Table of Contents:
The incidence of head and neck squamous cell carcinomas has steadily increased during
the last years. This tumor entity is characertized by its invasive character, its tendency
for early metastasis and its poor prognosis. To achieve an optimal treatment one needs
to understand the complex biological mechanisms of invasion and metastasis. The
family of zinc-dependent endopeptidases, the matrix metalloproteinases, represent a
group of proteins which are capable of remodelling the extra cellular matrix. They
contribute to the intra- and extravasation and to the local invasion.
In this thesis the co-expression of several MMPs in HNSCC-cell lines and tissue
samples was investigated using western blot and immunohistochemical methods. The
tumor-stroma-interference of tumor tissues as well as HNSCC-cell lines was regarded
with special interest by means of immunohistochemistry and immunozytochemistry. In
addition, the influence of the in vitro invasion on the MMP-9 expression and activity
was examined. For this purpose co-cultures of two cell lines were applied. The gelatine
zymography was used to determine the activity of the gelatinase MMP-9. The in vitro
invasion of cell lines was quantified by means of electrical resistance breakdown assay
(ERBA). The experiment has been carried out on 12 HNSCC-cell lines and 15 HNSCCtissue
The expression of MMP-1, -3, -7, -9, -11 und -15 in HNSCC-cell lines and tissue
samles was shown by immunohistochemical staining and WB analysis. The
immunohistochemical analysis revealed MMP expression in tumor cells as well as in
fibroblasts from the surrounding tissue. Healthy oral mucosa served as control tissue
and displayed in immunohistochemical staining less MMP expression compared to the
analyzed tumor tissues. Immunohistochemical staining with MMP-9 clearly indicated
an increased expression at the tumor-stroma-interference. Based on this the role of the
gelatinase MMP-9 in HNSCC was further investigated. Immunozytochemical analysis
of the tumor-stroma contact zone, which was imitated using a co-culture of fibroblasts
and tumor cells in vitro, followed. Partially NIH-3T3 fibroblasts expressed MMP-9
stronger than tumor cell lines. The immunofluorescence staining sporadically detected
an increased MMP-9 expression at the tumor-stroma-interference site. However, WB
analysis did not show an increased MMP-9 expression after co-culturing the HNSCCcell
lines with NIH-3T3. In addition, the zymography revealed no significant
gelatinolytic activity in any of the samples. The invasiveness of single cell-lines was
compared to invasiveness of co-cultured cells by the means of ERBA. The co-culture
with fibroblasts did not result in a change of invasiveness. It was observed that
HNSCC-cell lines altogether grew less invasive in vitro than in vivo.
The gained results suggest that an induction of the MMP-9 expression at the invasive
tumor front is present in vivo. This observation indicates an interaction between tumor
cells and the surrounding tissue, which seems to play a role in the MMP secretion.
However, it could not be managed to induce the MMP-9 expression in cell lines by
means of in vitro invasion. Presumably, numerous factors from the surrounding tissue
play a role in the invasion, metastasis and in the induction of the MMP-9 expression,
which cannot be copied sufficiently by co-culture.
The present study contributed to the elucidation of the importance of MMPs in HNSCC.
In summary, MMPs are expressed in HNSCC-cell lines as well as in tumor tissues. A
difference in the MMP expression between tumor- and control tissue with a greater
expression in tumor tissues could be observed. Furthermore, MMP-9 production seems
to be intensified at the tumor margins. This can be interpreted as an evidence for an
interaction between tumor cells and the surrounding tissue. The degradation of the
adjacent tissue and tumor growth takes place at the aggressive tumor front. The
influence of the MMP-9 expression at the invasive tumor front could - although only
sporadically - be shown for co-culture of fibroblasts with tumor cells. This circumstance
implies that the interaction between multiple factors from the tumor cells themselves as
well as from the surrounding tissue is essential for invasion and metastasis.
In order to improve diagnostic tools and therapies it is essential to explore the tumorstoma
interference in further studies. The development of a feasible, lifelike model
suitable for HNSCC, which is capabale of mimicing the invasion of tumor cells into
surrounding healthy tissue, is desirable to increase our understanding of this process.