Clarifying protein atomic structures for drug discovery
24 November 2006 (Volume 1 Issue 11)
To aid drug discovery, researchers at RIKEN's Structural Biophysics Laboratory are clarifying protein structures at the atomic level. Using bovine rhodopsin as a mammalian model, the team was the first to elucidate the atomic structure of a very important target in drug discovery—the so-called G-protein coupled receptor (GPCR). Recently, the team also succeeded in elucidating the catalytic mechanism by which an enzyme, named sphingomyelinase, snips off a lipid. These findings are expected to pave the way to new drug discoveries.
Crystal structure analysis of bovine rhodopsin
The finding on GPCR was published in August 2000 in the US magazine Science1 and has since been cited in some 2,000 international scientific papers. The paper was published by a collaborative team at the Structural Biophysics Laboratory of the SPring-8 Center and Washington University in the US.
Team leader, Masashi Miyano, Chief Scientist of the Structural Biophysics Laboratory, says that the attractiveness of the paper lies in the importance of GPCRs as targets for drugs to cure various diseases. “More than 50% of the drugs used currently exert therapeutic actions via stimulation or reduction of the function of GPCRs induced by the drug binding to this receptor,” he explains. “Thus, pharmaceutical companies are eager to know GPCR atomic structures ahead of their competitors.”
As GPCRs are embedded in the cell membrane, they can interact with the environment inside and outside the cell. As such, they play a crucial role in receiving communication signals from other cells. This occurs when signalling molecules, called ligands, bind to the active site of the receptor outside the cell, which in response, activates certain cellular responses. Much drug development today is focused on finding chemicals that modulate the ability of ligands to bind with GPCRs, thereby either inhibiting or accelerating cellular processes.
To date, about 1,000 kinds of GPCRs have been identified in the whole human genome. Their corresponding ligands vary with about half of them being odorant receptors. Hormones, metabolites, proteins, lipids, sugars, amino acids, and viruses are also recognized by specific GPCRs. Due to the biological and medical importance of these receptors, various GPCRs have been studied since the 1980s to clarify how they transmit information.
In 2000, Miyano and colleagues elucidated the world’s first atomic structure of GPCR (Fig. 1) using rhodopsin, which exists in photoreceptor cells, or rod cells in retina. Its ligand is a photon. “The mechanism to transduce information is considered to be common among all GPCRs, which means they should have a common structure for the shared common activation mechanism of G-proteins,” Miyano explains. “Clarifying the atomic structure of rhodopsin must have had a significant influence on the development of novel GPCR-targeted drugs. In fact, pharmaceutical company researchers have often told us directly that ‘it was really beneficial to our R&D’.”
Figure 1: Three dimensional structure of bovine rhodopsin
Bovine rhodopsin is one of the g-protein coupled receptors (GPCRs). Rhodopsin is embedded in the cell membrane of a visual cell, and has the seven transmembrane α-helix structure. A newly-discovered, short helix 8 structure (indicated by the dotted circle) follows the seven transmembrane helix structure in the membrane surface inside the cell.
A mature and functional protein has its own defined spatial shape—a three-dimensional structure—in which the polypeptide chain coded by the DNA of the gene, as an amino acid sequence in the genome, is folded properly. In other words, a protein with an incorrectly folded polypeptide chain does not function and is even toxic in spite of having the same amino acid sequence. Miyano’s teams’ dogma is: ‘Function and structure are two faces of the same thing in a biological system’. Thus, they aim to understand living organisms from the standpoint of the three-dimensional structures of proteins.
Miyano has two guiding principles in doing research. “One is to observe the three dimensional structures of proteins at the atomic level,” he says. “We will be able to interpret and regulate the function of proteins only by using the three dimensional structure of the protein at the atomic level.”
Miyano utilizes x-ray crystallographic analysis, not nuclear magnetic resonance analysis or electron microscopic analysis, to determine the protein structure at the atomic resolution since it is most economic and feasible. “Crystallization of a target protein was one of the biggest obstacles in analyzing the atomic structure using x-ray crystallography,” notes Miyano. “Some proteins took several years to crystallize. However, continuous and extended efforts on protein crystallization have been making it easier with the exception of some specific proteins.”
The teams’ second guiding principle is ‘focus on drug discovery’. Proteins are functional when their active sites bind molecules such as other proteins and lipids. This means it is possible to modulate protein function if researchers understand the structure of a target protein at an atomic level. Researchers can also artificially synthesize chemicals that are designed to bind the active site of the target protein. “Designing those molecules will lead to the development of new drugs,” says Miyano.
Membrane proteins are a difficult research target
Since clarifying the structure of bovine rhodopsin in 2000, using the world-leading Synchrotron Radiation Facility, at SPring-8, researchers around the world have failed to clarify the crystal structure of other mammalian GPCRs. “Some groups have succeeded in analyzing the crystal structure of bovine rhodopsin,” comments Miyano. “However, they have clarified bovine rhodopsin at the same inactive state as we did. Nothing is new except for achieving a higher resolution. We are also trying to analyze the crystal structure of other GPCRs, but it really is a big challenge.”
The reasons why the crystal structure of bovine rhodopsin alone has been analyzed are readily identifiable. In general, a large amount of protein is needed to grow crystals. Researchers often use the bacterium Escherichia coli for over-expression of protein. However, this method cannot be applied easily to proteins like GPCRs that are embedded in the cell membrane. As bovine rhodopsin is concentrated in cow retinas, it can be obtained in bulk from meat-processing plants. Unfortunately, this is the only natural source of GPCR so readily available.
The crystallization of membrane proteins is also extremely difficult because they are embedded in the cell membrane. In the case of rhodopsin, Tetsuji Okada at Washington University spent five years growing crystals of bovine rhodopsin. He started the research when he was a special post-doctoral researcher at RIKEN. Okada is now the Chief Scientist at the Japan Biological Information Research Center, National Institute of Advanced Industrial Science and Technology.
“If we know the atomic structures of GPCRs from various signal transduction systems or from different living organisms, we can compare their structural differences and similarities. And this would be a significant step toward the elucidation of the structural basis of the signal transduction mechanism through GPCRs,” says Miyano. To this end, his team is also working on the over-expression of membrane proteins using yeast and insect cells to crystallize membrane proteins.
Further explaining why it is so difficult to crystallize membrane proteins Miyano says: “To grow a protein crystal, we have to determine the conditions for the precipitating agents and pH (proton concentration in water) for each protein. We also have to take surfactants into account which makes the task more difficult. Cell membranes are made of lipid molecules. To extract membrane proteins from cell membranes, we need surfactants that can make oily lipids dissolve in water. It still usually takes a few years to crystallize a membrane protein. Yet we often fail to grow high-quality crystals that can be used for structure analysis.”
Worldwide, scientists are competing fiercely to overcome the challenges in analyzing the structure of GPCRs. In Europe, a project called MePNet (Membrane Proteins Network) is proceeding. Under such circumstances, how can Miyano come out a winner?
“The next two or three years will be critical. What is necessary is not a major breakthrough, but a succession of small things integrated together step by step,” Miyano says. “I am sure that we will succeed in clarifying the structure of new GPCRs again—by making the most of the dexterity and good teamwork for which Japanese people are characterized. Also the SPring-8 facility houses one of the world’s best instruments to look at such things.”
Lipids, another research target
Lipids are another axis of research target in Miyano’s laboratory. Lipids, along with proteins, are a major component of living bodies, and are also the main component of the cell membrane. “We pay attention to lipids because lipids themselves are ‘mediators’,” comments Miyano.
Many types of mediators are produced in living organisms. They are substances that bind to proteins to transmit signals and control the functions of living organisms. “What is interesting about lipid mediators is the fact that a single lipid changes into various derivatives, each of which exerts different functions,” Miyano says. For example, a lipid is broken down into segments by enzymes in a variety of ways, or it can be modified by various sugars and phosphoric acid. Some lipid mediators function through the specific GPCRs as selective modulators. “We advance our research activities based on the assumption that assuming that we may be able to control various functions acting in living organisms by clarifying the interaction between lipids and proteins.”
The Structural Biophysics Laboratory, in cooperation with Tokushima Bunri University, have clarified the crystal structure of a protein called sphingomyelinase, and elucidated its catalytic mechanism. This protein is an enzyme that serves as scissors to snip the lipid called sphingomyelin.
Hideo Ago, a senior research scientist at the Laboratory explains the background to this research. “The functions of some proteins are controlled by the number and kind of metal ions that bind to the proteins. It is also known from biochemical experiments in the 1970s that there are two types of metal ions; one enhances enzymatic activity significantly, and the other enhances enzymatic activity slightly. However, the mechanism had remained a mystery,” Ago says.
When sphingomyelinase snips sphingomyelin, it produces ceramide and phosphocholine that serve as mediators. Sphingomyelin is also considered to form a structure on the cell membrane, called a ‘raft’, upon which the GCPRs are floating. “We began this research assuming that if the crystal structure of sphingomyelinase is clarified and its catalytic mechanism is elucidated at the atomic level, we can expect a new interesting approach that links lipids, cell membranes, and even GPCRs,” says Ago.
Figure 2: Two types of metal ion binding structures at the active site of sphingomyelinase.
Water-bridged divalent metal ions binding structure
Enzymatic activity is significantly increased when both glutamic acid and histidine bind a divalent metal ion and the two metal ions are bridged by a water molecule. Cobalt ions are used for this structural analysis. Within an organism, magnesium ions work the same way.
Calcium ion binding structure
Enzymatic activity of this model is not much enhanced because the calcium ion is larger than the magnesium ion or cobalt ion. Thus the bound calcium ion at the glutamic acid prevents binding of another metal ion at the histidine.
Figure 2 shows the metal ion binding structures at the active site of sphingomyelinase that the team has clarified to date. Enzymatic activity is greatly enhanced when both glutamic acid and histidine bind a metal ion each, and a water molecule is placed in between the two metal ions (Fig. 2, left). In contrast, enzymatic activity is not much enhanced when a calcium ion binds to the glutamic acid (Fig. 2, right). The water-bridged double divalent cations are considered to be essential in enhancing enzymatic activity significantly. When a calcium ion has bound to the glutamic acid, there is no room left for another ion to bind to the corresponding histidine because the size of a calcium ion is larger than a metal ion which enhances the enzymatic activity. “The mechanism is simple and convincing when you look at the structures,” notes Ago. “We succeeded in clarifying the structure because we managed to observe the structure at the atomic level.”
In this study, the team used sphingomyelinase from Bacillus cereus. However, humans have similar enzymes. “It has been found that the function of human sphingomyelinase is related to apoptosis in nerve cells and lung emphysema caused by cigarette smoking,” says Ago. “Furthermore, ceramide, which is produced when sphingomyelin is broken down, are related to skin moisture retention. Thus we think the findings of this research will lead to the discovery of various drugs in the future.”
In considering the future of his teams’ research Miyano says, “We will do what we can do now. As a first step, we are trying to clarify the crystal structure of new GPCRs as soon as possible. We will also advance the use of SPring-8, and address the challenges to be faced, no matter how difficult.” Miyano also considers that now is a very interesting time to be involved in research. “Things we have never imagined are being clarified one after another—protein research is truly very interesting,” he notes. “However, although it is really interesting, we cannot afford to take enough time to appreciate and enjoy it. This, I think, is the tragedy of today’s life science.”