157 
DR. FREDRICKSON: I wonder if any member of the technical experts who 
are here would care to enlighten me a little bit further about Dr. Sedat's 
comment that these experiments that we are proposing guidelines for need not 
be done, that information could be obtained in a different way. Dr. Berg? 
DR. BERG: Unless John has some understanding or insight that I lack 
here, Arthur Kornberg's system for replicating DNA will replicate whatever 
you put in the test tube. Therefore it is necessary to start with something 
that is pure in order to amplify that component. So if you take the total 
DNA of a mouse you will make more of the total DNA of a mouse, which is no 
big trick because you can isolate as much DNA as you want. 
But if you are asking if you want to isolate a specific segment and 
produce large quantities of a specific segment that occurs with a frequency 
of one part per million, there is no way, using Tn vitro DNA replication 
that I know of, for doing this. 
This whole technique, if you really want to look at it roughly, is a 
magnificent column. It is a powerfully effective fractionating technique 
for taking individual segments of the chromosome and distributing them over 
a million test tubes. But you can't do that by starting with the total DNA 
and using a DNA replication system _in vitro . You will make more of what you 
have got, but if you started with a mess you end up with a mess. 
DR. SEDAT: Well, say you were to fragment with the restriction en- 
zymes, I mean, just dilute this sufficiently so that it would be equivalent 
of a single burst experiment, and if you had a sufficiently sensitive repli- 
cation system, you could well amplify it. I mean, the DNA replication sys- 
tem at Princeton actually will start with something with a fragment of FI, 
for example, and you will get something ten times the size of before. This 
represents a very substantial increase because it can be cleaved again with 
restriction enzymes, which really amplifies at least the DNA. 
Agreed, it is a little more difficult, and perhaps a lot more diffi- 
cult, but it is not inconceivable that this may well be the direction that 
one could go. 
DR. FREDRICKSON: One more minute. 
Mr. Madansky, very briefly, please. 
MR. MADANSKY: I believe also it is not just a matter of breaking up 
total DNA and putting it all into a DNA replication system. There are 
methods for pulling out those DNAs that you are interested in, even with 
control regions on either side. It is a technique developed by Ray White. 
If the DNA you are interested in has a messenger RNA or an RNA of some 
sort, you can put that on the column and then you can get a double-strand 
piece of DNA to make a triple helix under certain conditions with that RNA. 
[298] 
