| Abstract |
This thesis records the results of a series of experiments that were
designed to examine the biology of human malignant melanoma cells cultured
in vitro. The studies were so planned as to document phenotypic
differences that exist between melanomas, to define respects in which
melanoma cell differentiation could be modulated and to correlate
biochemical variability with in vivo behaviour as measured in the nude
mouse. Melanoma cell lines were established from biopsy material obtained
from 7 patients at Groote Schuur Hospital. Two of these lines synthesized
tyrosinase and melanin at a rate that was directly related to cell density.
The five remaining lines did not pigment. All of the lines showed
aneuploidy; 5 of the 7 showed anchorage-independent growth; and 6 of the 7
grew as lethal tumours in nude mice. As has been found with all other
melanomas studied, these cells released a plasminogen activator that was
chemically and immunologically identical to tissue activator. One of the
lines proved to be an exception to this general rule in that it
synthesized urokinase-type enzyme. Unlike most other human cells cultured
in vitro, melanoma cells proved to be relatively refractory to hormonal
stimuli. Addition of estrogen, progesterone, testosterone, dexamethasone
or melanocyte-stimulating hormone to the culture medium had very little
effect on cellular release of plasminogen activator, upon cell growth, or
upon cellular morphology. Although remarkably resistant to hormonal
influences, cellular release of plasminogen activator did appear to be
inhibited to a striking degree by cocultivation with normal skin
fibroblasts. This observation led to the discovery of a phenomenon in
which fibroblasts of many types bound tissue-type plasminogen activator
and so removed it from the medium. This was accompanied by an apparent
change in molecular weight of the melanoma cell enzyme from 72K daltons to
approximately 115K daltons, suggesting the presence of a 40-50K binding
molecule. In an attempt to influence in vitro differentiation, the tumour
promoter tetradecanoylphorbol acetate, and the differentiation-inducing
retinoid, retinoic acid, were added to the two pigmenting cell lines. The
effects of these compounds on induction of tyrosinase activity,
morphological change or plasminogen activator release differed. In the one
cell line, tetradecanoylphorbol acetate caused morphological maturation
with a decrease in the rate of plasminogen activator release and no
obvious effect upon pigmentation. This line was relatively resistant to
the action of retinoids. The other pigmenting line responded hardly at all
to the tumour promoter. Retinoic acid, on the other hand, inhibited the
induction of tyrosinase activity, yet caused an inhibition of growth and
plasminogen activator release. A number of interesting observations could
be made in experiments in which melanoma cells were inoculated into nude
mice. Firstly, the growth rate of the tumours in vivo correlated poorly
with the doubling times of the corresponding cells cultured in vitro.
Secondly, despite a marked inhibitory effect of fibroblasts on plasminogen
activator in vitro, coinjection of fibroblasts and melanoma cells in vivo
greatly enhanced tumour growth when small tumour cell inocula were used
and shortened the latent period for tumour appearance with larger inocula.
Thirdly, melanomas growing in nude mice differed strikingly in their
ability to elicit a desmoplastic response. Tumours in which large amounts
of host connective tissue were deposited tended to be heavily contaminated
with murine fibroblasts when re-established in vitro. This contamination
was not seen with tumours that contained very little connective tissue.
These results point to the existence of a melanoma-associated fibrogenic
factor. Finally, by excision of the primary tumour, it was possible to
avoid death of the animal from local complications and so allow time for
metastases to develop. In three mice, metastatic melanoma deposits could
be detected by this device, so establishing a protocol for the use of nude
mice as valid models for the experimental study of metastatic spread of
human tumours.
|
|---|
