It has been thirteen years since Prasanna "AP" de Silva and his colleagues at Queen's University Belfast published their first paper in the international science journal Nature, outlining how they hoped to convert small molecules into the kind of logical units that could carry out computations. Now, the team has developed a logical technology that allows them to generate millions of unique tags to track sub-microscopic objects, such as polymer beads in combinatorial chemistry or biological cells in medical diagnostics without damaging them, for the first time. The tags can be used in parallel with a simple 'wash & watch' protocol for identifying each unique bead or cell.

"Now, we've truly hit the teenage coming-of-age in our field of molecular logic and computation," de Silva told Reactive Reports, "This is the first application of molecular logic that conventional semiconductor logic and computers cannot do—the identification of small objects, of nano or micrometer size such as polymer beads, biological cells and, hopefully, even large molecules such as proteins, in a large population." Adding an ID tag to a nano/micrometric object has until now been an unwieldy process, yet the solution could give chemists in drug discovery, catalyst design, and nanomaterials science the key to the door. "What we have devised is molecular computational identification (MCID)," adds de Silva.

The first application of MCID will probably be in combinatorial chemistry. Combichem uses a set of chemical building blocks to synthesize a vast library of new molecules by building them up in all possible combinations of the building blocks. A solution of building block one attached to millions of microscopic beads is split in two. Building block two is reacted with the first batch and building block three with the second. Then these two batches are recombined and split again. Building block two is added to one and building block three to the other batch and so on. Any number of reaction steps can be carried out in this split and mix approach and any number of building blocks can be added. Each new molecule is slightly different from the previous one but knowing which molecules are on which beads is difficult to find out, without complicated analytical chemistry.

When combinatorial chemistry came to the fore as something of an inexact science in the 1990s, chemists hoped that the technology would allow them to create vast molecular libraries from which useful individual molecules might be plucked like an interesting book from a shelf. Unfortunately, the tiny beads used to support each molecule in the library are all but impossible to label and so picking out the right molecule became a little more involved than a quick pluck. "MCID could now offer a solution to the combichem blockage," he told us, "it could also find application in diagnostics by allowing cells rather than combichem beads to be tagged."

Logic on beads. The tube with the glowing beads on the left contains acid. The less bright tube on the right contains alkali.

So, how does de Silva's MCID system work?

The team use a series of fluorescent dyes that only glow under certain chemical conditions, such as below a certain pH or when a metal ion is present. By adding a different fluorescent molecule to each batch at each step of the split and mix method, not only will a new range of molecules be built from the chemical building blocks, but a combinatorial tag will be added to each batch of beads so that the new molecular library is fully and uniquely labeled. Reading the library tag on each set of beads involves adding different ions one after the other to make the tags glow and reading off their "colors" with a spectrometer, so that specific molecules can be separated out and their properties tested.

De Silva points out that he and his colleagues have numerous fluorescent molecules that respond to different chemical environments by glowing only when the conditions follow a logical rule—NOT, YES, AND, and PASS, for instance. This means that they can produce thousands if not millions of combinations of tags for even the biggest combinatorial chemistry experiment.

Since the tagging method is general for many objects in the size domain of nano/micrometers (provided that the tag can be attached to the object, and the object subjected to wetting or washing), we can imagine its use in drug discovery environments where polymer beads are used. Laboratories engaged in discovery of nanoparticulate materials with desired properties may benefit similarly. Other fascinating 'objects' worthy of tagging are cells (live or dead) and, even, large molecules (polymers such as proteins and polynucleotides).

Combinatorial chemistry expert Paul Bartlett of the University of California at Berkeley sees de Silva's approach as "conceptually interesting". However, he is not so sure of how easy it will be to implement. "I fear when the crispness of his logic circuitry comes up against the messiness of chemistry and biology, it won't prove to be practical," he told us, "For example, the fluorescent behavior of his tags are likely to be modulated, and non-uniformly, by the different compounds synthesized on the bead, for example, in a combichem application."

Bartlett also suggests that combinatorial chemistry is not itself hamstrung by an identity crisis of the kind that might be solved by MCID. "Library size is limited more by practicalities of design, synthesis, and screening than it is in compound identification," he adds. It will be interesting to see the logical conclusion of this research.

"The messiness of chemistry and biology are not in doubt," de Silva responds, "That's why it's useful to have flexibility within our method so that it can be tailored for a given task. For instance, the non-uniform modulation (quenching) of fluorescence by neighboring compounds can be tackled by choosing fluorophores with care." In addition, a degree of modulation of the fluorescence signal is tolerable within the error since the researchers are looking for digital signals. "Indeed, for smaller library sizes, we can stay with binary digital signals. As we know from the electronics world, digital signals are more robust towards modulation/fluctuation than a simple analogue measurement."

The messiness of biology alluded to by Bartlett was widely considered to be a major obstacle to the use of fluorescence-based molecular sensors until the early 1980s. Now, fluorescence sensors are commonplace in intracellular research. The ratioing of two signals proved to be a good way to reduce the variability introduced by local fluctuations (including quenching). "Our MCID method also uses the ratioing of two signals (in acid and alkali for instance) which adds to its robustness."

Tony Czarnik of the University of Nevada, Reno, agrees that the MCID technology may have potential. "AP's method has not been published previously and could, indeed, lead to the encoding of large numbers of beads (or other particles)," he told Reactive Reports.

"MCID is in a position to examine various populations of objects provided they are taggable, wettable, and watchable," adds de Silva, "Each of these situations will throw up its own practicalities which will need to be discussed with practitioners in that field so that problems can be overcome. Combichem is no different."

Nature Materials, 2006, online; http://dx.doi.org/10.1038/nmat1733

http://www.ch.qub.ac.uk/staff/desilva/index.html

http://www.chemsoc.org/pdf/LearnNet/rsc/combi.pdf