Three two-site apoA-I immunoassays using phage expressed detector antibodies - Preliminary clinical evaluation with cardiac patients.

High density lipoproteins (HDL) are a heterogenous group of subpopulations differing in protein/lipid composition and in their anti-atherogenic function. There is a lack of specific and robust assays which can target the functionality of HDL with respect to atherosclerosis. With recently generated CAD HDL targeted, single chain recombinant antibodies (scFvs) we set out to design and optimize apo A-I tests to compare it with conventional HDL-C and apo A-I analyses for diagnosis and risk assessment of coronary artery disease (CAD) and its outcome. Three highly sensitive two-site apo A-I assays: 022-454, 109-121 and 110-525 were optimized. A preliminary clinical evaluation of these assays, after proper sample dilution procedure, was performed using samples derived from 195 chest pain patients (myocardial infarction (MI), n = 86 and non-MI, n = 109), collected at the time of admission and at discharge from hospital (hospital stay ≤ 24 h). The clinical performance of the assays was compared with apo A-I measured with polyclonal anti-apo A-I antibody using conventional ELISA. Apo A-I data was in addition compared with HDL-C concentration of the samples. The concentration of apo A-I was significantly lower in MI patients than in non-MI individuals with assay 022-454 (admission and discharge samples, P < 0.0001 and = 0.004); assay 109-121 (admission and discharge samples, P = 0.04 and 0.0009), and, ELISA based apo A-I test (admission and discharge samples, P = 0.008 and < 0.0001). HDL-C (admission and discharge samples, P = 0.002 and P = 0.01) was also significantly lower in MI patients. In Kaplan- Meier analysis, two-site assay 109-121 assay predicted mortality from admission samples at 1.5 yrs (whole cohort, P = 0.01 and in MI patients, P = 0.05) and at 6 months (whole cohort, P = 0.04). Assay 110-525 predicted mortality at 1.5 yrs from admission samples of non-MI patients (P = 0.01) and at 6 months from whole discharge sample cohort (P = 0.04). Polyclonal anti-apo A-I based conventional assay predicted mortality at 1.5 yrs from admission samples of whole cohort (P = 0.03). Two-site apo A-I assay 022-454 and HDL-C provided no capability of predicting mortality in the whole cohort or any sub-group. In conclusion, two of the tested recombinant apo A-I antibody combinations (sc 109-121 and sc 110-525) display promising outcome to improve diagnosis and prediction of future cardiac events in cardiac patients over polyclonal apo A-I ELISA and HDL-C assays. The noted differences, while interesting, are preliminary and need however to be verified in extensive cohorts of pathological cardiac conditions and healthy controls.

Identification and analysis of anti-HDL scFv-antibodies obtained from phage display based synthetic antibody library.

OBJECTIVE: In epidemiological studies plasma high density lipoprotein cholesterol (HDL-C) levels are found to correlate inversely with atherosclerotic cardiovascular events. HDL consists of different subpopulations and they vary in their anti-atherogenic properties. The aim of this study is to isolate coronary artery disease (CAD) specific anti-HDL scFv-antibodies. DESIGN AND METHODS: To obtain CAD specific HDL binders, we used phage displayed synthetic antibody libraries to enrich specific antibodies against HDL isolated from CAD patients. The antibodies were affinity purified. Their capability to recognize apolipoproteins A-I and A-II, various HDL forms differing in lipid/protein ratios and plasma HDL, was studied using time-resolved fluorescence based immunoassay. RESULTS: Using different selection strategies and immunoassay based screening we obtained altogether 1200 clones displaying HDL binding activity. By sequencing 337, we identified 264 unique antibodies against HDL. A set of 61 antibodies were selected for further analysis. We found a variety of antibodies with different binding profiles, including apoA-I binding antibodies either in lipid-dependent or lipid-independent manner and binders against apoA-II. Several antibodies were able to discriminate between HDL derived from CAD patients and healthy controls. A majority of the antibodies were immunoreactive with HDL in plasma. CONCLUSION: The novel HDL recognizing antibodies isolated from synthetic antibody phage library have displayed interesting HDL-binding characteristics suggesting that, in addition to use as research tools, a part of them might be useful for the development of diagnostic methods for CAD risk assessment.