My odyssey in the world of molecules

(Part 1: Embarking on my journey)

"Man is simply a collection of molecules" and "can be understood in terms of molecules." – Linus Pauling

We have been fascinated by the various parts of our anatomy since time immemorial. As we developed more and more sophisticated tools, we learned more and more about ourselves and we began to make the connection between our actions and the components of our bodies. We are now probing deeper and deeper into our anatomy and we have begun to describe ourselves in terms of molecules and even of atoms. Nowadays, we ascribe not only our physical attributes, but sometimes also our mental capabilities and idiosyncrasies, to the molecules that make up our system. To be able to do this, we have found that we need to understand how those molecules work.

Key to the understanding of molecules and their function is structure. The visualization of the detailed structure of molecules did not become feasible until the early part of the last century. Since then, three-dimensional structures for all sorts of molecules, from simple ones comprised of only a few atoms to complex ones comprised of hundreds of thousands of atoms, have become available. This became possible when physicists brought their tools, in particular, spectroscopy and crystallography, to bear on the problems of biology. It is the continuing cooperation of biologists, chemists, and physicists – and, more and more, mathematicians – that might eventually lead us to a full molecular understanding of our selves.

This is a narrative in two parts of some of my adventures and misadventures in the world of molecules. In this part, I recount my early work. In Part 2, I describe how I am trying to contribute to the utilization of the structural knowledge that we now possess of some interesting biological molecules. Naturally (and I hope I will be forgiven for it), I will focus on my (few) successes and not dwell on my (many) failures.

My introduction to the world of molecules

Armed with a bachelor’s degree with a major in physics, earned from the University of the Philippines in 1960, I enrolled at the Johns Hopkins University in Baltimore, Maryland in 1964 to learn biophysics. I chose for my thesis adviser, Professor Warner E. Love, the crystallographer in the faculty. Warner, as we called him (we were not particularly formal at Hopkins and we called our professors by their first names, sometimes by their nicknames), was interested in the structure of proteins, their function, and their evolution. At that time, the three-dimensional structure of only two proteins, myoglobin and tetrameric hemoglobin, had been determined (both obtained by X-ray crystallography). The elucidation of the three-dimensional structure of a protein was quite an accomplishment then and John Kendrew and Max Perutz were awarded the Nobel Prize in 1962 for their work on those two proteins. Myoglobin and hemoglobin, whose primary function is to bind oxygen reversibly, turned out to have the same tertiary structure — soon termed the "myoglobin fold." But, those two are mammalian proteins. Many other organisms, including many invertebrates, some prokaryotes, and even some plants, have hemoglobin. Do those non-mammalian hemoglobins also have the "myoglobin fold"? Warner Love wanted to know.

For my Ph.D. thesis, I decided I would determine the three-dimensional structure of an invertebrate hemoglobin. Because I was young and foolish (and ambitious, as Dean Rhodora Azanza put it two years ago when she heard me describe my first venture into research), I chose to work on a giant hemoglobin – one with a molecular weight of four million. [Hemoglobins range in size from about 12 thousand (truncated myoglobins), to 17 thousand (myoglobins and monomeric hemoglobins), to 67 thousand (tetrameric hemoglobins like the one we have in our blood), to several million (the giants with 180 subunits).] I was going to use a combination of X-ray crystallography and electron microscopy. So, I picked a second thesis adviser, Professor Michael Beer, the resident electron microscopist. I proceeded to isolate the giant hemoglobin from a marine annelid [I bought a couple of thousand worms from a bait company, cut them up with scissors, squeezed the blood out from the carcasses, strained and centrifuged the debris out, and "purified" the hemoglobin by ammonium sulfate fractionation – a very crude protein purification procedure (we didn’t have chromatography then)] and set up the hemoglobin to crystallize. It didn’t. Even after weeks of waiting, all I got was "mud" (my term for the amorphous precipitate that formed in my crystallization vials). I isolated some more hemoglobin and tried again. Again, all I got was "mud."

But while waiting for the giant hemoglobin to crystallize, I had started working on another hemoglobin – from another marine annelid, Glycera dibranchiata – one that had the more manageable 17-thousand molecular weight. This hemoglobin quickly yielded beautiful crystals. And just as quickly, I dropped the giant hemoglobin project (and Michael Beer) and focused on Glycera hemoglobin. Soon enough, I had a three-dimensional structure at low resolution. This is an understatement. At that time, data collection took months and we had to write many of the computer programs we needed, and our computer was very slow (millions of times slower than present-day personal computers) and had scant memory (many millions of times less than what the computers of today have). Computer graphics was still many years in the future and we had to draw electron-density contours by hand and build our models by hand. Nevertheless, we managed. It was obvious, even at low resolution, that Glycera hemoglobin has the "myoglobin fold." (And so does every other hemoglobin whose structure has been determined to date.) Warner had his answer. We published the Glycera structure in Nature and it caught the attention of a science writer who featured our results in an article in the London Times. Needless to say, I enjoyed the (brief) publicity. All this happened in 1968. Some years later, we pursued the structure analysis of Glycera hemoglobin to "atomic resolution," which allowed us to assign a position to every atom in the molecule. The paper that we published on the atomic structure of Glycera hemoglobin did not go unnoticed. Shortly after that paper came out, I received a letter from Linus Pauling asking permission to include a figure in our paper in a book that he and his son, Peter, were writing. Of course, I consented. Receipt of that letter from the scientist I admire the most thrilled me no end – and it still does.

Early adventures in the world of antibodies

In 1971, I joined the group of David R. Davies at the National Institutes of Health (NIH). David (we were also not very formal at the NIH) had become interested in antibodies and his group had determined the structure of a human cryoglobulin (an antibody-like molecule produced in large amounts by a patient who was suffering from multiple myeloma and which crystallized in his serum when the weather turned cold, causing the clogging of blood vessels – the poor patient had to have extremities amputated before he mercifully died of the disease). When I arrived in the lab, David had just started a collaboration with Michael Potter of the National Cancer Institute (NCI). Mike Potter had a collection of many hundreds of mouse myeloma proteins, many of which had known ligand-binding specificity. Although Mike Potter’s proteins were produced by plasma cell tumors, the fact that many were found to bind common antigens lent support to the belief that they were actually true antibodies. Mike Potter provided David’s group with Fab (the antigen-binding fragment) from several of those mouse myeloma proteins and David Segal, a protein chemist in David Davies’ group and who was also a Hopkins Ph.D., proceeded to crystallize them. The first one to crystallize was McPC603 Fab, which has phosphocholine-binding specificity, and I was tasked to determine its crystal structure.

The crystallographic analysis of McPC603 Fab was not easy. Although large crystals were readily obtained, they were not very good diffractors, i.e. X-ray data from the crystals were not of high quality and were difficult to collect. I spent many hours taking X-ray photographs of crystals (on many occasions, I would expose the film for a week) and preparing heavy-atom derivatives (needed to determine the structure). And after suitable heavy-atom derivatives had been obtained, it took me several months to collect the necessary data for a structure analysis. All in all, it took me a year to determine the structure of McPC603 Fab at low resolution. We published the low-resolution structure of McPC603 Fab in Nature New Biology and it caught the attention of another science writer who featured our results in a short article in Nature.

We then pursued the crystallographic analysis of McPC603 Fab to higher resolution, with Dave Segal doing most of the data collection and analysis. We soon had an electron-density map that was of high-enough resolution to allow us to build an atomic model of the molecule. It took me six months to build the model of McPC603 Fab and another six months to measure the coordinates of the atoms. (The model now sits in the NIH museum.) We published our findings and were soon invited to several meetings to present our results.

The early 70s saw the flowering of the structural analysis of immunoglobulins. Structures of various myeloma proteins and fragments had become available from the laboratory of Roberto Poljak at Johns Hopkins, from Alan Edmundson’s at Argonne National Laboratories, and from Robert Huber’s at the Max-Planck Institute in Martinsried, Germany. The amino-acid sequence data on immunoglobulins, which were starting to accumulate, could now be correlated with three-dimensional structure. We were now able to visualize the hypervariable regions – the Complementarity-Determining Regions (CDRs) which Elvin Kabat and T.T. Wu had proposed to be the segments that form the antigen-binding site of antibodies. Since the structure of McPC603 Fab that we had determined had phosphocholine bound to it, we were able to confirm the Kabat-Wu proposal. The structure of antibodies was the subject of several science reports published in Nature, Science, the Journal of the American Medical Association, National Geographic, and various foreign newspapers. Those were heady days.

More structural information on antibodies and their interaction with antigen soon became available and it did not take long for the structural knowledge to be put to good use. The application of that structural knowledge to medicine is the subject of Part 2.

* * *

Eduardo A. Padlan has a Ph.D. in Biophysics and was a research scientist at the US National Institutes of Health until his retirement in 2000. He is currently an adjunct professor in the Marine Science Institute, College of Science, University of the Philippines Diliman. He is a corresponding member of the National Academy of Science and Technology, Philippines. He can be reached at [email protected] or at [email protected]