This may seem a stupid question, and might actually be one:
I'm familiar with cell culture and do experiments with cells. It is widely know that most mammalian cell lines cultured in the laboratory, as most cells, follow a growth curve with lag, log and plateau (and later death) phases. It makes sense for production to split them in the log phase, when they are actively dividing and colonizing the available space. But when it comes to plating them and performing experiments, it seems that we have inherited that rule: virtually all of the experimental procedures that can be found in papers are performed with cells between 60 and 85% confluency. Asking around the laboratories, everybody seems to think that using confluent cells for experiments is something that should be avoided.
Of course, the growth curve itself makes evident that confluent cells and log phase cells present very different behaviours, but I'm not able to recall what were the reasons that lead researchers to decide that the only relevant ensemble of cellular states was the log phase or that it was preferable to the others. In fact, we use cells with traits of cancer cells (when not directly cancer cells) because of practical needs (easy, sustained and humane production of cells for research) but we frequently use these systems to try and draw conclusions about general cellular biology. But most of the time cells in mammal bodies are not cells actively dividing and colonizing available space (those would be cancer cells and cells during development and wound healing) and one could think that cells between the end of the log phase and the onset of the death phase might be more reminiscent of cells in their physiological state. So why avoid using them? Or I am wrong and there is a significant body of research and knowledge on the biology of confluent cells that I haven't come across to?
Confluent cells are certainly used in cell culture experiments.
For example the majority of endothelial cell research, particularly dealing with aspects of shear stress response, utilizes confluent cells. For example, some very recent papers:
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0086249
http://onlinelibrary.wiley.com/doi/10.1002/jcb.24639/full
Further, it's common to serum starve the cells as well to ensure that they are quiescent.
Obviously, in vivo our endothelium forms a monolayer coating the interior of the vasculature (in healthy vessels anyways), so confluent, quiescent monolayers represent the best model and have been used for more then 30 years.
Now, it is important to note that when sub-culturing cells, it's typical to passage them during the log phase, before the cultures reach confluence. Thus way, the sub-cultures will be ready to grow.
Of course, it is always important to remember that cell culture represents just a model system. Depending on what one is interested in studying, it is important to consider such factors as growth phase, confluence, presence of absence of serum, etc. in order to provide the most physiologically relevant conditions. When drawing conclusions from cell culture studies it is important to keep in mind the many assumptions made and the limitations of the model.
Confluent cells are certainly used in cell culture experiments.
For example the majority of endothelial cell research, particularly dealing with aspects of shear stress response, utilizes confluent cells. For example, some very recent papers:
http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0086249
http://onlinelibrary.wiley.com/doi/10.1002/jcb.24639/full
Further, it's common to serum starve the cells as well to ensure that they are quiescent.
Obviously, in vivo our endothelium forms a monolayer coating the interior of the vasculature (in healthy vessels anyways), so confluent, quiescent monolayers represent the best model and have been used for more then 30 years.
Now, it is important to note that when sub-culturing cells, it's typical to passage them during the log phase, before the cultures reach confluence. Thus way, the sub-cultures will be ready to grow.
Of course, it is always important to remember that cell culture represents just a model system. Depending on what one is interested in studying, it is important to consider such factors as growth phase, confluence, presence of absence of serum, etc. in order to provide the most physiologically relevant conditions. When drawing conclusions from cell culture studies it is important to keep in mind the many assumptions made and the limitations of the model.
In my opinion, there also exists a tendency to copy conditions that were successful to other researchers without taking the time to set the conditions for our own experiments. We avoid performing the preliminary experiments to establish the best way to address the question we pose. For example, to choose a dose we should perform a dose-response curve, to choose a confluence condition we should try different options before and so on. The pressure and rush to obtain results to be published are against this “lost” of time.
Maybe you should perform your experiments in the two conditions and evaluate the differences of your results in one and another. I did it with my experiments (Cryobiology. 2013 Aug;67(1):76-83. doi: 10.1016/j.cryobiol.2013.05.005) and, although I did not find very sound differences there was a diverse behavior between log growing vs confluent cultures.
Thanks Peter and Maria for the references and your answers. i think the problem is that the models used too often are not explicit in papers, which in the context of rush that María describes leads to a sort of "popular lore" and you often get advised not to use confluent cells because "they behave different".
To put my case in a more specific frame, right now I'm working on interactions between proteins in the context of endocytosis. Since endocytosis is a rather general cellular activity I don't think I should look for a tissue model. And if any, epithelial cells, for their positions as boundaries in tissue extensively must exchange material and information with their surroundings, so they could be a good one. Then, if I try to co-immunoprecipitate proteins involved in endocytosis in confluent cells and find different interaction patterns or localization to different subcellular fractions when compared with the same cells not in confluency, that wouldn't invalidate what I might find about these interactions, but would only reflect the adaptation of the endocytic machinery to the conditions of the cell (an information that would be very interesting anyway). Since immunoprecipitations require a good amount of cellular extract, having higher cell densities also could prove useful. So I think that, unless I'm missing something, there is no valid reason to shy away from (also) using confluent cells for my experiments.
At any rate, I think some emphasis in expliciting the models used in the publications should be made. I for one will try to always include that information at least in the supplementary methods section in the future.
I think semiconfluent cells in culture are usually selected only when polarity and monolayer formation ( for epithelial cells as judged by Trans epithelial conductance/Resistance) do not play a significant role affecting the final read out of each experiment.
depending upon the working hypothesis the experimental designs are usually made and that can dictate some one to optimize each experimental conditions including which phase of the cells in culture could be used for a doable and meaningful experiments.
Confluent tight monolayers are definitely required for experiments to study electrical conductance/ resistance, paracellular transport of small molecules, ions and drugs and for studies involving proteins that only express in functional (gap junction/tight junction proteins) form whenever the cells are confluent where as semiconfluent cells are usually used for experiments targeted to measure protein and gene expression, microscopy, ratiometric fluorescence for measuring ion and pH which are not supposed to alter depending upon cells degree of confluency.
Studying the tight junction complex i usaually worked with confluent cultures.After couple of days in the confluent state the TJ developed maximun transepitehlial electrical resistance. Please concider the attachment.
The questions are: which cells do you use and what are you looking for ?
If you want to research healthy epithalial cells (or functions of healthy epithelia) you must use confluent cells but if you want to research wounded epithelia you should not use confluent cells (and the same for endothelial cells and others).
Therefore, there is no universal answer to your comment. Nevertheless, there a many, many studies using subconfluent cells for studies in which the researchers should use confluent cells - so it is on you to rely on these results.......
I agree with Patrick and Brandon. With few exceptions, studies with confluent cells reflect more physiological conditions than those with sub-confluent, still rapidly growing cells. However, since confluent cells in a dish are still very far from a "physiological normality", one should select for the exps the right time window after the confluence is reached – so that cell-to-cell interactions are maximized - and before cells start suffering and dying
Hi Andrés, there is also a cerebellar granule precursor cell model of the neurodegenerative disease JNCL where cells are grown in confluency for 7 days so they build the lysosomal storage material that is typically found in patients (Fossale et al 2004). In my opinion, this treatment mimics better the stress/condition they have in real tissue whereas still proliferating cells allow us to seperate early phenotypes from symptoms of the mutation.
It is a really intersting and important question Andres. If you passage at low confluency, your cells will be selected for faster growing cells over time. On the other hand, with high confluency the gene expression due to contact inhibition will be changed .
As long as you use the cells from simillar confluency for all experiments I do not see any problem using cells from high or low confluency .
Hi dear Andrés,
I do apologize for late answer.
I hope you got the proper answer from our collegues. In my point of view and for the sake of reproducibility always use about 80% confeleunce (or grow specific number of cells). The full confleunce cell culture will affect the cellular expressions.
Best wishes
Sarmad
- Badges
- Science topic
- Similar topics
- Biological Science
- Cell Biology