Tirapazamine (SR 4233) interrupts cell cycle progression and induces apoptosis

ELSEVIER

Cancer

Letters

105 ( 1996) 249-255

Tirapazamine (SR 4233) interrupts cell cycle progression and induces apoptosis Peck-Sun “Depurtment

Lina**, Kwei-Chi Hoa, Shung-Jun Yangb

of Radiution Oncology, Medical College of Virginia, Wrginia Commonwealth bDepamnent of Radiation Oncology. Methodist Hospital, Brooklyn,

Universiry. Richmond, NY 11215, USA

VA 23298,

USA

Received12April 1996;revision received29 April 1996;accepted30 April 1996

Tirapazamine (Tira), a bioreductive agent, is highly toxic to cells under low oxygen conditions. Since active investigations of this agent are focusing on its potential as an adjunct of radiotherapy to improve overall effects on radioresistant hypoxic tumor cells, understanding its toxic mechanisms under aerobic conditions is important to the clinical application of this agent. Tira-treated V79 Chinese hamster cells were tested for cytotoxicity by colony assay and growth inhibition by the MTT assay. The survival of V79 cells after being exposed to lOO,uM of Tira for 2 h was about 78% of untreated controls. The mitotic cell counts of V79 cells approached zero after 4 h treatment of Tira at 1OOpM or 3 h at 300pM. The fragmentation pattern of DNA isolated from cells 2 h after 300 PM Tira treatment showed characteristics of apoptotic cells. The induction of apoptosis by Tira was also detected by flow cytometric analysis and microscopic observation. These effects of Tira may be part of underlying toxic mechanisms to cells (including normal cells) under aerobic conditions. Keywords:

Aerobic cytotoxicity; Apoptosis; Mitotic index; lirapazamine; SR 4233

I. Introduction The growth of most solid tumors often leads to the development of regions with reduced oxygen tension. Tumor cells in these regions are generally more resistant to radiation and some chemotherapeutic

agents

than their counterparts at normal oxygen tension.

Manipulating

vascular function, hemoglobin

oxygen

affinity, or oxygen content of breathed air can lead to an improvement of tumor tissue oxygenation. Apply-

ing chemical sensitizers or other antitumor agents has ----. * Corresponding author.

Tel.:

+I

804 8288778;

fax:

8286042

0304-3835/96/$12.00 0 1996 El sevier Science Ireland Pi1 SO304-1835(96)04292-9

+I

804

been shown to enhance the effects of radiation on the control of some tumors. In comparison, using agents selectively toxic to hypoxic cells would be the most idea1 way to lessen the role of hypoxic cells in radiotherapy of tumors. Tirapazamine (Tira) (SR 4233, 3amino- 1,2,4benzotriazene 1,4-dioxide, WIN 59075) represents a mode1 for this class of agents by forming toxic oxidizing n&oxide radicals after a one-electron reduction process in hypoxic cells but not in the presence of oxygen (see [l] for review and references therein). Accumulating evidence suggests that Tira is consistently more toxic to hypoxic cells than to cells under aerobic conditions. Tira has already been tested in phase I clinical study [2] while preclinical

Ltd. All rights reserved

250

P.-S. L.in et al. I Cancer

study suggests that it may be potentially toxic to bone marrow [3]. Thus, understanding the toxic mechanisms of Tira under aerobic conditions is also important to the clinical application of this agent. In this report, we show that Tira interrupts the progression of V79 Chinese hamster cells through the cell cycle and induces apoptosis under aerobic culture conditions. These may be part of the underlying mechanisms of toxicity to bone marrow as reported by others.

2. Materials

and methods

2.1. Cell culture and cytotoxicity Monolayer V79 Chinese hamster cell culture and colony forming assay were performed as reported previously [4]. A V79 subline, V79S which was maintained as single-cell suspension culture in 250 ml size of spinner flasks was also used in this study. For the MTT assay, V79 cells were inoculated in 96-well plates at the densities of 5-10 X lo2 cells/well and incubated overnight before treatment. After indicated treatment times, cultures were returned to normal medium and incubated for 34 days, followed by incubation with MIT dye according to previously described procedures [5]. The mean + SD of the cell clonogenic and MTT assays were calculated. 2.2. Mitotic index V79 cultured in 35 mm dishes were treated with 100 or 300pM Tira for l-5 h and in some cultures colcemid (0.2pglml) was added during the last hour of incubation. Because there was no substantial accumulation of mitotic cells with colcemid in Tiratreated cultures, mitotic indices obtained in the absence of colcemid were presented in most of the experiments. At the end of treatment, cells were removed by trypsinization. The preparation of microscopic slides for cytological analysis proceeded according to a routine method [4]. The number of mitotic cells was counted under a 40x objective lens from each sample of not less than 2000 cells; To ensure the effect of Tira was not modified by trypsinization or colcemid treatment, mitotic indices from V79 cells cultured on coverglasses or V79S cells in

Letters

105 (1996)

249-255

suspension treated by Tira only were also included in this study. 2.3. Flow cytometry V79S cells were fixed in 70% ice-cold ethanol immediately after predetermined Tira treatment time. Monolayer cultured V79 cells were fixed in the same manner after trypsinization. Fixed cells were washed once in phosphate-buffered saline with 0.1% Triton-X and then stained with 2Opg/ml of propidium iodide [6] (Sigma Chem. Co., St Louis, MO) and 200 U/ml RNase A (Worthington Biochem. Co., Freehold, NJ). A FACScan Flow Cytometer (Becton Dickinson, San Jose, CA) was used to measure the propidium iodide signal which was analyzed using CellFIT (version 2.0) software (Becton Dickinson). 2.4. DNA extraction and gel electrophoresis Cell samples were lysed in a solution containing 50mM Tris-HCl (pH 8.0), 10 mM EDTA, 10 mM NaCl, 0.5% SDS (w/v), RNase (final concentration 25O,ug/ml), and proteinase K (final concentration 5OOpg/ml). The lysate was extracted with equal volume of phenol/chloroform (1: 1 v/v) and then precipitated with ethanol. The extracted DNA samples were dissolved in TE buffer and the amount of DNA equivalent to content of 3-5 X lo5 cells was electrophoresed at 50 V in 1.6% agarose gels in TBE (100 mM Tris-HCl, 90 mM boric acid, and 1 mM EDTA) buffer with ethidium bromide [7]. Gels were photographed under UV illumination. 3. Results 3.1. Colony and MlTassays cells

of Tira-treated

V79

The average survival for V79S cells treated by 100pM of Tira for 2 h from three experiments was 60.2%. The same treatment of V79 cells from monolayer cultures yielded an average of 77.9%, ranging from 98.96 f 5.08 to 48.90 + 2.94 (17 experiments). Increase of treatment time reduced the survival percentage further, e.g. survival after a 4 h treatment was ~30%. The relative growth of Tira-treated (2 h,

P.-S. Lin ei al. I Cuncer

0

1

2 Time

3

4

5

(Hour)

Fig. 1. The changes in mitotic indices of V79S cells as a function of time of Tim treatment (100pM). Colcemid was added in the last hour of treatment. 0, control; A, Tira-treated; bar, SD if greater than symbol.

Letters

IO5 (1996)

249-255

251

Since caffeine is known to modify the effect of radiation on lethal damage and the mitotic progression, the effect of Tira with caffeine on mitotic index was also studied. Caffeine at the concentration of 2 mM was added in two types of protocols: addition of caffeine at the sametime as Tira at the concentration of 3OOpM for 3 h and followed by drug-free medium for 2-24 h (Fig. 3A), and addition of caffeine for 3-6 h after Tira treatment (Fig. 3B). When caffeine was added along with Tira, no influence on recovery was observed (Fig. 3B). In contrast, caffeine added after Tira treatment enhanced the mitotic indices’ recovery even though the levels of mitotic indices did not overshoot that of the control cultures (Fig. 3B). Whether the inhibition of mitosis induced by Tira resulted from alteration in intracellular calcium levels was studied by adding the calcium channel blocker, nifedipine, with Tira. Fig. 4 shows that 1,uM of nifedipine during Tira treatment did not affect the mitotic indices of V79 cells.

100,&I) V79 cells after 3 days was 88.8% of untreated controls by the M’IT assay. .1..?.The .$ed

of Tira on mitotic index

The mitotic cell counts of logarithmically growing V79 monolayer and suspensioncells treated with Tira at the concentration of 100 or 3OOpM reduced progressively with treatment time. Fig. 1 shows that the changes in mitotic index of V79S cells in the presence of colcemid for 1 h as a function of Tira (100 ,uM) treatment time. The mitotic indices were approaching 0.1% for cells treated for more than 3 h. To determine whether the effect of Tira on the mitotic index of V79 cells is reversible, V79 cells cultured on glass coverslips were first treated with 3OOpM Tira for 3 h and incubated in drug-free medium for 2-24 h. The cells on the coverslips were fixed without colcemid pre-treatment. The mitotic indices recovered as the time of incubation in Tirafree ~mediumtime increased. However, even 24 h after Tira treatment, the mitotic activity had not returned to normal level (Fig. 2).

0.01

I 0

I

I

I

I

6

12

18

24

Time (hr)

Fig. 2. V79 cells cultured on glass coverslips were treated with Tira (3OOpM for 3 h, no colcemid) and the Tira-containing medium was replaced with normal medium. The recovery of mitotic indices as a function of post-treatment incubation time (2-24 h) in the absence of colcemid. A, control; A, Tira-treated; bar, SD.

P.-S. Lin et al. I Cancer

Letters

105 (1996)

249-2.5.5

0.1

t

0.011 0

6

12 Time

16

24

o.ol I 0

(hr)

6

12 Time

16

24

(hr)

Fig. 4. The effect of nifedipine (1 yM) present during Tira treatment (3OOyM. 3 h) on the mitotic indices of V79 cells cultured on glass-coverslips in the absence of colcemid. A, nifedipine treatment only; A, nifedipine added at the same time as Tira; bar, SD.

3.3. DNA content analysis by flow cytometry Flow cytometry of V79S cells treated with Tira (100 or 300,~~M for 3-6 h) showed a population of cells with hypodiploid DNA content (Fig. 5) which is associated with apoptotic cells. Increase in Tira concentration and treatment time can increase the percentage of cells with apoptotic DNA content measurement (Table 1). Parallel to the development of apoptosis, Tira treatment may have different effects on cells’ progression through the cell cycle and produce G2 block accompanied by a decrease in the percentage of S-phase cells (Table 1). 3.4. DNA fragmentation Fig. 3. The effect of caffeine (2 mM) on the mitotic indices of Tira-treated (300pM) V79 cells cultured on glass coverslips in the absence of colcemid. (A) Caffeine was added during the Tira treatment period (first 3 h) and (B) caffeine was added after Tira treatment period. A, caffeine treatment only; A, caffeine was added during Tira treatment; 13, caffeine was added after Tira treatment; bar, SD.

pattern

Because DNA fragmentation into oligonucleosome-length fragments is one commonly used marker for cells undergoing apoptosis, agarose gel electrophoresis of total V79S DNA from control and cells treated with Tira for 2, 3, 4, and 5 h was performed. Fig. 6 exhibits the ladder pattern, which is character-

P 3. tin et 01. I Cancer

Letters

105 (1996)

249-255

253

DNA content Fig. 5. DNA frequency histograms of V79S cells: untreated control (top) or treated with Tim (IOOpM) for 3 h (bottom), showing the appearance of apoptotic cells with hypodiploid DNA content (arrow) following staining with the propidium iodide, analyzed hy flow cytometn, istic of apoptosis, in the lanes of Tira-treated cells,

but not in the control lane.

Table

I

The percentage of cells in different phases of the cell cycle (excluding non-apoptotic cells) as well as the percentage of apoptotic cells following treatment with Tira from three groups of experiments Cell cycle

Treatment

Exp. gmup

Apoptosis

Cl

s

G2M

32.0

55.3 S7.8

12.7

1.9

1 I.1

12.01

38.8

46.1 40.2

17.0 21.0

1.2 13.1

38.3

46.3

15.3

37.4

39.0

23.6

1.7 24.1

A

Control ‘!‘~a (lOO,uuhl. ? h) E.rp group 1) Control 7-h (300,&i. 5 h)

31.1 36.9

Iixp. ~~roup i

Gontrol Tim (300,uM. 6 h)

Fig. 6. Agarose gel electrophoresis of DNA extracted from control V79S cells (lane 1) and cells treated with 300,uM of Tira for 2 h (lane 2), 3 h (lane 3), 4 h (lane 4). and 5 h (lane 5).

4. Discussion Aerobic cytotoxicity of Tira has been observed in several studies, showing that the cytotoxicity may be associatedwith the generation of active oxygen species [8,9] or depletion of ATP [lo]. Our observation of apoptosisin Tira-treated cells under aerobic conditions adds to the list of known effects of Tira. It should be emphasizedthat these Tira effects are not mutually exclusive and may all be part of different phasesof the processleading to cytotoxicity. Cytotoxicity of about 22-40% in our colony formation assay experiments was obtained from 20 identical experiments. Although the mechanisms have not yet been established,low pH of the medium has been shown to affect the cytotoxicity of cells, including V79 cells, under aerobic conditions [I I]. The extent of pH influence in our study was not determined but no obvious changesin phenol red indicator color in media were noted. Our results cannot rule out the possibleinfluence of other factors. There are, however, notable differences in Tira concentration used in those studies: 5OOpM was used in cytotoxicity at low pH [ 1I] or depletion of ATP [lo]

254

P.-S. tin et al. I Cancer

studies. The growth inhibition measured by the MIT assay was similar to that reported in other studies

[IQ. Although Tira disrupts progression of cells through the cell cycle, our result does not show that the cytotoxicity of this agent is cell cycle-specific. Because the M-phase of V79 cells is about 1 h long [ 13,141, the gradual reduction of mitotic cells, reaching maximal effect after 3 h, suggests that the G2 transition to M is sensitive to Tira. Affected cells may resume G2 to M transition shortly after removal of the agent but the process is slow and none of the mitotic indices obtained 3-24 h in Tira-free medium overshot the value of the control culture. In addition, the effect of Tira on mitotic index was not altered with the addition of 2 mM of caffeine either during or after the Tira treatment period. Caffeine has been shown to release radiation-induced S and G2 phase arrest in V79 cells [15] and it may enhance radiation effect on cytotoxicity [ 161. Likewise, the reduction of S-phase suggests that the Gl transition to S or perhaps the S-phase cells themselves may be sensitive to Tira. Thus, the disruption of cells’ progression through the cell cycle by Tira appears not to share the same mechanism(s) as for ionizing radiation. It is possible that cells in these sensitive cell cycle progression steps may be responsible for the cytotoxicity by apoptosis. This notion is supported by the fact that this agent depletes ATP [ 101 that will likely affect the phosphorylation status for normal S and G2 progression. The DNA fragmentation pattern suggests Tiraaffected cells undergo apoptosis. How or whether this is related to the disruption of cell cycle progression is not known at present. Our initial study failed to demonstrate any effect of the calcium channel blocker nifedipine on the reduction of mitotic index induced by Tira. Since Tira shows cell cycle-blocking activity, it is possible that this agent may exert effects on other targets which are important to cell proliferation. The ability of Tira to disrupt the cells’ progression through the cell cycle could sensitize cells to ionizing radiation. A previous toxicological study shows that mice developed leucopenia after being injected with Tira in a 6 week application protocol [3]. This toxicity to bone marrow is consistent with our finding that proliferating cells are susceptible to cell cycle blockage and apoptosis when exposed to Tira. If the ob-

Letters

105 (1996)

249-255

served Tira toxicity in cultured cells also occurs to other proliferating cells in vivo, extrapolating the toxicity patterns from experimental animal models to human applications may not be accurate. Because the size of stem cell pools and/or their turnover rates may be quite different between testing animals and humans (e.g. crypt cells; [ 17]), the level of toxicity may not be equal. Following this line of reasoning, at least before relevant data are obtained, the scheduling of Tira in radiotherapy protocols should take the compensatory proliferation of normal tissues into consideration. Acknowledgements We thank Sterling Winthrop Inc. for providing Tirapazamine and Ms. Kay Endriss for helping in the manuscript preparation. References VI Brown,J.M. PI

(1993) SR 4233 (Tirapazamine): a new anticancer drug exploiting hypoxia in solid tumor. Br. J. Cancer, 67, 1163-l 170. Doherty, N., Hancock, S.L., Kaye, S., Coleman, C.N., Shulman, L., Marquez, C., Naruscak, C., Rampling, R., Senan, S. and Roemeling, R.V. (1994) Muscle cramping in phase I clinical trials of Tirapazamine (SR 4233) with and without radiation. Int. J. Radiat. Oncol. Biol. Phys., 29, 379382.

131

[41

PI

Spiegel, J.F., Spear, M.A. and Brown, J.M. (1993) Toxicology of daily administration to mice of the radiation potentiator SR 4233 (WIN 59075). Radiother. Oncol., 26,79-81. Lin, P.S., Quamo, S., Ho, K.C. and Baur, K. (1985) The diethyldithiocarbamate concentration effects and interactions with other cytotoxic agents on Chinese hamster cells (V79). Radiat. Res., 102, 271-282. Lin, P.S., Ho, K.C., Sung, S.J. and Tsai, S. (1993) Cytotoxicity and manganese superoxide dismutase induction by tumor necrosis factor-a and ionizing radiation in MCF-7 human breast carcinoma cells. Lymphokine Cytokine Res., 12, 303-308.

PI Carson, W.E., Haldar, S., Baiocchi,

r71

R.A., Croce, CM. and Caligiuri, M.A. (1994) The c-kit ligand suppresses apoptosis of human natural killer cells through the upregulation of bcl2. Proc. Natl. Acad. Sci. USA, 91,7553-7557. Gunji, H., Hass, R. and Kufe, D. (1992) Intemucleosomal DNA fragmentation during phorbol ester-induced monocytic differentiation on Go/Cl arrest. J. Clin. Invest., 89, 954960.

PI

Herscher, L.L., Krishna, MC., Cook, J.A., Coleman, C.N., Biaglow, J.E., Tuttle, S.W., Gonzalezx, F.J. and Mitchell, J.B. (1994) Protection against SR 4233 (Tirapazamine)

I’ 3. fin et ul. I Cancer

191

[IO]

aerobic cytotoxicity by the metal chelator desferrioxamine and Tiron. Int. J. Radiat. Oncol. Biol. Phys., 30. 879-885. Zeman, E.M., Brown, J.M., Lemmon, M.J.. Hirst. V.K. and Lee. W.W. (1986) SR 4233: a new bioreductive agent with high selective toxicity for hypoxic mammalian cells. Int J. Radiat. Oncol. Biol. Phys., 12. 1239-1242. Am. G , Coleman C.N. and Teicher. B.A. (1994) SR-4233

(Tirapazamine) acts as an uncoupler of oxidative phosphorylatton in human MCF-7 breast carcinoma cells. Cancer Len.. xs. I’)5203 i I I ] Skatsgard. L.D.. Vinczm. A.. Skwarchuk, M.W. and Chaplin, 0.1. (1994) The effect of low pH and hypoxia on

] I?]

the cytotoxic cffccts of SR4233 and mitomycin Int. J Rndiat. Oncol. Biol. Phys., 29, 363-367. Stratford, I.J. and Stephens, M.A. (1989) The hypoxic cytotoxicity vnru hy the MTT

16. 47J-~l76

I.etters

105 (19%)

255

I I31

Cleaver. I.E. (1969) DNA repair in Chinese hamster cells of different sensitivities to ultra-violet light. Int. J. Radiat. Biol.. 16.277-285. [ 141 Sinclair, W.K. and Morton. R.A. (1966) X-Ray sensitivity during the cell generation cycle of cultured Chinese hamster cells. Radiat. Res., 29. 450-474. [ 151

[ 161

C in vitro.

Hain. J., Crompton, N.E.A.. Burkart, W. and Jaussi, R. ( 1993) Caffeine release of radiation induced S and G2 phase arrest in V79 hamster cells: increase of histone messenger RNA levels and ~34~~~~ activation. Cancer Res., 53, 1.5071510. Griffths, T.D., Carpenter, J.G. and Dahle, D.B. (1978) DNA synthesis and cell cells treated with

survival caffeine

after X-irradiation of mammalian or adenine. Int. J. Radiat. Biol..

33,493-405.

differential

of bioreductive agents determined in assay. Int J. Radial. Oncol. Biol. Phys..

249-255

[ 171

Potten, C.S. (1992) The significance induced apoptosis in the gastrointestinal cer Metastasis Rev.. I I. 1799195.

of spontaneous tract of mice.

and Can-