American Scientist

Imagine you’re barreling downhill on a backcountry ski slope. Atop the skis, you’re buffeted back and forth—by the wind, by the iciness of the snow pack, by the changing direction that you’re leaning to stay upright. All these factors will affect your path down the slope and your destination. During chemical reactions, molecules similarly can be affected by circumstances. Along the way to products, they reach a point of peak energy. Their journey forward then tracks downward, like a ski slope. And the path of that downward chemical reaction can fork, in a split known as a post–transition state bifurcation. Chemists have started to map the multitudes of available molecular trajectories heading downhill, and these maps are leading them to new ways of controlling the outcome of reactions.

In undergraduate courses, students typically learn that chemical reactions follow straightforward pathways over a hill from start to finish. A reactant (a starting molecule or collection of molecules at a low point in potential energy) converts to a product (another molecule, or a collection of molecules) at a low point in potential energy. That path goes through a transition state—the structure with the highest potential energy—which is the peak separating reactants and products (see sidebar, below).

In situations in which a reaction produces more than one product, chemists had generally assumed that separate pathways with different transition states led to each: The transition states had different total energies, and therefore the reactions faced different barriers and occurred at different rates. The reaction with the lower-energy transition state occurred more rapidly and produced more product. But these principles don’t explain all that chemists observe in the laboratory. Some reactions produce ratios of products that can’t be understood using these simple rules.

Recent studies using quantum chemistry have furnished a more complicated picture of reactions: Many don’t have simple one-track, downhill pathways. Instead, after reaching transition states, some pathways can bifurcate, splitting toward two or more low points in potential energy, which results in multiple products.

The revelation that a single transition state can produce more than one product has significant implications for chemists and others working with molecules. The synthesis of drug compounds and other important chemicals, the synthesis of polymers and other materials, the engineering of enzymes, and more, all rely on chemists’ ability to control reaction processes and produce desired products while minimizing or eliminating undesired ones.

This revelation also complicates a chemist’s traditional view of reactions —that a reactant morphs smoothly in structure and energy to result in a product. Instead, chemists must consider the energy of vibrations in reacting molecules and the ways that those oscillations alter the distances and angles between their atoms. As Richard Feynman famously said, “everything that living things do can be understood in terms of the jigglings and wigglings of atoms.” But only recently have researchers had sufficient computational resources to model these vibrations in more complex molecules, using quantum chemistry.

Modeling the Mayhem

Chemists have discussed the connection between molecular momentum and reaction mechanisms for approximately half a century. In 1970 theoretical chemist Lionel Salem of the French National Center for Scientific Research predicted that study of the dynamic behavior of reacting species would play an increasingly important role in how chemists rationalize chemical reactivity. Over the past three decades, organic chemist Barry Carpenter of the University of Bristol in the United Kingdom has argued convincingly that chemical reactions should be viewed and analyzed by considering atoms’ masses and vibrations alongside their positions, known as a phase-space perspective.

As Richard Feynman famously said, “everything that living things do can be understood in terms of the jigglings and wigglings of atoms.”

• Facebook
• Twitter

The notion of phase-space places the importance of atoms’ velocities and masses on equal footing with their positions (including the probable locations of electrons). For a reaction with a post–transition state bifurcation, the momentum associated with atoms that are not involved in breaking bonds or making new ones can control which product forms. A molecule’s vibrations, as it moves from an energy valley following a transition state to an energy ridge separating products, connect directly to which path it follows toward lower energies. That selection determines where the chemical “skier” ends up and which product is formed.

Applied theoretical chemists now run reaction simulations that distribute energy among vibrational modes of a transition state, activating particular bond stretches and angle bends to various degrees. Such calculations are known as dynamics trajectory simulations. When a transition state moves toward a product, its structure varies. Not only do those changes make it more like the product, but they also prompt other parts of the structure to jiggle and wiggle.

These calculations have already provided insights into why some chemical reactions produce unwanted side products. For instance, for many years synthetic chemists have been producing substances called β-lactones in the laboratory. Many β-lactones occur naturally and can have antimicrobial activity, which makes them interesting drug candidates. These compounds include 4-membered rings comprising three carbon atoms and one oxygen atom, with a second oxygen atom protruding from one of the neighboring carbon atoms. When carried out in the lab, many β-lactone–forming reactions produce a mixture of products. For example, the reaction described in the figure at the top of page 25 yielded only 18 percent of the desired β-lactone product and 50 percent of an undesired product.

Dynamics trajectory calculations for this process showed that the wiggling of molecules after passing the transition state determined which product they formed. A “split” in the reaction pathway after the transition state—a post–transition state bifurcation—led to both products, but the vibrations of the reacting molecules led to one product or the other. In this example, the phase-space viewpoint provides an unexpected model of reactivity that can be used both to reduce unwanted byproduct and to boost the yield of the desired compound.

Overall, these ideas have been relatively slow to take off within the chemistry community. Although molecular motions influence all chemical reactions, relatively few transformations have been explored with dynamics trajectory calculations. Therefore, it’s unclear how many chemical processes involve post–transition state bifurcations. But if chemists can study more reactions and learn more about the rules at play, they can apply their knowledge to produce new molecules more efficiently in the laboratory and to harness the complex chemistry catalyzed in nature.

Nature as Dynamic Director

Some biological reactions catalyzed by enzymes also involve the actions of post–transition state bifurcations, and these processes can have important implications for understanding activities in living organisms.

The plant Salvia miltiorrhiza (also called Chinese sage or tan shen, 丹參) has long been used in traditional Asian medicine to treat a range of conditions, from heart disease to hepatitis. This plant produces compounds called tanshinones through a series of enzyme-catalyzed reactions. It’s unclear how these molecules benefit the plants that produce them, but recently researchers have been testing some tanshinones in the clinic against heart disease and cancer. A single enzyme controls the reaction network that establishes the carbon core of these natural products. This network has the greatest number of linked post–transition state bifurcations yet discovered (a small subset of these are pictured in the last figure at the bottom of this article).

We’re still at the beginning of understanding the intricate details of how the dynamics of molecules influence their trajectories.

• Facebook
• Twitter

The tanshinone precursor, called miltiradiene, is a complex natural product that includes multiple rings composed of carbon atoms. Like many reactions used by chemists and nature to produce these types of complex structures, enzyme-catalyzed miltiradiene formation involves a positively- charged intermediate known as a carbocation. Such intermediates can rearrange their structures, reshuffling the positions of carbon atoms, to form a variety of products.

The rearrangement reaction in the figure at the bottom of this article was predicted to involve multiple sequential post–transition state bifurcations, allowing a single transition state to produce many products.

But even though this reaction pathway can form many products, in our simulations of molecular wiggling and jiggling we observe only two products that form in substantial quantities—the two cations highlighted in green and blue in the figure. These results suggest that the initially formed carbocation intermediate has an inherent tendency to form products that result from simple motions, such as the rotation around a single bond between two carbon atoms, shown in red in the figure. In nature, an enzyme assists in the reaction to form miltiradiene, but our dynamics calculations did not include the enzyme. Therefore, the enzyme catalyst isn’t solely responsible for determining which product is formed: The substrate has an inherent dynamical bias that’s expressed during the reaction.

Full of Potential

These results point to ways of redefining the questions that biochemists ask about enzymes and their roles. Once they have information about post–transition state bifurcations and their implications, biochemists can ask how these factors contribute to reaction rates, product selectivity, and even the evolution of enzymes.

In our computational study of miltiradiene formation, we observed a plethora of previously unknown post–transition state bifurcations. At first glance, that result might seem to suggest that many different products should form in roughly equal amounts. But our dynamics trajectory calculations tell a different story: Instead, the enzyme needs to actively suppress the formation of only one possible byproduct. With that vital information, researchers can focus their work on understanding how the enzyme blocks that undesired path, rather than chasing down false assumptions.

We’re still at the beginning of understanding the intricate details of how the dynamics of molecules influence their chemical trajectories. It’s a complex challenge. In some cases, the dynamic effects can be dwarfed by other competing factors, and reaction models that neglect or de-emphasize molecular momentum can still be useful. But not checking whether dynamic effects contribute is dangerous.

We have the privilege of exploring this molecular mystery and uncovering a new fundamental understanding of how molecules react. If we know that a reaction includes a post–transition state bifurcation and understand how dynamic effects control the ways molecules navigate past it, we can design new experiments to increase selectivity for a desired product. Knowing how an enzyme modulates substrate dynamical tendencies can lead directly to the design of enzyme mutants that produce useful molecules not found in nature. With insights gained from exploring this relatively uncharted and complicated territory—the back side of the downhill ski run—comes opportunities to influence how new molecules are made.