Hi Yong, by its "expression" I take it you mean at the RNA level by PCR/qPCR?

The correlation between RNA levels and protein levels can often be quite poor (for example a protein with a long half life can be present in large amounts in a cell but the RNA in low amounts - as it is not required to be translated very often).

Also when you performed IHC did you titre your antibody and try several antigen retrieval techniques (microwave, alkaline, trypsin and citrate)? Also did you include a negative control (non-specific antibody + secondary or secondary alone).

It could be that the IHC was sub-optimal and/or what you are seeing is not a "genuine" protein signal.

To confirm your IHC you could perhaps lyse the cells and IP the protein out and run a western blot. This should get around any epitope/IHC related problems (in a denaturing gel).

Hope this helps,

Gary

To add to the previous correct reply, human tumors are all different. One can always find a deviating one. No correlation is 100%, it is always quantitative. I am not surprised that a tumor would deviate from your “rule”. Finally, IHC is not a quantitative method. I would side again with Gary to run a Western, which may be made semi-quantitative with LiCor detection method.

Hi Gary and Alexander, Thank you for your help. In fact, my IHC results showed there is a obvious difference between adjacent normal tissues and cancer tissues. However there is no difference between in situ and metastatic tumors. So how to explain the gene's protein expression(IHC score) correlates with patients' metastasis and supression of this gene inhibit the migration. western blot is good suggestion, I will try. But is there any explaination for the no difference between in situ and metastatic tumors if the "no difference" is true? 

Hi I am getting a little confused

 In your original question you state: "But when I check the expression of this gene in in situ and metastatic tumor by IHC, it seems no difference. " and above you say "In fact, my IHC results showed there is a obvious difference between adjacent normal tissues and cancer tissues. However there is no difference between in situ and metastatic tumors"

In your original question I assumed you were referring to the RNA level (qPCR) compared to the protein level (by IHC) n the cancer itself.

What tissue are you performing the IHC on if not tissue from the tumour site of a patient (in situ)? Are you perhaps referring to the IHC results between the primary cancer and a secondary (metastasis) site?

If so the primary cancer can be composed of cells carrying many different mutations (I think they analysed singe cells from a cancer recently and fund up to 13 different cancer cells in one tumour).  So the primary cancer can be very heterogeneous - which can make it hard to treat as sub-populations can emerge.

Secondaries form when cancer cells, or small groups of cells, break off from the main cancer and migrate through the blood to another (secondary) site. One of the main ideas is that to do this they have to assume a "Mesenchymal" phenotype (to get through blood vessels and endothelium etc). Thus undergoing EMT if the original cells were epithelial. This is generally associated with changes in various proteins (cytokeratins for example). These cells can then revert to epithelial-like cells after engraftment at the secondary site.

Due to the heterogeneity of the primary cancer and changes that can occur during metastasis it is no surprise that he two cancer sites can express different, or different levels, of various proteins. 

A second factor (even assuming the cells are genetically the same at both tumour sites) is the nature of the surrounding tissue.

For example it is now well established that normal tissue can have a great influence on cancer cell growth - surrounding mesenchyme (such as fibroblasts) can "feed" growth factors and promote vascular  growth in the cancer (i.e. CAFs).

If your genes higher expression is not due to a mutation (promoter, negative regulatory elements) its higher expression my be due to the environment of the primary site - this could then fall down to "normal" levels in the new integration site.

If you could clarify your question  a little more it would help!

Gary

Hi Gary, sorry for the unclear expression and thanks again for your kindly and useful reply.

In my oringinal question  I want to express: T(primary tumor)>N(adjcent normal) both in mRNA and IHC, ie. the expression is consistent; also, M(metastatic tumor)>N. clinical data and in vitro experiment showed the expression in primary tumor correlates with patients' metastasis and migration, respectively. Then I think this gene may be functional in metastasis, so its expression maybe more higher in M than T. However, they are similar. That's my question.

Your explaination about heterogeneity and "MET" are good to explain my question.

Metastasis tumor maybe a subgroup cancer cell which have a strong "Mesenchymal" phenotype and can be a dominative group in metastatic tumor. If my gene(A) is functional in migration, the A+ cancer cell maybe enriched in metastatic cancer. And that's why I expected A may be higher in M than T. When a cancer cell completes its migration and reside at the secondary site, the "MET" may be the reason why A expression in M is identical to primary tumor. Also, a new microenviroment do inflence A expression.

Clear!

Really appreciate.

Hi Young

EMT may be a good explanation of why a marker regarded as important for the process of metastasis may not be expressed in the metastasic site.

Also remember the "seed" and "soil" concept... maybe that protein is important to disperse the "seed" but not necesary for survival in the new "soil". Depending on the type of tumor, the primary site offers a microenvironment quite different from a metastasic site, so the things needed to survive/disperse in/from one place may be downregulated to colonize another. This changes could be part of a rapid changes in cell metabolism, independent of genetic changes, or may indeed be subject to selection (influenced by the founder efect)  as tumors/metastasis are diverse and ever evolving.

Thanks! Martin.

It could also be that your protein is undergoing post-translational modifications in the metastatic tumours, such as phosphorylation or glycosylation, and your antibody doesn't detect it anymore even though it's there. Have you tried different antibodies binding to distinct epitopes? That might be your answer.