The effects of serum deprivation on several general cellular biochemical processes ("pleiotypic response") related to the growth of normal fibroblasts can be mimicked by treatment of these cells with prostaglandin El in the presence of serum. N8,02'-Dibutyryl adenosine 3': 5'-cyclic monophosphate and theophylline inhibit the membrane transport processes without much effect on other pleiotypic reactions such as overall protein and RNA synthesis and protein degradation. The amount of intracellular cyclic AMP increases during serum starvation and returns to the initial concentration in unstarved cells when growth is initiated again upon addition of serum. Fibroblasts transformed by simian virus 40 have a lower cyclic AMP content than their untransformed parents. Serum deprivation neither increases cyclic AMP content nor significantly affects the pleiotypic reactions in transformed cells. Cycloheximide causes a decrease in cyclic AMP content of normal fibroblasts coincidentally with the ability of this inhibitor to stimulate uridine transport and slow protein degradation in cells deprived of serum.In normal cells, several biochemical parameters, including membrane transport, overall rates of protein and RNA synthesis, and protein degradation, fluctuate coordinately with changes in cell-growth rate. These mechanistically unrelated processes and their coordinated response to environmental changes comprise what we have defined as the "pleiotypic program" and "pleiotypic control," respectively. In malignant cells, the pleiotypic program and the rate of growth are relatively insensitive to changes in cultural conditions; we thus picture transformed cells as having a defect in the pleiotypic control mechanism (1).In cultured fibroblasts, insulin as well as serum activates the pleiotypic program (1, 2). Moreover, the same set of reactions is regulated in many other cell types by specific hormones or growth factors. Thus, certain general features of hormonal regulation may also be exerted through the pleiotypic mechanism (1).We have postulated that a "mediator", probably formed at the cell membrane, coordinates the responses of the different reactions under pleiotypic control. Because of the resemblance of this regulatory program to "stringent control" in bacteria, we inquired whether guanosine 5'-diphosphate, 2'-or 3'-diphosphate (ppGpp) (a possible mediator of stringent control) (3, 4) could be involved in control in animal cells. However, this nucleotide was not detected in cultured 3T3 cells under conditions that should have elicited its appearance (5).In the meantime, numerous reports have implicated cAMP as a regulator of growth, morphology, and contact inhibition in cultured cells (6-11). We report here that altering the inAbbreviations: Bu2cAMP, N6,02'-dibutyryl adenosine 3':5'-cyclic monophosphate; PGE,, prostaglandin El; SV40, simian virus 40. tracellular cAMP content by various means can affect all the reactions under pleiotypic control in a way, suggesting that cAMP is very likely the pleiotypic mediator...

Previous studies have shown that exogenous dibutyryl cyclic AMP inhibits the uptake of uridine, leucine, and 2-deoxyglucose by cultured mouse fibroblasts. 3':5'-cyclic GMP is shown here to counteract these inhibitory effects as well as the inhibition ofprecursor transport and leucine incorporation into proteins produced by prostaglandin E1. We conclude, therefore, that cyclic GMP antagonizes the "pleiotypic" effects of cyclic AMP in these cells.Colcemid and vinblastine, but not cytochalasin B., reverse the transport inhibition caused by cyclic AMP without affecting the intracellular concentrations of cyclic AMP. These results suggest the possibility that cyclic AMP regulates the membrane transport of certain substrates by influencing the organization of microtubules.Recent work from various laboratories has implicated cAMP in the regulation of cellular growth. In a preceding communication (1) we presented evidence that cAMP exerts this control by coordinately influencing several biochemical processes related to growth. This set of reactions was defined as the "pleiotypic" program, and its regulation as pleiotypic control (2). Earlier we had singled out this coordinated control as being of general significance since the same set of reactions is affected in various cell types by specific hormones and growthpromoting factors (2). The fact that cAMP controls the pleiotypic reactions provides a basis for relating differentiation, malignant transformation, and hormonal control of growth (1). We report here some observations pertaining to the mechanisms of cAMP control of the pleiotypic parameters.cGMP antagonizes the cAMP stimulated synthesis of flgalactosidase in cell-free bacterial extracts (3, 4). In certain animal cells as well, cGMP appears as if it might also directly or indirectly counteract the effects of cAMP (5-9). We have, therefore, investigated the interaction between these cyclic nucleotides on the pleiotypic reactions and conclude that here, too, cGMP overcomes the actions of cAMP. We also suggest, on the basis of studies with inhibitors of microtubule assembly, that the cyclic nucleotides may influence some membrane transport processes by their effects on colcemid-and vinblastine-sensitive structures, possibly the microtubular apparatus itself. MATERIALS AND METHODSThe Balb/c 3T3 mouse fibroblasts and their SV 40-transformed derivatives were grown in MEM plus 10% calf serum.The rates of precursor transport and incorporation into macromolecules were measured as described (2) In 3T3 cells, the rates of transport of uridine, leucine, and 2-deoxyglucose are depressed during serum deprivation and stimulated by the addition of serum again. We have reported(1) that (Bu)2cAMP + theophylline or PGE1 added to cells maintained in serum-containing medium mimic the response to serum starvation. These effectors also inhibit serumstimulated uptake of precursor by previously starved cells (1). We now show that cGMP antagonizes the effects of both (Bu)2cAMP + theophylline and PGE1 on these biochemical processes...

The nuclear DNA of D. melanogaster contains DNA sequences that are repeated between ten and a hundred times more often than the next class of redundant DNA. This DNA, as a renatured duplex isolated on the basis of its renaturation kinetics, has a buoyant density of 1.691 (g/ml). In its native state it bands within the unique nuclear DNA peak (p = 1.701). These sequences have been localized by "in situ" hybridization in the chromocenter of the chromosomes of the salivary gland. The properties of centromeres are discussed in terms of the occurrence of repeated sequences at this locus.At present, our understanding of the organization of DNA sequences in the higher organism is very poor. The presence of multiple copies of certain sequences is one feature that has been established through the study of renaturation kinetics of denatured DNA (1, 2). Gene classes have been defined (1) The satellite DNAs of rodents represent a discrete set of sequences, with the highest extent of repetition (3-5) that is found in their genomes. The other repeated sequences belong to an intermediate class, with a continuous distribution of moderately repeated families, until a second discontinuity in the extent of repetitions presumably reveals sequences that represent single-copy DNA. Though representing 10% of the total DNA, the mouse satellite has a complexity of 300 base pairs. The basic unit probably contains a sequence of from 8 to 13 base pairs, distributed in some manner between sequences that have been derived by mutation from this basic sequence (6). There is evidence that this class of highly-redundant DNA is not transcribed (4). Recently, the mouse satellite DNA has been located in the centromeric heterochromatin of the metaphase chromosomes (7,8). This observation is consistent with the earlier demonstration that the mouse satellite DNA is associated with the less soluble fraction of metaphase chromosomes (9) (12), with the addition of a 4-hr digestion with 0.5 mg pronase per ml (self-digested for 2 hr) after the RNase and a-amylase treatments. As discussed in the results, cytoplasmic DNA contaminants were found in crude nuclear preparations (600 X g pellet). Nuclear DNA with less than 0.3% cytoplasmic DNA contamination was obtained from nuclei purified by sedimentation through 2.2 M sucrose for 1 hr at 45,000 X g at 4°C. Contamination with either bacterial or yeast nucleic acid was ruled out because 3H-labeled whole Escherichia coli and yeast cells added to the starting material did not add any labeled DNA to the final extracts. X 10-10 g sec2/cm5 was used to calculate (15) density differences; 99% E. coli [15N] DNA was used as a marker of buoyant density 1.727 ± 0.0004, relative to a primary standard, unlabeled E. coli DNA, assigned a buoyant density of 1.710. The cytoplasmic peaks contaminate this nuclear DNA sample by 3%.

We have investigated the arrangement of the highly reiterated centromeric DNA sequences in Drosophila melanogaster by studying the physical properties of sequences obtained after the hydroxyapatite isolation of rapidly renaturing molecules. It was found that the fraction of material recovered as partially renatured varied with the molecular weight of the starting material, although conservative length corrections on the chosen Cot values were made. On banding in a CsCl gradient a bimodal distribution for the renatured molecules was found. This bimodal distribution was also sensitive to the molecular weight of the single strand material. The peaks were shown to consist of respectively: h.a.r.r.DNAt sequences and renatured h.a.r.r.DNA sequences with attached unrenatured "spacer" sequences. From melting curve data, RNA-DNA hybridization and kinetic analysis other explanations for the increased yield and bimodal distribution of the renatured molecules could be ruled out. Computer simulation of a random shear process on a DNA sequence which alternates an h.a.r.r.DNA sequence from one to three times larger than the spacer sequence can be seen to fit the experimental situation. We conclude that the constitutive heterochromatin is composed of blocks of extremely simple sequences stretching from 1.5 to 6 million daltons interspersed by a more complex sequence of molecular weights of approximately 2 to 4 million daltons. A similar interspersion of sequences for the mouse satellite DNA was found, though preliminary data suggest that the stretches of mouse satellite may be much longer. The occurrence of a more complex DNA at this locus of the Drosophila chromosome raises some interesting possibilities for both the genetic potential of this area and for chromosomal "housekeeping" functions.

No abstract