Biology Reference
In-Depth Information
blood vs. lymphatic vessels (Vainionpaa et al.
2007
). BM constituents in normal and
angiogenic vessels also can differ quantitatively and, maybe more importantly,
qualitatively, e.g., can include new proteinaceous molecules (Cao
2005
; Vainionpaa
et al.
2007
). This broad compositional array suggests that more than a few MMPs
could be essential for degradation of any number of BM components to create the
points of entry for intravasating cells.
Since endothelial lumen and tube formation within collagenous matrices in vitro
critically depend on MT1-MMP, the notion that MT1-MMP governs proteolysis of
vascular BM in tumors in vivo has been widely accepted. Despite overwhelming
evidence for the role of MT1-MMP-mediated localized matrix proteolysis and
guided invasion of tumor and endothelial cells through fibrillar collagens, there is
little evidence that the proteolytic activity of MT1-MMP expressed by either tumor
or stromal cell types is responsible for the breaking of vascular BM in vivo.
Although essential components of migration/invasive machinery are expressed in
aggressive tumor cells, causal links between these markers and proteolytic func-
tions of specific MMPs precisely during intravasation have been addressed in rather
limited number of metastasis studies.
The mammalian in vivo model of intravasation, employing intravital multipho-
ton microscopy and advanced optical technology, has allowed for direct observation
of tumor chemotaxis toward blood vessels and rare events of tumor cell entry into
blood vessels in real time (Wyckoff et al.
2000
; Condeelis and Segall
2003
), but did
not provide data on the contribution of MMPs to this process. The avian, chick
embryo CAM model represents a unique quantitative intravasation model system
(Fig.
7.5
), where MMPs and serine proteases have been repeatedly proved to play
critical roles by demonstrating that the broad-range MMP and serine protease
inhibitors significantly inhibit intravasation and subsequent dissemination of fibro-
sarcoma as well as carcinoma cells (Kim et al.
1998
; Deryugina et al.
2005
; Madsen
et al.
2006
; Conn et al.
2009
). However, it remains controversial which MMPs are
critically required for cancer cells to penetrate though the BM and endothelial layers
of the CAM blood vessels. Thus, downregulation of MT1-MMP by specific siRNA
constructs either led to complete abrogation (Sabeh et al.
2004
; Li et al.
2008
) or was
without any negative effects on intravasation of the HT-1080 fibrosarcoma in the
chick embryo CAM model of spontaneous metastasis (Deryugina et al.
2005
;
Partridge et al.
2007
). Moreover, specific downregulation of MMP-2 and MMP-9
by siRNA did not inhibit intravasation of HT-1080 cells, but unexpectedly resulted
in a several fold increase in their dissemination (Deryugina et al.
2005
; Partridge
et al.
2007
). Making it more complicated, the intravasation of several carcinoma cell
types, including breast, prostate, and epidermoid carcinomas, was initially reported
to critically depend on MMP-9, but in uPA/uPAR-dependent fashion (Kim et al.
1998
). It is also important to keep in mind that the proteolytic breaching of vascular
BM by tumor cells during their intravasation has never been formally demonstrated
by any group employing the CAM model system. In contrast, the initial invasion of
tumor cells through intact chorionic epithelium was shown to be accompanied by
the active retraction of the epithelial BM into compact islands with active invasion
of cells into the mesoderm occurring between the islands (Armstrong et al.
1982
).
