Oliver Smithies:

[00:00:00] And this is book little t, begins on July 12th, 1978 and runs through the end of December in 1978.  The book is largely concerned with establishing the [30-5A?] as correct.  So begins with pBR322 recloning of the EcoRI fragment from [30-5A?] to try to get a probe that was reliable.  Not feeling very happy about it.  July 13th, page five.  [00:01:00] The ligation is minimal.  The enzyme is suspect due to freezer error.  Add some more enzyme and see.  And retest the conclusion.  There’s a slight increase in ligation from the previous.  And Fred Ho eventually settled on ligation conditions that — because in those days we still were not very good at isolating clones of hybridizing to phages, hadn’t learned the techniques which were later used.  New human hemoglobin candidate [5-1-1A?] [00:02:00] and 20 randoms from a mouse shotgun pool on page nine saying that one looked rather good but [30-5A?] is 4.8 kilobase pairs whereas [51-1A?], one of the new ones, is only 2.7.  Nonetheless the first human EcoRI complete single gene clone on unfractionated DNA.  So this that you could get a clone from a random shotgun experiment.  [00:03:00] [00:04:00] Some double digests on page 19, EcoRI, BglII, etc., EcoRI, BamHI, with a comment on which are positive and which are negative on (inaudible) there’s beginning to get Balb/c DNA on page 21 from livers.  t21 Balb/c DNA 11 milligrams.  And then some minis on random samples from the shotgun just to see what they contain.  And [00:05:00] mouse shotgun and human shotgun experiments [51-1A, 1A, and 1B?], two different things.  And all three are the same.  Gels on the minis Saturday, July 22nd.  Mouse shotgun 28 random phages and human shotgun it looks like another 20 phages also there.  Human shotgun random phages 20 and [00:06:00] mouse shotgun random phages onto the gels to see what is in them.

One or two (inaudible) 2 out of 19 are positive in the mouse shotgun experiment which can be seen as having different bands.  One of them appears to be 9 and the other one is 2 in the mouse shotgun, etc.  So one can see them, 4 of 5 are positive [00:07:00] out of 15 in the human shotgun.  So they were quite satisfactory isolations from random ones.  And repeating the human minis random shotgun [HSA?] on page 27.  And a new human candidate July 25th, it’s [37-hemoglobin-9A2?] mini prepared as usual.  And the conclusion it is a new clone about 1.4 kilobases long.  And hybridization shows it’s probably a globin but only [00:08:00] small homology.  Work this out at a later date.  So it’s a different clone that doesn’t hybridize quite so well as it should.  So it’s hemoglobin but not hybridizing well to the probe, which contained about 180,000 counts per minute of precipitable material from the human cDNA clone [00:09:00] pretreated with DEPC diethylpyrocarbonate, etc.  So this was with a human hemoglobin cDNA.  So this new candidate gene which was called [37-hemi-9A2 nor 9A2?] is being looked at again, or a different one is being looked at, [27-A2?] is being looked at on page 31, 2.4-kilobase insert but it’s probe-negative so not a correct one.  Some histone candidates looked at [00:10:00] on page 33.  HpaI digest and EcoRI digest of a whole bunch of clones on page 35.  And some strongly hybridizing bands but no comment on any.  [00:11:00] So some histone EcoRI clones for nick translation on page 37.  So continuing to look at the new clones on page 39, Sunday, August 6th.  And the result confirms it was a comparison of hemoglobin [Hb51?] versus [Hb4?] and [Hb117?].  Confirms that [Hb51?] is betalike.

[00:12:00] Going on with histone clones on August 9th.  Multiple tests on [Sb202?] and [Hb?] cDNA on page 43.  Rather dirty set of hybridization but nonetheless specific bands are clear.  So the conclusion is that [00:13:00] various ways of getting the blots cleaner.  But [28-2A?] goes from plus plus to approximately zero with poly — using hemoglobin cDNA whereas [Hb4?] and [5-1-1?] are not effective, meaning that those clones are real.  [00:14:00] The [Sb202?] test on page 45, Thursday, August 10th with a comment.  It’s crazy.  They should be histone fragments.  So some mix-up of some sort.  Tests of [PSV?] versus histone fragments on Friday, August 11th.  The results show no effect of poly(A) but that there’s a considerable loss of signal on 0.1 times [S of C?] but [26-4A1?] is still there.  So [26-4A1?] looks to be a real clone histone fragment.  [00:15:00] With a comment on page 49 that it appears that only some of the histones withstand 0.1 times [S of C?].  Therefore rewash them, etc. to pick a better human clone than [39-11?].

And some comments on Tuesday, September 5th and Thursday, September 7th on [51-1?] identification.  Beta can be eliminated by its size, 2.7 kilobases.  Delta looks OK.  [00:16:00] But A gamma and G gamma need to be excluded.  So this is the beginning of some fetal globin clones I think.  G gamma has an MboI site in the coding region.  Beta does not.  A gamma is not known.  So testing therefore to see whether there is an MboI site.  [00:17:00] Check of histone identifications again on page 55.  Some papers are sent off for publication on October 18th.  I don’t know which ones they were.  Probably find out if it was necessary.  Recloning of the 51.1 2.7-kb fragment in pBR322 and the 4.7 fragment of 30.5.  Tested the ligations again on the following page.  October 19th.  The ligation looks good on both samples.  But it is rather a poor test.  Should have had marker.  Proceed anyway.  [00:18:00] Well, having the idea of turning around the fragment in 3A delta lac to turn around the 2.7-kb piece.  Presumably I wanted it in the opposite orientation.  Unreadable result.  Forgot the [N blot?].  Let’s stop it a moment.

(inaudible) Thursday, October 19th, page 67 beginning to realize that one can [00:19:00] make hybridization more efficient by removing the terminal phosphate group in a vector.  So bacterial alkaline phosphatase on vector pBR322 was carried out on October 19th and then used for ligation afterwards after adding diethylpyrocarbonate to kill the enzyme and heated later on to destroy the diethylpyrocarbonate.  So there are the ligations being carried out. Apparently they didn’t work, at least on the next page, page 68.  It says the result of a check [00:20:00] on Friday, October 20th, BAP didn’t work.  The conclusion, the ligation is OK, but perhaps (inaudible) higher concentration since mainly linear polymers appear to be produced.  So (inaudible) so simpler recloning procedure being talked about on page 71.  And ligations 73.  One out of 36 were positive in one test and out of 114 in another test.  So only very [poorly obviously?].  Trying to get [00:21:00] 51.1 into 4A with all orientations.  The aim is to reclone the 2.7-kilobase piece into 4A to get all orientations.  Not clear why that was desirable.

Continuing to do that type of experiment.  Charon 3AHs51.1 digest for [class?] on page 81, Friday, October 7th.  Aimed to show that the intervening sequence size can be seen using a BamHI and RI digest.  So that was at the time when intervening sequences were quite exciting.  [00:22:00] Repeating the clones, 2.7 reclone on Sunday, October 28th, and so on.  Repeating the digests on page 87.  HindIII worked this time.  But should have had a delta lac control.  More ligations.  RI HindIII fragment.  And 4A RI fragment.  [00:23:00] No ligation.  Page 89.  Page 91 ligation is OK.  Use for tests.  [Seed?] at the page 89 ligation gave a high yield of phages, hundreds of hybridization-positive plaques were obtained, 24 were picked, 6 have been run as mini lysates.  And four of the six are probably 4A and 2.7-positive and two of these are probably 3A.  [00:24:00] More tests with pBR322 and 30.5 4.7-kb mini.  Was happy about this experiment with the excellent ligation on Friday, November 3rd.  Well worth testing.  If unsuccessful go to BAP for Charon products.  Seventy-nine candidates obtained.  And a comment on my instrument rating at 6:30 p.m. Thursday, November 9th.  [00:25:00] Instrument rating [OS?].

Still struggling with BAP page 97.  Calf intestinal alkaline phosphatase on pBR322 November 13th.  [00:26:00] And on page 102 a sad comment that my sister Nancy has inoperable cancer.  It was melanoma.  So that was in 1978 so she would have been 48 years old.  [00:27:00] Continuing with the [struggles?], new pBR322 RI on page 107.  Big cross through everything on page 109.  One calf intestinal alkaline phosphatase didn’t evidently do it.  And continuing to try to get the clones.  Not having much success.  What would be very easy now was really quite difficult in those days.  pBR322 doesn’t ligate as expected, 2.7 and 4.7 pieces do mainly into lambda.  But some expected products.  However, transfection yielded nothing.  [00:28:00] The ligations were thought to be OK.  Wednesday, November 15th.  And page 115, went to England to nurse Nancy.  And she died peacefully on December 1st, so that was very rapid.  I remember being with her in the last days.  I think it was quite some comfort to her.  Her husband was too upset to help.  [00:29:00] Still continuing.  Conclusion.  The ligations look fine on page 117.  I think we probably hadn’t realized that some fragments don’t clone well into pBR322 and was only later that we removed some sequences from pBR322 to make it easier to clone things.

Monday, December 11th immunoglobulin clones into lambda.  Immunoglobulin into lambda control.  Badly need an immunoglobulin in 4A delta lac as a positive control for shotgun.  [00:30:00] So try to get a positive control with cDNA.  Trying for blunt-end ligation.  Page 123 Charon 4A preparation for gamma 2b.  [00:31:00] Another partial shotgun on page 129, Wednesday, December 13th, human DNA for partial shotgun.  Was a little bit uncertain that it was good DNA.  Although on the following page 130 [00:32:00] following a repeat of HindIII digestion the conclusion was it was OK all the time.  See three pages later for other ideas.  So that was the control DNA.  A good [shell on?] t131.  And t131 plus HindIII cutting out some bands.  [00:33:00] So let’s review that again that Wednesday, December 13th it was just a test of the HindIII digestion on the ligation carried out on t125.  Still trying to get the strategy to work.  On page 133 pBR322 reclone strategy.  The various self-closure problems with pBR322 and RI pieces suggest the use of unequal end strategy.  [00:34:00] Where possible have two different ends.  Which later on was very much — we used very regularly, even managing to do ligations where there were as many as five different ends to get assembly.  But this beginning of understanding the best techniques.  So to use an RI and a Bam digest to get an RI end and a Bam end which won’t self-ligate.  That was the recloning strategy.

[00:35:00] Attempt at partial digest on page 135.  And quite clearly DNA looks fine.  As comment on page 134.  Query double the enzyme and repeat or use less DNA per slot.  But it’s quite clear that the digest can be varied by varying the time with the enzyme.  Partial digests are really quite possible.  [00:36:00] The obvious reason being to try to get bigger clones that include more than one enzyme site.  So I decided to do a 15-minute digest and a 10-minute digest cloning into delta, into 3A delta lac.  Looking for a 12-kb fragment, trying to get bigger pieces.  Page 137 Friday, December 15th, with the digest looked at [00:37:00] on page 139 looking quite nice.  So here on page 141 preparation of DNA for gradients.  Each gradient can handle about 200 micrograms on the SW41 rotor, etc., etc.  Set up that material.  And then the results on page 143 that — very good gradients and good digests with different sizes of fragments.  [00:38:00] Was an insert which must have been used for making a slide that [pool for?] 4A cloning [2, 3, and 4?], which are partial digest fragments ranging in size from about 20 kb to roughly about 5 or 6 kb.  An attempt to get bigger clones.  4A delta lac was rather heavily contaminated with E. coli DNA (inaudible) phage DNA page 145.  [00:39:00] Wednesday, December 20th.  And Jerry Slightom found that.  So trying to clean it up.  Some DNA that didn’t seem worthwhile on page 147 and 149.  Coming to the end of the book.  Recent low levels of hybridization to megaplates suggest the need for changes.  Looking back at the autoradiograph on s105 and s128 shows that the washing off of label is not temperature-dependent.  Nor is it salt concentration-dependent.  At least at 0.6 times [S of C?].  [00:40:00] The general conclusion being that it was the on reaction rather than the off reaction which was trouble.  So try hybridization in 20 times [S of C?] or 20 times [S magnesium C?].  That’s [SDS?] sodium citrate or [SDS?] magnesium citrate.  Jerry tested the idea that it works better [with the probe or something else is off?].  Going back to new hybridization retest Wednesday, December 27th.  The conclusion is [00:41:00] the usual technique looks very good.  But there were some problems with the experiment.  And so a whole bunch of new solutions were made on page 153 and 152.  Ready for further work in the coming days.  And that is the end of book t.