Cross-talk between thyroid hormone and specific retinoid X receptor subtypes in yeast selectively regulates cognate ligand actions.

Thyroid (T3) hormone beta1 (TR) and 9-cis retinoic acid (9c-RA) retinoid X receptors (RXR) can form heterodimer complexes that bind to hormone response elements (HREs) in target genes to either activate or repress transcription. However, the action of each cognate ligand and the accessory cellular factors that can differentially regulate the transcriptional responses of a heterodimer-DNA complex are not well understood. Studies in most mammalian cell lines have demonstrated that 9c-RA cannot bind or transactivate TR/RXR-T3 response element (TRE) complexes. In contrast, when identical heterodimer complexes were coexpressed in the yeast (Saccharomyces cerevisiae) with single copy typical TREs [i.e., DR+4 (direct repeat), F2 (everted repeat), or PAL (inverted repeat) DNA response elements] we observed that i) unliganded TRbeta1 homodimers had constitutive action on F2 and PAL but not DR4 TREs; ii) TRbeta1 homodimer responsivity to T3 ligand was relatively weak (less than twofold) and was only demonstrable on F2 but not PAL or DR4-TREs, whereas TRbeta1 heterodimers responded to T3 when RXRgamma but not RXR alpha was the heterodimeric partner; iii) RXR responsivity to 9c-RA (three- to sixfold) could be demonstrated only on palindromic TREs that could be enhanced by TRbeta1 on all TREs; iv) T3 + 9c-RA ligands increased (additively or synergistically) transactivation when RXRgamma but not alpha heterodimerized with TRbeta1 on both typical as well as atypical (DR1, DR3, DR5, and F2M) TREs. Substitutions for wild-type TRbeta1 of C-terminus mutants deficient in dimerization with RXRs abrogated the anticipated single and dual cognate ligand-induced effects on TRbeta1/RXRgamma transactivation of DR4 TREs, whereas mutants with preserved dimerization function but impaired T3 transactivation regions could maintain an enhanced 9c-RA response but were devoid of the anticipated T3 and dual (T3 + 9c-RA) cognate ligand-induced effects. Thus, the ligand-inducible response of TR and RXR homodimers expressed in yeast are relatively weak but can be further enhanced by TRbeta1 cross-talk with specific RXR subtypes in the presence of both cognate ligands.

Identification of USF as the ubiquitous murine factor that binds to and stimulates transcription from the immunoglobulin lambda 2-chain promoter.

To study the specificity and identity of NF-lambda 2, a ubiquitous murine nuclear factor that interacts specifically with the promoter of the lambda 2-chain gene and stimulates its transcription, competition experiments were carried out using DNA fragments from various immunoglobulin regulatory elements. The results showed that a fragment containing the H-chain enhancer competed efficiently for the binding of NF-lambda 2. Dissection of the H-chain enhancer revealed that the microE3 motif contributed the competing ability. Additionally, a regulatory region found in the adenovirus major late promoter, which interacts with the human general transcription factor USF, competed very efficiently for binding to NF-lambda 2. This region contains a sequence, CACGTGAC, which is identical to a region within the NF-lambda 2 motif. The pattern of complexes formation using oligonucleotide probes corresponding to the NF-lambda 2 and USF motifs were identical, and they both differed from that displayed by the E3 probes. Antisera against different domains of USF also react specifically with NF-lambda 2 showing that this factor is antigenically related, if not identical, to USF. Furthermore, the activity of the lambda 2 promoter in an in vitro transcription assay was significantly reduced when the nuclear extract used was USF-depleted. Addition of exogenous USF to this extract restored the transcription activity. Therefore, we conclude that NF-lambda 2 is the murine homologue of USF.

Interaction of a nuclear protein with a palindromic sequence of the mouse immunoglobulin lambda 2-chain gene promoter is important for its transcription.

By using a gel mobility retardation assay, we detected the formation of three major complexes from the binding of nuclear proteins to the promoter of the immunoglobulin lambda 2-chain gene. Two of the complexes were generated by the presence of an unidentified nuclear factor(s) called herein NF-lambda 2. Although the sequences between lambda 2- and lambda 1-chain gene promoters are very similar, the lambda 1-chain promoter did not compete for the binding of NF-lambda 2 efficiently. The binding site of NF-lambda 2 was localized by DNase I footprinting to a 14-bp region which is about 30 bp upstream of the immunoglobulin octamer motif. This region, referred to as the NF-lambda 2 motif, is within an 18-bp region of twofold rotational symmetry. Experiments with oligomers containing either the NF-lambda 2 or the octamer motifs as competitors for binding and DNase I footprinting, showed that the third complex is the product of the simultaneous binding of an octamer-binding protein and NF-lambda 2. Changing the sequence of the NF-lambda 2 motif to that of the lambda 1-chain counterpart abolished the binding ability of NF-lambda 2. Concomitantly, the level of chloramphenicol acetyltransferase expression driven by the mutated lambda 2 promoter decreased by two- to fivefold when compared with that of the wild-type promoter. It is therefore concluded that the interaction of NF-lambda 2 with the NF-lambda 2 motif stimulates transcription of the lambda 2-chain gene.

Analysis of the expression of murine lambda genes transfected into immunocompetent cell lines.

Hybridoma cell lines were transfected with plasmids containing either a rearranged lambda 1 or a rearranged lambda 2 mouse gene. The levels of lambda-chains synthesized by these transfectants were very low or undetectable. Activation of the expression of the lambda 2 gene was achieved artificially by deleting a portion of the region upstream of the promoter. Analogous deletions in the fragment containing the lambda 1 gene did not result in gene activation suggesting that the upstream sequences of lambda 1 and lambda 2 genes have diverged enough to allow differential regulation of their expression. However, both genes were activated by insertion, at a position upstream of the promoter, of a fragment containing the K-chain gene enhancer. These results suggest that the complete set of sequence elements that mediate lambda gene activation during normal B-cell differentiation are not all contained in the fragments of genomic DNA cloned so far, and thus, at least some of them must be located at a considerable distance from the promoter.

Carbon dioxide effects on ethanol production, pyruvate decarboxylase, and alcohol dehydrogenase activities in anaerobic sweet potato roots.

The effect of varied anaerobic atmospheres on the metabolism of sweet potato (Ipomoea batatas [L.] Lam.) roots was studied. The internal gas atmospheres of storage roots changed rapidly when the roots were submerged under water. O(2) and N(2) gases disappeared quickly and were replaced by CO(2). There were no appreciable differences in gas composition among the four cultivars that were studied. Under different anaerobic conditions, ethanol concentration in the roots was highest in a CO(2) environment, followed by submergence and a N(2) environment in all the cultivars except one. A positive relationship was found between ethanol production and pyruvate decarboxylase activity from both 100% CO(2)-treated and 100% N(2)-treated roots. CO(2) atmospheres also resulted in higher pyruvate decarboxylase activity than did N(2) atmospheres. Concentrations of CO(2) were higher within anaerobic roots than those in the ambient anaerobic atmosphere. The level of pyruvate decarboxylase and ethanol in anaerobic roots was proportional to the ambient CO(2) concentration. The measurable activity of pyruvate decarboxylase that was present in the roots was about 100 times less than that of alcohol dehydrogenase. Considering these observations, it is suggested that the rate-limiting enzyme for ethanol biosynthesis in sweet potato storage roots under anoxia is likely to be pyruvate decarboxylase rather than alcohol dehydrogenase.