Download Limited Penetration of Methotrexate into Human

[CANCER RESEARCH 40, 3665-3668.
0008-54 727 80 /0040-OOOOS02.00
October 1980]
Limited Penetration of Methotrexate into Human Osteosarcoma Spheroids
as a Proposed Model for Solid Tumor Resistance to Adjuvant
Chemotherapy1
Gary W. West, Ralph Weichselbaum,
and John B. Little
Harvard School of Public Health, and Joint Center lor Radiation Therapy. Department ol Radiation Therapy. Harvard Medical School. Boston. Massachusetts
02115
ABSTRACT
It is generally accepted that oxygen has a limited ability to
diffuse into solid tumor masses. However, the question of the
ability of chemotherapy agents to penetrate solid tumor masses
has not been evaluated. This clearly would have an impact on
the ability of chemotherapy to control microscopic disease
during the "avascular" phase of growth. An attempt was made
to evaluate the ability of methotrexate to penetrate solid tumor
masses when grown in three dimensions (spheroids). Since
methotrexate is used in the clinical management of human
osteosarcoma, we chose this drug-tumor combination for our
studies. This was done by growing human osteosarcoma cells
into spheroids and exposing spheroids of various sizes to
tritiated methotrexate. Audioradiographs were obtained from
sections through the center of spheroids of various sizes. Our
findings suggest that methotrexate has a limited ability to
penetrate into avascular tumor masses when grown in three
dimensions. This is most evident when the tumor masses are
approximately 250 /im and larger in diameter. In addition, we
compared the degree of penetration of methotrexate to the
growth fraction of the tumor, as measured by tritiated thymidine, and found that the growth fraction was much greater than
the fraction of cells reached by methotrexate. We conclude
that the limited ability of methotrexate to penetrate solid tumor
masses offers an alternative explanation for the limited effec
tiveness of methotrexate when used as an adjuvant for osteo
sarcoma. We question whether the established biochemical
mechanisms for methotrexate resistance are comprehensive
explanations for its limited clinical effectiveness.
INTRODUCTION
The goal of cancer chemotherapy is to eradicate clinically
evident or microscopic métastases.The use of cytotoxic drugs
to treat microscopic métastases after complete surgical re
moval of the primary tumor is referred to as "adjuvant chemo
therapy." MTX is used in combination with other drugs as an
adjuvant to surgery in patients with osteosarcoma and appears
to increase survival (3, 8). Nevertheless, 50% of patients with
osteosarcoma die from widespread metastatic disease despite
adjuvant treatment (19). The limited effectiveness of chemo
therapy is frequently attributed to the known ability of tumor
cells to develop resistance to a particular drug (11). To date,
3 mechanisms for tumor resistance to MTX' have been identi' This work was supported in part by Grant CA-2184802
from the National
Cancer Institute.
1 The abbreviations used are: MTX. methotrexate; DHFR, dehydrofolate reductase; MTS. multicellular tumor spheroid(s): dThd, deoxythymidine; EBSS.
Earle's balanced salt solution.
Received October 1, 1979; accepted July 14. 1980.
OCTOBER
fied: (a) alterations in the carrier-mediated active transport of
MTX across cellular membranes occur (1 7); (b) alterations in
the structure of the target enzyme of MTX, DHFR occur, such
that the high affinity for MTX is lost (6, 7); and, more recently,
(c) selective gene amplification develops which results in an
increase in DHFR production.
Despite their value for providing biochemical bases for MTX
resistance, it is important to recognize that these mechanisms
were discovered by growing cells under conditions which per
mit uniform exposure of cells to increasing concentrations of
drug for prolonged periods of time. However, this is not the
case for the clinical administration of MTX for adjuvant treat
ment in which high doses of MTX are given for relatively short
periods of time. Although in vitro or in vivo conditions of drug
exposure can be manipulated to select or produce resistant
cells, the number of resistant cells obtained are relatively rare
compared to wild-type cells (12). Therefore, if these transport
and enzymic mechanisms were the sole determinants of MTX
resistance, one might predict adjuvant MTX therapy to prove
more successful. Moreover, many other clinical studies on
various cytotoxic agents for a variety of solid tumor cell types
have been less effective than preliminary in vitro studies pre
dicted (22). Therefore, other mechanisms of drug resistance
probably exist.
It is known that solid tumors require adequate diffusion of
nutrients for tumor growth (2, 18), but the diffusion (whether
passive or facilitated) of cytotoxic agents into solid tumor
masses has not been investigated. Certainly, for solid tumors,
adequate drug diffusion to all cells is fundamental for cytotoxicity. The recent development of simple techniques for the
growth of MTS in vitro has permitted us to investigate drug
diffusion as a possible limitation to chemotherapy (5, 21 ). Since
MTX is used extensively in the treatment of human osteosar
coma, we chose this drug-tumor combination for our investi
gations. We report here experimental evidence to support the
hypothesis that diffusion gradients to cytotoxic agents develop
relatively early during the growth of solid tumors in vitro (i.e.,
with a diameter of approximately 250 firn), and we propose that
either passive or facilitated diffusion is an important variable in
determining the effectiveness of cytotoxic chemotherapy.
MATERIALS
AND METHODS
Differences of MTX Diffusion in Human Osteosarcoma
Cells Grown in 2 Versus 3 Dimensions. Extracellular MTX
enters cells by actively traversing the cell membrane in asso
ciation with a carrier (facilitated diffusion) (4, 20). Because of
its high affinity for DHFR, all intracellular MTX is bound to
DHFR until its binding sites are saturated. Initially, the drug
moves only intracellularly (influx), and its rate of entry is de-
1980
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1980 American Association for Cancer Research.
3665
G. W. West et al.
pendent on both extracellular drug concentration and mem
brane permeability. However, once all DHFR binding sites are
saturated, exchangeable or (nonbound) intracellular MTX ac
cumulates, and an equilibrium is established. Since DHFR is
concentrated in the cytoplasm, little or no MTX enters the
nucleus (16). This characteristic of MTX allowed us to identify
the diffusion patterns of MTX and to distinguish its diffusion
from that of thymidine. In all experiments, a standard exposure
concentration of 0.33 x 10~6 M high-specific activity (30 Ci/
mmol) [3H]MTX at 10 ¿iCi/mlwas used. The exposure concen
tration of [3H]dThd (49.5 Ci/mmol) at 10 /iCi/ml was 0.2 x
10~6 M. We chose this concentration and specific activity to
simulate therapeutic concentrations and to insure adequate
autoradiography.
Human osteogenic sarcoma tumor cells were obtained from
Dr. E. Lloyd of the Argonne National Laboratory with permission
from Dr. Jörgen Fogh of the Sloan-Kettering Memorial Cancer
Center (13). Cells were grown in Eagle's minimal essential
medium supplemented with 10% fetal calf serum (Microbiolog
ical Associates, Bethesda, Md.), 9 mg-D-glucose per liter, 0.66
mg sodium pyruvate per liter and 50 fig gentamicin per ml
(Schering Corp., Kenilworth, N. J.) at 37°in an atmosphere of
95% air:5% CO2. For the [3H]MTX uptake in monolayer exper
iments, exponentially growing stock cultures were treated with
0.25% trypsin in calcium- and magnesium-free EBSS for 3 min.
Two-hundred and fifty dissociated cells were counted in a
hemocytometer, plated in 35-mm Falcon tissue dishes, allowed
to grow to colonies of various sizes, and then exposed to 0.33
x 10~6 M [3H]MTX (30 Ci/mmol; Moravek Biochemicals, City
of Industry, Calif.) at 10 ftCi/ml for 48 hr. Radioactive medium
was safely discarded, and colonies were washed 5 times with
EBSS at 4°. This procedure should remove all extracellular
MTX but should have no effect on intracellular (enzyme-bound)
MTX (9). The effectiveness of the wash procedure was deter
mined by measuring the amount of radioactivity in a sample of
the final wash. No radioactivity (above background levels) was
detected. Colonies were fixed for 10 min with methanoliacetic
acid (3:1), dried for 4 hr, and then dipped into NTB-2 (Kodak)
nuclear track emulsion diluted 1:1 in complete darkness. Emul
sion-covered plates were dried overnight, exposed at 4°in a
desiccator light-sealed box for 21 days, developed with Kodak
D19-B (5 min), and then fixed, and photomicrographs were
taken.
For spheroid formation, approximately 105 cells were har
vested from monolayer stock cultures and seeded with 15 ml
of complete medium (i.e.. Eagle's minimum essential medium
supplement as in the monolayer experiments) into 100-mm
Falcon dishes that had been previously base coated (2 to 3
mm) with 0.75% Noble agar(Difco Laboratories, Detroit, Mich.)
in complete medium to prevent cell attachment (21). These
dishes were then incubated (95% air:5% CO2 at 37°) for
approximately 2 weeks for spheroids 50 /im in diameter and
approximately 5 weeks for speroids 250 /im in diameter. Five
ml of medium were exchanged twice per week for feeding.
After appropriate incubation times, all spheroids were approx
imately 50 or 250 jum in diameter. Approximately 15 to 20 of
these spheroids were transferred to 35-mm Falcon dishes base
coated with Noble agar and complete medium as described
above for exposure to 0.33 x 10~6 M [3H]MTX for 48 hr.
Similarly, approximately 15 to 20 of the 250-fim spheroids
were exposed to 0.2 x 10~6 M [3H]dThd for 48 hr. After these
3666
exposures, the respective radioactive media were safely dis
carded, and the spheroids were washed 9 to 10 times with
EBSS at 4°to ensure removal of all extracellular [3H]MTX or
[3H]dThd. The effectiveness of the wash procedure was deter
mined in a manner similar to that for the monolayer [3H]MTX
uptake experiments. The spheroids were then fixed in 2.5%
gluteraldehyde for 30 min, washed again 5 to 6 times with
EBSS at 4°to prevent negative chemography during autora
diography, dehydrated in 70, 80, 90, and 100% ethanol for 15
min in each dilution, and embedded in a Beem capsule, in
accordance with the instruction for the Polysciences JB-4
embedding kit. A Sorvall JB-4 glass microtome was used to
cut 2-/im sections from the center of the spheroids. These
sections were mounted on glass slides and processed for
autoradiography, as in the monolayer experiments. The autoradiographs were then stained with hemotoxylin and eosin, and
photomicrographs were taken.
RESULTS
Fig. '\A is an autoradiograph
of a small colony of human
osteosarcoma cells grown in monolayer after exposure to 0.33
x 10~6 M [3H]MTX for 48 hr. As expected for cells grown in
monolayer, [3H]MTX appears to be distributed uniformly in the
cytoplasm of all cells. Since the cells were washed thoroughly
after exposure to [3H]MTX, all extracellular [3H]MTX should be
absent; therefore, we interpret this autoradiograph to be an
accurate representation of intracellular MTX bound to DHFR.
Fig. 1/\ indicates that cells exposed to this concentration of
[3H]MTX maintain their ability to transport and take up MTX,
which is an important requirement for drug diffusion when
tumor cells are grown in 3 dimensions. If the cells are allowed
to grow to larger colony sizes (data not shown), there is more
crowding of cells but a similar distribution of [3H]MTX. Thus, in
2-dimensional growth, there are no limitations to drug diffusion
since the surface of each cell is always exposed to equal
concentrations of MTX. Under these conditions, MTX would be
expected to have a marked cytotoxic effect.
Fig. 1S is an autoradiograph of a 2-p.m section taken from
the center of an MTS of human osteosarcoma cells grown to
approximately 50 ¿imin diameter and exposed, as in Fig. 14,
to 0.33 x 10~6 M [3H]MTX for 48 hr. Although [3H]MTX is not
homogeneously distributed throughout the MTS, evidence of
uptake can be seen both peripherally and centrally within the
spheroidal mass. Thus, MTX reaches all cells comprising tumor
spheroids of this size, and all cells within the mass are vulner
able to its cytotoxic effect. Although cytoplasmic uptake is not
as distinct as in Fig. 1/\, [3H]MTX appears to surround the
nuclei of the cells comprising the tumor mass, and no evidence
of [3H]MTX within the nuclei can be seen.
If the MTS are allowed to grow to approximately 100 /im and
then exposed to [3H]MTX, the drug tends to concentrate toward
the periphery of the tumor mass, but evidence of central
[3H]MTX can still be seen (data not shown). However, if the
MTS are allowed to continue to grow in size before exposure
to [3H]MTX, the difference between central and peripheral
distribution of the drug increases until almost all the drug is
concentrated to the outermost peripheral cell layers, as dem
onstrated in Fig. 1C.
Fig. 1C is an autoradiograph of a 2-p.m section taken from
the center of an MTS allowed to grow to approximately 250
CANCER
RESEARCH
VOL. 40
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1980 American Association for Cancer Research.
Limited Diffusion and Drug Resistance
jum in diameter before exposure to [3H]MTX for 48 hr. In this
size range of MTS, drug diffusion is essentially limited to the
outer 2 to 3 cellular layers of the spheroidal mass. In fact, on
higher magnification (data not shown), we can detect no evi
dence of [3H]MTX above background levels in the center of the
spheroid. Also, as in Fig. 15, the drug appears to surround the
nuclei of the peripheral cells, which we interpret to be cytoplasmic concentration. Thus, in spheroids of this size, the
peripheral cells are most vulnerable to the cytotoxic effects of
MTX, while the central cells, which comprise the largest part of
the spheroid mass, have either a limited vulnerability to the
drug or may even be completely protected from the cytotoxic
agent.
Since it can be argued that the central cells of MTS in the
250-jum range are not viable, the fact that MTX does not diffuse
to reach these cells has little relevance to its ability to eradicate
métastasesof this size. To answer this question, we exposed
human osteosarcoma MTS in this size range to 0.2 x 10~6 M
[3H]dThd (10 /xCi/ml) for 48 hr to ensure continuous labeling
over approximately 2 doublings.
Fig. 1D is an autoradiograph of a 2-fim section taken through
the center of a human osteosarcoma MTS grown to approxi
mately 280 /im in diameter and then exposed to [3H]dThd.
When this autoradiograph is compared to that in Fig. 1C, it can
be seen that both the inner and outer cells of the MTS are
capable of DNA synthesis, as indicated by the marked nuclear
uptake of [3H]dThd throughout the MTS. Thus, the inner cells
not reached by MTX because of its limited diffusion (either
passive or facilitated) are viable. Stated differently, MTX does
not reach all the cells that make up the growth fraction of the
tumor mass.
DISCUSSION
chemotherapy should be very effective and perhaps curative.
However, once tumor cell populations develop 3-dimensional
growth configurations (linear growth phase), MTX is no longer
uniformly distributed to all cells within the tumor mass (Fig.
1B). The cytotoxic effect of MTX would be expected to become
increasingly ineffective with increasing tumor size until a diffu
sion gradient develops (Fig. 1C) during the late "linear" or
"dormant" phase of tumor growth, at which time only the most
peripheral outer layer of cells
effects of MTX. Since Fig. 1D
are also viable, MTX would not
tumors in this phase of growth.
ized and "breaks-out"
of the
is vulnerable to the cytotoxic
indicates that the central cells
be expected to be curative for
If the tumor becomes vascular
dormant phase, growth again
increases rapidly, and the effectiveness of cytotoxic agents
may then be limited by the magnitude of the number of tumor
cells. However, in this circumstance, the disease usually be
comes clinically evident and patients are no longer considered
candidates for standard "adjuvant chemotherapy"
regimens.
The maximum size of tumors during the "avascular" period of
tumor growth is well below the level that is detectable clinically
by available technology. Nevertheless, tumors of this size have
heterogeneous growth characteristics and develop heteroge
neous microenvironments which not only regulate their growth
but also affect the ability of cytotoxic agents to diffuse uniformly
to all viable cells. Our experiments indicate that MTX diffusion
(either passive or facilitated) into solid tumors of micrometastatic size is dependent on the relative size of the tumor mass.
Although the transport and enzyme mechanisms already iden
tified for MTX resistance are important biological findings with
profound clinical implications, the passive or facilitated diffu
sion of MTX into solid tumors of micrometastatic size may be
of equal clinical importance or at least offers another explana
tion for the variation of the efficacy of MTX adjuvant chemo
therapy in patients with primary osteosarcoma.
We propose that micrometastases can be thought of as the
prevascular or avascular period of tumor growth, first described
ACKNOWLEDGMENTS
by Folkman (1 ). This period of early tumor growth is divided
into 3 phases: (a) a brief phase of exponential growth (0 to 150
The authors wish to thank John Nove and Anne Schmidt for their expert
/im in diameter) during which all the cells have equal access to technical assistance. Dr. Hatsumi Nagasawa provided superb technical direction
as well as stimulating discussion.
nutrients; (6) a second phase of linear growth (150 to 500 /¿m
in diameter) beginning with the appearance of central necrosis,
probably due to a limitation of oxygen diffusion; and (c) a
REFERENCES
"dormant phase" which begins when growth plateaus (1 to 2
mm in diameter) and beyond which a tumor cannot grow unless
it becomes vascularized. Without vascularity, tumor growth in
all of these phases is "self-regulated"
by the centrifugal diffu
sion (passive or facilitated) of catabolites (growth inhibitors)
and the centripetal diffusion (passive or facilitated) of nutrients
(2). This description of early tumor growth can be used as a
model for micrometastases in which the regulation of cell kill is
greatly determined by how well cytotoxic agents can diffuse
(passive or facilitated) centripetally to the center of the tumor
mass, just as growth is greatly determined by the passive or
facilitated diffusion of nutrients.
Our observations of MTX diffusion are significant in that they
may represent drug tumor interactions during the different
phases of early or "prevascular"
metastatic tumor growth in
vivo. For example, cells grown in monolayer (Fig. 1/\) may be
analogous to the very earliest phase (exponential growth) of
métastases during which MTX is uniformly distributed to all
cells and is highly cytotoxic. In this circumstance, adjuvant
1. Folkman, J Tumor angiogenesis: therapeutic implications. N. Engl J. Med..
285. 1182-1186,
1971.
2. Folkman, J., and Hochberg, M. Self-regulation of growth in three dimensions.
J. Exp. Med., 138. 745-753, 1973
3. Gehan, E. A., Sutow, W. W., Uribe-Botero. G.. Romsdahl, M.. and Smith, T.
L. Osteosarcoma: the M. D. Anderson experience. 1950-1974. Prog. Can
cer Res. Ther.. 6. 271-282, 1978.
4. Goldman, l D.. Lichtenstein, N. S., and Oliverio, V. T. Carrier-mediated
transport of the folie acid analogue, methotrexate, in the L1210 leukemia
cell. J. Biol. Chem.. 243 5007-5017,
1968.
5. Haji-Karim, M.. and Carlsson. J. Proliferation and viability in cellular sphe
roids of human origin. Cancer Res., 38.1457-1464,
1978.
6. Hanggi, U.. and Littlefield, J. W. Isolation and characterization of the multiple
forms of dihydrofolate reducÃ-ase from methotrexate-resistant
hamster cells
J. Biol. Chem.. 249 1390-1397. 1974.
7. Harding. N. G. L.. Martelli, M. F., and Huennekens, F. M. Amethopterininduced changes in the multiple forms of dihydrofolate reducÃ-ase from
L1210 cells. Arch. Biochem. Biophys., 137. 295-296. 1970.
8. Jaffe, N.. and Frei, E., III. Osteogenic sarcoma: advances in treatment
Cancer (Phila.), 26. 351-359, 1976.
9. Johnson. L. F., Fuhrman. C. L.. and Abelson. H. T. Resistance of resting
3T6 mouse fibroblasts to methotrexate cytotoxicity. Cancer Res. 38. 24082412, 1978.
10. Kellems. R. E.. Alt. F. W.. andSchimke, R. T. Regulations of folate reducÃ-ase
synthesis in sensitive and methotrexate-resistant
sarcoma 180 cells J. Biol.
OCTOBER 1980
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1980 American Association for Cancer Research.
3667
G. W. Westetal.
Chem.,257:6987-6993,1976.
11. Mihich, E. (ed). Drug resistance and selectivity. Biochemical and cellular
bases. New York: Academic Press, Inc., 1973.
12. Nakamura. H., and Littlefield, J. W. Purification, properties, and synthesis of
dihydrofolate reducÃ-ase from wild type and methotrexate-resistant
hamster
cells. J. Biol. Chem.. 247. 179-187, 1972.
13. Porter, J., and Saksela, E. Two established in vitro cell lines from human
mesenchymal tumors. Int. J. Cancer, 2. 434-447,1967.
14. Schimke. R. T., Randal. J. K., Alt, F. W., and Kellems. R. F. Gene amplifi
cation and drug resistance in cultured murine cells. Science (Wash. D. C.).
202. 1051-1055,
1978.
15 Shields, R. Methotrexate resistance by gene amplification Nature (Lond.),
273. 269. 1978.
16. Sirotnak, F. M., and Donsbach. R. C. The intracellular concentration de
pendence of antifolate inhibition of DNA synthesis in L1210 leukemia cells.
Cancer Res., 34: 3332-3340,
1974.
17. Sirotnak, F. M.. Kurita. S.. and Hutchison, D. J. On the nature of a transport
alteration determining resistance to amethopterin in the L1210 leukemia.
Cancer Res.. 28: 75-80, 1968.
18. Thomlinson, R. H.. and Gray, L. H. The histological structure of some human
lung cancers and the possible implications for radiotherapy. Br. J. Cancer.
9. 539-549, 1955.
19. Uribe-Botero, G., Russell, W. O.. Sutow, W. W.. and Martin. R. G. Primary
osteosarcoma of bone: a clinicopathologic investigation of 243 cases, with
necropsy studies in 54 (Abstract). J. Am. Med. Assoc., 238. 1576, 1977.
20. Wilbrandt, W., and Rosenberg, T. The concept of carrier transport and its
corollaries in pharmacology. Pharmacol. Rev.. 13. 109-183, 1961.
21. Yuhas. J. M., Li. A. P.. Martinex, A. O., and Ladman, A. J. A simplified
method for production and growth of multicellular tumor spheroids. Cancer
Res., 37. 3639-3643,
1977.
22. Zubrod, C. G. Selective toxicity of anticancer drugs: Presidential Address.
Cancer Res., 38: 4377-4384,
1978.
v*.
***
¿r•
/
T
*s
"•
-'
- -v •
. •*'
•< > *
»*•*>
,
*•
200JJ
Fig. 1. A, autoradiograph of a small colony of human osteogenic tumor cells grown in monolayer and then exposed to 0.33 x 10~6 M [3H]MTX at 10/iCi/ml for
48 hr. 8 and C. autoradiographs of 2-jim sections taken from the center of human osteosarcoma tumor cells grown into MTS with diameters of approximately 50 and
250 (im. respectively, and then exposed to 0.33 x 10~6 M [3H]MTX at 10 /iCi/ml for 48 hr. D, autoradiograph of a 2-fim section taken from the center of a human
osteogenic sarcoma MTS grown to a diameter of approximately 280 ¡anand then exposed to 0.2 x 10"6 M [3H]dThd at 10 /iCi/ml for 48 hr. Except for exposure to
[3H)dThd, these spheroids were grown and processed identically to those in ßand C.
3668
CANCER
RESEARCH
VOL.
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1980 American Association for Cancer Research.
40
Limited Penetration of Methotrexate into Human Osteosarcoma
Spheroids as a Proposed Model for Solid Tumor Resistance to
Adjuvant Chemotherapy
Gary W. West, Ralph Weichselbaum and John B. Little
Cancer Res 1980;40:3665-3668.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/40/10/3665
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1980 American Association for Cancer Research.