Ribosome precursor particles in nucleoli.

Ribonucleoprotein (RNP) particles containing the precursors of ribosomal RNA were extracted from L cell nucleoli and analyzed under conditions comparable to those used in the characterization of cytoplasmic ribosomes. Using nucleoli from cells suitably labeled with (3)H-uridine, we detected three basic RNP components, sedimenting at approximately 62S, 78S, and 110S in sucrose gradients containing magnesium. A fourth particle, sedimenting at about 95S, appears to be a dimer of the 62S and 78S components. When centrifuged in gradients containing EDTA, the 62S, 78S, and 110S particles sediment at about 55S, 65S, and 80S, respectively. RNA was extracted from RNP particles which were prepared by two cycles of zonal centrifugation. The 62S particles yielded 32S RNA and a detectable amount of 28S RNA, the 78S structures, 32S RNA and possibly some 36S RNA, and the 110S particles, a mixture of 45S, 36S, and 32S RNA's. When cells were pulsed briefly and further incubated in the presence of actinomycin D, there was a gradual shift of radioactivity from heavier to lighter particles. This observation is consistent with the scheme of maturation: 110S --> 78S --> 62S. The principal buoyant densities in cesium chloride of the 110S, 78S, and 62S particles are 1.465, 1.490, and 1.545, respectively. These densities are all significantly lower than 1.570, which is characteristic of the mature large subunit of cytoplasmic ribosomes, suggesting that the precursor particles have a relatively higher ratio of protein to RNA, and that ribosome maturation involves, in addition to decrease in the size of the RNA molecules, a progressive decrease in the proportion of associated protein.

Aberrant intranucleolar maturation of ribosomal precursors in the absence of protein synthesis.

To help elucidate the role of protein in the maturation of ribosomal RNA in cultured L cells, we have studied the effects of cycloheximide upon the maturation process and upon the intranucleolar ribonucleoprotein particles containing the "preribosomal RNA's." Five parameters of these particles were analyzed: (a) extractability, (b) sedimentation characteristics in sucrose gradients, (c) RNA composition, (d) buoyant density in CsCl gradients, and (e) effects of increased ionic strength on the buoyant density. When protein synthesis is inhibited, the rate of conversion of the precursor 45S ribosomal RNA is rapidly diminished, falling to less than 30% of the control rate within 1 hr. Nevertheless, in terms of the first three parameters there is no difference between control and cycloheximide nucleolar particles. However, the cycloheximide particles have a lower and more heterogeneous buoyant density and a more variable response to increased ionic strength. The results imply that the protein composition of the cycloheximide particles is different from that of particles from control cells, and that the entire protein complement is not necessary for the first cleavages in the maturation process, although it is necessary for the normal rate of processing and for the eventual appearance of both 18S and 28S rRNA in mature ribosomes.

Hybridization properties of DNA sequences directing the synthesis of messenger RNA and heterogeneous nuclear RNA.

The relationship of the DNA sequences from which polyribosomal messenger RNA (mRNA) and heterogeneous nuclear RNA (NRNA) of mouse L cells are transcribed was investigated by means of hybridization kinetics and thermal denaturation of the hybrids. Hybridization was performed in formamide solutions at DNA excess. Under these conditions most of the hybridizing mRNA and NRNA react at values of D(o)t (DNA concentration multiplied by time) expected for RNA transcribed from the nonrepeated or rarely repeated fraction of the genome. However, a fraction of both mRNA and NRNA hybridize at values of D(o)t about 10,000 times lower, and therefore must be transcribed from highly redundant DNA sequences. The fraction of NRNA hybridizing to highly repeated sequences is about 1.7 times greater than the corresponding fraction of mRNA. The hybrids formed by the rapidly reacting fractions of both NRNA and mRNA melt over a narrow temperature range with a midpoint about 11 degrees C below that of native L cell DNA. This indicates that these hybrids consist of partially complementary sequences with approximately 11% mismatching of bases. Hybrids formed by the slowly reacting fraction of NRNA melt within 4 degrees -6 degrees C of native DNA, indicating very little, if any, mismatching of bases. Hybrids of the slowly reacting components of mRNA, formed under conditions of sufficiently low RNA input, have a high thermal stability, similar to that observed for hybrids of the slowly reacting NRNA component. However, when higher inputs of mRNA are used, hybrids are formed which have a strikingly lower thermal stability. This observation can be explained by assuming that there is sufficient similarity among the relatively rare DNA sequences coding for mRNA so that under hybridization conditions, in which these DNA sequences are not truly in excess, reversible hybrids exhibiting a considerable amount of mispairing are formed. The fact that a comparable phenomenon has not been observed for NRNA may mean that there is less similarity among the relatively rare DNA sequences coding for NRNA than there is among the rare sequences coding for mRNA.

Nuclear transcripts of mouse heavy chain immunoglobulin genes contain only the expressed class of C-region sequences.

In plasmacytoma cells producing IgG, IgA, or IgM immunoglobulin heavy chains, the large precursors of the heavy chain messenger RNA's contain nucleotide sequences that specify only the expressed class of constant region. This indicates that the switch from one class of heavy chain to another during B cell ontogeny does not occur by altered processing of a complex gene transcript.

The regulated production of mu m and mu s mRNA is dependent on the relative efficiencies of mu s poly(A) site usage and the c mu 4-to-M1 splice.

The relative abundance of the mRNAs encoding the membrane (mu m) and secreted (mu s) forms of immunoglobulin mu heavy chain is regulated during B-cell maturation by a change in the mode of RNA processing. Current models to explain this regulation involve either competition between cleavage-polyadenylation at the proximal (mu s) poly(A) site and cleavage-polyadenylation at the distal (mu m) poly(A) site [poly(A) site model] or competition between cleavage-polyadenylation at the mu s poly(A) site and splicing of the C mu 4 and M1 exons, which eliminates the mu s site (mu s site-splice model). To test certain predictions of these models and to determine whether there is a unique structural feature of the mu s poly(A) site that is essential for regulation, we constructed modified mu genes in which the mu s or mu m poly(A) site was replaced by other poly(A) sites and then studied the transient expression of these genes in cells representative of both early- and late-stage lymphocytes. Substitutions at the mu s site dramatically altered the relative usage of this site and caused corresponding reciprocal changes in the usage of the mu m site. Despite these changes, use of the proximal site was still usually higher in plasmacytomas than in pre-B cells, indicating that regulation does not depend on a unique feature of the mu s poly(A) site. Replacement of the distal (mu m) site had no detectable effect on the usage of the mu s site in either plasmacytomas or pre-B cells. These findings are inconsistent with the poly(A) site model. In addition, we noted that in a wide variety of organisms, the sequence at the 5' splice junction of the C mu 4-to-M1 intron is significantly different from the consensus 5' splice junction sequence and is therefore suboptimal with respect to its complementary base pairing with U1 small nuclear RNA. When we mutated this suboptimal sequence into the consensus sequence, the mu mRNA production in plasmacytoma cells was shifted from predominantly mu s to exclusively mu m. This result unequivocally demonstrated that splicing of the C mu 4-to-M1 exon is in competition with usage of the mu s poly(A) site. A key feature of this regulatory phenomenon appears to be the appropriately balanced efficiencies of these two processing reactions. Consistent with predictions of the mu s site-splice model, B cells were found to contain mu m precursor RNA that had undergone the C mu 4-to-M1 splice but had not yet been polyadenylated at the mu m site.

DNA-binding and chromatin localization properties of CHD1.

CHD1 is a novel DNA-binding protein that contains both a chromatin organization modifier (chromo) domain and a helicase/ATPase domain. We show here that CHD1 preferentially binds to relatively long A.T tracts in double-stranded DNA via minor-groove interactions. Several CHD1-binding sites were found in a well-characterized nuclear-matrix attachment region, which is located adjacent to the intronic enhancer of the kappa immunoglobulin gene. The DNA-binding activity of CHD1 was localized to a 229-amino-acid segment in the C-terminal portion of the protein, which contains sequence motifs that have previously been implicated in the minor-groove binding of other proteins. We also demonstrate that CHD1 is a constituent of bulk chromatin and that it can be extracted from nuclei with 0.6 M NaCl or with 2 mM EDTA after mild digestion with micrococcal nuclease. In contrast to another chromo-domain protein, HP1, CHD1 is not preferentially located in condensed centromeric heterochromatin, even though centromeric DNA is highly enriched in (A+T)-rich tracts. Most interestingly, CHD1 is released into the cytoplasm when cells enter mitosis and is reincorporated into chromatin during telophase-cytokinesis. These observations lend credence to the idea that CHD1, like other proteins with chromo or helicase/ATPase domains, plays an important role in the determination of chromatin architecture.

Characterization of the expressed gene and several processed pseudogenes for the mouse ribosomal protein L30 gene family.

Five cloned genes encoding the mouse ribosomal protein L30 were isolated from a recombinant DNA library and characterized by restriction mapping and nucleotide sequence analysis. Only one of these genes has introns and is expressed; the others are inactive processed pseudogenes. The expressed gene consists of five exons and four introns spanning 2,723 nucleotides. Transcripts of this gene are processed into the mature L30 mRNA by pathways that exhibit both constraints and flexibility with regard to the order of intron excision. The L30 mRNA which is 457 to 468 nucleotides in length excluding the polyadenylic acid tail, exhibits some microheterogeneity at its 3' end and encodes a basic protein of 115 amino acids. The 5' portion of the rpL30 gene has some novel features which are remarkably similar to the previously characterized mouse rpL32 gene. These include homologous sequences in the -60 to -340 region, the absence of a good TATA consensus sequence, and the presence of a palindromic pyrimidine sequence that spans the cap site.

Importance of introns for expression of mouse ribosomal protein gene rpL32.

The importance of intronic sequences for expression of the mouse ribosomal protein gene rpL32 was evaluated by transfection experiments with a series of mutant constructs in which one or more of the three rpL32 introns was totally or partially deleted. When transiently transfected into monkey kidney (COS) cells or stably transfected into mouse L cells, a mutant that lacked all three introns was completely inactive. Constructs that contained intron 1, either alone or in combination with another intron, were expressed as efficiently as was the normal intact rpL32 gene. Constructs that lacked intron 1 but contained another spliceable intron, even one from a foreign gene, were expressed at about 10 to 20% of the maximum level. These results indicated that intron 1 contains an element that increases the level of expression by 5- to 10-fold. A comparison of internal deletion mutants localized the element to within the first 27 base pairs of intron 1. Nuclear run-on experiments with stably transfected COS cells demonstrated that this element functions at the transcriptional level. The element was inactive when translocated to a position upstream of the transcriptional start site or to a position within intron 3, which indicated that it does not have the properties of a typical enhancer. From these and other results, we conclude that introns have both a general and a specific role in rpL32 expression. The general role, which can be satisfied by any spliceable intron, is to ensure an efficient yield of RNA transcripts. The specific role is uniquely attributable to intron 1, which contains a transcriptional regulatory element near its 5' end.

Characterization of the multigene family encoding the mouse S16 ribosomal protein: strategy for distinguishing an expressed gene from its processed pseudogene counterparts by an analysis of total genomic DNA.

Two genes from the family encoding mouse ribosomal protein S16 were cloned, sequenced, and analyzed. One gene was found to be a processed pseudogene, i.e., a nonfunctional gene presumably derived from an mRNA intermediate. The other S16 gene contained introns and had exonic sequences identical to those of a cloned S16 cDNA. The expression of this gene was demonstrated by Northern blot analysis of nuclear poly(A)+ RNA with cDNA and unique sequence intron probes. Each S16 intron contains a well-preserved remnant of the TACTAAC motif, which is ubiquitous in yeast introns and known to play a critical role in intron splicing. A sequence comparison with two other mouse ribosomal protein genes analyzed in our laboratory, L30 and L32, revealed common structural features which might be involved in the control and coordination of ribosomal protein gene expression. These include the lack of a canonical TATA box in the -20 to -30 region and a remarkably similar 12-nucleotide pyrimidine sequence (CTTCCYTYYTC) that spans the cap site and is flanked by C + G-rich sequences. The nature of the other members of the S16 family was evaluated by three types of experiment: a DNase I sensitivity analysis to measure the extent of chromatin condensation; an analysis of the thermal stability of cDNA-gene hybrids to estimate the extent of divergence of each gene sequence from that of the expressed gene; and a restriction fragment analysis which distinguishes intron-containing genes from intronless processed genes. The results of these analyses show that all genes except the expressed S16 gene are in a condensed chromatin configuration associated with transcriptional quiescence; that most of the genes within the S16 family have sequences greater than 7% divergent from the expressed S16 gene; and that at least 7 of the 10 S16 genes lack introns. We conclude that the ribosomal protein S16 multigene family contains one expressed intron-containing gene and nine inactive pseudogenes, most or all of which are of the processed type.

Glucocorticoids selectively inhibit translation of ribosomal protein mRNAs in P1798 lymphosarcoma cells.

When P1798 murine lymphosarcoma cells are exposed to 10(-7) M dexamethasone, there is a dramatic inhibition of rRNA synthesis, which is completely reversible when the hormone is withdrawn. In the present experiments we examined whether dexamethasone treatment causes any alteration in the accumulation or utilization of mRNAs that encode ribosomal proteins (rp mRNAs). No effect on the accumulation of six different rp mRNAs was detected. However, the translation of five of six rp mRNAs was selectively inhibited in the presence of the hormone, as judged by a substantial decrease in ribosomal loading. Normal translation of rp mRNA was resumed within a few hours after hormone withdrawal. In untreated or fully recovered cells, the distribution of rp mRNAs between polyribosomes and free ribonucleoprotein is distinctly bimodal, suggesting that rp mRNAs are subject to a particular form of translational control in which they are either translationally inactive or fully loaded with ribosomes. A possible relationship between this mode of translational control and the selective suppression of rp mRNA translation by glucocorticoids is discussed.

Localization of transcriptional regulatory elements and nuclear factor binding sites in mouse ribosomal protein gene rpL32.

The DNA sequences required for expression of the ribosomal protein gene rpL32 were identified by transient-expression assays of chimeric rpL32-chloramphenicol acetyltransferase genes. These studies showed that maximal rpL32 expression requires sequences in a 150- to 200-base-pair region spanning the transcriptional start site. Three discrete regions of importance were identified: one between positions -79 and -69 and two others located downstream of the transcriptional start site. Progressive 5' or 3' deletions caused stepwise decreases in expression, which suggested a complex interplay of redundant or compensatory elements. Gel mobility shift assays were used to identify trans-acting nuclear factors which bind to segments of the rpL32 promoter that are known to be important for transcription. Evidence for several distinct nuclear factors is presented. The binding sites for these factors were localized to the following regions: -79 to -69, -36 to -19, -19 to +11, +11 to +46 in exon I, and within the first 31 base pairs of intron 1. One of these factors may bind to multiple sites within the promoter region. Interestingly, the factor that binds to a sequence motif in the first exon also binds to similar motifs in a comparable region of the c-myc gene.

Characterization of productive and sterile transcripts from the immunoglobulin heavy-chain locus: processing of micron and muS mRNA.

An analysis of the sizes and sequence content of nuclear RNA transcripts of the heavy-chain locus in two B-cell lymphomas, 70Z/3 and 38C-13, and in selected hybridoma derivatives of 38C has led to the identification of two distinct precursors of the mRNAs encoding the membrane and secretory forms of mu chain. These precursors, termed Pm1 and Ps1, extend from a common 5' terminus (presumably the cap site) to alternative polyadenylation sites located 3' of the membrane and secretory tailpieces, Pm1 and Ps1 are present in similar amounts in lymphomas, indicating roughly equivalent usage of the two polyadenylation sites, whereas Ps1 much greater than Pm1 in hybridomas, indicating that mature plasma cells produce a trans-acting factor which enhances cleavage at the proximal (muS) site. The lymphomas also synthesize several nonproductive or sterile mu (Smu) transcripts from the second H allele. One class of sterile mu transcripts appears to be initiated about 1 kilobase downstream from the JH4 element. In 70Z, in which the nonproductive H allele has undergone a D1J2 fusion, another initiation site was located about 0.3 kilobase upstream of the D1 element. The sterile mu transcripts exhibit the same regulated termination at alternative polyadenylation sites as the mu mRNA precursors, although their rate of production is not necessarily coupled to that of the productive allele. This analysis has also defined probable processing pathways for productive and sterile components in which there is a 5' leads to 3' order for the excision of the large introns.

The developmentally regulated shift from membrane to secreted mu mRNA production is accompanied by an increase in cleavage-polyadenylation efficiency but no measurable change in splicing efficiency.

To determine whether there are any developmental changes in the efficiencies of cleavage-polyadenylation or splicing reactions that could affect the usage of weak (suboptimal) processing signals and thus provide a basis for the regulated production of mu m versus mu s mRNA during B-lymphocyte maturation, we studied the expression of transfected mu genes in which the natural competition between cleavage-polyadenylation and splicing was replaced by alternative usage of tandem weak and strong poly(A) sites or by competition between suboptimal and optimal 5' splice junctions. Our results indicate that there is a 50 to 100% increase in cleavage-polyadenylation efficiency but no measurable change in splicing efficiency as maturation proceeds from the B-cell to plasma cell stage.

The coupling between enhancer activity and hypomethylation of kappa immunoglobulin genes is developmentally regulated.

Previous studies have indicated that immunoglobulin enhancers are essential for establishing transcriptional competence but not for maintaining the activity of constitutively transcribed genes. To understand the basis for this developmental shift away from dependence on enhancer function, we have investigated the relationship between transcriptional activity and methylation status of the immunoglobulin kappa light-chain genes (kappa genes) in mouse cell lines representing different stages of B-cell maturation. Using pre-B-cell lines in which the level of a critical kappa enhancer-binding factor, NF-kappa B, was controlled by the administration or withdrawal of lipopolysaccharide and plasmacytoma lines that either contain or lack this factor, we studied the properties of endogenous kappa genes and of transfected kappa genes which were stably integrated into the genomes of these cells. In the pre-B cells, the exogenous (originally unmethylated) kappa genes, as well as endogenous kappa genes, were fully methylated and persistently dependent on enhancer function, even after more than 30 generations in a transcriptionally active state. In plasmacytoma cells, the endogenous kappa genes were invariably hypomethylated, whereas exogenous kappa genes were hypomethylated only in cells that contain NF-kappa B and are thus permissive for kappa enhancer function. These results indicate that the linkage of hypomethylation to enhancer-dependent activation of kappa transcription occurs after the pre-B-cell stage of development. The change in methylation status, together with associated changes in chromatin structure, may suffice to eliminate or lessen the importance of the enhancer for the maintenance of the transcriptionally active state.

Regulation of ribosomal protein mRNA content and translation in growth-stimulated mouse fibroblasts.

When resting (G0) mouse 3T6 fibroblasts are serum stimulated to reenter the cell cycle, the rates of synthesis of rRNA and ribosomal proteins increase, resulting in an increase in ribosome content beginning about 6 h after stimulation. In this study, we monitored the content, metabolism, and translation of ribosomal protein mRNA (rp mRNA) in resting, exponentially growing, and serum-stimulated 3T6 cells. Cloned cDNAs for seven rp mRNAs were used in DNA-excess filter hybridization studies to assay rp mRNA. We found that about 85% of rp mRNA is polyadenylated under all growth conditions. The rate of labeling of rp mRNA relative to total polyadenylated mRNA changed very little after stimulation. The half-life of rp mRNA was about 11 h in resting cells and about 8 h in exponentially growing cells, values which are similar to the half-lives of total mRNA in resting and growing cells (about 9 h). The content of rp mRNA relative to total mRNA was about the same in resting and growing 3T6 cells. Furthermore, the total amount of rp mRNA did not begin to increase until about 6 h after stimulation. Since an increase in rp mRNA content did not appear to be responsible for the increase in ribosomal protein synthesis, we determined the efficiency of translation of rp mRNA under different conditions. We found that about 85% of pulse-labeled rp mRNA was associated with polysomes in exponentially growing cells. In resting cells, however, only about half was associated with polysomes, and about 30% was found in the monosomal fraction. The distribution shifted to that found in growing cells within 3 h after serum stimulation. Similar results were obtained when cells were labeled for 10.5 h. About 70% of total polyadenylated mRNA was in the polysome fraction in all growth states regardless of labeling time, indicating that the shift in mRNA distribution was species specific. These results indicate that the content and metabolism of rp mRNA do not change significantly after growth stimulation. The rate of ribosomal protein synthesis appears to be controlled during the resting-growing transition by an alteration of the efficiency of translation of rp mRNA, possibly at the level of protein synthesis initiation.

Nonproductive kappa immunoglobulin genes: recombinational abnormalities and other lesions affecting transcription, RNA processing, turnover, and translation.

Six nonproductive kappa immunoglobulin genes (kappa- alleles) were cloned and sequenced. The structural abnormalities discerned from sequence analysis were correlated with functional lesions at the level of transcription, RNA processing, turnover, and translation. Four kappa- alleles, three containing V kappa genes and one not, are transcribed at normal or even greater than normal rates, the defects in these genes being expressed at various posttranscriptional levels. The other two kappa- alleles, both of which lacked V genes, exhibited greatly depressed yet clearly detectable transcriptional activity. These results are consistent with a hierarchical relationship between enhancer and promoter elements in which the enhancer establishes transcriptional competence at the kappa locus and the promoter (or pseudopromoter) determines the relative level of transcriptional activity. One of the structural abnormalities discovered in this study, a large deletion which removes the entire J kappa region, also provides new insight into the mechanism of VJ and VDJ recombination.

Transcription factor RFX1 helps control the promoter of the mouse ribosomal protein-encoding gene rpL30 by binding to its alpha element.

The factor that binds to the most upstream element (alpha) of the mouse rpL30 promoter was identified as RFX1, a novel 105-kDa protein that recognizes an important element of MHC class-II promoters. Identification was based on competition between rpL30 alpha and an RFX1-binding site for nuclear protein complex formation and on the ability of RFX1 antibody to supershift the electrophoretic mobility of the DNA-protein complexes. A mutation in the alpha-element that abolished its interaction with RFX1 reduced rpL30 promoter activity to about 43% of the wild-type level, indicating that RFX1 plays an important role in determining the strength of the rpL30 promoter. A search of a eukaryotic promoter database revealed candidate RFX1-binding sites in a variety of other promoters, suggesting that this protein may be implicated in the transcriptional regulation of a wide variety of genes.

Cell-free transcription of a mouse ribosomal-protein-encoding gene: the effects of promoter mutations.

The mouse ribosomal protein-encoding gene, rpS16, was accurately transcribed in vitro with a high-efficiency nuclear extract prepared from HeLa cells. An analysis of the relative activities of rpS16 templates containing deletions or deleterious mutations of various promoter elements indicated that the in vitro transcription system can recognize all of the promoter elements that were previously identified by in vivo transfection experiments. The importance of a polypyrimidine initiator, which spans the cap site, and an element termed C, which is located in the region conventionally occupied by a TATA-box, was also assessed with an appropriate set of mutant templates. In agreement with earlier in vivo studies, our in vitro results indicated that the initiator is the primary determinant for selecting the transcription start point. Interestingly, however, the in vitro system exhibited a strong bias for features that are not present in the natural rpS16 gene. Mutant templates that contained a purine rather than a pyrimidine at or near the cap site or that had the C element replaced by a canonical TATA-box were six- to tenfold more active than the wild-type (wt) rpS16 gene in vitro system, whereas in vivo, these mutants were expressed equivalently to the wt gene. A double mutant containing both a cap site substitution and the TATA-box replacement was about 18-fold more active than the wt gene in the in vitro transcription system. These results suggest that the in vitro system may preferentially build a functional transcription complex with components that do not normally interact with the wt rpS16 gene.(ABSTRACT TRUNCATED AT 250 WORDS)