Affiliations: [a] Department of Public Health and Caring Sciences/Molecular Geriatrics, Rudbeck Laboratory, Uppsala University, Uppsala, Sweden | [b] Mabtech AB, Nacka Strand, Sweden
Correspondence:
[*]
Correspondence to: Frida Ekholm Pettersson, Department of Public Health and Caring Sciences/Molecular Geriatrics Rudbeck Laboratory, Dag Hammarskjölds väg 20 751 85 Uppsala, Sweden. Tel.: +46 (0)18 471 48 02; E-mail: [email protected].
Note: [1] These authors have contributed equally to the work.
Abstract: Amyloid-β (Aβ) oligomers of different sizes and forms have recently been the focus for many Alzheimer's disease (AD) researchers. Various immunoassays have been used to detect low concentrations of these elusive Aβ species in different forms of human samples using little or no sample dilutions. However, the possibility that positive results may be caused by interference from heterophilic antibodies (HA) is often overlooked. HA, which recognize immunoglobulins from other species, are present in human plasma and cerebrospinal fluid (CSF) and may cause interference in sandwich immunoassays like enzyme-linked immunosorbent assays (ELISAs) by cross-binding the capture and detection antibodies of the assay. They thus may generate a false positive signal. Here we show that when assessing the Aβ oligomer content in plasma samples from 44 individuals with a sandwich ELISA, none of the 21 positive signals remained when the assay was repeated in the presence of factors blocking HA. Similarly, in CSF samples from 104 individuals, the signals from the 22 positive samples were strongly reduced when analyzed after anti-HA treatment. Taken together, HA interference is a problem that needs to be addressed when measuring low levels of an antigen in human plasma and CSF samples.
