Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Water in Biology

  1. Aug 30, 2008 #1
    I found this site the other day on the topic of water in biology by Nature writer Philip Ball (who's also written a variety of non-fiction, including a well-regarded book about water a couple of years ago) where he highlights recent papers on the issue of water in biological systems about twice a month on the average:

    Water in Biology

    It's made for fascinating reading/skimming thus far, and - as a bonus - Ball's review on water in biological systems is freely available here (at least for the time being). The citation is:

    Philip J. Ball (2008) "Water as an Active Constituent in Cell Biology." Chemical Reviews. 108(1): 74 -108. DOI: 10.1021/cr068037a

    Still working through it myself, but it's a good reminder that even something as ubiquitous as water is still an object of widespread study since we don't know nearly as much as we'd like to think we know.

    It reminded me of a discussion I once had with a former coworker (he was moving onto a crystallography lab, I had since moved into a NMR lab) over dinner one night. He asked about whether or not working with solids was really such a great idea (I think he was having an introspective moment about spending the next three years of his life having to grow crystals every five seconds, or at least look like he was doing so). We didn't really come to any sort of definitive conclusion, but given that a number of protein crystal structures show quite a bit of hydration, it's not as if he'd be working with completely dry samples. I of course pointed out that there are such things as solid state NMR, which can take place in a hydrated, non-ordered "solid" (where the notion of solid is more based upon rotational correlation times than in some idea of ordered packing as one might see in a crystal of some sort) and that MRI - which of course involves in vivo studies of nuclei in living cells - has people working with quadrupolar nuclei and getting results. That they're able to get anything out of it shows us that cells aren't some sort of isotropic aqueous solution, as the quadrupolar interaction in liquids are typically reduced to zero, but are still extant in anisotropic media.

    Anyway, I thought it might be of interest and any comments/feedback are welcomed.

    P.S. - I figured since it was "Water in Biology" I'd put it here instead of the Chemistry forum, but if the moderators see fit to move it, I can't really argue. A lot of the literature in this area does seem to come from chemistry departments..... :wink:
  2. jcsd
  3. Aug 31, 2008 #2


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I haven't followed the link to the article yet, but pondering the value of using crystal structures to understand protein structure/function in biological systems is a valid concern. Of course, it's the best information we can get at this time, so is better than nothing, no question there. But, biologists certainly recognize that this is not the state these proteins are in within the aqueous environment of the body, and some of the conformation and molecular interactions are likely different in the functioning biological system than in the crystallized structure. There are physicists aware of these limitations too, working toward other ways to study proteins in solution, but that's all technology in development. And, yes, you bring up the other limitation, that getting proteins to crystallize in a pure form at all is difficult, and can only be successfully done with very few of them.

    So, yes, this challenge of better understanding structure/function relationships in an aqueous environment of cells in a living organism is a hurdle yet to be overcome. So, while crystallography gives us more information than we would have without it, perhaps the apt metaphor is that we should take it with a bit of a grain of salt before hanging firm conclusions on those findings.
  4. Aug 31, 2008 #3


    User Avatar
    Science Advisor
    2015 Award

    One thing to keep in mind about protein crystals is that they are usually >50% water (proteins don't pack very efficiently into crystals. That's one reason why it's so damn hard to crystallize them!). In fact, one of the ways that crystallographers use to determine whether their crystals are protein (yay) or salt (boo), is by poking the crystals to see whether they get crushed easily or not.

    A few lines of evidence argue that crystal structures do represent the biologically relevant structures of proteins. First, for those proteins that have been crystallized in multiple different crystal forms with different crystal packing, the proteins tend to adopt the same or very similar structures. Of course there are examples of crystal packing introducing artifacts into the structure, but for the most part, the structures of crystallized proteins seem to be dominated by the thermodynamics of folding to their native structure, not the thermodynamics of crystal packing. Good crystallographers, though, will try to validate features of the crystal structure that occur near crystal packing interfaces. However, crystal packing does become a significant problem when examining multiprotein complexes and protein-protein interactions. Since proteins in crystal are probably at close to 1M concentration, crystal structures will capture even the weakest protein-protein interactions. Thus, if one sees a particular protein-protein interaction in a crystal structure, one must always be skeptical and perform additional experiments to verify that the interaction is real.

    Luckily for us, crystal structures very often lead to testable hypotheses that we can use to validate the crystal structures. For example, if the crystallographer observes a salt bridge between two residues in a complex, she can perform binding assays to determine whether this salt bridge occurs in reality. Let's say that the salt bridge occurs between a glutamate (negatively charged) and a lysine (positively charged). Our researcher would expect to see a loss of binding when the glutamate is mutated to a lysine or when the lysine is mutated to a glutamate. However, we should also see a rescue of the binding activity if we flip the identities of the amino acids (e.g. in a mutant containing both the glu->lys and lys->glu substitutions).

    My biggest issue with crystallography is that crystal structures give very little information about the dynamics of the molecules. As is being increasingly appreciated in protein biochemistry, the thermally-driven motions of a protein structure are important for understanding their functions. Crystallography frames structural biology in the wrong way. When we seek to solve the structure of a protein, we should seek the native ensemble of structures, not the structure of the protein. NMR is currently the best technique for examining these motions.

    NMR still does not represent proteins in their native, cellular environments. Cells are not dilute aqueous solutions, but rather they are very crowded environments. This crowding in the cell has some interesting implications (some of which are just beginning to be understood), for example, it has been recently shown that the degree of crowding can affect the shape of a protein (open-access article). Here's the citation:

    Homouz D, at al. (2008) Crowded, cell-like environment induces shape changes in aspherical protein. Proc. Natl. Acad. Sci. U.S.A. 105(33): 11754-11759.
  5. Sep 1, 2008 #4
    Thanks to Moonbear and Ygggdrasil for chiming in on this thread!

    Lysozyme, as I recall, has been the subject of a number of studies so as to test the effect/role of the hydration shells, with some of the structures having 20 to 25 percent solvent content. These are the lowest numbers that I can remember seeing for a protein crystal structure, having been grown under controlled humidity conditions if memory serves. It always struck me as pretty remarkable - especially given that I've heard numerous inorganic chemistry talks where they talk about their small-molecule structures - that even when attempting to produce a relatively "dehydrated" crystal structure, one would still be working with about one-quarter solvent in the end.

    I came across this paper while in grad school -

    Bertil Halle (2005) "Biomolecular cryocrystallography: Structural changes during flash-cooling." Proc. Nat. Acad. Sci. USA. 101(14): 4793-4798

    - which proposes that while cryocrystallography tends to get the backbone fold correct, its ability to accurately capture ambient(lab)-temperature/physiological conditions with respect to solvent, side-chain conformations, and ligand binding can be seriously questioned. (Disclaimer - in my dissertation research, I had a low-temperature crystal structure as a reference for my own investigations which had the ligand positioned too far from the catalytic cofactor for chemistry to occur at any reasonable timescale.) While not a crystallographer myself, I would like to think that we're going to see a renewed interest of ambient(lab)-temperature crystallography one of these days, at least as an adjunct to the low-temperature studies.

    I have a very simple personal criterion for assessing most solution NMR studies that come out in the literature. If they use different conditions for, first, the sample characterization/assay after purification and, secondly, the collection of NMR spectra, I sigh and shed a small tear. Usually if the protein is otherwise well-characterized in the literature (especially if there's either in vivo data or some sort of cell-free extract data), you have a guideline to find sample conditions where the protein is still functional. Unfortunately, there is the strain of thought that getting narrower linewidths is worth any sort of biochemical abomination. Assessing solid state NMR studies is a bit more difficult, but that's perhaps a rant saved for another day.

    It may be interesting to bring up this paper -

    Philipp Selenko, Zach Serber, Bedrick Gadea, Joan Ruderman, and Gerhard Wagner. (2006) "Quantitative NMR analysis of the protein GB1 domain in Xenopus laevis egg extracts and intact oocytes." Proc. Nat. Acad. Sci. USA. 103(32): 11904–11909.

    which serves as a simple but promising proof of principle for collecting NMR data from labeled proteins in intact cells. There are the expected effects from cellular viscosity (slower rotational correlation times) and in the relaxation behavior of amide groups, as well as also concentration effects. Mention was also made of studies done with an "artificially-frustrated" in vitro system - the GB1 protein in a solution containing high concentrations of BSA - which demonstrated comparable effects from viscosity. The overlay between the 1H-15N correlation spectra of the purified protein and inside the cell showed that the chemical shifts were quite comparable for the most part, although some peaks did split due to the move to an intracellular environment.

    Mostly, though, I'm just glad others thought enough of the topic to contribute, I was afraid I might be the only one.....
  6. Sep 2, 2008 #5

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    Although these posts tend to focus on crystallography/NMR, I should point out there are lots of people doing single-molecule studies. Carlos Bustamante (and coworkers) in particular, is doing a lot of really cool studies involving protein folding and the energy landscape.

    My understanding is that crystallized proteins, as Ygggdrasil points out, lose all dynamic information- in fact, I've seen structures with grey-ed out blobby areas that don't crystallize well (read: don't sit still), and unsurprisingly, those regions seem to correlate with ligand-specific binding sites. Even so, rigid structures like alpha helices, beta sheets, etc. can be 'imaged' and useful biological information extracted.

    As for the biological activity of water, per se, there's a lot of pseudo-science out there to watch out for. I won't name individuals becasue they can't defend themselves here, but there's no real evidence that water has some sort of functional role, other than as a solvent. To be sure, there are intriguing facts- the enormous electric field gradient at a membrane (nearly 1 MV/m) surely has effects on the water's dipole moment.
  7. Sep 3, 2008 #6


    User Avatar
    Gold Member

  8. Sep 3, 2008 #7


    User Avatar
    Gold Member

    See the chapter: Water p 43-65
  9. Sep 6, 2008 #8
    Andy Resnick - Indeed, there is a lot of really interesting research outside of the crystallography & NMR areas in single molecule biophysics. Looking at mechanical folding/unfolding of proteins is really fascinating, and being able to examine kinetics/enzymology at the single-molecule level is quite impressive. Outside of the single molecule people, there's also the soft condensed matter approach - here I'm thinking firstly of David Weitz and his lab looking at cell mechanics and rheology. There's also a lot of interest in these topics from the theoretical/computational side, not surprisingly. Seems like most anyone who does even the tiniest bit of interest in stat mech has a paper or two in this area.....

    I agree there is a lot of pseudoscience being tossed around out there with respect to the physics and chemistry of water, especially with regard to the marketing of various products of highly dubious value. However, I think Ball kept to the reasonably well established roles of water in biology, and I'd say that that there definitely are places where it serves as more than just a convenient medium/solvent (critical water molecules involved in the catalytic mechanisms of various enzymes, its likely role as "electrostatic shielding" in DNA-protein interactions, potential "water wires" observed in a couple of proteins, regulating ligand binding, and such). It's never presented as something completely fanciful, at least in my skimming of the article.

    somasimple - thanks for the article, I just gave it a brief run-through and it certainly seems like an interesting-enough paper. It makes perfect sense that focusing on water diffusion (as in diffusion fMRI) instead of relying upon either radiotagged water (as in PET imaging) or perturbed spin relaxation properties (as in BOLD fMRI) opens up new avenues in magnetic resonance imaging, especially as water is the major component of not only the brain but also the body in general. I'm not really sure why you quoted that part of the abstract that you did, though - the paper makes the case that by exploiting these differences one might be able to extract more information of potential diagnostic value. The paper notes that the spatial resolution in MRI is at the millimeter level, MRI really isn't something one uses to directly probe the behavior of molecules at the molecular scale. Any conclusions it makes about the behavior of water is, by its resolution limits, going to be something statistical in nature. You might find the review by Ball to be to your liking.
  10. Sep 8, 2008 #9


    User Avatar
    Gold Member


    I brought this paper because it makes a good review about water in biologic processes.
    There is many references that are in free full text and are a must read.
  11. Sep 8, 2008 #10
    I went back and saw that there is, in fact, a brief discussion about water in biology in and of itself, and not just in the context of either MRI or neurobiology. I must have glossed over it earlier.

    It does make me wonder about their observations of cell volume changes and see if there are parallels in areas in other areas of biology. Purely for example's sake (I am no plant biologist by any stretch of the imagination), it's known that plants and other organisms have cell walls that prevent them from bursting when there's too much water inside the cell. But perhaps there's enough flexibility in the cell wall to permit necessary processes to occur that depend on cellular volume changes. Something to think about, perhaps.
  12. Sep 9, 2008 #11


    User Avatar
    Gold Member

  13. Sep 9, 2008 #12
    I just thought that given my throwaway comment about plants was worth a quick Google search:

    Life under pressure: hydrostatic pressure in cell growth and function.

    And also apparently entire books are written about the physiology of cell volume regulation:

    Cellular and molecular physiology of cell volume regulation.

    It might be interesting to see what differences exist between organisms (for instance, mammalian versus plant) and tissue types (neurons as mentioned above, versus, say, renal cells as are mentioned in that Google Books find). It makes one think of all sorts of questions about the role of the cytoskeleton, the presence or absence of a cell wall, different regulatory mechanisms, and so on. Something to keep in mind, perhaps.
  14. Sep 9, 2008 #13


    User Avatar
    Gold Member

  15. Sep 10, 2008 #14

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    One large class of research questions in biology is "how is 'X' regulated?" For example- the cytoskeleton is highly dynamic, and changes in response to environmental cues. Same for nearly everything in a cell- what membrane proteins are in the membrane, where proteins are, what genes are being expressed,

    One principle in biology is 'homeostasis'- something is held constant over time. In my research, it's total body water and salt, and regulated by kidney function (which I study). Cell volume (via Na/K/Cl cotransporter) in the macula densa is used as a input (among several) to regulate the glomerular filtration pressure- how much fluid enters the kidney- and additional regulatory pathways monitor how much fluid is being resorbed back into the blood and how much is excreted as urine.
  16. Sep 25, 2008 #15
    I remember hearing somewhere that "heavy water" (D2O), containing the deuterium isotope of hydrogen, is poisonous to most life forms, despite being chemically identical to ordinary water. Can anyone here tell me if this is true, and if so why? If it is, it's certainly a suprise to me...
  17. Sep 26, 2008 #16
    Look up deuterium isotope effects - messes up hydrogen bond networks, can slow down the rate of enzymatic reactions, and so forth. The threshold at which effects are seen can vary, if at all - for example, bacteria can grow in 100% D2O, but it's just much slower than in regular growth media. Generally 25% will permit toxic effects to be seen, and at higher concentrations the organism will die. The main culprit - as i remember coming across in a class years ago - is that cell division in eukaryotes can't process due to the spindle can't form properly.

    However, any naturally occurring D2O is nothing to worry about, given the expected isotope distributions in nature.
  18. Sep 28, 2008 #17
    Something like one part in a thousand if memory serves... thanks for that!
  19. Sep 28, 2008 #18
    You're welcome.

    There have been some proposals for the usage of heavy water in certain therapeutic applications, although I'm not sure how far any of these ideas have progressed. I imagine that since we excrete excess fluids it would be a somewhat expensive treatment method, not to mention inherently inefficient to some degree since you'd constantly be flushing the heavy water away (literally).

    As an addendum to the earlier post, bacterial growth in high concentrations of deuterium oxide usually requires that the bacteria first be grown in less concentrated media first. The exact protocols will vary, depending on species and what kind of research you're trying to do.
  20. Sep 29, 2008 #19


    User Avatar
    Gold Member

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Water in Biology
  1. Biological Apatite (Replies: 1)

  2. Understanding Biology (Replies: 19)

  3. Biology calculations (Replies: 20)