I originally wrote this in Feb 2008, and later updated it for the old Lay Science. While making sure that this website was up-to-date, it occurred to me that this post would have disappeared with the rest of the Lay Science site. I have forgotten what updates I made when it made the move, and if I were setting out to write this article today it would no doubt be a completely different style, but here, for the archives, is the original version.

If you have ever worked in a molecular or medical biology research laboratory, chances are one of the first things you learnt was tissue culture (or the microbiology equivalents). Even if you know nothing about biology, you’ve probably heard mentions of “cell culture” on the news, or at the very least heard about the results of studies in tissue culture. If you hear about “cell lines”, you’ve got culture. If you hear about a “laboratory study” showing that your favourite chemical is carcinogenic, you’ve probably got culture. If you hear about new trials on a miracle cancer cure that has been shown to be effective in “preliminary laboratory tests”, you’ve probably got culture. Everything from zombie epidemics to £10,000 animal-free beef is cell culture. Knowing a little bit about what tissue culture is, and what its uses and limitations are, is therefore important when answering such questions as “is my baby’s bottle poisonous?”, “is stem-cell research ethical?” and “is vitamin C an effective cure for colds/cancer/HIV?”

So. What is tissue culture (TC)? It’s when you take specific cells from a multi-cellular animal and grow them in a dish full of nutrients (a mimic of your blood serum). The point of doing this is to create a system on which to experiment which does not require growing and killing lots of individuals — something that is, for some reason, considered unethical. Especially when it’s humans you propose using. Typically, human or other mammalian tissues are used — especially “model organisms” such as mice. You can use healthy or diseased cell lines from all sorts of different organs. Once you’ve grown up a nice batch of cells in your dish, you can see how they respond to your cancer drug, environmental contaminant, or new junk food ingredient. You can see exactly how the behaviour of your cells changes over the minutes, hours and days of exposure; how they recover after the chemical has been flushed away; how your cancer drug works in dozens of different tumours; how your junk food ingredient works in the old and young, male and female, fit and fat; and how your environmental contaminant interacts with other environmental contaminants. It’s great. If you work hard enough, you can know everything you want to know about your chemical within a week. Wipe out cancer and save the world by next Monday. At least, that’s what the animal-rights movement would have you believe. And the tabloid press fall for it daily.

Trouble is, it’s very easy to get superficially interesting answers using TC. Which makes it very easy to convince a journalist that you have important results, but very difficult to convince a scientist. That’s not to say that TC is not important. But everything that we measure in TC is an estimate of what happens in real life situations. It’s a model that uses surrogate measures from which we can develop hypotheses about what happens in reality. A bad analogy is in order, I think. Suppose you are building a car. You want to protect your future drivers from side-on impacts. Very early on in the design process, you have an engineer conduct strength tests on different materials and designs for doors. From this, you can narrow down the field of designs, and make hypotheses about which designs will perform best on the road. But you can not be sure that the strongest material will provide the best protection against injury and death. You would want play with the crash test dummies, before putting the car on the road. And once the car is on the market, you would analyse incidents. Because when the door is attached to the car and put on the road, a huge number of other variables comes into play. And so it is with, er… what was the topic again? Tissue culture.

Cells did not evolve for growth in a dish. They evolved in the context of cooperation with a vast number of other specialist cells in a body. They are not fine-tuned for survival in the absence of skin, an immune system, a digestive system, liver and kidneys. They are not supposed to live like barnacles on plastic. But if you’ve worked with research quality cell lines, you’ll know that it’s surprisingly easy to make them grow in a dish. Feed them every couple of days, and they’ll happily live for many months. Well go and say that to the post-docs and technicians who made it that way. They were up until midnight processing disgusting lumps of freshly excised tumour. They spent months trying out different combinations of nutrients and fungicides in an attempt to make the cells survive longer than a week. They may be easy to grow now, but don’t think there wasn’t any effort involved. Billions of cells died in the process of making those few grow. Under these circumstances, you can hardly expect the cells not have evolved a little. You are introducing them to a vast number of novel mutagens by taking them away from the protection of skin. And putting anything into a new environment is going to mean new selection pressures. When you finally manage to immortalise your cell line, is it because you’ve perfectly adapted the conditions to the cells, or because the cells have adapted to the conditions?

So. There are all sorts of reasons why TC can not be anything more than an approximation of what is happening in real life. A useful approximation, but unreliable in the absence confirmatory evidence from in vivo and population studies. But these are only the intrinsic limitations of TC. When judging the merits of TC based research, you must also take into the account the fact that TC is easily misused and misrepresented, and that charlatans are doing it all the time. TC is a favourite of cargo-cult healers and nutritionists — those who like to keep up a superficial appearance of having a scientific basis for their quackery. Take, for example, the shamen who pedal vitamin C as an HIV/AIDS drug (Patrick Holford, for example) or as a cancer therapy. They will tell you that in TC, vitamin C has been shown to kill tumour cells, or those cells that are infected with HIV. Therefore, the reasoning goes, we should abandon proven therapies, in favour of taking some vitamin supplements. Trouble is, you can chuck a big lump of any chemical in a dish of cells and the cells will die. I could pour a bag of vitamin C into a dish of healthy cells. They will die. Conclusion: those vitamin supplements are deadly poisonous. Except that your cells will never be exposed to a bag of vitamin C, because you have skin, a digestive system, and kidneys. And because people just don’t go around pouring bags of vitamin C down their throats. I could spit in a dish of cells and tell you that spit is a killer. It’s not.

But it’s not just charlatans that abuse TC. Many legitimate scientists bend the rules a little. They may not even be aware that they are doing it. Take the case of Bisphenol A (BPA), something I did a little work on a couple of years ago. BPA is a component of some plastics, notably bottles. It is known to very slowly leach out of the bottles and into your drink. There is a little bit of evidence (mostly from rats) to show that consuming BPA may be harmful. And there are a lot of TC experiments on the chemical. BPA is a xenoestrogen, meaning that it mimics the activity of estrogens. Estrogen, of course, regulates prolactin release, and cell division (particularly in the breasts). We know that BPA mimics estrogens because when we put some in our dish of tumour cells, we see that within seconds the estrogen receptors have been activated, and all the other effects of estrogen follow. There are loads of results to confirm this because there are a lot of experiments into the effect of estrogen (there’s plenty of money in breast cancer research). If you’re doing the experiment anyway, it’s hardly any more effort to look at BPA. And you can pretend that your research has another potential medical application. Since it’s not the primary aim of your research, the journal’s reviewers won’t notice that you’re using it at a thousand times the concentration that you would find it in the body. So even if enough BPA does leach out of your bottle, and even if BPA does do interesting things in the body, a large proportion of the TC studies will be irrelevant to understanding how it does those things, because they look at inappropriately large concentrations and inappropriately small timescales.

So, next time you are flicking through the health pages of the Daily Mail — which I know all of you like to do — engage healthy skepticism when they update the list of miracle cures and carcinogens. Like statistics, tissue culture is incredibly useful — whether you’re searching for the truth, or a convincing lie.