Genetic variants in the enhancer region of the thymidylate synthase gene in the Chilean population

Abstract

Aims

Thymidylate synthase (TYMS) is an important target enzyme for the fluoropyrimidines. The TYMS gene enhancer region possesses tandemly repeated (TSER) sequences that are polymorphic in humans and different among ethnic groups. The aims of this study were to estimate the frequencies of the TSER variants in two hospital samples located in the northern (HSJ) and eastern (CLC) parts of Santiago, Chile, and compare them with the frequencies in other populations of different ethnic origin.

Methods

Genotyping of TSER variants in 368 Chilean subjects (HSJ = 178 and CLC = 190) by polymerase chain reaction; products of amplification were electrophoresed, obtaining fragments of 250 bp for allele TSER*3 and 220 bp for allele TSER*2.

Results

The two hospital samples had different degrees of Amerindian admixture (HSJ 34.5%; CLC 15.9%), which was not reflected in the observed frequencies of the CLC TSER*3: 56.8% and HSJ TSER*3: 53.4%.

Conclusions

Our results are unexpected, considering that genetic markers in the Chilean population generally show allele frequencies between those observed in European Caucasians and Amerindians and that the percentage of Amerindian admixture in CLC is lower than in HSJ. Both hospitals should have had greater frequencies of TSER*3 than were found and the frequency should have been greater in HSJ than in CLC; the only logical explanation of our results is that the frequency of this allele in aboriginal Chilean people is much lower than the 80% estimated for Mongoloid populations.

Thymidylate synthase (TYMS) catalyses the intracellular conversion of deoxyuridine-5′-monophosphate (dUMP) to 2′ - deoxythymidine - 5′ - monophosphate (dTMP), which is essential for DNA replication. TYMS is an important target for cancer chemotherapy drugs such as 5-fluorouracil and raltitrexed. These drugs prevent DNA synthesis by forming stable complexes with TYMS and its folate cofactor, thus blocking the conversion of dUMP to dTMP [1]. Therefore, the in vivo regulation of TYMS is important both in normal tissue biology and in cancer therapeutics [2]. The human TYMS promoter has recently been characterized, identifying several important mechanisms for gene regulation [2, 3]. Tandem repeat sequences were identified near the initiation site. Analysis of the 5′-untranslated region of the TYMS gene identified a tandem repeat sequence that is a cis-acting enhancer element. This TYMS enhancer region (TSER) was shown to be polymorphic, containing either two or three 28-bp tandem repeats. In vitro expression studies suggest that the presence of a triple repeat (TSER*3) results in a 2.6-fold increase in mRNA expression compared with a double repeat (TSER*2) [4]. In vivo studies in human gastrointestinal tumours have shown a significant increase in TYMS protein levels and functional activity in patients with TSER*3 compared with individuals with TSER*2[5].

There are few population studies of the TSER locus in the literature. The frequencies reported for allele TSER*3 in Japanese [6], Chinese [7], South-west Asians [8], British Caucasians [8], American Caucasians [9, 10], African-Americans [9], Ghanaians [9], Kenyans [9], Northern Irish [11], Italians [12] and Australians [13] were 0.81, 0.82, 0.62, 0.60, 0.54, 0.54, 0.52, 0.56, 0.49, 0.52, 0.50 and 0.56, respectively. The highest population frequencies of TSER*3 appear in the Mongoloids and the lowest in the Negroids. Negroid and Mongoloid populations also have alleles with TSER*4, TSER*5 and TSER*9 repetitions, with frequencies between 0.18% and 3.76%[7, 9].

In Chile, allele frequencies for genetic markers are known to vary according to socioeconomic stratum. For example, the ABO*A allele of the ABO blood group and the RHD*d allele of the Rh system are more frequent in the higher socioeconomic stratum, which has more genes of European origin, while native Amerindian alleles are more prevalent in the lower socioeconomic stratum [14]. This is explained by the origin of the Chilean population, which is a mixture of Spanish and Chilean aborigines. Since the period of colonization (1541), aborigines and mestizos were considered to belong to a lower social stratum than the Spanish, for whom the government, property and occupations with the highest prestige and power were reserved. Hard labour was given first to aborigines and then to mestizos. These factors led to a cline of mixture between the two extreme classes (Aborigines and Spanish) [14].

The objective of this study was to estimate the allelic frequencies of TSER in two urban subpopulations of Santiago, Chile, according to socioeconomic stratum and to compare these frequencies with those of other populations. There are no studies for the TSER locus in Chilean populations. Our results provide an important starting point for pharmacogenomic analysis in cancer chemotherapy.

Materials and methods

EDTA-blood samples were obtained from healthy individuals from the San José Hospital (SJH), which belongs to the National Health Service and serves mainly low socioeconomic classes, and from Clínica Las Condes (CLC), a private hospital serving principally high socioeconomic classes. The SJH and CLC are located in the northern and eastern parts of Santiago, respectively. The purpose of the study was explained to all individuals and each participant signed an informed consent. The study was approved by the Ethics Committee of the Medicine Faculty of the University of Chile. To determine the TSER variants, DNA was isolated by routine methods and amplified by polymerase chain reaction (PCR) [15]. Two oligonucleotide primers were synthesized specifically to amplify the polymorphic region of TSER (sense primer 5′-GTGGCTCCTGCGT TTCCCCC-3′ and antisense primer 5′-GCTCCGAGC CGG CCACAGGCATGGCGCGG-3′) [6]. Amplification reactions were performed in a 25-µl volume containing 0.2 µg genomic DNA, 0.2 mm dNTP_s, 0.5 µm of each primer, 1.0 units of Taq polymerase (New England BioLabs, Ipswich, MA, USA) and 2.5 µl of 10 × Taq PCR buffer. Cycling conditions were: one cycle at 94 °C for 5 min with hot start; 30 cycles of 40 s at 94 °C, 1 min at 62 °C and 40 s at 72 °C; and one elongation cycle for 5 min at 72 °C. Reaction products were analysed on 4% agarose gels; the 28-bp difference in allele size (250 bp for TSER*3 and 220 bp for the TSER*2) discriminated between alleles [6]. Allele frequencies were calculated by allele counting. No technical problems were encountered in these processes, and less than 2% of the 374 samples could not be genotyped. The fact that the genotypic frequencies observed did not differ from Hardy–Weinberg equilibrium is indirect evidence against technical errors.

The χ² test was used to compare genotype distributions and allele frequencies between the studied groups and to test for Hardy–Weinberg equilibrium in the genotype distributions [16, 17]. For both samples the percentage of aboriginal admixture was estimated (Bernstein) [18], using ABO and Rh (anti-D serum) systems as genetic markers, and published estimates of their frequencies in the parental populations: ABO*O= 0.65, RHD*d= 0.41 for the Spanish population [19], and ABO*O= 0.98, RHD*d= 0.0 [20, 21] for Chilean aborigines. The nominal level of significance for all analyses was P < 0.05 [17].

Results

Table 1 presents the phenotype and allele frequencies for the variants of TSER, ABO and Rh loci and the Amerindian component in the two hospital samples, based on the ABO and Rh systems. The results reveal that the degree of Amerindian admixture in SJH was higher than in CLC, as expected. The TSER*3 allele was more frequent in both groups. The frequencies of the genotypes and alleles were not significantly different between the samples (P > 0.05 in all cases). None of the genotype frequencies of TSER or ABO deviated significantly from those predicted by Hardy–Weinberg equilibrium.

Table 1.

Genotype or phenotype distribution, allele frequencies for TSER, ABO and Rh loci, and Amerindian component in the San José Hospital (SJH) and Clínica Las Condes (CLC) samples

TSERGenotype	HSJNo	Freq. ± CI (95%)	CLCNo	Freq.± CI (95%)
TSER*2/TSER*2	40	22.47 ± 6.13	38	22.00 ± 5.69
TSER*2/TSER*3	86	48.31 ± 7.34	88	46.32 ± 7.09
TSER*3/TSER*3	52	29.21 ± 6.68	64	33.68 ± 6.72
Total	178		190
Allele
TSER*3		53.37 ± 5.18		56.84 ± 4.98
ABO
Phenotype
AB	2	1.40 ± 1.93	2	1,31 ± 1.80
A	50	34.97 ± 7.82	68	44.44 ± 7.87
B	11	7.69 ± 4.37	14	9.15 ± 4.57
O	80	55.94 ± 8.14	69	45.10 ± 7.88
Total	143		153
Allele
ABO*A		20.26 ± 4.66		26.49 ± 4.94
ABO*B		4.66 ± 2.44		5.40 ± 2.53
ABO*O		75.08 ± 5.01		68.10 ± 5.22
Rh
Phenotype
DD+Dd	130	94.89 ± 3.69	137	90.73 ± 4.63
dd	7	5.11 ± 3.69	14	9.27 ± 4.63
Total	137		151
Allele
RHD*d		22.60 ± 4.95		30.45 ± 5.19
Amerindian component		34.51 ± 16.30		15.87 ± 17.60

Table 2 compares the allele frequencies of TSER*3 in SJH and CLC samples with those of previous studies in other populations. The frequency of allele TSER*3 was lower in SJH, but not significantly different (53 ± 2.6%) from that in CLC (57 ± 2.5%); and the overall Chilean sample frequency (55.2 ± 1.8%) was similar to that of American Caucasians [9, 10] and African groups [9], and significantly lower than Chinese (81 ± 1.1%) [7] and Japanese (81 ± 6.1%) [6]. The TSER*3 frequency observed in CLC was significantly higher (P ≤ 0.05; Z-test) than for Italians [12] and Kenyans [9]; however, its frequency in SJH was significantly lower (P ≤ 0.05; Z-test) than that of South-west Asia [8]. Curiously, the frequency reported by Marsh et al.[9] in American Caucasians was between the CLC and SJH (CLC 57%; USA 54%; SJH 53%). Although TSER*4, TSER*5 and TSER*9 repetitions were not found in HSJ and CLC samples, it is possible that these alleles exit at frequencies too low to be observed in a study of this scale.

Table 2.

Frequencies of the TSER * 3 allele in different populations

Population	Alleles, N	TSER*3(% ± SE)	Reference
American Caucasian	208	(54.0 ± 3.4)	[9]
American Caucasian	2280	(54.1 ± 1.0)	[10]
British Caucasian	192	(60.0 ± 3.5)	[8]
Italian	272	(50.0 ± 3.0)	[12]
Spanish	ND	(57.0 ± ND)	[24]
Hungarian	200	(53.5 ± 3.5)	[25]
North Irish	784	(52.2 ± 1.8)	[10]
Australian	182	(56.0 ± 3.7)	[13]
South-west Asian	190	(62.0 ± 3.5)	[8]
Chinese	192	(82.5 ± 2.7)	[8]
Chinese	1336	(81.0 ± 1.1)	[7]
Japanese	42	(81.0 ± 6.1)	[6]
African-American	184	(52.0 ± 3.7)	[9]
Kenyan	196	(49.0 ± 3.6)	[9]
Ghanaian	496	(56.0 ± 2.2)	[9]
Chilean
HSJ	356	(53.0 ± 2.6)	Present study
CLC	380	(57.0 ± 2.5)	Present study
Total	736	(55.2 ± 1.8)	Present study

ND, No data.

Discussion

Considering that genetic markers in the Chilean population have generally shown allele frequencies between those observed in European Caucasians and Amerindians (of Mongoloid origin) [14, 22, 23] and that the percentage of Amerindian admixture in CLC is lower than in SJH, our results for the TSER locus are surprising; we expected to find a greater frequency of allele TSER*3 in HSJ than in CLC. In order to estimate expected frequencies, we must suppose that: (i) the Spanish population which mixed with the indigenous group had a TSER*3 allele frequency similar to that found in the current Spanish population (57%) [24]; (ii) the indigenous Chilean population had a TSER*3 allele frequency similar to that found in current Mongoloid groups (81%) [7]; and (iii) the estimates of admixture are close to the real values. Using Bernstein’s formula [18] and 34.51% and 15.87% as the estimate of degree of Amerindian admixture for SJH and CLC, respectively (Table 1), the expected frequencies would be 65.3% for HSJ and 60.8% for CLC. The frequency of the allele TSER*3 in the Chilean population appears to be incompatible with its ethnic origin. By contrast, all previous population studies of genes in the Chilean population have produced results consistent with the hypothesis of ethnic mixture [14, 22, 23]. Which factors may have been important in producing the current observed frequencies? At present we have no explanation for our results, but we suspect that the frequency of TSER*3 in aboriginal Chilean populations, when estimated, will prove to be much lower than the 80% found in Mongoloid populations [7], which might have occurred by genetic drift during the early colonization of South America.