I attended a small local conference a few years ago at which a presentation on practical low oxygen cell culture was given. It seemed like the perfect way to pep up our fetal hippocampal neuron cultures. But it never happened because of, um, "research culture"
It's actually not that expensive to do:
http://brincubator.com/
(Disclaimer: I have links to these folks, so treat that as unsubstantiated marketing drivel until you prove otherwise :) )
That's an answer which hasn't had much thought applied to it. In the groups I've worked in, the problems with culture conditions were long appreciated. The factors entrenching the existing practices are largely down to: standardisation between laboratories, being able to publish (non-standard practices can be a large negative for reviewers), and cost (DMEM is cheap in liquid or powder form).
I've used specialist media for endothelial cells, and stem cells of various types (human and mouse, embryonic and mesenchymal). It could be up to £200 for 500ml of culture media, with a shelf life of just a week once you'd added all the frozen supplements. This makes things very costly once you scale things up.
But culture media are only the beginning of the problem. Another source of big problems are the culture conditions. Standard incubator conditions used everywhere are 37.5°C, 5% CO₂, 100% humidity. However, this is not remotely realistic, and is actively damaging to most cell types. The 5% CO₂ keeps the pH in balance (the media contains buffers). But the other 95% is regular atmosphere containing ~20.9% O₂, so the effective concentration is ~19.9% O₂. But the physiological oxygen concentration in the body's tissues is much less, a fraction of a percent.
One culture phenomenon is that "primary" cultures of isolated cells don't proliferate much, but after several "passages" they eventually become "established" and more proliferative (essentially, somewhat cancerous, and losing some of the characteristics of the original cell type). There is also a limit on the total number of passages the cells can withstand (all except the most aggressive cancer lines). This is largely down to cumulative DNA damage from oxygen. If you culture the same cells under low oxygen conditions, the cells don't become established, retain their characteristics, and can be cultured nearly indefinitely. Much more realistic. This can be done with an incubator with additional gas feeds and an oxygen sensor, which can purge the excess oxygen with N₂ then add CO₂ to 5%.
Another problem is that the culture medium is static. The cells use up its nutrients and then sit in their own waste. It needs regular replacement. But this cycle of nutrient excess and deficit and waste accumulation isn't realistic either, and may also have profound effects upon cellular physiology. Tissues in the body are continually perfused by blood plasma, maintaining nutrient levels and removing waste. This is something else which researchers should be developing; in fact, developing such systems was a part of my PhD (an unsuccessful part, I should add; it's actually quite technically difficult to perfuse patterned 3D scaffolds). But doing this for flasks and plates is a much simpler problem.
So these three major areas (media, conditions, perfusion) all contain major deviations from physiological conditions which all have the potential to grossly affect research outcomes. The fact that we have all collectively stuck with what worked for over 50 years is somewhat outrageous, particularly when for the first two there are commercial solutions which could be used today, but most people ignore. That's lazy. None of it is a "money-making scheme", though there would certainly be opportunities for media and equipment vendors.
I cultured immortalised and primary cells during my PhD, and i came away with one firm decision about my postdoc: i would work on a model organism where i didn't need to do this cringe-inducingly unphysiological practice!
There were groups at the university working on zebrafish and Xenopus tropicalis (a little frog), both of which are pretty sweet experimental systems.
Zebrafish are really easy to raise, and are transparent, so you can do live microscopy on them without having to do surgery or take explants. Someone i knew worked on a project where you zap them with lasers to cause wounds, and then watch immune cells respond to the damage.
Trops are more work to raise, and the embryo is not transparent, but they are the only vertebrate for which we have powerful classical genetic tools - it's naturally diploid (two copies of each chromosome, like humans, but unlike many other animals you might consider using), but if you produce embryos using sperm which has been zapped with UV light to destroy its DNA, they're haploid, and then if you block the first cell division with a cold shock, they're diploid, but with two identical copies of each chromosome [1], so sort of functionally haploid in a way which lets you investigate the effect of mutations really easily.
Absolutely agreed. My primary area of research was vascularisation, and I read a lot of zebrafish-based work and saw some awesome microscopy work at conferences. I would have loved to have worked with tools like this. It made my own efforts (3D collagen scaffolds with co-cultured HUVECS and smooth muscle) look utterly unrealistic and deeply flawed from the outset (which, in retrospect, they were).
The thing i was thinking of is a rival frog, _Xenopus laevis_. That has been a classic model organism for research since the 1930s, and so ideally you would use that. But it's tetraploid - four copies of each chromosome. So are other related species of frog which are similarly useful. But not _X. tropicalis_ - that is basically "the Xenopus you can do genetics with!".
Zebrafish are diploid (i think), but their genome underwent a duplication during evolution, so they have two copies of each gene on each chromosome (more or less), which makes them hard to do genetics on in the same way as being tetraploid does.
As the other comment mentions, the embryos are, but also, you can get amelanistic zebrafish.
This might be an interesting addition. You can't usually just buy zebrafish from Petco and use them for research. There are standards in animal facilities that determine whether your fish must be "specific pathogen free" or various other levels, and so you have to generally buy them from one of a few sources if you can't bum some from a beneficent lab. When we went to purchase embryos to set up a zebrafish colony (at something like $400 for 2 females and 1 male), the facility in Europe had suffered a fire that required them to stop production for 4-6 months.
As an aquarist I find this hilarious. For $399 you could hire an expert in fish pathology to check that the fishes are free from diseases and spend the other dollar into buying the real fishes, and put it in a quarantine tank then.
> the facility in Europe had suffered a fire that required them to stop production for 4-6 months.
This sounds very strange, unless you want to artificially create a scarcity in the supply of an animal that anybody can breed by thousands in their living room. You don't need almost any space to breed this fishes, and aquariums can be moved right?. 6 months seem too much.
I wonder why nobody has tried with another of the 100000 species of fish with transparent embryos and bypass this money-grabbing structure.
The issue is the genetic 'cleanliness' of the animals. Some zebrafish will only glow red when you feed them special stuff, some glow green, some only glow green in certain parts of their yolks/chorion. Sometimes this only happens with every other generation of females.
It's very complicated and quite 'messy' when it comes to the phylogeny/genetic trees. Most of the time, researchers only need those 20 fish or so once every few years, if that. Sometimes they need to interbreed these different fish to glow green and red. There are literally tens of thousands of types of fish that have to be kept. As an example: Jackson Labs has mice ready to order to the specific genotype that you want and it gets very specific [0]; zebrafish aren't quite at that level yet, but are getting there.
But the effort is worth it. Zebrafish are a great animal for studying neuro development and many other things. Many of the proposed treatments for ALS, Alzheimer's, Parkinson's, etc. got their start in zebrafish and in studying how those little vertebrates function. The effort will save countless human lives.
Sure, you can hire the fish pathology expert (or send samples out to a lab), but our facility requires us to have "clean" fish coming in, so we can't do the exercise of breeding, treating, and certifying fish in our own facility. The same is true of mice and rats. We have many pathologists on staff who could certify such a thing, but it would require setting up a "dirty" facility completely separate from the "clean" ones (complete with a separate animal care staff to change cages and such). Overall, the price is justifiable to avoid the trouble.
You wouldn't even need a different specie to bypass the money-grabbing, you only need a new facility, your own certified processes, and some clout in the industry (read citations). Hell, you could start with the fish from the expensive vendor and sell the offspring (I think, unless there's a licensing agreement involved).
They are very easy. I had some long time ago and they even lay eggs and breed in my aquarium, but is not my favourite kind of fish for a community aquarium. Having evolved in clean and rapid rivers in India they are fast and nervous. Always darting and ready to engulf any food scrap. Can be a real source of stress for the other tank inhabitants and also for the owner (all you see most of the time is a blur when you try to observe your fishes).
I wonder how they can pick up a male and two females in a larvae pool. The sexual dimorfism on young zebrafishes is close to inexistent.
I had breeded fishes more transparent that those larvae, but must admit that the macrophage video is very cool.
While it is a concern that formulation of cell culture media has been static, as you point out, it's not really the greatest concern, nor is it a new concern.
For a lot of fields, culture medium is not that important as long as you can run controlled experiments reproducibly. The data produced still must be tested for any greater hypothesis than "Drug X causes an upregulation of gene Y in 293 cells." Not including that rigor, or at least that precise caveat, is sloppy science. That happens some, but unfortunately, a lot of science journalism enhances these opportunities for imprecision. That's how you get to headlines like "Drug X prevents cancer!" or later, "Drug X actually CAUSES cancer!"
This is hard to solve. On the one hand, it is important to be effective communicators with the public (who do fund most of this work) and not to sit closed-lipped in an ivory tower, but on the other hand, most people are not prepared to understand the majority of the science that goes on in academic labs, or appreciate the importance or nuance. What are we supposed to do in the end? (I may be slightly off topic here).
Anyway, yes, DMEM, F12K, RPMI, etc are not perfect, no, but it's important not discount the good science that has been achieved with them. With anything, it is important to have good experimental design.
So -- ignorant question here -- why not just use human blood, instead of trying to emulate it? Obviously this has consistency and reproducibility problems of its own, but are they worse than the effects of using a completely alien medium?
Human blood is limited in availability, and would be phenomenally expensive as well as having the reproducibility problems you mentioned. We use FBS (fœtal bovine serum) as a substitute. It's routinely added to culture media such as MEM and DMEM; 10% v/v is typical. This is still pretty expensive though, as well as having ethical concerns (the extraction process is horrific).
It's also still rather unrealistic. Most cells aren't directly exposed to blood plasma and all of its proteins. They are perfused by interstitial fluid which is blood plasma filtered through the capillaries.
One of the goals of some of the more modern media (as in this article) is to supplement every known necessary component of blood plasma to avoid the need for any animal-derived sera, and to make the media completely defined and standardised. This isn't new; people were doing this over a decade back when I was working on my PhD. However, the number of components in blood plasma makes this difficult, if not intractable. Most of these new media had a small amount of FBS (≤1%) to make up for the unknown components. That notwithstanding, the efforts to better understand the exact requirements are laudable. However, proving that all the needed components are present and in the correct concentrations is difficult if not impossible, since it might appear to work while causing some subtle physiological change. We might have to settle for "appears to be good enough for all common cell types". But even that would be a huge improvement upon the status quo.
You joke, but I work in a pharmaceutical company, and we have a program specifically for employees to donate blood for research purposes. It's complicated so that the donors and researchers are blinded to where the blood comes from and who it goes to.
On top of what others said, adding human samples increases the degree of regulations rightfully. Accidentally coming in contact with these reagents means a visit to the lab to confirm you didn't get any blood borne disease
Blood contains immune system elements which would likely reject cancer samples from a different donor. Having to obtain blood and cancer sample from the same donor would complicate things quite a bit.
In addition, it doesn't seem as though they are able to recycle the medium. Frequently draining and refilling with fresh blood seems rather wasteful.
The immune cells are trivial to remove; a few minutes in a centrifuge is all that is required. Even antibodies can be separated out pretty easily (though I haven't done it myself). The problem is the quantity and quality of the product. And blood alone isn't sufficient; it doesn't contain enough glucose and other metabolites to sustain cells for long (since it's continuously circulating and levels are maintained by the body, which isn't true for a culture flask).
The idea that scientists from the Beatson, an illustrious research institute run by a charity, would be hawking some snake oil is absurd. I suggest you tell your cancer researcher friend that they need to have a word with themselves.
I used to be a cancer researcher. "used to" mainly because of people like this cancer researcher you showed, who often have no clue (or have archaic preconceived notions) about a lot of things that affect their experiments such as this. The vast majority of them cannot tell the difference between one and two tailed t tests, and have little regard to the degree to which the artificiality of their experimental conditions screws up their results.
This is precisely the reason "50% of cancer research is irreproducible" is a conservative statement.
In practice, this could be really bad for machine learning when applied to cell images (I work for a company that grows modified human cells and collects the images and applies DNNs to them as classifier). DMEM (or FBS, which we use) would effectively cause a systemic bias that would be hard to correct for.
This makes me mad. Billions are being poured into medical research and I wonder how much of it is being wasted. Several weeks ago I heard that some study concluded that moderate amounts of alcohol and coffee are good for you[0]...Argh.
How much is being wasted? A lot of it. But media is not one of the big problems.
Media are formulated for various reasons. The first reason is to get something to grow in culture at all. This is hard. Most cells from mammals won't survive more than a few hours outside the body. Most bacteria won't grow in liquid culture. So the first thing a medium is formulated for is to get something to grow. Often you fail at this entirely. Leprosy is still cultured in armadillo footpads because we don't have a medium it will grow in. Penny Boston is probably the best in the world at this, and her artistry is basically thinking of what weird thing might put the right bits in medium to get some funny bug to grow. Rusty nails, for example. Getting a culture medium is a big deal.
Next is getting a repeatable culture medium. Old microbiology papers regularly used beef heart broth. That's made by boiling beef hearts and adding a few things to it. The amounts in it are completely uncontrolled. If you try the same experiment two weeks later with stringier beef hearts, and you get different results, is that the broth? Maybe. So we get defined media, media that are repeatable. Weirdly, that may still involve ground up critters. Luria-Bertani broth, which is ubiquitous, includes ground up yeast. We know how to culture yeast into a repeatable ingredient, though.
Then you have specialized broths. Can this organism produce its own ascorbic acid? Try to grow it in a medium without ascorbic acid. Once you get through a bunch of experiments like that, you end up with a minimal medium. Minimal media are incredibly useful for metabolism studies. Let me rephrase: they are essential for metabolism studies.
And then there are selective media, such as 7H9 used for mycobacteria. Tuberculosis grows in it okay. Very little else does. That's really helpful because almost every culturable organism grows faster than tuberculosis, so without a selective medium, you pretty much always get contaminated cultures.
And there are logistical issues. I know that if I grow my culture in a rich medium and then take cells from it and grow them in a minimal medium, I will get different results than if they were passaged through a minimal medium. That's not genetic, it's just physiological adaptation. But in tuberculosis, rich medium takes weeks to grow a culture in. Minimal medium takes months. Better to do a bunch of experiments in the rich medium to minimal medium, see if I find anything, then check if it's an artifact being held over from the rich medium. Cell culture studies aren't as long, but the same argument applies.
Plus every time I start working in a new medium, I have weeks or months of experiments to calibrate it, make sure I know how it affects what I'm studying, and get the data to be able to compare it to my previous media. It's a big investment, so most labs will have a handful that they use.
Meanwhile the organisms are also evolving under the selection pressures you're applying by growing them in these crazy environments. Formulating lab experiments that remain relevant to the world beyond is a difficult, detail driven subject and can go horribly wrong in many fascinating, mind bending ways. So there's a lot of wasted money in research, but it's not because of this stuff. This is table stakes, because biology is hard.
I think the concern is that there are many studies that all contradict each other. I've read about how wine is good for the heart, while others show that any amount of alcohol leads to negative health effects. Maybe both are right because they measured very specific effects.
People get tired of public health recommendations constantly flip-flopping (health effects of butter is another).
The human body is incredibly complex, and most studies look at incredibly specific effects. "increases lifespan by x years" takes decades to measure, but "increases blood cholesterol by x" can be measured much faster. So in the interest of getting papers published in reasonable timeframes most studies take the latter approach.
Then comes media, sees "X causes blood cholesterol to rise, cholesterol is bad, so X is bad" which is a deeply flawed conclusion. Most things aren't universally bad (high blood pressure is bad, but if you raise your blood pressure through exercise that actually lowers blood pressure in the long run), and many negative effects are offset by positive effects not looked at in the study. I wouldn't read too much into anything that isn't a metastudy and took less than a decade to test.
I think the current consensus about red whine is somewhere along the lines of: red whine is good for you, the more you drink the more that effect is offset by negative effects of alcohol, but in the amounts that were always recommended (about 1 glass a day) it's not a huge factor.
But of course those are the harmless examples, there's also interference from industry groups interacting with the current publishing environment (see the whole "fat is bad for you" that we had going for decades for no good reason). So there is a lot of room of improvement.
Studies on alcohol can show better results for moderate users because people who don't drink may have other medical issues or problems related to alcohol.
It's such a big part of culture that people who never drink often have a specific reason not to.
The most common specific reason for someone not to drink is religious. Religions with alcohol restrictions tend to also have other restrictions (both lifestyle and diet) that confound things. Thus it is really hard to know if any difference is from alcohol restriction or from something else.
Another reason people abstain from drink is because they used to drink heavily until it caused noticeable harm. If some of that damage is permanent, it can make moderate alcohol use seem comparatively healthy even if it's also slightly harmful.
The control group is obviously people who do not drink, period -- they don't say drinking moderately is better compared to someone who doesn't drink because they already have alcohol related issues.
No, that's not an obvious or safe assumption at all, and in fact I believe this is from the study that inspired the article linked in the root comment:
> The lack of objective confirmation of actual amount of alcohol consumed, lack of information concerning alcohol intake before baseline, and inclusion of ex-drinkers with never drinkers in the non-drinker category, as well as lack of assessment of drinking patterns place reservations on the observed findings.
I think the moderate is the more sighworthy part because in that context it is a tautology if it is an ammount that causes problems it isn't moderate by definition.
That and those sorts of food studies tend to be bullshit marketing pushes to drive consumption in the first place.
Don't let it make you mad - this is science. And this kind of refining progress is precisely how science should work. This is a good thing. Start with a toehold, and move inward. Cycle back when you realize your toe wasn't in quite the right place. But you couldn't have even gotten there without that toehold. We need controlled model systems to do science, even if the models are imperfect. And that's okay - the (good) scientists know this.
If we had started with 'full' media from the beginning we'd still be describing and making media variants rather than getting any cancer research done. And we've done a hell of a lot of good cancer research since these media were defined. It's a little like the 'Standards' xkcd comic - if you make too many standards, the standards cease to be useful as a standard.
Further, the effect of the media in this case here is likely more profound when doing 'biology' rather than 'biochemistry' experiments. And until recently, the cells themselves were toeholds onto real human biology; HeLa cells are very unusual, but we used them because we could reproduce the work between labs because the cells proliferated indefinitely.
Who cares about a little unnatural media when the genomes of the cells are completely messed up. (And again, the use of HeLa cells taught us a LOT even if they were very not-normal).
Normal human cells just don't last outside of a body - so if one insisted we only use normal human cells there would be zero research, and even less reproducible research because no one could use the same stocks.
It's only and precisely because we do have a (admittedly flawed) standard that such a different and likely more natural media even can be objectively compared against.
Biology has a lot of variables. You can't hold everything constant the way you'd like. So you pick just those that you expect will make the biggest differences. And we should certainly go back and check ourselves every once in a while to ensure we don't get lazy and assume a particular protocol should remain the same in the face of changes in the field. But I don't think that's the problem here. We're just better scientists now - and we're learning that our early assumptions can now be improved on.
The article, especially in its final paragraphs is actually quite even handed. The title, however, is very much hyperbole.
Well said, the shortcomings of EMEM and its variants are pretty well known. The only reason they have not been phased out seems to have more things to do with cost and institutional inertia.
Microbiology used to be in a similar place not very long ago, with people trying (and failing) to grow every species in lysogeny broth, a formulation from the 1950s that is not even remotedly close to optimal for any bacterium including lab strains of E. coli. Improvements happened gradually and by the mid 2000s things have been turned around, with more efficient techniques as well as the ability to study things previously considered impossible to culture in a petri dish. I think a similar shift will eventually come to mammalian cell culture, however we probably won't realise until it has already happened.
Yes, it's clearly bad news, but the world isn't quite so simple. The conclusions you draw depend upon the questions you ask. Suppose for a moment that what you say is completely true, and that there is a direct linear relationship between alcohol consumption and cancer incidence. Drinking more would clearly be detrimental. But would drinking moderately be an acceptable risk? It really depends upon whether alcohol has advantages which make it worth taking that risk. Given that it has its uses in social situations, and might help alleviate stress and help ones general wellbeing, it might well have advantages for promulgating the species and longevity. None of which is to say that being a carcinogen with a lot of other nasty effects is good, just that biology and human behaviour are very complex, and that there may be a host of subtle and not-so-subtle factors which also need to be taken into consideration.
That could be both true and irrelevant when the other health related factors, plus the quality of life (social experience, etc) alcohol in moderation can enable is taken into account. Almost everything is a carcinogen.
So we need to consider Amdahl's law and profile. He die from so many things that we do wrong (e.g. sedentary lifestyle, stress, lack of proper healthcare) that alcohol/coffee could be a very premature optimization.
People in some of the so-called "blue zones", the places where people live most than everybody else statistically, drink a few glasses of wine a day.
I haven't read the article yet, but the title sounds like "by putting salt water fish in fresh water, scientists may have skewed mortality rates". I hope it's not that bad.
However, until you show that these differences matter, nobody cares.