I want to engineer a genetic circuit to a chromosome of a mammalian cell line and want the function of the circuit to be stably expressed across extended period of cell culturing (>100 generations).
Provided that, in knock-in mice, a mutation introduced at the fertilized egg can still demonstrate the desired phenotype, say GFP expression, in adulthood, the stable maintenance of an engineered phenotype should be feasible in propagating cell lines.
Is my reasoning reasonable? Thanks!
Hi Xiao, you are facing a number of problems.
1. Cells in culture (in vitro) are under a constant "selective pressure". This means that any faster growing cells will come to dominate the culture over time.
Cells expressing your circuit or protein (GFP for example) will be using more "energy" in the terms of amino acids/RNA to make this extra protein.
Thus if (after several passages) any cell has been able to switch off the gene (either by deletion or silencing its promoter) you will get a gradual loss of the % of cells expressing your protein from your culture over time.
2. Chromosomal instability/polyploidy.
Many transformed cell lines (such as cancer cells) , whilst being able to grow in vitro indefinitely often have "scrambled" genetics.
For example whilst the majority of the cells in the culture may have a chromosome number of 46 (i.e. 87%) there will also be some cells containing either more (i.e. 5% with 47 - 50) or less chromosomes (i.e. 7% N=42).
As well as this trisomy's (i.e. 3 copies of chromosome 21) etc. of certain chromosomes can occur. In normal cells these problems (often due to problems that occur during mitosis) would prevent further cell division and propagation of aberrant cells.
The result of this is that if you select a cell line with these problems the cell population containing your circuit will become heterogeneous over time - some cells may have multiple copies of the circuit and many may have none (depending upon where the circuit has integrated)..
Obviously "normal" cells (often called primary cells) tend not to have these problems (so if your circuit integrates on chromosome 12 it is likely to stay there and be present in daughter cells with the sae copy number and location). However primary cells tend not to be immortal and lose the ends of their chromosomes (telomeres) over time. Therefore after a certain number of cell divisions they will cease to divide and grow (the Hayflick limit).
3. Homogenous expression and location.
Unless you can live with a population of cells that have different numbers of your circuit integrated indifferent chromosomal areas then you will need not only to generate cells stably express the circuit but you will also have to generate "clone".
In other words you will have to introduce your circuit into several million cells in a vector that contains a selectable marker (such as resistance to a toxin or anti-biotic).
You will then have to grow the cells in the presence of this drug/antibiotic to select for the cell where the gene has "stably" integrated into the genome.
These stable cells will have the circuit integrated randomly in different places from one cell to another (some may even have multiple copies).
You will then have to isolate separate single cells and grow them up so that each population has identical integration sites and circuit expression (i.e. a clonal population). To do this the cells must not only be capable of growing without the presence of other cells (which many cells do not like to do i.e. they are density dependent) but they must also be cable of maintaining the expression of the circuit until sufficient cell numbers are obtained (and these clonal cells will have to sustain expression over many, many population doublings (whist under the selective pressure of in vitro growth).
Therefore the selection of the target cell line has to include many desirable characteristics (stable ploidy. density independent growth, maintaining circuit expression of a discrete number of circuits many cell doublings etc).
This I quite a tall order.
As well as the cell you will also have to consider what promoter your circuit/gene will be under. Some promoters are constitutive (which adds to the pressure of sub-populations arising), some are weak, some tissue specific, some inducible etc.
Some promoters are more prone to switching off ("silencing") in some cell lines than in others (ie. CMV promoter in H9C2 cell lines).
As this is getting rather long I will post it and try to give you some solutions in a next message (also got to work!)
Gary
Thanks a lot, Gary! That really gives me the precise sense of what it takes to do this experiment. Admittedly, I've only worked with bacterial and yeast. The papers published in synthetic biology in mammalian cell lines give me the general impression that it is not that hard to work with mammalian cell lines. Obviously, I am wrong.
Hi Xiao it can be tricky! Each cell line is almost like an individual person requiring different media, plating density, feeding regimes and all have different morphologies. They will also change over time in culture so generally people freeze down large numbers at early passages - use a certain passage for experiments - then go back to the early stock and re-grow to the same passage for other experiments (to ensure the experiments are performed on cells with the same morphology/characteristics.
Adherent cells can even display different responses depending on their density in the 2D culture at the time of experimentation (ie immune-modulation ability of MSC changes with the density of the cells)
The main problem (which is a constant bug-bear when trying to perform very accurate measurements) is the "scrambled" genomes the different cells have.
I like Jurkat E6 cells myself - they have a pseudo-diploid chromosome set
http://www.atcc.org/products/all/TIB-152.aspx#characteristics
i.e. 46 chromosomes in most cells. However these cells are non-adherent.
Could you use non-adherent cells for your circuit? If so these might be good.
They are easier to grow (no need for trypsinisation) and can grow to a high number per flask (1x10e6 per ml).
Not sure how easy they are to transfect or how density dependent they are.
- Badges
- Science method