In this study, we demonstrated that RPPA combining near-infrared (NIR) dyes can be applied to human cumulus cells to measure potential biomarkers of oocyte competence. ERK2 proteins were detectable in individual human cumulus. Moreover, after validation by the siRNA approach in HEK293 cells of the antibodies targeting VCL and SRC, we were able to detect these two proteins of interest with the same sensitivity.
Defining embryo quality with non-invasive methods continues to be the major goal in ART while morphological criteria of early embryo development to predict further development or implantation remain poorly predictive . Various approaches involving transcriptomics, proteomics and metabolomics on the embryo or its cellular or non-cellular environment have been proposed (see  for review). The cumulus cells surrounding the oocyte may be targeted to develop a non-invasive method to investigate the oocyte developmental competence and several studies have already identified biomarkers by transcriptomic analysis . While the proteins remain the final effectors in the cells, no link between the product of these mRNAs (i.e. proteins) and the oocyte competence has been established to date. This is due, in large part, to a lack of technology sensitive enough to analyse specific proteins from the minute amount of cumulus cells available.
Protein microarray-based methods, particularly RPPA, are very sensitive and allow the detection of multiple specific targeted proteins in small quantities of biological material compared to Western blotting. Indeed, WB needs protein from 5 × 105 cells whereas RPPA requires nanoliters of protein lysate (picograms to femtograms of protein) from the equivalent of 200 cells . Compared to mass spectrometry, which may need further identification of the highlighted proteins, RPPA targets only specific proteins. Moreover, mass spectrometry requires heavy and expensive equipment compared to RPPA. RPPA has been successfully applied to biomarker profiling in cancer diseases (see  for review). In this context, we thought that RPPA could be used to evaluate the level of protein expression of specific proteins as biomarkers of oocyte developmental competence in cumulus cells. In order to achieve a proof of concept, we used two potential biomarkers RGS2 [19, 26] and SRC (data not shown) and a housekeeping protein VCL, which was used as a reference protein.
As a first step, we used siRNA targeting the proteins of interest in HEK293 cells as a mean to validate the antibodies for their use in RPPA . The knockdown efficiencies estimated by RPPA were close to those observed by Western blotting for SRC, VCL and GAPDH, thus allowed us to validate these antibodies for RPPA. In contrast, the knockdown efficiencies measured by WB and by RPPA for RGS2 did not allow us to validate this antibody for RPPA, even though no dominant non-specific band was observed for this antibody by WB. This might be explained by the experimental changes between the techniques. Indeed, the interactions between the epitope of the proteins and the paratope of the antibody might be modulated by the physico-chemical conditions such as the temperature, the pH of the buffers or the concentration of the detergent. Furthermore, the protein conformation might modify these interactions. Finally, the antibody targeting RGS2 was designed against a fragment of recombinant RGS2, which might lead to non-specific binding.
RPPA sensitivity was assessed on pooled then individual human CCs using the validated antibodies. The study by Dupuy et al. showed that RPPA combined with amplified NIR dyes allow the detection of phosphorylated proteins in less than 1 ng of total cell extract. Moreover, using NIR dyes allowed them to detect ERK2 protein in the equivalent of two HEK293 cells printed per spot. This work showed the extreme sensitivity of the techniques that could be applied to the research on a small amount of biological sample such as cumulus cells. In line with this work, we showed that ERK2 protein can be detected for a dilution corresponding to less than one cell of an individual human cumulus printed per spot. Furthermore, VCL and SRC proteins were detected with the same range of sensitivity.
Even though antibody validation remains a challenging problem, RPPA seems to be a useful technology for biomarker detection for clinical diagnostics in individual cumulus cells (sensitive, high throughput, and cost- and material-effective) compared to WB, 2D electrophoresis and mass spectrometry, which have already been tested on human cumulus cells. The siRNA approach has shown its efficiency for antibody validation [27, 28]. However, this approach is limited by the protein characteristics (i.e. turn-over of synthesis, half-life and cellular function) and requires specific development for each target. In this context, siRNA is not optimal for rapid and high throughput validation of a large collection of antibodies. Alternatively, overexpression of the targeted protein by plasmid transfection might be a better alternative for future antibody validations.