Main

Nasopharyngeal carcinoma (NPC), one of the most common malignant tumors in Southeast Asia and southern China, shows regional and familial clustering as other human cancers.1 Epidemiological studies suggest that 5–10% of this familial aggregation derives from inherited susceptibility, which implies that genetic factors play an important role in the pathogenesis of NPC. The crucial etiologic factors involved in NPC include the Epstein–Barr virus, chemical carcinogens, radiation, structural or functional mutation of oncogenes and tumor suppressor genes and chromosomal aberration.2–5 We have performed linkage analysis of high-heterozygosity microsatellite (STR) loci on Chromosome regions of 3p, 6q and 9p in high-risk families from Hunan Province of south China,6–9 and a locus on 3p21 was identified as a putative NPC susceptibility locus.1

To further verify the linage of chromosome 3p and NPC, in the present study, family-based association analysis was performed to test the transmission disequilibrium in 18 families for 15 markers on this chromosome region.

MATERIALS AND METHODS

Subjects

18 high-risk NPC families are recruited from Hunan Province in southern China. Most of these families were collected by Xiangya Hospital of Central South University and Hunan Tumor Hospital, Changsha, Hunan, China. All patients were diagnosed by pathologic examination, and the age at diagnosis of NPC was confirmed from medical records or other independent sources. Thirty-six affected and 93 unaffected individuals were used in this study (Table 1). Written informed consent was obtained from all studied participants. The study was approved by the ethical review committees of the appropriate institutions. Five to ten milliliter peripheral blood was obtained from each individual. Genomic DNA was extracted according to the routine phenol-chloroform procedure, and diluted to the final concentration of 20 ng/μl.

Table 1 Characteristics of families with NPC genotyped in this study
Full size table

Genotyping analysis

Primers sequences of 15 loci used in this study were obtained either from ABI PRISM® Linkage Mapping Set v2.0, or from GDB database (http://www.gdb.org), synthesized and labeled with different fluorescent dyes (FAM or HEX) by Shanghai Bioasia biological company.

PCR was performed in a volume of 5 μl, containing 10 mmol/L Tris-HCl, pH8.3, 50 mmol/L KCl, 1.5 mmol/L MgCl2, 0.2 mmol/L each of dNTPs, 1 μmol/L of each primer, 20 ng genomic DNA and 0.25 unit Hotstar Taq (Qiagen), in a GeneAmp PCR system 9700 (PE Applied Biosystems). The thermocycling began with a first stage of 95°C for 15 minutes, followed by 10 cycles of 94°C for 30 seconds, 64°C for 60 seconds (declines 0.5°C after each cycle), and 72°C for 45 seconds, then 25 cycles of 94°C for 30 seconds, 58°C for 60 seconds and 72°C for 45 seconds, completed with a final stage of 72°C for 10 minutes. 1.5 PCR products, 2.5 μl deionized formamide and 0.5 μl GeneScan® 400HD[ROX] Size standard were added to 0.5 μl Gel Loading Dye. After mixing and denaturing on 95°C for 3 minutes, the mixture was moved onto ice. 0.8 μl of the mixture was loaded on 5% polyacrylamide gel, which contained 7.5 M urea, electrophoreses on ABI PRISM TM377 DNA Sequencer, at 2000V for two hours.

Data collection and analysis

Software ABI PRISM® 377 XL Data Collection 3.0 was used in data collection. Program ABI PRISM® GeneScan® analysis 3.7 was used for adjusting the traces of electrophoresis, correcting the molecular markers and measuring the size of amplification products. Genotyping was performed by using Genotyper® 2.1, and Transmission Disequilibrium Test (TDT) was carried out by using Genehunter 2.1.10,11

RESULTS

Genotyping

All microsatellite loci used in this study were dinucleotide repeat polymorphism. A serial of fragments with two different basepairs were found by genotyping for each locus. For convenience, we gave each allele a serial number as their genotype according to their length by increasing 2 bp. The heterozygosity (H) of these loci was from 0.906 to 0.210 and the polymorphism information content (PIC) of these lici was from 0.898 to 0.188. D3S3560 had only two alleles, Therefore the H and PIC of this locus were low, and the H and PIC of other 14 loci were all more than 0.6. Genotype and allele distribution in subjects of some loci were mentioned in Xiong et al.6

Transmission disequilibrium test

Transmission disequilibrium test of single locus or haplotypes, which were consisted by two to five adjacent loci, was performed by using programs TDT, TDT2, TDT3, TDT4 and TDT5 of software Genehunter respectively.

One hundred fourteen alleles in 15 loci were observed, after performing single locus TDT analysis by program TDT, 5 alleles have positive correlation and four alleles have negative correlation with NPC, reached significant different (P < 0.05). In these alleles, 2 alleles (Allele 8 of D3S1298 and Allele6 of D3S3624) reached high significantly different level (P < 0.01) (Table 2).

Table 2 Loci and alleles reached significantly different level in single locus TDT analysis
Full size table

A Log Odds (lods) or an NPL score peak was obtained between D3S1298 and D3S1289 analyzed by two-point or multi-point, and parametric or non-parametric linkage analysis using Linkage or Genehunter software in our previously study.1 In this study, single-locus TDT analysis showed that located in this chromosome region, D3S1568 (alleles 5 and 10), D3S1289 (alleles 7 and 2) and D3S3624 (alleles 4 and 6) possessed two alleles correlating to NPC, respectively, and one of alleles of each locus, D3S1298 and D3S3624, reached high significantly different level. These results indicated that the TDT analysis was matched our previous linkage analysis.

Sixteen haplotypes were found associated with NPC (Table 3) by two loci TDT analysis using program TDT2. As a matter of convenience, each locus was assigned a serial number according its genetic distance to 3p-ter (Table 4), and their serial numbers represented the haplotype in Table 3.

Table 3 Haplotypes reached significantly different level in two loci TDT2 analysis
Full size table
Table 4 TDT analysis results of fifteen loci
Full size table

The results of program TDT2 showed that the eighth locus, D3S1568, and its adjacent locus comprised most of the haplotypes, D3S3624-D3S1568 (7–8) and D3S1568-D3S3560, (8–9) associated with NPC (5 haplotypes), and haplotype D3S3624 (allele 6)—D3S1568 (allele 11) reached high significant different level (++, P < 0.01).

Haplotypes consisting of three adjacent loci were found by the TDT analysis with program TDT3, 11 haplotypes comprised of seven loci correlated to NPC (P < 0.05), but no haplotype was found to reach a highly significant different level (P < 0.01) (Table 5).

Table 5 Haplotypes reached significantly different level in three loci TDT3 analysis
Full size table

Similarly, D3S1568 and its adjacent loci possessed most NPC associated haplotypes including D3S3624-D3S1568-D3S3560 (7-8-9) and D3S1568-D3S3560-D3S1289 (8-9-10).

Three haplotypes which consisted of four adjacent loci reached significantly different levels, including haplotypes comprised of D3S1568 (D3S3624-D3S1568-D3S3560-D3S1289) (Table 6). But no haplotype, which comprised of five loci, reached significantly different levels (date not shown).

Table 6 Haplotypes reached significant difference in four loci TDT4 analysis
Full size table

As a matter of convenience and obviousness, we summarized the above-mentioned results of TDT analyses of 15 loci in one table (Table 4), range them by their chromosome position. In this table, positive or negative correlation between allele or haplotype and NPC was not considered. If one allele or one haplotype correlated to NPC, an asterisk (*) was marked on the loci or the central position of the haplotype, if an allele or a haplotype reached high significant difference (P < 0.01), two asterisks were marked. If several alleles or haplotypes correlate to NPC, the asterisk would be added up.

As shown in Table 4, the distribution of the asterisks appeared as three peaks: A major TDT peak was observed near D3S1568, which possessed 20 alleles or haplotypes in six loci spanning 12.4 cm from D3S1298 to D3S1289. 3 of 20 alleles or haplotypes reached high significant difference. These 20 alleles or haplotypes reflected the linkage disequilibrium between this chromosome region and NPC, and these results provided further evidence that a novel NPC susceptibility gene may be located in this chromosome region. In addition, two minor TDT peaks were observed near D3S1489 and D3S1300, respectively, and it was consistent with our previous results of linkage analysis also.1

DISCUSSION

Because of its broad distribution on human genome, and its high heterozygosity as well as polymorphism information content, microsatellite is widely used in fine genetic linkage mapping, locating disease genes, individual distinguishing, parentage identification, etc. It also made great progress in locating and cloning the susceptibility genes of malignant tumors.1,12–14

In this study, we genotyped the 15 microsatellite loci on the short arm of chromosome 3 in 18 NPC families collected from Hunan Province, and TDT analysis was performed. The results of TDT analysis were consistent with our previous linkage analysis. Near D3S1568, spanning from D3S1298 to D3S1289, 20 NPC-associated alleles or haplotypes were observed, and the locus D3S1568 is exactly the region containing the lods score peak of our previously linkage analysis.1 It reflected that one or more novel NPC-associated susceptibility genes, especially correlating with Hunan familial NPC, locates in the region near locus D3S1568, in which there was a 630 Kb homozygous deletion of cancer, and many candidate tumor suppressor genes were identified.15

We also found alleles or haplotypes in other loci, such as D3S1297-D3S1489-D3S1266 and D3S1300-D3S1285-D3S3681, were associated to NPC. In our previous linkage analysis, in addition to the major peek on D3S1568, two minors lods score peaks on D3S1489 and D3S1300 were observed. Though the lods or NPL scores of these loci did not reach a significant level, it implies that one or more genes contribute minor effects on the development of NPC that may be located in this region. Near D3S1489, NPC associated gene NAG7 was cloned. Preliminary function analysis of this gene shows that it may play a certain role in the development of NPC.16–18

Recently, a high lods score was obtained on chromosome 4p15.1-q12 by employing a Genome-wide scan in 20 Cantonese nasopharyngeal carcinoma families, which indicated that there exists a susceptibility gene in this region.14 In addition, previous studies suggested that loci HLA-A2, HLA-B17, HLA-BW46,19,20 D6S15817 and polymorphisms of CYP450,21 NGX6,22 UBAP1,23 NOR124 genes were associated with NPC. In our opinion, considering the complicated pathogenesis resulted from heterogeneity and the environmental factors as well as multiple genes involved in NPC, it is acceptable that different results were obtained in the studies of different geography region and subjects.