How do we create reference materials for assays of glycoproteomic targets?

A new article from CPTAC researchers raises some interesting future questions.

Interest has been increasing in protein post translational modifications for diagnostic and therapeutic purposes.

(e.g. https://www.sciencedirect.com/science/article/pii/S0304416516300630

"While the underlying causes of cancer are genetic modifications, changes in cellular states mediate cancer development. Tumor cells display markedly changed glycosylation states, of which the O-GalNAc glycans called the Tn and TF antigens are particularly common. How these antigens get over-expressed is not clear. The expression levels of glycosylation enzymes fail to explain it.")

The CPTAC paper further supports this

"Protein glycosylation is arguably the most diverse and sophisticated form of protein modification which drastically escalates protein heterogeneity to facilitate functional plasticity"

A key point of the heterogeneity is that it is not directly observable via the genome (the modifications are 'post translational' i.e. after the genetic information has been transcribed)

"The cellular machinery for O‐linked glycosylation, located primarily in the Golgi apparatus, is believed to operate stochastically in response to changes in a wide range of both intrinsic and extrinsic factors"

This raises an interesting question, if the observable process that is 'diagnostic' is the change in the Golgi apparatus processing then this can result in; different glycan forms, different attachment points, the level of sialylation of the attached forms etc. but may leave protein abundance relatively unchanged.

Post translational modifications can drastically change the function of a protein transcribed from exactly the same gene sequence.

If we consider one of the results from the paper

"Among these heavily O‐linked glycoproteins, VCAN contained the highest number of sites reaching 165 sites with distinct peptide sequences surrounding the sites, whereas MUC1 contained 161 sites, the second highest, but composited from only six distinct sequence repeats. "

If there are 165 'stochastically' populated sites then what does a targeted approach focus on?

In a standard sandwich approach a polyclonal antibody could isolate the protein of interest but what would the reporter monoclonal be targeted at? How do you target increased glycan heterogeneity?

A related and important question is how do we create reference materials for assay validation and standardisation?

The standard approach to assess an assay via an 'analyte standard' is well understood and can be implemented easily via dilutions and method transfer but how do we create a clinical range standard for a protein experiencing alteration in its PTM complement? Is there such a smooth and defined transition sequence?

A possible approach would be to collect patient samples across some defined range and create a standard from pools of such samples in a similar manner used for the ERM DA 470k serum protein reference but it is not clear if such a set of pools would provide a stable reference across analytic techniques.

Targeted techniques such as MS or monoclonal antibodies may give different results to 'whole protein' analyses that separate on ensemble physical properties such as protein isoelectric points.

Protein glycan sialylation is known to create 'chains' in 2D gel electrophoresis. A spot in a chain may contain many proteoforms resultant from heterogeneous glycan attachment sites and as such gives more of an overview of the 'distribution' of proteoforms and as such may be more suited to measurement of aberrant stochastic attachment.

The issue is also highlighted by Bateman et al in "Human Central Nervous System (CNS) ApoE Isoforms Are Increased by Age, Differentially Altered by Amyloidosis, and Relative Amounts Reversed in the CNS Compared with Plasma" 

"Furthermore, ELISA-based assays lack the ability to measure isoforms independently, detect truncations or fragments, and identify post-translational modifications. Finally, because the accuracy of an ELISA is entirely predicated by the standard with which it is calibrated, these assays lack the necessary concordance across platforms, which ultimately hinders comparisons between studies." 

Unfortunately, I don't have the answer .... we are working on it!