Analysis of 8 X-chromosomal markers in the population of central Croatia

Aim To analyze 8 X-linked short tandem repeat (STR) markers in the population of central Croatia and to evaluate their forensic efficiency. Methods We carried out a statistical analysis of the data from previously performed genetic analyses, collected during routine forensic work by the Forensic Science Centre ‘‘Ivan Vučetić.’’ Mentype® Argus X-8 PCR amplification kit was used for typing the data of 99 unrelated healthy women and 78 men from central Croatia. Haplotype frequencies were calculated only in male samples. Arlequin 3.5 software was used to assess Hardy-Weinberg equilibrium (HWE), linkage disequilibrium (LD), observed and expected heterozygosity. Power of discrimination (PD) for men and women, polymorphism information content (PIC), power of exclusion, and mean exclusion chance for deficiency cases, normal trios, and duos were determined using online database ChrX-STR.org. Results In female samples, deviations from HWE (P = 0.006) for each locus were not found. LD test performed both on female and male samples revealed no significant association between markers (P = 0.002). DXS10135 was the most polymorphic locus (PIC = 0.931). PD varied from 0.692 to 0.935 in male and from 0.845 to 0.992 in female samples. Combined PD reached 99.999999% in men and 99.9999999999% in women. Conclusion Performed analyses revealed that the studied marker set contained polymorphic markers with high power of discrimination. We can conclude that Mentype® Argus X-8 PCR kit is suitable for application in the population of central Croatia. Results of this study, together with collected allele and haplotype frequencies, are the first step in establishing a national reference X-STR database based on 8 X-STR loci.

The X-chromosome is 155 million base pairs (Mb) long and carries approximately 1250 known genes (1). Men are hemizygous for all X chromosomal markers, while women carry two copies of X chromosome. In the process called X-chromosome inactivation in women, one of their X chromosomes is transcriptionally silenced in a complex and highly coordinated manner according to the Lyon hypothesis (2). A compact structure called Barr body is formed as the inactivated X chromosome condenses. The X chromosome is then stably maintained in a silent state (3). Mutations on the X chromosome occur less frequently than on autosomes due to several times lower nucleotide mutation rate in women than in men, which also reduces the X chromosome genetic diversity (4)(5)(6)(7). The X chromosome also shows faster genetic drift than the autosomes as a consequence of the smaller population size (8). This leads to a more pronounced population structure. In women, the overall rate of recombination is higher than in men and it varies along chromosomes in both sexes (9). Recombination occurs extremely rarely around the centromere and at the region Xq13.3-Xq21.3, which is distant from the Xp telomere (10,11). The X chromosome recombines only in women, therefore, a stronger linkage disequilibrium is observed than in autosomes.
Fathers transmit their X chromosome to daughters as haplotypes. X chromosome-linked short tandem repeat (X-STR) loci analysis is used in paternity testing, more complex defciency paternity cases, when half-sisters and/or grandmothers are to be examined; paternity testing including blood relatives; and rare cases of maternity testing (12). Its advantages are also confirmed in cases when female DNA traces have to be analyzed against a male background. In our study, we used Mentype® Argus X-8 kit which includes markers that are clustered into 4 linkage groups with 2 closely linked markers per group (DXS10135 and DXS8378; DXS7132 and DXS10074; HPRTB and DXS10101; DXS10134 and DXS7423) (13). Therefore, two markers of each group have to be handled as haplotype for genotyping. Tightly linked STR clusters, which segregate as stable haplotypes, are important for solving complex kinship cases. The usefulness of such clusters is determined by the stability against recombination. It has been already documented that using four tightly linked X-STRs provides stable haplotypes and they have been evaluated for forensic work in several studies (14)(15)(16)(17)(18)(19).
To our knowledge, there are no published population data on X-STR diversity in central Croatia. Therefore, the aim of this study was the detailed genetic characterization of 8 X-STR markers in the population of central Croatia. Also, our goal was to define the population structure of this region and to evaluate the forensic efficiency of the used X-STR markers. The association between those 8 X-STR markers was also analyzed. Finally, we performed a population comparison between central Croatian and European and non-European populations.

Study sample
In this study, we followed the guidelines on the use of STRs in forensic analysis created by DNA Commission of the International Society of Forensic Genetics (20)(21)(22)(23).
We carried out a statistical analysis of the data from previously performed genetic analyses, collected during routine forensic work by the Forensic Science Centre ''Ivan Vučetić. '' Data for 99 healthy women and 78 men from the following counties of central Croatia were used: Zagrebačka, Sisačko-moslavačka, Karlovačka, Bjelovarsko-bilogorska, and the city of Zagreb. Participants from all central Croatian counties were included in an attempt to account for any subpopulation variations. The sample size was selected according to the size of the studied population (approximately 2 million) and according to sample sizes in previous studies (16,17,19). The participants were not related and the samples were of sufficient quality and quantity to be included in statistical analysis. The study was approved by the Ethics Committee of the Institute for Medical Research and Occupational Health, Zagreb, Croatia.

DNA analysis
Genomic DNA from all samples from the materials expertise was extracted from Filter Technology Associates (FTA) cards (Whatman, Maidstone, Kent, UK) and buccal swabs (Whatmann) using Chelex (24). Mentype® Argus X-8 PCR amplification kit (Biotype AG, Dresden, Germany) (13) was used for the amplification of 8 X-STRs: amelogenin for sex determination, DXS7132, DXS7423, DXS8378, DXS10074, DXS10101, DXS10134, DXS10135, and HPRTB according to manufacturer's instruction. Mentype® Argus X-8 PCR amplification kit is a highly informative tool for kinship testing, because each of the 4 STR clusters spans less than 0.5 cM and represents a stable haplotype (14,25 CA, USA). Data analysis was performed using Genemap-per® software (version 3.2, Applied Biosystems). Amplicon sizing was performed using an internal size standard (DNA Size Standard 550 -ROX, Biotype AG), and the amplicons were compared with the Mentype® Argus X-8 allelic ladder (Biotype AG) for unambiguous allele designation.

Statistical analysis
The allele and haplotype frequencies were determined by counting. Arlequin 3.5 software (26) was used to assess population parameters, perform statistical inference, and compare allele frequencies between different populations. Hardy-Weinberg equilibrium (HWE) exact test, including observed (Ho) and expected heterozygosity (He), was performed for female samples. Linkage disequilibrium (LD) pair-wise loci test was performed both for female and male samples. Haplotype frequencies were calculated only in male samples. For HWE and LD tests, Bonferroni correction was used to determine the significance level. Significance level for HWE testing was 0.006 and for LD testing 0.002. Genetic heterogeneity within population was estimated as gene diversity for male haplotype data. Power of discrimination (PD) for men and women was calculated according to Desmarais et al (27). Other forensic parameters, such as polymorphism information content (PIC), power of exclusion (PE), mean exclusion chance (MEC) for deficiency cases (Krüger's formula), normal trios (Kishida's formula) and duos (Desmarais' formula) were determined using ChrX-STR.org online database that calculates population-genetic data (28). The haplotype frequencies from central Croatia were compared with those from German, Japanese, and Ghanaian population samples by applying exact test of population differentiation. For comparison with Polish population, exact test based on allele frequencies was used, because no haplotype frequencies were available in ChrX-STR.org online database (28). Exact test based on allele frequencies was also used for comparison with the population of Bosnia and Herzegovina for the loci DXS8378, DXS7132, HPRTB, and DXS7423. Arlequin 3.5 software was used for allele and haplotype frequencies comparison. Significance level for interpopulation comparison was set to 0.025 after Bonferroni correction (26).

ReSultS
We determined allele frequencies, Ho and He, and P values for the HWE of 8 X-STR in the population of central Croatia ( All P-values obtained by HWE test were higher than the significance level that was 0.006 after Bonferroni correction. LD test, which was performed both on female and male samples, revealed no significant association between the studied markers, because all P-values were higher than the significance level after applying Bonferroni correction (0.002) ( Table 2). Although LD test did not found a significant linkage between studied markers, this could be explained by the small sample size.
We compared the allele frequencies from our study to those from the population of southern Poland (29), which is also Slavic in origin. Also, we compared them to the closely located population of Bosnia and Herzegovina (30).  No significant differences were found for both populations; significance level was 0.025 after Bonferroni correction (Table 3).
Haplotype frequencies of four linkage groups (LG) were counted in 78 men. The linkage groups 1, 2, 3, and 4 revealed 37, 30, 35 and 30 haplotypes, respectively (Table   4). Moreover, gene diversity in central Croatian population was higher than 0.95 for each LG ( Table 4). The most frequent haplotypes were 25-12 and 24-12 for LG1; 13-16 for LG2; 13-31.2 for LG3; and 35-15 for LG4. Haplotype frequencies for each LG were compared with haplotype frequencies from Japan, Ghana, and Germany (14) ( Table 5). Significant differences were found for every LG for Japa- nese and Ghanaian population, P-values were lower than the significance level (0.025 after applying Bonferroni correction).
DXS10135 was the most polymorphic locus (with 25 alleles, PIC = 0.9306), whereas the lowest values were observed for DXS7423 and DXS8378 (both with 5 alleles, PIC 0.6316 and 0.6447, respectively). PIC for the whole marker set was 0.999998. PD ranged from 0.6922 to 0.9345 in male samples and from 0.8447 to 0.9918 in female samples. Combined PD reached 99.999999% in men and 99.9999999999% in women (Table 6).

DiSCuSSioN
We found variant alleles 15.3 at the DXS10135 and 30.2 and 37.1 at the DXS10134 loci, which is the first time these alleles were found in the population of central Croatia using Mentype® Argus X-8 PCR amplification kit. A previous study on the population from Bosnia and Herzegovina (30) investigated 4 mini X-STRs, divided into 7 pairs of markers (GATA172D05-GATA31E08, DXS7132-DXS7423, DXS9902-HPRTB, DXS7130-DXS6803, DXS6789-GATA172D05, DXS8378-GATA172D05, and DXS7424-DXS130) but since they did not use Mentype® Argus X-8 kit, we could compare only four common loci, DXS8378, DXS7132, HPRTB, and DXS7423. It is interesting that allele 8 at the HPRTB locus was found only in the population from Bosnia and Herzegovina and allele 11.2 was found only in central Croatia.
Genetic heterogeneity within the population of central Croatia was very high for each LG. Gene diversity for female (equivalent to expected heterozygosity for diploid data) and male data in the population was relatively high and close to 1 for the most informative markers, DXS10135 and DXS10101. The fact that no gametic association was proven between the loci, both in female and male samples, could be explained by high mutation rates for X-STR, which has already been observed in other populations (14,20). Obtained haplotype frequencies should be taken into account when a set of more than a single meiosis is considered (21).

Significant differences in all
LGs between central Croatia, Japan, and Ghana could be explained by a relatively large genetic distance between these populations. Although the population of central Croatia is genetically closer to German population, significant difference in haplotype frequencies was found in LG4. This could be due to large differences in the sample size between our (N = 78 men) and German population (N = 439 men) (14).
Forensic parameters for 8 X-STR markers in central Croatian population were comparable to those published elsewhere (14)(15)(16)(17)(18)(19)(20)22,29). DXS10135, DXS10101, and DXS10134  with the PIC value close to 1 (0.9306, 0.8962 and 0.8478, respectively) increased the discrimination power of 8 X-STRs in central Croatia. Other forensic parameters, such as PD, PE, and MEC indicate that Argus X-8 PCR amplification kit is suitable for forensic and kinship analysis in the population of central Croatia, in cases when autosomal STR markers do not provide the needed information. The X-STR data for central Croatia obtained in this study were submitted to ChrX-STR.org online database (28).
In a complex kinship testing, X-STR genotyping can supplement the analysis of autosomal, mitochondrial, and Y-chromosomal markers. In the last few years, the need for commercially available and validated X-STR kits has increased due to a growing number of complex kinship cases. Intensive studies of the X chromosome discover a lot of closely linked X-STR markers, which can be included in the commercially available kits (31-35).
Recent forensic casework, population genetics, and anthropological studies have used the relatively new commercially available Investigator Argus X-12 kit (36). Investigator Argus X-12 presents an improvement compared to Mentype® Argus X-8 in the sense of increased discriminatory power due to four linkage groups with three markers per group (Amelogenin; DXS10148, DXS10135, DXS8378; DXS7132, DXS10079, DXS10074; DXS10103, HPRTB, DXS10101; DXS10146, DXS10134, DXS7423) (36). In addition, in genotyping three markers of each group should be handled as a haplotype. The kit was successfully validated by Edelmann et al (37). More and more X-STR markers are appearing on the forensic scene. Even though the number of cases that require X-STRs analysis is still quite small, forensic community needs to have an appropriate genetic tool to solve any complex kinship case.
To the best of our knowledge, this is the first population study of 8 X-linked STRs in Croatia. All the analyzed markers were in HWE; therefore Hardy-Weinberg laws could be applied for match probability calculation. The limitation of this study is the relatively small sample size, which could be further extended to test LD.
X-linked markers in Mentype® Argus X-8 PCR amplification kit proved to be highly polymorphic with a high power of discrimination. Mentype® Argus X-8 was shown to be a robust kit that could be used as an additional marker panel for forensic identification, paternity testing, and kinship determination. The collected allele and haplotype frequencies data could help to establish X-STR kin-ship and identification analysis in central Croatia. The results of our study will be included in the Croatian national reference X-STR database based on 8 loci. Further studies are planned to get an overview of the X-STR variability in all Croatian regions, and there are plans for inclusion of 12 X-STR loci in the database. The implementation of additional marker set included in would increase discriminatory power as a very important prerequisite for further enlargement of Croatian X-STR database.