The FokI family of short interspersed repetitive elements (SINEs) has been found only in the genomes of charr fishes (genus Salvelinus). In an analysis of the insertion of FokI SINEs using PCR, we characterized six loci at which FokI SINEs have been inserted into the genomes of Salvelinus alpinus (Arctic charr) and/or S. malma (Dolly Varden). An analysis of one locus (Fok-223) suggested that a sister relationship exists between S. alpinus and S. malma and the SINE at this locus might have been inserted in a common ancestor of these two species, being fixed in all extant populations examined. By contrast, SINEs at two other loci (Fok-211 and Fok-206) were present specifically in the genome of S. alpinus, with polymorphism among populations of this species. Moreover, the presence or absence of the SINEs of the other three loci (Fok-214, Fok-217, and Fok-600) varied among populations of these two species. The most plausible interpretation of this result is that SINEs, which were ancestrally polymorphic in the genome of a common ancestor of these two species, are involved in an ongoing process of differential sorting and subsequent fixation in the various populations of each species.
A retroposon is defined as a nucleotide sequence, present initially as a cellular RNA transcript, that has been reincorporated into the genome via a cDNA intermediate. This process is called retroposition (Rogers 1985; Weineret al. 1986). Short interspersed repetitive elements (SINEs; Singer 1982) form one group of retroposons, and they are often present at more than 105 copies per genome (Okada 1991a,b; Okada and Ohshima 1995).
SINEs can be divided into two classes according to their origins. One class of SINEs is derived from the 7SL RNA (Weiner 1980; Ullu and Tschudi 1984) in the signal recognition particle (SRP) that is involved in the secretion of polypeptides during protein biosynthesis (Walter and Blobel 1982). The primate Alu family and the rodent type 1 (B1) family belong to this class (Schmid and Maraia 1992; Deininger and Batzer 1993). Members of the other class of SINEs originated from specific tRNAs. All known SINEs, with the exception of the primate Alu and rodent B1 SINEs, are members of the second class of SINEs (Okada 1991a,b; Okada and Ohshima 1995).
It is generally accepted that, in contrast to DNA transposable elements, a single unit of a SINE is never subsequently excised precisely except in cases of gross deletions. Moreover, it is believed that the insertion sites of SINE units are almost random (Kidoet al. 1995). These features allow SINEs to serve as excellent evolutionary and phylogenetic markers (Okada 1991b). Indeed, we propose that SINE insertion analysis is one of the best methods for the determination of phylogenetic relationships among closely related species (Murata et al. 1993, 1996; Shimamuraet al. 1997).
SINE insertion analysis is also useful for the studies of structures of populations of a species. In cases where SINEs were retroposed recently on an evolutionary time scale, they have not yet been fixed in populations of the species. For example, several Alu elements that were amplified recently have not yet been fixed within the human genome, and the distribution of these elements varies among geographically distinct groups of the world's population (Batzer et al. 1991, 1994; Batzer and Deininger 1991; Pernaet al. 1992; Hammer 1994; Kasset al. 1994). Batzer et al. (1994, 1996) demonstrated the African origin of Homo sapiens by using these elements as phylogenetic markers.
In a previous study, we characterized three different families of SINEs in the genomes of salmonid fishes (Kidoet al. 1991). We designated these SINEs the salmon SmaI family, the charr FokI family and the salmonid HpaI family. The SmaI family of SINEs is restricted to the genomes of chum salmon (Oncorhynchus keta) and pink salmon (O. gorbuscha). The charr FokI family of SINEs is present only in species that belong to the genus Salvelinus. The salmonid HpaI family of SINEs is present in all species in the family Salmonidae but not in other species (Matsumotoet al. 1986; Kido et al. 1991, 1994, 1995; Koishi and Okada 1991; Murata et al. 1993, 1996). Recently, we found another family of SINEs in coregonid fishes. This family of SINEs is almost identical to the SmaI family and it was designated the SmaI-cor family (SmaI family in cor egonids; Hamadaet al. 1997). Among these four different families of SINEs, it seems likely that the SmaI family of SINEs was amplified relatively recently because of its restricted distribution and the limited sequence divergence of members of this family (Kidoet al. 1991). The discovery that all the SmaI SINEs in the genome of chum salmon are polymorphic and not fixed among populations of the species supports this hypothesis (Takasakiet al. 1997).
Charr species have attracted the interest of many evolutionary biologists because of their highly variable life-history strategies, phenotypic plasticity, and potential for sympatric morphological divergence. Behnke (1980) and Cavender (1980) proposed that the genus Salvelinus includes six major morphologically distinct species, namely, Salvelinus fontinalis (brook trout), S. namaycush (lake trout), S. confluentus (bull trout), S. leucomaenis (white-spotted charr), S. malma (Dolly Varden) and S. alpinus (arctic charr). The continuous circumpolar distribution of S. alpinus has been demonstrated in the Arctic, and S. malma occurs sympatrically with S. alpinus in the northern Pacific. A sister relationship between S. alpinus and S. malma was confirmed in numerous studies by reference to both morphological and biochemical markers (Phillips et al. 1992, 1995; Stearley and Smith 1993).
S. alpinus and S. malma are recognized as distinct species (Behnke 1972; Morrow 1980; Reistet al. 1997). However, the relationships among populations of S. alpinus and S. malma in the North Pacific, as well as the relationships between S. alpinus and S. malma in the North Pacific and the other Asian charrs, have been the subject of controversy because of the variable morphology and life histories of these fishes (Behnke 1980, 1984; Cavender 1980; Savvaitova 1995). Savvaitova (1980) performed a morphological study, and claimed that S. malma should be considered synonymous with S. alpinus. She also proposed that the designation “species-complex” provides a more appropriate description of the complex structure of these two species, claiming that all Arctic charr and Dolly Varden should be included in one superspecies, namely, the S. alpinus complex (Savvaitova 1995).
To elucidate the complex structure of populations of these two species, we attempted to analyze the insertions of FokI SINEs in S. alpinus and S. malma. As is the case for the SmaI SINEs in the genome of O. keta (Takasakiet al. 1997), FokI SINEs were found to be highly polymorphic in the genomes of S. alpinus and S. malma, and to be useful for elucidation of the complex evolutionary history of these two species. The FokI SINEs are the second example to date of highly polymorphic SINEs.
MATERIALS AND METHODS
DNA samples: Individuals from each of the six species of charr from various locations were examined, as shown in Table 1. Total genomic DNA of each species was extracted as described by Blin and Stafford (1976) for large-scale preparation for the establishment of genomic libraries. For the analysis of populations, DNA was extracted from samples of individual fish as follows. One-half gram of liver, muscle or a whole fry was homogenized on ice in TNE solution, which contained 10 mm Tris-HCl (pH 8.0), 100 mm NaCl, and 1 mm EDTA. Then lysis buffer, which contained 500 μg/ml proteinase K, 2% sodium dodecyl sulfate (SDS), 10 mm Tris-HCl (pH 8.0), 150 mm NaCl, and 10 mm EDTA, was added to the solution, with incubation at 50° for 2 to 3 hr. DNA was extracted with phenol and chloroform, washed with chloroform and isoamyl alcohol, and collected by ethanol precipitation.
Construction and screening of genomic libraries, subcloning and sequencing: Total genomic DNA from S. alpinus and S. malma was separately digested with EcoRI for construction of a genomic library for each species. Digests were size-fractionated by sucrose gradient (10 to 40%, w/v) centrifugation. DNA fragments of 2 to 4 kb were ligated with λgt10 arms (Stratagene, La Jolla, CA) and then packaged in vitro. Screening was performed with an end-labeled oligonucleotide, designated FokI (see positions 20–40 in Figure 1), as the probe. Hybridization was allowed to proceed at 42° overnight in a solution of 6× SSC (SSC is 0.15 m NaCl, 0.015 m trisodium citrate, pH 7), 1% (w/v) SDS, 5× Denhardt's reagent [1× Denhardt's reagent is 0.02% (w/v) Ficoll 400, 0.02% (w/v) polyvinylpyrrolidone, and 0.02% (w/v) bovine serum albumin], and 100 μg/ml herring sperm DNA. Washing was performed in 2× SSC plus 1% SDS at 50° for 30 min. Positive phage clones were isolated and their inserts were subcloned into pUC18 or pUC19. Then the inserts were sequenced with primers that corresponded to or were complementary to the consensus sequence for the FokI family.
Amplification by PCR: When a unit of the FokI family appeared to have been integrated at a single locus within a genome, we synthesized 5′ and 3′ oligonucleotide primers (Oligo1000 DNA synthesizer; Beckman, Fullerton, CA). The sequences of primers are shown in Figure 2. The reaction mixtures for amplification by PCR contained Tth buffer (TOYOBO, Tokyo, Japan), 0.2 mm dNTPs (Pharmacia, Uppsala, Sweden), 100 ng of each primer, 1 μg of genomic DNA, and 2 units of Tth DNA polymerase (TOYOBO) in a final volume of 100 μl or 50 μl. The thermal cycling involved 30 repeats of denaturation at 93° for 1 min, annealing at 55° for 1 min, and extension at 72° for 1 min. The products of PCR were analyzed by electophoresis in 2% (w/v) NuSieve GTG and 1% (w/v) Seakem GTG agarose gels (FMC BioProducts, Rockland, ME).
Southern hybridization: Products of PCR were transferred from gels to GeneScreen Plus membranes (New England Nuclear Research Products, Boston) in 0.4 m NaOH and 0.6 m NaCl and then dried. For detection of a SINE unit of the FokI family, hybridization was performed with the Fok1 oligonucleotide as the probe, as described above and under the same conditions as those used for screening. For detection of orthologous loci of Fok-217, an end-labeled oligonucleotide that contained the flanking sequences of the SINE unit at that locus was used as a probe (Fok-217 flan: AGCCCTGCAGTTGCAGACGGTGCAGTTGCT). For subsequent rehybridization with a different probe, the first probe was removed by incubation in 0.4 m NaOH at 42° for 30 min.
Characterization of FokI SINEs in the genus Salvelinus: Previous studies in our laboratory showed that the charr FokI family of SINEs appeared to be restricted to species in the genus Salvelinus, such as S. malma, S. leucomaenis leucomaenis, S. leucomaenis pluvius, and S. namaycush. This family of SINEs was presumed to have been amplified at the time at which the genus Salvelinus diverged from other genera (Kidoet al. 1991). To confirm this hypothesis, we performed a dot hybridization experiment using the FokI sequence as a probe, including samples of DNA from other fishes in the genus Salvelinus (S. fontinalis, S. confluentus, and S. alpinus). DNAs from all fishes in the genus Salvelinus gave strong hybridization signals, confirming that they all contained the FokI family of SINEs (data not shown).
We constructed genomic libraries for S. alpinus from Overvatn Salangen and S. malma from Montana Creek and screened them for FokI SINEs. Five clones that contained the sequence of a FokI SINE were isolated from the genomic library of S. alpinus and they were designated Fok-(SA) 206, Fok-(SA) 211, Fok-(SA) 214, Fok-(SA) 217, and Fok-(SA) 223 (SA stands for S. alpinus). One clone was isolated from the genomic library of S. malma; it was designated Fok-(SM) 600 (SM stands for S. malma). An alignment of the SINE sequences of these clones, together with those obtained from a genomic library of S. leucomaenis leucomaenis (Kidoet al. 1991), is shown in Figure 1). The ten sequences of FokI SINEs were very similar, and we failed to find a distinct subfamily structure of the type reported for other SINE families from salmonids (Kido et al. 1994, 1995; Takasaki et al. 1994, 1997; Hamadaet al. 1997).
To estimate the times of insertion of the FokI SINEs, we performed an analysis by PCR. We determined the 5′- and 3′-flanking regions of each FokI sequence and synthesized a set of primers that flanked each unit, as shown in Figure 2. We then performed PCR using genomic DNA from S. alpinus, S. malma, S. leucomaenis, S. confluentus, S. namaycush, and S. fontinalis as templates.
As shown in Figure 3, a SINE at the Fok-206 and Fok-211 loci was found specifically in the genome of S. alpinus (Figure 3, A and B). We also found that a SINE at another locus, namely Fok-600, was specific to S. malma (Figure 3F). SINEs at three other loci, namely, Fok-214, Fok-217, and Fok-223, were found in the genomes of both S. alpinus and S. malma (Figure 3, C–E). In Figure 3, A–F, the upper DNA fragments (black arrowheads) indicate the presence of a FokI SINE, while the lower DNA fragments (white arrowheads) indicate the absence of such SINEs. To confirm that orthologous loci have been faithfully amplified in these cases, we determined the sequences of the orthologous loci, which do not contain a SINE unit, of S. malma (Fok-206 and Fok-211) and S. leucomaenis (Fok-214, Fok-217, Fok-223, and Fok-600; Figure 2).
To examine whether these species-specific insertions and the group-specific insertions of FokI SINEs were fixed or dimorphic in other populations, we analyzed DNA from 59 specimens of S. alpinus from 5 regions and 95 specimens of S. malma from 12 regions. The results are summarized in Table 2.
The FokI SINEs specific to S. alpinus are dimorphic among populations of this species: At the Fok-206 and Fok-211 loci, the FokI SINEs were not fixed among populations of S. alpinus (Table 2). For example, Figure 4 shows the results of PCR for the Fok-206 locus with specimens of S. alpinus from various populations. Individuals were scored as homozygous (+/+ or −/−) or heterozygous (+/−) for the presence of the 400-bp fragment (+) or the 238-bp fragment (−). The insertions were fixed in individuals from Maine in the United States, whereas, in the specimens from Loch Garry, Scotland, one individual was homozygous (+/+), three individuals were heterozygous (+/−), and five individuals were homozygous (−/−). Moreover, two specimens from Lake Inari, Finland, were homozygous (+/+) and five were heterozygous (+/−). No insertions were observed in the specimens from Lake 103 in the western Arctic of Canada and from Lake Hazen in the eastern Arctic of Canada.
As summarized in Table 2, in the case of the Fok-211 locus, insertion of a FokI SINE was only detected, with a heterozygous pattern (+/−), in the genomes of specimens of S. alpinus from Lake Inari, suggesting that the SINE was inserted very recently at this locus in an individual in this population or a closely related population. All specimens of S. malma from all populations had no insertion at either locus.
S. alpinus and S. malma form a monophyletic group: We next examined whether the FokI SINEs that have been found to be commonly inserted into the genomes of S. alpinus and S. malma in the pilot experiment (Fok-214, Fok-217, and Fok-223; Figure 3) were fixed among populations of both species.
A SINE at the Fok-223 locus was fixed in every specimen from 17 populations examined (Table 2). Although we cannot exclude the possibility that, in the other remaining populations of both species, the SINE at this locus is dimorphic, the present results favor the conclusion that the SINE at this locus is fixed in all individuals of both species. This locus provides the first evidence, from SINE insertion analysis, that S. alpinus and S. malma form a monophyletic group.
The FokI SINEs common to S. alpinus and S. malma are not fixed among populations of the two species: In the case of Fok-217 (Figure 5), insertion of the SINE in S. malma was observed only in a few populations, namely, those of the Klutina River/Lake and the Firth River [Figure 5A(a)], whereas insertion of the SINE in S. alpinus was observed in every population except for that in Lake Hazen [Figure 5B(a)]. In the case of Fok-214, insertion of the SINE were fixed in all the populations of S. alpinus examined and were presented in the population of S. malma in the Klutina Lake/River, but not presented in the populations of other S. malma (Table 2).
In the case of Fok-217, we confirmed the presence of the SINE in the upper fragment in Figure 5 and the validity of amplification by PCR of the orthologous locus by Southern hybridization with the FokI sequence and the flanking sequence of the SINE, respectively, as probes [Figure 5, A (b and c) and B (b and c)].
In the case of the Fok-600 locus, we demonstrated that insertion of FokI units was dimorphic among populations of S. malma, with the exception of the population from the Yanbetsu River (Table 2). Moreover, we also found insertion of a SINE at the Fok-600 locus in individuals from one population of S. alpinus, namely, the population from Lake Hazen. Thus, the insertion of a FokI SINE at the Fok-600 locus was not specific to S. malma but was common to S. malma and S. alpinus, and it was not fixed among the populations of the two species.
Our results indicated that several insertions of FokI SINEs that were common to S. alpinus and S. malma were not fixed among populations of the two species. We found that the insertions were not only intraspecifically polymorphic but also interspecifically variable. Such intraspecific polymorphism and interspecific variation of insertions of SINEs might reflect complex processes of speciation in which two closely related (sub)species diverge by subdivision into genetically distinct populations to form the more-distinguishable species (see below).
SINEs can be used as efficient tools for the determination of phylogenetic relationships among species: It is believed that a SINE is amplified in the germ cells of one individual and then spreads within a population through sexual reproduction and random genetic drift in the same manner as other changes in the DNA (Kimura 1983). As the frequency of the SINE increases gradually in the population, the genomes of members of the population must necessarily be polymorphic. However, among the many SINEs examined to date, intraspecific polymorphism of SINEs is very rare. Most SINEs, including the HpaI SINEs in salmonids, have been fixed in all populations of a given species (Takasaki et al. 1994, 1996). These results suggest that the period between the divergence of any two salmonid species and the present time has been long enough for SINEs to become fixed in all populations of each species. On the basis of this observation and the synapomorphic character of SINEs, phylogenetic relationships among species in the genus Oncorhynchus were determined in our laboratory (Murata et al. 1993, 1996). In this study, the existence of the SINE at the Fok-223 locus in S. malma and S. alpinus suggests a monophyletic relationship among the two species of Salvelinus (Figure 6).
Are populations of S. alpinus or S. malma monophyletic? In this study, we found intraspecific polymorphism and interspecific variation when we examined the insertion of SINEs in a large number of populations of the two charr species. It is possible that such patterns of insertion directly reflect the phylogenetic relationships among the populations. For example, can the results for Fok-214 be interpreted to indicate the monophyly of the population of S. malma in Klutina Lake and of all the populations of S. alpinus? Does the population of S. malma in Klutina Lake need to be reclassified as S. alpinus? In general, species have been identified from morphological characteristics, but it is well known that morphological characteristics are plastic and can respond to variations in the environment. Such phenotypic plasticity has caused many problems in the taxonomic analysis of charr (for review, see Behnke 1980, 1984; Cavender 1980; Savvaitova 1995; Phillips and Oakley 1997). Thus, the possibility exists that the present taxonomic status of several populations of the two species might be incorrect.
The interpretation, in terms of phylogeny, of the results for Fok-214 is, however, inconsistent with the interpretation of the results for Fok-600, which indicated that all populations of S. malma, with the exception of the populations in the Yanbetsu River, and S. alpinus from Lake Hazen have a SINE insertion at this locus. In addition, the interpretation of the results for Fok-600 is inconsistent with the interpretation of the results for Fok-217. Furthermore, it is very unlikely from a taxonomic perspective that the population of S. malma in Klutina Lake/River actually belongs to S. alpinus (Reistet al. 1997).
Accordingly, we must conclude that the patterns of insertions of FokI SINEs that we found do not necessarily reflect the actual phylogenetic relationships among the populations of these species.
Ancestral polymorphism is the most plausible explanation for variations in the presence or absence of a SINE at a given locus: To date, only two examples of intraspecific polymorphism of SINEs have been reported, namely, Alu SINEs in human populations and SmaI SINEs in populations of chum salmon, and each example has been shown to be useful for population analysis (Batzer et al. 1994, 1996; Takasakiet al. 1997). SINEs that are polymorphic among the populations of S. alpinus and S. malma provide a third example. In addition, these SINEs provide the first example of a useful tool for clarification of the processes of dispersion and sorting of ancestrally polymorphic SINEs among populations during speciation, as discussed below.
In this study, we clearly showed that the presence or absence of insertions of FokI SINEs (Fok-214, Fok-217, and Fok-600) varied in populations of S. alpinus and S. malma. The most plausible explanation for these intraspecific and interspecific variations is that they are the result of ancestral polymorphism.
Insertions of FokI SINE at the Fok-214, Fok-217, and Fok-600 loci might have occurred in a common ancestor of S. malma and S. alpinus. Before fixation of each SINE at these loci in the ancestral species, the speciation of S. malma and S. alpinus must have occurred and the polymorphic SINEs were inherited by and sorted to populations of the two new species. After speciation, the SINEs at the Fok-214, Fok-217, and Fok-600 loci were fixed, lost, or became polymorphic as a result of random genetic drift in each population. There is a possibility of such a situation having occurred if the population size of the ancestral species was too large, or the period between the divergence of these two species and the present time was too short to be fixed in all populations of each species. Therefore, the present observations might reflect the ongoing processes of sorting of ancestrally polymorphic SINEs toward fixation or loss in populations of the two new species during speciation. The complete fixation of the SINE at the Fok-223 locus in S. malma and S. alpinus might indicate that the insertion of this SINE at this locus occurred relatively soon after divergence of the ancestral species of S. malma and S. alpinus from other species of Salvelinus (Figure 6).
The significance of trans-species polymorphism in evolution was first indicated by Jan Klein and his colleagues (Klein 1987; Figueroaet al. 1988) in the major histocompatibility complex loci of the mouse and rat. Since SINEs described here appear to be neutral, the present analysis provides a more general scheme of how trans-species polymorphism behaves during the speciation of two closely related species.
Hybridization is another possibility: Hybridization between S. malma and S. alpinus, with successive introgression of nuclear information from one species to the other, provides another explanation for these variations, although this explanation appears less likely.
There is no good evidence of intermediates formed by hybridization between S. malma and S. alpinus (McPhail 1961; Behnke 1980). S. malma and S. alpinus are now considered valid species in view of the identification of several lakes in which the two species live sympatrically with no signs of hybridization (DeLacy and Morton 1943; McPhail 1961). However, there have been suggestions of the existence of hybrids between S. malma and S. alpinus (Volobuyevet al. 1979), and some evidence for gene flow from S. alpinus to S. malma has been presented (Gharrettet al. 1991). In the present case, we cannot yet completely rule out the possibility of hybridization because we did not analyze sympatric populations of the two species. However, introgression seems less likely than ancestral polymorphism, at least in the case of the populations of S. alpinus in Lake Hazen. Lake Hazen, at the northern end of Ellesmere Island in the Canadian high Arctic, is the largest body of freshwater in the world that is located entirely north of the Arctic Circle (Johnson 1990). Because the only species of fish in this lake is S. alpinus (Reistet al. 1995), it is unlikely that polymorphism of SINE insertions could be a result of hybridization between S. alpinus and S. malma. The addition of sympatric populations of the two species to our analysis will allow us to clarify and perhaps even eliminate the possibility of hybridization.
SINEs will be a useful tool for distinguishing one population from another: In this study, we found that the distribution of FokI SINEs at distinct loci varied among remote populations. As described above, such patterns of insertion do not necessarily reflect the phylogenetic relationships among the populations. However, we can use them as genetic markers to distinguish genetic variations. For example, at the Fok-600 locus, among the populations of S. malma that we showed, only the population from the Yanbetsu River had no insertion of a FokI SINE. The population of S. malma in Lake Shikaribetsu and its inlet stream, the Yanbetsu River, on Hokkaido Island, Japan, is thought to be a subspecies of S. malma (S. malma miyabei). S. malma miyabei has the most numerous gill-rakers, which range in number from 23 to 29 with a mode of 26, of all species of Salvelinus (Maekawa 1977). In addition, from an analysis of isozymes, S. malma miyabei was found to differ markedly from other populations of S. malma (Mitsuboshiet al. 1992). The results for Fok-600, as well as the results of other morphological, ecological, and biochemical studies, also support the hypothesis that the species in one distinct population of S. malma, which became landlocked in Lake Shikaribetsu during the recent Ice ages, might have adapted intrinsically to the environment in the lake (Maekawa 1977, 1985). The population size of the ancestral population of S. malma miyabei may have been small enough to lose the SINE insertion in all individuals.
Further screening of FokI SINE loci and analysis of more samples from various regions should allow us to clarify the structures of populations and the evolutionary history of S. alpinus and S. malma.
The authors are grateful to T. Yoshikawa, T. Masuda, and M. Tsuji for their help in isolation of SINE clones. We thank N. Davis (Bering Sea), F. Kircheis (Maine), J. Vuorinen (Finland), and M. Saneyoshi (Japan) for providing those samples and Parks Canada for field assistance at Lake Hazen. North American field work was supported by Fisheries and Oceans Canada and Alaska Department of Fish and Game. This work was supported by a Grant-in-Aid for Specially Promoted Research from the Ministry of Education, Science, and Culture of Japan.
Communicating editor: N. Takahata
- Received March 9, 1998.
- Accepted April 30, 1998.
- Copyright © 1998 by the Genetics Society of America