Interview with Marshall Nirenberg
     Tuesday, August 242002
Topic: Research,Genetics,SociaVPolitical Activism
      Interview by Eric Boyle

Eric Boyle (EB): Since a lot of the Digital Manuscripts Program collection covers the
latter portion of your career, I'd like to discuss briefly the general trajectory of your
research from the neuroblastoma and hybrid cells and chick retina work to Drosophila
and homeobox genes in the 1980s and beyond. Now in your interview with Jim Tabery
last year you touched on your earlier work with Drosophila and homeobox genes. Can
you discuss this transition and how or why it took place.

Marshall Nirenberg (MN): Well the first transition really took place between the time
we deciphered the genetic code and the time I decided to go into neurobiology. And the
logic to that was that there are two systems in biology that process information, genes and
the nervous system. And I was always interested in the nervous system anyway, although
I didn't know much about it. So, that was really one of the big factors that made me
interested in it is that I didn't know much about it. So, that was the logic for switching
from basically genes and protein synthesis in the genetic code to neurobiology. Now
when you make a change like this it was a big change. It was so easy to do the work in
molecular biology and protein synthesis related to messenger RNA and the code. It was
like turning a crank. I thought I could do it with one hand tied behind my back. But I
gave all of my postdoctoral fellows and all of the problems we had going to Tom Cassidy
who was a former postdoctoral fellow. He took over all of that work so I was free then to
explore neurobiology. Basically the thinking behind it is that it is really fun to explore
and the only way to really discover things is to jump in and ask question and begin to
explore the field. It's a risky business though. I mean you're giving up all of the things
that took you a long, long time to set up-all of the procedures, the methods, an
established laboratory, basically. So you do it with the knowledge that there is a lot of
risk involved in doing it. Nobody wants to fail. I don't want to fail. And yet there is a
real possibility of doing that. So you weigh different thoughts and ideas in doing it.

So, anyway when I decided to go into neurobiology it was a fantastic and wonderful
experience because I didn't know anything. You're starting from the beginning, really
from the beginning, and it was just a wonderful sense of learning, exploration, of
freedom. You know I didn't have the responsibilities then because I had given away all
of my responsibilities. It was a wonderful learning experience. I decided, eventually, to
explore two different paths. One was neuroblastoma cells. Neurons are non-dividing
cells and I thought it would be nice, it would be wonderful, to have a simple system of
clonal cells lines that had the properties of neurons and which grew synapses in culture.
And so I switched to, started to study, a tumor of neurons called the neuroblastoma--a
mouse neuroblastoma transplantable tumor. We did a lot of work with the neuroblastoma
cells. Gordon Sato, also about the same time cloned the same mouse neuroblastoma, the
C-1300 that we did. And I also began to explore C. elegans, nematodes, as a simple
system. It turned out that the only invertebrate that you could culture in a known


medium, a defined medium, was a nematode. There were a number of species of
nematodes that had been used at the time and this was before I knew that Sydney Brenner
had also picked nematodes to study. So, we worked a couple of years on nematodes
making mutants and so forth. I never published anything about the nematodes and
eventually I thought that both systems were very promising but Sydney Brenner was
doing very beautiful work with the nematode system and I thought that I could only
handle one or the other, not both simultaneously.

So I switched to neuroblastoma cells and our initial objective was to clone mouse
neuroblastoma cells, characterize there neural properties, and determine whether they
could form synapses. If they had sufficient neural properties, they made neural
transmitters for example, to be able to form synapses and use it as a model system for
studying synapse formation and information processing by neurons. We spent a lot of
time doing this. It proved to be a very fruitful thing. Takehiko Amano came to the lab
about that time. He was a superb tissue culturist from Japan and he was the one who
basically cloned all the lines of neuroblastoma cells. And we had many, many lines of
neuroblastoma cells. And then I began to set up with some postdoctoral fellows who
came to the lab who wanted to work in this area, set up biochemical assays for various
neural properties, like choline acetyltransferase that catalyzes the synthesis of
acetylcholine, or acetylcholinesterase which catalyzes its breakdown, or a number of
other assays, also some enzymes of catecholamine biosynthesis.

We also began to study the electrophysiologic properties of the cells and we soon found
that with the many different clones that we had that some clones synthesized
norepinephrine, whereas other clones synthesized acetylcholine. This was the first time
that acetylcholine had ever been found as a neurotransmitter in neuroblastoma cells and
that was exciting. Later it was found that in humans, also, some neuroblastoma cells
synthesize acetylcholine. So we had both cholinergic and ademergic neuroblastoma cells
and many neuroblastoma lines that would synthesize either transmitter. And we explored
the properties of these cells in great detail and we found that the neuroblastoma lines that
would synthesize acetylcholine could form synapses with striated muscle cells in culture.
Later we found a clonal line of muscle cells that we could use with clonal neuroblastoma
cells to form synapses. We could innervate every muscle cell on the plate. We found that
many of the neural properties were regulated by the level of cyclic AMP in the cells. We
found that if we elevated cellular cyclic AMP for a number of days, like five days, we
could shift the cells from a relatively undifferentiated state to a differentiated state where
they behaved like neurons. They acquired voltage sensitive ion channels, sodium
channels, potassium channels, calcium channels, and this then could from action
potentials. We showed that they could release transmitters and they innervated muscles
cells. If you reduced the level of cyclic AMP you could reduce the effectiveness of the
synapses. You could regulate the efficiency of the synapses by the level of cyclic AMP
that the cells possessed. We used a number of techniques to elevate cellular cyclic AMP.
Like we found that the cells had prostaglandin E-l receptors that were linked to activation
of adenylate cyclase when they were stimulated. So, we would add prostaglandin E-l to
the culture medium that they would then activate the endogenous adenylate cyclase,


which would catalyze the synthesis of cyclic AMP. You could raise cyclic AMP levels
over a period of days. These effects were slowly acquired and long lived.

I've always thought we were looking at some form of memory, some type of memory,
although I couldn't prove it and still can't prove it. But I think the ability to turn synapses
on and turn synapses off at will and the length of time, the slow acquisition of the ion
channels and the other machinery that are necessary for transmitter release and effective
synapse communication, all indicated that this was a form of memory. That if you
lowered the cyclic AMP levels gradually this cell became less able and finally unable to
communicate synaptically. This was a very interesting type of system to work in. You
could use these cells to study the biochemistry as well as the electrophysiology of the
process-a good model system. And that's what my aim was, to make model systems
that were useful for neurobiology.

We used another technique to rescue gene expression. Many of the cell lines were
defective in synapse formation and so we fused the neuroblastoma cells to other cell types
from the normal nervous system hoping to rescue gene expression for neural properties.
We made somatic cell hybrids by fusing cells of different kinds and we made many kind
of somatic hybrid cells that proved to have interesting properties. We found that many
cell lines had specific defects in synapse formation and we characterized those defects.
These are useful system, I've sent these cell lines to investigators all over the world for
many years.

EB: One of the things that was most striking from looking through your papers is that
you have letters from Japan and Europe, all over the world.

MN: We sent hundreds and hundreds, thousands probably, of cell lines to thousands of
investigators all over the world over many, many years. They've been used in many ways
so that's, I think, really been a very useful system.

One of the things we did I did with Werner Klee who was an investigator at the NIB and
a superb biochemist. Werner was interested in opiates and he wondered if the
neuroblastoma or the somatic hybrid cells had any opiate receptors. We screened many
of the lines, actually I've got thousands of cell lines, and really it's really a cell bank. It's
probably more neuroblastoma or hybrid cell lines than exist anywhere else in the world.
And so we screened many of cell lines looking for opiate receptors. And lo' and behold
we found one of our most well studied cell lines, the NG-10815 cells, which is a
neuroblastoma-glioma hybrid cell line, was loaded with opiate receptors. Once we found
that and we were able to count the density of receptors on the cells we wondered what
would happen if we cultured the cells in the presence of morphine. For example, would
they become addicted, would they become dependent on morphine. This led to some
really interesting experiments. We found that when you culture the cells in the presence
of morphine that morphine inhibits adenylate cyclase activity, so the level of cyclic AMP
drops precipitously. But then gradually if you continue to incubate the cells over the
course of a day the level gradually comes back to the normal level. Now, the reason that


it comes back to the normal level is because the activity of adenylate cyclase increases
during this day's incubation. But then the cells are dependent upon morphine to inhibit
the enzyme because if you withdraw the morphine, the level of adenylate cyclase activity
shoots way up and the amount of cyclic AMP synthesized really goes way up and then
only gradually falls back to the normal level. So, we proposed this dual regulation of
adenylate cyclase-morphine inhibits adenylate cyclase and the cells gradually become
dependent upon inhibition of adenylate cyclase and the basal level comes back to
normal-and this is equivalent to withdrawal phenomenon of the opiates. Other have
confirmed these findings and we also extended the findings to other neurotransmitters.
We found exactly the same thing with the muscarinic acetylcholine receptors. If you treat
with carbachol, it inhibits adenylate cyclase. Gradually the level of adenylate cyclase will
come up to the normal level. Then the cells are dependent upon carbachol to inhibit the
enzyme. The same thing with the alpha receptors. I think it is a general phenomenon.
We don't truly understand everything about this phenomenon even today because we
asked the question: are we synthesizing more enzyme or simply activating the enzyme?
The results showed that it's mixed. Mostly you're activating the enzyme but there is
some synthesis of the enzyme, increased synthesis that does occur that counts for about a
quarter of the increase in adenylate cyclase activity. Three quarters of the increase is
simply activation of the enzyme. I don't know how the enzyme is activated like that, that
remains to be determined.

EB: That's what I was going to ask you.

MN: I don't know. And I really wondered at the time. That work really raised questions
in my mind about possible treatments for opiate addiction and dependence. I never
pursued the clinical aspects of the findings but I always regretted not having pursued that
because I think that might have been a very fruitful pharmacologic approach to the
problem of opiate addiction. That problem really interested me for a number of years.

Then I became interested in a different kind of a problem. In the sixties it was proposed
by the grandfather of neurobiology, Roger Sprerry, in a theoretical article in PNAS,
strictly theory, but an interesting theory, that you could label, you could give every cell in
the retina an address, a molecular address with two kinds of gradients of molecules in
right angles to one another. That was a very intriguing suggestion. The reason he
suggested this is because it's known that in the brain there are various retinal topographic
maps that topographically reproduce the position of neurons in the retina so that you can
recreate a cohesive picture of the outside world. There are multiple topographic maps
like this that are produced and nobody knows how they're produced, or what the
molecular bases of the maps really are.

So, I thought of a way of asking: is there a molecular topographic map that exists in the
retina? And to study this we used monoclonal antibodies. Monoclonal antibodies are
produced with a clonal cell where each clone makes a different kind of an antibody in
culture, and so we used as an antigen cells from dorsal retina, basal retina, temporal
retina, or ventral retina--different topographic regions of the retina--and made monoclonal


antibodies and then made thousands of different clones, screened these clones and asked
the question does this antibody recognize an antigen that is preferentially distributed in
one region of the retina rather than the other retina (what Sperry had predicted, had
suggested). This was a test of Sperry's prediction. And lo' and behold we found an
antibody that recognized a molecule that we called the topographic that was distributed in
a dorsal-ventral gradient all the way across the retina. And it was a big gradient, like a
thirty-five fold gradient.. We purified the antigen, the molecule that the antibody was
recognizing, and found it was a small membrane protein that's present in highest
concentration in the dorsal retina and lowest concentration in the ventral retina.

This was really a fascinating affair and the interesting thing to study, because the way the
retina develops is concentrically, the oldest part of the retina is the center of the retina and
then you have rings of neurons that are laid down, concentric rings as development
proceeds. And how you form a dorsal-ventral gradient with that kind of developmental
history was really, I didn't understand it, couldn't understand it at all. We also found out
that the cells remember the amount of protein that they're supposed to synthesize because
we could trypsanize the retina, separate the cells (when you treat them with trypsan, a
protealytic enzyme that destroys the antigen) and then we would culture the cells and we
would take cells from dorsal retina or ventral retina or in between and do the separating
we found that they remembered how much protein to synthesize. And it didn't require
cell-cell contact, it didn't require contact with cells that had less or more of the antigen.
Nor was it regulated by mixing cells from ventral and dorsal retina-cells had a memory
of how much protein to synthesize. That was really exciting because this was the first
time a protein had ever been found that was distributed in a concentration gradient across
the entire tissue. I didn't understand and still don't truly understand how a grade of
synthesis or accumulation of this protein could occur, what the mechanism regulating the
synthesis and or degradation of this protein was. What did I understand? What the
function of the protein was. We tried to clone DNA for this protein. We purified the
protein. That was quite a job, to get enough retina to purify the protein and identify the
protein.

One interesting thing, during the course of this study we found a chick embryo with three
eyes. Two eyes in the regular place and the third eye in the middle of the forehead and
we asked whether the third eye also formed a gradient and it did. All three eyes had
exactly the same gradient. But we failed to clone DNA for this protein and that was an
important question because if we could have cloned it we could have identified the amino
acid sequence of the protein and would have given us a tool to use, a very important tool,
for further studies. For some reason we were not successful in cloning it.

At that time, soon afterward actually, there were a lot of reports of gradients of proteins
that had been found in Drosophila. I thought that to really understand this problem you
have to go to a simpler system where you have genetics that you can use. You can use
genetics as a tool. Drosophila has been studied for a hundred years almost, ninety years,
and there is a tremendous amount of genetic information that is known and wonderful
genetic tools that can be used with Drosophila. And that's the reason I switched to


Drosophila. Plus the fact that I saw a paper by Michael Levine and his co-workers on
evenscate, which is a homeobox protein that was distributed quite remarkably in some
neurons in the developing embryo and not in other neurons (highly specific expression of
the gene regulation of homeobox genes), genes that bind to DNA and recognize the
sequence in DNA that turn genes on or off, so they're gene regulators. So this is an
important class of genes and to find them quite specifically distributed in specific sets of
neurons was quite a remarkable observation. At that time there were seventeen
homeobox genes that were known, that had been found in Drosophila and it was a
burgeoning field of study.

I was really intrigued by this expression of a specific gene regulator in a specific subset of
neurons because we had been trying to find things like this with monoclonal antibodies as
a tool and here it was in Drosophila. I had never worked with Drosophila before but
when Yongsok Kim came to my lab as a postdoctoral fellow immediately after he got his
Ph.D. in Korea, I suggested to him that we look for new homeobox genes in Drosophila.
I designed a whole set of all the nucleotides that could be used as hybridization probes to
screen the library for new homeobox genes. These were oligonucleotides to highly
conserved regions of the homeobox. The homeobox is a hundred and eighty base pair
region that encodes a conserved sequence of sixty amino acid residues that fold in a
characteristic way that have a turned helix confirmation. And this is the part of protein
that actually binds to DNA and recognizes the sequence of DNA. And so they are
relatively constant, conserved regions within the homeobox. And so Yongsok is a superb
scientist and in a very short time he had discovered four new homeobox genes in
Drosophila which he called NK-1,2,3, and 4. And they all proved to be extremely
interesting.

I've continued to work with NK-2 because first of all it was distributed primarily in the
developing nervous system and it was expressed so early that I thought it might turn on
neural development, might even be the first step, the commitment step and that later was
shown to be true. It turns on neural development in the ventral part of the neuroectoderm
in a stripe, an anterior-posterior stripe of cells, and it initiates neural development in this
stripe. It's really interesting the way the nervous system is formed in Drosophila. There
are four anterior-posterior stripes which are next to one another, of cells, and they all
originate differently, in different mechanisms of origin. Eventually they form neurons
that will form the ventral nerve cord, which is like the spinal cord in higher organisms.
It's a really interesting story how they arise, how they regulate one another. NK-2
initiates neural development in the most ventral portion of the neuroectoderm. Another
homeobox gene, a different one, initiates neural development in the intermediate
neuroectoderm. Still a third homeobox gene, a different one, we don't know what
initiates neural development in the dorsal neuroectoderm but MSH is required for
specification of the identity of some neuroblasts in the dorsal neuroectoderm. So, there
are four different genes or proteins, they're all gene regulatory proteins, that are required
to initiate neural development to form the ventral nerve cord. In higher organisms, in the
early neural development, the homologs of these genes are found in the same relative
position during early development in what gives rise to the spinal cord. I think the basic


mechanism has been conserved in evolution. So, anyway, that's how I got into
Drosophila.

EB: It seems like Drosophila, again, is giving you that model system that you were
talking about earlier.

MN: That's right, that's exactly right.

EB: So, you mentioned sequencing a couple of times with the oligonucleotides and again
some sequencing was involved with the homeobox genes themselves that were
discovered in Drosophila.

MN: Yes, right.

EB: So, I'm wondering if there is a relationship between this sequencing work and what
we've been hearing so much about in the genome project and their sequencing in the
1990s.

MN: Well, I think the genome project and the project on other organisms are extremely
important. My goodness, the only instructions for how you build an organism are really
present in the genome. The Drosophila genome has been sequenced also. This
information is stimulating research in all directions, all fields, so it's really an amazing
time right now where there are thousands and thousands of genes that have been
discovered and in many cases nobody knows what the function of these genes are. So,
the functional genomics is really important now. Currently we're screening the
Drosophila genome using RNA interferons which is a new method which will give you
the equivalent of oligonucleotides by injecting double stranded RNA corresponding to a
specific messenger RNA. This results in the destruction of that messenger RNA and it
looks like a mutant phenotype then. We're looking for genes that are affecting the
development of the nervous system and we've found quite a few genes thus far.

EB: How do you feel about some of the forecasts that have been made for biomedical
applications in genomic research? Do you think the distance between the research that is
taking place now and those biomedical applications to come is perhaps greater than some
people promise?

MN: Well, when you're doing basic research, to translate that into useful products that
have therapeutic usefulness takes a lot of work and a lot of effort. There's a big gap
between the two. But I think, as I said, with the sequencing of the genome that
information is available now is stimulating work all across the board. Virtually
everybody's work is stimulated by it, by the availability of all this information on the
genome, and so I think that this is having a very large effect on research that is being done
currently throughout the world. And it will continue to do so.


EB: Another interesting parallel that I think there is between your own work and some of
the discussion related to the genome project is these concerns about the direction of
research and potential ethical and moral implications, reminiscent of your editorial from
Science in 1967. One of the striking things about recent genetic research seems to be,
again, these ethical questions. From what I've seen, you've taken a firm stand on some of
these key debates in recent years. In 1988, you signed a letter against human cloning
from the American Society for Cell Biology, to the President and Members of Congress.
In 2001, you also supported a statement in favor of stem cell research to George W. Bush.

MN: Absolutely. I think that stem cell research has tremendous possibilities for future
research. It's not quick. Progress won't be tremendously fast, but in the long run I think
that it holds enormous promise for future research. And I think that Bush compromised
on this by allowing research to be conducted with federal funds on the cell lines that were
already established at that time. That's not enough by any means. I think that that kind
of restriction ties the work down tremendously. We need many, many more cell lines.
First of all, there aren't sixty cell lines really available and useful, which he thought that
there was at the time. And you need many more cell lines. I think that's an artificial
restriction that should be done away with. And I think that this approach, using stem
cells, offers a lot of hope as potential therapies for various kinds of diseases-Parkinson's
disease, for example, Alzheimer's disease, even repair of heart attacks, infarcts, that kill
cardiac muscles cells, and other things as well. Ron McKay here at the NIH has shown
that he can culture pancreatic cells that will form islets and will release insulin on
demand, although at a much lower rate than is found in mice. But it works. This has to
be improved, the efficiency has to be improved, even for diabetes. For repair of all kinds
of cells, potentially, they could be done by this stem cell research. But the therapies will
be a long time coming, it's not going to be a rapid crack or something, in fact I think it is
going to be a slow, gradual, incremental thing. But I do think it has tremendous promise.
And I do think that research should be pushed and that the government should support it,
without the restrictions that are currently on that. In terms of human cloning, nobody I
know is in favor of human cloning. I think that's a very bad approach on the instinct and
I don't know anybody who is in favor of human cloning.

EB: Could you talk a little about the difference between human cloning versus stem cell
research.

MN: Sure, I mean to get stem cells you take very early embryos that are only a few cells,
very small, small cells, very early, early embryos and these are usually obtained as a
byproduct of in vitro fertilization therapies typically. And most of them are ordinarily
destroyed. They are routinely produced and most of them, the ones that aren't used for in
vitro fertilization to help parents that can't have children, are destroyed. Normally. And
so to take some of these tiny embryos consisting only of less than a hundred cells, fifty
cells or something like that, that would ordinarily be destroyed and could be used for
research purposes, stem cell research. Stem cell research is trying to use cultured cells
and allow them to differentiate in different ways that they wouldn't normally
differentiate, to make muscle, or nerves, or pancreatic cells, or what have you. To find


out all the steps that are needed for normal differentiation of the cells you need local
hormones, stimuli of various kinds, and many other reactions that we know very little
about right now. Basically, you're asking how do you form an embryo. What are the
molecular reactions that occur between the first cell and the final baby that is forming, the
entire organism. There is a tremendous amount of unknown information in there that
really has to be obtained to have effective stem cells which could ultimately be used for
replacement tissues. You know, I think that whether the work proceeds in this country or
elsewhere in the world it will proceed. It's going on in other countries right now that
don't have the restrictions that Bush has placed on stem cell research in this country. I
think the United States should be the leader in this work instead of being unduly
restricted.

EB: Another interesting thing that I think pops up that is related to a lot of these ethical
or moral questions related to science specifically, but maybe exist on a broader level in
some of the correspondence in the collection at DMP, is that you seem to receive a lot of
letters from a lot of different organizations supporting a lot of different causes-
humanitarian organizations, animal rights organizations, human rights coalitions, things
like this-requesting your support. I was wondering if you could talk a little bit about
your role, or participation, or support of some of those groups and what some of your
ideas are about that relationship between scientists and social or political activism.

MN: Well, I think I've always tried to support causes that I believe in, that I think are
good causes. So, when I receive requests of this kind, or to join with others in favor or
opposed to specific ideas or things, I always try to participate. And I think it's important
to do so. I normally receive lots of letters from various organizations to support different
things. Those that I believe in I always support, I always try to lend my support. I think
that's important to do that. Some people I think, like Linus Pauling, for example, who
was very outspoken in trying to put the ideas that he believed in forth to the public, and I
think that's really important to do this. At the time that he was doing it, for example,
concerning the radioactivity and the bomb, things of this sort, he took at that time what
was really an unpopular view and publicized the reasons why he believed in what he was
saying. And I think he did a tremendous service to the country and to the world actually
by doing this. I think it's very important to do it.

EB: Do you think, possibly at times there have been expectations that might have been
put on you as a result of your being a Nobel laureate as far as your involvement in these
types of things.

MN: Well, sometimes. I mean sometimes I get requests that are completely out of my
expertise and if I really don't understand the topic then I won't either join it or oppose it
because it's really truly out of my area of expertise. But some of the questions really are
social questions and I don't think anybody can say that they're right all of the time. I
mean I know that I've been wrong some of the time but you do the best you can.


EB: Do you think this role of being the expert or having an expertise is the only reason
why people sometimes look for the support of scientists in particular, or is there a greater
sense of responsibility that you feel a scientist has outside of just expertise.

MN: Well, I think that if you're knowledgeable about a particular area, more
knowledgeable than most people about this area, then you should use your knowledge to
impart that knowledge to the public and also it may be necessary for politicians to
understand your thoughts and the reasons for your thoughts, for your particular stand on
things because I think that studying science you have special knowledge that maybe the
politicians don't have that would be useful to them. So, that's the reason to support or
oppose particular projects.

EB: One of the things that is impressive about the body of work that is represented in the
collections is how you are able to balance your responsibilities as a lab chief, as a
practicing scientist, as a figure who is involved with work within the community, and also
different humanitarian organizations.

MN: Well, there is never enough time to do everything I think you want to do and you
have to make priorities, you have to set priorities. But within the constraints of time,
certainly the bulk of my time is spent in trying to do research in the lab and in working
within the lab but these other things sometimes don't take much time and if they're useful
we try to be useful and we try to be helpful wherever we can.

EB: Great, thank you. It's been very helpful. I really appreciate you taking the time to
meet with me today.

MN: Thank you.