Selection of targets based on M. tuberculosis biology
Identify the essential genes of M. tuberculosis by transposon mutagenesis, allelic
exchange, and a novel essentiality test system.
Genes that are essential for growth represent attractive drug
targets. We intend to test the essentiality of every gene of
M. tuberculosis using a combination of transposon mutagenesis and
allelic exchange. The strategy is simple. Make a mutation in every
gene of the M. tuberculosis genome and test if it is essential for
growth. We have developed a revolutionary technology -specialized
transduction - that has allowed our lab to disrupt 24 genes in
M. tuberculosis in the last 6 months. This represents three times
more allelic exchanges for M. tuberculosis than the number of
successful allelic exchanges published by the rest of the world
combined. We intend to extend this technology to test the
essentiality of all 4000 genes of M. tuberculosis. The knowledge that
a gene is essential for growth immediately implies that the gene is an
attractive drug target as its inhibition should lead to the death of
the tubercle bacillus.
Identify M. tuberculosis mutants that are
unable to grow in vivo (giv) in mice using signature
tagged mutagenesis
Rationale. In other propsals, we propose to study the molecular
mechanisms by which M. tuberculosis infects, grows, and persists
within the mouse. Specifically, we will take a genetic approach to
identify M. tuberculosis genes that are necessary for the organism to
grow and survive within the host. To identify mutants unable to
survive within the mouse, we have adapted the signature-tag
mutagenesis (STM) system, originally developed Dr. David Holden
(Hensel et al., 1995), to the mouse model of tuberculosis.
Target prioritization based on expectation of novel fold and family information.
The fold assignment methods of Fischer & Eisenberg (1997) and Mallick
et al. (2000) have been applied to the non-membrane-bound
protein-coding ORFs of the M. tuberculosis genome (see "Preliminary
Results"). For each genome-encoded protein, these methods detect the
known 3D fold with which it is most compatible, and supply a
confidence level for the prediction. Where no known fold seems
compatible, a confidence level that the protein is a "novel fold" is
estimated. The purpose of this work is to provide some guidance to
Consortium scientists in targeting of proteins for structural
studies. Final versions of be entered into the STOP TB Internal
Database (see below). Based on these results and on our priority
scheme, which will include novel fold and structure, we estimate that
at least 10% (40) of the structures we determine will contain novel
folds. Methods of fold assignment are being improved, by combining
the procedure of Mallick et al. (2000) with the procedure of Bowie et
al. (1991). Furthermore, although the present round of predictions has
excluded membrane proteins from the predictions, recent work by Bowie
(1999) suggests that fold assignment for membrane proteins may soon
become an effective procedure. We have carried out a similar analysis
of the probability that each of the protein structures we determine is
likely to be in a new structural family (Fischer, 1999). Based on
this analysis, we expect that about 50% (200) of the structures we
solve will be from a novel family of structure.
Membrane proteins.
We anticipate that approximately 20-25% of the proteins from M.
tuberculosis that we target based on our genetic and informatics
screens will be membrane proteins (Boyd, Schierle et al. 1998; Wallin
and von Heijne 1998). These will be identified by the presence of
transmembrane segments using the program MOMENT (Eisenberg, Weis &
Terwilliger, 1984) and the methods described below under ?Membrane
Protein Crystallization? will be applied to obtain structural
information.