Profile of Christopher Miller

For Christopher Miller, looking at a CLC transporter protein conjures images of peering down a double-barreled shotgun. With its twin channels, the evocative structure is characteristic of an unusual molecular family. CLC transporters are proteins, structurally similar to each other, that move chloride ions across cell membranes via surprisingly different mechanisms, either by ionic “leaks” through aqueous channels or by “pumping” the ion uphill by exchange with protons. Miller, elected to the National Academy of Sciences in 2007, has been studying these CLC proteins since before x-ray crystallography made it possible to see their structures. In his Inaugural Article (1) in this issue of PNAS, Miller presents evidence from a step-by-step investigation to determine whether the twin subunits in a CLC exchanger pump operate independently or cooperatively. Miller “straitjacketed” the protein, immobilizing the subunits relative to one another with disulfide linkages. This straitjacketing would cause failure of the pump if the two sides cooperated, but Miller shows that the pump continues to operate despite the restrictions.

A Sputnik Start

During his childhood in Greenwich, CT, Miller, now professor of biochemistry at Brandeis University (Waltham, MA), did not have an inordinate interest in science. “I liked the usual science things that little kids like, telescopes and chemistry sets,” he says. He did, however, have a passion for numbers. “I really did like math all through as long as I can remember.” His interest in science sparked during his junior year of high school, when he took his first science class, a physics course. Although he thought the teacher was terrible, Miller, 61, remembers the Sputnik-inspired, hands-on curriculum as scintillating. “I really lit up on that one,” he says.

When Miller entered Swarthmore College (Swarthmore, PA) in 1964, he remembers, “I thought I would like to be a math/physics geek of some sort.” Miller enjoyed his classes, but “What I learned during the course of that four years was that I was a middling to mediocre theoretical physicist,” he says. “I was never going to invent quantum mechanics myself.” Exposure to other physics students, like his friend John Mather, who would win the Nobel Prize in Physics in 2006, made Miller rethink life as a physicist. Miller branched out with other courses, including a biophysics class intended for those with no knowledge of biology. That course changed Miller's trajectory. “It gave me the career that I have,” he says. The class read research papers of biophysicists in the Philadelphia area and then visited their laboratories, which Miller recalls as his first encounter with research scientists. One of the scientists, molecular biologist Gilbert Ling at the University of Pennsylvania (Philadelphia, PA), would turn out to be Miller's future thesis advisor.

Christopher Miller

After graduating with a B.A. in physics, Miller knew he wanted join a doctoral program, but first he wanted to indulge a passion, teaching math. Miller took a position at Episcopal Academy (Marion, PA). “I knew that I wouldn't have any other time in my life to do this,” he says. He enjoyed the mix of math, performance, and control that teaching provided. Miller believed that being so close in age to his students gave him an advantage. “I knew all their little tricks,” he says. Still, he found it hard work. “Professoring is a whole lot easier than high school teaching.”

By the end of his first year, Miller had applied to multiple graduate schools, but he knew that he wanted to work with Ling at Penn and entered the Molecular Biology Program in 1969. “I came into this field knowing nothing biological,” he recalls. “I was a pure math/physics guy.” Ling appealed to Miller because “He was a very good teacher himself.” Miller also explains that Ling's “crackpot” theory that cell membranes do not exist appealed to Miller's sense of rebelliousness and desire to approach biology with the same purity as physics. Miller recalls that he learned valuable lessons about research when interpretations of his doctoral work on sugar transport and insulin action in skeletal muscle led him to disagree with Ling, who Miller says would not sign off on his thesis. Ling did allow Miller to continue working in the laboratory, a gesture Miller recounts as “honorable.” The experience “taught me a lot about science and how it works and what you have to keep your eye on when trying to make inferences from raw experimental data,” he explains.

Channeling the Unexpected

After receiving his doctoral degree in 1974, Miller began a postdoctoral position with Efraim Racker at Cornell University (Ithaca, NY). Miller had heard Racker talk at a Federation of European Biochemical Societies meeting in Atlantic City, NJ, in 1973 and liked his approach to science, which stood out as different from Miller's thesis research experience. “Instead of trying to test abstract theories, Ef was just trying to figure out what's going on,” explains Miller. Racker was the first to apply the biochemical reductionist approach to membrane proteins, purifying them and then reconstituting their function in artificial membranes. Undertaking a project funded by a grant from the Muscular Dystrophy Association, Miller used Racker's reconstitution methods to try to reconstitute calcium ion pump proteins in muscle cells to study how the sarcoplasmic reticulum releases calcium into the cytoplasm.

Soon Miller discovered, unexpectedly, that the rabbit sarcoplasmic reticulum membranes he was working with had potassium channels too (2). He found himself thinking that he could do electrical recording of ion channels without the fine dissections and manipulations required for traditional cellular electrophysiology. “I couldn't see myself doing the nerve cell dissections that card-carrying electrophysiologists do,” he says. “I had clumsy hands.” The artificial membrane technique allowed him to ask the same questions with a less demanding method for his shaky hands. During his time with Racker, Miller developed a system to put biological membranes into a reduced system for study (2), but the initial project eluded him. “I couldn't get results with the calcium pump,” he says. Miller remembers thinking, “Am I going to work on what I said I would work on or with what nature threw in my lap?” He decided to stay with the potassium channel studies, taking the work with him to his assistant professorship at Brandeis University in 1976 and picking up the electrophysiology background he needed as the project progressed. The potassium channel that he had discovered was unusual, he explains, in that it was “extremely simple in its mechanism,” unlike the neuronal channels under study at the time. According to Miller, the potassium channel provided a good, basic model system for studying the function of a protein channel opening and closing. Miller recalls that others expressed doubt about his work and thought that the channel was an artifact. He dubs this reaction “proper skepticism” and says that he “rather liked it” because it meant less competition and resulted in “five years of elbow room.” Over the next 20 years, Miller would continue to explore the field of ion channels.

In the late 1970s, neither protein structures nor DNA sequences were available. Placing a single protein in an artificial membrane and measuring its activity with electrodes, a technique known as single-channel reconstitution, provided a method for visualizing channel protein action (3, 4). “This technique was extremely powerful for watching the electrical function of single molecules in a chemically defined system,” Miller says. Throughout the 1980s, he continued to use single-channel recording to look at the details of how channels behave. Because the technique allowed him to produce low-resolution physical pictures of channel proteins, for example the double-barreled character of CLC chloride channels (5), Miller once titled a paper “Feeling around inside a channel in the dark” (6).

By the end of the 1980s, channel genes were being cloned, and Miller could combine the recording technique with genetic manipulation. He proceeded with what he recalls as a “satisfying series of experiments” with postdoc Jacques Neyton. Together they determined what kind of environment ions encountered inside a type of channel known as the calcium-activated BK potassium channel. BK channels are active in key physiological processes such as smooth muscle contraction and rhythmic nerve activity. Miller primarily studied the channels from rat muscle, but BK channels were ubiquitous. “Everywhere we ground it up, it seemed to be,” Miller says. He recalls that he once went to the supermarket and bought “super-cheap, week-old ground beef” and tried to record the channel, and “there it was.” Now that there are techniques more advanced than grinding up hamburger, he explains, the BK channel is known to be ubiquitous. With the single-channel recording technique, Neyton and Miller developed a cartoon picture of three K⁺ ions lined up in a pore of the BK channel. Miller recalls the satisfaction that came from that picture, as well as the satisfaction that came 10 years later when crystal structures came along and showed the ions to be just where his cartoon had predicted (7).

Miller enjoying a lobster.

Backburner Heats Up

A project that had simmered on the backburner in Miller's laboratory for most of the 1980s grew to overtake his work into the 1990s and beyond. Miller now exclusively studies chloride channels, work that began unexpectedly in the late 1970s. Amid skepticism over the validity of the sarcoplasmic reticulum potassium channel, Miller had decided to use his artificial membrane technique to show that he could find channels known to be there. He used the technique on what he thought would be a sure thing—the acetylcholine receptor from the electroplax organ of the electric ray Torpedo californica. As Miller explains, “the organ is like a high-powered battery,” replete with acetylcholine receptor. But the technique did not pull out the intended receptor. Instead, Miller found a chloride channel that had never been seen before (8). He says that he felt like nature again threw something in his lap. Miller would continue to study this chloride channel off and on over the next few decades. Using results from single-channel recordings, he drew another cartoon, this one depicting the protein's “idiosyncratic properties: two pores as a double-barreled shotgun” (5). Usually, Miller explains, channels are built with only one pore. Multiple subunits wrap around a single axis to form the pore, but the Torpedo channel appeared different and unprecedented. Miller became hooked, eager to explore this channel in depth. “I'm always a little attracted to the oddball thing to work on,” says Miller.

By 1990, Thomas Jentsch's group at Hamburg University (Hamburg, Germany) had cloned the gene for the Torpedo channel (9). Further work by Jentsch showed a profusion of homologs in other organisms, including humans (10). The Torpedo channel, named CLC-0, turned out to be the founding member of a large molecular family known as the CLC chloride channels. Biologically speaking, the proteins were important, and their malfunctions were implicated in a variety of diseases, but Miller remained interested in the biochemistry of membrane proteins. Through the 1990s, Miller engaged in more work on CLC-0, showing that its gating is an unusual, nonequilibrium process driven by ion movement through the protein itself (11), and purifying the protein to show the functional channel to be a two-pore homodimer (12, 13). “I ignored the biological side,” he says. “I was more interested in the protein's structure and mechanism.” In 1998, Rod MacKinnon (The Rockefeller University, New York, NY) crystallized and solved the structure of a bacterial K⁺ channel homolog (14). “It changed the whole face of the field,” Miller says. “At that point, I started to look for bacterial homologs of the CLC chloride channels.” He found them and, with his Brandeis colleague Niko Grigoieff, determined a low-resolution structure with electron crystallography (15). In 2002, MacKinnon's group published a high-resolution x-ray crystal structure of this same bacterial CLC homolog; it was, as Miller had predicted, double-barreled.

An Ambiguous Interface

Soon after the x-ray structure was solved, Miller found yet another “unexpected thing.” In 2004, the same year he garnered a John Simon Guggenheim Fellowship for a visit to Rod MacKinnon's Laboratory (MacKinnon had been Miller's undergraduate student in the 1970s and postdoc in the late 1980s), Miller showed that the bacterial CLC homolog is not an ion channel but rather a kind of “pump” known as an exchange transporter (16). There are two broad classes of proteins that cells use to move ions across membranes—pumps and channels. Pumps require energy to operate as they move ions uphill against concentration gradients. The bacterial CLC uses a proton gradient to transport chloride across membranes (16). Channels, Miller explains, are “nothing but a stupid old hole.” Ions leak through the pore with no energetic cost. The CLC family, it turns out, has two subfamilies, split almost in half between channels and pumps. The channels, such as the Torpedo CLC-0, are mostly found in outer, plasma membranes, whereas the pumps are mainly found in intracellular membranes. The pump members are important in regulating the pH across the lysosome membrane, for instance. “This was a very satisfying discovery because it was so weird,” he says. The channels and pumps are both built on the same molecular architecture, and Miller is currently trying to figure out how these structurally similar proteins do functionally different things. Miller explains that, traditionally, channels and pumps had been considered to be completely different kinds of proteins, but now there is a conference scheduled for 2008 about “the ambiguous interface between channels and pumps.”

Miller describes his Inaugural Article as a small part of this investigation, conducted over the last two years, to understand how the CLC Cl⁻/H⁺ exchanger, a pump, works (1). We are “generating various interesting mutants in hopes of dissecting the mechanism,” he says. The Inaugural Article addresses whether the two halves of the transporter, a homodimer, cooperate or act independently. If the subunits cooperate, they would have to move relative to each other. “The interface would have to wiggle,” he explains. To attack this question, “we straitjacketed that interface.” Using up to four covalent cross-bridges, Miller immobilized the two barrels of the bacterial transporter CLC-ec1, relative to each other. When reconstituted in an artificial membrane, the transporter still functioned, showing that the two sides act independently (1). The work provides a “small glimpse into the behavior,” and Miller hopes that “The accumulation of these bits of mechanistic information will, if we're lucky, eventually bring insight” into the larger question of how structurally similar proteins can have such different functions.

“I'm always a little attracted to the oddball thing to work on.”

Miller, a Howard Hughes Medical Institute Investigator since 1989, plans to continue to work on the junction of structure and function at this “ambiguous interface” and has added x-ray crystallography to the electrophysiological techniques in use in his laboratory. Miller maintains a small laboratory and stays personally involved at the bench. “I'm a kind of annoying micromanager,” he explains. When he is not teaching, Miller can be found in the laboratory, doing experiments himself. As a tenured professor, he says, “Why would you not do the thing you like to do best?”