Genetic Variation and Molecular Evolution

1

Genetic Variation and MolecularEvolution

Werner ArberBiozentrum, University of Basel, Klingelbergstrasse 70, Basel, Switzerland

1 Introduction 3

2 Principles of Molecular Evolution 42.1 Evolutionary Roles of Genetic Variation, Natural Selection,

and Isolation 42.2 Molecular Mechanisms of the Generation of Genetic Variation 6

3 Genetic Variation in Bacteria 7

4 Local Changes in the DNA Sequences 8

5 Intragenomic DNA Rearrangements 95.1 Site-specific DNA Inversion at Secondary Crossover Sites 105.2 Transposition of Mobile Genetic Elements 11

6 DNA Acquisition 12

7 The Three Natural Strategies Generating Genetic VariationsContribute Differently to the Evolutionary Process 13

8 Evolution Genes and Their Own Second-order Selection 15

9 Arguments for a General Relevance of the Theory of Molecular Evolution forAll Living Organisms 16

10 Conceptual Aspects of the Theory of Molecular Evolution 1710.1 Pertinent Scientific Questions 17

2 Genetic Variation and Molecular Evolution

10.2 Philosophical Values of the Knowledge on Molecular Evolution 1810.3 Aspects Relating to Practical Applications of Scientific Knowledge on

Molecular Evolution 20

Bibliography 21Books and Reviews 21Primary Literature 21

Keywords

Biological EvolutionA nondirected, dynamic process of diversification resulting from the steady interplaybetween spontaneous mutagenesis and natural selection.

DNA RearrangementResults from mostly enzyme-mediated recombination processes, which can be intra- orintermolecular.

Evolution GeneIts protein product acts as a generator of genetic variations and/or as a modulator ofthe frequency of genetic variations.

Gene AcquisitionResults from horizontal transfer of genetic information from a donor cell to a receptorcell. With bacteria, this can occur in transformation, conjugation, or phage-mediatedtransduction.

Natural SelectionResults from the capacity of living organisms to cope with the encounteredphysicochemical and biological environments. Largely depending on its genetic setupand physiological phenotype, each organism may have either a selective advantage or aselective disadvantage as compared to the other organisms present in thesame ecosystem.

Spontaneous MutationDefined here as any alteration of nucleotide sequences occurring to DNA without theintended intervention of an investigator. The term mutation is used here as a synonymof genetic variation.

TranspositionDNA rearrangement mediated by a mobile genetic element such as a bacterialinsertion sequence (IS) element or a transposon.

Genetic Variation and Molecular Evolution 3

Variation GeneratorAn enzyme or enzyme system whose mutagenic activity in the generation of geneticvariation has been documented.

� The comparison of DNA sequences of genes and entire genomes offers interestinginsights into the possible evolutionary relatedness of genetic information of livingorganisms. Together with a relatively rich database from experimental microbialgenetics, conclusions can be drawn on the molecular mechanisms by which geneticvariations are spontaneously generated. A number of different specific mechanismscontribute to the overall mutagenesis. These mechanisms are here grouped into threenatural strategies of the spontaneous generation of genetic variations: local changesof DNA sequences, intragenomic rearrangement of DNA segments, and acquisitionof foreign DNA by horizontal gene transfer. These three strategies have differentqualities with regard to their contributions to the evolutionary process. As a generalrule, none of the known mechanisms producing genetic variants is clearly directed.Rather, the resulting alterations in the inherited genomes are more random. Inaddition, usually only a minority of resulting variants provide a selective advantage.Interestingly, in most of the molecular mechanisms involved, the products of so-called evolution genes are involved as generators of genetic variation and/or asmodulators of the frequencies of genetic variation. Products of evolution genes workin tight collaboration with nongenetic factors such as structural flexibilities andchemical instabilities of molecules, chemical and physical mutagens, and randomencounter. All of these aspects contributing to the spontaneous generation of geneticvariations together form the core of the theory of molecular evolution. This theorybrings neo-Darwinism to the molecular level. In view of the increasing evidencecoming particularly from microbial genetics, knowledge of molecular evolution canbe seen as a confirmation of Darwinism at the level of biologically active molecules,in particular, nucleic acids and proteins. Philosophical and practical implications ofthis knowledge will be briefly discussed.

1Introduction

Evolutionary biology has traditionally de-voted its major attention to the comparisonof phenotypical traits of higher organ-isms, both of those actually living andof those that have been extinct (paleon-tological fossils). The resulting theory ofdescent is, together with other criteria, atthe basis of the systematic classification

of living organisms. Darwin’s theory ofnatural selection brought a new elementinto the understanding of the long-termdevelopment of forms of life. Natural se-lection is the result of organisms copingwith the encountered living conditions thatare dependent on both the environmentalphysicochemical conditions and the activ-ities of all living forms in a particularecological niche. The Darwinian theoryof evolution also postulated that intrinsic


properties of life are not entirely stableand principally identical for all organ-isms of a given species. In the so-calledmodern synthesis, in which evolutionarybiology and genetics became integrated,transmissible phenotypic variations repre-senting the substrate of natural selectionwere explained as due to genetic varia-tions (or mutations). Shortly thereafter,deoxyribonucleic acid (DNA) was identi-fied as the carrier of genetic information.DNA is thus also the target for mutagen-esis. Within the last few decades a rapiddevelopment of molecular genetics withnovel research strategies leading to ge-nomics, sequencing of entire genomes,and functional studies of genes and theirproducts paved the way for hitherto in-accessible knowledge on the basis of lifeand its multiple manifestations. This alsorelates to the process of molecular evolu-tion. A synoptical insight into the variousmolecular mechanisms contributing to thegeneration of genetic variations representsa molecular synthesis between the neo-Darwinian theory and molecular genetics.This synthesis can confirm the Darwinianevolution at the molecular level.

2Principles of Molecular Evolution

The principles of molecular evolution tobe outlined here are founded on

1. the neo-Darwinian theory of evolutionwith its three pillars of genetic variation,natural selection, and isolation;

2. the solidly established microbial genet-ics database;

3. DNA sequence comparisons with bioin-formatic tools;

4. physicochemical knowledge on the re-activity, conformational flexibility, and

chemical stability of biologically ac-tive molecules.

2.1Evolutionary Roles of Genetic Variation,Natural Selection, and Isolation

The long-term maintenance of any formof life requires a relatively high stabilityof its genetic information. However, rareoccasional genetic variations occur in allorganisms. This gives rise to mixed pop-ulations of organisms with the parentalgenome and organisms having one ormore alterations in their genome. Thesepopulations are steadily submitted to nat-ural selection. The experience shows that,in general, favorable genetic variations areconsiderably less frequent than unfavor-able ones. The latter provide a selective dis-advantage. Indeed, genetic variants withunfavorable genetic alterations will sooneror later get eliminated from propagatingpopulations, which will become enrichedfor organisms carrying favorable geneticvariations. It should be noted that by farnot all alterations in the nucleotide se-quences of a genome will lead to a changein the phenotype of the organism. How-ever, such silent and neutral mutationsmay later become physiologically relevantin conjunction with still other, upcomingDNA sequence changes.

Natural selection is by no means aconstant element. It varies both in timeand in space. This is due to variationsin the physicochemical environmentalconditions as well as to variations of thelife activities of all the different organismspresent in a particular ecological nicheand forming an ecosystem. Since a geneticvariation may also affect the influencethat the organism exerts on the otherorganisms present in the same ecosystem(e.g. think of weeds and pathogenicity


effects, but also of beneficial, synergisticeffects), any novel mutation may notonly influence the life of the concernedorganism itself but also the lives of othercohabitants of the same ecological niche.

The third pillar of biological evolu-tion – besides genetic variation and naturalselection – is isolation. Evolutionary biolo-gists define two different aspects of isola-tion. One of these is geographic isolation,which may seriously reduce the numberof potential habitats for an organism. The

other type of isolation is called reproductiveisolation. For example, two distantly re-lated diploid organisms may not be fertilein sexual reproduction. But reproductiveisolation can also be seen in a wider def-inition to seriously limit the possibility ofhorizontal transfer of segments of geneticinformation between two different kindsof organisms.

As summarized in Fig. 1, genetic vari-ation drives biological evolution. Com-plete genetic stability would render any

Sources of genetic diversity

Limitation of genetic diversity

Modulatesbiological evolution

Drives biological evolution

Directsbiological evolution

Organisms

Ecosystems

Geographic

Reproductive

Strategies Mechanisms

Isolation

Mutation

Genetic variation

Natural selection

Local change ofDNA sequence

DNArearrangement

DNAacquisition

Replicationinfidelities

Mutagens- internal- environmental

Recombinationalreshuffling of genomicDNA segments

Horizontalgene transfer

Physico chemical environment

Biological environment

Carrier capacity of biosphere

Fig. 1 Synoptical presentation of major elements of the theory of molecularevolution. A number of specific mechanisms, each with its own characteristics,contribute to the four groups of mechanisms of genetic variation listed. Each ofthe specific mechanisms follows one and sometimes more than one of the threeprincipal, qualitatively different strategies of genetic variation.


evolutionary process impossible. A veryhigh frequency of genetic variation wouldrapidly lead to the extinction of the con-cerned organisms because of the statedprevalence of unfavorable mutations in thespontaneous generation of genetic varia-tions. It is natural selection together withthe available sets of genetic variants thatdetermines the direction of biological evo-lution, or in other words, the directionsin which the branches of the tree of evo-lution grow. Finally, the geographic andreproductive isolations modulate the evo-lutionary process.

2.2Molecular Mechanisms of the Generationof Genetic Variation

The concept to be presented here re-quires the reader to question some long-established textbook knowledge, such asthe claim that spontaneous mutationswould largely result from errors, accidents,and illegitimate processes. We defend herethe alternative view that living nature ac-tively cares for biological evolution

1. by making use of intrinsic propertiesof matter such as a certain degree ofchemical instability and of structuralflexibility of molecules, and

2. by having developed genetically en-coded systems, the products of whichare involved in the generation of ge-netic variations and in modulating thefrequencies of genetic variation.

It might be relevant to mention here thatthe term mutation is differently defined inclassical genetics and in molecular genet-ics. In classical genetics, a mutant is anyvariant of a parental form, showing inits phenotypic properties some alterationthat becomes transmitted to the progeny.

In contrast, it has become a habit inmolecular genetics to call any alterationin the parental nucleotide sequence of thegenome a mutation, whether it has pheno-typic consequences or not. There is goodreason to assume that in most sponta-neously occurring mutagenesis events, thespecific mechanism involved will not payattention to whether the sequence alter-ation at the involved target site will causea phenotypic alteration or not. Therefore,for studies on mechanisms and on thestatistics of their occurrence, it is indicatedto follow the molecular genetics definitionof the term mutation, which we use hereas a synonym of genetic variation. We willuse the term spontaneous mutation to labelany type of DNA sequence alteration unin-tended by the investigator. This definitionsays nothing about whether a mutationrelates to a phenotypic change.

At present, the research on genetic vari-ation mainly follows two strategies. First,the increasing availability of entire andpartial genome sequences offers excellentmeans to compare regulatory sequenceelements, specific domains of genes, en-tire genes, groups of genes, and entiregenomes with regard to DNA sequencehomologies and genome organization.Within a given species, this can reveala genetic polymorphism. Between more orless related species, it can give a reliablemeasure for evolutionary relatedness. Forexample, the molecular clock – an indica-tor of evolutionary relatedness – is basedon single nucleotide alterations. Sequencecomparisons can often suggest how se-quence alterations could have occurred inthe course of past evolution. Secondly,a more reliable insight into the genera-tion of genetic variations can be gainedby the observation of individual processesof mutagenesis. In view of the large sizeof genomes and of the rare and random


occurrence of spontaneous DNA sequencealterations, this approach is relatively dif-ficult. However, a rich database is alreadyavailable from microbial genetics, particu-larly from bacteria and their viruses andplasmids. Their relatively small genomesare haploid so that phenotypic effectscaused by genetic variation normally be-come rapidly manifested. With appropriateselection and screening techniques, thisallows one to identify occasional, function-ally relevant mutations in populations. Onthe other hand, investigations on structuralalterations in the genomes of individualbacterial colonies, for example, by thestudy of restriction fragment length poly-morphism, can reveal when and where onthe genome a DNA rearrangement musthave occurred.

A quite solid, general result of this type ofexperimental investigations reveals that inthe spontaneous generation of genetic vari-ations, it is not just a single mechanism atwork. Rather, a number of mechanisticallydifferent processes contribute to the over-all mutagenesis. We will discuss selectedexamples below. Interestingly, a criticalevaluation of the situation shows that thespecific mechanisms of mutagenesis of-ten depend both on nongenetic elementsand on activities of specific enzymes, theproducts of the so-called evolution genes.The multitude of thus identified, distinctmechanisms contributing to the forma-tion of genetic variations can be groupedinto three qualitatively different naturalstrategies (Fig. 1). These are

1. local changes in the DNA sequences,2. rearrangement of DNA segments

within the genome,3. the acquisition by horizontal transfer of

a DNA segment originating in anotherkind of organism.

Selected examples for each of thesestrategies will be discussed in Sects. 4, 5,and 6.

3Genetic Variation in Bacteria

Several seminal discoveries, largely basedon work carried out with microorganismsbetween 1943 and 1953, were essentialfor the later development of molecu-lar genetics.

1. It was realized that bacteria and bacte-riophages have genes that can mutate,and that spontaneous mutations inmicroorganisms normally arise inde-pendently of the presence of selectiveagents. It was also learned that thegenetic information of bacteria and ofsome bacteriophages is carried in DNAmolecules rather than in other biologi-cal macromolecules such as proteins.

2. The newly discovered phenomena ofDNA transformation, bacterial con-jugation, and bacteriophage-mediatedtransduction revealed natural means ofhorizontal gene transfer between differ-ent bacterial cells.

3. It was seen that horizontal gene trans-fer has natural limits, including sys-tems of host-controlled modification,which are today known as DNA restric-tion–modification systems.

4. Mobile genetic elements were identi-fied as sources of genetic instabilityand were seen to represent media-tors of genetic rearrangements. Whilesuch rearrangements are often causedby transposition, they can also resultfrom the integration of a bacteriophagegenome into the genome of its bac-terial host strain, which is therebyrendered lysogenic.


It is at the end of this fruitful period ofdiscoveries, in 1953, that structural anal-ysis of DNA molecules led to the double-helix model. The filamentous structure ofDNA molecules made it clear as to howgenetic information could be contained inthe linear sequences of nucleotides. Thedouble helical nature of the model alsooffered an understanding of semiconser-vative DNA replication at the molecularlevel and thus of information transferinto progeny.

Many classical microbial genetic inves-tigations were carried out with Escherichiacoli K-12. Its genome is a single circu-lar DNA molecule (chromosome) of about4.6 × 106 bp. In periods of growth, the rateof spontaneous mutagenesis is about 10−9

per bp and per generation. This representsone new mutation in every few hundredcells in each generation. E. coli has severalwell-studied bacteriophages and plasmids.This material facilitates investigations onlife processes in these bacteria.

Under good growth conditions, the gen-eration time of E. coli, measured betweenone cell division and the next, is very short,on the order of 30 min. Upon exponentialgrowth, this leads to a multiplication fac-tor of 1000 every 5 h. Thus, a populationof 109 cells representing 30 generationsis reached from an inoculum of a singlecell in only 15 h. This rapid growth rategreatly facilitates population genetic stud-ies and thus facilitates investigations onthe evolutionary process.

On the filiform DNA molecules of E. coliand its bacteriophages and plasmids, thegenetic information is relatively denselystored as linear sequences of nucleotidesor base pairs. Genes depend on thepresence of continuous sequences of basepairs (reading frames) that encode specificgene products, usually proteins, and ofexpression control signals that ensure

the occurrence of gene expression at therelevant time with the needed efficiency.Mutations can affect reading frames aswell as control signals, both of whichrepresent specific DNA sequences. Inaddition, some specific DNA sequencesrelate to the control of the metabolism ofthe DNA molecules, in particular, theirreplication.

In the following Sects. 4, 5, and 6, we willdescribe selected examples of mechanismscontributing each in its specific manner tothe formation of genetic variants. We willgroup them into the already mentionedthree major natural strategies (Fig. 1),although, as we will later see, some of thespecific mechanisms involve more thanone of these strategies.

4Local Changes in the DNA Sequences

The process of DNA replication is oneof the important sources of geneticvariation, which may depend both onstructural features of the substrate DNAand on functional characteristics of thereplication fork.

Some of the ‘‘infidelities’’ of DNA repli-cation are likely to depend on tautomericforms of nucleotides, that is, a structuralflexibility inherent in these organic com-pounds. Base pairing depends on specificstructural forms. Conformational changesof nucleotides can result in a mispairingif short-living, unstable tautomeric formsare ‘‘correctly’’ used in the synthesis ofthe new complementary strand upon DNAreplication. For this reason, we do not con-sider mutations resulting in this processas errors and call them infidelities. DNAreplication is indeed one of the sources ofnucleotide substitution, and this plays an


important role in the evolutionary devel-opment of biological functions.

An inherent low degree of chemical in-stability of nucleotides represents anothersource of mutagenesis. For example, cyto-sine can undergo oxidative deamination tobecome uracil. Upon replication, this givesrise to an altered base pairing and resultsalso in nucleotide substitution.

Other replication infidelities that mayrelate to slippage in the replication forkcan result in either the deletion or theinsertion of one or a few nucleotides in thenewly synthesized DNA strand. If suchmutations occur within reading framesfor protein synthesis, the phenotypic effectmay be drastic. This is the case when,in the protein synthesized from a geneaffected by a frameshift mutation, theamino acid sequence downstream of thesite of mutation strongly differs from thatof the nonmutated product. In addition,the size of such a mutated protein isusually altered depending on the chanceoccurrence of an appropriate stop codonin the new reading frame.

Proofreading devices and other enzy-matic repair systems prevent replicationinfidelities from producing mutations atintolerably high rates. Generally, they actrapidly after replication by screening forimperfect base pairing in the double helix.Successful repair thereby requires specificmeans to distinguish the newly replicatedDNA strand from its template, the comple-mentary parental strand. Because of thesecorrection activities, many primary mis-pairings are removed before they have theopportunity to become fixed as mutations.DNA repair systems modulate the frequen-cies of mutagenesis.

Genetic information of some viruses,and sometimes also segments of geneticinformation of chromosomal origin, maypass through RNA molecules, which may

later become retrotranscribed into DNA.No efficient repair systems are knownto act at the level of RNA. Indeed,RNA replication shows a higher degreeof infidelity than DNA replication. Inconsequence, genetic information thatbecomes replicated as RNA moleculesgenerally suffers increased mutation rates.

A relatively large number of internal andenvironmental chemicals exert mutageniceffects by means of molecular mechanismsthat in many cases are well understood andoften cause local DNA sequence changes.Some intermediate products of the nor-mal metabolic activities of a cell may bemutagenic and may thus contribute tospontaneous local mutagenesis. The muta-gens of the environment include not onlya multitude of chemical compounds butalso ultraviolet radiation and some physic-ochemical constraints such as elevatedtemperature, which influence the chem-ical stability of nucleotides. Each of thesemutagens and mutagenic conditions con-tributes in a specific way to the generationof genetic variations.

Again, many of the potential sequencealterations brought about by internal andenvironmental mutagens are efficiently re-paired by enzymatic systems. However,since the efficiency of such repair israrely 100%, evolutionarily relevant mu-tations persist.

5Intragenomic DNA Rearrangements

Various recombination processes are wellknown to mediate DNA rearrangements,which often result in new nucleotidesequences.

While in haploid organisms generalrecombination is not essential for prop-agation, it influences genetic stability at


the population level in various ways asa generator of new sequence varieties.For example, it can bring about sequenceduplications and deletions by acting atsegments of homology that are carriedat different locations in a genome. Gen-eral recombination is also known to act inthe reparation of damage caused to DNAby ionizing radiation. In this case, an in-tact genome can become assembled fromundamaged segments of sister copies ofthe chromosome by homologous recom-bination. The best-known contribution ofgeneral recombination to genetic diversityis meiotic recombination bringing aboutrandom recombinants between the pater-nal and the maternal chromosomes indiploid organisms.

Two other widely spread types of recom-bination systems are dealt with in moredetail in Sects. 5.1 and 5.2: site-specific re-combination and transposition. Both areknown to contribute to genetic variation.Still other recombination processes, suchas the one mediated by DNA gyrase can, forthe time being, perhaps best be groupedas illegitimate recombinations. This groupmay contain several different molecularmechanisms that act with very low effi-ciency and have remained at least in partunexplained.

5.1Site-specific DNA Inversion at SecondaryCrossover Sites

Genetic fusions represent the results ofjoining together segments of two genes(gene fusions) or of two operons (operonfusions) that are not normally together. Anoperon is a set of often functionally relatedgenes that are copied into messenger RNA(i.e. transcribed) as a single unit. As aresult of this organization, those genesare coordinately regulated; that is, they

are turned on or off at the same time.Therefore, in an operon fusion, one ormore genes are put under a differenttranscription control, but the genes perse remain unchanged. In contrast, genefusion results in a hybrid gene composedof sequence motifs and often of functionaldomains originating in different genes.

In site-specific DNA inversion, a DNAsegment bordered by specific DNA se-quences acting as sites of crossing-overbecomes periodically inverted by the actionof the enzyme DNA invertase. Dependingon the location of the crossover sites, DNAinversion can give rise to gene fusion orto operon fusion. The underlying flip-flopsystem can result in microbial populationscomposed of organisms with different phe-notypic appearances: if, for example, theDNA inversion affects the specificity ofphage tail fibers, as is the case with phagesP1 and µ of E. coli, phage populations withtwo different host ranges will result.

Occasionally, a DNA sequence that de-viates considerably from the efficientlyused crossover site, a so-called secondarycrossover site, can serve in DNA inversion,which thus involves a normal crossover siteand a secondary crossover site. This pro-cess results in novel DNA arrangements,many of which may not be maintained be-cause of lethal consequences or reducedfitness; but if a few new sequences arebeneficial to the life of the organism, thesemay be selectively favored. This DNA re-arrangement activity can thus be looked atas evolutionarily important. Since manydifferent DNA sequences can serve inthis process as secondary crossover sites,although at quite low frequencies, site-specific DNA inversion systems act asvariation generators in large populationsof microorganisms. I have thus postulatedthat this evolutionary role of DNA inver-sion systems may be more important than


their much more efficient flip-flop mecha-nism, which can at most help a microbialpopulation to more readily adapt to twodifferent, frequently encountered environ-mental conditions. As a matter of fact,other strategies could be used as well forthe latter purpose.

Computer-aided comparison of DNAsequences quite often reveals that inde-pendent genes may consist of a particulardomain with high homology and of otherDNA sequences showing no significantsigns of relatedness. DNA inversion us-ing secondary sites of crossing-over is apotential source of such mosaic genes.DNA inversion can span over relativelylarge distances in DNA molecules and hasthe advantage of not loosing any DNA se-quences located between the two sites ofcrossing-over. Deletion formation repre-sents another source for gene fusion, butit has the disadvantage that the DNA se-quences between the sites of crossing-overare usually eliminated.

5.2Transposition of Mobile Genetic Elements

Nine different mobile genetic elementshave been found to reside, often in severalcopies, in the chromosome of E. coli K-12derivatives. This adds up to occupationof about 1% of the chromosomal lengthby such insertion sequences, also calledIS elements. At rates on the order of10−6 per individual IS element andper cell generation, these mobile geneticelements undergo transpositional DNArearrangements. These include simpletransposition of an element and morecomplex DNA rearrangements such asDNA inversion, deletion formation, andthe cointegration of two DNA molecules.Because of different degrees of specificityin the target selection upon transposition,

the IS-mediated DNA rearrangements areneither strictly reproducible nor fullyrandom. Transposition activities thus alsoact as variation generators. In additionto DNA rearrangements mediated by theenzyme transposase, which is usuallyencoded by the mobile DNA elementitself, other DNA rearrangements just takeadvantage of extended segments of DNAhomologies at the sites of residence ofidentical IS elements at which generalrecombination can act. Altogether, ISelements represent a major source ofgenetic plasticity of microorganisms.

Transposition occurs not only in grow-ing populations of bacteria but also inprolonged phases of rest. This is readilyseen with bacterial cultures stored at roomtemperature in stabs (little vials containinga small volume of growth medium in agar).Stabs are inoculated with a drop of a bacte-rial culture taken up with a platinum loop,which is inserted (‘‘stabbed’’) from the topto the bottom of the agar. After overnightincubation, the stab is tightly sealed andstored at room temperature. Most strainsof E. coli are viable in stabs during sev-eral decades of storage. That IS elementsexert transpositional activities under thesestorage conditions is easily seen as follows.

A stab can be opened at any time, a smallportion of the bacterial culture removed,and the bacteria well suspended in liquidmedium. After appropriate dilution, bacte-ria are spread on solid medium. Individualcolonies grown upon overnight incuba-tion are then isolated. DNA from suchsubclones is extracted and fragmentedwith a restriction enzyme. The DNA frag-ments are separated by gel electrophoresis.Southern hybridization with appropri-ate hybridization probes can then showwhether different subclones reveal re-striction fragment length polymorphisms


(RFLPs), which are indicative of the occur-rence of mutations during storage.

If this method is applied to subclonesisolated from old stab cultures, andif DNA sequences from residential ISelements serve as hybridization probes,an extensive polymorphism is revealed.None or only little polymorphism is seenwith hybridization probes from uniquechromosomal genes. Good evidence isavailable that transposition represents amajor source of this genomic plasticityobserved in stabs, which at most allowfor a very residual growth at the expenseof dead cells. One can conclude thatthe enzymes promoting transposition aresteadily present in the stored stabs. Indeed,the IS-related polymorphism increaseslinearly with time of storage for periods aslong as 30 years. In a culture of E. coli strainW3110 analyzed after 30 years of storage,each surviving subclone had suffered onthe average about a dozen RFLP changes asidentified with hybridization probes fromeight different residential IS elements, ofwhich IS5 was the most active. Lethalmutations could of course not be identifiedin this study.

Lethal mutations that affect essentialgenes for bacteriophage reproduction canbe accumulated in the prophage stateof the phage genome in its lysogenichost. Such mutants can be screenedfor their inability to produce infectivephage particles upon induction of phagereproduction. Experiments were carriedout with E. coli lysogenic for a phageP1 derivative grown in batch cultures at30 ◦C for about 100 generations allowingfor alternative periods of growth and rest.Most of the independent lethal mutantscould thereby be identified to be causedby the transposition of an IS elementfrom the host chromosome into the P1prophage that is maintained in its host

as a plasmid. In these experiments, IS2was the most active element and it mainlyinserted into a few preferred regions of theP1 genome, but each time at another site.The used insertion targets did not showany detectable homology or similaritywith each other. This is another goodexample for an enzymatically mediatedvariation generator. The experiment assuch identifies IS transposition as a majorsource of lethal mutagenesis.

There is no evidence available that bacte-rial mobile genetic elements would play anessential role in the bacterial life span ex-tending from one cell division to the next.However, these elements are major playersin the evolution of bacterial populations.As we have seen here, they contribute tointragenomic DNA rearrangements. De-pending on the target sequences involved,the resulting mutations may often be lethalby interrupting essential reading frames orexpression control regions. Favorable mu-tations may be relatively rare, but these cancontribute to evolutionarily advantageousdevelopments of the genome. In Sect. 6,we will see that mobile genetic elementsalso play important roles in the naturalstrategy of DNA acquisition.

6DNA Acquisition

While the mutagenesis mechanisms be-longing to the strategies of local changesin the DNA sequences (Sect. 4) and of in-tragenomic DNA rearrangements (Sect. 5)are exerted within the genome and canaffect any part of the genome, an addi-tional strategy of spontaneous sequencealterations depends on an external sourceof genetic information. In DNA or geneacquisition, genetic information indeedoriginates from an organism other than


the one undergoing mutagenesis. In bac-teria, DNA acquisition can occur by meansof transformation, conjugation, or virus-mediated transduction. In the latter twostrategies of horizontal gene transfer, aplasmid or a virus, respectively, acts asnatural gene vector.

The association and dissociation of chro-mosomal genes with natural gene vectorsoften arises from transpositional activitiesand from general recombination acting atIS elements that are at different chromo-somal locations. These mechanisms havebeen well studied with conjugative plas-mids and with bacteriophage genomesserving in specialized transduction. Forexample, composite transposons, whichare defined as two identical IS elementsflanking a segment of genomic DNA (of-ten with more than one gene unrelatedto the transposition process), are knownto occasionally transpose into a naturalgene vector and, after their transfer into areceptor cell, to transpose again into the re-ceptor chromosome. Hence, together withother mechanisms, such as site-specificand general recombination, transpositionrepresents an important promoter of hori-zontal gene transfer.

Several natural factors seriously limitgene acquisition. Transformation, con-jugation, and transduction depend onsurface compatibilities of the bacteria in-volved. Furthermore, upon penetration ofdonor DNA into receptor cells, the DNAis very often confronted with restrictionendonucleases. These enzymes cause afragmentation of the invading foreignDNA, which is subsequently completelydegraded. Before fragments become de-graded, however, they are recombinogenicand may find a chance to incorporate all orpart of their genetic information into thehost genome. Therefore, we interpret therole of restriction systems as follows: they

keep the rate of DNA acquisition low, andat the same time they stimulate the fixationof relatively small segments of acquiredDNA to the receptor genome. This strat-egy of acquisition in small steps can bestoffer microbial populations the chance tooccasionally extend their biological capaci-ties without extensive risk of disturbing thefunctional harmony of the receptor cell byacquiring too many different functions atonce. These considerations have their rel-evance at the level of selection exerted onthe hybrids resulting from horizontal DNAtransfer. This selection represents one ofthe last steps in the acquisition process.

DNA acquisition by horizontal genetransfer is a particularly interesting sourceof new genetic information for the receptorbacterium because the chance that theacquired DNA exerts useful biologicalfunctions is quite high – most likely, it hasalready assumed the same functions in thedonor bacterium.

An interesting hypothesis links theuniversality of the genetic code with theimportant role played by horizontal genetransfer in the evolutionary developmentof the living world. According to thisview, those organisms using the mostcommon genetic language would, in thelong term, be able to profit best fromthe increasing worldwide pool of geneticfunctions under the pressure of adaptingto changing living conditions.

7The Three Natural Strategies GeneratingGenetic Variations Contribute Differently tothe Evolutionary Process

Biological evolution is a systemic process.As outlined above, many different specificmechanisms contribute to generate ge-netic diversity that represents at any time


the substrate for natural selection. Thebuilding up of functional complexity is astepwise process, in which many randomattempts of genetic alterations becomerapidly rejected, while relatively few novelsequences are approved as favorable bynatural selection and are maintained andamplified. The genome can thus be seen asa cabinet in which key information from fa-vorable historical developments is stored.Stepwise, additional favorable informationis added. In the context of changing selec-tive conditions, stored information, havinglost its functionally beneficial relevance,may be deleted or favorably altered. As wehave seen, a multitude of different mech-anisms are behind this dynamic process.For a better understanding of the events,we have grouped the identified mecha-nisms into three major natural strategiesof genetic variation: local changes in theDNA sequences, intragenomic DNA rear-rangements, and DNA acquisition. Thesethree strategies have different qualitieswith regard to their contributions to bi-ological evolution.

The local DNA sequence change isprobably the most frequently involvedstrategy of genetic variation. Indeed, itsfrequency, which depends primarily onintrinsic properties of matter, chemicalinstability and conformational flexibility,would be intolerably high if it would not bemodulated by efficient enzymatic systemsof DNA repair. Local sequence changesbring about nucleotide substitution, thedeletion and the insertion of one or a fewbase pairs, or a local scrambling of a fewbase pairs. These sequence changes cancontribute to a stepwise improvement ofa biological function. It must be kept inmind that the functional test for suchimprovement is carried out by naturalselection. In principle, a long series ofstepwise local sequence changes could

also be expected to bring about a novelbiological function. However, this kind oflong-term process can gain efficiency onlyonce natural selection starts to be exertedon such upcoming function.

In contrast, the reshuffling of DNAsegments within the genome can beconsidered as a tinkering with existingelements, whereby favorable gene fusionsand operon fusions may occasionallyresult. DNA rearrangement can also bethe source of gene duplication and higheramplification, which are widely recognizedcontributions to the evolutionary progress.

In Sect. 6, we have already pointed to theevolutionarily high efficiency of horizontalgene transfer. As a matter of fact, DNAacquisition allows the recipient organismto share in the success of evolutionary de-velopments made by others. In drawingthe evolutionary tree of bacteria, DNA ac-quisition should be accounted for by moreor less randomly adding temporal horizon-tal shunts between individual branches. Itmust be kept in mind that usually onlysmall DNA segments flow through suchshunts in horizontal gene transfer.

Several of the specific mechanisms ofgenetic variation employ, strictly speak-ing, more than one of the three strategiesshown in Fig. 1. In transposition of ISelements, for example, a chromosomalDNA segment consisting of the mobilegenetic element can undergo a translo-cation and thereby become inserted ata new target site. As a rule, the targetsequence thereby gets duplicated, whichusually involves a few nucleotides. Thus,this transposition event will consist of botha DNA rearrangement and a local sequencechange.

As far as we know, most of the well-studied microbial strains use in paralleleach of the three natural strategies forthe generation of genetic variations. In


addition, bacteria very often use not onlyone, but several different specific mech-anisms for mutagenesis by each of thestrategies. Dissimilar specific mechanismsoften work with different efficiencies asreflected by their contribution to the over-all mutation rate. For any given strategy,it might be less relevant which specificmechanism is at work than the fact thatthe particular strategy finds its applicationwith an evolutionarily useful efficiency.In other words, specific mechanisms maysubstitute for each other within a strategy,at least to some degree. This rule doesnot apply between the strategies becauseof the difference in the qualities of theirevolutionary contributions.

The efficiency displayed by a givenspecific mechanism of spontaneous mu-tagenesis may depend on both internal(e.g. availability of enzymes that medi-ate mutagenesis) and external factors (e.g.environmental stress). It is also to benoted that some mechanisms may actmore or less randomly along a DNAmolecule, while other mechanisms mayshow regional or site preferences for theiractivities. In view of these considerations,we tend to assume that an evolutionar-ily fit (or well prepared) organism shouldbest be able to use a few specific muta-genesis mechanisms for each of the threestrategies to generate genetic variations. InSect. 8, we will explain what we mean byevolutionary fitness.

8Evolution Genes and Their OwnSecond-order Selection

The attentive reader will have seen in thedescription of some specific mechanismscontributing to the spontaneous mutagen-esis that besides a number of nongenetic

factors, specific products of genes are veryoften at work. These gene products canbelong to systems for repair of DNA dam-age and will, in this case, modulate thefrequency of mutagenesis. Similarly, re-striction enzymes seriously reduce boththe chance of DNA acquisition and thesize of a DNA segment that may even-tually be acquired by the recipient cell.Other gene products such as transposasesand other mediators of DNA recombina-tion act as generators of genetic variations.Since variation generators and modula-tors of the frequencies of genetic variationare evolutionary functions, we call theunderlying genetic information evolutiongenes. In the microbial world, these evolu-tion genes generally play no essential rolein the physiology of individual lives go-ing by cell division from one generationto the next. Under laboratory conditions,neither restriction–modification systemsnor enzymes for DNA rearrangements areneeded for the propagation of bacteria. Therole of such enzyme systems is primarilyevolutionary and becomes manifest at thelevel of populations.

We assume that evolutionary genes arethemselves submitted to selective pres-sure. However, such selection cannotfollow the rules of direct selection forimprovements of essential functions suchas those of housekeeping genes. Rather,the selection for the presence and im-provement of a variation generator will beexerted at the level of populations. Clearly,it will also be an individual that may oneday undergo a mutation, which improvesan evolutionary function. This functionwill also be exerted in its progeny in whichappropriate genetic variants of genes for di-rectly selected products will be either moreor less abundant. Any genetic alterationthat affects an evolution gene and proves inthe long term to be of higher evolutionary


value will be maintained and will providean evolutionary benefit to the carrier ofthe involved gene. In the long run, thiswill lead to fine-tuning of the evolutionaryfunctions of both variation generators andof modulators of the frequency of geneticvariation. The underlying indirect selec-tion based on the cells ability to providegenetic variants at a well-balanced level iscalled second-order selection.

We must be aware that some geneproducts may exert their essential func-tions for the benefit of both the lifeof the individual and the evolutionaryprogress of the population. In these cases,we assume that evolutionary selectionis exerted for both kinds of functionsand will eventually bring the gene to afine-tuned state to optionally carry outits functions for each of the differentpurposes. However, as we have alreadymentioned, a number of gene products in-volved in genetic variation are inessentialfor the lives of individual bacterial cells.Similarly, the products of many house-keeping genes are inessential for biologicalevolution.

9Arguments for a General Relevance of theTheory of Molecular Evolution for All LivingOrganisms

Largely on the basis of evidence from mi-crobial genetics, we have so far postulatedthat the products of a number of evolu-tion genes contribute, each in its specificway, to the generation of genetic variants atevolutionarily useful frequencies. Thereby,the sources of mutagenesis may relate ei-ther to the activity of the evolution geneproduct itself (e.g. a transposase) or to anongenetic factor (e.g. a chemical muta-gen or an intrinsic structural flexibility of

a nucleotide). In many cases, nongeneticfactors and products of evolution genes co-operate in the formation of genetic variantsat physiologically tolerable and evolution-arily beneficial levels. This is, for example,the case in spontaneous mutagenesis by anenvironmental mutagen when some of theprimary damage on the DNA gets success-fully repaired while some other damageleads to a fixed mutation.

The theory of evolution postulates thatlife on Earth started almost four billionyears ago with primitive, unicellular mi-croorganisms. It is in the first two billionyears that microbes must have developedthe basis for the actual setup of evolution-ary strategies and the underlying evolutiongenes. One can postulate that the acquiredevolutionary capacities could have allowedsome microbial populations to undergo adivision of labor in more and more sta-ble associations of cells. This developmentmight later have led to multicellular or-ganisms. In this kind of development,the evolutionary fitness of the involvedorganisms might have been an impor-tant precondition. At still later stagesof further evolutionary development, thethree natural strategies for generating ge-netic variations (Sects. 4 to 7) must havecontinued exerting their evolutionary in-fluence together with some other factorssuch as the formation of symbiotic asso-ciations. As a matter of fact, we attachconsiderable evolutionary relevance to en-dosymbiosis of higher organisms withbacteria. Such situations of cohabitationmay form an ideal condition for occasionalhorizontal gene transfer between the closeassociates.

A scientifically justified quest for fur-ther experimental proof of the postulatesof the theory of molecular evolution re-mains quite difficult to be answered.Clearly, there is a need for research on


the spontaneous generation of geneticvariation in higher organisms, ideally atthe level of the genomes. While thisis already quite difficult in microorgan-isms, it is of increased perplexity withthe much larger genomes of higher or-ganisms. However, sequence comparisonsnow offer fruitful ways to search for se-quence homologies, sequence similarities,single nucleotide polymorphisms, as wellas for traces of intragenomic DNA rear-rangements and of horizontal transfer ofgenetic information. Data so far availableare in support of the principles of molec-ular evolution outlined in Sect. 2, whichare likely to be valid for any kind of livingorganism.

Some of the general evolutionary strate-gies developed in microorganisms musthave also turned out to be useful for the de-velopmental and physiological processesat somatic levels of higher organisms.The generation of antibody diversity inthe immune systems of vertebrates bygenetic rearrangements and so-called so-matic mutagenesis is a good example.Another example is the enzymatic re-pair of DNA damage caused in somaticcells by external mutagens such as UVirradiation.

These considerations illustrate thatwhatever gene function may prove tobe useful for whatever particular pur-pose, it may be evolutionarily retainedand in the course of time further fine-tuned. We have already encountered thisprinciple in the microbial world, wherewe have postulated multifunctional en-zymes (such as working both for thephysiology of the cells and for an evo-lutionary task) to become evolutionarilyimproved both by direct and by second-order selection for the various functions.This may also be the case in higherorganisms.

10Conceptual Aspects of the Theory ofMolecular Evolution

With reference to Sect. 2.2, it is fair to againexplicitly state that for the time being,evolution genes and evolution functionsare a concept rather than a fully proven fact.This concept is based on a particular way tointerpret numerous available experimentaldata. We will briefly analyze the difficultiesto clarify the situation in a scientificdebate. This will be followed by pointing tophilosophical and more practical values ofa deeper understanding of the molecularprocesses that drive biological evolution.

10.1Pertinent Scientific Questions

In the history of scientific investigations,biologists have often searched for evidencethat living organisms would be able tospecifically modify, or adapt, their geneticinformation in order to better cope withupcoming changes in the living condi-tions. Most of these attempts have failed togive the expected response. In other cases,where a certain degree of specific adap-tation could be observed, specific causalexplanations have sometimes been foundupon deeper investigation. However, thereis at present no good scientific evidencefor a general rule that genetic alterationswould always be directed toward a specificgoal. This situation favors the view thatspontaneous mutations affect DNA moreor less randomly, which is in line with thegeneral observation that only a minority ofspontaneous mutants prove to be favorableunder the encountered living conditions.

The postulate of evolution genes that actas generators of genetic variations may bea surprise in this context. This has to dowith a widely followed concept of genetic

Documents

Genetic Variation and Molecular Evolution