Cell cycle parameters of adult rat hepatocytes in a defined medium. A note on the timing of nucleolar DNA replication

Hepatocytes, isolated from adult (250-350 g) rats, attached and survived well in primary culture on highly diluted (less than 1 microgram/cm2) collagen gel in a synthetic medium without serum or hormones. About 20% of the cells "spontaneously" entered S phase during the first 4 days of culturing, and mitoses were easily demonstrated at the near physiological concentration (1.25 mM) of Ca++ prevailing in the medium. Cultures given 9 nM epidermal growth factor (EGF) and 20 nM insulin 20 h after inoculation showed vigorous DNA synthesis and mitotic activity. Autoradiography of such cells exposed to [3H]thymidine allowed the determination of the following cell cycle parameters: Lag period from EGF/insulin stimulation till onset of increased DNA synthesis, 17 h; rate of entry into S phase (kG1/S), 0.028/h; duration of S phase, 8.4 h; duration of G2 phase, 2.7 h. The peak DNA synthesis (pulse labelling index, 24%) and peak mitotic activity (mitotic index, 1.7%) occurred 35 and 43 h, respectively, after the stimulation with EGF/insulin. These values are comparable to those reported during the in vivo compensatory hyperplasia following partial hepatectomy of adult rats. A marked variation of the intranuclear [3H]thymidine pulse labelling pattern was noted: During the first 1.5 h of the S phase, the labelling was extranucleolar and during the last 1.5 h chiefly nucleolar. The cells survived well in the absence of glucocorticoid, whose effect on cell cycle parameters therefore could be studied. Dexamethasone (25-250 nM) did not appreciably affect the durations of S phase and G2 phase or the pattern of preferential extranucleolar and nucleolar DNA synthesis within the S phase.     

STUDIES ON THE REGULATION OF LYMPHOCYTE PRODUCTION IN THE MURINE THYMUS AND SOME EFFECTS OF A CRUDE THYMUS EXTRACT

Cell proliferation in the murine thymus was studied in vivo under normal conditions and from 0 to 24 hr after a single injection of a water-soluble extract from mouse thymus, mouse spleen, and mouse skin. The thymus extract reduced during the first 24 hr the mitotic activity 40%; the spleen extract had a weaker inhibitory effect. The skin extract had no such effect. The thymus extract and spleen extract inhibited the flux of cells into the S phase 0–8 hr after the injection of the extract. Initial labelling index was also reduced in this period. Eight hours after injection of the thymus or spleen extracts the inhibited cells initiated DNA synthesis. The rate of progression of blast cells through the cell cycle was normal 24 hr after the injection of the extracts. It was deduced from the analysis that the thymus extract inhibits processes triggering Go/Gi cells into DNA synthesis, the inhibition of G₂ efflux being of minor importance. Finally a model for the regulation of proliferating thymic blast cells and the emigration of small lymphocytes from the thymus is proposed.     

DNA synthesis in BalB/C-3T3 ts 2 cells is restricted by a temperature-sensitive function of late G1 phase

ts 2 BalB/C-3T3 mouse fibroblasts are cdc mutants, which arrest late in G1, at or near the G1/S traverse, upon full expression of the heat-sensitive lesion. The kinetics of temperature inhibition of DNA synthesis in logarithmically growing cultures reveal three stages of heat inactivation. During the first generation time equivalent, normal semiconservative, semidiscontinuous replication proceeds but is reduced as cells exit and do not reenter S phase. During a second such period, a minimal rate of normal DNA synthesis is maintained. Thereafter, as the cells move into a third aborted cell division cycle, the rate of DNA synthesis increases. However, all semiconservative synthesis is then replaced by DNA repair replication. Temperature inactivation of the ts 2 protein results in shutdown of nuclear DNA synthesis. In contrast, normal replication of mitochondrial DNA proceeds at control rate throughout the first stage of temperature inactivation. Synthesis of this organellar genome is quantitatively reduced as the cells move into the second phase of heat inhibition. Titration of chromatin-bound DNA with ethidium bromide revealed that wild-type cells exhibit a changing DNA topology as the temperature is raised. Temperature-inactivated ts 2 cells behave as though their DNA has been topologically frozen in the configuration of control cells at or near entry into S phase.     

Metabolic effects of nucleosides in rat thymus cells in vitro

The incorporation of pyrimidine nucleosides into deoxyribonucleotides by rat thymocytes in vitro was decreased by the addition of any one of several nucleosides. In cells incubated with deoxyguanosine, the decrease was apparently caused by interference with the ribonucleotide reductase reaction, and at least part of the effect of deoxyadenosine was produced in the same way. While ara-C also produced a decrease in labelling of deoxynucleotides, this was quantitatively less than the effect on DNA synthesis, and may have resulted from an indirect effect on the reductase by a deoxyribonucleoside triphosphate which accumulated due to a direct effect of an ara-C derivative on the DNA polymerase. Cells incubated in the presence of adenosine showed a decreased labelling of deoxynucleotides and DNA due to inhibition of earlier steps in utilization of the labelled precursor, its uptake and phosphorylation. Guanosine, or a derivative, apparently reduced labelling of DNA and RNA even at concentrations which produced no alteration in uptake and phosphorylation of the precursor. An effect on the reductase was indicated. Uptake and phosphorylation of the pyrimidine ribonucleosides seem to be separate processes, since each can vary independently of the other when cells are incubated in the presence of various concentrations of exogenous nucleosides.     

Abnormal DNA synthesis in polyamine deficient cells revealed by bromodeoxyuridine—flow cytometry technique

Abstract. Chinese hamster ovary cells were seeded in the absence or presence of the polyamine synthesis inhibitor 2-difluoromethylornithine (DFMO). At 14 days after seeding, the cells were labelled for 15–120 min with the thymidine analogue bromo-deoxyuridine (BrdUrd) and they were then fixed directly after the labelling period. In addition, cells were labelled for 30 min and they were then allowed to progress in BrdUrd-free medium during a defined post-labelling time before fixation. An indirect immunofluorescence technique, using the monoclonal BrdUrd antibody and the intercalating stochiometric DNA stain, propidium iodide, was applied to enable quantification of cellular BrdUrd and DNA contents, respectively, by flow cytometry (FCM). By comparing the mean DNA content of BrdUrd-labelled cells to the mean DNA contents of G₁ and G₂ cells, a relative measure of the position of the BrdUrd-labelled cells was obtained (relative movement). Relative movement data, obtained from control and DFMO-treated cells fixed directly after BrdUrd labelling, indicated that DFMO-treated cells entered S phase at a normal rate, while their progression through S phase was impaired. DNA histograms of BrdUrd-labelled control cells fixed directly after labelling showed that most cells were found in early and late S phase, while DNA histograms of BrdUrd-labelled DFMO-treated cells showed that most cells were in early S phase, indicating a delayed progression through S phase. Analysis of relative movement of cells that were allowed to progress in BrdUrd-free medium after labelling showed that DFMO treatment resulted in a significant lengthening of the DNA synthesis time. Labelling index was significantly higher in DFMO-treated, growth-inhibited cells than in early plateau phase control cells indicating an S phase accumulation in the former cells.     

REGENERATIVE PROLIFERATION OF MOUSE EPIDERMAL CELLS FOLLOWING APPLICATION OF A SKIN IRRITANT (CANTHARIDIN)

The strong skin irritant cantharidin dissolved in benzene was applied to the back of hairless mice. Single cell suspensions of epidermal basal cells were obtained and flow microfluorometric measurements of cellular DNA content were made. Smears were made for autoradiography, and the [³H]TdR labelling index (LI) and mean grain count (MGC) were assessed up to 3 days after cantharidin application. Three successive peaks of cells with S phase DNA content accompanied by three LI peaks were observed. The first two peaks were follwed by peaks of cells in G₂ phase, indicating that after the acute cell injury caused by cantharidin the cells traversed the cell cycle in partial synchrony through two subsequent cell cycles, each of 10–12 hr duration. During this phase of rapid proliferation the LI reached the proportion of cells in S phase, contrary to what is observed in untreated mouse epidermis, where the labelled cells contribute to about half the proportion of cells with S phase DNA content. The first two peaks of cells in S phase and LI coincided with an increased MGC, whereas the third peak was accompanied by a MGC significantly below control values. This indicates that this latter peak is due to a longer DNA synthesis time rather than to a partially synchronized and increased cell proliferation. The duration of the G₁, S and G₂ phases seems to be reduced initially in rapidly proliferating epidermis.     

RADIOAUTOGRAPHIC STUDIES IN SYNCHRONIZED CULTURES OF OEDOGONIUM CARDIACUM

Synthesis of deoxyribonucleic acid (DNA) in synchronized cultures of Oedogonium cardiacum has been studied by radioautography. The germinated zoospores are pulse-labelled for 15 min with thymidine-2-C¹⁴. Use of penicillin in the medium reduces the background in the radioautographs by suppressing the growth of bacteria on Oedogonium cells. Incorporation of labelled precursor is greatly enhanced by growing the cells in a conditioned inorganic medium. The length of the DNA synthesis period (S), as estimated from the curve of percentage of labelled cells versus age during the first cell cycle is about 5-7 hr. The rate of labelling in the nucleus is non-uniform showing a dip during the mid S period. Concomitant with the doubling of DNA in the nucleus a fourfold increase in the nuclear volume is observed.     

CYTOKINETICS OF RAT TRACHEAL EPITHELIUM STIMULATED BY MECHANICAL TRAUMA

Mild abrasion of rat tracheal epithelium results in irreversible damage to the superficial cells and stimulates the viable basal cells to participate in a nearly synchronous wave of DNA synthesis and mitosis. For the growth population as a whole, DNA synthesis started at 14 hr after injury and persisted for 16 hr. The duration of S in individual cells was determined autoradiographically by identifying the time at which a second pulse of DNA precursor (¹⁴C-TdR) was no longer incorporated by cells labelled with ³H-TdR at the onset of S. S was found to be 8–9 hr long. It was also determined that cells entering S at later times synthesized DNA for the same 8–9 hr period. T_G2 was calculated to be 21/2–31/2 hr by subtraction of T_s and 1/2T_M from the period from onset of DNA synthesis to metaphase. By making a second denuding lesion adjacent to the first injury, the cells were stimulated through at least another period of S. At the peak of the second wave of DNA synthesis (50 hr after injury) ¹⁴C-TdR was present in the same cells which had incorporated ³H-TdR administered at the mid-point of the preceding synthetic phase. The 28-hr interval between these two peaks of synthesis is the measure of cell cycle duration for these regenerating tracheal epithelial cells.     

EGF induces cell cycle arrest of A431 human epidermoid carcinoma cells

The human carcinoma cell line A431 is unusual in that physiologic concentrations of epidermal growth factor (EGF) inhibit proliferation. In the presence of 5-10 nM EGF proliferation of A431 cells is abruptly and markedly decreased compared to the untreated control cultures, with little loss of cell viability over a 4-day period. This study was initiated to examine how EGF affects the progression of A431 cells through the cell cycle. Flow cytometric analysis of DNA in EGF-treated cells reveals a marked change in the cell cycle distribution. The percentage of cells in late S/G2 increases and early S phase is nearly depleted. Since addition of the mitotic inhibitor vinblastine causes accumulation of cells in mitosis and prevents reentry of cells into G1, it is possible to distinguish between slow progression through G1 and G2 and blocks in those phases. When control cells, not treated with EGF, are exposed to vinblastine, the cells accumulate mitotic figures, as expected, and show progression into S, thus diminishing the number of cells in G1. In contrast, no mitotic figures are found among the EGF-treated cells in the presence or absence of vinblastine, and progression from G1 into S is not observed, as the number of cells in G1 remains constant. These results suggest that there are two EGF-induced blocks in cell cycle transversal; one is in late S and/or G2, blocking entry into mitosis, and the other is in G1, blocking entry into S phase. After 24 hours of EGF treatment, DNA synthesis is reduced to less than 10% compared to untreated controls as measured by the incorporation of [3H]thymidine or BrdU. In contrast, protein synthesis is inhibited by about twofold. Although inhibition of protein synthesis is less extensive, it occurs 6 hours prior to an equivalent inhibition of DNA synthesis. The rapid decrease in protein synthesis may result in the subsequent cell cycle arrest which occurs several hours later.     

An artifact in measurement of S phase initiation and its implication for the kinetics of S phase-specific enzyme activities

It is shown that the different onset of S phase as measured by autoradiography vs cumulative thymidine uptake is an artifact. We consequently propose that S phase-specific enzyme activities may accumulate a few hours prior to the actual initiation of DNA synthesis. A “pre-S” DNA synthesis that can be readily detected only by autoradiography has been proposed. Published data show that DNA synthesis in cultured animal cells is initiated approx. 2 h later when measured by cumulative incorporation of [³H]thymidine ([³H]TdR) as compared with autoradiography. We show here that the difference is in reality an artifact, owing to not taking into account both gradual, asynchronous entry of cells into S phase, as well as time-dependent accumulation of radioactivity into each cell after it has entered S phase. Combination of these two factors leads to the conclusion that [³H]TdR should be incorporated approximately as the square of time following entry of the first cell into S. Taking this into account, the two methods then are in agreement, as predicted. This argument also applies to the enzyme activities shown to increase with DNA synthesis in synchronized cultures. Such an enzyme accumulation really could begin some time earlier than indicated by conventional plots of cumulative enzyme activity vs time and may, in fact, precede the onset of S by a few hours.     

DNA replication in asynchronous and synchronous ehrlich ascites cells in different conditions of growth

Ehrlich ascites tumour cells were labelled for DNA fibre autoradiography within the peritoneal cavity of a tumour-bearing mouse. The generation and the evaluation of the autoradiographic patterns is described and discussed. To study possible changes of the autoradiographic patterns during a natural S phase the labelling was performed in the mouse or in culture with asynchronous cells which were afterwards separated into synchronous subpopulations by zonal centrifugation. The subpopulations obtained were characterized by flow cytofluorometry in connection with the thymidine labelling index. We compared the DNA fibre autoradiographic patterns of several synchronous and asynchronous cell populations growing in the mouse or under different conditions in culture: The replicon size distributions of all populations examined were virtually the same. The fork movement rate was found to depend mainly on the metabolic condition of the cells. In culture it was significantly slower than in the mouse although a shortened S phase and therewith an increased DNA synthesis rate occurred. During a natural S phase it increased slightly, at most, while the DNA synthesis rate was considerably enhanced at the end of S. The changes in the rate of total DNA synthesis cannot account for the changes in the rate of chain growth. We conclude that the DNA synthesis rate is regulated almost exclusively by changing the replicon initiation frequency, while the fork movement rate is limited by the actual metabolic condition of the cells.     

In vitro bromodeoxyuridine-labelling of single cell suspensions: effects of time and temperature of sample storage

The present study was aimed to explore how the in vitro BrdUrd-labelling of rat thymocytes might be affected by both the time elapsed between obtaining the sample and the beginning of the labelling (0, 15, 30 or 60 min) and the effect of the temperature of storage (4°C versus room temperature). Single cell suspensions obtained after in vivo labelling with BrdUrd were used as controls. The S phase fraction was calculated by flow cytometry both according to BrdUrd-immunolabelling and DNA content. Immediate incubation with BrdUrd after the sample was obtained resulted in a slight decrease of the proportion of S phase cells analysed either according to DNA content or to BrdUrd-immunolabelling. Regardless of storage-temperature, the S phase fraction decreased in samples kept for 15 min or more before BrdUrd incubation. No BrdUrd-positive cells were detected in samples stored for 60 min at room temperature. This effect was related to temperature since positive cells were found when the samples were kept at 4°C during the same time period. Our results suggest that during in vitro incubation a relative loss of S phase cells exists and that a delay beyond 15 min between obtaining the sample and the in vitro labelling seriusly compromises the results of this technique.     

Comparison of isotopic and immunohistochemical methods of studying epithelial cell proliferation in hamster tongue

Our purpose was to validate different approaches to the study of cell proliferation in stratified squamous epithelia, using oral mucosa as a model. Dorsal and ventral tongue from the hamster were examined following in vivo labelling with tritiated thymidine and bromodeoxyuridine (BrdUrd), and in vitro labelling with BrdUrd. These were compared with direct immunolabelling of fixed tissue sections with monoclonal antibody PC10. For the former methods S phase cells were quantified following autoradiography or immunohistochemistry. We conclude that the proliferative status of simple, flat, lining mucosae such as ventral tongue can be derived by all three prelabelling methods and, on average, 18–19 cells per surface millimetre length were in DNA synthesis. On the other hand dorsal tongue epithelium, which is thicker, has an undulating morphology and a complex cell renewal pattern, gives different results with the three labelling methods. In both sites the proliferating cell nuclear antigen (PCNA) index was fourfold that obtained by nucleotide labelling. This is consistent with PCNA marking proliferative cells in other phases of the cell cycle in addition to the S phase. Thus, there are potential differences between the information on proliferative status derived by PCNA immunohistochemistry and other established cell cycle markers, which need to be taken into account in the interpretation of epithelial cell kinetic data in health and disease.     

Requirement for proliferating cell nuclear antigen expression during stages of the Chinese hamster ovary cell cycle

Proliferating cell nuclear antigen (PCNA/cyclin) is a nuclear protein that can stimulate purified DNA polymerase delta in vitro, and its synthesis correlates with the proliferation rate of cells. We have attempted to determine whether synthesis of PCNA/cyclin in Chinese hamster ovary cells is necessary to regulate entry into S phase. We have measured cellular PCNA/cyclin concentration of the mRNA or protein throughout the cell cycle. Cells were separated by centrifugal elutriation into populations enriched for G-1, S, and G-2/M phases. Quantitative Northern hybridization analysis was performed on RNA isolated from each cell population by using a cDNA clone of PCNA/cyclin as a probe. Results demonstrated that although intact PCNA/cyclin mRNA is present during all phases of the cell cycle, an induction of about 3-fold occurs during S phase. Two-parameter staining for PCNA/cyclin and DNA, and analysis by flow cytometry, confirmed that the quantity of PCNA/cyclin protein in the cells increases severalfold in G-1 or early S phase but generally is invariant in S and G-2/M phases. This cell cycle dependence of PCNA/cyclin expression suggests that the observed synthesis is a prerequisite for initiation of DNA replication. Introduction of an antisense oligonucleotide complementary to the PCNA/cyclin mRNA to inhibit PCNA/cyclin synthesis effectively prevented entry of G-1 phase cells into S phase. A complementary sense oligonucleotide used as a control did not have an inhibitory effect. This result suggests that a threshold concentration of PCNA/cyclin is necessary for entry into S phase.     

RELATIONSHIP OF NUTRITIONAL FACTORS TO IN VITRO TUMOR CELL GROWTH AND CYTOTOXICITY PRODUCED BY CYTOSINE ARABINOSIDE

The in vitro relationship between nutritional factors, proliferative status of tumor cells, and the cytotoxic action of cytosine arabinoside (ara-C) was investigated. The reduction in the concentration of only one essential amino acid, L-isoleucine, in the growth medium of A(T₁)Cl-3 hamster fibrosarcoma cells decreased DNA synthesis in this cell population and slowed the rate of progression of G₁ phase cells into S phase of the cell cycle. The complete omission of isoleucine from the growth medium blocked the progression of G₁ phase cells into S phase and prevented the cytotoxic action of ara-C. The addition of isoleucine to the isoleucine-deprived cells permitted these cells to enter the S phase and restored their sensitivity to the cytotoxic action of ara-C. When G₁ phase cells were placed in a medium containing reduced levels of all the amino acids and vitamins there was a prolongation of the G₁ phase. Since medium with low levels of amino acids produced a delay in the entry of G₁ phase cells into the S phase, the time interval in which these cells were most sensitive to the cytotoxic action of ara-C was different for G₁ phase cells placed in medium with adequate levels of all the amino acids. These in vitro data indicate that nutritional factors can markedly effect the proliferation of tumor cells and the cytotoxic action of ara-C.     

Quantification of the cellular proliferation on freshly dispersed cells from rat anterior pituitaries after in vivo and in vitro labelling with bromodeoxyuridine.

Summary The labelling index i.e., the proportion of cells in S phase of the cell cycle, has been calculated in cytospin preparations of rat anterior pituitary cells after labelling eitherin vivo orin vitro with the thymidine analogue bromodeoxyuridine (BrdU). The aims of this work were (1) to check whether enzymatic digestion interferes with the incorporation of BrdU into S phase cells and/or whether it has any deletereous effect on the immunohistochemical detection of cells that have already incorporated BrdU, and (2) to check the viability of simultaneous staining for BrdU and markers for the different types of pituitary cells in the cytospins. No statistical difference was found between the labelling index afterin vivo orin vitro labelling with BrdU. Identification of doubly-immunostained cells was straightforward and up to 40% of BrdU-labelled cells were immunopositive for pituitary hormones. It is suggested that cytospin preparations from biopsy samples may be used to study cellular proliferation without exposing the patient to the hazardous effects of BrdU infusion and without the interference of cell culture methods.     

MEASUREMENT OF THE KINETIC STATUS OF BONE MARROW PRECURSOR CELLS: THREE CAUTIONARY TALES

Measurements to determine the kinetic status of the morphologically unrecognizable haemopoietic precursor cells in the bone marrow are frequently carried out using techniques which inhibit or destroy cells in the DNA-synthetic (S) phase of the cell cycle. For example, tritiated thymidine (³H-TdR) has for many years been recognized as a highly specific label for DNA synthesis and, as such, administration of large doses of ³H-TdR has often been used, both in vitro and in vivo, to kill cells in S. Assay of the surviving cells has then given a measure of the proportion of the total cells which are in the S-phase of the generation cycle. Other compounds which have been used for the same purpose are: ¹²⁵Iodo-deoxyuridine (¹²⁵I-UdR), another S-phase specific label, or hydroxyurea (HU) which prevents entry of cells into S and inhibits or kills cells already in S (Sinclair, 1965). For a variety of reasons, different laboratories tend to make different choices of the agent to be used for this purpose. As a result, it has sometimes proved difficult to marry data obtained from different sources. In the course of using ³H-TdR, tritiated uridine (³H-Ur), ¹²⁵I-UdR and HU in attempts toevaluate the kinetic status of bone marrow stem cells, it has become clear that their use is not straightforward and this paper presents data which illustrate some of the pitfalls associated with their use.     

THE RELATIONSHIP BETWEEN DIURNAL VARIATION OF THE NUMBER OF CELLS IN MITOSIS AND OF THE NUMBER OF CELLS SYNTHESIZING DNA IN THE EPITHELIUM OF THE HAMSTER CHEEK POUCH

1. The numbers of cells in mitosis and in DNA synthesis in the epithelium of the hamster cheek pouch have been studied at different times of the day and night. 2. By accumulation of mitotic cells using colcemid, both the rate of entry of cells into mitosis and the duration of mitosis have been estimated at two different times of day. 3. A diurnal variation has been demonstrated in both the mitotic index and in the tritiated thymidine labelling index. Although these variations are of different amplitude and timing, the experimental data fit closely to the hypothesis that the diurnal mitotic variation is the result of a partially synchronous population moving through the DNA synthetic period. No direct action on the mitotic process need be postulated. 4. From the results of mitotic accumulation, it is clear that the rate of entry of cells into mitosis depends on the time of day at which this is studied. There is also the possibility that the duration of mitosis is slightly longer when the mitotic index is high. 5. It is concluded that, at least in the epithelium of the hamster cheek pouch, the diurnal rhythm in the number of mitoses present is a reflection of the diurnal variation in the number of cells synthesizing DNA at some time earlier. Small fluctuations in the mitotic pattern imposed by this partially synchronous population moving from S into mitosis, could be caused by slight variations in the duration of mitosis.     

Fibroblast growth regulatory factor inhibits DNA synthesis in BALB/c 3T3 cells by arresting in G1

A cell surface macromolecular component from quiescent BALB/c 3T3 mouse cells (designated fibroblast growth regulatory factor, FGRF) inhibits DNA synthesis and cell division in growing 3T3 cells. Addition of FGRF to synchronized populations of growing 3T3 cells in the late G1 or early S phase did not inhibit DNA synthesis in the immediate S phase. However, a significant inhibition was observed in the S phase of the next round of cell cycle. Cells exposed to the regulatory factor in late S/early G2 or early G1 showed reduced DNA synthesis in the upcoming S phase; the late S/early G2 cells were more sensitive to inhibition than the cells in the G1. Further, the regulatory factor delayed the progression of G0/G1-arrested cells into the next S phase. These results suggest that the physiological effect of FGRF is to arrest cells in early G1, thus preventing their entry into a new round of cell cycle. In contrast to untransformed 3T3 cells, mouse cells transformed by SV40 were not subjected to growth-arrest by the regulatory factor, although the transformed cells contain active FGRF that inhibits DNA synthesis in growing 3T3 cells.     

THE EFFECT OF POLYCYTHAEMIC SERUM ON THE PROLIFERATION OF RAT BONE MARROW CELLS IN VITRO

The effect of experimental polycythaemia on the rate of proliferation of erythrocytic precursor cells was investigated by means of an in vitro technique. The serum obtained from polycythaemic rats was found to inhibit significantly ³H-thymidine incorporation in normal rat bone marrow cells in vitro, as compared with normal serum. Autoradiographic analysis revealed that this inhibition resulted from a reduction in the number of labelled bone marrow cells. The inhibition proved to be specific to the erythrocyte precursor cells; the labelling index was reduced in the erythrocytic cell population by 21–50% (P < 0.001) at different incubation times, while the effect on the granulocytic cell population was negligible. It is deduced that an inhibitor substance responsible for the effects observed is present in polycythaemic serum. It is proposed that this factor is the ‘erythrocytic chalone'. The results support the general view that triggering of stem cells is not the only mode of regulation of erythropoiesis, but that the rate of proliferation of the precursor cells in the erythron is also regulated.