Search

mRNA turnover is an important regulatory component of gene expression and is significantly influenced by ribonucleoprotein (RNP) complexes which form on the mRNA. Studies of human alpha-globin mRNA stability have identified a specific RNP complex (alpha-complex) which forms on the 3' untranslated region (3'UTR) of the mRNA and appears to regulate the erythrocyte-specific accumulation of alpha-globin mRNA. One of the protein activities in this multiprotein complex is a poly(C)-binding activity which consists of two proteins, alphaCP1 and alphaCP2. Neither of these proteins, individually or as a pair, can bind the alpha-globin 3'UTR unless they are complexed with the remaining non-poly(C) binding proteins of the alpha-complex. With the yeast two-hybrid screen, a second alpha-complex protein was identified. This protein is a member of the previously identified A+U-rich (ARE) binding/degradation factor (AUF1) family of proteins, which are also known as the heterogeneous nuclear RNP (hnRNP) D proteins. We refer to these proteins as AUF1/hnRNP-D. Thus, a protein implicated in ARE-mediated mRNA decay is also an integral component of the mRNA stabilizing alpha-complex. The interaction of AUF1/hnRNP-D is more efficient with alphaCP1 relative to alphaCP2 both in vitro and in vivo, suggesting that the alpha-complex might be dynamic rather than a fixed complex. AUF1/hnRNP-D could, therefore, be a general mRNA turnover factor involved in both stabilization and decay of mRNA.

The immortalized rat calvarial bone cell line RCT-1 responds to treatment with retinoic acid (RA) by increased expression of osteoblast phenotype-related features, including the induction of liver/bone/kidney alkaline phosphatase (ALP) activity. ALP mRNA could not be demonstrated in unstimulated cells, but was first detected in cells treated for 6 h with 1 microM RA. Cycloheximide failed to block the RA induction of ALP mRNA, indicating that de novo protein synthesis was not a requirement for the RA effect and that the ALP gene may be a direct target for RA action. This was confirmed by nuclear run-on assays, which demonstrated a 2.5-fold increase in the abundance of ALP transcripts after 6 h of RA treatment. To determine whether the RA responsiveness was mediated by a specific segment of the ALP promoter, RCT-1 cells were transfected with a series of plasmids containing deletions of the 5'-flanking sequence of the human ALP gene fused to the chloramphenicol acetyl transferase (CAT) gene. CAT activity was measured in cells cultured in the presence of RA or vehicle. All but the smallest construct, which contained 44 basepairs up-stream of the initiation of transcription, were found to mediate a 2- to 3-fold increase in the expression of CAT activity in response to RA. Furthermore, when the region -108 to -45 of the human ALP gene was inserted into the expression vector pBLcat2, in a position immediately up-stream of the herpes simplex virus thymidine kinase promoter, the construct was found to mediate a 2-fold enhancement of CAT activity in response to RA. In gel retardation assays, a major band was present corresponding to the formation of a complex between the 32P-labeled probe containing the -108 to -45 sequence and proteins present in nuclear extracts of RCT-1 cells stimulated for 3 h with RA. These data suggest that the sequence of 64 basepairs (-108 to -45) 5' to the transcription start site is involved in the RA inducibility of the human ALP gene.

It has been proposed that the helix-loop-helix (HLH) protein Id serves as a general antagonist of cell differentiation by inhibiting bHLH (HLH with an adjacent stretch of basic amino acids) proteins specifically required for developmental programs (such as MyoD). We show here that ectopic expression of Id represses in vivo activity of the bHLH protein E2-5 (encoded by the E2A gene) and of both the immunoglobulin heavy-chain (IgH) and kappa-light-chain gene enhancers to which E2-5 binds. Id does not affect the activity of the bHLH-zip protein, TFE3, which also binds these enhancers. We examined a large panel of B-cell lines that represent different stages of lymphoid development and found only two that express Id mRNA. The cell lines Ba/F3 and LyD9 have been categorized previously as early B-lymphoid-cell progenitors. Unlike their more mature B-lymphoid-cell counterparts, Ba/F3 and LyD9 cells do not express I mu sterile transcripts, which are indicative of IgH enhancer activity. Moreover, Ba/F3-derived nuclear extracts lack E2-box-binding activity, indicating the absence of free bHLH proteins, and transfected Ba/F3 cells fail to support the activity of the IgH enhancer. Hence, expression of Id correlates inversely with bHLH protein activity and enhancer function in vivo. These results suggest that Id may play a role early in B-lymphoid-cell development to regulate transcription of the IgH locus.

Activity of the immunoglobulin heavy and kappa light chain gene enhancers depends on a complex interplay of ubiquitous and developmentally regulated proteins. Two complementary DNAs were isolated that encode proteins, denoted ITF-1 and ITF-2, that are expressed in a variety of cell types and bind the microE5/kappa 2 motif found in both heavy and kappa light chain enhancers. The complementary DNAs are the products of distinct genes, yet both ITF-1 and ITF-2 are structurally and functionally similar. The two proteins interact with one another through their putative helix-loop-helix motifs and each possesses a distinct domain that dictates transcription activation.

We have investigated the effect of polymerizing defined segments of the immunoglobulin heavy-chain enhancer on the activity of a single, linked transcription unit. Transient assays in lymphoid cells have led to the following observations. First, polymerizing the entire enhancer led to an increase in overall transcription. Second, polymerizing defined DNA segments revealed two distinct functional domains within the enhancer. Although each domain alone possessed only partial enhancer activity, greater than wild-type levels of activity could be obtained upon polymerization. One of these domains contains three regions thought to be involved in protein binding in vivo and in vitro (E motifs E1, E2, and E3). The other domain contains the fourth E motif (E4) and the conserved octanucleotide, ATTTGCAT. We have tested the functional importance of these motifs by determining the effect of mutating these elements singly or in combination in the context of the isolated domains. Although E2, E3, E4, and the octanucleotide are clearly important for enhancer function, mutation of the E1 motif did not appear to have an effect on enhancer activity in our assay. Transient assays in mouse L cells indicate that nonlymphoid cells are able to use a distinct subset of these motifs.