Recurring Themes in the Study of ECF s Factors − Rakyat Biologi

As more ECF s factors have been investigated, several common regulatory features have emerged. While none of these features is universal, in most cases ECF s factors coordinate transcriptional responses to extracellular signals, the signal transduction pathway often involves a membrane-localized anti-s factor, and ECF s factors often positively regulate their own synthesis. In addition, numerous examples have now been documented of overlapping promoter recognition and potential functional redundancy.

As defined originally for E. coli s^E and the closely related P. aeruginosa s^E protein, ECF s factors are often cotranscribed with one or more negative regulators (Hughes and Mathee, 1998). The immediately adjacent gene often encodes an anti-s factor, typically membrane-localized, that binds and stoichiometrically inactivates the ECF s factor. The transmembrane localization of the anti-s factor allows the cell to control the activity of an intracellular transcription factor in response to signals present outside the cytoplasmic membrane. Examples include the induction of the E. coli s^E regulon in response to unfolded proteins in the periplasm (Ravio and Silhavy, 2001) and the induction of several iron transport systems in response to the presence of the cognate ferri-siderophore complex (Braun, 1997). In this respect, ECF s factors together with their trans-membrane anti-s factors are functionally analogous to classical two component regulatory systems in which a transmembrane sensory protein (a histidine protein kinase) controls the activity of a cytoplasmic response regulator.

In general, ECF s factors are often quite divergent in sequence not only from other s factors, but from each other. Sequence analysis alone is sufficient to predict that a protein functions as an ECF s factor, but only in rare cases does it allow one to predict function (hopefully this will change as more and more ECF s factors become characterized). For example the role of M. tuberculosis s^H in controlling a disulfide stress regulon was a reasonable hypothesis based on the close similarity to s^R (Manganelli et al. 2001a). Our ability to make functional inferences based on the sequences of known and putative anti-s factors is even more limited. One of the best studied anti-s factors is RsrA, the redox-sensing regulator of S. coelicolor s^R. This protein contains a bound zinc ion and is the prototype for an emerging family of Zn-containing anti-s factors (The ZAS sub-family; Paget et al., 2001a). However, many anti-s factors have sequences suggestive of a metal-binding site and many of these regulators do not respond to oxidation. Therefore, the mere presence of a ZAS-type anti-s factor linked to a s factor gene is not a strong predictor of function.

A second recurring feature of ECF s factors is positive autoregulation: most genes encoding ECF s factors are preceded by a promoter recognized by the corresponding s. This presumably serves to amplify the signal produced by release of active s from the inactive s-anti-s complex. Despite their limited sequence identity, promoters recognized by ECF sub-family s factors often share characteristic sequence features, including the common occurrence of an “AAC” motif in the –35 consensus region. Inspection of the DNA sequence upstream of ECF s factor genes for possible autoregulatory sites can provide a preliminary indication of sequence selectivity for newly identified s factors of this class.

A third emerging theme in the analysis of ECF s factor regulons is regulon overlap. Particularly in those organisms containing many ECF s factors, some promoter sites appear to function as targets for regulation by more than one ECF s factor. The overlap between the B. subtilis s^X and s^W regulons appears to be due, at least in part, to overlapping recognition properties in the -10 consensus element (Qiu and Helmann, 2001). There is also limited overlap between the M. tuberculosis s^E and s^H regulons: both s factors recognize the same promoter for the sigB gene (Manganelli et al., 2001a). Despite the presence of regulon overlap, it is often possible to observe phenotypic consequences of mutations in ECF s factor genes. This lack of functional redundancy may be due, at least in part, to the fact that so many regulons are likely to be silent under normal growth conditions. In other words, the potential functional redundancies among ECF s factors is masked by the presence of anti-s factors that prevent the expression of many of these regulons unless appropriate inducing conditions are present.

Finally, we can consider prospects and strategies for understanding the roles of ECF s factors in the post-genomic era. Clearly, transcriptional profiling will continue to play an important role in defining ECF s factor regulons, particularly if conditions are employed that activate the regulon in question. In the absence of detailed knowledge of the physiological inducers, induction can be artificially elicited by either deletion or repression of the cognate anti-s or by overexpression of the s factor itself.

A complementary approach is provided by the computer-based identification of candidate target genes using the "consensus-search" method. This has worked quite well in those cases where the recognition sequence seems quite highly conserved (e.g. B. subtilis s^W and S. coelicolor s^R) but would not work very well if the consensus is not well defined (e.g. E. coli s^E). Analysis of the s^W regulon provides an instructive example of the bioinformatics of promoter recognition. In the whole genome there are exactly 27 perfect matches to the P_W autoregulatory sequence (Huang et al., 1999), TGAAAC N₁₆ CGTA, and 16 of these sites (all of those positioned upstream of reading frames) function as promoters (Huang et al., 1999). What about the other 11 sites? Most of these are in the middle of transcription units and/or are inappropriately oriented to serve an obvious role as promoters. It seems likely that many of these sites are “false positives”. How does the s^W holoenzyme distinguish the functional promoters from these other sites? We hypothesize that sequences in addition to the –35 and –10 elements are also important in promoter recognition. Candidates for such discriminatory elements include extensions of the –35 or –10 sequence elements (e.g. s^W promoters frequently have a T-rich segment adjacent to the -35 element and a -10 element of CGTAta; Table 7) and the upstream promoter (UP element) region (Huang et al., 1999). It is likely that similar considerations will pertain to other attempts to identify target genes using consensus-search methods. Clearly, we are still just beginning in our attempts to use computer-assisted analyses to extract biological information from genome sequences.