Fig 1.
Postulated mechanism of amyloid β (Aβ)-mediated modulation of the alternative complement cascade [19, 20].
Schematic shows the alternative complement pathway. Magenta text and cross- interaction of Aβ with complement factor I (CFI) causes reduction of its enzymatic activity and a reduction of the conversion of C3bH to iC3b. Light Blue text- Degradation pathway of C3b to C3d via C3bH, modulated by complement factor H (CFH) and CFI. Black text- conversion of C3b to C3 convertase. Orange-Amplification loop C3 to C3a + C3b. Red-Termination phase resulting in conversion of C5 to C5b78(9)n (membrane attack complex [MAC]). Aβ directly and indirectly produces a local inflammatory environment in retinal pigment epithelial (RPE) cells by modulating release of MCP-1, which leads to recruitment of macrophages and microglia locally, and production of TNF-α and IL1β. The action of these factors on the RPE leads to the release of complement factor B (CFB, green), which in turn is the second mechanism leading to activation of the alternative complement cascade [20].
Fig 2.
Schematic representation of phase I clinical study of Alzheimer’s disease (AD).
Please refer to (BA106006) NCT00459550 and to Andreasen et al. (2011)[23]) for more details of complete study. AD patients received either anti-amyloid β monoclonal antibody, GSK933776 6 mg/kg or placebo. The schematic shows the samples that were tested in the CFI bioactivity assays described in the text. Only samples from days 29–31 (1st and 2nd dose periods), and days 57–59 (2nd and 3rd dose periods) from subjects receiving GSK933776 or placebo were analyzed. A total of 5 subjects per group were assayed. Only 3 samples were available at certain time points. Red arrows indicate samples measured. Reprinted from Andreasen et al. [23] under a CC BY license, with permission from PLOS ONE, original copyright 2015.
Fig 3.
Preincubation of CFI with amyloid β (1–42) (Aβ) reduces CFI bioactivity.
CFI (10–30 000 ng/ml) was incubated with Aβ (200 μM) for 1 hour at room temperature. CFH (10 μg/ml) and C3b (80 μg/ml) were added to the reaction and incubated at 37°C for 30 min. Aliquots of the reaction mixture were analyzed for iC3b by ELISA. Pre-incubation with Aβ causes a statistically significant shift of the curve to the right and approximately, a 5-fold decrease in CFI bioactivity. Data are presented as means ± SEM in triplicate wells. A non-linear mixed effects model was used to estimate the ratio of EC50 for the dose response curves using a reparametrized 4 parameter logistic equation [25]. Number in parentheses indicate 95% confidence intervals for the estimated EC50 ratio or EC50 in ng/ml.
Fig 4.
Anti-amyloid β (Aβ) antibody, GSK933776, effectively blocks Aβ’s ability to inhibit CFI bioactivity.
Increasing doses of the antibody were preincubated with Aβ (30 μM) for 10 min at room temperature. CFI (1 μg/ml was added and incubated for 20 min at 37°C. CFH and C3b (80 μg/ml) were added and the mixture was incubated for an additional 30 min. Aliquots of the reaction mixture were analyzed for iC3b using ELISA. The curve in magenta shows the global fit of the two curves. The right dark red and dark blue bars show the effect of CFI + IgG1 (non-specific antibody control) on production of iC3b. The left 2 bars in lighter colors show the effect of CFI and IgG1 in presence of Aβ, i.e., reduction in CFI bioactivity. GSK933776 reduced Aβ’s ability to inhibit CFI in a dose-dependent manner (EC50 ~ 20 μg/ml). Data are presented as means ± SEM of triplicate determinations for two experiments conducted over 2 separate weeks. A non-linear mixed effects model was used to estimate the EC50 of the dose response curves using a reparametrized 4 parameter logistic equation [25]. Numbers in parentheses indicate 95% confidence intervals for the estimated EC50.
Fig 5.
Measurement of CFI bioactivity.
CFI bioactivity can be measured in plasma samples using an adapted assay that uses diluted plasma as the source of CFI. CFI bioactivity is temperature sensitive as heating the samples to 60° C abolishes the ability of the sample to generate iC3b (green symbol). Under the configuration of the assay, one can estimate the concentration of plasma that achieve half the amount of iC3b (i.e., 2.3 nl). We define this volume as 1 mU of CFI bioactivity. CFI bioactivity can be measured by determining the concentration of CFI protein in the same sample with an ELISA and correct the activity values by the amount of protein present in the sample, hence CFI bioactivity per unit of protein.
Fig 6.
CFI bioactivity in a cohort of subjects with staged AMD and non-AMD normals.
Subjects with AMD were staged according to the AREDS classification [24]. CFI bioactivity was determined by measuring the conversion of C3b to iC3b using ELISA for iC3b. Plasma dilutions (0.01–25 μl) were exposed to 80 μg/ml of each CFH and C3b and incubated at 37°C for 2 hours. Levels of iC3b were measured using ELISA. Samples from the same individual were run in the same assay as singlets. To control for the multiple assays conducted, each assay included a control sample. Data were fitted to a 4-parameter logistic equation and EC50s estimated. These values were converted to bioactivity in U/ μ g of CFI (1 mU = 1 / EC50 in pg/ml). The logarithm of the bioactivity for the experimental samples and control plasma samples were analyzed by a mixed-model ANCOVA, using the control sample as a covariate to correct for inter-run variability. Data are shown as the fitted means in the original scale ± SE estimated from the fitted values using the Taylor series expansion. N = 4 to 47 patients per group. Legends: AREDS, Age-Related Eye Disease Study stages I through III, GA, geographic atrophy, Wet-Inactive, wet active AMD.
Table 1.
Age distribution of subjects staged for AMD§.
Table 2.
CFI levels in a cohort of subjects with AMD and non-AMD normal measured by ELISA*.
Fig 7.
Plasma levels of free and total amyloid β (1–42) (Aβ) in Alzheimer’s disease (AD).
Plasma samples were collected from the phase I study of AD, receiving anti-Aβ antibody GSK933776 at 6 mg/kg (see Fig 2). Yellow transparent boxes indicate time intervals where CFI bioactivity measurements were performed. Following 3 doses of GSK933776, total Aβ levels do not appear to have reached steady state. Adapted from Andreasen et al., 2015, Fig 3A, p10 [23].
Fig 8.
CFI bioactivity in plasma samples of Alzheimer’s disease (AD).
Plasma was collected from the phase I study of AD receiving anti-amyloid β antibody GSK933776 at 6 mg/Kg. CFI bioactivity was determined by measuring the conversion of C3b to iC3b using ELISA for iC3b. Plasma dilutions (0.01–25 nl) were exposed to CFI and CFH (80 μg/ml each) and incubated for 2 hours at 37°C. Samples from same individuals were run in singlets within the same assay. Each assay included a control sample. Data were fitted using a non-linear mixed model that used a reparametrized[25] 4-parameter logistic equation and EC50s were estimated. The values were converted to bioactivity of CFI-Units/μg of CFI (1mU = 1/EC50 in pg). The logarithm of CFI bioactivity for study and control samples were analyzed by a non-linear mixed effects model using the control sample as a covariate to correct for inter-run variability. Data are presented as the geometric means ± SE estimated using the Taylor series expansion (n = 3–5 per group). Percentages shown in the graph were calculated from the integrated areas under the curve for each dosing period using the trapezoidal rule on the geometric means.