00:00:01 ...director of the Medical Research Council Unit for Molecular Biology at Cambridge, England,

00:00:08 past chairman of the European Molecular Biology Organization,

00:00:14 and he holds honorary degrees of Doctor of Philosophy from the University of Vienna and the University of Edinburgh.

00:00:22 Dr. Perutz and I first met when we were both students at the University of Vienna,

00:00:29 where we shared an interest in chemistry, mountaineering, and skiing.

00:00:36 We both have kept up our three interests throughout the years,

00:00:42 and yesterday we had the pleasure of showing Dr. Perutz Mount Rainier in all its glory during a hike in Mount Rainier National Park.

00:00:52 This is Dr. Perutz's first visit to Seattle,

00:00:55 and we are indeed pleased and honored that we should have him as the first Philip E. Wilcox Memorial Lecturer.

00:01:04 His lecture is entitled, The Molecular Pathology of Human Hemoglobins.

00:01:09 Ladies and gentlemen, Dr. Perutz.

00:01:12 Thank you.

00:01:34 I'm not showing this until afterwards.

00:01:39 Ladies and gentlemen, I'm very pleased to be here in Seattle,

00:01:48 and greatly honored to be asked to give a lecture in memory of a biochemist who spent much of his working life

00:01:58 trying to design methods that would help protein crystallographers to solve protein structures

00:02:06 and to get to know more about the structure and mechanism of the active sites of enzymes.

00:02:19 I want to talk today about a different kind of problem, the one of congenital diseases

00:02:29 and the contributions which protein crystallography was able to make in gaining a deeper understanding of them.

00:02:38 As you know, these are a widespread cause of human suffering,

00:02:42 and by its discovery of the genetic code and of the apparatus used for translating the base sequence of DNA into amino acid sequence,

00:02:56 molecular biology has provided us with an intellectual framework for understanding the nature of genetic diseases.

00:03:07 It appears that point mutations in the genes coding for protein chains lead to amino acid substitutions in enzymes or other proteins

00:03:19 which render these proteins inactive, so inhibiting their normal function,

00:03:30 and that this leads to the breakdown of one particular catalytic mechanism in the organism, say.

00:03:41 Now, can we go further than this?

00:03:45 The difficulty is that, apart from hemoglobin, there is yet not a single human protein of which we know both the amino acid sequence and the three-dimensional structure.

00:03:58 So, the only one where we can try and get a picture of the kind of damage that mutations do to function,

00:04:16 the only protein available so far then, is hemoglobin.

00:04:20 Now, at the same time, more variants have been discovered for this protein than for any other,

00:04:27 so that a study of the hemoglobin diseases offers us a unique opportunity for getting a better understanding of genetic effects on protein structure generally.

00:04:41 So, what do we want to do?

00:04:44 In this talk, I shall try to show you first what hemoglobin looks like, how it works,

00:04:50 and then discuss the effects of certain mutations on its structure and function.

00:04:57 Now, hemoglobin is a tetramer made up of two alpha chains, each containing about 141 amino acid residues each,

00:05:11 and two beta chains, each containing 146 residues, and four hemes.

00:05:18 And, as you know, each of these hemes is capable of combining reversibly with molecular oxygen.

00:05:25 Now, let's first look at what kind of structure one of these subunits has.

00:05:37 Tell me, have you got a pointer somewhere?

00:05:46 Oh, I'm afraid I forgot to ask.

00:05:55 Oh, here comes somebody with a device.

00:06:09 Thank you.

00:06:10 Thank you.

00:06:22 Lovely. Thank you very much.

00:06:30 Can we have these front lights off, please?

00:06:35 Now that's better.

00:06:39 This picture shows the tertiary structure of one of these globin subunits.

00:06:49 Never mind whether alpha or beta, we'll discuss that later.

00:06:53 And here is its amino end, and up there its carboxyl end.

00:06:58 Here is its amino end, and up there its carboxyl end.

00:07:02 Now, to make the journey from one end to the other, we start here,

00:07:07 then go around an alpha helix of 16 residues, turn a right-angled corner, go around a complex sort of loop,

00:07:16 then travel down one helix and up another, go through a non-helical segment,

00:07:24 and then go around this hairpin bend, and finally we arrive at the other end.

00:07:31 Now, then, the beta chain is made of eight helical segments and the same number of non-helical ones.

00:07:41 The only difference in structure between the alpha and the beta chain

00:07:45 is a deletion of two residues in this corner in the beta chain,

00:07:52 a deletion of seven residues, the C helix in the alpha chain,

00:07:58 and then finally the beta chain has one additional residue here.

00:08:02 Otherwise, these two types of structures, despite their many amino acid substitutions, are closely similar.

00:08:10 Now, this model, this drawing here, shows only the alpha-carbon positions of the chain.

00:08:18 If you look at it in three dimensions, what strikes you immediately is the exclusion of non-polar residues from the interior.

00:08:26 The interior of the chain is oily or waxy.

00:08:30 It's filled with hydrocarbons like phenylalanine, leucines, valines, tryptophanes, tyrosines,

00:08:37 and the surface is studded with charge groups, aspartic and glutamic acids, lysines, arginines, and so on,

00:08:45 which is soluble in water.

00:08:48 So, you will see the great importance of this non-polar interior later when we come to the unstable hemoglobins,

00:08:56 and you will see also how important it appears to be that all gaps in the surface are plugged, that they are closed.

00:09:04 There must not be a hole in these loops through which water can get in.

00:09:11 As soon as it does, the structure becomes unstable.

00:09:15 Now, why is this?

00:09:19 If you look at other protein structures, you'll find, generally, that the loops of the chains are strongly cross-linked

00:09:28 by hydrogen bonds between beta-plated sheets or, say, by hydrogen bonds between tyrosines and aspartates,

00:09:37 by cysteine bridges, and so forth.

00:09:40 The hemoglobins are not unique but rare in that they lack all this.

00:09:47 There are no strong bonds linking these helical regions together.

00:09:53 They just stick together by Van der Waals interactions or non-polar ones, if you like,

00:09:59 so that the hemoglobin molecules are relatively soft as protein goes.

00:10:06 You will see this has other important consequences because any misfit that a mutation produces

00:10:15 can give rise to dislocations which run right through the molecule and produce functional disturbances.

00:10:23 Now, the heme fits into this waxy mass so that its non-polar ends stick inwards and its propionic acids outwards.

00:10:36 The heme, then, is in contact with about 60 atoms of the globin,

00:10:43 and all these are Van der Waals contacts except for the one covalent link to the heme-linked histidine,

00:10:50 residue F8, which you see there.

00:10:53 And then on the other side, there's another histidine which may or may not form a hydrogen bond with the oxygen.

00:11:01 We don't really know exactly what its purpose is.

00:11:05 And then you see here the penultimate residue is a tyrosine,

00:11:10 and this tyrosine in deoxyhemoglobin, at any rate, forms a hydrogen bond with a carbonyl group of this residue there,

00:11:20 which is a valine.

00:11:21 And that, you will see, is also a very important link for stabilizing the deoxy form of this structure.

00:11:33 Now, any other points that we ought to consider while we look at this picture?

00:11:40 I don't think so.

00:11:43 So let's now look at the next one and see how two of these chains combine.

00:11:51 So here we have an alpha chain and a beta chain forming an alpha-beta dimer,

00:11:59 such as you get if you dissociate hemoglobin in strong salt solution.

00:12:05 And you see that this dimer sticks together along these helical regions here, which are the helices G.

00:12:16 The amino end of the chain is now there, the carboxyl end of the alpha chain there.

00:12:21 The amino end of the beta chain here, this carboxyl end there.

00:12:25 And here where that marker is, there's a two-fold symmetry axis,

00:12:29 which means that a rotation by 180 degrees about this vertical axis

00:12:35 would bring these two subunits into congruence with the other two that lie in front.

00:12:42 Now, between them lies a cavity, which is filled with water.

00:12:49 And between the beta chains, especially in the deoxy structure,

00:12:53 you will see there's a large hole into this 2,3-DPG fits.

00:12:59 So the next slide then shows you the complete molecule, two alpha chains, two beta chains.

00:13:13 And the form that this takes is the oxy structure,

00:13:20 where there's hardly any contact between the two alpha chains or the two beta chains,

00:13:30 but the tetramer is held together by contacts between alpha and beta here and there.

00:13:40 So there are two different kinds, alpha 1, beta 1, we call them, and alpha 1, beta 2.

00:13:48 Right. Now, what's the good of this molecule?

00:13:54 What purpose does it serve, and why does it have to be so complicated?

00:13:58 Why doesn't a single chain, say the beta chain,

00:14:03 why isn't that enough as an oxygen carrier in the red cell?

00:14:08 The answer is to be found—can we have the lights on for a moment, please?

00:14:14 The answer is to be found in the complex cooperative properties which hemoglobin has

00:14:22 and which are needed for it to fulfill its respiratory function.

00:14:27 As you know, the binding of oxygen of a hemoglobin solution is cooperative

00:14:32 in the sense that the oxygen affinity rises with increasing oxygen saturation,

00:14:39 and so it falls with decreasing oxygen saturation,

00:14:45 which seems a truism, but I say it because that is the physiologically important point.

00:14:53 It's easy to have an oxygen carrier that gets fully saturated with oxygen in the lungs,

00:15:00 where oxygen is plentiful,

00:15:03 but because the difference in partial pressure of oxygen between the lungs and the tissues

00:15:09 is only a factor of 10, it's very difficult to design an oxygen carrier

00:15:14 which will also dissociate a significant fraction of the oxygen it carries in the tissues.

00:15:22 And the cooperative properties of the hemoglobin, the sigmoid oxygen equilibrium curve,

00:15:29 allows for this.

00:15:31 So it allows a much larger fraction of the oxygen to be dissociated

00:15:36 than a hyperbolic oxygen dissociation curve such as you would get from a single chain would get.

00:15:43 So now there are several other curious properties.

00:15:48 The oxygen affinity of hemoglobin falls with falling pH.

00:15:54 It also falls with increasing concentration of various inorganic anions,

00:16:03 and especially phosphates, and especially certain organic phosphates,

00:16:11 and end of CO2.

00:16:20 So these chemically completely different agents, hydrogen ions, organic phosphates, and CO2,

00:16:31 all produce the same physiological effect.

00:16:35 They lower the oxygen affinity, and that seems like an extremely baffling property to any chemist.

00:16:42 How can three such different things produce the same effect?

00:16:49 Now, what purpose do they serve?

00:16:53 All these mechanisms help to make the discharge of oxygen yet more efficient.

00:17:02 So the presence of lactic acid and of CO2 in the tissues helps the discharge of oxygen.

00:17:11 Conversely, the uptake of hydrogen ions by hemoglobin on loss of oxygen

00:17:17 neutralizes the carbonic acid formed on combination of water and CO2,

00:17:23 and so helps to bring the CO2 into solution in the form of bicarbonate,

00:17:28 in which it can be transported back to the lungs.

00:17:31 And, of course, again, the lowering of the oxygen affinity by CO2 helps the discharge of oxygen.

00:17:39 And finally, when we come to organic phosphates, there's one specific one in human red cells,

00:17:46 2,3-diphosphoglycerate, which acts as a regulator of the oxygen affinity,

00:17:51 so that under stress, through loss of blood or going to high altitude,

00:17:59 there's a feedback mechanism producing more 2,3-diphosphoglycerate,

00:18:04 which allows for better discharge of oxygen.

00:18:11 It's the combination of these remarkable cooperative effects,

00:18:15 which have made hemoglobin such a fascinating field of study

00:18:19 for physiologists and biochemists and even physicists,

00:18:24 and these aspects, the mechanism of these, I shall be discussing in the subsequent lectures.

00:18:33 But in this talk, I just want to give you a brief outline of how it works,

00:18:38 because we need to know this in order to understand

00:18:42 how the abnormal hemoglobins interfere with the workings of hemoglobin.

00:18:48 Now, all these cooperative effects disappear if you dissociate hemoglobin into alpha-beta dimers.

00:18:57 So they're a specific property of the tetramer,

00:19:02 and not shown by either the alpha chains or the beta chains

00:19:07 or a combination of one alpha and one beta chain.

00:19:10 And that seems remarkable, that, for instance,

00:19:13 the oxygen affinity of the dimer should be independent of pH,

00:19:17 but when you assemble it into a tetramer, it becomes pH-dependent.

00:19:22 And the explanation is this, that hemoglobin exists in two different forms,

00:19:29 an arterial one with a high oxygen affinity

00:19:34 and a low affinity for hydrogen ions, CO2 and 2,3-DPG,

00:19:41 and a venous form in which these affinities are reversed.

00:19:46 So not only the color of hemoglobin changes when you go from the arteries to the veins,

00:19:52 the structure actually changes.

00:19:56 So, how does it change?

00:20:00 Let's have that slide again now, please.

00:20:08 The easiest thing to see is the change of what protein chemists call the quaternary structure

00:20:15 that is the arrangement of the four subunits relative to each other

00:20:19 and to the twofold symmetry axis.

00:20:22 So if this is oxyhemoglobin, if you take the oxygen off,

00:20:27 these two subunits move this way,

00:20:34 so they swing out forward and to the left,

00:20:37 and the other two subunits go that way,

00:20:41 they swing out backwards and to the right.

00:20:44 So on deoxygenation, the distance between this ion atom and the one at the back

00:20:50 increases by 7 angstroms.

00:20:53 But very important, in this contact here,

00:21:00 you see the white chain moves backwards,

00:21:04 the black chain moves forwards,

00:21:07 so that there's a large relative movement between these two subunits,

00:21:12 and you will see in a moment how the contact here is dovetailed

00:21:18 and designed very cleverly so that it can click backwards and forwards

00:21:24 only between two alternative positions.

00:21:27 Now, before showing you that,

00:21:30 I want to point one or two other things out in the model

00:21:35 so that you can visualize it when I show it in diagrammatic form.

00:21:39 I told you that in the oxy structure, there are no links between the alpha chains,

00:21:44 but if you go to deoxy, then these C-terminal arginines form salt bridges

00:21:51 with complementary groups on the other alpha chain.

00:21:54 Similarly, well, in the internal cavity,

00:22:00 there are no links between the beta chains,

00:22:03 but as you will see in the deoxy structure,

00:22:07 they are cross-linked by 2,3-DPG.

00:22:10 And then I'll be showing you an interesting contact here

00:22:16 between the C-terminal histidine of the beta chain

00:22:20 and a lysine which comes down from this alpha chain

00:22:25 and an aspartate which sits just there.

00:22:29 And it's this contact here which is responsible

00:22:36 for at least half the pH dependence of the oxygen affinity.

00:22:41 So, there we go, then.

00:22:43 The first thing I'm going to show you is this contact

00:22:46 in very diagrammatic form looked at from the top.

00:22:49 Right. I can have the next slide now.

00:22:53 So there it is.

00:22:55 Here it is in the deoxy structure,

00:22:58 and you see it's dovetailed, as I've said,

00:23:01 and there are about 34 amino acids in contact there,

00:23:08 and these contacts are all Van der Waals interaction.

00:23:11 But there's one important hydrogen bond

00:23:14 between a tyrosine in the alpha chain

00:23:17 and an aspartate in the beta chain,

00:23:20 and you will see later on the role it plays.

00:23:24 So, when the oxygen comes on

00:23:27 and the subunits click over to the other form,

00:23:31 this gets broken, the dovetailing clicks to the other side,

00:23:36 and now this asparagine joins up with that aspartate,

00:23:40 as you see down there.

00:23:44 So, this is one change.

00:23:47 Now, the next slide shows you the change in that corner

00:23:53 by the C-terminus of the beta chain,

00:23:55 the bottom left-hand corner of the slide today,

00:23:58 of the whole model that you looked at a moment ago.

00:24:02 So, what I've tried to draw here

00:24:07 is a sort of composite

00:24:10 where I've shown those regions

00:24:13 which hardly change their structure in white.

00:24:17 Now, the positions of the C-terminal residues

00:24:21 in oxyhemoglobin in red

00:24:24 and in deoxyhemoglobin in blue.

00:24:26 So then, let's move this to one side

00:24:32 in case you can't see it all.

00:24:36 Here's the ion atom and the helix to which it is linked,

00:24:40 and here's the reactive sulfadryl group.

00:24:43 Now, here's the C-terminal helix, which we call H,

00:24:47 and this red part indicates that in oxyhemoglobin,

00:24:53 these two C-terminal residues are mobile.

00:24:57 They don't always probably stick out like that.

00:25:00 Some of the time, the cyrosine is probably

00:25:02 sort of half in this pocket,

00:25:05 and for some of the time,

00:25:07 this probably forms a salt bridge

00:25:10 with the N-terminal amino group of the neighboring beta chain,

00:25:14 but certainly the imidazole here is free in the oxy form.

00:25:19 Now, on deoxygenation, several new bonds are formed.

00:25:25 First, this tyrosine forms a hydrogen bond

00:25:29 with a carbonyl group,

00:25:31 which I showed you actually in the very first slide,

00:25:34 with a carbonyl group of valine Fg5,

00:25:37 and is firmly fixed in the pocket between helices F and H.

00:25:42 Then this C-terminal carboxyl forms a salt bridge

00:25:46 with the lysine of 40α,

00:25:49 which sort of swings in on deoxygenation,

00:25:53 and then there's this salt bridge formed

00:25:56 between the imidazole and the aspartate.

00:25:59 Now, that gives the clue to the pH dependence.

00:26:06 The imidazole is a weak base,

00:26:09 with a pKa normally of about 7, or slightly below.

00:26:14 So, in this form, in oxyhemoglobin,

00:26:18 at physiological pH,

00:26:21 more than half of these histidines would be uncharged.

00:26:28 But when you combine this weak base with a strong acid,

00:26:34 it becomes a stronger base,

00:26:36 and so it takes up hydrogen ions,

00:26:38 becomes positively charged,

00:26:40 so that in deoxyhemoglobin,

00:26:42 its pKa rises to 8.1,

00:26:45 which means that at physiological pH,

00:26:48 nine-tenths of all these histidines

00:26:51 would carry a positive charge.

00:26:54 Now, what does this mean?

00:26:56 If you make the solution more acid,

00:27:01 then a greater fraction of these histidines

00:27:05 would be positively charged,

00:27:07 so that this bond would be strengthened.

00:27:10 Conversely, it means that these histidines

00:27:14 take up hydrogen ion when the oxygen comes off,

00:27:18 and so, as I said, neutralize the hydrogen ion,

00:27:22 set free in the reaction of CO2 and water,

00:27:25 and help the transport of carbon dioxide.

00:27:28 So then, so much for—

00:27:31 and this whole complex effect,

00:27:34 as all physiologists, biochemists, you know,

00:27:37 is known as the Bohr effect.

00:27:39 Now, the next slide shows you

00:27:43 the sort of scheme of these salt bridges

00:27:51 in a very diagrammatic way.

00:27:53 Here you see the oxy structure,

00:27:56 in which, say, the C-terminal—

00:28:02 the C-terminal arginines of the chains would be free.

00:28:07 The tyrosines would not be hydrogen-bonded.

00:28:10 The C—sorry, C-terminal arginines,

00:28:13 I should call these here.

00:28:15 Their complementary groups would be there.

00:28:18 The C-terminal histidines of the beta chains are free,

00:28:22 and the hemes are flat.

00:28:25 They've got oxygen bond.

00:28:27 Now, when the oxygen comes off,

00:28:29 in the next slide,

00:28:33 the C-terminal arginines are cross-linked

00:28:37 to complementary groups on the other alpha chains,

00:28:40 as you see there.

00:28:41 Here, lysine-42-alpha comes down,

00:28:45 forms a salt bridge with the C-terminal carboxyl.

00:28:48 Here's the histidine linked to an aspartate,

00:28:52 and there is the 2,3-DPG,

00:28:55 which you'll see in detail now in the next slide.

00:29:00 This is Arthur Arnon's marvelous drawing

00:29:03 of the stereochemical complementarity

00:29:07 between this regulator and the two beta chains.

00:29:11 So here is the amino terminus of one beta chain.

00:29:15 Here is its EF corner,

00:29:17 and there's the symmetrically related one.

00:29:20 Okay?

00:29:21 So here in the middle, on the 2,4-symmetry axis,

00:29:25 sits the 2,3-DPG,

00:29:27 so that its phosphates can form salt bridges

00:29:31 with the alpha-amino group,

00:29:33 with histidine-2,

00:29:35 and with histidine-143,

00:29:38 and its carboxyl group there

00:29:42 can combine with a lysine,

00:29:46 as you see here.

00:29:50 So there's a stereochemical and electrical complementarity

00:29:56 between 2,3-DPG and the globin chains.

00:30:00 Now, on oxygenation,

00:30:02 these alpha-amino groups move further apart,

00:30:05 and these groups move closer together

00:30:07 so that the 2,3-DPG is squeezed out.

00:30:10 And you'll see that in the next slide,

00:30:13 which is at low resolution,

00:30:15 shows you in plane lines

00:30:18 the positions of helices H and A

00:30:22 in deoxy and in the dotted curves in oxyhemoglobin.

00:30:27 And you see how these complementary groups

00:30:31 pull apart here, here, and here,

00:30:34 and move together there and there

00:30:36 so that the regulator is squeezed out.

00:30:39 Now, as you see, this is a marvelous model

00:30:42 of the way any metabolic regulator

00:30:45 might act on an allosteric enzyme,

00:30:49 or, say, an inducer might act on a genetic repressor,

00:30:56 changing the quaternary structure of the protein

00:30:59 by its complementarity

00:31:02 to one of two alternative allosteric forms,

00:31:07 and so changing the catalytic function

00:31:12 or repressor action,

00:31:14 or whatever it may be in this structure.

00:31:19 Now, can you switch backwards and forwards

00:31:24 between two slides?

00:31:25 If you can, can I have the last one back again, please?

00:31:31 Thank you.

00:31:32 I thought while I'm about it,

00:31:33 I might tell you about the new abnormal hemoglobin,

00:31:37 news of which is to appear in Nature next Saturday.

00:31:42 And this is hemoglobin little rock,

00:31:46 in which this histidine here is replaced by glutamine.

00:31:52 And that, as you may guess,

00:31:56 has an increased oxygen affinity

00:31:58 because its affinity for 2,3-DPG is lowered.

00:32:05 So this is what you would have expected,

00:32:08 but the authors were very puzzled

00:32:10 because it has an increased oxygen affinity

00:32:14 even in the absence of phosphate,

00:32:19 even in strict hemoglobin.

00:32:21 And can I have the next slide back again?

00:32:24 The explanation for this was that the glutamine,

00:32:29 which sits here in oxy,

00:32:32 can form a hydrogen bond with an asparagine,

00:32:35 which comes out from the opposite H helix of the beta chain,

00:32:41 so that this mutation, you see,

00:32:44 raises the oxygen affinity by two mechanisms.

00:32:48 One is its diminished affinity for 2,3-DPG.

00:32:52 The other is a stabilizing interaction,

00:32:55 which links the two beta chains together in the oxy,

00:32:59 in the high affinity structure,

00:33:01 so that on loss of oxygen,

00:33:04 the bond energy of two additional hydrogen bonds

00:33:08 first has to be overcome.

00:33:11 Right.

00:33:14 Now, I think I've discussed sort of in a general way

00:33:28 what the change of structure implies,

00:33:33 but I haven't said a word about the way how it's brought about.

00:33:37 How does the oxygen manage to click the structure

00:33:40 from one form to another?

00:33:42 And, of course, the obvious answer seemed to be

00:33:45 that it must be something to do

00:33:47 with the spatial effect of the oxygen,

00:33:49 but it turns out that this is only part of the story,

00:33:52 and there's another important mechanism at play,

00:33:55 to which I shall just draw your attention now,

00:33:58 and which I shall talk about at length

00:34:01 in my subsequent lectures, and that is this.

00:34:05 Can I have the next slide, please?

00:34:08 That by a strange quantum transition,

00:34:14 the size of the ion atom changes

00:34:17 depending on whether it's coordinated

00:34:20 to six nearest neighbors, as in oxyhemoglobin,

00:34:24 or to only five, as in deoxyhemoglobin.

00:34:28 And, in fact, the ion atom swells.

00:34:32 The ion-nitrogen bond distances increase

00:34:35 when the oxygen comes off,

00:34:37 and, as a result, well, the heme is so designed

00:34:42 that the six-coordinated ion just fits into the hole

00:34:46 between the four nitrogen atoms,

00:34:48 but the five-coordinated ion atom does not.

00:34:51 So, as a result, when the oxygen comes off,

00:34:53 the ion is squeezed out of the plane of the porphyrin,

00:34:56 and the hemelink histidine is pushed away

00:35:00 from the plane of the porphyrin ring

00:35:03 so that the distance increases from 2 to 2.9 angstroms.

00:35:07 And it is this, apparently, which provides

00:35:10 the main leverage for changing the structure

00:35:14 from the oxy to the deoxy form.

00:35:19 Now, these abnormal hemoglobins.

00:35:24 Shall we have light for a moment, please?

00:35:34 Some of you may be surprised

00:35:36 that about 180 different variants are known today.

00:35:41 This is because the analysis

00:35:50 of an electrophoretic pattern of the hemoglobins

00:35:54 is now routine in many hospitals

00:35:56 where some unexplained hematological symptoms are seen.

00:36:01 And what are they all?

00:36:07 In most of them, there's just a single amino acid

00:36:14 in either the alpha or the beta chain replaced by another.

00:36:19 And all these substitutions can be accounted for

00:36:23 by single base changes in the messenger RNA

00:36:27 coding for the globin chain,

00:36:30 assuming that the genetic code derived from E. coli

00:36:36 is also applicable to man.

00:36:38 So the first interesting and biologically important result

00:36:44 is this, that the abnormal hemoglobins

00:36:47 provide the strongest evidence we yet have

00:36:50 for the universality of the genetic code,

00:36:53 for the code being the same in microorganisms

00:36:56 as it is in man.

00:36:59 Now, other variants are deletions

00:37:03 of one or more amino acids

00:37:05 or additions of amino acids at the end of the chain,

00:37:08 such as hemoglobin constant strings,

00:37:11 constant spring,

00:37:13 which has an additional 31 residues on the alpha chain

00:37:17 due to a mutation chain termination residue.

00:37:21 And then there are crossovers.

00:37:26 There's a minor component of beta-like chains,

00:37:32 the delta chains,

00:37:33 at the locus which is closely linked to the beta chains.

00:37:37 And you can get variants due to crossover

00:37:40 between beta and delta,

00:37:41 so that the chain, say, has a delta-like sequence

00:37:45 in the first half

00:37:46 and a beta-like sequence in the second half,

00:37:49 or vice versa.

00:37:52 Now, clinically, the most important of the abnormal hemoglobins,

00:37:56 as you know, is sickle cell hemoglobin

00:37:58 because it is the one that presents

00:38:04 the most serious health problem.

00:38:06 It causes the deaths of about 80,000 children a year.

00:38:13 And many people are working on ways

00:38:21 in which these deaths could be prevented.

00:38:24 Now, sickle cell disease

00:38:28 is a disease inherited in a strictly Mendelian manner,

00:38:33 and it is due to the replacement

00:38:36 of glutamic acid, 6-beta,

00:38:41 at the surface of the molecule by a valine.

00:38:50 This replacement has remarkable consequences.

00:38:54 It renders the deoxygenated form of the hemoglobin insoluble,

00:38:59 but it has no effect on the properties of the oxygenated form.

00:39:05 So this results in a precipitation of the hemoglobin in the red cell

00:39:16 every time it is deoxygenated,

00:39:18 and a resolution when it is oxygenated.

00:39:22 The precipitate, in fact, turns out to be semi-crystalline,

00:39:28 so that the deoxy form crystallizes

00:39:31 in the form of thin, long fibers,

00:39:34 which, as you will see in a moment,

00:39:37 grow like bamboo shoots,

00:39:39 extending right through the length of the red cell

00:39:41 and pulling it out into bizarre shapes.

00:39:46 In my next slides,

00:39:48 I'll show you some electron micrographs.

00:39:51 First, some thin sections,

00:39:54 taken by Jack Gertels and Johanna Dobler

00:39:59 at St. Luke's Hospital in New York.

00:40:01 So here you see a transverse section

00:40:03 through a sickle-grade cell,

00:40:05 and I hope that at least those who sit fairly near

00:40:10 can see that it's full of black dots.

00:40:13 Now, these dots are about 180 angstroms apart,

00:40:18 and, as you will see, represent sections

00:40:22 through these long fibers,

00:40:24 which are close-packed in the red cell.

00:40:26 They fill it entirely, as you see.

00:40:29 And then you see fibers running,

00:40:31 wrapping themselves around the surface of the cell.

00:40:34 Now, the next slide, you see a longitudinal section,

00:40:38 and here you see these fibers running

00:40:40 right from one end to the other.

00:40:42 And then, at the edges, you see the little dots.

00:40:46 So those are the fibers which, you see,

00:40:49 wrap themselves around the cell that way.

00:40:52 Now, Beatrice Magdov, also in New York,

00:40:58 got some beautiful X-ray fiber diagrams

00:41:02 of the sickle-cell precipitate

00:41:06 and also of red cells,

00:41:08 which showed a repeat of pattern

00:41:10 along the cells every 64 angstroms.

00:41:13 And John Finch and I, in Cambridge,

00:41:18 wondered what the mode of aggregations

00:41:22 of the hemoglobin molecules might be

00:41:26 which gives rise to the growth of these fibers.

00:41:31 So the first experiment we did was this.

00:41:36 Can I have the next slide, please?

00:41:40 We took a 10% solution of hemoglobinase,

00:41:45 put it in a nitrogen-filled glove box,

00:41:48 deoxygenated it,

00:41:50 put a drop of the solution on an EM grid,

00:41:54 and dropped, washed it,

00:41:57 and dropped a little phosphotangstic acid on it

00:42:01 as a negative stain.

00:42:03 And this is the picture which we obtained,

00:42:06 where the high magnification,

00:42:10 where the little white blobs

00:42:13 represent hemoglobin molecules.

00:42:15 And here you see a characteristic,

00:42:18 rather regular pattern,

00:42:20 where you can actually see three rows of black dots.

00:42:24 And you can also see some...

00:42:26 I hope you can see this at the back.

00:42:29 Can you...

00:42:31 The striations that run across...

00:42:33 Can you see... Is that visible?

00:42:36 Oh, good.

00:42:38 I was afraid it would be too small, you see.

00:42:42 So you see these striations,

00:42:44 and they are 62 angstroms,

00:42:47 so correspond to the repeat

00:42:49 that these people saw in the X-ray diffraction work.

00:42:54 Now, one thing which you see occasionally

00:42:57 is the fraying of single filaments.

00:43:02 So, suggesting that the primary mode of aggregation

00:43:06 actually is an aggregation of single hemoglobin molecules

00:43:10 to form a string of beads, as it were.

00:43:14 Now, John Finch wasn't satisfied with this picture

00:43:20 because the thickness of the fibers seemed quite irregular,

00:43:24 whereas in the red cell pictures of Bertels,

00:43:27 they were absolutely regular.

00:43:29 So, the next thing he tried

00:43:32 was to take a suspension of red cells

00:43:35 and actually lyse them on the EM grid

00:43:38 with hypotonic saline.

00:43:40 And that gave much better results.

00:43:43 So the next picture shows you...

00:43:46 Next slide, please.

00:43:48 ...shows you this negatively stained preparation

00:43:52 of the fibers,

00:43:54 now beautifully uniform in thickness, as you see,

00:43:58 and corresponding in diameter exactly

00:44:02 to the ones Bertels had seen.

00:44:05 Again, you see these characteristic striations

00:44:08 at 62 angstroms,

00:44:10 and then you see an interesting alternation.

00:44:14 Here, the fiber consists of two rows,

00:44:18 three rows, two rows,

00:44:20 three rows, two rows, and so on.

00:44:23 And this alternation of twos and threes

00:44:26 suggested a helical structure.

00:44:29 But the difficulty is, if you just see this,

00:44:33 the analysis is rather subjective.

00:44:36 Now, Klug and his colleagues in Cambridge

00:44:40 have developed an objective method

00:44:43 for analyzing such electron micrographs

00:44:47 by...

00:44:50 by getting the optical diffraction picture

00:44:56 of one of these fibers.

00:44:58 So you make yourself a little mask

00:45:01 in which you screen off everything

00:45:04 except one length of fiber,

00:45:06 and then put it in a little diffraction machine,

00:45:09 illuminate it with a laser,

00:45:11 and the next slide shows you what you obtain.

00:45:16 You see, you get a picture

00:45:18 with distinct maxima in the diffraction pattern.

00:45:24 So, what does this mean?

00:45:27 These two spots are at a spacing of 1 over 62 angstrom,

00:45:32 so they just correspond to the striation,

00:45:35 which you see anyway.

00:45:37 And that's nothing new.

00:45:39 But the telling ones are these four reflections here,

00:45:44 because their distance apart

00:45:48 tells you the repeat of pattern

00:45:51 along the helical repeat along the fiber axis,

00:45:55 and their distance from the center

00:45:58 tells you the radius from the center of the fiber

00:46:04 to the center of mass of the protein.

00:46:07 So, with the help of this information,

00:46:12 it was possible to construct a model

00:46:16 of the sickle cell precipitate,

00:46:20 which you see in the next picture.

00:46:23 So there it is,

00:46:25 and it consists of six filaments

00:46:31 of single hemoglobin molecules

00:46:33 wrapped around a hollow tube.

00:46:37 And the repeat from this molecule

00:46:41 to one that lies exactly over the top of it

00:46:49 is 520 angstroms,

00:46:53 and the repeat between layers is 62 angstroms.

00:46:56 So here you see the plan.

00:46:58 Another way of looking at it is

00:47:00 that you can think of it as a succession of hexagonal rings,

00:47:04 each ring being displaced relative to its predecessor

00:47:07 by 7.3 degrees.

00:47:14 So, then,

00:47:17 see, the exciting thing about this structure is this.

00:47:21 If we could resolve the nature of the contacts

00:47:24 between the subunits,

00:47:27 then we might be able to design some drug, say,

00:47:31 which competes with the site here

00:47:34 and actually prevents the aggregation.

00:47:37 So, we then ask ourselves,

00:47:43 what are the sites of contact

00:47:45 where these molecules stick together?

00:47:48 We don't know yet,

00:47:50 but the optical dichroism gives us a clue

00:47:54 at least to the orientation of the molecules,

00:47:57 and that you see in the next picture.

00:48:00 Now, you'll see a hemoglobin model

00:48:03 oriented in the way it would be in the fiber.

00:48:07 So, there are the alpha and here are the beta chains,

00:48:10 and the hemes are approximately normal to the fiber axis,

00:48:14 and so the next molecules would be there and there.

00:48:19 But now, where are the sickle cell sites?

00:48:22 One is here, and the other one is there.

00:48:26 And in this picture, it sort of looks okay.

00:48:29 You think, all right,

00:48:31 this valine might stick to some site in the next molecule,

00:48:39 but in fact, this is only because the photograph is foreshortened.

00:48:44 If you look at the actual model,

00:48:46 the valines are in here and there, you see,

00:48:50 because these things come further in,

00:48:53 and it's quite impossible to bring them into contact

00:48:57 with the next molecule above or below.

00:49:00 So, clearly, if these play a part,

00:49:04 they must be linking the molecules within the hexagon side by side,

00:49:11 and there is an observation which suggests that this is really so,

00:49:15 because having sort of found this,

00:49:18 John Finch and I wondered how unique this precipitate was.

00:49:23 So he examined oxygenated normal cells, deoxygenated normal cells,

00:49:30 oxygenated sickle cells, oxygenated and deoxygenated fetal cells,

00:49:36 and they were all empty except deoxygenated normal cells,

00:49:43 in which he found fibers looking exactly like

00:49:51 the sort of aggregates of sickle cell filaments

00:49:55 that we had seen in free solution,

00:49:58 only at a much lower concentration.

00:50:01 So the next picture shows you this.

00:50:04 Can I have...

00:50:06 Oh, can you push it up, please?

00:50:10 Something's gone wrong.

00:50:13 No, still down there.

00:50:17 Well, I'll tell you in the meantime what it is.

00:50:20 You find aggregates of these filaments in the normal cells,

00:50:25 but they are irregular.

00:50:28 You don't get the regular hexagonal structure

00:50:35 that the hemoglobin S fibers form.

00:50:39 So telling you that hemoglobin A as such also has a tendency

00:50:44 to aggregate end-to-end,

00:50:46 but apparently not having the stabilizing interaction

00:50:50 to form the 6-fold fiber, it's much less soluble.

00:50:54 But look, never mind, let's go on to the next slide,

00:50:57 if this is so recalcitrant.

00:51:00 Yes.

00:51:02 So, now, I spent a lot of time on sickle cell hemoglobin

00:51:06 because of its great importance,

00:51:08 but now I want to tell you about some of the others,

00:51:15 which admittedly occur mainly in isolated families,

00:51:19 in heterozygotes,

00:51:21 but they are most illuminating about the sort of interplay

00:51:25 between protein structure and function.

00:51:29 Perhaps I should still say this,

00:51:32 that while the sickle cell gene is very frequent

00:51:37 in some parts of West Africa,

00:51:39 it occurs in about 40 percent of the population,

00:51:43 there are only three other variants

00:51:46 which are sufficiently frequent to occur in homozygotes,

00:51:50 and none of these produce severe disease.

00:51:58 All the other 180-odd are rare,

00:52:03 occur only in isolated families,

00:52:05 so that the probability of finding heterozygotes,

00:52:09 finding homozygotes, is very small.

00:52:12 But some of them produce severe clinical symptoms

00:52:17 even in heterozygotes,

00:52:19 and I shall now show you some of the reasons why.

00:52:24 Now, the first group,

00:52:29 go back to this argument

00:52:35 about the nonpolar interior of the globin chains,

00:52:39 which has to be sealed off from water

00:52:42 in order to function properly.

00:52:44 So, there is a mutation, hemoglobin Hammersmith,

00:52:50 in which a phenylalanine,

00:52:52 which hangs down here and seals off the heme pocket,

00:52:56 is replaced by a serine that leaves it open.

00:52:59 And this causes an instability of the beta chains,

00:53:05 causes the globin to be denatured in the red cell

00:53:10 and precipitate in the form of Heinz bodies,

00:53:12 which make the cell rigid,

00:53:14 and thereby liable to early destruction.

00:53:19 So that it results in a severe anemia.

00:53:22 They actually have a girl,

00:53:24 I think about an 18-year-old girl,

00:53:27 at Hammersmith, who has this disease.

00:53:29 She has a hemoglobin of only 6%.

00:53:32 Her urine is black because the hemolytic anemia is so severe.

00:53:39 But nevertheless, apparently, she's relatively fit

00:53:43 and actually goes swimming and plays tennis,

00:53:46 which is quite remarkable.

00:53:50 Now, another mutant of the same sort of physiological effect is this.

00:53:59 Not very well visible here.

00:54:02 It's a close contact between the helices B and E.

00:54:08 This one and that one.

00:54:10 There's one point where they're so close together

00:54:13 that there's not room for a side chain between them.

00:54:18 So both the residues of the contact are glycines.

00:54:22 In hemoglobin savannah, one of these glycines,

00:54:25 in position B24, is replaced by a valine,

00:54:30 which has a bulky side chain

00:54:32 and obviously squeezes these two helices apart.

00:54:38 And again, you have the same symptom,

00:54:40 enhanced by this severe hemolytic anemia.

00:54:45 Now, another very interesting mutation is hemoglobin Wien for Vienna.

00:54:51 In here, in this crevice, lies a tyrosine,

00:54:54 130, which is part of the helix H,

00:54:57 whose phenyl group is in...

00:54:59 whose sort of benzene ring is internal,

00:55:02 but the phenolic hydroxyl makes a hydrogen bond just outside.

00:55:05 Now, in hemoglobin Wien, this is replaced by aspartate,

00:55:09 which is short and produces a polar group

00:55:12 in the nonpolar interior of the protein,

00:55:14 and this produces a severe instability

00:55:18 so that the protein precipitates.

00:55:21 So the moral then is...

00:55:23 Well, there are others.

00:55:24 There's another mutation where a leucine is replaced by an arginine,

00:55:28 which leaves the holes through which water can get in.

00:55:31 So the moral of all this is

00:55:34 if you let water into the interior,

00:55:38 you produce a deleterious mutation

00:55:41 which affects the stability and function of the protein.

00:55:46 Now, quite another class of proteins

00:55:49 producing hematological symptoms

00:55:52 are those that affect the oxygen affinity,

00:55:55 and they do this often

00:55:57 merely by affecting the relative stability

00:56:01 of the venous and the arterial form,

00:56:03 and I'll show you an interesting example.

00:56:05 Now, the next slide

00:56:08 shows us a diagram you've seen before

00:56:12 with these alternative positions

00:56:15 of the bond between the contact

00:56:18 between the alpha and the beta subunit

00:56:21 and these hydrogen bonds.

00:56:23 Now, in hemoglobin Kempsi,

00:56:25 this aspartate is replaced by an asparagine

00:56:29 so that this single hydrogen bond here

00:56:33 becomes inactive.

00:56:38 As a result,

00:56:41 this hemoglobin crystallizes in the oxy form

00:56:45 even when you have completely deoxygenated it

00:56:49 because you have shifted the equilibrium

00:56:53 of these two allosteric forms

00:56:56 by the equivalent of two hydrogen bonds

00:56:59 which may be, say, 6 kilocalories,

00:57:01 something of that order.

00:57:03 We don't know exactly what.

00:57:05 There's another one

00:57:08 should be close to your local heart's

00:57:13 hemoglobin Yakima,

00:57:15 in which this is replaced by histidine

00:57:18 and has similar physiological properties.

00:57:22 Now, so, then,

00:57:25 the patients who have this

00:57:27 have a high oxygen affinity

00:57:29 and this means

00:57:33 that the organism has difficulty

00:57:36 in drawing off oxygen

00:57:38 so through the sort of feedback mechanism

00:57:41 which you have in the kidneys,

00:57:43 the lack of oxygen

00:57:46 stimulates the synthesis of erythropoietin

00:57:51 which in turn stimulates red cell synthesis

00:57:55 resulting in polycythemia.

00:57:57 So then these patients have polycythemia

00:58:00 merely through a disturbance

00:58:03 of the allosteric equilibrium

00:58:05 between these two forms.

00:58:07 Now, the converse you see there,

00:58:09 there's a hemoglobin Kansas

00:58:11 in which this asparagine

00:58:13 is replaced by a threonine

00:58:16 which cannot make that hydrogen bond.

00:58:19 So hemoglobin Kansas

00:58:22 has this structure

00:58:24 even when it's fully oxygenated

00:58:26 and the oxygen affinity is low therefore

00:58:31 and resulting in aplastic anemia.

00:58:34 So again, the anemia is due merely

00:58:37 to too much oxygen being available,

00:58:40 oxygen being drawn off too easily

00:58:43 so that the synthesis of erythropoietin

00:58:46 is repressed

00:58:48 and the anemia results.

00:58:57 There are two other hemoglobin,

00:59:01 abnormal hemoglobins

00:59:05 discovered here.

00:59:08 Let's say one of them discovered here,

00:59:10 both of them analyzed in great detail here.

00:59:13 These are hemoglobin Rainier and Bethesda.

00:59:17 Can I have the next slide, please?

00:59:20 They both have very high oxygen affinities.

00:59:25 Here you see the oxygen equilibrium curves

00:59:28 of normal hemoglobin

00:59:30 at three different pHs.

00:59:33 So this would be at low pH

00:59:37 and that one.