Oliver Smithies:
[00:00:00] This is physical Book 19 — XIX, December 1958 through April 1959. (pause) Having difficulty in deciding how to freeze the gels for the autoradiograph. It was necessary to freeze the gels, because, if you just stained them and tried to detect the ion, the ion was no longer bound to the transferrin. So it had to be done with a frozen gel or with an unfrozen gel. Frozen’s obviously better. And freezing gels, the second page of this book says the best way yet is [00:01:00] to have a metal plate polished and lower the gel down gently, without bubbles, and lower it (laughs) fairly slowly, about 20 seconds from a — with the metal, downwards. “Worked perfectly on three tests.” I was getting better about that.And Monday, December 22nd, 23rd, very interesting. This is my first entry that I remember of a sample from Verna [Callow?], tests on cholinesterase, typical and atypical genes. Because Verna Callow was the first person to find what we now have many examples of, pharmacogenetics, where some individuals respond strangely to [00:02:00] different pharmaceuticals. In this case, this was an inhibitor of cholinesterase, which was — which was used to inactivate the diaphragm, and for surgery of various types. Of course, the patient would be on artificial respiration. And it was usually temporary and the respiration would return after the choline esterase broke down the anesthetic, which was a choline ester. But Verna Callow found that, with some individuals, who responded by — or who did not recover from the anesthetic until much longer [00:03:00] time period — quite frightening. And these were the first examples of individuals who had genetic differences to different pharmaceuticals, as I was saying. And so this is — oh, mm-hmm — typical and atypical cholinesterase genes versus haptoglobin and transferrin genes. And they were over to Verna Callow for cholinesterase. And in other words, I was testing some of my transferrin variants and some of my haptoglobin variants, to see if they looked different, as far as cholinesterase was concerned. So it was Bob Crawford, for example, a straightforward CC and type 1‑1 haptoglobin — and [00:04:00] a 2‑1 and a 2‑1 modified and a 2‑2, with different transferrin types. And I tested in the opposite direction. I tested some sera from Verna Callow’s sera, whether they were presumed homozygote cholinesterase AA, heterozygote TA, and normal, cholinesterase TT. So TT was the normal cholinestera‑‑ and AA was the homozygote for the abnormal cholinesterase.
The “apparent degree of poorness” is [00:05:00] a comment on the left-hand side of this page, very poor, extremely poor, poor, and then fairly poor and OK — fairly poor, OK. But the conclusion is, “There’s no clear cholinesterase difference but the samples are too poor to be sure.” I don’t… My samples, Bob Crawford, etc., Len Jacob, all show usual cholinesterase only. But the samples were rather poor samples, I think, at that point.
Some frozen gels. December 29th, the first [00:06:00] gel, two hours in the deepfreeze. Etc., etc.
High-voltage — the v– Tuesday, Wednesday, December 31st, rabbit serum — with a rather obscure graph. Rabbit samples on Tuesday, Wednesday, January 7th. [00:07:00] A couple of samples looked rather messy.
At last. Here’s the autoradiograph that I was looking for. Wednesday, Thursday, January 7th, 8th, transferrin types from Elo Giblett. And that was a whole range of them. CC, [name redacted]; B1C, [name redacted]; B2C, [name redacted]; CD1, [I.K., 769?]; CD2, [name redacted]; and CD3, [name redacted]. And it says I took 0.3mls of 0.3% hemoglobin and added 0.3ml of Fe‑59 from before. [00:08:00] Etc., etc. And the gel was on for radiography at 12:00 and off at 1:00 a.m. Repeated on Monday, January 12/13. And there’s a beautiful — gel images and autoradiograph, which were eventually published, the protein stain on the left and the Fe‑59 autoradiograph on the right, showing absolutely unequivocally what the variations were in the transferrin. That was used in a publication with, [you know?]… Very beautiful result, Wednesday, Thursday, January 17.
[00:09:00] Testing other things — and benzidine, on January 8th, hemoglobin, haptoglobin stains, benzidine, and also dianisidine — just stability — and with a comment on the following page that “Dianisidine is very — is very poor and benzidine, as usual, very good.” And clearly, benzidine is much better. Therefore abandon, also, dianisidine and continue with benzidine tests, to get the benzidine, better…Further hemoglobin tests, following page, [00:10:00] trying to get rid of a problem of the gels gassing. For example, “2% hydrogen peroxide, very much gassing and fades greatly.” And by the time I got to 0.2% hydrogen peroxide, “No gassing and no fading,” and double-underlining. 0.1%, ditto. But these gels had not been precipitated with [50:50:10?]. And so there’s a comment, “50:50:10 less good by a factor of about 5.”
Have some confirmation of a different gel. [00:11:00] “Giblett CD2 is equivalent to the old D and now called CD1. (Giblett’s B2C is equal to the old BC and now called B2C.)” The notation had to be changed a little bit. Benzidine stains on January 11. Working out technique all the time.
Oh, Tuesday, Wednesday, January 13th, 14, looking for the pregnancy zone. A bunch of tests with 1159, Mrs. [00:12:00] [name redacted]. November 13th, ’58, haptoglobin 1‑1. And that sample was too aged. “Get a new sample from [name redacted], IV and IV plus hemoglobin.” And I think I’ll pause there for a moment, to the find the reference to the pregnancy zone.
(break in audio)
I recently found an old publication, “Alpha 2-Macroglobulin and Pregnancy Zone Protein.” Alpha-2-macroglobulin is the one that I used to call — and frequently called slow alpha‑2. And the pregnancy zone protein, I’d referred to earlier in our account. [00:13:00] And it’s a large review, about 50 pages long, “Alpha 2-Macroglobulin and the Pregnancy Zone Protein,” by [Claus?] Monk Petersen, published in the Danish Medical Bulletin, November 1993, I think it says. It’s not too well reproduced in my copy of it. And it starts off by talking — when it talks about the pregnancy zone, my description of it in the Advances of Protein Chemistry, Volume 4, [00:14:00] his reference 63. My reference is Advances of Protein Chemistry, not Volume 4, Volume 14, 1959. And it was in the Danish Medical Bulletin. It’s rather interesting to look at it. Very detailed account of those proteins.
So going on, this is January 13th, 14th, Tuesday, Wednesday, as I mentioned pregnancy zone experiments. And they’re quite easy to see in the — in the samples, [00:15:00] as they’re in good shape.
More autoradiography, on Wednesday, Thursday. That was Wednesday, Thursday, January 14th, 15. Another one from Harry Harris. And nice to see the letters from old friends. “Harry Harris, CD3 (CD1), vs. Elo Giblett, CD3 (CD1). Albert Levis,” etc. And then here is the letter from Harry, sa‑‑ “Dear [00:16:00] Oliver. Many thanks for your letter,” etc., “We are pleased that you will join us in the paper. And we’re quite happy to wait till you finish your review.” Because I was still writing the review for Advances of Protein Chemistry. Harry included four sample and some pictures of his three-dimensional gel and the paper that they had already published, Harry, Elizabeth Robson, and — I think it was Marcello Siniscalco. I think it must have… Almost forgotten. Mm-hmm. So the other copy of their paper — and a copy of the gel that I ran, which is a good gel, with the conclusion, [00:17:00] “Thus Harris’s CD1 is equivalent to the old CD and Giblett’s CD1 must now be termed CD0. And Harris CD3 is Giblett’s…” So it got more and more complicated, as more types were found. There was a three-dimensional gel sketched by Harry.
More gels, checking on various people, no particular comments about them. Just routine, continuing, getting more and more samples, more and more gels — [00:18:00] each page with a large gel, with several samples on it. Checks again, large checks, Monday, February 16th, rechecking, on multiple gels — on gels with multiple samples, that’s to say. Monday, February 23rd, to Tuesday, 24th, gel to check the BC versus CC, 2‑1 [00:19:00] versus 2‑1 modified, and a typical gel — with a little floating insert talking about Mrs. [name redacted] and Mrs. [name redacted], etc., just a li‑‑ a little sketch, of a little piece of paper. It’s clear, on Wednesday, Thursday, February 26… I have a gel there that’s labeled with ceruloplasmin. And I don’t know what made me know that that was ceruloplasmin. But it’s a benzidine stain. [00:20:00] And I think there must be some cross-reactivity. The benzidine can give a positive reaction with ceruloplasmin. This is a sample from a 2‑1 modified, that has that label on it. Since they are clearly stained, they obviously have the hemoglobin, haptoglobin stains. The protein-stained gel is there, as well. Working on this usual type of thing. (pause)
[00:21:00] I think I’m trying to understand the best way to get a contact between solutions of different concentration, the gel being the usual gel, pretty usual, 0.025-molar boric acid, 40% neutralized, with sodium hydroxide, and then various concentrations, 0.5-molar and 0.025-molar. A gel with lithium, sodium, potassium. Not at all clear what I was thinking about in this but…Oh. (laughs) [00:22:00] Presumably, the… Not… Worrying about shrinkage of gels at the negative slot surface — is the motivation. Continue with this on March 4. The en‑‑ the ends of the gels in contact with filter paper with buffer in it were always a little bit of a problem. That contact would either shrink the gel or fragment it and was never completely [00:23:00] satisfactory. But it was usable.
And trying to get an elution, preparative electrophoresis, on March 5th or thereabouts, with a gel and a cross-flow at the bottom of the gel, one drop, ten minutes. And a lot of breakage — or “Very much gel breakage on the outflow. Something wrong. Remake all the solutions.” And repeated on Monday, March 16, cross-flow, etc. (pause) [00:24:00] Obviously giving me trouble and not very understood. (long pause) Still worrying about gel breakage, with different solutions. Trying to understand whether it was — whether it was the cation or the anion that was [00:25:00] giving the trouble, with a comment that “All the exits are indistinguishable, frozen and some breakage. And all the entries are indistinguishable and perfect,” when an ordinary borate gel had potassium fluoride, potassium chloride, potassium bromide, potassium iodide. And then, “This suggests, as before, that the cation is the responsible ion.” Because all of them were the same, when the anion was varied, fluoride, chloride, bromide, iodide. Made note of the… And so this checking it, at March 19th, with gels with lithium, sodium, potassium, and cesium. They’re all swollen in — swollen in 30 [00:26:00] minute. (pause) Still not getting anything very convincing…
Friday, March 20th, “Accurate standardization of sodium borate exit concentration.” So that’s… Point-oh-two-five gel (inaudible), neutralized with inserts of 20% sodium hydroxide solutions, 0.4‑, 0.3‑, 0.2‑, 0.1-molar — [00:27:00] to look for gel breakdown. Tuesday, March 24, shows some sketches of what the sort of problem was, with the fragmented gel and the constricted gel. (pause)
Here, Wednesday, Thursday, March 25th, a regular vertical, air-cooled gel, trying to look for [00:28:00] ceruloplasmin, apparently. On the left-hand side it says, “Ceruloplasmin detection.” So benzidine, without hydrogen peroxide, is a good ceruloplasmin reagent. You don’t need to have hydrogen peroxide to check the ceruloplasmin. (pause) No, I don’t think it’s that. It has to be… No, that’s plus hydrogen peroxide, minus hydrogen [00:29:00] peroxide. I test with hydrogen peroxide and phenylenediamine and benzidine reagent, with and without hydrogen peroxide. And a checkmark indicating positive with phenylenediamine without hydrogen peroxide. And a checkmark meaning, correct, benzidine without hydrogen peroxide. (pause) The images — there are two photographs. One, although it isn’t labeled, is presumably just a higher exposure than the other. [00:30:00] But ceruloplasmin bands are easy to see. (pause) So it’s Emily Atwood, with and without hemogl‑‑ with and without hemoglobin, and the various tests — that ceruloplasmin shows in the gels that don’t have any hemoglobin. But the hemoglobin, haptoglobin is very clear, of course, when there’s hemoglobin there. So it was possible to detect ceruloplasmin the gels and where it migrated. [00:31:00]
And Thursday, March 26th, still this gel shrinkage. “Gels inspected 12 hours later.” And the final comment is, “Thus the three cations,” which are sodium, potassium, and lithium, “are very similar. Sodium is so little different from lithium as not to be worth changing, at present. Use 0.18-molar for entry and exit.” Just picking the concentration.
Monday, Tuesday, March 31st, a ceruloplasmin experiment. [00:32:00] “All the same in ceruloplasmin. And also checked with esterase. And the same for both.” One image is an image of the esterase result and the other is an image of the ceruloplasmin. And they’re all the same in both. (pause) Still messing around with ceruloplasmin detection on Monday, Tuesday, March 30th, 31st, testing for reagents. And the images are on the following page. “All four reagents are adequate, with sufficient time,” and the four reagents [00:33:00] being the usual benzidine reagent and pH 4.7 acetate buffer, saturated with benzidine, etc., pH 5.7, benzidine, and pH 5.7, paraphenyldiamine. “All four reagents are adequate, with sufficient time.” But the pH 4.7, benzidine, is the fastest and most intense. So it became a stain for ceruloplasmi‑‑ I don’t remember ever publishing it. But maybe I did somewhere.
Dr. Walker gave me some samples that were somewhat interesting. Tuesday, Wednesday, April 28th, we g‑‑ “These twins have respectively [00:34:00] all girls and all boys.” Hence the interest. But I doubt if it meant anything, as far as our haptoglobins or transferrins were concerned.
Going on with fluorescent fat dyes. (pause) And trying an inhibitor, a new idea to suppress aqueous fluorescence with thiocyanate. “Good with ER [conc.?].” I must have got this from some other place, that idea. And various comments on what happened, [00:35:00] on the next few pages. Uvitex U, for example. “Completely soluble in water. Partly soluble in toluene.” “This is what Ciba said,” is a comment — these fluorescent dyes or lipoproteins. And not very happy with it. Thursday, April 2nd, “Abandon. All negative.” (pause) Trichloroacetic acid fat trial. [00:36:00] Gels run with trichloroacetic acid and fat, tests with Amido Black and with Oil Blue N. And, “OBN and [Azo BN?],” etc. (long pause)
[00:37:00] And some more haptoglobin tests, around about April 10th or thereabouts, to try to localize haptoglobin zones for preparative purposes. Trying to migrate dye through the protein solution, to see if stained on the way, with a comment that “Amido Black and [Bromtino?] Blue migrating and crossing a haptoglobin type 2‑2 serum. The dye passed the haptoglobins but did not stain them,” so, “Dissociable in alkaline solution,” etcet‑‑ [00:38:00] (pause) With a gel that looks perfectly straightforward.I’m beginning to try to dissociate the haptoglobin from the polymers, April 14th. This is probably the beginning — or maybe not the beginning but another test trying to think about how to get the two polymers to break up into something simpler. And so April 14th is labeled, “Hp tests with benzoate,” [00:39:00] and which is supposed to be a good dissociating agent. But I got the usual 2‑2 pattern. So, “Decided to repeat [SS?] reduction in 8-molal urea.” So beginning to think about getting that critical step in working out what the haptoglobin polymers really were.
And the next page is a fairly important gel, though it’s quite ugly. Alkali tests of 0.05-molar sodium hydroxide, with a comment, the “breakdown and smear.” The 0.01-molar sodium hydroxide, “Considerable but incomplete breakdown and smear.” And the 0.002-molar [00:40:00] sodium hydroxide, “Usual pattern.” And 0.001. So take one look on the left-hand s– And, “Take one drop of haptoglobin solution…” This is with purified haptoglobin, I should have said, not with serum. “Take one drop of haptoglobin solutions and one drop of 0.05-molar sodium hydroxide. Leave for 30 minutes. And apply via starch grains thereafter.” And it just says, “The results suggest splitting and differences.” But there’s a sketch which shows the usual 1‑1 pattern. It looks to me, the sketch, it does. But the photographs show that the haptoglobin polymers have been destroyed. So in the presence of sodium hydroxide, [00:41:00] at least higher concentrations, the haptoglobin is degraded into something, that migrates not as multiple bands in the usual way — but causes a breakdown and smear, is a proper description of it.
And then urea and SS experiment, on the following page. Tells me, “With thioglycolic acid and 8‑molal urea…” “Molal urea” is [00:42:00] underlined. And the result is very clear, that the splitting occurred, that 8‑molal urea in the presence of thioglycolic acid breaks down the polymers and gives a different pattern, in the 1‑1 and the 2‑1 and the 2-2. The 2‑2 give two bands and the 1‑1 gives a single band. And the 2‑1 gives the single band with a faint 2‑2 but the polymers have gone. And during the gel preparation, there’s a note that “Pink color was formed, suggesting some oxidizing groups in the starch.” But there is no question at all that, by having [00:43:00] 8‑molal urea and the reducing agent, thioglycolic acid, the polymers have gone. The results are quite clear — and was the beginning of my beginning to understand the nature of the polymers.
Oh, the next day is — return to fat dye experiments. And a repeat of the alkaline tests on Wednesday, April 15, 16th. One-one, 2‑1, and 2‑2, with various amounts of sodium hydroxide and water, alkali. Alkali and water. And [00:44:00] very clear breakdown of the haptoglobins into complex and smeary patterns. Just sodium hydroxide was not a good dissociating reagent, in the sense of giving something that was easy to understand. The patterns with haptoglobin 1‑1 were difficult to understand — and with 2‑1 and 2‑2 also. Though there are faint smeary differences, it was not very satisfactory.
Repeating urea and thioglycolic acid, on April 15, 16th. “Poor result. Possibly the gel is oxidizing. Witness pink in the presence of thioglycolic acid and smeared zones.” So the zones are very smeary and spread out all over the place. [00:45:00] Urea — and that was with sodium hydroxide — but urea and thioglycolic acid is again beginning to gi‑‑ something that’s a little bit clearer. So this was probably April 16th or thereabouts. And it’s clear that the haptoglobins are, in fact, dissociated into something that migrates faster than the 1‑1, even when the [00:46:00] sample is a 2‑1 or 2‑2. And then this result, which we’ll probably see in the next book, was eventually published in Ciba’s symposium or something of that sort. I’ll check to see where the publication was. And this is the end of Book XIX. [00:46:27]