But a new method for identifying and classifying bacteria by imaging the radioisotope distribution has been developed by the Department of Reproductive Physiology at St. Bartholomew’s Hospital Medical College in London.
A team headed by Robert Silman was investigating the translation of mRNA labeled with 35S while researching the ACTH peptide hormones of the pituitary gland. They were identifying the translation products by polyacrylamide gel electrophoresis and autoradiography, when on one occasion they unknowingly laced their samples with bacteria. Many unexpected bands, resembling the bar codes on supermarket merchandise, appeared in the autoradiograph.
With help from the Department of Medical Microbiology at St. Bartholomew’s, the team learned that the band patterns were specific to individual strains of bacteria. Within a few weeks it became clear that the accidental contamination had spawned a technique capable of discriminating between organisms with a rare degree of sophistication.
Silman’s group then constructed a database for identifying bacteria through pattern recognition software, and the equipment was subsequently developed by Brian Pullan, who was at that time a professor of medical biophysics at Manchester University.
Now known as the AMBIS Radioanalytical Imaging System (AMBIS Systems Inc., San Diego, Calif.) the instrument produces an image 20 times faster than autoradiography. It is accurate over a wide range spanning 106 levels of activity. Any flat sample with dimensions of up to 20 X 20 cm can be analyzed and the system detects most of the radioisotopes commonly used in molecular biology laboratories.
The data are stored on an IBM PC/AT. The software analyzes the data and produces a graphic readout on the screen at the same time the image of the scan is displayed. it is possible to select the strip on the image to be analyzed, and print it using the printer included in the AMBIS system.
The instrument can be used for scanning any beta-emitting isotope except tritium. It can also be used for gamma emitters, but so far only 59Fe enzymes and 125I compounds have been examined satisfactorily. Scientists can examine samples from any biological source including molds, fungi, and algae, as well as enzyme assays.
Conventional autoradiography requires a densitometer to measure the optical density of the various bands, a method that is tedious and time-consuming with difficult data analysis. The dynamic range of X-ray film limits the procedure, whereas the radioanalytical imaging system not only automates the process, but measures the radiation directly by proportional counting.
The positional information of each band is also included in the data collection process so that this important information can be com pared quickly to a database. It is then much easier to store the data in a computer and subsequently analyze the information for image reconstruction.