The down-regulated in adenoma (DRA) gene is significantly down-regulated in adenomas and adenocarcinomas of the colon as well as in colon cancer cell lines. It is also mutated in the disease congenital chloride diarrhea, which is characterized by loss of chloride transport and diarrhea. We now show a second function for DRA relevant to colon tumorigenesis, i.e., growth suppression. Transfection of full-length DRA into various cell lines (DLD-1, HT-29, HCT-15, SW837, SW480, MCF-7, NIH3T3, CaSki, and HeLa) that lack endogenous DRA expression results in a reduced number of drug-resistant colonies compared with vector control, suggesting growth suppression by DRA. In addition, expression of DRA under the control of an inducible promoter reduced the growth rate of DLD-1 cells compared with cells not expressing DRA. The COOH-terminal cytoplasmic domain of DRA is required for growth suppression, but an in-frame deletion (ΔVal317) that causes congenital chloride diarrhea and results in a loss of anion transport had no effect on growth suppression, indicating that anion transport and growth suppression are independent functions of DRA. One cell line, adenovirus-transformed HEK293, exhibited significant resistance to DRA-induced growth suppression, whereas the human papillomavirus-transformed cell lines, CaSki and HeLa, did not. E1A is an adenoviral protein required to transform HEK293 cells. DLD-1 cells that stably express 12S E1A are resistant to growth suppression by DRA, similar to HEK293 cells.

Cancers of the colon and rectum will account for ∼148,000 new cases in 2002 (1). Colorectal cancer is the third leading cause of mortality from cancer for both men and women, following only lung and breast or prostate cancers. The average person has a 1 in 18 lifetime chance of contracting colon cancer (2).

Vogelstein and colleagues (3, 4, 5) have proposed a model outlining the genetic events responsible for colon cancer. Mutations in these genes can be inherited or arise spontaneously through somatic mutations. Among the genes playing a prominent role in colorectal cancer are the tumor suppressor genes APC, p53, and DCC and oncogenes such as K-ras and c-myc. Other types of genes known to play a role in colon cancer are those responsible for DNA repair, such as hMSH2 and hMLH1(3, 6). Although mutations of these genes are often the cause for the development of neoplasms, epigenetic changes such as hypermethylation of genes also occur frequently in colon cancers (2, 7, 8).

We have previously described a gene, DRA,5 that is transcriptionally down-regulated early in the neoplastic process. DRA was found by subtractive hybridization to be underexpressed or absent in colon cancer samples compared with their normal counterparts (9, 10, 11). Expression of DRA is primarily on the apical surface of the columnar epithelial cells of the normal colon. Other areas of expression include small intestine, intraprostatic seminal vesicle (12), proximal tubule of the kidney, and eccrine sweat gland (13).

DRA shares significant homology with a family of sulfate transporters. Functional assays in both Xenopus laevis oocytes and Sf9 insect cells confirmed that DRA is capable of facilitating transport of sulfate, chloride, and oxalate in vitro(14, 15, 16, 17, 18). Mutations in DRA were found to be responsible for a rare disease known as CLD (19). This disease presents as early as in utero as voluminous, watery diarrhea with an extremely high chloride content. This pathology is attributable to lack of chloride uptake in the colon, a major site of chloride uptake in the body (20).

Although the role of DRA in CLD is readily interpreted, its role in colon cancer is not yet clear. The observed underexpression or loss of DRA in colon cancer could be explained by one of two rationales. Down-regulation of DRA could simply be a result of the dedifferentiation process that occurs during cancer progression. Alternatively, down-regulation of DRA might play a role in the stepwise process by which colon cancer progresses. We present data in this report suggesting that loss of DRA does play an active role in colon cancer progression.

Cell Culture.

DLD-1 and HCT-15 cells were grown in RPMI 1640 supplemented with 10 and 20% fetal bovine serum, respectively. MCF-7, NIH3T3, HT29, SW837, SW480, CaSki, HeLa, and HEK293 cells were cultured in DMEM supplemented with 10% fetal bovine serum. Incubation was carried out at 37°C in a 5% CO2 atmosphere.

Colony Suppression Assay.

Transfections were by the Lipofectin, Lipofectamine (Life Technologies, Inc.), or FuGene 6 (Roche, Indianapolis, IN) manufacturers’ protocols. Typically, 2 μg of DNA were used to transfect one 35-mm well when cells were approximately 50–80% confluent. For stable selection, cells were trypsinized and distributed to two 100-mm culture dishes (10 and 90%) 48–72 h after transfection. Drug selection in G418 (400 μg/ml for DLD-1, HT29, HCT-15, and SW837; 600 μg/ml for MCF-7; 800 μg/ml for SW480, HeLa, and HEK293; 80 μg/ml for CaSki) continued for 2 weeks. Colony counts were carried out by first washing the plates once with PBS. They were fixed and stained with 0.2% crystal violet/20% methanol and then counted. DLD-1 cells were selected in 200 μg/ml zeocin for stable E1A expression.

Expression Vectors.

The pSGneo vector was constructed by adding the bacterial neomycin resistance gene cassette to the mammalian expression vector, pSG5 (Stratagene, La Jolla, CA).) An AatII-DraIII fragment from pSG5 was removed and replaced with an AatII-DraIII fragment from the pOP13CAT vector that contained the neomycin resistance gene.

The following deletion mutants of DRA cDNA in the pSGneo vector were constructed: (a) DraΔ606–764 was made by removing an EarI-BglII fragment from the pSGneoDRA construct, filling in the 3′ recessed DNA ends with Klenow fragment and blunt religation of the plasmid. This removes the codons for amino acids 606–764 (end) and adds 16 amino acids (Ile-Leu-Leu-Lys-Gln-Asn-Leu-Phe-Ile-Ala-Ala-Tyr-Asn-Gly-Tyr-Lys-STOP) encoded by the vector; (b) DraΔV317 was made by mutagenic PCR (21) using the mutagenic primer pTAAAAACAGGTTTAAAGTGGCTGTGGGGGACATG (5′ phosphorylated with T4 polynucleotide kinase; New England BioLabs, Beverly MA), along with the exterior primer pair 5′-CATTTGCTCTGCTGGTCGACATTC (forward) and 5′-CAGTCCCAAGGCTATTAACTCCTG (reverse). PCR was performed with 10 ng of template DNA using a two-step protocol (94°C for 5 min, followed by 30 cycles of 94°C for 30 s, 65°C for 5 min, and then 65°C for 15 min) using Deep Vent DNA polymerase (New England BioLabs, Beverly, MA). The PCR product was digested with PflMI, purified by agarose gel electrophoresis, and cloned into the two PflMI sites within DRA cDNA in the pSGneo vector. The mutation was verified by direct sequence analysis.

For tetracycline-repressible expression (22) in DLD-1 cells, DRA cDNA was cloned into the BglII site of the vector pUHD10-3hygro (pUHD10-3 containing a hygromycin resistance gene) to generate the plasmid pUHD10-3hygroDRA. This construct was transfected into a DLD-1 clone that had previously been stably transfected with the vector pUHD15-1neo (pUHD15-1 containing a neomycin/G418 resistance gene), which constitutively expresses the tetracycline repressor. Stable clones were selected in 200 μg/ml G418 and 200 μg/ml hygromycin and maintained in the presence of 1 μg/ml tetracycline to keep the DRA cDNA from being expressed. Individual clones were tested for DRA mRNA expression upon removal of tetracycline. Most clones showed a dramatic increase in DRA mRNA levels.

E1A in pSGzeo was constructed by first removing the neomycin resistance gene from pSGneo with a RsrII-MluI digest. The zeocin resistance cassette was removed from pcDNA3.1/Zeo (Invitrogen, Carlsbad, CA) with a BstYI digest and isolation of the 556-bp fragment. DNA ends were filled in with Klenow enzyme, and blunt-end ligation was carried out. 12S E1A (a gift from David Kurtz, Medical University of South Carolina) was removed from pBluescriptKS with a HindIII/ApaI digest and inserted into the HindIII/ApaI site of pSGzeo.

Sulfate Transport Assay.

Sulfate transport assays were conducted as described previously with some modifications (23). At confluency, cells were rinsed twice with low-ionic strength buffer (1 mm MgCl2, 300 mm sucrose, 10 mm Tris-HEPES, pH 7.5) containing various concentrations of Na2SO4 (0, 50, 100, and 400 μm SO4−2). They were next incubated for 10 min at 37°C in 1 ml of buffer with various levels of unlabeled Na2SO4 as above plus 0.13 μm of carrier-free [35S]sulfate (Amersham Pharmacia, Piscataway, NJ). Cells were then washed twice with 3.5 ml of ice-cold wash solution (100 mm sucrose, 100 mm NaNO3, 1 mm MgCl2, and 10 mm Tris-HEPES, pH 7.5) and lysed in 1 ml of NETN [0.5% NP40, 20 mm Tris (pH 8.0), 100 mm NaCl, and 1 mm EDTA]. Lysates were subjected to scintillation counting, and protein concentration was analyzed by the BCA method (Pierce, Rockford, IL). DIDS sensitivity was assessed using 1 mm DIDS.

Immunofluorescence.

Fluorescence immunohistochemical staining was performed with the polyclonal antibody I20F to the COOH terminus of DRA as described previously (10).

Growth Curves.

For growth curve analysis, ∼300,000 cells were seeded into 35-mm wells in the presence or absence of 1 μg/ml tetracycline. Cells were fed every day. Each day for 7 days, cells were trypsinized and counted. Growth assays were performed in triplicate.

SDS-PAGE and Western Blot.

Twenty μg of total cell protein/lane were resolved by SDS-PAGE on 10 or 12% polyacrylamide gels. Proteins were then transferred to polyvinylidene difluoride or nitrocellulose membrane. Membranes were blocked in 5% milk in TBS with 0.1% Tween 20 (TBST) for 30 min. Incubation with DRA and E1A (Transduction Laboratories, Lexington, KY) primary antibodies diluted to 1 μg/ml in blocking solution was carried out at room temperature for 1 h, followed by three 5-min washes with TBST. Secondary antibodies (goat-antimouse-horseradish peroxidase for E1A and goat-antirabbit-horseradish peroxidase for DRA; DAKO Corp., Carpinteria, CA) were diluted 1:1000 in blocking buffer and incubated with membranes for 1 h at room temperature. Detection with ECL Plus (Amersham Pharmacia), followed according to the manufacturer’s protocol.

Expression of Full-Length DRA in Mammalian Cell Lines.

To delineate the function of DRA, we transfected both colon and non-colon cell lines (DLD-1, HT-29, HCT-15, SW837, SW480, MCF-7, and NIH3T3) with a selectable mammalian expression vector containing full-length DRA. Before transfection, each cell line was confirmed by Northern blot to lack endogenous DRA expression (data not shown). After a 2-week drug selection, G418-resistant colonies were counted. Fig. 1 shows the percentage of colonies obtained from the DRA transfection relative to the number obtained with vector control in the colony suppression assay. In each case, the DRA transfection produced significantly fewer stable transfectants than the vector control. Nonetheless, some G418-resistant colonies from the DRA transfection would usually appear. Some of these colonies were expanded and screened for DRA expression, but none were found to express DRA protein, suggesting that the colonies were simply G418 resistant. This phenomenon of reduced colony numbers was seen with colon cancer cell lines (DLD-1, HT-29, HCT-15, SW837, and SW480), a breast cancer cell line (MCF-7), and immortalized mouse fibroblasts (NIH3T3), suggesting a role for DRA in growth control.

Inducible Expression of DRA in DLD-1 Cells.

To further investigate the possibility that DRA played a role in growth control and because we could not obtain stable expression of DRA, we made an inducible construct, pUHD10-3hygroDRA, to be used in the tetracycline-repressible system. This construct was transfected into a DLD-1 cell line that stably expressed the tetracycline repressor plasmid. Independent clones were picked and expanded. The effect of DRA expression on the growth rate of this clone is shown in Fig. 2. The inset shows immunofluorescence of a clone that expresses DRA protein upon removal of tetracycline. DRA protein expression caused a significant decrease in the growth rate of the clone when compared with the same clone grown in tetracycline (i.e., no DRA expression.) These data provided further evidence that DRA may act as a growth suppressor. The growth rate remained fairly constant between days 2 and 6 when the DLD-1 cells approached confluence. Expression of DRA caused growth suppression without causing cell death, and restoration of tetracycline resulted in resumed growth (data not shown).

Expression of DRA Mutants in Mammalian Cell Lines.

Several deletion mutants were constructed in an attempt to map the functional region(s) of DRA responsible for growth suppression. The mutant DRAΔ606–764 removes the last 158 amino acids from the protein. The mutant DRAΔV317 is a disease-causing mutation in CLD that has been shown to prevent anion transport (16, 19). The growth-suppression capabilities of these mutants were assessed in a colony-suppression assay using DLD-1 cells as described for Fig. 1. Fig. 3 shows that the DRAΔ606–764 mutant gave colony counts similar to those for vector control, as did the mutant containing only the COOH-terminal amino acids 606–764. This suggests that the COOH-terminal tail of DRA is required but not sufficient for growth suppression. In contrast, the DRAΔV317 mutant gave colony counts similar to those obtained for full-length DRA, suggesting that the growth control and anion transport functions of DRA are independent. Similar results were obtained when the transfections were performed in NIH3T3 cells (data not shown).

Expression of DRA in HEK293 Cells.

In a paper reporting that DRA was responsible for Cl/HCO3 exchange, Melvin et al.(17) also showed stable expression of mouse Dra in HEK293 cells. Therefore, we determined whether human DRA could also be stably expressed in HEK293 cells. Fig. 4 shows stable expression of DRA protein in HEK293 cells by Western blot (A) and immunofluorescence (B). To demonstrate that the DRA protein expressed in the HEK293 cells was also functional, we performed a sulfate transport assay on a stable clone (DRA #11). Fig. 4 D shows that the clone expressing DRA exhibited DIDS-sensitive sulfate transport, whereas HEK293 cells do not show sulfate transport.

Colony Suppression by DRA in CaSki and HeLa Cells.

Constitutive, stable expression of DRA had previously been unattainable in our laboratory using various cancer-derived cell lines. HEK293 cells are of human embryonic kidney origin and have been transformed by introduction of sheared genomic DNA from adenovirus type 5. This was the first virally transformed cell line into which we had attempted to express DRA. Therefore, we wanted to assess whether we could express DRA in other virally transformed cell lines. For this, we chose two cell lines transformed by HPV, CaSki (HPV-16) and HeLa (HPV-18). Fig. 5 shows colony-suppression experiments using CaSki, HeLa, and HEK293 cells. CaSki and HeLa cells exhibited levels of DRA-induced colony suppression consistent with the previous cell lines tested. HEK293 cells, however, exhibited a decreased level of DRA-induced growth suppression (∼60% in HEK293 cells versus 10–30% in other cell lines). Therefore, partial abrogation of growth suppression by DRA was observed in HEK293 cells and suggested that one or both of the adenovirus-transforming proteins (E1A and E1B) may be involved.

Adenovirus Type 2 12S E1A Abrogates DRA-induced Growth Suppression in DLD-1 Cells.

HEK293 cells express two transforming proteins, E1A and E1B. To assess whether expression of the E1A protein contributed to the observed abrogation of DRA-induced growth suppression, we constructed stable cell lines expressing the 243-amino acid 12S E1A protein in DLD-1 cells. Expression of E1A was confirmed by Western blot analysis (Fig. 6,A). Nuclear localization of E1A was confirmed by immunofluorescence (data not shown). We then performed a DRA colony-suppression transfection assay. Fig. 6,B shows colony counts from two independent DLD-1/E1A clones. The number of G418-resistant colonies in DLD-1 cells expressing E1A is significantly higher than in parental DLD-1 cells (compare 10% in Fig. 1versus ∼50% in Fig. 6,B). Furthermore, these stable transfectants express the DRA protein in a Western blot assay (Fig. 6 B, inset). The addition of E1B expression to DLD-1 cells did not alter the effect attributable to E1A alone (data not shown). This suggests that the pathway by which DRA suppresses cellular growth may be interrupted by E1A expression.

The function of DRA was initially investigated by attempting stable transfections of full-length protein into cell lines that lacked endogenous expression. However, these experiments consistently resulted in fewer drug-resistant colonies from the DRA transfection than that of a vector control. Among the few colonies that survived G418 selection, Northern and/or Western blot analyses were performed. Some of these colonies did express DRA mRNA, but none were found to express protein (data not shown). Therefore, DRA protein expression apparently caused growth arrest (or death), preventing colony formation. This pattern of DRA-induced growth suppression was observed in cell lines of various origins and cell types. Several colon cancer cell lines (DLD-1, SW480, SW837, HT-29, and HCT-15), the breast cancer cell line MCF-7, as well as the nontransformed mouse fibroblast cell line NIH3T3 were all apparently growth suppressed upon transfection with full-length human DRA cDNA. These results suggest that growth suppression by DRA is a general phenomenon, rather than a characteristic of a specific cell line.

Inducible expression of DRA provided further evidence that it is a growth suppressor. Upon DRA induction (by tetracycline removal) in a stable transfectant, its growth rate was dramatically decreased as seen by cell counts over a 7-day period (see Fig. 2). This could conceivably be attributed to cell death rather than growth arrest. However, there were no obvious signs of cell toxicity or apoptosis. No increased amount of cell debris or floating cells indicative of cell death was observed. Also, no obvious signs of apoptosis, such as membrane blebbing, were observed. Furthermore, the cells survived several rounds of growth in tetracycline, followed by growth cessation in the absence of the tetracycline (data not shown.) This indicates that cell death was not a major factor contributing to the flat growth curve during DRA protein expression. Importantly, the observed growth suppression by DRA is precisely consistent with our previous report showing that the only cells in the colonic mucosa to express DRA are those that have already withdrawn from the cell cycle (10).

We designed several deletion mutants to map the functional regions of DRA. ΔV317 is a disease-causing mutation in CLD (19) confirmed to be defective in both chloride and sulfate transport (16). Results from the colony-suppression assay show that ΔV317 is still capable of inhibiting colony formation. This suggests that the anion transport and growth-suppression functions of DRA are independent. However, it does not rule out indirect effects on neoplasia attributable to loss of anion transport function. Indeed, patients with CLD attributable to the ΔV317 mutation appear to be more likely to develop colorectal cancer than the general population (24). One possible mechanism for this may be attributable to lower levels of sulfation of heparan sulfate proteoglycans. Heparan sulfate proteoglycans modulate a number of receptor tyrosine kinase signaling pathways (including those pathways involving the ligands FGF, transforming growth factor β, Wnt/Wingless, and Hedgehog), the aberrant regulation of which contributes to neoplasia (reviewed in Ref. 25). A Drosophila mutant, sulfateless (sfl), which completely lacks sulfated disaccharides from heparan sulfate (26), is defective in its FGF signaling but can be rescued by a constitutively active form of an FGF receptor (27). This suggests that proper sulfation is obligatory for FGF signaling at the ligand/receptor level. In human colon cancer cell lines, altered patterns of sulfation and decreased overall sulfation levels of heparan sulfate occur in carcinoma versus adenoma (28). These changes result in a 10-fold decrease in the affinity of carcinoma-derived heparan sulfate for basic FGF and correlate with a reduced biological response to basic FGF in the carcinoma cells (29).

The mutant that removes the COOH terminus of DRA (Δ606–764) failed to suppress growth in the colony-suppression assay. The current structural model of DRA indicates that this mutation removes at least one transmembrane domain from DRA as well as the entire cytoplasmic tail. This may indicate that the COOH terminus of DRA contains one or more functional region(s) required for growth suppression. A construct containing only the amino acids removed in this mutant (DRA606–764) was tested in a colony-suppression assay. The COOH terminus alone did not suppress growth (see Fig. 3). This indicates that, although the COOH terminus is necessary to facilitate growth control, it is not sufficient to do so. It suggests that interaction with other parts of the DRA protein may also be required for growth suppression, perhaps involving some of the transmembrane architecture. Indeed, the transmembrane architecture of pendrin, a protein most closely related to DRA, cannot substitute for DRA in a colony-suppression assay.6

While some of the experiments described here were in progress, an intriguing report appeared describing the stable expression of mouse Dra in HEK293 cells (17). We were able to repeat this result using our human DRA cDNA. Both DRA protein expression and DRA protein function could be demonstrated in stably transfected HEK293 cells (see Fig. 4). Immunofluorescence of one of these clones (see Fig. 4) revealed strong staining in the Golgi in addition to the membrane where DRA is normally expressed. The Golgi localization is consistent with earlier transient transfections performed in COS-1 cells (10) and most likely represents the nascent polypeptide undergoing glycosylation. However, the demonstration of sulfate transport indicates that the DRA localized to the membrane was functionally active. We conclude that the observed reduction in DRA-induced growth suppression is not likely attributable to misexpression or a nonfunctional protein.

The demonstration of stable DRA expression in HEK293 cells causes us to consider possible explanations for the partial abrogation of growth suppression. HEK293 cells are human embryonic kidney cells that were transformed by introduction of adenovirus type 5 DNA. Specifically, this cell line expresses two transforming proteins, E1A and E1B. To determine whether the reduction of DRA-induced growth suppression was attributable to viral transformation, we transfected DRA into two other virally transformed cell lines, CaSki and HeLa. CaSki and HeLa cells are transformed by the HPV types 16 and 18, respectively. Both cell lines express two HPV-transforming proteins, E7 and E6. Although these proteins transform cells by a mechanism similar to E1A and E1B, respectively, we did not observe a reduction in DRA-induced growth suppression in CaSki and HeLa cells. It is unclear whether the failure to partially abrogate DRA-induced growth suppression in CaSki and HeLa cells is attributable to subtle differences in the mechanisms of the transforming proteins involved or may be attributable to cell line differences independent of the transforming proteins. However, we could demonstrate that the partial abrogation of DRA-induced colony suppression by E1A could be duplicated in DLD-1 cells, a cell line susceptible to growth control by DRA. This indicates that the 243-amino acid 12S E1A product is capable of mediating the abrogation.

There are numerous E1A-mediated effects resulting in cellular transformation (reviewed in Ref. 30). Because E1A is an early viral gene product, it is responsible for trans-activation of later viral genes, as well as controlling host cell growth upon infection. The latter is accomplished by targeted disruption of the host cell’s growth control machinery. This involves activation of positive growth regulators and inactivation of negative regulators. The most notable effect of E1A expression is the binding and effective inactivation of the retinoblastoma tumor suppressor protein, pRb. pRb negatively regulates cell growth by binding to the transcription factor E2F, which is responsible for transcription of genes involved in cell cycle progression (31). E1A disrupts this interaction by direct binding to pRb, thus freeing E2F and leading to unchecked cell cycle progression. Although one might speculate that DRA controls growth through an Rb pathway, this begs the question of why abrogation was not observed in CaSki and HeLa cells, where E7 ostensibly blocks pRb function in a manner similar to E1A. Furthermore, if DRA failed to suppress growth in HEK293 cells because of E1A’s disruption of pRb function, it would follow that cell lines that do undergo DRA-induced growth suppression might maintain functional (hypophosphorylated) pRb. However, at least for some cell lines, this is not the case. For example, DLD-1 and SW480 cells have been reported to lack the hypophosphorylated form of pRb (32). Presumably, with no hypophosphorylated pRb, E2F is constitutively unbound and free to transactivate positive cell cycle regulators. In addition, DLD-1, SW480, and HT-29 cell lines lack p16 expression and exhibit high levels of type D cyclins (32). These conditions result in cyclin-dependent kinase 4-mediated phosphorylation of Rb, thus inactivating it. Therefore, a role for pRb is unclear at this time.

It has been reported that the other Rb family members, p130/Rb2 and p107, have an aberrant phosphorylation pattern in HEK293 cells. This is a direct result of E1A expression (33). E1A disrupts the interaction between p130 or p107 and their respective E2F family members while preventing their hyperphosphorylation. The phosphorylation of pRb, however, is not blocked by E1A expression. For this reason, it is believed that p130/Rb2 and p107 may have roles in addition to those defined for pRb. There is some evidence that p130/Rb2 can act as a cyclin-dependent kinase inhibitor (34). One could envision a scenario wherein DRA depends on p130/Rb2 function as a cyclin-dependent kinase inhibitor to effectively inhibit growth. If E1A expression in DLD-1 cells interferes with this function of p130/Rb2, then a reduction in DRA-induced growth suppression might be expected. However, this explanation for the observed reduction is confounded by an important observation; E1A blockage of p130/Rb2 hyperphosphorylation occurs in HeLa cells (33), a cell line that we confirmed to be susceptible to DRA-induced growth suppression. Therefore, an E1A abrogation mechanism mediated through a pathway involving Rb family members is either a much more complicated matter or does not apply to DRA-induced growth suppression.

Other E1A-mediated events with widespread effects on the cell cycle exist. This leaves more avenues to be investigated to uncover the explanation for the observed reduction in DRA-induced growth suppression observed in HEK293 cells and in DLD-1 cells expressing E1A. Direct interaction between DRA and E1A is unlikely because DRA is a membrane protein and E1A functions in the nucleus. This would rule out E1A possibly binding to and masking functional regions of the COOH terminus of DRA. Alternatively, E1A might inhibit expression of a downstream effector of DRA through its trans-activation region. This could be assessed by mutation or deletion of this region in an E1A construct, followed by expression in DLD-1 cells and “challenge” by DRA transfection. If DRA-induced growth suppression was restored in the absence of E1A trans-activation, then proteins induced or inhibited by E1A expression would be candidates for members of the pathway by which DRA controls cell growth. Although the E1A-mediated abrogation of DRA-induced growth suppression cannot yet be definitively explained, it will hopefully provide clues to the mechanism of this protein with such diverse roles as anion transporter and growth suppressor.

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

Supported by Grant RPG-96-098-03-CSM from the American Cancer Society and Department of Defense Contract N00014-96-1-1298 (to C. W. S.).

5

The abbreviations used are: DRA, down-regulated in adenoma; CLD, congenital chloride diarrhea; DIDS, 4,4′-diisothiocyano-2,2′-disulfonic acid stilbene; HPV, human papillomavirus; FGF, fibroblast growth factor; pRb, retinoblastoma protein.

6

J. M. Chapman and C. W. Schweinfest, unpublished results.

Fig. 1.

Transfection of full-length DRA into various cell lines. The indicated cell lines were transfected with 2 μg of DRA DNA and subsequently cultured in the presence of G418 for 2 weeks. G418-resistant colonies were fixed, stained, and counted. Columns, means; bars, SD.

Fig. 1.

Transfection of full-length DRA into various cell lines. The indicated cell lines were transfected with 2 μg of DRA DNA and subsequently cultured in the presence of G418 for 2 weeks. G418-resistant colonies were fixed, stained, and counted. Columns, means; bars, SD.

Close modal
Fig. 2.

Effect of inducible DRA expression on cellular growth. A DLD-1 transfectant expressing an inducible DRA plasmid (pUHD10-3hygroDRA) and the tetracycline repressor was grown in the presence (▴) or absence (•) of tetracycline to measure growth rate. Growth curve data represent triplicates from one experiment; bars, SD. Inset, immunofluorescence (IMF) staining of DLD-1/DRA clone expressing pUHD10-3hygroDRA and the tetracycline repressor using a polyclonal DRA antibody in the presence (+Tet) or absence (−Tet) of tetracycline. Phase-contrast (P-C) of the field is shown.

Fig. 2.

Effect of inducible DRA expression on cellular growth. A DLD-1 transfectant expressing an inducible DRA plasmid (pUHD10-3hygroDRA) and the tetracycline repressor was grown in the presence (▴) or absence (•) of tetracycline to measure growth rate. Growth curve data represent triplicates from one experiment; bars, SD. Inset, immunofluorescence (IMF) staining of DLD-1/DRA clone expressing pUHD10-3hygroDRA and the tetracycline repressor using a polyclonal DRA antibody in the presence (+Tet) or absence (−Tet) of tetracycline. Phase-contrast (P-C) of the field is shown.

Close modal
Fig. 3.

Transfections of DRA mutants into DLD-1 cells. Cells were transfected with 2 μg of the indicated DRA mutants. G418-resistant colonies were counted after a 2-week selection in G418. Columns, means; bars, SD.

Fig. 3.

Transfections of DRA mutants into DLD-1 cells. Cells were transfected with 2 μg of the indicated DRA mutants. G418-resistant colonies were counted after a 2-week selection in G418. Columns, means; bars, SD.

Close modal
Fig. 4.

DRA expression in HEK293 cells. HEK293 cells were transfected with 2 μg of DRA cDNA. Stable clones were picked and expanded after a 2-week G418 selection. A, Western blot detection of DRA protein in stably transfected HEK293 cells. Lanes 1–4 contain 20 μg of protein from clones 293/DRA#6, 293/DRA#11, 293/DRA#14, and untransfected HEK293 cells, respectively. B, HEK293-DRA clone #11 was fixed with 4% formaldehyde in PBS, and immunofluorescence was performed using DRA I20F polyclonal antibody. C, immunofluorescence of parental HEK293 cells (as a negative control) with DRA I20F polyclonal antibody was performed as in B. D, a sulfate transport assay of HEK293-DRA clone #11 and parental HEK293 cells was performed as described in “Materials and Methods.”

Fig. 4.

DRA expression in HEK293 cells. HEK293 cells were transfected with 2 μg of DRA cDNA. Stable clones were picked and expanded after a 2-week G418 selection. A, Western blot detection of DRA protein in stably transfected HEK293 cells. Lanes 1–4 contain 20 μg of protein from clones 293/DRA#6, 293/DRA#11, 293/DRA#14, and untransfected HEK293 cells, respectively. B, HEK293-DRA clone #11 was fixed with 4% formaldehyde in PBS, and immunofluorescence was performed using DRA I20F polyclonal antibody. C, immunofluorescence of parental HEK293 cells (as a negative control) with DRA I20F polyclonal antibody was performed as in B. D, a sulfate transport assay of HEK293-DRA clone #11 and parental HEK293 cells was performed as described in “Materials and Methods.”

Close modal
Fig. 5.

Transfections of DRA into virally transformed cell lines. Cells were transfected with 2 μg of DRA DNA and selected in G418 for 2 weeks. After selection, colonies were fixed, stained, and counted. Columns, means; bars, SD.

Fig. 5.

Transfections of DRA into virally transformed cell lines. Cells were transfected with 2 μg of DRA DNA and selected in G418 for 2 weeks. After selection, colonies were fixed, stained, and counted. Columns, means; bars, SD.

Close modal
Fig. 6.

E1A abrogation of DRA-induced growth suppression. DLD-1 cells were transfected with 2 μg of adenovirus type 2 E1A 12S DNA and selected for 2 weeks in zeocin. Individual colonies were picked and expanded. A, Western blot showing stable E1A protein expression in DLD-1 cells. B, DLD-1/E1A clones #1 and #4 were transfected with 2 μg of DRA cDNA and selected for 2 weeks in G418. Colonies were fixed, stained, and counted or picked for expansion. Columns, means; bars, SD. Inset, stable DRA expression in two clones (E1A1 and E1A4) obtained from this transfection, negative controls of DLD-1/E1A cells and DLD-1 cells and a positive control of HEK293 cells that express DRA.

Fig. 6.

E1A abrogation of DRA-induced growth suppression. DLD-1 cells were transfected with 2 μg of adenovirus type 2 E1A 12S DNA and selected for 2 weeks in zeocin. Individual colonies were picked and expanded. A, Western blot showing stable E1A protein expression in DLD-1 cells. B, DLD-1/E1A clones #1 and #4 were transfected with 2 μg of DRA cDNA and selected for 2 weeks in G418. Colonies were fixed, stained, and counted or picked for expansion. Columns, means; bars, SD. Inset, stable DRA expression in two clones (E1A1 and E1A4) obtained from this transfection, negative controls of DLD-1/E1A cells and DLD-1 cells and a positive control of HEK293 cells that express DRA.

Close modal
1
Jemal A., Thomas A., Murray T., Thun M. Cancer statistics, 2002.
CA Cancer. J. Clin.
,
52
:
23
-47,  
2002
.
2
Ahuja N., Li Q., Mohan A. L., Baylin S. B., Issa J. P. Aging and DNA methylation in colorectal mucosa and cancer.
Cancer Res.
,
58
:
5489
-5494,  
1998
.
3
Kinzler K. W., Vogelstein B. Landscaping the cancer terrain [see comments].
Science (Wash. DC)
,
280
:
1036
-1037,  
1998
.
4
Kinzler K. W., Vogelstein B. Lessons from hereditary colorectal cancer.
Cell
,
87
:
159
-170,  
1996
.
5
Fearon E. R., Vogelstein B. A genetic model for colorectal tumorigenesis.
Cell
,
61
:
759
-767,  
1990
.
6
Jiricny J., Nyström-Lahti M. Mismatch repair defects in cancer.
Curr. Opin. Genet. Dev.
,
10
:
157
-161,  
2000
.
7
Herman J. G., Merlo A., Mao L., Lapidus R. G., Issa J. P., Davidson N. E., Sidransky D., Baylin S. B. Inactivation of the CDKN2/p16/MTS1 gene is frequently associated with aberrant DNA methylation in all common human cancers.
Cancer Res.
,
55
:
4525
-4530,  
1995
.
8
Gonzalez-Zulueta M., Bender C. M., Yang A. S., Nguyen T., Beart R. W., Van Tornout J. M., Jones P. A. Methylation of the 5′ CpG island of the p16/CDKN2 tumor suppressor gene in normal and transformed human tissues correlates with gene silencing.
Cancer Res.
,
55
:
4531
-4535,  
1995
.
9
Schweinfest C. W., Henderson K. W., Suster S., Kondoh N., Papas T. S. Identification of a colon mucosa gene that is down-regulated in colon adenomas and adenocarcinomas.
Proc. Natl. Acad. Sci. USA
,
90
:
4166
-4170,  
1993
.
10
Byeon M. K., Westerman M. A., Maroulakou I. G., Henderson K. W., Suster S., Zhang X-K., Papas T. S., Vesely J., Willingham M. C., Green J. E., Schweinfest C. W. The down-regulated in adenoma (DRA) gene encodes an intestine-specific membrane glycoprotein.
Oncogene
,
12
:
387
-396,  
1996
.
11
Antalis T. M., Reeder J. A., Gotley D. C., Byeon M. K., Walsh M. D., Henderson K. W., Papas T. S., Schweinfest C. W. Down-regulation of the down-regulated in adenoma (DRA) gene correlates with colon tumor progression.
Clin. Cancer Res.
,
4
:
1857
-1863,  
1998
.
12
Yang H., Jiang W., Furth E. E., Wen X., Katz J. P., Sellon R. K., Silberg D. G., Antalis T. M., Schweinfest C. W., Wu G. D. Intestinal inflammation reduces expression of DRA, a transporter responsible for congenital chloride diarrhea.
Am. J. Physiol.-Gastrointest. Liver Physiol.
,
275
:
G1445
-G1453,  
1998
.
13
Haila S., Saarialho-Kere U., Karjalainen-Lindsberg M. L., Lohi H., Airola K., Holmberg C., Hästbacka J., Kere J., Höglund P. The congenital chloride diarrhea gene is expressed in seminal vesicle, sweat gland, inflammatory colon epithelium, and in some dysplastic colon cells.
Histochem. Cell Biol.
,
113
:
279
-286,  
2000
.
14
Byeon M. K., Frankel A., Papas T. S., Henderson K. W., Schweinfest C. W. Human DRA functions as a sulfate transporter in Sf9 insect cells.
Protein Expression Purif.
,
12
:
67
-74,  
1998
.
15
Silberg D. G., Wang W., Moseley R. H., Traber P. G. The down regulated in adenoma (DRA) gene encodes an intestine-specific membrane sulfate transport protein.
J. Biol. Chem.
,
270
:
11897
-11902,  
1995
.
16
Moseley R. H., Höglund P., Wu G. D., Silberg D. G., Haila S., de la Chapelle A., Holmberg C., Kere J. Downregulated in adenoma gene encodes a chloride transporter defective in congenital chloride diarrhea.
Am. J. Physiol.-Gastrointest. Liver Physiol.
,
39
:
G185
-G192,  
1999
.
17
Melvin J. E., Park K., Richardson L., Schultheis P. J., Shull G. E. Mouse down-regulated in adenoma (DRA) is an intestinal Cl/HCO3 exchanger and is up-regulated in colon of mice lacking NHE3 Na+/H+ exchanger.
J. Biol. Chem.
,
274
:
22855
-22861,  
1999
.
18
Rajendran V. M., Black J., Ardito T. A., Sangan P., Alper S. L., Schweinfest C., Kashgarian M., Binder H. J. Regulation of DRA and AE1 in rat colon by dietary Na depletion.
Am. J. Physiol. Gastrointest. Liver Physiol.
,
279
:
G931
-G942,  
2000
.
19
Höglund P., Haila S., Socha J., Tomaszewski L., Saarialho-Kere U., Karjalainen-Lindsberg M-L., Airola K., Holmberg C., de la Chapelle A., Kere J. Mutations of the down-regulated in adenoma (DRA) gene cause congenital diarrhoea.
Nat. Genet.
,
14
:
316
-319,  
1996
.
20
Kere J., Lohi H., Höglund P. Genetic disorders of membrane transport III. Congenital chloride diarrhea.
Am. J. Physiol.
,
276
:
G7
-G13,  
1999
.
21
Michael S. F. Mutagenesis by incorporation of a phosphorylated oligo during PCR amplification.
BioTechniques
,
16
:
410
-412,  
1994
.
22
Gossen M., Bujard H. Tight control of gene expression in mammalian cells by tetracycline-responsive promoters.
Proc. Natl. Acad. Sci. USA
,
89
:
5547
-5551,  
1992
.
23
Hästbacka J., de la Chapelle A., Mahtani M. M., Clines G., Reeve-Daly M. P., Daly M., Hamilton B. A., Kusumi K., Trivedi B., Weaver A., Coloma A., Lovett M., Buckler A., Kaitila I., Lander E. S. The diastrophic dysplasia gene encodes a novel sulfate transporter: positional cloning by fine-structure linkage disequilibrium mapping.
Cell
,
78
:
1073
-1087,  
1994
.
24
Hemminki A., Höglund P., Pukkala E., Salovaara R., Järvinen H., Norio R., Aaltonen L. A. Intestinal cancer in patients with a germline mutation in the down-regulated in adenoma (DRA) gene.
Oncogene
,
16
:
681
-684,  
1998
.
25
Selva E. M., Perrimon N. Role of heparan sulfate proteoglycans in cell signaling and cancer.
Adv. Cancer Res.
,
83
:
67
-80,  
2001
.
26
Toyoda H., Kinoshita-Toyoda A., Fox B., Selleck S. B. Structural analysis of glycosaminoglycans in animals bearing mutations in sugarless, sulfateless, and tout-velu. Drosophila homologues of vertebrate genes encoding glycosaminoglycan biosynthetic enzymes.
J. Biol. Chem.
,
275
:
21856
-21861,  
2000
.
27
Lin X., Buff E. M., Perrimon N., Michelson A. M. Heparan sulfate proteoglycans are essential for FGF receptor signaling during Drosophila embryonic development.
Development (Camb.)
,
126
:
3715
-3723,  
1999
.
28
Jayson G. C., Lyon M., Paraskeva C., Turnbull J. E., Deakin J. A., Gallagher J. T. Heparan sulfate undergoes specific structural changes during the progression from human colon adenoma to carcinoma in vitro.
J. Biol. Chem.
,
273
:
51
-57,  
1998
.
29
Jayson G. C., Vives C., Paraskeva C., Schofield K., Coutts J., Fleetwood A., Gallagher J. T. Coordinated modulation of the fibroblast growth factor dual receptor mechanism during transformation from human colon adenoma to carcinoma.
Int. J. Cancer
,
82
:
298
-304,  
1999
.
30
Gallimore P. H., Turnell A. S. Adenovirus E1A: remodelling the host cell, a life or death experience.
Oncogene
,
20
:
7824
-7835,  
2001
.
31
Sherr C. J. Cancer cell cycles.
Science (Wash. DC)
,
274
:
1672
-1677,  
1996
.
32
Tominaga O., Nita M. E., Nagawa H., Fujii S., Tsuruo T., Muto T. Expressions of cell cycle regulators in human colorectal cancer cell lines.
Jpn. J. Cancer Res.
,
88
:
855
-860,  
1997
.
33
Parreño M., Garriga J., Limón A., Mayol X., Beck G. R., Jr., Moran E., Graña X. E1A blocks hyperphosphorylation of p130 and p107 without affecting the phosphorylation status of the retinoblastoma protein.
J. Virol.
,
74
:
3166
-3176,  
2000
.
34
Woo M. S., Sánchez I., Dynlacht B. D. p130 and p107 use a conserved domain to inhibit cellular cyclin-dependent kinase activity.
Mol. Cell. Biol.
,
17
:
3566
-3579,  
1997
.