Overexpression of the Amplified Pip4k2b Gene from 17Q11–12 In

Overexpression of the ampliﬁed Pip4k2b gene from 17q11–12 in breast cancer cells confers proliferation advantage

Shiuh-Wen Luoh*,1,2,4, Natarajan Venkatesan3 and Reshimi Tripathi1

1Department of Medicine, Division of Hematology and Oncology, University of Cincinnati, College of Medicine, Cincinnati, OH 45267, USA; 2Cincinnati Veterans’ Affairs Medical Center Cincinnati, Cincinnati, OH 45220, USA; 3Division of Hematology and Oncology, Department of Medicine, UCLA Medical Center, Los Angeles, CA 90095, USA

Gene amplification is common in solid tumors and is amplification maximum distinct from that of HER-2/Neu associated with adverse prognosis, disease progression, and serve as an independent target for amplification and and development of drug resistance. Asmall segment from selective retention. Pip4k2b amplification is associated chromosome 17q11–12 containing the HER-2/Neu gene is with overexpression at the RNAand protein level in breast amplified in about 25% of breast cancer. HER-2/Neu cancer cell lines. Stable expression of Pip4k2b in breast amplification is associated with adverse prognosis and cancer cell lines with and without HER-2/Neu amplifica- may predict response to chemotherapy and hormonal tion increases cell proliferation and anchorage-indepen- manipulation. Moreover, HER-2/Neu amplification may dent growth. The above observations implicate Pip4k2b in select patients for anti-HER-2/Neu-based therapy with the development and/or progression of breast cancer. Our Herceptin. We and others recently described a common study suggests that Pip4k2b may be a distinct target for sequence element from the HER-2/Neu region that was gene amplification and selective retention from 17q11–12. amplified in breast cancer cells. In addition, most, if not Oncogene (2004) 23, 1354–1363. doi:10.1038/sj.onc.1207251 all, of the amplified genes from this region display Published online 22 December 2003 overexpression. This raises the intriguing possibility that genes immediately adjacent to HER-2/Neu may influence Keywords: tumor biology; breast cancer; gene the biological behavior of breast cancer carrying HER-2/ amplification; gene expression; HER-2/Neu Neu amplification and serve as rational targets for therapy. By extracting sequence information from public databases, we have constructed a contig in bacterial artificial chromosomes (BACs) that extends from HER-2/ Introduction Neu to a phosphotidylinositol phosphate kinase (PIPK), Pip4k2b from 17q11–12. Although a role of PI-3-kinase Gene amplification is common in solid tumors and is and AKT in cancer biology has been previously described, associated with adverse prognosis, disease progression, PIPK has not been previously implicated. We show that and development of resistance to chemotherapy (Save- Pip4k2b, initially known as Pip5k2b, is amplified in a lyeva and Schwab, 2001). Amplification of N-Myc in subset of breast cancer cell lines and primary breast neuroblastoma (Seeger et al., 1985; Rubie et al., 1997) cancer samples that carry HER-2/Neu amplification. Out and amplification of HER-2/Neu in breast cancer are of eight breast cancer cell lines with HER-2/Neu associated with shortened survival (Slamon et al., 1987, amplification, three have concomitant amplification of 1989). Amplification of the DHFR gene enables cancer the Pip4k2b gene – UACC-812, BT-474 and ZR-75-30. cells to acquire resistance to the chemotherapeutic Similarly, two out of four primary breast tumors with agent, methotrexate (Savelyeva and Schwab, 2001). HER-2/Neu amplification carry Pip4k2b gene amplifica- Breast cancer with HER-2/Neu amplification may tion. Intriguingly, one tumor displays an increase in the exhibit a relative resistance to nonanthracycline-based gene copy number of Pip4k2b that is significantly more therapy but display retained or increased sensitivity to than that of HER-2/Neu. Moreover, dual color FISH anthracycline-based treatment (Muss et al., 1994; Paik reveals that amplified Pip4k2b gene may exist in a distinct et al., 1998; Thor et al., 1998). Amplification of structure from that of HER-2/Neu in ZR-75-30 cell line. chromosome 17q11–12 involving the HER-2/Neu gene These studies suggest that Pip4k2b may reside on an has been described in breast (Slamon et al., 1987), ovarian (Slamon et al., 1989), head and neck (Press et al., *Correspondence: S-W Luoh, Division of Hematology and Oncology, 1994), prostate (Ross et al., 1997), endometrial (Rolitsky Oregon Health Sciences University, Portland VA Medical Center, 3710 et al., 1999), gastric and esophageal cancers (Ishikawa SW US Veterans Hospital Road, RAD 47, Portland, OR 97239, USA; et al., 1997; Nakajima et al., 1999; Nessling et al., 1998; Fax: 503-273-5351 Brien et al., 2000; Walch et al., 2000). As is the case with 4Supported in part by American Cancer Society, Ohio Chapter, Ohio Cancer Research Associates, and VA Merit Award gene amplification in general, an increase in the copy Received 26 June 2003; revised 4 September 2003; accepted 29 September number of HER-2/Neu gene leads to overexpression of 2003 HER-2/Neu transcript and protein. Pip4k2b amplification in breast cancer S-W Luoh et al 1355 Retention of amplified sequences in cancer cells In this study, we show that Pip4k2b is amplified and indicates evolutionary advantage conferred by these overexpressed in a subset of breast cancer cell lines. amplified genes. Gene amplification may involve vari- Amplification of Pip4k2b is associated with overexpres- able amount of sequence. At one extreme, the amplified sion on the transcript and protein level. An increase in sequence involving N-Myc in some neuroblastoma cells the gene copy number of Pip4k2b that is significantly is pruned to a very limited segment of which N-Myc may more than that of HER-2/Neu has been observed in one be the only contributing genetic element (Amler and primary breast tumor. Moreover, dual color fluores- Schwab, 1989). On the other hand, amplification of cence in situ hybridization (FISH)reveals that amplified chromosome 11q13 may involve large, discontinuous Pip4k2b gene may exist in structure not accompanied by blocks of chromosomal DNA (Gaudray et al., 1992; that of HER-2/Neu in ZR-75-30 breast cancer cell line. Szepetowski et al., 1993). The extent of the sequence These studies suggest that Pip4k2b may fall on an that is amplified appears to be distinct among various amplification maximum distinct from that of HER-2/ tumor types. The pattern of overexpression of genes from Neu and serve as an independent target for amplification the amplified segment differs as well (Theillet et al., 1990; and selective retention. Overexpression of Pip4k2b in Lammie et al., 1991; Lammie and Peters, 1991; Proctor breast cancer cell lines confers proliferation advantage et al., 1991; Szepetowski et al., 1992; Roelofs et al., 1993; and promotes anchorage-independent growth. Pip4k2b Bringuier et al., 1996). This nonuniformity provides may therefore play important roles in the development evidence that distinct gene(s)are targets of 11q13 and/or progression of breast cancer. amplification in various tumor types. Similar complexity has been described for amplification events involving 17q22–23. Several specific and independent amplification Results maxima from this region were identified in breast cancer cell lines and breast tumors (Barlund et al., 2000; Wu A 17q11–12 contig from HER-2/Neu to Pip4k2b et al., 2000; Monni et al., 2001; Wu et al., 2001). The gene content of the HER-2/Neu amplified By extracting sequence information from the public sequences has recently been elucidated. Multiple genes databases, we assembled a contig in BACs that extended from the chromosomal segment containing HER-2/Neu from Hs.374824 through the HER-2/Neu gene to a lipid gene are coamplified with HER-2/Neu (Kauraniemi kinase, Pip4k2b from chromosome 17q11–12 (Figure 1). et al., 2001; Luoh, 2002). Interestingly, there is near Overlap was established based on identity of sequenced perfect correlation between increased gene dosage and portions of different BACs. This was verified with overexpression for these genes in breast cancer cell lines multiple PCR assays using primers derived from the (Kauraniemi et al., 2001; Luoh, 2002). This observation regions of overlap (data not shown). Dual color FISH raises the intriguing possibility that overexpression of using two differentially labeled BAC probes derived multiple coamplified genes, rather than amplification of from this region revealed colocalization of hybridization HER-2/Neu alone, may contribute to the development signals in breast cancer cells (not shown). A similar and/or progression of breast cancer that carries 17q11– contig using many of the BACs reported here has 12 amplification. Moreover, these genes may serve as recently been reported by NCBI (NT_010755), in additional targets for therapy for this subset of breast confirmation of our result. The whole region has been cancer patients who have a more aggressive form of completely sequenced. The distance between Hs.374824 disease. and HER-2/Neu is about 220 kb and that between HER- By extracting sequence information from public data- 2/Neu and Pip4k2b is about 900 kb. bases, we have constructed a contig in bacterial artificial chromosomes (BACs)that extends from HER-2/Neu to a phosphotidylinositol phosphate kinase (PIPK), Pip4k2b from 17q11–12. Initially known as Pip5k2b it is now reclassified as Pip4k2b (Rameh et al., 1997). The importance of phosphoinositides (PI)as lipid signaling molecules in eukaryotic cells has been well characterized. A number of phosphotidylinositol phosphate Figure 1 Physical linkage of HER-2/Neu and Pip4k2b.A kinases have been described that can phosphorylate chromosomal segment from 17q11–12 extending from HER-2/ different positions of the inositol ring leading to the Neu to Pip4k2b was cloned in a set of overlapping BACs. A total of formation of PIP2 and PIP3 that have various signaling 13 BACs were used to assemble this contig. BAC #129 was functions (Fruman et al., 1998; Anderson et al., 1999). partially sequenced (dotted lines). BAC #131, #906, #56, #386, Known PI-3-kinases and PI-4-kinases share sequence #2206, #387 #62, #2019, #390, #94 and #58 were completely sequenced (thick solid lines). BAC #98 has its insert sequenced homology over distinct domains. Pip4k2b belongs to a from both ends (thin line with circles at ends). The location and novel family of PI-4-P 5-kinases, which do not share direction of transcription for Pip4k2b HER-2/Neu, and Hs.374824 significant homology to other known protein or lipid are indicated. The direction of the centromere is towards the right. kinases. PIPKs have been implicated in many important Since the assembly of this contig, the whole region has been sequenced. The coordinates for Pip4k2b HER-2/Neu, and Hs. functions such as vesicular trafficking and cytoskeleton 374824 are 680854–645841, 1580230–1610078, and 1808550– assembly (Anderson et al., 1999). A role for PIPKs in 1801192, respectively (Genbank number NT_010755.13, April cancer has not been previously reported. 2003 release)

Oncogene Pip4k2b amplification in breast cancer S-W Luoh et al 1356 Pip4k2b amplification in breast cancer cell lines (Figure 2a). Five different EcoRI fragments were expected based on sequence information: a doublet at Southern blotting analysis revealed that Pip4k2b was 6.0 kb (6.0 and 6.0); a doublet around 3.0 kb (2.9 and amplified in three of the 15 breast cancer cell lines tested 3.0); and a band at 4.0 kb. The signal intensity of the

Figure 2 Amplification and overexpression of Pip4k2b in breast cancer cell lines. (a)Southern blotting analysis of genomic DNA isolated from 15 breast cancer cell lines using the open reading frame from the full-length Pip4k2b cDNA as the probe. The positions of the 6.0, 4.0 and 3.0 kb signals that represent fragments derived from Pip4k2b genomic locus after EcoRI digest are marked with arrows. The amount of digested genomic DNA loaded for each sample on the gel is shown with ethidium bromide staining at the bottom. (b) Northern blotting analysis of total RNA isolated from representative breast cancer cell lines using the same Pip4k2b probe as above. The position of the 6.2 kb transcript from Pip4k2b is marked with an arrow. MCF-7/HER-2 is derived from parental MCF-7 cell line and overexpresses HER-2/Neu transcript as a result of transfection. The amount of RNA loaded for each sample on the gel is shown with ethidium bromide staining at the bottom. (c)Immunoblotting analysis of total cellular protein isolated from representative breast cancer cell lines using an affinity-purified antipeptide antiserum against Pip4k2b. The position of the 48 kDa Pip4k2b protein signal is marked with an arrow. Left: no expression is noted in parental MCF-7 or SKBR-3 cell lines. Signals with the predicted size are seen in MCF-7 and SKBR-3 cells transfected with a full-length Pip4k2b cDNA expression vector, that is, MCF-7::GFP-IRES-Pip4k2b and SKBR-3::GFP-IRES-Pip4k2b. Signals of appropriately increased molecular weight (B75 kDa)are seen in the SKBR-3 and MCF-7 cells that are trasnfected to express a GFP and Pip4k2b fusion protein, that is, MCF-7::GFP-Pip4k2b and SKBR-3::GFP-Pip4k2b.No signal is seen with preimmune serum (not shown). Right: Pip4k2b expression in a panel of breast cancer cell lines by immunoblotting with affinity-purified antibody with validated specificity

Oncogene Pip4k2b amplification in breast cancer S-W Luoh et al 1357 4.0 kb band was much lower due to the small size of In ZR-75-30, all amplified HER-2/Neu signals (red)were exons contained in this genomic fragment. Of eight accompanied by amplified Pip4k2b signals (green). breast cancer cell lines with known HER-2/Neu ampli- Intriguingly, clusters of amplified Pip4k2b signals were fication (HCC-1954, SKBR-3, BT-474, UACC-812, observed without concomitant amplification of HER-2/ UACC-893, MB-453, ZR-75-30 and MB-361), three Neu (Figure 3). These results indicate that Pip4k2b can (ZR-75-30, BT-474 and UACC-812)displayed amplifi- be amplified and/or selectively retained independently cation of Pip4k2b We did not observe Pip4k2b from HER-2/Neu. In BT-474 cells, on the contrary, there amplification in any breast cancer cell line without is colocalization of hybridization signals from amplified HER-2/Neu amplification. Pip4k2b and HER-2/Neu genes (data not shown).

Concurrent Pip4k2b overexpression and amplification in HER-2/Neu and Pip4k2b amplification – analysis of breast cancer cell lines primary breast tumor samples Northern blotting analysis was performed to study the We then used Southern analysis to examine the expression of Pip4k2b from different breast cancer cell amplification status of both HER-2/Neu and Pip4k2b lines (Figure 2b). A transcript of 6.4 kb was detected genes in primary breast tumor samples. DNA samples corresponding to the full-length Pip4k2b transcript. from 12 randomly selected primary breast tumors were High level of Pip4k2b expression was seen only in the analysed. Three probes were used – HER-2/Neu, three cell lines that displayed amplification of the Hs.374824, and Pip4k2b Hs.374824 is located on the Pip4k2b gene – BT-474, UACC-812 and ZR-75-30. telomeric side of HER-2/Neu gene while Pip4k2b is on Modest expression without significant gene amplification the centromeric side (Figure 1). As can be seen in was noted in MB-361 cell line. Figure 4, samples #1,#3, #6, and #12 had an increase in In order to examine the expression of Pip4k2b protein the copy number for HER-2/Neu and Hs.374824genes. in different breast cancer cell lines, we first evaluate the Out of these four, only samples #3 and #12 showed a specificity of our affinity-purified antiserum. Parental MCF-7 and SKBR-3 cells display no amplification or expression of Pip4k2b gene. As expected, no Pip4k2b protein signal was seen. A predicted 48 kDa protein product was seen in the SKBR-3 and MCF-7 cells trasnfected with a full-length Pip4k2b cDNA expression vector, that is, MCF-7::GFP-IRES-Pip4k2b and SKBR- 3::GFP-IRES-Pip4k2b (Figure 2c, left). A B75 kDa protein product, expected of a Pip4k2b and GFP fusion protein, was seen in the SKBR-3 and MCF-7 cells that were trasnfected to express a GFP and Pip4k2b fusion protein, that is, MCF-7::GFP-Pip4k2b and SKBR- 3::GFP-Pip4k2b (Figure 2c, left). Immunoblotting with this affinity-purified antipeptide antiserum with validated specificity showed that the 48kDa Pip4k2b protein was readily detectable in BT- 474, ZR-75-30 and UACC-812 cells but not in several other breast cancer cell lines without gene amplification or overexpression (Figure 2c, right). The above study indicates that there is very good correlation between gene amplification and overexpression for Pip4k2b in breast cancer cell lines. Of note, overexpression of HER- 2/Neu in MCF-7 cell line did not significantly alter the expression of Pip4k2b.

HER-2/Neu and Pip4k2b amplification – a dual color FISH study Figure 3 Differential HER-2/Neu and Pip4k2b gene amplification in ZR-75-30 breast cancer cell. Dual-color FISH was performed on We used dual color FISH study to examine the structure metaphase spreads with a pool of two differentially labeled BAC of HER-2/Neu and Pip4k2b amplified DNA in breast probes. BAC#62, which contains a portion of HER-2/Neu gene, was labeled in top red. BAC#58, which contains Pip4k2b gene, was cancer cell lines. BACs #62, which contains a portion labeled in green. Top Left: signals from amplified HER-2/Neu gene of the HER-2/Neu gene, and BAC #58, which contains are in red. Top Right: signals from amplified Pip4k2b gene are in the Pip4k2b gene in its entirety, were differentially green. Bottom Middle: signals from both amplified HER-2/Neu labeled. These probes were then pooled and hybridized and Pip4k2b genes are visualized simultaneously with appropriate filter setting. Yellow signals are indicative of superimposition of simultaneously to metaphase spreads of breast cancer green and red signals. Arrows indicate colocalization of hybridiza- cells. Appropriate filter setting was adjusted to visualize tion signals generated by both BAC#62 and BAC#58. Arrowheads each hybridization signal either individually or together. indicate hybridization signals produced by BAC#58 only

Oncogene Pip4k2b amplification in breast cancer S-W Luoh et al 1358 Western blotting (Figure 2c, left). These cells were maintained in a logarithmic phase of growth in G418. The proliferation of different transfectants was then measured directly. A fixed number of transfectants were plated out on day 0 in duplicate. The cell number of one set of cells was counted on day 1 and that for the remaining set on day 3 for MCF-7 transfectants and day 5 for SKBR-3 transfectants, respectively. Growth rate was determined by the ratio of the cell number derived on day 3 or 5 over that on day 1. The average fold of increase was 4.49 for MCF-7::GFP cells and 6.06 for MCF-7::GFP-IRES-Pip4k2b cells, respectively. The difference in the fold of increase between MCF- 7::GFP cells and MCF-7::GFP-IRES-Pip4k2b cells (mean difference ¼ 1.57)was statistically significant (paired t ¼ 3.12; df ¼ 14; and P ¼ 0.0075 ). The average fold of increase was 6.35 for SKBR-3::GFP cells and 8.28 for SKBR-3::GFP-IRES-Pip4k2b cells. The difference in the fold of increase between SKBR-3::GFP Figure 4 Differential Pip4k2b and HER-2/Neu gene amplification cells and SKBR-3::GFP-IRES-Pip4k2b cells (mean in primary breast cancer samples. Genomic DNA from 12 primary difference 1.92)was statistically significant as well breast cancer samples was analysed with Southern blotting for ¼ amplification of HER-2/Neu, Pip4k2b and Hs.374824, respectively. (paired t ¼ 2.64; df ¼ 11; and P ¼ 0.023). The amount of DNA loaded for each sample on the gel is shown To corroborate the above observation, we used with ethidium bromide staining at the top CellTiter assays (Promega)to measure the proliferation of MCF-7 and SKBR-3 transfectants, respectively. Replicate samples of different transfectants were plated parallel increase in the copy number of the Pip4k2b at a fixed number in 96-well plates (n ¼ 8). Proliferation gene. Most notably, there was an increase in the gene of plated cells was then serially determined at days 1, 3 copy number of Pip4k2b in sample #5 without and 5. For both MCF-7 and SKBR-3 cells, Pip4k2b significant increase for either HER-2/Neu or Hs.374824 transfectants displayed statistically increased prolifera- gene. The latter indicates that Pip4k2b may constitute tion compared with their GFP expressing control part of an amplification maximum whose increase in (Figure 5). Taken together, the above shows that in gene copy number and/or subsequent selective retention the presence or absence of HER-2/Neu amplification, is not exclusively driven by HER-2/Neu. overexpression of Pip4k2b in breast cancer cells may Taken together, analysis of both breast cancer cell confer proliferation advantage. lines and primary breast tumor samples indicates that amplification and/or subsequent retention of amplified Forced expression of Pip4k2b protein causes Pip4k2b can proceed independently of HER-2/Neu. This morphological changes in MCF-7 cells supports the notion that Pip4k2b may be an additional target for amplification that confers selection advantage Expression of wild-type Pip4k2b in MCF-7 cells was from chromosome 17q11–12. associated with morphological changes (Figure 6).

Forced expression of Pip4k2b protein confers proliferation advantage in breast cancer cells with or without HER-2/Neu amplification To understand the functional outcome of Pip4k2b protein overexpression in breast cancer cells, breast cancer cell lines that stably overexpressed Pip4k2b protein were established by transfection of a Pip4k2b cDNA expression vector. Two breast cancer cell lines were studied – MCF-7 and SKBR-3. MCF-7 cells do not Figure 5 Proliferation advantage of MCF-7 and SKBR-3 breast carry HER-2/Neu amplification whereas SKBR-3 cells cancer cells with Pip4k2b overexpression. The growth rates of have high level of amplification. Cells expressing both MCF-7 (left)and SKBR-3 (right)cells expressing GFP (open boxes)and GFP and Pip4k2b (filled box)were compared. Identical GFP and Pip4k2b protein were made via transfection numbers of cells were seeded on day 1 and cell proliferation was with a bi-cistronic expression vector. Cells expressing measured with CellTiter assays (Promega)on days 1, 3 and 5. The GFP alone served as a control. FACS was performed to number of cells was measured with absorbance at 490 nM. derive a homogeneous polyclonal population of trans- Experiments were performed eight times for each time point (n ¼ 8)and repeated at least three times with identical results. The fectants that expressed GFP, which could function as a result of one representative experiment is shown here. Data are surrogate marker for Pip4k2b expression. The expression expressed as mean7s.d., and statistical significance is assessed by of Pip4k2b in stable transfectants was verified by two-tailed t-test for independent groups. **Po0.05; ***Po0.01

Oncogene Pip4k2b ampliﬁcation in breast cancer S-W Luoh et al 1359

Figure 6 Induction of morphological changes in MCF-7 breast cancer cells by Pip4k2b overexpression. Left: MCF-7 breast cancer cells that express GFP-MCF-7::GFP. Parental MCF-7 cells exhibit the same morphology (not shown). Right: MCF-7 breast cancer cells that express GFP and Pip4k2b-MCF-7::GFP-IRES-Pip4k2b

MCF-7::GFP-IRES-Pip4k2b cells are more refractile and rounded up, suggestive of loss of adhesiveness to the plastic surface in tissue culture. MCF-7::GFP-IRES- Pip4k2b cells also appear to lose their tendency for glandular formation in culture that is common for MCF-7::GFP cells (Figure 6)and parental MCF-7 cells (not shown). These morphological changes are not observed in SKBR-3 cells (not shown).

Forced expression of Pip4k2b protein promotes anchorage-independent growth in breast cancer cells with or without HER-2/Neu amplification Figure 7 Promotion of anchorage-independent growth in both MCF-7 and SKBR-3 breast cancer cells by Pip4k2b overexpres- MCF-7::GFP-IRES-Pip4k2b cells readily formed visible sion. The ability of breast cancer cells to form colonies in soft agar colonies in soft agar assays, which was not observed in was examined. Both non-HER-2/Neu amplified cells – MCF-7 (a), and HER-2/Neu amplified cells – SKBR-3 (b), were tested. MCF- MCF-7::GFP cells (Figure 7, upper). SKBR-3::GFP 7::GFP-IRES-Pip4k2b and SKBR-3::GFP-IRES-Pip4k2b are cells could produce a small number of small colonies in breast cancer cells that overexpress GFP and Pip4k2b. MCF- soft agar. By contrast, SKBR-3 cells expressing Pip4k2b 7::GFP and SKBR-3::GFP cells express GFP only and serve as and GFP showed enhanced colony formation with an controls. These experiments have been done at least four times with increase in both the number and size of colonies in soft identical results. The result from one representative experiment is shown here agar (Figure 7, lower). The above indicates that expression of Pip4k2b protein can promote anchorage- independent growth in breast cancer cells in the presence as well as the absence of HER-2/Neu amplification. likely have tumor-related functions and their roles in breast cancer development and/or progression are worthy of further study. Moreover, breast cancer with HER-2/Neu amplification is a more virulent disease. Discussion Current anti-HER-2/Neu-based therapy with Herceptin benefits only about half of the patients (Slamon et al., Amplification and overexpression of HER-2/Neu proto- 2001). Additional effective therapy is thus needed to oncogene and its association with poor clinical outcome battle this more aggressive form of breast cancer. have been well established (Slamon et al., 1987, 1989; Continued work, therefore, is necessary to assess Press et al., 1993; Seshadri et al., 1993; Ravdin and whether other coamplified and overexpressed genes from Chamness, 1995; Ross and Fletcher, 1999; Pauletti et al., the region may serve as rational targets for therapy. 2000). Overexpression of HER-2/Neu is thought to be We report in this study that Pip4k2b, a PIPK, is responsible for many of the characteristics of breast physically linked to HER-2/Neu from 17q11–12. Initi- cancer with HER-2/Neu amplification (Di Fiore et al., ally known as Pip5k2b it is now reclassified as Pip4k2b 1987; Hudziak et al., 1987; Chazin et al., 1992; (Rameh et al., 1997). Pip4k2b is amplified in a subset of Guy et al., 1992; Pietras et al., 1995). Contribution breast cancer cells with HER-2/Neu amplification. from additional coamplified and overexpressed genes Amplification is associated with overexpression on both from 17q11–12 cannot be ruled out. The core region from RNA and protein levels. Interestingly, amplification 17q11–12 that is commonly amplified in breast cancer and/or selective retention of Pip4k2b can occur inde- cell lines has recently been described (Kauraniemi et al., pendent of HER-2/Neu, indicating that Pip4k2b can be 2001; Luoh, 2002). Intriguingly, there is a striking an additional target for amplification from 17q11–12. correlation between amplification and overexpression of The retention of Pip4k2b amplified sequences indicates genes neighboring HER-2/Neu from this region (Kaur- selection advantage conferred by an increase in the copy aniemi et al., 2001; Luoh, 2002). Several of these genes number of the Pip4k2b gene. In support of this notion,

Oncogene Pip4k2b amplification in breast cancer S-W Luoh et al 1360 overexpression of Pip4k2b gene in breast cancer cell association of Pip4k2b with p55 TNF receptor remains lines can indeed increase proliferation rates and enhance to be elucidated. anchorage-independent growth. This growth promoting Amplification of the gene encoding the catalytic function is observed in the presence as well as the subunit of the PI-3-K has been reported in about 40% absence of concurrent HER-2/Neu amplification. Our of ovarian cancer (Shayesteh et al., 1999)and non-small- study therefore establishes an important functional role cell lung cancer (Massion et al., 2002). There is for Pip4k2b in the development and/or progression of association between an increase in the gene copy number breast cancer that carries 17q11–12 amplification and and the PI-3-K enzyme activity (Shayesteh et al., 1999). overexpression involving this gene. In addition, AKT2, which is a downstream effector of In mammalian cells, two types of PIPKs have been PI-3-K, is amplified in ovarian and pancreatic cancers characterized, type I and type II and each has three (Cheng et al., 1992; Bellacosa et al., 1995; Cheng et al., isoforms that are named a, b, and g (Fruman et al., 1996). A role of PIPK in cancer has not been reported in 1998; Anderson et al., 1999). The diversity of PIPKs the past. We describe here for the first time that Pip4k2b allows them to use different substrates and to generate is a target for amplification from 17q11–12. Moreover, multiple signaling molecules. Although related in overexpression of Pip4k2b is associated with a broad sequence, type I and type II PIPKs localize to different range of cancer-related functions such as morphological cellular compartments, have different substrate specifi- changes, increased proliferation and enhanced ancho- city and are functionally nonredundant (Homma et al., rage-independent growth. Our observations implicate 1998; Kunz et al., 2000). Initially identified as a PI-4-P 5- Pip4k2b in the development and/or progression of a kinase, type II PIPK has been shown to be predomi- subset of breast cancer that carries its amplification. nantly a 4-kinase (Rameh et al., 1997). Type I PIPKs are In summary, we report amplification and overexpres- dual specificity kinases and can phosphorylate both the sion of Pip4k2b in breast cancer. Pip4k2b is amplified in 4- and 5-hydroxyls of the inositol ring (Rameh et al., a subset of breast cancer with HER-2/Neu amplification 1997; Zhang et al., 1997; Tolias et al., 1998). More but it can be a target for amplification and/or selective recently, a novel type III PIPK, which is the mammalian retention independent of HER-2/Neu. Overexpression of homologue of the yeast Fab1, has been identified (Gary Pip4k2b in breast cancer cells leads to morphological et al., 1998; McEwen et al., 1999; Sbrissa et al., 1999). change and enhances proliferation and anchorage- In mammalian cells, type I PIPKs are found at the cell independent growth. Our study unravels unexpected membrane, in the nucleus and Golgi. Type II PIPKs genetic complexity and indicates gene(s)other than localize to the cytosol, the nucleus, the ER and the actin HER-2/Neu from 17q11–12 may confer selection ad- cytoskeleton but not primarily at the plasma membrane. vantage and serve as target(s)for amplification. Our Type I PIPKs have been implicated in secretion, study reveals that Pip4k2b may play important roles in vesicular trafficking, apoptosis, ion channel activity the development and/or progression of breast cancer. and regulation of actin assembly in eukaryotic cells Further study will focus on whether breast cancer with (Hay et al., 1995; Shibasaki et al., 1997; Huang et al., Pip4k2b amplification has distinct clinical behaviors and 1998; Ishihara et al., 1998; Arneson et al., 1999; whether blockage of its activity is of therapeutic benefit Mejillano et al., 2001). The function of type II PIPKs in breast cancer with Pip4k2b amplification. is largely unknown. Recent report suggests a role for the type II PIPKs in calcium ion induced granule secretion in platelets (Rozenvayn and Flaumenhaft, 2001). Materials and methods Involvement of type II PIPK in aspects of tumor BAC and plasmid constructs necrosis factor a-mediated signaling has been proposed (see below)(Castellino et al., 1997). The roles of PI and BAC and plasmid clones used in this study were obtained PIPKs in nuclear PI signaling remain speculative at this from Research Genetics (Huntsville, AL, USA). The follow- point. ing BACs have been sequenced in their entirety: BAC#131 Recently the crystal structure of Pip4k2b has been (Genbank accession number AC005288), BAC#906 (AC004408), BAC#56 (AC006441), BAC#62 (AC087491), solved (Rao et al., 1998). The tertiary structure of BAC#2019 (AC040933), and BAC#390 (AC009283), Pip4k2b consists of two identical subunits that interact BAC#58 (AC006449), BAC#2206 (AC091178), BAC#387 at the N terminus. The dimer is an elongated flat, disc (AC090844)and BAC#386 (AC007901).The following BAC like structure with clustering of charged amino acids on has been sequenced in draft: BAC#129 (AC021102). BAC#98 one surface. The charge and flatness suggest that it (AQ320284 and AQ320287)has been end sequenced. The serves as an interface for membrane association via identities and overlap of BACs were verified by PCR assays electrostatic interaction with the phospholipid head- based upon available sequence information. Preparation of groups. The N-terminal portion of Pip4k2b is required BAC DNA was performed with QIAGEN Plasmid Midi Kit for its association with the p55 TNF receptor. The (QIAGEN, Inc., Valencia, CA, USA)except that twice as portion of the p55 TNF receptor that is required for this much buffer was used in alkaline lysis. Full-length cDNA of human Pip4k2b was derived from association is the juxta-membrane domain of the IMAGE clone #2412369. Sequence analysis of the insert of receptor (Castellino et al., 1997). Many of the PI IMAGE #2412369 revealed that it contained the entire coding signaling enzymes directly interact with membrane region of the Pip4k2b gene identical to that described in receptors with the most notable ones being receptor Genbank number U85245. Pip4k2b open reading frame was tyrosine kinases. The functional significance of the released as a SmaI and XhoI fragment and inserted into the

Oncogene Pip4k2b amplification in breast cancer S-W Luoh et al 1361 Eco47III and XhoI sites of pIRES2-EGFP vector (BD SDS–polyacrylamide gel and transferred onto Hybond ECL Bioscience Clontech, Palo Alto, CA, USA)to make the membranes (Amersham). These membranes were then blocked GFP-IRES-Pip4k2b expression vector. The same SmaI and at room temperature with gentle rocking for 1 h with 5% w/v XhoI fragment was inserted into the ScaI and XhoI sites of nonfat dry milk in 1 Â Tris-buffered saline and 0.05% Tween pEGFP-C3 to make the GFP-Pip4k2b expression vector. The 20 (TBST). Affinity-purified antipeptide antiserum at appro- former will concurrently express full-length Pip4k2b protein priate dilution was incubated with the membranes in the and GFP from a bi-cistronic message. The latter will express presence of TBST with 5% w/v nonfat dry milk overnight at Pip4k2b protein that is fused in frame, which is verified by 41C with gentle rocking. After washing and incubation with sequencing, to the carboxyl terminus of GFP. secondary antibody, signals were detected using the LumiGLO Hs.374824 is encoded by IMAGE #2156561 (Genbank reagents (Cell Signaling)and exposure to X-ray film number AF373101). Sequence analysis reveals that it encodes a (X-OMAT, Eastman Kodak Co.). protein that is evolutionarily conserved (S-W Luoh, unpub- lished observation). Both IMAGE #2412369 and #2156561 Antiserum and affinity purification clones were obtained from Research Genetics. Rabbit polyclonal antibody was generated against an isoform- Cell lines and cultures specific peptide derived from a nonhomologous region at the N-terminal portion of the human Pip4k2b protein. This region The following breast cancer cell lines were obtained from was chosen to minimize crossreactivity against other members the American Type Culture Collection (Manassas, VA, USA): of this family of lipid kinase (PIPKs). The peptide sequence BT-20, MB-MDA-468, MB-MDA-435, MB-MDA-453, is as follows: Ser-Ser-Asn-Cys-Thr-Ser-Thr-Thr-Ala-Val- MB-MDA-361, MCF-7, HBL-100, ZR-75-1, ZR-75-30, Ala-Val-Ala-Pro-Leu-Ser-Ala-Ser-Lys. The internal cysteine HCC-1954, SKBR-3, BT-474, T-47D, and UACC-812. residue was used for conjugation to Keyhole Limpet Hemo- MCF-7/HER-2 is MCF-7 cell overexpressing a stably trans- cyanin carrier protein. Serial boosting was performed and the fected full-length HER-2/Neu cDNA. All cells except UACC- titer of antiserum monitored with ELISA. Rabbit antiserum 812 are grown in RPMI supplemented with 10% fetal bovine was subsequently affinity purified using a column conjugated serum at 371C in a humidified atmosphere with 5% CO2. with the same peptide antigen used for immunization. UACC-812 is grown in L-15 medium supplemented with 10% fetal bovine serum without CO2. DNA transfection and establishment of polyclonal stable transfectants Southern analysis Pip4k2b expression and control vectors were transfected into Genomic DNA was isolated from cell lines following MCF-7 and SKBR-3 cell lines with TransIt-LT-1 reagent proteinase K digestion, phenol–chloroform extraction and (PanVera)according to the manufacturer’s recommendation. isopropanol precipitation. Genomic DNA from UACC-893 Cells were then selected in medium containing G418 (Life was obtained from the American Type Culture Collection Technologies, Inc.)at 400 mg/mm (active form). Polyclonal G418- (Manassas, VA, USA)directly. Breast tumor genomic DNA resistant cells were expanded and grown in logarithmic phase in was a gift from Dennis Slamon at UCLA Medical Center. A medium containing G418 and split for passage when appropriate. measure of 10 mg of genomic DNA was digested with EcoRI, separated in agarose gel, and transferred to nylon membrane. Flow cytometry and cell sorting Inserts from IMAGE clones #2412369 and #2156561 were used as probes. Membranes were hybridized with probes High GFP expressing transfected cells were sorted on a labeled with random hexamer priming, washed and subject to FACSVantaget SE (BD Biosciences, San Jose, CA, USA) autoradiography backed with intensifying screens. equipped with a Spectra-Physics argon-ion laser emitting at 488 nm. GFP was detected in the FL1 channel with a 530 nm band-pass filter. Analysis was performed using CELLQuestt Northern analysis software (BD Biosciences, San Jose, CA, USA). Flow Cytometry Total RNA was isolated from breast cancer cell lines with (Coulter XL)analysis was used to monitor the percentage of cells TRIzol based upon the manufacturer’s recommendation (Life that were expressing GFP with the FL-1 channel. Technologies, Gaithersburg, MD, USA). In all, 20 mg of total cellular RNA was denatured by heating in the presence of Cell proliferation assays formamide, separated by electrophoresis in 1% agarose gel containing formaldehyde, transferred to nylon membranes, (1)Sorted SKBR-3 or MCF-7 transfectants were seeded in and hybridized with a 32P-labeled Pip4k2b probe as described duplicate at a density of 50 000 per well in 12-well plates on day 0. above. The intensity of the 28S and 18S rRNAs of each lane Cells from one of the wells were harvested and counted manually after ethidium bromide staining prior to transfer was used as with a counting chamber on day 1. Cells from the remaining wells loading control. were counted on day 3 for MCF-7 transfectants and day 5 for SKBR-3 transfectants, respectively. The fold increase for different transfectants was compared for each experiment. These Protein isolation and Western analysis experiments were repeated multiple times as indicated for both Cultured cells were first washed twice with ice-cold phosphate MCF-7 and SKBR-3 transfectants. buffered saline (pH 7.4). and then lysed on ice in a buffer (2)Sorted SKBR-3 or MCF-7 transfectants were plated into containing 50 mM Tris-HCl (pH7.4), 150 mM NaCl, 1% NP- flat-bottomed 96-well plates at a density of 1 000 cells per well. 40, 0.25% sodium deoxycholate, 1 mM EDTA, 1 mM NaF, Replicate cultures (n ¼ 8)for each cell lines were plated. These 1mM PMSF, 1 mM sodium vanadate, 1 mg/ml aprotinin, cells were cultured for 5 days and their proliferation 1 mg/ml leupeptin, and 1 mg/ml pepstatin. After centrifugation determined by a CellTiter assay (Promega)on days 1, 3, in cold room at maximal speed for 30 min, supernatants and 5 according to the manufacturer’s recommendation. were collected and stored in À801C until use. For Western The absorbance of the colorimetric assay was determined blotting analysis, 20 mg of protein was separated on a 10% at a wavelength of 490 nm using a multichannel plate

Oncogene Pip4k2b amplification in breast cancer S-W Luoh et al 1362 spectrophotometer (Biotek Instruments, Winooski, VT, USA). nuclei in a solution containing sheared human DNA, 50% Each experiment was done at least three times with identical formamide, 10% dextran sulfate, and 2 Â SSC. Specific hybri- results. dization signals were detected by incubating the hybridized slides in fluoresceinated antidigoxigenin antibodies as well as Soft agar assay avidin Texas red followed by counterstaining with DAPI. RPMI medium (2 Â , Life Technologies, Inc.)supplemented Statistical analysis with 10% fetal bovine serum, 1 mML-glutamine, 100 IU/ml penicillin, 100 mg/ml streptomycin and 0.25 mg/ml amphoter- The fold increase in cell numbers was analysed by a two-tailed icin B was mixed with an equal volume of 0.6% Noble Agar paired t-test between the Pip4k2b expressing cells and the GFP (Cellgro, Mediatech Inc.). A measure of 4 ml of this mixture expressing control cells. For CellTiter assays, means of was first laid down as a bottom layer per 60-mm Petri dish. absorbance at 490 nM7s.d. were compared between the For top layers, 50 000 cells were well dispersed in the same Pip4k2b expressing cells and the GFP expressing control cells mixture and layered on top of the bottom agar. These cells (n ¼ 8)at days 1, 3 and 5 using two-tailed unpaired t-test. A were then allowed to grow for 4–6 weeks in a 371C tissue P-value of less than 0.05 is considered statistically significant. culture incubator supplied with 5% humidified CO2. All analyses were performed with statistical software (StatView version 5.0.1; SAS Institute Inc., Cary, NC, USA). Fluorescence in situ hybridization Chromosome spreads of trypsinized BT-474 and ZR-75-30 Acknowledgements cells were prepared according to standard procedures includ- I (S-WL)thank Drs Dennis J Slamon and Owen N Witte for ing hypotonic treatment. Purified BAC DNA from BAC#58 mentorship and support. We thank Drs Bob Franco, Tom was labeled with digoxigenin dUTP and BAC DNA from #62 Clemans, Jeff Kanuf, and Peter Stambrook for careful review was labeled with biotin dUTP by nick translation. Two labeled of the manuscripts. S-WL was a recepient of Fellow Scholar- probes were combined as differentially labeled pairs and ship from American Society of Hematology. This paper is hybridized to metaphase chromosomes and interphase dedicated to the loving memory of late Dr Ching-Shuenn Luoh.

References

Amler LC and Schwab M. (1989). Mol. Cell. Biol., 9, 4903–4913. Hay JC, Fisette PL, Jenkins GH, Fukami K, Takenawa T, Anderson RA, Boronenkov IV, Doughman SD, Kunz J and Anderson RA and Martin TF. (1995). Nature, 374, 173–177. Loijens JC. (1999). J. Biol. Chem., 274, 9907–9910. Homma K, Terui S, Minemura M, Qadota H, Anraku Y, Arneson LS, Kunz J, Anderson RA and Traub LM. (1999). Kanaho Y and Ohya Y. (1998). J. Biol. Chem., 273, J. Biol. Chem., 274, 17794–17805. 15779–15786. Barlund M, Monni O, Kononen J, Cornelison R, Torhorst J, Huang CL, Feng S and Hilgemann DW. (1998). Nature, 391, Sauter G, Kallioniemi O-P and Kallioniemi A. (2000). 803–806. Cancer Res., 60, 5340–5344. Hudziak RM, Schlessinger J and Ullrich A. (1987). Proc. Natl. Bellacosa A, de Feo D, Godwin AK, Bell DW, Cheng JQ, Acad. Sci. USA, 84, 7159–7163. Altomare DA, Wan M, Dubeau L, Scambia G, Masciullo V, Ishihara H, Shibasaki Y, Kizuki N, Wada T, Yazaki Y, Asano Ferrandina G, Panici PB, Mancuso S, Neri G and Testa JR. T and Oka Y. (1998). J. Biol. Chem., 273, 8741–8748. (1995). Int. J. Cancer, 64, 280–285. Ishikawa T, Kobayashi M, Mai M, Suzuki T and Ooi A. Brien TP, Odze RD, Sheehan CE, McKenna BJ and Ross JS. (1997). Am. J. Pathol., 151, 761–768. (2000). Hum. Pathol., 31, 35–39. Kauraniemi P, Barlund M, Monni O and Kallioniemi A. Bringuier PP, Tamimi Y, Schuuring E and Schalken J. (1996). (2001). Cancer Res., 61, 8235–8240. Oncogene, 12, 1747–1753. Kunz J, Wilson MP, Kisseleva M, Hurley JH, Majerus PW Castellino AM, Parker GJ, Boronenkov IV, Anderson RA and and Anderson RA. (2000). Mol. Cell, 5, 1–11. Chao MV. (1997). J. Biol. Chem., 272, 5861–5870. Lammie GA, Fantl V, Smith R, Schuuring E, Brookes S, Chazin VR, Kaleko M, Miller AD and Slamon DJ. (1992). Michalides R, Dickson C, Arnold A and Peters G. (1991). Oncogene, 7, 1859–1866. Oncogene, 6, 439–444. Cheng JQ, Godwin AK, Bellacosa A, Taguchi T, Franke TF, Lammie GA and Peters G. (1991). Cancer Cells, 3, 413–420. Hamilton TC, Tsichlis PN and Testa JR. (1992). Proc. Natl. Luoh SW. (2002). Cancer Genet. Cytogenet., 136, 43–47. Acad. Sci. USA, 89, 9267–9271. Massion PP, Kuo WL, Stokoe D, Olshen AB, Treseler PA, Cheng JQ, Ruggeri B, Klein WM, Sonoda G, Altomare DA, Chin K, Chen C, Polikoff D, Jain AN, Pinkel D, Albertson Watson DK and Testa JR. (1996). Proc. Natl. Acad. Sci. DG, Jablons DM and Gray JW. (2002). Cancer Res., 62, USA, 93, 3636–3641. 3636–3640. Di Fiore PP, Pierce JH, Kraus MH, Segatto O, King CR and McEwen RK, Dove SK, Cooke FT, Painter GF, Holmes AB, Aaronson SA. (1987). Science, 237, 178–182. Shisheva A, Ohya Y, Parker PJ and Michell RH. (1999). Fruman DA, Meyers RE and Cantley LC. (1998). Annu. Rev. J. Biol. Chem., 274, 33905–33912. Biochem., 67, 481–507. Mejillano M, Yamamoto M, Rozelle AL, Sun HQ, Wang X Gary JD, Wurmser AE, Bonangelino CJ, Weisman LS and and Yin HL. (2001). J. Biol. Chem., 276, 1865–1872. Emr SD. (1998). J. Cell. Biol., 143, 65–79. Monni O, Barlund M, Mousses S, Kononen J, Sauter G, Gaudray P, Szepetowski P, Escot C, Birnbaum D and Theillet Heiskanen M, Paavola P, Avela K, Chen Y, Bittner ML C. (1992). Mutat. Res., 276, 317–328. and Kallioniemi A. (2001). Proc. Natl. Acad. Sci. USA, 98, Guy CT, Webster MA, Schaller M, Parsons TJ, Cardiff RD 5711–5716. and Muller WJ. (1992). Proc. Natl. Acad. Sci. USA, 89, Muss HB, Thor AD, Berry DA, Kute T, Liu ET, 10578–10582. Koerner F, Cirrincione CT, Budman DR, Wood WC,

Oncogene Pip4k2b ampliﬁcation in breast cancer S-W Luoh et al 1363 Barcos M and Henderson IC. (1994). N. Engl. J. Med., 330, Savelyeva L and Schwab M. (2001). Cancer Lett., 167, 115–123. 1260–1266. Sbrissa D, Ikonomov OC and Shisheva A. (1999). J. Biol. Nakajima M, Sawada H, Yamada Y, Watanabe A, Tatsumi Chem., 274, 21589–21597. M, Yamashita J, Matsuda M, Sakaguchi T, Hirao T and Seeger RC, Brodeur GM, Sather H, Dalton A, Siegel SE, Nakano H. (1999). Cancer, 85, 1894–1902. Wong KY and Hammond D. (1985). N. Engl. J. Med., 313, Nessling M, Solinas-Toldo S, Wilgenbus KK, Borchard F and 1111–1116. Lichter P. (1998). Genes Chromosomes Cancer, 23, 307–316. Seshadri R, Firgaira FA, Horsfall DJ, McCaul K, Setlur V and Paik S, Bryant J, Park C, Fisher B, Tan-Chiu E, Hyams D, Kitchen P. (1993). J. Clin. Oncol., 11, 1936–1942. Fisher ER, Lippman ME, Wickerham DL and Wolmark N. Shayesteh L, Lu Y, Kuo WL, Baldocchi R, Godfrey T, Collins (1998). J. Natl. Cancer Inst., 90, 1361–1370. C, Pinkel D, Powell B, Mills GB and Gray JW. (1999). Nat. Pauletti G, Dandekar S, Rong H, Ramos L, Peng H, Seshadri Genet., 21, 99–102. R and Slamon DJ. (2000). J. Clin. Oncol., 18, 3651–3664. Shibasaki Y, Ishihara H, Kizuki N, Asano T, Oka Y and Pietras RJ, Arboleda J, Reese DM, Wongvipat N, Pegram Yazaki Y. (1997). J. Biol. Chem., 272, 7578–7581. MD, Ramos L, Gorman CM, Parker MG, Sliwkowski MX Slamon DJ, Clark GM, Wong SG, Levin WJ, Ullrich A and and Slamon DJ. (1995). Oncogene, 10, 2435–2446. McGuire WL. (1987). Science, 235, 177–182. Press MF, Pike MC, Chazin VR, Hung G, Udove JA, Slamon DJ, Godolphin W, Jones LA, Holt JA, Wong SG, Markowicz M, Danyluk J, Godolphin W, Sliwkowski M Keith DE, Levin WJ, Stuart SG, Udove J, Ullrich A and and Akita R. (1993). Cancer Res., 53, 4960–4970. Press MF. (1989). Science, 244, 707–712. Press MF, Pike MC, Hung G, Zhou JY, Ma Y, George J, Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton V, Dietz-Band J, James W, Slamon DJ and Batsakis JG. (1994). Bajamonde A, Fleming T, Eiermann W, Wolter J, Pegram Cancer Res., 54, 5675–5682. M, Baselga J and Norton L. (2001). N. Engl. J. Med., 344, Proctor AJ, Coombs LM, Cairns JP and Knowles MA. (1991). 783–792. Oncogene, 6, 789–795. Szepetowski P, Courseaux A, Carle GF, Theillet C and Rameh LE, Tolias KF, Duckworth BC and Cantley LC. Gaudray P. (1992). Oncogene, 7, 751–755. (1997). Nature, 390, 192–196. Szepetowski P, Perucca-Lostanlen D and Gaudray P. (1993). Rao VD, Misra S, Boronenkov IV, Anderson RA and Hurley Genomics, 16, 745–750. JH. (1998). Cell, 94, 829–839. Theillet C, Adnane J, Szepetowski P, Simon MP, Jeanteur P, Ravdin PM and Chamness GC. (1995). Gene, 159, 19–27. Birnbaum D and Gaudray P. (1990). Oncogene, 5, 147–149. Roelofs H, Schuuring E, Wiegant J, Michalides R and Thor AD, Berry DA, Budman DR, Muss HB, Kute T, Giphart-Gassler M. (1993). Genes Chromosomes Cancer, 7, Henderson IC, Barcos M, Cirrincione C, Edgerton S, Allred 74–84. C, Norton L and Liu ET. (1998). J. Natl. Cancer Inst., 90, Rolitsky CD, Theil KS, McGaughy VR, Copeland LJ and 1346–1360. Niemann TH. (1999). Int. J. Gynecol. Pathol., 18, 138–143. Tolias KF, Rameh LE, Ishihara H, Shibasaki Y, Chen J, Ross JS and Fletcher JA. (1999). Semin. Cancer Biol., 9, 125–138. Prestwich GD, Cantley LC and Carpenter CL. (1998). J. Ross JS, Sheehan C, Hayner-Buchan AM, Ambros RA, Biol. Chem., 273, 18040–18046. Kallakury BV, Kaufman R, Fisher HA and Muraca PJ. Walch AK, Zitzelsberger HF, Bink K, Hutzler P, Bruch J, (1997). Hum. Pathol., 28, 827–833. Braselmann H, Aubele MM, Mueller J, Stein H, Siewert JR, Rozenvayn N and Flaumenhaft R. (2001). J. Biol. Chem., 276, Hoﬂer H and Werner M. (2000). Mod. Pathol., 13, 814–824. 22410–22419. Wu G, Sinclair C, Hinson S, Ingle JN, Roche PC and Couch Rubie H, Hartmann O, Michon J, Frappaz D, Coze C, FJ. (2001). Cancer Res., 61, 4951–4955. Chastagner P, Baranzelli MC, Plantaz D, Avet-Loiseau H, Wu GJ, Sinclair CS, Paape J, Ingle JN, Roche PC, James CD Benard J, Delattre O, Favrot M, Peyroulet MC, Thyss A, and Couch FJ. (2000). Cancer Res., 60, 5371–5375. Perel Y, Bergeron C, Courbon-Collet B, Vannier JP, Zhang X, Loijens JC, Boronenkov IV, Parker GJ, Norris FA, Lemerle J and Sommelet D. (1997). J. Clin. Oncol., 15, Chen J, Thum O, Prestwich GD, Majerus PW and Anderson 1171–1182. RA. (1997). J. Biol. Chem., 272, 17756–17761.

Oncogene