A PRELIMINARY STUDY OF THE ALLOZYME VARIATION IN THE GREY HAMSTER , CRICETULUS MIGRATORIUS ( MAMMALIA : RODENTIA ) , FROM THE ASIAN PART OF TURKEY

The aim of the present study was to evaluate the degree of genetic variation and divergence by cellulose acetate gel electrophoresis between samples of Cricetulus migratorius, a cricetine rodent distributed in the Asian part (Anatolia) of Turkey. Out of twenty allozyme loci scored for fifteen enzyme systems, eleven loci were detected to be polymorphic in at least one locality of the Turkish C. migratorius. Indices of genetic variability (the percentage of polymorphic loci, mean number of alleles per locus, and mean observed and expected heterozygosities) were found to be P(95%) = 28, A = 1.3, Ho = 0.226 and He = 0.218, respectively. Nei’s unbiased genetic distances ranged from 0.000 to 0.153, with an average value of 0.069. The mean gene flow was calculated to be Nm = 0.7484. This is a preliminary study describing the allozymic variations of C. migratorius from Turkey. As there are no extensive data on the allozymic variations of C. migratorius from other regions, our results could not be compared, in detail with those of other populations of the species C. migratorius,

In Turkey, the genus Cricetulus is represented with only one species, the grey dwarf hamster, C. migratorius (Kryštufek andVohralik, 2001, 2009).C. migratorius is distributed in the dry grasslands, steppes and semi deserts of Asia, Europe and Africa and also inhabits agricultural land and gardens (Kryštufek et al., 2008).The grey dwarf hamster appeared in the southern Levant, where the geographical region borders with the Mediterranean, roughly between Egypt and Anatolia (the Asian part of Turkey) 80 000-70 000 years ago (see Musser and Carleton, 2005).In the European (Thrace) and Asian parts (Anatolia) of Turkey, the grey dwarf hamster is a very common and widespread species (Kryštufek and Vohralik, 2009).
Most studies on the grey dwarf hamsters are restricted to morphological and karyological aspects.Based on a number of morphological features, Doğramacı (1989) suggested that this species pos-sess two subspecies (C. m. vernula Thomas, 1917 andC. m. cinerascens (Wagner, 1848)) in Turkey.Contrary to Doğramacı's (1989) view, and based on morphological and geographical observations, Kryštufek and Vohralik (2009) stated that they are suspicious about whether the division of Turkish grey dwarf hamsters within one subspecies is meaningful.Pamukoğlu and Albayrak (1996) reported the external and cranial measurements and fur color of C. migratorius from the Kastamonu province in northern Turkey.The conventional karyotype of the Turkish grey dwarf hamster was first described by Doğramacı and Kefelioğlu (1991).The authors recorded that the diploid number of chromosomes was 2n = 22 for C. migratorius from Turkey.Arslan and Akan (2008) and Aşan et al. (2010) studied the cytogeny of the grey dwarf hamster, C. migratorius, from central Anatolia in Turkey, and they reported the nucleolar organizer regions (NORs), G-and C-banded karyotypes with 2n = 22 for this species.Gharkheloo (2006) added new information on the morphological and karyological peculiarities of C. migratorius in the Zanjan province of Iran and reported the Iranian karyotype with 2n = 22.
The performed molecular studies on the Turkish grey dwarf hamster, C. migratorius are still scarce and there is only a single paper dealing with the genetic variation inferred from molecular markers for this species.Neumann et al. (2006) focused on the molecular phylogeny of the subfamily Cricetinae using the mitochondrial cytochrome b and 12S rRNA genes, and the nuclear vWF gene, in which they evaluated the phylogenetic relationship among C. migratorius sampled from the Kayseri province of Central Anatolia in Turkey and other hamster species.
To our current knowledge, there is no data on the biochemical characteristics of C. migratorius from Turkey.In this study, we present the genetic variation of individuals of the grey dwarf hamster species, C. migratorius, as revealed by allozyme electrophoresis from the Asian part of Turkey.

Grey hamster sampling
A total of nineteen specimens of C. migratorius were randomly sampled from ten ecologically different localities in Turkey (Fig. 1).Specimens were caught with mouse-like traps.Morphologically, these samples were identified in accordance with Kryštufek and Vohralik (2009).Voucher specimens were deposited in the Mammalogical Collection at the Department of Biology, Faculty of Sciences, Erciyes University in Kayseri, Turkey.

Biochemical analysis
Homogenates for the present study were obtained from portions of muscle, liver, and kidney tissues.One g tissue samples were homogenized in 3 ml buffer [with 0.1 per cent (v/v) Triton X-100 in 50 mM Tris-HCl pH 8.0] (Searle, 1985).The homogenates were used for the fifteen enzyme systems (Table 1).In order to separate allozymes by electrophoresis, we used TITAN III cellulose acetate plates (Helena Laboratory, USA), following the protocols described in Richardson et al. (1986), Hebert andBeaton (1993), andSearle (1985).The bands on cellulose acetate gels were scanned using a scanner (ScanMaker 1000XL Microtek) and also checked visually.

Data analysis
The bands observed on the gels were identified according to their electrophoretic mobility relative to that of the fastest band, which was assigned mobility = A, and scored alphabetically for allele designation.These scores were used for further allozyme data analyses.
Data analyses were performed with BIOSYS2 [original release of Sworfford and Slander (1981) modified by William C. Black IV (1997)  The genetic diversity within each population was determined by calculating the allelic frequencies, the percent of polymorphic loci (P), the mean expected and observed (direct count) heterozygosity (H e, and H o ), the mean number of alleles per locus, and matrix of Nei's (1978) genetic distance (D) produced by BIOSYS2, which was used to construct "The Unweighted Pair-Group Method with Arithmetic Averages (UPGMA) and Neighbor-Joining (NJ)" trees by means of NTSYS-Pc 2.1 version (Rolf 2000).In order to estimate the heterogeneity among the populations, Wright's (1978) F statistics, F is (inbreeding coefficient at population level), F it (inbreeding coefficient at total sample level) and F st (proportion of differentiation among the populations), were calculated for polymorphic loci.The overall gene flow (N m ), was estimated according to Wright's (1943Wright's ( , 1951) ) using the equation F = 1/4N m +1 among populations studied.

RESULTS
Twenty loci were consistently scorable for nineteen individuals from ten localities (Fig. 1).Among the twenty loci scored, although nine loci (aat-1, aat-2, g3pd, idh-1, idh-2, lap, mdh-1, mdh-2 and gpi) were monoallelic for the same electromorph in all individuals of the Turkish grey hamster, eleven loci (ak, g6pd, hk, ldh-1, ldh-2, ldh-3, mpi, pgd, pgm, ck-1   The four indices [the mean number of alleles per locus, the percentage of polymorphic loci, the expected and observed (direct-count) heterozygosi- ties] of genetic variability Table 3.In the analyzed, the mean number of locus (A) ranged from 1.0 to 1.6, with an average value of 1.3.The percentage of polymorphic loci (P) varied from 5% to 50%, with an average of 28% under the 0.95 criterion.At each locality, the observed (direct-count) (H o ) and expected (H e ) heterozygosities both ranged from from 0.050 to 0.400, with average values of 0.226 and 0.218, respectively.
The grey dwarf hamster, C. migratorius, displayed a high level of genetic heterogeneity among the samples.Ten of the genetic loci analyzed exhibited negative values of F is (mean -0.4712), which express an increase in heterozygosity, except for g6pd.Nine of the genetic loci analyzed also revealed negative values of F it (mean -0.1028), except for g6pd and ck-1 (Table 4).The coefficients of hierarchical F st for the ten specimens of C. migratorius were higher than zero in all polymorphic loci.The values of F st among all the specimens of C. migratorius varied between 0.0984 (pgm) and 0.4722 (hk), with an average value equalivent to 0.2504.The mean value of F ST calculated for the ten specimens of C. migratorius was relatively high.According to this value, there was 25.04% of the total genetic diversity among the specimens studied (Table 4).
The overall gene flow (N m ), estimated according to Wright's (1943Wright's ( , 1951) ) by using the equation F = 1/4N m +1, which was the number of effective migrants, was calculated to be 0.7484.This value revealed that the gene exchange was relatively low.
The unbiased genetic distances and identities were calculated according to Nei (1978), based on the allele frequencies of the twenty loci observed in the specimens studied.Nei's (1978) genetic distances (D) were found to be relatively low for any pairwise comparison; although the lowest value of genetic distances was determined to be 0.000, the highest value was found to be 0.153.The mean value of Nei's (1978) genetic distances was 0.069 (Table 5).
UPGMA and NJ cluster analyses were based on Nei's (1978) unbiased genetic distances that were carried out to detect the genetic relationships among the specimens of the species C. migratorius in Turkey (Fig. 2, Fig. 3).Analysis of Nei's (1978) unbiased genetic distance (D) matrices, using the UPGMA method, displayed that the Konya-Akşehir sample was relatively distinct from the other samples of Turkish grey hamster.According to this analysis; the Konya-Beyşehir sample was the closest to the Konya-Akşehir sample, which was the most basal in the UPGMA dendrogram (Fig. 2).Analysis of Nei's (1978) unbiased genetic distance (D) matrices among the ten localities, using the NJ method (Fig. 3), showed the existence of two allozymic groups, (I) and (II).According to the NJ analysis, the Kay-Table 5. Genetic similarity and distance for Cricetulus migratorius samples from ten localities in Turkey (Below diagonal: Nei's (1978) unbiased genetic identity, above diagonal: Nei's (1978) unbiased genetic distance).(Codes are indicated in Fig. 1).seri-ERU Campus, Kayseri-Yahyalı, Yozgat-Merkez, Yozgat-Boğazlıyan, Aksaray-Ihlara, Kars, Erzincan and samples clustered together in (I), while Konya-Akşehir and Konya-Beyşehir samples clustered the second group (II) (Fig. 3).The topologies of both UPGMA and NJ trees  based on Nei's (1978) genetic distances were relatively (Fig. 2, Fig. 3).In both trees, the Konya-Akşehir sample was closer to the Konya-Beyşehir sample than to the remaining samples.

DISCUSSION
The subfamily Cricetinae has two genera (Mesocricetus and Cricetulus) in Turkey and identification of these genera is relatively easy, because they can be diagnosed largely by size (condylobasal length), skull of murine shape and proportions, and pelage peculiarities.There is a detailed taxonomic key on the basis of these characteristics for their identification (Kryštufek and Vohralik, 2009).
Genetic difference can be used to comprehend whether or not two or more populations belong to the same taxon (Limeira et al., 2009).Allozyme electrophoresis is a powerful technique for estimating the phylogenetic relationships among populations or taxa, and infers that the enzyme loci can be considered to be important molecular characters.Murphy et al. (1996) asserted that many molecular characters should be used in phylogenetic studies and, in this sense, the number of these characters inferred from allozyme electrophoresis sufficient for revealing the phylogenetic relationships among populations or taxa.
There is no study on the allozymic variations of C. migratorius to which the results found in this study could compare with.Therefore, our results were compared with those determined for other members of the genus Cricetulus and the subfamily Cricetinae or Palaearctic hamsters.Kartavtsev et al. (1984a) investigated the variability of 18-20 protein loci in populations of five species (Cricetus cricetus, Cricetulus barabensis, Tscherskia triton, Phodopus campbelli and Mesocricetus auratus) of the different genera of Palaearctic hamsters and found that the mean heterozygosity (H) of these species ranged from 1.7 to 6.0, with an average value of 3.5.Among the genus Cricetulus, this value was found to be 4.0 for C. barabensis.Kartavtsev et al. (1984b) determined that the values of genetic similarity varied from 0.423 to 0.657, with an average value of 0.524 for the five different species pair.Yiğit et al. (2007) analyzed twenty loci Fig. 3. NJ tree based on Nei's (1978) unbiased genetic distances among the Turkish grey hamsters trapped in different localities (see Fig. 1 for key to OTU abbreviation).
in the five populations of Mesocricetus brandti and M. auratus from Anatolia Iran.They reported that one locus differed between the M. auratus and M. brandti populations.Observed and expected heterozygosities were reported by these authors to be 0.080 and 0.029, and 0.098 and 0.024 for M. auratus and M. brandti, respectively.According to Yiğit et al. (2007), while Nei's (1978) unbiased genetic distances among the populations of M. brandti were found to be relatively low, ranging from 0.006 to 0.025, the genetic distances between M. brandti and M. auratus ranged from 0.102 to 0.122.The level of genetic differentiation inferred from allozyme electrophoresis was higher in the Turkish samples of C. migratorius than those of Palaearctic hamsters.Mezhzhzerin (2001) performed allozyme variation analysis of 22 loci in two subspecies of C. migratorius, C. m. isabellinus de Fillippi, 1897 andC. m. bellicosus Charlemagne, 1916, from Kopetdagh and the steppe zone of Ukraine, respectively.The author observed fixation between the two subspecies at two loci, adh-1 and ck-2.Unfortunately, Mezhzhzerin (2001) did not report on the level of genetic differentiation and genetic distances for the populations of these subspecies.
Finally, the present work is the first study to describe the allozymic variations of C. migratorius.Our results could not be compared, in detail, with those of other populations of the species C. migratorius.At present there are no extensive data on the allozymic characteristics of C. migratorius from other regions.Furthermore, genetic studies can be accomplished in more detail using different genetic markers, i.e. nuclear and mitochondrial DNA.In this sense, these markers may assist in explaining the genetic differentiation between the Turkish populations of C. migratorius.

Table 1 .
Electrophoretic conditions for the fifteen enzyme systems *L: Liver; M: Muscle; **Not scored

Table 2 .
Allele frequencies at the twenty loci observed for Cricetulus migratorius from Turkey (Codes are indicated in Fig.1).

Table 2 .
Continued and est) allele frequencies calculated for the twenty loci analyzed are exhibited in Table2.

Table 3 .
Genetic variability for twenty loci of Cricetulus migratorius from Turkey (Codes are indicated in Fig.1).

Table 4 .
F-statistics analysis for polymorphic loci of Cricetulus migratorius from Turkey.