Capecchi Mario

Mario Capecchi

University of Utah, Howard Hughes Medical Institute

Evans, Martin

Martin Evans

Cardiff University

Smithies, Oliver

Oliver Smithies

University of North Carolina at Chapel Hill

For the development of a powerful technology for manipulating the mouse genome with exquisite precision, which allows the creation of animal models of human disease.

This year’s Lasker Basic Medical Research Award honors three scientists who developed a powerful technology that allows scientists to create animal models of human disease. With this technology, researchers have engineered mice with conditions such as atherosclerosis, cancer, high blood pressure, and cystic fibrosis, allowing the study of many debilitating disorders. And the same technology is uncovering the secrets of normal biological processes as well, revealing, for example, how the nervous system develops or how immune cells collaborate to quash microbial invaders.

This progress depends on the ability of researchers to manipulate the genetic material of mice with exquisite precision. Scientists can disable — or knock out — a gene, and they can also change its properties in more subtle ways. These procedures generate animals that carry specific genetic alterations, and recreate the underlying cause of a human disorder or uncover the role of any gene of interest.

Inactivating a mouse gene takes two steps. In the first, scientists snip out its middle in a test tube using conventional tools of molecular biology. The challenge then is to replace the intact, functional gene in the mouse chromosome with the modified version created in the test tube. This genetic swap requires the introduced DNA to find the corresponding DNA in the chromosome. Less than two decades ago, conventional wisdom held that this task was impossible in mammalian cells, and that DNA could insert only at random sites.

At the time, knocking out a particular gene was about as successful as tossing a dart into a heap of spaghetti and hoping it hits a particular strand in a certain spot. Scientists could add DNA to mouse cells, but it landed in random places. They couldn’t yet target DNA to a particular site in the chromosome — in other words, to a specific gene. Today, hitting a bull’s-eye is a standard maneuver.

By building on several decades of fundamental studies in mouse embryology and molecular genetics, Mario Capecchi, Martin Evans, and Oliver Smithies brought this work to fruition and delivered the technology of gene targeting in living mice to labs around the world. Scientists can now create and breed mice with particular pathologies, study them systematically, evaluate the functions of genes in an intact mammal, and dissect even the most complex processes.

How cancer helped

The seeds of this technology grew out of a bizarre form of cancer of the gonads, called a teratocarcinoma. Most tumors possess characteristics of their specific cell of origin, but teratocarcinomas display features of many different tissues. Skin, muscle, hair, and even teeth compose these masses. Researchers showed in the 1960s and 1970s that this chaotic mixture of tissue types is all derived from an immature type of cell with many potential fates, called an embryonic carcinoma (EC) cell.

In this respect, EC cells from teratocarcinomas resemble embryonic stem (ES) cells from normal animals, which scientists knew existed, but could not isolate. ES cells develop soon after fertilization, and they, too, have not yet committed to any particular tissue type. A number of labs showed that EC and ES cells are close cousins in other ways as well. Early embryos turn into teratocarcinomas instead of mice if they are implanted into sites in an adult mouse other than the uterus. And most importantly for the enterprise of gene targeting, either type of cell injected into very young embryos grows into normal parts of the fully developed mouse. The resulting ‘chimeric’ animals carry some cells from their ‘normal’ mother and father, and some cells from the mouse that donated the EC or ES cells.

In principle, if scientists could alter the DNA of a cell that would contribute to a live mouse, they could create an animal with any desired genetic change. The process would involve altering the cell’s DNA in a culture dish and generating a chimeric mouse that incorporates that cell’s new genetic information into its sperm and eggs. But EC cells didn’t lend themselves to this task. They contributed to most parts of the mouse, but it proved difficult to produce animals with the EC DNA in egg or sperm, an accomplishment that was essential to breed animals with the genetic change. ES cells seemed more promising, but they presented a different problem. Manipulating genes and then sorting through many cells to find the desired alterations wasn’t possible with these cells because scientists couldn’t grow them in the culture dishes where such maneuvers would take place.

Martin Evans, then at the University of Cambridge and now at Cardiff University in the UK, considered several explanations for the inability to culture ES cells from the early embryo. Such cells could be rare, he reasoned, and they might not survive well outside of the mother. Furthermore, they might exist for only a very short time at an exact phase of embryonic development.

Evans zeroed in on the time period when he thought ES cells existed in the embryo. To increase the number of potential cells, he and his colleague Matt Kaufman delayed implantation of the early embryo into the mother’s uterus. This procedure allowed the embryos to accumulate cells at the right stage.

In the early 1980s, Evans and Kaufman harvested the early embryos, grew them in culture dishes, and picked out cells that closely resembled EC cells. They had hit gold. These cells formed teratocarcinomas when injected into mice. Furthermore they could be maintained for long periods of time in culture and specialized magnificently well in the Petri dish. They possessed all the features that identified them as the sought-after embryonic stem cells.

Evans had isolated the much-pursued multi-purpose cells. In principle, their DNA could be manipulated to create any mutation of interest. But how were scientists going to accomplish this task with the exquisite precision they desired? The only way to manipulate an organism’s DNA at the time was to integrate genes randomly. This ability to produce so-called transgenic mice has proved extraordinarily powerful in many ways, but because the newly introduced gene (transgene) adds to a chromosome at a random position, it leaves the resident copy of the gene intact. As a result, the transgene technique allows scientists to add genes; it does not permit them to subtract genes or to alter them at will.

Gene targeting

While Evans was tracking down embryonic stem cells, Mario Capecchi, at the University of Utah in Salt Lake City, and Oliver Smithies, first at the University of Wisconsin in Madison and later at the University of North Carolina in Chapel Hill, were trying to target genes by a process called homologous recombination — “homologous” because the incoming DNA sequence lines up with its twin target sequence in the chromosome and “recombination” because the incoming and target molecules break and rejoin with each other. Depending on the technical details, the process replaces one version of the gene with another or adds another copy of the gene in tandem. In 1982, Capecchi showed that mammalian cells contain efficient enzymatic machinery for mediating homologous recombination between DNA molecules injected into the cell. With the thought of exploiting this machinery to carry out homologous recombination between newly added DNA and the corresponding chromosomal sequence, Capecchi and Smithies had the bold idea that they could alter any gene in the cell by replacing it with a modified version. Many scientists doubted that this notion would pan out. How could a unique incoming gene successfully sift through the vast amount of DNA in mammalian chromosomes and correctly home to the proper site? But Capecchi and Smithies persevered.

Smithies devised a laborious — but extremely sensitive — method to find cells in which gene integration occurred at a chosen location. He aimed to detect such rare cells among the many more in which the gene would occur at random sites. The winning cell would contain a single piece of DNA that carried unique features of both the input gene and the resident gene — characteristics that had resided on two different molecules at the beginning of the experiment, but had become next-door neighbors through homologous recombination.

Smithies sequentially divided a population of 4400 cells into smaller and smaller pools, tracking the ones that contained pieces of DNA with diagnostic features of both the donor and recipient molecules. Eventually, he found a cell that carried the critical combination of DNA sequences on a single molecule, and published the work in 1985. His discovery showed that specific planned modification of native genes was possible. The DNA dart had found its corresponding sequence in the vast tangle of chromosomal DNA; he’d targeted a specific gene.

Capecchi took a different tack. First, his lab generated mammalian cells that contained insertions of a defective drug resistance gene. In a second step, the researchers introduced target DNA that carried a different defective version of the gene into the same cells. Homologous recombination between the two sequences could create an operational gene, and only cells in which the two defective versions recombined would resist the lethal effects of a toxic drug. This approach allowed scientists to easily find cells in which homologous recombination occurred, and helped them work out experimental conditions that made this process occur more efficiently.

These advances were big wins — but the game wasn’t over. How would scientists harness this capability to inactivate genes? And how would they extend the approach from cultured mammalian cells to the whole mouse?

Knocking out genes

Because the efficiency of gene targeting was low, the idea was to find the rare cultured ES cells that contain the new genetic alterations, and then implant those cells into early embryos. The resultant mice would breed to produce offspring that harbored the genetic change in every cell.

By 1987, both Capecchi and Smithies had shown that they could target a native gene in ES cells. They chose the same gene, hprt, for practical reasons. Because of the peculiarities of the gene’s function, the teams could isolate cells that contain defective or normal versions by exposing cells to different chemicals. This aspect of hprt allowed them to find knockout cells and also cells in which a mutation in the gene had been corrected. These experiments established the paradigm of knocking out a gene.

But most genes do not share these special properties of hprt. Capecchi wanted to develop a system that would be of general use in altering genes. Random recombination was 1000 times more frequent than homologous recombination so he needed to figure out a way not only to identify cells that carried the new DNA, but also to eradicate cells that carried it in the wrong site in the chromosome.

In 1988, Capecchi formulated a general strategy to enrich for cells in which the homologous targeting event had occurred. The scheme that he devised, now called the “positive-negative” method, not only enriches for recipient cells that have incorporated the DNA; it eliminates those that have allowed it to integrate at random sites. Most importantly, this technique makes it possible to replace virtually any gene of interest.

Making designer mice

In the meantime, Evans had been pursuing mouse embryonic stem cell technology — developing its use as a way to make designer mice. He and his student Allan Bradley and postdoctoral fellow Elizabeth Robertson had demonstrated that ES cells could generate chimeric animals capable of transmitting their genetic material to offspring. Furthermore, they had modified ES cells in culture dishes to create mice that would breed to produce baby mice with the genetic alterations in every cell.

Diagram of genetically engineered mice

More than 4000 genetically engineered mice owe their existence to the technology developed by Capecchi, Evans, Smithies, and their colleagues—the procedure is pictured above in a simplified manner.

With this combination of Evans’s accomplishments in ES cell technology and Capecchi’s and Smithies’s achievements in directed gene targeting, one could now dream of altering any single gene in ES cells growing in culture, and then creating a mouse with the genetic change in all of its cells. In 1989, four groups realized this dream, each disabling a different gene.

A legacy

More than 4000 genetically engineered mice owe their existence to the technology developed by Capecchi, Evans, Smithies, and their colleagues. In addition to knocking out genes, the same method allows scientists to engineer precise changes in a particular gene. Insight into the processes to which these genes contribute is exploding as researchers assess the effects of the alterations, and pharmaceutical companies are using the resulting mouse models of disease to develop drugs. Refinements of the gene targeting technique are even allowing investigators to create a desired mutation in one particular tissue or at one particular time in the animal’s life. This ability allows scientists to study, for example, the effects of a gene in an adult mouse that is required for embryonic development. They can allow the animal to develop normally, and then inactivate the gene when desired.

Because genes influence nearly all biological phenomena, this technology is impacting the analysis in fields as diverse as cancer, immunology, neurobiology, human genetic disorders, endocrinology, and neurobiology. Capecchi, Evans, and Smithies have revolutionized the study of human health and disease.

by Evelyn Strauss

Key publications of Mario Capecchi

Capecchi, M.R. (1980). High efficiency transformation by direct microinjection of DNA into cultured mammalian cells. Cell. 22, 479–488.

Folger, K.R., Wong, E.A., Wahl, G., and Capecchi, M.R. (1982). Patterns of integration of DNA microinjected into cultured mammalian cells: Evidence for homologous recombination between injected plasmid DNA molecules. Mol. Cell. Biol. 2, 1372–1387.

Thomas, K.R., Folger, K.R., and Capecchi, M.R. (1986). High frequency targeting of genes to specific sites in the mammalian genome. Cell. 44, 419-428.

Thomas, K.R. and Capecchi, M.R. (1987). Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell. 51, 503–512.

Mansour, S.L., Thomas, K.R., and Capecchi, M.R. (1988). Disruption of the proto-oncogene int-2 in mouse embryo-derived stem cells: A general strategy for targeting mutations to nonselectable genes. Nature. 336, 348–352.

Capecchi, M.R. (1994). Targeted gene replacement. Sci. Am. 270, 54–61.

Capecchi, M.R. (1995). A personal view of gene targeting. Accomplishments in Cancer Research 1994, J.G. Fortner and J.E. Rhoads, ed,. Philadelphia: J.B. Lippincott, pp. 67–68.

Key publications of Oliver Smithies

Smithies, O. (1955). Zone electrophoresis in starch gels: group variations in the serum proteins of normal human adults. Biochem. J. 61, 629–641.

Smithies, O., Connell, G.E. and Dixon, G.H. (1962). Chromosomal rearrangements and the evolution of haptoglobin genes. Nature. 196, 232–236.

Slightom, J.L., Blechl, A.E., and Smithies, O. (1980). Human Fetal Gy- and Ay-globin genes: complete nucleotide sequences suggest that DNA can be exchanged between these duplicated genes. Cell. 21, 627–638.

Smithies, O., Gregg, R.G., Boggs, S.S., Koralewski, M.A., and Kucherlapati, R.S. (1985). Insertion of DNA sequences into the human chromosomal B-globin locus by homologous recombination. Nature. 317, 230–234.

Doetschman, T., Gregg, R.G., Maeda, N., Hooper, M.L., Melton, D.W., Thompson, S., and Smithies, O. (1987). Targeted correction of a mutant hprt gene in mouse embryonic stem cells. Nature. 330, 576–578.

Smithies, O., Kim, H.-S., Takahashi, N., and Edgell, M.H. On the importance of quantitative genetic variations in the etiology of hypertension. Kidney International (in press, 2000).

Award presentation by Ira Herskowitz

The release of the human genome sequence in draft form makes this a landmark year in the history of biology. Now we know that we have 30,000 or so genes (or is it 50,000?). We are now faced with several important questions, which include: First, what are the functions of these genes and the proteins that they code for? And, second, how can we use this information to improve human health?

Until the ability to knock out genes in the mouse was developed, determining the function of human genes seemed largely out of reach, tantalizingly so. For example, we might know of a human protein that is found only in certain cells of the brain and suspect what it might do, but how can we find out? Or, we might know of a gene in the fruit fly that is necessary for its development and see that humans have a very similar gene. Does it perform a similar function in humans? A powerful way to link a gene to function is to study the behavior of a mutant that lacks that gene and then see what the mutant can and cannot do. It’s somewhat like disabling an automobile by removing one part and then inferring the function of the part that was removed. But we can’t knock out genes in a human, so how can such mutants be produced?

The mighty mouse has come to the rescue. Its genes are typically 95 percent identical in sequence to ours, and we share the vast majority of our genes with the mouse.

Despite the obvious differences between human and mouse in morphology and in some physiological processes, these differences are greatly outweighed by our similarities: they have kidneys and brains like ours; they have an immune system and develop a lot like humans; and they get diseases such as cancer and others that affect their cardiovascular and nervous systems like us. In some respects, mice are pocket-sized humans. The bottom line is that the mouse provides the opportunity, dreamed about for decades, to make the link between a mammalian gene and its function. How is this done?

Building on more than one hundred years of genetic and embryological studies of the mouse, Mario Capecchi, Martin Evans, and Oliver Smithies have created a magic wand by which it is possible to modify any mouse gene with exquisite precision — to completely delete it or to produce a specifically altered form of the gene.

The same technology also makes it possible to go the other direction — instead of knocking out a mouse gene, it’s possible to restore function to a gene that is defective.

Let’s now look at the process by which a mouse knockout is constructed.

A key piece of starting material is a mouse gene that’s already been cloned: it might be a mouse gene corresponding to a human gene or a mouse gene corresponding to a fruit fly or nematode gene. The goal is to construct a mouse that lacks this gene. The second key piece of starting material is a special mouse cell line where the gene is going to be knocked out.

There are three steps for constructing a mouse knockout. In the first, a cloned gene is manipulated in a test tube to delete all or part of the gene. This is routine molecular biology. In step two, the mutated DNA is introduced into special mouse cells, where the mutated DNA replaces a normal gene copy in the chromosome. The crucial aspect of this process is that the mutant gene has to find the related sequences in the chromosome, so-called homologous DNA sequences, and then undergo recombination to switch places with the good gene. The ability of the introduced DNA to find the homologous DNA sequences is called ‘gene targeting’. There was no evidence for gene targeting in animal cells growing in culture and great doubt about whether this could be done. This is where Capecchi and Smithies made their most important contributions. In the third step, the cells with the targeted, inactivated gene are grown into a mouse that has this inactivated gene. It was Martin Evans who isolated the cell lines that made this possible and showed that genetic changes introduced into these cells in culture could be transmitted through the germ line and into mutant, progeny mice.

Let’s now look at our awardees.

Verona, Italy, has given us not only Romeo and Juliet, but Mario Capecchi. His early days as a child included living in orphanages and on the street in war-torn Italy from four to nine years of age, then growing up in a nurturing Quaker environment in Pennsylvania. I refer interested people to articles that are available on the Internet. Capecchi did his graduate work at Harvard with Jim Watson and was enormously productive, making textbook discoveries on molecular mechanisms underlying protein synthesis. This was a golden age of molecular biology. Mario learned his lessons well, and when he established his own laboratory at the University of Utah in 1973, he sought to bring molecular genetics to animal cells growing in culture and to learn how to manipulate the genes of these cells. This led him to undertake a series of studies beginning in 1977 that demonstrated gene targeting in animal cells and culminated in the construction of one of the first knockout mice in 1989. His first indications of homologous recombination in animal cells were published in 1982 and fueled a series of logical and remarkable studies that provide the standard methods for knocking out mouse genes.

Oliver Smithies was trained as a biochemist, but throughout his scientific career, homologous recombination kept on cropping up, and he came to think about how it could be used to fix defective genes. Smithies was born in Halifax, England, and raised in the United Kingdom. After studying at Oxford University, he came to the University of Wisconsin for postdoctoral studies in 1951 and was on the faculty there for 28 years, from 1960 to 1988. He is presently at the University of North Carolina at Chapel Hill, and may well have flown here in his own little plane to attend this luncheon. After important early contributions springing from his development of a method for fractionating proteins, he became intrigued by the structure and evolution of mammalian genes, which meant that he became involved in cloning these genes.

In the early 1980s, Smithies began to wonder whether homologous recombination — gene targeting — could be carried out experimentally to correct a defective gene, for example, a mutant globin gene. For this type of genetic correction to occur, exogenously introduced DNA would have to target to the homologous chromosomal DNA sequence and recombine with it. But was this possible? No one had demonstrated gene targeting in animal cells.
In 1985 Smithies and colleagues demonstrated that they could introduce a DNA segment containing part of the globin gene into cells and then find cells in which this DNA segment had targeted to the chromsomal globin gene. This was a tour de force of sophisticated molecular genetics. His strategy was completely different from that used by Capecchi and though laborious, the demonstration of targeting was unequivocal.

These studies from the Capecchi and Smithies laboratories provided one of the essential ingredients for constructing gene knockouts in mice, the ability to target genes in cultured animal cells. The crucial next step was to take mouse cell lines modified in this manner and produce mice from them.

The history of mammalian embryology is intellectually rich and filled with great practical applications. It was nurtured by the agricultural industry, among others, and involved important work with rabbits and mice. The United Kingdom can lay claim to many important contributions in this area, and thus Martin Evans is part of a distinguished tradition. Evans was born in the United Kingdom and graduated from Cambridge in 1963. He then went to University College London, where he studied vertebrate development using frogs. After working with a certain type of cancer cell line that could differentiate in cell culture and be used to generate whole mice, Evans set out to isolate normal cells from an early mouse embryo that would have similar properties. Work from Richard Gardner argued for the existence of such cells, but culturing them had been elusive. In 1981, Martin Evans and Matt Kaufman and, independently, Gail Martin, in the United States were successful in isolating such cells, which have become known as embryonic stem cells, ‘ES cells’. Evans then carried out an important series of experiments with his students Allan Bradley and Elizabeth Robertson that demonstrated that these ES cells could contribute to the mouse germ line. They further showed that genetically manipulated ES cells could transfer their genetic changes to progeny mice. The importance of ES cells was immediately recognized by Capecchi and Smithies, who learned how to grow ES cells and demonstrated that they could carry out targeted genetic alterations with them.

The first knockout mice constructed by gene targeting were published in 1989, and the rest is history. More than 4000 different knockout mice have been constructed in the last dozen years, and many more are in the works! To keep on top of this fast-moving field, I suggest looking at the Jackson Laboratory’s website, where you can find columns called “It’s a Knockout!” and “KO of the Month.”

The ability to modify the genetic makeup of a mouse by design provides a wealth of information on the function of the gene that is knocked out. Every aspect of mammalian physiology is being penetratingly analyzed by this technique. Particularly notable are the discoveries made on how the immune system functions, which have enormous implications for human health. Knockout mice made it possible to demonstrate unequivocally the molecular basis for prion diseases such as mad-cow disease. Knockout technology is also used to create mice that have versions of human diseases such as cystic fibrosis, muscular dystrophy, atherosclerosis, and many others. These mice make it possible to follow the course of a disease and provide an opportunity to identify and test drugs to ameliorate or cure these diseases.

The ability to precisely tailor mouse genes has completely revolutionized the practice of biomedical science for the last decade and is likely to become even more important in the decades to come. We are certain to reap an enormous bounty of information from knockout mice and reap great benefits for the improvement of human health.

Acceptance remarks

An interview with Oliver Smithies by Raju Kucherlapati

Raju Kucherlapati, Scientific Director of the Harvard-Partners Center for Genetics and Genomics, interviews Oliver Smithies, whose work on homologous recombination led him to provide the groundwork to manipulate the mouse genome with exquisite precision.

Date of interview: September 20, 2001

Kucherlapati: I’m Raju Kucherlapati from the Harvard Medical School, and here with me today is Dr. Oliver Smithies. He’s the 2001 Albert Lasker Award winner for Basic Medical Research. Oliver, welcome.

Smithies: Thank you, Raju.

Kucherlapati: Oliver, first I want to congratulate you on this great accomplishment of being the Lasker Award winner this year.

Smithies: Thank you.

Kucherlapati: Oliver, it would be wonderful if you could tell everybody a little bit about your educational background.

Smithies: Yes, I would enjoy doing that because I have enjoyed my education. My primary education was of course in a little school in a country village near a big town. But then from that school I got to go to the local grammar school, as it was called in those days in England. A very distinguished school that was first opened in 1596, so it was a longstanding school and with a great scholastic tradition. And in that school when you got to the end of your time in it, you had the opportunity to be in the sixth form and to apply for scholarships.

So I applied for a scholarship under the guidance of my high school or grammar school teachers who were really marvelous people, and I got a scholarship to go to Oxford. Of all the colleges in Oxford — there are many colleges in Oxford, and you can apply for different colleges — I applied to go to Balliol, which is a college which was a very scholastic college. Now this was during World War II, so it was quite a small college then and you were taught by two levels. You were taught by the university professors and lecturers, and you were also taught by your tutor. He would be a very personal individual who you had personal contact with once a week.

They didn’t have very many students at that time, so the total college was only 90-odd students and my tutor probably had ten students. You would meet with him once a week and write an essay on a topic of his choice, not yours. This was a rather remarkable type of education because we were expected to read textbooks for one day of the week and then reviews for the second day, and then after that it had to be all original literature. So you got to enjoy learning from the original literature very early in your career and understanding the basic parts of science.

That was the undergraduate education. I took a degree in physiology, animal physiology, and I was actually a medical student. In actual fact, I dropped out of med-school and went to do chemistry, and then took a couple of years of chemistry and started to work with the same tutor now as my PhD advisor. And I did a PhD with him for two or three years, working on developing what we would now call a tool of biophysics. But that hadn’t been invented, there’s a name then and it was called “physical biochemistry” when I was doing it. So that was my primary education.

I should add that I lived in his house part of the time. We became very great friends; I felt part of the family. Because in World War II there weren’t enough places to live, so I stayed in their house and messed up their sink with my chemicals and got all of that by his wife, who was a marvelous person.

Anyway, then I went to the University of Wisconsin as a post-doc in physical chemistry and did a rather undistinguished couple of years of post-doc. Then, having been at a very susceptible age, I fell in love with an American woman — or girl I would have called her at that time — and we decided that it would be a good idea to get married. She didn’t want to go to England, so I went to Canada because of a visa problem.

I went to Canada and stayed in Canada for about seven years, and it was there where I did some work we might talk of later in relation to starch gel electrophoresis, and became a geneticist. I came back to University of Wisconsin at Madison again, but now as a geneticist on the faculty of genetics, and then became a professional geneticist. So that’s a rather long-winded account of what happened.

Kucherlapati: Where do you work now?

Smithies: I work in the University of North Carolina in the Department of Pathology and Laboratory Medicine. I always get these mixed up, and so I’m really a pathologist now, officially, although I’m in actual fact — have been for a long time — and I’m now a molecular geneticist… is the best description of what I am.

Kucherlapati: Oliver, were there any early incidences in your life that sort of attracted you to science?

Smithies: Oh, I think probably it was a comic strip. There was a comic strip that had an inventor in it, and I thought that was neat stuff and I’d be an inventor. And that’s what I’ve been.

Kucherlapati: When was this?

Smithies: Probably seven or eight, something like that. As long ago as that, ever since I can remember, I wanted to be an inventor.

Kucherlapati: So you were always interested in science.

Smithies: Yeah, I think that’s absolutely true.

Kucherlapati: Are you still interested in science?

Smithies: Oh yes, I can’t stop it. I have a little story about science that I think tells the story. A scientist loves the work but shouldn’t be so engrossed in it that there isn’t anything else. So a scientist should have three things in life. Should have his or her science, should have a good hobby and should have a family life. So it comes, now we’re talking about Saturday, it’s the weekend, you wake up, what do you want to do? These three competing things: you want to go and do an experiment, you want to go flying, and you want to take your wife out to lunch. And if you’re lucky you can do all three; I did all three last Sunday, or maybe the Sunday before that I think.

Kucherlapati: Oliver, one of the early things that you did in science is to develop a method for separating proteins by starch gel electrophoresis. Can you tell me a little bit about how the ideas came about developing this method?

Smithies: Yes. I mentioned that I went to Canada for my first job now after my post-doc. And the person I went to work with was named David Scott. He’d been one of the earlier investigators of insulin and the purification of insulin, which was discovered in Toronto, which is where he was and where I went to work. And he said I could work on anything that I liked when I went to work with him, as long as it had something to do with insulin. I decided I would look for a precursor of insulin… in trying to develop a method of detecting what I thought would be very like insulin but would not be insulin because I didn’t think insulin was the natural product.

I was right, but never discovered the product. The real thing… I had a problem with the insulin… it wouldn’t migrate in my electrophoresis medium, the medium I used to separate proteins, which was filter paper soaked with a buffer. You put the protein on this filter paper and pass it in an electric current. It should migrate down the filter paper and separate from other things. But insulin stuck to the paper, terribly stuck to the paper, and it would just unroll like a carpet. So if you had a small amount it would unroll to here, if you had more it would unroll to here. And it was hopeless.

Then I heard of some guys who in the Hospital for Sick Children were using starch grains for a supporting medium for electrophoresis. And to give you an idea of what that’s like, it’s like taking a sand pie and filling it full of sea water so it’s sort of semi-solid, but there really is liquid in it and your protein can migrate all around the different grains and separate. Well they were doing this with starch grains and they had no problems with things sticking to the starch grains. But in order to find the proteins, they had to cut this sand pie, if you like, or this starch block into 50 chunks and measure the protein in each chunk.

That meant 50 protein determinations to do just one electrophoresis. Well, I didn’t even have a dishwasher, I didn’t have a technician, I couldn’t afford anything like that if I was going to do science, but I thought this was neat that it didn’t stick to starch.

Then I remembered a key thing, and this is part of my history that all of this work stems from in a way: I remembered helping my mother to do the laundry when I was about 12, and when she starched my father’s clothes she made the starch by boiling water and whatever. And then after you tied it up at the end it was a jelly, and I thought well, if I just cook the starch into a jelly, then I won’t have to slice it like this and I can stain it. And so I can cut out all this 50 protein determinations… and so it was a lazy man’s approach, and it invented this new method of electrophoresis that proved to be quite powerful.

Kucherlapati: Oliver, you said that you initially trained as a physiologist, and now you’re a geneticist. How did you transition from being a physiologist to a geneticist?

Smithies: Well, I always had this problem with deciding what I wanted to be. So when I went to college I couldn’t decide whether I wanted to be a physicist. I won my scholarship basically on the basis of my physics and math, or whether I wanted to go to med school. And I dropped out of med school and went into physiology, and then did chemistry and combined the two. In a way I think I’m very fortunate because I was never frightened of biology and I was never frightened of chemistry. Often enough people are good at one or the other and frightened of the other one.

So it gave me this feeling of being able to do both, which I’ve always treasured. Think perhaps when I mention my tutor again, Sandy Ogston was his name, I didn’t say his name, but Sandy Ogston always had a very broad view of science and taught that way, and got you to think that way. So whether it was physiology or biochemistry or molecular or DNA or protein didn’t make any difference — you just think the same sort of thoughts, transfer the medium.

Kucherlapati: So Oliver, you continue to do experiments now?

Smithies: Oh yes, every day pretty well.

Kucherlapati: You enjoy doing that?

Smithies: Oh very much. It’s a marvelous thing about science. I think that people, maybe people who aren’t in science don’t understand that science isn’t really achieving some great aim — it’s a lot of little steps. And you have to enjoy the little steps every day so you have to enjoy the experiments you do. It’s not so critical that they work, only that you felt that you did a good job and you’ve got… I know it didn’t work but now I think I know why it didn’t work. So you change the experiment and you do the experiment and if you’re lucky, you get lots of experiments and have the enjoyment. I like to do experiments that I get an enjoyment each day. So most of my experiments are relatively short. They might have a three-year or five year total span, but each little experiment is short.

Kucherlapati: Oliver, you’re really very well known as a toolmaker. Can you explain why people consider you as a toolmaker?

Smithies: Well, I suppose I really think I am a toolmaker; the starch gel electrophoresis, which was the one we were talking about with cooking the starch, turned out to be very powerful. It was the first of the high-resolution gel electrophoresis systems for separated proteins. And molecular biologists, I don’t think any of them would function as well as they can today without polyacrylamide gel electrophoresis, which was the immediate following of that tool. It enables you to solve the problem you have, but it gives other people a tool also.

So you get enjoyment from not only your own use of the tool but other people’s use of it. Probably the most important tool that I helped to develop was the means of altering genes and cells, which is essentially what the Lasker Award is about.

Kucherlapati: Oliver, you’re being honored for the development of a powerful technology for manipulating the mouse genome with exquisite precision. So can you tell us about what this technology is?

Smithies: The easiest way to describe it is to think of DNA as being a set of instructions. The DNA in each cell in your body has a whole set of instructions, and these are written out in an alphabet. It happens that the alphabet of the gene is only four letters and we normally use an alphabet in English anyway of 26 letters. But you can convey this information in a lot of letters. So it’s a very linear set of messages.

In fact, it’s rather a lot of messages. If you think of it in terms of a book, it would take a thousand books, each with a thousand pages to represent every letter that’s present in the message of the human genome. And my thought was that we could match DNA coming from outside a cell with the right place in this book, if you like, or in these thousand books, on the basis of one piece of DNA being able to recognize another.

I knew about that from the work that I’d done with the gel electrophoresis. Quite a bit of that work had been related to a protein called haptoglobin, and there were genetic differences in some people, not the same in others. That turned out to be related in part to DNA finding a partner and what we call recombining, exchanging partners, if you like. So here’s how I think of it in trying to tell people the sort of thing that we’re doing: Supposing we have a book that has in it, somewhere in the book, “I think we should go out to lunch in New York today.” That probably only occurs once in the book, maybe a long book.

If you wanted to find that sentence and change its meaning, you take a piece of DNA and you synthesize it or make it in a certain way that it says, “I think we should not go out to lunch today in New York today.” Now you put a new DNA into the mixture and ask it to find the partner. It goes along and it looks around everywhere, and there’s only one partner it can find. It finds, “I think we should go out to lunch in New York today.” But it’s got a new twist to it because it’s put in a little bit extra, it’s put in “not.” So if those two exchange places now, then this book says I think we should not go out to lunch today in New York. It’s completely changed its meaning only by introducing one piece of DNA that you find the target and then exchanging partners, as it were. That’s the principle of what we have been doing.

Kucherlapati: Why did you think of doing these types of experiments, Oliver?

Smithies: Well, you know, it was a slow process really, in a way, because my early work with the starch gel electrophoresis had been working with proteins. And then later as we learned how to do these things and were taught by other scientists, we learned to start thinking about DNA instead of proteins. So I began to study and my people in my lab began to study the differences between different genes now at the DNA level.

We found out when we examined genes that quite often we could find evidence that one gene had given the path of its sequence to another related gene, so it was as if there were two genes and one said, “I’m not going through, the New York business,” and the other one said the slightly different message, and yet they exchanged information. This happens naturally and we found evidence of this. So it seemed to me that it was a constant, continuing sort of thought that we ought to be able to use this process to help cure genetic… help cure genes, if you like. To take what we would now call gene therapy.

So I thought that if I can introduce DNA into a cell, I might eventually be able to alter a gene that’s in the cell. My initial thought was to try to do this for sickle cell anemia so that the gene I was interested in, therefore, was the same gene that affects… that is different in individuals who have sickle cell anemia, the betaglobin gene, the gene that codes for a part of hemoglobin. I’d always been thinking it would be good to be able to do this. We had plenty of supply of the good gene, as it were, and we knew there were individuals who had a gene that was damaged, and surely we ought to be able to put these together using this message searching that I was talking about, this information searching. But I couldn’t think how to do it, and what would make it work.

Then I was teaching a class in genetics in Wisconsin, and I was teaching about a paper that had been published — I think it was in something like April of 1982 — and in this paper, the first author was Goldfarb and the last author was Michael Wigler. They described a method that they had used to isolate one of the first genes related to making a cell into a cancer type of cell. It was very neat and very complicated, but I suddenly realized I could use this same principle to see if I could make one piece of DNA find a partner. The principle that that came out of that, and I wrote a page in my notebook about it, was fairly simple in idea.

But let’s go back to the “not” example. If I have “not” in the incoming DNA and there is no “not” in the target, then if I can get that “not” into the target and show that “not” is there, I know I’ve hit the gene. I used a piece of DNA, which says — if I can remember my metaphor well enough — I used a piece of DNA that said that “I think we should not go out to” and didn’t go on. But that’s enough to find “I think we should go out to” to change it to “not.” And now if I can find a “not” next to “lunch today in New York,” I know that I hit the target. I have to find that piece of DNA that now has “not” next to “New York.” Because the “not” was on the incoming DNA and the “New York” was on the target, and they were separate.

But if I’ve hit the target, then “not” would now be near “New York,” and that was the principle of the method that I invented. It took three years and it took your help, Raju — you know very well — I’m going to tell whoever is listening that Raju and I have been collaborating in science for I think 30 years. Is it about 30 years, Raju? I think it might be more. Anyway, we’ve been collaborating a long time, and so when I wanted to do this experiment, the first person I turned to for help was Raju. So together we did this experiment.

Kucherlapati: Oliver, do you remember the day when you were able to have convincing evidence that this event took place?

Smithies: Yes, I do very well; it was it was a spectacular day. The page I told you about that I wrote in my notebook after reading the Goldfarb and Wigler paper was some time in ’82, and then we took about three years of work to get to the proof of the correctness of the idea. And the final proof was to have some cells that we thought by indirect means — we could assume by indirect means that we had achieved this targeting. And finally, we got down to these cells and purified the DNA from these… my post-doc Ron Gregg purified the DNA from these cells. He ran a gel to separate the DNA and we knew that if we had achieved our aim we would find a piece of DNA that, if I remember correctly, would be seven kilobases long, seven units long. Whereas if we hadn’t been successful, it would be 11 long.

So this gel had to be run, and then I got the privilege of developing the film that told us what had happened in that experiment. I remember very well in the dark room thinking “now we’ve been working for three years on this with indirect evidence and now we’re going to know the truth.” This is like being a pilot flying in the clouds. I do that, as an instrument-rated pilot, and you fly a certain set of indirect indications, needles and things which let you know where you are, and you fly with the needles in such a way that you keep the needles exactly crossed like this. If you do that, you go down this line and then you come out of the clouds at the end of this line and there’s a moment of truth. Either the runway is there or it’s not there, and so this film, in a way when it was developed, it was like finding your runway.

And there it was, it was right and I went and told people at the scientific meeting and everybody at that meeting whenever they had their final result they would point to it and say, “And there is my runway.” So I remember very well.

Kucherlapati: Oliver, you’re being honored for applying this technology for manipulating the mouse genome. So how did that come about?

Smithies: Well, in a way it came about because we were not successful. We were not successful in achieving these modifications of genes with a high frequency; it was really very rare. So instead of what will be needed for gene therapy, let’s say a ten percent success rate at least or something like that, or maybe one percent, we were only succeeding one in a million times. And so that wasn’t really very useful for gene therapy.

But then I went to a meeting — a Gordon conference — I heard an investigator talk about these new cells that had been isolated by Martin Evans. These cells were called embryonic stem cells. Now many people now know what embryonic stem cells are, but at that time it was still very early days, back in the ’80s. I heard about these cells that Martin Evans had isolated, embryonic stem cells.

I thought well, my gosh, even if it’s very rare, we can do one of our experiments with embryonic stem cells and find the one in a million that has been altered and use that to make a mouse that has now a gene altered in a way that we wanted. And it doesn’t matter that it’s rather rare and rather difficult, because you can in a sense sift through all of the cells and find the one that’s been correctly altered and then return it back to a mouse, as Martin Evans taught us how to do, and get a mouse with the altered gene. So that’s how we came to be doing that, because it was an obviously useful thing to do with this technology which didn’t quite solve the problem I set out to do. I’m still working that, but maybe one day.

Kucherlapati: Oliver, the paper describing the first gene targeting was published in 1985. What has been the impact of this discovery over the last 16 years?

Smithies: Well, I suppose the end fact of it has been rather remarkable to me. But I don’t want people to think that I consider this just my invention as it were, my tool, because Martin Evans was there and Mario Capecchi, who’s being recognized, he made huge advances in making it much easier to do this than my methods have been. But together, as it were, we made a contribution to this. It’s enabled people to do many things that they had not been able to do before. It enables them to alter genes in such a way as to replicate human conditions, human genetic conditions that are harmful or for the individuals maybe even much more than harmful. So we could replicate some human conditions by making these changes in genes.

For example, one of the early ones, which we did, was to make a model of cystic fibrosis. This is a result of one gene being defective. So we made a mouse in which we made this gene defective. Bev Koller, my post-doc at the time, was a valiant worker in this, worked very hard to make that mouse… It’s led people to have a method of understanding what genes do when they don’t know anything about them. Here we’ve isolated a piece of DNA, and we’ve got this thing that looks like a gene there, but we don’t know what it does.

So one of the best experiments to do under those circumstances is to find out what happens when it’s missing, when it’s not working. And this gives one the capability of doing that. So many people use it to find out what genes do, because if you take it away what goes wrong? Well, if you take off the steering wheel of a car, what will happen to it? And you find out pretty soon that the steering wheel is essential to steer the car. The same way if you take out some essential gene, you find out that the animal won’t do or can’t do whatever was appropriate for that particular gene. Many, many several thousand of papers have been published using this technique since.

Kucherlapati: Oliver, all the techniques that are used to manipulate the mouse stem cells — can they also be applied to manipulate human stem cells?

Smithies: In principle the answer is yes, but in practice the answer would be “it would be foolish.” Because as I mentioned, we have to sift through millions to find the right one. Usually we get what we expect, but we don’t always get what we expect. And that means that it’s not really a fully controlled process, and it would be very foolish in my mind to try to use that in embryonic stem cells in humans at this stage. At least embryonic stem cells, unless they were being used for what we might call local effects, something that you could be sure if you weren’t right it wouldn’t cause harm. And I think maybe I’m modifying my statement, you see — if it’s used appropriately it could be valuable, but if it’s used inappropriately, it will be foolish.

Kucherlapati: Without thinking about modifying the human genome line, would this technology have the ability to affect the health of humans by modifying the stem cells?

Smithies: Yes. I think the answer is it might have that affect in the end, if one learns how to get stem cells — not necessarily embryonic stem cells, but stem cells from an individual who has something wrong. Let’s say they have something wrong with muscle, muscular dystrophy perhaps. I don’t want to take that too literally, but let’s say that might be an example. Then if one could isolate the stem cell which would replenish muscle cells, or replace, or help prepare, then repairing that stem cell and returning it to that individual would be rather marvelous because one would have avoided many of the difficulties of this type of thing. But it’s rather a long way away, I think — it doesn’t mean we shouldn’t work hard to get there, but don’t expect it tomorrow.

Kucherlapati: Oliver, you were saying earlier that, those particular technologies, if they’re used appropriately, could have great significance or benefit to human health. So what, what do you think the ethical implications of this gene-targeting work in human stem cells?

Smithies: I think so much of this is tied up with a person’s religious beliefs as to what’s appropriate and what is not appropriate, that I can’t give an answer that I think is other than personal. And maybe somebody else might not think the same way. I think that if we can help people, then it’s appropriate. If we do harm to people, then it’s not appropriate. I don’t think there are any moral implications in doing it; it’s whether we help or we hurt. If we help it’s appropriate; if we hurt, it’s not appropriate.

Kucherlapati: Oliver, thank you for taking the time to talk to me about your experiences and how you’ve gotten into this research, and how excited you are about the work that you do. I think that you would be excellent model for young scientists to follow in the footsteps of people like yourself. You’ve always been excited about science and have the ability to transmit that excitement to everybody.

Smithies: Thank you, Raju.

Key publications of Oliver Smithies

Smithies, O. (1955). Zone electrophoresis in starch gels: group variations in the serum proteins of normal human adults. Biochem. J. 61, 629–641.

Smithies, O., Connell, G.E., and Dixon, G.H. (1962). Chromosomal rearrangements and the evolution of haptoglobin genes. Nature. 196, 232–236.

Slightom, J.L., Blechl, A.E., and Smithies, O. (1980). Human Fetal Gy- and Ay-globin genes: complete nucleotide sequences suggest that DNA can be exchanged between these duplicated genes. Cell. 21, 627–638.

Smithies, O., Gregg, R.G., Boggs, S.S., Koralewski, M.A., and Kucherlapati, R.S. (1985). Insertion of DNA sequences into the human chromosomal B-globin locus by homologous recombination. Nature. 317, 230–234.

Doetschman, T., Gregg, R.G., Maeda, N., Hooper, M.L., Melton, D.W., Thompson, S., and Smithies, O. (1987). Targeted correction of a mutant hprt gene in mouse embryonic stem cells. Nature. 330, 576–578.

Smithies, O., Kim, H.-S., Takahashi, N., and Edgell, M.H. On the importance of quantitative genetic variations in the etiology of hypertension. Kidney Int. (in press, 2000).

Interview with Mario Capecchi, Martin Evans, and Oliver Smithies