CG&D
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cell Growth & Differentiation
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Farina, A. R.
Right arrow Articles by Mackay, A. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Farina, A. R.
Right arrow Articles by Mackay, A. R.
Cell Growth & Differentiation Vol. 10, 353-367, May 1999
© 1999 American Association for Cancer Research

Transcriptional Up-Regulation of Matrix Metalloproteinase-9 Expression during Spontaneous Epithelial to Neuroblast Phenotype Conversion by SK-N-SH Neuroblastoma Cells, Involved in Enhanced Invasivity, Depends upon GT-Box and Nuclear Factor {kappa}B Elements1

Antonietta R. Farina2, Antonella Tacconelli2, Alessandra Vacca, Marella Maroder, Alberto Gulino and Andrew R. Mackay3

Laboratory of Molecular Pathology, Department of Experimental Medicine, University of L’Aquila, Coppito II, 67100 L’Aquila [A. R. F., A. T., A. R. M.]; Neuromed Institute, 80077 Pozzilli [A. R. F., A. G., A. R. M.]; Department of Experimental Medicine and Pathology, University of Rome "La Sapienza," 00161 Rome [A. V., A. G.]; and Institute of Pathology, University of Palermo, 90141 Palermo [M. M.], Italy


   Abstract
TOP
 Abstract
Introduction
 Results
 Discussion
 Materials and Methods
 References
 
Spontaneous epithelial (S) to neuroblast (N) conversion enhanced the capacity of SK-N-SH neuroblastoma (NB) cells to invade reconstituted basement membrane in vitro. This involved a switch to matrix metalloproteinase (MMP) activity, in particular MMP-9, and was associated with the induction of MMP-9 expression. N-type-specific MMP-9 expression was herbimycin A inhibitable tyrosine kinase (possibly c-src) dependent and was regulated transcriptionally through GT-box (-52), and nuclear factor {kappa}B (NF{kappa}B; -600) elements within the MMP-9 gene. GT-box function was associated with elevated levels of specific nuclear GT-box binding complexes in N-type cells. NF{kappa}B function was associated with specific p50- and p65-containing nuclear NF{kappa}B binding complex(es). No function could be attributed to the proximal AP-1 (-79) element, and minimal function was attributed to the SP-1 (-560), ets (-540), or distal AP-1 (-533) elements. This was despite elevated levels of specific junD/fra-1 containing proximal AP-1 element binding complex(es) in N-type cells. Our data highlight a pivotal role for the GT-box, in concert with the NF{kappa}B element, in the transcriptional up-regulation of MMP-9 expression during spontaneous S to N phenotype conversion by SK-N-SH cells involved in enhanced basement membrane invasivity.


   Introduction
TOP
 Abstract
 Introduction
Results
 Discussion
 Materials and Methods
 References
 
NB4 is a highly malignant, neural crest-derived, childhood tumor characterized by a complex pattern of cellular phenotypes exhibiting partial differentiation along one or more neural crest lineage pathways (1) . NB cell lines in vitro reflect this heterogeneity, and three interchangeable phenotypes have been described: epithelial (S), neuroblast (N), and intermediate (I), which in addition to morphology, exhibit differential patterns of gene expression and differences in tumorigenicity and resistance to apoptosis (2, 3, 4, 5, 6, 7, 8) . These differences have prompted the hypothesis that cells exhibiting the N phenotype are more malignant than their S counterparts (5, 6, 7, 8) .

Uncontrolled ECM degradation, a hallmark of tumor progression, plays a central role in the loss of BM during the transition from in situ to invasive tumor behavior. The breaching of BM structures by tumor cells is, therefore, considered a reliable sign of malignancy in vivo and a relevant index of malignant behavior in vitro (9 , 10) . The involvement of ECM-degrading proteases in BM degradation leading to tumor invasion is now well established, and it is generally accepted that a proteolytic cascade involving enzymes of the plasmin system and family of MMPs is involved (9 , 11) . Debate exists, however, as to which enzymes of those expressed are directly involved in invasion (12 , 13) . This is of considerable importance in the future development of specific antiproteases as potential antimetastatic therapies.

MMP-9, a Mr 92,000 gelatinolytic type IV collagenolytic metalloproteinase implicated in the degradation of BM associated with tumor invasion (9 , 14, 15, 16, 17, 18) that confers malignant behavior to embryo cells (19) , is overexpressed in advanced stage metastatic NB (20) , other human malignancies, and malignant cell lines (20, 21, 22, 23) . MMP-9 expression can be induced by oncogenes, inflammatory cytokines, growth factors, and PKC-activating phorbol esters (9 , 14 , 17 , 20, 21, 22, 23, 24, 25, 26) .

The 5' flanking region of the MMP-9 gene has been cloned and shown to possess constitutive and inducible promoter activity when linked to reporter genes (17 , 24, 25, 26, 27, 28, 29) . Sequence analysis has revealed several potential regulatory regions involved in basal and induced MMP-9 transcriptional responses. An AP-1 element at position -79 relative to the transcriptional start site is necessary but insufficient for phorbol ester- and cytokine-enhanced MMP-9 transcription in fibrosarcoma cells and cooperates with NF{kappa}B and SP-1 elements at positions -600 and -560, respectively (17) . The ets element at position -540 has been implicated in MMP-9 promoter activation by ras (24) , EIAF (27) , fibroblast cell contact (28) , and EGF (29) and may function in concert with a distal AP-1 element at position -533 (24) . A GT-box at position -52, homologous to the retinoblastoma control elements of other genes (25 , 30, 31, 32) and the AP-1 element at position -79, are involved in MMP-9 promoter activation by v-src in human fibrosarcoma cells (25) , basal MMP-9 promoter activity in human squamous carcinoma cells (26) , and in H-ras/v-myc-dependent MMP-9 promoter activation in rat embryo cells (24) . It is generally accepted that full activation of the MMP-9 promoter depends upon the concerted action of several elements and that the AP-1 (-79) element is indispensable (17 , 24, 25, 26, 27, 28, 29) .

In this study, we have further examined the relationship between NB phenotype and malignant behavior by examining phenotype-specific differences in the capacity of NB cells to invade reconstituted BM in vitro and MMP-9 expression. For this purpose, we have used a model of spontaneous S to N phenotype conversion exhibited by the human NB cell line SK-N-SH (2 , 3) . We report that spontaneous conversion of SK-N-SH cells to the N phenotype results in an enhanced capacity to invade reconstituted BM in vitro, involving the induction of MMP-9 transcription and secretion. We characterize the cis and trans elements required for MMP-9 promoter activation involved in N-phenotype-specific MMP-9 transcription.


   Results
TOP
 Abstract
 Introduction
 Results
Discussion
 Materials and Methods
 References
 
Characterization of S and N Phenotypes Exhibited by SK-N-SH Cells.
SK-N-SH cells exhibited an epithelial (S) morphology with no neurite processes, high vimentin but low (NSE), and bcl-2 expression as determined by indirect immunofluorescence (Fig. 1A)Citation . Cells could be maintained in the S phenotype under normal passing conditions (recommended by American Type Culture Collection). Elongation (3–4-fold) in the time to passage and medium change resulted in the spontaneous conversion of SK-N-SH cells to an N phenotype characterized by a neuroblast morphology with neurite extensions of >1 cell body, low vimentin expression, but constitutive expression of NSE and bcl-2 (Fig. 1A)Citation , parameters reported previously to differentiate between S and N phenotypes (2, 3, 4 , 7 , 8) . This confirms previous reports of the capacity of SK-N-SH cells to interconvert between S and N phenotypes (2 , 3) .



View larger version (57K):
[in this window]
[in a new window]
  		Fig. 1. S to N conversion by SK-N-SH cells altered morphology, vimentin, bcl-2, and NSE expression and the capacity of cells to invade reconstituted BM in vitro. A, phase and light micrographs depicting morphology, vimentin, Bcl-2 and NSE expression in populations of N- and S-type SK-N-SH cells derived from the identical cell stock. B, histogram depicting the percentage difference in invasion of reconstituted BM exhibited by S- and N-type SK-N-SH cells in the presence (+) or absence (-) of anticatalytic anti-MMP-9 antibody (100 µg/ml) or recombinant human TIMP-1 (10 µg/ml). Results are provided as the means (n = 6) of relative invasion (%) with respect to untreated S-type SK-N-SH cells (100%); bars, SD.

 
For experimental purposes, SK-N-SH populations containing >95% of one or the other phenotype were analyzed. Direct comparisons were made between N- and S-type cells from the same stocks.

Spontaneous N-Type SK-N-SH Convertants Exhibit Enhanced BM Invasive Capacity Compared with S-Type Counterparts.
A comparison of the BM invasive capacity of independently derived N-type SK-N-SH cell populations and their respective S-type counterparts revealed a significant 3.6 ± 0.9 (P > 0.001, n = 6) increase in capacity to invade reconstituted BM in vitro (Fig. 1B)Citation . In contrast to S-type SK-N-SH BM invasion, which was not inhibited by anticatalytic anti-MMP9 antibody (100 µg/ml) or by recombinant human TIMP-1 (10 µg/ml), N-type invasion was significantly inhibited by both agents by between 30 and 40% (P > 0.001, n = 6 for both inhibitors; Fig. 1BCitation ). Neither S- nor N-type SK-N-SH invasion was inhibited by BSA (10 µg/ml) or by a monoclonal antibody to MT-MMP1 (100 µg/ml), used as control reagents. MT-MMP-1 was not expressed by either S- or N-type SK-N-SH cells (data not shown). These data indicate that enhanced invasion associated with conversion to the N phenotype involved TIMP-1-inhibitable MMP activity and in particular, MMP-9.

N-type SK-N-SH cells exhibited significantly (1.9 ± 0.3-fold) enhanced motility (P > 0.001, n = 6) across uncoated 8 µm Millipore filters when compared with S-type cells (data not shown).

N- but not S-Type SK-N-SH Cells Constitutively Express MMP-9.
S-type SK-N-SH cells constitutively expressed MMP-2 but not MMP-9 mRNA and protein, as determined by gelatin zymogram and Western and Northern blots (Fig. 2ACitation ; Refs. 12 and 21 ). N-type SK-N-SH cells constitutively expressed both MMP-2 and MMP-9 mRNA and protein (Fig. 2A)Citation . Both MMP-2 and MMP-9 were secreted in latent zymogen form. S- and N-type SK-N-SH cells constitutively expressed TIMP-1 and TIMP-2 mRNA and proteins. Expression of both TIMPs reduced upon conversion to the N phenotype (Fig. 2A)Citation .



View larger version (35K):
[in this window]
[in a new window]
  		Fig. 2. S to N SK-N-SH conversion induces MMP-9, reduces TIMP expression, and enhances net secreted levels of gelatinolytic MMP activity. A, representative zymograms, reverse zymograms, Western and Northern Blots demonstrating MMP-2, MMP-9, TIMP-1, and TIMP-2 mRNA, and protein levels expressed by representative S and N-type SK-N-SH cultures. Protein and activity levels were assessed in 48 h in serum-free conditioned media from equal cell numbers; mRNA levels were assessed after 24-h serum-free culture (2 µg/lane). A cDNA probe for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used to control RNA loading. B, histogram depicting gelatinolytic MMP activity in unconcentrated 48-h, serum-free conditioned medium from equal numbers of six independent cultures of S- and N-type SK-N-SH cells assayed in the presence (+) or absence (-) of 1 mM APMA, 10 mM EDTA, or anticatalytic anti-MMP-9 antibody (100 µg/ml). Results are expressed as the mean relative gelatin degradation (%) with respect to S-type SK-N-SH conditioned media (100%); bars, SD.

 
N-type SK-N-SH serum-free, 48-h conditioned media, activated with 1 mM APMA, exhibited a significantly elevated capacity to degrade type I gelatin in solution assay when compared with serum-free, 48-h conditioned media from equal numbers of S-type SK-N-SH cells (P > 0.05, n = 6; Fig. 2BCitation ). Enhanced gelatinolytic activity was completely inhibited by anticatalytic anti-MMP9 antibody (100 µg/ml). The anti-MMP-9 antibody did not significantly inhibit gelatinolytic activity in S-type conditioned media. Gelatinolytic activity in both S- and N-type SK-N-SH conditioned media was almost completely inhibited by EDTA (15 mM), confirming activity as MMP (Fig. 2B)Citation .

Elucidation of Potential Transcriptional Elements Involved in N Phenotype-specific MMP-9 Transcription.
In contrast to S-type SK-N-SH cells, N-type cells transactivated an MMP-9 promoter reporter gene construct (-670MMP9) in transient transfection assays (see Fig. 7ACitation ) and exhibited constitutive MMP-9 expression (Fig. 2)Citation , suggesting transcriptional up-regulation of MMP-9 expression in N-type cells.



View larger version (34K):
[in this window]
[in a new window]
  		Fig. 7. MMP-9 promoter is activated by N-type SK-N-SH through the GT-box and NF{kappa}B elements. A, top left, schematic representation of the 5' region of the MMP-9 gene used in this study. Right, histograms of CAT assays performed after the transient transfection of intact MMP-9 promoter reporter gene constructs (-670MMP-9) in S (hatched) and N-type SK-N-SH cells (A), intact and deleted MMP-9 promoter reporter gene constructs into N-type SK-N-SH cells (B), and intact and mutated MMP-9 promoter reporter gene constructs in N-type cells (diagrammed to the left; C). Mutated elements are represented by filled-in symbols. Results are expressed as either the mean ± fold induction of CAT activity for A or the percentage of difference in CAT activity with respect the intact MMP-9 promoter construct, -670MMP-9, in B and C. Constructs were transiently transfected with the control plasmid pRSVlacZ and CAT activity, normalized according to ß-galactosidase activity to account for differences in transfection efficiency. All transfections were performed in duplicate (n = 10 for intact constructs, 6 for deletion and mutant constructs; bars, SD).

 
To identify potential transcriptional elements involved in N phenotype-specific MMP-9 transcription, the 5' flanking sequence of the MMP-9 gene (1 to -670) was subjected to in vitro footprint analysis using nuclear extracts purified from S- and N-type SK-N-SH cells. As shown in Fig. 3Citation , two major regions mapping to positions -610 to -553 and from -79 to -28 of the MMP-9 promoter relative to the transcriptional start site revealed protected sequences. The first region (-610 to -553) contained a minimally protected area, exhibiting hypersensitive degradation sites, corresponding to an SP-1 element and a more notable protection corresponding to an NF{kappa}B element. The partial protection observed may reflect the relatively low levels of specific binding complexes detected by EMSA or differences in contact point sensitivity. No phenotype-specific differences were noted in the extent of protection of these two elements. The second region (-79 to -28) contained phenotype-specific differences in the protection of two areas, the first corresponding to the proximal AP-1 element that was not protected by S-type extracts and minimally protected by N-type extracts with the appearance of a hypersensitive degradation site. The second area corresponded to the GT-box, and although protected by S- and N-type nuclear extracts, it exhibited greater protection by N-type extracts. Protection of ets (-540) or distal AP-1 (-533) elements was noted (data not shown).



View larger version (69K):
[in this window]
[in a new window]
  		Fig. 3. Elevated protection of the MMP-9 GT-box in N-type SK-N-SH nuclear extracts. An autoradiograph of 8% polyacrylamide DNA sequencing gel depicting regions -610 to -553 (left) and -79 to -28 (right) of the 5' promoter region of the MMP-9 gene protected by nuclear proteins contained in 30 µg of N- and S-type SK-N-SH nuclear proteins is shown. C/A nucleotide sequence for the region -553 to -610 and G/T nucleotide sequence for the region -79 to -28 are displayed to the left of each respective lane. The putative transcriptional elements SP-1, NF{kappa}B, AP-1, and GT-box are indicated by sequence (boldface), position relative to the transcriptional start site and position in the gel (underlined). Sequences protected in the presence (+) of S- or N-type nuclear extracts appear lighter than the corresponding regions digested by DNAse in the absence (-) of nuclear extracts.

 
N-type SK-N-SH Nuclear Extracts Contain Elevated Levels of Specific GT-Box and AP-1 Element Binding Complex(es).
S- and N-type SK-N-SH nuclear extracts were assayed for the presence of specific DNA binding proteins by EMSA using double-stranded oligonucleotides corresponding to the MMP-9 GT-box (-52), AP-1 (-79 and -533), NF{kappa}B (-600), SP-1 (-560), and ets (-540) elements. As shown in Fig. 4Citation ., in contrast to S-type nuclear extracts, N-type nuclear extracts contained markedly elevated levels of specific GT-box binding complexes. Specificity of binding was confirmed by competition assays, in which excess unlabeled GT-box, but not NS or mutated GT-box, SP-1, or E2F oligonucleotides competed with labeled probe (Fig. 4B)Citation . Specific complexes did not react with anti-SP-1, E2F1, E2F4, DP-1, or DP-2 antibodies (data not shown), used to detect the possible presence of Rb-regulated transcription factors.



View larger version (33K):
[in this window]
[in a new window]
  		Fig. 4. N-type SK-N-SH nuclear extracts contain elevated levels of specific GT-box binding complex(es). A, representative EMSA comparing levels of GT-box binding complex(es) in nuclear extracts prepared from S- and N-type SK-N-SH cells of the same stock origin. B, GT-box binding specificity in N-type extracts was determined by competition with (+) or without (-) unlabeled GT-box, mutated GT-box, MMP-9 SP-1, E2F, and NS oligonucleotides. Arrows, specific shifts.

 
The proximal AP-1 (-79) element oligonucleotide bound elevated levels of nuclear complexes in N-type cells, judged to be specific by competition assay using cold AP-1, mutated AP-1, and NS oligonucleotides (Fig. 5A)Citation . Specific complexes reacted with antibodies against jun and fos families (pan-jun and pan-fos), which abrogated binding and antibodies against junD and fra-1, which induced complex supershifting. Complexes did not react with antibodies against c-jun, junB, c-fos, or fra-2 (Fig. 5B)Citation . N-type nuclear extracts contained elevated levels of fra-1 but not junD or cfos proteins when compared to S-type extracts, as judged by Western blot (Fig. 5C)Citation . It is unclear whether the subtle differences observed in electrophoretic mobility of immunoreactive fra-1, c-fos, and junD proteins reflect differences in phosphorylation state.



View larger version (49K):
[in this window]
[in a new window]
  		Fig. 5. N-type SK-N-SH nuclear extracts contain elevated levels of specific junD/fra-1-containing AP-1 site binding complex(es). A, representative EMSA demonstrating MMP-9 AP-1 element binding complex(es) in nuclear extracts prepared from S- and N-type cells of the same cell stock. Competition of specific AP-1 binding by cold MMP-9 AP-1 oligonucleotide but not by NS or a mutated MMP-9 AP-1 site oligonucleotide (MMP-9/mAP-1) in N-type extracts. Arrow, specific complex(es). B, representative EMSAs demonstrating reactivity of specific MMP-9 AP-1 element binding complex(es) in N-type nuclear extracts with antibodies against the entire jun family (pan jun), the entire fos family (pan fos), junD, and fra-1 but lack of reactivity with anti-cJun, JunB, Fra-2, cfos, or fosB antibodies. Arrows, specific complex(es) and reactive supershifted complexes. C, representative Western blots depicting levels of fra-1, cfos, and junD proteins in untreated S- and N-type nuclear extracts and extracts prepared from N-type cells treated with TPA (6 ng/ml for 4 h; anti-fra-1 and cfos only) prepared from equal numbers of SK-N-SH cells. TPA treatment was used to demonstrate relative anti-fra-1 and cfos specificity, and reactivity against both antibodies was demonstrated on the same blot, stripped and then rehybridized. Recombinant cJun and JunD (rcJun and rJunD) were used to confirm anti-junD antibody specificity.

 
The NF{kappa}B oligonucleotide bound similar levels of specific nuclear complexes in both S- and N-type cells. Complexes were judged to be specific by competition assay using cold NF{kappa}B oligonucleotide, mutated NF{kappa}B, or by an MMP-1 AP-1 oligonucleotide used as a NS competitor (Fig. 6Citation ., data shown for N-type cells only). Specific complexes in both S- and N-type nuclear extracts reacted with anti-p50 and -p65 antibodies, which induced complex supershifting, but not with anti-cRel antibody (Fig. 6Citation ; data shown only for N-type cells). Anti-p50, -p65, and cRel antibodies alone did not bind to the oligonucleotide probe in the absence of nuclear extracts (data not shown).



View larger version (67K):
[in this window]
[in a new window]
  		Fig. 6. Nuclear extracts from N-type SK-N-SH cells contain specific NF{kappa}B element binding complex(es) that contain p50 and p65 but not cRel proteins. A representative EMSA depicting specific NF{kappa}B site binding complexes (arrows) in N-type SK-N-SH nuclear extracts. Specific complexes were competed by excess unlabeled NF{kappa}B oligonucleotide but not by a mutated NF{kappa}B oligonucleotide (mNF{kappa}B) or by an MMP-1 AP-1 oligonucleotide, used as a NS competitor (NS). Specific shifts reacted with anti-p50 and -p65 but not cRel antibodies, which induced complex supershifting.

 
Specific binding to SP-1 (-560), ets (-540), or distal AP-1 (-533) elements was low or undetectable in both S- and N-type nuclear extracts (data not shown).

GT-Box, NF{kappa}B, and SP-1 but not AP-1 Elements Function in N Phenotype-specific MMP-9 Transcription.
To study the function of the various cis elements within the MMP-9 promoter, a series of deleted and point-mutated MMP-9 promoter reporter gene constructs were prepared (defined in Fig. 7, B and CCitation ) and transiently transfected into N-type SK-N-SH cells. The basic construct (-670MMP9) contained 670 bp of 5' regulatory sequence from the MMP-9 gene, including the putative TATA box, which covers the major regulatory regions described previously for constitutive and inducible MMP-9 responsiveness (17) . This construct was used to generate deletion and point mutants.

As demonstrated in Fig. 7ACitation , N-type SK-N-SH cells transactivated the MMP-9 promoter reporter gene construct -670MMP9, resulting in an 8.5 ± 2.25-fold induction in CAT activity, in transient transfection assays. S-type cells failed to transactivate this reporter gene construct.

As demonstrated in Fig. 7BCitation , deletion of the NF{kappa}B element (-586MMP9) significantly reduced CAT activity in N-type cells by 60%. Deletion of SP-1, ets, and distal AP-1 elements (-460MMP9) further reduced CAT activity by 25%. The deletion of the CA repeat region (-88MMP9) reduced CAT activity further by a statistically nonsignificant level. Deletion of the proximal AP-1 element (-70MMP9) and the GT-box (-49MMP9) did not significantly reduce CAT activity further.

As demonstrated in Fig. 7CCitation , mutation of the NF{kappa}B element (-670mNF{kappa}B) or the GT-box (-670mGT) significantly reduced CAT activity in transiently transfected N-type SK-N-SH cells by 60 and 90%, respectively, compared with intact promoter (-670MMP9). Mutation of the SP-1 element (-670mSP1) or the ets element (-670mEts) reduced CAT activity minimally by between 20 and 30%. Mutation of the proximal AP-1 element (-670mAP1) did not reduce CAT activity. All mutations were identical to those confirmed as unable to bind specific complexes in either S- or N-type SK-N-SH nuclear extracts by EMSA.

MMP-9 GT-Box and NF{kappa}B/SP-1 but not AP-1 Elements Confer Transcriptional Activation to a Heterologous Promoter in N-Type SK-N-SH Cells.
To further examine the transcriptional activity of GT-box, NF{kappa}B/SP-1, and AP-1 element binding complexes, elements were subcloned upstream of a heterologous cfos promoter ({Delta}56CAT) and transfected into S- and N-type SK-N-SH cells. Transcriptional activation of constructs containing the GT-box (one copy) and the NF{kappa}B/SP-1 elements (one copy) but not MMP-9, MMP-1, or TIMP-1 AP-1 elements (three copies) was observed in N-type SK-N-SH cells, confirming GT-box, NF{kappa}B/SP-1, but not AP-1 transcription factor function. In S-type cells, transcriptional function was observed for the NF{kappa}B/SP-1 element composite but not the GT-box or AP-1 elements (Fig. 8)Citation .



View larger version (56K):
[in this window]
[in a new window]
  		Fig. 8. The MMP-9 GT-box confers transcriptional activation to a heterologous cfos promoter in N-type but not S-type cells. Representative autoradiograph of three independent experiments demonstrating CAT activity in S- (A) and N-type (B) SK-N-SH cells transiently transfected with heterologous cfos promoter construct {Delta}56 alone (CON), containing the MMP-9 GT box (one copy; GT), the MMP-9 NF{kappa}B and SP-1 elements (NF{kappa}B/SP1), the MMP-9 AP-1 element (three repeats), the MMP-1 AP-1 element (three repeats), and the TIMP-1 AP-1 element (three repeats). CAT assays were adjusted for differences in transfection efficiency using a cotransfected ß-galactosidase expression vector.

 
Transient transfection of the same constructs into human fibrosarcoma HT-1080 cells, which constitutively express MMP-9 (17 , 21) , indicated function for the proximal AP-1 element but not GT-box in basal transcription, confirming a previous report (Ref. 17 ; data not shown), suggesting cell-specific usage of either GT-box or AP-1 elements for constitutive MMP-9 transcription. Specific AP-1 site binding complexes in HT-1080 cells differed from those in N-type SK-N-SH cells in cjun content alone (data not shown).

Herbimycin A Inhibits N-Type SK-N-SH MMP-9 Expression, and N-Type SK-N-SH Cells Exhibit Elevated c-src Kinase Activity.
Herbimycin A, at c-src inhibitory concentrations of 0.1 and 1 µM (33) , completely inhibited MMP-9 protein expression by N-type SK-N-SH cells, without affecting MMP-2 (Fig. 9A)Citation .



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 9. Constitutive MMP-9 expression by N-type SK-N-SH cells is inhibited by herbimycin A and associates with enhanced c-src kinase activity. A, representative gelatin zymogram demonstrating MMP-2 and MMP-9 activity in 48-h, serum-free N-type SK-N-SH conditioned medium treated with herbimycin A at concentrations ranging from 0 to 1 µM. B, representative Western blot depicting levels of c-src protein in total cell extracts from equal numbers of S- and N-type SK-N-SH cells. C, representative autoradiograph demonstrating the levels of p62GST c-src substrate kinased by c-src immunoprecipitated from equal numbers of S- and N-type SK-N-SH cells by anti-c-src antibody or preimmune IgG. The dotted line represents a crack in the gel to explain why bands appear at different levels.

 
Cell extracts from N-type SK-N-SH cells contained elevated levels of c-src protein, when compared with equal numbers of S-type SK-N-SH cells, as determined by Western blot (Fig. 9B)Citation , and exhibited elevated c-src kinase activity in immunoprecipitation kinase assays (Fig. 9C)Citation .

MMP-9 protein expression by N-type SK-N-SH cells was not inhibited by calphostin C, at PKC inhibitory concentrations (1–100 nM; Ref. 34 ), or by the EGF receptor kinase inhibitor lavendustin A (1 nM–1 µM; Ref. 35 ; data not shown).

N-Type SH-SY5Y, LA-N-1, SK-N-AS, and NB Cells Exhibit Similarities to Spontaneous N-Type SK-N-SH Convertants.
N-type SH-SY5Y cells, originally cloned from the SK-N-SH cell line (3) , exhibited significantly elevated invasive capacity (P > 0.001), constitutive MMP-9 expression, elevated levels of specific nuclear GT-box binding complexes, and c-src protein and kinase activity, when compared with S-type SH-EP cells also cloned from the SK-N-SH cell line (Ref. 3 ; Fig. 10Citation ).



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 10. SH-EP and SH-SY5Y cells exhibit similarity to S- and N-type SK-N-SH cells. A, phase contrast micrographs demonstrating S-type morphology of SH-EP and N-type morphology of SH-SY5Y cells. B, histogram depicting differences between SH-EP and SH-SY5Y invasivity through Matrigel. Results are expressed as the mean percentage difference in invasion compared with S-type SH-EP cells (100%) in six independent experiments. C, a representative Western blot demonstrating levels of MMP-2 and MMP-9 expressed by SH-EP and SH-SY5Y cells. D, representative EMSA demonstrating specific elevated levels of GT-box binding complex(es) in SH-SY5Y compared with SH-EP nuclear extracts. Arrows, specific complexes. E, representative Western blot and kinase assays depicting levels of c-src protein in total extracts and p62GST src substrate kinase by immunoprecipitated c-src from equal numbers of SH-EP and SH-SY5Y cells.

 
Two additional NB cell lines, LA-N-1 and SK-N-AS, also exhibited S to N conversion upon long-term culture without passage (Fig. 11A)Citation . Both cell lines exhibited N phenotype-specific MMP-9 expression and enhanced invasivity (Fig. 11, B and C)Citation .



View larger version (52K):
[in this window]
[in a new window]
  		Fig. 11. S to N conversion by LA-N-1 and SK-N-AS cells enhances invasivity and induces MMP-9 expression. A, phase contrast micrograph demonstrating S and N morphological phenotypes exhibited by LA-N-1 and SK-N-AS cells. B, representative gelatin zymogram demonstrating relative levels of MMP-2 and MMP-9 expressed by S- and N-type LA-N-1 and SK-N-AS cells. C, histogram depicting relative invasivity (%) of N-type LA-N-1 and SK-N-AS cells with respect to their S-type counterparts (100%). Results are expressed as mean relative invasion (%) of three independent experiments; bars, SD.

 

   Discussion
TOP
 Abstract
 Introduction
 Results
 Discussion
Materials and Methods
 References
 
In this study, we report that spontaneous S to N conversion by human SK-N-SH NB cells resulted in an enhanced capacity to invade reconstituted BM in vitro. Enhanced N-type invasion involved a switch to TIMP-1-inhibitable MMP activity and in particular that of MMP-9 and was associated with up-regulated MMP-9 expression, involving tyrosine kinase (possibly c-src) activity, and dependent upon transcription through the GT-box and NF{kappa}B elements within the 5' regulatory region of the MMP-9 gene.

The capacity of SK-N-SH cells to interconvert spontaneously between S and N phenotypes (2 , 3) , confirmed in this study, may represent an important endogenous mechanism involved in generating phenotype heterogeneity within NBs in vivo (1) . The enhanced capacity of N-type SK-N-SH cells to invade BM in vitro, an index of enhanced malignant behavior (10) , adds to other reports of enhanced N-type tumorgenicity and resistance to apoptosis (7 , 8) that fuel the hypothesis and may help to explain why the N phenotype is more malignant than its S-type counterpart in NB (2, 3, 4, 5, 6, 7, 8) .

Enhanced N-type invasion involved a switch to the involvement of MMP activity, concluded using the specific MMP inhibitor TIMP-1, which inhibited N- but not S-type SK-N-SH invasion. This confirms our previous reports of MMP-independent invasion and degradation of BM type IV collagen exhibited by S-type SK-N-SH cells (12 , 36) and demonstrates that phenotype-associated changes regulate the involvement of MMPs in BM invasion exhibited by this NB cell line in vitro.

The involvement of MMP-9 in enhanced N-type invasion was demonstrated using an anticatalytic anti-MMP9 antibody and supported by constitutive MMP-9 expression by N- but not S-type cells and by elevated net gelatinolytic MMP activity secreted by N-type cells, also abrogated by anticatalytic MMP-9 antibody. Because MMP-9 was secreted by N-type cells in zymogen form, its involvement in N-type invasion probably involved a Matrigel-mediated activation mechanism (37) . These data add to the increasing body of evidence that MMP-9 is involved in tumor invasion under certain circumstances (9 , 13 , 14 , 16 , 19) .

Of the other MMPs generally accepted to be involved in tumor invasion (9) , the MMP-2 activator, MT-MMP1, has been implicated in NB invasivity (38) . Both S- and N-type cells expressed similar levels of MMP-2 but did not express MT-MMP1, nor did an anti-MT-MMP1 antibody influence either S- or N-type SK-N-SH invasivity. MMP-2 is not involved in S-type SK-N-SH invasion (this study and Ref. 12 ). TIMP-1 inhibited N-type SK-N-SH invasivity but not by more than that observed with anti-MMP-9 antibody, which does not inhibit MMP-2 activity (39) . This would suggest a minimal role for MMP-2 in N-type SK-N-SH invasion. This may reflect the absence of an effective MMP-2 activation mechanism.

MMP-9 activity was responsible for only a proportion of the enhanced invasivity exhibited by N-type cells, suggesting the involvement of additional factors. Tumor invasion results from adhesive, degradative, and motile responses. A recent report has suggested that enhanced NB invasivity involves enhanced motility (40) . In this study, N-type SK-N-SH cells exhibited a 1.9-fold increase in random motility, confirming a role for enhanced motility in enhanced invasivity of N-type cells.

Conversion to the N phenotype by SK-N-SH cells was also associated with reduced expression of TIMP-1 and TIMP-2, supporting the hypothesis that an imbalance in protease/inhibitor expression may underlie tumor invasion (41) . This suggests that differential expression of MMP-9 and its specific inhibitor TIMP-1, which are generally coordinately expressed (9 , 21 , 22) , is a property of the N phenotype of SK-N-SH cells. Reduced TIMP expression is a feature of advanced stage NB in vivo (42) .

N-type SK-N-SH cells exhibit transcriptional up-regulated MMP-9 expression, concluded by the capacity of N- but not S-type cells to activate an MMP-9 promoter reporter gene construct. Analysis of the potential transcriptional elements involved N phenotype-specific MMP-9 transcription by in vitro footprint, and EMSA analyses revealed N phenotype-specific elevation in the protection of, and binding to, the GT-box (-54) and proximal AP-1 (-79) elements within the MMP-9 5' regulatory sequence. Both elements are reported to function in MMP-9 transcription (17 , 24, 25, 26 , 28) . However, point mutation of the GT-box but not the proximal AP-1 element abrogated N-type-specific MMP-9 promoter activity. This suggests transcriptional function for the GT-box but not the proximal AP-1 element in N phenotype-specific MMP-9 transcription, thus providing a novel function for the GT-box as a substitute for the proximal AP-1 element in MMP-9 transcription. To date, the proximal AP-1 element has been generally considered indispensable for MMP-9 transcription (17 , 24, 25, 26 , 28) .

The GT-box alone did not confer transcriptional activation to the MMP-9 promoter, determined using deleted MMP-9 promoter reporter gene constructs, confirming that this element is essential but insufficient for MMP-9 promoter activity in N-type SK-N-SH cells. GT-box transcription factor function was confirmed in N-type but not S-type cells, using a heterologous cfos promoter reporter gene construct bearing a single copy of the MMP-9 GT-box, further differentiating between S- and N-type cells. Partial characterization of GT-box binding complexes failed to detect the presence of SP-1- or Rb-regulated transcription factors of the E2F or DP families, reported previously to bind similar elements (25 , 30, 31, 32 , 43) . Identity with previously uncharacterized GT-box binding transcription factors (30, 31, 32) was not tested.

Lack of AP-1 transcription factor function in N-type cells was confirmed using heterologous cfos promoter reporter gene constructs bearing either MMP-1, MMP-9, or TIMP-1 AP-1 elements. This lack of function may also help to explain the lack of MMP-1 expression or elevated expression of TIMP-1 in N-type SK-N-SH cells, because both genes are sensitive to AP-1-mediated transcriptional up-regulation (44, 45, 46) . It is unlikely that complex composition alone was responsible for lack of AP-1 function, because junD and fra-1, detected within complex(es), transactivate AP-1 elements within MMP genes (44) . It is perhaps more likely that posttranslational modifications and/or cofactors required for full AP-1 transcriptional activity (47, 48, 49, 50, 51) were absent in N-type cells. This was supported by the failure of calphostin C to reduce N-type-specific MMP-9 expression at PKC-inhibiting concentrations (34) . PKC activates AP-1 complexes (47) . A function for the proximal MMP-9 AP-1 element was confirmed in basal MMP-9 transcription by human HT-1080 fibrosarcoma cells (this study and Ref. 17 ) and bound complex(es) in HT-1080 cells that differed in c-jun content alone to those observed in N-type SK-N-SH cells. This would suggest tumor cell-specific usage of the GT-box or AP-1 element in basal MMP-9 transcription.

The GT-box cooperated largely with the NF{kappa}B (-600) element in N-type SK-N-SH-specific MMP-9 promoter activity, as judged by deletion and mutation analysis. This adds to an increasing body of evidence that the NF{kappa}B element is a major modulator of MMP-9 transcription (17 , 24, 25, 26, 27) but requires an initiating element, considered previously to be the proximal AP-1 element (17) , which we now extend to include the GT-box. Despite N-type-specific MMP-9 expression, both S- and N-type cells exhibited similar degrees of NF{kappa}B site protection; levels of specific p50/p65-containing NF{kappa}B site binding complexes and both phenotypes exhibited the capacity to activate a cfos promoter reporter gene construct bearing a single copy of the NF{kappa}B/SP-1 element. This suggests constitutive NF{kappa}B transcription factor function common to both phenotypes. Because the NF{kappa}B element did not function independently of the GT-box in N-type cells, these data confirm the essential role of the GT-box and its cognate transcription factor(s), which cooperate with the NF{kappa}B element, for N phenotype-specific MMP-9 transcription.

Other elements involved in MMP-9 transcription include the SP-1 element (-560), the ets element (-540), closely spaced with the distal AP-1 element (17 , 24 , 27 , 28) . Deletion of all three elements (-460MMP9) reduced transcription by between 20 and 30%, as did mutation of either the SP-1 or ets elements. This suggests a minimal role for SP-1, ets, or concerted ets/distal AP-1 elements in N phenotype-specific MMP-9 transcription. This was further supported by the observations that SP-1, ets, and distal AP-1 elements were minimally protected by in vitro footprint assay and did not bind specific complexes in either S- or N-type nuclear extracts. Furthermore, N-type cells did not constitutively express MMP-1 or exhibit enhanced TIMP-1 expression, two genes that are sensitive to ets/AP-1 transcriptional up-regulation (44 , 45) . It is unlikely, therefore, that E1AF, an ets binding transcription factor associated with NB invasion (38) that induces MMP-9 transcription through the ets element (27) , or EGF, which induces MMP-9 transcription through the ets element (29) , was responsible for N phenotype-specific MMP-9 transcription.

v-src enhances MMP-9 transcription in HT-1080 cells through the GT-box (25) . Herbimycin A, at c-src-inhibitory concentrations (33) , abrogated MMP-9 expression by N-type cells, which expressed elevated levels of c-src protein and kinase activity compared with S-type cells. This indicates a role for tyrosine kinases (possible c-src) in N phenotype-specific MMP-9 expression, consistent with GT-box usage, and also confirms an association between enhanced c-src expression and neuronal NB differentiation (52) . The EGF tyrosine kinase inhibitor lavendustin A (35) failed to influence N phenotype-specific MMP-9 expression, further suggesting independence from EGF activity.

The N-myc gene is amplified in up to 20% of NBs (53) , and its expression has been associated with enhanced NB invasivity (38 , 40) . Invasivity of N-myc-amplified NB cells has also been correlated with MT-MMP-1 but not with MMP-9 expression (38) . Because S- and N-type SK-N-SH cells did not exhibit N-myc gene amplification or expression, nor did they express MT-MMP1, our data would suggest an alternative mechanism for enhanced NB invasivity that may be more relevant to non-N-myc-amplified disease.

The association between N phenotype, MMP-9 expression, and enhanced invasivity was also observed in SH-SY5Y, LA-N-1, and SK-N-AS NB cell lines, suggesting that phenotype-dependent regulation of invasivity associated with up-regulated MMP-9 expression may be a relatively common phenomenon in NB.

Our data implicate MMP-9 in the regulation of NB progression. Indeed, elevated MMP-9 levels are associated with advanced stage metastatic NB in vivo (20) . However, a stromal origin for MMP-9 has been hypothesized. Stromal MMP-9 staining may, however, also reflect stromal binding of tumor enzyme (54, 55, 56, 57) . Indeed, the reverse has been reported for MMP-2 (58) . Thus, localization of MMP-9 mRNA transcripts is required to clearly identify the MMP-9-expressing elements within NBs. Unfortunately, NB tumor samples were not available for this study. N-type SK-N-SH cells did not, however, stain positively for membrane or cytoplasmic MMP-9, despite secreting the enzyme (data not shown), supporting the possibility that tumor elements may express but not stain for MMP-9 in vivo. Furthermore, caution should be taken when evaluating MMP-9 transcription in vivo, because MMP-9 is a rare message detected only by reverse transcription-PCR or in highly purified mRNA (21 , 59) , even when high levels of protein are expressed. Therefore, regular in situ hybridization may underestimate tumor MMP-9 expression and should be evaluated, and in some cases reevaluated, by reverse transcription-PCR in situ hybridization.

In conclusion, we propose that spontaneous S to N phenotype conversion exhibited by SK-N-SH cells regulates malignant behavior by enhancing invasivity. Enhanced invasivity involves enhanced motility and MMP-9-dependent invasivity, associated with the transcriptional up-regulation of MMP-9 expression through the GT-box and NF{kappa}B elements. This report extends our knowledge concerning the mechanisms involved in phenotype-dependent regulation of malignant NB behavior to include regulation of invasivity and is the first report to demonstrate a direct role for MMP-9 in malignant NB behavior and the first to examine transcriptional regulation of MMP-9 in NB cells. Our data suggest a novel mechanism in the transcriptional up-regulation of MMP-9 expression dependent upon reciprocal cooperation between GT-box and NF{kappa}B elements, independent of the proximal AP-1 element, suggesting that the GT-box can substitute the AP-1 element as the essential but insufficient regulator of MMP-9 transcription in this cell line. This may also provide a mechanism for the differential regulation of MMP-9 and TIMP-1 and suggests tumor-specific usage for the GT-box or AP-1 elements for basal MMP-9 transcription. Finally, we propose that GT-box transcription factors, critical for N phenotype-specific MMP-9 expression in SK-N-SH cells, must be considered potential regulators of malignant NB behavior.


   Materials and Methods
TOP
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
References
 
Cells, Media, and Reagents.
The human NB cell lines SK-N-SH, SH-EP, SH-SY5Y, LA-N-1, and SK-N-AS cell lines were grown in Eagle Minimal Essential Medium (EMEM) (S-K-N-SH, SH-EP, and SH-SY5Y) or RPMI (LA-N-1 and SK-N-AS) supplemented with 10% NBCS, antibiotics, glutamine, and sodium pyruvate. S to N conversion by SK-N-SH cells was promoted by elongating time to passage and maintaining cells at high density. SK-N-SH cells that were either >95% S or >95% N were used for assays. Comparisons were made between N- and S-type cells derived from the same stocks. Calphostin C, lavendustin A, and herbimycin A were purchased from Calbiochem (Nottingham, United Kingdom). Anti-MMP-9 anticatalytic antibody and recombinant human TIMP-1 were purchased from Oncogene Science (Cambridge, MA). Antihuman plasminogen, vimentin, NSE, and Bcl-2 antibodies were purchased from Sigma Chemical Co., St. Louis, MO).

Matrigel Invasion Assay.
Invasion through reconstituted BM Matrigel was performed by a modification of a method described previously (10) . Briefly, 8 µm Millipore filters were used to assay tumor cell invasion. Cells were grown to subconfluence, detached, washed in PBS, and resuspended at a concentration of 0.5 x 106 cells/ml in invasion medium (Eagle Minimal Essential Medium (EMEM)/0.1% BSA with antibiotics, without serum). Cell suspensions were added to the upper well of invasion chambers and incubated at 37°C for 20 h. Cells that had traversed the filters were counted by light microscopy, after staining with hematoxylin and the removal of surface adherent cells. Motility assays were essentially the same as invasion assays except that uncoated 8 µm Millipore filters were used as a barrier to motility.

Preparation of Nuclear Extracts.
Nuclear extracts from SK-N-SH cells were prepared according to the procedure of Dignam et al. (60) , with the following modification: buffer C contained 20% glycerol, 0.45 M NaCl, and 20 mM HEPES (pH 7.5). Protein inhibitors leupeptin (4 µg/ml) and pepstatin A (4 µg/ml) were added to all buffers. The protein concentration of extracts was determined by Bio-Rad assay and adjusted to 5 mg/ml.

In Vitro Footprinting.
32P-end labeled MMP-9 5' flanking region probes spanning from -637 to -435 and from -220 to 6, relative to the transcriptional start site, were generated by PCR using primer sets ARM10 (5'-CGT CGA CAA GAT TCA GCC TGC GGA AGA C-3') end-labeled with 32P and FA159 (5'-CCA TCC TTG GCC TTT TGC AAC ACC CCC T-3'); and FA158 (5'-GTC TGG GGT CTT GCC TGA CTT GGC AGT-3') and FA130 (5'-GTG TCT GAC TGC AGC TGC TGT TGT GG-3') end-labeled with 32P, respectively. PCR fragments were purified by 8% polyacrylamide gel electrophoresis. Purified end-labeled fragments (50,000 cpm) were incubated in a final volume of 50 µl and final concentrations of 70 mM NaCl, 2 mM MgCl2, 20 mM HEPES pH 7.5, 0.2 mM EDTA, 0.2 mM DTT, 10% glycerol, 1 µg poly(deoxyinosinic-deoxycytidylic acid), and 15–50 µg of nuclear extract for 30 min at room temperature. After incubation, 5 µl of 5 mM CaCl2/10 mM MgCl2 solution were added to reactions and incubated for 1 min at room temperature. Reactions were then digested with 1–3 units of DNase 1 (Pharmacia, Uppsala, Sweden), and diluted in 5 mM CaCl2/10 mM MgCl2 solution for 1 min at room temperature; reactions were stopped by adding 140 µl of a solution containing 194 mM NaOAC, 32 mM EDTA, 0.14% SDS, 64 µg tRNA, and 50 µg of proteinase K, with incubation for 30 min at 42°C. Reactions were phenol/chloroform extracted and ethanol precipitated; degradation fragments were analyzed by 8% denaturing PAGE.

EMSAs.
EMSAs were carried out as described previously (61) . Binding reactions were performed at room temperature for 25 min. The binding reactions contained 32P 5'-end-labeled, double-stranded, oligonucleotide probe, 2 µg of poly(deoxyinosinic-deoxycytidylic acid), 5 µg of SK-N-SH nuclear extract, and additional competitor DNAs or antibody as specified in the figure legends. The oligodeoxribonucleotides used in this study were as follows: MMP-9 AP-1 (5'-CCTGACCCCTGAGTCAGCACTTGCCTGT-3'); MMP9mAP-1 (5'-CCTGACCCCgtcaaCcGCACTTGCCTGT-3'); MMP-1 AP-1 (5'-TTAAGCATGAGTCAGACACCT-3'); MMP9-GT (5'-CTGTCAAGGAGGGGTGGGGTCACAG-3'); MMP9mGT (5'-CTGTCAAGGAGGGtTaacGTCACAG-3'); NS (5'-TCGATAGGGAATTTACACGC-3'); MMP9-NF{kappa}B (5'-TGCCCCAGTGGAATTCCCCAGCCTTG-3'); MMP9mNF{kappa}B (5'-TGCCCCAGTGGcAacCCggAGCCTTG-3'); MMP9-SP1 (5'-TGTCCTTCCGCCCCCAGATGAA-3'); MMP9mSP1 (5'-TGTCCTTCgtcaaCCAGATGAA-3'); MMP9ets (5'-TAGCAGAGCCCATTCCTTCCGC-3'); and MMP9ets2 (5'-TAGCAGGGAGAGGAAGCTGAGT-3'). All oligonucleotides were double stranded; the complementary affinity strands are not indicated; lower cases refer to mutated bases.

Antibodies.
Purified fos antibody generated from amino acids 129–153 of the c-fos protein was kindly donated by Dr. M. Iadarola (NIH, Bethesda, MD). Antisera against c-jun and Fra-2 were kindly provided by Dr. T. Curran (Hoffman-LaRoche Inc., Nutley, NJ). Antibodies against junB, cfos, fosB, and fra-1 were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and antisera against junD and jun family were a gift from Dr. R. Bravo (Bristol-Mayer Squibb, Princeton, NJ). The anticatalytic monoclonal and polyclonal anti-MMP9, TIMP-1, and TIMP-2 antibodies have been described previously (23 , 39 , 46) .

Northern Blot Analysis.
Total and messenger RNAs were isolated from SK-N-SH cells, and Northern blots were prepared. Briefly, total RNA was extracted using RNasol (Iso-Tex, Friendsville, TX) as directed. Poly(A)-selected mRNA was purified from total RNA by oligo(dT) affinity chromatography as described previously (62) . Purified mRNA (2 µg) was heat denatured and separated on a 1.4% agarose gel in 10 mM sodium phosphate buffer and transferred to nylon membranes (Hybond N+; Amersham); blots were hybridized overnight with 32P-labeled cDNA probes in 40% formamide, 10% dextran sulfate, 4 x SSC (1 x SSC = 150 mM NaCl, 15 mM sodium citrate) and 0.1% SDS at 42°C, as described previously (63) . The filters were then washed extensively in 0.2 x SSC containing 1% SDS at 60°C and exposed to XAR-5 film (Kodak, Rochester, NY). Quantitative loading of RNA was determined by hybridization to a glyceraldehyde-3-phosphate dehydrogenase control cDNA probe.

Transient Transfection Assays.
SK-N-SH cells were grown on 100-mm Petri dishes. Supercoiled plasmid DNA (7 µg) was transfected into SK-N-SH cells using the calcium-phosphate precipitation method (64) . All cells were cotransfected with a control plasmid pRSVßgal (1.5 µg; Pharmacia, Uppsala, Sweden) for the estimation of transfection efficiency. After transfection, cells were incubated for 48 h and harvested; proteins were extracted and prepared for CAT and ß-galactosidase assays, as described previously (65 , 66) . The data determined in CAT assay were quantified using a scanning densitometer (Bio-Rad Model GS670) and Molecular AnalystTM PC software. CAT activity was normalized for variation in transfection efficiency in accordance with data obtained from ß-galactosidase control assays.

Plasmid Constructs.
Construct -670MMP9CAT was kindly provided by Drs. Sato and Seiki (University of Kanazawa, Ishikawa, Japan) and contained 5' regulatory sequence of the MMP-9 gene from position 53 to -670 relative to the transcriptional start site (17) . To create the construct -586MMP9CAT, in which the NF{kappa}B site was deleted, primers FA125 (5'-AAC CCG GGT TGC CTA GCA GAG CCC ATT-3') and FA129 (5'-GAC TCT AGA GGA TCC CCA GGA GCA CCA GGA CCA G-3') were used to PCR amplify -670MMP-9CAT to obtain a fragment that terminated at nucleotide -586. SmaI and XbaI restriction sites added to PCR primers were used to ligate the PCR fragment into the SmaI and XbaI sites of the promoterless PGEM4ZCAT vector. The construct -460MMP9CAT, in which NF{kappa}B, SP-1, and ets sites were deleted, was prepared from a SmaI/EcoRV digest of -670MMP9CAT, the major SmaI/EcoRV fragment gel purified and religated to give -460MM9CAT. To create the constructs -88MMP9CAT (NF{kappa}B, SP1, ets, and CACA regions deleted), -75MMP9CAT (NF{kappa}B/SP1/ets/CACA and AP-1 regions deleted), and -49MMP9CAT (NF{kappa}B, SP1, ets, CACA, AP-1, and GT regions deleted), PCR products were generated from -670MMP9CAT using primer FA129 with primers FA93 (5'-CCTGACCCCTGAGTCAGCACTTGCCTGT-3'), FA168 (5'-CTGTCAAGGAGGGGTGGGGTCACAG-3'), and FA128 (5'-AACCCGGGTGGGGTCACAGGAGCGCCT-3'), respectively. PCR fragments were purified and blunt end/XbaI ligated into the promoterless PGEM4ZCAT vector for constructs -88 and -75MMP9CAT and SmaI/XbaI ligated into the same vector for construct -49MMP9CAT. Constructs bearing mutations in NF{kappa}B (-670MMP9mNF{kappa}B), SP-1 (-670mSP1), ets (-670mets), AP-1 (-670mAP-1), and GT (-670mGT) elements were prepared from PCR fragments by restriction digestion and ligation. For -670mNF{kappa}B, a PCR fragment spanning -670 to -598 obtained using primers FA157 (5'-AAAGAGCTCTAGAGGCTGCTACTGTCCCCT-3') and FA135 (5'-CTGTTGACACTGGGGCAACCCCCTGCCTTCCGCAGG-3') digested with HincII was ligated to a HincII-digested second PCR fragment spanning -597 to 28 of -670MMP9CAT generated using primers FA136 (5'-ACGTCAACCCAGAGCCTTGCCTAGCAGAGCCCATTCCTT-3') and FA137 (5'-ACTCTAGAGATCCCCAGGAGCACCAGGAGGAGCACCAGGACCAGGGGCTGCCAGAGGCT-3') to yield -670MMP-9CAT mutated in the NF{kappa}B site from 5'-GGAATTCCCC3' to 5'-GGcAaccCgg-3'. An XbaI digest of this fragment was ligated into the XbaI site of promoterless PGEM4ZCAT. A similar strategy was used to introduce mutations to SP-1, ETS, TRE, and GT sites using flanking PCR primers FA157 and FA137 for all reactions used with specific primers: FA140 (5'-ACGTCGACGAAGGAATGGGCTCTGCTAGGCAAGGC-3') and FA141 (5'-ACGTCAACCAGATGAAGCAGGGAGAGGAAGCTGAG-3') for -670mSP1; FA144 (5'-ACGTCGACCTCCCTGCTTCATCTGGGGGCG- GAAG-3') and FA145 (5'-ACGTCAACCTGAGTCAAAGAAGGCTGTCAGG GAG-3') for -670mets; FA133 (5'-GCGGTTGACGGGGTCAGGGTGTGTGTGTGTGTGTGT-3') and FA134 (5'-CCGTCAACCGCACTTGCCTGTCAAGGAGGGGTGGGGTCA-3') for -670mAP-1; and FA172 (5'-CCCGTTAACGTCACAGGAGCGCCTCCTTAA-3') and FA173 (5'-CCCGTTAACCCTCCTTGACAGGCAAGTGCT-3') for -670mGT. This strategy introduced the following mutation for SP-1 (-563 to -557) from CCGCCCC to CgtCaaC; for ETS (-540 to -534) from AGGAAGC to gtcgAgC; for AP-1 (-79 to -73) from TGAGTCA to gtcacCg; and for GT (-54 to -46) from GGGGTGGGG to GGGtTaacG.

Heterologous promoter reporter gene constructs containing three repeats of the MMP-9, MMP-1, or TIMP-1 AP-1 elements; and one repeat of the MMP-9 GT-box, SP-1, or NF{kappa}B element were created by ligating klenow blunt-ended, double-stranded, annealed oligonucleotides upstream of the cfos promoter into the SalI site of the reporter gene vector {Delta}56CAT (67) .

All constructs were sequenced by the dideoxy-termination method using Sequenase Version 2.0 kits as directed by the supplier (USB).

Immunoblotting.
Immunoblots were performed by a modification of a method described previously (68) . Briefly, nuclear extracts (80 µg) for jun and fos proteins, total cell extracts (maximum protein loaded, 100 µg; adjusted for equal cell numbers) for p60 c-src protein, or concentrated culture supernatants for MMP-9 were separated by 10% SDS-PAGE under reducing conditions. Gels were run in the presence of prestained, low-range molecular weight standards (Bio-Rad, Hercules, CA). The gels were electroblotted to nitrocellulose filters. Membranes were dried completely before further processing. NS protein binding sites were blocked on membranes for 2 h at room temperature using a 5% solution of nonfat dried milk in PBS. After blocking, membranes were washed in PBS and incubated initially with primary antibody (1–5 µg/ml) diluted in blocking solution for 1 h at room temperature. The membranes were then washed three times for 15 min in PBS and incubated for an additional hour at room temperature with secondary antibody conjugated to horseradish peroxidase diluted 1:3000 in blocking solution. After extensive washing, antigen reactivity was demonstrated by chemiluminescence reaction (Amersham International, Bedford, United Kingdom). Immunoreactive bands were visualized on Kodak XAR-5 film. Molecular weights were approximated by comparison to prestained molecular weight markers (Bio-Rad) using Molecular Analyst PC software for the Bio-Rad model GS-670 imaging densitometer.

Immunoprecipitation and c-src Kinase Assay.
Assays were performed by a modification of a method described previously (69) . S- or N-type cells (20 x 106) were washed with PBS and lysed by repeated aspiration through a 21-gauge needle in 3 ml of RIPA buffer containing PBS (1x), 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, 100 µg/ml phenylmethylsulfonyl fluoride, 0.15 unit/ml aprotinin, and 1 mM sodium orthovanadate. Cellular debris was removed by centrifugation (3000 rpm, 4°C for 15 min), and 1 µg of monoclonal anti-c-src antibody (mAb#327; Ref. 69 ) was added; controls received 1 µg of preimmune mouse IgG, and reactions were incubated for 1 h at 4°C. Twenty µl of protein A-Sepharose (Sigma Chemical Co., St. Louis, MO; 20% in RIPA buffer) were added to reactions and incubated for an additional 2 h at 4°C. Immunoprecipitates were collected by centrifugation (2500 rpm for 5 min at 4°C), washed four times in cold RIPA buffer, and resuspended in 20 µl of kinase reaction buffer [20 mM HEPES (pH 7.5) and 5 mM MgCl2]. Ten µl of ATP mix containing 1 µl of 32P-labeled {gamma}ATP, 0.4 µl of 1 M MgCl2, 0.3 µl of 10 mM ATP, 8.3 µl of kinase reaction buffer, and 1 µg of GAPp62 src tyrosine kinase substrate (Santa Cruz Biotechnology) was added to immunoprecipitates and incubated for 20 min at 30°C. After incubation, samples were subjected to reducing 10% SDS-PAGE, and the levels of GAP p62 substrate kinased were determined by autoradiography of dried gels.

Substrate Gel Electrophoresis.
Regular gelatin and reverse zymograms were prepared as described previously (21) . Briefly, samples were subjected to regular nonreducing SDS-PAGE in gels copolymerized with 0.1% gelatin (regular) or 0.1% casein and 30% v/v 72-h serum-free MDA-MB-231 conditioned media (reverse zymograms). After electrophoresis, gels were washed in 2% Triton X-100, rinsed in water, and incubated in a buffer containing 50 mM Tris, 0.2 M NaCl, and 5 mM CaCl2 (pH 8.0) overnight (regular) or for 72 h (reverse) at 37°C. Enzyme/inhibitor activity was visualized by staining with Coomassie Blue.

Statistical Analysis.
The Student’s t test was used for statistical comparison of data. A comparison of means giving t values with associated probabilities of difference <0.05 was considered to be statistically different.


   Acknowledgments
 
We thank Dr. M. Seiki (University of Kanazawa, Ishikawa, Japan) for the -670MMP-9 reporter gene construct, Dr. S. Alemà (Consiglio Nazionale delle Ricerche, Rome, Italy) for the monoclonal anti-c-src antibody, Dr. T. Curran (Hoffman-LaRoche Inc., Nutley, NJ) for anti-cJun and Fra-2 antibodies, Dr. R. Bravo (Bristol-Mayer Squibb, Princeton, NJ) for anti-junD antibodies, and Dr. M. Iadrola (NIH, Bethesda, MD) for anti-cfos antibody.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 This work was supported by grants from the Associazione Italiana per la Lotta al Neuroblastoma, the Associazione Italiana per la Ricerca sul Cancro, National Research Council (CNR) Biotechnology Project, and MURST (40%). Back

2 The first two authors contributed equally to this work. Back

3 To whom requests for reprints should be addressed, at Section of Molecular Pathology, Department of Experimental Medicine, University of L’Aquila, Via Vetoio, Coppito II, 67100 L’Aquila, Italy. Phone: 39-862-433542; Fax: 39-862-433523; E-mail: mackay{at}univaq.it Back

4 The abbreviations used are: NB, neuroblastoma; ECM, extracellular matrix; BM, basement membrane; MMP, matrix metalloproteinase; MT-MMP, membrane-type MMP; NF{kappa}B, nuclear factor {kappa}B; CAT, chloramphenicol acetyltransferase; PKC, protein kinase C; EGF, epidermal growth factor; TIMP, tissue inhibitor of metalloproteinase; NSE, neuron-specific enolase; EMSA, electrophoretic mobility shift assay; NS, nonspecific; APMA, aminophenylmercuricacetate. Back

Received for publication 11/30/98. Revision received 2/26/99. Accepted for publication 3/16/99.


   References
TOP
 Abstract
 Introduction
 Results
 Discussion
 Materials and Methods
 References
 

  1. Tischler A. L., Dichter M. A., Bialis B., Green L. A. Neuroendocrine neoplasms and their cells of origin. N. Engl. J. Med., 296: 919-924, 1977.[Medline]
  2. Rettig W. J., Spengler B. A., Garin-Chesa P., Old L. J., Biedler J. L. Co-ordinate changes in neuronal phenotype and surface antigen expression in human neuroblastoma cell variants. Cancer Res., 47: 1383-1389, 1987.[Abstract/Free Full Text]
  3. Ross R. A., Spengler B. A., Biedler J. L. Co-ordinate morphological and biochemical interconversion of human neuroblastoma cells. J. Natl. Cancer Inst., 71: 741-749, 1983.
  4. Ciccarone V., Spengler B. A., Meyers M. B., Biedler J. L., Ross R. A. Phenotypic diversification in human neuroblastoma cells: expression of distinct neural crest lineages. Cancer Res., 4: 219-255, 1989.
  5. Biedler J. L., Helson L., Spengler E. Morphology and growth, tumorigenicity and cytogenetics of human neuroblastoma cells in continuous culture. Cancer Res., 33: 2643-2652, 1973.[Abstract/Free Full Text]
  6. Biedler J. L., Spengler B. A., Chang T-D., Ross R. A. Transdifferentiation of human neuroblastoma cells results in co-ordinate loss of neuronal and malignant properties Evans A. E. D’Angio G. J. Knudson A. G. Seeger R. C. eds. . Advances in Neuroblastoma Research 2. Progr. Clin. Biol. Res., 271: 265-276, 1988.
  7. Hanada M., Krajewski S., Tanaka S., Cazals-Hatem D., Spengler B. A., Ross R. A., Biedler J. L., Reed J. C. Regulation of Bcl-2 oncoprotein levels with differentiation of human neuroblastoma cells. Cancer Res., 53: 4978-4986, 1993.[Abstract/Free Full Text]
  8. Lasorella A., Lavarone A., Isreal M. Differentiation of neuroblastoma enhances Bcl2 expression and induces alterations of apoptosis and drug resistance. Cancer Res., 55: 711-716, 1995.
  9. Cottam D. W., Rees R. C. Regulation of matrix metalloproteinases: their role in tumor invasion and metastasis. Int. J. Oncol., 2: 861-872, 1993.
  10. Albini A., Iwamoto Y., Kleinman H. K., Martin G. R., Aaronson S. A., Kozlowski J. M., McEwan T. N. A rapid in vitro assay for quantitating the invasive potential of tumour cells. Cancer Res., 47: 3239-3245, 1987.[Abstract/Free Full Text]
  11. Mignatti P. E., Robbins E., Rifkin D. B. Tumor invasion through the amniotic membrane: requirements for a proteolytic cascade. Cell, 47: 487-498, 1986.[Medline]
  12. Tiberio A., Farina A. R., Tacconelli A., Cappabianca L., Gulino A., Mackay A. R. Retinoic acid enhanced invasion through reconstituted basement membrane by human SK-N-SH neuroblastoma cells involves membrane associated tissue-type plasminogen activator. Int. J. Cancer, 73: 740-748, 1997.[Medline]
  13. Farina A. R., Coppa A., Tiberio A., Tacconelli A., Turco A., Colletta G., Gulino A., Mackay A. R. Transforming growth factor ß1 enhances the invasiveness of human MDA-MB-231 breast cancer cells by upregulating urokinase activity. Int. J. Cancer, 75: 721-730, 1998.[Medline]
  14. Ballin M., Gomez D. E., Sinha C. C., Thorgeirsson U. P. Ras oncogene mediated induction of a 92kDa metalloproteinase: strong correlation with the malignant phenotype. Biochem. Biophys. Res. Commun., 154: 832-838, 1988.[Medline]
  15. Wilhelm S. M., Collier I. E., Marmer B. L., Eisen A. Z., Grant G. A., Goldberg G. I. SV40 transformed human lung fibroblasts secrete a 92kDa type IV collagenase, which is identical to that secreted by human macrophages. J. Biol. Chem., 264: 17213-17221, 1989.[Abstract/Free Full Text]
  16. Bernhard E. J., Muschel R. J., Hughes E. N. Mr 92,000 gelatinase release correlates with the metastatic phenotype in transformed rat embryo cells. Cancer Res., 50: 3872-3877, 1990.[Abstract/Free Full Text]
  17. Sato H., Seiki M. Regulatory mechanism of 92kDa type IV collagenase gene expression which is associated with invasiveness of tumor cells. Oncogene, 8: 395-405, 1993.[Medline]
  18. Mackay A. R., Hartzler J. L., Pelina M. D., Thorgeirsson U. P. Studies on the ability of 65-kDa and 92-kDa tumor cell gelatinases to degrade type IV collagen. J. Biol. Chem., 265: 21929-21934, 1990.[Abstract/Free Full Text]
  19. Bernhard E. J., Gruber S. B., Muschel R. J. Direct evidence linking expression of matrix metalloproteinase-9 (92kDa gelatinase/collagenase) to the metastatic phenotype in transformed rat embryo cells. Proc. Natl. Acad. Sci. USA, 91: 4293-4297, 1994.[Abstract/Free Full Text]
  20. Sugiura Y., Shimada H., Seeger R. C., Laug W. E., DeClerck Y. A. Matrix metalloproteinase-2 and -9 are expressed in human neuroblastoma: contribution of stromal cells to their production and correlation with metastasis. Cancer Res., 58: 2209-2216, 1998.[Abstract/Free Full Text]
  21. Mackay A. R., Ballin M., Pellina M. D., Farina A. R., Nason A. M., Hartzler J. L., Thorgeirsson U. P. Effect of phorbol ester and cytokines on matrix metalloproteinase and tissue inhibitor of metalloproteinase expression in tumor and normal cell lines. Invasion Metastasis, 12: 168-184, 1992.[Medline]
  22. Sato H., Kida Y., Mai M., Endo Y., Tanaka J., Seiki M. Expression of genes encoding type IV collagen-degrading metalloproteinases and tissue inhibitors of metalloproteinases in various human tumor cells. Oncogene, 7: 77-83, 1992.[Medline]
  23. Festuccia C., Bologna M., Vicentini C., Tacconelli A., Miano R., Violini S., Mackay A. R. Increased matrix metalloproteinase-9 secretion in short-term tissue cultures of prostatic tumor cells. Int. J. Cancer, 69: 1-8, 1996.[Medline]
  24. Himelstein B. P., Lee E. J., Sato H., Seiki M., Muschel R. J. Transcriptional activation of the matrix metalloproteinase-9 gene in an H-ras and v-myc transformed rat embryo cell line. Oncogene, 14: 1995-1998, 1997.[Medline]
  25. Sato H., Kita M., Seiki M. v-Src activates the expression of 92-kDa type IV collagenase through the AP-1 site and the GT box homologous to retinoblastoma control elements. A mechanism regulating gene expression independent of inflammatory cytokines. J. Biol. Chem., 268: 23460-23468, 1993.[Abstract/Free Full Text]
  26. Gum R., Lengyel E., Juarez J., Chen J. H., Sato H., Seiki M., Boyd D. Stimulation of 92-kDa gelatinase B promoter activity by ras is mitogen-activated protein kinase kinase-1 independent and requires multiple transcription factor binding sites including closely spaced PEA3/ets and AP-1 sequences. J. Biol. Chem., 271: 10672-10680, 1996.[Abstract/Free Full Text]
  27. Higashino F., Yoshida K., Noumi T., Seiki M., Fujinaga K. Ets-related protein E1A-F can activate three different matrix metalloproteinase gene promoters. Oncogene, 10: 1461-1463, 1995.[Medline]
  28. Himelstein B. P., Lee E. J., Sato H., Seiki M., Muschel R. J. Tumor cell mediated transcriptional activation of the fibroblast matrix metalloproteinase-9 gene: involvement of multiple transcription factors including Ets and an alternating purine-pyrimidine repeat. Clin. Exp. Metastasis, 16: 169-177, 1998.[Medline]
  29. Watabe T., Yoshida K., Shindoh M., Mitsunori K., Fujikawa K., Sato H., Seiki M., Ishii S., Fujinaga K. The ETS-1 and ETS-2 transcription factors activate the promoters for invasion-associated urokinase and collagenase genes in response to epidermal growth factor. Int. J. Cancer, 77: 128-137, 1998.[Medline]
  30. Young P. P., Mendelson C. R. A GT box is essential for basal and cyclic adenosine 3',5' monophosphate regulation of the human surfactant protein A2 gene in alveolar type II cells: evidence for the binding of lung nuclear factors distinct from SP-1. Mol. Endocrinol., 11: 1082-1093, 1997.[Abstract/Free Full Text]
  31. Ikezu T., Okamoto T., Murayama Y., Okamoto T., Homma Y., Ogata E., Nashimioto I. Bi-directional regulation of c-fos promoter by an oncogenic gip-2 mutant of G alpha i2. A novel implication of retinoblastoma gene product. J. Biol. Chem., 269: 31955-31961, 1994.[Abstract/Free Full Text]
  32. Pai S. R., Bird R. C. Interaction of the Rb tumour suppresser protein with the c-fos promoter in c-fos transfected cells overexpressing c-fos and Rb. Anticancer Res., 17: 3265-3272, 1997.