Quantum Computing Spins Closer

The promise of quantum computing is that it will dramatically outshine traditional computers in tackling certain key problems: searching large databases, factoring large numbers, creating uncrackable codes and simulating the atomic structure of materials.

A quantum step in that direction, if you'll pardon the pun, has been taken by Stanford researchers who announced their success in a paper published in the journal Nature. Working in the Ginzton Laboratory, they've employed ultrafast lasers to set a new speed record for the time it takes to rotate the spin of an individual electron and confirm the spin's new position.

Why does that matter? Existing computers, from laptops to supercomputers, see data as bits of information. Each bit can be either a zero or a one. But a quantum bit can be both zero and one at the same time, a situation known as a superposition state. This allows quantum computers to act like a massively parallel computer in some circumstances, solving problems that are almost impossible for classic computers to handle.
Quantum computing can be accomplished using a property of electrons known as "spin." A single unit of quantum information is the qubit, and can be constructed from a single electron spin, which in this experiment was confined within a nano-sized semiconductor known as a quantum dot.

An electron spin may be described as up or down (a variation of the usual zero and one) and may be manipulated from one state to another. The faster these electrons can be switched, the more quickly numbers can be crunched in a quantum fashion, with its intrinsic advantages over traditional computing designs.
The qubit in the Stanford experiment was manipulated and measured about 100 times faster than with previous techniques, said one of the researchers, David Press, a graduate student in applied physics.
The experiments were conducted at a temperature of almost absolute zero, inside a strong magnetic field produced by a superconducting magnet. The researchers first hit the qubit with laser light of specific frequencies to define and measure the electron spin, all within a few nanoseconds. Then they rotated the spin with polarized light pulses in a few tens of picoseconds (a picosecond is one trillionth of a second). Finally, the spin state was read out with yet another optical pulse.

Similar experiments have been done before, but with radio-frequency pulses, which are slower than laser-light pulses. "The optics were quite tricky," Press said. The researchers had to find a single, specific photon emitted from the qubit in order confirm the spin state of the electron. That photon, however, was clouded in a sea of scattered photons from the lasers themselves.

"The big benefit is to make quantum computing faster," Press said. The experiment "pushed quantum dots up to speed with other qubit candidate systems to ultimately build a quantum computer."

Quantum computers are still years away. In the shorter term, Press said, researchers would like to build a system of tens or hundreds of qubits to simulate the operation of a larger quantum system.

The other authors of the Nature paper were Bingyang Zhang of the Ginzton Lab, and Thaddeus Ladd and Yoshihisa Yamamoto of the Ginzton Lab and the National Institute of Informatics in Tokyo.

Plants Engineered to Produce New Drugs

Humans have long taken advantage of the huge variety of medicinal compounds produced by plants. Now MIT chemists have found a new way to expand plants' pharmaceutical repertoire by genetically engineering them to produce unnatural variants of their usual products.

The researchers, led by Associate Professor Sarah O'Connor, have added bacterial genes to the periwinkle plant, enabling it to attach halogens such as chlorine or bromine to a class of compounds called alkaloids that the plant normally produces. Many alkaloids have pharmaceutical properties, and halogens, which are often added to antibiotics and other drugs, can make medicines more effective or last longer in the body.

The team's primary target, an alkaloid called vinblastine, is commonly used to treat cancers such as Hodgkin's lymphoma. O'Connor sees vinblastine and other drugs made by plants as scaffolds that she can modify in a variety of ways to enhance their effectiveness.

"We're trying to use plant biosynthetic mechanisms to easily make a whole range of different iterations of natural products," she said. "If you tweak the structure of natural products, very often you get different or improved biological and pharmacological activity."

O'Connor, graduate student Weerawat Runguphan and former postdoctoral associate Xudong Qu describe their engineered periwinkle plants in the Nov. 3 online edition of Nature. The research was funded by the National Institutes of Health and the American Cancer Society.

Engineering new genes into plants has been done before: In the 1990s, scientists developed corn that could produce an insecticide called Bt, which comes from a bacterial gene. However, O'Connor's approach, known as metabolic engineering, goes beyond simply adding a gene that codes for a novel protein. Metabolic engineers tinker with the series of reactions that the host organisms use to build new molecules, adding genes for new enzymes that reshape these natural synthetic pathways. This can lead to a huge variety of end products.

Most metabolic engineers use bacteria as their host organism, in part because their genes are easier to manipulate. O'Connor's work with plants makes her a rare exception. She doesn't believe one approach is better than the other, but one factor that drew her to engineer plants is that most plant synthetic pathways have not been completely revealed. "You can't reconstitute a whole plant pathway in bacteria unless you have all the genes," she said.

In previous experiments, O'Connor and her students induced periwinkle root cells to create novel compounds by feeding them slightly altered versions of their usual starting materials. In the new study, they engineered the cells to express genes that code for enzymes that attach chlorine or bromine to vinblastine precursors and other alkaloids.

The two new genes came from bacteria that naturally produce halogenated compounds. It's much more rare for plants to generate such compounds on their own, said O'Connor. It is also possible, though very difficult, to synthesize halogenated alkaloids in a laboratory.

To make alkaloids, plants first convert an amino acid called tryptophan into tryptamine. After that initial step, about a dozen more reactions are required, and the plants can produce hundreds of different final products. In the new genetically engineered plants, a bacterial enzyme called halogenase attaches a chlorine (or bromine) atom to tryptamine. That halogen stays on the molecule throughout the synthesis.

In future work, the researchers hope to engineer full periwinkle plants to produce the novel compounds. They are also working on improving the overall yield of the synthesis, which is about 15 fold lower than the plant's yield of naturally occurring alkaloids. One way to do that is to introduce the halogen further along in the process, said O'Connor.