Cytocentric Visionary: Jeffrey A. Stuart, Brock University

Oxygen: A Big Uncontrolled Variable in Cell Culture

Dr. Jeffrey A. Stuart is an Associate Professor of Biological Sciences at Brock University in St. Catherine’s, Canada. Dr. Alicia Henn interviews him here about his recent publication in BBRC, “Hydrogen peroxide production is affected by oxygen levels in mammalian cell culture. The transcript was edited for length.

Alicia Henn

You are a Cytocentric Visionary because you are working to improve the in vitro environment for cells. What made you look at hydrogen peroxide production in culture?

Jeffrey Stuart

I do mitochondrial work, focusing on aging and lifespan, so reactive oxygen species (ROS) receive a lot of attention. For a long time, it was thought to regulate aging. It certainly regulates mitochondrial processes including apoptosis, and respiration itself is a big source of ROS.

I’ve recognized for quite some time that oxygen is not well controlled in cell culture models. For over a decade, we have regulated oxygen in our incubators.  Typically, we’ll have an atmospheric oxygen incubator and an incubator that is set at 3 -5% oxygen. The missing part was the relationship between oxygen levels in culture and what cells were doing in terms of making reactive oxygen and nitrogen species.

We decided we had better start systematically measuring these things. First, we looked at hydrogen peroxide. We had methods in the lab for measuring it, and it’s got well-characterized roles in redox signaling. We knew that there was a biological story there.

AH

In this paper, you reported more hydrogen peroxide in cells cultured in room air incubators?     

JS

There are three key points in this paper. We looked at the relationship between incubator settings and hydrogen peroxide production by six commonly-used cell lines. In six out of six cases, we got higher rates of hydrogen peroxide production at higher oxygen levels.

It’s important to know that when you move from physiological oxygen to the standard tissue culture oxygen, you are impacting hydrogen peroxide production, and that has consequences. That’s the first point.

The second point is we’ve shown that it’s coming from an NADPH oxidase (apparently NOX1 and/or 4). That makes sense because there are a number of sources of reactive oxygen in cell. The mitochondrial KmO2 are quite low, so probably those are saturated, even at physiological oxygen levels. An enzyme like NOX4 has a KmO2 of 16-18% oxygen. The rate at which NOX4 produces hydrogen peroxide should on this basis be very sensitive to incubator oxygen.

AH

And the third point?

JS

The temptation would be to say; “We will do all of our experiments at 3% to 5% oxygen.” When we did that, we measured the oxygen level immediately outside of the cells.

We found that if you grow cells in a monolayer at one specific oxygen level, like 5%, it doesn’t mean that that’s the amount of oxygen that they see. That depends on how quickly they’re using the oxygen, the height of the medium column and how long the cells have spent undisturbed in the medium. You get big gradients of oxygen from the gas-liquid interface down to the cells.

We showed that with the incubator set to 5% oxygen, the cells can see more like 1%. It means that we’ve shown that oxygen matters, and at the same time shown that it’s hard to actually regulate the amount of oxygen that cells see by just regulating the gas phase oxygen in the incubator.          

AH

Did it change your protocols, when you realized how critical pericellular oxygen was?

JS

We have, for over a decade, been using a low oxygen incubation protocol because our primary lines really don’t like oxygen. What I learned in the course of this particular project was that, that was not getting us where I thought it was getting us.

I didn’t realize that if we have incubator gas at 5% oxygen, the cells would, over time, deplete so much oxygen that they would be seeing more like 1%. I didn’t realize how quickly and how commonly it’s happening, under our experimental conditions.

That means when we think our cells are seeing 5%, they are actually cycling between 1 and 5%, depending on when we’re changing media and how fast the cells are growing.

That’s too big of an uncontrolled variable. That’s too big of a range. So we are now developing some new solutions to prevent those standing gradients.

AH

 We call all those factors that go into deriving the pericellular oxygen levels from the environment “protocol-dependent factors.” By the way, you are my hero for using 18% oxygen. Many, many researchers choose 21% oxygen in closed chambers or cite their room air incubator as being at 21% oxygen.

JS

We didn’t choose it. Most of the incubators we use in the lab have an oxygen sensor. It’s close to 100% humidity in there, and CO2 is added to 5%, so of course you don’t stay at 21%O2. Under standard cell culture conditions incubator O2 is around 17 - 18%. 21% is relatively dry room air with no CO2 and low humidity.           

AH

Why doesn’t everyone monitor the O2 in their incubators? 

JS

It’s probably historical. When cell culture protocols developed, people were trying to simply grow cells. There was very little understanding of redox signaling; they just wanted the damn things to grow.

People used cancer cell lines that, in our experience, don’t care if they’re growing at high or low oxygen. People that now work with primary cell lines, and particularly stem cells, those people have a much better understanding of the importance of oxygen. Those types of cells really care. They’ll senesce, or maybe undergo differentiation.

So if you’re working with a primary cell line and you just want the things to grow without changing, you need that low oxygen. People do recognize that.

To look at redox proteomics, you need to have a good handle on oxygen levels. Otherwise, you’re going to have a lot of inherent variability in your data.

AH

You said that primary cells don’t like room oxygen. Hal Broxmeyer’s group showed in 2015 that isolation of cells in room air oxygen dramatically cut HSC yields and repopulation capacity. How much could this room air selection step during initial isolation affect everything we understand about cells in vitro?    

JS

Yes, you’ve not only got oxygen change, you’ve got temperature and CO2. Those are massive shocks all around. It doesn’t surprise me that that could change the nature of the cell populations. It certainly makes sense.

AH

You pre-equilibrate all of the fluids that you work with?

JS

Yes, it takes a long time. Not everybody realizes how slowly gaseous oxygen equilibrates with medium when it’s not stirred. It takes at least overnight to get there, so we’re always careful about that.

AH

If everyone was taught from the beginning that room air environment is detrimental to cells from the inside of the body, would it be such a big deal to control oxygen?

JS

I think it’s all just inertia. You get all this existing equipment that’s not designed to regulate oxygen. To ask people to change all that is tough. If they started that way, it’s no problem. 

A lot of our experiments end up on confocal microscopes. Even when people do put an effort into the cells’ environment, they’ll sometimes just take the dish and throw it on the microscope stage. They think that if they take their measurements within a half an hour, it’s all good.

The rate at which carbon dioxide leaves the media in atmosphere is just massive, so the cells rapidly become alkaline. The temperature’s changing, oxygen levels are changing and the cells are dying.

AH

Absolutely. You say, in this paper, that standard cell culture conditions are hyperoxic but that the extent to which this influences cellular physiology is probably under-appreciated.

JS

Nitric oxide is another really important small molecule in cellular biology. Two of the three nitric oxide synthase isoforms have Km(O2)s that make them sensitive to oxygen levels in the range of 1-18%, so they may produce more nitric oxide under standard cell culture conditions (hyperoxia) than they would under physioxic conditions.

There’s a real danger that you’re seeing something that would never occur in vivo if you’ve got these higher nitric oxide levels.

If you’re studying redox-regulated events they’re highly likely to be effected by different levels of nitric oxide or hydrogen peroxide under the different oxygen conditions. All of that is what I call cell physiology. Signaling is the way that cell physiology is regulated. That’s the foundation of it.

AH

How important is proper terminology when describing physiologic oxygen?

JS

Physioxia. It’s super important, right, but I think people are fairly entrenched in their idea of 3% oxygen in culture being hypoxic. The terminology is being used inappropriately, but it’s hard to change it once it becomes part of the lexicon.          

AH

“Hypoxia” is used in so many different ways that it’s meaningless.

JS

Well, to be honest, I can’t right now think of a single exception of this. Even when people are talking about hypoxia, and let’s say they think it’s at 3% or at 1% or whatever, they’re always measuring gas phase oxygen. That’s another level of problem.

So in the medium beside the cells, if they’ve been sitting in that environment for 48 hours undisturbed, which is not inconceivable as an experimental protocol by any means, they are certainly way below that 3% or 1% level.    

The whole idea that you can record oxygenation levels by just assuming that cells see what is in the gas phase is clearly not correct. If you think about hypoxia inducible factor signaling, it’s a massive thing in that field alone, isn’t it?

I do think that what have been missing in the literature are those straight forward empirical measurements of hydrogen peroxide, nitric oxide, and redox modifications. Then we can relate them directly to the oxygen levels.

Thank you for your time and your thoughts, Dr. Stuart. We look forward to your upcoming publications on oxygen and nitric oxide production.

If you would like to be featured in our Cytocentric Visionary Series, contact us. We would love to hear about your work.