NMR in Physics, Structural Biology and Medical Diagnosis
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Title of Series | ||
| Number of Parts | 340 | |
| Author | ||
| License | CC Attribution - NonCommercial - NoDerivatives 4.0 International: You are free to use, copy, distribute and transmit the work or content in unchanged form for any legal and non-commercial purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/45017 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
00:00
ChemistryNobeliumPhysical chemistryMagnetismChemistryChemical compoundMolecular geometryLecture/Conference
00:41
MagnetismMoleculeSolutionMoleculeSurface scienceMotion (physics)NanoparticleChemical structureRecreational drug useProteinArzneimittelforschungBallistic traumaBlue cheeseCrystallographyTrauma (medicine)ZunderbeständigkeitBase (chemistry)Wine tasting descriptorsComputer animation
02:30
ButcherPhysical chemistryLecture/Conference
03:34
ButcherPhysical chemistryLecture/Conference
04:00
SolutionNobeliumPhysical chemistryWaterAngiotensin-converting enzymeNanoparticleMoleculeMotion (physics)SoapChemical plantComputer animationLecture/Conference
06:17
NobeliumStokes shiftMoleculeDihydroergotamineSolutionMoleculeOptische AktivitätAzo couplingCell fusionNanoparticleMemory-EffektLecture/ConferenceComputer animation
07:37
MoleculeMuscle relaxantMemory-EffektNanoparticleOptische AktivitätPhysical chemistryOctane ratingMotion (physics)Lecture/Conference
08:11
ÜbergangszustandNuclear Overhauser effectPhysical chemistryMoleculeProteinNobeliumChemical structureMagnetismSoil conservationÜbergangszustandSea levelSample (material)LactitolWaterFiningsWine tasting descriptorsWursthülleElectronMaterials scienceColumbia RecordsDeath by burningSudden infant death syndromeProcess (computing)ButcherComputer animationLecture/Conference
12:15
ÜbergangszustandNucleolusCobaltoxide
12:43
CobaltoxideWaterMagnetismTool steelSeparation processCope rearrangementComputer animationLecture/Conference
13:28
Chemische VerschiebungTool steelGesundheitsstörungEmission spectrumMagnetismCombine harvesterRiver sourceLecture/Conference
14:16
NobeliumPolyethylenterephthalateTool steelPharmaceutical drugAgeingWine tasting descriptorsAusgangsgesteinMachinabilityPsychological traumaRiver sourceChemistryBase (chemistry)AageChemical propertyLecture/ConferenceMeeting/InterviewChemical experiment
15:47
Physical chemistryRiver sourceRadiologyAtom probeProteinBeerCheminformaticsMeeting/InterviewLecture/Conference
16:08
Biomolecular structureSolutionProteinNuclear Overhauser effectChemical structureÜbergangszustandMultiprotein complexAmmonium dihydrogen phosphateChemistrySolutionMacromoleculeHydrogenSystemic therapyRiver deltaEmission spectrumShear strengthChemical structureConcentrateWaterChemische VerschiebungPeptideFunctional groupNobeliumAzo couplingÜbergangszustandSpectroscopySunscreenWine tasting descriptorsSoil conservationAtomic numberTransformation <Genetik>Base (chemistry)Physical chemistryCell membraneLactitolElectronic cigaretteProteinMalerfarbeMuscle relaxantMoleculeCheminformaticsMultiprotein complexAmino acidCytochromeMethylgruppeDipol <1,3->NanoparticleCrystallographyRecreational drug usePotenz <Homöopathie>X-ray crystallographyStickstoffatomIsotopenmarkierungZunderbeständigkeitPolymorphism (biology)GezeitenküsteTranslation <Genetik>Ring strainOperonTool steelDiet foodElektronentransferMedicalizationDyeingHope, ArkansasClick chemistryActivity (UML)Gum arabicISO-Komplex-HeilweiseStockfishComputer animation
25:28
MedicalizationX-ray crystallographyRecreational drug useCrystallographyLecture/ConferenceMeeting/Interview
25:51
Multiprotein complexLife expectancyRecreational drug useWaterWalking
26:43
Multiprotein complexBase (chemistry)Physical chemistryLecture/Conference
27:10
Angular milLecture/ConferenceComputer animation
Transcript: English(auto-generated)
00:15
Well, thank you, Astrid. I think I should say right at the outset that we are now going to talk about down to the
00:26
earth's science. NMR stands for nuclear magnetic resonance, and I'm going to discuss how this physics
00:42
phenomenon is presently used in structural biology and structural chemistry, as well as in medical diagnosis. I spent the first few minutes talking, just trying to give you an impression of what we
01:01
can do with nuclear magnetic resonance today, on the one hand from studying intact bodies, on the other hand studying molecules. Here you have a protein structure in light blue, a drug molecule attached to it.
01:24
You take the drug away, you can study the surface of the protein and try in a rational way to improve existing drugs or to discover new drugs. Now, the special thing about using NMR for such studies is that we work in solution.
01:48
You will see in the following talks this morning that crystallography, or more recently cryo-electron microscopy, makes pictures of such molecules in some ways similar to
02:08
taking photographic pictures. For us, the question arises, how can we make a short picture if those particles are in solution and undergo random stochastic motions on the nanosecond timescale?
02:29
Another application of the NMR principle helps to see bodily harm. Now, you see, this is my right knee in 1989, so about 20 some years ago, and I had to
02:52
have a look at my knee, and I still have to have a look at my knee. You see, Pete DeBatch took a lot of photographs of me once during a training session some
03:05
10 months ago, and my knee still looks okay. Comparing the knee in 1989, and now in February 29 of this year, shows that the
03:22
knee is still okay, so I continue to play football, and you also see the advance made from 1989 to 2016 in the quality of the pictures that we get. Now here, the physics question arises, why do we get contrast?
03:43
Why do we get good contrast in these images? Why can we distinguish between bone and tendons and muscles and water? Actually, it is so that we only see the water, and all the remainders of the image are
04:03
inferred from observing the water. How can we, what is the physics basis for this? Well, this goes a long way back. In 1827, an English botanist, Mr. Brown, made the following observation.
04:26
He carelessly dropped pollen from his plants. He was a botanist, so he had pollen, and he dropped pollen into a beak with water, and the pollen broke, so he had pieces of variable size, and then he looked at this
04:43
mess under the microscope, and in contrast to his expectations, those smaller and larger particles didn't stand still in the water. They moved laterally, and they moved in random ways, and he published this in 1827
05:05
and had no idea as to the background of this observation, and now you will be surprised that this is Robert Brown who solved the problem, Einstein. This is hardly ever mentioned among physicists.
05:21
What happened in 1905 is that Einstein started this year with a publication that analyzed the lateral diffusion as observed under the microscope by Mr. Brown, about
05:41
100 years earlier, and that was also, in many ways, the start of statistical mechanics. I mean, the key is that he understood that the thermal motion, the thermal energy of the solvent, consisting of much smaller molecules than these particles that Mr. Brown
06:04
observed are the reason for the stochastic translational diffusion that Brown observed. Then I think Einstein must have been distracted by other work, much less important for me,
06:20
such as the special relativity theory, which he published in July, and then the magneto-optic effect, which he published in August, and then he sat down again, did serious work, and in December, he published a second paper. It must have taken him a few months to realize that if he had this coupling between
06:45
the thermal agitation of the solvent molecules and the solute, that this must induce not only translational diffusion, but also rotational diffusion, and so he published the second paper on the rotational Brownian motion in December of 1905.
07:04
So that was the number four paper of Einstein's production of that year, and that is the key to understanding what we see in NMR and how we can conduct and continuously improve
07:22
the NMR techniques in solution. And the relation that goes down to very simple principles, we have the Stokes-Einstein relation, which essentially says that there is a correlation time that describes the length during which
07:42
a particle keeps memory of its past, and that this correlation time is proportional to the third power of the radius of the particle, if you represent the particle with an equivalent sphere, and then though the size of the particle determines all the relevant NMR parameters,
08:06
longitudinal transverse relaxation time, and the nuclear overall effect. Now it is so that depending on the relative frequency of the rotational Brownian motion
08:21
and the Larmor frequency used to observe the NMR signal, we either are in the spin physics of conservative transitions or dissipative decisions, and the reason why we can determine NMR structures of proteins is that we are in this regime.
08:45
In this regime it would be literally impossible, or at least very much more difficult, that is the regime of small molecules, I mean water would fall into here. And so we now understand why we see contrast in imaging.
09:04
So water molecules in our body still move rapidly and give a sharp signal, bones, muscles, tendons are far beyond good and bad, they don't move at all, the signals are gone, and that's why we can do magnetic resonance imaging, observing only the water molecule.
09:26
On the other hand, when we want to use NMR in structural biology, we have to make sure that we are in the regime of conservative transitions between the spin levels.
09:41
Okay, now you have actually, that was the most important thing I had to say today. And let me now briefly remind you of what NMR is all about. The simplest case is when you look at the spin one half, you have two eigenstates,
10:02
and in the absence of a magnetic field, the two states of a spin one half are degenerate, cannot be distinguished. Now in 1896, Zeeman observed in optical spectra that there was a fine structure splitting,
10:24
and it's a very small energy, and he actually lost his job, that's important for the students here. Zeeman was a graduate student, and for some reason that I think nobody can really understand, he had to study the optical spectra of some materials, and his thesis advisor had strictly
10:46
prohibited the use of magnetic field for these experiments. He disobeyed, put the sample into a magnetic field, and observed the Zeeman effect. Then he lost his job, and six years later he got the Nobel Prize, then his thesis advisor
11:04
lost her job, and he became director of the institute. A very important message. All right, so Zeeman observed the splitting. Now this splitting is of very low energy, it's in the radio frequency range, and
11:25
there was no way to directly observe transitions between this level until after World War II where hordes of scientists came back from the U.S. Army, having worked on radar developments,
11:41
and they had the techniques available to observe these very low energy transitions between the Zeeman levels of nuclear spins, and in parallel also between the corresponding levels of electron spin. So here you see Peter Zeeman, it's perhaps also interesting to notice that Albert Einstein
12:06
visited Zeeman in his lab sometime around 1920, so you see networking was important a long time ago. Then, as I mentioned, after the war, Felix Bloch and Edward Purcell, who got the physics
12:27
Nobel Prize in 1952, were able to directly detect the transitions between the two energy levels of spin 1.5 nuclei.
12:44
Now if you look at water, you see only one line, let's assume that oxygen has no spin for our purpose, we have a spin 1.5 in two equivalent hydrogen atoms of the water, and that gives a single line, that seems reasonably uninteresting.
13:05
I have worked several years in the 1960s looking at this line, and what actually is today our contrast agents in MRI, we didn't know this at the time, we would just add paramagnetic particles and observe how this affected parameters in the water.
13:26
Others had much smarter idea, just the following, they would apply, in addition to the external magnetic field, they would apply a magnetic field gradient across macroscopic
13:41
objects. So we always see only the water, that's too simple, so the spectrum had to be made more complicated, so you would take about 100 recordings across the head, but always slightly different field, that means when you go across the head, the resonance condition
14:05
changes, in print you get three lines at three different positions of the head, you do this in three dimensions, and then you develop some mathematical tools and you get
14:20
images, now you see it can also be done off the head, not only off the knee. And Paul Outterboer and Peter Mansfield got the medicine prize in 2003 for this achievement, and you can see such a magnet is quite frightening thing to look at, even today there is an
14:46
awful lot of noise when you are inside, it's like being subjected to the noise of a machine gun, I was amazed when I went back into it in February that this is not improved. Now when we worked in this field early on, we worked with newborn infants, and this
15:10
was in the 1980s, we didn't have a whole body magnet, we had a smaller magnet which could be used for babies up to about one year old, but usually we would study
15:22
babies that had between age two days and three weeks, usually with heavy trauma to the head, and we couldn't sedate these kids with chemicals, we had to ask the parents to come in and just hold the baby until it fell asleep and we could push it into
15:43
the magnet, and the parents were extremely frightened about the magnet, and so we built the fairy house, and this is Chris Bush, a physics student who is now head of radiology
16:04
at the University of Berlin. Now let's turn to structural biology. I give you a few years so that you see the chronological development. This picture is from 1969, it shows a particular
16:22
protein cytochrome C, there were no computers which could make a drawing, that was long in the future, so there was an artist Irving Geis who would prepare drawings and paintings of the molecules based on the atomic coordinates that had at that time
16:43
been determined by x-ray crystallography. There is a museum in New York with about 300 of these paintings. Now we have a different situation, before we were looking at the water, we had to make the water spectrum more complicated in three dimensions,
17:03
in each dimension 100 recordings, and that gives a reasonably complex spectrum to deal with. Now here the macromolecule presents us with a complex spectrum, and hundreds of lines overlap, you see it should be like this, but that's what it was.
17:25
So something, some other way had to be done, and that is to introduce artificial time axis, and to go from one dimensional to multi-dimensional NMR, and then you have to sell this, so we had in 1982, in collaboration with Richard Ernst, we had cosinosis
17:46
XE and FOXI. You see, this you can sell. XE stands for spin echo correlated spectroscopy. If you use this word 300 times in the publication, nobody will read it, so, and
18:02
that's how a two dimensional spectrum looks, you see, now you have added artificial time axis, a two dimensional Fourier transformation gets us into two dimensional frequency space, and now the lines that all used to be on a single frequency axis move
18:21
out into the plane, and are reasonably well separated, and Richard Ernst got the chemistry Nobel Prize in 1991. Then we had to do something with this well-resolved spectra, and the problem now is that the particle of which we want to get high resolution
18:48
structure moves at random frequencies, I'm not talking about momentum or similar, it is stochastic movements in the frequency range of 10 to the minus 11 to 10 to the
19:02
minus nine or thereabouts. So, the solution to the problem is to find parameters that determine the structure, but are invariant under rotational and translational motion, and one such quantity is the distance between two atoms. This is a scale, and as long
19:26
as the structure stays intact, it can translate and rotate at any frequency, so the distance will remain the same, and that was the solution to the problem. With the so-called nuclear overhauser effect, we can measure the inverse six power of the distance between atoms
19:47
inside a macromolecule, and then we have the correlation function which we could treat in ways that would not interfere with these distance measurements. And though, with the two-dimensional experiments, we will get plenty of peaks, each peak represents
20:07
the distance between two atoms in the structure, and all it needed was a few years of developing algorithms that could then transfer the high-dimensional distance space back into
20:24
three-dimensional space where we could present the structures. And that was the first structure that we had solved in 1984, and that was then also good enough for a trip to Stockholm
20:40
much, much later, of course. Okay, the final thing which is perhaps also of some interest for physicists is that when we went into the conservative regime of spin physics, we were in good shape to work with
21:03
small molecules such as the one that I had shown before. But then we would, this is an old picture, it shows the size distribution of the proteins for which the structure had been studied. It goes up to about 30 kiloDalton, and that was the end of it. And then,
21:23
now if we, you see, there are a lot of membrane proteins, you'll hear more about this from my colleagues who use crystallography for these studies. But in order to be able to apply the method to membrane proteins, we had to beat that size limit and be able
21:42
to work in the range from 60 kiloDalton. And for this, we would study systems of two coupled spins one half, typically 15N and one H. Or it could also be done with methyl
22:05
groups out on the periphery of amino acid side chains. But see, you need microbiology to introduce nitrogen-15 at high concentration into the polypeptide chain. Now when you have a two spin one half system, you get four transitions between the combined
22:28
eigenstate of which now is a spin system of spin one. And it turns out that only one of the four transitions is under certain circumstances coupled to the Brownian motion,
22:43
and one of the transitions can be uncoupled from the Brownian motion. And that then enables to beat the size limit. Here you see the fine structure of a peak of such a nitrogen-15 hydrogen group, and you can see that one of the four components, and I'm talking about
23:06
these four transitions, one of the four is hardly seen. One gives a sharp line, and two are sort of intermediate. So the point was to select this peak only, throw away 75%
23:21
of the intensity, and gain about a factor of 50 in the end. And that led to Drosie in 1997, transverse relaxation, optimized spectroscopy. I mean, you can just write out the evolution of single transition basis operators. Then you have a few relaxation
23:43
terms which destroy the spectrum when you go to large molecules. And there you have two of these relaxation terms which include the difference between a dipole-dipole coupling and the so-called chemical shift anisotropic coupling. Now it is so that the chemical
24:05
shift anisotropic coupling is very strongly dependent on the magnetic field. Here given as corresponding hydrogen resonant frequency, whereas the dipole coupling is essentially
24:21
invariant, a very good approximation invariant of the strength of the magnetic field. Now this had been recognized long ago in the 1960s, but then the highest available magnetic field was here. So there was no use for this theory. And then by 1997 we had the magnet
24:46
here, and then P minus delta vanishes, and one of the four components in the two-spin system is now uncoupled from the Brownian motion. And we could go as far as studying
25:00
particles of one million molecular weight in solution, such as the complex of GroEL with GroES, and we will get this sort of a spectrum. And I hope that even speaking about terrestrial experiments, I could give you an impression
25:23
of the importance of basic physics research for practical applications. I mean the use of NMR today has major impact in medical diagnosis. Football would be essentially impossible
25:43
today if the MRI were not available for the players after the game. And it plays an important role jointly with X-ray crystallography in providing the basis for rational drug design and improvement of existing drugs. And I'm impressed actually and very happy
26:07
to be in this field which shows that basic research that dates back 200 years to Mr. Brown, who dropped his pollen into a water beaker, led to applications which can now
26:24
day-to-day make improvements to the quality of human life. And I think we should remind our politicians how important it was that Einstein had the liberty of working on the
26:42
Brownian motion long before having an idea that there would ever be an NMR experiment. They had no idea that there would be an NMR experiment, not to speak of Mr. Brown. And I think you might want to take this along and keep enthusiastic about doing basic
27:02
research in physics. Thank you.
Recommendations
Series of 20 media