potential diagnostic value.—Gwyneth Dickey Zakaib, with reporting by Madolyn Bowman Rogers.
De Kimpe L, van Haastert ES, Kaminari A, Zwart R, Rutjes H, Hoozemans JJ, Scheper W. Intracellular accumulation of aggregated pyroglutamate amyloid β: convergence of aging and Aβ pathology at the lysosome. Age (Dordr). 2012 Apr 4. Abstract
De Kimpe L, Bennis A, Zwart R, van Haastert ES, Hoozemans JJ, Scheper W. Disturbed Ca2+ homeostasis increases glutaminyl cyclase expression; connecting two early pathogenic events in Alzheimer's disease in vitro. PLoS One. 2012;7(9):e44674. doi: 10.1371/journal.pone.0044674. Epub 2012 Sep 7. Abstract