Identification of a third autonomous activation domain within the human estrogen receptor.

Using a genetic selection system established in the yeast Saccharomyces cerevisiae, we have isolated, by random mutagenesis of the human estrogen receptor (ER), six mutants that display constitutive transcriptional activity. All of the mutants identified contained single base insertions or deletions leading to frameshift mutations, resulting in receptor truncations within the hormone-binding domain between amino acids (aa) 324-351. Interestingly, an ER mutant (aa 1-282) was transcriptionally inactive in yeast, suggesting that a domain important for transcriptional activity lies between aa 282 and 351 within human ER. Deletions representative of the mutants isolated in the yeast system were created in mammalian expression vectors and examined for transcriptional activity in animal cells to determine the physiological relevance of this domain. Receptors truncated at aa 282 were either weakly active or inactive; however, an ER deletion at aa 351 was approximately 50% as active as wild type ER (induced with estrogen). Furthermore, a chimeric receptor consisting of the DNA binding domain of GAL4 fused to aa 282-351 of the human ER was transcriptionally active on a GAL4 reporter. We conclude, therefore, that an autonomous activation domain (referred to as AF2a), functional in both yeast and mammalian cells, lies between aa 282-351 of the human ER.

Transcriptional activation and dimerization functions in the human vitamin D receptor.

The C-terminal domain of the human vitamin D receptor (hVDR) is essential for dimerization with retinoid X receptors and for transcriptional activation. To define the dimerization domain of the hVDR, a series of internal deletion mutants of the receptor were prepared beginning within the E domain and extending through the F domain to the C terminus. These mutant receptors were tested for dimerization and transcriptional activities by means of gel shift assay and beta-galactosidase assay, respectively, in a yeast system. The dimerization domain of the hVDR was localized to two separate but adjacent regions of the receptor molecule. In these experiments, the activation domain colocalized with dimerization. To more precisely delineate a relationship between these domains, region-specific random mutagenesis was carried out to detect mutants using error-prone PCR and a functional screen strategy employed using transformed yeast. Two classes of inactive receptors were identified: one in which both transcriptional activation and dimerization were compromised and a second in which only transcriptional activation was abolished. Most of the mutations responsible for these phenotypes were single. The studies suggest a separation between dimerization and transactivation domains. We reconstituted each of these hVDR mutants in a mammalian expression vector and evaluated them individually in COS-1 cells. All VDR mutants were transcriptionally active in this cellular background in response to 1,25-dihydroxyvitamin D3 although the potency of the hormone was reduced. The latter observation coincided with the observation that each mutant was compromised to some extent in binding affinity. These data clearly demonstrate the existence of an activation domain in hVDR that is separable from the domain involved in dimerization. Factors that couple hVDR to the general transcription apparatus in yeast through the activation domain in the hVDR, however, appear to be unrelated or dissimilar to those used in COS-1 cells.

Transactivation properties of retinoic acid and retinoid X receptors in mammalian cells and yeast. Correlation with hormone binding and effects of metabolism.

The binding affinities of 9-cis-retinoic acid (9-cis-RA) and all-trans-retinoic acid (t-RA) for retinoic acid receptors (RAR) alpha, beta, and gamma and for retinoid X receptors (RXR) alpha, beta, and gamma were determined using the recombinant receptor proteins and were compared with each hormone's ability to activate transcription through the receptors in mammalian and yeast cell systems. 9-cis-RA bound to both the RXRs (Kd values = 1.4-2.4 nM) and the RARs (Kd values = 0.2-0.8 nM). The ability of 9-cis-RA to bind to the RARs and RXRs correlated with its ability to produce similar transactivation profiles with these receptors in mammalian and yeast cell assays. t-RA bound to the RARs (Kd values = 0.2-0.4 nM) and activated transcription through the RARs in mammalian and yeast cells. In contrast, while t-RA did not bind to the RXRs, it did activate the RXRs, albeit less potently than 9-cis-RA, in mammalian cells. In yeast, however, the RXRs activated transcription only in the presence of 9-cis-RA, not with t-RA. While RAR gamma is activated in yeast by either t-RA or 9-cis-RA, the overall level of transcription was increased upon the addition of hormone-occupied RXR. Metabolism studies suggest that while there was no cell-dependent interconversion between t-RA and 9-cis-RA in yeast, there was cell-dependent conversion of 9-cis-RA to t-RA in mammalian cells [corrected].

The vitamin D-responsive element in the human osteocalcin gene. Association with a nuclear proto-oncogene enhancer.

A vitamin D-responsive element (VDRE) locus within the 5' region of the human osteocalcin gene promoter contains a steroid response-like half-site immediately proximal to a consensus site for the AP-1 nuclear oncogene family. In the studies described here, internal mutagenesis of the osteocalcin promoter coupled to functional assays reveal that the interaction of the vitamin D receptor is limited to the proximal region of the VDRE locus. Mutations within the distal AP-1 consensus site reduce the basal activity of the promoter but have little effect on vitamin D inducibility. The absolute level of promoter activity induced by hormone, however, is dramatically reduced in the absence of an intact AP-1 site suggesting a functional synergism between the receptor and AP-1-related proteins. In vitro receptor-DNA binding studies confirm the lack of requirement for the distal component in receptor binding. These results suggest that the osteocalcin VDRE is limited to 15 nucleotides closely juxtaposed to a distal functional AP-1 site. The close association of the two sites may lead to proto-oncogene and steroid receptor interactions that result in interesting functional consequences.

Sequence elements in the human osteocalcin gene confer basal activation and inducible response to hormonal vitamin D3.

Osteoblast-specific expression of the bone protein osteocalcin is controlled at the transcriptional level by the steroid hormone 1 alpha,25-dihydroxyvitamin D3. As this protein may represent a marker for bone activity in human disease, we examined the regulation of its expression at the molecular level by evaluating human osteocalcin gene promoter function. We describe regions within the promoter that contribute to basal expression of the gene in osteoblast-like cells in culture. Further, we define a 21-base-pair DNA element with the sequence 5'-GTGACTCACCGGGTGAACGGG-3', which acts in cis to mediate 1 alpha,25-dihydroxyvitamin D3 inducibility of the osteocalcin gene. This response element bears sequence similarity with other short DNA segments, particularly those for estrogen and thyroid hormone, which act together with their respective trans-acting receptors to modulate gene transcription.

Functional domains of the human vitamin D3 receptor regulate osteocalcin gene expression.

The human vitamin D receptor (VDR) has been cloned recently. Two cDNAs comprising the full-length VDR were spliced, cloned into a mammalian expression vector, and transiently expressed in COS-1 cells. The protein product exhibited properties consistent with that observed for receptor in human cells. A series of 5'- and 3'-deletions of the full-length VDR cDNA was prepared and evaluated. Native DNA binding was localized to a peptide fragment (residues 1-114) whose most prominent feature is the cysteine rich region proven to represent the DNA binding domain in other steroid receptors. Steroid binding-competence required synthesis of a peptide that initiated C-terminal to the DNA-binding domain at residue 114 and which contained the remaining 313 residues. To determine the location of elements within the receptor necessary for transcription, an osteocalcin gene promoter-chloramphenicol acetyltransferase reporter gene was cotransfected together with wild type or mutant VDR cDNAs and the latter's effect on chloramphenicol acetyltransferase activity was assessed. Cotransfection of wild type receptor alone resulted in efficient transcription of the reporter plasmid. However, synthesis of a peptide containing the DNA binding domain as well as 76 residues carboxy terminal to this region exhibited some degree of activity, albeit constitutive. These results suggest that the functional domains of the VDR are similar to that of other steroid receptors and that these domains participate in the transcriptional regulation of the human osteocalcin gene.

Human creatine kinase: isolation and sequence analysis of cDNA clones for the B subunit, development of subunit specific probes and determination of gene copy number.

cDNA clones for human B creatine kinase were isolated from human brain and placenta libraries. The entire coding and 3' untranslated regions, as well as 23 bp of the 5' untranslated region were sequenced. Complete sequence identity was found among the clones, with the exception of an area of heterogeneity among the 3' untranslated region of the brain and placenta clones. A 77.7% nucleotide sequence identity was found between the coding region of human B creatine kinase and our previously reported human M creatine kinase. In contrast, no homology was found in the 3' untranslated regions. Probes were constructed from the nonconserved 3' untranslated regions of human M and B creatine kinase and were shown to be highly specific. Southern transfers of total genomic DNA derived from human placenta and digested to completion with several restriction enzymes were probed with the MCK and BCK specific probes producing single hybridization bands. These results suggest that creatine kinase M and B are single copy genes in the human genome.

Isolation and sequence analysis of a full-length cDNA for human M creatine kinase.

A full length cDNA for human M creatine kinase has been isolated and sequenced. The cDNA contains 77 bp of 5' untranslated, 338 bp of 3' untranslated sequence and the entire coding region (1146 bp) for human M creatine kinase. The M creatine kinases from different species share considerable sequence homology within the coding region (77-91%) and in amino acid sequence (82-97%). Little or no sequence homology is observed in the 3' untranslated sequence of the mammalian M creatine kinases, although canine and human creatine kinase share overall 80% sequence homology in 5' untranslated sequence. A unique 8 bp sequence was identified in the 5' untranslated regions of mammalian M creatine kinase but is not present in B creatine kinase cDNA. The degree of sequence conservation observed implies an evolutionary constraint on M creatine kinase structure beyond that which would be expected for the maintenance of enzymatic function.