Gene trapping identifies chloride channel 4 as a novel inducer of colon cancer cell migration, invasion and metastases.

BACKGROUND: To date, there are few reports on gene products contributing to colon cancer progression. METHODS: We used a gene trap comprised of an enhanced retroviral mutagen (ERM) cassette that includes a tetracycline-responsive promoter upstream of a haemagglutinin (HA) tag and a splice donor site. Integration of the ERM within an endogenous gene yields a tetracycline-regulated HA-tagged transcript. We transduced RKO colon cancer cells expressing a tetracycline trans-activator-off with the ERM-encoding retrovirus and screened for enhanced migration. RESULTS: One clone showed fivefold enhanced migration with tetracycline withdrawal. Rapid amplification of cDNA ends identified the trapped gene as the chloride channel 4 (CLCN4) exchanger. Stable expression of a CLCN4 cDNA enhanced motility, whereas cells knocked down or null for this transcript showed reduced migration/invasion. CLCN4-overexpressing RKO colon cancer cells were more resistant than controls to proton load-induced cytotoxicity, consistent with the H(+)-extruding function of this antiporter. Intra-splenic delivery of RKO-CLCN4 transfectants, but not controls, yielded liver metastases, and transcript levels were higher in colon cancer metastases to the liver when compared with primary tumours. CONCLUSIONS: CLCN4 is a novel driver of colon cancer progression.

Identification of an histone H3 acetylated/K4-methylated-bound intragenic enhancer regulatory for urokinase receptor expression.

The transcriptionally regulated urokinase-type plasminogen activator receptor (u-PAR) contributes to cancer progression. Although previous studies have identified multiple 5' regulatory elements, these cis motifs cannot fully account for u-PAR expression prompting a search for hitherto uncharacterized regulatory elements. DNase I hypersensitivity and chromatin immunoprecipitation assays using u-PAR-expressing colon cancer cells indicated a hypersensitive region (+665/+2068) in intron 1 enriched with acetylated histone 3 (H3) and H3 methylated at lysine 4, markers of regulatory regions. The +665/+2068 region increased transcription from a u-PAR-promoter in an orientation- and distance-independent manner fulfilling the criteria of an enhancer. Optimal stimulation of the u-PAR promoter by phorbol ester required this enhancer. Systematic truncations combined with DNase I footprinting revealed two protected regions (+1060/+1099 and +1123/+1134) with deletion of the latter practically abolishing enhancer activity. The +1123/+1134 region harbored non-consensus activator protein-1 and Ets1 binding sites bound with c-Jun (and/or the related JunD/JunB) and c-Fos (and/or the related FosB/Fra-1/Fra-2) as revealed with chromatin immunoprecipitation. Further, nuclear extract from resected colon cancers showed elevated protein binding to a +1123/+1134-spanning probe coordinate with elevated u-PAR protein. Thus, we have defined a novel intragenic enhancer in the u-PAR gene required for constitutive and inducible expression.

Comparison of growth requirements of two human intratumoral colon carcinoma cell lines in monolayer and soft agarose.

The human colon, intratumoral subpopulations HCT 116 and HCT 116a were established in chemically defined medium supplemented with transferrin, insulin, epidermal growth factor (EGF), triiodothyronine, hydrocortisone, and sodium selenite. The responsiveness of the adapted cell lines to these growth factors was compared in anchorage-dependent and -independent assays. HCT 116 cells maintained in serum-free conditions were further adapted to growth factor deprivation, and the effects of these polypeptides were determined in anchorage-independent assays. In monolayer, HCT 116 cells adapted to grow in serum-free medium responded to transferrin but not to EGF or insulin. Similarly adapted HCT 116a cells were, however, insensitive to transferrin addition but manifested a 300 and 500% increase in growth rates with EGF and insulin, respectively. Optimal growth of HCT 116 cells was seen in the presence of insulin and transferrin, while maximum proliferation of HCT 116a cells depended on combined insulin, transferrin, and EGF. In soft agarose, both HCT 116 and HCT 116a subpopulations showed a stringent requirement for transferrin. No combination of growth factors without transferrin supported colony formation. These data suggest that (a) these colon tumor subpopulations may be subject to separate growth controls, and (b) there may be an important role for transferrin in anchorage-independent growth and possibly in the maintenance of malignant characteristics.

In vivo measurements of lung corrections for fast-neutron therapy.

Silicon diodes were inserted into the esophagus and bronchus of anesthetized rhesus monkeys in order to measure the corrections to tumor dose resulting from intervening lung tissue during fast-neutron therapy. The derived corrections were applied to tumor doses for patients being treated for cancer of the esophagus on the fast-neutron beam at TAMVEC. In vivo dosimetry performed on these patients using silicon diodes in the esophagus confirmed the accuracy of the lung corrections. The measured dose and calculated dose agreed to within 4% for four different patients. Tha magnitude of the correction is on the order of 16% for the typical esophageal cancer patient. These studies were also done with Cobalt-60 in order to test, against other data, the results obtained with this animal model.

Expression cloning of novel regulators of 92 kDa type IV collagenase expression.

Overexpression of the 92 kDa type IV collagenase (MMP-9) contributes to cancer progression. However, to date, there are few known regulators of expression of this metalloproteinase. We employed an expression library comprising 500,000 cDNA clones to screen for novel regulators of MMP-9 expression. HT1080 cells were transiently co-transfected with an MMP-9 promoter-luciferase reporter and pools of the cDNA expression library. Positive-scoring pools were subdivided in secondary and tertiary screens, after which the regulatory cDNAs were identified by DNA sequencing. This brief review illustrates the utility of expression cloning in identifying specific regulators of MMP-9 expression.

Insensitivity of laminin degradation directed by receptor bound urokinase to PAI-1 in cultured colon cancer.

This laboratory recently reported that laminin degradation by cultured colon cancer was plasminogen dependent and reflected the presence of urokinase bound to cell surface receptors. (Schlecte, W.; Murano, G.; Boyd D. Cancer Res., 49:6064-6069; 1989). The present study was undertaken to determine the sensitivity of urokinase receptor directed proteolysis to the type I plasminogen activator inhibitor (PAI-1). Colon cancer cell types, that were highly effective in degrading laminin in vitro, elaborated into their conditioned medium an inhibitor which was indistinguishable from PAI-1 on the basis of its performance in reverse zymography, western blotting, and immunoprecipitation assays. A fraction of this PAI-1 was active, as evidenced by complex formation with the active site of radioactive urokinase. Laminin degradation by the colon cancer cells, however, did not appear to be affected by the endogenous inhibitor, since an antibody to the inhibitor, which blocked urokinase-PAI-1 interactions, had little effect on laminin turnover. Further, addition of exogenous PAI-1, activated by guanidine hydrochloride pretreatment, to the colon cancer cells did not perturb laminin degradation. Because laminin degradation by colonic cells was a function of receptor bound urokinase, presumably immobilized plasminogen activator escaped the neutralizing effect of the inhibitor. These data suggest either a shielding effect of the receptor on the plasminogen activator or a physical separation of activator and inhibitor. Either way, for cultured colon cancer at least, laminin degradation directed by urokinase receptor bound plasminogen activator appeared unaffected by the presence of this inhibitor.

Neurohumoral control of esophageal epithelial electrolyte transport.

The neurohumoral control of epithelial esophageal electrolyte transport was investigated by studying the effect of various hormones and neuroeffector agents on the potential difference (PD) in vivo or on the electrical parameters of electrolyte transport in vitro. The rabbit esophagus, which has no submucosal esophageal glands, demonstrated no effect of pentagastrin, cholecystokinin octapeptide, or synthetic secretin in vivo, and no effect of these hormones or of vasopressin, aldosterone, carbachol, epinephrine, or cAMP in vitro. The rabbit esophagus did respond to metabolic substrates (glucose) in vitro by increasing sodium absorption. In contrast, the opossum esophagus, which contains extensive submucosal glands, had a lower electrical resistance, PD, short-circuit current, and sodium absorption with higher chloride secretion. This esophagus responded to carbachol and epinephrine by sodium and chloride secretion. We believe that only the submucosal glands of the esophagus are under significant neurohumoral control while the sodium transporting function of the stratified squamous epithelium of this organ is important in maintaining its barrier function.

KiSS-1 represses 92-kDa type IV collagenase expression by down-regulating NF-kappa B binding to the promoter as a consequence of Ikappa Balpha -induced block of p65/p50 nuclear translocation.

The 92-kDa type IV collagenase (MMP-9) plays a critical role in tissue remodeling. We undertook a study to determine whether the KiSS-1 gene, previously shown to suppress cancer spread (metastases), negatively regulates MMP-9 expression. Six cell lines positive for MMP-9 mRNA were deficient in KiSS-1 mRNA. One of these cell lines, HT-1080, stably transfected with a KiSS-1 expression construct, demonstrated substantially lower MMP-9 enzyme activity/protein and in vitro invasiveness. The lower MMP-9 enzyme activity reflected reduced steady-state mRNA levels which, in turn, was due to attenuated transcription. Activation of ERKs and JNKs by phorbol 12-myristate 13-acetate and tumor necrosis factor alpha, respectively, leading to increased MMP-9 amounts was not antagonized by KiSS-1 expression, suggesting that MAPK pathways modulating MMP-9 synthesis are not the target of KiSS-1. Although MMP-9 expression is regulated by AP-1, Sp1, and Ets transcription factors, KiSS-1 did not alter the binding of these factors to the MMP-9 promoter. However, NF-kappaB binding to the MMP-9 promoter required for expression of this collagenase was reduced by KiSS-1 expression. Diminished NF-kappaB binding reflected less p50/p65 in the nucleus secondary to increased IkappaBalpha levels in the cytosols of the KiSS-1 transfectants. Thus, KiSS-1 diminishes MMP-9 expression by effecting reduced NF-kappaB binding to the promoter.

Water and electrolyte transport by rabbit esophagus.

The nature of the transmural electrical potential difference and the characteristics of water and electrolyte transport by rabbit esophagus were determined with in vivo and in vitro studies. The potential difference of the perfused esophagus in vivo was -28 +/- 3 mV (lumen negative). In vitro the potential difference was -17.9 +/- 0.6 mV, the short-circuit current 12.9 +/- 0.6 muA/cm2, and the resistance 1,466 +/- 43 ohm-cm2. Net mucosal-to-serosal sodium transport from Ringer solution in the short-circuited esophagus in vitro accounted for 77% of the simultaneously measured short-circuit current and net serosal-to-mucosal chloride transport for 14%. Studies with bicarbonate-free, chloride-free, and bicarbonate-chloride-free solutions suggested that the net serosal-to mucosal transport of these two anions accounts for the short-circuit current not due to sodium absorption. The potential difference and short-circuit current were saturating functions of bathing solution sodium concentration and were inhibited by serosal ouabain and by amiloride. Thus active mucosal-to-serosal sodium transport is the major determinant of the potential difference and short-circuit current in this epithelium.

Involvement of urokinase and its receptor in the invasiveness of human prostatic carcinoma cell lines.

We have investigated the role of urokinase (UK) and its cell-surface receptor in determining the invasiveness of prostate cancer cells. Three human cell lines, DU-145, PC-3 and LNCaP, that differ in androgen-responsiveness and growth characteristics, were tested. Analysis of the conditioned medium by an enzyme-linked immunosorbent assay showed secretion of UK by DU-145 (63 ng/mL/10(6) cells/48 hr) and PC-3 (682 ng/mL/10(6) cells/48 hr), but absence of secretion by LNCaP cells. Western blot analysis and enzyme activity assay of the conditioned medium confirmed these results. Scatchard analysis of radioligand binding with acid pretreated cells showed the presence of a single population of high affinity UK receptors on DU-145 cells (93,000 sites/cell, Kd = 0.9 nM) and PC-3 cells (25,000 sites/cell, Kd = 1.0 nM) but not on LNCaP cells. DU-145 and PC-3 cells were found to be highly invasive in in vitro invasion assays: 4.5 +/- 0.5% and 6.5 +/- 0.5%, respectively, of total tumor cells (approximately 2 x 10(5)) had penetrated reconstituted basement membrane (Matrigel) in a 72 hr incubation in serum-free growth medium. Under similar conditions, less than 0.25% LNCaP cells invaded Matrigel. The data indicate that androgen unresponsive, aggressive prostate tumor cells of high metastatic potential, DU-145 and PC-3, secrete UK and display cell-surface UK receptors, fully charged with the protease. Conversely, relatively indolent LNCaP cells of low metastatic potential do not secrete UK nor do they possess its binding sites. UK receptor antagonists, UK 12-32 and UK 6-135, which compete with labeled UK for binding to prostatic cells but do not inhibit cellular proliferation or UK secretion, markedly reduced DU-145 and PC-3 cell invasion (80-85% inhibition), thereby suggesting an important role of receptor-bound UK in prostate tumor cell invasion.

Acute and late effects of 16- and 50-MeVd leads to Be neutrons on the oral mucosa of rhesus monkeys.

Twenty-five rhesus monkeys were randomly assigned to one of five treatment schedules: 1. control group, no irradiation, 2. 60Co five times weekly, 3. 60Co twice weekly, 4. 16-MeVd leads to Be neutrons twice weekly, and 5. 50-MeVd leads to Be neutrons twice weekly. Although the acute reactions were similar in the four irradiated groups, the late sequelae were more severe in the animals irradiated twice weekly with 60Co or neutrons. All of the animals irradiated with 60Co twice weekly or with 16 MeVd leads to Be neutrons exhibited oromucosal necrosis, whereas none of those irradiated five times weekly with 60Co did. The difference in the effect of photon fractionation on early and late radiation sequelae may be related to different patterns of redistribution of surviving cells through the division cycle in tissues responsible for early and late damage.

Regulation of urokinase-type plasminogen activator expression in squamous-cell carcinoma of the oral cavity.

We undertook a study to determine the role and regulation of expression of urokinase-type plasminogen activator in squamous-cell carcinoma of the oral cavity. The contribution of urokinase to the invasive process was clearly demonstrated in experiments in which in vitro invasion by a cultured squamous-cell carcinoma cell line was substantially reduced by a monoclonal antibody directed at the catalytic site of urokinase. The conditioned medium from 2 invasive cell lines contained high levels of urokinase. Examination of resected tumors for urokinase revealed (a) localization of the antigen to the tumor cells and (b) higher levels of the plasminogen activator in tumor tissue than in adjacent non-malignant tissue. These results suggested that elevated expression of urokinase is a common feature of this malignancy. The mRNA half-lives of cell lines expressing high and low levels of urokinase were similar, indicating that elevated levels of mRNA for the plasminogen activator were not a consequence of a more stable transcript. No evidence of u-PA gene amplification was obtained by Southern blotting of DNA derived from the cell lines expressing high urokinase levels. Transfection of squamous-cell carcinoma cells, which express high levels of urokinase, with a construct bearing the chloramphenicol acetyl transferase gene driven by the full-length (2345 bp) urokinase promoter indicated activation of the urokinase promoter in trans. In conclusion, our results suggest that transcriptional activation of the urokinase gene, in squamous-cell carcinomas of the oral cavity, leads to elevated levels of urokinase mRNA/protein which, in turn, can promote the invasive phenotype.

Transcriptional induction of the urokinase receptor gene by a constitutively active Src. Requirement of an upstream motif (-152/-135) bound with Sp1.

Since c-src overexpression increases colonic cell invasiveness and because both Src activity and urokinase receptor protein are elevated in invasive colon cancers, the present study was undertaken: 1) to determine if a constitutively active Src regulates urokinase receptor expression and 2) to identify required cis-elements and trans-acting factors. SW480 colon cancer cells transfected with an expression plasmid (c-srcY527F) encoding a constitutively active Src protein manifested increased urokinase receptor gene expression and Src activity. Treatment of the src transfectants with a Src-inhibitor (PD173955) reduced urokinase receptor protein levels and laminin degradation. Inasmuch as we recently implicated an upstream region of the urokinase receptor promoter (-152/-135) in constitutive urokinase receptor expression, we determined its role for the induction by src. Whereas the activity of a CAT reporter driven by this region was stimulated by c-srcY527F, the u-PAR promoter mutated at the Sp1-binding motif in the -152/-135 region was not. Nuclear extracts from the src transfectants demonstrated increased Sp1 binding to region -152/-135 compared with those from SW480 cells. Finally, endogenous urokinase receptor protein amounts in 10 colon cancers and corresponding normal colon correlated with Src specific activity. These data suggest that urokinase receptor gene expression is regulated by Src partly via increased Sp1 binding.