Cell-cycle phase and proliferation state-dependent radiation and chemotherapeutic agent toxicity in vivo

Cell-Cycle Phase and Proliferation State-Dependent Radiation and Chemotherapeutic Agent Toxicity In Vivo Dietmar W. Siemann and Peter C. Keng

t is now well accepted that the cycle phase and proliferation state of a cell may significantly affect response to radiation and cytotoxic drugs. The experimental basis for this belief comes primarily from in vitro studies using synchronized cells and/or exponential or plateau-phase cell cultures. Indeed, using a variety of synchronization techniques, extensive studies have been carried out to determine the cycle phase dependence of the killing action of radiation and a wide variety of anticancer drugs. E,2 Most of these investigations have described the age response function of a particular cell line treated with a given agent. Typically in such studies, cell subsets, synchronized at the various phases of the cell cycle, are treated with a fixed dose of the cytotoxic agent and cell survival as a function of cell cycle phase is determined (Fig 1). For radiation, most, although not all, cell lines have shown greatest radiation sensitivity in the M and late G2 phases and maximum resistance in the S phase. 35 In contrast, cell age response patterns for chemotherapeutic agents vary considerably, although most do show differential efficacy in some parts of the cell cycle. 1,2,6 To evaluate the impact of proliferation state on anticancer therapies, many investigators have compared the treatment efficacy of the test agent in log or plateau phase cell cultures. These growth phases are chosen to mimic in vitro the nutrient deprivation that can occur in tumors and cause the formation of subpopulations with distinct proliferation states. 7 The cell populations include a cycling or proliferative subpopulation (P cells), a noncycling or quiescent subpopulation (Q cells), and a nonproliferating sub-

I

From the Tumor Biology Division and Department of Radiation Oncology, UniversityofRochesterCancer Center,Rochester,NY. Supported by National Institutes of Health Grants No. CA-11051 and CA-36858. Address reprint requests to Dietmar W. Siemann, PhD, University of RochesterCancer Center, 601 EImwoodAve, Boa"704, Rochester,N Y 14642. Copyright 9 1993 by IA~B.Saunders Company 1053-4296/93/0302-0005505.00/0

population destined for death. The P and Q cell subpopulations hold considerable interest because they manifest different sensitivities to a variety of treatment modalities, including radiation and anticancer drugs I (Fig 2). Although Q cells may be more radiosensitive,~-I~in the case of the chemotherapeutic agents, depending on the drug and the cell line, the therapeutic efficacy may be greater, less, or equivalent in P and Q cell populations.11-17However, comparing only the inherent sensitivities of P and Q cells may be a somewhat tenuous approach given the considerable importance of the repair of potentially lethal damage (PLD). 18,19The latter phenomenon is known to occur to a far greater degree in nonproliferating than in proliferating cells. 18,19 Consequently, the repair of PLD needs to be considered when attempting to extrapolate observations made in tissue culture to the therapeutic outcome in situ. Despite these reservations, in vitro investigations have provided essential mechanistic information concerning the impact of cell age and proliferation state on the efficacy of anticancer agents.

In Vivo Cell Age Response To Anticancer Therapies Unlike the results of tissue culture investigations, the influence of cell cycle position and proliferation state on the in vivo outcome of therapies has been more difficult to ascertain. In the past this has been primarily a consequence of difficulties associated with isolating synchronized cells directly from solid tumors. 2~ However, with the development of effective cell isolation techniques, it now has become possible to determine the in situ drug or radiation response across the cell cycle in a number of solid tumor models? 3 An important consideration in such investigations is the well established fact that treatment with anticancer agents can lead not only to cell-cycle phase-specific cell killing but also to significant perturbations in the distribution of cells within

Seminars in Radiation Oncology, Vol 3, No 2 (April), 1993:pp 99-104

99

100

Siemann and Keng

Because clinical radiation therapy is typically given as a multiple-course fractionated-dose regimen, it was also of interest to evaluate whether prior treatment influenced the cell age response to subsequent radiation exposures. Tumor-bearing mice were irradiated with a test dose of 10 Gy following pretreatment with a dose of 10 Gy administered either as a single fraction or as five daily 2 Gy fractious. The results showed a constant cell age response in tumor cells surviving previous irradiation (Fig 4), implying that the cell cycle response observed in untreated tumors would be predictive of the response of tumors undergoing treatment. Information on the treatment-dependent survival in the isolated cell subpopulations, such as that shown in Figs 3 and 4, not only establishes the importance of the in situ cell age response to a given agent but also provides a means to predict which treatments might be applied successfully in combinationY 4,25 The ability to study cell cycle and proliferatiou effects directly in situ may be of considerable value because tissue culture data may not always predict in vivo tumor responses. For example, when mouse K H T tumor cells were evaluated for their radiation response across the cell cycle, ~6 cells grown and irradiated in vitro showed a fairly conventional response but the cell age response of K H T cells derived from solid tumors and irradiated in vitro was different (Fig 5). In particular, the response of mid S through Ge + M phase cells from tumors was quite unlike that seen for tissue culture cells. These data

o c~

10-I

-

-

Q: G1

S

I 0

GzM

I

I

I

CELLVOLUME

500

1000

1500

2000

2500

(p3]

Figure 1. Age-response function of human hypernephroma cells irradiated with 6 Gy. (Modified from Chang and Keng. 5) the cell cycle. The extent of such perturbations can readily be determined by flow cytometric (FCM) analysis, z4 but FCM data alone do not provide information concerning the viability of the treated cells. Yet, from a therapeutic standpoint, it is clearly important that the cell cycle position of the surviving tumor cells be determined. To overcome this difficulty we have used centrifugal elutriation in conjunction with FCM analysis. This approach allows the isolatiou, from treated tumors, of cells in the various phases of the cell cycle with the same degree of purity as from untreated tumors. 24 Consequently, it is possible to determine, with respect to cell cycle phase, the in situ survival characteristics of tumor cells treated with drugs or radiation (Fig 3).

r

~

.=

O ,i-, O

.

.

----

Exponential - 1 hr

II

Unfed ed p l aplateau t e a u l- h l r h r

I1

O1 r, >

>

.01.1 .01

Figure 2. Dose survival

B

A .001

9

r

10

9

|

20

,

i

i

30

40

9

.001

i

VP-16

9

0

50 (gg/ml)

|

10

9

=

20

9

i

30

9

|

40

9

i

50

curves of human (A) Mel80 and (B) A431 cells in treated in exponential, fed, or unfed plateau phase with VP-16 for l hour. (Data from Chapuis et al. 15)

Cell Kinetics and In Vivo Cancer Therapy

101

to0

"r

&1 &/*

161 r Z

"

"

& " & " &T - &-A" T T ~:/ •

lit ~5a

I

I

n

n

I

I

I

I

,

,

,

n

,

,

r

,

,

I

V

T

I

I

T--T *"

i

I

I

i0-~

I

I

I

I

i%r

~GF

163

I

-& &

3. Clonogenic cell

survival as a function of cell cycle phase for KHT sarcomas treated in vivo. Results shown are (a) 6 mg/kg MIT C, (b) 7.5 mg/kg CCNU, (c) 50 mg/kg CTX, and (d) 15 Gy radiation. (Data from Sicmann and Keng. 94)

I

,

_.--T-T-. T,

Y Figure

I

i

l

J

6

~' i

i

l

S l

8

show that the direct extrapolation of in vitro data to the in vivo situation may not always be valid.

Impact of Proliferating and Nonproliferating Cells on In Situ Therapeutic Response Although tissue culture studies suggest that nonproliferating (Q) cells may be more radiosensitive than proliferating (P) cells, a-m little is known of the actual radiation sensitivities of Q cells in situ. Yet, even if found to be radiosensitive in vivo, it could be argued that Q cells still may pose a significant detriment to radiation therapy. This is because of data suggesting that Q cells from in vitro plateau phase monolayer cultures have an increased capacity for the repair of PLD, relative to P cells, la,m Furthermore, factors that reduce metabolic activity, such as low pH and low glucose, have been shown in vitro to favor a quiescent state and potentiate the capacity for repair ofPLD. 27,2a Although such factors can readily arise in vivo, it should be recognized that the repair of PLD is not a universal phenomenon in tumors. Indeed, preclinical evidence exists indicating that the repair of PLD occurs in some but not all solid tumorsfl 9 Consequently, the ultimate relevance of Q cells in clinical radiotherapy remains unclear. As with radiation, many of the clinically useful anticancer drugs show profound proliferation-dependent cytotoxicities in vitro, l,ulJ5 Thus, tumor cells grown as plateau phase cultures are generally more

i

l

"

~ G n i

l

~

~

I

6 Fraction

" a

a 8

|

S I

" '

I0

'

12

GEM" ' ' 14

16

number

resistant to these drugs than those in the exponential phase of growth (Fig 2). Although the mechanisms responsible for this resistance have not been clearly identified, it has been shown that, at least for some agents, drug-enzyme interactions such as the DNAo topoisomerase II complex, as well as intracellular thiols may be involved? 5J6 To fully comprehend the io o

I

I

]

I

DIIo0(GY)D2(GY)IoIO

t

i

T T r

00n

9

9

o

L" J-""=" -T-- .T IP',,j

l

"r

i

T

o,.o

i r

l ', l

6

GL I

99 |

8

i

S i

!

, oGzM-, I

I

I0 12 Fraction number

I

14

4. Tumor cell survival across the cell cycle in air-breathing mice resulting from a second dose of radiation D2 (10 Gy) given 24 hours after a priming dose D1 (10 Gy). The dashed curve represents survival attributable to D2 alone in the absence of a priming dose. Figure

102

Siemann andKo~g

0.1 In vitro

0.01

In vivo

0 o

-~

0.001

H

G=§

Gz+M

from cycling tumor cells, we have used FCM with acridine orange staining in conjunction with a cell separation technique based on the diffusion properties of the fluorochrome Hoechst 33342. 34 This approach makes use of the fact that after intravenous injection of the Hoechst dye, cells close to the blood vessels stain more brightly than cells further away. 35,3G When such sorted populations were evaluated by flow cytometry using acridine orange, it was possible to identify the presence of Q cell subpopulations in these tumors (Fig 6). The data further indicated a

Figure 5. Cell age response to 10 Gy for KHT sarcoma cells derived from tissue culture (0) or solid tumors (IJ). (Data from Keng et al. 26) impact of the cellular proliferation state on the efficacy of clinical cancer chemotherapy, and to develop more effective drugs against these potentially treatment refractory quiescent tumor cells, it is necessary to study these cells directly in situ. A major difficulty in carrying out Q cell-related therapeutic investigations directly in tumors is that the 3H-thymidine labeling technique, most commonly used in such studies, cannot distinguish tumors composed of uniformly proliferating and nonproliferating subpopulations from tumors that show a wide range of cell proliferation. 3~ In addition, this technique provides no information regarding the viability of the P and Q cell subpopulations. 7,3~However, more recently, a variety of approaches including physical cell separation, acridine orange staining coupled with dual parameter flow cytometry, and the use of monoclonal antibodies against antigens associated with cellular proliferative activities, have shown promise in the identification and isolation of Q cells from cell cultures, multicell spheroids, and solid tumors. 9,31-33 For example, Bauer et a131 used a combination of centrifugal elutriation and flowcytometry to separat e populations of Q cells from multicell EMT6 tumor spheroids. The proportion of Q cells was monitored by flow cytometry with acridine orange staining. Using these techniques, enriched Q cell fractions ( > 82%) were collected. In subsequent experiments the same procedures were used to show that the Q cells isolated from tissue culture or three-dimensional spheroids were more sensitive to ionizing radiation than proliferating cells. ~l~ A slightly different approach has been applied to study Q cells in K H T sarcomas. Because in vivo quiescent tumor cells may be located far away from blood vessels in areas that are relatively hypoxic or otherwise differ metabolically or morphologically

A

[

I

I

I

I

I

I

. i ~

.

i

.

9

.

Red fluorescence

Figure 6. Two-parameter analysis of acridine orange stained KHT tumor-derived cell suspensions obtained after cell sorting on the basis of Hoechst 33342 fluorescence intensity. Results shown are (A) cells from the 20% brightest sort fraction, and (B) cells from the 20% dimmest sort fraction.

103

CeUKinetics and In Vivo Cancer Therapy

greater proportion of nonproliferating cells to be associated with the hypoxic cell subpopulation (Fig 6A v B). In the future, such studies may make feasible the isolation of quiescent cells directly from solid tumors.

,

,

,

,

,

,

,

,

,

,

f ~

~-'~

Targeting Cell-Cycle Phase-Specific Resistance Preclinical investigations using in vivo tumor models in conjunction with effective cell separation and sorting techniques may allow the development of mechanism-based future directions for optimizing treatment strategies. For example, Fig 3 shows the cell-cycle phase-dependent survival characteristics of KHT sarcomas treated in situ with a variety of anticancer agents. Following irradiation, late S and G2 + M cells were observed to be sensitive but GI cells were preferentially spared. The early Gl phase resistance was in part attributable to the presence of hypoxic cells in this phase of the cell cycle.34,37For all chemotherapeutic agents evaluated, survival of cells at the GI/S boundary 24 hours after drug exposure was at a nadir, whereas cells in late S or G2 + M were always drug resistant. 24 However, KHT tumors treated with mitomycin C (MIT C) showed sensitivity in the G1 phase, which likely reflected the reported preferential toxicity of this agent in hypoxic cells.38 Because of the survival characteristics of the tumor cell subpopulations following treatment with M1T C and radiation (Fig 3), the combination of these two agents was evaluated in detail. MIT C was administered 24 hours before radiation therapy to allow for full expression of damage and to avoid the complication of MIT C sensitizing the hypoxic cells to radiation. The resultant cell survival was found to be uniform throughout all cell subpopulations, 24 indicating that this combination reduced the importance of both hypoxia and cell cycle-specificityin the overall treatment response (Fig 7). Isoeffect plot analysis applied to the MIT C-radiation protocol further indicated that this treatment combination led to supra-additive cell kill in the tumor. 24

,

m

w

MITC

~,m

Radiation

~ O 0 - - A T- - . -T- T - - O - - .Q- - O - - a - - t ~tO t

MITC i

24 hr t Radiotloo

-K~

Gi

, 9

S

-- :

i

i

L

I

6

6

K)

R

Fraction

I

GzM.-~ I

14

I

i

16

number

Figure 7. Clonogenic cell survival of KHT sarcoma cells separated by centrifugal elutriation after an in situ treatment consisting of 6 mg/kg MIT C administered 24 hours before a 15 Gy dose of radiation. The dashed curves are redrawn from Fig 3. (Data from Siemann and Keng34) their rapidly proliferating counterparts. Thus, the existence of such tumor cell subpopulations may lead to therapeutic resistance. Alternatively, knowledge of the presence of such tumor cell subpopulations with different properties may allow the development of therapeutic strategies aimed at specifically targeting the different cell subpopulations. In either case it is essential that the in situ impact of cell cycle phase and proliferative state on tumor response to treatment be well understood.

References Conclusions Tumor cells may be preferentially spared from some commonly used anticancer therapies by their cell cycle state at the time of treatment. Similarly, quiescent tumor cells may show intrinsically different sensitivities to drugs or radiation compared with

1. Steel GG: Growth kinetics of tumours. Oxford, UK, Clarendon, 1977 2. Tannock I: Cell kinetics and chemotheraW. A critical review. Cancer Treat Rep 62:117-133, 1978 3. Terasima T, Tolmach LJ: Changes in X-ray sensitivity of HeLa cells during che division cycle. Nature 190:1210-12 l 1, 1961 4. Sinclair WK: X-ray sensitivity during the cell generation cycle

104

S iernann and Keng

of cultured Chinese hamster cells. Radiat Res 29:450-474, 1966 5. Chang AYC, Keng PC: Inhibition of cell growth in synchronized human hy-pernephroma cells by recombinant interferon D and radiation.J Interferon Res 3:379-385, 1983 6. Madoc-Jones H, Bruce WR: Site of action ofcytotoxic agents in the cell life cycle, in Sartorelli AC, Johns DG (eds): Antineoplastic and immunosuppressive agents. Berlin, Germany, Springer-Verlag, I974 7. Dethlefsen LA: In quest of the quaint quiescent cells, in Meyn RE, Withers HR (eds): Radiation Biology"in Cancer Research. New York, NY, Raven, 1980, pp 415-435 8. Luk CK, Keng PC, Sutherland RM: Quiescent cell isolation in multicellular tumor speroids using centrifugal elutriation. Cancer Res 42:72-78, 1982 9. Ng CE, Keng PC, Sutherland RM: Characterization of radiation sensitivity of human squamous carcinoma A431 cells. BrJ Cancer 56:301-307, 1987 10. Wilson KM, Keng PC: Radiation-induced DNA damage and repair in quiescent and proliferating human tumor cells in vitro. IntJ Radiat Biol 55:385-395, 1989 11. Bhuyan BK, Fraser TJ, Day KJ: Cell proliferation kinetics and drug sensitivity of exponential and stationary populations of cultured LI210 cells. Cancer Res 37:1057-1063, 1977 12. Clarkson B: The survival value of the dormant state in neoplastic and normal cell populations, in Clarkson BD, Baserga R (eds): Proliferation in Animal Cells. Cold Spring Harbor, NY, Cold Spring Harbor Laboratory, 1974, pp 945972 13. Dexter DL, Calabresi P: Intraneoplastic diversity. Biochem Biophys Acta 695:97-1 I2, 1982 14. Chapuis JC, Keng PC, Siemann DW: Activity of etoposide (VP-16) in human tumor cells under different growth conditions. IntJ Radiat Oncol Biol Phys 20:373-375, 1991 15. Chapuis JC, Keng PC, Siemann DW: Activity of etoposide (VP-16) and teniposide (VM-26) in exponential and plateau phase human tumor cell cultures. Anti-Cancer Drugs (3:245252, 1992) 16. Sullivan DM, Glisson BS, Hodges PK, et al: Proliferation dependance of topoisomerase II mediated drug action. Biochem 25:2248-2256, 1986 17. Chow KC, Ross WE: Topoisomerase-specific drug sensitivity in relation to cell cycle progression. Mol Cell Biol 7:3 119-3123, 1987 18. Hahn GM, Rockwell S, Kallman RE, et al: Repair of potentially lethal damage in vivo in solid tumor cells after X-irradiation. Cancer Res 34:351-354, 1974 19. Phillips RA, Tolmach LJ: Repair of potentially lethal damage in X-irradiated HeLa cells. Radiat Res 29:413-432, 1966 20. Grdina DJ, Peters LJ,Jones s, et al: Separation of cells from a murine fibrosarcoma on the basis of size. i. Relationship between cell size and age as modified by growth in vivo or in vitro.J Natl Cancer Inst 61:209-214, 1978 21. Meistrich MI, Grdina DJ, Meyn RE, et al: Separation of cells from mouse solid tumors by centrifugal elutriation. Cancer Res 37:4291-4296, 1977

22. Keng PC, Wheeler KT, Siemann DW, et al: Direct synchronization of cells from solid tumors by centrifugal elutriation. Exp Cell Res 134:15-22, 1981 23. Keng PC, Siemann DW, Lord EM: Separation of malignant cells from host cells using centrifugal elutriation, in Pretlow TP, Pretlow TG (eds): Cell Separation: Methods and Selected Applications. Orlando, FL, Academic, 1987, pp 51-74 24. Siemann DW, Keng PC: Responses of tumor cell subpopulations to single modality and combined modality responses. NCI Monographs 6:101-105, 1988 25. Siemann DW, Maddison K, Wolf K, et al: In vivo interaction between radiation and 1-(2-chloroethyl)-3--cyclohexyl-l-nitrosourea in the presence or absence of misonidazole in mice. Cancer Res 45:198-202, 1985 26. Keng PC, Siemann DW, Wheeler KT: Comparison of tumor age response to radiation for cells derived from tissue culture or solid tumors. BrJ Cancer 50:519-526, 1984 27. Varnes ME, Dethlefsen LA, BiaglowJE: The effect of pH on potentially lethal damage recovery in A549 cells. Radiat Res 108:80-90, 1986 28. Koch CJ, Meneses JJ, Harris JW: The effect of extreme hypoxia and glucose on the repair of potentially lethal and sublethal damage by mammalian cells. Radiat Res 70:542-551, 1977 29. Siemann DW: Do in vitro studies of potential lethal damage repair predict for in situ results? Int J Radiat Biol 56:567-571, 1989 30. Dethlefsen LA, Bauer KD, Riley RM: Heterogeneous cells in solid tumors measured by acidic acridine orange staining. Cytometry 1:89-97, 1980 31. Bauer KD, Keng PC, Sutherland RM: Quiescent cell isolation in multicellular tumor spheroids using centrifugal elutriation. Cancer Res 42:72-78, 1982 32. GerdesJ, Lemke H, Baisch H, et al: Cell cycle analysis of a cell proliferation-associated human nuclear antigen defined by the monoclonal antibody Ki-67. J Immunol 133:1710-1715, 1984 33. Sasaki K, Murakami T, Kawasaki M, et al: The cell cycle associated change of the Ki-67 reactive nuclear antigen expression..] Cell Physiol 133:579-584, 1987 34. Siemann DW, Keng PC: Characterization of radiation resistant hypoxic cell subpopulations in KHT sarcomas II Cell sorting. BrJ Cancer 58:296-300, 1988 35. Chaplin DJ, Durand RE, Olive PL: Cell selection from a murine turnout using the flouorescent probe Hoechst 33342. BrJ Cancer 5 t :569-572, 1985 36. Olive PL, Chaplin DJ, Durand RE: Pharmacokinetics, binding, and distribution of Hoechst 33342 in spheroids and murine tumors. BrJ Cancer 52:739-746, 1985 37. Siemann DW, Keng PC: Characterization of radiation resistant hypoxic cell subpopulations in KHT sarcomas I. Centrifugal elutriation. BrJ Cancer 55:33-36, 1987 38. Sartorelli AC: The role of mitomycin antibiotics in the chemotherapy of solid tumors. Biochem Pharmacol 35:67-69, 1986