Oliver Smithies:

[00:00:00] I’ll just go and get my aqua — this is Book F, 1961.  Continuing, trying to improve the purification using an ECTEOLA-Sephadex cellulose with a sodium chloride gradient, bulk experiment.  With a gel, we commented on, but no images of it.  A rather poor yield even when 0.04, 0.08M, must elute in bulk, a comment on page five. [00:01:00]

Bulk Ecteola, next day, April, page five, yes.  On page seven, bulk Ecteola preparation, 1-1 and 2-2.  No gels still, see the result.

Although on page 11, I’m evidently very pleased at using 250mLs of 50% Dowex hydrochloride wash, cold and reacted with deionized 2-1 [00:02:00] and acetic acid at pH 4.7, under various attempts.  Eluted into 2mL fractions, testing then different concentrations of sodium acetate from 0 up to 0.005M.  And, very promising indeed, it says, page 11.

(inaudible) over the page, the results show very marked decrease in impurities as the sodium acetate increased without a parallel loss of haptoglobin.  Impurities appear to be down to about 5% of usual by adding 5,000M sodium acetate, but the haptoglobin [00:03:00] is only down to about 50%, and that may be just spread out, so rather encouraged.

Test of hemoglobin elution, and with and without sodium chloride.  With the result on page 17, the results are essentially the same as with sodium acetate.  So, these are hemoglobin elution on one gel, and multiple pictures of it, and then [00:04:00] that’s the gel with, a gel 1 and gel 2.  The comments are that the results are essentially the same as sodium acetate, doubtful if there’s any loss of haptoglobin.  But it comes out in the second fraction also, et cetera, etcetera, still trying to improve things.

Friday, April 28th, testing of the concentration for effects for splitting.  Some comments on different gels. [00:05:00]

Wednesday, May 3rd, trying deionized urea.  Just different concentrations of starch, formic acid, pH about 3.8.  Results very satisfactory for 2-1S separations with the deionized urea.  95% starch is better than 100, and 105 is very much better. (laughter) For a while, it didn’t make sense. [00:06:00]

May 4th, test of DEA hydrochloride, plus or minus sodium acetate.

Splitting on the recent two-dimensional tests on Monday May 8th, page 39.  And, with a comment that it looks very much as if the first diffuse 2-2 zone is lacking part of the molecule [00:07:00] with the two-dimensional gels being in one dimension.  TBV gel, Tris-Borate-Versene gel.  I don’t know when this was first tested, but it became a very good buffer later for use for hemoglobin that I found.  So the first gel was fraction 2-2, was number one fraction of 2-2 from F7 with 4% hemoglobin onto Tris-Borate-Versene gel and run, and then that was cut out, that those samples were cut out and run with 0.21M [00:08:00] and it says 14M, but I doubt that it’s 14M, mercaptoethanol.  Might be 100%.  No, there’s something wrong.  Two-dimensional gels are quite pretty.  With clearly the polymers are disappear, from the gels with the second dimension having no mercaptoethanol, and then by the time [00:09:00]

Looking at page 38, the bottom gel in the image appears to be haptoglobin, hemoglobin-haptoglobin run without any mercaptoethanol, and it’s just a diagonal gel of all the polymers.  And then, as the mercaptoethanol increases, then rather smeary results are obtained, a little bit difficult to interpret, but clearly something is going on.

And it’s [00:10:00] surprising how often I have the comment results “very promising,” page 43 is “results v. promising.” So many times, (inaudible) I must have been a perpetual optimist.

FS typing.  So Tuesday, May 9th, page 45, FS typings. [00:11:00] [phone conversation] So on Tuesday, May 19th, page 45, I’m FS typing, as it says.  So 2mLs of distilled water, 2mLs of plasma, [00:12:00] and mixed bed resins stirred for 20 minutes, and filtered, et cetera, on to cylinders, and eluted, so this is a quick way of making small amounts of material.

And so on the next page is says, page 47, the results show good recovery on all of the sample.  And then overnight gel, with 8M urea, all 5M formic, deionized 8M urea.  This gel is not so good.  Return to plus-mercaptoethanol.  Within the following page, the bulk result is there.  2-1 [00:13:00] with 14M mercaptoethanol, and 1-1 with 14M mercaptoethanol means undiluted.  But it’s of course not a large volume of it.  The experiment was typically eluted 5mL, et cetera.  Pooled FS, added hemoglobin, et cetera.  2-1, and 1-1, 2-1 gave two bands, and 1-1 gave probably two bands.  But, [00:14:00] they were run for two-dimensional splitting.

So a repeat of the first dimension on page 51.  The first dimensional strips were exposed to 40M mercaptoethanol.  So evidently, that’s what I was doing, treating the strips with high concentrations of mercaptoethanol.  Confirming the results, and page 51 confirmed that 2-2 fast zone lacks a component that the 2-1S compact, et cetera, et cetera.  Still trying to understand the [00:15:00] depolymerized products.

Micro-tests of DEAE on May 11th to develop a method that wouldn’t require so much material.  Various trials to improve things. [00:16:00] Nothing very remarkable.

And now, Friday May 26th, new standard DEAE micro-procedure, starts with mixed bed resin to deionize the plasma, and then mixed with 1% DEAE in distilled water, et cetera, added sodium acetate, pH 4.3, et cetera. [00:17:00] Dry and take up, and here is a comment which later, I knew I used it at a later stage, that, “Clear that ammonium acetate is volatile at 100 degrees centigrade,” because I was drying and taking it back up into solution.

Attempting different elutions, page 75, and looking at the product in urea formic acid gel.  Just different elution procedures. [00:18:00] Trying to get DEAE Sephadex for gamma-globulin.  So clear that the DEAE — that the Sephadex very nicely prepares gamma-globulin free from the alpha-globulins or from albumin.  So [00:19:00] some micro-tests, apparently 3mLs of distilled water, and only about 1mL of serum, 12 samples, from Red Cross typing gel, tried to see if I could get material purified with only using 1mL of serum.  So this was 3mLs of distilled water, and less than or equal to 1mL of 12 serum samples, et cetera, mixed bed resin, and filter, and then DEAE, pH 4, and then set up any column and elute it with 0.125M ammonium acetate, and dried at 100 degrees Centigrade.  This is the first really, beginning of what later was a very useful — not exactly the first, but [00:20:00] development of a procedure which was later very useful, which was to dry the samples down with a salt solution ammonium acetate, and then rely on the evaporation of the salt so that you got the purified protein without any salt.

This was a gel run on page 83, May 31st, and then June 1st, the gels are shown.  All of the gels for page 83 is not immediately clear, but it says that all the [00:21:00] samples were typeable, all 12 samples.

The next page, however is not that same gel, but another gel.  The beginning of the procedure that later I used very extensively for typing small samples, as far as 1F and 1S were concerned, and that Walter Nance used, here at the beginning of the column that was eventually made, which was Thursday, June the 1st, where there was a small column at the bottom [00:22:00] with a — into which the DEAE would eventually settle, and a larger, wider diameter column at the top.  So the larger, diluted sample could be placed, and then the DEAE cellulose concentrated at the bottom.  So it’s a very useful technique.  Whether, I’m fairly sure that this was published, but this was the first written example of it, page 85, June 1st, Thursday.  So we’re looking at page 85, Thursday, June the 1st, serum micro-test.  The same procedure was used as yesterday, but all the columns were allowed to drain, et cetera, before distilled water was added, so [00:23:00] it was a rather neat invention, you might say, where a column was made that would accept quite a large volume of diluted DEAE cellulose and then concentrated into a small portion at the bottom, the diagram in the bottom right hand side of this page is exactly the method that was eventually published, with the paper that I — with George Connell and Gordon Dixon, published in 1961, on inheritance of haptoglobin subtype, where the little sketch at the bottom, that is part of [00:24:00] the paper’s Figure 1, a special column without dead space for subtyping haptoglobins.  This is the first gel testing it, and it’s quite a nice gel, as a start that the final paper, which we no doubt will come across, the gel, is not this gel.  But the beginning of a good method of being able to subtype haptoglobins without using a lot of material, only a mL of plasma, serum. [00:25:00] The eventual publication was, the method could be used with only half a mL of serum, so it was really quite a good, efficient way of making haptoglobin. [00:26:00] The eluates were 2mL, and then they were evaporated through dryness, and the ammonium acetate was volatilized by heating the beakers in an oven at 100 degrees for approximately one hour until they no longer smelled of ammonium acetate.  It was really quite a unique procedure that we enjoyed using.

[00:27:00] Preparing more gamma-globulin for Lois Kitze, on the following page, 87, 89.

Page 91, repeat serum micro-tests.  This one looks closer to the gel that was eventually used, but I’ll just check which one eventually was used.  It was just not a particular test.

New DEAE Sephadex [00:28:00] column was tried on June 4th, why, I don’t know.

Complete change of topic on June the 6th, Australian Aborigine typing, referring to a pedigree for the [Alpine?] family in collected data Book 1, with a — these are various members of the [name redacted] family, which were being tested, in this case for HP subtypes, so they were 2-1, and 1-1, and later in red ink, in red crayon, the subdivision was made that they were [00:29:00] in this family.  Everybody was, when the 1 gene was there, they were all 1S.  Even though the samples were in poor shape, they were worth testing, and the results were interesting; the pedigree was perfectly internally consistent.

And here were the samples tested with 1-1 and 2-1 and 2-2 haptoglobin types, with the micro-test of haptoglobin subtype, and it can be seen that all of 1 genes are 1S, but with a comment, that it was wise [00:30:00], in future always have a haptoglobin 2-1 pool sample on the gel, as a marker because when they were all 1S and no 1F, so you couldn’t tell which of it —

Repeat then, on page 103, of these samples, and all checked perfectly, quite beautiful gels, and a joy to look at even at this late stage.  Wednesday, June 7th, page 103. [00:31:00]

Continuing the same sort of test, with another family, the [name redacted] family, [name redacted] family, and so on.  More pedigrees on Friday, June 9th, [name redacted] family.

And, a set of gels on page 115 that were run on Friday, June the 9th, pretty gel, with very clear subtyping.  Again, in this case, it was a 1F standard there, and [00:32:00] 2-1F standard to help make sure that one was looking at the difference between 1S and 1F.

Continuing to test other families.  Here’s the [name redacted] family, that all 1S, but, the [name redacted] family and [name redacted] family.  The [name redacted] family had the 1F and [00:33:00] the 1S in the family, as did the [name redacted] family, and the results are shown in a very pretty gel on page 120-121.

George’s family, George Connell’s family tested on page 123, where father Connell was 1F/1F, and mother was 2-1S, and Mary Louise was F/S, and so was George, 1F/1S, so whenever we ran their sample, we got two bands, which is the same as you get with a pool. [00:34:00] It’s one of the reasons why we didn’t spot what was going on earlier, because George and his sister were both 1F/1S.

Pretty gel again on page 125.  Continuing to yield a quick result now with this procedure. [00:35:00] Just curious to find the actual gel that was used in the publication of the inheritance of the haptoglobin subtype in 1961, July 1961, so it’s somewhere in here, in this book, probably, because here we are in June of 1961.  So I think probably the gel has already passed, and I haven’t been able to spot it. [00:36:00]

Still continuing, many gels of the same general character testing for 1F and 1S.  Page 135 is another pretty gel.

I’m going to pause for a moment and go backwards and see if I can find the gel that corresponds to the paper.  I think it must have been already passed.

So here’s page 135 again, a very pretty gel, but not the one that was published, I don’t think. [00:37:00] Members of two families.  So 135, very pretty gel.

More pedigrees on June 16th, and so on.  Checking the types, all haptoglobin types to be checked on page 143, and more gels from the micro method, one gel broken on page 147-148.  More gels of the same type, continuing. [00:38:00]

Checking on the gamma globulins, and as we come to the end of this book with tests of the high voltage sample, 30V/cm, still every now and then being kept to look at those, and ending the book on July 22nd, Saturday, with some photographic tests of gels, again to get the best images.  Red filter, high test, and no filter, and red filter, et cetera, et cetera.  And that’s the end of this book.

There’s a particularly pretty gel on page 131, or two gels, which was run [00:39:00] from two families.  The [name redacted] family, and the [name redacted] family, and this gel was, in fact, used in the publication, so page 131, the two gels were run, and the lower gel on the page is the basis for figure two in my paper with George Connell and Gordon Dixon on the inheritance of the haptoglobin subtype.  Oliver Smithies, George E. Connell, and Gordon Dixon, received July 17th, 1961, and the gel on which it was based was run [00:40:00] on June 14th, so it didn’t take long to submit the paper that the gel was run one month before the paper was received by the journal, so we didn’t waste any time.  We knew what we had.  So that’s a pretty gel on page 130-131 of book F, which is rather enjoyable to look back and see.  Those are gels where it’s such a joy, and the method was so simple that it could be used very easily, with minimal amount of serum, half a mL.