76
D.C. Hill et al.
have been extracted in a range of solvents, or simply filtered to remove high
molecular weight material. In the latter case, samples will contain a range of
medium components, may be highly coloured and may have a low or high pH or
ionic strength. Microbial screens must therefore be able to detect active metabolites above a background of potential interference.
Secondly, they must be sensitive. Assuming that active metabolites are
produced in fermentations at a concentration of 1-10 jag ml- 1, have an average
molecular weight of 500 Da, and are diluted in the assay 20- to 100-fold, the
required detection limit of the screen is in the 20-200 nM range.
Thirdly, to enhance metabolite detection, the assay should be highly specific
for the molecular or cellular target. To aid the determination of selectivity, data
generated from the screening programme can be compared with that from
similar and unrelated assays before progressing samples to metabolite isolation
and structure elucidation.
Finally, the throughput of the microbial screen should be carefully considered before commencing. Most assays can be converted to high throughput
screens, but cell-based assays are inherently more complex and can take significantly more resources and time to operate. Miniaturisation of assay formats and
the use of robotics can have a considerable impact on resource management for
screening.
Effective management of the data generated from advanced screening programmes is critical for success. Microbial screening cannot progress without
complete integration of input from scientists from the disciplines of microbiology, biochemistry and natural products chemistry to facilitate effective analysis
of their data. Information from all three disciplines needs to be interrogated
thoroughly when progressing samples through to metabolite isolation. Microbiological information such as organism source, description and fermentation
data should be recorded in detail. New taxonomic techniques should be
introduced where possible and their outputs incorporated into the data analysis
process. In the screening group, use of advanced data analysis packages should
be exploited to identify true hits and determine selectivity. This should include
use of suitable controls to check data validity and assess assay performance
continuously. Finally, in natural products chemistry, the use of chromatographic techniques to separate and purify the active compounds, and sensitive
spectroscopic techniques to characterise them, should be coupled with the
searching of a range of spectral libraries and natural products databases to
recognise and eliminate known compounds. Valuable resources and expertise
can then be focused on the isolation of metabolites with potential as lead
templates for development into new drugs.
Development of a fully integrated bioinformatics system is highly desirable
to facilitate rapid detection and identification of lead compounds. Such a system
combines new methods, existing tools, hardware and developed networks, with
basic computer technology such that data from all disciplines is linked together
in an apparently seamless unit. This allows rapid and detailed analysis of data
from all disciplines prior to selection of hits for progression to and during
D.C. Hill et al.
have been extracted in a range of solvents, or simply filtered to remove high
molecular weight material. In the latter case, samples will contain a range of
medium components, may be highly coloured and may have a low or high pH or
ionic strength. Microbial screens must therefore be able to detect active metabolites above a background of potential interference.
Secondly, they must be sensitive. Assuming that active metabolites are
produced in fermentations at a concentration of 1-10 jag ml- 1, have an average
molecular weight of 500 Da, and are diluted in the assay 20- to 100-fold, the
required detection limit of the screen is in the 20-200 nM range.
Thirdly, to enhance metabolite detection, the assay should be highly specific
for the molecular or cellular target. To aid the determination of selectivity, data
generated from the screening programme can be compared with that from
similar and unrelated assays before progressing samples to metabolite isolation
and structure elucidation.
Finally, the throughput of the microbial screen should be carefully considered before commencing. Most assays can be converted to high throughput
screens, but cell-based assays are inherently more complex and can take significantly more resources and time to operate. Miniaturisation of assay formats and
the use of robotics can have a considerable impact on resource management for
screening.
Effective management of the data generated from advanced screening programmes is critical for success. Microbial screening cannot progress without
complete integration of input from scientists from the disciplines of microbiology, biochemistry and natural products chemistry to facilitate effective analysis
of their data. Information from all three disciplines needs to be interrogated
thoroughly when progressing samples through to metabolite isolation. Microbiological information such as organism source, description and fermentation
data should be recorded in detail. New taxonomic techniques should be
introduced where possible and their outputs incorporated into the data analysis
process. In the screening group, use of advanced data analysis packages should
be exploited to identify true hits and determine selectivity. This should include
use of suitable controls to check data validity and assess assay performance
continuously. Finally, in natural products chemistry, the use of chromatographic techniques to separate and purify the active compounds, and sensitive
spectroscopic techniques to characterise them, should be coupled with the
searching of a range of spectral libraries and natural products databases to
recognise and eliminate known compounds. Valuable resources and expertise
can then be focused on the isolation of metabolites with potential as lead
templates for development into new drugs.
Development of a fully integrated bioinformatics system is highly desirable
to facilitate rapid detection and identification of lead compounds. Such a system
combines new methods, existing tools, hardware and developed networks, with
basic computer technology such that data from all disciplines is linked together
in an apparently seamless unit. This allows rapid and detailed analysis of data
from all disciplines prior to selection of hits for progression to and during