The Alzheimer's disease-associated amyloid beta-protein is an antimicrobial peptide

Abstract

Background: The amyloid beta-protein (Abeta) is believed to be the key mediator of Alzheimer's disease (AD) pathology. Abeta is most often characterized as an incidental catabolic byproduct that lacks a normal physiological role. However, Abeta has been shown to be a specific ligand for a number of different receptors and other molecules, transported by complex trafficking pathways, modulated in response to a variety of environmental stressors, and able to induce pro-inflammatory activities.

Methodology/principal findings: Here, we provide data supporting an in vivo function for Abeta as an antimicrobial peptide (AMP). Experiments used established in vitro assays to compare antimicrobial activities of Abeta and LL-37, an archetypical human AMP. Findings reveal that Abeta exerts antimicrobial activity against eight common and clinically relevant microorganisms with a potency equivalent to, and in some cases greater than, LL-37. Furthermore, we show that AD whole brain homogenates have significantly higher antimicrobial activity than aged matched non-AD samples and that AMP action correlates with tissue Abeta levels. Consistent with Abeta-mediated activity, the increased antimicrobial action was ablated by immunodepletion of AD brain homogenates with anti-Abeta antibodies.

Conclusions/significance: Our findings suggest Abeta is a hitherto unrecognized AMP that may normally function in the innate immune system. This finding stands in stark contrast to current models of Abeta-mediated pathology and has important implications for ongoing and future AD treatment strategies.

Figure 1. Growth of E. faecalis is inhibited by Aβ42.

E. faecalis were cultured alone (circle) with 25 µg/ml of Aβ42 (triangle) or LL-37 (diamond). Panel A; Bacterial growth with time was monitored by inoculation of agar with diluted incubants and counting CFU. Representative data from six experiments is shown as mean signal of four replicates ± s.e.m. Panel B Incubants were monitored for Aβ42 and LL-37 by Western blot with mAb 6E10 or anti-LL-37. The figure shows representative signal for Aβ42 (odd lanes) or LL-37 (even lanes) incubants from six replicate experiments.

Figure 2. E. faecalis pre-incubated with Aβ42 are mAb 6E10 immunoreactive.

Bacteria were incubated (1 hr at 37°C) with (panel A) or without (panel B) Aβ42 (25 µg/ml). Following repeated washes, the bacteria were fixed onto glass slides and immunostained with the HRP conjugated anti-Aβ antibody (mAb 6E10-HRP).

Figure 3. AD brain homogenates have increased antimicrobial activity against C. albicans.

AD and non-AD brain samples were tested for Aβ-mediated inhibition of C. albicans. Samples of temporal lobe (Temp. L.) and cerebellum (Cereb.) from AD (n = 32) and age-matched control subjects (n = 13) were homogenized in culture broth. Panel A; Homogenates were inoculated with log-phase C. albicans and microbial growth determined by alamar blue viability assay. Data is shown as percentage of signal for C. albicans alone (average of four replicates) ± s.e.m. Panel B; Homogenates were assayed for Aβ40 and Aβ42 by commercially available ELISA. Graph shows Aβ signal (sum of Aβ40 and Aβ42) against C. albicans growth for temporal lobe homogenates from combined AD and non-demented cohorts (n = 42). Probability analysis used unpaired two-tailed t-tests (p). Correlation was determined by calculating the Pearson r correlation coefficient (r).

Figure 4. Immunodepletion of Aβ from AD brain homogenates attenuates C. albicans inhibition.

Homogenates of temporal lobe (Temp. L.) and cerebellum (Cereb.) were prepared from AD (n = 32) or non-demented (n = 13) subjects. AD (AD) or non-demented (non-AD) homogenates were pooled and then incubated with Magno-beads pre-loaded with rabbit IgG (IgG) or a polyconal rabbit anti-Aβ antibody (α-Aβ). Following bead removal samples were analyzed for Aβ signal by Western blot and assayed for C. albicans growth by alamar blue viability assay. Panel A shows C. albicans growth in treated homogenates as a percentage of signal in culture broth alone. Immunodepletion of AD temporal lobe homogenates with α-Aβ restored microbial growth to levels equivalent to non-demented control samples. Graph shows average of five replicates ± s.e.m. Panel B; Untreated and immunodepleted homogenates (1∶16 dilution) were Western blotted and probed with the Aβ-specific mAb 4G8 antibody. Analysis confirmed Aβ signal was reduced in temporal lobe homogenate incubated with anti-amyloid β-peptide antibody (Lane 1) compared to sample incubate alone (Lane 2) or with rabbit IgG (Lane 3). Aβ in dilutions of cerebellum homogenate is below the level of detection for our experimental conditions (Lanes 4–6). Blots included synthetic Aβ42 (Aβ42) standard (Lane 7). Statistical probability analysis (p) of data used unpaired two-tailed t-test.