Molecular Cloning and Characterization of Human MAWD, a Novel Protein Containing WD-40 Repeats Frequently Overexpressed in Breast Cancer¹ Free

Accumulating evidence suggests that a cancer arises through accumulation of a number of genetic lesions. In addition, loss of sensitivity to negative growth regulators is thought to be an important step in the development of neoplastic lesions. TGF³-β is one of the most elucidated negative growth regulators of cells. However, the exact mechanism(s) for signaling from the receptors for TGF-β remain to be further elucidated. To date, several types of signaling molecules have been identified as downstream mediators of the TGF-β receptors. Among them, TRIP-1 has unique function and structure containing WD-40 motif (1).

The WD repeats proteins are characterized by a cluster of repeated sequences, each repeating around 40 amino acids in length and usually ending with tryptophan-aspartate (WD), hence the name. The WD-40 repeats are found in proteins involved in a wide variety of cellular processes ranging from signal transduction to RNA processing(2). Proteins containing WD repeats are often physically associated with other proteins and are believed in many cases to act as scaffolds upon which the multimeric complexes are built(3). The crystal structure of the heterotrimeric G protein revealed that the WD repeats of the subunit fold into a highly symmetrical, propeller-like structure with the conserved region of each repeat forming a part of the propeller blade reviewed by Neer and Schmidt (4).

These exciting findings prompted us to try to discover a new WD-40 repeats protein that might be involved in cell growth or cancer progression. In the present study, we report the cloning and sequencing of a full-length cDNA that codes for a novel WD-40 repeats protein similar to TRIP-1. As described later, we name the gene product hMAWD,standing for a putative human MAPK activator with WD repeats. The possible roles of the gene are also discussed.

To find a new member of the human WD-40 repeats protein, the WD-40 motifs of TRIP-1 were used to search homologues in gene data bases stored at the National Center for Biotechnology by using the tBLASTn sequence alignment program. w313 cDNA, which was identified by the similarity search, was amplified by RT-PCR procedure from the mRNA of human normal tissues.

A human HepG2 cDNA library constructed using lambda ZAP vector(Stratagene) was purchased and screened by using standard conditions. Briefly, 1 × 10⁶ clones in total were screened with the ECL-labeled cDNA fragment. Positive clones were selected, and their insert cDNAs were excised in vivo in pBluescript II, following the supplier’s recommendation. The longest cDNA thus constructed was sequenced to completion of both strands using the Thermo Sequenase Dye Termination Cycle Sequencing pre-mix kit(Amersham) on an A.L.F. DNA sequencer II (Pharmacia LKB). The search for the sequence analysis and the comparison to related sequences were carried out using tBLASTn program.

Chromosomal assignment of the hMAWD gene was accomplished by PCR analysis of genomic DNAs of the Genebridge 4 RH panel (Research Genetics, Huntsville, AL) as well as those of Stanford G3 RH panel(Research Genetics). The Ums (sense, 5′-TGAATTAGCTCCAGTG) and Dma(antisense, 5′-ATAACAGGCCACTGTA) PCR primers were designed specifically to amplify the hMAWD sequence in the rodent background genes. PCR consisted of 35 cycles (94°C for 40 s, 56°C for 40 s, and 72°C for 40 s) after the initial denaturation step (94°C for 8 min). The data for hMAWD of Genebridge 4 RH panel were as follows:111111111000111101011010000100001011111111000110111110010010-0111111110011101111101111.

COS-7 cells, ECV304 cells, and the derived cells were cultured in DMEM with 10% fetal bovine serum and antibiotics.

Cells (1 × 10⁴) were suspended in 2 ml of 0.4% agar in DMEM containing 10% FBS and overlayed on a layer of 0.8% agar in medium. After 10 days, the number of colonies <1.0 mm in diameter was counted.

Fresh surgical specimens of 46 patients diagnosed histologically as having breast cancer (46 cases of invasive ductal adenocarcinoma) and their adjacent normal tissues were obtained from patients undergoing surgery. The materials, obtained immediately after the surgical procedure, were frozen in liquid nitrogen and stored at −80°C.

For the purpose of producing specific antibodies, the partial hMAWD(amino acids 173–350) and the full-length hMAWD were expressed and purified as GST fusion proteins using bacterial expression vector pGEX5X-1 (Pharmacia Biotech, Inc.). Rabbits were immunized, and specific antibody was prepared using the fusion protein.

These methods were basically performed as described previously(5, 6). Then, the cells were inspected with a confocal scanning microscopy and standard microscopy.

Total cell lysates were prepared by suspending 100-mg samples in RIPA buffer [40 mm Tris-HCl (pH 7.5), 1% Triton X-100, and 150 mm NaCl]. The total cell lysate was resuspending in a standard Laemmli sample preparation buffer and loaded onto a 10%polyacrylamide gel. Proteins were immunoblotted to Immobilon filters,and the filters were incubated sequentially with anti-hMAWD polyclonal antibody, alkalyphosphatase-conjugated antirabbit IgG, and developed with the Promega NBT-BCIP detection system suggested in the manufacturer’s instructional protocol.

To determine whether there might be additional members of the human WD-40 repeat protein family lurking in one of the sequence databases, the tBLASTn program (7) was used to query each database with the amino acid sequences of the human TRIP-1 protein. The expressed sequence tag database revealed several clones, apparently from the same gene, that shared significant amino acid sequence similarities to the WD-40 repeat motif of TRIP-1. The 300-bp cDNA fragment was amplified by RT-PCR from a human placenta cDNA library and identified as a clone (w313) that has significant homology to TRIP-1. The w313 fragment was then used as a probe for the screening of a full-length cDNA against a human HepG2 cDNA library. Using a photo-labeled w313 probe, a total of four putative positive clones were isolated after screening about 1 × 10⁶ plaques of the cDNA library (8, 9). Sequence analysis indicated that all of these clones contain cDNAs from the same encoding a full-length corresponding to w313. The entire sequence of the clones encodes a 1.8-kb nucleotide sequence with a single ORF of 1050 nucleotides. A typical polyadenylation signal(AATAAA) is located downstream from the first in-frame stop codon(10). The ORF sequence of the clone predicts a protein of 350 amino acids with a calculated molecular weight of M_r 39,000. We applied computer analysis, and the deduced protein sequence had six repeats of WD-40 motif, most closely related to mouse STRAP with 97% amino acid identity over the entire sequence. STRAP is the recently identified novel protein that is associated with TGF-β receptors from mouse cDNA library by the two-hybrid system (11). From the alignment and the highly conserved amino acid sequences, the identified cDNA is supposed to be a human homologue of the mouse STRAP gene. To determine the expression patterns of hMAWD mRNA, Northern blot hybridizations and cycle-limited RT-PCR were carried out. It revealed that the hMAWD gene is ubiquitously expressed in all human tissues examined (data not shown). The hMAWD messages are about 2.0 kb,matching the size of the cloned cDNAs. These data are consistent with that of STRAP (11).

DNA from a panel of rodent-human hybrids, together carrying most human chromosome regions, was tested for presence of the hMAWDgene locus by PCR amplification (12; Fig. 1). Oligonucleotide primers Ums and Dma, corresponding to the 3′untranslated region of the hMAWD gene, were used to amplify specifically hMAWD but not its rodent homologue. To refine the localization of the gene, both of the Genebridge 4 RH panel and the Stanford Gene 3 panel were tested (12). The results of these data were equal and showed that the hMAWD locus lies near the region to the marker D12S1596 (LOD, 5.5). These findings led to the placement of the hMAWD gene at 12q11–12.

To generate cells overexpressing the cDNA for assessing biological effects and for evaluating the function of the hMAWD protein in cell growth, we engineered four different expression vectors containing the full-length untagged, His-tagged, Xpress-tagged, and green fluorescent protein-tagged cDNA in the sense in-frame orientation. These were first tested in transient transfection assays with COS-7 cells and were found to significantly express protein, compared with mock-transfected controls (data not shown). The intracellular localization of both endogenous and overexpressed hMAWD was mainly shown in cytoplasm (Fig. 2, A and B ). Then, to derive stable cell lines overexpressing the protein, we used the well-established human ECV304 cell line (13). After plasmid gene transfection by Lipofection, drug-resistant colonies were selected and screened. The third panel in Fig. 2,C shows a Western blot analysis of the expression in three independent ECV304 clones harboring Xpress-tagged hMAWD sense vector (termed P1, P2, and P3), vector only (C), and parental ECV304 cells (N). All of these cells overexpressing exogenous hMAWD were unexpectedly formed colonies in Petri dishes (data not shown) and exhibited growth in soft agar (Fig. 3, A and B). Src-transformed cells served as a positive control, and parental ECV304 cells and ECV304 cells harboring vector alone served as negative controls for these assays. These results suggest that targeted up-regulation of hMAWD protein expression disrupts contact inhibition and leads to anchorage-independent growth.

To approach the mechanisms by which overexpression of hMAWD promotes a loss of contact inhibition and anchorage-independent growth, we used a variety of phospho-specific antibodies that have been generated against the activated forms of well-known signal transducer molecules (data not shown). Using antibody detected against activated MAPK among them, we found that MAPK was constitutively activated in all three hMAWD-overexpressed clones (Fig. 3,C, second panel). Blotting with MAPK phospho-independent antibody and with Tululin antibody was performed as controls for equal loading (Fig. 3,C, top and bottom). Furthermore, it was confirmed by immunostaining that the cells overexpressing hMAWD were just the anti-phospho-MAPK immunoreactive cells (Fig. 2 C). Thus, hMAWD overexpression could lead to constitutive activation of MAPK.

It seems unlikely that activation of MAPK depends on the hMAWD-stimulated autocrine growth factors such as fibroblast growth factor or platelet-derived growth factor because the supernatant of the hMAWD-overexpressing clones could not activate the MAPK of parental cells (data not shown). We have also found that the amount of tyrosine-phosphorylated protein (14) was not increased in hMAWD-overexpressing cells compared with the parental cells (data not shown). Treatment with the specific MAP/extracellular signal-regulated kinase inhibitor (PD98059, 50 mm) reduced the level of activated MAPK to normal levels and blocked the ability to undergo anchorage-independent growth in soft agar; however, ras activation was not found in the hMAWD-overexpressing cells (data not shown). These findings may indicate the Ras-independent MAPK activation by hMAWD. Several Ras-independent pathways have been shown to induce MAPK activation (15, 16, 17). To gain insight into possible mechanisms of MAPK activation by hMAWD, we are looking for upstream proteins that might interact with hMAWD.

These findings prompted us to investigate whether the hMAWDgene was also overexpressed in human cancers. We found preliminarily that the PCR fragments corresponding to hMAWD mRNA transcripts were well amplified in breast tumors relative to patient-matched normal breast tissues (data not shown). We next evaluated whether the increased expression of hMAWD was paralleled by an increase in immunoreactive protein. Forty-six primary tumor breast tissues that were examined histologically for the presence of tumor cells and patient-matched adjacent normal breast tissues were obtained and examined to quantify the contents in breast carcinomas. None of the patients included in this study had received radiation therapy or chemotherapy before surgery. As shown in Fig. 4, where representative 15 cases were presented, Western blot analyses revealed a immunoreactive band of approximately M_r 39,000 in the tumor tissues that were undetectable in the normal tissue counterpart. In 21 of 46(45.6%) breast cancer tissues compared with normal tissues, higher expression of hMAWD was detected by immunoblotting analyses. It seemed unlikely that the increases in hMAWD expression in breast cancer tissues merely reflected an increased amount of applied protein,because SHP2, a protein tyrosine phosphatase, expression was not changed when compared with normal tissue (Fig. 4). Furthermore, by immunohistochemical studies using specific anti-hMAWD antibody, the immunoreactive staining in cytoplasm was observed in the tumor cells (Fig. 2 D). No signal was seen in the normal cells of these specimens, whereas an intensity-variable diffuse cytoplasmic staining in at least 50% of the tumor cells was detected in all breast cancers that had been formerly examined to be hMAWD positive by Western blot analyses, suggesting that the staining was not attributable to fixation artifacts. In this screening approach, hMAWD expression levels were thus independently up-regulated during mammary tumorigenesis.

Breast cancer, which is one of the most common life-threatening diseases in women, occurs in hereditary and sporadic forms. However,because malignant tumor progression is a complex series of sequential steps that incorporate specific properties of tumor cells and host, the mechanism is still not well understood. Our present findings may indicate that the hMAWD gene is a frequent target for alterations in human breast cancers. Similar to the examples of erbB2 or cyclin D, hMAWD may also be useful clinical tumor markers for managing patients with breast cancer (18, 19, 20). Of course these studies have to be repeated in the future before definitive conclusions can be drawn; however, the novel finding in the present study provides valuable information for understanding the WD-40 repeats proteins and to find the next target of TGF-β receptor and cancer signals.

We thank Ryoko Miyabe for excellent technical assistance.