| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics |
Institute for Molecular and Cellular Biology, Humboldt University Berlin, Max Delbrück Center of Molecular Medicine, 13122 Berlin, Germany [K. B., A. P., M. S., M. G.]; Department of Medicine and Therapeutics, University of Glasgow, Western Infirmary, Glasgow G11 6NT, United Kingdom [A. H. B.]; Institute for Pathology, University Clinic Benjamin Franklin, Free University Berlin, 12200 Berlin, Germany [A. P-C.]; and Biomedical Research Campus Berlin-Buch, 13122 Berlin, Germany [W. A.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50% of hepatocytes resulted in 95% reduction in
metastasis after tumor challenge compared with controls. Furthermore,
TIMP-2 gene transfer into livers with preexisting
metastatic spread resulted in a 77% reduction in tumor cell growth.
Our data imply that MMP activity of metastatic cancer cells is required
for spread and subsequent tumor growth and that enhancing
antiproteolytic defense mechanisms in target organs represents a novel
form of cancer gene therapy. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The metastatic spread of a primary tumor to distant organs involves a series of events that include the detachment of the cells from the primary tumor, migration and invasion through the basement membranes of blood and lymph vessels, adhesion to the endothelium of secondary organs, and extravasation and intravasation of the secondary organ. Metastatic tumor cells are known to produce a variety of proteolytic enzymes that are required to degrade ECMs,2 promoting intravasation and extravasation. Among these are matrix MMPs, a family of at least 20 zinc-dependent endopeptidases that collectively have the capacity to degrade all of the major components of the ECM (4) and that have been extensively implicated in cancer progression (5, 6, 7) . Moreover, MMPs appear to be involved in the process of angiogenesis, mediating the remodeling and penetration of the ECM by new capillaries (8) . Once, they are activated, MMPs are regulated by a group of specific endogenous inhibitors known as TIMPs. Under normal circumstances, such as tissue remodeling during embryogenesis, there is a tight regulation of MMP activity that is lost upon tumor growth and metastasis, leading to deregulated MMP activity in many solid tumors (7) . Several studies have demonstrated elevated levels of MMP-1, MMP-2, MMP-9, and MMP-11 in colorectal cancer (9, 10, 11, 12, 13) , suggesting a role for MMPs in this type of disease. MMP-2, the Mr 72,000 gelatinase A that degrades substrates including gelatins, fibronectins, and type IV collagens (4) , is among the best characterized of the MMPs. The enzyme is specifically inhibited by TIMP-2 (5 , 14, 15, 16) . TIMP-2 has been demonstrated to reduce tumor cell growth as well as metastasis (17 , 18) . Overexpression of TIMP-2 reduced primary tumor growth as well as hematogenous metastasis of transformed rat embryo fibroblasts (17 , 18) and primary tumor growth but not metastases of metastatic human melanoma (19) . In these studies, the general problem of efficient in vivo gene transfer into tumors was not addressed explicitly because TIMP-2 was transferred in vitro, or the vectors were preincubated with tumor cells before tumor induction. Conversely, in this study, we show for the first time that efficient in vivo transfer of TIMP-2 is possible by targeting the nontumorous tissue of a target organ, which circumvents the need to efficiently transduce the tumorous deposits themselves. We evaluated the potential of adenovirus-mediated overexpression of TIMP-2 to protect mouse livers from metastasis from a highly invasive cell line of colorectal origin (LS174T cells). We show that expression of TIMP-2 by hepatocytes protects from challenge by metastatic cells and, furthermore, reduces the growth of well-established metastases.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Western Blotting for Quantification of Secretion of Recombinant
TIMP-2.
A2 cells were infected in DMEM without serum for 1.5 h with
viruses and MOI as indicated. Supernatants were collected after 24 and
48 h and centrifuged for 5 min, and 10 µl were electrophoresed
on 12% polyacrylamide/SDS gels. After transfer onto nitrocellulose
membranes, TIMP-2 was detected with a mouse monoclonal antibody against
human TIMP-2 (T2–101, Ab-1; Dianova, Hamburg, Germany), biotinylated
goat antimouse antibody, and streptavidine-POD conjugate, followed by
enhanced chemiluminescence.
Detection of the Secretion of Recombinant TIMP-2 by Reversed
Gelatin Zymography.
Production of TIMP-2 by infected A2 cells was assessed by reversed
gelatin zymography as described (20)
. Briefly, A2 cells
were infected in DMEM without serum for 1.5 h with viruses and MOI
as indicated. Supernatants were collected after 48 h, centrifuged
for 10 min, and concentrated using Amicon filters (Lexington, MA).
NaN3, Brij-35, and
CaCl2 were added to 0.1%, 0.05%, and 5
mM, respectively. Samples were mixed with nonreducing
buffer and loaded onto 10% polyacrylamide/SDS gels impregnated with 1
mg/ml gelatin (porcine skin type I, bloom 300, G2500; Sigma,
Deisenhofen, Germany) and 107 ng/ml active MMP-2 protein
(Oncogene, Cambridge, MA). After electrophoresis, SDS was removed by
incubating for 2 h in 2.5% Triton X-100. The gel was then
incubated overnight in 50 mM Tris-HCl (pH 8.0), 50
mM NaCl, 10 mM CaCl2,
0.05% Brij-35, and 0.02% NaN3 at 37°C. Gels
were stained with 0.5% Coomassie Brilliant Blue (R250; Sigma),
and bands of gelatinase inhibitory activity representing TIMPs appeared
dark against the digested background.
Detection of Endogenous MMP-2 by Gelatin Zymography.
Production of MMP-2 by tumor cells was assessed by gelatin zymography
as described (20)
. Briefly, subconfluent cells were grown
for 48 h in DMEM without serum. The supernatant was removed and
centrifuged for 10 min at 1500 x g. If
necessary, supernatant was concentrated using Amicon filters. Gels
(7.5%, but without addition of activated MMP-2) were electrophoresed
and incubated as for reversed zymography. Bands of lysis representing
gelatinase activity were visualized against a dark background.
Detection of Recombinant Human TIMP-2 in the Serum of Nude Mice
by ELISA.
Adenoviruses were administered i.v. into the tail vein of nude mice. At
various time points, blood was drawn, allowed to coagulate, and
centrifuged for 10 min at 15,000 x g, and
serum was collected. Samples were analyzed for TIMP-2 levels at three
different dilutions using a quantitative ELISA assay (RPN 2618;
Amersham Buchler, Braunschweig, Germany).
Immunohistochemistry and X-gal Staining.
Adenoviruses were administered i.v. into the tail vein of nude mice.
After 3 days, liver sections were snap frozen in liquid nitrogen. For
immunohistochemical detection of TIMP-2 in liver tissue, a mouse
monoclonal antibody against human TIMP-2 (1:10; T2–101, Ab-1;
Dianova), followed by an FITC-conjugated sheep antimouse as the
secondary antibody, was used. No staining was seen if an unspecific
isotype antibody was used instead of the TIMP-2 antibody. Livers from
mice infected with Ad-ßgal were also stained with X-gal
according to standard procedures.
For determination of angiogenesis, proliferation, and apoptosis, paraffin sections from the livers of both efficacy experiments (preventive and therapeutic approach) were deparaffinated and stained with the respective antibodies. For angiogenesis, a mouse-anti-CD31 antibody (JC/70A; Dako, Hamburg, Germany) was used, and for proliferating cells, a mouse-antihuman MIB-1 (Ki-67) antibody (dia 505; Dianova) was used. Detection was carried out according to the alkaline phosphatase anti-alkaline phosphatase (APAAP) method using a rabbit-antimouse secondary antibody (Z0259; Dako), a mouse monoclonal APAAP complex (Dako), and the Fast Red chromogenic substate system (Dako). For detection of apoptosis the ApopTag fluorescein in situ apoptosis kit (Intergen, Oxford, United Kingdom; TUNEL method) was used according to the specifications of the manufacturer, and propidium iodide was used as a counterstain. Mitoses were also counted on H&E-stained slides. Ten microscopic fields/animal were examined at x400 (MIB, mitoses, and CD31) or with an oil immersion objective (x1000; TUNEL) using a Zeiss Axioskop fluorescence microscope (Carl Zeiss, Jena, Germany), and means were calculated. Means from the data of all tumor-bearing animals in the individual groups were calculated, and the Mann-Whitney test was applied.
In Vivo Antitumor Experiments.
Viruses and doses as indicated were applied i.v. into the tail vein of
nude mice. Induction of liver metastases was performed according to
Giavazzi et al. (25)
. Briefly, a small upper
quadrant incision was used to expose the spleen, and 2 x 106 LS174T cells in 50 µl were injected
into the lower splenic pole with a 27-gauge needle. The spleen was
returned to the abdominal cavity, the peritoneum was closed by suture,
and the skin was closed with wound clips. For the time sequence of
virus and tumor cell applications, see the figure legends. Five weeks
after tumor cell inoculation (which had been determined previously as
the week when 50% of the animals had to be euthanized because of their
tumor burden) was used as the end point of experiments. Animals that
had to be euthanized prematurely because of tumor burden were also
included into data acquisition.
The weight of liver metastases was determined as follows. The total mass of metastatic livers was measured, and for each animal, the percentage of tumor mass versus normal liver tissue was quantified macroscopically. Then, the livers were examined histologically in a blinded fashion by preparation of leveled sections. To this extent, livers were fixed in 4% neutral-buffered formalin and after thin sectioning were completely embedded in paraffin. One to 2-µm thin slides were prepared from each block and were stained with H&E. The percentage of tumor mass as compared with normal liver tissue was estimated for all liver pieces and slides, and the mean for each liver was calculated. The macroscopic and histological determinations of the tumorous fraction were performed independently and were within a 10% deviation between the methods. The mean of the two values was then taken. This value, expressed as a decimal, was multiplied with the weight of the tumorous livers, and the result was determined to be the actual tumor weight. For statistical analysis, the Mann-Whitney test was applied.
For survival experiments, a dose of 0.5 x 106 LS174 cells was used. Animals were euthanized when they appeared moribund. For statistical analysis, the Kaplan-Meier log rank test was applied.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Gelatin zymography revealed that LS174T cells produced high levels of
gelatinolytic MMP activity compared with other cell types, and this was
predominantly MMP-2. We therefore used the LS174T cell line for
in vivo analysis.
|
21,500) was detected by Western blotting in the supernatants of
Ad-TIMP-2-transduced cells but not in Ad-ßgal or PBS-treated
cells (Fig. 2a)
. No significant effect on cell survival was seen after
infection at MOI 100 or 400 compared with control adenovirus-infected
cells (data not shown).
|
. As expected, bands of MMP-2 inhibitory
activity representing TIMP-2 were detected in the supernatants of
Ad-TIMP-2-infected A2 cells but not in the supernatants of uninfected
or Ad-ßgal-infected cells.
We next evaluated adenovirus-mediated TIMP-2 gene transfer
into mouse livers in vivo. Mouse livers have been shown to
be highly susceptible to transduction by adenoviruses, even after
systemic delivery of the vectors (26, 27, 28)
. For
transduction of 50 and 100% of hepatocytes, 3 x 1010 or 6 x 1010 pfu were, respectively, sufficient (Fig. 3, B and C)
as determined by X-gal staining 3 days
after delivery of Ad-ßgal.
|
. No TIMP-2 immunoreactivity was observed
in the uninfected (Fig. 3D)
or Ad-ßgal-infected (not
shown) controls. TIMP-2 immunoreactivity was distributed throughout the
cytoplasm. No adverse animal behavior was noted at any viral dose.
Histopathology revealed dose-dependent disseminated areas of necrosis
(data not shown) with no differences between Ad-TIMP-2 and Ad-ßgal.
Taken together, these data demonstrate highly efficient adenoviral gene
transfer of TIMP-2 to the liver of nude mice and consequent expression
of TIMP-2.
To determine whether TIMP-2 would, as expected, be secreted into the
serum in vivo, we measured levels of TIMP-2 3 days after
systemic application of increasing doses of Ad-TIMP-2 by ELISA. Whereas
the serum levels in uninfected or Ad-ßgal-infected controls were
below the detection level of 8.0 ng/ml, application of 0.75 x 1010 pfu of Ad-TIMP-2, which results in
<10% infected hepatocytes in vivo (data not shown), led to
highly elevated serum levels (Fig. 4a)
. A substantial further increase in systemic TIMP-2 levels
was seen at higher doses of Ad-TIMP-2 but not Ad-ßgal (Fig. 4a)
. To evaluate the duration of transgene expression, we
monitored TIMP-2 levels over a time of 5 weeks after application of
3 x 1010 pfu Ad-TIMP-2 (Fig. 4b)
. Within the first 14 days, the levels remained stable.
After 2 weeks, however, a marked decrease in TIMP-2 expression was
observed. Four or 5 weeks after virus application, which was the time
when the efficacy experiments were terminated, the levels had decreased
50- or 100-fold but were still clearly above the background (Fig. 4b)
. In summary, these data demonstrate the efficacy of
adenovirus-mediated gene transfer to hepatocytes in vivo for
production of secreted diffusible inhibitors.
|
. The majority of the livers of control
animals were nearly entirely replaced by tumor (Fig. 5b)
,
and the weights of the tumorous livers were up to three times above
normal liver weights. In contrast, in the Ad-TIMP-2 group, three of
five animals were macroscopically tumor free, and the other two animals
had only small, marginally located metastases (data not shown and Fig. 5c
). The histopathological examination revealed only
occasional metastases (one to three per slide) in the three animals
that were macroscopically tumor free. These results indicate that the
development of highly invasive metastases to the liver can be nearly
entirely prevented by TIMP-2 overexpression.
|
. Comparison of treatment groups that received the same
viral doses of Ad-ßgal and Ad-TIMP-2, respectively, revealed a
statistically significant superiority only for low-dose treatment
(P = 0.038). In the high-dose groups, there
was a trend toward smaller volumes.
|
, in all three therapeutic settings, Ad-TIMP-2-treated
mice appeared to show an increased survival than Ad-ßgal-treated
mice. In addition, 4 of 15 animals (27%) in the TIMP-2 groups survived
the up-to-now 10-month observation period without disease, whereas all
Ad-ßgal-treated mice (n = 15) died within
4.5 months. These differences were, however, not significant because of
the small number of animals in this pilot survival experiment. Also,
some animals suffered from involvement of extrahepatic organs, such as
the spleen and peritoneum, which may have outbalanced the liver
preventive effect of TIMP-2. These combined results indicate that
transfer of TIMP-2 can reduce the growth of well-established liver
metastases and prolong the survival of treated animals.
Assessment of Proliferative, Angiogenic, and Apoptotic Activity in
Metastatic Livers.
To further characterize tumors histologically, we assessed
proliferative and apoptotic indices of tumors from animals in both
preventive and therapeutic experiments. In livers pretreated with
adenoviral vectors prior to tumor challenge, Ad-TIMP-2-treated
animals had lower proliferative indices than both Ad-ßgal-treated and
uninfected controls (Fig. 7)
. This result was reproduced for animals treated after tumor challenge,
except for mitoses at high-dose viral infection (Fig. 7)
. Evaluation of
CD31 staining in sections revealed that Ad-TIMP-2 treatment had no
significant effect on angiogenesis in either experimental design (Fig. 7)
. However, these experiments revealed an induction of CD31 staining
in high-dose Ad-ßgal-treated animals (Fig. 7)
. Interestingly, TUNEL
staining was increased in Ad-TIMP-2-treated mice compared with both
Ad-ßgal-treated and uninfected animals and in both experimental
protocols (Fig. 7)
.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We were surprised about the extremely high efficacy of metastasis prevention achieved, with three of five animals remaining macroscopically tumor free as compared with extensive metastasis with nearly entire replacement of the liver tissue with tumor in the control groups. Each of the steps of the metastatic cascade after adhesion of the tumor cell to the endothelium may be inhibited by TIMP-2, i.e.: (a) the initial extravasation either from the sinusoids through the Dissé space or from larger portal vessels through the basal membrane; (b) the invasion and degradation of the ECM within the parenchymal tissue; and (c) the inhibition of angiogenesis. The superiority of the preventive approach over the therapeutic approach argues for the importance of TIMP-2 in the inhibition of the earlier stages of metastasis. This is underscored by the fact that preventive expression of TIMP-2 by the host tissue is obviously sufficient for metastasis prevention, whereas the antitumor effect noted in the treatment of established metastases might have been attributable in part to direct adenoviral infection of the tumorous deposits themselves.
Available in vitro data demonstrate that adenovirally transferred TIMP-2 and other TIMPs can reduce the invasion of MMP-2-secreting melanoma cells through reconstituted basal membranes (33) . It remains to be shown, however, whether this mechanism is the most relevant in vivo. Recently, it was shown that mice transgenic for TIMP-1 and SV40 T antigen developed much smaller liver tumors than controls, although TIMP-1 overexpression could not prevent metastatic challenge (34) . Although this effect was interpreted as inhibition of angiogenesis, it argues for a minor importance of the inhibition of extravasation by TIMP-1 and for the diversity of effects that can be expected from different classes of TIMPs. In our model, the efficacy against metastasis formation as well as further growth argues for a participation of several mechanisms in the invasive cascade. Because MMPs are involved in all these steps and the inhibitory effect of TIMP-2 on MMP-2 and other MMPs has been well described, this interaction is probably at least partially responsible for the antimetastatic effects observed. However, we have not presented clear evidence for this assumption, and anti-invasive or other growth-inhibitory effects of TIMP-2, independent from its effects on MMPs, may also be involved. Our extensive histological examination revealed that the observed growth inhibition by Ad-TIMP-2 is paralleled by a statistically significant inhibition of proliferation and a stimulation of apoptosis. These effects could be the indirect result of the inhibition of invasion by TIMP-2, but it is also possible that the effects are exerted by TIMP-2 more directly. Further experimentation is required to elucidate the mechanisms underlying these phenotypic changes. The pronounced histological differences between the treatment groups, 4 or 5 weeks after virus application, were somewhat surprising because serum TIMP-2 levels had already declined 50- or 100-fold by the time the animals were scrutinized for histological examination. One reason for the marked effects on tumor histology, despite low levels of TIMP-2 in the blood, could be that the concentration of TIMP-2 was markedly higher in the liver sinusoids or the hepatic interstitium than in the peripheral blood, and that the local TIMP-2 levels were still sufficient for exerting therapeutic effects at this time. Alternatively, overexpression of TIMP-2 early in tumor development may lead to a severe and long-lasting disturbance of the balance of proliferation and apoptosis, which may be similar to the well-described and long-lasting inhibitory effects of single doses of chemotherapeutic agents.
Several clinical scenarios for our approach are imaginable:
(a) The i.v. injection of Ad-TIMP-2 might serve as a prophylactic adjunct to conventional modalities in cases of nonresectable primary tumors with the risk of metastasation or alternatively before surgery of the primary tumor and during other maneuvers that could lead to blood-borne shedding of cancer cells.
(b) The outgrowth of disseminated micrometastases, which can remain dormant for extended periods of time, may be prevented. This concept will be particularly dependent on the broad availability of vectors with the potential for long-term expression. In this context, it may be advantageous that, other than for the treatment of monogenetic diseases, immunogenicity of the transferred transgene product does not have to be expected, because it is an endogenous gene that becomes overexpressed.
(c) Established metastases could be the target of our approach, although further improvements in treatment efficacy would be desirable to justify a clinical trial.
(d) Other organs besides the liver that are at high risk for metastatic deposits could be targeted if the promising modifications in vector tropism are further extended.
A very helpful assessment of the significance of a new approach can be provided by a comparison with more established therapies. In the case of protease inhibitors, clinical trials with synthetic inhibitors of MMPs, such as batimastat and marimastat, have displayed antitumor efficacy as measured by reduction of the levels of cancer-specific antigens, and some of these substances are already in Phase II/III clinical trials (35) . Any success of synthetic inhibitors underscores in principle the utility of systemic inhibition of certain proteases. A gene therapeutic approach could add additional useful features, such as: (a) the achievement of high local concentrations by using the natural tropism of vectors or by modifying the viral surface structures leading to increased efficacy and reduced toxicity; (b) the continuous release of substances over extended periods of time with potentially lower costs for long-term treatment than generated by synthetic inhibitors; and (c) the combination of two or more highly specific protease inhibitors. In this context, it will be of crucial importance to determine which of the mechanisms, the local secretion of TIMP-2 in the target organ or the secretion of TIMP-2 into the blood with the potential for systemic effects, will be of higher relevance or how both principles can be efficiently combined.
In summary, we have described an approach that allows efficient treatment of colorectal liver metastases by targeting and, thereby, protecting normal organ tissue from cancer cell invasion, which may obviate the need for targeting of tumor deposits themselves.
|
FOOTNOTES |
|---|
1 To whom requests for reprints should be
addressed, at Institut für Molekulare Zellbiologie,
Humboldt-Universität Berlin, Max-Delbrück-Haus,
Robert Rössle Strasse 10, D-13122 Berlin, Germany. Phone:
49-30-9406-3309; Fax: 49-30-9406-3306; E-mail: kbrand{at}mdc-berlin.de 
2 The abbreviations used are: ECM, extracellular
matrix; MMP, metalloproteinase; TIMP, tissue inhibitor of MMP; MOI,
multiplicities of infection; Ad-TIMP-2, Adeno-TIMP-2;
Ad-ßgal,
Adeno-ß-galactosidase; X-gal,
5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside;
TUNEL, terminal deoxynucleotidyl transferase-mediated nick end
labeling; pfu, plaque-forming unit(s).
Received 8/ 3/99. Accepted 8/16/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. A. Fernandez, C. Butterfield, G. Jackson, and M. A. Moses Structural and Functional Uncoupling of the Enzymatic and Angiogenic Inhibitory Activities of Tissue Inhibitor of Metalloproteinase-2 (TIMP-2): LOOP 6 IS A NOVEL ANGIOGENESIS INHIBITOR J. Biol. Chem., October 17, 2003; 278(42): 40989 - 40995. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. W. Spurbeck, C. Y. C. Ng, T. S. Strom, E. F. Vanin, and A. M. Davidoff Enforced expression of tissue inhibitor of matrix metalloproteinase-3 affects functional capillary morphogenesis and inhibits tumor growth in a murine tumor model Blood, October 16, 2002; 100(9): 3361 - 3368. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. H. Baker, D. R. Edwards, and G. Murphy Metalloproteinase inhibitors: biological actions and therapeutic opportunities J. Cell Sci., January 10, 2002; 115(19): 3719 - 3727. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
0.01; and
***, P