« A Word We Didn't Know We Needed |
Main
| Vibrational Scent Wins A Round? »
February 4, 2013
Single-Cell NMR? How About Single-Protein NMR?
Posted by Derek
Two different research teams have reported a completely different way to run NMR experiments, one that looks like it could take the resolution down to cellular (or even large protein) levels. These two papers in Science have the details (and there's an overall commentary here, and more at Nature News).
This is not, as you've probably guessed, just a matter of shrinking down the probe and its detector coil. Our usual method of running NMR spectra doesn't scale down that far; there are severe signal/noise problems, among other things. This new method uses crystal defects just under the surface of diamond crystals - if a nitrogen atom gets in there instead of a carbon, you're left with a negatively charged center with a very useful spin state. It's capable of extraordinarily sensitive detection of magnetic fields; you have a single-atom magnetometer.
And that's been used to detect NMR signals in volumes of a few cubic nanometers. By comparison, erythrocytes (among the smallest of human cells) have a volume of around 100 cubic micrometers. By contrast, a 50 kD protein has a minimal radius of 2.4 nm, giving it a volume of 58 cubic nanometers at the absolute minimum. This is all being done at room temperature, I might add. If this technique can be made more robust, we are potentially looking at MRI imaging of individual proteins, and surely at a detailed intracellular level, which is a bizarre thought. And there's room for improvement:
By implementing different advanced noise suppression techniques, Mamin et al. and Staudacher et al. have succeeded in using near-surface NVs to detect small volumes of proton spins outside of the diamond crystal. Both authors conclude that their observed signals are consistent with a detection volume on the order of (5 cubic nanometers) or less. This sensitivity is comparable to that of the cryogenic MRFM technique and should be adequate for detecting large individual protein molecules. Both groups also project much smaller detection volumes in the future by using NVs closer to the diamond surface. Staudacher et al. expect to improve sensitivity by using the NV to spin-polarize the nuclei. Mamin et al. project that sensitivity may eventually approach the level of single protons, provided that the NV coherence time can be kept long enough.
I love this sort of thing, and I don't mind admitting it. Imagine detecting a ligand binding event by NMR on an individual protein molecule, or following the distribution of a fluorinated drug candidate inside a single cell. I can't wait to see it in action.
Comments (10)
+ TrackBacks (0) | Category: Analytical Chemistry
POST A COMMENT
- RELATED ENTRIES
- › AstraZeneca's Move To Hot, Happening Cambridge
- › Stealing A Compound, To Set Up in China
- › Tecfidera's Price
- › Novartis Loses the Glivec Patent Fight in India
- › ChemBark Closing Down (Update: Not Really!)
- › Fake Journals - But They'd Like Real Money
- › Chemical Probes Versus Drugs
- › Two New Books
1. luysii on February 4, 2013 11:59 AM writes...
I think you meant 'radius' instead of 'minimal spherical volume'
as
4/3 * pi * 2.4^3 = 58
Permalink to Comment2. Derek Lowe on February 4, 2013 12:38 PM writes...
I did indeed - fixed! Thanks. . .
Permalink to Comment3. ptm on February 4, 2013 1:18 PM writes...
Very exciting, I hope it will live up to it's promise.
Permalink to Comment4. MJ on February 4, 2013 2:46 PM writes...
Neat!
Amusingly, this comes 20 years after Wrachtrup (and coworkers) published a paper on optically detected magnetic resonance of a single molecule.
Of course, it's kind of interesting to see this from a "how science is done" standpoint - you have the more physically oriented researchers pushing the state-of-the-art to the molecular level, while there's been a lot of neat work in the opposite direction for examining proteins in near-native environments and in whole cells. You have to wonder when they'll converge in the end....
Permalink to Comment5. Curt F. on February 4, 2013 3:46 PM writes...
Thanks to Derek for blogging these papers -- they look really neat.
One tiny niche application that could perhaps be enabled by the new NMR is inline HPLC detection. HPLC-NMR has been done before but I think a few orders of improvement in sensitivity might help it tremendously. Imagine being able to get NMR and high-res MS data on peaks in a chromatograph without having to do preparative purification of anything.
Obviously nano-NMR still has a long way to go before that is a reality, but I hope it's an area that developers consider (or have considered).
When will we get a 96-well (or 384 or 1536)-well format NMR reader?
Permalink to Comment6. Anonymous on February 4, 2013 4:18 PM writes...
NMR of the future! But can it core a apple?
Permalink to Comment7. Imaging guy on February 4, 2013 6:23 PM writes...
Cells and tissues usually express thousands of different proteins (basic- 20,000 proteins and if different splice variants and post translational modifications included- widely believed to be about 100,000 proteins) and it is currently impossible to identify the presence and position a specific protein inside the cells or tissue sections without using either antibody based methods such as immunohistochemistry or transgenic methods using fluorescent proteins. Both have many disadvantages. I was always wondering how I could indentify a specific protein inside the cells and tissues without using labels. Recently mass spectrometry imaging (MSI) was suggested as a possible label-free method (PMID: 20397170). Now it seems that this molecular MRI method could also be another label-free alternative in the future.
The article you mention (Size and Shape of Protein Molecules....) describes the calculation of radius and volume of a protein of a given mass based on the assumption that the density of proteins is 1.37 g/cubic cm. However I think that the equation 2.1 in the second page is wrong. Anyone care to check?
Permalink to CommentBy the way, you wrote “By contrast, a 50 kD protein has a minimal radius of 2.4 nM". I should think 2.4 is not nM but nanometer (nm).
8. dave on February 4, 2013 10:21 PM writes...
Equation 2.1 looks OK to me. first term is V-bar (inverse of density), second term converts cm^3 => nm^3. divided by Avogadro's to convert from molar mass (g/mol). Article looks like a great primer for cell/molecular biologists.
Permalink to Commentcheers
9. Imaging Guy on February 5, 2013 4:22 AM writes...
Here are my calculations.
Given - the density of protein= 1.37 g/ cubic cm
1 cubic cm= 1.37 g (protein occupying 1 cubic cm volume will weigh 1.37 g)
(Substitute 1 cubic cm with 1*10^21 cubic nm and multiply 1.37 g with 6.02*10^23 Da)
1* 10^21 cubic nm = 8.24* 10^23 Da
1 cubic nm= ?
1 cubic nm= 824 Da
(the reciprocal) 1 Da= 1.21* 10^ (-3) cubic nm
You can see his values (equation 2.1) are the reverse of what I obtain.
Calculating the volume of a protein of given molecular weight from my equation
1 Da= 1.21* 10^ (-3) cubic nm
50,000 Da (50 kDa) =?
The volume (V) of 50 kDa protein= 50, 000* 1.21* 10^ (-3) = 60.5 cubic nm
[Compare with luysii calculation using the radius of 2.4 (given in the Table 1) = 58]
If you want to calculate the radius
Permalink to CommentVolume (V) = 4/3* pi* r^3
r^3= 3V/4 pi
= 3* 60.5 cubic nm/4* 3.14
= 14.45
r = cube root of 14.45
r= 2.44 nm
10. Morten G on February 5, 2013 8:25 AM writes...
If you want to look at a single molecule inside a cell you could express and purify it labelled and then use laser photoporation to get it into a cell.
http://www.st-andrews.ac.uk/~ds50/photoporation.html (see reference 15)
Permalink to Comment