Introduction

Familial Mediterranean fever (FMF, MIM249100) is an autosomal recessive disorder characterized by recurrent attacks of fever and inflammation in the peritoneum, synovium, or pleura, accompanied by pain. Destructive oligoarthritis and potentially life-threatening secondary amyloidosis are the major long-term complications associated with the disease.1, 2, 3, 4, 5, 6, 7, 8

FMF is encountered more frequently in the people from the Mediterranean region (non-Ashkenazi Jews, Armenians, Turks and Arabs), which are considered as four classically affected populations. The illness is less common in other populations of Mediterranean ancestry, with the carrier rate and the severity of the manifestation of the disease varying considerably both among and within different ethnic groups.

The discovery in 1997 of FMF causing gene9 (MEditerranean FeVer – MEFV) has created possibilities to study the distribution of various mutations in geographically and ethnically different populations.1, 3, 4, 6, 7, 8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 The gene locates on chromosome 16p13.3 and includes 10 exons, and it encodes 781-amino-acid protein, named pyrin or marenostrin. The protein is likely to assist normally in keeping inflammation under control by deactivating the immune response; without this control, an inappropriate full-blown inflammatory reaction occurs.9, 23, 24, 25

To date, 142 mutations have been identified in the MEFV gene (http://fmf.igh.cnrs.fr/infevers/), most of which are substitutions (78 of them are missense, one – nonsense, 39 – silent mutations, 17 are located in introns, two – in UTS), one is duplication, two are insertions and two are deletions. Of these mutations, five account for more than 70% of FMF cases – V726A, M694V, M694I, M680I and E148Q7, 26, 27 and have different frequencies in classically affected populations. Forty-eight of the MEFV mutations so far identified are found in exon 10.

A wide range of studies have shown different rates of clinical manifestation of the disorder, which are caused by different MEFV mutations and compound heterozygotes.2, 5, 14, 17, 28, 29, 30, 31, 32, 33, 34, 35 The association is not strong, and this indicates the presence of different modifying factors, which alter the pattern and the rate of clinical manifestation of the disorder. Numerous observations3, 7, 11, 28, 36, 37, 38, 39, 40, 41, 42, 43, 44 demonstrate that FMF phenotype is mainly controlled by a number of factors: the MEFV itself, other genes, the patient's sex and so far undetermined population-specific factors. This supports the view42 that genetic typing alone cannot be accepted as a final diagnosis owing to the absence of a strong correlation between the clinical picture of the disorder and the MEFV mutations. Moreover, some results indicate that FMF may be provoked in carriers of a single mutation by factors unrelated to the other MEFV allele.40

From population genetic perspective, the most intriguing modifiers accounting for the clinical manifestation of the various MEFV mutations are those that should be regarded as population-specific factors. These include genetic structure and demographic history of populations, as well as a complex combination of climatic and geographic features specific to the settlement regions of ethnic groups considered.

Populations differ from each other according to the rate of prevalence of FMF and unequal frequencies of the most common mutations.7, 42, 45, 46, 47 Differing rates of the manifestation of FMF was shown not only for ethnically different populations6, 36, 42 but also among regional groups of the same ethnicity.11, 14, 36 To what extent abovementioned differences are due to modifier genes or environmental and population-specific factors remains to be established. Currently, this area of research is developing very rapidly and new populations different from classically affected ones are being studied: Anglo-Saxons and Germans,31 British,48 Afghans,49 Indians and Chinese,12 Italians,11, 46, 50 Spanish,15 French, Greeks46, 51 and Cypriots.52

In this review, we consider some principal population genetic features of FMF relating to: (1) the identification of the area and the time (the genesis) of different MEFV mutations' origins, and the role of ancient migrations in FMF spatial spreading; (2) the influence of environmental factors on the rate of FMF clinical manifestation; (3) the impact of cultural traditions on FMF incidence and clinical manifestation; (4) the hypothesis of selective advantage in MEFV mutations' carriers.

The origins of MEFV mutations

The study of the genesis of different MEFV mutations will contribute greatly to our understanding of the role of ancient contacts between the populations of the Mediterranean and Near East regions in the spatial spread of FMF. To date, little research has been carried out in this area, and their results show only indirect evidences on the time and the place of MEFV mutations' origins.11, 23, 53, 54, 55, 56, 57

Two MEFV mutations, M694V and V726A, are thought likely to date back at least to biblical times: they were seen in association with specific microsatellite haplotypes in populations separated for many centuries (eg North African and Iraqi Jews, Ashkenazi Jews and Arab Druze).9, 11, 58 Moreover, four different haplotypes bearing the M694V mutation converged on one single-nucleotide polymorphism (SNP) haplotype within MEFV. In the case of the North African Jewish population, more than 80% of FMF carrier chromosomes bear the M694V mutation in association with one microsatellite haplotype, which strongly support genetic drift as an explanation for the high frequency of the mutation in this ethnic group.9, 28 The effect does not exist in the Arab population of Jordan, which was successively governed by different conquerors.59 The mutation M694V is usually observed in association with specific microsatellite haplotype in the Iraqi Jews, who were virtually isolated from the majority of the Jewish population since the Babylonian captivity (2500 years ago) until the reestablishment of the state of Israel. The presence of the V726A mutation and its corresponding microsatellite haplotype in the Armenian, the Ashkenazi Jewish and the Druze FMF patients indicate the age of this mutation to be about 2000 years.9

It is suggested that the M694V mutation in the Arabs of North Africa might have been introduced by the Jewish population that emigrated there from the Middle East after the destruction of Solomon's Temple and also from Spain since the end of 15th century CE. During the second wave of migration, which continued more than two centuries, the Jewish population settled mainly in Morocco.21

The M694I mutation also might be considered as ancient, as it is not a recurrent mutation and is present in the Berbers, the indigenous population of the Maghreb. In other words, this mutation was present in North Africa before Arabization and Islamicization started in the 7th century CE.21 Consequently, in this particular case, historical records are able to explain the presence of dissimilar patterns of the M694V and M694I mutations' distribution in the Arabs of North Africa.

Available SNP haplotype data support the view that chromosomes with E148Q mutation from different ethnic groups probably share a common progenitor, thus implicating a founder effect.11 Bernot et al49 identified several microsatellite haplotypes linked with the E148Q mutation. A novel founder haplotype was observed only among North African Jewish carriers that allowed the E148Q to be considered as a recurrent mutation with a relatively ancient origin – most likely more than 1500–2000 years ago. Interestingly, the E148Q mutation frequency is 15% in Chinese and 21% in Indians,12 possibly indicating that the mutation arose far from the Mediterranean basin. Otherwise, these data may support the hypothesis of the recurrent nature of E148Q mutation.

Rare mutations, such as P369S and K695R, could possibly be inherited from Ashkenazi Jews through migrations to Europe and intermarriages with the Sephardic population.11

The observed similarity between Jordanians and Turks seems to reflect the trace left by the latter after years of dominance in Jordan.59 Although the population of Jordan is almost entirely Arab, with minorities of Circassians and Armenians, each of whom account for less than 1% of the population, the different populations that came into this land left significant traces in their genetic legacy.

Some results indicate a possible founder effect for the spatial scattering of MEFV's principal mutations in North African Jews, and might suggest multiple sources and origins of the Arab and Iraqi Jewish populations. They derive from the observed high variability of mutations among Palestinian Arab and Middle Eastern Jewish FMF patients, as compared with North African Jews.60

There is little information about the genetic homogeneity/heterogeneity of FMF in the Turkish patients.45 Preliminary data suggest that the majority of the cases may be attributable to the same disease locus responsible for the disorder in other ethnic groups, although the origin of the mutations in the Turks is likely to be heterogenous.23, 53, 56

Summarizing, we can conclude that currently the most productive way to provide a temporal and spatial framework for the genesis of MEFV mutations is to analyze a set of microsatellites linked with different mutations in ethnically homogenous groups.

Influence of environmental factors

Different findings indicate that the environment may have notable impact on the rate and pattern of FMF clinical manifestation in patients of the same ethnicity. It is worthwhile stressing that all four classically affected populations have big Diaspora communities in different regions of the world, which allows studying the influence of various environmental factors on the clinical picture of the disorder.

It was found61 that the number of attacks per year in Armenians living in Armenia was significantly higher than in Armenians living in the USA (although both groups have the same genotype distribution). Additionally, the incidence of amyloidosis in the Armenian patients depends on the place of domicile and the samples studied: 0% incidence among those living in the United States,62 and an incidence between 2563 and 48.5%4 among those living in Armenia. In Israel, the number of attacks per year and other FMF characteristics were found to be similar in both North African and other Jewish populations, even though the mutation/genotype distribution was different.61

FMF inflammatory attacks can be triggered by stress and extreme physical exercise.64 In general, the effect of environment on the inflammatory attacks in FMF is not surprising and is also seen in other cyclic conditions, such as sickle cell anemia. However, in contrast to the latter, where only one mutation exists, in FMF, predisposition to the influence of environment is dependent on which mutations are present. This is confirmed by stronger predisposition in the M694V homozygotes.61

Impact of cultural traditions

Among different population-specific cultural practices, consanguinity is considered as one of the main features influencing the rate of any inherited disorder. Specifically, consanguinity increases the frequency of homozygotes in a population, which results in higher than expected FMF incidence.59 For example, the consanguinity rate in Jordan varies between 50 and 64%,65, 66 with the prevalence of the disease 1:2600.67, 68 Furthermore, 52% of the patients had familial FMF history and consanguineous marriages were present in 60% of the families.59 It means that the cultural practice of consanguinity should be taken into account when conducting population genetic study of FMF in communities where this tradition is relatively common. In those populations, the frequency of homozygotes is higher and subsequently the proportion of compound heterozygotes is lower. This type of apportionment may notably change the pattern of FMF clinical manifestation in the populations with expressed rate of consanguinity.

It is reasonable to suggest that other ethnically linked cultural traditions and practices, for example, peculiarities of diet, life style and so forth, may also have impact on the rate of FMF clinical manifestation and the age of onset of the disorder.

Possible selective advantage of heterozygotes

Selective advantage in MEFV heterozygotes remains an attractive assumption when attempting to explain the observed level of carriers in the populations of the Mediterranean area. The strikingly high mutation frequencies raise speculation that heterozygotes may have some selective advantage, perhaps manifested by increased resistance to a yet unidentified infectious agent.9 As the MEFV mutations' frequency is much lower in the East European Ashkenazi Jewish population, it seems plausible that carriers of FMF gene mutations might have a selective advantage, possibly because of heightened resistance to a pathogen endemic to the eastern Mediterranean area. Recent findings suggest that the putative survival advantage of MEFV mutations' carriers may derive from an increased innate immune response to a broad class of bacterial pathogens, rather than to a single agent, as has often been assumed.69

Accepting the hypothesis, we would expect low rates of morbidity by infectious diseases in heterozygotes and high rates of mortality in embryos bearing two or more severe MEFV mutations. Available data do not support this hypothesis unambiguously: Brenner-Ullman et al70 proposed a reduced rate of asthma in FMF heterozygotes, but the differences between groups was of borderline statistical significance; in a special study,47 there was no difference in morbidity between Ashkenazi carriers and non-carriers of mutations in MEFV.

The results of a recent study43 argue against the hypothesis of an increased embryonic death of zygotes with two severe MEFV mutations. Under such an assumption, the number of patients carrying these mutations would be much lower than would be expected from Hardy–Weinberg equilibrium, a situation reported in other conditions.71 It is also suggested that MEFV mutations in Middle Eastern populations may be a balanced polymorphism protective against brucellosis. This would be similar to the presence of high levels of sickle cell anemia in sub-Saharan Africa as a result of its protective effect against malaria. This hypothesis could be tested in a case–control study looking at the prevalence of MEFV mutations in contemporary Middle Eastern people with brucellosis, compared to healthy controls.72

Practically, as argued by Aksentijevich et al,11 the precise nature of FMF selective advantage may not be so easy to reveal, as antibiotic therapy and modern public health measures are likely to obscure the effects of infectious diseases, and even a small advantage compounded over many generations may give rise to high carrier frequency.

Conclusion

Population genetic study of FMF in different ethnic groups of Mediterranean ancestry brings to a range of fundamental outcomes, which go beyond the limits of clinical medicine and applied medical genetics. Basic findings of genetic studies of FMF in various populations contribute: (a) to the identification, in the short-term, of the place and the time of founder mutations' origins, (b) to the understanding of the role of ancient migrations and contacts between Mediterranean and Near East populations in the spatial spreading of the disorder, (c) to the clarification of the impact of environmental factors and cultural traditions on the rate of FMF incidence, (d) to the testing of the possible selective advantage in the carriers of MEFV mutations.

One might also expect to see some extent of geographic stratification of classically affected populations according to the distribution of MEFV mutations. As a result, this may contribute to the understanding of the roles of genetic and epigenetic factors influencing FMF in shaping the patterns of genetic structure of regional groups of the same population. For example, in our study,73, 74 a marked geographic structure of the Armenian population according to the Y chromosome markers was shown. Regional stratifications based on various genetic markers have been found for the other three populations too.