Newsletter - Edition 28

Sensors and Surfaces

Oxford Innovation Society Lecture - March 1999
Professor Allen Hill, New Chemistry Laboratory

Sensors

At the beginning of the eighties, we were seeking uses for the method we had discovered just a few years earlier which allowed one to obtain the rapid electrochemistry of some important proteins. We found that these proteins, such as cytochrome c or the copper protein, azurin, could act as excellent mediators of electron transfer to many interesting enzymes but, of course, these were expensive and, before genetic engineering became such a ubiquitous technique, difficult to 'tune' to the exact requirement of the enzyme. It struck us that the simple organometallic compound, ferrocene, had many of the features exhibited by, say, cytochrome c but our organometallic colleagues had not been idle over the previous decades and there were literally thousands of ferrocenes with different characteristics, foremost amongst which was a different redox potential implying that they could be matched to a wide range of enzymes. They indeed proved to be excellent mediators to a wide range of enzymes, of which most interest was aroused by glucose oxidase. We obtained¹ a good, linear response to glucose over the range appropriate to diabetic patients.

In those days, there was little help for academics who sought to exploit their discoveries - unlike now - and there followed a frustrating and dispiriting few months. Eventually, we were given a little support for the work whilst others made attempts to raise sufficient capital to take the system through to manufacture. Though it was in our mind to make a device capable of use in vivo, the company, now called MediSense, decided to make a home-use device. The premises were opened in Abingdon in 1984 and the device was launched on to the market in 1989. One of the key features of the method was that the system could be used in whole blood and therefore no associated steps were called for, which is useful for diabetics. That has proved to be a key feature of the system and has, no doubt, accounted for its success. The company had much help from Dr. David Mathews, who now heads the Diabetic Clinic at the John Radcliffe Hospital in Oxford. MediSense, Inc., is now a division of Abbott Laboratories and may make a billion of the 'strips', used in the sensor, in 1999.

What has happened since then to this 'Sixth Sense'? Though we showed that the method was suitable for use in say, cholesterol sensing, and there has been an enormous effort all over the world which has given rise to a vast range of methods suitable for a host of analytes, the field is dominated, to the extent of 95%, by the glucose market. It occurred to us sometime ago that, just as protein electrochemistry had given rise to the glucose sensor, so the establishment in 1989 of direct enzyme electrochemistry², i.e., without the use of mediators, could allow the development of a whole new class of sensors. What advantage could they have over the sensors that incorporated mediators? The most important one was that, since they do not require a mediator between enzyme and electrode, they behave independently of one another and thus one can easily envisage a multi-electrode that contains five, ten ……, different enzyme configurations responsive to different analytes. Of course, one isn't envisaging each electrode being of the same size as that incorporated in conventional sensors. We, and many others, have investigated the properties of microelectrodes. With Professor Peter Dobson, we have studied³ many such configurations and, provided the distance between each microelectrode is of the order of ~50mm, a diameter of ~3mm would be sufficient. Sensors suitable for environmental, medical, home-use can be envisaged: it remains to be seen whether such configurations will be produced in the first decade of the new millennium.

Surfaces

All electrochemical phenomena take place at, on or nearby an electrode surface yet there is very little information about the molecular arrangement thereon. Of course, there have been many studies in which an interpretation of the nature of the electrode surface has been derived from macroscopic properties but few in which detailed molecular images have emerged. One cannot fail to be inspired and challenged by the wealth of information revealed over the past ten years or so by Scanning Probe Microscopy: whether one seeks information on the atomic structure of surfaces or the movement of atoms and molecules to order, these have been revealed in marvellous detail provided one requires such information about systems studied in vacuum. There were attempts to study the structures of biological molecules and systems fairly early on and, whilst useful information was obtained, there were two principal concerns: were the structures significantly altered as a consequence of these studies in vacuo or in air, as opposed to water? What was the consequence of adsorption of the biomolecules on the surface: were they static or capable of movement? In recent work, we attempted to address both these issues as the study of proteins and enzymes on electrode surfaces is at the heart of bioelectrochemistry.

We first studied⁴ a gold surface, modified, by self-assembly, with a compound that facilitated the electrochemistry of redox proteins. Of course, a single crystal face of gold had to be used to ensure that it was sufficiently flat for detailed examination. Figure 1 shows one of the arrangements, determined by Scanning Tunnelling Microscopy (STM), of the molecule, 2-thiopyrimidine adsorbed on to Au(III): individual molecules can be seen though the atomic structure is not observed. [STM reveals the tunnelling pathway between tip and surface and thus will not necessarily reveal the atomic details of an adsorbed molecule.] To increase the chances of restricting the movement of proteins on the surface, we made use of the advantages that the techniques of genetic engineering can bring to this work, viz., by studying a redox protein that had, on its surface, a cysteine residue to aid attachment to the gold surface. The STM of the copper-containing protein, azurin, containing the surface cysteine, was studied⁵ under a variety of conditions but the most intriguing observations came (Figure 2) from the examination of the structure of a single molecule: the STM images seem to reveal structure within the molecule. There had been some work suggesting that, the heavier the atom, there would be an enhancement of the tunnelling current. One protein provided an excellent example to test this idea: metallothionein that contains seven zinc ions. Its structure was already known, by diffraction studies and by NMR spectroscopy and so an appropriate comparison was possible. The STM of zinc7-metallothionein was obtained⁶ and showed (Figure 3) essentially the same features as were apparent by the other methods: the size and shape of the molecule were very similar, the two domains were apparent and, most importantly, the tunnelling current was enhanced in the positions of the seven zinc ions and their attendant cysteine ligands. This molecule was adsorbed on to the gold: since no genetically derived forms exist, at the moment, it may adhere via interactions with the surface methionine though there is evidence that, as has long been expected for most proteins on surfaces, it is somewhat 'flattened' on adsorption.

Another major variant of SPM methods is Atomic Force Microscopy (AFM). Whereas STM depends on the surface being a conductor, AFM makes no such demands and is sensitive to the different atomic and molecular interactions between the tip and the surface. With Professor T. Powell, we have studied⁷ the features revealed by the AFM of single cardiac cells Figure 4). This has proved most encouraging because the structural features are revealed below the outer membrane and are strikingly similar to those apparent in images obtained from Transmission Electron Microscopy (TEM).

My colleague, Professor C.M. Dobson, has described in a previous Isis Lecture, exciting work concerning the formation on amyloid fibrils from otherwise innocent proteins and enzymes. As a result, it seemed reasonable to attempt their investigation by AFM. A sequence of images have been obtained⁸ that illustrate (Figure 5) in detail the formation of structures, from proto-fibrils to plaques formed from a the SH3 domain of phosphatidyl inositol-3'-kinase It remains to be seen how these images compared with those produced in an environment more like that in the various organs of the body that are affected in, say, Alzheimer's disease.

Conclusions

Electrochemistry is undergoing its greater change for the past few decades in the sense that, for the first time, one is able to gain information on the true molecular nature of redox active adsorbates. Not only may this aid our understanding of electrochemical phenomena but may eventually lead to the control of the disposition of such adsorbates and the relationship that has to reactions that ensue, even, perhaps, those of analytical interest.

TEM has been of enormous importance in revealing atomic and molecular information on a wide range of surfaces. It has the disadvantage that it has to be used in vacuo but the related method, cryo-EM is capable to revealing knowledge of biological materials in biologically-relevant solutions at low temperature. AFM offers a complementary method of obtaining a similar understanding of a wide range of materials of biological and medical relevance. Indeed, the subject of SPM, applied to such systems has only begun: there continue to be a panoply of SPM methods capable of yielding structural and kinetic information provided that the entity adsorbs on to a surface. I am reminded of the first experiments on three-dimensional NMR spectroscopy of water present in some living systems: who would have thought that Magnetic Resonance Imaging would, some thirty years later, be of widespread importance. Perhaps, SPM, in some form or other, will bear similar fruit.

Acknowledgements. I would like to thank my colleagues, particularly Professors G. Canters, P.J. Dobson, C.M. Dobson, T. Powell, K. Uosaki and B.L. Vallee and Dr. L-L. Wong and their associates. My group has been most helpful, especially Dr. J.J. Davis. I am grateful for the support of the EPSRC and British Gas and, especially MediSense Inc., and Abbott Laboratories.

References

A.E.G. Cass, G. Davis, G.D. Francis, H.A.O. Hill, W.J. Aston, I.J. Higgins, E.V. Plotkin, L.D.L. Scott and A.P.F. Turner, "Ferrocene-mediated Enzyme Electrode for Amperometric Determination of Glucose", Anal. Chem. 1984, 56, 667-671.
L.H. Guo, H.A.O. Hill, D.J. Hopper, G.A. Lawrence and G.S. Sanghera, "Direct Un-mediated Electrochemistry of the Enzyme, p-Cresolmethylhydroxylase", J. Electroanal. Chem., 1989, 266, 379-390.
M.G. Boutelle, H.A.O. Hill, M. Berners, P.J. Dobson and P. Leigh, "New Technologies for Amperometric Biosensors", J. Mol. Recogn., 1996, 9, 664-671.
J.J. Davis, H.A.O. Hill, R. Yamada, H. Naohara and K. Uosaki, "Scanning Tunnelling Microscopy Study of the Self Assembly of 2-mercaptopyrimidine and 4,6-dimethyl-2-mercaptopyrimidine on Au(111)", J. Chem. Soc., Farad., 1998, 94, 1315-1319.
J.J. Davis, C.M. Halliwell, H.A.O. Hill, G.W. Canters, M.C. van Amsterdam and M. Ph. Verbeet, "Protein Adsorption at a Gold Electrode Studied by in situ Scanning Tunnelling Microscopy", New Journal of Chemistry, 1998, 1, 1119-1124.
J.J. Davis, H.A.O. Hill, A. Kurz, C. Jacob, W. Maret and B.L. Vallee, "A Scanning Tunnelling Microscopy Study of Rabbit Metallothionein" PhysChemComm. 1998, 1 20.
J.J. Davis, H.A.O. Hill and T. Powell, to be submitted.
J.J. Davis, H.A.O. Hill, A. Chamberlain and C.M. Dobson, to be submitted.