I am working with Bovine Aortic endothelial cells and in those cells we want to check the effect of growth factors on a particular gene at the mRNA levels. There are human primary cells available but I did not find any human endothelial cell line. I wanted to know what are the advantages of using a primary cell over cell line ( I know that cell lines are transformed and are a deviation from normal cells). Also, in terms of the experiment when should we use cell line and when should we check in primary cells?
This is a very pertinent question that I am struggling with after 20 years in research. Very simply put, cell lines are easy to work with, you can have them forever, and the scientific community tends to accept your findings more readily because cell lines have been considered the standard for similar research in the past. Do the cell lines represent what is occurring in vivo? - no, not at all. For long-term studies of basic biology, cell lines are convenient as evidenced by labs using the same cell lines over decades. Primary cell lines have the following limitations - they are not well characterized, have limited life span, slow in proliferation and finally nothing to compare with. And frustratingly, journal reviewers will have more annoying questions about the cells and more experiments to validate, not because of your work, but due to their ignorance...........Primary cells derived from different patients can behave differently in culture conditions depending on the genetics and age of individuals from whom the tissue was derived. Also, fibroblasts derived from different parts of the body may have different characteristics...the beauty of biology!!!!!! Moreover, with every passage, the characteristics of primary cells change. I would therefore say that for consistency of results from experiment to experiment and from year to year cell lines would be preferable as long as they are handled properly. After all, as some suggested before, every time you subculture primary cells, over its life time, you are selecting for a sub population of cells continuously until they die out. If you really think about it you are selecting for cells that are capable of proliferating on plastic substrate, a surface the cells are not used to inside the body. These same cells fail to proliferate on softer substrates that approximate human tissue stiffness. More importantly, the population of cells you are dealing with at passage 20 is not the same as the ones at passage 5. Obviously, this can skew your results. Therefore in the case of primary cells, it is imperative that you work with the same cell population for the same experiment. So you may want to expand enough cells at one time and freeze them so that you can go back to the same passage for all your experiments to ensure consistent results (Shhhhh another secret - don’t share these cells with anybody and don’t trust cells others give you….basically you have to know what you are working with). The same goes true for cell lines, but less so. If you allow cells to grow to very high confluency, over a period of time, you are going to select for a sub-population of cells adapted to out maneuver slow growing cells. One more important note -make cell line identification a standard procedure in your lab. Clinical trials have occurred based on results obtained using the wrong cell lines; and journals are starting to ask researchers to validate their cell lines before publishing their data. There have been a number of publications recently based on these issues. Hope this answers some of your concerns.
If you do decide to perform your experiments with a cell line, firstly ensure that the cell line is what you think it is (ideally obtain it from a cell line respository, rather than accept it as a gift from another lab), and karyotype it. There are a number of papers warning about the pitfalls of cell lines, and I can attach some here, if you wish.
Whatever results you obtain, you should probably try to replicate them in primary cells, as they are a superior model of the in vivo situation. This should help to ensure that your results are not just an quirk of the particular cell line that you are using.
Does anybody have any particular rules for deciding when to use cell lines and when to use primary cells? For many cell types, there are no protocols available for primary cultures, or there are no cell lines established, so there is no choice available. But, when both primary and cell lines are available, is it mainly a choice based on convenience?
Good luck,
Stuart
Stuart, could you attach the references for those papers about the pitfalls of cell lines? That would be helpful - trying to avoid all the dark chasms around.
Hi Shambhavi, I agree with Stuart and second evrything he has said. Cell lines are a good place to start as access to primary cells is more difficult (gerenally). Also, since people have worked with characterised cell lines, you can verify your data and people find it easier to put a perspective to your data since they have experience with those cells. Then again, as stuart says, it is a good idea to test out your results in primary cells because they present an environment closer to in vivo setting than ones presented by cell lines.
Something to be kept in mind is that you should, ideally, not carry primary cells beyond a certain passage. Cells tend to change with every passage, depending on cell type, conditions, etc. In my experience, I would not take them beyond passage 3 or 4, for example, but this would depend on your cell type.
All the best
Saurabh
Each time the cells are passaged they become more homogenous and less similar to the cells in situ. I would always choose primary cells over a cell line if they were available.
It really depends.
I work with both human primary cells and their supposingly equivalent immortalized cell line.
Broadly speaking, the primary cells and the immortalized cells expressed the expected 'generic' cell markers by immuno-fluorescence. However when we went deeper in depth (microarray), we found their gene signatures to be quite different.
Isolated primary cells from the tissue sample you get your hands on will be the closet thing you can work with, but be weary of potential sample-to-sample variation.
Next, if your primary cells are difficult to grow, difficult to come by, and/or have very limited proliferative capacity, the next viable alternative is to use immortalized cell line 'equivalent' to carry on with your research. But once the optimization work or characterization work is done on the cell line, you will have to validate the results in the primary cells. To be honest, if I have unlimited supply of primary cells, I'd work with primary cells all the way...
Good Luck with your research!!
Gary
Stuart is right, I can add my experience: I am working with epithelial primary cells and passaging does change them a lot: less growth, shape changes, less mitosis and cell markers decreased. The appropriate Medium and supplements can help keeping them stable throughout the culture period. And if you are using trypsin, consider other less-aggressive agents for cell dissociation. Damaged cells will suffer more the subculturing.
All comment are correct
Usually first you will reach to biology data by primary cells and also cell lines. But when you decied to show molecular event base on your biology data, you need to use cell lines because your primary cell may be will not be enough but cell line, you can expand..
It depends on what you want to do! Primary cells are difficult to culture and they die and also some tend to differentiate within first few passages. I used primary cells from aorta and they used to differentiate to fibroblast like phenotype.
Though difficult, experiments using primary cell lines yield data which are definitely far more acceptable than using cell lines.
Primary cell culture is better than cell line for experiments. but as other told, culture of primary cells is difficult but if you use medium calcium free such as RPMI1640 is best. If you want to use cell line, you must sure that cell line has no infection with other cell line and use cell line in primary passages.
Yes it is true the primary cultures are difficult and this makes researchers use cell lines. You can make a a few passages of primary culture. I am using most of my research in primary cells. But you have to examine exactly how long it to retain their properties in vivo. That depends on the length of life in vivo. Your cells do not have a such a short period of life such as the lute al cells (14 days).
Find itself some secreted factor or expression of which is characteristic of these cells, passaged 3-4 timesand determine its expression in each of its passages. You can then use the cells from passage 1 to 4 for example.
Ideally all work is best if done in primary cells because that is what the tissue is closest to. However, primary cells are difficult to isolate, grow and passage and above all expensive, as it involves sacrifice of animals and tissue extraction. That is where cell lines, which are immortalized primary cells come in. However the process of immortalizing causes changes in cellular morphology and characterestics. So you have to be sure that what you wish to determine, is not affected. Primary cells do not replicate very well, nor give you many passages as they tend to differentiate and stop multiplying. Also, like immortalized cell lines that should not be passaged over 40- 60 passages, primary cell lines should not be passaged once you find them slow down considerably in multiplication or get differentiated to 60% or more. Hope that helps!
Primary cells are a more heterogenous population of cells; a cell line is a more homogenous population although every trypsinisation you exert a selective pressure to your cells, so the cells with the best proliferation rate will survive
Hi Mark,
As requested, here are a few papers with advice about using cell lines. I can't get the 'add publication' feature to work, unfortunately, so I've just pasted the authors and titles in.
I've just submitted a paper where I stress the benefits of primary cells over cell lines, and I used these refs.
I would quote my summary here, but the paper hasn't been accepted yet...
Hope these help,
Stuart
36. American Type Culture Collection Standards Development Organization Workgroup ASN-0002 Cell line misidentification: the beginning of the end. Nat. Rev. Cancer 10, 441–448 (2010).
37. Nardone, R.M. Curbing rampant cross-contamination and misidentification of cell lines. BioTechniques 45, 221-227 (2008).
38. Buehring, G.C., Eby, E.A. & Eby, M.J. Cell line cross-contamination: how aware are Mammalian cell culturists of the problem and how to monitor it? In Vitro Cell. Dev. Biol. Anim. 40, 211-5 (2004).
39. Rojas, A. Cell Line Cross-Contamination: Who Wins? J. Biol. Chem. 286, le20 (2011).
40. MacLeod, R.A.F. et al. Widespread intraspecies cross-contamination of human tumor cell lines arising at source. Int. J. Cancer 83, 555-63 (1999).
41. Freshney, R.I. Cell line provenance. Cytotechnology 39, 55-67 (2002).
42. Hughes, P., Marshall, D., Reid, Y., Parkes, H. & Gelber, C. The costs of using unauthenticated, over-passaged cell lines: how much more data do we need? BioTechniques 43, 575-586 (2007).
@ Stuart, if your paper is accepted, please send me a copy. Otherwise I hope you will upload it to ResearchGate. Thanks for your insightful comments.
This is a very pertinent question that I am struggling with after 20 years in research. Very simply put, cell lines are easy to work with, you can have them forever, and the scientific community tends to accept your findings more readily because cell lines have been considered the standard for similar research in the past. Do the cell lines represent what is occurring in vivo? - no, not at all. For long-term studies of basic biology, cell lines are convenient as evidenced by labs using the same cell lines over decades. Primary cell lines have the following limitations - they are not well characterized, have limited life span, slow in proliferation and finally nothing to compare with. And frustratingly, journal reviewers will have more annoying questions about the cells and more experiments to validate, not because of your work, but due to their ignorance...........Primary cells derived from different patients can behave differently in culture conditions depending on the genetics and age of individuals from whom the tissue was derived. Also, fibroblasts derived from different parts of the body may have different characteristics...the beauty of biology!!!!!! Moreover, with every passage, the characteristics of primary cells change. I would therefore say that for consistency of results from experiment to experiment and from year to year cell lines would be preferable as long as they are handled properly. After all, as some suggested before, every time you subculture primary cells, over its life time, you are selecting for a sub population of cells continuously until they die out. If you really think about it you are selecting for cells that are capable of proliferating on plastic substrate, a surface the cells are not used to inside the body. These same cells fail to proliferate on softer substrates that approximate human tissue stiffness. More importantly, the population of cells you are dealing with at passage 20 is not the same as the ones at passage 5. Obviously, this can skew your results. Therefore in the case of primary cells, it is imperative that you work with the same cell population for the same experiment. So you may want to expand enough cells at one time and freeze them so that you can go back to the same passage for all your experiments to ensure consistent results (Shhhhh another secret - don’t share these cells with anybody and don’t trust cells others give you….basically you have to know what you are working with). The same goes true for cell lines, but less so. If you allow cells to grow to very high confluency, over a period of time, you are going to select for a sub-population of cells adapted to out maneuver slow growing cells. One more important note -make cell line identification a standard procedure in your lab. Clinical trials have occurred based on results obtained using the wrong cell lines; and journals are starting to ask researchers to validate their cell lines before publishing their data. There have been a number of publications recently based on these issues. Hope this answers some of your concerns.
Primary cells are ALWAYS better than a cell line. All cell lines need careful assessment of phenotype before using - does the phenotype match that seen in the tissue. Also, cells change their phenotype with passage and substantial stocks of early passage should be frozen: cells should not then be used beyond a small number of additional passages.
I agree with a lot of what has been said about the merits of primary cells, particularly Dr. Ambadys excellent comments; that they often more closely resemble the tissue they are taken from and that they are more representative of the human population, i.e. heterogeneous ... but that they are time-consuming, periodically frustrating and often require a much larger N in order to see biological effects.
But I also think it is important to note that that cell lines provide a very useful tool for modeling specific processes or identifying smaller changes which might be difficult to see in primary cells due to donor variability. There are definitely a lot of problems with the cell lines, including the aforementioned cross-contamination, selection in culture and the fact that many cell lines do not express particular receptors or other proteins which similar types of primary cells express. However, cell lines are more malleable for use as knock-outs and knock-ins, with reporter assays or to grow proteins or virus stocks. Overall, cell lines are just another tool and if experiments are well-researched and thoughtful, they can provide very valuable information about the process being studied.
Ideally, I think that a good experiment proves that their phenomenon occurs in primary cells, perhaps refines their understanding of the mechanism in easier to use cell lines, and then returns to primary cells to demonstrate the newly-defined mechanism in a more biologically relevant system.
I agree with Peter Gaskill. If you want to do any sort of manipulations such as RNAi, transfection etc, cell line is the way to go. By the time you finish clonal selection in primary cells the cells would be approaching close to the point of senescence.
We used to correct by the specific activity of each G418 batch (I personnaly used either Sigma or Gibco, either powder or solutions. After this correction, try to go up to 1.25 mg/ml (you already testsed2) and this takes time14-20 days to kill cells. Good luck
Dear Shambhavi,
We had the same problem in the early 1980's when the small group I was working in were interested in Endothelial Derived Relaxing Factor (EDRF), later identified as Nitric Oxide.
Ideally we tried to work with Human endothelial cells. The problem with that the amount of human material we could obtain. I isolated plenty of Human Umbilical Vein endothelial cells but the number of cells that I could isolate was never nearly enough to do the experiments we wanted to perform.
So we looked at both Bovine and Porcine Thoracic aortic Endothelial cells. The obvious advantage was the huge amounts of tissue that was available from abattoirs. We grew cells from both sources and could easily get the number of cells for our experiments.
It turned out that the Porcine derived cells gave us the desired result i.e. the release of EDRF by an agonist gave us a large relaxation of our tissue but was short lived. The bovine derived cells gave us a completely different profile i..e. a very small release of EDRF but over a long period.
The cells were also only used at passage 1. The reason for this was that at passage 2 and onwards NO EDRF release could be shown to be released from the cells.
We also observed changes in the Prostenoid's released from the cells. Again at Passage 1 (P1), Prostacyclin was the main prostanoid released. Prostaglandin E2 was undetectable at P1. At P2 this prostenoid ratio was completely reversed.
So the take home message is that cell culture is just a model which you use to your advantage. Primary cells WILL change dramatically over just one passage.
Cell lines have been immortalised and therefore are nothing like normal cells but can be grown in huge numbers
Everyone posting on this topic is doing cell culture in a completely false environment i.e. at 21% Oxygen. We purchased an Hypoxic chamber 10 years ago so we could look at cells in culture in varying oxygen concentrations, nearing tissue oxygenation levels.
Cell lines are usually clonal and are very homogenous in their cell growth and genetic characteristics and as you said are transformed and hence do not truly represent the cell types or tissues they are said to represent. Primay cells rarely grow in culture long enough to do lengthy experiments and each time we grow fresh cells from a tissue, you should understand that they are not essentially the same cells as the previous one. That is why it is essential to show a result in more than one patient when using primary cells while it is not so with cell lines (which did originate from one person at some point of time in the past). Simply put, cell lines are like the prototype models that we test space flight on. Primary cells are like the actual space flight in space. Most of the stuff that we test out on the prototype will work in space but some of the unique environments that only space can offer cannot be replicated in a prototype. Cell lines are used only for the reason that this can be maintained for fairly easily without the addition of any external growth factors or feeder layers etc. required for many primary cells and they are very repeatable. It just makes life easier for scientists. Most of the data that we get from cell lines are indicative of what will happen in primary cells. It ismostly a time saving method for researchers to find out what is going inside a cell, but it is limited by the fact that they are not exact representations of the untransformed primary cell. SO make your pick.
I totally agree with Dr. Sakthikumar Ambady. I have a lot of frustration using primary cell cultures since they sometimes will not give you exactly the same amount of change. I have o repeat several times to be sure about my findings.
The beauty of cell lines for most part is the reproducibility since they are clones. This is just like to use inbred animals which you can expect they will respond pretty much the same way. I keep asking about whether the finding from cell line, or inbred animal, can be count as n=1? Not surprsingly, there are reports using different cell line gave us tottaly opposite result.
I had a lot of discussion with researchers about some protein functions that are different in cell culture as compare to the original finding from purified protein runs in test tubes (a lot of them published in high impcat factor journals). Should we trust the findings in cells or the one from the test tube. My belief is the response will be highly variable in population and to use only one cell line will be dangerous in predicting the true response. That is one concern in translational research.
One more thing. Yes, the primary cell culture does change phenotype and morphology after a few passages and the cells will not be possible 100% pure. The problem using primary culture is that you can not use passage 1 or 2 since you can not have enough number of cells and it will be extremely expensive to use them in that way. I do know one of my colleague doing his research by human pulmonary artery microvascular endothelial cells (I was told it is from human fetus though) from one vendor (you can Google it). But, I do not know how good they are when pass the 6th passage.
I am glad to learn from Neale about the experience in NO research. My question is the phenotype of HUVEC. This type of cell will disappear within 2 weeks of life after birth and I wonder how should we classify them. Some told me HUVEC are like pulmonary endothelial cells but arterial or venous? I agree that a lot of us culture the EC in wrong environment with 21% O2 that was why we tried our experiment in both 3% and 21% environment to see the difference. At least in some assay the results are almost the same. We have used pulmonay endothelial cells from fetal sheep and was able to see production of EDRF, PGI2, TXB2, etc even at 6th passage. I guess different phenotypes determine the study results. I have to agree that we chosed the cell type of convenience.
I agree with most of the comments: primary cells resemble more the tissue where they are taken from. However, you should be aware that once you put them in culture you can change a lot of things in these cells.... concentrations of oxigen, availability of nutrients, etc... and the fact that they will change with each passage (mentioned by several of the people replying to you) is not unique to the primary cells. It also occurs in cell lines. During my PhD, my supervisor asked me to freeze 50 vials of a cell line I used for gene expression analyses, to make sure that everytime I did an experiment I had genetically identical cells.... not new, although most people seem not to be aware of it (primary cells usually are free from mycoplasma, which does not always happen on cell lines, even if you get them from repositories).
Usually is easier to work with cell lines. If you find something interesting, it will be necessary that you prove it either in vivo or in primary cells.... as you have seen from all the comments it will depend of the availability of appropriate cell lines and/or genetic models that are useful.
Yes, passaging changes the biological features of primary cells. Use primary cells if you have access to them. Most people cannot perform colony/clonogenic assays with primary cells. Cell lines are easier to use for long term experiments (like transfections, clonogenic assays, differentiation studies). primary cells are easy to use for short term studies, biochemistry, proteomics, genomics, apoptosis assays, confocal imaging.
It has been touched upon through out these comments, but primary cells give you a tangible and meaningful link to human disease. So whilst in vitro mechanistic work is powerful and achievable in the context of a cell line, its link to in vivo pathophysiology will always be in question. Therefore, targeted validation in primary cells of results obtained from cell lines, suggests that the biology in question is relevant to human health.
Furthermore, there are instances when the fundermental characteristics of a cell line are in opposition to the phenotype of a human disease, and in these situations the use of cell lines is highly questionable. An example would be chronic lymphocytic leukaemia (CLL), where a CLL cell lines, defined by a proliferation phenotype, is in stark contrast to the in vivo situation. Many CLL researchers would claim that a CLL cell line in a contradiction in terms, and therefore meaningless in the context of disease biology.
I agree with all above who carry out studies with cell lines to gain insight, but confirm their findings as best as possible with primary cells.
I am also fully agree with all above Researchers specially Sakthikumar Ambady, Jayakumar Nair...Thankyou sir for ur Valuable discussion
If you need to work with an endothelial cell line you may use EA.hy926. I use them for difficult experiments ( for example trasnfection of plasmids) or preliminary experiments. Hope this helps.
Have you considered immortalizing your cells (or other primary cells in which you are interested- say human aortic endothelial cells) with telomerase? Although this does not always work, cells immortalized in this fashion often retain characteristics more like the primary cells for a longer period of time than cells immortalized by other means. Of course, since the primary cells vary over time, it can be important when you add your TERT and the characteristics the cells had at that stage.
I agree with all the interesting comments above. Primary cells are ideal, but if one must use cell lines..its essential to do proper controls. Choosing a + and - control is worthwhile. It really helps in validating results. If one wants to find mechanisms, one can also avoid overexpressing genes-it does nor give you physiologicaly meaningful data. Knocking out genes, Transfecting mutant genes is also a good option.
Good Luck!
Some cells under passaging my change their gene expression characteristics such as chondrocytes. Carfull should be to be sure your primary cells have not be changed to other cell types.
Hello! I agree with most statements above.
In addition, I think it could be useful to optimise the experimental conditions you want to use with cell lines first and then move on to primary cells. It is also important to consider whether to use 2D or 3D cell culture with e.g. matrigel or collagen as scaffold.
Passaging primary cells changes their behaviour and I personally wouldn't use them beyong passage 5. Ideally you can get so much material, that you can start experiments already at passage 2 or 3 - which are necessary to get rid offf unwanted cell types (e.g. pericytes, immune cells, epithelial cells; fibroblasts are hard to avoid though).
Good luck with experiments
Yes, Passaging primary cells does alter cell characterisitcs...but cell density when plating also matters. We used primary human mammary epithelium and found that their differentiation potential and their response to drugs was influenced by cell density. We did not do systematic studies on this. Cell density also hugely affects the response of cell lines to drugs etc...Again, more studies needed!
Very good comments from the panel of scientists. Will be very helpful since i am in the lab culturing SCC 25 cells from ATCC
it's very good answer and comment.. Good for my study. Thank for all
I've read a few of the recent comments. I agree with quite a few of them, in that primary cells are more physiological, and that cells do change with each passage. I used to work on human endothelial cells, and had to use primary cells, as the cell lines available didn't express our protein of interest, which is specific to the endothelium. Primary cells are obviously much harder to work with, but I just thought I would add that it is very important to check the levels of your protein of interest with changes in passage numbers. I found that my protein of interest would decrease with increasing passage, which to me indicated that the cells were changing and had fewer 'endothelial-like' properties. We always use the cells at very early passages, but we're also lucky that we've got a good source of cells with a good collaboration with a local hospital (we use HUVEC from umbilical cords). Although you can buy them, extracting them yourself isn't hard, but you do need to treat them 'nicely' and be gentle with them. There are many protocols available on-line, many which are good, I am sure. Hope this helps, and good luck with your study!
As an alternative to "primary" cells you could also use (murine) embryonic stem cells/stem cell lines (e.g. V6.5). Advantage is that they are non-transformed, yet straightforward to cultivate (you can easily adapt them to feeder-free conditions, i.e. growing them on gelatinized plates instead of a feeder support of MEFs), and clonal of origin; moreover, they possess great proliferative capacity with doubling times around 10 to 12 hours when handled appropriately, thereby outperforming most if not all established tumor and/or transformed cell lines. In fact, we nowadays mostly use them as cell culture "workhorses" and even prepare nuclear extracts from them (starting from > 10^9 cells...). Their self-renewal capacity as well as their pluripotency (you can easily differentiate them into many different cell lineages) make them attractive cellular models for various applications. Of course, all of the above suggestions and comments concerning good lab practice when culturing/passaging/controlling cells apply also in this case...
Cell line is easier to work with and you save your time, but just be sure the cells not transformed in culture, I have experience with cartilage cells, they changes into fibroblast like cells and stop experssing genes characeterstic of cartilage when cultured in vitro
Yes, Yaseer Ahmed is correct. Primary Cartilage cells become fibroblastic in culture and stop expressing collagen type 2. One can get better data with explant cutures of cartilage. We could show anti-arthritic activity of glucosamine sulphate and some other drugs using these explant cultures. The C3H10T1/2 cartilage cells developed by Dr. M. Goldring are supposed to be good-but I have not used them myself.
I agree with most what has been said above only to add that I worked with primary hepatocytes cells and they too change into fibroblasts. I don’t really have direct experience with other primary cell types but I think at least for cells that conform solid organs, it is not only cell-to-cell contact that matters but also the extracellular matrix is important to maintain their identity. My speculation in the case of hepatocytes is that cells are normally blocked into G0 cell cycle but after isolation they enter into G1 phase (when in plate they are already in G1) and the only thing a researcher can do is to delay it. So, experiments with this kind of cells should be done immediately after cells isolation and one should avoid any cell passaging unless on purpose.
Also, last time I was sent DNA from normal mouse mammary gland, tumor tissues, cell lines derived from tumors and from a mammary gland epithelial Cell line that is used as a control in a wide range of signaling study to use it as control for DNA methylation analysis. The results indicated that the control epithelial cell line has an altered global DNA methylation pattern comparable to transformed cells at least at the level of all major DNA repeats I have analyzed. I was sent again the samples and again the same results discarding any artifact. I think a cell line can be validated to study a process that could be extrapolated to its physiological setting within the organism but not others. With transformed cell, the problem is far more complex. Each case has to be considered separately and so the process being studied. I think all scientists agree that passaging is an important parameter. Some cells may lose some chromosomes in the process, etc…
I think it will be more helpful if more scientists join in to share their experiences on this important thread.
Dr. Boukaba's comment is interesting. Methylation state will change as a cell goes through cell cycle-correct? So, how would one get controls and test cells with similar methylation status?? Perhaps using synchronized populations of both cell types (controls and test) will help.
Dear Venil,
I didn't mean that. That is not correct. I will try to comment on that later. I am really sorry if the way by which I have formulated my response to the thread has been misunderstood. Thanks
Dear Venil,
Sorry for the misunderstanding. DNA methylation is considered to be epigenetic because it is conserved through mitosis. Patterns of DNA methylation are normally “faithfully” copied during DNA replication and inherited by the daughter cells.
Only to make my previous two statements a bit clear:
1/ Referring to my experience (Not very extensive) with hepatocytes is that in monolayer cultures it is not possible to prevent them from dedifferentiating and that was the idea I have got from the lab who has a life time experience with that (Isolating them by collagenase perfusion of the liver). I have no experience with 3D cultures and how they can preserve hepatocytes cell identity. It is true that cell identity is defined by its epigenetic landscape such as DNA methylation pattern but to my knowledge the difference in DNA methylation map of cultured hepatocytes before and after their dedifferentiation has not been done.
2/ What I meant is that sometimes we may be using a cell line with the premise that it has all characteristics of a primary cell or even the characteristic of the tissue but that only hold true for a validated process. Maybe it is not sufficient from a system biology approach whereby all cellular processes are interconnected and hence may influence each others. It comes to consensus among scientists to accept or not a cell line as model for any given studied process.
In my case, even though as most of us have worked with breast cancer cell lines, I have no experience with mammary epithelial cell lines. What I said is that it was a surprise for me that an NON-TRANSFORMED mammary epithelial cell line that is used as a model to study some signaling pathways in a natural context, presented the same DNA hypomethylation profile at the level of the DNA repeats as a transformed cell.
Normally a cell line is validated by the presence or absence of specific markers. Is that sufficient? I don’t know. I think it is far more complicated than just that. But it might be OK for most studied processes. There might be differences from cell to cell of the same lineage that are unaccounted for by this approach. A pathway may be more activated in a clone than in another, the same could be said about an enzyme, transcription of a gene, etc..Single cell analysis techniques will provide very interesting answers too.
Thanks for clarifying Dr. Boukaba, I understand what you mean. Still, I find epigentics confusing. I learned a lot from your reply-best Regards!
Dear Dr Sumantran,
Here are two links on RG debating on the topic of the definition of epigenetics that you may find interesting.
1/
https://www.researchgate.net/post/Though_histone_modifications_are_currently_considered_epigenetic_is_this_the_correct_label_How_many_generations_can_they_be_inherited
2/ https://www.researchgate.net/post/Where_does_the_boundary_lie_between_epigenetic_control_and_non-epigenetic_gene_regulation-particularly_with_respect_to_the_role_of_modified_histones
Best regards
We work with epithelial respiratory cells and don't assume anything, so we repeat our experiments in HEp-2 and A549 cells and in primary respiratory epithelial cultures to compare resuts. So if you don't have endothelial lines just work with primary cells, since these are more similar to natural models, main advantage of cell lines is that they grow continuously and you always have biological material for your experiments, not more.
about cells passage...I am always wondering what this really mean...lots of us have cell lines...say HELA passage 40...is it simply a number on a plate or have these cells only been passage 40times....could not believe this...hela were established ages ago!
Hi Didier! The genomic sequence of HeLa cells has just been released. I still have to read the paper @ (http://www.g3journal.org/content/early/2013/03/11/g3.113.005777.abstract). But authors are questioning experimental models validated on this cell line. More will follow....It seems that HeLa is a complete mess....
yes, I have seen the paper...and a quick reading (see also the news and view in nature) suggest..that indeed it is a weird cell line and it should be use cautiously.
sayning that, from my experience there is as much hela cells than they are lab on this earth:-)...they sequence the kyoto isolate which show some specific features (they are flat and people use these for their imaging works).
I am niot sure they have assemble the genome and I am wonderingif one can perform blast searches on the genome.
If your cell are infinite cells they wouldn't change as long as your are passaging them using the same media and culturing conditions, your aortic endothelial cells would become arterial endothelial cells if changes made to the culture medium. Thanks
If your cell are infinite cells they wouldn't change as long as your are passaging them using the same media and culturing conditions, your aortic endothelial cells would become arterial endothelial cells if changes made to the culture medium. Thanks
If your cell are infinite cells they wouldn't change as long as your are passaging them using the same media and culturing conditions, your aortic endothelial cells would become arterial endothelial cells if changes made to the culture medium. Thanks
If your cell are infinite cells they wouldn't change as long as your are passaging them using the same media and culturing conditions, your aortic endothelial cells would become arterial endothelial cells if changes made to the culture medium. Thanks
If your cell are infinite cells they wouldn't change as long as your are passaging them using the same media and culturing conditions, your aortic endothelial cells would become arterial endothelial cells if changes made to the culture medium. Thanks
I alway first use cell line then primary cell later. At least cell line is much eays to culture, However after passaged again and again there are some deviation in cell line,so if possible i always try to verify it by primary cells
In general, primary cell will be used for distinguished the biology effect of one thing. After this finding, by use the several cell lines that finding should be confirm and for describe molecular mechanism of this event, cell lines are used because primary cell not be enough for several molecular experiments.
In vitro culture conditions definitely change the characteristic of primary cultured cells. During my study on primary colorectal cancer cells I found that chromosomal changes theme was changed within passages (you can see my article in my publications about subpoulation of cells).
This is a great question. It's comforting to see that we all have essentially the same concerns about the best ex vivo/in vitro models to use. I think everyone has made valid points. This healthy discussion clearly highlights the fact that with all of our fancy high tech equipment and our ability to sequence any genome at will. We still don't have the foggiest idea what's going on in our cell culture flasks. That said, based on my personal biases, I would like to add my two cents to the discussion.
1. All cells (primary and cell lines) change over time in culture. Thus all cells should only be used to passage 5-8. Remember, all cell lines were derived from primary cells that were grown in culture, reached a crisis stage, collapsed, an a few hardy souls with the correct darwinian characteristics (mutations) survived to become cell lines.
2.Remember to use good cell culture practices. Such as, check your cells for mycoplasma (it still amazes me how many labs don't do this), Get your cells from a cell repository like ATCC, If not quarantine your cells until you have confirmed they are bug free and they who they say they are. Passage your cells when they are 60%-70% confluent.
3. When using cell lines for your experiments, please, please, please use 3-5 different cell lines. This will at least give everyone an idea of the variability between the cell lines.
4. Cells don't normally grow on plastic. So at the very minimum, flasks should be coated with some form of extra cellular matrix. They maintain their phenotype better.
5. Cells do not grow in 2D, so we should all be thinking in 3D (IMAX for cells). As the literature clearly shows that gene expression of 2D vs 3D is very different.
6. No cell is an island (think ES cells). All cells are in constant contact with other cell types. As an example, when primary CLL or MCL cells are cultured in media alone they quickly die. However, when cultured with stromal cells they are relatively easy to maintain. The down side, the amount of expertise and time and $ required to maintain them increases along with the experimental design.
7. Yes, cell lines are easier to use than primary cells but remember, easier is not always better. The variability of primary cells most likely represents the variability often observed in patients disease.
8. When reviewing a paper that only used primary cells, please keep in mind that experiments using primary cells will be more variable that those using cell lines and were much more difficult to execute. Continue to be critical of the research, but remember, they took the road less traveled so cut them some slack.
My final point is, we should not be satisfied with our current models, we should all be working on improving our cell cultures models to better mimic the in vivo environment. In the mean time all of us need to keep in mind the limitations of our models and that are results may or may not reflect the real world.
Culture ON!
This is a great question to ask, and the answers on this thread offer great advice. Daniel offers great practical advice. I would like to add a few musings to the debate. Cell culture over a number of passages will change the cell if it is a primary or established cell line, I have seen many different clones of the same cell line behave differently depending on where they came from and how long they have been in culture. However primary cells can of often do change very very quickly. So if we want a model of a particular cell type then we really need to think about the cells that we are using, and try an mimic the natural situation as closely as possible-Daniel points out a number of things such as substrate coatings and stromal cells which could help your cell feel more at home and so less likely to be stressed into a phenotypic change. one great advantage of primary cell culture is that they do vary from batch to batch, so whilst the data may not be a tight as with a cell line, it may actually be much more robust.
I assume that you have good access to bovine aortic tissue, and one thing that you do not mention, but which could be really useful is ex-vivo culture.
I have worked with BAEC's in the past. It would be interesting to use them again and assess endothelial nitric oxide synthase under the conditions which we grow some of our cells in the COY hypoxic chamber (i.e. at 01-1.0% oxygen).
As people have previously stated cell culture is a flawed method for looking at what happens to cells in vivo. I am sure that growing cells at 21% oxygen as we all do is not a good starting point.
I only partly agree with of my predecessors deductions. From physiological point of view cell line are not useful . For 30 years I worked with coculture of theca and granulosa cells. I insulates them from different sloughterhause animals but always in the same period of the follicular phase. VERIFYING for me is the secretion of estradiol in primary cultures. It showing me that my model is good. However I used one rule. Always are pooled cells from the same size of follicles from 5-6 animals. Then I avoid intra-individual differences. Answering for questions interesting for me cell line of eg granulosa cells is not appropriate.
I agree with most of the answers but my recommendations are:
1. Primary cell culture is better for your work but don't forget the control culture
2. If you have the ability to do 3D culture, please go ahead
3. Primary cell culture is very closely related to In vivo testing
4. if you design to perform your work after passaging the primary cells, try to do that on cells with the same Passage number
Stability of long term cell ljne depends on the origin of that cell line. This stability usually seen in transformed cell lines (apparantly normal cell line). No cell line can be gown in vitro for long term unless the cells are genetically transformed that escape division control mechanisms. However cancer cell lines usually show different genetic change including chromosomal changes. When we examine colorectal cancer cells for example we found no similarity in chromosome change among cells of the same colorectal cancer type. Milions of changes can be seen in the same cancer type. Primary cell culture is more close to the original cells in vivo, within time the invitro culture conditions will form a pressure on the cells that makes cells change their characteristics to live under those in vitro culture. Any researcher can test this truth by studying the chromosomal situation in any cell line during many serial passages, tgen the results will definitely show changing in the pattern of chromosomal changes indicating characteristic cha nges.
Interesting topic and a great conversation that came up!
I agree on all the points made.
We work with living organisms and therefore it almost is thier nature to change when exposed to selective conditions as cell culture really is.
I worked with a great variety of cells so far and sometimes even when you work with the said "well characterized" cell lines that are generally accepted (also by journal reviewers) as model system (in my case bone biology) you can not believe your results until you could also show it in primary cells. AND if you find something that speaks strongly against the use of a certain cell line as model system, don't be too shy to publish it and share this knowledge. We had results that were compromising one of the most used cell lines in our field and it was indeed hard to convince people of our findings.
I'm aware of the long list of said disadvantages of primary cells, but they are the better model to mimic the in vivo conditions.
@Shambhavi...I agree with all of the above points, so I will just let you know what we do to address the "cell line vs primary cells" issue. Because the cells we work with are very infrequent, and the assays we use require a significant number of cells to obtain all the results we want....we first do our experiment with a cell line (several batches from different frozen stocks). If possible, we then repeat the experiment in primary isolated cells that we collect ourselves. We only use fresh isolated cells and they are only reliable for ~2 weeks. We try to use them on the same day. I realize that this may not be possible for your studies, but as the great comments made in this post have already pointed out, time is your enemy with cells out of their context (the body). When the cells are removed from the body or even culture conditions that are more biologically appropriate, I have seen changes in their phenotype within hours!!
Hi All to risk further muddying the waters does anyone have any thoughts on the relationship, pros and cons of explant tissue cultures compared to primary/established cell line lines as ex vivo models?
Michael, in my experience, explants of articular hyaline cartilage are preferred to articular chondrocytes for studies on osteoarthritis . We and others published data showing that explant cultures work v. well for evaluating anti-arthritic and anti-inflammatory drugs. This suggests that explants are giving physiologically meaningful results which in turn means that the paracrine signalling in intact. This is not surprising because the ECM of explants is undisturbed. In the case of Articular chondrocytes, there is no authentic ECM and they rapidly de-differentiate in vitro. Hope this helps!
Hi there, I agree with Daniel, all cells are part of a system and if their reaction to specific conditions are to be tested surely they have to be tested within a system which is as close to their natural environment as possible. I've been looking at some of the 3D systems for some of the stuff we hope to be doing and it looks like this is the way to go especially when drug testing and clinical trials may be a follow on.
If you can use explants that may help but as pointed out by most folk you have to get your conditions right very quickly so that you don't lose them, and if the tissue is from human donors they will be rare (make sure that ethics, consent etc is all taken care off.) so practice on the closest animal equivalent is essential to get your technique sorted.. If you can use the 3D systems and get the correct proportions of cells (so cell ID would be important) you could probably grow a nice system which you can use in situ for testing then process for IHC, Immuno or whatever.
Agree with the points above. But one may always keep their minds open: No perfect models exist, regardless of whatever cell lines, primary cells, animal models or even certain groups of patients. Simply because of variation between individuals, species etc combined with environmental factors. Nature is complex. And science is expenditure, we have to start somewhere depends on where you are. And it works, as you see the knowledge gained in the past by our predecessors. Just some common-sense comments.
In addition to the strong effect of in vitro conditions on cancer cell lines grown in vitro, the freezing down process exert another side effect on those cells. Freezing cancer cell in liquid nitrogen may alter the proportion of cell subclones within the same cancer cell population of the same tumour origin. In spite of all those disadvantages of in vitro cancer cell lines, they remain the best tool for most biological and medical research work.
Immortalized endothelial cells may be fine for basic routine stuff (like learning to grow and maintain cell culture) but they are not viable in long term. The advantage of primary are that they are close from in vivo however they differentiate rapidly and loose their original phenotype. There are some described in the litterature but their origin are a bit "exotic" such as the EAhy.926 (hybridoma) or the ECV306 (that some people consider as bladder carcinoma mislabelled as sponatenously immortalized HUVECs).
I am working with brain endothelial cells and so far the primary rat brain endothelial cells isolated from rat are the best in terms of material access, if you can access to bovine and porcine, it is also fine.
For primary cell cultures from companies, except HUVECs, I had bad experience overall with them.
Primary culture can be used to assess the previous in vivo effect of something targeted on the cultured cells while long term culture can be used to assess different in vitro impacts on that cells because primary culture may lives for short term and cannot give the results of in vitro studies like peripheral blood lymphocyte in short term culture. Long term cell lines can be used to assess such as cytotoxicity , cytogenetic changes, physiology and other biological test because these cells stay living for long term and can be followed up easily.
Once the primary cells are transformed into normal cells, there will be certain changes from their function point of view. Therefore, checking the effect fo growth factors would be best on the normal cells than on primary cell line. But then it all depends on which is the gene you are trying to study.
Cell line data alone is hard to believe, even when many cell lines of the required tissue are used. Primary cultures data is essential in all in vitro studies even though it is difficult to get and maintain. If no primary cells are available, one can use immortalized cells (intermediate between normal and transformed). We and others have reported that immortalized MCF10 cell line gives different responses when compared with normal and transformed human breast cancer cell lines like MCF-7 etc. Unfortunately, it is not easy to find immortalized cell lines for many tissues like ovarian and colon epithelium. Hope this helps!
I would like to add the additional comment about cell lines, that even if they are obtained from reputable cell banks and are found not to be contaminated, they may behave differently and have a different genetic profile. This is what our group has found using the Walker 256 breast carcinoma cell line from two different cell banks:
Characterisation of Walker 256 breast carcinoma cells from two tumour cell banks as assessed using two models of secondary brain tumours.
Kate M Lewis, Elizabeth Harford-Wright, Robert Vink, Mounir N Ghabriel
Cancer Cell International (impact factor: 1.97). 02/2013; 13(1):5. DOI:10.1186/1475-2867-13-5
Source: PubMed
I would say that choosing one over another depends on what kind of experiments are you planning to do with them. Cell lines obtained from cell repository are either "altered" or cancerous cells, in which they have high number of mutations. This may not go well if you are planning to do some gene expression analysis, due to the eQTL difference caused by some mutations. On the other hand, using cell lines have the upper hand on reproducibility, which is kind of the pitfall of using primary cells (as they behave differently with each passage).
As mentionned before, whatever you observe in a cell line must be confirmed in a different cell line or ideally in primary cells to ensure that whatever you observed is not linked to a specific mutation that occured in your cell line.
Most (if not all) cell lines are cancer-derived cells and the more they are passaged, the more mutations they will accumulate. It's always wise to test several "similar" cell lines for whatever you are looking for and then primary cells if it is possible.
We recently published a paper on dendritic cell lines (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3491238/pdf/fimmu-03-00331.pdf) and how usefull they were to screen different treatments, activations,... and then primary cells were used to confirm whatever interesting observation we made in these cell lines. These cell lines showed a significant interest in the fact that we could grown millions without problem, wheras it is difficult to get a lot of primary dendritic cells (linked to the fact that they don't survive very long ex-vivo). It was therefore possible to test a lot of parameters and get a lot of material for large preliminary experiments, before focusing on much lower cell number from primary DC and confirming our results.
Unless you are working on the specific behavior of a cell line towards a given treatment with increasing passage number (which also would need controls), primary cells are the best tool to confirm your preliminary observations made in the cell lines.
Adding all of above answers, it is important to know that in personalaized medicine we should use primary culture for each person and also the most similar results to in vivo studies can be made with primary culteres which are not passaged.
The origin of cell lines in some cell banks is not from the originators and they prerpared them from the researchers in their countries. unfortunately some of them are missidentified( recently I had received two missidentified cell line from CLS cell bank), so you should order the cell lines from the most reliable cell banks.
I agrre with Parvaneh Farzaneh.
Some years ago I examined the effect of different chemoteraupetyków used in patients with ovarian cancer in primary cultures. I used a test Alama Blue and caspase-3 activity. The answer to the individual used alone or in many cases, two at the same time was different. In some cases it was indeed visible inhibition of proliferation apoptotic effect action, but unfortunately 30% was seen not only the lack of inhibition of proliferation but even stimulation.
Unfortunately, the number of cases (30) was so small that I could not publish the results, but I have the results in my database
Yes, Ewa is correct. We also got respondesr and non-repsonders to Anti-inflammatory drugs in explants of cartilage from arthritis patients. We had small sample numbers and did statistics on these small groupss and showed significant results. When the reviewer rejected the data, we pointed out that there is real heterogeneity in patient response, and how bichemical proof is absent in the literature. Then, our paper got accepted! Ewa, I hope you publish your data...such info is needed for ovarian cancer too ! All the Best!
Thank you for this statement. Please send me a pdf of your publication. I will be able cite your research in the discussion.
Dr. Ewa, Thanks for your interest in our paper. We worked on herbal drugs and published in a good journal (ECAM). I have attached the pdf. Citing this may not help you for a cancer paper..correct? If you find any on ovarian cancer where they talk of CR, PR (Complete response and Partial response), it may be worth citing them. What about the samples you used for primary cultures? Do you have any patient details on CR and PR for them. That will help a lot. Hope this helps!
Dr Venil,
The aim of our study was to find the both cheap and quick method for selecting the best chemotherapy for individual patient. I had a medical history and histopathological diagnosis.
Tumor derived tissue were obtained from 11 ovarian cancer patients (not much). Cells were obtained from tumor explants by enzymatic dissociation and incubated with different combination of chemotherapy drugs for 6 days. At the end proliferation and caspases activity was evaluated. 50% had been resistant for both single and multidrug chemotherapy. We received enough cells that can be carried out all the combinations in each female patient
What's that? used in patients had no effect. But they were also female patient who have had single drug activity. So why give 2-3 simultaneously? I think that 200-300 patients tested so would give chance to take into consideration such examinations prior to treatment.
I have to once more to look at this set aside publications and try to publish somewhere
Yes, it is worth doing more samples and getting reproducible data. Can you use explants without enzyme digestion?? Enzymes may kill some cells-right? As you know, there are companies selling such technology--Oncoprint, Mammaprint, histoculture assay etc...but I do not think they are validated for clinical use -especially in ovarian cancer (OC). If you can choose the most popular drugs for OC, and do expts with 50-100 samples and show response or lack of response...it will be great ! Pl. try for this!
My colleagues believe that primary cells are preferred by reviewers and in high impact journals especially with human primary cells. That's why researchers have trouble in choosing cell types for their research and use primary cells as a first choice for higher impact journal publication.
As I am in tissue engineering and regenerative medicine, a feasibility study of new method need to be produced with cell lines. Reproducibility is main issue at first step and cell lines are better for this purpose. However, cell lines are different from primary cells, it looks the results need to be confirmed with primary cells.
It looks that clinical translation, primary cell based study should be a choice, but proof of concept or other basic research may start with cell lines.
Venil,
I have I applied trypsin to obtain single cells. The amount of material obtained during surgery is small to perform all variants in several repetitions. In the case of trypsin, cell viability is realy good (90-95%). I trypsinized several times in a very short time. Trypsin activity easily inactivate.
Chang,
I do not agree with you. The cell line is only one cell type. In the organs are important co-operation of all the cells making up him and secreting various factors acting on auto-paracrine manner.
Right Dr. Ewa, I hope to see your pdf on this OC work soon. Its very important from personalized medicine point of view. I think Dr. Chang's point re: working out proof of concept in cell lines is understandable because (as you say), we get so little primary tissue. Once we have an interesting result from cell line, we must verify in Primary culture.
I am totally agree with sakthikumar and stuart. If you want your answer try to read both answer. They have taken care of all the issue. So i dont want to write anymore. Good luck
- Badges
- Science topic