BIOELECTRONICS Wiring - up ion channels

In his famous Neuromancer trilogy1–3, science-fiction author William Gibson equips his main characters with biochip implants that allow them to connect their brains to computer networks and ‘upload’ new knowledge or capabilities as needed. For instance, a hero of the story instantly becomes a jet pilot by simply loading the corresponding software into his brain. An interface between biological cells and transistors — in the extreme case between a human brain and a computer — has been the dream not only of science-fiction authors, but also of scientists for many years now. Information transfer between the biological and the electronic realms would not only be important for futuristic applications such as the one mentioned above, but also for prosthetics and biosensing in general. However, there is a significant problem with this: our brain works with ions, whereas computers use electrons to process information, and it is far from straightforward to marry these two worlds. Communication requires ‘translation’ of ions to electrons through electrochemical processes or capacitive coupling of ionic and electronic signals. Reporting in the Proceedings of the National Academy of Sciences, Nipun Misra and co-workers4 now show how this coupling might be achieved using lipid-coated silicon nanowires. Misra et al. covered an electrically contacted nanowire with a lipid bilayer — a biomembrane resembling the one that surrounds biological cells — and inserted protein pores into these membranes (Fig. 1). Cell membranes are densely packed with specialized proteins that are responsible for interaction and signal exchange with the cell’s environment. Some of these membrane proteins contain tiny channels through which water and ions can flow — often in a highly specific manner. These so-called ion channels can switch between a conducting and a non-conducting state depending on the binding of small molecules or the voltage across the membrane, a property called ‘gating’. For this reason, ion channels have been previously referred to as “life’s transistors”5. Misra and colleagues were able to incorporate two model ion channels into the lipid bilayer surrounding their nanowires — the well-studied peptide pores gramicidin A and alamethicin — and use them to detect pH changes in the solution. Without the bilayer, the solution pH directly affects the protonation state of the nanowires’ silicon dioxide surface, which in turn influences their conductance — similar to traditional ionsensitive field-effect transistors. The dense lipid bilayer, however, makes the nanowires unresponsive to pH. The pH sensitivity is regained only when protons can flow through the ion channels inserted into the bilayer. Importantly, Misra et al. showed that this effect can be ‘gated’. The gramicidin A pore can be blocked by calcium ions, and the researchers could verify that the pH response of the nanowires was indeed rendered calcium-sensitive by the ion channels. Alamethicin, on the other hand, is known to be a voltage-gated pore. When a bias was applied to the nanowire with respect to an electrode in solution, the pH sensitivity was shown to be voltagedependent as expected. As noted above, there have already been many attempts to couple whole cells or reconstituted ion channels in membranes to electronics. In fact, in a series of seminal papers, Peter Fromherz and his group were able to demonstrate bidirectional information transfer between neurons and transistors6,7. Here, the coupling between the weak neural signals and the transistors proved to be notoriously difficult, mainly because the large salt-solution-filled ‘cleft’ between cells and chip prevented their direct interaction by electrical polarization. Recent advances towards cell-based biosensors were made using genetically modified cells overexpressing a serotonin neuroreceptor channel8. The first use of reconstituted — that is, cell-free — ion channels as biosensors had already been shown in 1997 by Cornell et al.9, but this did not result in a large burst of research activity at that time. Today, many researchers believe that chemically synthesized nanowires could revolutionize the field. Their small dimensions perpendicular to the direction of electrical transport make their conductance exceedingly sensitive to the nanowires’ surface, and therefore chemically modified nanowires seem to be ideally suited as components for biosensors10,11. In the case of the work reported by Misra et al., an extra advantage is the dense and smooth coverage of the lipid bilayer over the nanowires, resulting in an only 4-nm-thick water-filled gap between membrane and wire. Owing to the high BIOELECTRONICS