00:00:01 Well, all right.

00:00:03 Now, at this point, I've said enough.

00:00:09 We'll let Bob Woodward give us a lecture, which incidentally has the same title.

00:00:15 This has the same title as the paper he gave here in 1961.

00:00:20 But the chemistry is not the same.

00:00:22 The number of advances is huge.

00:00:24 It's a delightful experience,

00:00:27 and it's a wonderful opportunity to personally welcome Bob Woodward, my old mentor.

00:00:31 Bob.

00:00:32 Thank you.

00:00:33 Thank you.

00:00:42 Yes, it's okay.

00:01:10 Well, Dr. Trost and Dr. Zimmerman, ladies and gentlemen, can you hear me in the back?

00:01:18 You cannot.

00:01:19 All right, we'll have to see what we can do.

00:01:37 All right, can you hear me up there now?

00:01:39 You can.

00:01:41 Now, I tend to get a little absent-minded sometimes when I'm lecturing,

00:01:45 and so if my voice should fall so that you can't hear me, please shout.

00:01:49 Just say louder or something like that.

00:01:52 That's the only courteous thing to do,

00:01:55 because it obviously doesn't make any sense for me to be here speaking if you can't hear me.

00:02:04 Well, it's a very, very special pleasure for me to be back here at Wisconsin after much too long a time.

00:02:13 I'm sorry that I haven't been here, had a chance to visit this very active department in almost 20 years.

00:02:22 It's a very special pleasure also to be here on this occasion

00:02:26 when the laboratories are being dedicated to Professor McElvain,

00:02:32 because I had for him a great deal of respect and admiration.

00:02:37 I admired his work, and beyond that I can add the testimony of someone

00:02:42 who did not have the privilege of working directly with him at any time,

00:02:47 that he was, in the relations which I did have with him from time to time,

00:02:52 invariably a warm, kind, and best of all, most of all, highly encouraging in every way.

00:03:01 Really a wonderful man.

00:03:03 He did a great deal for Wisconsin, of course, as you heard and know.

00:03:08 He did a great deal for chemistry, and he was, as I said, most encouraging to other young chemists,

00:03:17 those with whom he didn't have close, direct relations,

00:03:21 but rather whom he knew just as colleagues as chemists in the world outside Wisconsin.

00:03:33 Now when Professor Trost introduced Professor Zimmerman and said we were about to have

00:03:40 the longest session or the longest part of the proceedings,

00:03:43 I didn't quite know what he was referring to.

00:03:48 And as Professor Zimmerman arose and then spent some time up here, I began to think I knew.

00:03:56 Well, by and large, considering the opportunities he had, he was quite kind,

00:04:11 and I thank you very much, Howie.

00:04:18 Now this afternoon I'm going to take the opportunity given me

00:04:25 to describe some of our work in the area of the synthesis of macrolides.

00:04:34 Now macrolides comprise a very large class of naturally occurring substances,

00:04:42 each member of which is characterized by certain structural features,

00:04:49 one of which is invariably a large lactone ring, hence the name of macrolides.

00:04:59 These lactone rings contain many carbon atoms in the rings,

00:05:05 and many of those carbon atoms in the naturally occurring macrolides are chiral sites,

00:05:13 that is, centers of asymmetry.

00:05:19 This situation presents to the synthetic chemist a peculiar challenge.

00:05:27 That is, the challenge is that of devising methods for creating asymmetric systems

00:05:41 in situations in which one does not have the kind of control

00:05:49 which is well known and well worked out in rigid systems,

00:05:54 such as, for example, in cyclohexane derivatives,

00:05:57 or more particularly in fused polycyclic systems,

00:06:01 which have a considerable degree of rigidity,

00:06:05 and within which systems the modes of creation of asymmetry

00:06:11 and the use of asymmetry which may be present to direct the creation of new asymmetry,

00:06:17 all those things have been very well worked out in these cyclic systems.

00:06:24 But now, to exemplify a little further what I mean about the situation in the case of the macrolides,

00:06:30 let me say that I'm going to devote my time mainly to work directed towards the synthesis of a particular macrolide,

00:06:38 namely erythromycin.

00:06:41 Our work has been carried out on a very broad front, particularly in its earlier stages,

00:06:48 and so I shall have to select from the work that we have done a fairly narrow portion,

00:06:56 and that portion will be mainly concerned with a path in the direction of erythromycin.

00:07:06 And if I may have the first slide.

00:07:11 And now, there we are.

00:07:14 Here I have erythromycin.

00:07:17 The slide is devoted mainly to showing what erythromycin is,

00:07:21 and there's one representation of it over here on the left,

00:07:25 and you will see the large lactone ring here.

00:07:31 It is, in this case, a 14-membered lactone ring,

00:07:35 and the stars, I'm not sure you can see that they're stars,

00:07:38 but those little things which appear are various of the atoms in the periphery.

00:07:44 Each such starred atom is an asymmetric atom.

00:07:49 That is, there would be a possibility that the groups attached at each of those points might be inverted,

00:07:56 and of course, were that to be the case at any one of the centers,

00:07:59 the molecule would be something else and not erythromycin.

00:08:03 Why did I say there was a peculiar stereochemical problem?

00:08:06 Well, this is a large lactone ring,

00:08:11 and we might be tempted to say that the elements of rigidity present in smaller rings

00:08:19 are not going to be present in such a large ring,

00:08:23 and indeed, that we're dealing here with a structural situation

00:08:29 more reminiscent of that which obtains in open-chain compounds than in ring compounds.

00:08:38 And in open-chain compounds, we might be inclined to think

00:08:42 that the open chain of carbon atoms will be rather floppy

00:08:47 and that there will not be elements of rigidity

00:08:50 which enable us to do stereochemical things in a very well-planned or well-defined way.

00:08:58 Well, now the actual situation in a molecule such as erythromycin

00:09:02 is probably not quite as near that of an open-chain compound

00:09:11 as might be implied by that first way of looking at the position.

00:09:16 And indeed, while I've shown here at the upper left

00:09:21 just a planar representation of the erythromycin structure,

00:09:25 below it there's a rather crude attempt to present the actual disposition of the atoms in space

00:09:33 in a three-dimensional sense.

00:09:35 It's, I say, a very crude representation of the three-dimensional situation.

00:09:40 But if one examines that lower formula,

00:09:43 one will find that, in fact, there are structural arrangements

00:09:50 which are rather reminiscent of those which obtain in the rigid systems

00:09:55 found in fused polycyclic compounds.

00:10:00 For example, let's just look at this chain.

00:10:04 You find that we have here a series of antiperiplanar carbon-carbon bonds.

00:10:12 There are about five of them there.

00:10:14 And the groups attached, in fact, have dispositions in space

00:10:19 which are highly reminiscent of equatorial and axial groups

00:10:24 attached to a cyclohexane ring.

00:10:27 And indeed, on the other side, you'll find another system of five carbon atoms

00:10:34 of which the same may be said.

00:10:36 Namely, these bonds are rigidly antiperiplanar.

00:10:39 And there's a group, hydrogen, it isn't written there,

00:10:42 but that little line is intended to represent the position of a hydrogen atom,

00:10:46 which you can see clearly looks rather like an axially disposed hydrogen atom,

00:10:51 whereas over here there's an OR group

00:10:56 which looks as though it were equatorially disposed.

00:11:01 Now, how do we know that erythromycin looks like this in space?

00:11:06 And that, of course, comes from X-ray crystallographic determinations

00:11:10 which show accurately the precise positions of the atoms

00:11:14 in the erythromycin molecule in a crystalline derivative of erythromycin.

00:11:20 Now, you may well say, all right, that's the way it looks in the crystal.

00:11:25 Is it going to be that way in solution?

00:11:28 Now, it does turn out, and it's a matter of much interest,

00:11:32 that in many cases, not only here in the case of erythromycin,

00:11:36 but in many, many cases it turns out

00:11:39 that the disposition of atoms in space of a molecule within a crystal

00:11:44 mirrors quite accurately the situation which the same molecule enjoys in solution.

00:11:52 In other words, one should not assume that when one dissolves a molecule,

00:11:58 it immediately just takes a very much more random orientation

00:12:02 of the atoms within its molecules.

00:12:06 And nuclear magnetic resonance studies on erythromycin and its derivatives

00:12:11 carried out, of course, in solution,

00:12:14 have led to a picture of the structure in solution

00:12:17 which is very similar to the one derived from the X-ray studies.

00:12:23 So, one might be tempted then to say,

00:12:26 well, should we not just make some kind of study

00:12:32 or develop some way of dealing with the stereochemistry

00:12:36 of these at least quasi-rigid large ring systems,

00:12:40 just as we earlier studied the stereochemical situation

00:12:46 in smaller six-membered rings?

00:12:50 Well, it was my feeling at least at the time we were planning this work

00:12:56 that we should not make the assumption, or make the attempt if you wish,

00:13:02 to devise such parallelisms, shall I say,

00:13:07 to these synthetic methods that had been worked out in the small ring systems.

00:13:12 I felt that even though in the case of erythromycin and its derivatives

00:13:17 it had been found that there was a considerable degree of conformational stability,

00:13:23 that that conformational stability would probably not be maintained

00:13:28 through very large changes in the nature of the attached groupings.

00:13:33 And therefore, it seemed to me that I was back at the situation I outlined in my opening remarks,

00:13:40 namely that one really should forget for the moment at least

00:13:44 the possibility of special 14-membered ring stereochemistry

00:13:49 and just go ahead to solve the problems associated with building a long chain of carbon atoms

00:13:54 with the many asymmetric centers incorporated within it.

00:13:59 Before leaving this slide, let me just mention almost parenthetically

00:14:04 that the macrolides also are characterized by having certain rather unusual sugar units

00:14:12 attached to oxygen atoms, which are themselves attached to the main chain.

00:14:16 Here they are in this case in erythromycin R1, which is cladinose,

00:14:21 this rather unusual sugar-like unit, and R2, desosamine, another sugar unit.

00:14:29 All right, now may we have the next slide.

00:14:33 So we are going to be dealing with a complicated stereochemical situation,

00:14:40 and I showed two representations of the erythromycin molecule on the earlier slide,

00:14:47 but I'd now like to show you a rather formal one, which, however, is very useful,

00:14:52 namely just the way in which we might designate the 14-membered ring of erythromycin

00:14:58 by the so-called Fischer Convention, which is one of the oldest of stereochemical conventions

00:15:03 and, in my opinion, in this field, probably the most useful convention.

00:15:08 The actual convention is indicated up here.

00:15:11 If one has any group of three atoms, N minus one, N and N, plus one,

00:15:18 and then two groups attached to the chain, one is attached to the left, the other to the right,

00:15:24 and these are designated as L and D groups.

00:15:27 And so you see here the chain of the erythromycin lactone system

00:15:34 consisting of 13 carbon atoms and one oxygen to make up the entire ring of 14 members,

00:15:41 and that the various ten of the carbon atoms along this chain, which are asymmetric,

00:15:48 and let's look, for example, at carbon atom two.

00:15:51 Here's a methyl group which might be on the right, as it happens to be,

00:15:55 or might have been on the left. That would be a different compound.

00:15:58 And you see from this Fischer Convention very easily and clearly

00:16:02 what the orientation at each asymmetric center is.

00:16:07 And I again add parenthetically that the widely used and popular R and S system

00:16:16 of Kahn, Engel, and Prelog is absolutely to be discarded in discussions of compounds

00:16:25 within this macrolide field. It leads only to utter confusion.

00:16:31 So we throw that away and stick to the simple Fischer Convention,

00:16:37 and I shall use this convention as a kind of monitor of our progress as we make it such as it is.

00:16:48 So on the next slide you can see more or less the epitomize of the conclusions

00:17:00 from the sort of rambling account I have given you in my introductory remarks.

00:17:06 Namely, we felt that we weren't going to be afraid of the 14-membered ring,

00:17:10 nor were we going to be tempted by it. We were in fact going to ignore it for some time

00:17:16 and just make a properly substituted 13-hydroxytribecanoic acid

00:17:22 with all the asymmetric centers present in the right orientation and with opposite substituents.

00:17:28 And we should leave then until some subsequent stage the problem of converting the hydroxy acid

00:17:34 or derivative of the hydroxy acid into the corresponding lactone.

00:17:40 Now I might say that that strategy has to some extent been shown in the intervening years

00:17:49 to be not entirely unwise in the light of the fact that other people in other research groups

00:17:57 most notably Massimuni in his synthesis of methimicin

00:18:02 and Professor Corey at Harvard in his synthesis of erythronolide B

00:18:08 have indeed found that very long-chain omega-hydroxy acids

00:18:14 can be converted to the corresponding lactones indeed in some cases quite smoothly.

00:18:24 I do have to say also though at this point

00:18:28 that I think that problem is one which deserves still more attention

00:18:34 and that the methods which have been used successfully

00:18:39 do not really represent special solutions which are needed

00:18:46 to the problem of favoring lactone formation over the alternative of course in such cases

00:18:52 which is polyester formation.

00:18:56 And so further work is needed in that area

00:19:01 though there have been conspicuous successes already.

00:19:05 Now our next major decision is shown on the next slide.

00:19:13 Having decided that we have very difficult problems,

00:19:15 ones that we didn't know how to deal with

00:19:19 because we decided we had to construct a straight chain,

00:19:23 long chain of carbon atoms containing many asymmetric centers

00:19:26 in a particular one of the many possible stereochemical arrangements,

00:19:32 we decided to obliterate all these intractable stereochemical problems

00:19:40 by what I've called there, perhaps a little presumptuously,

00:19:44 the creative introduction of elements of rigidity

00:19:47 which permits stereospecific operations of these types.

00:19:50 Well there's known of course, predictable, that's somewhat dubious word.

00:19:56 The third one I rather like, desired whether or not known or predictable.

00:20:02 We do like the third one.

00:20:05 And by the way it happens to the most sophisticated chemists.

00:20:10 Now what did I mean by creative introduction of elements of rigidity

00:20:15 which permits stereospecific operations of these types?

00:20:18 And perhaps I can show you what I mean by that on the next slide.

00:20:25 Now this is a wholly hypothetical or theoretical slide at the moment

00:20:30 and at the top on the left you'll see a bicyclic system.

00:20:36 This is a di-thia-decalin system.

00:20:38 It has two fused six-membered rings, not carbocyclic rings.

00:20:42 Each of them is heterocyclic having one sulfur atom.

00:20:46 There it is.

00:20:48 Now there is the creative introduction of elements,

00:20:54 elements literally if you wish, the elements happen to be sulfur atoms in this case.

00:21:00 You see this could be regarded as a straight chain

00:21:05 and notice if the sulfur atoms weren't there we'd have something like this.

00:21:10 And so what we've done is take this chain of atoms, carbon atoms,

00:21:15 to each other one of which there is attached a methyl group.

00:21:20 And if you happen to notice along the main chain of erythromycin

00:21:24 every other atom had a methyl group.

00:21:29 And we imagine that those methyl groups are bridged by sulfur atoms

00:21:32 then we would get to this di-thia-decalin that I've shown you on the left.

00:21:38 So indeed we can imagine this process in either direction

00:21:42 and so far it's purely an intellectual process

00:21:46 but we do know that chemically it is possible in many cases

00:21:50 to effect the direct removal of sulfur atoms from organic molecules using rainy nickel.

00:21:56 And were we to treat such a di-thia-decalin with rainy nickel

00:22:00 we might expect that we would remove the sulfur

00:22:03 and obtain this substance having this array of methyl groups.

00:22:08 So in other words this is a scheme so far hypothetical

00:22:12 which makes use of the recurring methyl groups on every other carbon atom

00:22:18 along the erythromycin chain essentially as the stereochemical determiners.

00:22:26 In other words we decided to use ring systems of this kind.

00:22:35 You will observe that we've now reduced the problem in a sense to

00:22:40 one the solution of which is known.

00:22:43 Since I've already mentioned that we know how to carry out stereospecific

00:22:48 or stereoselective operations in cyclohexane systems

00:22:52 or fused cyclohexane systems

00:22:55 and we could hope that the introduction of sulfur atoms

00:22:59 would not distort the situation which obtains in the carbocyclic systems very much

00:23:05 and that results might be achieved with the sulfur systems

00:23:09 comparable to those known for the carbon systems.

00:23:13 Now I've just gone a little further here to indicate that

00:23:18 in a dithiodecal system of this sort, the one that I've shown

00:23:22 which happens to be cis-fused,

00:23:26 one has a situation of two possible conformational arrangements

00:23:33 which are shown here.

00:23:35 These are the two conformations which that cis-dithiodecal system may assume

00:23:43 and you can see all the usual things that we associate with fused ring systems here

00:23:49 with in this case this group is an axial hydroxyl group,

00:23:54 an equatorial and of course in the other possible conformational arrangement

00:24:00 those change position in respect that this now becomes equatorial in that axis.

00:24:07 And the chain that I've been using here in this so far hypothetical exercise

00:24:11 is in fact part of the erythromycin chain

00:24:14 and here is the Fisher Convention representation

00:24:18 of the 9, 10, 11, 12, 13, 14 system of atoms

00:24:24 at one terminus of the erythromycin chain.

00:24:28 And you see what I'm suggesting is that we incorporate that chain

00:24:32 into a cyclic system which would enable us to carry out

00:24:36 stereospecific or stereoselective reactions.

00:24:40 With the intention ultimately of removing the sulfur atoms

00:24:43 which had provided the necessary rigidity for the direction of stereochemical operations

00:24:49 which would at the same time give us the needed structural elements

00:24:53 namely the array of alternate methyl groups.

00:24:59 So there's the plan, now let's go on to its execution.

00:25:03 With the next slide where we start with a simple and readily available

00:25:08 known substance, the 4-thiacyclohexanone.

00:25:13 Now can you still hear me in the back?

00:25:16 Right.

00:25:18 This substance is transformed through some rather simple operations

00:25:22 into the dimethyl ketal.

00:25:31 What might that be?

00:25:34 Yeah, forget it.

00:25:37 420?

00:25:39 How often will it happen?

00:25:47 Does it, for example, increase in frequency?

00:25:57 Alright, so we've got to this ketal.

00:26:00 We haven't gotten very far and the time has gone by.

00:26:05 We've got to this ketal.

00:26:07 And here we see something else which deserves special mention.

00:26:11 You see I've indicated that the sulfur was just put in

00:26:14 in order to provide some rigidity and enable us to work

00:26:19 not with open chain but with cyclic substances.

00:26:23 But as we shall see, and I think it's fair to say again and again,

00:26:26 the sulfur also served a number of other functions.

00:26:30 And here the function is that it activates the carbon atom next to it

00:26:35 so that substitution could be carried out very smoothly here

00:26:38 using N-chlorosuccinimide and carbon tetrachloride at zero degrees.

00:26:43 That introduces chlorine alpha to the sulfur in a very smooth reaction.

00:26:47 The chlorine can be replaced using thiourea which makes a carbon-sulfur bond

00:26:52 and then the extra atoms can be removed by hydrolysis

00:26:55 to give this cyclic diphyr-hemiacetal,

00:27:02 an interesting compound as we'll see in a moment,

00:27:05 which is readily obtained by the process shown.

00:27:10 Now, on the next slide,

00:27:16 we begin to develop some more atoms.

00:27:21 Actually, we're dealing now with the 9-10 carbon atom system

00:27:26 of the erythromycin main chain.

00:27:29 And what I've shown here is simply the gamma-benzo-oxy derivative

00:27:36 of butyric methyl butyrate,

00:27:39 which can be formulated in the alpha position,

00:27:43 at the 10 position here.

00:27:45 This carbon atom is introduced using methyl formate

00:27:49 and a strong base of lithium diisopropyl amide at minus 78 degrees.

00:27:55 And then, in a fairly standard reaction,

00:27:59 using methyl orthoformate, methanol, and sulfuric acid,

00:28:05 the formyl group, shown in enolized form here,

00:28:09 which it actually enjoys,

00:28:11 is transformed into corresponding dimethylacetal.

00:28:15 Now, the only reactive group present here is the ester group,

00:28:19 carbomethoxy group, which is reduced with lithium aluminum hydride and ether

00:28:24 to give a carbonyl grouping.

00:28:26 The carbonyl is then activated by conversion into the mesyl derivative

00:28:30 using mesyl chloride and pyridine.

00:28:32 All fairly standard reactions.

00:28:35 Now we may proceed to the next slide.

00:28:38 And on the extreme left,

00:28:41 we'll see the dithiacyclic hemiacetal,

00:28:45 whose preparation I showed earlier.

00:28:49 And now we see the substance that we've just made.

00:28:52 And here I show the combination of the two.

00:28:56 And you see we have the highly nucleophilic mercaptan group,

00:29:02 and here the reactive electrophilic reagent,

00:29:05 this methane sulfonyl derivative,

00:29:09 and in the presence of sodium methoxide, these combine,

00:29:13 and one gets then the combination products.

00:29:16 Now, I show here just two structures,

00:29:20 and even at that, this slide is simplified for purposes of explication.

00:29:26 Why so many isomers?

00:29:28 Well, for the moment, this substance is, of course,

00:29:32 one which possesses an asymmetric center and is racemic,

00:29:35 and so is this one.

00:29:37 And therefore, when we combine the two,

00:29:40 we must inevitably obtain diastereomers.

00:29:44 And I've shown two diastereomers here, you see,

00:29:48 which we can identify, if you wish, by the dispositions of the hydrogen atoms.

00:29:52 I've chosen to show hydrogen atom up here and here,

00:29:57 and of course the other hydrogen atom in one case,

00:30:00 then at the top here I've written it down,

00:30:03 in the other isomer it's up.

00:30:05 And of course, since these are racemic compounds,

00:30:07 the two mirror images of these are also produced.

00:30:10 Just bear that in mind,

00:30:12 but it would just make it a little complicated to show them all.

00:30:16 Now, after the combination of these two units,

00:30:19 the materials are not separated,

00:30:22 but rather subjected to further operations,

00:30:24 which are shown on the next slide.

00:30:29 Now, here's that same mixture reproduced again,

00:30:33 which has not been separated,

00:30:34 but simply treated with aqueous acetic acid,

00:30:38 the purpose of which is to remove all these protecting groups.

00:30:41 You see, this is both a ketal and an acetal,

00:30:44 as is its partner below,

00:30:46 and the aqueous acetic acid removes all those extra methoxy groups

00:30:50 and gets us to the ketoaldehyde,

00:30:53 keto here, aldehyde here.

00:30:56 That's also the case in the other diastereomer.

00:31:03 Now, these ketoaldehydes undergo a reaction,

00:31:07 which I think could be described as an expected one,

00:31:11 carried out when we first studied it, this reaction,

00:31:15 using a silica, silica gel, as catalyst.

00:31:21 Now, the reaction is an aldol reaction

00:31:25 in which a hydrogen is removed from this methylene group

00:31:30 adjacent to the ketonic carbonyl group of the ring,

00:31:33 and then an addition takes place to the aldehyde carbonyl group

00:31:37 to generate a new carbon-carbon bond

00:31:39 and, of course, a second ring, as shown here.

00:31:44 And that takes place also with the other isomer.

00:31:50 There's a new carbon-carbon bond formed.

00:31:52 Here it is shown here.

00:31:54 And so we have gone from the monocyclic compounds to bicyclic compounds,

00:31:59 which you may remember are reminiscent of those that I showed some slides back

00:32:04 in discussing more or less the planning of our work.

00:32:08 Here we are now with some actual diphyodecalin compounds

00:32:13 formed in these aldol cyclizations.

00:32:17 Now, you will notice that

00:32:20 in the conversion of each of these ketoaldehydes

00:32:23 to the bicyclic compound, two new asymmetric centers are generated.

00:32:28 Those are shown with hatched circles here.

00:32:31 There's a new center here and a new center here.

00:32:37 Now, interestingly enough,

00:32:40 that, of course, would mean that from each of the substances shown in the middle,

00:32:44 it would be possible, a priori, to get four possible cyclization products.

00:32:48 But, in fact, only one is produced.

00:32:52 So here is a reaction in which two asymmetric centers are generated.

00:32:59 They are generated under the influence of the one asymmetric center

00:33:02 which is actually present, namely this one,

00:33:06 and they are generated stereospecifically, each of them.

00:33:11 So, you see, here we have our first example of the control

00:33:16 over the stereochemistry of a reaction

00:33:18 which can be obtained when one is dealing with cyclic substances.

00:33:23 As I say, you get from this substance only one cyclization product.

00:33:29 Likewise, from this substance, only one cyclization product.

00:33:34 Now, these are both racemic, so I have to add, to be quite precise,

00:33:38 the qualification that the mirror image of each of these is also produced.

00:33:44 But the reactions are completely stereospecific.

00:33:48 And the diastereomeric, the racemic diastereomeric compounds

00:33:54 are, at this stage, very readily separable by chromatography.

00:34:01 Now, let us see why, or if we can give at least a rationalization,

00:34:10 of why this reaction takes place in so highly stereospecific a way.

00:34:15 And I think we might be able to discern that if we take a look at the next slide.

00:34:21 And over on the left here, I've tried to show a very crude representation

00:34:25 of what goes on in the cyclization which leads to the new carbon-carbon bond

00:34:32 with the generation of two asymmetric centers.

00:34:35 And what I have here is the intermediate ketoaldehyde shown in another form.

00:34:41 I have enolized the carbonyl group of the six-membered ring.

00:34:45 This is the enol in the direction that we want.

00:34:50 And now I've drawn this dotted line which more or less represents

00:34:55 a line drawn orthogonal to the plane of each double bond,

00:35:00 namely the double bond of the cyclic enol and the double bond of the carbonyl group.

00:35:09 Now, I think we have here yet another example of a role,

00:35:14 a special role played by the sulfur atoms.

00:35:18 In that, in this substance which has now two sulfur atoms,

00:35:24 if we got that sulfur atom in sulfur compounds of this kind,

00:35:30 one should observe the so-called anomeric effect,

00:35:34 the same anomeric effect that exists in the case of oxygen analogs,

00:35:39 then we would expect the sulfur atom attached alpha to a ring sulfur

00:35:46 to have a quasi-axial disposition.

00:35:51 Now, I know that some theorists say that there is no anomeric effect in sulfur,

00:35:56 but they should get their pencils out again because there isn't any doubt that there is

00:36:00 on the basis of many experimental observations, not only our own.

00:36:04 And I think this is one of our own at least,

00:36:07 which certainly provides some evidence that the anomeric effect is operative

00:36:12 in the case of sulfur compounds.

00:36:15 Now, you see, if this adopts the quasi-axial orientation as I have shown it here,

00:36:21 then, of course, we can only attack this ring in the carbon-carbon bond formation reaction

00:36:28 at the bottom, that is on the same side as the sulfur atom,

00:36:32 and that will lead necessarily to the production of compounds having a cis-ring juncture

00:36:37 as shown in these more or less three-dimensional representations.

00:36:46 Now, that takes care of this carbon atom.

00:36:50 One may expect that the hydrogen atom in an asymmetric center generated at that point

00:36:56 will be on the same side, I'm sorry, it's here,

00:36:59 will be on the same side as the one already present between the two sulfur atoms,

00:37:04 likewise up here.

00:37:07 There's the one between the two sulfur atoms and on the same side of the bicyclic system,

00:37:11 the one that's formed in the aldol reaction.

00:37:13 How about the hydroxyl group, which is also produced stereospecifically?

00:37:18 Now, here we think the answer is even simpler,

00:37:23 namely, this aldehyde-carbonyl group might be disposed for reaction in the sense shown,

00:37:29 or it might be twisted about this carbon-carbon bond

00:37:32 so that the carbonyl group would be pointing back more or less in that direction,

00:37:36 and of course, theoretically, that would be much less favorable

00:37:39 than the disposition in which I've shown it.

00:37:42 And therefore, we feel that accounts for the fact

00:37:45 that the hydroxyl group comes up in both cases.

00:37:51 And it would appear that these very simple considerations

00:37:55 are responsible for the fact that the two centers are formed

00:38:00 in a completely stereospecific way.

00:38:03 Now, of course, you will recall I've abbreviated this a bit.

00:38:06 We had both isomers, those in which the two R groups,

00:38:09 the R group was on one side of this molecule or on the other,

00:38:14 and I've just shown them as the same thing here for simplification.

00:38:18 But now, in the products, you see, these are the same

00:38:24 insofar as hydrogen here, oxygen there, hydrogen here, oxygen there.

00:38:31 But the large group, the large attached group, is the one shown here.

00:38:38 And of course, that's on a different side of the bicyclic system

00:38:41 in each of these two cases.

00:38:44 And incidentally, and this is of some importance in our later discussion,

00:38:48 of course, that large group will, by preference,

00:38:53 take up the equatorial orientation as shown here or here in either case.

00:39:00 And of course, that determines the choice of which of the two

00:39:05 a priori possible cis-decalin conformations one might have.

00:39:09 They will be different in the two cases because the large groupings

00:39:13 are on opposite sides of the ring systems in the two cases.

00:39:19 All right, now, the two, either of these aldol products,

00:39:29 hydroxyketones, can be carried further by simple reactions.

00:39:33 The experimental details are not quite the same,

00:39:36 but the method is, in principle, the same.

00:39:39 The hydroxyl group is mesilated here, and then, using allox,

00:39:47 the mesial group is removed to give the alpha-beta unsaturated ketone,

00:39:52 a very simple reaction.

00:39:54 And likewise, with the other isomer, again, the hydroxyl group is mesilated.

00:40:00 When that is mesilated, in this case, under the conditions of the reaction,

00:40:04 the elimination takes place to give the unsaturated ketone spontaneously.

00:40:09 And so we can get both or either of the unsaturated ketones

00:40:15 shown here on the right in that simple way.

00:40:21 Now, it probably will not be apparent to you right at the moment

00:40:27 that...

00:40:34 It seems to be speeding up just a little.

00:40:37 Perhaps I can speed up a little.

00:40:39 It probably won't be apparent to you immediately at this moment,

00:40:43 but one of these two isomers is one that we want.

00:40:47 It's namely this one and this one we don't really want for synthetic purposes.

00:40:52 So, for a while at least, my discussion, the ensuing discussion,

00:40:56 will be concentrated upon the upper isomer,

00:40:59 that which has the hydrogen atom here

00:41:02 and at what we hope someday is going to be C10

00:41:06 on opposite sides of the dithiodecal system.

00:41:12 Now, let us see on the next slide how we go further

00:41:17 and how we make use of this general plan of achieving stereospecificity

00:41:22 through having a relatively rigid system,

00:41:24 such as that which one has in the dithiodecal.

00:41:28 A series of rather straightforward transformations.

00:41:31 In the alpha-beta unsaturated ketones,

00:41:34 the same one shown in the earlier slide,

00:41:36 with hydrogen here and here on opposite sides of the ring system.

00:41:41 The carbonyl group is simply reduced with sodium borohydride.

00:41:46 That takes place as one would expect stereospecifically

00:41:50 to give only one alcohol with the hydroxyl group down.

00:41:55 The alcohol group is then protected by acetylation.

00:42:07 The acetyl derivative is then subjected to osmylation

00:42:15 at the carbon-carbon double bond.

00:42:18 Many of you who would look at a model of this would realize

00:42:21 that the osmylation can take place only in one sense, and it does.

00:42:25 The two hydroxyl groups are added on the top face of the molecule

00:42:29 using osmium tetroxide,

00:42:32 and then those are subjected to protection

00:42:36 using acetone dimethyl ketal and a trace of acid

00:42:40 to give this protected substance.

00:42:45 Now, we've built up then a considerable chain

00:42:49 of asymmetrically substituted centers,

00:42:53 here, here, here, and here, four of them.

00:42:57 Now let's look at the next slide

00:43:00 in which I show that molecule that was just shown earlier.

00:43:07 And we were able at this point

00:43:11 to have a full three-dimensional X-ray crystallographic

00:43:15 structure determination of that acetate

00:43:19 carried out by Mrs. G. Reese in Basel,

00:43:23 a kind of captive X-ray crystallographer

00:43:26 who's extremely able.

00:43:29 We'll see more of her work as we go on.

00:43:32 And that X-ray determination verified,

00:43:35 and here's an attempt to show more or less in three dimensions

00:43:40 the result obtained by Mrs. Reese in the X-ray crystallographic study,

00:43:44 and it verifies the stereochemistry,

00:43:48 which, if you wish, up until this point

00:43:52 had been based simply on argumentation.

00:43:55 There was other evidence, of course.

00:43:58 We had nuclear magnetic resonance evidence and so forth,

00:44:01 but really the final and most conclusive evidence

00:44:05 for structure is an X-ray crystallographic structure determination.

00:44:09 And here it is in this case.

00:44:12 And we had built up, then, by the relatively simple sequence of reactions

00:44:17 which I've shown you,

00:44:20 the 10, 11, 12, and 13 centers

00:44:25 of the carbon chain of erythromycin,

00:44:28 shown here in the Fischer projection.

00:44:31 I don't expect you to remember back to the full projection that I showed you,

00:44:35 but if you were able to compare them,

00:44:38 you would find that, in fact, now,

00:44:41 this acetate that we've prepared by synthesis

00:44:44 is a substance in which the stereochemical problem

00:44:47 presented by this terminus of the erythromycin chain

00:44:51 has been solved.

00:44:54 Now let's go on to the next slide.

00:44:58 All of the work that I have described so far

00:45:01 was carried out with racemic substances,

00:45:04 and I've shown in each case

00:45:07 only one of the two mirror image enantiomers,

00:45:11 just for purposes of simplification.

00:45:15 But, of course, erythromycin is an optically active substance,

00:45:19 and one needs to have, then, only, ultimately,

00:45:23 the substance of the correct absolute configuration,

00:45:27 and therefore the intermediates, at some stage,

00:45:30 have to be substances which are optically active

00:45:33 and have the right absolute configuration,

00:45:36 as well as having proper

00:45:39 relative stereochemical relationships

00:45:42 within each molecule.

00:45:45 And here, at this point, we come to grips with that problem.

00:45:49 And at the left here, you will see

00:45:52 the now familiar dithiacyclic hemiacetal.

00:45:57 Which has a free sulfhydryl group,

00:46:00 which could be esterified very easily.

00:46:04 And now this case, I've shown both enantiomers

00:46:07 of that mercaptan.

00:46:10 There's the mirror plane, you see.

00:46:13 And we have one mirror image isomer here,

00:46:16 and the other one here.

00:46:19 These two make up together the racemic mercaptan.

00:46:23 Now, therefore, when we treat

00:46:26 with an optically active chloride,

00:46:29 and we used the levorotatory camphanyl chloride,

00:46:33 which is readily obtainable from the

00:46:36 dextrorotatory camphor,

00:46:39 which is a well-known form of camphor.

00:46:43 Since this is optically active, this is racemic,

00:46:46 we do, of course, get two diastereomers.

00:46:49 And here they are.

00:46:52 One compound each from one of the

00:46:55 mirror image enantiomers.

00:46:58 Two diastereomers. Now those diastereomers

00:47:01 are readily separated by crystallization.

00:47:04 And what we have done then,

00:47:07 essentially, note,

00:47:10 the left-hand parts of these diastereomers

00:47:13 are still mirror images, but the right-hand parts

00:47:16 are the same. They derive from this optically active material.

00:47:19 What we have done then, essentially, is we've resolved

00:47:22 the mercaptan

00:47:25 into its optical enantiomers.

00:47:28 How do we know which is which?

00:47:31 Well, of course, that's very easy.

00:47:34 We asked Mrs. Reese to do an x-ray

00:47:37 crystallographic structure determination, which she did

00:47:40 with great skill and great speed

00:47:43 on this isomer, which we guessed was the one that we wanted,

00:47:46 and we guessed correctly. I won't tell you how we guessed.

00:47:49 Now, you see, this is a very nice way

00:47:52 of determining the absolute configuration

00:47:55 of an organic substance, that is, have a full

00:47:58 x-ray crystallographic study done of a substance

00:48:01 into which you have inserted something whose absolute configuration

00:48:04 is known. That is, since we knew

00:48:07 the configuration here, the x-ray study then

00:48:10 tells us the configuration at the other part of the molecule.

00:48:13 I should add, by the way, that unless

00:48:16 special measures are taken which complicate x-ray crystallographic

00:48:19 determinations, they do not

00:48:22 give one the absolute configuration. But this is a

00:48:25 simple little trick which enables one

00:48:28 from an x-ray structure determination to determine

00:48:31 absolute configuration. And this

00:48:34 was the substance of the desired absolute

00:48:37 configuration. That isn't an obvious

00:48:40 relationship at the moment, but I think it will become so

00:48:43 as we go along. And so this represented, if you wish,

00:48:46 a resolution of the

00:48:49 mercaptan into the optical

00:48:52 enantiomers and an establishment of which of the ones

00:48:55 we wanted. Now, this statement

00:48:58 that I've just made is true only

00:49:01 in the light of what can be regarded, I think,

00:49:04 as a remarkable observation which we made.

00:49:07 Namely, that the ester,

00:49:10 the camphanylthioester,

00:49:13 on treatment with sodium methoxide and methanol

00:49:16 underwent methamolysis

00:49:19 of the thioester grouping to give the

00:49:22 fremercaptan. And, this is a surprising thing,

00:49:25 the fremercaptan is entirely

00:49:28 configurationally stable. In fact, we don't know how

00:49:31 to maximize it. Now, that is in striking

00:49:34 contrast to the situation that one would have

00:49:37 if one had oxygen here and here in place

00:49:40 of the sulfur atoms. Those substances,

00:49:43 the hemiacetals, cyclic

00:49:46 hemiacetals of that sort, have very little

00:49:49 in the way of configurational integrity.

00:49:52 The classic case, of course, is that of the sugars which are

00:49:55 of, do have that structural element. They undergo

00:49:58 a so-called mutarotation, which is nothing other than

00:50:01 a loss of stereochemical integrity

00:50:04 at the carbon atom corresponding to this one.

00:50:07 That's the case for the oxygen analogs, but

00:50:10 to our great good fortune, in the case of the

00:50:13 sulfur analogs, that is

00:50:16 not so. And these mercaptans are

00:50:19 configurationally perfectly stable, and so we

00:50:22 could obtain here the resolved

00:50:25 optically pure mercaptan

00:50:28 of the desired absolute configuration.

00:50:31 Now, just for the fun of it, I'm going to

00:50:34 point out that at this moment, of course, this substance

00:50:37 has only one asymmetric center, and it's not an asymmetric

00:50:40 center that's present in erythromycin at all.

00:50:43 So you might say, well, how on earth could it then have

00:50:46 the desired conformation? Well, perhaps that will become

00:50:49 clear as we proceed.

00:50:52 To the next slide.

00:50:55 Earlier slide.

00:50:58 And here I show both forms,

00:51:01 of course, the two mirror images. This is a racemic substance

00:51:04 and therefore is an equal mixture of the two enantiomers

00:51:07 shown here. Now we combine that

00:51:10 with the optically active mercaptan

00:51:13 of the desired absolute configuration. The combination

00:51:16 reaction is the same, the displacement by the highly

00:51:19 nucleophilic mercaptan group of the mesialate group

00:51:22 in either case. And that, of course,

00:51:25 leads to the optically active diastereomers

00:51:28 just as previously it had

00:51:31 led to a mixture of

00:51:34 racemic diastereomers. Again,

00:51:37 they're not separated in the sequence at this stage,

00:51:40 but we go on to the next slide where we see

00:51:43 that mixture of now optically active

00:51:47 diastereomers, which is again treated with

00:51:50 aqueous acetic acid. This is the same change

00:51:53 giving the ketoaldehyde,

00:51:56 two of them. And in this case,

00:51:59 using these optically active substances,

00:52:02 we use the catalyst for the aldol reaction.

00:52:05 We now wish to carry out this carbon-carbon bond forming

00:52:08 aldolization, which I've already described,

00:52:11 and which in the racemic case was brought about using

00:52:14 silica gel. We now carry that reaction out

00:52:17 using as catalyst D-proline.

00:52:20 And aldolization

00:52:23 occurs. It's entirely stereospecific aldolization

00:52:26 as it was in the earlier described case.

00:52:29 The new carbon-carbon bond is formed

00:52:32 and these two substances are

00:52:35 obtained and are very readily separable by chromatography.

00:52:38 The upper one, the desired one, being a less

00:52:41 polar substance, they're both easily obtained

00:52:44 in the pure state, optically pure.

00:52:47 And it is the upper one of the two

00:52:50 which, as we shall see, we desire.

00:52:53 So here we have, in this relatively still

00:52:56 simple sequence, a way of obtaining now

00:52:59 a substance of known absolute configuration,

00:53:02 known relative stereochemistry,

00:53:05 and it provides the basis for our further discussion.

00:53:08 So we go to the next slide

00:53:11 where we see again the aldol on the upper left,

00:53:14 now only the desired one, the one that has

00:53:17 the hydrogen here between the sulfur

00:53:20 and the hydrogen over here on opposite sides

00:53:23 of the diethyde-decalin system. This is now a resolved

00:53:26 substance. It's optically active and

00:53:29 it has the desired absolute

00:53:32 configuration and the desired

00:53:35 relative configuration

00:53:38 at some of the centers.

00:53:41 And now we proceed through a series of reactions

00:53:44 very reminiscent of the ones I've already described.

00:53:47 The dehydration of the beta-hydroxyl group

00:53:50 using mesothloride pyridine followed by

00:53:53 alox to give the alpha-beta unsaturated ketone.

00:53:56 The reduction of sodium borohydride.

00:53:59 This is now just being carried out in the optically active series.

00:54:02 But now a slight change which

00:54:05 turns out to be a convenient

00:54:08 matter of convenience. Hitherto we had

00:54:11 acetylated that compound. Now we

00:54:14 transform this hydroxyl group into the corresponding

00:54:17 methoxy methyl group through treatment

00:54:20 with iodomethyl ether and potassium hydride

00:54:23 under the right conditions. And now on the next

00:54:26 slide, that methoxy methyl

00:54:29 compound appears again on the left.

00:54:32 It's osmolated. Again the

00:54:35 osmolation takes place of course stereospecifically to

00:54:38 deliver two hydroxyl groups from the top of this molecule.

00:54:41 Here they are in the next intermediate.

00:54:44 And again this substance was protected.

00:54:47 The two hydroxyl groups within that substance were protected

00:54:50 by the formation of a cyclic ketal in the

00:54:53 usual and earlier described way.

00:54:59 Now, let us

00:55:02 go to the next slide

00:55:05 where on the left

00:55:08 we see that same intermediate.

00:55:11 The methoxy methyl ether

00:55:14 you see. It's the same substance that I showed before.

00:55:17 And now we have to put to the

00:55:20 test one of the basic

00:55:23 elements of our

00:55:26 whole plan.

00:55:29 Namely, we have in this substance

00:55:32 built up a chain of atoms

00:55:35 having the desired relative

00:55:38 stereochemistry at each of four centers

00:55:41 along that chain. We've used the sulfur atoms

00:55:44 to confer upon the systems the rigidity

00:55:47 needed to achieve the desired stereospecificity.

00:55:50 And ultimately we had to get rid of the sulfur

00:55:53 atoms and here we are. At this point

00:55:56 this compound is desulfurized

00:55:59 using W2 rainy nickel in ethanol

00:56:02 and the exclamation

00:56:05 mark is there just to indicate

00:56:08 our pleasure that the reaction in fact did take place

00:56:11 without any complications whatsoever.

00:56:14 That isn't always the case with one's chemical plan.

00:56:17 But very fortunately in this case

00:56:20 it is so. This is a very smooth desulfurization

00:56:23 indeed and you see if you

00:56:26 take those sulfurs away you're left with

00:56:29 the three methyl groups which one needs

00:56:32 attached to alternate carbon atoms.

00:56:35 Now

00:56:38 this chain

00:56:41 needs to be modified and I show here a few

00:56:44 of the steps used to modify it.

00:56:47 The first reaction is one with

00:56:50 orthonitrophenylselenosyanate

00:56:53 in the presence of

00:56:56 tributylphosphine

00:56:59 which converts this alcohol

00:57:02 this primary alcohol very smoothly into the

00:57:05 corresponding orthonitrophenyl

00:57:08 seleno compound. Very clean

00:57:11 and smooth reaction. And now

00:57:15 again a simple and smooth reaction.

00:57:18 The resulting selenium

00:57:21 compound is oxidized with hydrogen peroxide

00:57:24 and tetrahydrofuran. One does not see

00:57:27 the selenoxide which is

00:57:30 the presumed initial product of that oxidation

00:57:33 because elimination takes place spontaneously

00:57:36 to give the olefin. So you see

00:57:39 the result of those two steps has been simply a dehydration

00:57:42 of this alcohol to give an olefinic

00:57:45 group here which is forthwith cleaved

00:57:48 using ozone

00:57:51 to give the corresponding aldehyde.

00:57:54 There it is. Now we've built up

00:57:57 this substance which now

00:58:00 you see is an open chain substance containing

00:58:03 four asymmetric centers.

00:58:06 It's optically active and has the right relative

00:58:09 configuration at

00:58:12 four centers as well as the right

00:58:15 absolute configuration.

00:58:18 Now let's

00:58:21 on the next slide go back to this

00:58:24 same

00:58:27 substance which was the point of departure for the sequence

00:58:30 shown on the just preceding slide.

00:58:33 It's the same compound but now we do something a little

00:58:36 bit different to it.

00:58:39 The methoxymethyl ether here, you remember we put that on as a

00:58:42 protecting group. We now remove it using

00:58:45 carefully chosen conditions which involve the right

00:58:48 amount of trifluoroacetic acid and methylene chloride

00:58:51 that frees the hydroxyl group

00:58:54 which now is very readily oxidized using

00:58:57 trifluoroacetic anhydride and dimethyl

00:59:00 sulfoxide followed by

00:59:03 isopropyl ethylamine and methylene chloride.

00:59:06 That simply is a nice way of oxidizing

00:59:09 this ketone, this alcohol

00:59:12 to the corresponding ketone.

00:59:17 Now let's go to the next slide.

00:59:24 At this time it is, I think,

00:59:27 desirable to point out

00:59:30 a remarkable and

00:59:33 in the light of our plans

00:59:36 highly significant feature

00:59:39 of the stereochemical arrangements

00:59:42 within the erythromycin molecule.

00:59:45 And to introduce that topic I show here on the right

00:59:48 again this main chain

00:59:51 from 1 through

00:59:54 13 carbons and then the 14th being oxygen

00:59:57 with its 10 asymmetric centers

01:00:00 and I've put these brackets here

01:00:03 these brackets to show

01:00:06 that the relative and absolute

01:00:09 stereochemical relationships at

01:00:12 C10, C11, and C12

01:00:15 and remember those are the ones that we've just built up in this

01:00:18 aldehyde intermediate whose preparation I have described.

01:00:21 There's C10, C11, C12,

01:00:24 C13 and if you translate this diagram

01:00:27 into our formal Fischer diagram you'll see that we have

01:00:30 here in this aldehyde intermediate

01:00:33 we have exactly what we want in respect of

01:00:36 the stereochemistry in an absolute and relative sense

01:00:39 at C10, C11, C12, and C13

01:00:42 of the erythromycin chain.

01:00:45 Now the special feature to which I alluded a moment ago

01:00:48 is that the stereochemical circumstances

01:00:51 at C10, C11, and C12

01:00:54 in the erythromycin main chain

01:00:57 are both in a relative and absolute sense

01:01:00 precisely the same as the stereochemical circumstances

01:01:03 at C4, C5, and C6.

01:01:06 So if you look at

01:01:09 C4 here

01:01:12 and look at C10. So you have

01:01:15 hydrogen to the left, methyl to the right at C10

01:01:18 hydrogen to the left at 4, methyl to the 10

01:01:21 that indicates that the configurational

01:01:24 relationships at 4 and 10 are precisely the same.

01:01:27 Well what does that mean from the synthetic point of view?

01:01:30 Just this. I've already mentioned that the aldehyde

01:01:33 whose preparation I've described

01:01:36 gives us the C9, 10,

01:01:39 11, 12, 13 part of the erythromycin chain

01:01:42 and the ketone whose preparation

01:01:45 I've described gives us the 4, 5, 6

01:01:48 part of the erythromycin molecule.

01:01:51 That is to say by the same route and indeed from one point

01:01:54 from the same intermediate we can make building blocks

01:01:57 namely the ketone which represents the

01:02:00 4, 5, 6 part of the erythromycin molecule

01:02:03 and this aldehyde which represents

01:02:06 the 10, 11, 12, and 13

01:02:09 parts of the erythromycin molecule.

01:02:12 And as luck would have it the functionality

01:02:15 in the two building blocks is complementary

01:02:18 in that you see the ketone

01:02:21 representing the 4, 5, 6

01:02:24 system has here

01:02:27 what would become C8, an activated

01:02:30 methylene group

01:02:33 and of course that provides an opportunity for a

01:02:36 condensation with the aldehyde group which represents

01:02:39 C9 in the other intermediate

01:02:42 of aldol condensation in other words

01:02:45 which would make a C8-C9 bond

01:02:48 that's over here

01:02:51 C8-C9 would join these two chains together

01:02:54 so that in fact was our

01:02:57 next operation and if we look on the next slide

01:03:00 you'll see on the upper left

01:03:03 the product of the aldol reaction

01:03:06 very smooth reaction between the ketone which was shown on the top

01:03:09 of the previous slide

01:03:12 represented here on the right and the aldehyde

01:03:15 you see there's the aldehyde oxygen

01:03:18 on the left. Now

01:03:21 that aldolization is not stereospecific

01:03:24 it leads to a mixture of aldols but that's a matter

01:03:27 of no consequence because the asymmetry

01:03:30 generated in this aldolization is

01:03:33 immediately destroyed as you see

01:03:36 in the very next step because

01:03:39 without isolation of a pure aldol

01:03:42 the hydroxyl group is

01:03:45 oxidized

01:03:48 using trifluoroacetic anhydride

01:03:51 dimethyl sulfoxide and diisopropyl

01:03:54 ethylamine. The aldehyde is oxidized

01:03:57 and that gets us to a carbonyl compound

01:04:00 the state of dicarbonyl compound of course is readily enolizable

01:04:03 it does exist to a considerable

01:04:06 extent in the ketonic form

01:04:09 and we lose

01:04:12 the asymmetry of course at this center

01:04:15 when that oxidation takes place

01:04:18 this center becomes labile the other one

01:04:21 of the two produced and as we'll see there's

01:04:24 every reason to suppose that it will assume only one of the two

01:04:27 possible confirmations

01:04:30 sorry, configurations, more of that in a moment

01:04:33 now that

01:04:36 beta dicarbonyl compound

01:04:39 is then converted into an

01:04:42 enol acetate shown here at the bottom

01:04:45 now this really is a simple but very very interesting reaction

01:04:48 the study of

01:04:51 the enolization and acylation and so forth

01:04:54 of beta diketones has been

01:04:57 studied for a century by organic chemists

01:05:00 but there are still things to be learned about it

01:05:03 for one thing I might point out

01:05:06 that from such a beta diketone

01:05:09 you can imagine that at least three possible

01:05:12 monoacetates might be formed

01:05:15 that is there are three possible enols

01:05:18 that one can write, monoenols

01:05:21 and correspondingly three possible acetates

01:05:24 and the whole story would be too long

01:05:27 one to tell here, an interesting one

01:05:30 this is the acetate we wanted and the one we were able to make

01:05:33 using the right conditions

01:05:36 namely treatment of the diketone with potassium hydride

01:05:39 and hexamethylphosphoramide THF mixtures at zero

01:05:42 cooling to minus 78 degrees

01:05:45 treating with acetyl chloride

01:05:48 this one of the possible

01:05:51 monoacetoxic compounds

01:05:54 and you see it is the one in which

01:05:57 the carbonyl

01:06:00 previous carbonyl oxygen and this bond are

01:06:03 disposed in a trance sense

01:06:06 now when the acetate had been obtained

01:06:09 it was reduced with sodium borohydride

01:06:12 it's the carbonyl group here which is reduced

01:06:15 with the corresponding carbonyl shown here

01:06:21 now if we go to the next slide we'll see its fate

01:06:24 that carbonyl

01:06:27 was treated with mesyl chloride

01:06:30 and pyridine and then with gamma

01:06:33 dimethylamino pyridine and methylene chloride methanol

01:06:36 mesolate this hydroxyl group

01:06:39 and then treating with this pyridine derivative

01:06:42 effect elimination and cleavage to give now

01:06:45 the alpha beta unsaturated ketone

01:06:48 shown here on the right

01:06:51 now why did we go through all that

01:06:54 well because we wanted to remove an oxygen

01:06:57 atom from this chain

01:07:00 that was the oxygen atom of the ketone building block

01:07:03 of course that oxygen atom was useful in order to

01:07:06 make the bond that we needed to but now we needed to get rid of it

01:07:09 and it is the processes that I'm now describing

01:07:12 which were useful in the elimination

01:07:15 of the unwanted oxygen atom

01:07:18 now in fact at this point all we need is a CH2

01:07:21 group that is there's no substituent in erythromycin

01:07:24 at that group and you might say well why not just

01:07:27 reduce that carbon-carbon double bond

01:07:30 that looks like a simple operation but in fact we were unable

01:07:33 to do that but here sulfur came to our aid again

01:07:36 this is an alpha beta unsaturated

01:07:39 carbonyl system so it readily undergoes 1,4 addition

01:07:42 of mercaptans

01:07:45 the best choice turned out to be

01:07:48 a benzyl mercaptan in the presence of

01:07:51 normal butyl lithium in tetrahydrofuran

01:07:54 the benzyl mercaptan adds undergoes addition

01:07:57 very smoothly to that

01:08:00 alpha beta unsaturated carbonyl system

01:08:04 now here's an interesting point

01:08:07 to which I alluded very briefly just a moment ago

01:08:10 namely we're still dealing with

01:08:13 molecules which have a di-thiodecal system

01:08:16 present here

01:08:19 and here and so now here's

01:08:22 what I think is one of the rather nicer stereochemical

01:08:25 points of our scheme

01:08:28 just consider this intermediate

01:08:31 in general terms as shown over here on the left

01:08:34 in which I've given a quasi-stereochemical

01:08:37 or three-dimensional representation of a di-thiodecal

01:08:40 system and you notice here at what

01:08:43 corresponds to C4 of the erythromycin chain

01:08:46 we have a very large side chain

01:08:49 which we'll want of course to be equatorial

01:08:52 and we'll tend to force the

01:08:55 di-thiodecal system to occupy or to assume

01:08:59 one of the two a priori possible conformations

01:09:02 now having done so then

01:09:05 the group at the other position over here at the other side of the molecule

01:09:08 C8 will also wish to adopt the equatorial

01:09:11 disposition and it may do so because

01:09:14 of course the hydrogen next to it is enolizable

01:09:17 and this is then an invertible center

01:09:20 so in other words what I'm saying is that in this di-thiodecal

01:09:23 system when you have two large groups

01:09:26 one over on this side at what's going to be C4

01:09:29 the other at the other side at what's going to be C8

01:09:32 the stable orientation of groupings is that

01:09:35 which has hydrogen up here and hydrogen down here

01:09:38 and that's exactly the relative orientation

01:09:41 that one wants within the erythromycin molecule

01:09:44 so we could be quite confident

01:09:47 that the conformation here

01:09:50 was the one that I've shown

01:09:53 continuing with this hydrogen up at C8

01:09:56 and down at C4

01:10:04 now let's go on to the next slide

01:10:07 in which essentially I just

01:10:10 summarize the point that we've reached

01:10:13 and indicate what remains to be done

01:10:16 over on the left you see

01:10:19 the molecule that we've just reached with this

01:10:23 addition of benzoyl mercaptan

01:10:26 and within that molecule

01:10:29 we now have

01:10:32 we feel C4, C5

01:10:35 C6 in the right

01:10:38 relative configurations and right absolute

01:10:41 configuration on the basis of the argument I've just

01:10:44 presented C8 in the right relative

01:10:47 and absolute configuration and likewise

01:10:50 10, 11, 12 and 13

01:10:53 so one might have

01:10:56 some ground for being confident that at this point

01:10:59 8 of the 10 centers

01:11:05 4, 5, 6, 8

01:11:08 10, 11, 12 and 13 were all in place

01:11:11 in an absolute sense and in a relative sense

01:11:14 and the remaining task

01:11:17 was to build

01:11:20 from this carbon atom

01:11:23 which represents C3 the rest of the chain

01:11:26 namely here we are we need to modify

01:11:29 C3 and add two carbon atoms

01:11:32 add 2 and 1 so we have to add two carbon atoms

01:11:35 in the chain and we have to affect

01:11:38 the construction of the stereochemistry in the proper

01:11:41 sense of C3 and C2

01:11:44 so let's see whether there was anything

01:11:47 that we could do about that

01:11:50 and on the next slide

01:11:59 we see some further rather simple reactions

01:12:02 namely the reduction

01:12:05 of the carbonyl group

01:12:08 not much going on in this slide the formulae look fairly

01:12:11 formidable but the reactions are simple

01:12:14 the carbonyl group is reduced with lithium aluminum hydride

01:12:17 and ether at minus 30 degrees

01:12:20 a couple of interesting things about that reaction

01:12:23 simple though it may be namely it takes place entirely

01:12:26 stereospecifically only one product is formed

01:12:29 and I have to tell you that I have no idea why that's the case

01:12:32 it was not predicted and I don't care to rationalize it

01:12:35 after the fact I'll just leave it as a fact

01:12:38 it is reduced

01:12:41 to give only one of the two possible alcohols

01:12:44 now I've shown a particular one of them

01:12:47 and in that sense I'm anticipating a bit because I've just

01:12:50 told you and it's true that we did not predict it

01:12:53 and we cannot or do not rationalize it

01:12:59 it's also interesting by the way that carbonyl group isn't reduced at all

01:13:02 by sodium borohydride under any conditions

01:13:05 it's a rather peculiar carbonyl group

01:13:08 now the next reaction is very simple

01:13:11 the hydroxyl group is simply acylated

01:13:14 more or less standard way to give the corresponding

01:13:17 acetoxy compound

01:13:20 and now we can go to the next slide

01:13:23 and here

01:13:26 you see an even bigger exclamation point

01:13:29 because again we have to

01:13:32 come to this key element in our planning

01:13:35 the sulfur element

01:13:38 and we have to remove it by rainy nickel from this half of the molecule

01:13:41 and indeed that takes place very smoothly

01:13:44 generating now another three of the methyl groups

01:13:47 based alternately along this

01:13:50 building chain

01:13:53 now I may say the sulfur atom has really done us

01:13:56 a great deal of good in this whole

01:13:59 story

01:14:02 this could be called a kind of exercise in the chemistry of sulfur

01:14:05 because of course in this desulfurization

01:14:08 we not only remove these two sulfurs

01:14:11 but we remove this benzyl mercapto

01:14:14 I'm sorry this benzyl diol group

01:14:17 to deliver the corresponding hydrogen to this carbon

01:14:20 giving us a CH2 group there which is what we need

01:14:23 so we use sulfur even to effect the reduction

01:14:26 of the double bond which I mentioned a little bit earlier

01:14:29 in alpha beta unsaturated ketone

01:14:32 whose double bond we couldn't reduce by more obvious or simple methods

01:14:35 so here's yet another use of sulfur in our sequence

01:14:38 and this desulfurization carried out

01:14:41 under proper conditions proceeds very smoothly

01:14:44 removing those two sulfurs

01:14:47 and likewise this one with no complications

01:14:50 so now we proceed

01:14:53 to the next slide

01:14:56 and at the top on the left you'll see

01:14:59 the desulfurization product

01:15:02 and on this slide you'll see a series of reactions

01:15:05 which is entirely reminiscent of

01:15:08 in fact identical with

01:15:11 those reactions used to make

01:15:14 the aldehyde building block for the lower terminus

01:15:17 of the erythromycin chain

01:15:20 in other words here we have a primary alcohol

01:15:23 and by the way I didn't call attention to it

01:15:26 but that desulfurization reaction is very much

01:15:29 the more helpful in that

01:15:32 during the course of the desulfurization the benzyl protecting group

01:15:35 is removed here to give this free alcohol group

01:15:38 here we have then in the desulfurization product

01:15:41 a free hydroxyl group and again using

01:15:44 orthonitrophenyl selenosyanate

01:15:47 the corresponding orthonitrophenyl selenocompound

01:15:50 is produced by the same kind of reaction

01:15:53 the same oxidation with hydrogen peroxide

01:15:56 and tetrahydrofuran

01:15:59 giving the olefin and again oxidation

01:16:02 to give the aldehyde this is just the change

01:16:05 that we've already established in the simplest series

01:16:08 the transformation of this side chain into the next lower aldehyde

01:16:11 in a very smooth series of reactions

01:16:15 now on the next slide

01:16:18 we see that aldehyde produced there again

01:16:21 on the left and now

01:16:24 we come to the problem that I mentioned that we must

01:16:27 add two further carbon atoms of the chain

01:16:30 including I might say parenthetically

01:16:33 one of these methyl groups attached

01:16:36 to it and we must generate two

01:16:39 new asymmetric centers the last two

01:16:42 present in the erythromycin main chain

01:16:45 and here we see how that

01:16:48 in the end turned out to

01:16:51 be possible and here again we see some sulfur

01:16:54 you see sulfur is again and again popping up in this

01:16:57 sequence we simply use the anion generated

01:17:00 from the tertiary butyl

01:17:03 thioester of propionic acid

01:17:06 there's that anion which undergoes

01:17:09 at minus 110

01:17:12 degrees

01:17:15 for eight minutes

01:17:18 a very smooth reaction

01:17:21 addition to the carbonyl group of the aldehyde

01:17:24 to give the product shown below here

01:17:27 it's just a beta hydroxy thioester

01:17:30 and interestingly enough

01:17:33 this reaction takes place in an almost

01:17:36 specific way it's at least very very very highly

01:17:39 stereoselective and in such

01:17:42 wise that at this

01:17:45 carbon atom the hydroxyl

01:17:48 is disposed as shown in the desired

01:17:51 configuration

01:17:54 whereas at

01:17:57 carbon atom number two

01:18:00 the stereochemistry of the methyl

01:18:03 group

01:18:06 is just the opposite of that

01:18:09 that one wants

01:18:12 so we've done pretty well here

01:18:15 we had one chance in two

01:18:18 in generating this center and we won

01:18:21 we had one chance in two in generating the other one

01:18:24 and for the moment we lost

01:18:27 but half the job was done and this small

01:18:30 minority was shall we say

01:18:33 adjusted

01:18:36 in the next stage which can be seen on the next slide

01:18:39 where we have this compound

01:18:42 with the right configuration at C3

01:18:45 and the wrong configuration at C2

01:18:48 now before going on

01:18:51 I perhaps should say that the sulfur

01:18:54 we used the atom sulfur again

01:18:57 and that this reaction was possible or this reaction sequence

01:19:00 that I'm now describing was possible only with the

01:19:03 thiol ester and not with a normal ester

01:19:06 as we'll see in a moment

01:19:09 or as I'll mention in a moment

01:19:12 we had the wrong configuration here how did we take care of that

01:19:15 well we simply made the anion at that center using in this case

01:19:18 mesityl lithium in the presence of

01:19:21 tetramethyl ethylenediamine

01:19:24 in tetrahydrofurane at minus 78 degrees

01:19:27 that generates the anion at the alpha position

01:19:30 to the carbonyl group of the thiol ester

01:19:33 and then that anion is quenched

01:19:36 using

01:19:39 2,6-ditertiary-butyl-4-methyl-phenol

01:19:42 in tetrahydrofurane and lo and behold when it's quenched

01:19:45 the methyl group comes out on the right side

01:19:48 as shown here

01:19:51 obviously we're very pleased with that

01:19:59 now this sequence

01:20:02 you see a very very simple sequence for adding the last

01:20:05 two asymmetric centers

01:20:08 it is very pleasing indeed and it deserves a couple of comments

01:20:11 as I said

01:20:14 it's not possible with a normal ester

01:20:17 as distinguished from a thiol ester

01:20:20 for the simple reason that

01:20:23 one can get to this stage

01:20:26 but then when one tries to make the anion

01:20:29 instead of getting the anion you just get dealdolization

01:20:32 and go back to the aldehyde which is no good at all

01:20:35 but the thiol ester contains a more activated hydrogen

01:20:38 that makes it possible

01:20:41 to prepare from this the anion at the alpha position

01:20:44 and do the quenching reaction which gives the correct

01:20:47 stereochemistry

01:20:50 now there's a rather amusing

01:20:53 story here

01:20:56 I've alluded a number of times to the fact

01:20:59 that this chain is one which

01:21:02 has methyl groups at every other one

01:21:05 of the carbon atoms

01:21:08 along the main chain

01:21:11 and of course that is a consequence

01:21:14 of the fact that the biogenesis

01:21:17 of erythromycin involves

01:21:20 the combination of seven propionic acid units

01:21:23 there's quite a bit known about that

01:21:26 biogenesis but the fundamental feature of it is

01:21:29 that essentially seven propionic acid units

01:21:32 are built up into a chain

01:21:35 and that of course these

01:21:38 methyl groups represent each one of them

01:21:41 one of the propionic acid units

01:21:44 by which

01:21:47 the erythromycin chain is built up

01:21:50 in nature. Now not long ago

01:21:53 Professor Dorothy Hodgkin was

01:21:56 visiting Cambridge and we had a pleasant conversation

01:21:59 about chemistry and crystallography and she said

01:22:02 Bob does your method for building up the erythromycin chain

01:22:05 resemble that

01:22:08 used in nature at all. I said oh no no no

01:22:11 it has nothing to do with it whatsoever

01:22:14 and then being suspicious of everything I say

01:22:17 I thought to myself can that be true

01:22:20 and I thought well really it isn't true at all

01:22:23 the last step here of course is almost

01:22:26 biomimetic if you wish since we are in fact using

01:22:29 propionic acid derivatives to add the last two centers

01:22:32 and for that matter a thioester and it's thioesters

01:22:35 that are used in nature. So quite inadvertently

01:22:38 this very simple solution to the

01:22:41 final stereochemical problem

01:22:44 could be called if you want biomimetic

01:22:47 but I didn't think of that until after we'd done it.

01:22:50 Alright now

01:22:53 you might be justified

01:22:56 in saying that some of the stereochemical points

01:22:59 that I have presented to you

01:23:02 and I call to mind that we've now built up the entire

01:23:05 chain and I have stated that we have

01:23:08 now all ten of the

01:23:11 asymmetric centers present

01:23:14 properly oriented in a relative sense as well as

01:23:17 in an absolute sense

01:23:20 you might regard that as

01:23:23 an argumentative statement

01:23:26 so of course we call upon Mrs. Reese and on the next slide

01:23:30 we show

01:23:33 just one of the pictures that she gave us

01:23:36 she carried out an x-ray crystallographic study

01:23:39 of the thioester which was on the preceding slide

01:23:42 and I dare say you'd have to study it for some time

01:23:45 to convince yourself that it represents the same thing that

01:23:48 was represented in planar form on the preceding slide

01:23:51 but it does. Mrs. Reese's determination

01:23:55 confirms the structure

01:23:58 of the thioester in all respects

01:24:01 and does show beyond any question

01:24:04 that we have succeeded in building up

01:24:07 the entire erythromycin main chain

01:24:10 with each of the ten asymmetric centers

01:24:13 oriented with respect to all the others

01:24:16 in a proper way in a relative sense

01:24:19 as well as in an absolute sense

01:24:23 Mrs. Reese by the way once she had the data

01:24:26 which took a little longer a few days

01:24:29 completed the actual determination in one day

01:24:37 on the next slide

01:24:43 I show again on the left the thioester

01:24:46 it's the same structure that was shown on the previous slide

01:24:49 we can now regard it as established

01:24:52 and just for comparison again

01:24:55 the monitor the Fisher diagram on the left

01:24:58 if you translate this structure into this one

01:25:01 you will find that all of the carbon atoms

01:25:04 two, three, four, five, six

01:25:07 eight, ten, eleven, twelve, thirteen

01:25:10 are properly substituted

01:25:13 and properly oriented

01:25:16 in all respects

01:25:19 and so the first phase of our work

01:25:22 at this point may be

01:25:25 described as completed

01:25:28 and we now must enter the

01:25:31 second phase

01:25:34 namely we must

01:25:37 now come to grips with

01:25:40 that which I mentioned earlier

01:25:43 we now have a derivative

01:25:46 of a 13-hydroxytridecanoic acid

01:25:49 properly substituted and oriented

01:25:52 at all the asymmetric centers

01:25:55 and we now must convert that hydroxy acid derivative

01:25:58 into the corresponding lactone

01:26:01 that involves making a bond

01:26:04 from that oxygen to this

01:26:07 carbonyl group over here

01:26:10 and probably some changes in nature

01:26:13 of protecting groups and so forth

01:26:16 I say our work has entered that phase

01:26:19 and we're actively engaged upon it at the present

01:26:22 I have nothing to report on it here this afternoon

01:26:25 I'm going to conclude with the construction

01:26:28 of the main chain and the solution of the

01:26:31 stereochemical problems associated with the main chain

01:26:34 we shall have to deal with the problems of lactonization

01:26:38 at some future time

01:26:41 now this work has gone on for

01:26:44 considerable time

01:26:47 it has been to me highly interesting and exciting

01:26:50 it has

01:26:53 involved many things that I haven't been able to tell you today

01:26:56 this has been a highly condensed version

01:26:59 you may regard that as a remark

01:27:02 out of place but it is a highly condensed version

01:27:05 of what we've done

01:27:08 and I've chosen to emphasize some of the high points

01:27:11 as I say this work

01:27:14 was carried out on a broad front

01:27:17 and of course it involves the devoted and skillful work

01:27:20 of my collaborators

01:27:23 I don't have anything to do about this

01:27:26 I just have the privilege of coming and speaking to you about it

01:27:29 and on the next slide

01:27:32 this is a kind of historical survey

01:27:35 of the collaborators

01:27:38 you see we started back in 72-73

01:27:41 and have proceeded

01:27:44 as I said on a broad front

01:27:47 and the very great skill and devotion of these men

01:27:50 are responsible for

01:27:53 such results as we have been able to

01:27:56 obtain

01:27:59 and I want to express here my thanks

01:28:02 to all of them

01:28:05 Dr. Blodichick is here in the audience

01:28:08 he's one of your boys and a very good one indeed

01:28:11 so I can express my appreciation to him in person

01:28:14 I admire very much the patience

01:28:17 the skill, the devotion

01:28:20 and the spirit that my collaborators have shown

01:28:23 in carrying out the work that I've told you about

01:28:26 and if you've enjoyed what I've had to say this afternoon

01:28:29 it is to them that you are indebted

01:28:32 so I thank you for your patience

01:28:56 I should just simply point out

01:28:59 that it is not traditional

01:29:02 in these named lectureships

01:29:05 to have a question period

01:29:08 I suppose if there are any questions

01:29:11 we might have some very informal ones up front

01:29:14 very briefly since we're dragging Bob off

01:29:17 but I do think we should conclude

01:29:20 by expressing special appreciation

01:29:23 for Bob Woodward coming west

01:29:26 it's about the only trip I think he's

01:29:29 making these days within the country

01:29:32 this is a superb lecture Bob

01:29:35 it's great to have you back again

01:29:38 and we'll hope to have Woodward again

01:29:41 long after all of you students are gone