MODIFICATION OF OSCILLATORY BEHAVIOUR OF PROTEIN TYROSINE KINASE AND PHOSPHATASE DURING ALL-TRANS RETINOIC ACID-INDUCED DIFFFERENTIATION OF LEUKAEMIC CELLS

MODIFICATION OF OSCILLATORY BEHAVIOUR OF PROTEIN TYROSINE KINASE AND PHOSPHATASE DURING ALL-TRANS RETINOIC ACID-INDUCED DIFFFERENTIATION OF LEUKAEMIC CELLS

Cell Biology International 2002, Vol. 26, No. 12, 1035–1042 doi:10.1006/cbir.2002.0963, available online at http://www.idealibrary.com on MODIFICATIO...

366KB Sizes 0 Downloads 12 Views

Cell Biology International 2002, Vol. 26, No. 12, 1035–1042 doi:10.1006/cbir.2002.0963, available online at http://www.idealibrary.com on

MODIFICATION OF OSCILLATORY BEHAVIOUR OF PROTEIN TYROSINE KINASE AND PHOSPHATASE DURING ALL-TRANS RETINOIC ACID-INDUCED DIFFFERENTIATION OF LEUKAEMIC CELLS J. L. CALVERT-EVERS1,* and K. D. HAMMOND2,3 Departments of 1Surgery and 2Haematology and Molecular Medicine, Medical School, University of the Witwatersrand, 7 York Road, Parktown, Johannesburg, 2193 South Africa; 3Department of Biochemistry, Faculty of Medicine and Health Sciences, University of the United Arab Emirates, P.O. Box 17666, Al-Ain, United Arab Emirates Received 11 March 2002; accepted 21 August 2002

Granulocytic maturation of human acute promyelocytic leukaemic (HL-60) cells was induced using all-trans retinoic acid (ATRA). Time-dependent changes in the enzyme activities of protein tyrosine phosphatase (PTP) and protein tyrosine kinase (PTK), and the total extractable protein content were monitored in proliferating and differentiating cells. The existence of periodicity was demonstrated clearly in both PTP and PTK enzyme activities and in the amount of protein extracted from the cells. Following ATRA treatment, differentiation-induced changes in rhythmic characteristics such as period and amplitude were evident. A noticeable effect was that of ATRA on the enzyme activity of PTP, for which four distinct patterns of oscillatory behaviour were identified. This study examines these changes, in an attempt to gain insight into the role which biochemical oscillators may play in the regulation of molecular control  2002 Elsevier Science Ltd. All rights reserved. mechanisms. K: protein tyrosine kinase; protein tyrosine phosphatase; all-trans retinoic acid; oscillations; proliferation; differentiation; acute promyelocytic leukaemia cells.

INTRODUCTION Reversible phosphorylation plays a dominant role in co-ordinating complex cellular activities. In the past, protein phosphatases have been thought of as enzymes that served merely to counteract the action of kinases, whether this resulted in the activation or inhibition of a target enzyme. However, it soon became clear that the cyclic interconversion of these two regulatory enzymes represents the dynamic process in which the steadystate equilibrium between active and inactive forms varies depending on parameters determined by the metabolic state of the system (Averna et al., 2001; Zhang, 2001). Earlier studies show that phosphorylation processes are highly dynamic, even under resting conditions (Hammond et al., 1988; *To whom correspondence [email protected] 1065–6995/02/$-see front matter

should

be

addressed:

E-mail:

Hammond, 1998; Calvert-Evers and Hammond, 2000). When these intracellular processes were disturbed by differentiating agents such as hexamethylene bisacetamide (HMBA) and ATRA, distinct time-dependent changes in both PTP and PTK enzyme activities were detected (Hammond et al., 1998; Calvert-Evers, 2000). In this study, we investigated further the activities of these enzymes in the same cell preparation both before and after ATRA treatment, so that a direct comparison between these two key regulatory enzymes could be made. In addition, evidence was sought to confirm the long-standing hypothesis that regulators of replication initiate or inhibit oscillatory behaviour by affecting the concentrations of cellular constituents (Gilbert, 1974a, 1974b). This may occur as a result of changes in the rate and kinetics of the reactions forming the intracellular control system. Although high frequency  2002 Elsevier Science Ltd. All rights reserved.

1036

oscillations and rhythmic activities associated with reversible protein phosphorylation have been reported (Hammond et al., 1998), it was our intention to determine whether any distinctive features such as phase shifts or changes in the frequency, amplitude or phasing of a particular set of rhythms could be distinguished. If this is indeed the case, then these rhythmic characteristics may play a vital role in modulating cellular behaviour. METHODS HL-60 cell culture: maintenance, differentiation and extraction HL-60 (ATCC CCL 240) cells were maintained as single cell cultures suspended in RPMI medium supplemented with 10% foetal calf serum and antibiotics (penicillin 0.52 g/l, streptomycin 0.86 g/l). Having reached a cell density of 1106 cells/ml, HL-60 cells were divided equally into two tissue culture flasks, labelled ‘control’ and ‘test’. Differentiation of HL-60 ‘test’ cell cultures was initiated using ATRA (1 M). After a 24 h incubation period, 50 ml samples of control and induced HL-60 cultures were extracted at 5 min intervals over a 1 h time period. The extracts were centrifuged (600g, 10 min) and washed twice with 0.9% saline. The washed pellets were resuspended in 150 l lysing buffer (30 mM Tris-HCl, pH 6.8, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS). Following two cycles of freeze/ thawing, the homogenized extracts were centrifuged (2000g, 20 min) and the supernatant stored at 20C. Determination of total protein content and enzyme activity The amount of total protein extracted from each cell lysate was determined using the Lowry method (Lowry et al., 1951). Protein values were expressed as g/l. Quantitative assessment of PTP and PTK enzyme activity was determined using nonradioactive photometric enzyme-linked immunoassays (ELISA; Boehringer Mannheim, Germany). The determination of PTP enzyme activity has been described elsewhere (Calvert-Evers and Hammond, 2000) and is reported as pmol substrate dephosphorylated/min/10 l of cell extract. The enzyme activity of PTK was detected by monitoring the transfer of the -phosphate group from ATP to a tyrosine of a biotin-labelled substrate peptide corresponding to amino acid sequence 1–17 of gastrin.

Cell Biology International, Vol. 26, No. 12, 2002

After quenching the enzyme reaction using a specific inhibitor (final concentration 3 mM piceatannol, 10% v/v DMSO), phosphorylated and dephosphorylated substrate was immobilized by binding to a streptavidin-coated microtitre plate. The fraction of phosphorylated substrate was determined immunochemically using a highly specific anti-tyrosine antibody, directly conjugated to peroxidase. PTK enzyme activity was determined by calculating the substrate turnover from a phosphopeptide standard curve and was reported as pmol phosphate incorporated into substrate/ min/10 l of cell extract. Data values were plotted as a function of time using a fourth power polynomial fit through the data points. Analysis of oscillations The analysis of cellular oscillations and the relationship between PTP and PTK enzyme activities included linear correlations and determination of the rhythmic characteristics, period and amplitude. These characteristics were determined using a computer programme (PERAMP) designed by D. A. Gilbert. An estimate of the period could be obtained manually by calculating the number of successive peaks (or troughs) seen over a given time interval; similarly, the amplitude could be determined from the vertical distance between a peak (or trough) to the line joining adjacent troughs (or peaks). RESULTS Temporal variations in total extractable protein, PTP and PTK enzyme activities Ten individual time-course experiments were carried out. Temporal variations were observed in the amount of total protein and enzyme activities of PTP and PTK extracted from both proliferating and differentiating HL-60 cells. ATRA treatment significantly altered the pattern of the oscillating waveforms. For PTP enzyme activity, the effect of ATRA gave rise to four distinct patterns of oscillatory behaviour: partial phase shift (3 out of 10 experiments); dampening (4 out of 10 experiments); phase shift (1 out of 10 experiments) and aperiodic, random behaviour (2 out of 10 experiments). Representative results of the specific changes in oscillatory behaviour from four separate experiments are presented. Partial Phase shift (Fig. 1). Following treatment with ATRA, a partial phase shift in PTP enzyme

Cell Biology International, Vol. 26, No. 12, 2002

Fig. 1. Partial phase shift in PTP enzyme activity following ATRA treatment. Aliquots of control (——) and ATRAinduced (– – – –) HL-60 cell cultures were extracted at 5 min intervals over a period of 1 h. Graph A depicts the data values obtained for extractable protein in g/l; graph B, tyrosine phosphatase enzyme activity in pmol/min/10 l of cell extract; and graph C tyrosine kinase enzyme activity in pmol/min/10 l of cell extract. Graphs A, B and C can be directly compared, as the data values were obtained using the same cell preparations.

activity was observed in experiments that initially showed periodic oscillatory patterns. A sharp increase in PTP enzyme activity occurred at different times in different experiments. In Figure 1B, the increase in PTP enzyme activity was seen after 50 min; in other cases the change was seen after 10 or 25 min. In contrast to PTP, other than a slight increase in activity, the overall pattern for PTK remained essentially the same in the presence of ATRA (Fig. 1C). Dampening (Fig. 2). In some instances, ATRA had a dampening effect on the oscillatory behaviour of PTP enzyme activity (Fig. 2B). A slight increase in PTK enzyme activity (Fig. 2C) was demonstrated

1037

Fig. 2. Dampening of PTP enzyme activity following ATRA treatment. Aliquots of control (——) and ATRA-induced (– – – –) HL-60 cell cultures were extracted at 5 min intervals over a period of 1 h. Units of measurement are the same as those in Figure 1. Graphs A, B and C can be compared directly as the data values were obtained using the same cell preparations.

in corresponding ATRA-treated HL-60 extracts. Of particular interest in this graph is a sudden decrease in PTK enzyme activity after 35 min. Such ‘abnormal’ data values were not dismissed, as they may indicate the convergence of several rhythms of different amplitude, frequency and period. Phase shift (Fig. 3). A most striking change in oscillatory behaviour may be seen in Figure 3B, in which ATRA induced an ‘out-of-phase’ modulation in PTP enzyme activity. In contrast, there was much less deviation between the corresponding PTK enzyme activity patterns (Fig. 3C). Aperiodic, random behaviour (Fig. 4). An interesting response to ATRA may be observed in Figure 4B, where the oscillatory pattern of PTP

1038

Fig. 3. Phase shift in PTP enzyme activity following ATRA treatment. Aliquots of control (——) and ATRA-induced (– – – –) HL-60 cell cultures were extracted at 5 min intervals over a period of 1 h. Units of measurement are the same as those in Figure 1. Graphs A, B and C can be compared directly as the data values were obtained using the same cell preparations.

enzyme activity in untreated cells was aperiodic and complex. However, following treatment with ATRA, there was an overall suppression of fluctuations. Dampening of PTK enzyme activity occurred following ATRA treatment (Fig. 4C). Protein/enzyme activity relationships Linear correlation studies were carried out to determine whether the amount of extractable protein and PTP and PTK enzyme activities were inter-related. Generally, for all experiments, the degree of correlation between PTP or PTK enzyme activities and the corresponding total protein

Cell Biology International, Vol. 26, No. 12, 2002

Fig. 4. Aperiodic, random behaviour in PTP enzyme activity following ATRA treatment. Aliquots of control (——) and ATRA-induced (– – – –) HL-60 cell cultures were extracted at 5 min intervals over a period of 1 h. Units of measurement are the same as those in Figure 1. Graphs A, B and C can be compared directly as the data values were obtained using the same cell preparations.

concentration in untreated and treated HL-60 cells was poor (correlation coefficients ranged between 0.018 and 0.788).

PTP/PTK enzyme relationship Time sequence values of PTP enzyme activity were plotted against the corresponding time sequence values for PTK enzyme activity, to determine the time-relationship between the two enzyme activities. In most cases, PTP-PTK relationships were poorly correlated (correlation coefficients ranged between 0.013 and 0.723). Following ATRA treatment, the linear relationship worsened.

Cell Biology International, Vol. 26, No. 12, 2002

Fig. 5. Period (min) and amplitude (arbitrary activity units) variations in PTP (A and B) and PTK (C and D) enzyme activities in untreated control (——) and ATRA-treated (– – – –) cells. Data values correspond to those of Figure 1, representing a partial phase shift in PTP enzyme activity following ATRA treatment.

1039

Fig. 7. Period (min) and amplitude (arbitrary activity units) variations in PTP (A and B) and PTK (C and D) enzyme activities in untreated control (——) and ATRA-treated (– – – –) cells. Data values correspond to those of Figure 3, representing a phase shift in PTP enzyme activity following ATRA treatment.

of the oscillatory patterns, as depicted in Figure 4, prevented computation. PTP activity showed timedependent changes in both period and amplitude. Variations between the values for PTP period were greater in the presence of ATRA than in its absence. PTP amplitude was modified by ATRA, with considerable decreases seen in the experiments depicted in Figures 6 and 7. For PTK, both period and amplitude showed variations with time, and again the patterns were modified in the presence of ATRA.

DISCUSSION Evidence of dynamic oscillatory processes Fig. 6. Period (min) and amplitude (arbitrary activity units) variations in PTP (A and B) and PTK (C and D) enzyme activities in untreated control (——) and ATRA-treated (– – – –) cells. Data values correspond to those of Figure 2, representing a dampening in PTP enzyme activity following ATRA treatment.

Variations in rhythmic characteristics The rhythmic characteristics of period and amplitude are shown in Figures 5, 6 and 7 and summarized in Table 1. Only three sets of data are presented, since the aperiodic and irregular nature

Temporal organization and time-dependent changes in the pattern of oscillatory behaviour were evident in both extractable total protein content and in PTP and PTK enzyme activities, though no obvious relationship could be distinguished between the protein concentration and enzyme activities. The poor degree of correlation between both PTP and PTK enzyme activities and the corresponding protein concentrations in proliferating and differentiating HL-60 cells suggests relative independence of these sets of oscillators. If temporal changes in PTP and PTK enzyme activities were caused simply by fluctuations in the amount of protein extracted from the cells, the timing

1040

Cell Biology International, Vol. 26, No. 12, 2002

Table 1. A comparison between the period and amplitude of protein tyrosine phosphatase and protein tyrosine kinase enzyme activities in untreated and ATRA-treated HL-60 cells Period (min)

Amplitude (arbitrary activity units) Untreated Treated

Untreated

Treated

PTP enzyme activity Partial phase shift Dampening Phase shift

10–12 (11) 5–20 (12) 10–12 (11)

8–18 (13) 8–13 (10) 8–18 (13)

0.5–1.4 (1.0) 0.5–10 (5) 0.5–1.4 (1.0)

0.1–0.8 (0.4) 0.1–1.0 (0.5) 0.1–0.8 (0.4)

PTK enzyme activity Partial phase shift Dampening Phase shift

20–22 (21) 10–14 (12) 12–22 (17)

10–25 (17) 3–20 (11) 14–26 (20)

2–4 (3) 0.2–1.2 (0.7) 0.5–4 (2)

1–11 (6) 0.5–4 (2) 1–2.8 (1.5)

The range of values obtained in the individual experiments representing the different patterns of behaviour is given. Estimates of mean value are shown in parentheses.

relationship would not be apparent, as the enzyme activities should vary in unison with the protein changes. If the activities were interdependent, phase plane plots should be linear and have a positive slope. As no particular relationship could be detected for these periodicities, it is possible that the frequency of the rhythms are different, or that multiple interacting rhythms are involved (see Gilbert and Tsilimigras, 1981). These findings are consistent with previous reports (Calvert-Evers and Hammond, 2000; Gilbert and Tsilimigras, 1981; Ferreira et al., 1994). Differentiation-induced changes in dynamic behaviour In untreated HL-60 cultures, the cells may have adapted to each other on a temporal basis so that an overall periodicity exists. It has been suggested that normal disturbances during cell culture, such as feeding the cells at a particular time every day or cell-cell signalling, may cause partial entrainment of cellular oscillations and produce a certain degree of metabolic synchrony (Hammond, et al., 1998). This ‘steady’ state may result from interactions between the individual regulatory processes within the cell and between the metabolic control systems of different cells (Gilbert, 1984). Although it is doubtful whether any cell is in a true stable state, since they are continuously being affected by both random and deterministic disturbances of low magnitude, the overall pattern of rhythm behaviour is periodic and no metabolic switches are irreversibly triggered.

Following treatment with ATRA, differentiation-induced changes in the pattern of oscillatory behaviour were observed in PTP and PTK enzyme activities. It has been reported that regardless of the inducer used, specific proteins that are tyrosine-phosphorylated in proliferating HL-60 cells undergo gradual dephosphorylation 12–72 h after induction of differentiation (Gineitis et al., 1999). Some of our results are in agreement with this finding, as suppression of PTP enzyme activity was observed in 4 out of the 10 experiments. However, the varied response of the oscillating rhythms to ATRA warrants further analysis and draws attention to some of the complexities associated with the study of cellular oscillations. Although differentiation-induced changes in the oscillatory pattern of behaviour were observed for PTK enzyme activity, the response in relation to changes seen in PTP enzyme activity was not as dramatic. Modulation of rhythmic characteristics The results show that the period and amplitude of the rhythmic characteristics of PTP and PTK enzyme activities varied with time (Table 1). Furthermore, distinct changes in these characteristics were seen in HL-60 cell extracts following differentiation. Because of the wide range of values, the mean period was difficult to estimate. It appeared, however, to be of a similar order of magnitude in both proliferating and differentiating cells, and for both enzymes. Differences in the amplitude values obtained for PTP enzyme activity in untreated and induced cells were very obvious.

Cell Biology International, Vol. 26, No. 12, 2002

As might be expected, a noticeable decrease (10-fold) was observed in experiments showing dampening or suppression of PTP enzyme activity following ATRA treatment. Such an observation provides for an extra dimension of metabolic control through differential modulation of rhythmic characteristics such as amplitude. It should be remembered that communication can exist between certain rhythms with the same periodicity in other cells that may lead to gradual metabolic synchronization of the population. An increase in amplitude for PTK following induction of differentiation occurred in all of the experiments considered. Evidence of periodic modulation of rhythmic characteristics occurring in phosphoprotein phosphatases has been reported in earlier studies. In studies of phosphotyrosine phosphatase (PTPase) activity in murine erythroleukaemic cells (MEL), it was found that insulin decreases the frequency of the rhythm of PTPase activity, while the inducer of differentiation, HMBA, decreases its amplitude (Ferreira et al., 1994). In subsequent studies, time-dependent changes were assessed in the activities of phosphoserine, phosphothreonine and phosphotyrosine phosphatases in MEL cells (Ferreira et al., 1996). When the cells were treated with the differentiation inducers HMBA and dimethyl sulphoxide (DMSO), differences in the frequency and phasing of rhythms were seen in the treated cells, compared with controls. It was speculated that the period of the primary activity oscillations was in the order of 10 min, or even less. Further studies of temporal variations in the expression of PTK in HL-60 cells using HMBA as the inducing agent showed striking disturbances after the cells had been exposed to the inducer for 24 h. The mean enzyme activity was considerably higher than for untreated cells, and there was a distinct waveform with a period of about 40 min, of varying amplitude (Hammond, 1998). A crucial limiting factor in the analysis of biological rhythms is the timing of the samples (Gilbert and Ferreira, 2000). The sampling interval in these earlier studies was between 10 and 15 min and, since the frequency of the periodic rhythm may be influenced by the sampling frequency, it was suggested that the true period of the rhythms remains to be determined (Hammond et al., 1998). Factors which influence dynamic behaviour of phosphorylating systems The wide variety of modulatory effects following ATRA treatment reflects the dynamic nature of

1041

cells and the complex time-dependent interactions between the individual regulatory processes within the cell, between the metabolic control systems of different cells, and with the environment. The HL-60 cell culture used in this study constitutes a heterogeneous population, with cells distributed unequally throughout the cell cycle. Since the metabolic state of the cell is phase-dependent, one would expect the response of cells in the quiescent state to be different to those that are actively replicating. In addition, the relative timing of endogenous rhythms responsible for changing the metabolic state of the cell will have an effect on the nature of the cell’s response to external stimuli (Gilbert, 1974a, 1974b, 1984). The immediate effect of an inducing agent such as ATRA on the oscillating system may be a rapid increase in one experiment, a decrease in another, or there may be no response at all. A point worth mentioning is that individual ‘abnormal’ data values, which differ markedly from the rest of the data, should not be dismissed as random experimental errors. Such irregularities in periodic oscillations have been simulated in theoretical studies involving the combined effect of two or more modulating periodicities (cf. Gilbert and Ferreira, 2000). In conclusion Evidence of periodic modulation of rhythmic characteristics is demonstrated clearly in this study. However, the precise interpretation and significance of these modulations in relation to changes in cellular behaviour is speculative at this stage, although it is possible that they may play an essential regulatory role. Apart from the unpredictable nature of non-linear dynamic systems, the amplitude and frequency of oscillations triggered through surface receptors vary from cell to cell over time, resulting in a complex mixture of stimulus waveforms (cf. Gilbert and Lloyd, 2000). In addition, surface receptors are often coupled to multiple signalling pathways, making it difficult to ascribe downstream effects to any one particular oscillator. It is suggested that differences in frequency and amplitude may increase the efficacy of biochemical processes and specificity of gene expression through the periodic modulation of oscillatory rhythms, resulting in subsequent changes in cellular behaviour. It is hoped that an understanding of oscillatory phenomena on a biochemical and molecular level will lead to an understanding of the origin, as well as the physiological function, of

1042

these rhythms, and the conditions under which defective enzyme regulation may give rise to malignant transformation. ACKNOWLEDGEMENTS This work was supported by the National Research Foundation, South Africa. Dr Don Gilbert is thanked for providing computer programmes and helpful advice. REFERENCES A M, D R, P M, S F, P S, M E, 2001. Changes in intracellular calpastatin localization are mediated by reversible phosphorylation. Biochem J 354: 25–30. C-E JL, 2000. Temporal variations in protein tyrosine phosphatases and kinases during cell proliferation and differentiation. Ph.D thesis, Faculty of Science, University of the Witwatersrand, Johannesburg, South Africa. C-E JL, H KD, 2000. Temporal variations in protein tyrosine phosphatase activity during cell proliferation and differentiation. Cell Biol Int 24: 559–567. F GMN, H KD, G DA, 1994. Insulin stimulation of high frequency phosphorylation dynamics in murine erythroleukaemia cells. Biosystems 33: 31–43. F GMN, W¨  H, H KD, G DA, 1996. High frequency oscillations in the activity of phosphotyrosine phosphatase in murine erythroleukaemic cells: action of insulin and hexamethylene bisacetamide. Cell Biol Int 9: 599–605. G DA, 1974a. The temporal response of the dynamic cell to disturbance and its possible relationship to differentiation and cancer. S Afr Sci 70: 234–244.

Cell Biology International, Vol. 26, No. 12, 2002

G DA, 1974b. The nature of the cell cycle and the control of cell proliferation. Biosystems 5: 197–206. G DA, T CWA, 1981. Cellular oscillations: relative independence of enzyme activity rhythms and periodic variations in the amount of extractable protein. S Afr J Sci 77: 66–72. G DA, 1984. Temporal organisation, reorganisation and disorganisation in cells. In: Edmunds LN Jr, ed. Cell Cycle Clocks. New York and Basel, Marcel Dekker Inc. 5–25. G DA, L D, 2000. The living cell: a complex autodynamic multi-oscillator system? Cell Biol Int 24: 569–580. G DA, F GMN, 2000. Problems associated with the study of cellular oscillations. Cell Biol Int 24: 501–514. G A, T G, S J, S VP, T T, 1999. Long-term changes in tyrosine phosphorylation of the abundant nuclear proteins during granulocytic differentiation of HL-60 cells. Cell Mol Life Sci 55: 317–26. H KD, C L, M B, G DA, 1988. Temporal changes in phosphoamino acid phosphatase activities in murine erythroleukaemic cells. Int J Biochem 21: 197–201. H KD, 1998. Oscillations and cancer: control of protein kinases during proliferation and differentiation of leukaemic cells. 17th International Cancer Congress, Rio de Janeiro (Brazil), August 24–28, 1998. H KD, B R, B U, G DA, 1998. Dynamic cells: Temporal organisation and control of phosphorylation. Trends Com Biochem Physiol 4: 75–88. L OH, R NJ, F AL, R RJ, 1951. Protein measurement with the Folin phenol reagent. J Biol Chem 193: 265–275. Z ZY, 2001. Modulation of protein kinase signaling by protein phosphatase and inhibitors. Cell Mol Biol Lett 6: 487–488.