**3. Functions of** *LAPTM4B***-encoding proteins and the mechanisms thereof**

Functional studies of *LAPTM4B*-encoding proteins and the mechanisms thereof have been performed via transgenic and gene knockdown techniques. *LAPTM4B* cDNA is introduced into plasmids with or without a FLAG tag, or into an E1/E3-deleted replication-deficient adenovirus type 5 (Ad5) vector. The LAPTM4B-35 expressing plasmids and cells are all designated in our publications as "AF" if they have a FLAG tag or as "AE" if they do not have a FLAG tag. Similarly, LAPTM4B-24 expressing plasmids and cells are designated as "BF" or "BE". LAPTM4B-35 knockdown is performed by transfection of shRNA-expressing plasmids (Yang et al., 2010c). These functional studies of LAPTM4B-35 and LAPTM4B-24 have determined that these two isoforms of LAPTM4B proteins have antagonistic functions, namely up-regulation of LAPTM4B-35 and LAPTM4B-24 respectively promotes and suppresses hepatocarcinogenesis and metastasis.

#### **3.1 Over-expression of LAPTM4B-35 promotes carcinogenesis and metastasis of hepatocellular carcinoma**

The effect of LAPTM4B over-expression on carcinogenesis was first explored using mouse NIH3T3 fibroblast cell line (He et al., 2003). Two cell lines stably over-expressing both LAPTM4B mRNA and LAPTM4B-35 were obtained from transfection of *LAPTM4B* cDNA integrated in plasmids. The LAPTM4B-35 over-expressing NIH3T3-AE cells generated a palpable mass by day 7 in all sites of inoculation in NIH3T3 mice, while in the control groups inoculated with NIH3T3-Mock cells or NIH3T3 parent cells, no tumor appeared

Notably, LAPTM family shows similarity in some functional characteristics to "tetraspanin" family, but difference in the structural characteristics, including lacking some conserved amino acid residues and having a smaller EC2 domain. (Maecker et al., 1997; Martin, 2005). Evolutionarily, *LAPTM* is an ancient and conserved gene family in mammalian species as well as in lower eukaryotic organisms, including zebrafish (Danio rerio) and Drosophila. The phylogenic tree shown in Figure 6 is constructed according to multiple sequence alignment results. It shows that *LAPTM4B* and the other two genes in the *LAPTM* family, *LAPTM4A* and *LAPTM5*, are distributed in clusters. Human *LAPTM4B* shows the closest similarity to Bos Taurus, and also shows some similarity to its zebrafish homolog. In

Fig. 6. The evolutionary lineage tree for LAPTM family (This figure was prepared by

**3. Functions of** *LAPTM4B***-encoding proteins and the mechanisms thereof** 

**3.1 Over-expression of LAPTM4B-35 promotes carcinogenesis and metastasis of** 

The effect of LAPTM4B over-expression on carcinogenesis was first explored using mouse NIH3T3 fibroblast cell line (He et al., 2003). Two cell lines stably over-expressing both LAPTM4B mRNA and LAPTM4B-35 were obtained from transfection of *LAPTM4B* cDNA integrated in plasmids. The LAPTM4B-35 over-expressing NIH3T3-AE cells generated a palpable mass by day 7 in all sites of inoculation in NIH3T3 mice, while in the control groups inoculated with NIH3T3-Mock cells or NIH3T3 parent cells, no tumor appeared

Functional studies of *LAPTM4B*-encoding proteins and the mechanisms thereof have been performed via transgenic and gene knockdown techniques. *LAPTM4B* cDNA is introduced into plasmids with or without a FLAG tag, or into an E1/E3-deleted replication-deficient adenovirus type 5 (Ad5) vector. The LAPTM4B-35 expressing plasmids and cells are all designated in our publications as "AF" if they have a FLAG tag or as "AE" if they do not have a FLAG tag. Similarly, LAPTM4B-24 expressing plasmids and cells are designated as "BF" or "BE". LAPTM4B-35 knockdown is performed by transfection of shRNA-expressing plasmids (Yang et al., 2010c). These functional studies of LAPTM4B-35 and LAPTM4B-24 have determined that these two isoforms of LAPTM4B proteins have antagonistic functions, namely up-regulation of LAPTM4B-35 and LAPTM4B-24 respectively promotes and

addition, *LAPTM4B* is more related to *LAPTM4A* than *LAPTM5*.

suppresses hepatocarcinogenesis and metastasis.

**hepatocellular carcinoma** 

Shuang Shi)

within an 80 day period following inoculation. Half of the tumor masses generated from LAPTM4B-overexpressing NIH3T3-AE cells showed growth and were histochemically identified as malignant fibrosarcoma; the other half of these masses regressed and were finally identified as liquid lymphoid tissue. Further, the L02 cell line which originated from normal human liver was then used to generate a LAPTM4B-35 over-expressing cell model by infection with replication deficient adenovirus Ad-AE containing LAPTM4B-35 full length cDNA (Lily et al., 2011). Inoculation of LAPTM4B-35 over-expressing L02-AE cells can result in rapidly growing carcinoma xenografts in 100% (6/6) of inoculated sites in the left axilla of nude mice (Figure 7a and 7b) and shorten their live span (Figure 7d) , as

Fig. 7. Upregulation of LAPTM4B-35 promotes tumorigenesis in nude mice. (a) Xenograft tumors formed from L02-Ad-AE cells, in which LAPTM4B-35 is overexpressing, and tumors formed from L02-Ad-Null cells which are indicated by white arrows. (b) Tumor growth curves from L02-Ad-AE cells and L02-Ad-Null cells. (c) Top panel: H&E staining of tumor formed from L02-Ad-Null cells (left) and L02-Ad-AE cells (right). Black arrows indicate the cancer cells in blood vessel. Bottom panel: immunohistochemical evaluation of LAPTM4B-35 expression in xenografts. Tumors derived from L02-Ad-AE cells showed high expression of LAPTM4B-35 (right), but tumors derived from L02-Ad-Null cells showed very low expression of LAPTM4B-35 (left). An anti-LAPTM4B-N10 pAb, which specifically reacts with LAPTM4B-35, was used for IHC (original magnification 200X). (d) Kaplan–Meier survival curves plotted with SPSS 16.0 (n=6). LAPTM4B-35 up-regulation shortened the live span of mice challenged with L02-Ad-AE cells, as compared to the control group infected by Ad-null, the empty Ad vectors. P=0.049 L02-Ad-AE versus L02-Ad-Null. (Lily et al., 2011).

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 15

growth and metastasis (Yang et al., 2010c). The xenografts in nude mice originating from HepG2-AE and BEL7402-AE cells, in which LAPTM4B-35 has been up-regulated, grow significantly faster (Figure 8a, 8b and 8c left pannel) and with greater numbers of more widespread metastases in the lymph nodes and lungs than the control group transfected by Mock (empty) plasmids (Figure 8d upper panels). Conversely, the xenografts in nude mice originating from HepG2-RNAi and BEL7402-RNAi cells, in which endogenous LAPTM4B-35 has been knocked down, grow significantly slower (Figure 8a, 8b and 8c right panel). At the same time, metastases in the lymph nodes and lungs in the HepG2-shRNA and BEL7402-shRNA groups were also fewer and smaller than in the corresponding control groups (Figure 8d lower panels). Overall these experiments demonstrate that up-regulation of LAPTM4B-35 promotes HCC growth and metastasis, while down-regulation has an

**3.2 Over-expression of LAPTM4B-35 induces deregulation of proliferation** 

and malignant transformation (XL.Liu et al., 2009).

proliferation resulting from LAPTM4B-35 over-expression.

Promotion and inhibition of proliferation by LAPTM4B-35 up-regulation and downregulation respectively has been demonstrated conclusively using L02-AE, BEL7402-AE, HepG2-AE and HLE-AE cell lines, together with BEL7402-RNAi and GepG2-RNAi cells (L. Li et al,. 2011; XR. Liu et al., 2009; H.Yang et al., 2010c). Except the L02 cell line was originally derived from normal liver, the BEL7402, HepG2 and HLE cell lines were all originally derived from human HCCs. It is noteworthy that up-regulation of LAPTM4B-35 not only accelerates cell proliferation (Figure 9a, 9b and 9c), but also induces deregulation of proliferation, which is a characteristically neoplastic phenotype. HLE-AE, BEL7402-AE and HepG2-AE cells in which LAPTM4B-35 has been up-regulated, generate colonies which are significantly larger and greater in number; whereas HepG2-RNAi and BEL7402-RNAi cells in which the endogenous LAPTM4B-35 has been down-regulated produce markedly smaller and fewer colonies in soft agar (Figure 9d), respectively demonstrating enhancement and diminution of anchorage-independent growth which is a fundamental criterion for evaluation of proliferative deregulation in malignant transformation (XR.Liu et al., 2009; H.Yang et al., 2010). In addition, growth of HLE-AE cells is less dependent on exogenous growth factors derived from serum supplement as compared with control HLE-MOCK cells, representing auto-secretion of growth factor by LAPTM4B-35 overexpressing cells which is another criterion of proliferative deregulation

Moreover, LAPTM4B-35 over-expression alters not only the proliferation-associated malignant cellular phenotype, but also the proliferation-regulatory proteins encoded by oncogenes and tumor suppressor genes (He et al., 2003; XR. Liu et al., 2009; H. Yang et al., 2010c). We have found that cell cycle-promoting proteins including cyclin D1, cyclin E and p-Rb are up-regulated in LAPTM4B-35 over-expressing HepG2-AE, BEL7402-AE and HLE-AE cells. At the same time, the transcription factors c-Myc, c-Fos and c-Jun that positively regulate expression of cyclins, are also significantly up-regulated. Conversely, the cell cycleinhibiting proteins of cyclin-dependent kinase inhibitor (CKI) family, including p16, p21 and in particular p27 are markedly down-regulated in these LAPTM4B-35 over-expressing HCC cell lines. At the same time, knockdown of endogenous LAPTM4B-35 gives reverse effects (Figure 10). These studies serve to identify the molecular basis for deregulated malignant

inhibitory effect.

Fig. 8. LAPTM4B-35 promotes tumor growth and metastasis in nude mice. (a) Representative xenograft tumors formed from HepG2-AE cells over-expressing LAPTM4B-35 and HepG2-shRNA cells with LAPTM4B-35 knockdown, and corresponding controls (HepG2-mock1 cells and HepG2-mock2 cells), (b) Tumor growth curves from HepG2-AE and HepG2-mock1 control cells (left), and from HepG2-shRNA and HepG2-mock2 control cells (right). (c) Average tumor weight in nude mice sacrificed 7 weeks after inoculation with HepG2-AE and HepG2-mock1 control cells (left), and with HepG2-shRNA and HepG2 mock2 control cells (right). n = 6. (P < 0.001: HepG2-AE vs. HepG2-mock1 and HepG2 shRNA vs. HepG2-mock2). For all data the mean and standard deviation represent the average of three independent experiments. (d) LAPTM4B-35 promotes tumor metastasis in nude mice: Macroscopic photos and microscopic photos shown by H&E staining of tumor metastases in lymph nodes and lungs of mice subcutaneously inoculated with BEL7402-AE cells (upper panels), BEL7402–shRNA cells (lower panels) or corresponding control cells. (Yang et al., 2010c)

compared to the control group infected by Ad-null, in which very small tumors were formed in the right axilla of nude mice (Figure 7a as indicated by the white arrows). The xenografts, which were generated from L02-AE cells, expressed high levels of LAPTM4B-35 as shown by immunohistochemistry (Figure 7c bottom right panel), and these tumors were very well vascularized and tumor blood vessels showed invasion of cancer cells by H&E staining (Figure 7c top right panel). These findings demonstrated that LAPTM4b-35 overexpressing liver cells have potential for hepatocarcinogenesis and metastasis. In addition, the HepG2 cell line which originated from human hepatoblastoma and the BEL7402 cell line which originated from human hepatocellular carcinoma were used for generating upregulated and down-regulated cell models to study the effects of LAPTM4B-35 on tumor

Fig. 8. LAPTM4B-35 promotes tumor growth and metastasis in nude mice. (a)

(Yang et al., 2010c)

Representative xenograft tumors formed from HepG2-AE cells over-expressing LAPTM4B-35 and HepG2-shRNA cells with LAPTM4B-35 knockdown, and corresponding controls (HepG2-mock1 cells and HepG2-mock2 cells), (b) Tumor growth curves from HepG2-AE and HepG2-mock1 control cells (left), and from HepG2-shRNA and HepG2-mock2 control cells (right). (c) Average tumor weight in nude mice sacrificed 7 weeks after inoculation with HepG2-AE and HepG2-mock1 control cells (left), and with HepG2-shRNA and HepG2 mock2 control cells (right). n = 6. (P < 0.001: HepG2-AE vs. HepG2-mock1 and HepG2 shRNA vs. HepG2-mock2). For all data the mean and standard deviation represent the average of three independent experiments. (d) LAPTM4B-35 promotes tumor metastasis in nude mice: Macroscopic photos and microscopic photos shown by H&E staining of tumor metastases in lymph nodes and lungs of mice subcutaneously inoculated with BEL7402-AE cells (upper panels), BEL7402–shRNA cells (lower panels) or corresponding control cells.

compared to the control group infected by Ad-null, in which very small tumors were formed in the right axilla of nude mice (Figure 7a as indicated by the white arrows). The xenografts, which were generated from L02-AE cells, expressed high levels of LAPTM4B-35 as shown by immunohistochemistry (Figure 7c bottom right panel), and these tumors were very well vascularized and tumor blood vessels showed invasion of cancer cells by H&E staining (Figure 7c top right panel). These findings demonstrated that LAPTM4b-35 overexpressing liver cells have potential for hepatocarcinogenesis and metastasis. In addition, the HepG2 cell line which originated from human hepatoblastoma and the BEL7402 cell line which originated from human hepatocellular carcinoma were used for generating upregulated and down-regulated cell models to study the effects of LAPTM4B-35 on tumor growth and metastasis (Yang et al., 2010c). The xenografts in nude mice originating from HepG2-AE and BEL7402-AE cells, in which LAPTM4B-35 has been up-regulated, grow significantly faster (Figure 8a, 8b and 8c left pannel) and with greater numbers of more widespread metastases in the lymph nodes and lungs than the control group transfected by Mock (empty) plasmids (Figure 8d upper panels). Conversely, the xenografts in nude mice originating from HepG2-RNAi and BEL7402-RNAi cells, in which endogenous LAPTM4B-35 has been knocked down, grow significantly slower (Figure 8a, 8b and 8c right panel). At the same time, metastases in the lymph nodes and lungs in the HepG2-shRNA and BEL7402-shRNA groups were also fewer and smaller than in the corresponding control groups (Figure 8d lower panels). Overall these experiments demonstrate that up-regulation of LAPTM4B-35 promotes HCC growth and metastasis, while down-regulation has an inhibitory effect.

#### **3.2 Over-expression of LAPTM4B-35 induces deregulation of proliferation**

Promotion and inhibition of proliferation by LAPTM4B-35 up-regulation and downregulation respectively has been demonstrated conclusively using L02-AE, BEL7402-AE, HepG2-AE and HLE-AE cell lines, together with BEL7402-RNAi and GepG2-RNAi cells (L. Li et al,. 2011; XR. Liu et al., 2009; H.Yang et al., 2010c). Except the L02 cell line was originally derived from normal liver, the BEL7402, HepG2 and HLE cell lines were all originally derived from human HCCs. It is noteworthy that up-regulation of LAPTM4B-35 not only accelerates cell proliferation (Figure 9a, 9b and 9c), but also induces deregulation of proliferation, which is a characteristically neoplastic phenotype. HLE-AE, BEL7402-AE and HepG2-AE cells in which LAPTM4B-35 has been up-regulated, generate colonies which are significantly larger and greater in number; whereas HepG2-RNAi and BEL7402-RNAi cells in which the endogenous LAPTM4B-35 has been down-regulated produce markedly smaller and fewer colonies in soft agar (Figure 9d), respectively demonstrating enhancement and diminution of anchorage-independent growth which is a fundamental criterion for evaluation of proliferative deregulation in malignant transformation (XR.Liu et al., 2009; H.Yang et al., 2010). In addition, growth of HLE-AE cells is less dependent on exogenous growth factors derived from serum supplement as compared with control HLE-MOCK cells, representing auto-secretion of growth factor by LAPTM4B-35 overexpressing cells which is another criterion of proliferative deregulation and malignant transformation (XL.Liu et al., 2009).

Moreover, LAPTM4B-35 over-expression alters not only the proliferation-associated malignant cellular phenotype, but also the proliferation-regulatory proteins encoded by oncogenes and tumor suppressor genes (He et al., 2003; XR. Liu et al., 2009; H. Yang et al., 2010c). We have found that cell cycle-promoting proteins including cyclin D1, cyclin E and p-Rb are up-regulated in LAPTM4B-35 over-expressing HepG2-AE, BEL7402-AE and HLE-AE cells. At the same time, the transcription factors c-Myc, c-Fos and c-Jun that positively regulate expression of cyclins, are also significantly up-regulated. Conversely, the cell cycleinhibiting proteins of cyclin-dependent kinase inhibitor (CKI) family, including p16, p21 and in particular p27 are markedly down-regulated in these LAPTM4B-35 over-expressing HCC cell lines. At the same time, knockdown of endogenous LAPTM4B-35 gives reverse effects (Figure 10). These studies serve to identify the molecular basis for deregulated malignant proliferation resulting from LAPTM4B-35 over-expression.

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 17

Fig. 10. LAPTM4B-35 alters levels of proliferation-regulating proteins analyzed by Western blot: (a) Level of p-Rb, c-Myc, cyclinD1 and cyclinE protein was significantly increased by over-expression and decreased by knockdown of LAPTM4B-35, respectively. Conversely, level of tumor suppressor gene products, p16 and p27, was significantly decreased by over-expression and increased by knockdown of LAPTM4B-35, respectively.

But level of CDK2, CDK4 and p21 was not altered, (b) Restoration of LAPTM4B-35 expression in HepG2-AE(M)R cells reversed the decrease in c-Myc and cyclinD1 proteins resulting from knockdown of endogenous LAPTM4B-35 in HepG2-shRNA cells. (H. Yang

**3.3 Over-expression of LAPTM4B-35 enhances resistance to induced apoptosis** 

Resistance to apoptosis is one of fundamental characteristics of cancer cells. LAPTM4B-35 over-expressing L02-AE, HepG2-AE and BEL7402-AE cells show marked resistance to druginduced apoptosis (Figure 11-13). However, down-regulation of endogenous LAPTM4B-35 can restore sensitivity of cancer cells to drug-induced apoptosis (L.Li et al., 2010; H.Yang et al., 2010c). In addition, up-regulation of LAPTM4B-35 inhibits activation of the apoptosis executive caspase 3, up-regulates the anti-apoptotic gene bcl-2, bcl-xL and phosphorylated Bad, and also down-regulates the pro-apoptotic gene Bax and Bid (Figure 12, L.Li et al., 2011; L. Zhou, 2011). Moreover, PI3K/AKT, a fundamental signaling pathway for cell survival is activated by up-regulation of LAPTM4B-35, and is inhibited by down-regulation

et al., 2010c).

Fig. 9. Over-expression of LAPTM4B-35 promotes cell proliferation and induces deregulation of proliferation. (a) Growth curves determined by MTT assay. Left panel: overexpression of LAPTM4B-35 promotes rapid increase in cell viability/proliferation compared with the control. Middle panel: knockdown of LAPTM4B-35 inhibits increase in cell viability/proliferation as compared with the control. Right panel: restoration of LAPTM4B-35 expression in HepG2(M)R cells, in which the RNAi target was mutated, so that the siRNA produced from shRNA can not bind to the target mRNA, reverses inhibition of cell viability/proliferation resulting from knockdown of LAPTM4B-35. (b) DNA synthesis analyzed by BrdU incorporation assay. P < 0.05: HepG2-AE vs. HepG2-mock1, and HepG2 shRNA vs. HepG2-mock2. (c) Cell cycle analyzed by FACS. P < 0.05: HepG2-AE vs. HepG2 mock1, and HepG2-shRNA vs. HepG2-Mock2. (d) LAPTM4B-35 promotes colony formation in soft agar. Left panel: overexpression of LAPTM4B-35 promoted colony formation (upper); knockdown of endogenous LAPTM4B-35 inhibited colony formation (down). Right panel: a histogram showing colony numbers larger than 50 µm that were counted 4 weeks after seeding. \*P < 0.05: HepG2-AE vs. HepG2-mock1; \*\*P < 0.001: HepG2-shRNA vs. HepG2 mock2. (H. Yang et al., 2010c)

Fig. 9. Over-expression of LAPTM4B-35 promotes cell proliferation and induces deregulation of proliferation. (a) Growth curves determined by MTT assay. Left panel: overexpression of LAPTM4B-35 promotes rapid increase in cell viability/proliferation compared with the control. Middle panel: knockdown of LAPTM4B-35 inhibits increase in cell viability/proliferation as compared with the control. Right panel: restoration of LAPTM4B-35 expression in HepG2(M)R cells, in which the RNAi target was mutated, so that the siRNA produced from shRNA can not bind to the target mRNA, reverses inhibition of cell viability/proliferation resulting from knockdown of LAPTM4B-35. (b) DNA synthesis analyzed by BrdU incorporation assay. P < 0.05: HepG2-AE vs. HepG2-mock1, and HepG2 shRNA vs. HepG2-mock2. (c) Cell cycle analyzed by FACS. P < 0.05: HepG2-AE vs. HepG2 mock1, and HepG2-shRNA vs. HepG2-Mock2. (d) LAPTM4B-35 promotes colony formation in soft agar. Left panel: overexpression of LAPTM4B-35 promoted colony formation (upper); knockdown of endogenous LAPTM4B-35 inhibited colony formation (down). Right panel: a histogram showing colony numbers larger than 50 µm that were counted 4 weeks after seeding. \*P < 0.05: HepG2-AE vs. HepG2-mock1; \*\*P < 0.001: HepG2-shRNA vs. HepG2-

mock2. (H. Yang et al., 2010c)

Fig. 10. LAPTM4B-35 alters levels of proliferation-regulating proteins analyzed by Western blot: (a) Level of p-Rb, c-Myc, cyclinD1 and cyclinE protein was significantly increased by over-expression and decreased by knockdown of LAPTM4B-35, respectively. Conversely, level of tumor suppressor gene products, p16 and p27, was significantly decreased by over-expression and increased by knockdown of LAPTM4B-35, respectively. But level of CDK2, CDK4 and p21 was not altered, (b) Restoration of LAPTM4B-35 expression in HepG2-AE(M)R cells reversed the decrease in c-Myc and cyclinD1 proteins resulting from knockdown of endogenous LAPTM4B-35 in HepG2-shRNA cells. (H. Yang et al., 2010c).

#### **3.3 Over-expression of LAPTM4B-35 enhances resistance to induced apoptosis**

Resistance to apoptosis is one of fundamental characteristics of cancer cells. LAPTM4B-35 over-expressing L02-AE, HepG2-AE and BEL7402-AE cells show marked resistance to druginduced apoptosis (Figure 11-13). However, down-regulation of endogenous LAPTM4B-35 can restore sensitivity of cancer cells to drug-induced apoptosis (L.Li et al., 2010; H.Yang et al., 2010c). In addition, up-regulation of LAPTM4B-35 inhibits activation of the apoptosis executive caspase 3, up-regulates the anti-apoptotic gene bcl-2, bcl-xL and phosphorylated Bad, and also down-regulates the pro-apoptotic gene Bax and Bid (Figure 12, L.Li et al., 2011; L. Zhou, 2011). Moreover, PI3K/AKT, a fundamental signaling pathway for cell survival is activated by up-regulation of LAPTM4B-35, and is inhibited by down-regulation

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 19

Fig. 12. Up-regulation of LAPTM4B-35 effects activation or expression of apoptosisassociated molecules in L02-AE cells. (a) Inhibition of caspase-3 activation in LAPTM4B-35 over-expressing L02-Ad-AE cells measured by the DEVD cleavage assay. Datum in each point represents the mean SD of three independent experiments. \*\*\*P < 0.001: L02- Ad-AE vs. L02-Ad-Null. (b) Cleavage of procaspase-3 and PARP was inhibited by upregulation of LAPTM4B-35, demonstrating the inhibition of apoptosis. (c) Up-regulation of the anti-apoptotic bcl-2 protein and down-regulation of the pro-apoptotic Bax protein in L02-Ad-AE cells analyzed by Western blot. (d) Increase of phosphorylated AKT and Bad in L02-Ad-AE cells, suggesting the activation of PI3K/AKT signaling and inhibition

of apoptosis. (Li et al., 2011).

of endogenous LAPTM4B-35 (L.Li et al., 2010, 2011; H. Yang et al., 2010c). This data serves to establish the cellular and molecular basis of resistance to apoptosis promoted by overexpression of LAPTM4B-35.

<sup>\*</sup> P<0.05: Ad-Null vs. Ad-AE

Fig. 11. Up-regulation of LAPTM4B-35 protects L02-AE cells from adriamycin induced apoptosis. (a) LIVE/DEAD Viability/Cytotoxicity Kit assay. Green-stained cells are viable cells; cells with red stained nuclear are late apoptotic and dead cells. The LAPTM4B-35 overexpressing L02-Ad-AE cells show less apoptotic cells as compared with L02-Ad-Null control cells. (b) Upper panel: Flow cytometry analysis of apoptosis by Annexin V and PI staining. The number of apoptotic L02-Ad-AE cells appeared in early-, late-phase of apoptosis and dead phase were all less than L02-Ad-Null cells. Lower panel: The histograms showing the cell percentage of dead cells (left) and Annexin V-positive cells in early plus late apoptotic phases (right). (c) Apoptosis shown by transmission electron microscopy. Black arrow indicates the apoptotic L02-Ad-Null control cell induced by adriamycin; this was not seen in L02-Ad-AE cells (L. Li et al., 2011).

of endogenous LAPTM4B-35 (L.Li et al., 2010, 2011; H. Yang et al., 2010c). This data serves to establish the cellular and molecular basis of resistance to apoptosis promoted by over-

Fig. 11. Up-regulation of LAPTM4B-35 protects L02-AE cells from adriamycin induced apoptosis. (a) LIVE/DEAD Viability/Cytotoxicity Kit assay. Green-stained cells are viable cells; cells with red stained nuclear are late apoptotic and dead cells. The LAPTM4B-35 overexpressing L02-Ad-AE cells show less apoptotic cells as compared with L02-Ad-Null control cells. (b) Upper panel: Flow cytometry analysis of apoptosis by Annexin V and PI staining. The number of apoptotic L02-Ad-AE cells appeared in early-, late-phase of apoptosis and dead phase were all less than L02-Ad-Null cells. Lower panel: The histograms showing the cell percentage of dead cells (left) and Annexin V-positive cells in early plus late apoptotic phases (right). (c) Apoptosis shown by transmission electron microscopy. Black arrow indicates the apoptotic L02-Ad-Null control cell induced by adriamycin; this was not seen in

expression of LAPTM4B-35.

\* P<0.05: Ad-Null vs. Ad-AE

L02-Ad-AE cells (L. Li et al., 2011).

Fig. 12. Up-regulation of LAPTM4B-35 effects activation or expression of apoptosisassociated molecules in L02-AE cells. (a) Inhibition of caspase-3 activation in LAPTM4B-35 over-expressing L02-Ad-AE cells measured by the DEVD cleavage assay. Datum in each point represents the mean SD of three independent experiments. \*\*\*P < 0.001: L02- Ad-AE vs. L02-Ad-Null. (b) Cleavage of procaspase-3 and PARP was inhibited by upregulation of LAPTM4B-35, demonstrating the inhibition of apoptosis. (c) Up-regulation of the anti-apoptotic bcl-2 protein and down-regulation of the pro-apoptotic Bax protein in L02-Ad-AE cells analyzed by Western blot. (d) Increase of phosphorylated AKT and Bad in L02-Ad-AE cells, suggesting the activation of PI3K/AKT signaling and inhibition of apoptosis. (Li et al., 2011).

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 21

Fig. 14. Over-expression of LAPTM4B-35 promotes migration and invasion. The left panels: (a) Cell migration measured by Boyden chamber assay was promoted and inhibited, respectively, by over-expression (upper panel) and knockdown (lower panel) of LAPTM4B-35 in HepG2 cells. (b) Cell invasion measured by Boyden chamber assay in the presence of Matrigel is promoted and inhibited, respectively, by over-expression (upper panel) and knockdown (lower panel) of LAPTM4B-35 in HepG2 cells. The right panels are the histograms showing number of migratory and invasive cells. \*P < 0.05: HepG2-AE vs.

Multi-drug resistance is a significant obstacle in cancer chemotherapy. We have demonstrated (L Li et al., 2010) that drug efflux and resistance to multiple drugs including doxorubicin, paclitaxel and cisplatin, are enhanced in LAPTM4B-35 over-expressing L02-AE cells that were originally derived from normal liver (Figure 15a), whereas these phenomena are reversed in LAPTM4B-35 down-regulated HeLa-RNAi cells (Figure 15b). In addition, the drug retention in HeLa-RNAi cells was significantly more than in HeLa-Mock control cells (Figure 15c). These findings imply multi-drug resistance is promoted by over-expression of LAPTM4B-35 (L Li et al., 2010). At the same time, LAPTM4B associated multi-drug resistance has also been demonstrated in breast cancer (Hu, 2009; Y.Li., 2010) and ovarian cancer (Yin, 2011a) by other research groups. Y Li et al. (2010) demonstrated that amplification of LAPTM4B and YWHAZ, which was shown by fluorescence in situ

HepG2-mock1; \*\*P < 0.001 HepG2-shRNA vs. HepG2-mock2.

**3.5 Over-expression of LAPTM4B-35 motivates multi-drug resistance** 

Fig. 13. Up-regulation and down-regulation of LAPTM4B-35 respectively inhibits and promotes HepG2 cells apoptosis induced by adriamycin. (a) Adriamycin (15 µg/ml) induced apoptosis analyzed by FACS in HepG2-AE, HepG2-shRNA, HepG2-AE (M)R cells and corresponding controls. P < 0.05: HepG2-AE cells vs. HepG2-mock1, and HepG2-AE (M) R vs. HepG2-shRNA, P < 0.001: HepG2-shRNA vs. HepG2-mock2. (b) Analysis of cleaved caspase-3 and PARP by Western blot, demonstrating activation and inhibition of the apoptotic pathway respectively by up-regulation and down-regulation of LAPTM4B-35 in HepG2 cells.

#### **3.4 Over-expression of LAPTM4B-35 enhances migration and invasion**

Cell migration and invasion are prerequisites for metastasis. Using LAPTM4B-35 overexpressing HCC cells, including HLE-AE, HepG2-AE and BEL7402-AE cells, it has been demonstrated that the cells over-expressing LAPTM4B-35 display enhanced capacity for migration and invasion (XR. Liu et al., 2009; H. Yang et al., 2010c). Conversely downregulation of endogenous LAPTM4B-35 by RNAi inhibits migration and invasion of HCC cells (Figure 14, H. Yang et al., 2010c). In addition, the matrix metal proteases (MMP2 and MMP9) and urine plasminogen activator (uPA) which are key proteases for cancer cell invasion, are up-regulated and/or activated by over-expression of LAPTM4B-35 (L. Zhou et al., 2010).

Fig. 13. Up-regulation and down-regulation of LAPTM4B-35 respectively inhibits and promotes HepG2 cells apoptosis induced by adriamycin. (a) Adriamycin (15 µg/ml) induced apoptosis analyzed by FACS in HepG2-AE, HepG2-shRNA, HepG2-AE (M)R cells and corresponding controls. P < 0.05: HepG2-AE cells vs. HepG2-mock1, and HepG2-AE (M) R vs. HepG2-shRNA, P < 0.001: HepG2-shRNA vs. HepG2-mock2. (b) Analysis of cleaved caspase-3 and PARP by Western blot, demonstrating activation and inhibition of the apoptotic pathway respectively by up-regulation and down-regulation of LAPTM4B-35 in

**3.4 Over-expression of LAPTM4B-35 enhances migration and invasion** 

Cell migration and invasion are prerequisites for metastasis. Using LAPTM4B-35 overexpressing HCC cells, including HLE-AE, HepG2-AE and BEL7402-AE cells, it has been demonstrated that the cells over-expressing LAPTM4B-35 display enhanced capacity for migration and invasion (XR. Liu et al., 2009; H. Yang et al., 2010c). Conversely downregulation of endogenous LAPTM4B-35 by RNAi inhibits migration and invasion of HCC cells (Figure 14, H. Yang et al., 2010c). In addition, the matrix metal proteases (MMP2 and MMP9) and urine plasminogen activator (uPA) which are key proteases for cancer cell invasion, are up-regulated and/or activated by over-expression of LAPTM4B-35 (L. Zhou

HepG2 cells.

et al., 2010).

Fig. 14. Over-expression of LAPTM4B-35 promotes migration and invasion. The left panels: (a) Cell migration measured by Boyden chamber assay was promoted and inhibited, respectively, by over-expression (upper panel) and knockdown (lower panel) of LAPTM4B-35 in HepG2 cells. (b) Cell invasion measured by Boyden chamber assay in the presence of Matrigel is promoted and inhibited, respectively, by over-expression (upper panel) and knockdown (lower panel) of LAPTM4B-35 in HepG2 cells. The right panels are the histograms showing number of migratory and invasive cells. \*P < 0.05: HepG2-AE vs. HepG2-mock1; \*\*P < 0.001 HepG2-shRNA vs. HepG2-mock2.

#### **3.5 Over-expression of LAPTM4B-35 motivates multi-drug resistance**

Multi-drug resistance is a significant obstacle in cancer chemotherapy. We have demonstrated (L Li et al., 2010) that drug efflux and resistance to multiple drugs including doxorubicin, paclitaxel and cisplatin, are enhanced in LAPTM4B-35 over-expressing L02-AE cells that were originally derived from normal liver (Figure 15a), whereas these phenomena are reversed in LAPTM4B-35 down-regulated HeLa-RNAi cells (Figure 15b). In addition, the drug retention in HeLa-RNAi cells was significantly more than in HeLa-Mock control cells (Figure 15c). These findings imply multi-drug resistance is promoted by over-expression of LAPTM4B-35 (L Li et al., 2010). At the same time, LAPTM4B associated multi-drug resistance has also been demonstrated in breast cancer (Hu, 2009; Y.Li., 2010) and ovarian cancer (Yin, 2011a) by other research groups. Y Li et al. (2010) demonstrated that amplification of LAPTM4B and YWHAZ, which was shown by fluorescence in situ

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 23

compaired with L02-Mock control cells. (b) Upper panel: The LAPTM4B-35 knocking down HeLa-RNAi cells show slower drug efflux as compaired with HeLa-Mock cells. The drug efflux was measured by Laser Scanning Confocal Microscopy (LSCM). Lower Panels

of (a) and (b): The fluorescence intensity in each cell was monitored and recorded separately as each curve. (c) The Rhodamine-123 retention in HeLa-RNAi cells is significantly greater than that of HeLa Mock control cells (\*P<0.05). The time course of Rhodamine-123 retention in LAPTM4B-35 knocking down HeLa-RNAi cells and HeLa-Mock cells was analyzed by a FACS. (d) Left panel: LAPTM4B-35 is co-localized with P-gp (MDR1) mainly at endomembrane organelles in HeLa cells (Bar, 8 mm). Right panel: LAPTM4B-35 is co-localized with P-gp mainly on plasma membrane in PC-3 cells which are spreading onto fibronectin substrate (Bar, 20 mm). Co-localization of LAPTM4B-35 and P-gp was analyzed by Triple-staining immunofluorescence and LSCM. Yellow color is the overlapping of green (P-gp) and red (LAPEM4B-35) signals. (e) Interaction of LAPTM4B-35 with P-gp analyzed by Co-immunoprecipitation (Co-IP). The antibodies used for immuno-precipitation were anti-LAPTM4B-35-N10pAb (left panel) and anti-P-gp pAb (right panel). The antibodies used for immuno-blotting (Western blot) were anti-P-gp

mAb (upper panel) and anti-LAPTM4B-35-N10pAb (lower panel). Interaction of

with red line. (L Li et al., 2010)

chemotherapy by targeting LAPTM4B-35.

LAPTM4B-35 with P-gp is demonstrated by Co-IP analysis of both sets marked with a box

hybridization (FISH), contributes to chemotherapy resistance of breast cancer. Hu et al (2009) reported activation of *MTDH* and *LAPTM4B* (which are localized at the same gene locus) determined by gain of 8q22 via comparative genomic hybridization (CGH) promotes chemoresistance and metastasis of breast cancer. These observations are consistent with our current results. All these studies indicate that LAPTM4B plays an important role in multidrug resistance. Our study on the mechanism indicates that there is a molecular interaction and co-localization between LAPTM4B-35 and P-gp (MDR1). P-gp is a classic transporter that results in multi-drug resistance by enhancing drug efflux from cancer cells. Interestingly, the co-localization presents at the plasma membrane when cancer cells are spreading onto extracellular matrix component fibronectin, but at intracellular membrane compartments (this may mainly involve endosomes and lysosomes based on the distribution pattern) when cancer cells are not spreading onto extracellular matrix component (Figure 15d). Since P-gp trafficking between lysosomes and plasma membrane plays a critical role in multi-drug resistance (Fu et al., 2004, 2007), our results suggest that LAPTM4B-35 may be a significant factor that is involved in the trafficking of LAPTM4B-35 and P-gp between plasma membrane and intracellular compartments in giving rise to multi-drug resistance. The detailed mechanism for this combined trafficking remains to be further studied. Additionally, increasing numbers of experiments have recently shown that activation of the PI3K/AKT signaling pathway can regulate or enhance multi-drug resistance (Abdul-Ghani et al., 2006; Knuefermann et al., 2003; McCubrey et al., 2006; Tazzari et al., 2007). Our study indicates that PI3K/AKT signaling pathway is remarkably activated by over-expression of LAPTM4B-35 (see section 3.6). Accordingly, PI3K inhibitors can inhibit AKT phosphorylation/activation and increase the sensitivity of cancer cells over-expressing LAPTM4B-35 to doxorubicin, paclitaxel and cisplatin (L Li et al., 2010). Overall, our data provide new insight into the molecular mechanisms of multidrug resistance and open a novel avenue for overcoming multi-drug resistance in

Fig. 15. LAPTM4B-35 motivates multi-drug efflux by interaction with P-gp. (a) Upper panel: The LAPTM4B-35 over-expressing L02-AE cells show faster efflux of Rhodamine-123 as

Fig. 15. LAPTM4B-35 motivates multi-drug efflux by interaction with P-gp. (a) Upper panel: The LAPTM4B-35 over-expressing L02-AE cells show faster efflux of Rhodamine-123 as

compaired with L02-Mock control cells. (b) Upper panel: The LAPTM4B-35 knocking down HeLa-RNAi cells show slower drug efflux as compaired with HeLa-Mock cells. The drug efflux was measured by Laser Scanning Confocal Microscopy (LSCM). Lower Panels of (a) and (b): The fluorescence intensity in each cell was monitored and recorded separately as each curve. (c) The Rhodamine-123 retention in HeLa-RNAi cells is significantly greater than that of HeLa Mock control cells (\*P<0.05). The time course of Rhodamine-123 retention in LAPTM4B-35 knocking down HeLa-RNAi cells and HeLa-Mock cells was analyzed by a FACS. (d) Left panel: LAPTM4B-35 is co-localized with P-gp (MDR1) mainly at endomembrane organelles in HeLa cells (Bar, 8 mm). Right panel: LAPTM4B-35 is co-localized with P-gp mainly on plasma membrane in PC-3 cells which are spreading onto fibronectin substrate (Bar, 20 mm). Co-localization of LAPTM4B-35 and P-gp was analyzed by Triple-staining immunofluorescence and LSCM. Yellow color is the overlapping of green (P-gp) and red (LAPEM4B-35) signals. (e) Interaction of LAPTM4B-35 with P-gp analyzed by Co-immunoprecipitation (Co-IP). The antibodies used for immuno-precipitation were anti-LAPTM4B-35-N10pAb (left panel) and anti-P-gp pAb (right panel). The antibodies used for immuno-blotting (Western blot) were anti-P-gp mAb (upper panel) and anti-LAPTM4B-35-N10pAb (lower panel). Interaction of LAPTM4B-35 with P-gp is demonstrated by Co-IP analysis of both sets marked with a box with red line. (L Li et al., 2010)

hybridization (FISH), contributes to chemotherapy resistance of breast cancer. Hu et al (2009) reported activation of *MTDH* and *LAPTM4B* (which are localized at the same gene locus) determined by gain of 8q22 via comparative genomic hybridization (CGH) promotes chemoresistance and metastasis of breast cancer. These observations are consistent with our current results. All these studies indicate that LAPTM4B plays an important role in multidrug resistance. Our study on the mechanism indicates that there is a molecular interaction and co-localization between LAPTM4B-35 and P-gp (MDR1). P-gp is a classic transporter that results in multi-drug resistance by enhancing drug efflux from cancer cells. Interestingly, the co-localization presents at the plasma membrane when cancer cells are spreading onto extracellular matrix component fibronectin, but at intracellular membrane compartments (this may mainly involve endosomes and lysosomes based on the distribution pattern) when cancer cells are not spreading onto extracellular matrix component (Figure 15d). Since P-gp trafficking between lysosomes and plasma membrane plays a critical role in multi-drug resistance (Fu et al., 2004, 2007), our results suggest that LAPTM4B-35 may be a significant factor that is involved in the trafficking of LAPTM4B-35 and P-gp between plasma membrane and intracellular compartments in giving rise to multi-drug resistance. The detailed mechanism for this combined trafficking remains to be further studied. Additionally, increasing numbers of experiments have recently shown that activation of the PI3K/AKT signaling pathway can regulate or enhance multi-drug resistance (Abdul-Ghani et al., 2006; Knuefermann et al., 2003; McCubrey et al., 2006; Tazzari et al., 2007). Our study indicates that PI3K/AKT signaling pathway is remarkably activated by over-expression of LAPTM4B-35 (see section 3.6). Accordingly, PI3K inhibitors can inhibit AKT phosphorylation/activation and increase the sensitivity of cancer cells over-expressing LAPTM4B-35 to doxorubicin, paclitaxel and cisplatin (L Li et al., 2010). Overall, our data provide new insight into the molecular mechanisms of multidrug resistance and open a novel avenue for overcoming multi-drug resistance in chemotherapy by targeting LAPTM4B-35.

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 25

or platform of signal molecules which functions at signal network upstream. Further investigation of this point will greatly improve our understanding on hepatocarcinogenesis, metastasis and recurrence of hepatocellular carcinoma, and will thus provide novel

Overall our studies demonstrate there is a relationship of LAPTM4B-35 over-expression with up-regulation of proliferation-promoting proteins and down-regulation of proliferation-inhibiting proteins in hepatocellular carcinoma cells. This relationship is

Fig. 16. LAPTM4B-35 over-expression activates a signaling network. (a) Over-expression and knockdown of LAPTM4B-35, respectively, activates and inhibits PI3K/AKT signaling

phosphorylating product PIP3 and two additional kinases. GSK3 and FOXO4 are both down stream effectors of AKT and thus phosphorylated by this activated kinase. (b) Phosphorulated c-Myc (p-Myc) is diminished and increased, respectively, by overexpression and knockdown in HepG2 cells. (c) Degradation of c-Myc is respectively

diminished by over-expression of LAPTM4B-35 and enhanced by knowdown of LAPTM4B-35. This experiment was performed via cycloheximide pulse-chase assay, in which stably transfected HepG2 cells were treated with 50 µg/ml of cycloheximide to inhibit biosynthesis of proteins and harvested at 0, 7, 15 and 30 min., then Western blot with anti-Myc Ab was performed to evaluate degradation of c-Myc. The results indicates that up-regulation of

**3.7 LAPTM4B-24 up-regulation induces apoptosis and autophagocytosis, and** 

LAPTM4B-24 is an isoform encoded by the *LAPTM4B* gene. It is a truncated form of LAPTM4B-35 by lacking a 91 amino acid sequence at the N-terminus (Shao et al., 2003). We have demonstrated that up-regulation of LAPTM4B-24 via transfection with plasmids

pathway shown by Western blot. AKT activation is generally evaluated by its phosphorylations at S 308 and S473 resulting from PI3K activation through its

LAPTM4B-35 enhances stability of c-Myc, and *vice versa*.

**abolishes carcinogenicity of hepatocellular carcinoma cells** 

strategies for targeted chemotherapy of hepatocellular carcinoma.

mediated by a signaling network.

#### **3.6 Over-expression of LAPTM4B-35 activates a signaling network associated with carcinogenesis**

Over-expression of LAPTM4B-35 in liver and hepatocellular carcinoma cell lines enhances a wide range of malignant cellular and molecular phenotypes. The mechanism for the wide variety of roles played by over-expression of LAPTM4B-35 is therefore of great interest and has provoked scientific attention. We believe the most likely mechanism which may account for the findings and phenotypic changes described above is the involvement of LAPTM4B-35 in a signaling network. Based on the striking similarity of the proline-rich motifs at the Nterminus of LAPTM4B-35 and at the PI3K p85α regulatory subunit, we first explored the effect of LAPTM4b-35 over-expression in HCC cells on activation of the PI3K/AKT signaling pathway. We found that LAPTM4b-35 up-regulation constantly activates PI3K/AKT signaling which has been shown by increased phosphorylation of AKT and its down stream signaling molecules (Figure 16a) in HepG2, BEL7402 HCC cells and HeLa cells (L Li et al., 2010; Yang et al., 2010). Furthermore, interaction between the proline-rich motif of LAPTM4B-35 N-terminal tail and the PI3K p85α regulatory subunit is demonstrated by site-specific mutations, GST-pull down and co-immunoprecipitation (Co-IP) (L Li et al., 2010; XR Liu et al., 2009; and unpublished data). It is believed that binding of the prolinerich motif of LAPTM4B-35 to the SH3 domain of PI3K p85α regulatory subunit may release the inhibitory effect of p85α regulatory subunit of PI3K on kinase activity of the p110 catalytic subunit of PI3K. We then found that the phosphorylation of GSK3 and FOXO4, both of which are downstream signaling molecules of phosphorylated/activated AKT, is promoted by over-expression and attenuated by knockdown of LAPTM4B-35 in HepG2-AE cells (Figure 16a). Since GSK3β (Glycogen synthase kinase 3 beta) is inactivated after phosphorylation by AKT, as a result the phosphorylation of c-Myc and cyclinD1, which are the GSK3β down stream effectors, is thus diminished and increased, respectively, in LAPTM4B-35 over-expressing and knocking down cancer cells (Figure16b). Sequentially, because diminished phophorylation of c-Myc and cyclinD1 can lead to decrease of degradation by proteosomes via ubiqutination (Diehl J et al., 1998), as a result, c-Myc and cyclinD1 become more stable and thus accumulate in LAPTM4B-35 over-expressing cancer cells (Figure 10a and 16c) playing roles in carcinogenesis. On the other hand, FOXO4 is a transcription factor of p27. It is known that phosphorylation of FOXO4 may result in sequestration in the cytoplasm with resultant loss of its function (Medema et al., 2000). This may be the reason that expression of p27 is dramatically diminished by overexpression of LAPTM4B-35, but restored by its knockdown via RNAi (Figure 10a). Based on the fact that p27 is a member of the cyclin-dependent kinase inhibitor (CKI) family, and that restored expression of p27 may inhibit cell proliferation and thus be a favorable prognostic indicator for patients with HCC (M.Fiorentino et al., 2000), it is possible that clinical therapy using LAPTM4B-35 specific RNA interference may improve the prognosis of HCC patients. In addition, the integrin/FAK, RTK/Ras/ERK and Wnt signaling pathways are also activated by LAPTM4B-35 over-expression (unpublished data). It has been demonstrated via GST-pull down, Co-IP and site-specific mutations that LAPTM4B-35 can interact with several signal molecules in cytoplasm and plasma membrane, such as PI3K p85α, FAK, integrins (α5 and α6) and RTKs (for example IGF1R), and so on. Taken together, it is evident that LAPTM4B-35 over-expression activates a signaling network, consisting of at least four signaling pathways which are closely associated with hepatocarcinogenesis, metastasis and multidrug resistance (Whittaker, 2010). It is proposed that LAPTM4B-35 may act as an organizer

Over-expression of LAPTM4B-35 in liver and hepatocellular carcinoma cell lines enhances a wide range of malignant cellular and molecular phenotypes. The mechanism for the wide variety of roles played by over-expression of LAPTM4B-35 is therefore of great interest and has provoked scientific attention. We believe the most likely mechanism which may account for the findings and phenotypic changes described above is the involvement of LAPTM4B-35 in a signaling network. Based on the striking similarity of the proline-rich motifs at the Nterminus of LAPTM4B-35 and at the PI3K p85α regulatory subunit, we first explored the effect of LAPTM4b-35 over-expression in HCC cells on activation of the PI3K/AKT signaling pathway. We found that LAPTM4b-35 up-regulation constantly activates PI3K/AKT signaling which has been shown by increased phosphorylation of AKT and its down stream signaling molecules (Figure 16a) in HepG2, BEL7402 HCC cells and HeLa cells (L Li et al., 2010; Yang et al., 2010). Furthermore, interaction between the proline-rich motif of LAPTM4B-35 N-terminal tail and the PI3K p85α regulatory subunit is demonstrated by site-specific mutations, GST-pull down and co-immunoprecipitation (Co-IP) (L Li et al., 2010; XR Liu et al., 2009; and unpublished data). It is believed that binding of the prolinerich motif of LAPTM4B-35 to the SH3 domain of PI3K p85α regulatory subunit may release the inhibitory effect of p85α regulatory subunit of PI3K on kinase activity of the p110 catalytic subunit of PI3K. We then found that the phosphorylation of GSK3 and FOXO4, both of which are downstream signaling molecules of phosphorylated/activated AKT, is promoted by over-expression and attenuated by knockdown of LAPTM4B-35 in HepG2-AE cells (Figure 16a). Since GSK3β (Glycogen synthase kinase 3 beta) is inactivated after phosphorylation by AKT, as a result the phosphorylation of c-Myc and cyclinD1, which are the GSK3β down stream effectors, is thus diminished and increased, respectively, in LAPTM4B-35 over-expressing and knocking down cancer cells (Figure16b). Sequentially, because diminished phophorylation of c-Myc and cyclinD1 can lead to decrease of degradation by proteosomes via ubiqutination (Diehl J et al., 1998), as a result, c-Myc and cyclinD1 become more stable and thus accumulate in LAPTM4B-35 over-expressing cancer cells (Figure 10a and 16c) playing roles in carcinogenesis. On the other hand, FOXO4 is a transcription factor of p27. It is known that phosphorylation of FOXO4 may result in sequestration in the cytoplasm with resultant loss of its function (Medema et al., 2000). This may be the reason that expression of p27 is dramatically diminished by overexpression of LAPTM4B-35, but restored by its knockdown via RNAi (Figure 10a). Based on the fact that p27 is a member of the cyclin-dependent kinase inhibitor (CKI) family, and that restored expression of p27 may inhibit cell proliferation and thus be a favorable prognostic indicator for patients with HCC (M.Fiorentino et al., 2000), it is possible that clinical therapy using LAPTM4B-35 specific RNA interference may improve the prognosis of HCC patients. In addition, the integrin/FAK, RTK/Ras/ERK and Wnt signaling pathways are also activated by LAPTM4B-35 over-expression (unpublished data). It has been demonstrated via GST-pull down, Co-IP and site-specific mutations that LAPTM4B-35 can interact with several signal molecules in cytoplasm and plasma membrane, such as PI3K p85α, FAK, integrins (α5 and α6) and RTKs (for example IGF1R), and so on. Taken together, it is evident that LAPTM4B-35 over-expression activates a signaling network, consisting of at least four signaling pathways which are closely associated with hepatocarcinogenesis, metastasis and multidrug resistance (Whittaker, 2010). It is proposed that LAPTM4B-35 may act as an organizer

**3.6 Over-expression of LAPTM4B-35 activates a signaling network associated with** 

**carcinogenesis** 

or platform of signal molecules which functions at signal network upstream. Further investigation of this point will greatly improve our understanding on hepatocarcinogenesis, metastasis and recurrence of hepatocellular carcinoma, and will thus provide novel strategies for targeted chemotherapy of hepatocellular carcinoma.

Overall our studies demonstrate there is a relationship of LAPTM4B-35 over-expression with up-regulation of proliferation-promoting proteins and down-regulation of proliferation-inhibiting proteins in hepatocellular carcinoma cells. This relationship is mediated by a signaling network.

Fig. 16. LAPTM4B-35 over-expression activates a signaling network. (a) Over-expression and knockdown of LAPTM4B-35, respectively, activates and inhibits PI3K/AKT signaling pathway shown by Western blot. AKT activation is generally evaluated by its phosphorylations at S 308 and S473 resulting from PI3K activation through its phosphorylating product PIP3 and two additional kinases. GSK3 and FOXO4 are both down stream effectors of AKT and thus phosphorylated by this activated kinase. (b) Phosphorulated c-Myc (p-Myc) is diminished and increased, respectively, by overexpression and knockdown in HepG2 cells. (c) Degradation of c-Myc is respectively diminished by over-expression of LAPTM4B-35 and enhanced by knowdown of LAPTM4B-35. This experiment was performed via cycloheximide pulse-chase assay, in which stably transfected HepG2 cells were treated with 50 µg/ml of cycloheximide to inhibit biosynthesis of proteins and harvested at 0, 7, 15 and 30 min., then Western blot with anti-Myc Ab was performed to evaluate degradation of c-Myc. The results indicates that up-regulation of LAPTM4B-35 enhances stability of c-Myc, and *vice versa*.

#### **3.7 LAPTM4B-24 up-regulation induces apoptosis and autophagocytosis, and abolishes carcinogenicity of hepatocellular carcinoma cells**

LAPTM4B-24 is an isoform encoded by the *LAPTM4B* gene. It is a truncated form of LAPTM4B-35 by lacking a 91 amino acid sequence at the N-terminus (Shao et al., 2003). We have demonstrated that up-regulation of LAPTM4B-24 via transfection with plasmids

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 27

 Fig. 17. Correlation of LAPTM4B-35 expression levels in HCC tissues with pathological grading, metastasis and postoperative survival time. (a) Level of LAPTM4B-35 protein in HCCs shows positive correlation with pathological grade. (b) LAPTM4B-35 expressions in HCCs from 65 patients were divided into three groups: ''Low,'' ''Mediate,'' and ''High.'' Level of LAPTM4B-35 in HCC shows significant positive correlation with intrahepatic and extrahepatic metastasis. \*P<0.05. (c) Levels of LAPTM4B-35 in HCCs show significantly negative correlation with overall (left) and disease-free (right) postoperative survival of

cancer patients. (H.Yang et al., 2010a, 2010b).

pcDNA-BE or infection with replication-deficient adenovirus Ad-BE induces apoptosis and autophagocytosis of HCC cells, as well as associated cellular and molecular alterations. At the same time LAPTM4B-24 up-regulated HCC cells lose its carcinogenicity (unpublished data). These studies indicate that LAPTM4B-24 plays an antagonistic role in hepatocarcinogenesis.

In summary, the LAPTM4B-35 plays pivotal roles in keeping cell survival, proliferation, migration and invasion, and so on; whereas LAPTM4B-24 plays critical roles in regulating programmed cell death, including apoptosis and autophagocytosis. The expressive equilibrium of LAPTM4B-35 and LAPTM4B-24 maintains physiological homeostasis of cells. Destroy of this equilibrium would lead to diseases. Up-regulation of LAPTM4B-35 leads to oncogenesis, while up-regulation of LAPTM4B-24 may plays a role in cancer regression.
