J. Cell Sci. 59, 1-12 (1983)Printed in Great Britain © Company of Biologists Limited 1983

EFFECTS OF TEMPERATURE ON THE REPLICATIONOF CHROMOSOMAL DNA OF XENOPUS LAEVIS CELLS

A. A. AL-SALEHDepartment of Zoology, King Saud University, Riyadh, Saudi Arabia

SUMMARY

DNA fibre autoradiography has been used to study the effects of temperature on the replicationof chromomosomal DNA of Xenopus laevis cells in tissue culture at 18, 23 and 28°C. Pulse/stepdown labelling shows that the DNA replicates bidirectionally. Origin-to-origin distances (initia-tion intervals) vary, but the range of and the mean initiation intervals at all three temperatures aremuch the same. The mean interval between initiation points is of the order of 60 to 66 ym. Stagger-ing of initiation is evident at all three temperatures. Evidence against the existence of replicationtermini is provided. The rates of progress of DNA replication forks are 6/im/h at 18°C, 10/un/hat 23 °C and 16/im/h at 28°C.

INTRODUCTION

Autoradiographic studies on whole cells have shown that chromosomes do notreplicate their DNA sequentially from end to end as a single unit, but instead replicateat many independent regions or sites (Taylor, 1960; Lima-de-Faria, 1961; Painter,1961; Stubblefied & Muellar, 1962; Moorhead & Defendi, 1963; Schmid, 1963;German, 1964; Painter, Jermany & Rasmussen, 1966). The most convincing in-formation from whole cells concerning multiple sites of DNA synthesis perchromosome comes from autoradiographic studies of the polytene chromosomes oflarval Diptera (Plaut & Nash, 1964; Rudkin, 1972).

Cairns (1962, 1963, 1966) invented an autoradiographic technique that makes itpossible to visualize labelled DNA fibres in the light microscope. This technique isbased on a gentle dialysis, which aims at obtaining long fibres of DNA from labelledcells. Cairns' observations showed that the chromosome of Escherichia colt consistsof one circular double-stranded molecule, which replicates from a single 'origin',whereas HeLa cell chromosomes replicate from many origins. The same conclusionwas reached by Huberman & Riggs (1966), for Chinese hamster cells in culture. It hasnow become well-recognized that the chromosomal DNA of eukaryotes replicatesbidirectionally from many initiation sites, which are tandemly arranged along theDNA double helices (Huberman & Riggs, 1968; Amaldi, Carnevali, Leoni & Mariot-ti, 1972; Callan, 1972; Hand & Tamm, 1972, 1973; Weintraub, 1972; McFarlane &Callan, 1973; Kriegstein & Hogness, 1974; Van't Hof, 1975).

In eukaryotic cells generally, replication of DNA occurs during a particular phaseof the cell cycle - the S phase (Swift, 1950; Walker & Yates, 1952; Howard & Pelc,1953; Lajtha, Oliver & Ellis, 1964). As this phase is one of the phases dependent ontemperature for its duration (Sisken, Morasca & Kibby, 1965; Rao & Engelberg,

2 A. A. Al-Saleh

1966; Watanabe & Okada, 1967; Saladino & Johnson, 1974; Al-Saleh, 1977), thereis a clear indication of some relationship between the overall rate of DNA replicationand temperature. One of the aims of this study has been to see whether differences inculture temperature, which affect S-phase duration (Al-Saleh, 1977), affect thereplication rate, initiation sites and initiation intervals of Xenopus somatic cells inculture.

MATERIALS AND METHODS

The experiments were carried out on an established cell line (1-6) of Xenopus laevis kidney cellsoriginally produced by Dr K. A. Rafferty (1969). Replicate cultures of cells were grown at 28°C,23°Cand 18 °C in a complete amphibian medium (Al-Saleh, 1977). Subcultures were started with5 X 104 cells/ml of medium in Petri dishes (60 mm in diameter). Each dish contained 4 ml of freshmedium. The dishes were arranged inside a desiccator with a small quantity of distilled water in thebottom to maintain humidity. The desiccator was sealed and incubated for 48 h at 28 °C and 23 °C,and twice as long at 18 CC. When cells had reached logarithmic growth, 5-fluorodeoxyuridine(FdUrd) and undine (Urd) were added to each dish to give a final concentration of 1 Jig/ml FdUrdand 0-5/ig/ml of Urd and the dishes were incubated for a further 20h at 28°C and 23°C, and fora further 40 h at 18 °C. This step was applied to arrest most of the cells at the beginning of S-phase.After the treatment with FdUrd/Urd, the medium was poured off all dishes and 2 ml of freshmedium containing 50/iCi/ml of [3H]thymidine (26Ci/mmol) were added to each dish and thecells were labelled for 1 h at 28°C or for 2h at 23 °C and 18 °C.

In experiments referred to as chased or pulse-stepdown labelling, after the cells had been labelledfor a determined period, sufficient non-radioactive thymidine was added to the original medium soas to reduce the specific activity of [3H] thymidine to one quarter of its original level and the cellswere left to grow for a further determined period. Cells grown at 28 CC were labelled for 1 h followedby a stepdown of 1 h. Cells grown at 23 °C were labelled for 2 h followed by a stepdown of 2 h, whilecells grown at 18 °C were labelled for 4 h followed by a stepdown of 4 h. After labelling, the cells weretrypsinized, harvested and DNA fibre autoradiographs were prepared as described by Callan(1972). The measurements were made directly from preparations using bright-field illuminationwith a Zeiss microscope, with 12-5X eyepieces, one containing a micrometer scale, Optivarat 1-25,and 40x planapochromat oil-immersion objective. This gave a magnification in which 50 oculardivisions were equal to 100 fim.

RESULTS

Initiation intervals

Initiation sites or origins are places where DNA synthesis begins and initiationintervals are distances between neighbouring origins. In pulse-stepdown labellingexperiments, initiation intervals are most simply measured from midpoint to mid-point of adjacent dense stretches of silver grains where both dense stretches areflanked by tails (i.e. regions of diminished grain density) at either end.Autoradiographs of this kind are shown in Figs 1, 4 and 6. Where DNA synthesis hadalready started prior to the provision of [3H]thymidine and continued after labelling,the resultant autoradiographs show two dense stretches of silver grains with tails ofdiminished grain density towards the 'outside', and a sharply demarcated gap betweenthe dense stretches of silver grains towards the 'inside'. Such a gap represents thelength of DNA replicated prior to labelling, and its midpoint may be taken as anorigin. This gives further possibilities for measuring initiation intervals and examplesare shown in Figs 2, 3 and 5 (see Fig. 7 for further illustration).

Eflects of temperature on DNA replication in Xenopus cells

4 A. A. Al-Saleh

Initiation intervals can be determined accurately only from pulse-stepdown label-ling regimes, because only in such preparations can one identify the direction ofreplication. This permits a clear distinction to be made between long labelledstretches that owe their length to fusion between neighbouring forks and those thatowe their length to replication having initiated when labelling started, and thenproceeded throughout the labelling. Examples are shown in the middle stretch of Fig.3 and the two replicating units of Fig. 1.

It is clear from Fig. 7 (A, B, C) that the initiation intervals vary, ranging from 26to 136Jim (A), 10 to 144jz (B) and 18 to 136 jzm (c), but the mean is more or less thesame, 55 to 66jzm, at the three different incubation temperatures. Callan (1972) hasmeasured the initiation intervals of this cell-line in culture at 25 °C. He found that theintervals range from 18 to 128 /im, about a mean of 60/im. Comparing the results ofCallan at 25 °C with mine at 18 °C and 23 °C, it seems reasonable to assume that themean initiation intervals at 28 °C is also around 60Jim, and if this is so then theinitiation intervals are evidently not influenced by culture temperature, despite theinfluence of temperature on duration of S phase (Al-Saleh, 1977).

Direction ofDNA replication

It has been generally accepted that pulse-stepdown labelling provides directevidence as to the direction in which DNA synthesis proceeds. DNA fibreautoradiographs from cells that have been subjected to two successive pulses of[ H]thymidine, the first high and the second low in specific activity, show dense graintracks (where replication started during the first pulse) flanked by tails of declininggrain density, as a result of dilution of the specific activity during the stepdown. Ifreplication were to proceed in one direction, the dense labelled tracks would show tailsat one end, and if replication is bidirectional the tracks would show tails at both ends.

The results of this study fully substantiate the view that the replication ofchromosomal DNA olXenopus proceeds bidirectionally (Figs 1, 4, 6). The majorityof the origins that one sees had initiated only after the FdUrd block had been removed,and these show the expected tails at both ends. There are, however, a minority oforigins, coming from cells that were already in S phase when the FdUrd was applied,where replication was already in progress prior to the provision of [3H]thymidine. Inthese, provided fusion between neighbours does not occur during the pulse period athigh specific activity, they show up in autoradiographs as dense tracks with a tail atone end only. Far more frequently than chance would allow, one sees such tracks withtails 'back-to-back'; i.e., with both tails directed towards the outside (Figs 2, 3, 5).Such tracks equally confirm that replication proceeds bidirectionally from the origins.

Further evidence that replication occurs bidirectionally from origins comes fromtracks that have undergone fusion, or better still near-fusion, between adjacent unitsduring the stepdown (Figs 1, 3, 6). These regularly show fusion occurring between'tails', towards the outside of which are dense tracks. The point to emphasize is thatthe fusions are tail-to-tail, and not between the tail of one track and the dense regionof an adjacent track. If the latter were found, this would indicate that at least one ofthe fusing tracks had originated from an initiation site from which replication had

Effects of temperature on DNA replication in Xenopus cells80

60

40

20

80

60

20

80

60

40

20

Initiation interval lengths

Fig. 7. A, B, c are comparable histograms of the frequency distributions of initiationintervals of measured DNA fibre autoradiographs of Xenopus cell cultures at 18 °C and28 C, respectively. The cells were pretreated with FdUrd and labelled for 4 h followed bya stepdown of 4 h at 18 °C (A) ; labelled for 2 h followed by a stepdown of 2 h at 23 °C (B) •and labelled for 1 h followed by a stepdown of 1 h at 28 °C (c).

proceeded in one direction only; that they are not found is strong evidence against thenotion that some origins initiate replication in one direction only.

Fig. 1 is a particularly clear example, in which fusion had evidently occurred justat the time of harvesting. One can infer the position of initiation site 1, while initiationsite 2 is marked by slight separation of the sister duplexes. The tails are symmetricalabout these two origins, so the less-dense track, which includes the point of fusion,must surely represent the fusion of tails, not the fusion of one exaggeratedly long tailcoming from, say origin 1, and the dense track coming from origin 2.

Pulse-stepdown autoradiographs show not only sister-strand separation after acertain time (Figs 1, 5), but also give a clear idea concerning the relative times ofinitiation of neighbouring origins and whether or not termini exist. Figs 2 and 3 showtracks where replication had already initiated before labelling and continued duringthe 8h of labelling (4h high + 4h low at 18°C), with unlabelled gaps bounded byheavily labelled tracks with tails to the outside. The arrows indicate the location ofinitiation sites, which are assumed to be at the middle of each gap. Sites 1 and 2 of Fig.3 appear as though they had initiated at the same time because the lengths of the twounlabelled gaps are the same. A particularly good example of staggering in the activa-tion of initiation sites can be seen in Fig. 5. This photograph was taken from a DNAfibre autoradiograph of Xenopus cells in culture at 23 °C. The cells were labelled with

6 A. A. Al-Saleh

[ H]thymidine for 2 h followed by a stepdown of 2 h. It shows that site number 1 hadinitiated before 2 and 3, and that site 3 had initiated before 2. Fig. 5 also shows a goodexample of sister-strand separation, bidirectional replication and fusions, which hadtaken place between replicating tracks proceeding from sites 1 and 2 and between 2and 3 during the first 2h labelling period, i.e. before the stepdown. Termini areassumed to be points where a replication fork waits until a neighbouring replicatingfork comes into its proximity, and there has been some discussion as to whether or notthey exist (Huberman & Riggs, 1968; McFarlane & Callan, 1973). If termini exist,then DNA fibre autoradiographs from cells subjected to stepdown labelling might beexpected to show, at least occasionally, dense labelled tracks where one has a tail andthe other none, or at least evident asymmetry. But generally the results of this studyargue against the existence of termini; a replication fork seems to proceed withoutinterruption until it meets a neighbouring fork coming from the opposite direction.Fig. 5 provides strong evidence against the existence of termini symmetrically spacedon either side of an origin, which was the early claim of Huberman & Riggs (1968).

Replication rate

The replication rate of DNA cannot be determined from pulse/stepdown prepara-tions because tails do not have precise limits. Therefore, replication rate must beestimated from pulse-labelled cells. In order to make a realistic estimate of the rateof progress of replication forks one must choose a period of labelling so restricted thatfew or no fusions have occurred between any of those neighbouring units that initiatedduring the labelling period. So, labelled tracks chosen for measurements were selectedfrom regions of the autoradiographs where the fibre density was low and where atandem arrangement of three or more tracks could be clearly recognized.

Figs 8, 9 and 10 are photographs of DNA fibre autoradiographs from cells labelledfor 2 h and incubated at 18 °C, 23 °C and 28 °C, respectively. These photographs showthat the lengths of the labelled tracks increase as a consequence of increasing theincubation temperature from 18 °C to 28 °C. In the same way Figs 11, 12 and 13 arephotographs of DNA fibre autoradiographs from cells labelled for 4h and incubatedat 18 °C, 23 °C and 28 °C, respectively. These photographs also indicate that labelledtrack lengths are temperature-dependent. It is clear in Fig. 13 that fusion has occurred

Figs 8-10. Photographs of DNA fibre autoradiographs prepared from Xenopus cells intissue culture, and chosen as being typical of tracks resulting from cultures labelled for 2 hat 18 °C (Fig. 8), at 23 °C (Fig. 9) and at 28 °C (Fig. 10). Labelling was in all cases precededby treatment with FdUrd and the exposure times were 8 to 11 months. The threephotographs show clearly how track lengths are dependent on culture temperature,provided the time of labelling chosen is sufficiently short to prevent fusions occurringbetween replicating segments that initiated during the labelling period.

Figs 11-13. Photographs of DNA fibre autoradiographs prepared from Xenopus cells intissue culture, and chosen as being typical of tracks resulting from cultures labelled for 4 hat 18°C (Fig. 11), at 23°C (Fig. 12) and at 28°C (Fig. 13). Labelling was in all casespreceded by treatment with FdUrd, and exposure times were 8 to 11 months. The verylong track in Fig. 13 (where arrows indicate sister-strand separation) is entirely typical offusions obtained after labelling for 4h at 28°C. Bar, 100/an.

Effects of temperature on DNA replication in Xenopus cells

8

c *

12

Figs 8-13

8 A. A. Al-Saleh

between at least two replication units, because of the presence of a long uninterruptedtrack with two sister-strand separated regions (see arrows).

Fig. 14 shows comparable histograms of the frequency distributions of labelledtrack lengths measured from pulse-labelled preparations. Fig. 14A (2h of labelling)shows a steep rise and an equally steep fall in the 18-24 /im tracks, with relatively few(20) tracks longer than this. These may be accounted for either by occasional fusionsbetween replicating units converging from particularly close origins, or as aconsequence of a replication fork rate that is higher than normal. However they maybe explained, I think that they can be ignored and that one should take the positionof the steeply falling right-hand shoulder (RHS) of the histogram as a realisticmeasure of two-way replication, which has progressed in 2h; this is 24/im, giving aone-way rate of 6/im/h. It is manifestly not valid to use the mean track length of thishistogram (17/im) as a measure of the two-way rate; first, because the histogramnecessarily includes one-way tracks and also two-way tracks that initiated late. Fig.14B (2 h of labelling) shows that the length of labelled tracks range from 10 to 96 /im,with a mean, labelled length of about 32/im. Choosing the RHS rather than the mean(for the reasons previously mentioned) for estimating replication rate, this wouldappear to be about 40 /im, i.e. the one-way replication rate at this temperature is about10/un/h. Fig- 14c (1 h of labelling) is well-peaked, and the RHS falls at about 32 /im.This gives an estimate for the one-way replication rate of 16/im/h.

100

o 60

sr 40Li.

20

i120

100

g 80

| 60

* 40

20

160

140

120

S" 100a3 80

s£ 60

40

20

:J1

^ n tr v co o)

r- CN •* in r» 35

Labelled lengths (/im)

Fig. 14. A, B, c are comparable histograms of the frequency distributions of labelled tracklengths as measured from DNA fibre autoradiographs prepared from FdUrd-treated cellscultured at 18 °C, 23 °C and 28 °C, respectively. The cells were labelled for 2h at 18 °C(A) and 23 °C (B), and 1 h at 28 °C (c).

Effects of temperature on DNA replication in Xenopus cells 9

The conclusion is that DNA replication rate is temperature-dependent and theestimated one-way rate is 6(im/h at 18°C, 10/xm/h at 23°C and 16/im/h at 28°C.

DISCUSSION

Pulse/stepdown preparations indicate that the mean initiation intervals inXenopuscells in culture do not vary significantly with temperature. The mean initiation inter-vals for cells cultured at 18 °C and 23 °C is the same, of the order of 66/im. Cellscultured at 28 °C have a mean of about 55 /im. Callan (1972) has found that the meaninitiation intervals for Xenopus cells in culture at 25 °C is about 60/im. Callan'sfindings have encouraged me to assume that the mean distance between adjacentactive sites is about 60/im.

The 5-phase durations at 18°C, 23 °C and 28°C are 29-5h, 15-5 h and 135 h,respectively (Al-Saleh, 1977). If the initiation intervals are the same at all threetemperatures, then either varying rates of fork progression, or varying degrees ofstaggering of the activation of origins, or both together, must be responsible for thetemperature-dependence of 5-phase duration.

Different tissues may well have different numbers of active sites in relation to thediffering durations of their 5 phases (Callan, 1972; Hand & Tamm, 1974). Callanproposed that variation in DNA/ histone packing within the chromosomes of differenttissues might be responsible for the variation in the number of active origins. Hepointed out, for example, that in Triturus spermatocyte nuclei, where much of thechromatin is densely packed, a smaller number of sites may be exposed to initiatorenzyme(s) than in typical somatic nuclei. Where the DNA/histone packing probablydoes not change, as in Xenopus tissue-culture cells over the range 18 °C to 28 °C, thenthe number of exposed initiation sites should remain constant.

Variable staggering in the initiation times of active origins might be another factorresponsible for variation in duration of 5-phase in the course of development or indifferent types of cells (Amaldie/ al. 1973; Callan, 1973; McFarlanefc Callan, 1973).Staggering of initiation times has been found for Xenopus cells at all three tem-peratures, 18 °C, 23 °C and 28 °C, and a particularly good example of this is shown inFig. 3.

The most likely single factor responsible for decreasing duration of 5 phase withincreasing temperature is the rate of progress of replication forks. Table 1, and Fig.15 show how 5-phase durations, fork progression rates and incubation temperatures

Table 1. A comparison of data concerning S-phase duration, replication rate andinitiation intervals in Xenopus cells cultured at different temperatures

S-phase Replication Range of Mean ofTemperature duration rate one-way initiation initiation

(°C) (h) (fjm/h) intervals (/an) intervals (/an)

18 29-5 6 26-136 6623 15-5 10 10-144 6628 13-5 16 18-136 55

CEL59

10 A. A. Al-Saleh

- 1 5 • 15 -

10

5

rat

ion

icat

RepI

10

5

18 23 28

Temperature (°C)

10 20S-phase duration (h)

30

Fig. 15. Shows the relationship between the S-phase duration (•) , replication rate (•)and temperature.

are related. Table 1 also shows that temperature does not effect the initiation intervals.Chibon's (1973) results show that in various tissues of Pleurodeles, the lengths of thecell-cycle phases are temperature-dependent. He proposed that the extension of 5phase at low temperature might be due to a decreased production of enzymes andprecursors necessary for DNA synthesis. Saladino & Johnson (1974) showed that inChinese hamster cells S phase is related to the rate of DNA synthesis and is a functionof incubation temperature.

Fork progression rate, the number of active sites, staggering in the sequence ofinitiation times, the rates of synthesis of proteins required as initiation factors and forchain elongation are all of consequence during the S phase. Also, the concentrationsof precursors, the relative diffusion rates of small and large molecules between thecytoplasm and the nucleus, and how these relate to temperature, all govern themetabolism of the cell and therefore the duration of the S phase. The results presentedhere show that it is the rate of fork progression that is influenced by temperature.Whether or not this effect is a direct expression of the altered mobility of the enzymesor whether it is because of one or several possible secondary effects that have not beendetermined.

I wish to thank Professor H. G. Callan for advice and guidance in this work.

REFERENCES

AL-SALEH, A. A. (1977). The temperature-dependence of cell cycle parameters and chromosomalDNA replication in tissue cultures of Xenopus laveis. Ph.D. thesis, St Andrews University.

AMALDI, F., CARNEVALI, F., LEONI, L. & MAKIOTTI, D. (1972). Replicon origins in Chinesehamster cell DNA. 1. Labelling procedure and preliminary observations. Expl Cell Res. 74,367-374.

Effects of temperature on DNA replication in Xenopus cells 11

CAIRNS, J. (1962). Minimum estimate for the length of the DNA of Escherichia coli byautoradiography. J. molec. Biol. 4, 407-409.

CAIRNS, J. (1963). The bacterial chromosome and its manner of replication as seen byautoradiography. J. molec. Biol. 6, 208-213.

CAIRNS, J. (1966). Autoradiography of HeLa Cell DNAJ. molec. Biol. 15, 372-373.CALLAN, H. G. (1972). Replication of DNA in the chromosomes of eukaryotes. Proc. R. Soc. B,

181, 19-41.CALLAN, H. G. (1973). Replication of DNA in eukaryotic chromosomes. Br. med. Bull. 29,

192-195.CHIBON, P. (1973). Cell proliferation in the late embryos and young larvae of the newt Pleurodeles

waltlii Michah. In The Cell Cycle in Development and Differentiation (ed. M. Ball & F. Billet),pp. 257-277. Cambridge University Press.

GERMAN, J. L. (1964). The pattern of DNA synthesis in the chromosomes of human blood cells.J. Cell Biol. 20, 37-55.

HAND, R. & TAMM, I. (1972). Rate of DNA chain growth in mammalian cells infected withcytocidal RNA virus. Virology 47, 331-337.

HAND, R. & TAMM, I. (1973). DNA replication: direction and rate of chain growth in mammaliancells. J. Cell Biol. 58, 410-418.

HAND, R. & TAMM, I. (1974). Initiation of DNA replication in mammalian cells and its inhibitionby reovirus infection. J. molec.Biol. 82, 125-183.

HOWARD, A. & PELC, S. R. (1953). Synthesis of deoxyribonucleic acid in normal and irradiatedcells and its relation to chromosome breakage. Heredity, Land, (suppl.) 6, 261-273.

HUBERMAN, J. A. & RIGGS, A. D. (1966). Autoradiographs of chromosomal DNA fibers fromChinese hamster cells. Proc. natn. Acad. Sci. U.SA. 55, 599-606.

HUBERMAN, J. A. & RIGGS, A. D. (1968). On the mechanism of replication in mammalianchromosomes. .7. molec. Biol. 32, 327-341.

KRIEGSTEIN, H. J. & HOGNESS, D. S. (1974). Mechanism of DNA replication in Drosophilachromosomes: Structure of replication forks and evidence for bidirectionality. Proc. natn. Acad.Sci. U.SA. 71, 135-139.

LAJTHA, L. G. OLIVER, R. & ELLIS, F. (1964). Incorporation of 32P and adenine HCin vitro. Br.J. Cancer 8, 367-379.

LIMA-DE-FARIA, A. (1961). Initiation of DNA synthesis at specific segments in meioticchromosomes of Melanoplus. Hereditas 47, 647-694.

MCFARLANE, P. W. & CALLAN, H. G. (1973). DNA replication in the chromosome of the chicken,Callus domesticus.jf. Cell Sci. 13, 821-839.

MOORHEAD, D. S. & DEFENDY, V. (1963). Asynchrony of DNA synthesis in chromosomes ofhuman diploid cells, jf. Cell Biol. 16, 202-209.

PAINTER, R. B. (1961). Asynchronous replication of HeLa S3 chromosomal deoxyribonucleic acid.jf. biophys. biochem. Cytol. 11, 485-508.

PAINTER, R. B. JERMANY, D. A. & RASMUSSEN, R. E. (1966). A method to determine the numberof DNA replicating units in cultured mammalian cells. J. molec. Biol. 17, 47-56.

PLAUT, W. & NASH, D. (1964). Localized DNA synthesis in polytene chromosomes and its implica-tion. In The Role of Chromosomes in Development (ed. M. Locke), pp. 133-135. New York,London": Academic Press.

RAFFERTY, K. D. JR (1969). Mass culture of amphibian cells: Methods and observations concerningstability of cell type. In Biology of Amphibian Tumours (ed. M. Mizell), pp 52-58. New York:Springer-Verlag.

RAO, P. N. & ENGELBERG, J. (1966). Effects of temperature on the mitotic cycle of normal andsynchronized mammalian cells. In Cell Synchrony (ed. T. L. Cameron & G. M. Padilla), pp.332-352. New York, London: Academic Press.

RUDKIN, G. T. (1972). Replication in polytene chromosomes. In Results and Problems in CellDifferentiation, vol. 4, pp. 59-85. Berlin, Heidelberg, New York: Springer-Verlag.

SALADINO, C. F. & JOHNSON, H. A. (1974). Rate of DNA synthesis as a function of temperaturein cultured hamster fibroblasts (V-79) and HeLa-S3 cells. Expl Cell Res. 85, 248-254.

SCHMID, W. (1963). DNA replication pattern of human chromosomes. Cytogenetics 2, 175-193.SISKEN, J. E., MORASCA, L. & KIBBY, S. (1965). Effects of temperature on the kinetics of the

mitotic cycle of mammalian cells in cultures. Expl Cell Res. 39, 103-116.

12 A. A. Al-Saleh

STUBBLEFIELD, E. & MUELLER, G. C. (1962). Molecular events in reproduction of animal cells. II.The localized synthesis of DNA in the chromosomes of HeLa cells. Cancer Res. 22, 1091-^1099.

SWIFT, H. (1950). The constancy of deoxyribonucleic acid in plant nuclei. Proc. natn. Acad. Sci.U.SA. 36, 643-653.

TAYLOR, J. H. (1960). Asynchronous duplication of chromosomes in cultured cells of Chinesehamster. J. biophys. biochem. Cytol. 7, 455-464.

VAN'T HOF, J. (1975). DNA fiber replication in chromosomes of higher plants (Pisum sativum).Expl Cell Res. 93, 95-104.

WALKER, P. M. B. & YATES, H. B. (1952). Nuclear components of dividing cells. Proc. R. Soc. B,140, 274-299.

WATANABE, I. & ODADA, S. (1967). Effects of temperature on growth rates of cultured mammaliancells (L5 178Y).J. CellBiol. 32, 309-323.

WEINTRAUB, H. (1972). Bi-directional initiation of DNA synthesis in developing chickerythroblasts. Nature, newBiol. 236, 195-197.

(Received 7 April 1982)