decreased gene expression during cell growth

[pichia123]

Hi,

recently I found the expression of my target gene was decreased during cell growth. when the cells were cultured for 24hr,48hr and 72hr without any stimulation, a time-dependent decrease of gene expression appeared. could anyone offer some suggestions why this happens?
Thanks in advance for your suggestions.

[Astarte]

There are many reasons why a gene may be downregulated, but without knowing more details on the function of the gene and the experiments you performed to measure the downregulation, it's just guessing. If you want a relevant answer to your question, you should provide more details.
One reason could be that your gene is involved in a pathway that is not active, for instance a virulence pathway only activated when in contact with the host cell or certain metabolic pathways for which the precursor is not present in the medium,...
It is also possible that after 72 hr your cells are losing their viability and everything except the apoptotic or necrotic pathways is downregulated.

[CrowSan]

Hi - as Astarte rightly says without more info it's hard to say.
I assume that the gene you looking at is a native one and not driven by a constitutative promoter (such as CMV).
I also assume it is not a house keeping gene (which, as it is normally essential for cell growth should not noticably decrease).

If the gene is a "stem cell" marker your cells may be differentiating and down-regulating this.

If the gene is a receptor for an external growth factor (and you are not stimulating the cells) the cells may be down-regulating the RNA as it is not needed.

If the cells are reaching confluence over this period perhaps confluency is affecting the gene (for example a gene involved in cell division may be downregulated if the cells stop dividing because they have reached confluency, become contact inhibited and entered G0).

Any number of reasons

[pichia123]

Thank you Astarte and CrowSan for your useful suggestions.
My target gene was a tumor marker and was overexpressed in cancer patients. I used lung cancer cells to detect gene expression by real time PCR. After 72hr cuture, i found the cell confluency was >90%. As you said, i thought it was very possible that my gene was involed in apoptotic pathway or cell division. But how can i prove that this hypothesis is true? Did you find some similar references? Thank you very much.

[Astarte]

Guessing the function based on the expression (or transcription?) during in vitro growth is really farfetched and will not give you any reliable conclusions.
As a starting point, if you have the sequence, would be to blast your target gene and see if it gives any significant hits (blastX would be best if the gene is not yet in any database) and/or a Prosite search to identify domains. Then you could use silencing or knock out experiments to see the effect.
Once you have an idea of the function, you should design an experiment to trigger this function and look at the expression after trigger. If you add a GFP marker, you can also localize the expressed protein, this can also be interesting depending on the suspected function (transcription factor should be located in nucleus, extracellular protein should be excreted,...).

[relaxin]

As you said, i thought it was very possible that my gene was involed in apoptotic pathway or cell division. But how can i prove that this hypothesis is true? Did you find some similar references? Thank you very much.

I do not think it is involved in apoptotic pathway, since it is overexpressed in cancer cells. To show it is involved in cell division, you can transfect the gene expression constructs (with either sense or antisense insert) into a cell line, select for stable expressing clones, and determine the proliferation rate of these cells.

[CrowSan]

I guess the gene could be involved in apoptosis if it either inhibited an apoptotic pathway or was a dominant-negative form of a pro-apoptotic protein. The fact that it appears to decline over time may link it to cell division I guess. It may also be that as these are (primary?) cancer cells plated in vitro it is just not in the right environment to make this protein (i.e. maybe needs factors released from surrounding cells such as fibroblasts). Care should be taken in assuming "cancer markers" are the same as "involved in cancer". Many markers are just abberant expression of things such as feotal markers or stem cell markers (cancer cells often display phenotypes reminiscent of stem cells or cells expressed in the embryo).
Probably the simplest was (as you know at least some of the sequence to perform qPCR) would be to clone it out and express it in a line that does not usually express this gene and see what effect it has (increase/decrease in prolifferation, resistance to apoptosis etc). Although not an ideal model system (as the cells you are putting the gene in do not normaly express the protein) it is a model that has been used extensively.

Allternatively you might also want to see if you can maintain the expression of the marker in your original cell type.
You could try:
a) spliting the cells pre-confluency and maintaining the cells in the exponential phase
b) adding various growth factors to see if this stimulates the cells
c) Adding a mild apoptotic inducer (if the protein is involved in preventing apoptosis you may be able to select for cells expressing it, or up-regulate it in the current cells)

Do you know if all your cells express this marker or is it a sub-set of cells?

[pichia123]

Thank you for all your suggestions which inspire me much and I will do some experiments to find more information.

CrowSan wrote: Do you know if all your cells express this marker or is it a sub-set of cells?

Is there any difference between sub-set of cell and original cell type?

[CrowSan]

Is there any difference between sub-set of cell and original cell type?

There is if only a small number (or <100%) of the original cells expressed this marker.The cancer may well be hetrogenous in it's cellular profile from one cell to another. For example cells on the inside of the tumour/cancer (in vivo) may have less nutrient/gas suppy so may be excreting factors to encourage new blood vessel growth, whilst cells on the outside of the tumour may be expressing markers/proteins regulated by their interaction with non-tumour cells. Thus (unless the cells you are working with are clonal) sub-sets can exist.

Also, removing cells from their 3D extra-cellular environment can easily change a whole host of gene expression profiles.
Imagine a gene that is up-regulated in response to a cytokine/growth factor released by surrounding stromal cells (such as fibroblasts which now seem to play a greater and greater role in cancer progression). Take the cancer cell away from this environment and this protein may just decrease with time. You may not even notice a decrease in the proliferation of the cancer cells in vitro as the up-regulated gene X may have a role in cancer invasiveness (a metalo-protease or a collagenase for example).

The suggestion has already been raised here (and it was a good one) that you might want to examine the sequence of the protein/DNA that you have for homologous ones that have already been described. This will give (probably) the greatest clue as to its function in the absence of experimental data.

Let us know how you get on and good luck
CrSn