Understanding the Colors of Life: A Norway-California Collaboration

FEATURED
Abhik Ghosh, University of Tromsø, Norway, and SDSC Senior Fellow
Peter Taylor, SDSC and UC San Diego
David F. Bocian, UC Riverside

ften called the "colors of life," porphyrins are responsible for the red color of blood and the green of plants. But besides being intensely colored, the ring-shaped compounds in the porphyrin family also play key roles in many vital processes of life. Porphyrins are the structures in proteins that carry around the oxygen, electrons, and stored energy to make these processes happen. Through the bloodstream, the protein hemoglobin carries oxygen on its porphyrin rings. A related protein, myoglobin, carries oxygen in the same way through the muscles. And even though porphyrins have been studied for a hundred years, computational chemists and experimentalists are even now uncovering new surprises about porphyrin chemistry.

"Porphyrin chemistry is Shakespearean, in that 'age cannot wither, nor custom stale,' its infinite variety," said Abhik Ghosh, an associate professor of Chemistry at the University of Tromsø, Norway, and an SDSC Senior Fellow. "In spite of more than a century of intense study, the vibrancy and variety of porphyrin chemistry shows no sign of waning."

Working in Norway and California, Ghosh has long sought to exploit recent advances in high-performance computing and quantum chemistry, especially density functional theory (DFT), to unravel key problems in the chemistry and biology of porphyrins. In this effort, Ghosh is leading a far-flung group of collaborators.

The work described here is, in part, a continuing collaboration between Ghosh and David F. Bocian, a distinguished professor of chemistry and experimental spectroscopist at UC Riverside. Ghosh is also collaborating with Peter Taylor, professor of chemistry at UCSD and deputy director of SDSC, with the aim of reexamining a number of issues of porphyrin chemistry with advanced non-DFT methods being developed by Taylor and others. Ghosh's coworkers in Norway include graduate students Torgil Vangberg, Emmanuel Gonzalez, and Tebikie Wondimagegn.

Ghosh recently ended a one-year stint as a visiting scientist at SDSC, which regularly hosts international researchers. Other current and recent visiting chemists include post-doctoral researchers Dan Jonsson from Sweden, Robert Polly from the University of Technology in Graz, Austria, and Kenneth Ruud from Norway, and visiting scientist Carlos Quintanar from Mexico.

A SURPRISING TILT

OLD PROBLEMS,
NEW LIGHT

Figure 1. Tilting and Bending

Potential energy contours (kcal/mol) for tilting (tau) and bending (beta) of Fe(II)CO (orange) and Fe(III)CN (blue) units in hemes.

A SURPRISING TILT

One key issue addressed by Ghosh and Bocian is the following: When heme, the iron-containing cofactor of myoglobin, is separated from the protein, it binds poisonous carbon monoxide much more strongly than it binds oxygen. How, then, does the protein protect us from suffocating on the carbon monoxide produced naturally in the body or inhaled while smoking? The conventional explanation, based on early crystallographic studies of myoglobin, was that the protein forces the bound carbon monoxide to adopt an unfavorable, highly bent orientation. For oxygen, however, such a bent orientation is natural. In other words, a respiratory protein such as myoglobin was thought to actively discriminate against binding with carbon monoxide, in favor of binding with oxygen.

In the early to mid-1990s, a number of spectroscopists interpreted their findings in terms of an essentially upright carbon monoxide bound to myoglobin, thus directly contradicting conventional ideas. Considerable publicity attended the death of the "bent-CO dogma" and the emergence of a straight-CO viewpoint. But it turns out that the straight-CO view was also based on a flawed interpretation of experimental results.

Ghosh and Bocian's surprising finding is that the iron-carbon monoxide unit, Fe(II)CO, is deformable when tilting (tau) and bending (beta) occur cooperatively. (The carbon atom tilts away from being perpendicular to the porphyrin ring, and the oxygen bends away from the straight line connecting the iron and carbon.) The tilted elliptical contours of the FeCO potential energy surface, calculated by Ghosh using supercomputers in Norway and at SDSC, reflect a large tilt-bend interaction constant (Figure 1). Note the relatively untilted elliptical contours for deformation of the Fe(III)CN group, which has a comparatively ordinary potential energy surface. (Here II and III refer to the oxidation state of the iron.) "The key insight from quantum chemistry has been that moderate deformation of bound carbon monoxide does not cost sufficient energy to be the main basis for myoglobin's discrimination against carbon monoxide," Ghosh says.

Initially greeted with some skepticism, the key feature of the Ghosh-Bocian potential surface--namely a large tilt-bend interaction constant--has now been confirmed by a number of laboratories. The current view of the problem is that myoglobin does not actively discriminate against carbon monoxide. Instead, it stabilizes the bound oxygen through hydrogen bonding, which accounts for the majority of the "discrimination energy."

Figure 2. Unexpected Orbital Overlap

Direct orbital overlap between porphyrin (pi) cloud and ortho bromine substituents on mesophenyl groups, seen from above (left) and from the side (right).

OLD PROBLEMS, NEW LIGHT

"A few years ago, the physical and theoretical aspects of porphyrin chemistry seemed to be gradually slipping away from the frontiers of chemistry research," Ghosh said, looking back on developments in his field. "Apparently, the key issues had been beaten to death. Today, I am convinced otherwise. Old problems are appearing in a completely new light, and with breathtaking advances in computational technology, there are many new problems to work on."

Recent computational studies by Ghosh and others have revealed a variety of orbital interactions in porphyrin-type molecules that had not been suspected previously. For instance, the effects of changing certain groups (ortho-substituents) in tetraphenylporphyrins, synthetic molecules widely used in biomimetic catalysis of hydrocarbon oxygenation, have long been interpreted only in terms of steric and inductive effects. Theoretical work by Ghosh's research group and spectroscopic work by Bocian's group have revealed an unexpected orbital overlap between the ortho substituents and the (pi)-cloud of the porphyrin (Figure 2). This finding may be relevant to the design and working of nanoscale electronic devices based on large multiporphyrin arrays.

Another example of an interesting orbital interaction studied by Ghosh and coworkers is the metal-ligand interaction in six-coordinate Fe(III) porphyrins, or ferrihemes. These molecules are models of the active sites of the cytochrome b proteins involved in biological electron transfer. Quantum chemical studies have allowed visualization of two distinctly different patterns of distribution of the unpaired electron in these molecules. Which distribution is preferred depends on the axial ligands and the particular porphyrin-type equatorial ligand.

Figure 3 depicts the open-shell orbitals for the two configurations. The computational studies also elucidated an important structural feature of these Fe(III) hemes: the porphyrin ring is significantly nonplanar or ruffled, which carries many implications for chemical behavior.

Adds Peter Taylor, "Translating these problems from the DFT environment into ab initio calculations of the kind my group specializes in has begun in collaboration with Ghosh and his group. These calculations represent a real challenge to the limits of computational chemistry and of computing power." --MM

REFERENCE
Ghosh, A., 1998: Accounts of Chemical Research 31, 189-198.

Figure 3. A Surprising Interaction

Two modes of iron(3d)-porphyrin(pi) orbital interactions in low-spin six-coordinate hemes, modeled after cytochrome b active sites. If the 4-cyanopyridine ligands are assumed to coordinate along the z direction, one interaction involves the xz/yz orbital, and the other involves the xy orbital. The latter interaction, a surprising one, is possible because of the nonplanar, ruffled geometry of the porphyrin macrocycle, shown in the stick diagram of the molecular geometry (right). Such a geometry may also strengthen the interaction.