Oliver Smithies:[00:00:00] This is the beginning of Book r, starting December 13th, 1977, and going through February of 1978. It begins with a map of pBR322, and a growth of pBR322. And inserting a dT tail, and a dA tail eventually, and a dA tail to try to get an ability to clone dA on dT-tailed cDNA. There was a very informative [00:01:00] blot there showing the actual results of these hybridizations to different replicas. It’s quite nicely preserved. It’s an Ann Blechl page, I think, and the film can be aligned very nicely with the plaques to show how the system worked. No doubt, why we scanned the book, that we have a copy of that. I will try to check. In the scanning, there is an autoradiograph with a row of heavy spots on the right-hand margin which matches [00:02:00] preceding image, go back. The preceding image being on the right-hand side, if one can — I can’t see the page number, but it’s probably defined. It’s this one, do we know what that is? We need a name for it. It’s there. One can, with a little ingenuity, one of the scans has a whole bunch of circles across the top, and along the right-hand margin, which are different plaques. On the left-hand side, there is a scan of a page that has yellow markings [00:03:00] on them. That is the one that these can be matched. That left-hand image can be matched to the radiograph on the right-hand side of the next page. So the right-hand side of the next page in the scan can be matched to the left-hand side of the image to superimpose and see the phage — it’s easy to see with the natural objects, which are inserted into the book anyway, at this place, so anybody interested would be able to see how it worked. It’s quite spectacular. Very heavy signals, which were markers, [00:04:00] which turned out later on to be actually a problem sometimes, that we used a marker which probably cross-contaminated things, as we might see later on.
So, continuing then, pBR322 is being worked with, single and double cuts on Tuesday, December 13th, with a couple of gels there. And HindIII digests, and so on. Learning my way around pBR322. Terminal transferase on pBR322. And, exonuclease, etc., [00:05:00] etc., technical methods of getting some label, with the conclusion that, “The lambda exo degraded too much. Go to Pst digestion and tail directly.”
So, the next one, at the next page is a Pst scheme with a conclusion that it worked fine. Terminal transferase then, on a page that’s been flagged, Pst on page 9, and terminal transferase on page 11. These, being the critical steps in getting bacteriophages, getting plasmids, pBR322 to work. [00:06:00]
For example, on page 10, the conclusion is that “The dT tails went on very well, estimate about 250 per end, assuming mobility equivalent to double-stranded DNA. The dCs were not so good, so, remake the dCTP and try again.” Anne Blechl had had some successes, as reported on Friday, December 16th in the cDNA department, reports a recent success in Allan Kinnibrough [sand?], in Jeff Ross’s lab, so [00:07:00] Anne Blechl is reporting success in Jeff Ross’s lab, who, Jeff Ross was always a happy collaborator and helper for me; he worked in McArdle Institute, and was a good friend to work with. A lot of work with cDNA and hemoglobin. And, this is one of this protocols, listed.
Some problem, evidently, on page 14, with a big comment, [00:08:00] “Nuts! Adding a fifth of the volume of 2M hydroxide overnight to denature,” but why it’s “Nuts!” is not indicated.
Because on the next page, it says the conclusion, “Good T tails, A tails much less. But virtually no annealed product.” This is HF, Harvey Faber 13[13?] DNA, etc.
Going on with these, trying to calculate specific activities on page 18. [00:09:00] And trying to determine how many radioactive ends had been added to a reaction that was analyzed on page 19.
Thinking, there’s some confusion about which shotgun was being used, evidently, because the handwritten ink is positive human shotgun plaques, phages on page [00:10:00] 21, and also on page 23, crossed out, and as it says, “Later [rat?],” so evidently, there was a confusion about which one was working. [Dave?] is devoted to trying to get the radioactive probes to hybridize to bacteriophage plaques containing fragments of human or mouse, or rat DNA.
A nice, handwritten note from Ann Blechl on page 26, talking about Ann Blechl filters there on [00:11:00] the third floor in prewash since Monday night at 70 degrees, etc. There are four bags, and she goes on for a whole two-sided note about this. “Each piece is coded, so you needn’t worry about mixing if you keep the probes separate.” So this is Ann Blechl.
A test of a globin clone on Wednesday [00:12:00] December 28th, a globin clone SPD-8. It might just be OK. And, diagram suggesting what the insert might look like. And going on with these technical problems, page 36, “Hinf seems to work OK; can’t be sure of RI. It probably did work, since it is tough enzyme, very doubtful if the insert could be seen. Think again on this one.”
Going on with this SPD-8 [00:13:00] globin clone, and with a conclusion on page 38 that, “The clone has a Bam site. It could be the left end.”
Working with this one particular clone on page 41, and 43. [00:14:00] Conference with Fred about it on page 43.
Conclusions with the following page, “No HindIII in the insert; no Bal cut in the insert. Can’t see [Kp and hyper p’s?] for sure, though it ought to be visible,” so not looking very good. A new clone test. These are called “R47,” and there’s another writing. [00:15:00] Looking for the DNA in these lysates.
Here’s the time when it became clear that what we thought was human was really rat, because on page 49, Friday December 30th, a comment says, “Check of the data shows that SPD-8 is Hg rat shotgun. But some new human shotguns, SPO, are positive anyway, so it didn’t matter. [00:16:00] With some thought that these clone lysates are giving such excellent DNA, try making bulk DNA for sequencing.” And then, new clones to test on Saturday, December 31st.
And, at the end of the year, here we are on Sunday, January 1st with a comment at the top of the page, “End of a good year, one of the best!!” It had indeed been a very productive year, with the flying, and learning bacteriophage, and getting colonies.
With, on page 50 opposite, the Sunday, January 1st entry, the conclusion that, “MP9, [00:17:00] 23, A, C, and D are clear clones; MP2 20 A looks to have a trace of a clone. Can’t see anything in MP7,” etc. Going on with this general type of work.
Retests of potential clones, and a beautiful gel on page 52 with a comment, “Transferred and cut into 4.” And how the transfer was done, before and after transfer. [00:18:00]
However, beginning to worry about a contaminant; I mentioned the possibility of a contaminant, on Sunday January 1st, “On the way home to supper, I had the thought that SPH-23 and SPD-8 are probably Charon 3AHb4 contaminants.” That was because of what turned out to be a silly use of a Charon DNA-containing hemoglobin insert as a marker, with the result that you could get the phages which were really replicas of the marker.
And, so on page 56, the conclusion is that, “Subject [00:19:00] to an internal check, SPH-23 is 3AHb4. It has no RI site.” In other words, we’d contaminated our system with a positive clone of cDNA.
Continuing this worry, that, with a re-check of SPD-8 and SPH-23 on Monday, January 2nd, “None of these clones have a RI site; therefore they all are equivalent to Hb4.” In other words, we again were isolating a contaminant. [00:20:00] I remember that we were somewhat, to say the least, disappointed, and realized that we had to be very careful about this in future work, that if one was isolating the same thing over and over again because it was contaminated, and it turned out when doing the Nobel Prize-winning work on gene targeting, that that was again a worry that there might be contamination with a positive test clone that was being recovered in the real experiment. We’ll come to that in due course.
So new and better clones of SPD-24 on [00:21:00] Tuesday, January 3rd. But the first tests were no clones in these, hybridization.
Worry about the probe. “This probe may be no good.”
So here is the hypothesis for the false positives on page 65, Wednesday, January 4, “[Wes?] hypothesis for the false positives: the slave plate pickups are loaded with Hb4. Some washes off and rehybridizes to Sharon 3A delta lac plaques, and these then give a positive [00:22:00] signal with the hemoglobin probe. And this, comments on how to test this idea.
So, retesting human clones, of which there are, one, two, three, four, about a dozen clones being tested. These [old/whole?] lysates are now very poor, but [#32?] for sure has something large in it, and #42 has a small clone, etc., etc.[00:23:00] Retesting these with fresh lysates on Friday, January 6th, and saying more or less the same, but 32 has a large clone; 42 has a small, etc. And I said, [dot blot?] example, photograph rather well on page 70. This is the rat growth hormone test, for the rat [Mp2 SPE 20A OS?] rat is a potential rat growth hormone clone. I don’t remember [00:24:00] that we ever followed that up very well, but the hybridization overlay is nicely illustrated on page 70, with Polaroid photographs of the film over the replica plaque.
So Saturday, January 7th, deciding to screen all of the [picture?] areas from human shotgun experiment. And the new array for the test was made. [00:25:00]
Hpa1 digests on Sunday, January 8, page 77. And these were 16 different samples, different clones, 16 in all. [MP1-1 through MP1-16?]. MP7-1 through -16, and MP1 again, now clones 17-32, [00:26:00] lysates on page 79.
So, on Thursday, January 12th, page 81, the [anneal individual?] probe hybridization of about r79 test arrays. [00:27:00] So probe 1 is globin cDNA, and probe 2 is for rat growth hormone; probe 3 is RI, and probe 4 is HCS.
Continuing with Ann’s clones from the r69, Book r, page 69 series, review [00:28:00] of the results on Saturday, January 14th. Provisional positives being listed on page 84. So, continuing, Saturday January 14th, a new hemoglobin probe test on all available DNAs, so the arrays were set up in six copies of r49 and r83 phages, and different roles. And array 87-1 samples, and another in the [00:29:00] reverse order, etc., etc., just details about how the arrays were made.
With the autoradiograph on page 88, and the superposition of the film with the array diagram, with a Polaroid image there, showing which were positive. [00:30:00] Which I can see three, four, five positives, strongly positive clones. The results after six days are the same, only r23, 1, 2, and 3, and r47 3, 4, and 6, are clearly positive after six days’ exposure.
And continuing with different experiments of the same general type. Rat growth hormone tests on page 95, tests of the plasmid probe on page 97. [00:31:00] And another array test on page 99, plasmid lambda clone test. Some with, this is looking for, it looks like looking for plasmid with insert, ones with lambda clones, and with Mini col E1 DNA, and pf bn3, different plasmids. [00:32:00]
Recheck of Mini col E1 and pf bn3 plasmids on page 103. Returning to the [00:33:00] hemoglobin-4 contamination on page 107, the second film, of SPD-8, and SPH-23, shows that one was the deliberate positive control, “Marked, but forgotten!” And the other was the accidental pick-up of an alignment mark, “Sad, errors.” So that, it was a good lesson.
Total number of plaques on the human MP-1 and MP-7 is about 400,000. Total on the rat, about 1,200,000. And of the rat, [yet finish?], the number is [00:34:00] about 400,000.
Here’s a comment of Natalie, that’s Natalie Bernstein. She was a longtime tech, and a great friend of our long-term secretary, Frances Mann. “Making good DNA, 113, delta lac DNA, 14mg.” [00:35:00] Made from that bulk lysate.
Tests of [minis?] from the macro experiment, showing that the method scales up perfectly, etc., etc., how to make larger amounts of material from starting plaques.
Nick translating plasmids on January 31st, page 117. And here we are, immunoglobulin gamma-2b, check, check. Jerry Slightom has prepared a [chem mark who’s?] PMB-9, immunoglobulin plasmid. This seems to be one of the first [00:36:00] mentions of Jerry Slightom, but he was a very productive postdoc that stayed with us for quite some time.
The conclusion from this experiment on page 118 is that, “PMB-9 has an RI site and a Bam site at pH 7. As expected, lost the RI site, but now has two Bam sites,” so things were looking reasonably well-understood.
Continuing on page 121, “Check on the Bam site and the [00:37:00] RI distances on PMB-9.” The conclusion is that, “The BamHI site is very close to RI, since the [PY?] is almost unmeasurable,” whatever that means.
Check on Thursday, February 7th, of KMs, dA and dT tail, PMB-9. And, conclusion that “The tail stuff is good.” Continuing with the same type of work. [00:38:00]
Repeat of the Hpa digest of r113, delta lac DNA. “Difficult to understand the results,” etc.
So, continuing with cesium chloride binding again on February 3rd to try to purify the DNA. On, [00:39:00] (inaudible) 2nd banding DNA from this, February 3rd, the experiment, “Recovered about 2mg total DNA from delta lac DNA, from page 113.”
The second banding check on the following day. But, some question about it being necessary to do the second. [00:40:00]
And trying to get away from having to do the banding on page 141. On 140, the conclusion that, “Differential precipitation works fine. Try it also instead of first banding,” in other words, trying to avoid having to do the banding in the ultracentrifuge with cesium chloride.
Full-scale precipitation and tests without banding the following day. The conclusion of, “Result, that page 144 had both cesium chlorine binding, and precipitation are needed, cesium chloride for slower impurities, and precipitation for the faster impurities.” [00:41:00]
On page 147, I obviously went to Australia. Now I remember when I was in Australia, renting an airplane and flying around a little bit in Australia. Had to pass the written test, which first time I tried, I was one point short, and someone illegally, the examiner said, “Why don’t you take it again tomorrow; you’ll probably pass,” which I did, and so I flew a little while while I was in Australia.
Now I’m beginning to think about dideoxy sequencing, on Monday, March 20th, instead of — this is the Sanger method, instead of the Maxam and Gilbert method of sequencing. [00:42:00] Checking it out with Bam/KpmI digests, on a gamma tube, MPC-11, pH 7 plasmid, which were called, gamma tube b, (inaudible) (inaudible) 11 in the future, which had 1100bp Bam/Kpm insert, and into the plasmid, which looks like, maybe not pBR322; it looks a bit big, but in a plasmid. [00:43:00] And with a conclusion that, “Kpm didn’t work. Repeat using Kpm, and then changing the buffer.”
So Kpm/Bam digest again on the following page, and with a conclusion that, “Kpm was working, but the yield of the 700bp piece is lower than expected.”
That ends this book, page 151, March 21st.