A Molecular Genetic Study of Hybridization in Four ...

1 Russian Jornal of Genetics, Vol. 38, No. 7, 2002, pp. 796-809. Translated from Genetica, Vol. 38, No. 7, 2002, pp. 950-964. Original Russian Text Copyright © 2002 by Ermakov, Surin, Titov, Tagiev, Luk'yanenko, Formozov

_______________________________________________________________________________ ANIMAL GENETICS _____________________________________________________________________________

A Molecular Genetic Study of Hybridization in Four Species of Ground Squirrels (Spermophilus: Rodentia, Sciuridae) O. A. Ermakov1, V. L. Surin2, S.V. Titov1, A. F. Tagiev2, A. V. Luk'yanenko2, and N. A. Formozov3 1

Department of Zoology and Ecology, Penza State Pedagogical University, Penza, 440026 Russia; e-mail: [email protected] 2 Hematological Research Center, Russian Academy of Medical Sciences, Moscow, 125167 Russia; e-mail: [email protected] 3 Department of Vertebrate Zoology, Lomonosov Moscow State University, Moscow, 119899 Russia Received June 26, 2001; in final form, December 24, 2001. Abstract—Four species of ground squirrel—yellow (Spermophilus fulvus), russet (S. major), small (S. pygmaeus), and spotted (S. suslicus)—occur in the Volga region. Between S. major and S. pigmaeus, S. major and S. fulvus, and S. major and S. suslicus, sporadic hybridization was reported. Using sequencing and restriction analysis, we have examined the mtDNA C region in 13 yellow, 60 russet, 61 small, 45 spotted ground squirrels, and 9 phenotypic hybrids between these species. It was shown that 43% of S. major individuals had "alien" mitotypes typical of S. fulvus and S. pygmaeus. Alien mitotypes occurred both within and outside sympatric zones. No alien mitotypes were found in 119 animals of the other three species, which suggests that only one parental species (S. major) predominantly participates in backcrosses. Phenotypic hybrids S. fulvus x S. major and S. major x S. pygmaeus) were reliably identified using RAPD–PCR of nuclear DNA. However, we could find no significant traces of hybridization in S. major with alien mitotypes. Analysis of p53 pseudogenes of S. major and S. fulvus that were for the first time described in the present study produced similar results: 59 out of 60 individuals of S. major (including S. major with S. fulvus mitotypes) had only the pseudogene variant specific for S. major. This situation is possible even at low hybridization frequencies (less than 1% according to field observations and 1.4 to 2.7% according to nuclear DNA analysis) if dispersal of S. major from the sympatric zones mainly involved animals that obtained alien mtDNA via backcrossing. The prevalence of animals with alien mitotypes in some S. major populations can be explained by the founder effect. Further studies based on large samples are required for clarifying the discrepancies between mitochondrial and nuclear DNA data.

INTRODUCTION By now, natural hybridization events have been recorded in most orders of mammals. Hence, the view of Mayr [1] on rarity of this phenomenon among the members of Mammalia seems obsolete. For instance, S.S. Schwartz [2] lists over 400 remote hybrids, many of them fertile. At the same time, natural hybridization in mammals is clearly less understood than that in fishes, amphibians, reptiles and birds. Many studies of mammalian hybridization do not go beyond reporting the occurrence of hybrids. Hybridization frequency and its effects on variability, including genetic variation, remain largely unclear. The possible evolutionary consequences of hybridization between sympatric species have been scarcely studied because such hybridization until recently was thought to be extremely rare [1]. Members of Marmotinae, in particular ground squirrels, have traditionally been used in cytogenetic

and microevolutionary studies [3-5] including those on interspecific hybridization. The Volga region and adjoining territories are inhabited by four ground squirrel species: yellow (Spermophilus fulvus Lichtenstein, 1823), russet (S. major Pallas, 1778), small (S. pygmaeus Pallas, 1778), and spotted (S. suslicus Guldenstaedt, 1770). The range boundaries of these species are located in this region; some of the areas are sympatric. During the 20th century the range boundaries and the sympatric areas have undergone substantial changes. In the second half of the century, a contact zone of S. major and S. suslicus has appeared on the right bank of the Volga River [6], the sympatric zone of S. major and S. pygmaeus on the left bank of this river has expanded [7], and a small sympatric area of three qround squirrel species (S. major, S. fulvus, and S. pygmaeus) has emerged in the Trans-Volga region near Saratov [8].

PDF created with pdfFactory Pro trial version www.pdffactory.com

2 In the Trans-Volga region, occasional hybridization of S. major with S. pygmaeus [9-12] and S. fulvus [9, 13] was recorded in the areas of overlapping ranges. On the right bank of the Volga River, introgressive hybridization between parapatric species S. pygmaeus and S. suslicus was reported [14], and more recently, hybrids between S. major and S. suslicus were found in that area [15]. According to the field observations, upon sporadic hybridization the proportions of S. major x S. pygmaeus and S. major x S. fulvus hybrids do not exceed respectively 1 [12] and 2-3% [9] of the total number of animals in the sympatric zone. Preliminary field observations indicate that the hybrids are fertile. Nevertheless, our studies that specifically addressed this issue, failed to reveal any effect of sporadic hybridization on morphological variability of parental species in the sympatric zone [12]. The present work was aimed at assessing effects of sporadic hybridization between the four ground squirrel species by molecular-genetic methods. The objectives included estimation of interspecific differences and

intraspecific variation in S. major, S. fulvus, S. pygmaeus, and S. suslicus at the level of the genome and fragments of mitochondrial (C-region) and nuclear (gene p53) DNA; detection of hybrids; and search for the traces of hybridization in the ground squirrel populations. MATERIALS AND METHODS The animals were collected in 1994 through 2000 in 36 localities of the Volga region and adjacent territories. The sampling localities are designated by numerals given in the table before their addresses and showing the corresponding points on the map (Fig. 1). Thirteen individuals of S. fulvus, 60 individuals of S. major, 61 individuals of S. pygmaeus, and 45 individuals of S. suslicus were examined. In addition, we studied 9 putative interspecific hybrids diagnosed using morphological, colorimetric, electroacoustic (S. major x S. pygmaeus, S. major x S. fulvus) [12, 15], and cytogenetic (S. major x S. suslicus; Bystrakova, unpublished data) analyses. The samples and molecular techniques are described in Table 1.

Fig. 1. Sampling localities: a, Spermophilus fulvus; b, S. major; c, S. pygmaeus; d, S. suslicus; e, S. major x S. fulvus hybrid; f, S. major x S. pygmaeus hybrid; g, S. major x S. suslicus hybrid.


3 DNA was isolated from liver tissue specimens stored in 96% ethanol. Small (about 50 mg) tissue pieces were mashed by scalpel, homogenized by grinding in a 1.5-ml test-tube incubated in 0.5 ml of STE buffer for 30 to 60 min, and centrifuged. DNA was isolated from the pellet using a standard procedure that included overnight treatment with sodium dodecylsulfate and proteinase K at 50˚C and subsequent phenol extraction. At first we amplified a 1.2-kb mt DNA fragment covering the full-size C-region flanked by genes for proline and phenylalanine tRNAs using universal primers L15926 and H00651 [16]. Then primer L15926 was substituted by the specific primer MDL1 (TCCACCTTCAACTCCCAAAGC) synthesized on the basis of the sequenced mtDNA gene for tRNA-Pro of the four ground squirrel species examined. Polymerase chain reaction (PCR) was conducted in the reaction mixture containing 50 Mm Tris-HCl (pH 8.9), 20 mM ammonium sulfate, 20 μM EDTA, 170 μg/ml of bovine serum albumin, 200 μM of each dNTP, 2 mM MGCl2, 0.6 μM of each primer, 0.1-0.2 μg DNA and 2 units of Taq polymerase. Amplification was conducted as follows: 1 min. at 94˚C, 1 min at 58˚C, 3 min at 72˚C (35 cycles) for the

pair of primers L15926 and H00651 and 1 min. at 94˚C, 1 min at 62˚C, 3 min at 72˚C (30 cycles) for the pair of primers MDL1 and H00651. The p53 gene fragment was amplified in a standard reaction mixture containing 2 mM magnesium chloride and 0.6 μM of each of primers SC3 and ASC3 [17]: SC3: GCGTGTGGAGTATTTGGATG; ASC3: TATTCTCCATCCAGTGGTTT. PCR was conducted in the following regime: 1 min at 94˚C, 1 min at 58˚C, and 3 min at 72˚C (30 cycles). The PCR fragments obtained were digested with restriction endonucleases RsaI (mtDNA C-region) or HaeIII (p53) for 2 to 4 h or at 37˚C overnight adding 2-4 enzyme activity units directly to the aliquots of the amplification mixtures (5-10 μl). The amplification mixtures (initial and after restriction) were analyzed by electrophoresis in 6% polyacrylamide gel (PAAG), stained with ethidium bromide and visualized in UV light. For sequencing, PCR fragments were eluted after fractionation in 4-6% PAAG. Sequencing was conducted by modified Sanger's method [18] using the above primers.

Table 1. Sampling sites and methods of investigation Number of animals examined Species and locality of collection (sample size given in parentheses) S. fulvus Saratov oblast 25. Rovenskii raion, Yablonevka village (3) 26. Krasnokutskii raion, D'yakovka village (4) Volgograd oblast 30. Bykovskii raion, Bykovo (6) S. major Kirov oblast 1. Malmyzhskii raion, Tatarskaya-Verkhnyaya Gon'ba village (4) Chuvashiya 2. Tsivil'skii raion, Molodezhnyi settlement (1) Ul'yanovskaya oblast 9. Kuzovatovskii raion, Smyshlyaevka village (3) 11. Nikolaevskii raion, Klin village (2) Samara oblast 12. Syzrankii raion, Novye Ozerki village (6) 13. Stavropol'skii raion, Musorka village (16) 14. Pokhvistnevskii raion, Aktivnyi settlement (3) Bashkortostan 15. Aurgazinskii raion, Tolbazy village (5) 16. Ishimbaevskii raion, Makarovo village (4) Saratov oblast 17. Vol'skii raion, Verkhnyaya Chernavka village (5) 20. Engelsskii raion, Lugovskoe village (3)

restriction sequencing analysis of of C-region C-region

RAPDPCR

restriction sequencing analysis of of pseudopseudogene gene p53 p53

1 1

3 4

3 3

1 -

3 4

-

6

6

1

6

2

4

-

1

4

-

1

-

-

1

1 2

3 2

1

1 -

3 2

1 1 -

6 16 3

6 16 3

-

6 16 3

1 1

5 4

2 2

-

5 4

2

5 3

5 3

-

5 3


4 Number of animals examined Species and locality of collection (sample size given in parentheses)

23. Ershovskii raion, Ershov town (4) 24. Dergachevskii raion, Stepanovka settlement (4) S. pygmaeus Saratov oblast 18. Perelyubskii raion, Rubtsovka village (2) 21. Saratovskii raion, Aleksandrovka village (1) 23. Ershovskii raion, Ershov town (2) 24. Dergachevskii raion, Stepanovka settlement (3) 27. Piterskii raion, Novotulka village (18) Volgograd oblast 29. Elanskii raion, Rodinskii settlement (12) 31. Dubovskii raion, Gornyi Balyklei settlement (8) 32. Bykovskii raion, Krasnoselets settlement (6) 33. Svetloyarskii raion, Dubovyi Ovrag settlement (4) Astrakhan' oblast 34. Enotaevskii raion, Nikol'skoe village (4) 35. Akhtubinskii raion, Nizhnii Baskunchak settlement (1) S. suslicus Chuvashia 2. Tsivil'skii raion, Molodezhnyi settlement (1) 4. Batyrevskii raion, Kozlovka village (4) Nizhnii Novgorod oblast 3. Sechenovskii raion, Novoselki settlement (3) Mordovia 5. Atyashevskii raion, Alasheevka village (2) Ul'yanovsk oblast 6. Surskii raion, Surskoe settlement (6) 7. Korsunskii raion, Urino-Karlinskoe village (2) 8. Tsil'ninskii raion, Timirsyany village (2) 10. Nikolaevskii raion, Kuroedovskie Vyselki village (15) 11. Nikolaevskii raion, Klin village (4) Saratov oblast 19. Arkadakskii raion, Sotszemledel'skii settlement (2) Lipetsk oblast 28. Zadonskii raion, Donskoe settlement (1) Ukraine 36. near Odessa (3) Putative hybrids S. major x S. suslicus Chuvashia 2. Tsivil'skii raion, Molodezhnyi settlement (7) S. major x S. fulvus Saratov oblast 22. Engelsskii raion, Voskresenovka village (1) S. major x S. pygmaeus Saratov oblast 24. Dergachevskii raion, Stepanovka settlement (1) Total (188)

For RAPD–PCR of nuclear DNA, we used pairwise combinations of three primers (GF1, L5, and C19 [19]) complementary respectively to fragments of

restriction sequencing analysis of of C-region C-region

RAPDPCR

restriction sequencing analysis of of pseudopseudogene gene p53 p53

1 -

4 4

4 4

-

4 4

1 1 -

2 1 2 3 18

2 2 3 18

-

-

1 1 2 2

12 8 6 4

3 3 3 3

-

-

1 -

4 1

3 -

-

-

-

1 4

-

-

-

2

3

-

-

-

-

2

-

-

-

1 -

6 2 2 15 4

3 4

-

-

-

2

2

-

-

-

1

-

-

-

1

3

-

-

-

3

7

-

-

6

1

1

1

-

1

1 33

1 188

1 109

1 5

1 83

genes for human glucose-6-phosphate and gammaglobin and the region of human Alu-repeat specific for primates. PCR was conducted in a standard


5 reaction mixture in the following regime: 1 min. at 94˚C, 2 min at 55˚C, 5 min at 72˚C (35 cycles). The reaction products were separated by electrophoresis in 6% PAAG (20 x 20 cm) in a unit Protean II (BioRad, United States). Interspecific and interpopulation differences was estimated from the number of nucleotide substitutions in aligned sequences not accounting for microdeletions/microinsertions. Distances between DNA sequences were computed using Kimura's mutation model [20].

downstream of the tRNA-Pro gene in 27 individual of four ground squirrel species from different populations (this number comprised four phenotypic hybrids). Based on mtDNA analysis in animals from allopatric populations isolated by geographic barriers, we identified four major species-specific mitotypes designated A (S. fulvus), B (S. major), C (S. pygmaeus), and D (S. suslicus). Figure 2 presents the complete nucleotide sequence of the mtDNA region studied in S. fulvus and sequence differences in 23 animals from several populations of each species and in four phenotypic hybrids. Note the nonuniform distribution of the differences within the mtDNA fragment: the presence of a rather large highly conserved region (positions 97-155) lacking any changes and a number of highly variable parts. A small region between positions 264 and 280 exhibited particularly high variation; this region did not show virtually any similarity in pairwise comparisons S. major, S. fulvus/S. pygmaeus, S.suslicus. Genetic distances between populations of the four ground squirrel species are given in Table 2. Figure 3 presents an UPGMA dendrogram showing phylogenetic relationships between the mitotypes of different species and within-species groups revealed in our study.

RESULTS AND DISCUSSION The nature of inheritance of the mitochondrial genome (maternal transmission) may promote more prolonged maintenance of hybridization "traces" in the form of alien mitotypes in the population. Being a regulatory mtDNA region, the C-region (also called the D-loop region) exhibit higher variability than, for instance, the cytochrome b gene, which is traditionally studied by genosystematics. This property of the Cregion seemed to us very important, as we were interested in estimating genetic variation both among several closely related species and within each of these species. We first established the primary structure of the 315-bp C-region fragment located immediately

Table 2. Genetic distances (below diagonal) and the number of nucleotide substitutions (above diagonal) between the mitotypes of different species and intraspecific groups of ground squirrels

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

S.f. 26; S.m. 20 S.f. 25 S.m. 13; S.m. x S.p. 24 S.m. 15; S.m. 11-1 S.m. 16; S.m. 1 S.m. 12; S.m. 11-2 S.p. 18 S.p. 23 S.p. 32-1 S.p. 32-2 S.p. 29 S.p. 31 S.p. 33-1 S.p. 33-2 S.p. 34 S.s. 3 S.s. 10 S.s. 2 S.s. 36 S.m. x S.f. 22 S.m. x S.s. 2-1 S.m. x S.s. 2-2

1 0.003 0.130 0.114 0.116 0.351 0.350 0.474 0.363 0.425 0.419 0.487 0.386 0.400 0.361 0.445 0.513 0.591 0.389 0.120 0.007 0.636

2 1 0.123 0.107 0.109 0.353 0.351 0.478 0.364 0.427 0.420 0.484 0.386 0.404 0.363 0.450 0.520 0.602 0.391 0.113 0.003 0.647

3 25 24 0.014 0.007 0.414 0.408 0.466 0.391 0.402 0.361 0.412 0.468 0.374 0.365 0.403 0.439 0.495 0.317 0.022 0.116 0.529

4 23 22 4 0.007 0.327 0.324 0.365 0.310 0.319 0.330 0.375 0.366 0.358 0.292 0.440 0.419 0.473 0.292 0.014 0.114 0.508

5 23 22 2 2 0.359 0.355 0.401 0.340 0.350 0.361 0.412 0.402 0.374 0.319 0.403 0.539 0.495 0.304 0.014 0.116 0.529

6 47 47 48 44 46 0.007 0.045 0.018 0.035 0.137 0.143 0.136 0.133 0.104 0.297 0.283 0.351 0.235 0.298 0.377 0.331


7 47 47 48 44 46 2 0.034 0.010 0.026 0.113 0.119 0.135 0.117 0.089 0.338 0.320 0.393 0.261 0.296 0.375 0.373

8 53 53 50 46 48 11 9 0.021 0.007 0.104 0.131 0.148 0.110 0.114 0.392 0.290 0.316 0.214 0.376 0.511 0.299

9 48 48 47 43 45 5 3 6 0.014 0.088 0.112 0.108 0.094 0.082 0.280 0.265 0.326 0.218 0.283 0.388 0.309

10 51 51 48 44 46 9 7 2 4 0.100 0.125 0.142 0.107 0.111 0.260 0.278 0.305 0.190 0.329 0.454 0.289

11 50 50 46 44 46 23 21 20 18 20 0.010 0.036 0.065 0.040 0.342 0.291 0.333 0.232 0.341 0.391 0.347

6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

S.f. 26 S.f. 25 S.m. 13; S.m. x S.p. 24 S.m. 15; S.m. 11 S.m. 16; S.m. 1 S.m. 12; S.s. 11 S.p. 18 S.p. 23 S.p. 32-1 S.p. 32-2 S.p. 29 S.p. 31 S.p. 33-1 S.p. 33-2 S.p. 34 S.s. 3 S.s. 10 S.s. 2 S.s. 36 S.m. x S.f. 22 S.m. x S.s. 2-1 S.m. x S.s. 2-2

12 52 52 48 46 48 24 22 23 21 23 3 0.052 0.084 0.056 0.419 0.352 0.403 0.279 0.391 0.446 0.419

13 49 49 50 46 48 24 24 25 21 25 9 12 0.082 0.035 0.368 0.318 0.325 0.256 0.332 0.411 0.335

14 51 51 48 47 48 25 23 22 20 22 15 18 17 0.082 0.446 0.386 0.372 0.268 0.347 0.384 0.385

The differences between S. major and S. fulvus, which I.M. Gromov et al. [21] assign to the subgenus Colobotis, were 6.7–7.6%; between S. pygmaeus and S. suslicus (subgenus Citellus), 12.1–15.2%; between species from different subgenera, 13.3–17.8%. These results confirm the view of systematics that species of the subgenus Colobotis are similar whereas Citellus species are more distant from each other [21]. Each of the species displayed differentiation at the subspecific and interpopulational level (Fig. 2b; Table 2). Maximum differences among populations were 0.6 in S. fulvus and 1.9% in S. major. These data conform to the current view on low intraspecific variability in S. fulvus and S. major. At the same time, populations of different S. pygmaeus subspecies from the right and left Volga banks different by 5.7–7.9% nucleotide substitutions. The southern (S. p. ellermani) and northern (S. p. brauneri) subspecies of S. pygmaeus of the right Volga bank differ by 2.9-5.7%. The S. suslicus subspecies from the Volga region (S. s. guttatus and S. s. suslicus) differ by 1.3–2.5%. However, according to our results, the 36-chromosome form of S. suslicus from the right bank of Dnepr (subspecies S. s. meridiooccidentalis), which several authors [22, 23] regard as an independent species, differ from S. suslicus from the Volga region by 6.7– 7.3%, which is comparable to the subspecific differences in S. pygmaeus from the Volga banks. In this part of our study, we found alien mitotypes in three individuals of S. major and in a phenotypic

15 48 48 46 42 44 21 19 22 18 22 10 13 9 18 0.307 0.267 0.286 0.218 0.268 0.386 0.302

16 51 51 48 49 48 41 43 41 40 39 42 45 44 47 41 0.014 0.024 0.122 0.433 0.457 0.030

17 53 53 49 48 49 40 42 41 39 40 40 43 42 45 39 4 0.026 0.114 0.481 0.530 0.031

18 55 55 51 50 51 44 46 43 43 42 43 46 42 44 41 7 7 0.171 0.532 0.616 0.003

19 48 48 43 41 42 36 38 35 35 33 36 39 38 38 35 24 23 29 0.302 0.406 0.188

20 24 23 6 4 4 43 43 47 42 45 45 47 45 47 41 49 50 52 42 0.121 0.566

21 2 1 23 23 23 48 48 54 49 52 49 51 50 50 49 51 53 55 49 24 0.662

22 56 56 52 51 52 43 45 42 42 41 44 47 43 45 42 8 8 1 30 53 56 -

hybrid S. major/S. suslicus. Two S. major animals from populations 11 and 12 had mitotypes C characteristic of S. pygmaeus; one animal of this species from population 20 and a hybrid had mitotype A typical of S. fulvus. To screen the total sample for alien mitotypes, we used restriction analysis. The mtDNA fragment examined had ten recognition sites for restriction endonuclease RsaI. Four of these sites (sites 3, 5, 9, and 10) were shared by all the four ground squirrel species whereas six sites were species-specific (Fig. 2a). In addition, two more variable RsaI sites were found in the 3' part of the Cregion. Despite of the fact, that restriction analysis permits detection of only a small part of nucleotide substitutions, this method proved to be highly informative and could reveal not only interspecific, but also intraspecific differences. In all, we identified nine different restriction patterns (Fig. 4). Fragments nonspecific for the given species were checked by sequencing. If nonspecific fragments occurred in several animals from one population, only one to two specimens (six specimens in all) were checked. In all cases the sequencing and restriction data coincided, which shows rather high reliability of restriction analysis. The major mitotypes variants detected by the latter method virtually completely coincided with the clades of the dendrogram (Fig. 3) constructed from Cregion sequences; parameters of the mitotypes variants are given below.


7

Fig. 2. (a) Sequence of the C-region fragment of S. fulvus mtDNA aligned relative to analogous sequences of the other three ground squirrel species. RsaI recognition sites are designated by arrows. RsaI sites resulting from nucleotide substitutions in other ground squirrel species are shown by asterisks. (b) C region variants of the four ground squirrel species. Only variable positions are given; their numbers correspond to the complete aligned sequences. Deletions and identical nucleotides are designated by hyphens and dots, respectively. Abbreviated designations and numbers of populations are given at the left (designations 32-1, 32-2, etc. refer to different individuals from the same population); mitotype variants according to restriction analysis data are presented in parentheses.

Fig. 3. The UPGMA dendrogram showing phylogenetic relationships between mitotypes of different ground squirrel species and intraspecific groups. Designations as in Fig. 2. Mitotypes detected by restriction analysis are given after brackets at the right.


8 A comparison of the results of restriction analysis and sequencing show that the differences in the nucleotide substitution number between different mtDNA variants in each ground squirrel species (3– 7%) significantly exceed the variability level within any variant (0.7–2%). Only in S. fulvus the same restriction variant was found in all of the 13 animals studied; we designated this variant A. Characteristic fragments of 126 and 60 bp are formed because of RsaI sites 2 and 7 specific for S. fulvus (Fig. 2a). The highest number of restriction variants (four) was observed in S. pygmaeus. All of them (C1, C2, C3, and C4) were derivatives of the basic mitotypes C. The 135-bp fragment typical of subtypes C1 and C2 and absent in C3 and C4 was characterized by a

polymorphic RsaI site in the 3' part of the C-region. The 68-bp fragment in C2 and 73-bp fragment in C3 appear due to the presence of sites 6 and 8, respectively, within a 80-bp fragment, which is seen as a continuous band (79-80 bp) (Fig. 4). Variants C1 and C2 were found only in S. pygmaeus from the TransVolga region; C3 and C4, in the right-bank populations of this species (Fig. 5). On the right bank of Volga, the ratio of variants C3 and C4 in S. pygmaeus colonies gradually changes. At the northern range boundary (populations 21 and 29), all animals examined had variant C3; variant C4 appears in more southern regions where it occurs at low frequencies (population 31) but it predominates in the colonies situated to the south of Volgograd (populations 33 and 34).

Fig. 4. Patterns of mtDNA C-region fragments in four ground squirrel species obtained by restriction analysis. All 9 RsaI restriction patterns are presented; their designations are given below the electrophoregram. Fragments sizes (bp) are given at the right.

In S. suslicus, two restriction fragments (D1 and D2) were detected. These variants differed by a fragment of about 200 bp, which was formed due to another polymorphic RsaI site in the 3' part of the Cregion. In our electrophoretic system the fragments of 80 bp (characteristic of D1) and 79 bp (characteristic of D2) migrate at virtually the same level. Variant D2 was found only in S. suslicus from the vicinity of Odessa (population 36). Animals from all other colonies of this species (populations 2-8, 10-12, 28) had exclusively the D1 variant (Fig. 5). In 34 out of 60 individuals of S. major (the data on other specimens of this species are given below), two restriction variants caused by a C–T substitution at position 158 in site 4 were detected. Without this substitution the 80-bp fragment splits into fragments of 44 and 36 bp (Fig. 4 does not show fragments of lengths greater than 200 and smaller than 60 bp). Variants of the major mitotypes B of S. major were

designated B1 (–80 bp) and B2 (+80 bp). Animals with variant B1 inhabit S. major colonies in Saratov (populations 17 and 24) and Kirov (population 1) oblasts, as well as a colony on the right bank of the Belaya River (population 16). This variant was also recorded in phenotypic hybrids S. major/S. fulvus (population 22) and S. major/S. pygmaeus (population 24). Individuals of S. major having variant B2 prevail in the Samara Trans-Volga region (population 14) and in the interfluvial area of rivers Belaya and Ufa (population 15). This variant was also found in the colony from the right bank of Volga in Ul'yanovsk oblast (population 11) (Fig. 5). The remaining 26 individuals of S. major (43.3% of the total sample of this species) had mitotypes characteristic of other ground squirrel species. In 11 and 15 of these 26 animals, respectively mitotype C2 typical of S. pygmaeus from the Trans-Volga region and A specific for S. fulvus were recorded (Fig. 5).


9

Fig. 5. The distribution of mitotypes in four ground squirrel species of the Volga region.

An example of screening different populations of S. pygmaeus and S. major is shown in Fig. 6. In particular, the electrophoregram in this figure demonstrates that one S. major animal from population 23 has the mtDNA variant characteristic of S. pygmaeus. Specimens of S. major having mitotypes untypical for this species were found both in S. major–S. pygmaeus and S. major–S. suslicus sympatric zones and outside them. In a "pure" S. major population 20 from the left bank of the Bol'shoi Karaman River, which was nevertheless situated in a S. major–S. pygmaeus sympatric zone, none of the three collected animals had the mitotypes typical of this species. Out of these three animals, two had the A mitotypes (specific for S. fulvus) and one, C1 mitotype (specific for S. pygmaeus). It is known that in the first half of the 20th century this territory was inhabited by S. fulvus [24] (at present the northern boundary of its range is situated 50 km to the south) and in the early 1980s, S. major started to spread from there southward along the Volga bank. However, S. major having the S. fulvus

mitotype were found also in the northern part of the S. major range. Specimens of S. major having the S. fulvus mitotype predominated (11 animals of 16) in population 13 located at the left Volga bank in Samara oblast (about 300 km to the north from the S. fulvus range boundary). This mtDNA variant was detected in S. major and in five out of seven S. major/S. suslicus hybrids in a colony from Chuvashiya isolated from the main right-bank area of S. major (population 2). On the left bank of Volga, in addition to population 20, the S. pygmaeus mitotype was recorded in one out of four individuals of S. major (Fig. 5) collected in a joint S. major–S. pygmaeus colony on the upper Malyi Uzen' River (population 23). Individuals of S. major having mitotype C1 specific for the leftbank S. pygmaeus predominate in the right-bank area of S. major from the city of Ul'yanovsk to Syzran'. The C1 variant was found in eight out of ten animals collected in this area. Interestingly, some of the rightbank S. major colonies (e.g., population 12) apparently consist only of animals with mitotype C1.


10

Fig. 6. Restriction (RsaI) screening of the mtDNA C-region of S. pygmaeus and S. major from different populations. The ordinal number of populations is given after the abbreviated name of the species. Mitotype variants are presented below. R is nondigested PCR fragment of mtDNA.

The presence of alien mtDNA has not been yet revealed in S. pygmaeus, S. fulvus, and S. suslicus. This may be explained by the fact that backcrosses of hybrids occurred (and still occur) mainly with one of the parental species, S. major. However, alien mitotypes may be found in these three species in samples of larger size taken from the contact zones. Interspecific differences at the nuclear DNA level were examined using RAPD–PCR as well as the tumor suppressor gene p53. We used the RAPD-PCR procedure developed by us earlier. It is distinguished by the reaction mixture containing two unrelated and relatively long (20–25 nucleotides) primers [19], which allowed us to obtain fingerprints with numerous characteristic PCR products. A RAPD fingerprint obtained using primers GF1 and L5 is presented in Fig. 7. The electrophoregram demonstrates that the rather complex PCR fragment patterns are generally species-specific, but have some differences that can be ascribed to variation among populations (e.g., see S. pygmaeus from populations 23, 24, 27) and among individuals (most pronounced in S. fulvus from population 30 and S. major from population 24). The S. major group from population 12 having the C1 mitotype appears to be the most homogeneous. The total similarity of both mitotypes and fingerprints of animals from this group indicates their close relatedness (the founder effect). The least

informative data for primers GF1 and L5 were obtained for S. suslicus. This species patterns contained only a small amount of short fragments, and it was problematic to identify among them even one fragment specific for the given species. RAPD–PCR conducted with the two other primer combinations (GF1/C19 and L5/C19) yielded similar results. In spite of a number of drawbacks (numerous PCR fragments characteristic of individual variability, anonymity of the revealed structures, and a possible discrepancy between the observed and actual number of differences, e.g., due to the presence of heteroduplex PCR fragments), the RAPD–PCR technique can provide valuable information in population studies similar to the present work. In addition, we used RAPD–PCR for more reliable identification of phenotypic hybrids S. major x S. fulvus (population 22) and S. major x S. pygmaeus (population 24). Analysis of the fingerprints obtained confirmed the hybrid origin of these animals. The electrophoregram in Fig. 8 distinctly shows fragments characteristic of both parental species in the pattern of a phenotypic hybrid S.major/S. pygmaeus. This hybrid inherited six out of seven fragments specific for S. pygmaeus and three out of seven fragments specific for S. major. Morphologic and electroacoustic examination also showed some predominance of the S. pygmaeus traits in this animal.


11

Fig. 7. RAPD-PCR patterns of nuclear DNA of four ground squirrel species from different populations (primers GF1 and L5). Marker HpaII: pUC19 digest.

Fig. 8. Nuclear DNA fragment patterns of S. pygmaeus, S. major, and their hybrids obtained by RAPD-PCR with primer pair C19/L5.


12 At the same time RAPD–PCR did not reveal any principal differences between nuclear DNA in S. major having typical and alien mitotypes. The next stage of nuclear DNA analysis in the four ground squirrel species was examination of differences in the tumor suppressor gene p53. A fragment of this gene located between exons 6 and 9 was amplified in PCR using the primers specific for mRNA of the human p53 gene [17]. One of these primers (ASC3) was completely, and the other (SC3), partly homologous to the known mRNA sequences of ground squirrel S. beecheyi [25].

When DNA specimens of S. major and S. fulvus were used as amplification template, in addition to the expected PCR product over 1 kb in size we observed a shorter major fragment of 378 bp (Fig. 9a), which was absent in S. pygmaeus and S. suslicus. Sequencing showed that this additional fragment was a processed copy of the corresponding p53 gene region, i.e., is a pseudogene derivative. The unlikely assumption that mRNA admixture in the DNA specimens served as a template for its synthesis is contradicted by the fact that a preliminary treatment of the specimen with DNase results in disappearance of the additional PCR product whereas a treatment with RNase does not affect it.

Fig. 9. PCR amplification (a) and restriction analysis (b) of a pseudogene p53 fragment of S. major, S.fulvus, and their hybrid. Fragments 1300 and 378 bp belong to gene and pseudogene p53, respectively.

Earlier, pseudogenes have been found only in some mouse and rat species [26-29]. In all cases it was represented by nonfunctional, greatly distorted processed copy of the original gene. By contrast, the nucleotide sequence of the p53 pseudogene that we detected in ground squirrels S. major and S. fulvus in its examined region practically did not differ from the corresponding mRNA. Apparently, this gene arose relatively recently and even may be expressed. A sequence comparison of S. major and S. fulvus pseudogenes p53 between codons 210 and 317 (codon numeration is given relatively to human pseudogene p53) revealed only one difference between them, i.e., a G–T substitution at position 3 of codon 249 involving the HaeIII recognition site, which significantly facilitated testing. Upon digestion with HaeIII the 378bp amplified pseudogene p53 region is cut into 233-bp and 145-bp fragments in S. major and remains intact in S. fulvus (Fig. 9b). Restriction analysis of DNA of S. fulvus, S. major, and their hybrid (sample sizes were 13, 60, and 1 animals, respectively) showed that the pseudogene fragment was not digested in all S. fulvus individuals, while only characteristic 233-bp and 145-bp fragments were observed in 59 S. major individuals. Pseudogene p53 was heterozygous in the S. major x S. fulvus

phenotypic hybrid from population 22 (Fig. 9b) and in one S. major individual from population 17 (right bank of Volga, Vol'sk region, Saratov oblast). The latter case is particularly interesting because that population, from which six animals were collected, is outside of the modern sympatric zone of S. major and S. fulvus. The presence of a heterozygous animal in a right-bank colony of S. major can be explained by the fact that the "Volsk area" of the species range was formed in the late 1970s by immigrants from the left bank of Volga [7], where in the first part of the 20th century a S. fulvus population was located right opposite of Vol'sk (see collection of Zoological Museum of Moscow State University, specimen no. S-32533). The southern boundary of the S. major range was in the same territory (along the Bol'shoi Irgyz River). Thus, a S. major–S. fulvus contact zone existed in that area, which made the appearance of hybrids very likely. Pseudogene p53 of S. major was found also in a S. major x S. pygmaeus phenotypic hybrid (population 24) and in six out of seven S. major x S. suslicus phenotypic hybrids (population 2). However, the lack of own pseudogenes in S. pygmaeus and S. suslicus did not allow us to establish homozygosity or heterozygosity of these hybrids with regard to this character.


13 The results of RAPD-PCR and the analysis of a particular nuclear DNA fragment (pseudogene p53) show that hybridization in species pairs S. major–S. pygmaeus and S. major–S. suslicus is currently sporadic. According to Mayr's classification [30], it falls into category b, i.e., "occasional or frequent appearance of hybrids between sympatric species with some of the hybrids capable of backcrossing with one or both parental species." RAPD–PCR analysis of 41 S. major individuals (11 animals from sympatric and 30 animals from allopatric populations) and 31 S. pygmaeus individuals (8 animals from sympatric and 23 animals from allopatric populations) showed coincidence of characters of the both species only in the phenotypic hybrid (1.37% of the total number of animals examined); reliable traces of hybridization have not been yet found. The restriction analysis of pseudogene p53 in the joint group of S. major and S. fulvus revealed two heterozygous animals (2.74% of the total sample of two species), one of which was a phenotypic hybrid. Thus, the results obtained by two different methods of nuclear DNA analysis in general coincide with the field data that testify that the proportion of hybrids in ground squirrel populations is less than 1%. This shows the possibility of using these methods for studying current natural hybridization and for identification of first-generation hybrids. According to the C-region mtDNA analysis, the proportion of animals with alien mitotypes in S. major populations is 15–30 times higher than the proportion of phenotypic hybrids found in the wild. It would seem that the number of hybrids is underestimated because of their more cryptic nature as compared to animals of the parental species. However, preliminary field observations testify against this assumption. The frequency of recording hybrid animals is not likely to be underestimated (given also the fact that we may additionally have recorded in the field secondgeneration hybrids, i.e., offspring of the backcrosses of the first-generation hybrids with one of the parental species). Thus, at present no accumulation of hybrids

and blurring of the species boundaries is observed in the sympatric zones of S. major, S. pygmaeus, and S. fulvus in the Trans-Volga region, i.e., hybridization of these species is not introgressive. Factors that ensure morphological and physiological discreteness of S. major preventing it from introgression of foreign genes require further research. The observed high proportion of alien mitotypes in populations of S. major, including those located far from the modern sympatric zones may result from interspecific hybridization that have occurred in the remote past on a larger scale than at present. On the other hand, this situation could occur relatively recently even at the currently observed low hybridization rate, if S. major populations with high proportion of animals that had acquired alien mtDNA from backcrossing predominantly colonized the area along Volga to the north of the sympatric zones. In the cases of a small number of colonizing animals, the founder effect could take place. Apparently, this was the mode of formation of S. major populations on the right-bank territory near Samara, the origin of which is dated back to the first half of the 20th century [31, 32]. It may well be that animals with the C1 mitotype prevailed among the small number of russet ground squirrels that appeared on the Volga right bank in this region. Later, when this group grew in number, these animals became predominant. The origin of population 2, which was apparently founded by a few S. major individuals with mitotype A, may be the similar. This reasoning may also explain the distribution of alien mitotypes in S. major populations on the left bank of Volga. Evidently, the recurrent fluctuation of the range size results from migration of animals from marginal populations. On the other hand, it is known that hybridization rate in marginal populations may be higher than in the center of the sympatric zone [33]. Thus, animals carrying an alien mitotype can actively participate in colonization of new territories. Upon recurrent range fluctuations, this must lead to their wide distribution outside sympatric zones.

ACKNOWLEDGMENTS We thank V.S. Lebedev and S.S. Zborovskii for their help in mathematical treatment of the data, V.Yu. Il'in, S.B. Luk'yanov, N.V. Bystrakova, and D.G.

Smirnov for assistance in collecting the material, V.A. Lobkov for providing material on S. suslicus from Odessa, and A.V. Misyurin for synthesis of oligonucleotide primers used in this work.

REFERENCES 1. Mayr, E., Populations, Species, and Evolution, Cambridge, Mass.: Harvard Univ., 1970. 2. Schwartz, S.S., Ekologicheskie Zakonomernosti Evolyutsii (Ecological Regularities at Evolution), Moscow: Nauka, 1980.

3.

Vorontsov, N.N. and Lyapunova, E.A., Chromosome Numbers and Speciation in Holarctic Ground Squirrels and Marmots (Sciuridae, Xerinae, Marmotinae), Byull. Mosk. O–va Ispyt. Prir., Otd. Biol., 1970, no. 3, pp. 112–126. 4. Mezhzherin, S.V., Brandler, O.V., Lyapunova, E.A., et al., Genetic Relationships and Differentiation of


14 Palearctic Ground Marmots Marmotinae Pocock, 1923 (Rodentia, Sciuridae), Genetika (Moscow), 1999, vol. 35, no. 6, pp. 756–764. 5. Steppan, S.J., Akhverdyan, M.R., Lyapunova, E.A., et al., Molecular Phylogeny of the Marmots (Rodentia: Sciuridae): Tests of Evolutionary and Biogeographic Hypotheses, Syst. Biol., 1999, vol. 48, no. 4, pp. 715–734. 6. Titov, S.V., Current Distribution and Changes in Number of Spotted Ground Squirrel Spermophilus suslicus in the Eastern Part of Its Range, Zool. Zh., 2001, vol. 80, no. 2, pp. 230–235. 7. Ermakov, O.A. and Titov, S.V., Dynamics of the Range Boundaries in Russet Ground Squirrel Spermophilus major (Rodenta, Sciuridae) from the Volga Region, Zool. Zh., 2000, vol. 79, no. 4, pp. 503–509. 8. Ermakov, O.A., Current Distribution of Yellow Ground Squirrel Spermophilus fulvus Licht. in the Trans-Volga Region, in Fauna i ekologiya zhivotnykh (Fauna and Animal Ecology), Penza: PGPU, 1997, no. 2, pp. 25–28. 9. Bazhanov, V.S., Ground Squirrel Hybrids (To the Problem of Interspecific Hybridization in Nature), Dokl. Akad. Nauk SSSR, 1944, vol. 12, no. 7, pp. 321–322. 10. Denisov, V.P., On Hybridization of Species of the Genus Citellus Oken, Zool. Zh., 1963, vol. 42, no. 12, pp. 1887–1889. 11. Formozov, N.A. and Nikol'skii, A.A., Sound Signal of a Hybrid of Russet and Small Groud Squirrels (Citellus major X C. pygmaeus), Vestn. Mosk. Univ., Ser. 16: Biol., 1986, no. 4, pp. 3–8. 12. Ermakov, O.A., Russet and Small Groud Squirrels of the Volga Region: Distribution and Interrelationships, Cand. Sci. (Biol.) Dissertation, Moscow: Moscow State Univ., 1996. 13. Ognev, S.I., Zveri SSSR i prilezhashchikh stran. Gryzuny (Animals of the Soviet Union and Adjacent Countries: Rodents), Moscow: Akad. Nauk SSSR, 1947, vol. 5. 14. Denisov, V.P., Relationships between Spotted and Small Groud Squirrels in the Contact Zone of Their Species Ranges, Zool. Zh., 1961, vol. 40, no. 7, pp. 1079–1085. 15. Titov, S.V., Relationships between Spotted and Russet Groud Squirrels in the Recently Formed Sympatry Zone, Cand. Sci. (Biol.) Dissertation, Moscow: Moscow State Univ., 1999. 16. Kocher, T.D., Thomas, W.K., and Meyer, A., Dynamics of Mitochondrial DNA Evolution in Animals: Amplification and Sequencing with Conserved Primers, Proc. Natl. Acad. Sci. USA, 1989, vol. 86, pp. 6196–6200.

17. Sugimoto, K., Hirano, N., Toyoshima, H., et al., Mutations of the p53 Gene in Myelodysplastic Syndrome (MDS) and MDS-Derived Leukemia, Blood, 1993, vol. 81, pp. 3022–3026. 18. Tagiev, A.F., Surin, V.L., Gol'tsov, A.A., et al., The Spectrum of Beta-Thalassemia Mutations in Azerbaijan Republic, Hum. Mutat., 1993, vol. 2, pp. 152–154. 19. Surin, V.L., Bannikova, A.A., Tagiev, A.F., et al., Molecular Systematics of Hedgehog (Erinaceidae, Insectivora) of Northeastern Palearctics: The Testing of a New Method, Dokl. Akad. Nauk, 1997, vol. 353, pp. 278–280. 20. Kimura, M., A Simple Method for Estimating Evolutionary Rates of Base Substitutions through Comparative Studies of Nucleotide Sequences, J. Mol. Evol., 1980, vol. 16, pp. 111–120. 21. Gromov, I.M., Bibikov, D.I., Kalabukhov, N.I., et al., Ground Marmots (Marmotinae) in the Fauna of the Soviet Union, Mlekopitayushchie (Mammals), Moscow: Nauka, 1965, vol. 3, no. 2. 22. Zagorodnyuk, M.V. and Fedorchenko, A.A., Allopatric Species among Rodents of the Group Spermophilus suslicus (Mammalia), Vestn. Zool., 1995, nos. 5-6, pp. 49–58. 23. Korablev, V.P., Distribution of Chromosome Forms of Spotted Ground Squirrel Spermophilus suslicus Guld., 1770, Tezisy soveshchaniya "Redkie vidy mlekopitayushchikh Rossii i sopredel'nykh territorii" (Proc. Conf. "Rare Species of Russia and Adjacent Regions), Moscow, 1997, p. 50. 24. Orlov, E.I., Yellow Ground Squirrel Citellus fulvus Licht., in Materialy k poznaniyu fauny Povolzh'ya (Materials to Study the Fauna of the Volga Region), Saratov, 1929, no. 4. 25. Rivkina, M.B., Cullen, J.M., Robinson, W.S., and Marion, P.L., State of the p53 Gene in Hepatocellular Carcinomas of Ground Squirrels and Woodchucks with Past and Ongoing Infection with Hepadnaviruses, Cancer Res., 1994, vol. 54, pp. 5430–5437. 26. Zakut-Houri, R., Oren, M., and Bienz, B., A Single Gene and a Pseudogene for the Cellular Tumor Antigen p53, Nature, 1983, vol. 306, pp. 594–597. 27. Tanooka, H., Ootsuyama, A., Shiroishi, T., and Moriwaki, K., Distribution of the p53 Pseudogene among Mouse Species and Subspecies, Mamm. Genome, 1995, vol. 6, pp. 360–362. 28. Weghorst, C.M., Buzard, G.S., Calvert, R.J., et al., Cloning and Sequence of a Processed p53 Pseudogene from Rat: A Potential Source of False 'Mutations' in PCR Fragments of Tumor DNA, Gene, 1995, vol. 166, pp. 317–322. 29. Prager, E.M., Boursot, P., and Sage, R.D., New Assays for Y Chromosome and p53 Pseudogene


15 Clines among East Holstein House Mice, Mamm. Genome, 1997, vol. 8, pp. 279–281. 30. Mayr, E., Animal Species and Evolution, Cambridge: Belknap, 1963. 31. Bazhanov, V.S., Current Distribution and History of Russet Ground Squirrel Citellus major, Izv. Akad. Nauk Kaz. SSR, Ser. Zool., 1948, vol. 63, no. 8, pp. 27–49. 32. Popov, V.A., Mlekopitayushchie VolzhskoKamskogo kraya (nasekomoyadnye, rukokrylye, gryzuny) (Mammals of the Volga–Kama Region: Insectivores, Bats, and Rodents), Kazan', 1960. 33. Panov, E.N., Species Borders and Hybridization in Birds, Gibridizatsiya i problema vida u pozvonochnykh. Trudy zoomuzeya MGU (Hybridization and the Problem of Species in Vertebrates: Proc. Zool. Museum of the Moscow State Univ.), Moscow: Mosk. Gos. Univ., 1993, vol. 30, pp. 53–96.
