The inside of a vial is not like the inside of a cell. That’s something that we don’t always think about while running assays, but it’s certainly true. Not many places, truth be told, are quite like the inside of a cell. It’s crowded in there. for starters, with all sorts of proteins, macromolecules, and cellular structures elbowing into each other. One study found that around half the proteins in a given cell are basically immobile (membrane bound, you’d assume for the most part). That one also found, though, that once you get away from the hydration layers around the cell constituents, that the water molecules present have rotational behavior just like bulk water, so it’s not like the whole cell is some sort of congealed mass. Arthur Kornberg’s famous line that “cells are gels” shouldn’t make you think of a jar of preserves – maybe something more like a bubble tea would be a better mental picture.
Two recent papers bring these mental models up. This one is looking at the balance between associative forces (hydrogen bonding, van der Waals forces, and all the rest) and depletion forces. Most of us are going to say “and what?” to that last part, but depletion forces, though lesser-known, are very important in crowded solutions. Imagine an ideal solution as made up of different-sized solute molecules, grapefruits and blueberries, in my own thought experiment (floating around in some medium that can solubilize both of them!)Assume that there are no real fruit-fruit interactive forces, just neutral spheres wandering around. There’s a so-called “exclusion volume” around each of the spheres (and yes, we will also consider grapefruits as perfectly spherical for the purposes of illustration). That exclusion volume is the region around each grapefruit where it’s impossible for the center of a blueberry to be, because they would physically bang into each other. But as things get more crowded (adding more grapefruits to the solution) they can overlap each other’s exclusion volumes (there’s an illustration of this in the Wikipedia link above). This actually increases the total volume available to the blueberries, compared to the possible state where the exclusion volumes aren’t overlapping, and that increases the entropy of the system by giving them more room to move around. So thermodynamically, that exclusion-volume-overlap can happen spontaneously, since it’s energetically favorable.
Proteins aren’t spherical fruit, though, and on that we can all agree. They have intramolecular forces of their own (some attractive and some repulsive), and this complicates the balance with the depletion force. This paper goes into detail about this situation, which is further complicated by the hydrogen-bond structure of liquid water. The new paper mentioned above is studying proteins in models of E. coli cell interiors, and finds that the depletion force can be overcome, especially with smaller macromolecules. And in a way, it has to be, because otherwise it might well be enough to irreversibly aggregate a wide variety of proteins. There are surely been selection pressure on this as well – the authors conclude that “intracellular macromolecular binding constants are finely tuned to exploit depletion forces while avoiding large scale aggregation“. This also suggests that some of the experimental methods used to model the crowding in cell interiors (adding PEG, dextran, or the like) can be inadequate.
The authors of this new paper, though, show that such artificial conditions can still be enough. They’re looking at the SNARE proteins, which are essential to vesicle fusion with the cell membrane. The problem with studying them ex vivo, though, is that they just don’t seem to behave correctly in artificial systems – low activity, for one thing, and some of them have to be cooled to just above freezing before running the assay, which is hardly physiological. This team found that adding artificial crowding agents like BSA and dextran suddenly made things fly right, however – particularly the stimulatory effect of some other proteins whose function had been unclear and unreplicable under standard assay conditions:
Two key conclusions can be drawn from the findings of our crowded fusion assays. First, the vesicle fusion machinery has been evolved to function optimally within the crowded environment of the cell. SNAREs and SM proteins actually take advantage of intracellular macromolecular crowding to efficiently drive vesicle fusion. Second, the SNARE-SM mediated vesicle fusion can be reconstituted and characterized only in the presence of macromolecular crowding agents.
So imperfect crowding agents can be a lot better than none at all. And we should all keep this sort of thing in mind as we set up our useful, optimized, high-throughput, but totally artificial protein target assays, because what’s going on in the cell might be quite different. . .