Roh JH, Holtzman DM.
Stealth attack: plaque-specific antibody allows for efficient Aβ removal without side effects.
Neuron. 2012 Dec 6;76(5):859-61.
Please login to recommend the paper.
To make a comment you must login or register.
Passive immunization is one of the central themes for AD therapy. The question is, What is the appropriate target? The paper by DeMattos and colleagues argues that using mE8-IgG2a, a monoclonal antibody (mAb) specific for the pyroglutamate Aβ (pGluAβ) found in plaques is a new tool for therapy. The main message is that the pGluAβ mAb reduces deposited amyloid without inducing microhemorrhage.
When we look at the history behind Aβ immunotherapy, we see that first the AD field believed that plaques are toxic and their removal beneficial for the patients. That turned out to be false and produced significant side effects like hemorrhages. Next, the field turned to soluble oligomers of different size, suggesting that they should be targeted. There is convincing evidence for that notion, in my view. Especially, N-truncated forms of Aβ starting with pyroglutamate proved to be one of the seeding oligomers.
Already in 2010, we generated a monoclonal antibody (9D5) that selectively recognizes oligomeric assemblies of pGluAβ and studied their potential involvement in disease. Passive immunization of 5xFAD mice with 9D5 significantly reduced overall Aβ plaque load and pGluAβ levels, and normalized behavioral deficits. These data indicated that 9D5 is a therapeutically effective monoclonal antibody targeting low-molecular-weight pGluAβ oligomers. These oligomers are likely to affect other Aβ species and act as seed for oligomerization in vivo (Wirths et al., 2010b). 9D5 does not react at all with monomers or dimers, and plaques are only very rarely detected with this antibody. This is in contrast to pGluAβ antibodies 2-48 and 1-57 we have previously made that react with all forms and, of course, detect all plaques in sporadic and familial AD and two mouse models (Wirths et al., 2010a).
Cindy Lemere has demonstrated that passive immunization of APPswe/PS1ΔE9 transgenic mice with a highly specific monoclonal antibody against pGluAβ significantly reduced total plaque deposition and appeared to lower gliosis in the hippocampus and cerebellum in both prevention and therapeutic studies. Insoluble Aβ levels in brain homogenates were not significantly different between immunized and control mice. Microhemorrhage was not observed with anti-pyroglutamate-Aβ immunotherapy (Frost et al., 2012).
Now we learn that plaques are again an interesting target, this time with an antibody that recognizes all forms of pGluAβ aggregated in plaques. No side effects, such as hemorrhages, were observed after passive immunization.
This work published by Ronald DeMattos and colleagues is interesting, but leaves the field with many questions unanswered. Mostly neuropathological and biochemical data were presented. The most important experiment, i.e., whether the treatment is beneficial for the mice or not, has not yet been done. Does the pGluAβ antibody rescue, or at least normalize, behavioral deficits as 9D5 did? What effects can be seen on synaptic deficits and neuron loss?
Such questions could be addressed using different available pGluAβ antibodies and studying their therapeutic potential in mouse models for AD. Our oligomer-specific antibody 9D5 is commercially available at Synaptic Systems, Goettingen, Germany.
This paper places an underestimated Aβ species center stage: the pyroglutamated Aβ (pEAβ).
In general, there are several possible approaches to reduce pEAβ species:
1. Preventing such modification by using small molecule inhibitors of glutaminyl cyclase (QC), the enzyme responsible for the modification (Schilling et al., 2004).
2. Clearing such already modified or continuously forming species by active or passive immunization.
According to the mechanism presented in this study, the pEAβ species are seen simply as plaque-specific docking points for immunotherapy and as such being indeed very effective. Work from other groups and ours suggests, however, that these pEAβ species play an important role in the genesis of pathology by way of their significant toxic and seeding potential (Wirths et al., 2009; Morawski et al., 2010; Hartlage-Rübsamen et al., 2011, Alexandru et al., 2011). This view is not represented in the discussion, nor have the available data been adequately acknowledged in the present study.
Moreover, we would like to point out that these pEAβ species cannot simply occur spontaneously. The half-life of the spontaneous oxoproline ring formation means that this process would, under physiological conditions, take years to decades (Seifert et al., 2009; Jawhar et al., 2011). Glutaminyl cyclase activity is required to generate these species readily in living beings, and this enzyme is upregulated early in AD (Schilling et al., 2008, Jawhar et al., 2011; De Kimpe et al., 2012a; De Kimpe et al., 2012b, Valenti et al., 2012), driving or driven also by inflammatory processes (Cynis et al., 2011).
Many studies have clearly shown that in human brain, abundant pE modification of Aβ speeds up aggregation of the peptide. This correlates with its occurrence in deposits, which are detectable by β amyloid-directed PET labels (Maeda et al., 2007), and with its accumulation with disease progression (Pivtoraiko et al., 2012). In contrast, pEAβ is rarely present in CSF and plasma. Moreover, several experiments indicate that diffusible ligands are formed from pEAβ in vitro (Schlenzig et al., 2012) and can be extracted from AD tissue (Piccini et al., 2005, Nussbaum et al., 2012). Diffusible oligomers containing pEAβ forms potently interfere with LTP and neuron viability, as has been recently demonstrated (Nussbaum et al., 2012).
The major driver of these characteristics is the increased surface hydrophobicity of these species—a feature that has also been linked to the toxicity of other amyloid peptides (Schlenzig et al., 2012). In that regard, the study by DeMattos at al. leaves open an important point. Are such oligomers recognized by the antibody, and does this result in functional improvement? Answering this question would strengthen the paper’s conclusions significantly.
The present study concludes that pEAβ does not serve as a species that provokes buildup of deposits. We would respectfully submit that this conclusion is inadequately supported by experimental evidence in this study. The preventive trial with the mE8 antibodies was stopped at an age when PDAPP mice were previously described just to start the deposition; hence, it was not possible to measure how the antibody would have affected further amyloid buildup in the months to come. In addition, the data presented on total Aβ in the prevention and in the therapeutic studies using the mE8 antibodies are difficult to interpret, since the extraction method used does not allow a differentiation between deposited and soluble material at the younger and the older ages of the treated PDAPP mice.
In contrast, preventive passive immunization of APP/PS1 mice has been shown to reduce early plaque load significantly (Frost et al., 2012). This result is further supported by novel double transgenic mouse lines (FAD42 and 5xFAD/hQC; Jawhar et al., 2011; Wittnam et al., 2012) that exhibit early pathology and memory impairment caused by QC-induced pEAβ formation. Also, our own studies applying QC inhibitors in a preventive manner ameliorated general Aβ pathology and behavior impairment late, but also early, during the progression of Aβ formation, aggregation, and deposition (Schilling et al., 2008).
Our lead QC inhibitor has nearly completed Phase 1 studies in Europe with excellent data on safety, tolerability, and target engagement.
Finally, contrasting our data with mE8, our pEAβ-specific monoclonal antibody shows plaque lowering, and this reduction appears to correlate with an improvement in behavior also early on (Lemere et al., personal communication). Our antibody specifically recognizes monomeric, oligomeric, fibrillar pEAβ, and mixed Aβ material in vitro. Accordingly, the antibody reduced Aβ in monotherapeutic prophylactic and therapeutic preclinical trials (Frost et al., 2012).
In conclusion, the presented study indeed offers novel and welcome perspectives for the field of immunotherapy. The data published to date remain, however, inconclusive with regard to pEAβ’s role in toxicity.
References: Lemere C. personal communication (2012).