CHAPTER 8 Changing Genes: Site-directed Mutagenesis And .

POGC08 12/8/05 8:47 AM Page 141CHAPTER 8Changing genes: site-directed mutagenesis andprotein engineeringIntroductionThe generation and characterization of mutants isan essential component of any study on structure–function relationships. Knowledge of the threedimensional structure of a protein, RNA species,or DNA regulatory element (e.g. a promoter) canprovide clues to the way in which they function butproof that the correct mechanism has been elucidated requires the analysis of mutants that haveamino acid or nucleotide changes at key residues(see Box 8.2).Classically, mutants are generated by treatingthe test organism with chemical or physical agentsthat modify DNA (mutagens). This method of mutagenesis has been extremely successful, as witnessedby the growth of molecular biology and functionalgenomics, but suffers from a number of disadvantages. First, any gene in the organism can be mutatedand the frequency with which mutants occur in thegene of interest can be very low. This means thatselection strategies have to be developed. Second,even when mutants with the desired phenotype areisolated, there is no guarantee that the mutation hasoccurred in the gene of interest. Third, prior to thedevelopment of gene-cloning and sequencing techniques, there was no way of knowing where in thegene the mutation had occurred and whether itarose by a single base change, an insertion of DNA,or a deletion.As techniques in molecular biology have developed, so that the isolation and study of a single geneis not just possible but routine, so mutagenesis hasalso been refined. Instead of crudely mutagenizingmany cells or organisms and then analyzing manythousands or millions of offspring to isolate a desiredmutant, it is now possible to change specifically anygiven base in a cloned DNA sequence. This techniqueis known as site-directed mutagenesis. It has becomea basic tool of gene manipulation, for it simplifiesDNA manipulations that in the past required a great··deal of ingenuity and hard work, e.g. the creation orelimination of cleavage sites for restriction endonucleases. The importance of site-directed mutagenesisgoes beyond gene structure–function relationshipsfor the technique enables mutant proteins with novelproperties of value to be created (protein engineering).Such mutant proteins may have only minor changesbut it is not uncommon for entire domains to bedeleted or new domains added.Primer extension (the single-primer method)is a simple method for site-directed mutationThe first method of site-directed mutagenesis tobe developed was the single-primer method (Gillamet al. 1980, Zoller & Smith 1983). As originallydescribed the method involves in vitro DNA synthesiswith a chemically synthesized oligonucleotide (7–20nucleotides long) that carries a base mismatch withthe complementary sequence. As shown in Fig. 8.1,the method requires that the DNA to be mutatedis available in single-stranded form, and cloningthe gene in M13-based vectors makes this easy.However, DNA cloned in a plasmid and obtainedin duplex form can also be converted to a partiallysingle-stranded molecule that is suitable (DalbadieMcFarland et al. 1982).The synthetic oligonucleotide primes DNA synthesis and is itself incorporated into the resultingheteroduplex molecule. After transformation ofthe host E. coli, this heteroduplex gives rise to homoduplexes whose sequences are either that of theoriginal wild-type DNA or that containing the mutatedbase. The frequency with which mutated clones arise,compared with wild-type clones, may be low. Inorder to pick out mutants, the clones can be screenedby nucleic acid hybridization with 32P-labeled oligonucleotide as probe. Under suitable conditions ofstringency, i.e. temperature and cation concentration, a positive signal will be obtained only withmutant clones. This allows ready detection of the

POGC08 12/8/05 8:47 AM Page 142142CHAPTER 8TAChemically synthesizedoligonucleotideAnnealA*A G G CCACGT G*TACGT GAASingle-strandedM13 recombinantTCAGGCT(1) DNA polymerase 4dNTPs(2) T4 DNA ligase ATPAA*A G G CACGT G*TTCTransform E. coliATCA G G CACGT G*Screen plaques with 32P-labeledoligonucleotide as hybridizationprobeIsolate mutantdesired mutant (Wallace et al. 1981, Traboniet al. 1983). It is prudent to check the sequence of themutant directly by DNA sequencing, in order tocheck that the procedure has not introduced otheradventitious changes. This was a particular necessity with early versions of the technique which madeuse of E. coli DNA polymerase. The more recent useof the high-fidelity DNA polymerases has minimizedthe problem of extraneous mutations as well asshortening the time for copying the second strand.Also, these polymerases do not “strand-displace”the oligomer, a process which would eliminate theoriginal mutant oligonucleotide.A variation of the procedure (Fig. 8.2) outlinedabove involves oligonucleotides containing insertedor deleted sequences. As long as stable hybrids areformed with single-stranded wild-type DNA, prim-AAMutantTAA G G CCACGT G*TTWild-typeFig. 8.1Oligonucleotidedirected mutagenesis.Asterisks indicatemismatched bases.Originally the Klenowfragment of DNApolymerase was used,but now this has beenlargely replaced withT7 polymerase.ing of in vitro DNA synthesis can occur, ultimatelygiving rise to clones corresponding to the insertedor deleted sequence (Wallace et al. 1980, Norranderet al. 1983).The single-primer method has a number ofdeficienciesThe efficiency with which the single-primer methodyields mutants is dependent upon several factors.The double-stranded heteroduplex molecules thatare generated will be contaminated both by anysingle-stranded non-mutant template DNA that hasremained uncopied and by partially double-strandedmolecules. The presence of these species considerablyreduces the proportion of mutant progeny. Theycan be removed by sucrose gradient centrifugation··

POGC08 12/8/05 8:47 AM Page 143143Changing genes: site-directed mutagenesis and protein engineeringFig. 8.2Oligonucleotidedirected mutagenesisused for multiple pointmutation, insertionmutagenesis, anddeletion mutagenesis.Multiple sisMutant oligonucleotidewith multiple (four)single base pairmismatchesMutant oligonucleotidecarrying a sequence tobe inserted sandwichedbetween two regionswith sequencescomplementary to siteson either sides of thetarget site in thetemplateMutant oligonucleotidespanning the region tobe deleted, binding totwo separate sites, oneon either side of thetargetor by agarose gel electrophoresis, but this is timeconsuming and inconvenient.Following transformation and in vivo DNAsynthesis, segregation of the two strands of theheteroduplex molecule can occur, yielding a mixedpopulation of mutant and non-mutant progeny.Mutant progeny have to be purified away fromparental molecules, and this process is complicatedby the cell’s mismatch repair system. In theory, themismatch repair system should yield equal numbersof mutant and non-mutant progeny, but in practicemutants are counterselected. The major reason forthis low yield of mutant progeny is that the methyldirected mismatch repair system of E. coli favors therepair of non-methylated DNA. In the cell, newlysynthesized DNA strands that have not yet beenmethylated are preferentially repaired at the positionof the mismatch, thereby eliminating a mutation. Ina similar way, the non-methylated in vitro-generatedmutant strand is repaired by the cell so that themajority of progeny are wild type (Kramer et al.1984). The problems associated with the mismatchrepair system can be overcome by using host strainscarrying the mutL, mutS, or mutH mutations, whichprevent the methyl-directed repair of mismatches.A heteroduplex molecule with one mutant andone non-mutant strand must inevitably give riseto both mutant and non-mutant progeny uponreplication. It would be desirable to suppress thegrowth of non-mutants, and various strategies havebeen developed with this in mind (Kramer, B. 1984,Carter et al. 1985, Kunkel 1985, Sayers & Eckstein1991).Another disadvantage of all of the primer extension methods is that they require a single-strandedtemplate. In contrast, with PCR-based mutagenesis··(see below) the template can be single-stranded ordouble-stranded, circular or linear. In comparisonwith single-stranded DNAs, double-stranded DNAsare much easier to prepare. Also, gene inserts are ingeneral more stable with double-stranded DNAs.The issues raised above account for the fact thatmost of the mutagenesis kits that are available commercially make use of multiple primers and doublestranded templates. For example, in the GeneEditorTMsystem (Fig. 8.3), two primers are used. One of theseprimers encodes the mutation to be inserted intothe target gene. The second encodes a mutation thatenhances the antibiotic resistance properties of theampicillin-resistance determinant on the vector byconferring resistance to ceftazidime as well. Afterextending the two primers to yield an intact circularDNA molecule, the mutated plasmid is transformedinto E. coli and selection made for the enhancedantibiotic resistance. Plasmids encoding the enhancedantibiotic resistance also should carry the mutatedtarget gene. In a variant of this procedure, the vectorhas two antibiotic resistance determinants (ampicillinand tetracycline) but one of these (AmpR) carries amutation. Again, two primers are used: one carryingthe mutation to be introduced to the target geneand the other restores ampicillin resistance. Afterthe in vitro mutagenesis steps, the plasmid is transformed into E. coli and selection made for ampicillinresistance.Methods have been developed that simplifythe process of making all possible amino acidsubstitutions at a selected siteUsing site-directed mutagenesis it is possible tochange two or three adjacent nucleotides so that