Oliver Smithies:

[00:00:00] So this is book upsilon, begins April 29th, 1991, and goes through September the same year.  It begins with a continuation of the ideas being developed in the previous book, and particularly on page 143 of book tau, preparation there was made as already described of the PGK Neo fragment, so this is a fragment of DNA [00:01:00] that contains a PGK promoter, driving Neo gene, and obtained by digesting with SalI, so it’s a SalI/SalI fragment.

And that material is then being used on page 1, Monday April 29th, to make polymers, the idea being to have multiple copies of the Neo gene to help make sure that one could get strong selection for [00:02:00] G418 resistance, and therefore use a higher concentration of G418 and get better discrimination between targeted and non-targeted cells.  This was, this first experiment on Monday, April 29th, page 1, shows that you could get many, many polymers, quite easily, conclusion, “Very rapid ligation; the gel shows the polymer in 2.5 minutes, it’s close to optimum.  Very little of monomer left.  And all the other numbers are represented.” 2.5 minutes of ligation. [00:03:00]

And more work on polymers on Tuesday, April 30th, page 5.

Beginning to worry about, or think about, what if the proteins were involved in the recombination of events protein-protein interactions might affect the frequency of targeting. [00:04:00] So page 7, Tuesday, April 30th, has a diagram of what was our standard correction of the E14TG2a deletion mutation from Martin Hooper, and which contained exons 3, 4, and 5 of the HPRT gene, but was missing the promoter in exons 1, 2, and 3, and PMP8, which was the standard correcting vector, which by ordinary omega-type recombination could give corrected gene with a promoter, exons 1, 2 and 3 now inserted [00:05:00] in front of the deletion, which was everything prior to exon 3.  EG2a had all of the exons after 3, 3, 4, 5, et cetera, but it was missing exons 1, 2 and 3 in the promoter, and thinking about if there was proteins involved, that might alter the frequency of recombination.

Another attempt at ligating PGK Neo, following page.  More minis, Friday, May 3rd, page 13. [00:06:00] “Abandon all.” (laughter) The last of the minis.

Probably, Saturday May 4th, page 50, υ4 and 5, are probably Neo dimers.  Nothing else is worth keeping, and 4 and 5 are not either.  So, re-checking minis on the next pages.

Following up an idea that had been described on page υ7, although I don’t think I commented on it, but I’ll go back to it, on page 7. [00:07:00] That was the idea of protein interaction.  And Frank Grosfeld had information regarding a locus activating region that, the locus activating region could be bound to the promoter region of the beta-globin gene with a protein-protein interaction of a preamble, as it were, of the work that is very common is of bringing together by protein-protein interaction, bringing DNA sequences together that were apart [00:08:00] by proteins.  Here’s a comment, assays for assemblies of complex, protein, transcript, used these days for looking at protein-protein interactions, the idea being to insert a locus activating region into the construct, and to see if it improved the transformation and production of protein by the beta-A gene, for — this was to help Ed Shesly with his work on beta globin. [00:09:00]

And so on page 23, or opposite page 22, is a note on the plasmid obtained from Tim Townes which contained the hypersensitive region, locus activating region of the beta globin gene.

And, Tuesday, June 4th, inserting the locus activating region into the beta construct.  This is the hypersensitive region [00:10:00] of DNA, DNAs activity showing lack of histones, and that generally assumed to mean a lack of histones in the region, or lack of heterochromatin at least, and a diagram of what, how we can insert the hypersensitive 2 region on Kpn-Cla fragment, and insert it into the poly-linker, with the construct being diagramed on page 27, Wednesday, June 5th, inserting the 1.9kb locus activating region [00:11:00] upstream of the Neo gene.

The continuing experiments to find the right product.  Usual technical problems, [00:12:00] Saturday June 8th, page 37, “Processed and ran the υ35 minis with the conclusion that quit; need to start again with a partial NcoI digest, because a complete digest had cut out the region of importance.”

So here we are partially digesting on Wednesday, June 19th, page 43.  And combining that with digestion with a second enzyme, SnaBI on page 49. [00:13:00]

So we’re on page 50. [00:14:00]

So, the work continues on Wednesday, June 26th, page 61.  The next steps are processing these minis, #2D, 4K, and 6D, “All cuts are fine.  These were various digests with ClaI, Kpn, and the two together,” et cetera.  “And, 2D looks the smallest, and better bets are 4K and 6D.  Let’s go with 4K for the first try.” [00:15:00] And this was Kpn/ClaI digest, “Need to digest Ed’s plasmid also with Kpn and ClaI in order to insert the fragment that is being done on the following page 63.  So, page 63 is cutting the recipient and page 65 is a prep gel for the fragment, the locus activating region fragment that’s going to be put into that site.

And the ligation then on page 67, Thursday, June 27th, putting the thing together, “Ligations are excellent.”  And transformation, [00:16:00] page 69, and various minis looking pretty good.

So, check on the constructs, Monday July 1st, page 71, and diagram of all the various things that are being made.  There’s a Neo emycin gene in front of the beta globin locus, and a Neomycin in front of the beta globin locus with a deletion in front of it, Neo-delta-beta, and LAR Neo-beta, and LAR Neo-delta-beta, the various constructs that could be made reasonably clear.  And looking at a gel on page [00:17:00] 73, Monday, July 1st, and various ones look to be OK.

More tests on page 75.  “Very bad gel.  Buffers were exhausted, repeated the gel.”

The view of what should be done next on page 77.  One buffer was no good. [00:18:00] And Saturday, July 6th, page 81, 6D and 4K are identical with two NcoI sites, and Neo-beta has two NcoI sites, et cetera, et cetera, but things are looking good.

2D still not correct for the following reasons, Saturday, July 6th.  Various interpretations of results with confirmation of these conclusions on Sunday, July 7th, page 85, but the 1.9kb locus activating region [00:19:00] fragment is present, but 2D now looks better, but that was an error.  And they are into 2D, Sunday July 7th, page 87.

And we will check on the products of this work on Tuesday, July 9th, page 91, the conclusion that the LAR went in fine, but the 2D is not correct.  This is a logical mistake. [00:20:00] But eventually, plan is, we’ll go to Ed Shesly, Tuesday, July 9th, page 93, υ93 Neo beta, υ3 LAR Neo beta, and then Neo delta-beta, LAR Neo delta beta, with IMAPs on the left hand side, with a note the change in the LAR position.  Four plasmids are there illustrated quite clearly. [00:21:00] [00:22:00]

So, on Thursday, August 22nd, page 97, I’m thinking of a new targeting scheme.  The idea is to use DNA specifying telomeres on the 5’ and 3’ ends of a targeting construct to reduce non-homologous targeting without affecting the homologous recombination, the idea being that if there were telomeres on the ends, there wouldn’t be recombination that was non-homologous. [00:23:00] Making a transformation to get a construct that would be called pBS-shuttle. [00:24:00]

Consider using a shuttle vector which could be expressed both in E. coli and in a mammalian cell.  This is pBS HUT ER pB shuttle, considering using that and making it.

And so on Saturday, August 24th, I’m making a Sal digest of this shuttle vector.  And, continuing with the idea that telomeric piece on Saturday, Sunday [00:25:00] August 24th, 25th, redigesting on page 105, and dimerizing the BS, the Bam-Sal piece of DNA with the telomere, telomeric sequence on it, and making a construct illustrated on page 109, with two telomeres pointing in opposite directions, which call B-tel2, with a Sal site in between. [00:26:00] So it’s a special type of telomeric sequence.

Retransforming on page 111, unlikely.

Doing more than one thing at once, of course, so Wednesday, August 28th, we go back to conversion of HPRT Bam-to-Sal, necessary to get the DNA that I wanted to change a Bam site to a Sal site, which in our days would be very easy, but in those days [00:27:00] was not as easy as it is now.

But here we are on page 117, trying to get this telomeric sequencing but the conclusion is that all are the vector.  So a whole bunch of different minis being studied on Sunday, September 1st, page 119, [00:28:00] photographed as usual, et cetera, #33 and 34 look somewhat promising.

And looking at the various tests of 12 plasmids for HPRT-Bam now, and all except #9 are OK, so proceed with 3 and 6, although later on, abandoned.  Just, soldiering on. [00:29:00]

Check on #3 and #6, page 125, Tuesday, September 3rd.  #55 and #57 also.

Trying again, Friday September 6th, 129 page to get the telomerase-2 repeating.  Going on with two candidates on page 135.  Neither of them are correct.  Neither have [00:30:00] Bam-Tel, S-Tel, Bam, because Sal didn’t cut either of them.  Most likely, I got a dimer of the vector.

So we see on page 137, Friday, September 13th, typical entry.  Tel2 again, again, again, just repeat and repeat until you get what you want.

And some thoughts on Friday, September 13th, page 139.  Some improvements possible.  Delta-left Neo by Neo delta right, very old [00:31:00] plasmid-by-plasmid recombination system.

And typical September 17th, Tuesday, 143, a bunch of candidates digested with Bam and Sal, the conclusion, “None are correct.  Start again with a new strategy.”

And a new strategy is listed on Wednesday, September 18th, page 145, just how to get to this product, which is telomerase, Sal site, HPRT-Neo [00:32:00] telomerase.  More oligos needed to change a Bam site to a Bam-Kpn-Bam site, introducing Kpn site into Bam, page 149.

But not very happy with the result, page 153, Friday September 20th, transformation HPRT Neo Tel, telomere and conclusion, “Looks very poor, but streak out #3 and try.” [00:33:00] So here’s a possible rescue of HPRT Neo Tel on Monday 23rd of September, page 155.  Conclusion is OK.  And so picture 3B, and more thoughts on introducing Kpn site into the Bam site of the construct.

So ending the book on Wednesday, September 25th, with a preparative digest to make [00:34:00] a fragment, Sal, #4, telomeric sequence, Kpn, et cetera, et cetera.  And, with satisfactory precipitate with ethanol, back-up for Kpn and complete digest.  But there is the gel, and no interpretation of the results on page 163, the end of book upsilon.