Oliver Smithies:

[00:00:00] This is Book q which extends from July 15th, 1977 to the end of December of the same year.  The beginning of the book is looking for cDNA plasmids in order to use these for the genome search, so this is making complementary DNA and inserting it into different materials; in this case, into chi1776.

So, the first page talks about rescuing some colonies and making replica filter papers, a total of, [00:01:00] from what I can see, something like 122 different colonies, replicated.  So one would be used for hybridization, and one is the master colony, which is then used to propagate and purify the particular clone.

Comments on the following page 3, “Results of the first immunoglobulin transformation carried out, and read to two possible positives into the medium,” etc.

Continuing in the same vein, on page 5, it’s talking about inspecting the old hemoglobin plasmids, etc.  And then, at the bottom of the page, is a very nice interruption of vacation, and flying; I got my [00:02:00] pilot license on Friday, July 15th, 1977.  So that’s, what, 48 years ago.  Not 48 — yeah, 48.  Forty-nine.  So, that was time off.

Coming back from work here, then, so there’s a jump from July 14th; the next entry is September 9th, of this, talking about the characterization of mouse immunoglobulin, the heavy chain lambda clones.  So these are clones that are, as illustrated on page 6.  For example, [00:03:00] #14, showing the cDNA, perhaps, 1.8 kilobases long, and talking about Lesley Furlong’s [Hpa1?] map, and one can see that the piece in, what I call the 12.0 piece is the clone-bearer, and indicated by its progressive increase in size.

And there is a map — a gel showing these pieces on the following page 7.  Hpa1 digests, on Monday, September 12.  And new preparations of the immunoglobulin clones, [00:04:00] which there are three in particular mentioned:  Ig#1, Ig#4, Ig#11.  There’s some difficulty with the digests apparently on page 11.  And repeated the digests on page 13, September 13th, and with a lot of results on page — tabulation of sizes on page 12, with the results all over the page, showing the gels, and the plot of the size of, where different bands might occur, and [00:05:00] the size range.

Hpa1 digests, a repeat of — on Ig#4, Wednesday, September 14th.  In summary, and it gives a summary of the pieces there.

Making some probes with radioactive phosphate probe from lambda phages, made, on Friday, September 16th. [00:06:00] With some difficulty of (inaudible), shown on page 23.  “Obviously it did not work.  Abandon these two and threw everything out.  Repeat on Monday.”  Typical of having to repeat things.

Now with BamHI digests on page 27, and then a very interesting page occurs on page 28, and this is the first sequencer that we are obtained. My recollection is, [nine?], I think I might try to call Lesley Furlong about this, that this was a sequence that Lesley Furlong got, and perhaps [00:07:00] the first one she was doing, the end-labeling, using the end-labeling method for determining DNA sequences, Walter Gilbert, Maxam and Gilbert method, where the end is labeled with phosphate, and then sequence is read.  I remember the excitement of getting these first sequences.  So, we tried to translate the sequences shown in this diagram, and didn’t recognize them.  Of course, t hey didn’t correspond to any immunoglobulin sequence, as it turns out, because with modern methods, it was very easy for me to find out where the sequences came from, and it turns out that both [00:08:00] of them are E. coli sequences, so this clone, the sequence, anyway, that was determined, was actually a piece of E. coli DNA, with 98% identity to E. coli.  (inaudible)1917 complete genome.  So, the clone was actually a little piece of bacterial DNA.  But, it didn’t prevent us from being very excited about it, because we didn’t know that.  We thought that we were getting what we wanted.  Of course, eventually we did, but not at that time.

Repeating the Hha digest for acrylamide, and [00:09:00] with troubles in the gel.  This is (inaudible), despite the poor run, everyone can see that HhaI is greater than Hha-delta lac; in other words, #1 has a piece of DNA in it.

And the repeat gel on page 34, and 35, quite an impressive gel.  This is 8% acrylamide gel, so an acrylamide gel, fairly small fragments, were very beautiful.

And I tried to make radioactive probe again, nick-translated on page 37, [00:10:00] Monday, September 26, checking the probe, etc., on the following pages.

And repeating the nick translation to get some hotter probe, and it’s going on over several pages, including Thursday, September 29th, page 43, full-scale remake of probes.  And, running a column to clean up the probes and get rid of the unbound radioactive dNTPs, and so, there were 2×10^6 counts per minute total representing about 2% incorporation.

And repeating [00:11:00] it on October 7th.  And this time, it was only a 2.5×10^5 counts per minute, “Very poor.  Increase the dNTP level.”  And the following day, a repeat, but still not very good.  “Must try phenol.”  So, that didn’t help.  The following page with phenol, Sunday, October 9th, “Total: 0.” Trying to get a probe to work.

Now, back to the use of the probes, plaque hybridization on page 57. [00:12:00] With a result that there’s nothing on any film, virtually no counts on the filters.  So, trying to make a better probe again on the following page, have to repeat and repeat and repeat.  “Nothing worthwhile.  The mRNA which we were using to make the probe, or something, must be wrong.  Especially see previous page.  Likely none of the material is cDNA.”

This problem is emphasized on page 61, where the head of the page is, “cDNA again, again, again, da-da-da-da-da!” Having to try to [00:13:00] get a probe.

But, using [Jeff Ruffs’?] enzyme on page 65, October 19th, I got 2.4×10^6 counts.  So it looks somewhat better, and the column results also better.

So, here’s some work of Ed.  I have to remember which Ed that is. [00:14:00] Had some DNA from each of three immunoglobulin cDNA clones, Ig2, and Ig17.  Each have enough DNA to, “see unconcentrated,” etc., etc.

With a rather strange conclusion on the following page, “None of the plasmids have enzyme sites, or else the enzymes are not active with these preparations.”  That’s BamHI, HpaI and EcoRI, so I think the latter explanation is the more likely.

And, a little comment [00:15:00] on the [flag?] here, on page 7.  — “see Charlie 29,” that’s [Maury?] Airport to Madison to [EBN?], and then an arrow [toe?] was Flight 22, and got my glider rating.  I’ve used it a great deal ever since.  So those are good days for non-work enjoyment.

Doing experiments still on Friday. [00:16:00] “Forgot to denature the probe,” on Saturday in big, red letters.  First test was negative on the following page.  Then rat DNA, the rat was a young, adult male, about six months old from the genetic lab.  Why I was wanting to do that DNA is not immediately clear.  So, I have, half to Harvey, and half to Fred Blattner, that rat DNA dialyzed on the following page.

(inaudible) some complementary [00:17:00] DNA, this is Ann — it says, “Ann’s complementary DNA,” that’s probably Ann Blechl.  6×10^5 counts total, yield only about a quarter of that obtained earlier in the book on page q65.  So, transferring the material, and different methods of doing this, Southern blots.  Comment that, zero weights work.

With, as followed up on Monday, October 21st, with a DNA [00:18:00] transfer that is very much better this time due to no weights, so these seem like the Southern blots which are so — or became so routine were still not so easy in those days when we were beginning.  Yes, and the “Ed” was Ed Sheldon there, because here, he just commented on, Ed Sheldon always uses SDS in his pre-wash.

Interesting that on page 87, [00:19:00] “I have an old container of HpaI believed to be lot 3 from Bio Labs,” so it was early days of Bio Labs too, if that was only their lot #3.  “Enzyme is OK, and it is Hpa,” is the conclusion of that experiment.

Beginning of restriction tests of human shared DNA, and Charon phage, and pCR1 clones, immunoglobulin clones. [00:20:00] But the conclusion was, from the November 2nd experiment, “No digestion.  Repeat with more enzyme.” (laughter) And then the next page, “RI was over-killed.  Repeat with less enzyme.”

And then, recheck of restriction sites on two of the immunoglobulin cDNA clones, Ig17 and Ig19.  There were some problems in distinguishing between an enzyme that nicks, and an enzyme that cuts, on the bottom of the page, November 4th. [00:21:00]

Looking at this, repeat of HpaI digests on Harvey’s clones with more enzyme and less DNA in some cases, with a conclusion that, many of them are probably just, the Charon 3A-delta lac; in other words, have nothing in them.  “Cannot tell, digestion only complete in three of them.  #1 may have an insert, but more likely, just more partials, partial digests,” which of course, always a confusion.  So not convinced by any [of them?]. [00:22:00]

Some work on rat insulin, rat growth hormone, human chorionic somatomammotropin, Howard Goodman nick translated for probes.  What I was trying to do at that stage is, again, as I have said often, not immediately obvious.

“A final test?” about the [find?] of Hpa on Harvey’s clones, on page 101, November 9th.  General test for inserts prior to packaging the phages.  But, the conclusion on page 100 is,  [00:23:00] “Confirms the previous gel.  All of them are probably delta lac.  But the diagnostic procedure looks good,” even though none of these particular clones were any good.

“Rescued one clone, Ig#2, but very little, if at all bigger than the control plasmid — the control pCR1.”  And on the next page, the conclusion that, “Ig#2 is a small deletion with no extra RI, BamI, or HpaI site. [00:24:00]

Some comments on shrinkage of film, and shrinkage of filters on November 21st.  We now must have had the sequencing going, because on Monday, November 21st, there’s a review of Charon 3A immunoglobulin whatever clones and Ig#1, as a Bam site [00:25:00] is expected, but [Phil Tucker?] had been unable to sequence it in the rightward direction.

And page 113, November 22nd, test of growth on two different bacterial strains, KHO2, which doesn’t have a suppressor, as I remember, versus growth on Dense P Sup F, 40 Sup F. [00:26:00]

And, a protocol for, from Harvey Faber on DNA from lysates.  More tests of restriction sites, HindIII, Kpn, Pst, and AvaII, so enzymes are now much more readily available, but the result looks much more complicated, as shown on page 117, where AvaII, PstI, and Kpn/HindIII mix are tested [00:27:00] with these four possible clones, Ig#1, 4, #11, and #14.  And the conclusion is comparing with Lesley’s gel, Ig#1 must be approximately equal to hemoglobin.  By my measurement, it’s 2%, therefore greater than or equal to a thousand base pairs, meaning that that has a good hemoglobin cDNA, and that’s the hemoglobin clones, approximately equivalent to hemoglobin, the immunoglobulin ones, compares with Lesley’s gel, Ig#1, must be approximately equivalent in length to the hemoglobin clone.  By my measurement, be greater than 1000 bases.  In other words, [00:28:00] we have, I’m concluding that we have a good cDNA for immunoglobulin in the phage.

Trying to get the mouse R1 genomic clones, 12 Charon 3A clones isolated to test whether there was insertion on Saturday, November 26th.  So this is a bunch of different clones, 1-12, digested with [00:29:00] HpaI, HindIII, and Kpn.  And conclusion, “10 were testable, and all have a clone segment.”  The, roughly 250-600bp inserts, (inaudible).

And remaking plasma probes on December 2nd.  Here is much better results, with total 3×10^7, and 3.7×10^7 [00:30:00] counts per minute, various fractions, so this was much better labeling. [00:31:00]

So these are, Friday December 2nd, on page 122, it’s talking about an R1 plasmid, and rat growth hormone, and a human — and “HCS,” [00:32:00] whatever that HCS is, a different material.  And then R1 is a mouse genome, R1 genomic clone.  So, trying to isolate good clones.

To check the unlikely possibility that these clones are all big, on Saturday, December 3rd, conclusion that “One of them is OK, but is delta lac,” and etc., “repeat.”  So this is looking at these different shotgun clones of [00:33:00] mouse, rat, and human shotgun clones, their method being to take RI digests or whatever, and ligate them to the pages, and then plate, and then look for hybridization of the probes to the different phage plaques, and no doubt, we’ll talk at some stage about the method which we use.  Using the “Tucker Box,” which was a rather nice combination of ideas, where —

So I’ve gone back to Saturday, December [00:34:00] 3rd, page 127, where plate lysates from mouse, rat, and human shotgun experiments were being considered, and in this case, basically the method was to take the Charon 3A bacteriophage, cut it with a suitable restriction enzyme, and then hybridize, just total RI digests of mouse, rat, or human DNA in the RI site, or whatever site we chose, and then grow up phages and plate them on, what we’ll call “mega-plates,” they were actually cafeteria tray which were large enough to allow the plating of many, many plaques, [00:35:00] many cloned phages.  And then, take a replica of the phages and hybridize it to the probe, and then using what we called the “Tucker box,” we would be able to identify which plaques corresponded to the radioactive signal.  And this was Phil Tucker’s idea, he’d been trained in a physics lab where they were doing X-ray crystallography, and they were used to assembling, super-imposing images, by looking at two images at the same time, projected with a half-silvered mirror, and so, as we said in the paper published later, that the Tucker box was [00:36:00] based on a technique widely used for crystallographic model building, which proved indispensible for picking candidates from the mega-plates requiring P3 containment; that’s to say grown under safe conditions as far as the safety aspects of cloning were considered at that time.  So, this is, on Saturday is talking about making the [phage?] lysate.

And 12 clones [00:37:00] being looked at on Sunday, December 4. Conclusion, “Still no RI cuts, or sticky-end problems,” so some problems were with the experiment.  So, for example, looking on again at digests from several different clones, human, rat, and mouse clones — no, total DNA plate lysates, hoping to see the phage bands, but [00:38:00] not enough phage-derived DNA in these plate lysates to be detectable above the E coli DNA, so direct [looking?] was not [possible?].

We’re concerned the way the different type of digestion with the EcoRI restriction enzyme, what they called “RI star cuts” were being considered on page 134. [00:39:00]

Back to the more important tests again, on Tuesday, December 6th, “Test of a human shotgun.  Twelve independent second-plating plaques grown by Ann Blechl, and processed by the preceding protocol,” etc.  So these were different plaques, which were looked at again with PstI digests, Ig #1, 4, 11, and 14, which we talked about before.  Conclusion that “The Pst results are confirmed, but can’t find any extra piece.  No evidence of any inserts.” [00:40:00]

An airplane comment again on page 141.  I remember this particular thing.  “Coronary arteriogram, and [mitro/mitral?] competence test completely normal.  Medical certificate (inaudible) (inaudible) time.”  That was because, with hypertension at that time, the Federal Aviation Administration required rather extensive tests to establish that one could get a medical certificate, and I had a coronary arteriogram, which was rather unusual for a normal person.  And I remember the cheers when the arteriogram showed that I had normal coronary arteries, but I also remember seeing that my heart stopped, [00:41:00] and the comments on the nurses there, “Cough, cough, cough, cough!” in order to restart the heart. (laughter) So it was quite an experience.  But anyway, I passed the test, and my medical certificate, only a matter of time.  I really did want to fly.

So test of the human plaques continued on December 8th.  The arteriogram was December 7th.  And, some modeling of labels, unfortunately.

Now looking at some phages which had been used with dA and dT tails. [00:42:00] More pages of gels and digests following.

Going back to the human shotgun again, December 11th, Sunday, size range of the shotgun clones is approximately 5%-15%.  So RI should work well.  And, the conclusion is there, that “#32 has about 5,000bp and cuts cleanly.  [One can?] see at about 2,560bp.” [00:43:00] So that it was a quite pleasing result.

And continuing Sunday, December 11th, with the dA/dT tail material.  Where, (inaudible) [mistakenly digest?].  The conclusion that, “The result is better in the sense of the enzyme activity, is more complete.  But still can find no evidence of proper annealed product.” [00:44:00]

Going Monday, December 12th, “Kate is right.  The dT is difficult to see, what has happened.  It is difficult to see what has happened, but there’s very little reconstituted delta lac in any of these,” etc.  So Kate must have been one of Fred’s students.

And that is almost the end of this book, which ends on thoughts on immunoglobulin work, Tuesday, December 13th.  “Since the immunoglobulin bacteriophages have gone so relatively slowly, and we really need a plasmid for genomic work, I should remake  [00:45:00] the immunoglobulin cDNA plasmids with a more up-to-date vector, and also,” etc., etc., so thinking about using pBR322 as the vector, subject to having the right size.  pBR313 is the next best, both available, but pBR322 with a deletion chosen to improve its growth proved to be the better choice.  And that ends Book q, a very productive book.


[00:00:00] This is a little addendum to Book q.  Later on in the books, we come across P plasmid HCS-1, and I mentioned that I don’t know what it is.  But in fact, it is described on page 97, Tuesday, November 8th, of Book q, and it stands for “Human Chorionic Somatomammotropin,” and it’s for a plasmid from Howard Goodman that we were using as a test of our ability to obtain probes that were specific.  So, he sent us rat insulin, pRI-1, rat growth hormone, pRGH-1, and human chorionic somatomammotropin, pHCS-1 [00:01:00], for testing our ability to make nick-translational probes, so it’s very straightforward in the end.


So it’s very clear, but it’s not significant.  And so we’ll stop that topic.