Cell Tissue Kinet. (1976) 9,157-165.

CHARACTERISTICS O F T H E ISOPRENALINE STIMULATED PROLIFERATIVE RESPONSE O F R A T SUBMAXILLARY G L A N D K. W. KOSCHEL,G . S. HODGSONA N D J . M. R A D L E Y Biological Research Unit, Cancer Institute, Melbourne (Received 20 May 1975;revision receiued 22 August 1975) ABSTRACT

A simple stochastic model has been developed to determine the cell cycle kinetics of the isoprenaline stimulated proliferative response in rat acinar cells. The response was measured experimentally, using 3H-TdR labelling of interphase cells and cumulative collections of mitotic cells with vincristine. The rise and fall of the fraction of labelled interphase cells and of metaphase cells is expressed by the product of the proliferative fraction and a difference of probability distributions. The probability statements of the model were formulated and then compared by an iterative fitting procedure to experimental data to obtain estimates of the model parameters. The model when fitted to the combined fraction labelled interphase (FLIW) and fraction metaphase (FMW,) waves gave a mean G,, transit time of 21-2 hr, mean G,, S transit time of 27-0 hr, and mean Gi, S + G , transit time of 35-8 hr for a single injection of isoprenaline, where G,, is the initiation to S phase time. When successive injections of isoprenaline were given at intervals of 24 and 28 hr the corresponding values after the third injection were 12.4 hr, 20.8 hr and 25.7 hr respectively. The variance of the Gi, phase dropped from 18.1 to 1.3 while the other variances remained unchanged. The estimated proliferative fraction was 0.24 after a single injection of isoprenaline, and 0.31 after three injections of the drug. Independently determined values of the proliferative fraction, obtained from repeated 3H-TdR injections, were 0.21 and 0.36 respectively.

+

+

A single injection of isoprenaline stimulates a wave of DNA synthesis amongst the acinar cells of the rat submaxillary gland, beginning about 15 hr later and reaching a peak at about 25 hr (Barka, 1965; Radley, 1968). Repeated administration of the drug at daily intervals results in further waves of DNA synthesis with peak 3H-TdR labelling occurring about 15 hr after each injection (Radley, 1968; Barka, 1970). Similar findings have been reported for mouse salivary glands (Whitlock, Kaufman & Baserga, 1968; Novi & Baserga, 1971). Correspondence: Dr K. W. Koschel, Biological Research Unit, Cancer Institute, 278 William Street, Melbourne, Australia. 157

158

K . W. Koschel, G . S. Hodgson and J. M . Radley

The shape of the wave of interphase labelling with 3H-TdR is dependent on two factors, the time taken for cells to enter the S phase after application of the stimulus (designated Gi, to indicate the gap between initiation and S phase), and the duration of the S phase. The question arose as to whether it might be feasible to estimate these parameters given the shape of the labelling index wave. Cell cycle parameters of exponentially growing populations can be measured using FLM analysis (Mendelsohn & Takahashi, 1971); however, in the present situation cells are stimulated to undergo only one division in response to an injection of isoprenaline and FLM analysis is not appropriate. A search for a suitable method led to the derivation of the model described in the appendix. The characterization of the kinetics of growth of stimulated submaxillary glands was achieved by coupling the model with iterative parameter estimation. The data from the early studies cited above were considered insufficiently detailed for analysis and so the time course of 3H-TdR labelling of submaxillary acinar cells following isoprenaline was re-examined, both after a single injection of the drug and after three injections at daily intervals. In addition, to better define the kinetic parameters, estimates of the frequency of mitosis were obtained using vincristine arrest of cells in metaphase (Radley & Hodgson, 1973). From these data, a simple model of the response of the acinar cells of the rat submaxillary gland was defined in terms of cell cycle parameters. MATERIALS A N D METHODS Canberra Black female rats, aged 8 weeks and weighing approximately 140 g, were used for all experiments. A single injection of isoprenaline was given intraperitonealiy at 9 p.m. The drug (isoprenaline sulphate) was freshly dissolved, and given as a dose of 10 mg in 0.2 ml saline. At 4-hourly intervals, beginning at 16 hr after the injection of isoprenaline, a group of seven rats were injected intraperitoneally with 50 pCi 3H-TdR (Radiochemical Centre, Amersham; specific activity 5 Ci/mmol) followed 10 min after by vincristine (0.5 mg/kg body weight; Oncovin, Lilly). The rats were killed 4 hr afterwards, and the submaxillary glands removed and fixed in 10 % neutral formalin. Autoradiographs were prepared from 3 ,um thick paraffin sections by the dipping method, using Kodak NTB2 emulsion, and after processing were stained with haematoxylin. The degree of labelling was determined by scoring lo00 acinar cells per rat. Each autoradiograph was also scored for the number of metaphases per lo00 acinar cells. The size of the proliferative fraction was obtained from autoradiographs derived from a group of seven rats which had been given repeated 3H-TdR injections for the duration of the DNA synthesis wave. The rats were given 25 pCi 3H-TdR intraperitoneally starting at 16 hr after the drug, and this was repeated every 4 hr until the rats were killed at 52 hr. Experiments were carried out to determine the extent of nuclear doubling following mitosis of the proliferating cells. Twenty rats were given a single injection of isoprenaline as before, and at the time of peak labelling, determined from the above studies, were given an injection of 50 pCi 3H-TdR. The rats were then divided into two groups of ten, one of which was killed 1 hr later, the other 24 hr later when labelled nuclei would be expected to have divided. Autoradiographs were prepared and the percentage of labelled acinar cells determined. Similar experiments to those described above were carried out after rats had been given three injections of isoprenaline. For these experiments the first two injections were given

Isoprenaline stimulated proliferative response

159

24 hr apart and the third injection 28 hr after the second. The third injection was always given at 9 p.m. to avoid the possibility of diurnal variation influencing the results (Burns, Scheving & Tsai, 1972). Details of these experiments are described in the results. RESULTS Experimental points describing the fraction labelled interphase wave (FLIW) and fraction accumulated metaphase wave (FMW,) are shown in Fig. 1, (a) and (b), for the single injection and three injection data respectively. Following a single injection of isoprenaline the fraction of labelled cells is highest after 24 hr with a peak in the corresponding metaphase accumulation data at 36 hr. Values obtained after three injections of isoprenaline are markedly different to those for a single injection, with peaks at 16 and 24 hr respectively. The data in Fig. 1 were used to derive parameter estimates of the cell cycle phases using the method outlined in the appendix. These parameter estimates are shown in Table 1 for both single and for three injections of isoprenaline. In each case, the data for FLIW and FMW, curves were combined and parameter estimates were determined in a single fitting procedure. The best fit model response curves associated with these estimates are shown in Fig. 1. An estimate of the proliferative fraction (PF) obtained by fitting the model to data is included in Table 1. For comparison, the table also shows two experimentally determined values, one obtained from a group of rats given repeated 3H-TdR injections, and the other

e

.-

LL c

o.21 0

4

8

1 2 1 6 2 Time (hr)

FIG.1. Experimentally determined fraction of 3H-TdR labelled acinar cells (x) and fraction of metaphase arrested cells (0) at times following (a) a single, and (b) three injections of isoprenaline (kSE). Best fit model curves are shown for the fraction labelled interphase wave (-) and for the accumulated fraction metaphase wave (----).

K . W . Koschel, G . S . Hodgson and J. M . Radley

160

TABLE 1. Best fit parameter estimates (hours) for FLIW and FMW. response curves (+SE) Doubling Isoof nuclei prenaline p~~~ $G,. at mitosis injections

kis ts

$G,,

+s

PF p~~~+ s + c2 P’G,, + s + c2 estiPF PF mated ’H-TdR mitosis

~~

No

One Three

Yes

One Three

21-2 18-1 +1-1 k5-5 12-4 1.3 k0.2 +1-4

27.0 +1-2 208 k1.5

68.3 512.1 70.4 k23.2

35.8 k0.4 25-7 53.7

27.2 +2-8 26.2

21.6 21.3 f1.2 k5.4 12.5 1.7 f0.2 +1-8

27-0 +1-3 209 +la3

71.7 k13.0 81.2

36.1 k0.4 26.2 20.7

27.1 +2-3

k9.5

f8.5

21.8

k4-5

0.24 0.21 f002 k0.03 0.31 0.36 a 0 6 k0.05

0.25

0-27 f0.03 0-34 k0.05

0.27

0.34

0.39

obtained by summation of mitotic fractions for all time points. The contribution of cells stimulated to enter DNA synthesis by the second isoprenaline injection to the proliferative fraction observed after the third dose of the drug was examined. A group of rats were given the first two injections of isoprenaline but not the third, and were then given repeated injections of 3H-TdR at the same times as the group given three injections of the drug. The percentage of labelled acinar cells in this group was found to be 0.9% (k0.26 SE) from which it may be concluded that cells stimulated to proliferate by the second isoprenaline injection did not contribute significantly to the labelling observed after the third injection of the drug. Another group of animals which did not receive isoprenaline at all but were labelled repeatedly with 3H-TdR as above had 1.0% (k0-25) acinar cells labelled. It will be noted in Table 1 that two sets of results have been calculated for the growth parameters, one assuming that no increase in the number of cell nuclei takes place as a result of mitosis, the other that nuclear doubling always occurs. The latter values were calculated after correcting each data point for the increase in the number of nuclei up to that time which would be expected to have occurred as a result of nuclear division. The similarities between the values of the proliferative fractions found experimentally with repeated injections of 3H-TdR compared with the total fraction of cells entering mitosis as measured by vincristine accumulation of metaphases suggests that the number of nuclei remains constant. A separate attempt to determine the extent of nuclear doubling was made by injecting 3H-TdR at the time of peak labelling index and then comparing the fraction of cells labelled between two groups of rats killed 1 hr and 24 hr later. Following the single isoprenaline injection the fraction of cells labelled at these times were 0.094 (kO.011 SE) and 0.130 (f0-017 SE) respectively. After the third isoprenaline injection the values were 0.282 +_ 0.023 and 0.269 & 0-014.In both cases the values were not significantly different, suggesting that little or no change in the number of acinar nuclei occurred in the glands. These results are consistent with previous reports that there is an increased frequency of polyploid cells in salivary glands following isoprenaline treatment (Schneyer, Finley & Finley, 1967; Radley, 1967). Whilst these results vindicate the use of a model in which the number of cell nuclei remains constant, it is of interest to note that the parameter estimates are little altered if the data are corrected to allow for nuclear doubling with every mitosis (Table 1). It would be anticipated that in situations where the proliferative fraction approached 1-0 the estimates would exhibit a greater disparity.

Isoprenaline stimulated proliferative response

161

The calculations of phase times show evidence of negative correlation between the G,, S phase and the combined phases Gi, S Gz, for gamma curve fitting procedures. The fiducial probability that P G ~ +, s, G , , + s + G~ < - H ~ G+~s ,/ ~ G ~+ ,s + GJ where and + + G2 is the phase transit time correlation coefficient and uGi,+ pGIs+ s, oG1,+ + G2 are the corresponding standard deviations, gives a value of >O-9 for fitting procedures on both sets of data and implies a high probability of negative correlation between the transit times of the G,, S and Gi, S Gzdistributions of cell cycle times. Whether this type of correlation would appear in non-vincristine treated rats remains to be determined. Comparing the parameters for the single injection and three injection responses and their respective variance-covariance matrices, we find a x2 > 1400 on 6 degrees of freedom for the difference between parameter sets. This difference is significant at the