• I diluted 16.66mcg of SYNTHETIC Oxytocin in 1000ml of Normal 0.9% Saline, which gives a concentration of 16 660pg/ml. Important to note here - no serum/plasma was used. It was pure oxytocin diluted in normal saline. 
• The results I obtained from the OXYTOCIN ELISA Kit gave me optical densities (OD) of between 2.2 - 2.7 - which equates to concentrations between 0 - 15pg/ml for that kit’s standard curve. This is very strange as the calculated concentration was more than 16 660pg/ml. I would expect a LOW optical density. 

If the synthetic oxytocin doesn’t contain any or very few residual solvents used in the process (for the reaction or purification) that could interfere with the antibody, I would not expect the binding to be significantly different.

If this is the case, the kit should still work with synthetic oxytocin.

Are you referring to whether anti-oxytocin mAbs will also recognize carbetocin? I would say it is likely, considering that the only modifications are the replacement of the disulfide bridge with a thiol (S-S -> S-CH3), removal of the amine group that facilitated the reduction of that bond, and conversion of the phenol side chain to a methoxy benzene. This does cause a minor shift in the conformational and relative spatial orientation so it really depends on the specificity of the mAb and whether the loss of an amine/sulfur and addition of two methyl groups impedes critical binding moieties or causes the spatial orientation of the antigenic determinant to become misaligned. My guess would be that the mAb recognizes and binds but with lower affinity.

I'm trying to fix this issue:

The currently available anti-oxytocin monoclonal antibodies have been, to my knowledge, unimpressive. Oxytocin is a cyclical 9-residue peptide so one could imagine that a single antibody would use the majority of the molecule as the antigenic determinant. In a sandwich assay, at least when I imagine it conceptually, would have a very difficult time due to two extremely large immunoglobulins and their glycans converging on a much smaller target. I would imagine the steric hindrance is a significant impediment to the accuracy of these types of immunoassays, which is what has been reported, at least in the literature that I've read thus far. I'm working on a project to develop a methodology for screening structure-switching aptamers that bind oxytocin with low-nM Kd and high specificity, in order to obviate the requisite dual antibody-based affinity interactions and replace it with a short ssDNA oligonucleotide that undergoes a significant shift between native folded state and the oxytocin-bound state. Engineering the construct design so that this property can be screened for, in addition to the canonical affinity/specificity screening, allows for the development of a [No Oxytocin->Quenched] [Aptamer->Oxytocin->Complex->Signal Elicited] system.

It very well could be possible then that the antibody would have a lower affinity if the interaction is dependent on the hydrogen bond donors afforded by the amine and the phenol which are absent in the synthetic form (or if the methyl groups could create additional hydrophobic interactions).

Is just one of the cysteines in this synthetic form methylated? There could be some interdisulfide binding (connecting the heavy/light chains of the Ab) by a free thiolate in the reduced oxytocin if only one of the cysteines is methylated. I'm not sure if then the antibody structure would be compromised (dissociation of one of the chains).