Circulating Levels of sFlt1 Splice Variants as Predictive Markers for the Development of Preeclampsia

Abstract

Angiogenic biomarkers, including soluble fms-like tyrosine kinase 1 (sFlt1), are thought to be predictors of preeclampsia onset; however, improvement is needed before a widespread diagnostic test can be utilized. Here we describe the development and use of diagnostic monoclonal antibodies specific to the two main splice variants of sFlt1, sFlt1-1 and sFlt1-14. These antibodies were selected for their sensitivity and specificity to their respective sFlt1 isoform in a capture ELISA format. Data from this pilot study suggest that sFlt1-1 may be more predictive of preeclampsia than total sFlt1. It may be possible to improve current diagnostic platforms if more specific antibodies are utilized.

Keywords: diagnostic; isoforms; monoclonal antibody (mAb); preeclampsia; soluble fms-like tyrosine kinase 1 (sFlt1); splice variants.

Figure 1

C-terminal peptide sequences used for generation of mouse mAbs specific to sFlt1-1 (solid line) and sFlt1-14 (dashed line). (A) A keyhole limpet hemocyanin (KLH) fusion protein preceded each sFlt1 peptide antigen used for immunizations while a Thioredoxin (TRX) or glutathione S-transferase (GST) tag preceded the sFlt1-14 and sFlt1-1 peptides, respectively, used for screening ELISAs. Full-length sFlt1-1 containing a C-terminal his tag was used to generate all other mouse mAbs and HuMAbs that recognize total sFlt1; (B) Schematic of sFlt1-14 and sFlt1-1 proteins identifying the unique epitopes antibodies were directed against.

Figure 2

Hybridoma cultures were screened by ELISA against (A) GST-1C or (B) TRX-Exon14 to isolate sFlt1-1 or sFlt1-14 specific mAbs, respectively. 1CKLH18 and Ex14-1 were selected for their favorable binding properties as isoform-specific mAbs, while the total sFlt1-specific mAb, 10ugR#9, which binds a shared domain of the full-length protein, was included as a negative control.

Figure 3

Mouse mAbs were specific for sFlt1 splice variants as measured by capture ELISA. Total sFlt1-specific mAb, 10ugR#9, recognized both recombinant sFlt1-1 (A) and sFlt1-14 (B) splice variants with similar sensitivity; however, mAb 1CKLH18 only detected sFlt1-1 while mAb Ex14-1 only detected sFlt1-14. In addition, each mAb specifically recognized endogenous sFlt1 present in amniotic fluid (C), while biological fluid that does not contain sFlt1 (normal human serum) was not detected (D). To assess interference of quantitation in biological fluids (E), 25 ng/mL of recombinant sFlt1-14 was spiked into normal human sera or amniotic fluid.

Figure 4

Mouse mAbs Ex14-1 (A) and 1CKLH18 (B) specifically detected endogenous sFlt1 isoforms from amniotic fluid on a Western blot. Each mAb recognized a single protein of the expected molecular weight (~115 kDa) when compared to recombinant standards. Included as a positive control (C) is a commercially available mAb (Sigma Cat.#V4262) that recognized total sFlt1.

Figure 5

sFlt1 isoform and VEGFR-1 quantitation from serum samples at three gestational windows (GW) during pregnancy. (A) sFlt1-1, (B) sFlt1-14 and (C) VEGFR-1 levels from all women included in the study and (D–F, respectively) a subset from women included in A–C diagnosed with chronic hypertension and/or diabetes mellitus (chtn_dm) are reported as the mean biomarker level ± SEM. * p ≤ 0.05; ** p ≤ 0.01.

Figure 6

Receiver operator curves generated from the sensitivity and specificity of sFlt1-1 and VEGFR-1 preeclampsia predictions at gestational windows 1 and 2 in (A) all samples measured and (B) a high-risk subset of these women with chronic hypertension and/or diabetes mellitus.