Heterozygosis deficit of polymorphic markers linked to the β-globin gene cluster region in the Iranian population.

OBJECTIVES: Iran is considered as one of the high-prevalence areas for β-thalassemia with a rate of about 10% carrier frequency. Molecular diagnosis of the disease is performed both by direct sequencing and indirectly by the use of polymorphic markers present in the beta globin gene cluster. However, to date there is no reliable information on the application of the markers in the Iranian population. Here we report the results of an extended molecular analysis of five RFLP markers, XmnI, HindIIIA, HindIIIG, RsaI and HinfI, located within the β-globin gene cluster region in four subpopulations of Iran.
MATERIALS AND METHODS: A total of 552 blood samples taken from the Iranian subpopulations including Isfahan, Chaharmahal-O-Bakhtiari, Khuzestan and Hormozgan were genotyped using PCR-RFLP and sequencing. The allele frequency, the expected and observed heterozygosity, and Shannon's information index (I) of these markers were calculated.
RESULTS: Distribution of the allele frequencies for XmnI, HindIIIA, HindIIIG, RsaI and HinfI polymorphic markers did not differ significantly among the subpopulations examined. Overall observed heterozygosity ranged from 0.1706 for HindIIIA to 0.4484 for RsaI. The Shannon index was <1 for all the polymorphic markers in the populations studied. The data indicated that heterozygosity of these markers was low in the Iranian population.
CONCLUSION: The results suggested that genotyping of these markers is not informative enough once used as single markers for prenatal diagnosis and carrier detection of β-thalassemia in the Iranian population. However, haplotyping of these markers may provide more useful data in linkage analysis and prenatal diagnosis as well as carrier detections for β-thalassemia in Iranians.

Introduction

The hemoglobinopathies are the most common inherited single-gene disorders worldwide, resulting from the heterozygote advantage against malaria (1). Among the hemoglobin disorders, α-thalassemias are found with high gene frequencies in African, Asian and the Mediterranean populations, whereas β-thalassemias are particularly prevalent in Asia and the Mediterranean natives (2). The β-thalassemias are a group of anemia which affect the rate of synthesis of normal β-globin chains. β-thalassemia occurs due to the mutations in the β-globin gene cluster located on short arm of chromosome 11 (3). More than 500 mutations have been characterized in β-globin gene region associated with the β-thalassemia disease (OMIM 141900). However, it is revealed that each population has its own spectrum of mutations (4). The beta globin gene cluster region contains five expressed genes with the same transcriptional orientation and one pseudogene (ψβ) between Aγ and δ genes in the order of 5’-ε-Gγ-Aγ-ψβ-δ-β-3’ (5-6). Moreover, the cluster region contains several polymorphic markers throughout chromosome 11, which could be used as useful tools for linkage analysis and molecular diagnosis of beta thalassemia disease (7).

β-thalassemia is one of the most common public health problems in Iran, having the frequency of about 10% in the Iranian population (8). More than two million carriers of β-thalassemia live in Iran (9). Therefore, molecular diagnosis of the disease is highly requested in most molecular diagnosis centers in Iran. The disease is diagnosed both by direct sequencing and indirectly by the use of polymorphic markers present in the beta globin gene cluster.

However, to date there is no reliable information on the application of the markers in the Iranian population. In this study, the aim was to investigate The presence or absence of the restriction sites are represented by (+) and (-) symbols in this table the population indices of five most commonly used polymorphic markers located in the beta globin gene cluster region including XmnI, HindIIIA, HindIIIG, RsaI and HinfI in the β-thalassemia carrier individuals from central and south provinces of Iran.

Materials and Methods

Genomic DNA samples and genotyping

Blood samples were collected from 552 unrelated β-thalassemia carriers referred to the Isfahan Medical Genetics Center, Isfahan, Iran. Informed consent was taken from all the individuals participated in the study. Also, the ethical committee of University of Isfahan has approved the study (approval number 790205). The selected individuals were from the provinces of Isfahan (n=176), Chaharmahal-O-Bakhtyari (n=201), Khuzestan (n=97) and Hormozgan (n=78). Genomic DNA from each individual was obtained from peripheral leucocytes by salting out procedure (10). Five RFLP markers including XmnI 5 to the Gγ gene, HindIII in the IVSII of Gγ and Aγ genes, RsaI 5 and HinfI 3 to the β-globin gene were genotyped using PCR-RFLP technique (11, 12). The primers and PCR-RFLP conditions are available on request.

Statistical analysis

The statistical analysis of data was performed by use of the Popgene32 software version 1.31 (available at http://www.ualberta.ca/~fyeh/-download.htm) and PyPop program (13). The allele frequency, the expected and observed heterozygo-sity, Shannon’s information index (I), Fixation indices and the Ewens-Watterson test were estimated. The D’ values between possible pairings of studied markers were obtained by means of the 2LD program (14).

Results

Five polymorphic markers in the beta globin gene region including XmnI, HindIIIA, HindIIIG, RsaI and HinfI were genotyped in two central populations of Iran, Isfahan and Chaharmahal-O-Bakhtyari, as well as two south populations, Khuzestan and Hormozgan (Figure 1). As shown in Table 1, allele frequency of the markers did not differ significantly among the four Iranian subpopulations. The expected and observed hetero-zygosity, Shannon’s information index (I), and Fis value of the markers are shown in Table 2. The overall observed heterozygosity ranged from 0.1706 (HindIIIA marker) to 0.4484 (RsaI marker). Negative Fis was observed for HinfI markers in Khuzestan and RsaI marker in Hormozgan, indicating an excess of heterozygotes. All these polymorphic markers showed the Shannon index <1 in the studied populations, revealing low heterozygosity of studied markers in these populations of Iran (Table 2).

Figure 1

A typical illustration of HinfI marker genotyping in four DNA samples from the Iranian population. Total genomic DNA was extracted and digested with HinfI restriction enzyme. The digestion products were resolved on 2% agarose gel and visualized following ethidum bromide staining. Lane 1: homozygous -/-, lanes 2 and 3: homozygous +/+, lane 4: heterozygous +/-. M represents 100 bp DNA ladder

Table 1

The frequency distribution of RFLP markers located within β -globin gene cluster region in four subpopulations of Iran

Population	Sample size (2n)	Allele	XmnI	HindIIIA	HindIIIG	RsaI	HinfI
Isfahan	176	-	0.7105	0.8553	0.5592	0.3947	0.2237
		+	0.2895	0.1447	0.4408	0.6053	0.7763
Chaharmahal-O-Bakhtyari	201	-	0.8564	0.8911	0.7228	0.3564	0.1584
		+	0.1436	0.1089	0.2772	0.6436	0.8416
Khuzestan	97	-	0.6915	0.9255	0.6489	0.3830	0.2766
		+	0.3085	0.0745	0.3511	0.6170	0.7234
Hormozgan	78	-	0.8571	0.8929	0.6964	0.5536	0.1071
		+	0.1429	0.1071	0.3036	0.4464	0.8929
Overall	552	-	0.7817	0.8869	0.6567	0.3948	0.1944
		+	0.2183	0.1131	0.3433	0.6052	0.8056

Table 2

The observed and expected heterozygosity, Shannon’s information index (I) and inbreeding coefficient (Fis) of five RFLP markers located within β -globin gene cluster region in four subpopulations of Iran

Population	RFLP marker	Obs. Het.	Exp. Het.	I	Chi-square	P-value	Fis
Isfahan	XmnI	0.3684	0.4141	0.6017	0.941879	0.331795	0.1044
	HindIIIA	0.2368	0.2492	0.4135	0.195192	0.658630	0.0434
	HindIIIG	0.4342	0.4963	0.6861	1.204216	0.272500	0.1192
	RsaI	0.4737	0.4810	0.6708	0.017856	0.893698	0.0087
	HinfI	0.3158	0.3496	0.5315	0.728893	0.393242	0.0907
Chaharmahal-O-Bakhtyari	XmnI	0.2475	0.2471	0.4114	0.000265	0.987015	-0.0066
	HindIIIA	0.1386	0.1951	0.3442	8.822704	0.002975	0.2859
	HindIIIG	0.2970	0.4027	0.5903	7.062978	0.007869	0.2588
	RsaI	0.3762	0.4611	0.6513	3.458935	0.062911	0.1799
	HinfI	0.2574	0.2680	0.4370	0.160687	0.688525	0.0346
Khuzestan	XmnI	0.3191	0.4313	0.6179	3.269645	0.070573	0.2520
	HindIIIA	0.1489	0.1393	0.2650	0.258220	0.611346	-0.0805
	HindIIIG	0.3617	0.4605	0.6481	2.221162	0.136131	0.2062
	RsaI	0.4681	0.4777	0.6655	0.019480	0.889000	0.0096
	HinfI	0.4681	0.4045	0.5897	1.200467	0.273228	-0.1697
Hormozgan	XmnI	0.2143	0.2494	0.4101	0.525135	0.468660	0.1250
	HindIIIA	0.1429	0.1948	0.3405	2.351020	0.125201	0.2533
	HindIIIG	0.2500	0.4305	0.6139	5.176500	0.022894	0.4087
	RsaI	0.6071	0.5032	0.6874	1.238710	0.265720	-0.2284
	HinfI	0.1429	0.1948	0.3405	2.351020	0.125201	0.2533
Total	XmnI	0.2937	0.3419	0.5247	5.060398	0.024479	0.1395
	HindIIIA	0.1706	0.2010	0.3529	5.846419	0.015609	0.1494
	HindIIIG	0.3452	0.4518	0.6432	14.078405	0.000175	0.2343
	RsaI	0.4484	0.4788	0.6709	1.021542	0.312154	0.0617
	HinfI	0.3016	0.3139	0.4926	0.390791	0.531884	0.0373

Fixation indices FIS, FST, FIT were used to analyze genetic variation. As shown in Table 3, the F varied among studied markers from 0.0064 (HindIIIA) to 0.0355 (XmnI) with a mean of 0.0231. The FIT values ranged from 0.0134 in RsaI marker to 0.2553 in HindIIIG polymorphic marker. However, mean FIS (0.1025) and FIT (0.1232) were both positive and greater than zero. The subscripts I, S and T refer to individual, population and total population, respectively. The mean number of migrants per generation among populations (Nm) was reported maximum for the RFLP marker HindIIIA (38.9648) and minimum for XmnI marker (6.7938).

Table 3

F-statistics for five RFLP markers located within β -globin gene cluster region in the Iranian population

RFLP marker	FIS	FIT	FST	Nm
XmnI	0.1350	0.1657	0.0355	6.7938
HindIIIA	0.1344	0.1399	0.0064	38.9648
HindIIIG	0.2422	0.2553	0.0172	14.2843
RsaI	-0.0114	0.0134	0.0245	9.9656
HinfI	0.0177	0.0438	0.0266	9.1389
Mean	0.1025	0.1232	0.0231	10.5792

In Ewens-Watterson test of neutrality for these studied RFLP markers, the observed homozygosity (F) lied inside the limit of 95% confidence region, and the normalized deviate of the homozygosity (Fnd) was estimated negative for all these polymorphic markers (Table 4).

Table 4

The Ewens-Watterson homozygosity test of neutrality for five RFLP markers located within β -globin gene cluster region in the Iranian subpopulations

RFLP marker	K	F		Fnd	L95	U95
XmnI	2	0.6588	0.8539	-1.1768	0.5045	0.9960
HindIIIA	2	0.7994	0.8539	-0.3287	0.5042	0.9960
HindIIIG	2	0.5491	0.8539	-1.8380	0.5053	0.9960
RsaI	2	0.5221	0.8539	-2.0009	0.5045	0.9960
HinfI	2	0.6867	0.8539	-1.0081	0.5035	0.9960

Pairwise linkage disequilibrium values were also estimated using the 2LD programs (Zhao 2004). The obtained results are illustrated in Table 5. The number of samples of Hormozgan population was insufficient, and linkage disequilibrium values were only assessed for Isfahan, Chaharmahal-O-Bakhtyari and Khuzestan populations. D’ values ranged from 0.01516 (HindIIIA-HinfI) to 0.81098 (XmnI-HindIIIG) in the studied Iranian subpopulations.

Table 5

Linkage disequilibrium values for possible pairings of markers located within β-globin gene cluster region in four Iranian subpopulations

Population	D’
		HindIIIA	HindIIIG	RsaI	HinfI
Isfahan	XmnI	0.524066	0.809143	0.026556	0.331211
	HindIIIA		0.691725	0.015976	0.999867
	HindIIIG			0.007165	0.218419
	RsaI				0.879942
Chaharmahal-O-Bakhtyari	XmnI	0.933551	0.832847	0.151087	0.111049
	HindIIIA		0.672099	0.149183	0.227101
	HindIIIG			0.001400	0.054807
	RsaI				0.554129
Khuzestan	XmnI	0.313749	0.707050	0.040465	0.365207
	HindIIIA		1.000000	0.970610	0.231064
	HindIIIG			0.120297	0.250281
	RsaI				0.823589
Total	XmnI	0.58026	0.81098	0.08158	0.27417
	HindIIIA		0.70023	0.09390	0.01516
	HindIIIG			0.02481	0.18810
	RsaI				0.77609

Discussion

The usefulness of the polymorphic markers in linkage study and prenatal diagnosis as well as carrier detection of genetic diseases depends on the frequency and informative status of the markers in each population. As shown in Table 1, the frequency of the minor alleles exceeds 0.3 for HindIIIG and RsaI markers. These data indicate that these two polymorphic markers, HindIIIG and RsaI, have a high degree of polymorphism in the Iranian population. Significant difference between the observed and expected heterozygosity was not observed for almost any of the RFLP markers in the studied Iranian population. The observed heterozygosity was lower than the expected heterozygosity for HindIIIG and RsaI in Hormozgan and Chaharmahal-O-Bakhtyari populations and for XmnI and HindIIIG in the Khuzestan population. These results indicate the presence of heterozygosis deficit of these markers in these populations. The observed heterozygosity of the RFLP markers was lower than 50% in all the studied populations, indicating that these markers are less informative in prenatal diagnosis and carrier detection of β-thalassemia. However, haplotyping of the markers could be more useful than separately genotyping them in the Iranian population.

Analysis of Shannon index indicated that the highest value was observed for RsaI polymorphism in all Iranian subpopulations studied. These data revealed that RsaI polymorphism was the most useful markers among the five RFLP markers investigated. The HindIIIA polymorphism in all the studied populations, and XmnI and HinfI polymorphisms in Hormozgan and Chaharmahal-O-Bakhtyari populations showed a very low Shannon index. These results reveal that these markers have little application in carrier detection and prenatal diagnosis of β-thalassemia disease in the Iranian population.

Moreover, the studied RFLP markers did not show significant deviations from Hardy Wein- berg Equilibrium (HWE) except HindIIIA and HindIIIG polymorphisms in Chaharmahal-O-Bakhtyari population and HindIIIG polymorphism in Hormozgan population. Once the subpopulations were considered as one population, the population only deviated from HWE at XmnI, HindIIIA and HindIIIG polymorphic markers. In the calculation of HWE, it should be noted that the selected individuals were unrelated with no consanguineous relationship (their parents were unrelated). However, whether their parental marriage was random or not, is not clear. Therefore, we cannot definitely exclude any population inbreeding and claim that all marriages in that population were random.

Genetic structure of these populations was analyzed through Wright’s F-statistics. F is the most suitable statistic for studying the overall level of genetic divergence (15). The data showed that mean F was 0.0231. This indicated low level of population differentiation and high inbreeding among the Iranian subpopulations. The values of mean FIS and FIT were both greater than zero. This suggested the presence of a heterozygosis deficit within the studied populations. The value of Nm was 10.5792 (Table 3). When Nm is higher than 1, gene flow is able to offset the differentiation among populations caused by isolation and genetic drift (16).

In the Iranian population, F value lied inside the lower and upper limits of the 95% confidence region of expected F values for all these RFLP markers (Table 4), signifying that these polymorphic markers in β-globin gene region were not under genetic hitchhiking and selection force operating on another locus could not change allele frequency and heterozygosity of these polymorphic markers in this population.

As shown in Table 4, the expected homozygosity () was higher than the observed homozygosity (F) value for all these polymorphic markers in the β-globin gene cluster. The normalized deviate of the homozygosity (Fnd) was also negative. These data could support the presence of balancing selection at the β-globin gene cluster region in the Iranian population.

The results obtained from LD estimation indicated that D’ values for three pairings of the markers, XmnI-HindIIIG, HindIIIA-HindIIIG and RsaI-HinfI, were higher than 0.5 in all three populations. These data indicate the presence of high linkage disequilibrium between these three paired markers. These LD values could provide useful information for selecting suitable markers in molecular diagnosis of beta thalassemia disease. The D’ values close to zero for other pairings of markers showed that they were in linkage equilibrium. Therefore, the two-marker haplotypes XmnI-RsaI, HindIIIA-RsaI, HindIIIG-RsaI, XmnI-HinfI, HindIIIG- HinfI and HindIIIA- HinfI should be used with caution in diagnosis of disease.

Conclusion

In summary, the Iranian population showed heterozygosity deficiency for the RFLP markers including XmnI, HindIIIA, HindIIIG, RsaI and HinfI, located within the β-globin gene region. Indeed, the haplotype phasing is more applicable than genotyping markers separately in the Iranian population. Although a large number of unrelated individuals from different provinces of Iran were included in the present study, one should note that the data cannot be applied to the whole Iranian population. More investigations on other parts of the Iranian population are required. The data obtained from this study could provide additional insight into the genetic structure of β-globin gene region in the Iranian population.