Soluble CD146, an innovative and non-invasive biomarker of embryo selection for in vitro fertilization

Abstract

Although progress was made in in vitro fertilization (IVF) techniques, the majority of embryos transferred fail to implant. Morphology embryo scoring is the standard procedure for most of IVF centres for choosing the best embryo, but remains limited since even the embryos classified as "top quality" may not implant. As it has been shown that i) CD146 is involved in embryo implantation and ii) membrane form is shed to generate soluble CD146 (sCD146), we propose that sCD146 in embryo supernatants may constitute a new biomarker of embryo selection. Immunocytochemical staining showed expression of CD146 in early embryo stages and sCD146 was detected by ELISA and Western-blot in embryo supernatants from D2. We retrospectively studied 126 couples who underwent IVF attempt. The embryo culture medium from each transferred embryo (n = 222) was collected for measurement of sCD146 by ELISA. Significantly higher sCD146 concentrations were present in embryo supernatants that did not implant (n = 185) as compared to those that successfully implanted (n = 37) (1310 +/- 1152 pg.mL-1 vs. 845+/- 1173 pg.mL-1, p = 0.024). Sensitivity analysis performed on single embryo transfers (n = 71) confirmed this association (p = 0.0054). The computed ROC curve established that the optimal sCD146 concentration for embryo implantation is under 1164 pg.mL-1 (sensitivity: 76%, specificity: 48%, PPV: 25% and NPV: 92%). Over this sCD146 threshold, the implantation rate was significantly lower (9% with sCD146 levels >1164 pg.ml-1 vs. 22% with sCD146 levels ≤ 1164 pg.mL-1, p = 0.01). Among the embryos preselected by morphologic scoring, sCD146 determination could allow a better selection of the embryo(s), thus improving the success of elective single embryo transfer. This study establishes the proof of concept for the use of sCD146 as a biomarker for IVF by excluding the embryo with the highest sCD146 level. A multicentre prospective study will now be necessary to further establish its use in clinical practice.

Fig 1. Description of embryos recruitment.

Description of informative transfers.

Fig 2. Distribution of couples and embryos.

Distribution of sCD146 concentrations in embryos, according to couple.

Fig 3. Presence of CD146 in embryos and embryo supernatants.

3a: Immunocytochemical staining of CD146 in early embryo stages (day 2 and day 3), morula stage (day 4) and blastocyst stage (day 5). Embryos were stained with anti-CD146 (S-Endo 1) or mouse IgG1 isotype control and then incubated with a secondary anti-mouse antibody associated to a fluorescent probe, Alexa Fluor 488^® (shown in green). Nuclear DNA was labeled with DAPI (shown in blue). Pictures were achieved using laser confocal microscopy (63x oil objective magnification, NA:1.4). TE, trophectoderm; ICM, inner cell mass. 3b: Western blot analysis of embryos supernatants by ELISA. The negative control (Neg) corresponds to culture medium without an embryo. The positive control corresponds to a lysate of HUVEC (mbCD146). MW, molecular weight; Neg, negative control; EM1, embryo culture medium 1 (sCD146 = 110 pg); EM2, embryo culture medium 2 (sCD146 = 100 pg); EM3, embryo culture medium 3 (sCD146 = 100 pg); mbCD146: membrane CD146 (10 mg protein). Gels were run under the same experimental conditions. 3c: Soluble CD146 concentrations from day 2 to day 5.

Fig 4. Boxplot of sCD146 concentrations between implanted (YES, n = 37) and non-implanted embryos (NO, n = 185) (p = 0.024).