Analysis of the message-sequence content of the pulse-labeled poly(A)+ heterogeneous nuclear RNA from HeLa cells by cDNA-excess hybridizations.

The message-sequence content of pulse-labeled poly(A)+ HeLa heterogenous nuclear RNA (hnRNA) has been examined by hybridizations to an excess of message cDNA. Control experiments show that the message cDNA accurately reflects the sequence distribution of the complex mixture of poly(A)+ messages present in the HeLa cytoplasm. Pulse-labeled poly(A)+ molecules in both the lamina-associated and shnRNA fractions contain message sequences, and approximately 65% of the poly(A)-adjacent hnRNA sequences are homologous to the 3' ends of mRNA. The majority of the pulse-labeled hnRNA molecules contain abundant message sequences. By use of these techniques it is also shown that some pulse-labeled polyadenylated message sequences are still synthesized in the presence of the adenosine analogue 5,6-dichloro-beta-D-ribofuranosylbenzimidazole under conditions where little or no new cytoplasmic mRNA is produced.     

Concentrations of individual RNA sequences in polyadenylated nuclear and cytoplasmic RNA populations of Drosophila cells.

Steady state concentrations of individual RNA sequences in poly(A) nuclear and cytoplasmic RNA populations of Drosophila Kc cells were determined using cloned cDNA fragments. These cDNAs represent poly(A) RNA sequences of different abundance in the cytoplasm of Kc cells, but their steady state concentrations in poly(A) hnRNA was always lower. Of ten different sequences analysed, eight showed some four-fold lower concentration in hnRNA mRNA, two were underrepresented in hnRNA relative to the others. The obvious clustering of mRNA/hnRNA ratios is discussed in relation to sequence complexity and turnover rates of these RNA populations.     

Nuclear precursors for ceruloplasmin-coding messenger RNA from rat liver

The distribution of the sequences coding for ceruloplasmin (CP) in rat liver heterogeneous nuclear RNA (hnRNA) was studied using highly specific CP cDNA as a hybridization probe. The content of CP-coding sequences in poly(A)-containing and poly(A)-free subfractions of hnRNA was shown to be respectively 1 and 27 equivalents of CP mRNA molecule per one hepatocyte. The gel electrophoresis of hnRNA under strongly denaturing conditions with the subsequent transfer of RNA to diazobenzyloxymethyl paper and hybridization with [32P]-cDNA probe showed that CP mRNA sequences were of multiple molecular weight distribution. In particular, 9.0, 6.6, 2.4 and 1.6 megadalton fractions of non-polyadenylate hnRNA carried CP-coding sequences while the only hand that hybridized to CP cDNA was detected in polyadenylated hnRNA. This band was of a molecular weight 1.1-1.2 megadaltons corresponding to that of cytoplasmic CP mRNA. The hybridization of high molecular weight hnRNA with full-length CP cDNA followed by the determination of the size of cDNA fragments protected against SI nuclease demonstrated that coding sequences of CP pre-mRNA are interrupted by intervening sequences.     

Biosynthesis and utilization of extensively undermethylated poly(A)+ RNA in CHO cells during a cycloleucine treatment.

The role of RNA methylations in the control of mRNA maturation and incorporation into polysomes has been investigated through a study of the effects in vivo of cycloleucine, a specific inhibitor of S-adenosyl-methionine mediated methylation. During the cycloleucine treatment, the rate of biosynthesis of hnRNA and its subsequent polyadenylation were only slightly reduced as compared with untreated cells. However a significant lag-time in the cytoplasmic appearance of poly(A)+ undermethylated molecules was observed, in parallel with a transient shift in the average size of hnRNA towards higher molecular weight. Nevertheless, the total amount of pulse-labelled poly(A)+ mRNA transferred to cytoplasm after a long chase time (3 h.) was approximately the same for both cycloleucine-treated and control cells. Extensively undermethylated poly(A)+ cytoplasmic RNAs, possessing a 5' terminal cap were incorporated into polysomes in proportions very similar to control messenger molecules. These results suggest that a normal level of methylation is not stringently required for the production of the functional mRNA molecules although it appears to be of importance for the kinetics of the maturational process.     

5'-terminal structures of poly(A)+ cytoplasmic messenger RNA and of poly(A)+ and poly(A)- heterogeneous nuclear RNA of cells of the dipteran Drosophila melanogaster.

Structures at the 5′ terminus of poly (A)-containing cytoplasmic RNA and heterogeneous nuclear RNA containing and lacking poly(A) have been examined in RNA extracted from both normal and heat-shocked Drosophila cells. ³²P-labeled RNA was digested with ribonucleases T₂, T₁ and A and the products fractionated by a fingerprinting procedure which separates both unblocked 5′ phosphorylated termini and the blocked, methylated, “capped” termini, known to be present in the messenger RNA of most eukaryotes.Approximately 80% of the 5′-terminal structures recovered from digests of poly(A)-containing Drosophila mRNA are cap structures of the general form m⁷G^5′ppp^5′X(m)pY(m)pZp. With respect to the extent of ribose methylation and the base distribution, the 5′-terminal sequences of Drosophila capped mRNA appear to be intermediate between those of unicellular eukaryotes and those of mammals. Drosophila is the first organism known in which type 0 (no ribose methylations), type 1 (one ribose methylation), and type 2 (two ribose methylations) caps are all present. In contrast to mammalian cells, the caps of Drosophila never contain the doubly methylated nucleoside N⁶,2′-O-dimethyladenosine. Both purines and pyrimidines can be found as the penultimate nucleoside of Drosophila caps and there is a wide variety of X-Y base combinations. The relative frequencies of these different base combinations, and the extent of ribose methylation, vary with the duration of labeling. The large majority of poly(A)-containing cytoplasmic RNA molecules from heat-shocked Drosophila cells are also capped, but these caps are unusual in having almost exclusively purines as the penultimate X base.Greater than 75% of the 5′ termini of heterogeneous nuclear RNA (hnRNA) containing poly(A) and greater than 50% of the termini of hnRNA lacking poly (A) are also capped. Triphosphorylated nucleotides, common as the 5′ nucleotides of mammalian hnRNA, are rare in the poly(A)-containing hnRNA of Drosophila. The frequency of the various type 0 and type 1 cap sequences of cytoplasmic and nuclear poly (A)-containing RNA are almost identical. The caps of hnRNA lacking poly(A) are also quite similar to those of poly-adenylated hnRNA, but are somewhat lower in their content of penultimate pyrimidine nucleosides, suggesting that these two populations of molecules are not identical.     

Rates of synthesis and turnover of 5′ cap structures of hnRNA and mRNA and their changes during sea urchin development

The relationship between heterogeneous nuclear RNA (hnRNA) and messenger RNA (mRNA) synthesis has been studied as a function of the development of the sea urchin embryo through the use of methyl incorporation. Several parameters in the metabolism of capped hnRNA and mRNA of early blastula and late gastrula stages have been investigated by measuring the kinetics of transfer of methyl groups from S-adenosylmethionine to the 5′ cap structures in nuclear and cytoplasmic RNA:

• 1 The rate constants for the decay of hnRNA caps and the synthesis of mRNA caps are equal to within experimental error. This equality indicates a flux of precursor hnRNA caps to mRNA caps with a very high degree of conservation of the hnRNA caps. This conservation holds for each embryonic stage.
• 2 From literature data on the labeling kinetics of GTP and mRNA, we have calculated the decay constant of a putative mRNA precursor component of hnRNA. The value of this constant is very close to that for the decay constant of hnRNA caps. Hence, all hnRNA caps and some portion of their associated hnRNA sequences behave kinetically as the pre-mRNA fraction. This kinetically ascribed pre-mRNA comprises approximately 30% of the hnRNA mass.
• 3 The part of the hnRNA which does not serve as precursor to mRNA turns over at least twice as rapidly as the pre-mRNA fraction.
• 4 During development from early blastula to late gastrula, the rate of hnRNA cap synthesis drops from 2 × 10³ molecules/min/cell to half of this value. This decline is parallel to the decline in total hnRNA synthesis and thereby confirms the constant degree of capping of hnRNA, as previously reported. We infer that the pre-mRNA fraction of hnRNA remains nearly constant during this developmental period.

     

Complexity and characterization of polyadenylated RNA in the mouse brain

The complexity of nuclear RNA, poly(A)hnRNA, poly(A)mRNA, and total poly(A)RNA from mouse brain has been measured by saturation hybridization with nonrepeated DNA. These DNA populations were complementary, respectively, to 21, 13.5, 3.8, and 13.3% of the DNA. From the RNA Cot required to achieve half-sturation, it was estimated that about 2.5–3% of the mass of total nuclear RNA constituted most of the complexity. Similarly, complexity driver molecules constituted 6–7% of the mass of the poly(A)hnRNA. 75–80% of the poly(A)mRNA diversity is contained in an estimated 4–5% of the mass of this mRNA. Poly(A)hnRNA constituted about 20% of the mass of nuclear RNA and was comprised of molecules which sedimented in DMSO-sucrose gradients largely between 16S and 60S. The number average size of poly(A)hnRNA determined by sedimentation, electron microscopy, or poly(A) content was 4200–4800 nucleotides. Poly(A)mRNA constituted about 2% of the total polysomal RNA, and the number average size was 1100–1400 nucleotides. The complexity of whole cell poly(A)RNA, which contains both poly(A)hnRNA and poly(A)mRNA populations, was the same as poly(A)hnRNA. This implies that cytoplasmic polyadenylation does not occur to any apparent qualitative extent and that poly(A)mRNA is a subset of the poly(A)hnRNA population. The complexity of poly(A)hnRNA and poly(A)mRNA in kilobases was 5 × 10⁵ and 1.4 × 10⁵, respectively. DNA which hybridized with poly(A)mRNA renatures in the presence of excess total DNA at the same rate as nonrepetitive tracer DNA. Hence saturation values are due to hybridization with nonrepeated DNA and are therefore a direct measure of the sequence complexity of poly(A)mRNA. These results indicate that the nonrepeated sequence complexity of the poly(A)mRNA population is equal to about one fourth that observed for poly(A)hnRNA.     

The relationship between hnRNA+-poly(A) and mRNA+-poly (A) in non-dividing human lymphocytes. Evidence for distinct synthetic pathways for mRNA precursor- and nonprecursor-hnRNA

The processing of hnRNA+-poly(A) to mRNA+-poly(A) has been studied in resting lymphocytes from human peripheral blood. In pulse-chase experiments, two types of hnRNA+-poly(A) have been distinguished: the first is labeled predominantly with exogenous radioactive precursors supplied during the pulse, and the second incorporates primarily scavenged labeled precursors made available during a chase incubation. When the disappearance of both types of hnRNA+-poly(A) was quantitatively compared with the appearance of stable and labile mRNA+-poly(A), only 10% of the anticipated cytoplasmic material was actually obtained. Statistically, 90% of the poly(A)-bearing hnRNA molecules processed were degraded. The two types of hnRNA+-poly(A) were found to be functionally different. Pulse-labeled material was processed to poly(A)-bearing mRNA; "chase-labeled" molecules did not leave the nucleus and never served as precursors for cytoplasmic mRNA. The data fit a model in which there are distinct pathways for mRNA precursor- and nonprecursor-hnRNA+-poly(A).     

Proteins associated with poly(A) and other regions of mRNA and hnRNA molecules as investigated by crosslinking

The proteins associated with poly(A) and other regions of mRNA and hnRNA molecules in mouse L cells were investigated with the aid of ultraviolet light-induced crosslinking of proteins to RNA. The poly(A)s of polyribosomal and free cytoplasmic mRNAs are associated with a protein, p78A. In contrast, the poly(A) of hnRNA is associated with a smaller protein, p60A, that differs from p78A in its partial peptide map. p78A occurs free in the cytoplasm, but p60A does not. There is a second 78 kd protein, p78X, associated with mRNA sequences other than poly(A). p78X differs from p78A in its partial peptide map. The total proteins crosslinked to polyribosomal and free cytoplasmic mRNAs are similar. However, the total proteins crosslinked to hnRNA are quite different from those crosslinked to mRNA. We suggest that newly synthesized mRNA molecules emerging from the nucleus into the cytoplasm shed the proteins with which they were associated in the nucleus and become associated with a new set of proteins derived from the cytosol. Furthermore, the cytoplasmic mRNA-associated proteins continue to exchange with free proteins.     

Properties of a 110000 molecular weight protein in rat liver hnRNP particles

Particles carrying heterogeneous nuclear RNA (30–40 S-particles) were isolated from rat liver nuclei and the particle proteins separated by sodium dodecylsulfate gel electrophoresis. Some properties of a 110000 molecular weight component (P 110/103) were studied in detail: (i) P 110/103 was labeled to a 4–5 times higher specific activity than the major particle proteins in the presence of [¹⁴C]-amino acidsin vivo. (ii) In nuclei incubated with [³H]- or [³²P]-nicotinamide adenine dinucleotide P 110/103 was labeled presumably by ADP-ribosylation. (iii) A protein with the same molecular weight as P 110/103 and isolated from the nuclear extract by affinity chromatography was phosphorylated in vitro.Abbreviations hnRNA heterogeneous nuclear RNA - hnRNP and mRNP ribonucleoproteins which contain hnRNA respectively mRNA     

Herpesvirus infection alters the steady-state levels of cellular polyadenylated RNA in polyoma virus-transformed BHK cells.

Polyoma-transformed BHK cells are permissive for the replication of herpes simplex virus type 1. The effect of herpes infection on the steady-state levels of bulk mRNA sequences in these cells was studied by using cDNA to polyadenylated cytoplasmic RNA from uninfected cells. The principal findings were: (i) herpes simplex virus type 1 infection caused a pronounced reduction in the cytoplasmic levels of moderately abundant mRNAs' (ii) after infection, increased amounts of RNA complementary to the cDNA were isolated as part of the nonadenylated cytoplasmic RNA fraction.     

Globin mRNA precursor. Cross-linking in situ of double-stranded segments with aminomethyltrioxalen.

The globin mRNA sequences present in duck erythroblast nuclei appear under non-denaturing conditions to be associated with heterogeneous nuclear RNA (hnRNA) molecules of various sizes. Under denaturing conditions, however, the bulk of the globin mRNA sequences associated with hnRNA are released as molecules of size close to that of the active globin mRNA. To find out whether hydrogen-bonded structures occur in situ or arise after RNA extraction, nuclei were treated with aminomethyltrioxalen and exposed to ultraviolet light. This treatment generates covalent links between opposite strands of double-stranded nuclei acids, which were visualised by electron microscopy. It appears that, after cross-linking, a fraction of the globin mRNA sequences present in nuclei is associated with high-molecular-weight hnRNA molecules by a link found associated with a band of 0.9 x 10(6) molecular weight approximately. It is suggested that within the erythroblast nucleus, globin mRNA sequences are associated by hydrogen bonds with RNA of high molecular weight. These structures may represent intermediate steps in globin mRNA processing.