М. Golubovsky, L. Kaidanov INVESTIGATION OF GENETIC VARIABILITY IN NATURAL DROSOPHILA POPULATIONS*

The paper describes the origin and development of population genetics, its main disco­ veries achieved in experiments with Drosophila before 1948 breakage and summarized the results of two series of long-term studies on genetic variability in Drosophila car­ ried out by the authors. The conceptual contribution of two main centers or school leaded by N. Koltzov and S. Chetverikov in Moscow and Yu. Filipchenko in St.-Pelersburg is considered. The ideas and methodology of these schools were introduced into American and European genetics by Th. Dobzhansky and N. V. Timofeev-Ressovsky. A new approach to the description of the eukaryotic genome organizations is des­ cribed. The authors argued that main source of the origin of various hereditary changes in nature is the interaction of environment with the facultative genomic elements.

to Dobzhansky [20] Koltzov was «a man of multifarious interests and knowledge, of imposing presence, and with the eloquence of a spellbin ding orator.His public lectures were events memorable to his peers and to beginners alike».In 1921 he invited in his institute S. S. Chetverikov (1880Chetverikov ( -1959) ) and A. S. Serebrovsky (1892Serebrovsky ( -1949) ) to develop the field of genetical studies.Their disciples make up the famous Moscow school of evolutionary and population genetics [2,5,20,65,109].
N. I. Vavilov (1887-1943) and Yu. A. Filipchenko (1882-1930) were leaders of the Leningrad (St.Petersburg) shool of evolutionary and population genetics.In 1921 Vavilov headed the Institute of Applied Bo tany (later the Institute of Plant Industry).He succeeded in collecting a cohort of an outstanding scientists.Among them were famous cytogeneticists G. A. Levitsky and G. D. Karpechenko, both of which together with their leader Vavrlov, were killed during Stalin's reign of terror.Le vitsky studyed the structure of the chromosome from the point of view of systematics and evolution.He coined the term «karyotype».Curiously, namely Levitsky, originally the professor of botany in Kiev accidentally met there, the graduate student Dobzhansky (he rented the room at flat of Levitsky) and attracted him to genetics.
In 1913 Filipchenko began teaching the first course in genetics in Russia at Petersburg University.He then formed the Department of Ge netics and Experimental Zoology (1919) and later in 1921, the Labora tory of Genetics in the Russian Academy of Sciences.With the sudden death of Filipchenko in 1930, Vavilov became ,the director of this labora tory.In 1933 it was moved to Moscow and transformed into the Institute of Genetics.Vavilov headed the institute up to his arrest in 1940.
Filipchenko was a brilliant biologist combining both a profound know ledge of general biology and the history of evolutionary though with vast experimental skills.His interests in genetics range from plant genetics to eugenics.He wrote six textbooks [30,31] and many reviews which, according to Dobzhansky [20] were used in institutions of higher lear ning «until Lysenko's pogrom».Filipchenko had great impact on the evolutionary and genetical thought in Russia.He was the first person who clearly distinguished between micro and macroevolution.
The ideas and methodology which developed in these Russian shools were introduced into European genetics by Timofeev-Ressovsky.His dra matic biography and scientific life were recently reviewed [88J.Certain traditions of Russian shool were transferred to America by Dobzhansky.Both emigrants are renown in their fields and together helped found the new discipline of population genetics.
Major contributions before 1948.Before embarking on a detailed description of Russian studies on the field of population and evolutio-nary genetics of Drosophila let us try to discriminate between these two related fields.Evolutionary genetics embraces the analysis of the orga nisation of genetic material, its comparison in different species and me chanisms of change, phenomenology of the expression of genetic factors, genetical aspects of speciation and events of intra and interspecies hyb ridization.Population genetics includes first of all an analysis of genetic heterogeneity and the polymorphism of populations together with the me chanisms responsible for its maintenance, the study of the genetic con sequences of selection and other evolutionary forces, genogeography and the phenetics of natural populations.
Hugo de Fries could well be named the father of evolutionary gene> tics.He rediscovered Mendel's laws of inheritance of genetic variants.He established the principles of the spontaneous mutation process, He described polyploidy and chromosomal speciation in Oenothera and he found non-mendelian peculiarities in the progeny of interspecies hybrids.
An the same time S. S. Chetverikov contributed to the foundation store of both theoretical and experimental population genetics.In his classical paper [16,17] he published important conceptual discoveries.He introduced the concepts of mutation pressure and genotypic milieu.He described material, methodology and tasks of population genetics.He linked the concepts and data of mendelian genetics with the theory of natural selection and ontogenesis.He also made important theoretical predictions which were consequently confirmed by his students.From the point of theory of knowledge and cognitology it was a real conceptual discovery [90].
Using the inbreeding method, Chetverikov's students established the predicted existence of hidden recessive mutations in the progeny of wild flies of different Drosophila species [36,94,104].Following on, Dubinin and his collaborators used the method of homozygotisation of autosomes of D. melanogaster with the help of dominant balanced markers (a mo dification of Muller's classical approach).Dubinin and his group con ducted a systematic comparative analysis of the concentration of lethal mutations in geographically remote populations.During 11 years of stu dy they described the lethal and visible mutation frequency in 14 popu lations.The study produced the first reliable data set allowing a compa rison of the genetic load of drastic mutations from distant populations [23][24][25].
The results obtained by Dubinin's group were confirmed by other po pulation investigators from the Leningrad school [12,86,87].Similar data were also obtained both on D. melanogaster and other Drosophila species in United States in the well known studies of Dobzhansky, Gor don, Ives and Spencer.Gershenson succeeded in carrying out simulta neous «isogenisation» of all autosomes of D. melanogaster in a sample of flies caught from a wild population.This approach revealed that practi cally every fly from nature carried some mutations.This fact permitted a more accurate evaluation of genetic heterogeneity in natural populati ons and gave a better insight into the mechanism of interaction between mutations located in different chromosomes.
The description of minor or physiological mutations distributed in natural populations was made simultaneously by Dobzhansky and Queal [21] and Muretov [85].The talented young including the ideas of pene trance, expressivity and specificity.In addition, Rokitsky [92] introduced the concept of a «field of gene action» or a stereotyped pattern of gene expression through studying inherited Drosophila bristle variations.All of these components or modules of a gene manifestation may be under the influence of genotypical and external factors.
Timofeev-Ressovsky [105,106] also conducted a series of obvious and simple experiments showing that the combination of two or three drastic mutations may increase variability.The methodology of this in vestigation was used by Dobzhansky in his classical studies on the effect of lethal mutations isolated from nature and on the viability of their he terozygous carriers.Some of the major contributions made by Russian studies to the field of Drosophila population genetics are presented in Table 1.
The Russian biologist I. I. Schmalhausen classically associated with the field of evolutionary morphology developed the theory of sta bilising selection.He was deeply involved in the study of evolutionary and population genetics.He was also the supervisor to the Doctor of Science degree for both M. M. Kamshilov and Raissa Berg.At the end of 30s he invited to his Institute in Kiev a brilliant geneticist, S. Gershenson.
Under the influence of Schmalhausen, Kamshilov completed a series of sophisticated experiments on the evolution of norms of reaction.He showed that the morphological manifestation of a mutation and the mo de of its expression after developmental action of different external fac tors can be completely transformed during natural selection.For instan ce, the mutation eyeless in D. melanogaster originally having quite a variable expression may be readily transformed by selection into a developmentally stable form [69,70].
Kamshilov also modelled the physiological effects and genetical adap tation of populations to drastic external factors.Populations of flies we re selected for tolerance to severe cold temperature treatment.After a successful result, it appeared that the action of weak cold treatment du ring development became a stimulating factor [71]! Thus, the conditions of selection are indirectly imprinted into the genotypically based norms of reaction with selection changing the role of different genes during de- The high concentration of recessive lethals in wild popur lations and differencies between them in their lethal gene pool content Inversion polymorphism in wild populations, its adaptive value and temporary fluctuations A detailed description of minor viabilty mutations spre ading in natural populations The viability advantage of some natural lethals in heterozygotes The genotypical control of mutability; the discovery of highly mutable genes in natural populations Transformation during selection, phenotypically or physi ologically expressed norms of reaction A synchronous rise and fall of mutability level in geogra phically isolated populations Chetverikov [16] Serebrovsky [95,96] Romashov [93], Dubinin [22] Timofeev-Ressovsky [103,105,106] Serebrovsky [98] H. A. and N. W. Timo feev-Ressovsky [104], Romashov, Balkashina, [94], Gershenson, [36] Dubinin et al. [25], Dubinin [23] Dubinin and Tinyakov [26,27] Muretov [85] Mazing [8>1] Tinyakov [105], Gershenson [38], Berg et al. [12], Berg [6,7], Duseeva [28] Kamshilov [69,71] Berg et al. [12], Berg [6,7] velopment and their contribution to resulting phenotype.Unfortunately the original works of Kamshilov practically disappeared becoming the unknown and unquoted predecessor to some of Waddington's major ex periments and ideas.
One of the major problems studied in Filipchenko's Department was the genetics of speciation and interspecies hybridisation.The possibility to investigate this problem arose when Sturtevant discovered the species D. simulans which is the sibling to D. melanogaster.After reviewing the long history of such studies, Provine [91] concluded that, after Sturte vant, «primary among these were the studies of Julius Kerkis of the La boratory of Genetics, USSR Academy of Sciences in Leningrad».Kerkis (1907Kerkis ( -1977) ) was involved in genetics with Dobzharisky after their cu rious accidental meeting in a forest near Kiev in 1920.Kerkis was then a boy collecting butterflies and Dobzhansky was a student collecting laf dy-birds.Later they met again in Filipchenko's Deratment in Leningrad.Kerkis became a student of Dobzhansky's and friend until death.Kerkis conducted remarkable experiments on the influence of temperature on the hybrids [72] and described the character of conjugation of polytene chro mosomes hybrids [73].
Profound ideas in the fields of evolutionary and population genetics were developed by Serebrovsky (1892Serebrovsky ( -1948)).He proposed the concept gene pool andof the gene pool and genegeography [96,97].In 1929 he was the first, together with his student Dubinin, to deduce the complex linear structure of the scute-achaete locus.Modern molecular data con firmed their principal conclusion, mentioning «the pioneering work of the Russian genetic school in the 1930s» [15].Then in 1938 Serebrovsky, summarising data on the genetical study of scute-achate locus, came to the profound conclusion that new genes arise in the process of evolution via the process of duplication and divergence.This idea was also ahead of its time and that of Ohno's.
The dramatic fate of the other original idea of Serebrovsky's about usage of induced translocations for insect pest control is also now known [98].The paper rested in obscurity till 1968.In 1968 Serebrovsky's paper was resurrected by C. F. Curtis and then an English translation was dis seminated throughout western literature.In 1941 Serebrovsky prepared a more detailed version of his paper «Theoretical foundations of the trans location method of pest control» which was not published until 1971.The impact of Serebrovsky's ideas on a conceptual basis for genetical control was thoroughly analysed by Whitten [111].
One of the peculiar achievements of Chetverikov's school is the syn thesis of data from population genetics and developmental genetics [5].Timofeef-Ressovsky established the main principles of the morphological manifestation of mutant genes.
Thus Russian population genetics developed constantly in productive union with evolutionary morphology.It corresponds that there is current interest in the phenotypically or physiologically expressed traits of orga nisms.As noted sarcastically by Scharloo [95] «after the Dark Age of Elektrophoresis, evolutionary biologists are now experiencing the Rena issance of the Phenotype».
The direction of 'studies during the modern period.After Lysenko's pogrom in 1948 all genetical studies in the USSR ceased.They were re newed only in the 1960's in three main centres: We briefly summarise some of the main results obtained after the 1960's.
The dependence of inversion polymorphisms on ecological factors discovered initially in D. funebris populations by Dubinin and his asso ciates was confirmed by Borisov [14] following his studies of D. funebris populations in the Moscow region.The existence of urban and rural ra ces of Drosophila differing in karyotype structure was confirmed together with established evident spreading of the rural race over a 20 year peri od from 25 to 800 square kms.
A remarkable series of experiments on interspecific hybridisation among the D. vtrilis group was conducted by N. N. Sokolov and his fol lowers.Sokolov [100] developed an exact cytogenetical method of chro mosome substitution for the study of nucleocytoplasmic incompatibility for interspecies hybrids.He showed that in the hybrids of D. virilisY^D.lumtnei somatic heteroploidy regularly occurs due to the elimination of one chromosome in the fi^st of the mitotic cleavage division.Then Evgeniev showed that the elimination of this chromosome from one species is result of presence of A -T rich region and its retarded replication.
Mitrofanov [83,84] continued the phenogenetic tradition of the Mos cow school of evolutionary genetics.He found that penetrance of domi nant Puffed mutation, which is located on the 2nd chromosome of D. vi rilis, preatly varied in intraspecies hybrids, depending on the origin of the wild strain and the presence and concentration of modifier genes.In interspecies hvbrids the expression of Pf was shown to be influenced by cytoplasmic factors.
Evgeniev and Zelentzova [29] described a new family of mobile ele ments in D. vtrilis group called pDv.In D. virilis individual copies of pDv are scattered over approximately 200 sites on all chromosomes.In other related species D. texana and D. novamexicana, the number of pDv sites is markedly lower, where the pDv copies are completely absent in D. littoralis.Localisation of pDv coincides with the sites of concentrati on of satellite DNA.In interspecific hybrids of D. virilis\D.littoralis and D. virilisY^D.lummei transposition of pDv from D. virilis chromoso mes to foreign ones was demonstrated.These data can be used to explain the drastic reshuffling of stDNA in D. virilis and other related species.Movable elements may well be functioning as a shuttle system.
Evgenjev and Lozovskaya afterwards discovered the phenomenon of hybrid dysgenesis in D. virilis connected with the mobilisation of Pdv and the occurrence of unstable mutations.
Korochkin and his associates [771 has been conducting comprehensi ve comparative genetic and biochemical analysis of the expression of a clusters of esterase genes in different philades of Drosophila.Thev found that species-specific differences in the expression and developmental re gulation of esterase genes are connected with the action of regulatory mutations.They discovered and localised different types of trans-acting regulators or modifiers which are selected during evolution.These inclu de (a) regulators controlling isozyme activity and the amount of mRNA, (b) temporal regulators which control the exact time of tissue-specific expression and (c) «architectural» genes controlling the ratio of free and membrane bound fractions.These result may be presented as the mole cular incarnation of the above mentioned phenogenetic concepts of Timofeev-Ressovsky and Rokitsky.
An exclusively important and original set of data were obtained by a Russian Drosophilists on the problem of the origin of spontaneous mu tations.As earlv as 1939, Gershenson [37] discovered the mutagenic ac tion of foreign DNA in Drosophila.These promising studies were inter rupted firstly be war and then by the collapse of Russian genetics in 1948.Since the 1960s Gershenson, together with his follower Yu.N. Alexandrov, undertook a systematic study of the peculiar mutagenic action of this natural source of variation.It was established [39,40] that fo reign DNA extracts, including DNA and RNA viruses added to Droso phila food or injected in haemolymph, raised the mutation rate and pro duced an unstable state of some host genes.The action of toreign and viral DNA and RNA differed in their locus specificity.Autosomal lethal mutations and unstable visible ones occurred preferentially in a definite group of loci characteristic for each foreign agent tested.
Multilocus damages, which result in so called multi-lethal chromo somes, have been observed regularly.On the other hand mu'lti-lethal chro mosomes were isolated in distant natural populations [41].Is it possible that the damage caused by multi-lethal chromosomes from nature is in duced by viral foreign DNA?
The possibility of an experimental approach to solve this problem appeared after the creation of two types of lethal collections: (a) lethals induced by viruses and exogenous DNA by Yu.N. Alexandrov in the la boratory of Gershenson and (b) lethals isolated in different years from distant natural D. melanogaster populations.The authors decided to com pare the allelic sets of lethals from these two collections [3].In became apparent that multilethal chromosomes from natural populations showed allelism with lethals induced by different viral DNA.Together with data collected by Golubovsky and Plus [53] it follows that (a) viruses may be a specific source of spontaneous mutations and (b) during virus-in duced mutagenesis in nature similar multiple chromosome lesions can occur repeatedly and independently in isolated populations as a result of a single mutation event.
There are additional pieces of evidence towards this conclusion.Firstly, that foreign viral agents may activate different host mobile ele ments and' induce unstable mutations [33].Secondly, that multiple chro mosomal lesions occur during P and Hobo-mediated hybrid dysgenesis [13,68,78].
The main conclusion from Gershenson's discovery is important.Ger shenson constantly stressed that if we bear in mind how universally di stributed viruses are, how often human beings come into contact with them after various prophylactic virus inoculations and when virus pre parations are applied in the struggle against harmful insects, etc. then we have to recognise that the mutagenicity of viruses is a very serious problem [40].
Another important conclusion was made by R. B. Khesin (1922Khesin ( -1985) ) -the last outstanding disciple of Serebrovsky.He published [75] о comprehensive volume, «Genome Inconstancy» where he collected and thoroughly analysed, from an evolutionary point of view, all data sets from viruses to high eukaryotes concerning mobile elements and about mobility of other genomic elements (amplification, ordered rearrange ments during development etc.).He put forward the idea that the eukaryotic genome must be viewed, as a population of self reproducing DNA and RNA molecules.Moreover, having in mind the regular horizontal transfer of transposons and plasmids in microorganisms, the increasing number of similar facts in Drosophila species and the occurrence of the same retroviral sequences among remote animal species, Khesin conclu ded that it is possible to conceive of a gene pool encompassing all living organisms [75].Thus Serebrovsky's gene pool concept exhibits an un predictable and colourful array.The search for the source of genetic va riability in nature leads to the study of the interaction between different DNA and RNA carriers and logically to biocenotic genetics (the syno nym of biocenosis in western literature is ecosystem).
Biocenotic genetics has being distinct aspects.One of them was de veloped in the Department of Genetics of St. Petersburg University [79,80].In the ecological system Drosophila melanogaster -Saccharomy c e s cerevisiae Drosophila is a consuming species and the yeast is a produ cing species.Drosophila and yeast are linked through the food chain in natural biocenoses: Drosophila shows an obligate dependence on sterol ISSN 0233-76'57.БИОПОЛИМЕРЫ И КЛЕТКА.1994.Т. 10, № 5 biosynthesis of yeast.Sterols are required by insects as the buildings blocks of membranes and as precursors of ecdysone.The steroid hormo nes control larval moults and metamorphosis.Luchnikova and Inge-Vechtomov decided to study the effect of a genetic block in the synthetic path way of the producing organisms on the genetic variation of the consu ming one.Mutant yeasts deficient in sterol biosynthesis were used.The feeding of Drosophila larvae on the mutant yeast blocked larval develop ment.Starving adult flies of sterols shortens their lifespan and decrea ses fertility i. e. the rates of oviposition and subsequent hatching.In ad dition female become sterile.The genetic consequences of the mutational alteration in the yeast-Drosophila food chain are an increased rate of in duced mutations in flies (aneuploidy, dominant lethals) and a decreased frequency of crossingover.
Long-term population studies on Drosophila.Now we are starting on our own series of long-term experiments with Drosophila connected with the genetic analysis of natural populations and the analysis of the gene tic consequences of selection.As noted by Dobzhansky [19], there are only a handful of cases in which genetical changes in populations were actually observed.The need for such studies on^Drosophila after the dis covery of global changes in the genetic structure and distribution of mo vable elements and after evidence of their recent horizontal transfer, are now becoming even more necessary [76].Long-term systematic studies of natural and laboratory populations yielded some unpredictable and un clear results.
Survival strategies of inbred long-selected lines: genetic mechani s m s.The evaluation of the genetic consequences of selection is a very important task for population genetics and the theo ry of selective breeding.To approach this problem, a long term experi ment was conducted in 1965.Flies from a D. metanogaster natural po pulation collected from Yessentuki (Crimea) were selected for low ma ting activity in males by close inbreeding [63].As a result, a low acti vity strain (LA) with a complex of morphological, physiological, beha vioural and genetic alterations, which greatly decreased its fitness, was established [66].In the LA strain the whole spectrum of changes pro bably resulted from a correlated response to negative selection for an adaptively important trait.
Up to 1993 the LA stock had been selected for 650 generations.The aim of this long term experiment was to study the nature of p;enetical variability in response to different generations of inbreeding and to es tablish the genetic mechanisms that promote survival in inbred lines un der strong selection.Three main mechanisms were discovered: (1) a highly increased level of mutability in inbred stocks and the accumulation of supervital mutations capable of suppressing lethals and semilethals; the maintenance of a high potential for genetic variability; (2) the non-accidental ordered transpositions of some copialikc mo bile elements accompanied with increased viability; (3) mediated by Яобо-elements, a considerable increase in genome instability of lines selected in the minus-direction.
The manifestation of the first mechanism was demonstrated by suc cessful reverse selection of flies from the LA stock in a plus direction.The possibility of successful selection in the plus and minus directions after many dozens or even hundreds of inbred generations demonstrates the high potential for hereditary variation in the strains studied Using the standard CyL/Pm procedure, we found in LA second chromosomes the whole spectrum of mutations from lethal to supervital one.The pro portion of various classes of mutations was repeatedly estimated during the selection protocol [64,67].
In the HA strain which was selected simultaneously for high mating activity, the genetic load was minimal, whereas in the LA strain nearly 50 % of 2nd chromosomes carried deleterious mutations.A considerable proportion of chromosomes in the highly active strain carried supervital mutations.This finding is consistent with comparisons of strains selected in the opposite direction.
By recombinational analysis we have demonstrated the presence of supressor mutations in the HA (high activity) inbred stock.The half-do minant and dominant supressor mutations can be viewed as elements of a compensation complex according to Strunnicov [99].
The study of behaviour of different movable elements in inbred stocks was accomplished in collaboration with V. A. Gvozdev and his colleagues in the Institute of Molecular Genetics of Russian Academy of Sciences (Moscow).The distribution of movable elements mdgl, mdg3 and mdg4 and copia was studied in the original inbred stocks and its derivatives selected in both plus and minus directions.Distinct interstrain differen ces were found with respect to the number and localization of mdg co pies.The original strains were quite stable over many years.Transposi tion frequencies of mdg elements according to rough estimates are close to Ы0- 4 .
The directed familial selection for increased adaptive properties re sulted in the segregation of certain rare families with an increased adap tive capacity.It turns out that in these families the genome is reconstruc ted: mdg elements appear at new sites and their number generally in creases.Copies usually vanish from old sites but in some cases a pro portion of them may remain at their previous locations.
It has been demonstrated for the first time that this type of genome restructuring involves a large number of MEs operating synchronously.The copies of MEs in new sites often appear together.The genome presu mably contains «hot spots» that are targets for highly effective transpositi on of MEs during selection for increased fitness.In the LA strain the num ber of mdgl movable elements was minimal.But in the substrains with increased viability from one to seven new sites of mdgl appeared.At least two mobile genes, mdg3 and copia y have an affinity to similar chro mosomal regions.This leads to the conclusion that their transposition is coordinated.Such trends have not been observed for other mdg families.
Copies of the mdg4 family never change their localisation in response to selection.
Gvozdev and Kaidanov [61] have put forward a hypothesis concer ning the existence of a system of adaptive transpositions of MEs which is based on the regulatory effect of MEs on the array of structural genes.The long terminal repeats of copia-like MEs contain both ter mination sites and enhancers.This gives them the capacity to regulate the gene.
Lethal mutations occurred in LA stocks at a high frequency.Some LA chromosomes with an apparent lethal mutation were maintained as Q/L/lethal heterozygotes in a balanced condition for nearly three years.Then they were screened for their disposition of mobile elements and for the presence of rearrangements.The result was striking.A considerable number (13 of 33 or 39,4%) of lethal containing chromosomes carried chromosomal rearrangements, predominantly paracentric and pericentric inversions.Most of the rearranged chromosomes (9 out 13) carried a combination of several inversions, and among them two also had trans positions and one a translocation.Apparently in crosses of CyL/Pm stra ins to LA, the induction of complex multisite rearrangements had taken place.They strongly resemble those appearing in crosses which induce H-E hybrid dysgenesis due to the transposition of /lofco-elements [78].It appears that one of the LA substocks contained about 30-35 copies of the hobo-element on all of its major polytene chromosome arms.
In addition to rearrangements the lethal chromosomes maintained in balancers (СуL/'lethal) showed an enormous level of mass transpo sition.In lethal chromosomes of LA strains less than 30 % of the origi nal sites of copia-like movable elements contain original disposition [68]!Such reshuffling of the genome is presumably the result of the activati on of Яобо-elements capable of inducing hybrid dysgenesis.
The discovery of the ordered or concerted transposition of mobile genetic elements made it necessary to re-evaluate earlier conventional views about the non-directional nature of hereditary variation.We pro bably have to recognise that selection can control the formation of pat tern of migration of mobile elements in the genome Hot spots of the ge nome are targets which quickly become occupied by mobile genes when selection is directed towards intensification of vital functions [61,65,68].
Long term-studies of gene pool dynamics, mu tation outbursts, and gene instability in Drosophila populations.Long term studies conducted since 1963 by Golubovsky and his associated resulted in a number of major conclusions: (1) the existence of parallel and synchronous changes in the gene pool in adjacent and distant natural populations of D. melanogaster; (2) the discovery of ubiquitous stable polymorphisms of the oncoge nic lethal mutation /(2)g/-lethal (2) giant larvae; (3) evidence that mutation outbursts in nature are connected with the appearance of unstable mutations and the activation of an mobile elements results in insertions and transpositions.
(4) a new understanding of the eukaryotic genome's organisation and the nature of spontaneous hereditary changes.
Dobzhansky and his colleagues have described synchronous fluctu ations in the frequency of defined inversions in distinct populations of D. pseudoobscura [4,19].Parallel results were found during long term studies on mutability and the dynamics of gene pools of lethals and vi sible mutations [7, 9-^11, 46, 49].Up to beginning of the 60s certain general features concerning the population concentration of recessive lethals and their allelic content were established.But what was needed was a detailed description of the population dynamics of the lethal pool and an estimation, in statu nascendi, the influence of the major microevolutionary factors (i.e. the mutation process, selection, drift, isolation etc).
Toward this it was decided to analyse the dynamics of the lethal ge ne pool (concentration, allelism and localization) simultaneously in two to three adjacent populations together with a genetic analysis of the geo graphically most distant ones.Populations from Uman were choosen (Uk raine, between Kiev and Odessa) because they had been initially studied by Berg and Olenov in the 30s.About 40 000 diallelic crosses were carri ed out and the fate of more than 100 lethal genes on 2nd chromosome was investigated.Comparisons of the patterns of lethality were made between the different seasons of fly reproduction (the beginning of spring, summer and the end of autumn) and in successive years.
We found two peculiar features of lethal pool dynamics: (a) a quasistationary state, and (b) parallelism in dynamics and allelic content of lethals in adjacent populations.The concentration of lethals accumu lated in heterozygotes was stably maintained in each population studied at a defined level.Allelism tests of intrapopulation samples (40-50 lethal containing chromosomes on average) showed that nearly half of the lethals sampled occurred singly, while other lethals were pereated, being represented by two, three or more alleles.Each year there was a consi derable turnover of the gene pool on account of the transition of rare (unique) lethals to the category of repeated ones and vice versa... hence the term quasistationary state.
Meanwhile a systematic and simultaneous comparison of lethal sets in adjacent populations revealed remarkable parllelism despite the con stant allelic turnover.Allelic sets of lethals from different populations studied in the same season usually showed greater similarity than sets of lethals extracted in successive periods from the same population.No mutation process and no genetic drift can explain such parallelism.
It has been suggested that the quasistationary state and parallel changes in allelic lethal content is a consequence of hostparasite selec tion with the possible tolerance of flies carrying some lethals in hetero-zygous condition to infectious microorganisms [8,49].The lethal alleles giving relative tolerance to their carrier provide a selective advantage as long as they are relatively rare.But after spreading in populations flies carrying such repeated lethals lose their advantage when new mutations of virulence appear in the infectious agent.The similar ordered flow of lethal alleles is thereby ensured in different host fly populations [49].
This picture of permanent turnover of the gene pool of lethals is complemented by the remarkable ubiquitous stable polymorphism of le thal alleles of the gene l(2)glAethal giant larvae (or simply Igl).
The ubiquitous distribution of l(2)gl alleles in natural populations of Ukraine, Crimea, Middle Asia and Far East was demostrated in two periods: 1963-1978 [43][44][45], and then in 1979-1989.Hundreds of ge nes are capable to mutate to the lethal condition on the 2nd chromosome of D. melanogaster and it is namely Igl alleles that mutate most frequent ly.We observed that each fly out of 20-50 is +/lgl in nature.What is the reason for their ubiquitous distribution?It was found that -\-/lgl ani mals have an advantage in stressful conditions: the action of cold or high temperature during development and infection of virulent picornavirus DCV [89,101,102].
During the middle of 60s Elisabeth Gateff made a fascinating disco very that blockage of metamorphosis in Igl homozygotes is due to inva sive neoplastic growth of brain tissue cells.It appears that Igl is the first example of a monogenically controlled neoplasm in Drosophila and the first case of a tumor-supressor gene found in animals [32].
Phenogenetic analysis of Igl alleles extracted from different popu lations showed their multiple allelism.So we are not dealing here with the single mutation having arisen once and then spread, but with a con tinuous mutation process at a given locus following selection of newly arisen lethal alleles [101].This conclusion was confirmed by the mole cular analysis of natural Igl alleles made by Bernard Mechler who sho wed that all lethal Igl alleles are intragenic deletions and insertions [82].The same «roo» or В104 movable element is inserted in Igl locus in distant populations.At the same time Green and Shepherd [59] found that Igl mutations are also frequent in a Californian population and that the Igl gene is the target of the action of a powerful MR mutator fac tor.In the presence of MRh-12 microdeletions in the telomeric region of the left arm of 2nd chromosome, where Igl is located, originate with a frequency of 10 -3 .
Thus it turns out that the selective gene in natural populations is itself capable to regulate the level of its mutability, maintaining in a population its specific mutator.From a general point of view it is essen tial to note that lethal oncogenic mutations may provide a heterozygous selective advantage.Can we exclude a similar situation occurring in hu mans where several tumor-supressor mutations have also been discovered (human neuroblastoma is very similar to neuroblastoma induced by 1(2)gl alleles)?
A third aspect of the long-term studies of genetic variability in na tural populations concerns the monitoring of the mutation process.The concept of specific periods during which the mutation rate sharply rises was first developed by Hugo de Vries but subsequently rejected.Longterm population studies definitely support this old idea.Particularly de monstrative are outbursts of mutability at the same loci observed practi cally simultaneously in distant populations.In 30s Russian population geneticists found high mutability in the sex-linked genes yellow and whi te [6,7,28,38].Afteh the 40s the mutability of yellow and white retur ned to a normal level [7].From the beginning of 80s the mutability le vel of the yellow gene in the Uman population once more sharply incre ased.But this time the «fashionable» alleles were not yellow 1 but yellow 2 [56,117].
In summary, the authors, up to 1973, have accumulated much expe rimental evidence for what they have called the phenomenon of «Mode of Mutations».They predicted that one definite mode «will pass giving way to other visible anomalies» [49].This proved to be the case.Namely in 1973 in two geographically isolated populations (Caucasus and Middle Asia region) an outburst of mutability of the sex-linked singed bristle locus was discovered [10,62].The sn mutability increased a hundred fold to (0,3-0,9) • 10~3.The peculiar thing about this mutation mode was the simultaneous appearance of multiple superunstable alleles in distant populations and the fact that not only the mutant but their normal deri vatives are quite unstable.
The detailed genetic analysis of unstable s/x-alleles lead to the con clusion that instability of the sn-gene is the result of insertions [42, 50,.Fig. 1.Natural genetic engineering.Transposon loading by two mutations in the X-chromosome isolated from a Far East population.The two genes s^-singed bristles and с/ш-club wing concomiantantly manifest and form one bimutant unstable system.Symbols src+, sn m , sn s and sn ex correspond to the phenotypes wild, moderate, strong and extreme mutant bristle.Arrows point to the direction of mutations of a given al lele.A scheme shows that mutation transitions are ordered in three clusters /, 2. and 3. Clusters differ by phenotypes of sn-unstable derivatives, the direction and frequency of their mutation transitions and the character of concomitant c/w manifestation.Inside of each cluster the rate of mutations is enormous, up to 2-5 % in average.Transitions between clusters are quite rare (10~4) and connected with the appearance of a new snallele.In clusters / and 3 sn-alleles mutate according to the principle «all or none» in both directions.But: (i) in cluster 1 c/w phenotype manifest only concomitantly with SAi-mutant expression and sn+ always c/w+\ (il) in cluster 3 c/w manifests both in sn ex and sn+ derivatives [112] 57].The behaviour of the unstable sn-gene was quite similar to the be haviour of the unstable mutation studied by B. McClintock.Thus her re markable hypothesis about the existence of movable controlling elements appeared to be valid not only for maize and for some laboratory alleles in Drosophila but also for natural Drosophila populations.At Stadlers T Symposium in 1978, M. Green commented that the discovery of wideworld appearance and the distribution of unstable sn-alleles «represen ted the first systematic recovery of mutable genes in Drosophila and po sed the novel question: what is the origin of unstable sn mutants?» [58].
As it is well known now, in 1977-78, mobile genetic elements (ME) were discovered in Drosophila independently by two groups of molecular geneticists in USA and Russia.The sn locus appeared to be the target for the now famous P element.More than 50 % of unstable sn alleles from natural populations appeared to be connected with P-insertion [54].P-mediated unstable s/i-alleles isolated from one or distant populations differed on their phenotype, rate and direction of mutation in germ in somatic cells.
Unstable sn alleles isolated at the same time from one population appeared to be connected with the insertion of different ME (not only P but also copia-like ME).So the conclusion is inevitable that the src-mutation outburst is the result of the simultaneous activatio-п of different mo bile elements [54].
In Table 2 we summarised certain new phenomena of genetic varia bility discovered in our genetic studies of insertion mutations isolated from natural populations.Of special interest is the first described case of natural genetic engineering.Two separate genes «singed bristle» and «club wing» on the X-chromosome isolaied from a Far Eastern population appeared to be joined under the control of one ME capable of transposition (pressumably mdg3).As a result of such engineering bimutant unstable construction occurred [51,52,112].This insertional bimutant construction was unstable and oave normal derivatives which once more reversed to the bimutant condition.Many of the genetic aspects of this new genetic construction mediated by transposon behaviour were clarified.The spectrum of ordered changes of this natural engineering construction was thoroughly studied (Fig. 1).
We now need a wide presentation of sources and factors inducing genetic variability.It is evident that classical mutations make up only part of the genetic variability found in eukaryotic genomes.It was suggested that the eukaryotic genome can be seen as a population of DNA and RNA self reduplicating elements [75].It is then useful to make a further division.We suggested [48] a distrimination in the eukaryotic genome between the two subsystems: the OBLIGATE and the FACUL-TATIVE.We believe this subdivision is a natural one.The obligate part includes the set of genetic loci contained on chromosomes and in the DNA of cellular organelles.In classical genetics each genetic locus has a definite position on a genetic map.It is the skeleton of the genome.But the genome also includes different kinds of facultative elements, varying in number and cell topography.The DNA and RNA elements may vary in number and topography from cell to cell and from individual to individual.The facultative part of the genome includes the hierarchy of elements from highly repeated and satellite DNA to plasmids, B-chronioso mes and certain (relatively stable) cytobionts present both in the nucleus and the cytoplasm.
This natural structural subdivision leads to the discrimination between changes within obligate elements (changes of their structure, num-

Level of organization
Character of variability and genetic events Genes and Gene instability due to insertions of P and other MEs activated in chromosomes natural populations [48].Allele specific character of instability in germinal and somatic cells [43,44,46].Increased rate of mutations in X-ehromosomes containing unstable sn-alleles [47].Paramutation phenomenon in the progeny of heterozygotes on two unstable sn-alleles [48].
Influence of homologous locus structure on the mutation rate of an insertional allele [113,114].Recombination in heterozygotes on two insertional alleles is abnormal, it is induced by the excision of MEs and occurs premeiotically [113,114].The complex control of insertional mutability: cytotype, mutator genes, interactions of MEs [52,55,108].Individual (De-Clusters of mutants in the progeny of one individual with unstable velopmental) alleles due to transpositions of ME during the first zygotic divisions [50,56].Genetic engineering in nature: two separate genes under the control of one ME are simultaneusly expressed and matate [112].Expression of insertional mutation may depend on temperature and the Y-ohromosome and is similar to position effect [1,15] Thus interactions between the components of the system ENVIRON MENT -FACULTATIVE -OBLIGATE ELEMENTS is the major sour ce of most genetical changes occurring de novo in nature.This scheme is shown in Fig. 2 where narrows indicate the nature of the links, and their width schematically correspond to their intensity or force.If mutations usually occur at a low frequency, unpredictable, variations may occur frequently and orderly.
The subsystem of the facultative elements is the first one which re acts to environmental changes.Many variations (changes in the number of high repeated or stD'NA sequencies, changes in the number and topo graphy of amplified sequencies and mobile elements, and the presence or absence of cytobionts) may have no visible or physiological effects detectable by usual genetic methods.At the same time such variations as the transposition of MEs may induce insertional mutations and rearran gements of chromosomes with definite physiological consequences as ste rility.Changes in the number of definite facultative elements, for instan ce the number of RNA containing sigma viruses in the cytoplasm may have such important physiological consequences as susceptibility to car bon dioxide.The observation and estimation of the frequency of variati on-creation events needs a special selective system as in the case of MRfactors [57,58], hybrid dysgenesis or, as in the case of transpositional bursts and concerted transpositions discovered by the Russian geneti cists [34,35,60,61].
In conclusion we would like to remind the reader that the direction of Russian population studies and Dobzhansky's work were often syn chronous and parallelic.This fact is rooted both in the inner logic of sci ence and that both their «foundation principles», originated from the sa me scientific schools.Dobzhansky regularly exchanged letters with his Leningrad Univer sity student and friend Ju.Ja.Kerkis.At the same time he was keenly interested in population studies in Russia.
In one of his last letters, Dobzhansky discussing the puzzling pheno menon of synchronous fluctuations of mutability and the appearance of multiple series of unstable sn alleles in distant populations, remarked 62 ISSN 0233-7057.БИОПОЛИМЕРЫ И КЛЕТКА.1994.Т. 10. № 5 ber and position) and changes within facultative elements.The first class of changes correspond to classical mutations.Various changes of diffe rent facultative elements may be called variations.Variations are the most frequent class of hereditary changes, because faculttive elements are quite susceptible to a wide spectrum of evironmental changes.Muta tions or direct damage of genetic loci occur mainly after the strong how wrong much of molecular geneticists thinking was in the 70s.Po pulation genetics to them seemed to be as exhausted field like a squeezed lemon, wrote Dobzhansky.Indeed the discovery of the fact that multiple instability in nature is connected with the activation of mobile elements marks the beginning the promising synthesis of molecular and popula tion genetics.

, К Zokharoy T A Gerasimoya
Acknowledgment.The authors are very grateful to Richard Newcomb for his comments and help in the preparation of this manuscript.We thank M. M. Green, John Gibson, A. Korol, C. Krimbas, A. Saura, Nadine Plus, C. Biemont and J. David for reading of the original version of the ma nuscript, critical remarks and advices.

( 1 )
The Institute of De velopmental Genetics in Moscow headed by B. L. Astaurov, an eminent evolutionary geneticist from Chetverikov's group; (2) The Department of Genetics in St. Petersburg (Leningrad) University, headed by M. E. Lobashov from Filipchenko's school and (3) The Institute of Cytology and Genetics organised in 1957 in Novosibirsk Academic City (Academgorodok).The first director of Novosibirsk Institute was Dubinin.He invited there Julius Kerkis and Zoia S. Nikoro who collaborated with Chetverikov from the mid 30s up to 1948.In 1963 Raissa Berg headed the first laboratory of population genetics at The Novosibirsk Institute where she renewed her long term studies on mutability in natural Drosophila populations.

Table 1
Major contributions of Droscphilists to the field of population genetics before 1948

Table 2
Fig.2.Two types of genetic elements in the eukaryotic genomes and two types of ge netic changes: mutations and variations action of environmental factors (radiation, chemical mutagenes, defects in repair system et cetera).But the predominantly weak action of environ mental factors (in a broad sense-physical, biotical and genetical) is recei ved and reacted to initially by the set of facultative elements.