What are the limitations of these methods?

Until now, anti-cancer therapies that inhibit MET signalling have given impressive results in patients selected for MET exon 14-skipping mutations. However, in fact, around 30-50% of patients still don’t respond to treatment and the other selection criteria of MET gene amplification and MET protein overexpression are also inadequate for predicting response to MET therapies that block kinase activity. Clearly, there is a need for better biomarker analysis.

What we (and others) have wanted to look at was actually MET kinase activity – the result of which is tyrosine phosphorylation. The initial part of MET signalling is autophosphorylation at MET tyrosines 1234 and 1235, which then in turn leads to phosphorylation of tyrosines 1349 and 1356 in the MET C-terminus, which binds downstream signalling molecules. Additionally, there is the question of whether cellular localization of MET (in the nucleus or at the plasma membrane) is also important. Until now, there have been many challenges with doing these studies.

Antibody specificity

Antibody specificity is a key component of these studies as many early commercial antibodies to MET cross-reacted with other signalling proteins such as Src. Previous studies have used anti-pY1349MET monoclonal antibodies, as well as anti-pY1234/1235MET monoclonal antibodies. However, no link between phospho-MET levels and patient survival was seen. We have also previously developed a chemiluminescence-based sandwich immunoassay using an antibody that reliably detected full-length pY1234/1235MET in lysate samples. However, this assay did not detect the biologically active pY1234/1235 intracellular C-terminal fragments of MET that are caused by proteolytic cleavage during signal transduction.

Sample collection

Phospho-MET is incredibly unstable in unprocessed tumor biopsies at room temp and sample collection methods need to involve flash-freezing to preserve this form. The lack of adequate sample collection methods has been a problem for previous attempts to assess phospho-MET levels.

Sample type

Assessment of phospho-MET using lysate-based samples used for immunoprecipitation (IP) results in a lack of information about the subcellular localization of these molecules. This is important to ascertain because previous studies have observed the nuclear localization of cleaved MET C-terminal domain, localization of full-length MET receptor to the nuclear envelope, and the plasma membrane. Currently, the evidence about the importance of cellular localization of MET receptor seems to vary according
to the different histologies being studied, so this needs to be assessed for different tumour types.

How can detection of enzymatically active MET be improved?

In our paper, “Multiplex immunofluorescence microscopy assays for pharmacodynamic assessment of MET tyrosine kinase activation in the plasma membrane and nucleus”, we show how we have built on previous studies to improve detection of enzymatically active MET and tackle the limitations described above.

Antibody specificity

As phosphorylation of MET Y1234 is not required for MET activation in all instances, we developed an immunofluorescence microscopy-based assay using our monoclonal antibody that specifically recognizes pY1235MET regardless of MET Y1234 phosphorylation status. This assay allows quantitation and localization of full-length MET as well as the phosphorylated C-terminal cleavage products that contains the activated kinase domain that result during signal transduction.

Sample collection

In a previous study we have described protocols for sample collection and processing that maximize stability and maintenance of the MET tyrosine phosphorylation (Srivastava et al., 2016). This involves flash freezing samples to keep the cold ischemic time to less than 5 minutes followed by normal tissue processing in formalin.

Sample type

The assay described in our paper shows how detailed information about the subcellular localization of activated MET can be ascertained using a small number of tissue sections from a single tumour biopsy core.

What are your thoughts on future perspectives of this field?

Our clinically validated, quantitative pY1235MET immunofluorescence assay shows that it is possible to achieve meaningful measurements of tumour MET activation and its subcellular localization. We found that tumour total MET levels do not necessarily represent activated MET levels (indicated by tyrosine 1235 phosphorylation), suggesting that pY1235MET may be an independent biomarker predictive of response to therapies that target MET activation.

Additionally, analysis of plasma-membrane-associated MET will also be useful to determine the accessibility of MET for antibody-based therapies such as antibody drug conjugates that are currently in clinical development.

Now that reliable quantitation of pY1235MET is possible, it will be of interest in future to determine the whole gamut of information about a patient’s MET status all at the one time – MET gene amplification, MET protein overexpression, MET gene mutations as well as now MET activation and subcellular localization – to try to understand exactly how MET is contributing to a patient’s tumour and how best to treat it.