Skip to main content

Biological News

Cells Like Turbulence. Well, Some Cells.

Let’s talk cell culture – specifically, how weird it is. There are an awful lot of ways to grow cells, naturally – different media of course, different scales, differences in how crowded you let them get, how often you split them, add nutrients, wash them, all those things. All of those make sense to me, but there’s another big variable that has always struck me as a bit freaky: the shape and size of the vessel you grow them in.

Experienced protein-production types will find nothing odd about that at all. They’re used to switching between roller bottles, shaker bags, T-flasks, all sorts of sizes, shapes, and agitation methods to see if these have an effect on protein production. They don’t always, but sometimes it can be dramatic. I recall a project where the yield of isolated protein went up from (qualitatively speaking) “useless” to “pretty darn good” just by switching the cells into shaker bags. Here’s a brief guide from Corning that goes into some of the options, which does not neglect their own product line, naturally. A big difference comes into play right at the start: do your cells want to adhere to some surface, or would they rather float around freely in solution? (Here’s ThermoFisher on that subject; they are also happy to sell you labware).

Agitation, in all its various forms, is particularly important for the suspension cultures, especially as the volume goes up. You run into exactly the same volume/surface area considerations and mixing problems as you do in scale-up chemistry. Heating and cooling, if they’re being done from the walls of the container, become increasingly distant from the bulk of the solution. And in the case of cell culture, small volumes can take in all the oxygen they need from the surface, but that gets harder and harder as well as the darn volume goes on increasing as the cubic function.

All that makes sense. But even once you’ve narrowed down to (say) suspension cell culture in a defined medium, mechanical agitation can still be a big factor. Here’s a particularly dramatic example: platelet production. It would be very useful to produce blood platelets in vitro from megakaryocytes instead of isolating them from donor samples, but that’s been difficult to realize on scale. The authors (a large team from Japan who clearly have put in a great deal of effort) have discovered that a key feature of natural in vivo platelet production involves turbulent flow.

The idea that blood-flow shear stress was important in platelet production had already been proposed, but adapting this to cell culture had (so far) not produced the desired results. A closer look showed that the megakaryocytes that were exposed to laminar flow did not release platelets, while the ones that experienced turbulent flow did. Enter the Reynolds number and the Navier-Stokes equations, once again. Those things are seriously important, and are a serious headache for physicists, engineers, and mathematicians, since turbulence itself is so poorly understood. The Clay Institute has a million dollars waiting for you if you can tell them some of the most fundamental things about the Navier-Stokes equations, for example, and it just goes on from there.

At any rate, megakaryocytes, it turns out, apparently care very much about turbulence themselves. This new paper goes on to describe the design and testing of a new cell culture reaction that deliberately exposes the cells to turbulence as well as to shear, and it does the job: production significantly increases of what seem (in animal models) to be functional platelets. The cells start releasing much larger amounts of crucial thrombopoietic mediators (such as IGFBP-2, MIF, and others) under these conditions, which seems to be a key part of the whole process. How exactly they’re sensing the turbulent flow and what the coupling mechanisms are for the enhanced protein production are questions that still seem to be open. You’d have to assume some sort of cell-surface mechanoreceptor system, I’d think, but we’ll see. This opens up some new frontiers for even more detailed cell culture conditions, though, and you can be sure that a lot of biotech folks are taking notice. . .


13 comments on “Cells Like Turbulence. Well, Some Cells.”

  1. Old Timer says:

    I don’t think you linked to the paper (from Japan?).

    1. Derek Lowe says:

      I hate it when I do that! Just fixed, thanks.

  2. Barry says:

    Curious that it’s platelet production that is so sensitive. Turbulent flow has a role in both angiogenesis and clotting. And clotting brings in platelets.

    1. Locketopus says:

      Well, in a normal vein or artery (where the flow is as close to laminar as evolution can make it) excessive platelets are bad, but in the case of a ragged wound (which would produce turbulent flow), platelets are good. Same with angiogenesis.

      Ragged wound -> turbulent flow -> must stop uncontrolled bleeding, then generate new vessels in the damaged tissue.

  3. Nesprin says:

    There’s several mechanosenor systems known in cells- this field is growing fast! A quick overview: several integrins and a number of the Trp channels are known to be mechanosensitive. Actin/Lats/Yap/Taz form a separate mechanosensitive system. And lastly the nuclear envelope and chromatin architecture are able to sense mechanical perturbations.

    1. lazybratsche says:

      Any hints on how these different mechanosensory pathways are combined to distinguish between laminar and turbulent flow? It seems to me that it must require some sort of derivation across time or space, to measure changes in local sheer stress over the scale of a cell.

      1. Nesprin says:

        Sure- I can think of a couple of different mechanisms off the top of my head.

        This work uses two different flow patterns to generate laminar vs. turbulent flow with no change in buffer viscosity. So this translates into faster overall flow in the turbulent condition compared to laminar ones. This could alter cell biology either by either activating different mechanoreceptors at different levels.

        Alternatively, turbulent flow is by definition unsteady- cells sitting under a turbulent flow would typically experience variable shear stresses instead of constant shears in laminar flows. Kinetics of application of forces is about as difficult to handle as kinetics of small molecule signaling, but there is good evidence that the rate at which forces are applied can matter as much as the total force.

        More simply turbulent flows have a lower ratio of flowrate to pressure drop than laminar flows, which suggests that cells experiencing turbulent vs. laminar flows could experience altered nutrient delivery and removal of waste. There’s decent evidence that slow flow rates are associated with formation of autocrine signaling gradients- since secreted factors would have a higher concentration at the end of the cell farthest from the flow input.

        Or alternatively, variable shears across a cell body could result in polarization of the cell body and all the biology inherent in that…

        1. Chemist with a litle engineering background says:

          “More simply turbulent flows have a lower ratio of flowrate to pressure drop than laminar flows, which suggests that cells experiencing turbulent vs. laminar flows could experience altered nutrient delivery and removal of waste. There’s decent evidence that slow flow rates are associated with formation of autocrine signaling gradients- since secreted factors would have a higher concentration at the end of the cell farthest from the flow input. ”

          Translation: turbulent flow is more efficient at removing concentration gradients at the cell surface that can inhibit signaling, etc.

          Is this correct? Because that is the (perhaps overly) simplistic way that I think about this effect. Similar to the difference between laminar flow and plug flow in a tubular reactor when considering wall effects and residence time distributions.

  4. Chris Phoenix says:

    Big whorls have little whorls
    That feed on their velocity;
    And little whorls have lesser whorls
    And so on to viscosity.
    – Lewis F. Richardson

    1. John Wayne says:

      This poem is just stunning.

  5. One of the best things I have ever read regarding viscosity and turbulence is Ed Purcell’s (of NMR fame) “Life at Low Reynolds Numbers” (linked in handle). It tells you how unintuitively weird things get at small (but still macroscopic) level.

    1. Scott says:

      Heck, the Wright Brothers ran afoul of Reynolds Numbers trying to scale bird airfoils up to something large enough to support a human. Bird wings and the Wright Flyer wings are thin, whole assembly is shaped more like a flattened C. But when you scale up to airplane wings, you need a much thicker center, proportionally, than bird wings have.

      And apparently the only way insect wings work is because the air acts more like an oil at that scale (someone scaled up fly wings and powered them in oil, generated lift/thrust). Hey, aviation is a fascination of mine. Everyone needs at least one fascination in life.

  6. 10 Fingers says:

    Thanks to Chris for the reference to Lewis F. Richardson and for his poem. Quite the interesting man, he was.

    And, I was really struck this morning, catching up on this month’s posts, by the remarkable conversation and community that manifests here. Thanks to Derek for nucleating something rare with the diversity of his observations and insights, and for protecting and maintaining the integrity and sanctity of this intellectual Commons.

Comments are closed.