Oliver Smithies:[00:00:00] This is Book Roman XXI, Roman XXI, 1959. It doesn’t say what the dates are. But it starts on July 29, with some typing gels. I’m beginning to think about adding iodoacetamide. So here is 8‑[molar?] urea, boric acid, sodium hydroxide, 0.2-molar mercaptoethanol, labeled, “U/B/ME.” And then 8‑molar urea, boric acid, sodium hydroxide, iodoacetamide, labeled, “U/B/IA.” So I’m going to be using iodoacetamide now.
So the Thursday, [00:01:00] Friday, August 7th, took some 1% hemoglobin 2‑‑ so, 1% haptoglobin 2‑1 in 8‑molar urea — and reduced in the ordinary way, with 8‑molar urea, pH 7. And added either iodoacetamide or iodoacetamide followed by mercaptoethanol or iodoacetamide, mercaptoethanol, different… Or does…? Not quite clear what this is doing. It… The control is iodoacetamide without mercaptoethanol. It’s not the order. It’s which is added. So the control is iodoacetamide alone. And the zero time is iodoacetamide and mercaptoethanol. [00:02:00] But presumably, the order would be the natural order. So here is the statement of the order. “At 9:54 a.m., added 0.1mls of urea, borate, mercaptoethanol to the hap‑‑ to the haptoglobin, which was already in urea and borate, mixed. And immediately, in two minutes, added iodoacetamide,” to get [2.5X?] iodoacetamide, relative to the mercaptoethanol. So the logical order — reduction. And then iodoacetamide afterwards.
And the gels were run on Saturday, August [00:03:00] 8th. It says the reaction of mercaptoethanol with iodoacetamide… “Tested by taking –” etc., etc. “Reaction complete in a few seconds,” is what it says. And on the left-hand page, it says, “The reaction is very rapid, complete in –” quotation — “zero minutes. But a secondary reaction appreciable at 8 hours.” So something is happening at a later time — which is quite noticeable at 26 hour, leads to doubling of some of the zones. So better be fairly quick.
Continuing to determine the effect of iodoacetamide [00:04:00] on reduced haptoglobin. (pause) An earlier sample. Still reduces iodine. So it’s still got thiol there. And testing with different levels of iodoacetamide. Gel is not — the photographs are not very pretty. And it’s rather faint. And the multiple bands of haptoglobin sample are converted into a single, slower zone by iodoacetamide. And some interpretation there. Further time [00:05:00] tests of still single components, with iodoacetamide. And…
This is for purified… That looks like fraction 4. It looks as if George had tried to purify and evidently had tried to separate the fast components from the slower ones. Not really very well done, and here. But the purified material, which is called #4, doesn’t have the slowly migrating constant region. So George had evidently been able to [00:06:00] give me some purified material, that just contained what we later on called the alpha chains of the haptoglobin, the ones that were varying with the genetic type. (pause)
Long-period time tests, on August 10th, 23 hours, 48 hours, 70 hours, trying to understand what was happening. Tested with the — of urea concentration on the iodoacetamide separations, on the plus iodoacetamide separation, Wednesday, Thursday, [00:07:00] October 12th. Eight-molar urea versus 4, 1.5, and 0, in formic acid gel. Ran rather too long. The fast fraction is now smeared. But complete entry into the 8‑molar gel. Getting there, slowly. Different variations on the — on the same theme. Thursday, Friday, October 13th, 14, haptoglobin solutions, 1 perce‑‑ [00:08:00] reduced with mercaptoethanol, in the presence of 8‑molar urea. The reaction is visible, in the 2‑2, because there’s a color formation, and very rapid. And the gel is a formic acid gel, with 1.5-molar urea. Continued with this type of result, using different haptoglobin types, including the Johnson, Elo Giblett’s. So there’s 1‑1 purified, 2‑1 modified, 2‑1 Johnson, and 2‑2, on one of these gels. And [00:09:00] although the pattern is not easy to interpret yet, there is a constant-region material that is running in this gel, 8‑molar urea, formic acid, staining very heavily, and then quite variable differences in the fast-migrating material, that is going to be later called the alpha chain.
Getting to some gels which were probably later published, now, a zero urea, formic acid gel. Now [poor serum?]. And it’s beginning to get results that were published. On [00:10:00] Wednesday, October 19th, there is a gel there where — mercaptoethanol-reduced, in the presence of 8‑molar urea, and then treated with iodoacetamide. Gel is illustrated. And though that was never published as such, there is a diagram of it, of the type of result, in the — in the paper that I published with George Connell and Gordon Dixon, in the Journal of Molecular Biology, 1966, on the “Gene Action in the Human Haptoglobins.” So that’s seven years later, this was published. Took a long time to [00:11:00] get published. But there is a gel showing the me‑‑ when a formic acid gel is run on mercaptoethanol-reduced, iodoacetamide-treated haptoglobins, all of that in 8‑molar urea, you could see material not entering the gel and a very clear difference in what we later called the alpha chain of the haptoglobin. One-one is different from 2– Have to look at these again. The gel is reversed here, so [00:12:00] there’s a bit of difficulty. Haptoglobin 1‑1 gave a faster-migrating band and a large amount of material that didn’t enter the gel. The 2‑1 modified had the faster one plus a moderate amount of a slower-migrating material. Two-one itself had more of the slower-migrating material, as expected, the gene 2 product. Johnson was much slower — the slower material. So it was the heterozygote. Johnson was a heterozygote for 1 and what later turned out to be triplication. And so the very clear demonstration. I don’t know whether that gel was ever published or not. Probably was and I have to look for it somewhere. But the result is very clear and very understandable, that the [00:13:00] variable material migrates in formic — in formic acid gels, whereas the constant part of the molecule hardly gets into the gel, at that point. And if the gel is made too molar in urea, then the rather smeary patterns of the formic acid gel are much improved and one can begin to see more clearly the differences, that 2‑2 gives a heavy band, the Johnson has a faster-migrating band and the heavy, slower band. But interestingly enough, though I now know what [00:14:00] the result is, that the — that the faster band in Johnson phenotype, which is a 1 — later became haptoglobin 1 gene product and haptoglobin gene 3, that the faster band is not the same as in the other individuals — and so, later on, that that was understood, but not commented on here and not really observed. I think it was an omission not to have observed that.
The next page is with 4‑molar urea gel, where similar patterns are being seen. And then [00:15:00] an 8‑molar urea, mercaptoethanol gel, the beginning of the final type of gel used to solve these problems, where everything migrates now into the gel — 8‑molar urea, formic acid gel, with mercaptoethanol-reduced, in the presence of urea, haptoglobin — and then treated with iodoacetamide — that the slow components now can be seen to be the same in all of the — and there is nothing left at the origin. And the fast components are different — the same as the difference in the 2‑molar urea gel or the 0‑molar. So one could choose which to use, [00:16:00] in order to see the differences in the faster-migrating material.
Again, a co‑‑ a complete set of gels, untreated haptoglobin, in 8‑molar urea gel — and, I think, at an alkaline pH. It’s not very clear. (pause) [00:17:00] (long pause)
Untreated haptoglobin in 8‑molar urea, alkaline gel, presumably, and with a comment that “The Johnson type is the only one [00:18:00] to show any effect of urea, on this length of treatment.” But the 2‑2 stored a long time was very much affected. (pause) Some comment on the use of iodoacetamide. “It’s dangerous at a high p‑‑ because of possible secondary reactions,” which, “Some also occur at lower pHs. Therefore, retesting, use a different protective groups.” Ethylmaleimide [00:19:00] and chlorobenzoic acid were thought of.
So there are some blocking tests, on Friday, August 21st, in alkaline solutions. The gels don’t look particularly good. But it says, “Select the best looking one for future tests.” So that’s iodoacetamide and ethylmaleimide and paramercuric benzoate.
Time tests going on. [00:20:00] pH tests — lower pH for all of the reactions. pH 7.2. And on Monday, Tuesday, 24, 25th of August, pH 7.2 experiment — with the comments that no addition is said to be rapid. Iodine decolorization. Oh, yes. No addition… And [00:21:00] actually, that’s just testing whether there is any reducing agent left, any thiol left. But the results of pH 7. It says, “Once again,” and, “ethylmaleimide completely wrecked — completely wrecked. All show very serious changes,” as… “PCMB wrecks slow.” (laughs) So not liking it very much. Maybe a mistake. Come back, stop here, [doing?]…
Went, on Sunday, Monday, experiment with “Considering fast reaction immediately with urea and then — and then fast removal with [Servatex?],” etc. [00:22:00] Little bit worried about time effects. And Tuesday, Wednesday, “Repeat of last week’s samples,” one with iodoacetamide. But they look pretty bad. As it says, “All are right. Indication of some non-staining material (query: carbohydrate?) ahead of the smear.” So couldn’t leave them around a long time.
Some cattle transfusion tests here, with a note from Charlie Hickman. (pause) [00:23:00] A Holstein sensitized with a previous sample. So attempt to see whether there was any effect of pre-injections of a different transferrin type.
Investigating the reaction of lower pHs. Wednesday, Thursday, 8‑molar urea with acetic acid, pH 5. And, etc. And [00:24:00] with a comment later on on the gels, “Clearly, the mercaptoethanol has not reacted completely in this time, 2 hours, intended to be 3 hours. — But by 5 hours, pH 8.05, some further reaction has occurred.” So not really happy with the results yet.
So the summary on Monday, August 31st, Tuesday, September 1st, “The effect of urea, pH 7.2, results are most promising.” Decided to try the effect of storing in the presence of urea, at this pH — and 7 hours and 25 hours stored material. With a big note on the left, [00:25:00] “Electrophoresis too poor. Repeat these experiments.”
Antibody tests. Never went anywhere. Some animals injected intracutaneously with rabbit serum — George Connell serum — trying to make antibodies against the haptoglobins, presumably. Quite a few pages devoted to antibody tests, in different rabbits.[00:26:00] And then back to haptoglobin again, Saturday, September 19th. Re-crystallizing the urea. Material stored for 48 hours. And results are over the page. “Results show only small differences, possibly not significant” — which mean starting material, and re‑crystallized urea, didn’t seem to make any different, whether the urea was crystallized or not. Perfectly understandable result. [00:27:00] And 72 hours with iodoacetamide. Tested again re‑crystallized urea. Re‑cr– No indication that it made any different. Re‑crystallized urea again, Thursday, September 24th, re‑crystallized from methyl alcohol. (pause) [00:28:00] (pause) An attempt to look at that, with a comment that “The re‑crystallization improves both slightly.” So there’s a slightly better result with re‑crystallized urea. (pause)
Material being [00:29:00] stored in different ways, room temperature or in the cold room. Running the gel. That’s not the storage. Running the gel at room temperature or running it in the cold room. With the comment that “The difference between warm and cold gels is very slight. Both show [excellent?] fractions.” So re‑crystallized urea looks good. And then there are two rather nice gels showing the results. (pause) [00:30:00] (pause) But still not noticing the difference in the 1 gene — and the product from haptoglobin 1 gene, that there are really two forms of that one product. No comment yet on what later turned to be a difference between what we called 1F and 1S. But hadn’t, and really understood yet. Hm. September 25th, “Attempted to re‑crystallize urea from boiling water,” and, “Serious evolution of ammonia,” [00:31:00] probably due to the formation of biuret. So re‑crystalli‑‑ from ethanol. (pause) Samples stored for 30 minutes and 10.5 hours and separated. Really not any obvious differences — Monday, September 28. Continuing to try to stabilize everything. Similar pages [00:32:00] coming together.
Tried methanol, Wednesday, September 30. “Attempted to break the hydrogen bonds with methanol.” A preparation of haptoglobin 1‑1 in sodium borate. And then up to — 0%, 20%, 40%, and 80% methanol, by volume. And then added mercaptoethanol. With a mistake. The comment that “Acetone [00:33:00] gives the usual fast products.” So I evidently tried it in acetone, as well. “Methanol gives a trace –” 64%. So in the presence… But dissociation with mercaptoethanol — sorry — with methano‑‑ with methanol did not work. But it appears that 64% acetone gives a — usual product, instead of using thiol. And I don’t understand that, at this point. See what happened. So acetone tests, Thursday, October 4. [00:34:00] Made haptoglobin in 0.1-molar borate buffer and added 0%, 10%, 30%, and 80% acetone, and then added mercaptoethanol. So it’s using acetone instead of urea. That’s the experiment. And followed by iodoacetamide. And it says then that “All are effective. The gel is run usual 8‑molar urea gel” and that the acetone tests, from Wednesday, showing the fast band and the new [00:35:00] acetone tests, at 30% acetone or 80%, worked. “All are effective. The usual two bands to one band, with iodoacetamide.”
Trying sodium sulfide, Na2S, a clean crystal, made up. And the comment then that “Sodium sulfide reduces haptoglobin more readily than mercaptoethanol. But mercaptoethanol gives a further reaction at a higher concentration.” So comments on it. Never really followed up. [00:36:00] A comparison of mercaptoethanol and sodium sulfide, alongside each other. (pause) But it’s clear tha the products are not quite the same at the higher concentrations. “The urea gel shows that 0.067-molar mercaptoethanol gives the clearest product, as good as any ever obtained with iodoacetamide, but still multiple slow and [00:37:00] double fast. The sodium sulfide is also multiple. So try higher concentrations of mercaptoethanol.” The problem here which was unrecognized is that the haptoglobin in these preparations was actually a pool of haptoglobin of type 1‑1 and type 2‑1 and 2‑2 and that there was a variation in the gene 1 product — which could have been recognized earlier, from the Johnson phenotype but has not yet been recognized. So the results were quite confusing. We got two bands with haptoglobin 1‑1, of the fast component, and only one with haptoglobin 2‑2. Because there were two forms of the gene 1, [00:38:00] not yet recognized. Some experiments on Sunday, October 4, of different concentrations of mercaptoethanol. And repeated on Tuesday, which is a nicer gel. All the samples had been a longer time but the result is nicer on the Tuesday gel than on the earlier one. Double bands for haptoglobin 1‑1, of the fast band. (pause)[00:39:00] Wednesday, Thursday, October 17, trying other denaturants. It turned out to be quite an important gel, trying 0.1-molar cetyltrimethylammonium bromide, CAB, in borate buffer, and trying sodium lauryl sulfate and a less — in borate buffer and acetone and urea, so a comparison of taking haptoglobin and reducing it with these different denaturants. And the interesting thing was, there, that… Actually, turned out to be quite important, though I don’t know whether this was the gel on which I saw it. But [00:40:00] the sodium lauryl sulfate appeared to precipitate all of the slower-migrating material. Not yet very clear. This was a formic acid gel, without urea. But nothing migrates in the presence of sodium lauryl sulfate. But of the other tests, urea and acetone, really very similar results. Although the comment was that the acetone preparation became very viscous and that cetyl ammonium bromide gave a precipitate, which later dissolved. So it was a beginning [00:41:00] of testing other material.
No, it was really the gel of — following day, Friday, October 9th, which really gave the clue to what was later used to purify the haptoglobin alpha chains. And here was a run starting with 5% haptoglobin 1‑1 — a pool — in mercaptoethanol and then — so that’s without the denaturant — and then varying levels of cetyl ammonium bromide or sodium lauryl sulfate, SDS, together. And a formic acid gel and an oxalic gel, with 8‑molar urea. And [00:42:00] the cetyl ammonium bromide samples all gave the alpha chain. But the sodium lauryl sulfate didn’t appear to give anything in the formic acid gel. But when the gel was 8‑molar urea, oxalic acid, then one could see, interestingly enough, that 8‑molar ur‑‑ 60% acetone gave basically the same result. But 0.05-molar sodium lauryl sulfate gave a precipitate at the origin. In fact, if anything, it ran backwards. But the alpha chains were migrated OK. So it would look as if it might be used to purify. [00:43:00] The comment, then, is that “The only promising one is sodium lauryl sulfate, with the slight –” and w– (pause) I don’t know at what stage it was realized that the sodium lauryl sulfate gave a bulk method of making alpha chains. But it later definitely was used. So this Friday, October 9th, gel was an important gel.
Trying dioxane, on October 13th, as well as sodium lauryl [00:44:00] sulfate at lower concentration. And acetone again. With a gel run on Tuesday, Wednesday, October 14th, lauryl sulfate and dioxane. Think the dioxane… With a comment here that this was “18 microliters of — and 20 microliters of haptoglobin, 5% –” in — and it’s here, “0.1-molar borate, with 0.5-molar mercaptoethanol,” etc., etc. And the — and the reaction had gone to completion even without denaturation. [00:45:00] So they all showed that the reaction had occurred without adding any denaturant, at these high concentrations. And perhaps that makes sense later on — not immediately, at this point. (pause) I liked the migration of the fast component, on Wednesday, October 14th. “No entry of the migration, in 20%. It’s almost Gaussian.” I was always worried by the front of the zones being sharp [00:46:00] and the back trailing. So the photograph of the 20% gel showed that it was quite nice migration, by adding dioxane.
And then trying dimethylformamide. And dimethylformamide, pH 3.5, 20%, 10%, or 0%. Gelled quite well. The 40% gave two phases, but no entry of any of the larger component. But migration in the 20%, again, is rather Gaussian, as judged by the stain.
Miscellaneous entry tests, on Friday, October 16th, trying [00:47:00] 50% glycerol, 5% boric acid, 10% boric acid, 50% ethylene glycol. Fifty-percent glycerol was… (pause) All of them gave entry into the gel. This was haptoglobin that had been reduced in boric acid, with mercaptoethanol, but [00:48:00] so that it had the thiol but no denaturant, to try to see what would be obtained. But this was haptoglobin 1‑1 and there’s no indication of any splitting at all.
A two-dimensional gel, run on Saturday, October 17th, with an attempt to see if the two bands in the… This was a two-dimensional gel. And one direction was a 8‑molar urea gel at an acid pH. And the other gel was without [00:49:00] urea, from the look of it. And shows the usual answer, that the constant component, the slowly-migrating constant component with the multiple band, doesn’t enter the formate gel without urea. But the other two don’t appear to be behaving as if they were different sizes, just as if they were different charges, for the 1‑1 double band. “…pro– An excellent result. Query: Proves that urea appears to [fully react?] and produces two zone.” But note, this haptoglobin was not reduced with iodoacetamide. So still having a great deal of difficulty understanding why haptoglobin 1‑1 gets two fast zones. [00:50:00] The clues were already there but have not been recognized.
Some ordinary, standard gels, typing gels. Testing acid urea reaction, on October 26. “No difference, plus or minus acid level. Probably the reaction between urea breakdown products or rearrangement products,” etc., etc. Obscure comment. Fifty-percent acetone — 50% acetone plus 8‑molar urea. And 8‑molar urea plus acid didn’t have any [00:51:00] effect. This was in adding the urea — or the formic acid after the reduction had been carried, with urea and mercaptoethanol. Trying dimethylformamide again. And that’s the end of this book. [00:51:37]