Mitochondrial DNA sequence analysis in Sicily

AMERICAN JOURNAL OF HUMAN BIOLOGY 13:576–589 (2001)

Mitochondrial DNA Sequence Analysis in Sicily ` ,1 L. VACCA,1 M. MEMMI`,2 AND L. VARESI2 G. VONA,1* M.E. GHIANI,1 C.M. CALO Department of Experimental Biology, Section of Anthropological Sciences, University of Cagliari, Monserrato, Italy 2 Faculty of Sciences and Technologies, University of Corsica, Corte, France

1

ABSTRACT This study reports data on the sequences of the first hypervariable segment of a sample of the Sicilian population from Alia (Palermo, Italy). The results show the presence of 32 different haplotypes in the 49 individuals examined. The average number of pairwise nucleotide differences was 4.04, i.e., 1.17% per nucleotide. The distribution of the nucleotide differences matches the theoretical distribution and indicates only one major episode of expansion that occurred between 20,732 and 59,691 years ago, between the Middle Paleolithic and Upper Paleolithic. Compared with the other populations, parameters of the Sicilian sample lie in an intermediate position between the eastern and western Mediterranean populations. This is due to numerous contacts that Sicily has had with the Mediterranean area since prehistoric times. At the same time, the singularity of some of the haplotypes present in the sample studied indicates the persistence of some characteristics caused by genetic drift and isolation that the population has endured in the course of its history. Am. J. Hum. Biol. 13:576–589, 2001. © 2001 Wiley-Liss, Inc.

Mitochondrial DNA has been fully sequenced by Anderson et al. (1981). A haploid DNA of 16,569 base pairs contains information for the codification of some proteins and various types of ribosomal and transfer RNA specific of mitochondrial metabolism. The mtDNA molecule has a noncoding part, called a D-loop, which contains the origin of the replication and information that controls it. One of the fundamental characteristics of mtDNA is that it has a mutation rate roughly 10 times higher than the rate in nuclear DNA so that short-term evolving phenomena can be analyzed (Ferris et al., 1981; Cann et al., 1984; Vigilant et al., 1989; Ward et al., 1991). This mutation level is substantiated by an accumulation of mutations transmitted through the maternal line, since another basic property of mtDNA is that it is inherited only maternally. Mutations are most frequently shown by base substitutions, while deletions and insertions rarely appear. Transitions are always more numerous than transversions (Brown et al., 1979). Many analyses show a preference for segment I of the non-codifying area of mtDNA, which seems to have the highest evolution rate of the molecule. This is the area that was analyzed in a sample from Sicily (Italy) in the context of populations originating mainly in the Mediterranean region. HISTORICAL BACKGROUND The most ancient traces of human presence in Sicily, the largest island in the west-

© 2001 Wiley-Liss, Inc.

PROD #M20045R2

ern Mediterranean, date back to Paleolithic, Neolithic, and the Iron Age and show influences of both the western and eastern areas of the Mediterranean. This suggests that even in prehistoric times Sicily was an important crossroads of culture and trade between the eastern and western Mediterranean. During the great Indo-European migrations, the eastern part of Sicily was occupied by the Siculi (1,200 B.C.), who pushed the resident populations, the Sicans and the Elimi, toward the central and western parts of the island. The 8th century B.C. saw the beginning of one the most intense colonizations: colonization by the Greeks affected most parts of the coastal area, and the existing Phoenician presence was limited to the northwestern area. Over the next few centuries, Sicily became completely “Hellenized.” Another important time for Sicily was the period under Roman rule. Confiscated lands were given to veteran Roman soldiers, and great estates were established where large

Contract grant sponsor: Ministero dell’Universita` e Ricerca Scientifica e Tecnologica (MURST, Italy), Programmi Ricerca Scientifica di Rilevante Interesse Nazionale (ex-40%); Contract grant number: 9805557360-003. *Correspondence to: Prof. Giuseppe Vona, Department of Experimental Biology, Section of Anthropological Sciences, Cittadella Universitaria, S.S. 554, Km. 4,500, 09042 Monserrato (Ca), Italy. E-mail: [email protected] Received 26 May 2000; Revision received 3 January 2001; Accepted 7 January 2001

mtDNA IN SICILY

numbers of slaves were used to work the fields. Both the landowners and slaves came from all corners of the Roman domain. The Romans, contrary to the Greeks, pushed their occupation deeply inland, transforming the island also ecologically. Intense deforestation took place, and the monoculture of wheat started. Sicily in effect became the granary of Rome. After the passage of several other groups, e.g., the Vandals from North Africa and the Byzantines, Arab domination began in 827 A.D. with the landing in Gela of 10,000 Muslims from North Africa and Spain. This transformed Sicily into the center of a wide confederation of Muslim nations. In 11th century A.D., Sicily was occupied by the Normans and experienced an extraordinary immigration of numerous groups, particularly of Italians, French, and Longobards. Spanish domination began in 1282, and Sicily went through one of the most disastrous periods of its history. Spanish domination was aggravated by the spread of banditry and the presence of a large number of mercenaries. Return to monoculture linked to deforestation and a series of terrible events such as famine, earthquakes, and outbreaks of plague and cholera forced many of Sicilians to emigrate. Before becoming part of the realm of Italy in 1860 Sicily was the stage for many wars, which brought powerful armies of foreign soldiers to the island. For example, the restoration of the Borbonnes (1820) cost Sicily an invasion of 12,000 Austrian soldiers. Sicily’s history over the last century was characterized by a series of formidable and irresolute problems, leading to a high level of emigration which peaked in the years after the two World Wars; 1,500,000 Sicilians abandoned the island after the First World War. Many of these events had important demographic implications, influencing the biologic history of Sicily and shaping its genetic structure (Vona et al., 2000). The effects of one of the events, which dramatically affected Sicily, was the discovery in 1995 of about 300 skeletons of individuals who had died of cholera in 1837 A.D. in the small commune of Alia, in the eastern part of the province of Palermo. Situated 800 meters above sea level on the southwestern slopes of the Madonie mountain chain, about 80 km from Palermo (Fig. 1), Alia today has a

577

population of about 4,000 inhabitants. In 1837 Alia, along with many other Sicilian villages, was hit by an epidemic of cholera and the population was reduced by at least 10%. Given the high number of deaths, the bodies were mostly buried in a cave on the outskirts of the village rather than in the traditional cemetery. At the end of the epidemic, entrance to the cave was closed and only in 1995 did an excavation bring the remains to light. This study reports the results of the analysis of the sequences of the control segment I of mitochondrial DNA taken from a sample of the present-day population of from Alia, Sicily. MATERIALS AND METHODS Population The sample includes 49 unrelated, apparently healthy, adults of both sexes from Alia whose families were born and lived in the same village for at least three generations. With prior informed consent, a blood sample was collected from each individual by venipuncture at the Municipal Outpatients Clinic at Alia. The sample was put in a sterile test tube containing EDTA, stored at −20°C, and then transported to the Department of Experimental Biology at Cagliari University.

Amplification and sequencing of DNA The DNA was subsequently extracted from the whole blood using the QIAamp Blood Kit威 (QIAGEN). The DNA was amplified starting from 10 ␮l of the final lysatum. The hypervariable segment I of the D-loop was amplified by polymerase chain reaction. The primer sequences were as follows: L15996, 5⬘-CTC-CAC-CAT-TAG-CAC-CCAAAG-C-3⬘, and H16401, 5⬘-TGA-TTT-CACGGA-GGA-TGG-TG-3⬘, thus producing an amplified fragment of 425 bases. PCR conditions were as follows: 15 min at 95°C in 2.5 U of HotStar Taq Polymerase威 for a total reaction volume of 50 ␮l (10 mM each of dNTP, 25 mM di MgCl2, 15 ␮l of the primer couples, 10 ␮l of DNA of the sample to be amplified); 35 cycles at 94°C for 45 sec, 56°C for 1 min and 74°C for 1 min. Chain elongation was continued after the last cycle for 10 min at 72°C. The products of the amplification were visualized by electrophoresis on agarose gels containing 2% ethidium bromide. The products of the PCR were then purified in Cen-

578

G. VONA ET AL.

Fig. 1. Geographical localization of the studied populations. 1 ⳱ Sicily; 2 ⳱ Sardinia; 3 ⳱ Corsica; 4 ⳱ Tuscany; 5 ⳱ Basques; 6 ⳱ Spain; 7 ⳱ Berbers; 8 ⳱ Turkey; 9 ⳱ Middle East; 10 ⳱ Asia; 11 ⳱ Africa.

tricon-100威 tubes (Perkin-Elmer). Subsequently, 10 ␮l of amplified and purified mtDNA were sequenced by an automated sequencer ABI 373. Each sample was sequenced with the H and L primers used for the PCR. Some of the samples were sequenced twice, and some were sequenced with the primer H or L. The results were compared in order to avoid possible ambiguities in readings. The sequences thus obtained covered a total of 388 bases, from base 16,023 to base 16,410, and were compared with Anderson’s (1981) reference sequence through the CLUSTAL V program.

Statistical analysis Internal genetic diversity. The statistical analyses were carried out using the packages PHYLIP 3.51 C (Felsenstein, 1989) and ARLEQUIN 1.1 (Schneider et al., 1997). The phylogenetic analysis of the sequences was performed in accordance with Kimura’s (1980) two-parameter model. The tree rela-

tive to the 49 sequences was drawn up after 100 permutations, according to the neighbor-joining method (Saitou and Nei, 1987) and the difference between clusters was tested through the transformed pairwise Fst method (Reynolds et al., 1983; Slatkin, 1995). The expected number of frequencies and the effective size of the population were estimated under the infinite allele model by means of the number of individuals and sequences (Hartl and Clark, 1989). Variability within the sample examined was analyzed by Shannon’s diversity index in its standardized form H⬘ (Magurran, 1988). Dating of expansion. The pairwise nucleotide difference distribution was fitted to the Rogers and Harpending (1992) model, and the standard errors were calculated by 1,000 bootstrap iterations. The parameters ␶ and ␪ were then obtained following the two-parameter model of Harpending et al. (1993). Expansion time was estimated in accordance with the model of Roger and

mtDNA IN SICILY

Harpending (1992) through the parameter ␶ using mutation level calculations taken from reported works (Vigilant et al., 1991; Ward et al., 1991) and a generation period of 20 years. Comparison between populations. The sequences of the Sicilian sample were compared with those of the following populations: Basque (Bertranpetit et al., 1995), Berber (Corte-Real et al., 1996), Corsican (Varesi et al., 2000), Middle East (Di Rienzo and Wilson, 1991), Sardinian (Di Rienzo and Wilson, 1991), Spanish (Corte-Real et al., 1996), Turkish (Comas et al., 1996), Tuscan (Francalacci et al., 1996), African (Vigilant et al., 1989), and Asian (Stoneking, 1993). A neighbor-joining tree (Saitou and Nei, 1987) was constructed using the genetic distances computed through the method proposed by Rao (1982). RESULTS Sequence diversity Table 1 shows the variable sites in the control region of mtDNA compared with the reference sequence. The 49 individuals show 32 different sequences with 36 variable sites and 108 mutations. With the sole exception of transversion C→A, the substitutions are all transitions. The transition/ transversion ratio is, therefore, 36:1; much higher than that reported by Tamura and Nei (1993) for human populations. Most of the transitions concern the pyrimidines and are equally distributed between the two possible types. Only 6 individuals have a sequence identical to that of Anderson et al. (1981). The highest number of variable sites present in any one individual (SA23) is 8, while individual SA34 showed 7. The site with the highest number of substitutions is 16,126 with a T→C transition in 12 individuals (24.49%), followed by sites 16,311 and 16,189 with the same type of transition in 8 (16.33%) and 7 (14.29%) individuals, respectively. Under the assumption of the infinite allele model, using the number of individuals and the number of sequences, it was possible to calculate the expected number of sequences (Hartl and Clark, 1989). The value taken of ␪ in the sample of 49 individuals and 32 different sequences is 38.93, which corresponds to an estimate of 128.33 sequences per 1,000 individuals. This value of ␪ is much closer to that calculated in Corsi-

579

cans (38.52) (Varesi et al., 2000) and in Basques (27.58) (Bertranpetit et al., 1995) than that shown by Tuscans (99.38) (Francalacci et al., 1996).

Phylogenetic analysis Sequence variability as expressed through the index of diversity H⬘ (Table 2) is 0.961. If the Asian sample is excluded, this value is one of the highest reported in Table 2. Values vary between the 0.891 in Berbers and 0.991 in the Middle East. The average number of pairwise differences in nucleotides between all possible couples of individuals is 4.040 ± 0.845, i.e., 1.17% per nucleotide. The number of steps needed to construct the parsimony tree is 1.122. Phylogenetic analysis of the sequences carried out with the DNADIST program of PHYLIP 3.51C and using the method proposed by Kimura (1980) allowed the construction of a neighbor-joining tree from the distances (Saitou and Nei, 1987). The tree shown in Figure 2 shows how the sequences split into three well-defined clusters. The sequences that do not show variations from the Anderson et al. (1981) sequence are in cluster B together with other sequences, which have a lower number of transitions and with the only sequence carrying a transversion. From the Fst calculation, clusters A and C appear the most similar (Fst ⳱ 0.053), whereas clusters B and C result the most differentiated (Fst ⳱ 0.1256). Nevertheless, the association of the sequences in the different clusters seems quite random, and analysis of the variance shows that there is no significant heterogeneity between the three clusters (F ⳱ 0,089; df ⳱ 2; 46). In the Alia sample, the number of steps, which is influenced by the number of different sequences present in a population, has one of the lowest values among those reported for the populations in Table 2. This suggests that, from a phylogenetic point of view, the sequences obtained are closely correlated. Nucleotide difference distribution Analysis of the nucleotide difference distribution shows that the distribution curve (Fig. 3) is bell-shaped, similar to that described for other populations, with only one peak at 4.04 (variance ⳱ 4.943) (Table 2). The distribution matches the theoretical

a

T C C C -

C T T T T -

1 6 0 9 3

-

C T -

1 6 1 1 4

C C C C C C C -

T C C -

1 6 1 2 6

A -

G A -

1 6 1 2 9

-

A G -

1 6 1 6 3

T T -

C -

1 6 1 6 8

A T -

C -

1 6 1 6 9

C C -

T -

1 6 1 7 2

-

C T -

1 6 1 8 6

C C C C -

T C -

1 6 1 8 9

T T

C -

1 6 1 9 2

T -

C -

1 6 1 9 3

-

T C -

1 6 2 0 9

T T T T T -

C -

1 6 2 2 3

-

T C -

1 6 2 2 4

G -

A -

1 6 2 3 5

-

C T -

1 6 2 5 6

T

C -

1 6 2 6 1

T T -

C -

1 6 2 6 4

-

A G -

1 6 2 6 5

T -

C T -

1 6 2 7 0

-

G A -

1 6 2 7 4

T T T -

C T -

1 6 2 7 8

G

A -

1 6 2 8 9

T T T T -

C -

1 6 2 9 2

Sequences are given in comparison to the references sequence of Anderson (1981). Only nucleotides differeing from the reference are shown.

AND SA15, SA21, SA64 SA32 SA35, SA19 SA37 SA38 SA39 SA40, SA26 SA41 SA42, SA9 SA43, SA52, SA11, SA13, SA18, SA27 SA45, SA46, SA49 SA51, SA14 SA53 SA57 SA58 SA62 SA67, SA1, SA7 SA25, SA6 SA44 SA10 SA22 SA24 SA28 SA2 SA33, SA3 SA34 SA36 SA4 SA5 SA8 SA23 SA12

SICILY N ⳱ 49

1 6 0 6 9

TABLE 1. Variable sites of the control region found in Sicilian individualsa

T T -

C T -

1 6 2 9 4

T T -

C -

1 6 2 9 5

T T -

C -

1 6 2 9 6

C -

T -

1 6 2 9 8

C C -

T C -

1 6 3 0 4

C C C C -

T C -

1 6 3 1 1

A A A A -

G -

1 6 3 1 9

T -

C -

1 6 3 2 0

-

C T -

1 6 3 5 5

C C C C -

1 6 3 6 2 T -

mtDNA IN SICILY

TABLE 2. Sequences divergence in Sicilian and in the comparison populationsa Sicily Sardinia Corsica Tuscany Basques Spain Berbers Turkey Middle East Asia Africa

N

k

a

b

d

H⬘

49 69 47 52 45 58 86 45 42 23 22

32 46 31 40 27 52 30 38 38 23 12

37 52 40 54 32 62 37 56 59 59 34

4.040 ± 0.845 4.221 ± 0.558 3.712 ± 0.822 5.032 ± 0.543 3.152 ± 0.721 5.083 ± 0.994 4.782 ± 0.872 5.386 ± 0.728 7.083 ± 1.228 8.283 ± 1.011 3.194 ± 0.898

1.122 1.826 1.774 2.025 1.481 1.846 1.700 2.001 2.211 3.261 4.500

0.961 0.930 0.903 0.899 0.912 0.980 0.891 0.987 0.991 1.000 0.946

a N ⳱ number of individuals; k ⳱ number of different sequences; a ⳱ number of variable nucleotides; b ⳱ mean nucleotide pairwise differences; d ⳱ mean number of steps per sequence in a maximum parsimony tree; H⬘ ⳱ index of Shannon.

distribution based on the model of Rogers and Harpending (1992) (␹2 ⳱ 18.421; df ⳱ 11; P ⳱ 0.072). This suggests that only one major episode of population expansion in the past which would have increased an initial population of roughly 300–900 to something between 12,000 and 35,000 individuals (Table 3).

Dating expansion The estimation of dating expansion of the population was calculated by the method of Rogers and Harpending (1992) and Harpending et al. (1993). From the values obtained for the population of Alia for the parameters ␶ ⳱ 3.089, ␽0 ⳱ 0.951, ␽1 ⳱ 36.401, and the different mutation rates reported in literature, expansion would have occurred between 20,732 and 59,691 years ago with an intermediate value of 39,679 years (Table 3). Comparison with other populations All of the variable sites found in the Alia population have already been described in other populations. Therefore, the Sicilian population has no specific substitutions of its own. The substitution frequency in site 16,126 obtained for Sicily lies midway between the eastern and western comparison populations (Table 4). On the contrary, variability of site 16,311 does not appear to be distributed according to a gradient as indicated by Comas et al. (1996). The lowest frequency is found among the Basques (11.11%) and the highest among the Tuscans (21.15%). For site 16,311 Sicily has a frequency similar to the Berbers, Sardinians, and Turks. Another two mutated sites, 16,069 (12.24%)

581

and 16,223 (10.20%), both with a C→T transition, have a frequency in Sicily midway between the east and west. The two sites also appear to vary with a decreasing gradient from east to west. For 16,069 the Sicilian frequency is closer to that of the Tuscans and Turks than to those of the other western populations, where the highest frequency (6.90%) is found in the Spanish. The reverse is true for the substitution frequency of site 16,223, which, in Sicily, is of the western type. Substitution at site 16,261, which has the highest frequency in the Middle East (19.51%) and is absent in Corsica and Sardinia, barely reaches 2.04% in Sicily. The comparison of pairwise differences in Alia and other populations (Fig. 4) shows that the peak in Alia falls to the left side, having one of the lowest average values of pairwise differences among those shown in Table 2. This position indicates that Alia’s expansion time is one of the most recent compared to the comparison populations. The Sicilian sample was compared with the other populations by the genetic distances of Rao (1982). The genetic tree in Figure 5, drawn from the matrix of distances in Table 5, is of the neighbor-joining type obtained with 100 iterations. The two samples from Asia and Africa, which lie in two separate branches, are clearly differentiated from the Euro-Mediterranean populations, which are all found in another branch. This stretches from the populations of the Middle East on one side, near the fork that separates Asia and Africa, to the Basques and Corsicans on the opposite side. The sample from Sicily lies between Tuscany and Sardinia but closer to the former. Table 6 shows the haplotype frequencies that emerge from the sequencing and that Sicily has in common with the other comparison populations. While there are no common haplotypes between Sicily and the African sample, there is one common sequence with Asia (SA32). The most frequent haplotype in Sicily is sequence SA43, which is identical to the reference sequence of Anderson et al. (1981), with an intermediate frequency to those of the comparison populations. The lowest values are shown by eastern populations and the highest by the Corsicans and Basques. This haplotype was not seen in the Middle East sample. Other than the Sicilian sample, haplotypes of the sequences SA33 and SA25 were

582

G. VONA ET AL.

Fig. 2. Neighbor-joining tree of the sequences in the Sicilian samples.

only found in the Basques, while haplotypes SA36 and SA37 are also present in the Spanish. The haplotype represented by sequence SA67, absent in the Middle East and Tuscan samples, has its highest frequency in the Berbers, with the Sicilian frequency located between those of the Berbers and the western populations. For sequence SA40, which was not found in Tuscany, Sardinia, Corsica, or the Basques, Sicily has a frequency very similar to that of the eastern countries and the Berbers. Sequences SA53 and SA58 seem to be exclusive to the Sardinians and Sicilians, while haplotype SA2 is shared only with the Tuscans. The haplotypes that characterize the remaining (44.9%) sequences of Alia are not found in the comparison populations.

DISCUSSION The vast literature concerning mtDNA highlights its importance in the study of evolutionary processes in humans. Sequence variation analysis of a control region seems to be useful and highly effective in the study of a population’s history, not only on a global level but also regionally. Many genetic studies have been done on the Sicilian population; they have concerned mainly proteins, and only recently has the molecular point of view been discussed (Vona et al., 2000). The results suggest several genetic particularities of the Sicilian population determined by complex cultural and genetic relations that have marked the island’s

mtDNA IN SICILY

583

Fig. 3. Pairwise differences distribution of Sicilians and other comparison populations.

evolutionary history. This study shows the sequences of the first hypervariable segment of mtDNA in a sample of the Sicilian population from Alia in the province of Palermo (Italy). The results for Alia reveal several fea-

tures of the region of the first hypervariable segment of mtDNA, some of which have been reported in other studies. A prevalence of transitions over transversions is noted but in a much higher proportion than found in other populations. There is also a marked

584

G. VONA ET AL.

TABLE 3. Demographic parameter estimates for the Sicilian population according to the model of Roger and Harpending (1992) for different proposed mutation ratesa

␮

N0

N1

Times of divergence (years)

1.490 × 10−3 (Ward et al., 1991) 7.785 × 10−4 (Vigilant et al., 1991) 5.175 × 10−4 (Vigilant et al., 1991)

319

12,215

20,732

610

23,380

39,679

919

35,170

59,691

a N0 and N1 are the effective population sizes before and after the expansion episode.

TABLE 4. Most frequent substitution in the Sicilian sample in comparison with other populations Site position 16,069 16,126 16,223 16,261 16,311 Sicily Sardinia Corsica Tuscany Basques Spain Berbers Turkey Middle East Asia Africa

12.24 5.80 2.13 13.46 4.44 6.90 3.50 15.56 19.51 0.00 0.00

24.49 17.39 8.51 23.08 6.67 18.97 8.20 26.67 46.34 9.52 0.00

10.20 7.25 8.51 13.46 4.44 10.34 17.60 28.89 26.83 47.62 86.67

2.04 0.00 0.00 5.77 2.22 3.45 0.00 8.89 19.51 9.52 0.00

16.33 17.39 19.15 21.15 11.11 12.07 16.50 17.78 14.63 19.05 100.00

predominance of pyrimidine transitions over purine. The analysis shows that Sicilians, as far as the sequencing of mtDNA is concerned, have some characteristics that can be considered midway among those found in comparison populations. This is particularly evident in the cline of the frequencies of some substitutions, on the average, and in the distribution of the nucleotide pairwise differences. The position taken in the genetic tree of the populations also leads to the same conclusion. The tree shows approximately an east to west gradient with the Middle East at one extreme and the Basques and Corsicans at the other. The mutation at site 16,126, which is most frequent in the Sicilian sample, shows a decreasing east–west variation of frequencies with extreme values in the Middle East (46.3%) and in the Basques (6.7%). The frequencies of the substitutions relative to sites 16,069 and 16,223 also appear to have an east→west gradient, the Sicilians having an intermediate frequency between the two extremes. The distribution of the nucleotide pairwise differences reflects the history of the

ancient population, especially the expansions and the periods during which they occurred. Rogers and Harpending (1992) demonstrated that the rapid growth of a population is signaled by a smooth mismatch distribution that has only one peak. The mismatch distribution in Alia sample suggests that the main episode of expansion took place between 20,732 and 59,691 years ago, with an intermediate indication of 39,679 years ago. The comparison populations also show bell-shaped distributions, all denoting a series of expansions that took place later. Expansions occurred first in the east and then in the west. This is consistent with an east to west migration. The population with the oldest expansion would be the Middle East, roughly between 47,000 and 135,000 years ago. On the contrary, the most recent expansion would have taken place among the Basques, about 14,000–41,000 years ago. The observations noted from the various analyses seem to be consistent with an east–west migration (Comas et al., 1996; Francalacci et al., 1996). Barbujani et al. (1995), in a study of mtDNA in Italy, suggested that the time of expansion of the Italian populations, including Sicily, was somewhere between 8,200 and 20,525 years ago, and the size of the population was estimated between 1,160 and 2,326 females. The authors also proposed two possible expansions: one after the maximum of the last glaciation in the Upper Paleolithic and the other at the beginning of the Neolithic. The data in the present study only partly agree with Barbujani et al. (1995), both in terms of the time of expansion and the increase of population numbers. As indicated by the calculated data and in Figure 4, Sicily seems to be among the populations with a more recent expansion, but this unique event falls during the Middle and Upper Paleolithic, and its temporal position seems to agree with archaeological dates (Chilardi et al., 1996). Further, the estimated increase in population is much more substantial than that indicated by Barbujani et al. (1995). The first presence of modern populations is placed in the Near East from where they expanded toward Europe (Stringer, 1989; Mellars, 1993) replacing the Neanderthals. Other expansions successively occurred, such as in the Neolithic, which stretched from western Asia toward Europe and included Sicily. Other invasions also occurred

mtDNA IN SICILY

Fig. 4.

585

Neighbor-joining tree relating eleven populations, according to the distance matrix shown in Table 5.

in Sicily, which, as already noted, became a crossroads of human, cultural, and commercial traffic. Thus, a question of interest is how Sicily’s numerous contacts with other

populations influenced the genetic structure with regard to mtDNA. There is no trace of all these events in the mismatch distribution. Most of the popula-

586

G. VONA ET AL.

TABLE 5. Pairwise differences among the Sicilians and the comparison populationsa Sicily Sardinia Corsica Tuscany Basques Spain Berbers Turkey Middle East Asia Africa

Sicily Sardinia Corsica Tuscany Basques Spain Berbers Turkey M. East

Asia

Africa

4.04 0.009 0.104 0.000 0.087 0.000 0.681 0.018 0.238 0.358 7.562

6.52 6.69 6.48 6.98 6.21 6.89 7.21 6.99 7.91 8.28 4.695

11.18 10.68 11.34 10.75 11.72 10.66 10.41 10.53 10.26 10.43 3.19

4.14 4.22 0.059 0.003 0.083 0.038 0.630 0.116 0.309 0.437 6.978

3.98 4.02 3.71 0.082 0.080 0.117 0.916 0.142 0.549 0.483 7.890

5.54 4.63 4.45 5.03 0.115 0.033 0.644 0.003 0.158 0.326 6.635

3.68 3.77 3.51 4.21 3.15 0.076 0.836 0.272 0.576 0.493 8.554

4.56 4.69 4.51 5.09 4.19 5.08 0.618 0.047 0.232 0.211 6.528

5.09 5.13 5.16 5.55 4.80 5.55 4.78 0.719 0.683 0.678 6.434

4.73 4.92 4.69 5.21 4.54 5.28 5.80 5.38 0.097 0.163 6.249

5.80 5.96 5.94 6.21 5.69 6.31 6.61 6.33 7.08 0.230 5.124

a Below the diagonal: intermatches-based genetic distances, D ⳱ Dij − (Di + Dj/2) (RAO). Above the diagonal: intermatches between pairs of populations. Diagonal: mismatches within populations (i.e., parwise differences).

tions considered show a bell-shaped distribution of the pairwise differences. Some of the populations, e.g., Sardinia and Corsica, were to some degree invaded by the same populations that invaded Sicily, but in these cases, too, no effects attributable to these invasions are evident in any of the curves. Nevertheless, it should be remembered that the genetic diversity found at the mtDNA level in contemporary human populations is due to the differences accumulated both before and after their expansion (Relethford, 1988). When the frequencies of substitutions found in the sequences analysis and the haplotypes present in the Sicilian population are examined, it is evident that the Sicilian gene pool has been influenced by other Mediterranean basin populations. A gene contribution from the Near East appears unquestionable. Contributions to the Sicilian gene pool also appear to come from other regions of the Mediterranean. In fact, some haplotypes that the Sicilian population has in common with the other comparison populations can be seen as traces left by populations passing through the island. In most cases these haplotypes seem to date to the Paleolithic and Neolithic. In view of Sicily’s prehistoric and historic past, it is difficult to say whether these existed prior to the population expansion, if they were from the Paleolithic and Neolithic migrations, or if they reached the island in more recent times through groups that had retained them in their gene pool. Further, the haplotype frequencies that the Sicilians have in common with each individual population are low and lead us to the suggestion that the external influences of each population are of little importance.

Analysis of the Sicilian mtDNA haplotypes in relation to the recent work of Richards et al. (2000) on founder lineages of mtDNA in Europe suggests that about 70% of the haplotypes from the sample of present-day Alia could date back to the Paleolithic, whereas about 10% could have originated during the Neolithic. These values are very close to those that Richards et al. (2000) have indicated for the central Mediterranean: around 80% for the Paleolithic component and 10% for the Neolithic. On the basis of these observations, it seems likely, therefore, that the various immigrations which took place in Sicily did have as much impact on the mtDNA genetic structure, as that which resulted after the population expansion about 26,000 years ago. About 45% of the sequences found in the Alia sample are of haplotypes that are not present in the comparison populations. This peculiarity could be linked to the partial isolation that Alia experienced in some periods and to genetic drift. For example, for the first part of the 19th century A.D., the population of Alia was characterized by a progressive process of isolation, interrupted only after the unification of Italy (1860). In 1901 the number of inhabitants from Alia who had emigrated to the United States amounted to more than 3,000. Analysis of surnames reveals, particularly in the middle part of the 19th century, an extremely low number of different surnames in Alia: roughly 13 out of 100 (Bigazzi, 2000). Thus, between 1830 and 1844, the period of the cholera epidemic, of the 219 surnames of the previous period, Alia retained only 127 different surnames. Further, Alia, situated on the edge of the main

SA2

SA58

SA53

SA4

SA32

SA37

SA36

SA40

SA25

SA15

SA33

SA44

SA67

Haplotype SA43

SICILY n ⳱ 49 6 0.122 3 0.061 1 0.020 2 0.041 3 0.061 2 0.041 2 0.041 1 0.020 1 0.020 1 0.020 1 0.020 1 0.020 1 0.020 1 0.020

CORSICA n ⳱ 47 11 0.234 3 0.064 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000

TUSCANY n ⳱ 52 9 0.173 0 0.000 0 0.000 0 0.000 1 0.019 0 0.000 0 0.000 0 0.000 0 0.000 1 0.019 0 0.000 0 0.000 0 0.000 1 0.019

BASQUES n ⳱ 45 9 0.200 2 0.044 1 0.022 2 0.044 3 0.067 1 0.022 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000

SPAIN n ⳱ 58 5 0.086 2 0.034 0 0.000 0 0.000 0 0.000 0 0.000 1 0.017 1 0.017 1 0.017 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000

BERBERS n ⳱ 85 8 0.094 7 0.082 0 0.000 0 0.000 0 0.000 0 0.000 4 0.047 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000

TURKEY n ⳱ 45 2 0.044 1 0.022 0 0.000 0 0.000 2 0.044 0 0.000 2 0.044 0 0.000 0 0.000 0 0.000 1 0.022 0 0.000 0 0.000 0 0.000

M. EAST n ⳱ 41 0 0.000 0 0.000 2 0.049 0 0.000 0 0.000 0 0.000 2 0.049 0 0.000 0 0.000 0 0.000 1 0.024 0 0.000 0 0.000 0 0.000

TABLE 6. Absolute and relative frequencies of the haplotypes common to Sicily and other populations SARDINIA n ⳱ 69 15 0.217 3 0.043 0 0.000 0 0.000 1 0.014 0 0.000 0 0.000 0 0.000 0 0.000 1 0.014 0 0.000 3 0.043 1 0.014 0 0.000

ASIA n ⳱ 21 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 1 0.048 0 0.000 0 0.000 0 0.000 0 0.000

AFRICA n ⳱ 15 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000 0 0.000

588

G. VONA ET AL.

trade routes and never having much to offer economically, has always been in a marginal and isolated position compared to flows within the island. The middle position occupied by the population of Sicily confirms its ancient origin, on the one hand, and the conservation of its genetic identity, on the other hand. The genetic divergence that emerges in some of the parameters analyzed and in the peculiarity of some sequences may have been caused by the most recent expansions, which increased local mitochondrial variability, which, in turn, has since been maintained by isolation. During the period of cholera about 30% of the surnames disappeared from Alia, but in the following periods a similar number of new surnames, mainly from the surrounding areas, were introduced. The characteristics of mtDNA offered by the population of Alia indicate, on the one hand, the influences of contacts which the island has had with various populations of the Mediterranean area, and on the other hand, some particularities which seem to have been shaped by the effects of genetic drift and isolation. These particularities are also evident in studies of the genetic structure of the Sicilian population, in general, and of Alia (Vona et al., 2000). This analysis also confirms a position for Sicily midway between the European and Middle East populations (Rickards et al., 1992). In work carried out with restriction enzymes on mtDNA in a sample of Sicilians, Semino et al. (1989) indicated the presence (4.4%) of the African complex HpaI-3/ AvaII-3 (40% in Senegal and in the Bantu of South Africa). The authors hypothesized a migration of genes from Africa to Sicily, estimated at about 10%, which was introduced into the Sicilian gene pool by Black slaves brought by the Phoenicians and the Romans and/or by Arab migrations. Results at the mtDNA sequencing level, however, show no Black African influence in the Sicilian population. The haplotypes that emerge from the sequences, some of which common to both Sicilians and Berbers, suggest the possibility of a certain amount of contact with the populations of north Africa. Results of the analysis of mtDNA sequences indicate that it is difficult to identify in the characteristics of the present-day population of Sicily genetic influences that could be attributed to any single group that

invaded the island and settled there for shorter or longer periods. Further study of mtDNA by applying the restriction enzymes to the present and earlier (from the time of the 1837 cholera outbreak) populations could shed light on the interactions between the Sicilian population and other populations of the Mediterranean area which needs to be examined more thoroughly. ACKNOWLEDGMENTS We are grateful to the mayor of Alia, Dr. G. D’Andrea, and the whole population for their helpfulness. LITERATURE CITED Anderson S, Bankier AT, Barrel BG, De Bruijn MHL, Coulson AP, Drouin J, Eperon IC, Nierlich DP, Roe BA, Sanger F, Schreier PH, Smith AJH, Staden R, Young IG. 1981. Sequence and organisation of the human mitochondrial genome. Nature 290:457–465. Barbujani G, Bertorelle G, Capitani G, Scozzari R. 1995. Geographical structuring in the mtDNA of Italian. Proc Natl Acad Sci U S A 92:9171–9175. Bertanpetit J, Sala J, Calafell F, Underhill PA, Moral P, Comas D. 1995. Human mitochondrial DNA variation and the origin of Basques. Ann Hum Genet 59:63–81. Bigazzi R. 2000. Analisi microevolutive sulla struttura biologica e demografica della popolazione di Alia (PA) del XIX secolo. Tesi di Dottorato di Ricerca in Scienze Antropologiche, Universita` di Firenze. Brown WM, Geroge JRM, Wilson AC. 1979 Rapid evolution of animal mitochondrial DNA. Proc Natl Acad Sci U S A 76:1967–1971. Cann RL, Brown WM, Wilson RC. 1984. Polymorphic sites and the mechanism of evolution in human mitochondrial DNA. Genetics 106:479–499. Chilardi S, Frayer DW, Gioia P, Macchiarelli R, Mussi M. 1996. Fontana Nuova di Ragusa (Sicily, Italy): southernmost Aurignatian site in Europe. Antiquity 70:553–563. Comas D, Calafell F, Mateu E, Pe´rez-Lezaun A, Bertranpetit J. 1996. Geographic variation in human mitochondrial DNA control region sequence: the population history of Turkey and its relationship to the European populations. Mol Biol Evol 13:1067–1077. Corte Real HBSM, Macaulay VA, Richards MB, Hariti G, Issard MS, Cambon-Thomsen A, Papiha S. 1996. Genetic diversity in the Iberian Peninsula determined from mitochondrial sequence analysis. Ann Hum Genet 60:331–350. Culotta P. 1995. L’architettura della Gurfa. In: La Gurfa e il Mediterraneo. Alia: Ed. Comune di Alia. Cumbo G. 1995. Le grotte della Gurfa ed altre emergenze archeologiche nella Sicilia centrale, zona cuscinetto, tra le idrovie Platani e Torto. Alia: Ed. Comune di Alia. Di Rienzo A, Wilson AC. 1991. Branching Pattern in the Evolutionary Tree for Human Mitochondrial DNA. Proc Natl Acad Sci U S A 88:1597–1601. Felsenstein J. 1989. PHYLIP-phylogeny inference package (version 3.2). Cladistics 5:164–166. Ferris S, Brown WM, Davidson WS, Wilson AC. 1981. Extensive polymorphism in the mitochondrial DNA of apes. Proc Natl Acad Sci U S A 78:6319–6323. Francalacci P, Bertranpetit J, Calafell F, Underi-Ill PA. 1996. Sequence diversity of the control region of mitochondrial DNA in Tuscany and its implications for

mtDNA IN SICILY

the peopling of Europe. Am J Phys Anthropol 100: 443–460. Harpending HC, Sherry ST, Rogers AR, Stoneking M. 1993. The genetic structure of ancient human populations. Curr Anthropol 34:483–496. Hartl DL, Clark AG. 1989. Principles of population genetics. Sunderland, MA: Sinauer Associates. Kimura M. 1980. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J Mol Evol 16:111–120. Magurran AE. 1988. Ecological diversity and its measure. Princeton: Princeton University Press. Mellars P. 1993. Archeology and modern human origins in Europe. Proc Br Acad 82:1–35. Rao CR. 1982. Diversity and dissimilarity coefficients: a unified approach. Theor Pop Biol 21:24–43. Reynolds J, Weir BS, Cockerham CC. 1983. Estimation for the coancestry coefficient: basis for short-term genetic distances. Genetics 105:767–779. Relethford JH. 1988. Mitochondrial DNA and ancient population growth. Am J Phys Anthropol 105:1–7. Richards M, Macaulay V, Hickey E, Vega E, Sykes B, Guida V, Rengo C, Sellitto D, Cruciani F, Kivisild T, Villems R, Thomas M, Rychkov S, Rychkov O, Rychkov Y, Go¨lge M, Dimitrov D, Hill E, Bradley D, Romano V, Calı` F, Vona G, Demaine A, Papiha S, Triantaphyllidis C, Stefanescu G, Hatina J, Belledi M, Di Rienzo A, Novalletto A, Oppenheim A, Nørby S, AlZahari N, Satachiara Benerecetti S, Scozzari R, Torroni A, Bandelt H-J. 2000. Tracing European lineages in the Near Eastern mtDNA Pool. Am J Hum Genet 67:1251–1276. Rickards O, Biondi G, De Stefano GF, Vecchi F, Walter H. 1992. Genetic structure of the population of Sicily. Am J Phys Anthropol 87:395–406. Rogers AR, Harpending H. 1992. Population growth makes waves in the distribution of pairwise genetic distances. Mol Biol Evol 9:552–569. Saitou N, Nei M. 1987. The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol Biol Evol 4:406–425. Schneider S, Kueffer JM, Roessli D, Excoffier L. 1997.

589

Arlequin 1.1. Software program for population genetic data analysis, User Manual Genetic and Biometry Lab., University of Geneva. Semino O, Torroni A, Scozzari R, Brega A, De Benedictis G, Santachiara Benerecetti AS. 1989. Mitochondrial DNA polymorphisms in Italy. III. Population data from Sicily: a possible quantitation of maternal African ancestry. Ann Hum Genet 53:193–202. Slatkin M. 1995. A measure of population subdivision based on microsatellite allele frequencies. Genetics 139:457–462. Stoneking M. 1993. DNA and recent human evolution. Evol Anthropol. 2:60-73. Stringer CB. 1989. The origin of the early modern humans: a comparison of the European and nonEuropean evidence. In: Mellars P, Stringer C, editors. The human revolution: behavioural and biological perspectives on the origin of modern humans. Princeton: Princeton University Press. p 232–244. Tamura A, Nei M. 1993. Estimation of the number of nucleotide substitutions in the control region of mtDNA im humans and chimpanzees. Mol Biol Evol 10:512–526. Vigilant L, Pennington R, Harpending H, Kocher TD, Wilson AC. 1989. Mitochondrial DNA sequences in single hairs from Southern African Population. Evolution 86:9350–9354. Vigilant L, Stoneking M, Harpending H, Hawkes K, Wilson AC. 1991. African populations and the evolution of mitochondrial DNA. Science 253:1503–1507. Varesi L, Memmı` M, Cristofari MC, Mameli GE, Calo` CM, Vona G. 2000. Mitochondrial control-region sequence variation in the Corsican population, France. Am J Hum Biol 12:339–351. Vona G, Chiarelli B, Ghiani ME, Sineo L. 2000. Genetic structure of Sicily: a review. In: Susanne C, Bodzsar EB, editors. Human population genetics in Europe, Vol 1. Budapest: Eotvos University Press. p 63–78. Ward RH, Frazier BL, Dew-Jager K, Pa¨a¨bo S. 1991. Extensive mitochondrial diversity within a single Amerindian tribe. Proc Natl Acad Sci U S A 88:8720– 8724.