Failure of wild-type p53 gene therapy in human cancer cells expressing a mutant p53 protein

The introduction of exogenous wild-type p53 into human cancer cells bearing p53 mutation does not necessarily result in inhibition of tumor growth. We have demonstrated this in MDA-MB468 breast cancer cells which are hemizygous for p53 mutation and also in KM12SM colorectal carcinoma cells which are heterozygous for p53 mutation. The wtp53 transfectants decreased three- to four-fold the number of colonies compared with controls. Most wtp53-expressing cells died by apoptosis at early passages, but some cells were able to form colonies and their proliferation rate was similar to control transfectants. This reversion was observed in three of the six MDA-MB-468 clones selected. When MDA-wtp53 transfectants were implanted orthotopically in nude mice only one clone showed prolonged tumor latency. No differences were found in either tumor proliferation or apoptosis in tumors. Integration and expression of exogenous wtp53 was assessed in early and late passages in vitro, and in tumors growing in vivo. Consistently, we found mutations in the exogenous wtp53 gene of MDA-MB468 transfectants. Excision of the exogenous gene was an alternative to abrogate the wtp53 function that was extremely efficient in KM12 cells, although they maintained resistance to geneticin. These results were corroborated by the functional assay in yeast. In conclusion, wtp53 is inactivated in these cancer cells by different mechanisms. The presence of mutated p53 may confer genome instability and mutator ability, which allows cells to escape the effects of the exogenous wtp53 and contributes to the failure of wtp53 gene therapy.

Detection of differentially expressed gelatinase A in metastatic and non-metastatic subpopulations of tumor cells by target RNA arbitrarily primed polymerase chain reaction (TRAP-PCR)

We have developed a novel procedure called Targeted RNA AP-PCR (TRAP-PCR) to quantitatively measure specific mRNA expression. The target mRNA is reverse transcribed using a specific primer and PCR is performed under low stringency conditions to generate a rich fingerprint-type band pattern. In this situation multiple sequences are coamplified with the targeted sequence. The amplification is carried out in a competitive fashion and is, in consequence, quantitative. We have applied this technique to determine Gelatinase A (Gel A) mRNA expression in the MXT mouse mammary carcinoma system. TRAP-PCR analysis using primers for Gel A produced a reproducible fingerprint including one major band whose identity was confirmed to be Gel A cDNA. Highly metastatic MXT subclones show an increased Gel A expression. Results were confirmed by Northern blot and protein activity (gelatin zymography). TRAP-PCR is a simple, sensitive and specific technique to comparatively quantify mRNA expression and requires less template than conventional methods.

Down-regulation of E-cadherin in mouse skin carcinoma cells enhances a migratory and invasive phenotype linked to matrix metalloproteinase-9 gelatinase expression

To assess the role of gelatinases in mouse skin tumor progression and their link to the expression of E-cadherin (E-CD), the cell-cell adhesion protein, we used the highly metastatic squamous HaCa4 cell line and several HaCa4-derived clones obtained by transfection of the mouse E-CD cDNA. Expression of matrix metalloproteinase-9 (MMP-9) mRNA and protein activity were present in E-CD (-) HaCa4 and control clones in culture, but they were strongly diminished in E-CD (+) clones (E24 and E62) at subconfluence. To explore the suppressive effect of the cell-cell contacts mediated by E-CD on MMP-9 expression, we introduced a plasmid encoding mouse E-CD antisense cDNA into the E24 cell clone. The transfectant P1-clones obtained with reduced or absent E-CD expression showed increased levels of MMP-9 gelatinase, motility in vitro, and metastatic potential in vivo. Expression of MMP-9 in the various cell clones was also negatively modulated by cell density, but this effect was much stronger in E-CD (+) cells, despite the fact that all of the cell clones analyzed maintained the expression of P-cadherin and made cell-cell contacts at high cell density. Our results indicate that in this cell system, the E-CD-mediated cell-cell contacts are involved in the down-regulation of MMP-9 expression. Thus, the loss of E-CD triggers a migratory and invasive phenotype in mouse squamous carcinoma cells.

Metastatic ability of MXT mouse mammary subpopulations correlates with clonal expression and/or membrane-association of gelatinase A

We have developed a novel murine mammary tumor system with variants representing different stages of tumor progression. The MXT-s parental cell line was established from a urethane-induced and hormone-sensitive mammary tumor. MXT-s parental cells are highly tumorigenic but poorly metastatic. MXT clones and variants were selected by either in vitro or in vivo procedures, and they differ in metastatic ability and 17 beta-estradiol dependency for tumor growth. The MXT-c1.1 and MXT-B2 cell lines produced lung metastasis after intravenous injection into 100% of syngenic mice, but only MXT-c1.1 cells were highly metastatic from intramammary tumors. The fingerprints obtained by arbitrarily primed-polymerase chain reaction demonstrated that the metastatic variants and clones had a common genetic background and resulted from clonal selection from the parental cell line. We studied whether the matrix metalloproteinase (MMP) profile is correlated with tumor progression and metastatic ability in the MXT tumor system. Gelatinases A and B were assayed in the cells, both by enzyme activity and mRNA expression. Gelatinase A was expressed in MXT-c1.1 cells, whereas MXT-B2 cells did not express either MMP. In contrast, the mammary fat pad tumors expressed both gelatinases. Membrane Type 1-MMP transcripts were also detected in MXT cells and tumors. Because the mRNA levels of gelatinase. A were low in MXT-B2 tumors, we suggested that exogenous gelatinase A bound the cell membranes of MXT-B2 cells in vivo. Indirect evidence was obtained in vitro by treatment of MXT-B2 cells with NIH/3T3 fibroblast-conditioned medium. After this treatment, we detected a gelatinolytic activity at M(r) 68,000 in the cell-membrane extract of MXT-B2 cells and an increase in migratory ability through type IV collagen matrices. On the other hand, Ha-ras gene dosage correlated positively with metastatic ability but not with either gelatinase A or gelatinase B expression. No significant differences were observed in the expression of stromelysin-1 and tissue inhibitors of MMP. Thus, in the MXT tumor system, the expression of gelatinase A or its cell association and Ha-ras gene dosage independently contribute to the metastatic phenotype.

Stimulation of angiogenesis as an explanation of Matrigel-enhanced tumorigenicity

Matrigel, a reconstituted extract of basement membrane, enhances the growth of different human cancer cell lines when transplanted into nude mice. Here that stimulation was confirmed in the BALB/c murine mammary-tumor cell line M3MC, as well as in human colon (SW948) and mammary (MDA-MB-468) carcinoma cell lines transplanted in nude and SCID mice, respectively. Subcutaneous and intra-mammary fat-pad inoculations of Matrigel alone generated an angiogenic response which was macroscopically evident by day 9. Histological analysis of the local host reaction occurring at the site of injection revealed an early peripheral fibroblast response, followed by mononuclear cell infiltration, solid and hollow fibroblast cords projections from the edge to the center of the Matrigel plug, and finally capillary ingrowths. Conditioned media obtained from the gels generated in vivo, acted as very strong chemoattractants for mouse lung capillary endothelial cells, stimulating their motility between 38 and 82 times with respect to the control. Our results suggest an important role of host cells recruited by Matrigel, which could favor angiogenesis of the area and thus facilitate the growth of tumor cells co-inoculated with the basement membrane extract.

Transforming growth factor release by human tumor cell lines and their interactions with other growth factors

We have studied the production of transforming growth factor (TGF) by several human tumor cell lines and their interactions with exogenously added epidermal growth factor (EGF) and insulin. TGF-like activities were present in all the conditioned media tested. The clonogenic capacity of the tumor cell lines had no correlation with the TGF-like activity production. EGF and insulin had a promoting colony-forming activity on tumor cells but this effect was not additive. Moreover, an inverse statistically significant correlation (-0.817, p less than 0.05) was found between the response to exogenous EGF and the EGF or TGF-alpha production by tumor cell lines. The EGF receptor (EGF-R) was not detected in any of the melanomas studied, nor in breast adenocarcinoma cell lines which were producers of EGF or TGF-alpha.