Oliver Smithies:

[00:00:00] This is book little u beginning on Friday, December 27th, 1978 and going through to the end of March in 1979.  So it starts with reconstruction experiments, etc. to find the right conditions for hybridization.  So Ann Blechl made a megaplate with about 290,000 plaques of 4A delta lac, containing about 520 [00:01:00] 4A-51.1 phages.  So it was 520 out of 290,000.  That would be roughly 1 in 500 or 1 in 600 phages.  And tests showed that the titer of 4A-51.1 is not altered.  About 11,000 4A total.  All plaques are very small as usual.  Abut half size of usual.  About a third the size of the usual plaques.  [00:02:00] And so we’re [lifting?] these and try hybridizing the plaque (inaudible) to try to make sure that we can find positives that we know are there in a big plate.  On a megaplate which was in the — [packed?] in the cafeteria trays.  And on film one of the comments that the 3A controls looked fine.  But no sign of the 4As yet at three and a half hours.  And here is the result at 53 hours.  [00:03:00] Which maybe will find the answer.  So January 1st, 1979 a new one.  Conclusion book u page three experiment is that 4A takes more probe.  Counting the number of counts per square centimeter, 200,000 counts per ml and 20 microliters per square centimeter.  And the probe, etc.  Calculation of how much probe to use.

So here the beginning.  [00:04:00] An experiment.  With a comment on flying from Miami to Pamplona to Sanguesa to Madrid with Field Morey.  Shotgun with sheared DNA on Tuesday, January 23rd.  Desirable to have a completely random 24A shotgun (inaudible) try shearing, binding, etc., etc.  Ideas of how to do that.  So a shear experiment was started on January 24th.  [00:05:00] Starting the shear.  It’s not listed about how the shear was done.  But the shearing was carried out for up to 120 minutes.  About 1 milliliter 6,000 rpm.  [00:06:00] This is a blade used for a [stirrer?] with a horizontal blade.  And the 6,000 rpm is the medium setting of the rpm of that particular [vertis?] shearing machine which could nominally go up to 23,000.  But here I say that about 18,000 seems to be the top speed.  [00:07:00] Blade is one supplied.  And in shearing one can indeed that it is progressing on the gel shown on page 12 but the conclusion is that need to increase the rpm, which is done then on the following page at high.  Sheared at 9,000 rpm in the cold room with a conclusion that not too much but the starting material was smaller than the overloaded gel suggests.  So try one hour at 7,500 instead of 9,000.  So shear was continued on January 26th, page 17 [00:08:00] with gel showing the very clear effect of the shearing.

And Saturday, January 27th, page 20.  Some comments on how to correct the ends.  That it might be better to continue filling in the DNA to the end with a polymerase and use an [00:09:00] S1 nuclease to get rid of any that was overhanging in the other direction.  So an overhang of 5’-phosphate could be filled in by polymerase.  And an overhang of a 3’-hydroxyl could be removed with S1.  So there’s DNA (inaudible) polymerase being talked about on page 21.  And some comments on what other people use.  [00:10:00] Some thoughts on what to use.  With a comment on page 23 of the various considerations.  This is very complex to cover the cases not covered by partial RI.  Therefore try straight blunt-end ligation to the phage and forget about RI [cutout?], etc.  I was feeling it was too complicated.  [00:11:00] Continue in this vein.  Test ligations on January 29th, page 31, with a comment there that I see 2 microliters of EcoRI JS number two 20 units per microliter.  That was an EcoRI preparation that Jerry Slightom made, that number two preparation, 20 units per microliter, which we used for at least 10 years afterwards.  It was a very good preparation of EcoRI that Jerry had made.

[00:12:00] Further ligation tests on Tuesday, January 30th with a Biolabs technical data sheet talking about the T4 DNA ligase which requires ATP.  And used for about 10 units per microliter, get cohesive ends, and 2 unis per microliter to ligate blunt ends.  [00:13:00] And so testing the ideas with — Jerry has some 4A largely cleaned of inserts that has been treated with RI and a gradient fractionated and tested for ligation with partial RI from the previous book and try these conditions.  And page 39 looking at the results that the ligation is working well.  The DNA polymerase degraded a lot.  Polymerase was — ligase didn’t work in this order, etc.  So attempt at this other experiment.  [00:14:00] Worrying about enzymes, checking on them on page 43, Wednesday, January 31st, because DNA polymerase appeared to digest the DNA.  Decide to check Biolabs [DNA1?] and also Boehringer phenol preparation.  And S1.  Conclusion S1 is about 30 units per microliter.  New and old ligation buffers are equivalent and 15 [1 5?] ligation for one hour is only a little poorer than overnight at 4 degrees and so forth.  Try it in the Biolabs.  [00:15:00] Pol I is good, try it again.  And so continuing to try to get clones.

Quick check on S1 February 1st.  Technical data on DNA polymerase I from Biolabs.  Ligation tests on the sheared material February 1st.  Continuing on page 51 that S1 degrades the DNA substantially.  There is a small but greater than 10% [extent?] of ligation with the Biolabs pol I plus ligase, etc., etc.  Attempting to get ligation.  [00:16:00] Effect of kinase with RI samples and shear samples on page 55.  Ligations with Charon 4A page 57.  Just more and more tests.  Ligation in capital letters on Monday, February 5th.  The three precipitated DNA samples taken up, etc. with distilled water and pooled at (inaudible) so on and so forth.  The conclusion is that DNA polymerase unstuck the ends of lambda.  Therefore ligate lambda plus RI before this reaction.  [00:17:00] Ligation was, however, poor anyway.  So try to understand it.

Ligation of lambda on page 61.  Ligation of cohesive ends went fine.  So some recovery of fragments (inaudible) on February 6th.  [Real ligation?] of that material.  Sheared DNA, etc.  Different tests.  With a disappointing result.  [00:18:00] Conclusion, no improvement.  Further rescue attempts.  Phenol-extract and try again.  Page 66.  Following day doing it.  Still no ligation.  Four days off for surgery on page 71.  I don’t remember what the surgery was.  Back to work on Friday, February 16th.  Remake of sheared DNA.  [00:19:00] Conclusion, it was sheared enough.  Perhaps a little too much.  Gradient ultracentrifuge on Saturday, February 17th.  Getting a pool from different fractions.  Test experiments for ligation in 4A with blunt ends and with RI ends on Sunday, February 18th.  Conclusion that all the enzymes worked apparently.  Although the polymerase wrecks the DNA fairly badly at 37 degrees.  [00:20:00] Continuing on testing various ligation functions on February 19th.

I don’t know the right spelling of Klenow.  On page 83 it’s Klebnow.  JS points out the DNA polymerase large fragment is equivalent to Klebnow which should be Klenow.  DNA polymerase I is what I need.  And so some schemes about what might happen.  [00:21:00] On rereading Kornberg it’s doubtful that DNA polymerase I will work, but try it anyway.  And also S1.  Meanwhile ligate 4A and clean up the – clean up further.  Test of DNA polymerase I as anticipated.  This is Tuesday, February 20th.  As anticipated the DNA pol I is no different from DNA pol large.  Test S1.  And then work on ligation conditions assuming the polymerase is working correctly.  Still trying to ligate and fractionate.  [00:22:00] Ethanol precipitation instead of gradient centrifugation being tried on page 93.  No detectable DNA in the supernatant.  [00:23:00] But it was precipitating the DNA all right.  It appears to have been precipitated OK.  Looking at the gel on page 92.

Final check on the 4A phage on page 95.  Post RI on u93 and post (inaudible) u45 material (inaudible) and post ligation, etc. tests.  [00:24:00] Continuing blunt-end tests on Sunday, February 25th.  Never tired of repeating experiments.  Overnight ligations on page 99.  Results confirm earlier tests.  No sign of blunt-end ligation.  Or no blunt ends.  The RI ligation is OK.  The pol I answer must be wrong in some way.  Test S1.  So doing so on the next page.  Looking at the S1 apparently working.  [00:25:00] Liking the idea of testing an enzyme with known blunt ends HpaI on page 103 and 102.  New blunt end substrate tested.  There is no — [Ousway?] test on page 105 and 104.  There’s no blunt-end ligating activity in this T4 ligase under these conditions.  Try other ligases or other conditions.  Not working.  Blunt-end ligation continued on page 107.  Not able to find the conditions.  Temperature tests on page 109 [00:26:00] with a comment that 37 degrees doesn’t work.

And similar results on — increase enzyme level and RNA ligase, etc. on page 111.  Conclusion.  Can’t do it with this T4 ligase alone or this buffer under these conditions.  In other words the experiment didn’t work.  Testing T4 RNA ligase.  No.  That’s DNA ligase.  There’s RNA ligase and DNA ligase.  Difficult to read.  It is actually an RNA ligase (inaudible) this is [00:27:00] possibly the increased concentration of T4 ligase which works.  It’s due to T4 RNA ligase impurity.  So try adding the T4 RNA ligase, etc.  Why I would have thought to test RNA ligase beats me.  But it doesn’t look as if there was anything very satisfactory.  Saturday, March 3rd something worth a little bit cheering up.  Page 115.  Higher DNA concentration test.  Lower enzyme.  [00:28:00] Where the control had nothing and then increasing amounts of — not clear here.  That can’t be — I’ll try that again.  Saturday, March 3rd higher DNA concentration being tested.  Yes.  So it is a concentration test.  But the samples are diluted so that the gel shows them at the same concentration.  So ranging from 125 micrograms per ml to 500 micrograms per ml.  And there are definite increase in ligation with the — because polymers are appearing.  So the conclusion is at last a visible effect, exclamation mark.  [00:29:00] Slight but significant ligation at the highest and next highest DNA concentration level.  So test an increase in the T4 ligase concentration.  So I could see at last some ligation of the larger fragments.  Smaller fragments don’t appear to have changed appreciably.

So trying a higher concentration of the ligase with a comment on conclusion on page 116 that the ligation is not much altered by increasing the concentration of the ligase.  But the results are so low level that it would be wise to try the methylation [theory?].  So rescuing the DNA.  [00:30:00] Test for testing whether IVS, intervening sequence S1, or intervening sequence 2 versus transcription.  And assume we have all of A gamma.  In other words now we’re talking about bacteriophage in which we think we have the whole of A gamma gene at this point.  Showing a map of exon [1 1 to 30?] and exon [2 31 to 121?] and exon [3 to 122?] onwards with an RI end [00:31:00] and what looks like an Mbo — maybe both ends are RI.  But anyway this is an mRNA of the structural gene.  So we have the mRNA of the structural gene with an MboI site, MboII site, and an RI site.  And then similar sites in the genomic DNA clones.  Testing for expression [the thought?].  The less elegant idea but probably a lot easier on the next page.  Since we can now buy a Bam-RI linker at last.  [00:32:00] Linkers (inaudible). Talking about the constructs.  These are relatively easy to construct on page 124.  But an assay for their function is much less available.

Well, a digression on Tuesday, March 20th, page 129.  Xenopus DNA prepared.  Next few pages are devoted to finding out what I’ve got with a spectrum and a comment that this is not a DNA spectrum.  So that the product [00:33:00] must be a protein complex.    Xenopus digestion of the Xenopus DNA within the [ditto?] with a comment that the DNA — page 133.  That the DNA is fine.  Large enough and digestible.  The concentration must be about 0.5 micrograms per microliter.  And it digested well with RI.

Now going back to DNA for Southern blots from 30.5 talking about some AKR female mice [00:34:00] and liver DNA.  In preparing DNA with sarkosyl and proteinase K.  And a higher temperature.  Series of RI digests on the Xenopus laevis DNA on page 137.  No digestion.  Going back to continuing with it, and [00:35:00] time series on RI on Xenopus laevis DNA.  Not understanding it.  So here was an experiment on page 139, Tuesday, March 27th, using Jerry Slightom’s buffer and his EcoRI with the Xenopus laevis DNA 5 minutes up through 60 minutes [00:36:00] 6.7 units per microgram.  And very clear digest even in five minutes.  So the conclusion is that I overshot the amount of enzyme.  Therefore decrease the enzyme two- to threefold.  There must be some inhibitor which is critically overcome.  Meaning that something in the preparation is inactivating the EcoRI and then when that inhibitor is removed by adding the RI then the reaction can proceed.

Repeat of the time series with Xenopus laevis [00:37:00] DNA but with the conclusion that this DNA is too sheared to give the best types of [particles?] but can go ahead with it.  So instead went back again to making DNA from Xenopus laevis red cells.  March 29th, page 143.  Half a ml of packed red cells available, etc., etc.  And preparing the DNA.  Sarkosyl and proteinase K followed by extraction with phenol, hydroxyquinoline, and precipitation.  Adding diethylpyrocarbonate.  [00:38:00] So this is now with this material.  Again Xenopus laevis partial digest.  This is still using the — on page 145.  That’s still using the earlier DNA.  The digestion went nicely.  Close to ideal digestion schedule.

[00:39:00] So this was Xenopus laevis partial digest on page 145 and they were onto a sodium chloride gradient.  And samples were taken from the gel and quite good fractionation is observed.  I’m dying.

Interviewer:

Some tea?  Do you want some more tea?

OS:

[00:40:00] I’m all right.  I got a little bit left.  I’m not really dying.  Sounds like it.  So continuing to purify the red blood cell Xenopus laevis DNA on page 149 and 151 which is — and on page 152, the end of the book.  The DNA are still OK but it will take 2 to 4 microliters of EcoRI to digest 10 microliters.  That’s the end of book u.  [00:40:55]