ELECTRON SPECTROSCOPIC IMAGING OF ORGANIC-COMPOUNDS USING PC-BASED ENERGY SEQUENCE IMAGING SOFTWARE

We have designed, characterised and tested Windows based software packages for rapid elemental microanalysis using a sequence of energy filtered transmission electron microscopic images which combines electron spectroscopic imaging with electron energy loss spectroscopy (Image-EELS). Image sequences containing up to 45 images of 512 x 512 pixels were recorded rapidly over an appropriate range of the energy spectrum for an analysis of boron, nitrogen, titanium and fluorine. Energy loss information was extracted from local variations in the energy loss spectrum with good energy resolution (5 eV) and a spatial resolution of 2 nm for the determination of the relative concentration of elements within a region of interest by grey level integration in consecutive images of an energy loss sequence, using the least mean squares method to calculate the parameters for an A . e(-tau) background model. The methods were used to define the optimum collections conditions and practical limitations of the system for qualitative and semi-quantitative elemental analysis relevant to biological specimens. Qualitative elemental analysis of images from Lowicryl HM23- embedded ultra-thin section of a boric acid polymer showed that boron was not distributed evenly. Semi-quantitative elemental analysis of the fluorine-containing organic polymer Nafion suggested a linear relationship between fluorine energy loss intensity, image area and specimen thickness. These techniques were applied to an examination of the reaction between purified RNA and the novel titanium-containing anti-cancer agent MKT-4, showing that ratios calculated from the cumulative electron energy loss intensity increase within regions of interest as a function of grey level intensity for titanium and nitrogen are directly proportional to the incubation concentration of the drug with RNA. Expression of data in this way is shown to be a useful indicator of the relative drug distribution within specific intracellular compartments of human cancer cells incubated with MKT-4 in vitro.