Department of Medical Biochemistry, Division of Molecular Biology

Thymidine Kinase as a Marker for the Malignant State of a Cell

Group Leader: Edgar Wawra

Postdocs:

PhD Student:

Diploma students:

Technician: Ingrid Mudrak (part time)

I. What Causes a Cell to Become a Tumour Cell ?
The replication cycle of eukaryotic cells may generally be considered as a series of transitions or checkpoints, where the cells decide what to do next: to synthesise DNA, to enter mitosis, simply continue to growth, to differentiate e.t.c. At one of these points, the cell has to decide "to divide or not to divide". This is dependent on the metabolic state of the cell (is it fat enough to be able to perform the energy consuming steps of DNA synthe sis and mitosis ?) but in a multicellular organism it is also a function of signals from the surrounding (is it desirable for the organism to have two cells instead of one ?). Before the cell enters division, it has to activate a number of genes - so-call ed "S-phase genes" - which are necessary for DNA synthesis, but remain silent during any other period of the cell cycle and also in resting or differentiated cells. Such S-phase genes obviously are the DNA polymerases and the enzymes of the DNA precursor metabolism which supply the ongoing synthesis with the necessary deoxynucleotide triphosphates. The activation of these (and probably of many other unidentified genes) is done few hours before the start of DNA synthesis.

Figure 1 presents the metabolic pathways leading to DNA synthesis and the involved enzymes. Take into consideration that there are two groups of reactions: the steps of the de novo synthesis, which are highly active during S-phas e, and the reactions of the so-called salvage pathway, which supplies the cellular metabolism with a low amount of triphosphates at any time - probably for repair synthesis. Deoxycytidine kinase is an example for a "salvage enzyme", it is constitut ively expressed in any cell. Thymidine kinase catalyses a similar reaction, but for reasons not yet understood, this enzyme is regulated exactly like other "de novo enzymes" and is obviously part of the S-phase machinery.

Tumour cells behave differently. Obviously, they have lost the checkpoint where the decision to divide is made and they grow independently and - if malign - as fast as they can. Consequently, they also have lost control over their S-pha se genes! We already have shown that genes encoding thymidine kinase, dihydrofolate reductase, and both subunits of nucleosidediphosphate reductase, which are well regulated on mRNA level in normal cells, are unregulated and are constantly enhanced in tum our virus transformed cells. Polyoma virus large T antigen (a protein necessary for polyoma virus-caused cellular transformation) alone is sufficient to change in otherwise normal cells the expression pattern of S-phase genes from the regulated phenotype to the unregulated one.

II. How do Oncogenes and Tumour Suppressors fit in this Picture?

Two families of regulatory proteins have been described: the cyclines, which fluctuate during the cellular cycle and are therefore able to act as cellular clocks, and the cyclin-dependent kinases (CDK), which - in concert with the cyclines - specifically phosphorylate other regulatory proteins. The current model invokes successive waves of different CDK activity, regulated in turn by different cyclins. One subgroup of these - the cyclines D - reac h their maximum intracellular level few hours before S-phase and are therefore most obviously involved in the onset of the replicative cycle.

The S phase regulated genes share common properties in respect to their promoters. DNA polymerase a, dihydrofolate reductase, thymidylate synthase and thymidine kinase are all regulated by the transcription factor E2F, which itself is regulated by the retinoblastoma gene product (pRb) (see Figure 2). pRb binds and inactivates E2F, but phosphorylation of pRb sets E2F free. This pRb is target for the tum our antigens of DNA tumour viruses, so that in virus transformed cells, pRb allows E2F to activate these genes.

But pRb is cell cycle regulated by the cyclin D/CDK system, and this is known to be a target for two of the most prominent tumour suppressor gene products p16 and p53. p16 is an inhibitor of the cyclin D/cdk4 kinase, p53 is an activator of p21 which inhibits cyclin D/cdk2 and cyclin E/cdk2. Therefore, the described deregulation of S-phase genes is an indicator for many different tumour causing mechanisms: presence of viral coded tumour antigen, defect in pRb, and defect in either p53, p 21 or p16. This explains why we found this effect in so many different neoplastic cell populations: in various tumours, in leukaemia samples, in cell lines derived from tumours or obtained after transformation by tumour virus.

III. Thymidine Kinase is a Convenient Marker for the Deregulation of S-Phase Genes

Thymidine kinase (TK) catalyses the ATP-dependent phosphorylation of thymidine and deoxyuridine. The activity of TK is strictly regulated during the normal cell cycle, peaking at the onset of DNA synthesis but remaining extremely low in resting cells. We already have been shown that the change in TK gene expression during the cell cycle is accompanied by a similar change in TK enzyme activity. Obviously, activity of TK is dominated by the amount o f its mRNA, and the difference between normal and malignant cells at the level of transcription is mirrored by the enzyme activity. Compared with other S phase regulated genes, TK appears to be the most convenient indicator for this effect: (i) the half l ive of the enzyme is short enough to reflect variations during the cell cycle, (ii) there is no interfering cell-cycle dependent regulation on protein level, and (iii) TK activity is easy to detect.

In cell populations from different phases of the cellular cycle (obtained by elutriation centrifugation), we measured enzyme activities and mRNA levels and could detect the differences between normal and malignant cells described above. However, it would be much desirable to measure this effect on a single cell basis too. Besides the statistical advantage, it would also enable us to analyse cellular mixtures and to identify few transformed cells in a surplus of normal ones. The idea for such an assay was to synthesise an artificial but fluorescent substrate for thymidine kinase, which is phosphorylated by the enzyme like thymidine. the accumulation of the phosphorylated products in the cells should therefore be visible by their fluoresc ence. Simultaneous measurement of this fluorescence together with that of a DNA sensitive stain should provide us with information for both thymidine kinase activity and phase of the cell cycle for each individual cell.

We succeeded to synthesise such a dye, and we found that the intracellular accumulation of its fluorescence was strictly correlated with the overall thymidine kinase activity of the cells. This implies that from the series of necessary events (uptake into the cell, phosphorylation to the mono-, di-, and triphosphate), the step catalysed by thymidine kinase is the bottle neck of the whole pathway and therefore rate limiting. Using a cytofluorometer, this assay allows to discriminate betw een normal growing cells, like diploid fibroblasts, and virally transformed cells or lines derived from tumours. But we also had identified leukemic cells in patients bone marrow or peripheral blood.

In order to show how our results have to be interpreted, a few examples are shown in Figure 3. The DNA distribution presented on top of each diagram represents the distribution of cells in G1, S and G2 phase (the G2 peak on the right co ntains cells with exactly twice the amount of DNA than the G1 peak on the left, between the two peaks are the S phase cells with intermediate DNA content.).The two dimensional presentation below shows simultaneous measurement of two fluorescences, reflect ing DNA amount and thymidine kinase (TK) activity for each cell. Normally growing cells have low TK activity in G1, this activity increases during early S and returns to about the original level in G2. (Resting cells have even lower TK activity.) Malignan t growing cells always exhibit more TK activity than S phases of normal cells do. These cells are therefore found further down the axis in the diagram.

The examples in Figure 3 are normal fibroblasts (extreme left) and HeLa cells (a classical tumour cell line, on the extreme right). The other diagrams represent cell lines obtained from three different subsequent biopsies from a patient suffering on Ewing tumour. The first biopsy (1) was taken from the benign growing tumour, the other samples (2, 3) were taken from the same tumour during its rapid progression to malignancy. Cytogenetic analysis did not indicate any alteration in the are as of the gene locus for thymidine kinase nor for pRb. But the later two biopsies show a deletion of the region where the p16 gene is localised. (Indeed, analysis on RNA levels confirmed that the lines 2 and 3 were p16 negative). Obviously, the defect in p16 caused the cells to become malignant. We further conformed this idea by genetically manipulating p16-negative cells, i.e. we transiently transformed the cells with the p16 gene, and were able to find the expected reversion in the thymidine kinase expr ession pattern when using our fluorescence assay.

When applied on artificial mixtures of logarithmically growing normal and transformed cells, this method enabled us to detect the malignant in a 10 000 fold excess of normal ones. The benefits of such a simple and general tumour marker are obvious: this opens the possibility to use this method widely for clinical diagnosis, but it may also serve as a tool in cell biology for the study of the mechanism of cell transformation.

Publications

1. Soucek, T., Pusch, O., Hengstschläger-Ottnad, E., Wawra, E., Bernaschek, G., and Hengstschläger, M. (1995). Expression of the cyclin-dependent kinase inhibitor p16 during the ongoing cell cycle. FEBS Let t. 373, 164-169.

2. Hengstschläger, M., Pusch, O., Hengstschläger-Ottnad, E., Ambros, P.F., Bernaschek, G., and Wawra, E. (1996). Loss of p16/MTS1 tumor suppressor gene causes E2F-mediated deregulation of essential enzymes of the DNA precur sor metabolism. DNA Cell Biol. 15, 41-51.

3. Hengstschläger, M., Pfeilstöcker, M., and Wawra, E. (1996). Identification of leukaemic cells in bone marrow and blood samples by a new cytofluorometric assay. Brit. J. Cancer 73, 1237-1240.

4. Hengstschläger, M., Hengstschläger-Ottnad, E., Pusch, O., and Wawra, E. (1996). The role of p16 in the E2F-dependent thymidine kinase regulation. Oncogene 12, 1635-1643.

5. Pusch, O., Soucek, T., Wawra, E., Hengstschläger-Ottnad, E., Bernaschek, G., and Hengstschläger, M. (1996). Specific transformation abolishes cyclin D1 fluctuation throughout the cell cycle. FEBS Lett. 385, 143-148

Collaboration Markus Hengstschläger, Department of Prenatal Diagnosis and Therapy, Obstetrics and Gynecology, AKH, University of Vienna

back

Division of Molecular Biology

Department of Medical Biochemistry

Division of Molecular Biology

Research Groups

Thymidine Kinase as a Marker for the Malignant State of a Cell