Making the First Recombinant DNA Molecule

 

                                        Overview

Dr. Berg reminisces about the ideas and experiments that led to the creation of the first recombinant DNA molecule.

                                       Transcript

00:00:11.14 I'm Paul Berg.
00:00:12.23 I'm a retired professor at Stanford Medical Center.
00:00:17.02 And I'm going to reminisce a little bit
00:00:19.16 about how the recombinant DNA idea emerged.
00:00:25.09 I had been carrying out experiments
00:00:27.20 with a colleague at Stanford, Charles Yanofsky,
00:00:31.17 and as part of the experiments we were doing
00:00:33.21 we were using phage-mediated transduction,
00:00:36.22 the delivery of genes from one strain into another
00:00:41.00 using a virus as the carrier
00:00:43.21 of that new genetic information,
00:00:47.21 a very important technology
00:00:50.03 that had been widely used,
00:00:51.25 and very important in the development of our current notions
00:00:54.07 of the molecular biology of microbes.
00:00:58.15 But, at the same time,
00:01:00.21 I had decided to move from my work in bacteria
00:01:03.22 to try to learn something about mammalian cells.
00:01:07.14 It was largely to try and test
00:01:09.15 whether the ideas that had prevailed
00:01:12.02 about gene function in microbes
00:01:14.22 was also applicable to mammalian cells.
00:01:17.01 And I took a year off and went to work
00:01:19.23 at the Salk Institute with Renato Dulbecco,
00:01:22.13 and to work on the tumor viruses.
00:01:24.23 And I chose the tumor viruses
00:01:27.01 because they were known to have very small genomes,
00:01:29.06 and they were known to transform normal cells
00:01:31.18 into cancer cells.
00:01:33.29 And that seemed like a nice place to start
00:01:36.28 because, again, using viruses
00:01:40.09 that have limited genetic information,
00:01:41.25 being able to follow the expression of their genes,
00:01:44.25 seemed more tempting than trying to study
00:01:48.01 the whole mammalian cell genome.
00:01:52.29 During the time I was down there,
00:01:56.00 it was reported that
00:01:58.23 when SV40, or polyoma, infects mammalian cells,
00:02:02.28 virus are produced,
00:02:05.04 but amongst the progeny there are virions
00:02:08.04 which contain only cellular DNA, and not viral DNA.
00:02:13.03 And that was very similar
00:02:15.09 to the bacterial transduction system P1,
00:02:19.03 which infects E. coli
00:02:21.09 and comes out as P1 particles,
00:02:23.05 but containing E. coli segments of DNA,
00:02:25.17 and it seemed to me interesting to think about the idea of
00:02:29.25 trying to develop a transduction system
00:02:32.19 for mammalian cells to facilitate
00:02:34.27 the genetic modifications of cells,
00:02:36.20 much as we had done with microbes.
00:02:39.03 But very quickly it became clear
00:02:41.15 that a single virus particle
00:02:43.22 could not contain very much DNA,
00:02:45.19 in fact the limit is about five kilobases.
00:02:48.14 And the prospect of being able to find a gene
00:02:51.23 that was present in the human genome
00:02:54.20 of three billion basepairs,
00:02:57.01 and find it in a virus particle
00:02:59.15 that contained five [kilobases],
00:03:01.19 seemed pretty small.
00:03:03.15 So... but I liked the idea that we could actually
00:03:05.26 introduce new genes into mammalian cells,
00:03:08.09 and we could do it without a virus particle.
00:03:10.14 We could actually take the SV40 DNA,
00:03:13.00 which was known to integrate into the cell's DNA that it infects,
00:03:17.18 and link up to it some foreign DNA,
00:03:21.12 anything we wanted to put into the mammalian cell.
00:03:24.00 And so the first idea was:
00:03:27.02 could we get two different DNAs,
00:03:29.07 join them together covalently,
00:03:31.12 and then use them as a way...
00:03:34.10 as a transducing agent?
00:03:36.07 And we had in the lab a plasmid,
00:03:40.13 lambdadvgal.
00:03:42.04 It was a small DNA molecule
00:03:44.08 about ten kb,
00:03:46.07 which had lambda genes
00:03:48.16 and appropriate genes that it could replicate in E. coli,
00:03:51.05 and linked to it were the three bacterial genes
00:03:54.00 that encode the gal operon...
00:03:56.01 that encode the three genes
00:03:59.17 necessary for metabolizing galactose.
00:04:01.11 And the idea was to take the two DNAs
00:04:03.26 and join them together.
00:04:05.22 And we had to develop a way
00:04:08.04 to join DNA molecules together,
00:04:10.11 and we employed what was already known as
00:04:12.29 cohesive ends in the bacteriophage lambda.
00:04:15.27 The bacteriophage has these ends
00:04:19.03 which are cohesive,
00:04:20.24 that is, the two ends can be joined
00:04:22.25 to each other to form a circle,
00:04:24.13 or they can be joined to different molecules
00:04:26.03 that have the same kind of ends.
00:04:27.19 And so the idea...
00:04:29.11 could we synthesize synthetic ends
00:04:31.12 onto the two molecules we wanted to join?
00:04:33.19 And that was already known how to do it.
00:04:37.00 There was an enzyme that would polymerize A or T
00:04:40.09 onto the ends of one molecule,
00:04:42.13 and A or T on the other,
00:04:44.29 so you have two molecules,
00:04:46.23 one which A ends and T ends,
00:04:48.24 and if you mix them they come together.
00:04:51.08 And that was the scheme we set out to do,
00:04:53.05 and that turned out to be pretty straightforward,
00:04:56.09 I think it was actually done within less than a year,
00:04:59.05 and so we were able to make the first recombinant DNA,
00:05:02.18 although it wasn't called that at the time.
00:05:04.25 The first recombinant DNA molecule
00:05:07.05 was part SV40 and part lambdadvgal.
00:05:11.03 And the idea was to be able to
00:05:13.19 propagate these molecules in E. coli,
00:05:15.20 and maybe make mutations in the SV40 sequence,
00:05:19.04 but also to transfect them into mammalian cells
00:05:22.21 and to test whether we could get expression
00:05:24.20 of the exogenous genes.
00:05:27.08 So, that was the first idea
00:05:30.15 and the first fulfillment of that idea,
00:05:33.00 and that of course led, ultimately,
00:05:35.05 to the evolution of the whole recombinant DNA technology,
00:05:38.11 except that it became much easier
00:05:40.21 to join DNA molecules together
00:05:43.01 through cohesive ends created by restriction enzymes.