Oliver Smithies:

[00:00:00] October 30th, 1986, the beginning of book nu(ν) continuing on this general type of work with first page being some tests of induction, all of which are negative, on the opposing page.  And in the following page again, all for induction.

Planning session, October 31st, more photographs of hoping to induce or detect induction.  Pages and pages of similar experiments. [00:01:00] All negative, Tuesday, November 4th, page 9.

Reconsideration of the status on Wednesday, November 5th, page 11, “A re-review of current status suggests only sensible EK experiments at the moment.  The only sensible EK experiment at the moment is HPRT neo fusion knockout, but we could still try to make HPRT-EK cells by gamma radiation, so it’s clear that I’m thinking here of knocking out the HPRT gene as a means of being able to get rid of the difficult experiments that had been done with the bacteriophage. [00:02:00] But it’s still knockout of HPRT, not correction yet.

So DNA on page 13, Tuesday, November 11th, preparing DNA from the 129 ES cells, or EK cells.  That’s not correct.  Let’s do that page again.  Page 13, Tuesday, November 11th, two things going on.  EK cell, beta-S cells for transfer, commented on and then [00:03:00] “Making DNA of EK-type DNA,” as I called it, on Monday, November 11th, just a preparation of DNA from 129, strain 129 cell.

And checking the DNA on the following page.  DNA is good, and that it can digest OK with Bam as a test.  Friday, November 14, page 15.

Back to blood smears, the male chimeras, seven male chimeras with the neo-beta-s plasmid, comments on what the degree of chimerism [00:04:00], not altogether happy with the antibody.

“A MISTAKE” in big capitals, page 16.  “The mouse serum reacts with anti-mouse immunoglobulin,” so it gives a serious background.  And another comment that “The animals never had offspring, ditto animal G.”  But there were, nonetheless, that still five chimeras that were being looked at.

And, these are looked at on the following page, Sunday November 16th, page 19, the results of chimerism tests.  And the data showed on CNG, a few positive cells, but they’re all white cells, so they’re presumably [00:05:00] mouse white cells making immunoglobulin, or with absorbed immunoglobulin.  Not true chimerism.  But the technique is good anyway, that’s correct.  And with very convincing images.  There’s only one cell with fluorescent on the smear.

Thinking about flow cytometry, flow cytometric separation of cells, FACS preparation, and Ron Jensen uses dimethyl suberimidate to induce production of the [00:06:00] cell, of the hemoglobin, if I understand correctly.  No, I’m not really sure that that’s correct, but he uses dimethyl suberimidate in the FACS sorting.

Trying to make the cells resistant to hemolysis without destroying their antigenicity, what was calling, “hard balls,” and new 21 hard ball.  These are [00:07:00] being — suberimidate is a cross-linker, I remember now.  So, here we are on page 23, preparing “OS hard balls”  “Bled into fetal–” and this is my blood, “Bled into fetal calf serum, spun down, washed,” etc.  “Undetectable hemolysis, resuspended, wash them,” etc., “They’re nice, red cells, and expose them to aqueous dimethyl suberimidate,” and they were still red after incubating for 20 minutes, “and then heat it [00:08:00] to 80 minutes — 80 degrees, and they turn brown as the hemoglobin was denatured into methemoglobin, and washed again, and stored, as a new 23, hard balls, with the thought that heat denaturation probably may — renders the antibody more effective.  “And Ron thought that it permeabilizes the cells.  So I wasn’t sure about that, so let’s try again without heat,” is the comment.

Continuing in this general way to make hard balls.  Here we are on page 27, Tuesday, November 18th, back to EK cells, some notes.  “About 40% good cells in [00:09:00] EK-beta-S, J1-transformed cells from mu-55 fetal layer,” etc.  One mouse, nine embryos.  Two chimeras were — one male, one female.

So intermingling these different experiments, hard balls, etc. on Thursday, November 20th, page 29.  Thursday, November 20th, trying various preparations of red cells, smears, [00:10:00] OS red cell, dimethyl suberimidate cells, cells also treated with methyl alcohol and so on.  But the results are very instructive.  Only the red blood cell spread from 1% BSA or anything light stained, and that wasn’t very good.  All the rest are, for all intents and purposes, negative.  So not a very strong staining.

A method of trying to get cells to stick to [00:11:00] slides, using Histostik, obtainable at 20mLs, could be bought for $15.  So, a bunch of slides made with that material, Friday, November 21st, page 33, but on page 35, it says, “The results are disappointing.”  Seeing different things in the following page.

And on the other hand, Sunday, November 29th, page 39, “A check on the induction of Phil’s material.”  Phil had made 20% fetal calf serum smeared slides [00:12:00] on Histostik.  ML cells, two days after induction, and the induction is fine.  And, quite a decent image of those cells.  So, when you have something that really does make hemoglobin, you can get positive.  So, here we go, doing similar experiments over the next many pages.

Different illumination being thought of, page 47, Sunday [00:13:00] November 30th, trying epi-illumination, for getting images.  And, page 49, Monday, December 1st, epi-illumination, and some very strong fluorescent images.  Let’s see what it says about them.  Yes, the conclusion is that “Epi-illumination is a really significant improvement on well-dispersed smears.”  With, the difference between different images is quite clear.  And, as the next page shows direct illumination and epi-illumination [00:14:00], a similar thing.

And so that results in Monday, December 1st, a re-review of the old slides, but now examined by epi-illumination on page 53.  And, as it can be seen, all of the, about 10 samples on the right-hand page are all negative, despite the fact that the controls are OK, meaning that the controls are positive, but the induction is negative.  Every single one, so it shows the same thing, but the control is not always OK.  But, none of the previous ones look to have been induced, except for 4.5, [00:15:00] there was evidently one not reviewed on page 53, 52.

But on Tuesday, December 2nd, page 55, 4.5 is being re-tested, which is x nu 3-5.  The control was strongly positive with epi-illumination and x nu1, but all four fill sculptures are negative, with HP2S.

Looking at the hard ball experiments again, December 3rd.  And the epi-illumination [00:16:00] was now so good that it’s starting to bleach the fluorescein, page 59.  An attempt with, I think this was eventually with Nobuyo Maeda, to get a golden hamster, embryonic fibroblasts.  There were some reasons for hoping that would be possible.  So two golden hamsters were sacrificed by anesthesia, or by anesthesia and decapitation, but only the first was pregnant.  Many reabsorptions, but managed to obtain something out of the first one.  So golden hamster embryonic fibroblasts [00:17:00] were obtained, growth quite good, combined as nu 61 fibroblasts.

Continue with the trying to get a good assay of red cell on as usual, persist, persist, persist doing the same type of experiment over and over again until either success or abandon.


[00:00:00] This is resuming book nu on page 69, Saturday, December 6, just trying to try different ways still of making cells of, identifying cells that are expressing beta-globin, either S or A.  Continuing in that way for several more pages, with a comment, Monday, December 8th, page 75, stain the hard ball of cells that the results are spectactular; not very dependent on the level of dimethyl suberimidate, [00:01:00] but whether that turned out to be really true or not, we’ll find out.

Going back to making ES cells on Tuesday, December 9th, page 77, “New attempt at getting ES cells,” notice ES, not EK anymore, by OS protocol, taking a large number of blastocysts from ICR, outbred mice 70 to 80 blastocysts, and setting them up in dishes about 30 to each of, two different wells [00:02:00] in — with a few fetal layer of mu-55 fetal layers.  I’m reminding myself that the EK cells are from 129-SV/EV mice, which are GPI 1A homozygotes.

So looking at the cultures on the following pages, Wednesday, December 10th, page 79, “Around half are still visible as blastocysts.  Trypsin didn’t work.  Continue the culture,” etc.

A little comment on page 78, “This one looks to be doing something.” [00:03:00] Thursday, December 11th.

Gene rescue around, on Wednesday, December 10th, page 81, thinking about gene rescue experiments for Paula Henthorn, and about to do sort of work again, and looking at Charon3A-delta-delta-xbar DNA, which has been from phage DNA from previously, getting ready to use this material.

So, the ligations [00:04:00] on Wednesday, December 10th, page 85, these looking at delta-xbar, and delta-delta-xbar, Charon 3A, and ligating various materials.

Packaging, again, looks like old pages for the gene targeting experiments, packaging with all C1A and 5.2-positive and IBS2-positive, etc., etc.  One very doubtful positive, picked in the area and plated; tiny plaques, so not satisfactory.  Typical type of problems with the old [00:05:00] lambda phage assay.

Saturday, December 13th, enter temps that I think, as I recollect, were primarily something that Nobuyo Maeda wanted to do.  And that is, Saturday, December 13th, 1989, tried to get golden hamster ES cells, and so here’s the beginning of getting fibroblasts, first of all, from golden hamsters.

Several progressing well, following day, on Monday, cloning wings, we used. [00:06:00]

And here’s one more single-strain packaging experiment, large scale, page 95, Tuesday, December 16, usual sort of diagram, as in the early days of gene targeting in a phage experiment.  No notes of any positives, however.  Back on film with the plaques are very painful. Is this the DNA in the mix, causing loss of signal, or whatever, omit the DNA. [00:07:00] A couple of exclamation marks, “Evidence of severe inhibition somewhere!!”  Not a good outcome.

Now, however, here we are on Wednesday, December 17th, the beginning of a new thought that eventually led to a very satisfactory conclusion, thinking that it would be possible to detect recombinants using the PCR reaction that had just been described by — using the PCR reaction, which had just been described by Kerry Mullis, thinking that this one could use the suitable probe that would allow [00:08:00] one to detect substitution or a deletion by using one probe corresponding to the input DNA, and another — one primer, that’s to say, corresponding to the input DNA, and another primer corresponding to some part in the target, so that when the two came together, one would be able to amplify a recombinant fragment, or as I call it on page 97, “a diagnostic fragment,” thinking that it ought to be within about a kilobase, because at that time one didn’t know how to amplify fragments much bigger than a kilobase long using PCR.  And as we shall see, this idea was then exploited quite considerably [00:09:00] at later times, so Wednesday, December 17th, important idea.

Looking at the colonies in the following pages.  Agar cloning, Monday, December 29th, page 111, and so on. [00:10:00] An example being Tuesday, December 30th, page 115, fetal layers and EK cells distributed on these fetal layers.  Try to learn how to do these experiments, small number of cells.

Attempts on colony lifting on Thursday, January 1st, New Year’s Day, 1987, thinking that one could isolate 100 a day [00:11:00] with practice, some ambiguous spelling of practice.  Single cells in Terasaki plates are feasible was the thought.  But that omits the thought of how important it is to have conditioned medium that will allow a single cell to grow in a huge dish which even a Terasaki dish is huge when considered in terms of a single-cell.

Page 118, it describes the Thursday [00:12:00] experiment, and by looking at them on Monday, January 5th, the results are, dead, dead, dead, dead, dead, except for F9, when five days, about 20 good-looking cells on January 1st, and among the January 5th, there’s several hundred cells in there, so F9 had it work.

Considering the whole situation, Friday, January 2nd, page 121, single-cell into Terasaki plates, method will work, fine judging from division already beginning. [00:13:00] How to pick up the cell, was the thought on this page.

Being continued, thought on Monday January 5th, page 127, colony picking, with a comment on the bottom that supports the earlier thought that Tuesday, January 6th, “Pretty clear that they don’t like this medium.  Try conditioned medium.”

Going back to some of the results of embryo transfer, seven male offspring [00:14:00] from EK-beta-S first clone were tested, but no colored offspring, so that there wasn’t any germ line transmission.  And here is some indication of Tom Doetschman’s involvement, because he says — he says Dumdee will flow out of this, and other ES cell lines of his own, and we would test in blastocyst from various mutants at the c-kit locus, or WWV and blastocyst, etc.

More colonies on Tuesday, January 6th [00:15:00] in a culture from Ron Gregg RGB-mu-1111, if we go back to that, which was an agar experiment, trying to clone male/ML cells, and RGB-11-6v as various cells, more colonies, just testing whether the colony — one can grow cells from — whether one can propagate single-cell, etc.  So, these were picked into conditioned medium, and a group of experiments called A, and a [00:16:00] group of experiments called B, with notes on them.

Crazy idea number two, Saturday, January 10th, page 135, but no good, how to pick colonies didn’t work.

And nothing terribly interesting in the next few pages, and on page 145, or 144, an insertion of a whole bunch of [00:17:00] thoughts on how experiments might go, Friday, January 16th, 1987.  Experiments that might be done with Ed, which is showing normal chromosome with beta-globin wildtype, delta-beta, and trying to get an insertion, a crossover with beta-S, so insert within a plasmid-by-plasmid experiment, and what primers one would use to detect that event.

Then, Vicki’s experiments are considered.  She was doing plasmid by plasmid, and we’re trying to find out [00:18:00] if one could get insertions at different lengths.  So, here are her experiments and the possible planned experiments, and possible PCR, assay.  Other rather wild ideas, of cytokeratin.

Back to simple problems, now is the first actual test of the December 17th idea of using PCR, because we have made a three-temperature water bath [00:19:00] controlled heated reaction block for, in other words, a PCR machine.  One couldn’t buy one of these, and it was several years before one — at least a couple of years before you could buy one.  So by making it, even though it took quite a lot of work, one saved a lot of time and was able to do the experiment.  So here is the design of a rather crude three-temperature system with a melt, anneal, and reaction water baths, and some relays which were the best I knew how to do that could have variable times, a couple of variable-time relays, and how to set the system up, and have a diagram of a panel that could be made to control [00:20:00] the temperatures, the delays, rather, and therefore the temperatures.  When the melt valve was open — the anneal valve, the reaction valves.

So on Saturday, March 14th, put in a 15-second delay on the valve openings to allow closure of the preceding valve, because it looks as if there was some overlap, and beginning to test this system.

So Monday, March 18th, assembly was completed, and tested out OK.  And the timing order is important, and talks about Taq polymerase obtained from Biolabs, and now.  And a diagram, multicolored diagram [00:21:00], the circuit diagram, just a simple, really, mechanical device, not digital at all.  I didn’t have any knowledge of what digital timers were available at that time.  Though some other people did, in fact, produce a digital machine which was quite simple, which I’ll refer to later.  And that’s the end of book nu(ν) which ends by looking at some of the red cell fluorescent cells, and comes to the conclusion that they probably are not red cells, but are cells that are IGR-positive and maybe [00:22:00] are, in fact, antibody-producing cells, rather than red cells with the right markers.  So we go to the next book.