Subtitles section Play video Print subtitles >> All right, well I think maybe we'll begin even if the last few people come in. So I want to talk about two related techniques today. They are related in pulse sequence. They're completely different in what they do. One of them is TOCSY and one of them is ROESY. So TOCSY is a correlation experiment. It stands for Total Correlation Spectroscopy. I'll put total in quotes because one doesn't get an infinite number of cross peaks. This technique was co-developed, developed at the same time with another technique that's the same pulse sequence called HOHAHA which always sounds good at Christmastime. It stands for Homonuclear Hartmann-Hahn Spectroscopy and they're the same technique but TOCSY has taken over, so the idea is that you get cross peaks with all and again I'm going to put this in quotes because there are limits, other spins in the spin system and so what this technique is like is it's like a super COSY. I mean, I'll give you a really simple example here. If we have propanol and that was sort of the sketch of the molecule when I gave the first [inaudible] in 2D spectra if you have propanol COSY will link your methyl group to your central methylene and it will link the central methylene to the next methylene and assuming the OH is exchanging rapidly that won't be linked in there. What TOCSY will do since all of these groups are in the same spin system is TOCSY will also link the methyl group to the terminal methylene and where TOCSY really, really shines, there are sort of two different situations that TOCSY really shines. In the small molecules realm where TOCSY really shines is in situations of overlap and I've been pretty good about giving you spectra where there's not too much overlap and you can walk your way through COSY but sometimes you'll end up with peaks overlapping on other peaks and you simply stop being able to walk your way through and so you look and you say, hey what's going on here I can't figure out my way all the way through, so I'll just give you a couple of hypothetical examples that maybe illustrate my thinking on this of a case where you might have overlap. So like if you take a molecule of pentanol and we think about where things typically show up at it we'd say all right plain vanilla methyl is.4 and methylene that's next to an oxygen say is at 3.5 parts per million and methylene that's beta to an oxygen that's going to be I don't know that's going to be somewhere around 1.9 parts per million but by the time you get down to this methylene chain-- bless you-- you're going to have your methylenes pretty much unperturbed both at about 1.4 ppm. In other words, if you go into the COSY spectrum you're going to get into a quagmire right here where you have trouble tracing your way through the COSY spectrum and so those types of issues, if you've got multiple spin systems in the molecule and you're really trying to figure out what your fragments are those types of issues can be problematic and so again let me give you a molecule and just sort of talk about typical chemical shifts, so if we'd say, all right an ester of methylene next to an ester, I usually think 4.1 ppm, plane vanilla methylenes I think about, methyls I think about 9 parts per million. A methyl group that's beta to a carbonyl I think about, say about 1.1 parts per million. In other words it's beta to an electron withdrawing group so it's at the plane position shifted down by a couple of tenth of a ppm but then if you think about a methine that's say beta to an oxygen, normally I think about a methine maybe at 1.9 ppm but by being beta to an oxygen maybe it will be about 2.4 ppm and normally I'll think of say a methyl group as next to an ester as being about 2 parts per million but if it's a methylene group it will go down by another about.4 parts per million so you could easily how a molecule like this might have two spin systems where when you try to trace your way through you get caught up and so if everything overlapped at 2.4 it would be really hard in a COSY of a molecule like this to distinguish this spin system from this one, in other words to try to walk your way through and see if this methyl was part of a different spin system than this methyl over here and TOCSY's extremely good at doing this. The other thing that TOCSY is good at, there's a parameter that's very important in TOCSY called the spin lock mixing time and so when I say all other spins in the spin system we're typically talking about within a limit maybe through about 7 bonds. If you vary the spin lock mixing time and you can use TOCSY as sort of a super COSY experiment where if you have a very short time you basically get just one jump just like from here to here but if you go a little longer two jumps starts to appear. If you go a little longer in your spin lock mixing time more and more go. So you can do a series of TOCSY experiments that will basically walk your way from one spin to the next to the next and what's good is if you have a region where there's overlap and then a region where there isn't overlap, so like the overlap might be at 2.4 in this molecule and the region that doesn't have overlap might be at say 1.1 and.9 you can go ahead and walk along those TOCSY tracks and see who the coupling partner is for each and then do longer and longer mixing times to pick them all out. So that's one really good use for TOCSY is overlap. The other really good use is I'll call it biopolymers which sounds like something intimidating but any sort of molecule that has spin systems that are units within, so there are many types of macro lactams and many types of antibiotic, cyclic esters that have unnatural amino acids that have a series of spin systems. I'll just show you some biopolymers for example, peptides and proteins and so if you think about it each amino acid in a peptide or a protein is its own spin system and so forth where each unit comprises a spin system and so you can pick out all of these spin systems and basically very quickly assign all of the resonances in a polypeptide and we'll do an example of this with a cyclic peptide. Sugars are another example. We just saw Professor Peng Wu's seminar and he was working with various types of oligosaccharides and so each oligosaccharide, each monosaccharide unit is an isolated unit and they're often very heavily overlapped and so I'm just giving sort of a generic cartoon of an oligosaccharide structure. And so each sugar unit is its own spin system and TOCSY they are very crowded together. They all end up having similar chemical shifts and TOCSY really shines at working with oligosaccharides. And the other area that works out very well with TOCSY is nucleic acids, DNA and RNA where again the basic unit is sugars and bases and each sugar is its own little spin system. And so you have a nuclear base and depending on if it's DNA or RNA you'll have an OH at these positions so I'll just put this in brackets. So again, all of these are representing kind of pieces of a biopolymer structure that might be useful for elucidating. [ Silence ] So as I was hinting at before, one of the limits of TOCSY is it doesn't go on forever and so the limits are I'm going to say, I always hate to put a hard number, about 7 bonds so for example, what do I mean? I mean let's say we look at the molecule lysine, so lysine is an amino acid with a four carbon chain. I'll put this as part of a biopolymer. So if you're talking about tracing your way from the epsilon carbon to the NH group you're going through one, two, three, four, five, six, seven bonds. That's about as far as you would go and so in other words you'd end up having this hydrogen, if you do it right, crossing with all of the methylenes along the way giving cross peaks as well as if you do right the NH group, unless you do the experiment in D2O in which case the NH group will have exchanged. So as I said the parameter is, the key parameter is a mixing time and typically this is one of the experiments, the experiments downstairs that are like a COSY experiment or an HMQC experiment or an HMBC experiment the parameters John Greaves gives you, if you take this default 10 hertz HMBC experiment that's going to be sort of one size fits all. In the case of a TOCSY experiment you actually have to think intelligently about the experiment. Typical values are about 75 to 100 millisecond spin lock mixing time and you'd obviously, you'd want to go to the high end to pick up longer correlations you might go up to say 200 milliseconds. If you go shorter, particularity if you're down sort of in the 25 to 75 range you'll be using the experiment as kind of a super COSY but one where you can walk your way from one bond to the next to the next. Now one of the implications for your own project is that with strychnine is because strychnine has some really extended spin systems you may not be able to trace your way through all of the spin systems but you'll be able to get part way. The other limitation, so obviously, so one limit is the number of bonds. The other limit is your coupling basically proceeds directly depending on how strongly things are coupled, in other words if you have a very small J that can lead to an absence of cross peaks so it's not so much an issue with a flexible chain but if you come down to strychnine and you're tracing your way through a spin system where one dihedral is close to 90 degrees. Your coupling constant is very small, a hertz or two or zero hertz. If you have a really small coupling constant TOCSY may not take you through. Basically, you need to have some reasonably large couplings, so you may see things behaving like as if they're isolated spin systems or nearly isolated spin systems so, now the nice thing about TOCSY as I said is it's very good at dealing with overlap and I'll show you in just a moment an example where you just would be struggling like crazy by COSY and we're going to assign a zillion different protons in one fell swoop. There's an alternative that's extremely powerful and we'll talk about it in the last week of class and that's the HMQC TOCSY. So TOCSY works as long as you can find some regions where there aren't overlap and you can get one resonance that isn't overlapping but if you've got really bad overlap you may even have trouble tracing your way through a TOCSY. HMQC TOCSY is a variant that's like TOCSY but it has the dispersion of the C 13 dimension. Remember C 13 resonance is because you have 200 ppm end up having very little overlap and so the dispersion can be very, very powerful. That's too much for us to assimilate at this point so let me just say I'll show you that in the future. All right, what I'd like to do now is to talk about, give us an example of one molecule. We're going to assign every resonance in this molecule and the molecule is gramicidin S. It's an antibiotic and it's a non-ribosomal peptide. What that means is that it's not synthesized by the traditional T RNA or DNA or messenger RNA, T RNA mechanism and its structure consists of five amino acids that are repeated trice and so I'm going to draw the structure of the molecule and I'll draw it kind of in a stylized fashion because I think that's actually useful for reflecting the confirmation of the molecule. So the molecule starts with a proline and we next continue with a valine and we next, it's a non-ribosomal polypeptide so we next have an, call it unnatural amino acid but it would be better to say non-proteinogenic or non-ribosomal amino acid and so the next one I'll label this as valine. The next one is ornithine. Ornithine is like lysine except instead of having a four carbon chain it has a three carbon chain, so one, two, three. The next amino acid is in the molecule is leucine and the final amino acid before we repeat ourselves is the unnatural enantiomer of phenylalanine so that's D-phenylalanine and so that's half the molecule and then the molecule repeats itself.