I would like to immunophenotype cancer cell lines based on the expression of cell surface markers. I have done three individual experiments with cells of different passage number, which is at least three generations apart. The composition of the sub-populations within the cell line varied greatly from one passage to another. Is this phenomenon commonly seen in cell immunophenotyping? Would anyone please suggest a way to address this issue. Thanks.
Hi Yuan, interesting question. It rather depends on what you are looking at and your cell culture protocol.
All cells in culture undergo a "selective pressure" and (over time) will gain or lose markers as cells that "grow faster" gradually dominate the culture. It is for this reason that most work on a cell line is performed at a certain passage number (i.e. on a "working bank").
There are certain factors that can cause the phenotypic drift to be small or large however (if you split the cells at a low density for example you are artificially generating a "founder effect" and if you split at high density you will maintain the diversity of the original population more (if that makes sense)).
Also if you grow the cells in poor quality FBS or media lacking certain nutrients/amino acids you will be selecting for cells that can grow in this media - this is an "adaptive process"
In fact what you are observing in your cancer cells is very similar to what occurs in vivo for cancers - they recently sequenced the genome of various cancer metastasis in one individual and found a great diversity between (and within) each metastasis (there are a couple of studies on this but here is one: http://www.bcr.org/content/cancers-heterogeneity).
How to address: a) use good quality FBS (such as Gibco, ATCC, Sigma etc) and avoid cheap FBS source b) maintain cells at a low split ratio (i.e. 1:3 rather than 1:5 or 1:10)
c) avoid any other stressors to the culture (such as prolonged time outside of incubator etc).
d) If it is an adherent cell culture, when you passage make sure that ALL of the cells have detached in the flask - if you leave cell behind you are in fact selecting for cells which are loosely adherent. e) Be consistent/precise when you passage - rather than use a 1:4 split, seed each new flask with the same number of cells (i.e. 7x10e5 cells per flask passage - see below)
Personally I would generate a working bank and examine cells at the same passage (rather than distant passages). To generate a working bank a) get the cells (i.e. from ATCC) b)Expand to ~ 10-20 million and freeze down at least half (these are your "master bank" i.e. 3-10 vials at 2-3 million per ml) continue expanding until you have sufficient cells for your experiments (i.e. 50 million and freeze down 25 vials at 2 million per vial - these are your "working bank"). Remove cells from working bank for experiments. If you run out get a vial of the "master bank" out and expand to the same passage level as your working bank (and use at the working bank passage).
If you want to be really accurate you can perform all experiments on cells of the same PD (population doubling). As "passage" is rather a vague concept and can depend upon
how many cells you seed per flask when passaging (i.e. if you have passage 4 cells and put 1000 cells in on T175 and 100 000 cells in another and grow these up to confluence they will both be passage 5 cells but one set has expanded a lot more than the other set).
PD is given by the equation: PD =Log (harvested/seeded)/Log 2 to do this you will have to know the number of cells seeded per flask as well as the harvested cell number.
Hope this helps,
Gary
Hi Gary, I appreciate your very informative answer. It makes sense to me and it really helps me proceeding to the next step. Thank you deeply.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Cell Biology
- Culture Cells
- Cell Line