Identification of candidate IgG biomarkers for Alzheimer's disease via combinatorial library screening

Abstract

The adaptive immune system is thought to be a rich source of protein biomarkers, but diagnostically useful antibodies remain unknown for a large number of diseases. This is, in part, because the antigens that trigger an immune response in many diseases remain unknown. We present here a general and unbiased approach to the identification of diagnostically useful antibodies that avoids the requirement for antigen identification. This method involves the comparative screening of combinatorial libraries of unnatural, synthetic molecules against serum samples obtained from cases and controls. Molecules that retain far more IgG antibodies from the case samples than the controls are identified and subsequently tested as capture agents for diagnostically useful antibodies. The utility of this method is demonstrated using a mouse model for multiple sclerosis and via the identification of two candidate IgG biomarkers for Alzheimer's disease.

Figure 1. Schematic representation of the strategy employed to identify synthetic molecules that capture antibody biomarkers

The Y-shaped figures represent IgG antibodies. The figure depicts hypothetical binding of an antibody present at high levels in an autoimmune serum sample, but not in a healthy serum sample, binding to two compounds on a microarray. After subsequent probing with a fluorescently-labeled secondary antibody, this would produce a much higher intensity at these two spots on the array (indicated in red scale) after exposure to the autoimmune serum sample than the healthy serum sample.

Figure 2. Identification and characterization of peptoids that capture antibodies present at high levels in Mog peptide-immunized mice

A. Raw images of peptoid arrays hybridized with serum obtained from CFA- or CFA +Mog peptide-immunized mice. About half of two arrays are shown at the top. The sections of the arrays boxed in blue are blown up to highlight a region displaying one of the peptoids (AMogP3) that clearly distinguished the CFA- and CFA+Mog peptide-immunized mice. Images were obtained by incubating serum from immunized mice with the array followed by addition of fluorescently labeled (Alexa-647) Goat-anti Mouse IgG antibody. The intensity of the fluorescence at each spot is displayed in a false-colored red scale in which a white spot means the intensity is beyond the linear range of the detector. The structure of AMogP3, the compound that is highlighted in the pink box, is shown as its free form. The molecule was tethered covalently to the array via the cysteine sulfur that is included in all of the molecules in the library. B. Quantitation of the fluorescence intensity measured at each of the three peptoid (AMogP1-3) features on the array that discriminate CFA + Mog peptide- from CFA-immunized mice. The bars depict the mean ± the standard deviation for three independent experiments. The general structure of the library employed to make the array shown in Supp. Fig. S1. The structures of the other two peptoids, AMogP2 and AMogP3, that distinguish control and EAE mice, are shown in Supp. Fig. S2

Figure 3. Validation of peptoids AMogP1-3 as capture agents for EAE-specific antibodies

A. “Sub-arrays” containing only AMogP1-3, the Mog peptide and a MBP-derived negative control peptide were created. Serum from seven Mog/CFA-immunized and seven CFA-injected mice not used in the previous experiments were analyzed in a blinded fashion. The fluorescence intensity observed at each feature is shown after unblinding the sample identities. Shown is the mean ± S.D. for samples run in triplicate. B. Raw images of sub-arrays containing the AMogP1-3, control peptide and Mog peptide that were incubated with serum from a Mog + CFA-immunized mouse (left) or a CFA-immunized mouse (right). C. Selectivity of peptoids for antibodies present in Mog peptide-immunized mice. Sub-arrays containing the AMogP1-3 peptoids, the Mog peptide, the Ova peptide and a control peptide were exposed to serum from three mice immunized with Ova peptide (Ova1-3) or three mice with SLE (SLE1-3) followed by a fluorescently labeled secondary antibody. The fluorescence intensities at each feature are shown. Mean ± S.D. for samples run in triplicate. Supp. Fig. S2 displays the peptide sequences and peptoid structures.

Figure 4. Peptoids AMogP1-3 capture anti-Mog peptide antibodies resulting from an adaptive immune response

A. Level of IgG antibody captured by the peptoids as a function of time after immunization. Sub-arrays displaying the molecules indicated were incubated with serum collected from mice at the indicated times after immunization with Mog peptide+CFA, followed by fluorescently labeled secondary antibody. The amount of fluorescence captured at each feature is shown. B. Effect of depletion of anti-Mog peptide antibodies on the amount of IgG antibodies captured by the peptoids. Serum from Mog peptide + CFA-immunized mice was passed over columns displaying either excess Mog peptide or a control peptide. These Mog depleted or mock depleted serum samples were then hybridized to a sub-array displaying AMogP1-3, Mog peptide and a control molecule. After subsequent hybridization with labeled secondary antibody, the signal intensities were recorded and plotted. Mean ± S.D. for samples run in triplicate.

Figure 5. Validation of peptoids identified as biomarkers of ovalbumin immunoreactivity

A. Sub-arrays” containing only AOvaP1-3, the Ova peptide and a control peptide were created. Serum from seven Ova peptide + CFA-immunized and seven CFA-injected mice not used in the previous experiments were analyzed in a blinded fashion. The fluorescence intensity observed at each feature is shown after unblinding the sample identities. B. Raw images of sub-arrays containing the control peptide, AOvaP1-3 and Ova peptide that were incubated with serum from Ova + CFA-immunized mouse (left) or a CFA-immunized mouse (right). C. Selectivity of peptoids for antibodies present in Ova peptide-immunized mice. Sub-arrays containing the AOvaP1-3 peptoids, the Mog peptide, the Ova peptide and a control peptide were exposed to serum from three mice immunized with Mog peptide (Mog1-3) or three mice with SLE (SLE1-3). The fluorescence intensities at each feature observed after probing with the fluorescently labeled secondary antibody are shown. Error bars represent the mean ± S.D. for samples run in triplicate. The structures of the Ova peptide antigen and the peptoids that distinguish Ova-immunized from control mice (AOvaP1-3) are shown in Supp. Fig. S2. Supp. Fig. S3 displays some of the primary data that led to the identification of AOvaP1-3 as discriminators of mice that were and were not immunized with Ova peptide. Supp. Fig. 4 demonstrates that peptoids AOvaP1-3 bind anti-Ova peptide antibodies.

Figure 6. Peptoids that retain antibodies from the serum of patients with Alzheimer's Disease

A peptoid library was screened for ligands to AD-specific IgG antibodies. The structures of the three best peptoids that were found to discriminate age-matched controls and AD patients are shown in the top right. The levels of antibodies retained from the indicated serum samples in subsequent sub-array experiments are shown on the left. The numbers indicate a patient identifier (for example, AD1 or NC9; only every other number is shown). The samples employed in the training sets are labeled as such (AD Train and NC Train) as are the samples employed in blinded test studies. AD = Alzheimer's Disease, NC = normal control. The error bars indicate the mean ± S.D. for samples run in triplicate. See text for details. Supp. Tables S2-S6 and Supp. Fig. S5 present a detailed statistical analysis of these data as well as those shown in Fig. 7A. Supp. Fig. S6 demonstrates that the intensities shown in this figure represent the high end of the linear range of the assay.

Figure 7. Peptoids ADP1-3 bind two different antibodies that are present in the serum of Alzheimer's patients, but not patients with Parkinson's Disease (PD) or lupus (SLE)

A. Comparison of levels of IgG antibodies captured by peptoids ADP1-3 from serum samples collected from a patient with AD (individual 1), a normal control (individual 23) or patients with Parkinson's Disease (PD) or lupus (SLE). B. Serum from an autopsy-confirmed AD patient was passed repeatedly over immobilized ADP1 or, as a control, an irrelevant peptide. The serum samples were then diluted and hybridized to sub-arrays displaying peptoids ADP1-3. The amount of antibody captured by each peptoid was measured. Shown is the Mean ± S.D. for samples run in triplicate. Supp. Tables S2-S6 and Supp. Fig. S5 present a detailed statistical analysis of these data as well as those shown in Fig. 6.