Evidence that GCD6 and GCD7, translational regulators of GCN4, are subunits of the guanine nucleotide exchange factor for eIF-2 in Saccharomyces cerevisiae.

Starvation of the yeast Saccharomyces cerevisiae for an amino acid signals increased translation of GCN4, a transcriptional activator of amino acid biosynthetic genes. We have isolated and characterized the GCD6 and GCD7 genes and shown that their products are required to repress GCN4 translation under nonstarvation conditions. We find that both GCD6 and GCD7 show sequence similarities to components of a high-molecular-weight complex (the GCD complex) that appears to be the yeast equivalent of translation initiation factor 2B (eIF-2B), which catalyzes GDP-GTP exchange on eIF-2. Furthermore, we show that GCD6 is 30% identical to the largest subunit of eIF-2B isolated from rabbit reticulocytes. Deletion of either GCD6 or GCD7 is lethal, and nonlethal mutations in these genes increase GCN4 translation in the same fashion described for defects in known subunits of eIF-2 or the GCD complex; derepression of GCN4 is dependent on short open reading frames in the GCN4 mRNA leader and occurs independently of eIF-2 alpha phosphorylation by protein kinase GCN2, which is normally required to stimulate GCN4 translation. Together, our results provide evidence that GCD6 and GCD7 are subunits of eIF-2B in S. cerevisiae and further implicate this GDP-GTP exchange factor in gene-specific translational control.

Cloning and characterization of cDNAs encoding the epsilon-subunit of eukaryotic initiation factor-2B from rabbit and human.

A rabbit reticulocyte lysate cDNA library was screened with a polyclonal antiserum directed against eukaryotic initiation factor eIF-2B (eIF-2B). A 2508 base pair cDNA (pA1) was isolated and determined to encode the epsilon-subunit of eIF-2B based on the immunoreactivity of the fusion protein expressed from the cDNA in Escherichia coli and the presence of two peptide sequences obtained from two V8 fragments of purified nonrecombinant eIF-2B epsilon in the deduced amino acid sequence of the cDNA. The open reading frame of the cDNA began with the third nucleotide of the cDNA with the first AUG codon at nucleotide 522. Mutational analysis of pA1 indicated that the cDNA did not code for full-length eIF-2B epsilon. Seven missing codons of the reading-frame and the 71 nucleotide 5' non-coding region of the eIF-2B epsilon mRNA were obtained by 5' RACE. A human eIF-2B epsilon cDNA fragment, which corresponded to a similar 2.3 kb fragment generated by digestion of the rabbit pA1 cDNA with EcoRI, was isolated from a human histiocytic lymphoma (U-937) cell cDNA library constructed in lambda gt10. The nucleotide and amino acid sequences were highly conserved between the rabbit and human cDNAs, showing approx. 90% sequence identity within the open reading frame. Northern and Western blot analyses of reticulocyte lysate and other rabbit tissue extracts indicated that the eIF-2B epsilon polypeptide has a similar apparent molecular weight in all tissues examined, and is coded for by a single approximately 2.8 kilobase mRNA species which is ubiquitously expressed.

1,25-Dihydroxyvitamin D3 increases nuclear vitamin D3 receptors by blocking ubiquitin/proteasome-mediated degradation in human skin.

1,25-Dihydroxyvitamin D3 (D3) exerts its effects by binding to and activating nuclear vitamin D3 receptors (VDRs) that regulate transcription of target genes. We have investigated regulation of VDR levels in human skin in vivo and in cultured human keratinocytes. Quantitative ligand-binding analysis revealed that human skin expressed approximately 220 VDRs per cell, which bound D3 with high affinity [(dissociation constant (Kd) = 0.22 nM]. In human skin nuclear extracts, VDR exclusively bound to DNA containing vitamin D3 response elements as heterodimers with retinoid X receptors. Topical application of D3 to human skin elevated VDR protein levels 2-fold, as measured by both ligand-binding and DNA-binding assays. In contrast, the D3 analog calcipotriene had no effect on VDR levels. Topical D3 had no effect on VDR mRNA, indicating that D3 either stimulated synthesis and/or inhibited degradation of VDRs. To investigate this latter possibility, recombinant VDRs were incubated with skin lysates in the presence or absence of D3. The presence of D3 substantially protected VDRs against degradation by human skin lysates. VDR degradation was inhibited by proteasome inhibitors, but not lysosome or serine protease inhibitors. In cultured keratinocytes, D3 or proteasome inhibitors increased VDR protein without affecting VDR mRNA levels. In cells, VDR was ubiquitinated and this ubiquitination was inhibited by D3. Proteasome inhibitors in combination with D3 enhanced VDR-mediated gene expression, as measured by induction of vitamin D3 24-hydroxylase mRNA in cultured keratinocytes. Taken together, our findings indicate that low VDR levels are maintained, in part, through ubiquitin/proteasome-mediated degradation and that low VDR levels limit D3 signaling. D3 exerts dual positive influences on its nuclear receptor, simultaneously stimulating VDR transactivation activity and retarding VDR degradation.

U1 snRNA is cleaved by RNase III and processed through an Sm site-dependent pathway.

Core snRNP proteins bind snRNA through the conserved Sm site, PuA(U)n>/=3GPu. While yeast U1 snRNA has three matches to the Sm consensus, the U1 3'-terminal Sm site was found to be both necessary and sufficient for U1 function. Mutation of this site inhibited pre-mRNA splicing, blocked cell division and resulted in the accumulation of two 3'-extended forms of the U1 snRNA. Cells which harbor the Sm site mutation lack mature U1 RNA (U1alpha) but have a minor polyadenylated species, U1gamma, and a prominent, non-polyadenylated species, U1beta. Metabolic depletion of the essential Sm core protein, Smd1p, also resulted in the increased accumulation of U1beta and U1gamma. In vitro, synthetic U1 precursors were cleaved by Rnt1p (RNase III) very near the U1beta 3'-end observed in vivo. We propose that U1beta is an Rnt1p-cleaved intermediate and that U1 maturation to the U1alpha form occurs through an Sm-sensitive step. Interestingly, both U1alpha and a second, much longer RNA, U1straightepsilon, were produced in an rnt1 mutant strain. These results suggest that yeast U1 snRNA processing may progress through Rnt1p-dependent and Rnt1p-independent pathways, both of which require a fun-ctional Sm site for final snRNA maturation.