Is colocalisation of a protein with a presynaptic marker sufficient evidence to say that the protein is a component of axon terminals?

Question

I am reading journal papers about the subcellular localisation of the insulin receptor (IR) in neurons.

I have read a paper stating that IR is highly enriched at synapses, localising to both the presynaptic axon terminal and the postsynaptic density compartments.

The above statement saying that IR localises to presynaptic axon terminals comes from immunofluorescence images of cultured neurons where IR colocalises with synaptophysin puncta.

^ https://pubmed.ncbi.nlm.nih.gov/16978790/

In contrast, another paper showed via subcellular fractionation of the rat brain that IR is enriched in the post-synaptic density (PSD) fraction (through analysing the IR levels in the PSD fraction via Western blotting).

I think this is stronger evidence compared to immunocytochemistry.

To say that a protein is a component of the presynaptic terminal is colocalisation with a presynaptic marker (e.g. synaptophysin) sufficient evidence? Or should there be other forms of evidence to back this statement up?

Answer 1

Colocalization of a protein with another structure is necessary but not sufficient to say that the protein is a component of that structure.

Mixed into the concept of a structure's existence is the concept of an observational timescale. A presynaptic terminal is a structure that exists over specific timescales; all of its constitutive (and identity-conferring) components must colocalize with it, over all of its lifetime. Otherwise it would be more apt to specify something which colocalizes over more limited timescales to be cofunctional structures with the presynaptic terminal.

Experimental techniques tend to probe specific timescales. For example, immunostaining is based on the assumption that the dissociation rate constant for a specifically bound state is small compared to nonspecific bound states (allowing for the 'washing' steps to have sufficient efficacy). So, in general, spatial correlation of two molecular structures, based on a single observational approach, is insufficient evidence to qualify the structures as parts of a whole.

Answer 2

Empirical science is not based on proof, but on evidence.

No one study is the final word on anything. "Sufficient evidence" is entirely up to the person operating on the evidence and what they plan to do with it. Sufficient for further study? Certainly. Sufficient for using insulin injections to cure every mental/cognitive issue? Of course not.

I think I'd focus less here on the different measurement techniques and more on the model systems used, in addition to Ryan's points about timescales.