Engineering immunity

Interview with David Baltimore, co-recipient of the Nobel Prize in Physiology or Medicine 1975.

David Baltimore at

DAVID BALTIMORE, the path-breaking American biologist and co-recipient of the 1975 Nobel Prize in Physiology or Medicine, was on a lecture tour in India during January 14-18. He came at the invitation of Cell Press, the Massachusetts-based publishers of the well-known biological research journal Cell, and TnQ Books and Journals, a Chennai-based company that supplies pre-press, design and software solutions to scientific, technical and medical publishers, Cell Press in particular. This was the first of a proposed Cell Press-TnQ annual lecture series as part of which Baltimore gave lectures in New Delhi, Bangalore and Chennai.

Baltimore won the Nobel Prize along with Howard Martin Temin and Renato Dulbecco for discoveries concerning the interaction between tumour viruses and the genetic material of the cell, which unravelled how cancer-causing ribonucleic acid (RNA) viruses managed to infect and permanently alter a healthy cell. The centrepiece of this work was Baltimores identification in 1970 of the enzyme reverse transcriptase in virus particles. This provided the evidence for the conversion of (single-stranded) RNA into (single-stranded) deoxyribonucleic acid (DNA), the reverse of what is known as the central dogma of biology, namely, that normal transcription of genetic information involves the synthesis of RNA from DNA, which in turn is translated into proteins. This work has provided the key to our understanding of the life cycles of a whole class of RNA viruses called retroviruses, which includes the human immunodeficiency virus, or HIV.

In the years following the Nobel Prize, Baltimore has extensively researched molecular mechanisms in the immune system and how they are controlled and has contributed a great deal to the understanding of cancer and acquired immune deficiency syndrome, or AIDS. His present research is focussed on the control of inflammatory and immune responses to infection. In recent years, he has launched a major programme called Engineering Immunity. The concept, which marks a paradigm shift from current approaches to treatment and vaccines, could pave the way for the medicine of the future.

Seventy-year-old Baltimore was a faculty member of the Massachusetts Institute of Technology (MIT) for nearly 30 years. He is currently President Emeritus and Robert A. Millikan Professor of Biology at the California Institute of Technology (Caltech). He served as Caltechs President between 1997 and 2006. He is the outgoing President of the American Association for the Advancement of Science (AAAS).

Excerpts from a wide-ranging interview Baltimore gave Frontline, first in New Delhi and then in Bangalore:

Prof. Baltimore, let me take you back to your initial days. You majored in chemistry?

I did.What made you switch to biology?

I didnt. I started off in biology. And I decided to do my degree in chemistry late in my academic time because the chemists would let me do research and the biologists didnt have that opportunity. Anyway, I had done most of the chemistry degree because actually chemistry is the best preparation for biology. But I started off with biology in high school and never deviated.

Why was it that the biologists could not do research?

I was in Swarthmore College, which was more or less an arts school. The chemists were able to arrange for me to do research at the University of Pennsylvania where I could meet people and choose problems to work on.

What was the subject of your research in chemistry?

It was on the Energy of hydrolysis of pyrophosphates, two phosphate groups bonded together. (Laughs mildly.)

That forms the basis of a lot of biological processes as well. Is that why you chose the topic?

Thats right. But the major driving force of biological energy is ATP, or adenosine triphosphate, which is adenosine with three phosphates on it. [Baltimore illustrates it on a sheet of paper as A-P-P-P.] And in certain reactions, like in the synthesis of DNA, you cleave here [shows it by drawing a vertical dividing line as A-P | P-P] until you get adenosine [mono] phosphate and that links to a polymer plus a pyrophosphate. If you want to know the energetics of this reaction, you have got to know how much energy there is in the pyrophosphate. We were trying to measure the equilibrium constant between pyrophosphates and two phosphates [P-P <==> P + P].

In your lecture, you also mentioned that a large part of your early research career was devoted to the polio virus. What aspects of the virus did you research?

Well, I entered polio virus research in 1961. By that time polio was cured as a medical problem because of the [Salk] vaccine. So the interest I had in the polio virus was not medical but rather in its basic biology.

The interest was in how an RNA virus will multiply because it became clear actually a few years earlier that you can get an infectious RNA from an RNA virus. And that meant that RNA was the genetic material, but in the whole rest of the biological world, DNA is the genetic material. So I devoted actually a lot of my early career to try to understand how RNA could act as the genetic material. And that involved the study of the polio virus.

Did that set the stage for you to enter the study of cancer RNA viruses and lead to the discovery of reverse transcriptase?

Actually, yes, it did. Once I discovered the reverse transcription reaction, I actually did a little diagram of the different genetic processes that underlay viral development and viral multiplication. And that became quite famous as a classification of the viruses based on the overall nature of their replication.

What specifically about cancer viruses got you interested in them?

It was just a challenge to understand cancer viruses. We understood a lot about non-cancer-inducing viruses but cancer viruses stood out as presenting a particular problem, and it looked like the solution to that problem might be very important. So I tried to resolve that.

Was it already known that certain RNA viruses were associated with cancer?

Oh, yes. The [knowledge of the] involvement of viruses in cancer goes back to the early part of the 20th century. But actually, at first we didnt know that they were RNA viruses, and then later it was discovered that chemically they were RNA viruses. But that presented a conundrum because RNA is generally a transient molecule in cells. And DNA, of course, has to be absolutely permanent. So, how does this transient molecule provide genetic stability? In particular, how could it transform the growth properties of the cell and turn it into a cancer cell? And Howard Temin had suggested that it could be through a DNA intermediate through the RNA transferring information to a DNA... but that idea received no good experimental proof and most people just ignored it. So I decided to test it.

What provided that proof?

We proved it by showing that the virus particle contained an enzyme that copied RNA into DNA and thats the reverse transcriptase. I say we, but I actually figured out the experiment myself. I am the only author of the paper. And its by implication because the existence of an enzyme doesnt actually prove it. And somebody else later proved it (laughs). showed that if you get an infectious DNA, that proved that the DNA was the intermediate. But by that time, the whole world was convinced. Its an interesting philosophical question why a certain kind of evidence would strike the scientific community as definitive even though logically it isnt definitive. In this case, the existence of the enzyme pretty much had everyone believing it even though Temin and I knew that there was a hole in the argument. We knew we had the enzyme. We knew it was in the virus particle. But we didnt actually know that there was an infectious DNA made from it.

Now there are a whole lot of these RNA viruses that go by the name retroviruses. What distinguishes one family from the other?

The genes that they have.

How does HIV differ from other viruses? Why are we not able to mount any immune response to it? After all, it is just another retrovirus.

HIV, for instance, has a set of genes that you dont find in other retroviruses, or in any other virus for that matter.

Does that alone make HIV unique? What is so special about those genes?

The genes encode proteins, and its those proteins that really have the special properties. We generally infer the sequence of the proteins from looking at the sequences of the DNA and the RNA [of the virus]. And HIV makes proteins that are hidden away from the immune system. And they are hidden by a number of tricks, one of which, encoded in the protein instructions, is that they should have carbohydrates added to them. The carbohydrates then form a sort of coating over the whole [virus] particle and hide it away from the immune system.

The immune system cant see the carbohydrates, actually self-carbohydrates. It physically cant get inside the carbohydrates to look at the proteins. So the proteins are protected from interaction with the antibodies. But then if it is completely surrounded, there would be no way for it to bind to the cell. So it has to have some patch and it does have a patch, but then there are all sorts of things about the detailed structure of that patch that make it inaccessible to the antibody. We try to find ways now, especially in my lab, to actually make smaller kinds of molecules that are not as big as the antibodies to try to get into those crevices that are available on the surface of HIV.

Is it merely the size of the antibodies that inhibits immune response?

Size is a big thing. Not merely but its a big [factor]. Vaccines have to protect against the big binding sites. And the only way, historically, that we know to make a vaccine is to elicit antibodies from the immune system, and since this [HIV] is largely resistant to antibodies, that has been futile.

How about the geometry of the binding site itself?

There are some aspects of the geometry of the binding sites that are important.

Is the geometry not well known by now?

Yes. We dont know it entirely. The problem is that the HIV surface protein is a trimer [called gp 120; see diagram]. So, we look down on the surface and we find it like this [Baltimore explains by drawing the top view on a sheet of paper as a circle with three equal sectors], and all three are identical. But it is very hard to make that trimer. So, what you get if you try to make that protein is that you get a monomer and it looks like that [essentially one sector]. The monomer [on the protein] has sugar all around it so that it cant be attacked, but these faces or crevices [between individual monomers and into the plane of the paper] do not have any sugar in them because they are ordinarily involved in these internal interactions so that they are protected. So, people have figured out the [three-dimensional] structure of the monomer; they havent figured out the structure of the trimer. And the exact structure, the way the faces meet each other, is critically important. So, we dont fully understand the surface of the virus, and there are people working on it, but its been very difficult because making stable trimers that you can then determine the structure of has been a real challenge. Now I do know one laboratory that thinks they have it now....

But without detailed knowledge of the geometry, you are attempting to make small molecules that can enter the surface crevices. How does that work?

Ah! Its a problem. But there are some aspects of it that we know well enough that we can try to model. If you look at the [virus] particle from the side, you see the trimer as a cap. There is in it another piece, which is actually a separate piece of protein [gp 41] ,and this is what sticks out... sort of pierces the virus [membrane] at the other end below to present a knob-like structure that sticks out from the virus.... The region in here [between the bottom of the cap and the virus surface] is pretty open. But the space in here is very narrow.

What we are trying to do is to make small molecules that will go in there. Also, the actual binding site that binds to the cell is up here [on the cap] and that is where we dont have the full knowledge of the structure. So we know some of whats going on in there. You could try to find how to get in there. But we are working down here because this is like a linear protein. So its easy to model.

Many of the drugs that we have for HIV today basically attack the earlier stage of virus multiplication, the reverse transcriptase or protease that is involved. But why is it proving so difficult to design a 100 per cent efficacious drug to inhibit these enzymes?

The problem largely is that the virus can mutate and make enzymes that are insensitive to the inhibitors. We then try to make new inhibitors that will inhibit the mutated form and the virus mutates again. But every time the virus mutates, it is making a less fit virus. So you can find combinations of small molecules that the virus really has difficulty getting around. And thats, in fact, what people use now, the multiple drug therapy.

Now that you have some idea of how to go about the problem, how far would you say you are from developing a vaccine?

We are nowhere in the development of a vaccine. Since to make antibodies work is such a difficult thing, we had hoped that we could use another arm of the immune system, T-cells, to protect against the virus. Much of the vaccine work thats been done over the last five to 10 years has been focussed on making a T-cell-based vaccine. Thats what Merck did and they just had a total failure of the vaccine, and now we are uncertain whether we can make a T-cell-based vaccine. I think there will be more tests of it. There are other people trying to do in somewhat different ways from Merck. But this is a huge setback.

How do T-cell-based vaccines avoid the difficulty encountered in making antibodies work, and why did the concept not work in this case?

T-cells dont see the surface of the virus particle. The way T-cells work is that they look at little peptides. There are cells in the body that can engulf the virus and chew it up into little pieces. And then there are the little pieces on the surface of the cell in a carrier molecule and thats what they present to the T-cell immune system. Those are perfectly naked little pieces of proteins. Carbohydrate or sugar on the virus doesnt matter. Thats one of the reasons we were hopeful that its sort of more carnal to the way antibodies work. We dont know why it didnt work. It may be that there wasnt enough T-cell immunity; it may be that it was the wrong kind of T-cell immunity; it may be that T-cells simply arent protective.

So how will the small molecules that you propose be produced and delivered to the target?

So when I say we are trying to make small protein molecules that go in there [to counter the virus protein], the immune system wont make them for us. So our idea is if we can find these molecules, to actually programme the immune system using gene therapy to make these protective molecules. And we are investigating that as a possibility. Actually, I got a big grant from the [Bill & Melinda] Gates Foundation to do this. And its a matter of desperation because nobody has ever done anything like what I am suggesting. But because HIV has proven to be so difficult, the idea of attacking it in some very different way is, I think, important.

Is this the engineering immunity that you have been talking of? What exactly does it mean?

The bodys immune system is able to handle most infections quite successfully, either by raising antibodies from B-cells or by the action of T-cells. But some important ones are able to evade the immune system and they pose the greatest challenge to medical research: HIV, malaria and tuberculosis. Cancer is another important disease to which the immune system rarely responds. But it ought to be able to do better. I had a notion about five years ago that you might be able to use gene therapy to protect the immune system against attack by HIV and other things and to allow the immune system to do things it cant normally do like making small proteins or is not able to do well enough. I call it engineering immunity, and the programme is related to HIV and its also related to cancer. Its truly a great challenge, but none of it requires brand new science. A lot of this is about engineering. It is a combination of gene therapy, stem cell therapy and immunotherapy, each one of which has been controversial and does not seem to work separately. But together it seems to work pretty well. The project was funded in its initial stages by the Skirball Foundation, and now we have a big grant [of about $14 million] from the Gates Foundation.

But gene therapy itself, though it showed a lot of promise, has not really succeeded, has it?

Yes. A lot of promise but it has not shown any real effectiveness. And when it has been effective, its been dangerous.

So do you think your concept could still be effective?

Yes, I do.What gives you this confidence?

First of all, we are using virus vectors different from the ones that have proven difficult to carry the genes, and our strategy is different. What we want to do is to put the genes into the stem cells of the immune system. We know how to do that. So I am moderately confident that we can do all of these. The real challenge is to be able to target the virus vectors to specific cells in the immune system.

How will this concept be implemented in the case of HIV?

So you look for genes that will make those small molecules and try to find one that will bind [to the viral protein]. The really difficult thing is finding the molecule that will do the trick. You can do that with fragments of antibodies. We are also now working with a new group that I have discovered who can make these kinds of binding molecules based on structures very different from antibodies. So thats nice, something very different.

So would you say we can hope for a vaccine in the next 10 years?

No. I have always said, since 1986 when I first got involved in this, that a vaccine is at least 10 years away. And I still say that it is at least 10 years away.... Its awful. Because here we are 20 years later, and we are still saying its 10 years off. But I dont know if there are many people who would disagree with me.

What kind of cancer research are you currently engaged in?

Well, actually most of the cancer work we are doing now is related to therapy, immunotherapy, melanoma and other cancers, but mainly melanoma. Its also part of the general programme on engineering immunity that I talked about.

For HIV, the idea is to programme the B-cells of the immune system to look for the antigenic proteins on the virus and make those small proteins that can bind to it. For cancer, we engineer the T-cells to look for a peptide to which they can bind. Cancer cells make proteins that are different from normal, in particular differentiation. Though the immune system is supposed to clear the tumour cells, it is unable to do so because of autoimmunity. It is inhibited from doing so by inhibitors.

So what we would like to do is to counter these inhibitors and target differentiation by making the immune systems response to a particular antigen more robust. So the melanoma programme that we have is trying to get genes for T-cell receptors that allow T-cells to particularly focus on melanoma. And this is a joint programme with a group at UCLA [University of California, Los Angeles].

How does the idea work?

The problem here is to get the right T-cell receptors. We can use protein design techniques to make the receptors with the structure that we want. Then make retroviral vectors to express these T-cell receptors and target them to modify the haematopoietic stem cells so that that the T-cells that these stem cells produce are specific to cancer antigens. You have to have the immune system react slowly over a long time.

Do you think that you will be able to engineer the immune system so that its ability to produce the right proteins to fight diseases is there as a permanent mechanism?

Yes, thats right. Its a permanent change in the genetic arrangement of the immune system.

How will the permanent change be brought about?

Its by integration of the virus that carries the gene. Its also thanks to reverse transcription. Its a retrovirus that carry genes into the cell and makes them part of the DNA of the cell.

It would be interesting if you could actually do it in humans.

We can do it. We know we can do it. We do it in mice, and it works all right.

But you need to be able to find the right kind of small proteins.

Yes. I am actually doing that in association with a structural biologist. We are not trying to do that ourselves. I have a little consortium of people working on this.

Are there any safety concerns with this kind of approach?

Safety issues became important following the experiment of retrovirus-based gene therapy in treating a certain immunodeficiency in children, called XSCID [X-linked severe combined immunodeficiency], which arises because the children lack [the cytokine receptor] the common gamma chain. The idea was to make a vector that will express the common gamma chain and insert it into haematopoietic stem cells, and then the kids make T-cells that have the common gamma chain. In most cases they do, maybe in some they dont.

And so there is a group in Paris that has been doing this and reported quite remarkable success in eight out of 10 kids. But then a few of the kids came down with cancer T-cell leukaemia and when they characterised the tumour cells, they found that in the tumour cells the virus that had carried the gene had integrated next to an oncogene, the gene known to be involved in cancer.

Baltimore in Chennai

This occurred only in two cases, but it was too remarkable owing to, first of all, that the gene happened to take hold over an oncogene; also it meant that the tumour cells were clonal, that they all had the same site for expression. So all this is bad news because it is just that whenever you do this kind of experiment, you run that risk. And that may be true. I dont know. But these were with vectors that were made with a mouse leukaemia virus. And thats genetically kind of similar.

We prefer to make vectors with an HIV-based lentivirus [slow acting virus] vector. And those vectors dont integrate and activate genes anywhere near as efficiently as the mouse leukaemia virus. So I think they are much safer. And what we are trying to do doesnt involve the common gamma chain. It turns out that the common gamma chain is an oncogene because it is such a strong stimulus to the growth of the T-cell, which is good for the immunodeficient child but ultimately bad. So, we now think that those two things working together caused those tumours. And if you have only one, you are not running that risk.

But is it not conceivable that there would be other kinds of side effects with other kinds of vectors?

We are not going to introduce the [gene for the] common gamma chain. The genes that we are interested are benign from that point of view, and by switching viruses, we get vectors that dont integrate and activate genes so easily. Those two protections added together really reduce the probability of that kind of danger.

And then, of course, when we are trying to deal with things like cancer, patients are going to run a small risk if there is one. Even then, common gamma chain deficient kids... if I had a kid like that I would want them to be transduced and take some risk because the consequences of not doing it are worse than the consequences of doing it.

Is this concept of being able to engineer the immune system itself emerging as a new paradigm in biological research?

Well. Its emerging in my lab. If you are asking if there are many people doing this kind of thing, there are a few.

Not just for HIV and cancer but being able to engineer the immune system in general to combat a whole lot of diseases

Well, most of the work is being done on cancer. There is of course work [going on] to correct genetic disease, which is totally separate, but that doesnt focus on the immune system necessarily although often it does. [For instance,] ADA [adenosine deaminase] deficiency can be cured by putting a gene for ADA protein into immune cells. That has been done very successfully in many places: Milan, England. Soon, I think, it could become a basic therapy.

So where are we exactly at present in engineering immunity research for these techniques to become an operational tool in health care?

The one thats farthest [along] is the cancer treatment because we are preparing for clinical trials, but you know doing clinical trials and having a treatment are two different things.

And this is for melanoma?Yes. Thats right. Melanoma treatment.

You said in the lecture that you would be treating patients in a months time.

In a few months time. I would love it to happen in a months time. But I think it will be a few months. Youve got to make materials in GMP [good manufacturing practice] conditions; youve got to get approvals from many different sides.

Malaria is another disease that is eluding a vaccine. Are there any engineering immunity approaches that could be valid here?

Thats interesting. I think there probably are. But there isnt anybody I know who is trying to do that. The other two big challenges are malaria and tuberculosis. But so far I dont think anybody is trying to do that.