Alzheimer's disease (AD) is a progressive neurodegenerative disorder characterized by the presence of classical lesions in different brain regions of the neocortex and hippocampus [1]. Among these lesions, the amyloid plaques formed by the aggregation of amyloid-β (Aβ) peptides are prominent. The pathogenesis of the disease is complex and is driven by both environmental and genetic factors. Although most of the cases are sporadic with an obscure etiology, some forms of the disease are inherited and several genes are implicated in familial Alzheimer's disease (FAD). The understanding of the molecular basis of the disease gained significant knowledge with the observation that mutations in the three genes linked to early-onset autosomal dominant FAD, increase the production of a highly insoluble isoform of Aβ. Together, these observations support the so-called amyloid hypothesis in the pathogenesis of Alzheimer's disease [2, 3]. Sequential endoproteolysis of the amyloid-β precursor protein (APP) by the aspartic protease β-secretase/BACE1 and by the γ-secretase proteolytic complex leads to the production of Aβ (see Figure 1A). In an alternative non-amyloidogenic pathway, APP is endoproteolyzed within the Aβ region by α-secretase to produce the secreted sAPPα fragment (Figure 1A) [4, 5].

Numerous primary and immortalized cell lines have been used to analyze APP processing and Aβ production in in vitro cultures. These cell culture systems have proven to be indispensable for the identification of pharmacological and genetic modifiers of APP metabolism prior to in vivo studies [6–8]. It is estimated that 15 to 35% of cell cultures in current use are infected with mycoplasmas [9]. With a diameter of about 0.2 – 0.4 μm, mycoplasmas are small, slow-growing bacteria that are generally unaffected by the antibiotics used against common bacteria and fungi. They can go undetected for long periods of time as the contaminated cells may grow normally and appear normal by light microscopy. Mycoplasma contaminations can however have negative effects, ranging from inhibition of metabolism and growth to induction of malignant transformation or apoptosis [10, 11]. In this study, we show that mycoplasmas can degrade extracellular Aβ. These results indicate that mycoplasma contaminations can introduce an unsuspected source of variability in Aβ level measurements in cultured cells.

Degradation of extracellular Aβ by mycoplasma contaminants

To first assess the effect of mycoplasmas on the levels of Aβ in cell culture, medium from mycoplasma-positive cells, detected by polymerase-mediated amplification of mycoplasma genomic DNA (Figure 2A), was diluted 1:1 with conditioned medium from APP-transfected HEK293 cells. The resulting mixture was then incubated at 37°C for different periods of time. Under these conditions, we observed a robust degradation of Aβ after 1 hr of incubation, while sAPPα levels were not affected (Figure 2B, lanes 3–6). Aβ degradation did not occur when fresh medium (Control DMEM) was mixed with Aβ-containing medium (Figure 2B, lanes 1 and 2). Importantly, reduction of the contaminant load by a 0.2 μm filtration of the mycoplasma-positive medium prior to incubation, significantly prevented Aβ degradation (Figure 2B, lanes 7–10). Together these data strongly indicate that mycoplasmas efficiently degrade Aβ during normal cell culture conditions.

Mycoplasma contamination prevents Aβ accumulation in APP-transfected cells

In order to confirm the detrimental effect of mycoplasma contamination on Aβ accumulation in cell culture, we directly analyzed Aβ levels produced by mycoplasma-positive APP-transfected HEK293 cells. Interestingly, no Aβ was detected in the conditioned medium of the mycoplasma-positive cells (Figure 3A, lane 2), while these cells strongly overexpressed APP (Figure 3A, upper panel, lanes 1 and 2).

Aβ peptides are degraded in cell culture systems and in vivo by at least four peptidases from the metallopeptidase family, neprilysin, endothelin-converting enzyme-1 and -2, and insulin-degrading enzyme [15]. Bacteria express numerous proteases with important biological activities, such as degradation of tissue matrix proteins or activation of zymogens through limited proteolysis. The vast majority of the contaminations in mycoplasma-positive cells are due to M. hyorhinis, M. orale, M. arginini, M. fermentans, M. hominis and Acholeplasma laidlawii [12, 13]. Partial sequencing of the M. hominis genome predicted the expression of at least three metallopeptidases, and M. penetrans, a mycoplasma species isolated from human, is predicted to express oligopeptidase O1, a metallopeptidase from the M13 neprilysin family. Here M. hyorhinis was identified in the contaminated APP-transfected cell lines. It is conceivable that M. hyorhinis also expresses metallopeptidases with homologies with known mammalian metallopeptidases, which degrade Aβ but not large secreted proteins, such as sAPPα. Contamination with mycoplasmas would therefore provide a very efficient peptidase-driven mechanism of Aβ clearance.

In conclusion, we show that (i) mycoplasmas from cell culture contaminations degrade in vitro Aβ produced in cell lines, (ii) mycoplasma-positive APP-transfected cells do not accumulate Aβ in the conditioned medium, and (iii) eradication of the mycoplasma contaminant (i.e. M. hyorhinis) by treatments with a quinolone-based antibiotic, restored extracellular Aβ accumulation in APP-transfected cells. Together these results demonstrate that mycoplasmas may represent a significant source of variability when comparing extracellular Aβ levels between different cell lines. We therefore recommend assessment of mycoplasma contamination prior to extracellular Aβ level measurements in cultured cells.