**2.1 Identification and cloning of the** *LAPTM4B* **gene**

To search for genes involved in deregulation of proliferation and aberrant differentiation for hepatocellular carcinoma, fluorescent-differential display was performed using human liver tissues, including adult liver, fetal liver, HCC and its paired samples of noncancerous liver from the same patient (paired noncancerous liver). A cDNA fragment, which was not a part of any known gene, was found from 110 differential display fragments obtained. This cDNA fragment was highly expressed in HCC, significantly expressed in fetal liver and the paired noncancerous liver, but showed very low expression in normal adult liver (JJ. Liu et al., 2000). Cloning the full length cDNA (NM\_018407, ID=55353, Figure 1) harboring this fragment was performed by Expressed Sequence Tag (EST) splicing, 5' rapid amplification

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 5

Bioinformatics shows that the full-length cDNA sequence of LAPTM4B contains two translational initiation codons (ATG) separated by an interval of 273 bp and therefore, encodes two protein isoforms with apparent molecular weights of 35 kDa containing 317 amino acid residues and 24 kDa containing 226 amino acid residues, which are designated LAPTM4B-35 and LAPTM4B-24. LAPTM4B-35 has a pI at 9.07 due to its high content of arginine residues, while LAPTM4B-24 has a pI at 4.65 resulting from a high content of acidic amino acid residues. Computer analysis shows that LAPTM4B is an integral membrane glycoprotein (Figure 2), with four transmembrane regions, two extracellular domains (EC1 and EC2), one small intracellular loop, together with one N-terminal and one C-terminal tail which reside in the cytoplasm. One potential N-glycosylation site is present in the EC1 domain and several O-glycosylation sites may also be present. LAPTM4B-35 contains six putative intracytoplasmic phosphorylation sites, including one tyrosine phosphorylation site at Try 285, and four N-myristylation sites. Compared to LAPTM4B-24, LAPTM4B-35 has an extra 91 amino acid sequence in its N-terminus that harbors a proline-rich domain, which serves as the binding site of the SH3 domain in some signaling molecules. A conserved YXXф (Y represents tyrosine, ф represents a large hydrophobic amino acid, X represents any amino acid) motif in the C-terminal tail is believed to function as a signal for targeting

**2.2 Bioinformatics analysis of LAPTM4B proteins** 

to lysosomal membranes and sorting.

Fig. 2. Topology of LAPTM4B-35 (left) and LAPTM4B-24 (right) proteins

**2.3 Expression of LAPTM4B mRNA and protein in normal tissues and cancer tissues**  Expression of LAPTM4B mRNA has been evaluated by Northern blot and RT-PCR, and found to be widely expressed in human tissues. Its expression is highest in the heart, skeletal muscle and uterus, and more high in testis and ovary. Expression is moderate in the kidney and pancreas, and low in the liver, spleen and thymus, but is lowest in the lung and peripheral leukocytes (Figure 3a) (Shao et al., 2003; Kasper et al., 2005). High expression is

of cDNA ends (FACE) and reverse transcription-polymerase chain reaction (RT–PCR). Based on the lysosome-targeting signal (YXXΦ) found on the encoded protein this gene was originally designated by the HUGO Gene Nomenclature Committee as "lysosomalassociated protein transmembrane 4 beta" (*LAPTM4B)* and this designation was then revised as "lysosomal protein transmembrane 4 beta" (*LAPTM4B)*. Importantly, it should be emphasized that reverse-transcription from total RNA to synthesize this cDNA must be conducted at 65o C using the ThermoScript RT-PCR Systems (Gibco-BRL) as there is very high content of GC at the 5' terminus (Shao et al., 2003), otherwise full length LAPTM4B cDNA cannot be obtained. BLAST program analysis shows that the *LAPTM4B* gene maps to chromosome 8q22.1, spanning at least 50 kb composed of seven exons separated by six introns, and contains a 951bp open reading frame (ORF). LAPTM4B mRNA is approximately 2.16 kb in length and which is consistent with the size of the mRNA observed in Northern blots. Two polyadenylation signal sites, AATAAA and AATTAAA are present at the 3' UTR (Figure 1). The former polyadenylation site (AATAAA) may result in another 1.42 kb mRNA variant (Shao et al., 2003).


Fig. 1. The nucleotide sequence of LAPTM4B cDNA and the deduced amino-acid sequence. The sequence numbers of the first nucleotide (left) and the final amino acid (right) in each row is indicated, respectively. Exon divisions are indicated by . The polyadenylation consensus sequences (AATAAA, AATTAAA) are underlined. Transmembrane regions are boxed. The putative N-glycosylation site is marked with a ring of hexagon; potential phosphorylation sites are circled, with the corresponding kinases indicated in italic characters. Stop codon or inframe stop codons are indicated as # or \*. The PPRP and YXXΦ motifs are boxed with red line of dashes. (Shao et al., 2003)

#### **2.2 Bioinformatics analysis of LAPTM4B proteins**

4 Hepatocellular Carcinoma – Basic Research

of cDNA ends (FACE) and reverse transcription-polymerase chain reaction (RT–PCR). Based on the lysosome-targeting signal (YXXΦ) found on the encoded protein this gene was originally designated by the HUGO Gene Nomenclature Committee as "lysosomalassociated protein transmembrane 4 beta" (*LAPTM4B)* and this designation was then revised as "lysosomal protein transmembrane 4 beta" (*LAPTM4B)*. Importantly, it should be emphasized that reverse-transcription from total RNA to synthesize this cDNA must be conducted at 65o C using the ThermoScript RT-PCR Systems (Gibco-BRL) as there is very high content of GC at the 5' terminus (Shao et al., 2003), otherwise full length LAPTM4B cDNA cannot be obtained. BLAST program analysis shows that the *LAPTM4B* gene maps to chromosome 8q22.1, spanning at least 50 kb composed of seven exons separated by six introns, and contains a 951bp open reading frame (ORF). LAPTM4B mRNA is approximately 2.16 kb in length and which is consistent with the size of the mRNA observed in Northern blots. Two polyadenylation signal sites, AATAAA and AATTAAA are present at the 3' UTR (Figure 1). The former polyadenylation site (AATAAA) may result in another

Fig. 1. The nucleotide sequence of LAPTM4B cDNA and the deduced amino-acid sequence. The sequence numbers of the first nucleotide (left) and the final amino acid (right) in each row is indicated, respectively. Exon divisions are indicated by . The polyadenylation consensus sequences (AATAAA, AATTAAA) are underlined. Transmembrane regions are boxed. The putative N-glycosylation site is marked with a ring of hexagon; potential phosphorylation sites are circled, with the corresponding kinases indicated in italic characters. Stop codon or inframe stop codons are indicated as # or \*. The PPRP and YXXΦ

motifs are boxed with red line of dashes. (Shao et al., 2003)

1.42 kb mRNA variant (Shao et al., 2003).

Bioinformatics shows that the full-length cDNA sequence of LAPTM4B contains two translational initiation codons (ATG) separated by an interval of 273 bp and therefore, encodes two protein isoforms with apparent molecular weights of 35 kDa containing 317 amino acid residues and 24 kDa containing 226 amino acid residues, which are designated LAPTM4B-35 and LAPTM4B-24. LAPTM4B-35 has a pI at 9.07 due to its high content of arginine residues, while LAPTM4B-24 has a pI at 4.65 resulting from a high content of acidic amino acid residues. Computer analysis shows that LAPTM4B is an integral membrane glycoprotein (Figure 2), with four transmembrane regions, two extracellular domains (EC1 and EC2), one small intracellular loop, together with one N-terminal and one C-terminal tail which reside in the cytoplasm. One potential N-glycosylation site is present in the EC1 domain and several O-glycosylation sites may also be present. LAPTM4B-35 contains six putative intracytoplasmic phosphorylation sites, including one tyrosine phosphorylation site at Try 285, and four N-myristylation sites. Compared to LAPTM4B-24, LAPTM4B-35 has an extra 91 amino acid sequence in its N-terminus that harbors a proline-rich domain, which serves as the binding site of the SH3 domain in some signaling molecules. A conserved YXXф (Y represents tyrosine, ф represents a large hydrophobic amino acid, X represents any amino acid) motif in the C-terminal tail is believed to function as a signal for targeting to lysosomal membranes and sorting.

Fig. 2. Topology of LAPTM4B-35 (left) and LAPTM4B-24 (right) proteins

#### **2.3 Expression of LAPTM4B mRNA and protein in normal tissues and cancer tissues**

Expression of LAPTM4B mRNA has been evaluated by Northern blot and RT-PCR, and found to be widely expressed in human tissues. Its expression is highest in the heart, skeletal muscle and uterus, and more high in testis and ovary. Expression is moderate in the kidney and pancreas, and low in the liver, spleen and thymus, but is lowest in the lung and peripheral leukocytes (Figure 3a) (Shao et al., 2003; Kasper et al., 2005). High expression is

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 7

Fig. 4. LAPTM4B mRNA expression in hepatocellular carcinomas (HCCs) tissues and HCCderived cell lines analyzed by Northern blot with a LAPTM4B prob (corresponding to nt 1876-2171 on NM\_018407). (a) Northern blot analysis of LAPTM4B. Upper panel: HCC tissues (T), paired noncancerous liver tissues (N), fetal and adult normal liver tissues (F- & A-liv); Lower panel: The human liver- and HCC-derived cell lines, and mouse HCC-derived cell lines. (b) Hybridization in situ was performed on tissue sections. LAPTM4B mRNA expressions in HCC tissue (i, ii), paired noncancerous tissue (iii), and BEL7402 cells (iv). Positive LAPTM4B mRNA expression was preferentially in hepatic cancerous cells as indicated by blue-purple staining. Original magnification, 40 (i, iii) or 100 (ii, iv). (c) Positive correlation between LAPTM4B mRNA levels and tumor grades. Each spot in the figure represents the ratio of tumor vs. paired noncancerous liver tissue (T/N) of the LAPTM4B mRNA expression evaluated by Northern blot. These spots derived from 55 HCC patients.

(Shao et al., 2003)

also seen in fetal heart, kidney and spleen (Figure 3b). Expression of LAPTM4B mRNA in HCC has been confirmed by Northern blot (Figure 4a), RT-PCR and hybridization in situ (Figure 4b). The frequency of LAPTM4B mRNA over-expression as demonstrated by Northern blot is as high as 87.3% (Shao et al., 2003); Expression of LAPTM4B-35 and LAPTM4B-24 proteins in tissues have been demonstrated by Western blot and immunohistochemistry (IHC). As shown by Western blot (Figure 5a), LAPTM4B-35 is much more abundant than LAPTM4B-24 in the liver and HCC. In other words, overexpression of LAPTM4B-35 but not LAPTM4B-24 presents in HCC, and the ratio of LAPTM4B-35 to LAPTM4B-24 is thus markedly increased in HCC (X. Liu et al., 2004). The frequency of LAPTM4B-35 up-regulation is also extremely high (Figure 5b), and it has been shown that T/N>1.5 [T=tumor)/N =paired non-cancerous liver] is found in 87.7% (57/65) of HCC, T/N>2 in 76.9% (50/65) of HCC and T/N>4 in 61.5% (40/65) of HCC (Yang et al., 2010a). More importantly, the levels of LAPTM4B-35 in HCC tissues are significantly associated with pathological grades, TNM staging, portal vein invasion and recurrence, but not with age, gender, viral status, tumor size and serum AFP levels (Table 1, Yang et al., 2010a). Moreover, LAPTM4B mRNA and LAPTM4B-35 over-expression also occurs in a wide range of hepatocellular carcinoma- and other cancer-derived cell lines (Shao et al., 2003; XR. Liu et al., 2004).

Fig. 3. *LAPTM4B* mRNA expression in normal human tissues analyzed by Northern blot with a *LAPTM4B* prob (corresponding to nt 1876-2171 on NM\_018407). (a) Expression of LAPTM4B mRNA in adult human tissues (b) Expression of LAPTM4B mRNA in fetal human tissues (Shao et al., 2003)

also seen in fetal heart, kidney and spleen (Figure 3b). Expression of LAPTM4B mRNA in HCC has been confirmed by Northern blot (Figure 4a), RT-PCR and hybridization in situ (Figure 4b). The frequency of LAPTM4B mRNA over-expression as demonstrated by Northern blot is as high as 87.3% (Shao et al., 2003); Expression of LAPTM4B-35 and LAPTM4B-24 proteins in tissues have been demonstrated by Western blot and immunohistochemistry (IHC). As shown by Western blot (Figure 5a), LAPTM4B-35 is much more abundant than LAPTM4B-24 in the liver and HCC. In other words, overexpression of LAPTM4B-35 but not LAPTM4B-24 presents in HCC, and the ratio of LAPTM4B-35 to LAPTM4B-24 is thus markedly increased in HCC (X. Liu et al., 2004). The frequency of LAPTM4B-35 up-regulation is also extremely high (Figure 5b), and it has been shown that T/N>1.5 [T=tumor)/N =paired non-cancerous liver] is found in 87.7% (57/65) of HCC, T/N>2 in 76.9% (50/65) of HCC and T/N>4 in 61.5% (40/65) of HCC (Yang et al., 2010a). More importantly, the levels of LAPTM4B-35 in HCC tissues are significantly associated with pathological grades, TNM staging, portal vein invasion and recurrence, but not with age, gender, viral status, tumor size and serum AFP levels (Table 1, Yang et al., 2010a). Moreover, LAPTM4B mRNA and LAPTM4B-35 over-expression also occurs in a wide range of hepatocellular carcinoma- and other cancer-derived cell lines

Fig. 3. *LAPTM4B* mRNA expression in normal human tissues analyzed by Northern blot

(a) Expression of LAPTM4B mRNA in adult human tissues (b) Expression of LAPTM4B

with a *LAPTM4B* prob (corresponding to nt 1876-2171 on NM\_018407).

mRNA in fetal human tissues (Shao et al., 2003)

(Shao et al., 2003; XR. Liu et al., 2004).

Fig. 4. LAPTM4B mRNA expression in hepatocellular carcinomas (HCCs) tissues and HCCderived cell lines analyzed by Northern blot with a LAPTM4B prob (corresponding to nt 1876-2171 on NM\_018407). (a) Northern blot analysis of LAPTM4B. Upper panel: HCC tissues (T), paired noncancerous liver tissues (N), fetal and adult normal liver tissues (F- & A-liv); Lower panel: The human liver- and HCC-derived cell lines, and mouse HCC-derived cell lines. (b) Hybridization in situ was performed on tissue sections. LAPTM4B mRNA expressions in HCC tissue (i, ii), paired noncancerous tissue (iii), and BEL7402 cells (iv). Positive LAPTM4B mRNA expression was preferentially in hepatic cancerous cells as indicated by blue-purple staining. Original magnification, 40 (i, iii) or 100 (ii, iv). (c) Positive correlation between LAPTM4B mRNA levels and tumor grades. Each spot in the figure represents the ratio of tumor vs. paired noncancerous liver tissue (T/N) of the LAPTM4B mRNA expression evaluated by Northern blot. These spots derived from 55 HCC patients. (Shao et al., 2003)

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 9

Low Middle High

Variables Patients (n) LAPTM4B-35 expression *Pa* 

Gender 0.434

Age (years) 0.544

Cirrhosis 0.013

Viral status 0.766

Tumor size 0.421

Invasive tumor <0.001

differentiation <0.01*<sup>b</sup>*

Serum AFP level 0.517

TNM stage <0.001

Table 1. Relationship between LAPTM4B-35 expression and Clinical-pathological features of

All cases 65 15 30 20

Male 53 11 24 18 Female 12 4 6 2

<60 42 8 21 13 ≥60 23 7 9 7

Yes 44 6 25 13 No 21 9 5 7

Hepatitis virus B 46 9 23 14 Hepatitis virus C 12 4 5 3 Both hepatitis virus B and C 2 0 1 1 Non-B, Non-C 5 2 1 2

<5 cm 34 10 15 9 ≥5 cm 31 5 15 11

Yes 17 0 4 13 No 48 15 26 7

Well Differentiation (WD) 18 6 8 4 Middle Differentiation (MD) 24 7 11 6 Poor Differentiation (PD) 23 2\* 11 10\*

<25 ng/ml 32 6 17 9 ≥25 ng/ml 33 9 13 11

I 16 11 3 2 II 21 2 14 5 III—IV 28 2 13 13

*b*: p<0.01: Low expression vs. High expression for PD group.

Histopathological

*a*: Chi-square test;

HCC (Yang et al., 2010a)

Fig. 5. Expression of LAPTM4B-35 and LAPTM4B-24 protein identified by Western blot in HCCs and paired noncancerous liver tissues (PNL) from the same one patient. (a) Expression of LAPTM4B-35 and LAPTM4B-24 was evaluated by Western blot with an anti-LAPTM4B-EC2 pAb, which reacts with both LAPTM4B-35 and LAPTM4B-24, and anti-LAPTM4B-N-pAb, which reacts with merely LAPTM4B-35. It is shown that LAPTM4B-35 is much more abundant than LAPTM4B-24 in the normal liver (NL) and HCC. (b) Expression of LAPTM4B-35 detected with Anti-LAPTM4B-N-10 pAb. The upper panel shows the expression of LAPTM4B-35 in HCC (T) and paired noncancerous liver tissues (N) tissues. Densitometry was normalized with -actin. The middle panel shows the levels of LAPTM4B-35 in HCC tissues derived from 65 patients. The lower panel shows the levels of LAPTM4B-35 in the paired noncancerous liver tissues derived from the same 65 patients. (Yang et al., 2010a).



*a*: Chi-square test;

8 Hepatocellular Carcinoma – Basic Research

Fig. 5. Expression of LAPTM4B-35 and LAPTM4B-24 protein identified by Western blot in

(a) Expression of LAPTM4B-35 and LAPTM4B-24 was evaluated by Western blot with an anti-LAPTM4B-EC2 pAb, which reacts with both LAPTM4B-35 and LAPTM4B-24, and anti-LAPTM4B-N-pAb, which reacts with merely LAPTM4B-35. It is shown that LAPTM4B-35 is much more abundant than LAPTM4B-24 in the normal liver (NL) and HCC. (b) Expression of LAPTM4B-35 detected with Anti-LAPTM4B-N-10 pAb. The upper panel shows the expression of LAPTM4B-35 in HCC (T) and paired noncancerous liver tissues (N) tissues.

LAPTM4B-35 in HCC tissues derived from 65 patients. The lower panel shows the levels of LAPTM4B-35 in the paired noncancerous liver tissues derived from the same 65 patients.

HCCs and paired noncancerous liver tissues (PNL) from the same one patient.

Densitometry was normalized with -actin. The middle panel shows the levels of

(Yang et al., 2010a).

*b*: p<0.01: Low expression vs. High expression for PD group.

Table 1. Relationship between LAPTM4B-35 expression and Clinical-pathological features of HCC (Yang et al., 2010a)

LAPTM4B: A Novel Diagnostic Biomarker and Therapeutic Target for Hepatocellular Carcinoma 11

unpublished data). In addition, allele *\*2* has been found to be associated with increased risk of lung cancer (Deng et al., 2005), stomach cancer (Liu et al., 2007), cervical carcinoma (Meng et al., 2011) and colon cancer but not of rectal cancer (Cheng et al., 2008). These results indicate that *LAPTM4B \*2* is also a potential risk factor for the development of some solid cancers.

(n=206)

\*2/2 21 (10.2) 32 (17.4) 0.038a

In addition to LAPTM4B, LAPTM4A (Hogue et al., 1996) and LAPTM5 (Adra et al., 1996) are also members of lysosome-associated protein transmembrane (LAPTM) family. LAPTM4A (27kDa) shows a 46% homology with LAPTM4B in amino acid sequences. In comparison with LAPTM4B-35 containing 317 amino acid residues, LAPTM4A containing 233 amino acid residues and LAPTM5 containing 262 amino acid residues both are of the Nterminus-truncated molecules of LAPTM family. The three members in the LAPTM family all localize at late endosomes and lysosomes, and play an important role in lysosomal function**,** including transporting structurally unrelated amphiphilic molecules between cytosol and lysosomes, and are involved in autophagocytosis. Moreover, LAPTM4B, LAPTM4A and LAPTM5 interact and co-localize with mucolipin 1, which is a lysosomal ion channel that belongs to the transient receptor potential (TRP) superfamily and their loss-offunction mutations result in mucolipidosis type IV (MLIV), a lysosomal storage disorder characterized by severe mental and psychomotor retardation (Vergarajauregui et al., 2011). LAPTM4A is involved in the subcellular compartmentalization of diverse hydrophobic small molecules and contributes to the inherent drug sensitivity or resistance of the mammalian cell (Hogue et al., 1999). The *LAPTM5* gene maps to chromosome 1p34 and is expressed mainly in hematopoietic cells and immune cells (Adra et al., 1996). LAPTM5 protein physically interacts with the B cell receptor (BCR) complex and promotes its degradation in the lysosomal compartment in mouse B cells, and thus negatively regulates cell surface BCR levels and B cell activation (Ouchida, 2010). The expression of LAPTM5 gene is usually down-regulated through DNA methylation in a neuroblastoma cell-specific manner, while over-expression of this gene may induce spontaneous regression of neuroblastomas. It is believed that caspase-independent lysosomal cell death due to lysosomal destabilization resulted from LAPTM5 up-regulation is closely associated with

\*1/1 97 (47.1) 77(41.8) \*1/2 88 (42.7) 75 (40.8)

\*1 0.68 0.62 \*2 0.32 0.38

Table 2. Distribution of *LAPTM4B* genotypes in HCC and Controls

a : when compared with the combined frequency of 1/1 and 1/2;

the spontaneous regression (Inoue et al., 2009).

LAPTM4B genotype

Allele frequency

**2.5 LAPTM family** 

OR: 1.855, 95%CI: 1.027 – 3.348

N (%)

*<sup>P</sup>* value Controls

HCC (n=184)

In addition, over-expressions of LAPTM4B mRNA and LAPTM4B-35 protein have also been found in other solid cancers. Kasper et al. (2005) reported that LAPTM4B mRNA tested by Northern Blot is over-expressed in 88% of lung cancer, 50.9% of breast cancers, 67% of colon cancers, 68% of uterus cancers, and 37% of gastric cancers. Peng et al. (2005) reported that LAPTM4B-35 protein as evaluated by immuno-histochemistry (IHC) is highly expressed in lung cancer, stomach cancer, colon cancer and breast cancer. Subsequently, the frequency of over-expression of LAPTM4B-35 in more cases of various cancers was determined, and was found in 76 % of gallbladder cancers (Zhou et al., 2007), 72 % of cholangiocarcinomas (Zhou et al.,. 2008), 63.5% of ovarian cancers (Yang et al., 2008; Yin et al., 2010), 72.57% of cervical cancers (Meng et al., 2010a) and 70.91% of endometrial cancers (Meng, et al., 2010b). However, the mechanism for over-expression of LAPTM4B in HCC and other cancers has not as yet been completely elucidated. Mutation and demethylation of *LAPTM4B* gene has not been found. Nevertheless, it has been reported in a large number of articles that the chromosome 8q region harboring *LAPTM4B* gene is amplified as shown by fluorescence in situ hybridization (FISH) or gained as shown by comparative genomic hybridization (CGH) in both hepatoblastoma and hepatocellular carcinoma (Buendia et al., 2002; Longerichet al., 2011; Marchio et al., 1997). More precisely, it has been recently reported that chromosome 8q 22 where the *LAPTM4B* gene localizes is amplified or gained in breast cancer (Hu et al., 2009; Y. Li. et al., 2010). Therefore, gene amplification may be the cytogenetic basis of LAPTM4B over-expression. However, the genomic DNA copy number alterations of most genes in general do not appear to parallel corresponding transcriptional expression (Huang et al., 2006). It is reasonable to propose that transcriptional up-regulation by transcription factors and/or microRNAs may also contribute to LAPTM4B-35 over-expression in cancers. A more in-depth analysis will be required to clarify these points.

#### **2.4** *LAPTM4B* **alleles and their significance in susceptibility for hepatocarcinogenesis**

Two alleles of the *LAPTM4B* gene have been identified in our laboratory and designated *LAPTM4B \*1* and *\*2*. Allele *\*2* differs from allele *\*1* in that it contains an extra tandemly arranged 19-bp (gcttggagctccagcagct) sequence at the 5'UTR in the first exon. A PCR-based method was established for genotyping of this gene using the primers 5' GCCGACTAGGGGACTGGCGGA 3' and 5' CGAGAGCTCCGAGCTTCTGCC 3' to amplify the partial sequence of exon 1, and using genomic DNA as the template. To investigate the relationship between the allelic variants of *LAPTM4B* and the susceptibility to HCC or esophageal squamous cell carcinoma (ESC), patients with HCC or ESC, and two control groups of normal individuals from the corresponding regions were analyzed. Significant differences in the frequency of genotype *LAPTM4B\*2/2* were found in patients with HCC (17.4%) as compared with the controls (10.2%) (p<0.05), indicating that individuals with the *\*2/\*2* genotype are more susceptible to HCC than *\*1/\*1* and \**1/\*2* individuals (Table 2). However, no difference was observed in the frequencies of *LAPTM4B* genotypes in patients with ESC as compared with corresponding controls. These results suggest that allele *\*2* may be associated specifically with susceptibility to HCC. In conclusion, our data suggest that *LAPTM4B\*2/\*2* is associated with susceptibility to HCC and the *LAPTM4B* genotype provides a new means for screening for people who are susceptible to primary hepatocellular carcinoma. This may be of importance for the assessment and prevention of developing hepatocellular carcinoma in high risk populations, in particular for the patients with liver cirrhosis of small liver nodule from HBV and HCV chronic infection (Shao et al., unpublished data). In addition, allele *\*2* has been found to be associated with increased risk of lung cancer (Deng et al., 2005), stomach cancer (Liu et al., 2007), cervical carcinoma (Meng et al., 2011) and colon cancer but not of rectal cancer (Cheng et al., 2008). These results indicate that *LAPTM4B \*2* is also a potential risk factor for the development of some solid cancers.


a : when compared with the combined frequency of 1/1 and 1/2; OR: 1.855, 95%CI: 1.027 – 3.348

Table 2. Distribution of *LAPTM4B* genotypes in HCC and Controls
