| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Biosciences Division, SRI International, Menlo Park, California, USA;
* Department of Medicine, Duke University Medical Center, Durham, North Carolina, USA; and
Institut fur Molekularbiologie der Universitat Zurich, Zurich, Switzerland
1 Correspondence: Biosciences Division, SRI International, Menlo Park, CA, 94025, USA. E-mail: brian.murphy{at}sri.com
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: MTF-1 • fibrosis • tumorigenesis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3
4)
that responds to a number of environmental perturbations including heavy metal load, hypoxia, and oxidative stress (5
6
7
8
9)
. Upon activation, MTF-1 is translocated to the nucleus where it is known to bind to a DNA sequence motif called the metal response element (MRE, core consensus sequence TGCRCNC). This protein is believed to cooperate with other transcription factors, including c-fos and upstream stimulatory factor (USF) -1 and -2 in response to heavy metals (10
, 11)
; our most current data demonstrate a direct interaction of MTF-1 with NF-
B p65 (RelA) under hypoxic conditions (Nagy et al. unpublished results). In addition to stress-induced activation of metallothionein (MT) transcription, we and others have shown that MTF-1 regulates the constitutive and stress-inducible expression of the proangiogenic and proinflammatory placenta growth factor (PlGF) (7
, 12)
, 
-glutamylcysteine synthase (GCS) (13)
, and zinc transporter-1 (14)
. The importance of this transcription factor is further highlighted by in vivo studies demonstrating lethality in MTF-1 null mice at embryonic day (E) 13.5 due to liver degeneration (13)
.
We have speculated that aberrant expression (and/or trans-activation) of MTF-1 could contribute to the development of malignant phenotypes (5
6
7)
. This hypothesis is supported by our ongoing work and by clinical studies that linked elevated expression of two MTF-1 target genes (MT-II and PlGF) with aggressive behavior in a variety of human tumors (15
16
17
18
19
20
21
22
23
24
25
26
27
28
29)
. Although little is known of the expression or trans-activation patterns of MTF-1 in human tumors, a recent DNA microarray study recorded a correlation between increased MTF-1 expression and the fraction of radio-resistant cells within human cervical tumors (30)
, possibly as a function of the antioxidant properties of MT (31)
.
Tissues respond to injury by mounting a wound-healing response, a three-phase, self-limited process characterized by formation of granulation tissue (composed of blood vessels, matrix, and inflammatory cells) and its subsequent resolution into scar tissue (32
, 33)
. Conversely, tumors have been categorized as "wounds that do not heal" (34)
as they subvert the host tissue repair pathways to generate stroma (matrix formation) and blood vessels (nutrition) for tumor expansion. In addition, tumor cells co-opt some signaling molecules of the innate immune system, such as chemokines, selectins, and their receptors, for invasion, migration, and metastasis (35)
. Tumor cells have, however, adapted to evade the final remodeling/scar phase since the acellular structure of scars cannot support tumor needs. Although this general strategy has been appreciated for 
20 years, the underlying molecular mechanisms involved in the escape from the host fibrotic response have remained largely undefined. Here we demonstrate for the first time that MTF-1 expression is critical for tumor expansion and that the loss of MTF-1 results in marked increases in ECM deposition, leading to fibrosis of the tumor mass. Molecular analysis suggests this phenotytpic change may be related in part to MTF-1 inhibitory control of transforming growth factor-ß1 (TGF-ß1) and tissue transglutaminase (tTG).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 13)
. Briefly, primary embryonic fibroblasts were isolated from 12.5-day-old mouse embryos and PCR genotyping for MTF-1 null cells was performed using genomic DNA prepared from yolk sacs and early passages of primary cells. Wild-type and MTF-1 null primary cells were then transfected with 10 µg of plasmid CMV-TAg and 10 µg of plasmid c-H-ras(A) per 100 mm-diameter cell culture plate according to Schonthal et al. (36)
. Cell foci were isolated and immortalized cell lines were derived. All cell lines (and new thaws) were genotyped by PCR analysis for the presence of the MTF-1 wild-type and null alleles as well as for genomic integration of TAg and oncogenic H-ras. In vitro proliferation rates were assessed using standard Alamar Blue staining and a CytoFluor 2350 Fluorescence Measurement System (Millipore, Bedford, MA, USA).
Xenograft studies
Male Balb/c nude athymic mice of an average weight of >25 g were selected for these studies. The animals were kept at 24°C on a 12 h light-dark cycle. TAg/Ras MTF-1+/+ and MTF-1–/– MEFs (4x106 cells/animal) were implanted by subcutaneous injection in the flank region of Balb/c nude athymic mice. Tumor volumes were measured every other day and body weights every week. MTF-1+/+ and MTF-1–/– MEF-derived tumors were harvested on days 7 and 16 postimplantation whereas MTF-1–/– tumors were harvested on day 35. For each experiment (total of three), each time point represented at least 10 animals. Harvested tumors were appropriately divided and saved for paraffin embedding, frozen sectioning, and protein analysis.
Histological analyses
Hematoxylin and eosin (H&E) and Masson’s trichrome staining were carried out on 5 µm-thick paraffin embedded sections. Mitotic count was evaluated on H&E-stained sections. Sections were scanned and three high-power fields (400x) showing a maximum number of mitotic figures were selected for counting (37)
. The data from all experiments were pooled and analyzed.
Collagen scoring was performed on Masson’s trichrome-stained sections. Three high-power fields in a section showing greatest degree of positive staining (blue strands) were scored from 0 to 3. 0 (negligible): 1 (mild), 2 (moderate), and 3 (strong). Microvessel density was calculated as described by (38)
using frozen sections immunostained with monoclonal CD 31 (blood vessel marker; BD PharMingen, San Diego, CA, USA). Three areas with highest visible blood vessels density per sample section were selected and the number of blood vessels was counted in a high-power field.
Immunoblotting analyses
Ras-transformed MTF-1+/+ and MTF-1–/– MEFs were lysed with an NP-40-based whole cell extract-buffered solution containing various protease inhibitors as described previously (7)
. Excised tumors were stored at –80°C and thawed on ice. Whole MTF-1–/– tumors were weighed and immediately placed in 1 mL lysis buffer (50 mM Tris pH 7.5, 150 mM NaCl, 2 mM EDTA, 0.5% NP-40, and 0.5% deoxycholate) containing 2 µg/mL aprotinin, 2 µg/mL leupeptin, and 120 µg/mL PMSF. MTF-1+/+ tumors were much larger than the MTF-1–/– tumors, and x0.2 g slice of each MTF-1+/+ tumor was used for Western analysis. The samples were kept on ice throughout the lysis procedure. All samples were homogenized (high speed) for 15 s, then sonicated (50% output) for 20 s. Samples were centrifuged at 14,000 RPM for 30 min (at 4°C). An aliquot was taken from the supernatant for protein analysis prior to Western blot. Detergent-soluble proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using NuPAGE Novex Bis-Tris 4–12% gels and a NuPAGE MOPS running buffer (Invitrogen, Carlsbad, CA, USA) and transferred to Immobilon P membranes (Millipore Corporation). In this separation system, proteins migrate faster than observed in the standard Tris/glycine running buffer. Blots were probed with the specific antibody and proteins were detected using an ECL-PLUS Enhanced Chemiluminescence detection system according to the manufacturer’s protocol (Amersham Life Sciences, Arlington Heights, IL, USA). Primary antibodies included a polyclonal anti-human MTF-1 and a polyclonal anti-tTG (NeoMarkers, Freemont, CA, USA). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Ambion, Austin, TX, USA) and Ponceau S (Sigma, St. Louis, MO, USA) staining verified equal protein loading and transfer.
ELISA analysis of latent and activated TGF-ß1
Tumor samples and fibroblast cultures were tested for latent and activated TGF-ß1 using a DuoSet ELISA kit for TGF-ß1 (Cat. DY240; R&D Systems, Minneapolis, MN, USA). MTF-1+/+ and MTF-1–/– tumor samples were homogenized, on ice, in lysis buffer (50 mM Tris pH 7.5, 150 mM NaCl, 2 mM EDTA, 0.5% NP-40, 0.5% deoxycholate, 2 µg/mL aprotinin, 2 µg/mL leupeptin, and 120 µg/mL PMSF), sonicated, and centrifuged at 12,500 x g at 4°C for 30 min. Aliquots of the lysates were treated to activate latent TGF-ß1 according to kit directions (plasma sample protocol) and kept overnight at 4°C until the ELISA was performed (39)
.
Supernatant and whole cell extract (WCE) samples from wild-type and MTF-1 knockout MEFs were prepared. Cells were plated in DMEM at 5% FBS and grown to confluence; at confluence the supernatants were harvested, protease inhibitors were added as above, and the samples were stored at –80°C. Cells were washed four times with PBS, scraped, pelleted, and stored at –80°C. The cell pellets were thawed on ice in lysis buffer (see above), resuspended, incubated on ice 30 min, and centrifuged at 9000 x g, 15 min at 4°C. Aliquots of WCEs and supernatants were treated to activate latent TGF-ß1 according to the manufacturer’s recommendations and kept overnight at 4°C until ELISA was performed. The TGF-ß1 ELISA was performed using an autoplate washer (Bio-Tek Instruments, Winooski, VT, USA) and a PowerWave 200 MicroplateScanning Spectrophotometer (Bio-Tek Instruments) with Kineticalc 4 software.
Plasmin assay
Plasmin activity was assessed according to the method of Le et al. (40)
. Briefly, 100 mm dishes were seeded with MTF-1+/+ or MTF-1–/– MEFs (5 million/dish) and incubated for 24 h in serum-free medium. Conditioned media was collected and cleared by centrifugation. The supernatant was concentrated 75-fold; 10 µL of the final volume was added to 490 µL assay buffer (100 mM Tris, pH 7.5, 10 mM CaCl2) and incubated for 60 min at 37°C. The fluorometer (Cytofluor 2350 Fluorescence Measurement System, Millipore) was set to excite at 400 nm; 20 µL of plasmin substrate (D-Val-Leu-Lys-AFC; ICN Biomedicals, Inc., Aurora, OH, USA) was added to the mixture and emission fluorescence was read at 500 nm over 1 h.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The results of a representative tumor xenograft growth study (one of three separate experiments) are presented in Fig. 2
. MTF-1+/+ tumors reached a mean size of 69 ± 7 mm3 by day 7 and grew rapidly to a mean volume of 957 ± 126 mm3 by day 16. Conversely, little or no tumor growth was observed in animals implanted with the MTF-1–/– MEFs on day 7 (average size of 9.5 ± 0.8 mm3); these tumors grew to a volume of only 28.2 ± 2.5 mm3 (P<0.001 compared with MTF-1+/+ tumors) by day 16. We repeated the study in female mice and found similar results (data not shown), implying this effect was not gender specific. A subset of MTF-1–/– tumor-bearing male mice was followed for an additional 20 days. Tumors in these mice resumed growth around day 22, and by day 35 (termination day) the average tumor volume was 698 ± 98 mm3. However, mean growth rate of the MTF-1–/– fibrosarcomas remained noticeably slower than that observed for wild-type tumors (Fig. 2A
). Western analysis of tumor lysates predictably found depressed levels of MTF-1 protein in day 7 and day 16 MTF-1–/– tumors compared with wild-type tumors (Fig. 2B
). The residual MTF-1 protein observed in these MTF-1–/– tumors most likely represents host cell expression. Total MTF-1 levels remained relatively depressed in the day 35 MTF-1–/– tumors sampled.
|
|
MTF-1 knockout results in enhanced matrix deposition and decreased tumor angiogenesis and proliferation
To better understand the role of MTF-1 in tumor progression, we performed histological examinations of representative tumor sections (Fig. 3
). Under light microscopy, H&E staining of paraffin-embedded blocks indicated major differences between the two tumor types. Day 16 MTF-1+/+ tumors displayed interlacing bundles and whorls of fusiform cells that were large, with minimally distinct cellular margins interspersed with foci of tumor necrosis. In contrast, the day 16 MTF-1–/– tumors displayed well-differentiated fusiform cells that appeared to be less invasive than the MTF-1+/+ tumor cells. Not surprisingly, fewer mitotic figures were found in day 16 MTF-1–/– tumors compared with MTF-1+/+ tumors (Fig. 3
, upper panel; Table 1
). By day 35 the MTF-1–/– tumors appeared more like day 16 MTF-1+/+ tumors, with increased mitotic counts (Table 1)
.
|
|
A distinctive feature observed in the MTF-1–/– tumors was the occurrence of abundant collagen fibers around the tumor mass in comparison to MTF-1+/+ tumors (Fig. 3
, middle panel). A specialized stain for collagen fibers (Masson’s trichrome) was used to semiquantitatively measure the collagen deposition and found 10-fold more collagen deposition in MTF-1–/– than in MTF-1+/+ tumors (Table 1)
. By day 35, collagen in the MTF-1–/– tumors decreased to levels approaching day 16 MTF-1+/+ tumors (Fig. 3
, middle panel; Table 1
). Since rapid tumor growth is accompanied by angiogenesis, we assessed microvessel density in the tumor tissues harvested at days 16 and 35. On day 16 the blood vessel count for MTF-1–/– tumors was found to be significantly lower (20%) than that observed for the MTF-1+/+ tumors (Fig. 3
, lower panel; Table 1
), but returned to near wild-type tumor values by day 35.
MTF-1 loss and enhanced matrix production correlate with increased activation of TGF-ß1
TGF-ß1 is a pluripotent cytokine with a well-documented role as the central regulator of extracellular matrix (ECM) synthesis and stabilization (41)
. It is synthesized as a latent form that is activated by cleavage of the latent binding protein, involving a cell surface complex (e.g., see refs 32
, 39
). We therefore evaluated the levels of activated (a)TGF-ß1 and total (t)TGF-ß1 (latent plus activated) in MTF-1+/+ and MTF-1–/– tumor tissue lysates (days 7 and 16) and day 35 MTF-1–/– tumors (Fig. 4
). Despite marked differences in growth rates, the levels of tTGF-ß1 in MTF-1–/– tumors remained constant (60 to 65 pg/mg tumor tissue) throughout the entire sampling period (sampling days 7, 16, and 35). The only significant differences in tTGF-ß1 levels were observed in day 7 (MTF-1+/+ tumors: 39 pg/mg tissue±3.6, n=4; MTF-1–/– tumors: 65 pg/mg tissue±4.4, n=4, respectively). aTGF-ß1 data provided clearer insight into the possible molecular mechanisms involved in MTF-1 control of tumor growth (Fig. 4A
). For example, at day 7 sampling MTF-1–/– tumor aTGF-ß1 was 300% higher (corrected value in relation to tTGF-ß1 present in tumors) than in wild-type tumors. This trend was more striking by day 16, when MTF-1–/– tumors displayed 530% (adjusted value) more aTGF-ß1 protein than day 16 MTF-1+/+ tumors. However, aTGF-ß1 levels returned to near wild-type levels by day 35.
|
Since both host and tumor cells can both synthesize TGF-ß1, we next evaluated the possible contribution of oncogenic MEFs to the enhanced aTGF-ß1 levels observed in the MTF-1–/– tumors. Consistent with the in vivo analysis (Fig. 4A
), aTGF-ß1 in the WCE of MTF-1–/– cultures was found to be higher than that of the wild-type MEFs (Fig. 4B, C
). In fact, aTGF-ß1 levels in the WCE of MTF-1+/+ MEFs (and in the media of both lines) were below the sensitivity of the DuoSet ELISA kit (see Materials and Methods). tTGF-ß1 levels were also significantly higher in the WCEs of MTF-1–/– MEFs compared with the wild-type cells.
MTF-1 loss correlates with increases in tTG expression but not plasmin activity
Since TGF ß1 is activated from its latent form predominantly by a surface complex [comprised of urokinase-type plasminogen activator receptor (uPAR), plasminogen, mannose-6-phosphate (M-6-P) receptor (also known as insulin-like growth factor II receptor), and tTG (42)
], we initiated studies to determine whether components of this pathway are controlled by MTF-1. Plasmin activity was readily detectable in the conditioned media of MTF-1+/+ and MTF-1–/– MEFs within 15 min of monitoring, but no significant differences were observed as a result of MTF-1 loss (data not shown). This finding was not surprising, as recent studies suggest only a minor role for plasmin compared with tTG and M-6-P receptor in this TGF ß1 activation pathway (43
, 44)
. We next explored the role of tTG expression as a function of MTF-1 loss since it not only activates latent TGF-ß1, but also has extensive matrix stabilizing effects (45
, 46)
. Western analysis indicated dramatic increases in the expression of full-length, active tTG (80 kDa) in day 7 and day 16 MTF-1–/– tumors compared with MTF-1+/+ tumors (Fig. 5
A). We observed some fragmentation of tTG, consistent with our earlier reports of degradation of tTG protein in inflamed and tumor tissues (32
, 45
, 46)
. tTG expression patterns in the day 35 MTF-1–/– tumors returned to near day 16 wild-type levels, a finding consistent with the tumor growth, collagen deposition, and TGF-ß1 data. Finally, in vitro analysis confirmed the tumor studies showing that loss of MTF-1 in ras-transformed MEFs resulted in increased expression of full-length tTG (Fig. 5B
).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 34)
, providing the essential scaffold (ECM) and nutrition (angiogenesis) for support, growth, and metastasis. In this model tumors employ only the first two phases of the healing cascade (inflammation and proliferation). Since remodeling/scar formation is the eventual result of the healing response, a condition not optimal for growth due to its acellular nature, tumors are believed to avoid this final phase and re-engage with the inflammatory phase (34)
. The underlying molecular mechanisms that control this extremely complex interaction have remained relatively undefined. Here we show for the first time that tumor cell-derived MTF-1 is involved in restricting the host response that attempts to limit tumor growth by fibrosis or scar formation.
Tumor xenograft studies demonstrated that MTF-1 null tumors were greatly restricted in their ability to maintain optimal growth. This finding is in direct contrast to the in vitro model, where the growth rates of the MTF-1–/– and MTF-1+/+ MEFs were essentially identical. These data imply unique tumor-specific interactions between tumor cells and host tissue that lead to the dramatically reduced tumor growth. The tumors derived from wild-type MEFs entered exponential growth by day 7 whereas growth of the MTF-1 null tumors was delayed until approximately day 22. From day 22 to day 35 (final sampling point), the MTF-1–/– tumors did grow, but at a markedly slower mean rate than observed for MTF-1+/+ tumors (Fig. 2A
). This type of slowed resumption of tumor growth has been reported for genetic knockout of other stress-related genes, including hypoxia-inducible factor-1 (HIF-1) (47)
. We speculate that the ability of these MTF-1–/– tumors to eventually bypass the requirement for protumorigenic MTF-1 may involve the recruitment of compensatory molecular pathways within the tumor cells, in situ tumor cell mutations, and/or host tissue contributions.
The normal balance of controlled ECM synthesis, deposition, and degradation is disrupted in a number of pathological conditions. For example, excessive matrix synthesis can result in fibrotic diseases, including pulmonary fibrosis, liver cirrhosis, and systemic sclerosis, whereas unbalanced ECM degradation contributes to tumorigenesis, rheumatoid arthritis, and osteoarthritis (48)
. The healing response is essential for tumor growth as it provides a stable matrix for attachment and migration of cells; the tumor ECM binds and stores cytokines and interacts with cell surface signaling cascade receptors that control cell cycle, proliferation, and apoptosis (49)
. But tumors avoid the natural conclusion of the healing response (tissue remodeling/scarring) by unknown mechanisms. Histological analysis of day 16 wild-type and MTF-1 null xenograft tumors clearly demonstrates that loss of MTF-1 expression in transformed cells causes dramatic increases in tumor collagen deposition. Not surprisingly, tumor vessel densities and mitotic indices were also decreased as a function of MTF-1 loss. In addition to the effects of fibrotic scarring, these decreases in vessel densities may involve the positive transcriptional control of MTF-1 over PlGF (7)
, a vascular endothelial growth factor family member thought to be intimately involved in tumor angiogenesis and inflammation (12
, 29)
. Taken together, these studies suggest a role for MTF-1 in the remodeling of tumor ECM, which promotes optimal malignant growth.
TGF-ß1 is a widely expressed cytokine involved in tissue remodeling and fibrosis (50)
, inhibition of cell proliferation, and induction of apoptosis (51)
. This protein is one of the most potent inducers of ECM deposition via several mechanisms including the stimulation of collagen synthesis, ECM receptors, protease inhibitors, and other profibrotic cytokines (52
, 53)
. TGF-ß1 is thought to have complex biphasic effects during tumorigenesis, acting as a tumor suppressor in the early stages of tumor growth and later as a stimulant of cancer progression (e.g., invasion, metastasis, and angiogenesis) by its autocrine and paracrine actions on tumor cells and their surrounding stromal environment (54
55
56)
. In these studies we evaluated TGF-ß1 expression as a function of tumor cell MTF-1. Our results demonstrated that genetic loss of MTF-1 in ras-transformed fibroblasts resulted in increased levels of total and activated TGF-ß1 in vitro. The tumor studies are more complicated in that only day 7 MTF-1–/– tumors displayed higher tTGF-ß1 levels compared with wild-type tumors. However, the in vivo data did reveal a strong correlation between TGF-ß1 activation and MTF-1 loss in day 7 and day 16 tumors. We believe that the inhibitory control of TGF-ß1 activation by functional tumor cell MTF-1 reveals an important underlying role for MTF-1 in the process of tumor progression. Furthermore, the finding that aTGF-ß1 levels in day 35 MTF-1–/– tumors dropped to near wild-type levels may explain how these tumors partially compensate for MTF-1 loss as reflected in resumed growth after day 22. The contribution of stromal cells to this phenomenon remains to be determined; however, a new study reports that loss of TGF-ß1 signaling in host fibroblast cells increases the oncogenic potential of adjacent epithelia (57)
.
tTG is a calcium-dependent, multifunctional enzyme that covalently cross-links a wide variety of ECM proteins, including fibronectin (58)
, collagen (59)
, fibrin (60)
, fibrinogen (61)
, vitronectin (62)
, and laminin/nidogen (63)
, producing a protease resistant matrix (64)
. Inappropriate overexpression of active (full-length) tTG is therefore thought to offer a formidable obstacle to growth of blood vessels, and this limits tumor progression as part of a host response mechanism (32)
. tTG is also involved in the conversion of latent to active TGF-ß1 whereas the tTG gene itself is induced by TGF-ß1, thus forming a positive feedback loop between the two proteins (65)
. We and others have suggested that tTG could play a major role in translating many effects of TGF-ß1 in the ECM such as modulation, remodeling, and stabilization of the ECM, and enhancement of wound tensile strength (e.g., see refs 39
, 46
). In earlier studies we observed coexpression of the tTG and active TGF-ß1 at sites within wounded regions and at sites of inflammation (32
, 46
, 66)
. Our group has shown that direct application of recombinant tTG to a mammary adenocarcinoma using a dorsal skin flap window chamber model resulted in increased levels of collagen around the tumor, resulting in fibrosis (32)
. The findings presented herein showing the levels of activated tTG and TGF-ß1 markedly increased in MTF-1–/– tumors support the argument that increases in aTGF-ß1 lead to higher production of active tTG (32
, 39
, 64)
, shifting the ECM balance of the developing tumor to enhanced matrix deposition and subsequent inhibition of growth. Consistent with the interdependence of tTG andTGF-ß1, the eventual escape of MTF-1–/– tumors from the anti-growth phenotype also appears to involve a return to near-normal levels of tTG.
Although the expression and trans-activation patterns of MTF-1 in human tumors have yet to be formally addressed, accumulating evidence suggests an oncogenic role for this transcription factor (6
, 7
, 30)
. Aberrant expression and/or trans-activation of MTF-1 may result from oncogenic ras transformation (7)
, germline or somatic mutations, and epigenetic factors such as transient hypoxia, oxidative stress, and metal accumulation (6
7
8
9)
. This protumorigenic behavior of MTF-1 may involve other cellular properties of MTF-1 in addition to its role as a transcription factor. For example, MTF-1 is now thought to act as a chromatin insulator (shielding specific transcriptionally active regions from the repressive effects of flanking chromatin; ref 67
) and to interact with p53 in the control of general protein translation during periods of cellular stress (68)
.
To our knowledge, this is the first report to highlight the importance of stress-inducible MTF-1 in tumor development. Specifically, our studies show that inhibition of tumor growth due to MTF-1 loss is linked to the development of excessive fibrotic scarring. This overproduction of tumor ECM most likely reflects increases in levels of activated TGF-ß1 and tTG, two proteins considered to be central components in ECM homeostasis and thus in normal and aberrant wound-healing processes. In light of these findings, we propose that tumor cells can adapt, in a MTF-1-dependent background, to evade the final remodeling/scar phase of the wound-healing cascade, as the acellular structure of scars cannot support tumor needs. The eventual return of MTF-1–/– tumor growth rates to near wild-type rates is consistent with decreased collagen deposition and decreases in levels of the activated forms of TGF-ß1 and tTG, implying that fibrosarcomas can partially compensate for MTF-1 loss by using molecular pathways yet to be elucidated. Future investigations should provide greater insight into these phenomena and into the interplay between tumor and host cells as functions of MTF-1 gain or loss.
|
ACKNOWLEDGMENTS |
|---|
Received for publication December 2, 2003. Accepted for publication April 15, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. J. Murphy, T. Kimura, B. G. Sato, Y. Shi, and G. K. Andrews Metallothionein Induction by Hypoxia Involves Cooperative Interactions between Metal-Responsive Transcription Factor-1 and Hypoxia-Inducible Transcription Factor-1{alpha} Mol. Cancer Res., March 1, 2008; 6(3): 483 - 490. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Wimmer, Y. Wang, O. Georgiev, and W. Schaffner Two major branches of anti-cadmium defense in the mouse: MTF-1/metallothioneins and glutathione Nucleic Acids Res., October 12, 2005; 33(18): 5715 - 5727. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
