Ap immunotherapy prevents and Reverses Alzheimers disease neuropathology

Dale Schenk1, Dora Games1, and Peter Seubert

By the late 1990s, research into the role of beta amyloid peptide (Ap) in Alzheimer's disease uncovered a number of important findings. It was a key pathological feature of the disease and nearly all known familial forms of the disorder appeared to either directly or indirectly increase levels of the more amyloidogenic Ap 42 in the brain tissue of affected individuals. The precise details of the Ap generating proteases, beta and gamma secretase, were not yet understood, but their existence was fairly certain. Most importantly, a few transgenic mouse models finally existed that demonstrated fairly robust progressive amyloidosis.

In the mid 1990s, we were fortunate enough to be working with one of these models, the PDAPP mouse (Games et al. 1995), and we asked the question of what experiments might be reasonable to do with the model to better understand beta amyloid plaque formation and neuropathology or, more ambitiously, if it was possible to intervene therapeutically in these pathogenic processes. Many ideas were put forth for possible experiments in this mouse model: for example, trying to understand whether the cholinergic system or glutamatergic systems were dysfunctional as well as testing a number of compounds, such as NSAIDS, for which there already existed compelling epidemiological efficacy data for AD. We certainly did not have enough animals to test all of the ideas put forth, since breeding mice has always been a challenge for large-scale experiments; thus we had to rank various experiments for their potential merit and likelihood of success. It was against this backdrop that the idea of immunizing the mice with the Ap peptide itself was suggested.

The rationale behind this idea as a potential therapeutic approach was that binding of antibodies to antigens could effectively reduce the free levels of unbound peptide. This idea, coupled with the fact that virtually all plasma proteins enter the brain at approximately 0.3% of the plasma levels, led to calculations that a circulating antibody titer of 1:50,000 would achieve a titer of 1:1,500 in the brain, possibly enough to exert a biological effect. Also, if antibody did reach the brain in sufficient quantities, it was conceivable that it might disrupt the formation of fibrils, as had been shown previously in vitro (Solomon et al. 1997). Despite supportive data at the time, one did have to make the leap to assume that the reduction in free levels of Ap would be sufficient to reverse or reduce plaque burden. This was indeed a reasonable assumption to prevent plaque formation, but perhaps far less likely to reverse any existing plaque. The hurdle of the blood-brain barrier, together with the insolubility of amyloid plaques and the frank probability of failure of such an experiment, resulted in the bestowing of the lowest

1 Elan Pharmaceuticals, 800 Gateway Bld., CA 94080 South San Francisco, USA

priority ranking by our internal review process to this effort, out of all of the numerous suggestions put forth to be tried in mice.

Fortunately, through tenacity, a few extra animals were secured and the experiments were performed (Schenk et al. 1999). Specifically, PDAPP mice were immunized with Ap 42 plus adjuvant beginning at a very young age (6 weeks), prior to any plaque deposition, and were boosted monthly with immunogen until the mice were sacrificed at the age of 13 months. As a control, we immunized a group of mice with a fragment of serum amyloid protein (SAP). At the conclusion of the experiment, the brains of these mice were examined for plaque burden, astocytosis and microgliosis. Surprisingly, the mice that had been immunized with Ap peptide were essentially devoid of amyloid plaques. The result was so striking that our first reaction was to reconfirm the transgenic status of the Ap-immunized animals. We also examined alternative plaque detection methods, such as Congo red and thioflavin stains, which were also negative on the Ap-immunized brains. Remarkably, the simplest conclusion for this first experiment was that the immunization with Ap peptide had resulted in antibodies that had somehow blocked the formation of amyloid plaques. Immediately upon seeing these results, we set out to do an even more critical and difficult experiment. Rather than immunizing very young mice that had no pathology at the beginning of the experiment, we initiated immunization with Ap peptide at the age of 11 months (i.e., plaque-bearing animals) and continued treatment until the age of either 15 or 18 months. The expectation was that this would be a much more stringent test of the therapeutic potential of Ap immunotherapy, since the low concentration of antibodies expected to enter the brain would now have to block elongation of existing fibrils of Ap rather than stop the initiation of new ones. Analysis of the brain tissue of the PDAPP mice that had been immunized again demonstrated that not only did immunization with Ap peptide block the increase in further plaque formation but it also appeared to have actually eliminated existing plaques. This impression, based on histological images from single time points, would be elegantly confirmed in vivo (Backsai et al. 2001).

From the perspective of the role of amyloid burden in the PDAPP mouse and its relevance to AD, we also observed in these studies that, when amyloid burden was reduced, both dystrophic neurites burden and astrocytosis were simultaneously reduced. Reassuringly, this remarkable early preclinical observation would also be found years later to hold true for patients suffering from the disease who were treated with Ap immunotherapy (Ferrer et al. 2004; Masliah et al. 2005a; Nicoll et al. 2003).

Perhaps the most unexpected finding from these studies was that, in the Ap-immunized mice, microglia appeared to have taken up Ap peptide in far greater amounts than previously seen in either the transgenic mice or in AD brain tissue itself. Another result suggested in this initial report, and demonstrated to be true in subsequent papers, was that Fc receptors on microglial cells engaged via antibodies that were bound to Ap plaques represent a powerful plaque-clearing mechanism.

It is fair to say that the two experiments described in the original publication inspired many further studies in laboratories around the globe that were interested in Ap peptide and AD (Arendash et al. 2001; Janus et al. 2000; Morgan et al. 2000; Schenk et al. 2004; Wilcock et al. 2004b). Perhaps most importantly, it was now possible to specifically reduce the burden of beta amyloid plaques in the APP transgenic mice and to assess a variety of outcomes. These studies also allowed numerous laboratories to better understand factors involved in the process of amyloidosis and its reversal.

Aß immunotherapy prevents and Reverses Alzheimer's disease neuropathology 421

Finally, they opened a potentially rich and complex therapeutic avenue for treatment or even prevention of AD that was unanticipated. Each of these areas will be briefly discussed here, though they are sufficiently complex that space will not allow a truly appropriate review.

Perhaps the most immediate question raised by the experiments described in the first paper was whether the reduction in plaque burden was attributable to anti-Aß antibodies or somehow caused by a T-cell response to Aß. This question was unambiguously resolved by showing that passive treatment of PDAPP mice with monoclonal antibodies to Aß could demonstrate reduction in plaque burden and related neu-ropathologies (Bard et al. 2003). This same paper also demonstrated that anti-Aß antibodies do enter the brain and bind to plaques and then engage Fc receptors that mediate plaque removal. In fact, the antibody-mediated process of Aß removal and elimination could be demonstrated in vitro using brain sections and microglial cells.

The question of what plaque burden might have to do with cognitive capabilities -a key burning question, and one that is dealt with elsewhere in these reviews - was quickly and accurately assessed by two groups simultaneously, with similar results, in 2001 (Janus et al. 2000; Morgan et al. 2000). Both groups nicely demonstrated that Aß immunization could reduce the loss of cognitive performance typically seen in a number of different APP/presenilin mouse models. These important findings provided further impetus for testing the immunotherapeutic approach in AD patients. Many immunotherapy-based papers have followed. Most of these have attempted to addresswhichformsofAß are exerting various cognitive impairments and to determine the mechanism by which various anti-Aß antibodies exert their beneficial effects. Use of different mouse models and various anti-Aß antibodies has proven valuable in unraveling these questions.

The question of precisely why plaques form initially has not yet been resolved, but the underlying mechanisms involved in the reduction of Aß plaque burden by immunotherapy have been partially resolved. Current findings suggest that, in vivo, at least three mechanisms are responsible for the ability of anti-Aß antibodies to reduce plaques. The first is simple destabilization of existing plaques by physically binding to them, as most directly shown in vivo by Backsai et al., who demonstrated that injection of Fab fragments of anti-Aß monoclonal antibodies removed existing plaques, as shown in living animals by dual photon confocal microscopy (Backsai et al. 2001). The precise biophysical principles of this phenomena are not fully understood, but it is conceivable that antibodies binding to the free C- or N-terminal regions of Aß, which are accessible even when the peptide forms a fibril, change the conformation of the peptide such that the fibril is destabilized and disassociates. The second mechanism, already discussed and cited, is Fc-mediated phagocytosis by microglia cells. This is likely to be catalytic in the sense that a few antibodies bound to a plaque are likely sufficient for the microglial cells to engage and engulf a significant fraction of the entire plaque, making the process fairly efficient. A third component is simple reduction in the free concentration of Aß peptide, as was initially anticipated in the rationale for the first immunization experiment. This mechanism has been coined the "sink" hypothesis, with the concept that since antibodies remain predominantly in the blood, they essentially serve to draw out the Aß from the brain through mass action (De Mattos et al. 2001). Though the concept is attractive, it is difficult to test in vivo since all antibodies tested enter the brain at some low level and the other mechanisms described above will likely also enter into the observed effects in vivo.

Perhaps the most important question underlying Aß immunotherapy is determining its potential clinical utility. The first clinical test of the potential of immunization of Aß as a possible treatment for AD used a synthetic version of Aß 42 termed AN 1792. This agent was used in a number of phase 1 clinical trials where its tolerability and safety were investigated (Bayer et al. 2005). Following these early studies, it moved into a relatively large, phase 2 multicenter, placebo-controlled, double-blind study to investigate additional safety and pilot efficacy of the approach, although it was not powered to test for efficacy on the standard clinical endpoints in Alzheimer's disease such as ADAS-COG. Early in the phase 2 study, after a large majority of patients had received two doses of AN 1792, two cases of meningo-encephalitis occurred followed rapidly by two more. Dosing was immediately halted, although the study remained blinded and was converted essentially into a monitoring safety study that would still investigate exploratory endpoints at the 12-month time point (Gilman et al. 2005). The results of the study have been extensively discussed elsewhere, but several biological signals occurred in this study that represent very original observations that have not been seen before in therapeutic trials in AD patients. For example, in patients that generated reasonable antibody titers to Aß (greater than 1:2,200), CSF levels of tau where reduced towards normal values, volumetric MRI values were reduced (Fox et al. 2005), a composite neuropyschometric battery of tests showed improvement and autopsy analysis showed plaque burden to be reduced and evidence of active amyloid clearance (Ferrer et al. 2004; Masliah et al. 2005a). Collectively, these results are consistent in many respects with what has been seen in APP transgenic mouse models of AD, with the notable exception that the meningo-encephalitis had not been predicted. The effects on cognition in AD patients were far too preliminary, given that drug dosing in the study was interrupted, to infer whether AN 1792 did or did not have a convincing effect overall on the patients' performance. Nonetheless, this initial clinical testing, in addition to all the progress since the initial preclinical report (Schenk et al. 1999), has resulted in a large number of ongoing clinical trials worldwide. The most advanced of these is Bapineuzumab, a humanized anti-Aß antibody, currently in phase 2 studies in the US and imaging studies in Europe, to investigate safety, tolerability and initial exploratory efficacy in mild to moderate AD.

The initial preclinical study (Schenk et al. 1999) has indeed initiated a great many studies in both the discovery and applied clinical fields of AD. Relative to the lon-g-term investigation of the role of Aß in AD, which is now almost 20 years of age, we should have an unambiguous answer regarding the clinical utility of Aß immunotherapy within the next five years.

The general concept of immunizing with an amyloidogenic protein or peptide for treatment of a disease has also been recently expanded to a number of different disease classes, such as prion protein biology (Heppner et al. 2001) and Parkinson's disease (Masliah et al. 2005b). It is earnestly hoped that this general approach will generate a number of additional new therapeutic strategies for a class of diseases that have remained refractory to a large number of potential treatments thus far.

Christian Haass (left) and Roger M. Nitsch (right)
All About Alzheimers

All About Alzheimers

The comprehensive new ebook All About Alzheimers puts everything into perspective. Youll gain insight and awareness into the disease. Learn how to maintain the patients emotional health. Discover tactics you can use to deal with constant life changes. Find out how counselors can help, and when they should intervene. Learn safety precautions that can protect you, your family and your loved one. All About Alzheimers will truly empower you.

Get My Free Ebook


Post a comment