"The future for chemists who are interested in biological problems is very bright because of our ability to make molecules that can enter inside the milieu of a living cell," says Stuart L. Schreiber of Harvard. "It's a terribly exciting time."

One of the lessons learned in the last decade of cell biology research is that protein-protein interactions constitute the key mechanism for regulating the steps in the signal pathways of cells. Imagine, then, the implications of being able to influence this process to dimerize molecules inside cells, turning on and off signal pathways or regulating gene expression at will. This feat has been demonstrated in the laboratory, and biologists have used words like "spectacular" to describe it. The research team that pulled it off was led by Stuart L. Schreiber, a Howard Hughes Medical Institute investigator and Harvard University chemist, along with his collaborator Gerald Crabtree of the Howard Hughes Medical Institute at Stanford University.

   For the past few years, Schreiber has been making a name for himself as an organic chemist working in cell biology and immunology. Just one indicator of his success was his appearance in the #4 position on Science Watch's list of the hottest scientists of 1993, with seven recent papers from his laboratory each attracting sufficient citations to merit inclusion in ISI's Hot Papers Database (see Science Watch, 4[10]:1-2, December 1993).

   In the early 1980s, Schreiber moved from chemistry into biology with his study of cyclosporine, an immunosuppressant agent now widely credited for the revolution in organ transplantation. In 1986, Schreiber and his collaborators identified the receptor for a newly discovered immunosuppressant known as FK506. Schreiber describes the years that followed as "a process of jumping into the pool and learning how to swim as quickly as possible." His laboratory has elucidated the mechanisms by which these immunosuppressants function; this research led him last year to the discovery, with Crabtree, of how to enlist FK506 to bring proteins together inside cells and regulate cellular pathways in a way that may have profound implications in both basic research and human gene therapy.

   Schreiber earned his bachelor's degree at the University of Virginia in 1977 and his doctorate in organic chemistry in 1981 from Harvard. After serving on the faculty at Yale from 1981 to 1988, he returned to Harvard, where he is now a professor in the department of chemistry. He is also a founder of ARIAD Gene Therapeutics.

From his office at Harvard, Schreiber recently spoke
with Science Watch correspondent Gary A. Taubes.

Even in your research reports you talk about the role that a chemist can play in cell biology. What was the connection for you, and how did you make that move across disciplines?

   Schreiber: For me, the progression was logical. What really fascinated me about organic chemistry was that it involves understanding the conformational and three-dimensional properties of molecules, understanding why they adopt the shape that they do, and applying organic chemistry principles to try to determine what those three-dimensional structures might be. I was attempting to use that knowledge to guide the design of organic reactions when I became aware of a fascinating problem in biology. My first logical step was to realize that proteins are in many ways the essence of biology, and that proteins are composed of amino acids, which are themselves not unlike the organic molecules with which I was very familiar. I rather naively thought that if I studied proteins from an organic chemist's perspective, using conformational analysis and applying it to proteins, perhaps my perspective would be interesting and different.

What was it about cyclosporine, in particular, that intrigued you?

   Schreiber: It was the three-dimensional conformation of cyclosporine. It's a cyclic peptide, which seemed like a logical bridge between simple organic molecules and proteins. Then, in reading about cyclosporine, I realized that there was a major mystery associated with it. It had specific effects on cells, but most of the mechanistic work that had been done on cyclosporine was performed using animals. And it seemed to me that the fundamental actions would have to be at the molecular level inside or on the surface of cells. The cells that cyclosporine seemed to inhibit were T cells.

How much did you know about immunology at the time?

   Schreiber: Let's just say that I was unprepared to study the cellular mechanisms of cyclosporine in 1984. I was completely naive about the fields of immunology and cell biology. I don't think I could even have articulated at that time what cell biology was.

Your recent work on controlling signal transduction emerged from your research on FK506 and cyclosporine. What were the key steps in that progression?

   Schreiber: Cyclosporine led us to study FK506, which is also a useful immunosuppressant agent. The structure of FK506 is completely unrelated to cyclosporine, but it seemed to have similar effects, although FK506 is much more potent. FK506 is also a macrocyclic compound, a large ring system that superficially would appear to be very floppy; but if you apply principles of conformational analysis, you could in fact predict that certain three-dimensional shapes would be preferred over others.
   To make a long story short, we found that FK506 and cyclosporine–remarkably, as it turns out–both have the same mechanism of inhibiting cellular function. What they ultimately do is block a signal transduction pathway from the surface of T cells into the nucleus, and they block it at the same place. But what's amazing is that their initial receptors are different. So here are these two different immunophilins, binding two different immunosuppressant agents, and yet subsequently binding to the same target–a protein called calcineurin. This was the finding that revealed that calcineurin is in fact a key signaling molecule along the T cell receptor signaling pathway.
   Cyclosporine and FK506 are very unusual in that they have two protein-binding surfaces within their structures. They bind to their immunophilin receptor, and that complex binds to calcineurin. So there is a receptor-ligand-receptor complex where the drug is contacting both proteins simultaneously. We were aware of that even before we discovered that calcineurin is the target. We knew there was a target and we found it. And then it just took a little while to realize that those molecules were the organic equivalent of the protein dimerizers in cells.

Just how important is protein dimerization in cellular regulation?

   Schreiber: Virtually every cellular process you can think of involves regulated protein dimerization. Not only signal transduction but transcriptional activation, for example, and protein degradation in cells all occur and are regulated by specific protein-protein interactions, by the ability of a molecule to suddenly become able to bind to another protein. More recently, it's been recognized in many instances that information can be transferred when a protein binds to another protein without even causing it to change its shape. Just the simple act of bringing one protein into close proximity to another protein can have a profound effect on its activity.

Does controlling the proximity effect give you an advantage over other ways of trying to control signal transduction?

   Schreiber: When you achieve a proximity effect–when you cause one protein to associate with another–you don't have to be geometrically precise the way you would if you wanted to make, for example, a mimic of neurotransmitter, which acts by inducing an allosteric change in its receptor.
   Together with Gerald Crabtree's lab, we've been making dimerizers–organic molecules of which one half can bind to one protein, the other half to another; this causes the two proteins to come together. If you pick the right two proteins, it's possible, for example, to intervene in the signal transduction pathway. You can bypass all the early events in the pathway, and pick an intermediate step, and make a molecule that will induce the binding of these specific proteins. Then you can activate or deactivate that signaling pathway in midstream.
   This has applications beyond signal transduction. For example, recent mechanistic investigations have revealed how transcription occurs–how genes are activated and transcribed into messenger RNA by proteins called transcriptional activators, which are nothing but dimerizers. In that case they have a DNA-binding surface and a protein-binding surface. They bind to DNA in the proximity of a specific gene and, with the protein-binding surface, bind to a specific protein required to transcribe that gene. So in that instance, one could consider them a molecular adaptor or dimerizer.
   Once you recognize that transcriptional activation involves dimerization, you use the same principle we used to make our original dimerizers. You make organic molecules that are cell-permeable, but composed of two different binding surfaces; one half binds to one target and one to the other, and they're covalently linked together to cause two proteins to come together.

What molecules are you using as your dimerizers?

   Schreiber: We started with what's called FK1012, which has essentially a piece of FK506 on one half and a piece on the other. But we can now mix and match all kinds of organic molecules. It's an incredibly simple concept and, frankly, a very obvious one, but no one had thought of it. But one does need the wherewithal to make such molecules.

Have any of these experiments been done in vivo?

   Schreiber: No, they've all been done in cell culture. The obvious next step is to see if we can induce dimerization in animals. We want to see if we can cause certain genes to be turned on or turned off at specific times, to cause certain proteins to be degraded at certain times at the sixth day of embryonic development, say, and have them to come back on the seventh day. That's a level of control not currently available. And that's where this research is currently headed. The aim is to do this kind of protein dimerization in the context of a living animal, and to do it with a kind of explicit control, where only specific cells are caused to die at a certain time, or only specific genes are activated at a specific time, or only a targeted protein is caused to degrade at a specific time.

Are there clinical applications for the research or does it seem as if it will be a tool for basic researchers only?

   Schreiber: The hope on the medical side would be to use dimerizers in conjunction with gene therapy. The idea would be that if you understand the etiology of a particular disease, and you can envision activating or inactivating a specific pathway by protein dimerization, then that defines the nature of the dimeric molecule you'd like to make.
   In a general sense, the hope is that by introducing a regulatory system into cells, you could create in vivo therapeutic proteins: ribozymes, anti-sense agents, and the like. The use of low-molecular-weight organic compounds, which are the kinds of molecules familiar to us as the ones that come from the pharmaceutical industry, holds out the possibility of a pharmaceutical approach to gene therapy. It offers the ability to regulate the synthesis of therapeutic entities like proteins. This is admittedly far in the future, because there are plenty of technical hurdles that have to be overcome.
   The other thing our labs are doing is gaining control over the signaling pathways that lead to programmed cell death, to growth inhibition, and to cell proliferation. There are many instances in human disorders where having that ability could be very useful. For example, using gene therapy to allow an organic molecule to kill only tumor cells would obviously be very significant. The new challenge is to get DNA into tumor cells, and the targeting of genes. But the problem is such that, if we can solve it once, then the organic molecule will work continuously. You need to get the gene in, and once it's there it can then respond to the organic molecule. It's a whole different set of challenges and problems compared to the classical approaches to a variety of disorders, but if we could overcome these challenges, it would have great potential in medicine.
   It's such a cliche, but I think we're in for a major revolution in medicine in the next decade or two because of the explosive nature of the field of cell biology, and the fact that it's converging with other fields such as chemistry and structural biology. It's a terribly exciting time.

Do you have a "take-home message" from the perspective of a chemist working in cell biology?

   Schreiber: The future for chemists who are interested in biological problems is very bright because of our ability to make molecules that can enter inside the milieu of a living cell, perhaps even in animals themselves. But what's required is the use of the powerful tools of chemistry–the theories behind conformational analysis and the techniques behind organic chemistry that allow us to synthesize these molecules. But you also have to know enough about biological problems that you can determine what molecules you should be making and what experiments you should be doing with them. I think there's no getting around the fact that it requires a multidisciplinary effort. It can't be done with chemistry alone, and one might even say that it can't be done with biology alone.