Oncogene (2010) 29, 5785–5795 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc ORIGINAL ARTICLE LAPTM4B: A novel cancer-associated motivates multidrug resistance through efflux and activating PI3K/AKT signaling

LLi1, XH Wei1, YP Pan1, HC Li, H Yang, QH He, Y Pang, Y Shan, FX Xiong, GZ Shao and RL Zhou

Department of Cell Biology, School of Basic Medical Sciences, Peking University, Beijing, China

LAPTM4B (lysosomal transmembrane 4 beta) Introduction is a newly identified cancer-associated gene. Both of its mRNA and the encoded LAPTM4B-35 protein are LAPTM4B (lysosomal protein transmembrane 4 beta) is significantly upregulated with more than 70% frequency a newly identified cancer-associated gene (NM_018407, in a wide variety of cancers. The LAPTM4B-35 level in Gene ID ¼ 55353) mapped to 8q22. 1, and cancer is evidenced to be an independent prognostic factor encodes a 35 kDa tetratransmembrance glycoprotein and its upregulation promotes cell proliferation, migration (Shao et al., 2003). It has been reported that the and invasion, as well as tumorigenesis in nude mice. LAPTM4B mRNA and the LAPTM4B-35 protein In contrary, knockdown of LAPTM4B-35 expression by detected by western blot or immunohistochemistry are RNA interference (RNAi) reverses all of the above markedly upregulated in a wide variety of cancers, malignant phenotypes. We herein reveal a new role of including HCC (more than 85%; Shao et al., 2003; Peng LAPTM4B-35 in promoting multidrug resistance of cancer et al., 2005; Yang et al., 2009), gallbladder carcinoma cells. Upregulation of LAPTM4B-35 motivates multidrug (76%; Zhou et al., 2006), extra-hepatic cholangiocarci- resistance by enhancement of efflux from cancer cells of a noma (72%; Zhou et al., 2008), lung cancer (88%), variety of chemodrugs with variant structures and proper- colon cancer (67%), uterus cancer (68%), breast cancer ties, including , paclitaxel and cisplatin (50%; Kasper et al., 2005) and ovary cancer (69%; Yang through colocalization and interaction of LAPTM4B-35 et al., 2008), as well as in adrenocorticotrophin-secreting with multidrug resistance (MDR) 1 (P-glycoprotein, adenomas and non-functioning pituitary adenomas P-gp), and also by activation of PI3K/AKT signaling (Morris et al., 2005). Also, LAPTM4B-35 is highly pathway through interaction of PPRP motif contained in expressed in a large number of cancer cell lines the N-terminus of LAPTM4B-35 with the p85a regula- (Shao et al., 2003; Liu et al., 2004). It has been revealed tory subunit of PI3K. The specific inhibitors of PI3K and that the levels of upregulated mRNA and LAPTM4B-35 knockdown of LAPTM4B-35 expression by RNAi eliminate protein are correlated significantly with pathological the multidrug resistance effect motivated by upregulation grades/differentiation of cancers and prognosis of of LAPTM4B-35. In conclusion, LAPTM4B-35 motivates patients with HCC, gallbladder carcinoma, extra-hepa- multidrug resistance of cancer cells by promoting drug tic cholangiocarcinoma and ovary cancer (Shao et al., efflux through colocalization and interaction with P-gp, 2003; Kasper et al., 2005; Zhou et al., 2006, 2008; Yang and anti-apoptosis by activating PI3K/AKT signaling. et al., 2008, 2009). Overexpression of LAPTM4B-35 These findings provide a promising novel strategy for by transfection of LAPTM4B complementary DNA sensitizing chemical therapy of cancers and increasing (cDNA) promotes deregulated cell proliferation, the chemotherapeutic efficacy through knockdown migration and invasion (Liu et al., 2009), as well as LAPTM4B-35 expression by RNAi. tumorigenesis of mouse NIH3T3 cells (He et al., 2003), Oncogene (2010) 29, 5785–5795; doi:10.1038/onc.2010.303; growth and metastasis of human HCC xenografts in published online 16 August 2010 nude mice (Yang et al., 2010). In contrary, knockdown of LAPTM4B-35 by RNA interference (RNAi) Keywords: LAPTM4B; multidrug resistance; drug reverses all of the above malignant phenotypes (Yang efflux; PI3K-AKT signaling et al., 2010). LAPTM4B-35 shows a 46% homology to a murine lysosomal tetra-transmembrane protein, mLAPTM4A (Hogue et al., 1996) at the amino acid level (Shao et al., 2003). It was reported that the mLAPTM4A Correspondence: Professor RL Zhou, Department of Cell Biology, functions in the transport of nucleosides and/or nucleo- School of Basic Medical Sciences, Peking University, 38 Xue Yuan side derivatives between the cytosol and the lumen of Road, Beijing 100191, China. an intracellular membrane-bound compartment and E-mail: rlzhou@ bjmu.edu.cn 1These authors contributed equally to this work. may confer multidrug resistance upon expression in Received 3 June 2009; revised 28 May 2010; accepted 14 June 2010; Saccharomyces cerevisiae (Hogue et al., 1996, 1999; published online 16 August 2010 Cabrita et al., 1999). LAPTM4B motivates multidrug resistance LLiet al 5786 MDR is most commonly associated with human LAPTM4B-35 has a role in enhancing multidrug cancers and is the major obstacle to cancer chemo- resistance. therapy (Ambudkar et al., 2003; Larsen et al., 2000). This study aimed to explore whether LAPTM4B Increasing Rhodamine-123 efflux in L02 cells by associates with MDR and to reveal the mechanism upregulation of LAPTM4B-35, but increasing underlying this phenotype. LAPTM4B-overexpressing Rhodamine-123 retain in HeLa cells by knock down and -knocking down cells were applied to compare their of LAPTM4B-35 expression potentials of drug resistance to doxorubicin, paclitaxel To compare Rhodamine-123 efflux in LAPTM4B-35- and cisplatin. The results demonstrated that overexpres- overexpressing and knockdown cells, L02 cell line, a cell sion of LAPTM4B-35 motivates multidrug resistance by line originated from human normal liver and expressed enhancement of efflux from cancer cells of various relatively low level of LAPTM4B-35, and HeLa cell cytotoxic drugs. The activation of PI3K/AKT signaling line, a human cervical cancer cell line which expressed pathway, another important mechanism for MDR high level of LAPTM4B-35, were used, respectively. (Lee et al., 2004), was also analyzed. The association The results showed that the loaded fluorescence dye lost of activation of PI3K/AKT signaling pathway with rapidly in L02-AE cells, but almost totally retained LAPTM4B-35 upregulation was revealed via increased in L02-MOCK cells (Figure 2a) during a period of p-AKT, and effects of PI3K inhibitors and dominant 120 min, indicating that the Rhodamine-123 efflux is negative mutation of PI3K p85a. Interaction between obviously promoted by LAPTM4B-35 overexpression. LAPTM4B-35 and PI3K p85a was evidenced by co- Adversely, the loaded fluorescence dye almost totally immunoprecipitation (Co-IP) and GST-pulldown ana- retained at the same level in HeLa-RNAi cells despite lyses. The binding sites were primarily identified by site- lost rapidly in HeLa-MOCK cells during a period of specific mutations. In conclusion, LAPTM4B motivates 120 min (Figure 2b), indicating that the Rhodamine-123 multidrug resistance by enhancement of efflux from efflux is prevented significantly by LAPTM4B-35 down- cancer cells of a variety of cytotoxic drugs and regulation. activation of PI3K/AKT signaling pathway. The time course of Rhodamine-123 efflux was also quantitatively detected by flow cytometry. The results (Figure 2e) showed that the fluorescence retention curve was significantly higher in HeLa-RNAi cell population Results than that in the MOCK cell population, indicating that LAPTM4B-35-knockdown by RNAi can increase drug Expression of LAPTM4B-35 in wild-type and retention inside cancer cells. transfected HeLa cell lines The expression of LAPTM4B-35 protein was analyzed by western blot using a purified polyclonal antibody Colocalization of LAPTM4B-35 and directed a 10 amino acid-peptide in EC2 extracellular MDR1/P-glycoprotein (P-gp) region of LAPTM4B-35. Figure 1a shows that the The most common ATP-dependent membrane efflux LAPTM4B-35 is upregulated in HeLa-AE cells, while it transporters responsible for hydrophobic drug resis- is markedly knocked down in HeLa-small hairpin tance are MDR1/P-gp and multidrug resistance-related RNA (shRNA) cells, when compared with their control protein (MRP) 1, which contain two ATP-binding HeLa-MOCK cells, respectively. cassettes in the cytoplasmic region (Ambudkar et al., 2003). As LAPTM4B-35 does not contain ATP-binding cassette, it is unlikely that it functions as an efflux Increasing drug-resistance in HeLa cells by upregulation pump itself. To investigate whether the MDR effect of LAPTM4B-35, but increasing sensitivity to motivated by LAPTM4B-35 upregulation is associated chemotherapeutic drugs by knock down of with the ATP-dependent membrane efflux transporter, LAPTM4B-35 expression Laser Scanning Confocal Microscopic analysis was To explore the effects of LAPTM4B-35 on drug applied with anti-LAPTM4B-35 antibody and anti- sensitivity, we first examined the viability of HeLa cells, MDR1 or anti-MRP1 antibody. The results showed which express either high or low level of LAPTM4B-35 that LAPTM4B-35 is colocalized with MDR1/P-gp in a panel of three chemotherapeutic agents with variant (Figure 2c), but not MRP1 (Supplementary data), at structures, properties and known different mechanisms plasma membrane and endomembrane organelles, of action. As compared with the control HeLa-MOCK1 suggesting these two may directly interact or cells, the LAPTM4B-35-overexpressing HeLa-AE cells be co-transported inside cell. showed a higher viability curve, indicating significantly increased resistance to doxorubicin, paclitaxel and Interaction of LAPTM4B-35 and MDR1/P-gp cisplatin (Figure 1b). In contrary, the HeLa-RNAi To further investigate the association of LAPTM4B-35 cells showed a lower viability curve, indicating signifi- with MDR1/P-gp, co-immunoprecipitation was per- cantly increased sensitivity to these three chemother- formed (Figure 2d). P-gp or MRP1 was immuno- apeutic agents (Figure 1c). Similar results were also precipitated with specific pAb, and western blots obtained from PC-3 prostate cancer cell line (shown of immunoprecipitates were carried out using a pAb in Supplementary data). These results suggest that directed against LAPTM4B-35. LAPTM4B-35 was

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5787

Figure 1 Overexpression of LAPTM4B-35 increases drug-resistance in HeLa cells. (a) Overexpression or knockdown of LAPTM4B- 35 was identified by western blot analysis. HeLa cells were transfected with pcDNA3-Mock1, pcDNA3-LAPTM4B-AE or pSilencer- Mock2, pSilencer-LAPTM4B-shRNA plasmids, and the transfectant was designated as Mock1, AE, Mock2, and RNAi, respectively. The transfected cells were harvested 48 h after transfection for western blot analyses using purified anti-LAPTM4B-EC2 antibody. (b) Viability curves of LAPTM4B-35-overexpressing HeLa-AE cells in the presence of various anticancer chemotherapeutic drugs. (c) Viability curves of LAPTM4B-35-knocking down HeLa-RNAi cells in the presence of various anticancer chemotherapeutic drugs. The HeLa-MOCK1, HeLa-AE, HeLa-MOCK2 and HeLa-RNAi cells were exposed to increasing concentrations of doxorubicin, paclitaxel or cisplatin, respectively, for 48 h. Cell viabilities were determined by acid phosphatase assay as described in Materials and methods. Results were expressed as a mean±s.d. of viable cell percentage in triplicates from three independent experiments. (*Po0.05 vs Mock; **Po0.01 vs Mock, Student’ t-test). detected in immune complexes immunoprecipitated with (Tazzari et al., 2007). PI3K/AKT signaling pathway P-gp antibody, but not MRP1 antibody. This experiment is activated in HeLa cells (Barancı´k et al., 2006). indicates that there is a physical interaction between P-gp To explore the additional molecular mechanism of and LAPTM4B-35. LAPTM4B-35 for MDR besides drug efflux, we detected the effects of PI3K inhibitors on LAPTM4B- 35-associated MDR. The results showed that in the The LAPTM4B-35-motivated MDR is eliminated presence of LY294002 or Wortmannin, the specific by PI3K inhibitors inhibitor of PI3K, the enhanced MDR effect of HeLa- PI3K/AKT signaling pathway, the most important AE cells was eliminated (Figure 3e and f compared mechanism for keeping cell survival, is also responsible with Figure 1b), implicating the involvement of PI3K for MDR and has been paid more and more attention signaling in LAPTM4B-35-motivated MDR.

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5788 Association of LAPTM4B-35 expression levels with showed that much higher levels of p-AKT appeared in the activation of PI3K/AKT signaling pathway HeLa-AE cells compared with HeLa-MOCK1 cells, but To explore the relationship between the expression levels no difference in the level of total AKT protein was found. of LAPTM4B-35 and the activation of PI3K/AKT Accordingly, much lower levels of p-AKT473 appeared signaling pathway, phosphorylated Akt (p-AKT) S473, in HeLa-RNAi cells compared with HeLa-Mock2 cells. the fully activated indication, was detected in HeLa-AE In addition, Figure 3a (right) showed that treatment and HeLa-RNAi cells by western blot. Figure 3a (left) of the cells with the PI3K specific inhibitors, either

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5789 LY294002 or Wortmannin greatly reduced the p-AKT473 negative p85a (DN-p85a) encoded by this mutant would in either HeLa-Mock or HeLa-AE cells, and eliminated be inactive for activating PI3K/AKT signaling. The HeLa the elevation of p-AKT473 in HeLa-AE cells as compared cells were co-transfected with pcDNA-AE plasmids and with HeLa-Mock cells. These results suggest that the p85a-DN plasmids. As shown in Figure 3d, the p-AKT473 activation of PI3K/AKT signaling pathway is associated was reduced significantly in the co-transfectants of with LAPTM4B-35 expression levels, in other words, the dominant-negative p85a with Mock or AE plasmids LAPTM4B-35 has a role in activation of PI3K/AKT when compared with the mono-transfectants with signaling. Mock or AE plasmids, indicating that the phosphoryla- tion/activation of AKT473 in either HeLa-Mock or HeLa-AE cells is dependent on PI3K activation The interaction between LAPTM4B-35 and PI3K p85a through interaction between LAPTM4B-35 and p85a regulatory subunit evidenced by GST-pulldown and Co-IP subunit. Structurally, the N-terminal tail in cytoplasmic region of LAPTM4B-35 contains a proline-rich motif PPRP, which is identical to that of the regulatory subunit Upregulation of LAPTM4B-35 inhibits chemodrugs- p85a of PI3K (Shao et al., 2003). The proline-rich motif induced apoptosis, but knockdown of LAPTM4B-35 can be recognized and bound by SH3 domain of a promotes apoptosis in HeLa-RNAi cells variety of cytoplasmic signaling molecules, including Failure of apoptosis is one of the mechanisms for MDR PI3K regulatory subunit p85a. The interaction between (Tsuruo et al., 2003). The results from fluorescence- LAPTM4B-35 and PI3K p85a was first evidenced by activated cell sorting analysis with annexin-V apoptosis GST-pulldown in vitro. The GST-N1-99 and PPRP- kit showed that the LAPTM4B-35-upregulated HeLa-AE mutated (P12, 13, 15A) GST-N1-99 (PA) were applied cells were resistant to apoptosis induced by doxorubicin to pull down the p85a-HA, which ectopically expressed (Figure 4a, up and Figure 4b, left), paclitaxel (Figure in HeLa cells, and the pull down p85a-HA was 4a, b, middle) and cisplatin (Figure 4a, bottom and identified by anti-HA antibody. The result (Figure 3b) Figure 4b, right). However, LAPTM4B-35-knocking showed that PI3K p85a-HA was detectable in GST-N1- down HeLa-RNAi cells were more sensitive to apoptosis 99 pulldown. Although significantly reduced PI3K induced by doxorubicin (Figure 4a, up and Figure 4b, p85a-HA was not detected in a PPRP-mutated GST- left), paclitaxel (Figure 4a, b, middle) and cisplatin N1-99 (PA) pulldown, suggesting that the LAPTM4B- (Figure 4a, bottom and Figure 4b, right). 35 interacts with p85a and this interaction contributes Caspase 3 is the effector of caspase-dependent partially to the activation of PI3K. To demonstrate the apoptosis. The caspase 3 activity can be measured by interaction between LAPTM4B-35 and PI3K p85a generation of cleaved poly (ADP-ribose) polymerase, in vivo, Co-IP analysis was performed in cells expressing the substrate of caspase 3. Our results (Figure 4c) herein wild-type LAPTM4B-35 and PPRP-mutated (P12, 13, showed that in the presence of various chemotherapeutic 15A) LAPTM4B-35 (PA). As shown in Figure 3c, the drugs the cleaved poly (ADP-ribose) polymerase interaction of LAPTM4B-35 and PI3K p85a existed in appeared in HeLa-Mock cells, but attenuated dramati- lysate from HeLa-AF cells, which expressed wild-type cally in HeLa-AE cells and recovered in HeLa-RNAi LAPTM4B-35, but reduced in lysate from PPRP- cells. These results indicate that upregulation mutated HeLa-AF (PA) cells. These results suggest that of LAPTM4B-35 resists caspase-dependent apoptosis the PPRP motif in N-terminus of LAPTM4B-35 may be induced by various chemotherapeutic drugs, whereas the binding site for PI3K p85a regulatory subunit. downregulation of LAPTM4B-35 by RNAi can reverse To further demonstrate the interaction of LAPTM4B-35 the anti-apoptotic effect exerted by LAPTM4B-35- and p85a regulatory subunit is associated with AKT upregulaton. These results provide further evidence that phosphorylation/activation, we constructed a p85a-DN upregulation of LAPTM4B-35 motivates MDR pheno- mutant, in which the PI3K p110 catalytic subunit- type by evasion from drug-induced apoptosis via binding site of p85a was deleted. As such, the dominant caspase-dependent pathway.

Figure 2 Overexpression of LAPTM4B-35 increase the efflux of Rhodamine-123 in HeLa cells. (a, b) Drug efflux assay by Laser Scanning Confocal Microscopy (LSCM). (a) The L02 cells were infected with Ad-Null or Ad-LAPTM4B-AE, and designated as L02-Mock and L02-AE, respectively. (b) The HeLa cells were transfected with pSilencer or pSilencer-LAPTM4B-shRNA plasmids, and designated as HeLa-Mock2 and HeLa-RNAi, respectively. Cells were loaded with Rhodamine-123 for 20 min at 37 1C, and then observed and photographed for 120 min (up) under a Laser Scanning Confocal Microscope at an excitation wavelength of 488 nm. The fluorescence intensity in each cell was monitored and recorded separately as each curve (down). (e) Time course of Rhodamine-123 retention in LAPTM4B-35 knocking down HeLa-RNAi cells. The cells were loaded with Rhodamine-123 and analyzed by a FACSCalibur flow cytometer with excitation using a 488 nm argon ion laser. The Rhodamine-123 retention in HeLa-RNAi cells is significantly higher than that of MOCK control (*Po0.05 vs Mock, Student’ t-test). (d) Interaction of LAPTM4B-35 with MDR1/ P-gp detected by Co-IP. The anti-P-gp pAb, anti-LAPTM4B-35-N10 pAb and anti-MRP1 pAb were used for immunoprecipitation, rabbit IgG pAb was used as a control antibody. The anti-LAPTM4B-35-N10 pAb, anti P-gp mAb and anti-MRP1 pAb were used for western blot. The LAPTM4B-35 can interact with P-gp but not MRP1 expressed in HCC tissue. (c) Triple-staining immunofluorescence and LSCM analysis of LAPTM4B-35 colocalization with P-gp. Yellow color is the overlapping of green (P-gp) and red (LAPEM4B-35) signals. Left, LAPTM4B-35 is colocalized with P-gp mainly at endomembrane organelles in HeLa cells (Bar, 8 mm). Right, LAPTM4B-35 is colocalized with P-gp mainly at plasma membrane in PC-3 cells when coverslides coated with fibronectin (Bar, 20 mm).

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5790 Discussion includes several hundred different transport proteins, which share a structural similarity consisting of four Resistance to different chemotherapeutic drugs (multi- domains: two transmembrane (T) domains and two drug resistance, MDR) is a major clinical obstacle in the cytosolic ATP-binding (A) domains. The drug efflux treatment of cancers. Classical MDR is the consequence mediated by multidrug transporters dependents on the of overexpression of transporter proteins belonging ATP-binding domains in transporter molecule and the to the ATP binding cassette (ABC) family, such as energy released by ATP hydrolysis, and thus reducing MDR1 (P-gp), MRPs and so on. The ABC superfamily intracellular drug concentrations to sublethal levels

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5791 (Larsen et al., 2000; Gottesman, 2002). Herein, we It is generally thought that the overexpression of evidenced that LAPTM4B-35-upregulation, which MDR1 (P-gp) encoded by mdr1 gene is a predominant occurs in a large number of solid cancers, can increase cause of the acquisition of an MDR phenotype (Tsuruo efflux of a variety of chemodrugs, including doxorubi- et al., 2003). The relationship between PI3K/AKT signaling cin, paclitaxel and cisplatin, whereas LAPTM4B-35- and some transporter proteins, such as P-gp and MRP, knockdown by RNAi can decrease efflux and increase have recently been revealed. For example, PI3K/AKT retention of drugs. Thus, LAPTM4B-35 is associated activity, analyzed by phosphatidylinositol trisphosphate with MDR. However, there is no ATP binding sites in production and p-AKT expression, is higher in the the molecular structure of LAPTM4B from bioinfor- resistant cell lines than in the sensitive one; and matics analysis. Therefore, it might be less possibility inhibition with PI3K inhibitor, Wortmannin or that LAPTM4B-35 itself is a kind of multidrug LY294002, promotes apoptosis in the resistant cell lines transporter, despite LAPTM4A is thought to function and inhibits P-gp function and drug efflux (Garcı´a et al., as a transporter protein (Cabrita et al., 1999; Hogue 2008). Our finding that LAPTM4B-35 overexpression et al., 1999). motivates MDR, supports the concept that the ATP- Another mechanism for MDR involves in alterations dependent drug efflux via P-gp is associated with anti- in the apoptotic response associated with PI3K/AKT apoptosis via activation of PI3K/AKT signaling. In signaling. More and more experiments have recently addition, PI3K activation leads to mdr1 gene expression revealed that the activation of PI3K/AKT signaling and subsequent chemoresistance in advanced prostate pathway can regulate or enhance multidrug resistance cancer cells (Lee et al., 2004); PI3K/AKT signaling (Knuefermann et al., 2003; Abdul-Ghani et al., 2006; controls MRP1 expression in human acute myelogenous Tazzari et al., 2007). This signaling pathway is activated leukemia blasts (Tazzari et al., 2007). The P-gp in cervical cancer cells and the PI3K/AKT signaling expression was decreased by treatment with LY294002 inhibitor LY294002 can efficiently inhibits drug resis- in mouse L1210 leukemia cell line (Choi et al., 2007) and tance and cell growth, and induced apoptosis in HeLa in human HeLa cell line, also the HeLa cell growth cells (Barancı´k et al., 2006). In this study, we demon- and induced-apoptosis were affected as well (Zhang strate via HeLa cell models that LAPTM4B-35-upregu- et al., 2008). However, the upregulation of P-gp in lation can strongly activate/phosphorylate AKT473 by LAPTM4B-35 overexpressing HeLa and PC-3 cells was directly binding to PI3K p85 a subunit via the PPRP not found. motif in the N-terminus of LAPTM4B-35 protein as A number of multidrug-resistant cancer cells have evidenced by GST-pulldown and Co-IP analyses; these been shown to have acquired an increased capacity to effects can be inhibited by PI3K inhibitors, LY294002 sequester weakly basic anticancer drugs in their acidic and Watmannin. Conversely, LAPTM4B-35-knock- vesicles, including lysosomes and endosomes, relative to down by RNAi can dramatically inhibit PI3K/AKT drug-sensitive counterparts. Lysosomal sequestration signaling as revealed in this study. In addition, has a significant role in the emergence of MDR as chemotherapeutic drugs induce a series of cellular lysosomes effectively sequester drugs away from their responses that impact on tumor cell proliferation and cellular targets (Duvvuri et al., 2005; Fu and Roufogalis, survival/apoptosis. In fact, several lines of evidence have 2007). The study on subcellular distribution of the basic suggested a direct correlation between alteration in drugs revealed that lysosome has the highest affinity survival pathways and chemoresistance (Tsuruo et al., for the basic drugs and responsible for biphasic uptake. 2003). And some components of these pathways have The study on intracellular traffic of human P-gp been pointed as critical targets for cancer intervention. revealed that a small fraction of the total P-gp-EGFP We also demonstrate herein that LAPTM4B-35- expressed was localized intracellularly and was present upregulation can inhibit drug-induced apoptosis by in early endosome and lysosome compartments (Fu and caspase-depentent pathway. Therefore, it is supposed Roufogalis, 2007). Membrane proteins may traffic to the that one of the mechanisms for LAPTM4B-35-upregu- cell plasma membrane via two distinct pathways. In the lation motivated MDR is anti-apoptosis by activating first, a constitutive/default pathway, membrane protein PI3K/AKT signaling pathway. is incorporated into vesicles and directly moves from

Figure 3 LAPTM4B-35 motivates MDR through activation of the PI3K/AKT signaling pathway. (a) LAPTM4B-35-enhanced AKT phosphorylation determined by western blot analysis. (b) Interaction of p85a subunit with wild-type LAPTM4B-35 detected by GST- pulldown. GST-LAPTM4B-N1-99 and the PPRP-mutated GST-LAPTM4B-N1-99 (PA) were applied to pull down the p85a-HA, which ectopically expressed in HeLa cells. The pull down p85a-HA was identified by anti-HA antibody. (c) Interaction of p85a subunit with wild-type LAPTM4B-35 detected by Co-IP analysis. HeLa cells were co-transfected with pcDNA3-p85a-HA plasmids and pcDNA3-LAPTM4B-AF (the pcDNA3-LAPTM4B-AE with FLAG tag) or pcDNA3-LAPTM4B-AF (PA) plasmids. After serum starvation for 18 h, the transfected cells were incubated in the presence or absence of fetal calf serum (FCS) for 15 min, and then lysed for Co-IP analysis. The anti-FLAG antibody was used for immunoprecipitation. The anti-HA antibody was used for western blot. The p85a-HA can interact with wild-type LAPTM4B-35 in HeLa-AF cells, but not the mutated LAPTM4B-35 (PA) in HeLa-AF (PA) cells. (d) Inhibition of AKT phosphorylation in HeLa-AE cells by PI3K p85a dominant negative strategy. HeLa cells were co-transfected with pcDNA3-AE plasmids and pcDNA3-DNp85a plasmids, in which the PI3K p110-binding site of p85a was deleted, and thus DN-p85a is inactive for activating PI3K/AKT signaling. After 18 h starvation by serum-withdrawing from the cell culture medium, the cells were stimulated with 20% FCS for 15 min, then phophorylated AKT was measured by western blot. (e, f) Viability curves of LAPTM4B-35- upregulating HeLa-AE cells in the presence of LY294002 and Wortmannin, respectively.

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5792

Figure 4 Upregulation of LAPTM4B-35 inhibits chemodrugs-induced apoptosis. (a) Flow cytometry analysis of apoptosis by Annexin V and PI staining. The HeLa-MOCK1, HeLa-AE, HeLa-MOCK2 and HeLa-RNAi cells were incubated with 1 mmol/l doxorubicin (up), 0.05 mmol/l paclitaxel (middle) or 3 mmol/l cisplatin (bottom), respectively. After 48 h incubation, the cells were harvested and analyzed by FACS Calibur flow cytometer. (b) Column diagrams of apoptotic cells in percentage. (*Po0.05 vs Mock; **Po0.01 vs Mock). (c) Western blot analysis of poly (ADP-ribose) polymerase (PARP).

Golgi to the plasma membrane along the actin cently, the study on human breast cancer MCF-7 cells cytoskeleton. In the second pathway, traffic is through revealed that P-gp-EGFP traffic was shown to occur via an intracellular endosomal system, with protein vesicles early endosome transport to the plasma membrane. being transported to endosome compartments to form Subsequent movement of P-gp-EGFP away from the an intracellular pool, followed by further transport to plasma membrane occurred by endocytosis to the early the plasma membrane (Fu and Roufogalis, 2007). For endosome and lysosome. However, the detailed mechan- example, Kim et al. (1997) suggested that there was an ism of P-gp traffic has been little studied and is still intracellular pool of P-gp and that P-gp recycled unclear (Fu and Roufogalis, 2007). As note, the between the plasma membrane and the intracellular LAPTM4B-35 protein possesses typical lysosome-tar- pool, indicating a possible endocytosis pathway. Re- geting signals at both N- and C-terminal cytoplasmic

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5793 tails, and also the endocytic adapter protein (AP) 1 et al., 2003) and Wei (Yang et al., 2010) in our lab, respectively. binding site (Shao et al., 2003). It has been evidenced by Transient transfection was conducted with pcDNA3, pcDNA3- Laser Scanning Confocal Microscopy in our lab that LAPTM4B-AE (AE) or pSilencer, pSilencer-LAPTM4B-shRNA LAPTM4B-35 distributed at plasma membrane, endo- (shRNA) plasmids (Yang et al., 2010). The resulted transfec- somes and lysosomes (data not shown). In this study, we tants are designated as HeLa-MOCK1, HeLa-AE, HeLa- evidenced the colocalization of LAPTM4B-35 and P-gp MOCK2 and HeLa-RNAi, respectively. After incubation for another 48 h, these cells were harvested for use in each on plasma membrane and endomembrane organelles, experiment. as well as the interaction of these two multi-transmem- L02 cells were infected by Ad-null and Ad-AE, an E1/E3- brane proteins. Thus, it is likely that LAPTM4B-35 deleted replication-deficient adenovirus type 5 vector, which may have important roles in driving endocytosis and contains the whole ORF of LAPTM4B cDNA, and labeled as trafficking of P-gp between plasma membrane and L02-MOCK and L02-AE, respectively. intracellular compartments. It is also possible that binding of LAPTM4B-35 to P-gp may contribute to Plasmid construction of PA mutant and p85a-DN mutant its activation by alternating P-gp conformation. Further The pcDNA3-AE and pcDNA3-AF plasmids that contains a studies to reveal the effects of LAPTM4B-35 on Flag-tag at the 30 end of pcDNA3-AE were used to construct trafficking or conformation of P-gp remain to be done. the mutant plasmids. Taken together, LAPTM4B-35, that promotes cancer For construction of pcDNA3-AF (PA) plasmids, growth and metastasis, also has critical roles in MDR LAPTM4B cDNA with P12, 13, 15A mutated (PA) DNA through increasing drug efflux by P-gp to reduce drug sequence was obtained by two-step PCR-amplification. First, two PCR-amplifications were carried out respectively using concentration in cytoplasm and activating PI3K/AKT 0 signaling to escape from drug-induced apoptosis. two pairs of primers Laptm4b-F-BamHI (5 -GGGATCCG CCACCATGACGTCACGGACTCGGGTC-30) and PA-1 Accordingly, LAPTM4B-35-knockdown by RNAi may (50-GAGGGCGCGGGCCGCACTCGGCCATGTGACCCG provide a promising novel strategy to overcome MDR AGT-30), PA-2 (50-CACATGGCCGAGTGCGGCCCGCGC in cancer . It is possible that LAPTM4B- CCTCCCCGTCC-30) and Laptm4b-R-EcoRI (50-CGGAATT 35 may be an important factor that involves in drug- CGGCAGACACGTAAGGTGGCGG-30). The two amplified transporter trafficking between plasma membrane and fragments were mixed as the template of the second PCR- intracellular compartments. amplification with Laptm4b-F-BamHI and Laptm4b-R-HindIII as primers. Then, the amplified fragment with P12, 13, 15A mutated sequence was cloned into pcDNA3-Flag plasmid and Materials and methods named as pcDNA3-AF (PA). For construction of pGEX-KG-N1-99 (PA) plasmids, Cell lines and antibodies N1-99 (PA) DNA sequence of LAPTM4B cDNA was PCR- 0 Human cervical carcinoma HeLa cell line was from ATCC amplified by using PA-F-BamHI (5 -CGGGATCCATGAC 0 0 (Manassas, VA, USA) and kept in our laboratory. L02 cell line GTCACGGACTCGGGTC-3 ) and PA-R-HindIII (5 -CCCA 0 was originally from human normal liver established by AGCTTTCACGTCCAGGGCGCGACCATC-3 ) as forward Institute of Biochemistry and Cell Biology, Shanghai Institutes and reverse primers, respectively, and pcDNA3-AF (PA) for Biological Sciences, Chinese Academy of Sciences, was used as the template. The amplified fragment was Shanghai, China. The cell lines were cultured in RPMI 1640 cloned into pGEX-KG plasmid and named as pGEX-KG- medium supplemented with 10% serum. Antigen-peptide N1-99 (PA). affinity chromatography purified rabbit anti-LAPTM4B-35 For construction of pcDNA3-p85a-HA and pcDNA3 p85a- polyclone antibodies were produced in our lab by immuniza- DN-HA plasmids, p85a cDNA was PCR-amplified by using 0 tion of a KLH-conjugated synthetic peptide composed of 10 p85a-F-BamHI (5 -GGGATCCGCCACCATGAGTGCTGA 0 0 amino acid residues located at EC2 or N-terminus. Anti-Akt, GGGGTACCAG-3 ) and p85a-R-XbaI(5-GCTCTAGAGC 0 anti-p-Akt and anti-poly (ADP-ribose) polymerase antibodies TTCCGCCGGGTCGCCTCTGCTGTG-3 )asforwardand were purchased from Cell Signaling (Boston, MA, USA). Anti- reverse primers, respectively, and the L02 cell line cDNA b-actin polyclonal antibodies were from Zhong Shan Biotech- was used as the template. The amplified fragment was cloned nology Co., LTD (Beijing, China). Anti-P-gp mAb C219 was into pcDNA3-HA plasmid and named as pcDNA3-p85a-HA. from Calbiochem (Merck KGaA, Darmstadt, Germany). Another two PCR-amplifications were carried out using two 0 Anti-P-gp pAb Mdr (H-241) was from Santa Cruz Biotechno- pairs of primers p85a-F-BamHI and p85a-DN-1 (5 -ATTGCC logy, Inc. (Santa Cruz, CA, USA). TTCACGGCGGCTTTGGATTTCCTGGGATGTGCGGG T-30), p85a-DN-2 (50-CAGGAAATCCAAAGCCGCCGTGA 0 Reagents AGGCAATGAGAAAGAAATA-3 ) and p85a-R-XbaI. The Doxorubicin was purchased from Wanle Pharmaceutical two amplified fragments were mixed as the template of the last (Shenzhen, China). Cisplatin was from XiLu Pharmaceutical PCR with p85a-F-BamHI and p85a-R-XbaI as primers. Thus, (Shandong, China). Paclitaxel was kindly provided by the amplified fragment, in which p85a479À513, the interacting Professor Wang Yongchao (Beijing Normal University, domain with p110a catalytic subunit of PI3K was deleted, Beijing, China). LY294002, Wortmannin, Rhodamine-123 and cloned into pcDNA3-HA plasmid and named as and acid phosphatase substrate were purchased from Sigma pcDNA3-p85a-DN-HA. (St Louis, MO, USA). Determination of viability curves Construction and transfection of the expressive plasmid Untransfected and transfected cells were exposed to various The expressive plasmids, pcDNA3-LAPTM4B-AE containing concentrations of chemotherapeutic agents and incubated LAPTM4B ORF cDNA and the RNAi plasmid containing for 48 h, the total viable cells were determined with the pSilencer-LAPTM4B-shRNA were constructed by Shao (Shao acid phosphatase assay (Yang et al., 1996). Cell viability was

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5794 then calculated by measuring the optical absorbance at Apoptosis assay 405 nm and normalizing the value with the value of the Cells were incubated with various chemotherapeutic reagents. untreated cells. Adherent and floating cells were both collected after 48 h incubation, apoptosis was analyzed using an Annexin V Time course of Rhodamine-123 retention determined Apoptosis Detection kit (Biosea, Beijing, China) with a by fluorescence-activated cell sorting FACSCalibur flow cytometer (Becton Dickinson). The LAPTM4B-35 knocking down HeLa-RNAi cells were loaded with Rhodamine-123 at a final concentration of Laser scanning confocal microscopy 1.5 mg/ml for 20 min at 37 1C. Aliquot of Rhodamine-loaded The HeLa or PC-3 cells were attached to coverslips coated cells was collected immediately or in every 20 min period with fibronectin or nothing, fixed with freshly prepared 4% of time and analyzed by a FACSCalibur flow cytometer formaldehyde in PBS for 10 min at room temperature. (Becton Dickinson, Franklin Lakes, NJ, USA) with excitation Permeabilized with 0.25% Triton X-100, blocked at room using a 488 nm argon ion laser. The percentages of fluorescent temperature for1 h in PBS containing 10% normal goat serum. intensity from Rhodamine-123 retained in the cells were Then incubated at 4 1C overnight with anti-P-gp mAb, washed plotted against time. three times for 5 min with PBS, then incubated at room temperature with FITC-conjugated goat anti-mouse IgG for Rhodamine-123 efflux assay 40 min. Following incubation, coverslips were washed three Briefly, cells were seeded on culture plates and incubated with times for 5 min with PBS, blocked with 5% bovine serum 1.5 mg/ml Rhodamine-123 for 20 min at 37 1C. Then the cells albumin, then incubated at 4 1C overnight with anti- analyzed by laser-scanning microscope at an excitation LAPTM4B35 pAb, washed three times for 5 min with PBS, wavelength of 488 nm. The photos were taken every 5 min. then incubated at room temperature with TRITC-conjugated goat anti-rabbit IgG for 40 min. Coverslips were then counter- stained for 10 min with DAPI. The observation was made In vitro GST-pulldown assay using a Leica TCS SP2 Laser Scanning Confocal Microscope Approximately 20 mg of the GST-GST-N1À99- and GST- (Leica Microsystems, Mannheim, Germany). N1À99(PA)-fusion proteins was incubated with 40 mlof glutathione beads and 200 mg of HeLa cell lysates at 4 1C with gentle rocking for 2 h. Then the glutathione beads were washed Statistical analysis four times with PBS containing NP-40 (0.5%), Triton X-100 Statistical analysis was performed using Student’s t-test. (0.5%), NaF (20 mM) and protease inhibitors to remove A P value o0.05 was considered statistically significant. nonspecifically bound proteins. Finally, the pulled down All experiments were repeated at least twice. proteins were subjected to SDS—polyacrylamide gel electro- phoresis and immunoblot analyses. Conflict of interest Co-IP assay Lysates from various transfected HeLa cells or HCC tissue The authors declare no conflict of interest. were pre-cleared with proteinA/G beads at 4 1C overnight before Co-IP. 500 mg of pre-cleared cell lysates were incubated with 1 mg of indicated antibody and 40 ml proteinA/G beads with gentle rocking at 4 1C overnight. After washing five times Acknowledgements with washing buffer (PBS buffer containing protease inhibi- tors), the precipitated proteins were subjected to SDS– This work was supported by grants from 211 and 985 polyacrylamide gel electrophoresis and immunoblot analysis Foundation of Peking University, National High-tech R&D using relative antibodies. Each experiment was repeated at Program (863 Program) (2006AA02A305), 248 major R&D least three times. program of Beijing (H020220020310) and NNSFC (90408018).

References

Abdul-Ghani R, Serra V, Gyo¨rffy B, Ju¨rchott K, Solf A, Dietel M Duvvuri M, Konkar S, Funk RS, Krise JM, Krise JP. (2005). A et al. (2006). The PI3K inhibitor LY294002 blocks drug export from chemical strategy to manipulate the intracellular localization of resistant colon carcinoma cells overexpressing MRP1. Oncogene 25: drugs in resistant cancer cells. Biochemistry 44: 15743–15749. 1743–1752. Fu D, Roufogalis BD. (2007). Actin disruption inhibits Ambudkar SV, Kimchi-Sarfaty C, Sauna ZE, Gottesman MM. (2003). endosomal traffic of P-glycoprotein-EGFP and resistance to P-glycoprotein: from genomics to mechanism. Oncogene 22: accumulation. Am J Physiol. Cell Physiol 292: 7468–7485. C1543–C1552. Barancı´k M, Boha´cov V, Sedla´k J, Sulov Z, Breier A. (2006). Garcı´a MG, Alaniz LD, Cordo Russo RI, Alvarez E, Hajos SE. (2008). LY294,002, a specific inhibitor of PI3K/Akt kinase pathway, PI3K/Akt inhibition modulates multidrug resistance and activates antagonizes P-glycoprotein-mediated multidrug resistance. Eur J NF-kappaB in murine lymphoma cell lines. Leuk Res 33: 288–296. Pharm Sci 29: 426–434. Gottesman MM. (2002). Mechanisms of cancer drug resistance. Cabrita MA, Hobman TC, Hogue DL, King KM, Cass CE. (1999). Annual Rev Med 53: 615–627. Mouse transporter protein, a membrane protein that regulates He J, Shao GZ, Zhou RL. (2003). Effects of the novel gene, cellular multidrug resistance, is localized to lysosomes. Cancer Res LAPTM4B, high expression in hepatocellular carcinoma on cell 59: 4890–4897. proliferation and tumorigenesis of NIH3T3 cells. J Peking Choi BH, Kim CG, Lim Y, Shin SY, Lee YH. (2007). Curcumin down- University (Health Sciences) 35: 348–352. regulates the multidrug-resistance mdr1b gene by inhibiting the Hogue DL, Ellison MJ, Young JD, Cass CE. (1996). Identification of a PI3K/Akt/NFkappaB pathway. Cancer Lett 259: 111–118. novel membrane transporter associated with intracellular mem-

Oncogene LAPTM4B motivates multidrug resistance LLiet al 5795 branes by phenotypic complementation in the yeast Saccharomyces Shao GZ, Zhou RL, Zhang QY, Zhang Y, Liu JJ, Rui JA et al. (2003). cerevisiae. J Biol Chem 271: 9801–9808. Molecular cloning and characterization of LAPTM4B, a novel gene Hogue DL, Kerby L, Ling V. (1999). A mammalian lysosomal upregulatively expressed in hepatocellular carcinoma. Oncogene 22: membrane protein confers multidrug resistance upon expression in 5060–5069. Saccharomyces cerevisiae. J Biolog Chem 274: 12877–12882. Tazzari PL, Cappellini A, Ricci F, Evangelisti C, Papa V, Grafone T Kasper G, Vogel A, Klaman I, Gro J, Petersen I, Weber B et al. (2005). et al. (2007). MRP1 & PI3K/Akt multidrug resistance-associated The human LAPTM4b transcript is up regulated in various types of protein 1 expression is under the control of the phosphoinositide 3 solid tumours and seems to play a dual functional role during kinase/Akt signal transduction network in human acute myelo- tumour progression. Cancer Lett 224: 93–103. genous leukemia blasts. Leukemia 21: 427–438. Kim H, Barroso M, Samanta R, Greenberger L, Sztul E. (1997). Tsuruo T, Naito M, Tomida A, Fujita N, Mashima T, Experimentally induced changes in the endocytic traffic of Sakamoto H et al. (2003). Molecular targeting therapy of P-glycoprotein alter drug resistance of cancer cells. Am J Physiol. cancer: drug resistance, apoptosis and survival signal. Cancer Sci Cell Physiol 273: C687–C702. 94: 15–21. Knuefermann C, Lu Y, Liu B, Jin W, Liang K, Wu L et al. (2003). Yang TT, Sinai P, Kain SR. (1996). An acid phosphatase assay for HER2/PI-3K/Akt activation leads to a multidrug resistance in quantifying the growth of adherent and nonadherent cells. Anal human breast adenocarcinoma cells. Oncogene 22: 3205–3212. Biochem 241: 103–108. Larsen AK, Escargueil AE, Skladanowski A. (2000). Resistance Yang Y, Yang H, McNutt MA, Xiong F, Nie X, Li L et al. (2008). mechanisms associated with altered intracellular distribution of LAPTM4B over expression is an independent prognostic marker in anticancer agents. Pharmacol Ther 85: 217–229. ovarian carcinoma. Oncol Rep 20: 1077–1083. Lee Jr JT, Steelman LS, McCubrey JA. (2004). Phosphatidylinositol Yang H, Xiong FX, Lin M, Yang Y, Nie X, Zhou RL. (2009). 30-kinase activation leads to multidrug resistance protein-1 expres- LAPTM4B-35 overexpression is a risk factor for tumor recurrence sion and subsequent chemoresistance in advanced prostate cancer and poor prognosis in hepatocellular carcinoma. J Cancer Res Clin cells. Cancer Res 64: 8397–8404. Oncol 136: 275–281. Liu XR, Zhou RL, Zhang QY, Zhang Y, Jin YY, Lin M et al. (2004). Yang H, Xiong F, Wei X, Yang Y, McNutt MA, Zhou R. (2010). Structure analysis and expressions of a novel tetratransmembrane LAPTM4B-35 promotes growth and metastasis of hepatocelluar protein, lysosome–associated protein transmembrane 4 B. World J carcinoma. Cancer Lett 294: 236–244. Gastroenterol 10: 1555–1559. Zhang XY, Zhang HY, Zhang PN, Lu X, Sun H. (2008). Elevated Liu X, Xiong F, Wei X, Yang H, Zhou R. (2009). LAPTM4B-35, a phosphatidylinositol 3-kinase activation and its clinicopathological novel tetratransmembrane protein and its PPRP motif play critical significance in cervical cancer. Eur J Obstetr, Gynecol, Reprod Biol roles in proliferation and metastatic potential of hepatocellular 139: 237–244. carcinoma cells. Cancer Sci 100: 2335–2340. Zhou L, Hea XD, Chen J, Cui QC, Qu Q, Rui JA et al. (2006). Morris DG, Musat M, Czirja´k S, Hanze´ly Z, Lillington DM, Korbonits Overexpression of LAPTM4B-35 closely correlated with clinico- M et al. (2005). Differential gene expression in pituitary adenomas by pathological features and post-resectional survival of gallbla- oligonucleotide array analysis. Eur J Endocrinol 153: 143–151. dder carcinoma. Eur J Cancer 43: 809–815. Peng C, Zhou RL, Shao GZ, Rui JA, Wang SB, Lin M et al. (2005). Zhou L, He XD, Cui QC, Zhou WX, Qu Q, Zhou RL et al. (2008). Expression of lysosome-associated protein transmembrane 4B-35 Expression of LAPTM4B-35: a novel marker of progression, in cancer and its correlation with the differentiation status invasiveness and poor prognosis of extrahepatic cholangio- of hepatocellular carcinoma. World J Gastroenterol 11: 2704–2708. carcinoma. Cancer Lett 264: 209–217.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene