No abstract

We have used a series of human estrogen receptor (ER) mutants to evaluate the cell- and promoter-specific transcriptional activities of the TAF1 and TAF2 transactivation regions within the human ER. We show that the manifestation of TAF1 or TAF2 function depends strongly upon promoter context; on certain promoters, both the TAF1 and TAF2 activators are required for wild-type transcriptional activity, whereas on other promoters, the TAF1 and TAF2 activators function independently. Using these constructs, we show that the antagonist activity of the triphenylethylene-derived antiestrogens, e.g. tamoxifen, arises from their intrinsic inability to activate ER TAF2 function. However, on certain promoters, these antiestrogens efficiently activate gene transcription through ER. Consistent with this observation, the TAF2 function of the ER is not required on all promoters. In these TAF2-independent promoter contexts, TAF2 function may be provided by a separate transcription factor bound to the promoter. These data suggest that 1) TAF1 may be the major transcriptional activator of the ER; and 2) TAF2 functions as a transcriptional facilitator. On promoters where TAF2 function is provided independently of the ER, the TAF1 function of the ER can function independently of TAF2 activity, allowing triphenylethylene-derived antiestrogens to demonstrate partial agonist activity. These observations provide a possible molecular explanation for the tissue-specific partial agonist properties of tamoxifen and related triphenylethylene antiestrogens observed in vivo.

Signal transducers and activators of transcription (STAT proteins) bind to palindromic sequence elements related to interferon y (IFN-y) activation sites, which were first identified in the promoters of IFN-'yinducible genes. Although the sequences of the natural palindromic STAT-binding elements vary considerably, they conform to the general structure TT(N)5AA. We have systematically examined the effects of the spacing between the TT and AA core half sites on the binding of the STAT complexes activated by IFN-y, interleukin (IL) 6, granulocytemacrophage colony-stimulating factor, and IL-4. We show that (i) as suggested earlier, a core palindromic TT-AA motifwith a 5-bp spacing displays general STAT binding, (ii) a palindromic motif with a spacing of 4 bp selectively binds to complexes containing Stat3, and (iii) a motif with a 6-bp spacing selectively binds the STAT complexes activated by IL-4. We have examined natural elements in the promoters of cytokine-responsive genes that differ in half-site spacing and found that they display binding properties predicted from the synthetic binding sites. Furthermore, the observed differential selective binding characteristics for the most part correlate with the ability to mediate transcriptional activation of transfected test genes in response to the cytokines tested. Our results thus demonstrate that the specificity of STAT-directed transcription in response to particular cytokines or cytokine families depends in part on the spacing of half sites within the conserved response element sequence.

CSA successfully allows most severe steroid resistant UC patients to retain their colons, and provides time for "elective" colectomy in others, especially if 6MP/aza are also given. Careful monitoring for side effects, including PCP prophylaxis, should be part of the treatment protocol.

No abstract