With the advent of recombinant DNA technology and the development of methods for the purification and sequencing of proteins, it became possible to clone the gene encoding any purified protein. With the cloned gene in hand, one can use any one of a number of techniques to introduce random mutations into a cloned sequence, or one can induce mutations in regions of interest, such as sites of putative functional motifs, using a variety of in vitro techniques. The mutant alleles are then tested in vivo for a change in phenotype. This approach is called reverse genetics. In the classic genetic approach one starts with a phenotype of interest and isolates mutants with an altered phenotype. No prior mechanistic understanding of the phenomena controlling the phenotype is required and a large array of genes is likely to be identified. The classic genetic approach casts a wide net and is capable of providing a broad view of a complex biological system. In the reverse genetic approach mutations are directed to a particular gene (protein) and the effects of these changes on the phenotype of the organism are then determined. Some of these phenotypic effects can be predicted based on the biochemical and/or biological information available about this protein. Some of the phenotypes might be novel and unexpected. Whatever the outcome, the reverse genetic approach provides detailed information regarding the function of a particular gene product.
Several methods of in vitro mutagenesis are available to the researcher. Some induce random changes; others are site directed and can induce specific changes in particular basepairs or short sequence regions. One should just be aware that there is tremendous flexibility in the type of mutation that can be induced. Protocols for these methods are available in the literature and are included with commercially available kits. These will not be discussed here.
Two very important considerations must be kept in mind when developing a strategy for testing the in vivo phenotype of the mutant alleles. First, mutations generated in vitro when tested in vivo may not have a detectable phenotype. If they do have a phenotype, the mutant alleles could be recessive. Therefore, analysis of the phenotype must be done in strains that carry only recessive mutant alleles of the gene under study. Preferably, the other copies of the gene present in the strain should be null alleles. This avoids any possibility of an interaction with the mutant gene product being tested. Construction of a null allele in Saccharomyces is extremely straightforward using the one-step gene replacement methods described in Chapter 1. Second, because subtle phenotypic variation among the mutant alleles is possible, it is important to test each mutant allele in isogenic strains. This is necessary in order to be able clearly to associate the phenotype of the mutant with that specific mutational alteration and not variations in the genetic background.
The need to use a null allele when testing the phenotype of a mutant gene becomes a problem if the gene product has an essential function. In Saccharomyces this difficulty is addressed quite simply, but in other organisms this presents a greater difficulty. To test whether a gene is essential in Saccharomyces, a null allele is constructed in a diploid strain. Tetrad analysis of this heterozygous diploid will give two viable and two nonviable spores if the gene is essential. If the diploid is transformed with an extrachromosomal plasmid carrying the wild-type allele of the gene of interest, all four spores should be viable. The two spores containing the chromosomal null allele are able to grow because of the plasmid copy of the gene and viability of these spores will be dependent upon continued presence of the plasmid.
To determine the phenotype of a mutant allele of the gene of interest, the heterozygous GENl/genlA diploid is transformed with the extrachromosomal plasmid carrying the mutant genl gene, sporulated under selective conditions to maintain the plasmid, and tetrad analysis done. If the two spores with the null mutation are not viable, then the plasmid-borne mutant allele is not functional. But, if all four spores are viable, then the mutant allele is capable of providing function, i.e. it is complementing the chromosomal null allele. The function of the genl mutant allele may not be entirely wild type and this can be determined by a detailed analysis of the phenotype of the mutant spores.
Another method is called the plasmid shuffle. Using the procedures described above, a strain containing the chromosomal null allele and a plasmid-borne wildtype allele (plasmid 1) is constructed. A second plasmid carrying the mutant allele of the gene and a different nutritional marker for selection (plasmid 2) is introduced into the same host cell. The doubly transformed host cell is grown under conditions that select for the maintenance of plasmid 2 but not plasmid 1. If plasmid 1 can be lost, then the mutant gene on plasmid 2 is functional and its phenotype, if any, can be studied. If the plasmid is never lost despite growth in nonselective conditions, then the mutant gene on plasmid 2 is nonfunctional. A very simple way of testing the ability to lose a plasmid uses the pink: white color change of ade2 versus ADE2 strains, respectively. This is called a colony sectioning assay. If the selective marker gene on plasmid 1 is ADE2 and the host strain is ade2, then the colony formed by this strain will be white so long as the cells contain plasmid 1. If the colony is allowed to grow on a medium containing adenine, then plasmid 1 can be spontaneously lost so long as the transformant is not dependent on the wild-type gene also carried by this plasmid. When plasmid 1 is lost the cell will be ade2, will produce the pink pigment, and all the progeny will be pink forming a pink sector in the white colony. Pink sectors are easily observed in colonies. The researcher can simply scan a large number of colonies growing on an adenine-containing medium for sectoring. If this is observed, the host cell is not dependent on plasmid 1 and the mutant gene on plasmid 2 is functional.
A third method uses tightly regulated promoters, such as CTR1, to control expression of the wild-type allele carried by plasmid 1 (Labbe & Thiele, 1999). If the mutant allele carried by plasmid 2 is not functional, then viability of the double transformant will be dependent on the expression of the wild-type gene, which can be determined by comparing growth under conditions where the gene is expressed and not expressed.
ADVANCED CONCEPTS IN MOLECULAR GENETIC ANALYSIS COLD-SENSITIVE CONDITIONAL MUTATIONS
Cold-sensitive mutations are especially useful because previous experience has demonstrated that cold-sensitive mutations occur most often in genes encoding protein components of complex multimeric structures such as ribosomes, kineto-chores, viral coats, and other large multiprotein complexes. Cold sensitivity often appears to be the result of a decreased ability of the components to assemble into functional structures at the lower temperature, while at normal growth temperature assembly occurs at adequate rates. Thus, if one were interested in isolating mutations in genes encoding components of such structures or investigating the assembly process, one should consider isolating cold-sensitive mutations. This is done using various methods of random mutagenesis on the gene carried on a plasmid vector. The mutagenized pool is transformed into a null mutant host strain and screened for a cold-sensitive phenotype. If the null allele is lethal, a method such as plasmid shuffle would have to be used.
As with temperature-sensitive mutations, cold-sensitive mutations are non-null mutations. Therefore, if one obtains allele-specific suppressors of a cold-sensitive mutation, this suggests a physical interaction between the protein with the cold-sensitive alteration and the suppressor gene product. If a cold-sensitive mutation weakens the interaction at the nonpermissive temperature, then the suppressor mutation is likely to strengthen the interaction at the nonpermissive temperature. It is impossible to predict the phenotype of the suppressor mutation in the absence of the original cold-sensitive mutation. If the interaction with the wild-type product is unaffected, then the suppressor mutation will have no mutant phenotype. On the other hand, if the interaction is disrupted, the suppressor mutation might have a mutant phenotype similar to the original mutation; or, if the interaction is strengthened, the suppressor mutation might have a novel phenotype. This information would have to be empirically determined.
Widespread use of dominant negative mutant alleles became possible with the advent of gene cloning and DNA technology (Herskowitz, 1987). A dominant negative mutation is one that disrupts the function of the wild-type allele. Thus, the use of dominant negative mutations has become a very powerful technique for studies of protein function in organisms where the usual methods of producing mutant alleles are difficult or impossible. The dominant negative is a way of eliminating the function of a particular gene without having to isolate loss-of-function mutations in that gene. One often sees this method used in studies with mammalian cells.
The term dominant negative seems contradictory because a negative mutation, which implies loss-of-function, should be recessive not dominant. Most often, the dominant negative mutation requires overexpression to be dominant but occasionally examples are discovered where this is not necessary. The need for overexpression is explained as follows. Many proteins bind to other proteins and form multiprotein complexes in which only one component has the catalytic activity while the others function as regulators of the catalytic subunit. A mutation in the catalytic domain of the protein will affect its catalytic activity but will not necessarily affect its binding to the regulatory components in the complex. If the other components are made in limiting amounts, then the overproduced mutant subunit will bind to and sequester (titrate) the available subunits and make them unavailable to the small amount of wild-type protein synthesized by the otherwise wild-type host cell. If the limiting regulatory subunit is a positive regulator a null-like phenotype will be produced. If the limiting subunit is a negative regulator a constitutive phenotype will be produced. In this manner, a loss-of-function mutation can dominate the phenotype of a wildtype host.
The Cdc28 protein is a cyclin-dependent protein kinase and binding of proteins called cyclins is necessary to activate the kinase activity of Cdc28p. Cdc28p kinase is an essential protein needed for the progress from G1 to S and G2 to M in the cell cycle. Cyclin proteins are present in limiting amounts at certain times in the cell cycle and if these are depleted the cell will be unable to enter the S phase or mitosis. If a mutation is introduced into CDC28 at a site encoding an essential residue of the kinase activity, such as in the ATP binding site, then the encoded protein will be inactive as a kinase and the mutation will be recessive. However, the ability of the altered Cdc28p to bind cyclin has not been affected by this mutation. Therefore, if this mutant cdc28 gene is overexpressed, the mutant protein product will be very abundant and the cyclin proteins will bind to this nonfunctional protein instead of to the much less abundant wild-type Cdc28p encoded by the chromosomal copy. By so doing, the mutant product blocks the activation of the wild-type Cdc28 protein thereby preventing it from functioning. Thus, when this recessive cdc28 mutation is overexpressed the nonfunctional phenotype becomes 'dominant'. In this way, the geneticist can explore the phenotype of a cdc28 null mutation without having to construct one. This is extremely useful in mammalian cells where constructing a deletion mutation is an extremely lengthy and expensive undertaking.
Occasionally, one finds a dominant loss-of-function mutation that does not need to be overexpressed. In these cases the gene product is part of a multiprotein complex of like subunits and the mutation has altered a functional activity of the protein but not its ability to form the multiprotein complex. The dominance results from the fact that even if one subunit of the complex is mutant the entire complex is nonfunctional. The tumor suppressor protein p53 is a noteworthy example (Ko & Prives, 1996).
The dominant negative mutation has additional uses. For example, one can use it to isolate the gene encoding the interacting protein using multicopy suppression. The mutant phenotype produced by the overexpression of a dominant negative mutant gene should be suppressed by the overexpression of the interacting protein. One would use some type of overexpression library, introduce this into the host cell carrying the overexpressed dominant negative mutation, and select/screen for wild-type-like suppressor-containing transformants. The suppressing plasmid can easily be recovered and the gene responsible for the suppression identified.
The dominant negative strategy can be used for structure-function analysis of the catalytic subunit of a multiprotein enzyme complex. It also could be used to characterize a multiprotein DNA-binding complex. Mutation of the DNA-binding domain but not the domains used for protein-protein binding should produce a dominant negative allele. Similar approaches could be used for studies of other multiprotein complexes.
Charged-cluster to alanine scanning mutagenesis is a 'semi-random' approach to choosing which residues to mutate so as to improve the probability of affecting the function in a meaningful way. Alterations that cause the gene product to be unstable are uninteresting and do not allow one to explore function. Charged-cluster to alanine scanning mutagenesis is a method that optimizes the generation of mutant alleles with stable gene products.
X-ray crystallography has been used successfully to reveal the three-dimensional structure of many proteins. Generally speaking, clusters of charged residues tend to be located at the surface of a protein while hydrophobic and nonpolar residues are often buried in the core of the protein. The charged-cluster to alanine approach to constructing in vitro mutations uses this finding to propose the following. First, if one scans a protein sequence those regions containing clusters of charged residues have a reasonable probability of being positioned at the surface of the folded protein. Second, alterations of surface residues should produce fewer of the types of structural abnormality that often make mutant proteins the target of proteolytic degradation. Third, changes in these charged clusters are likely to alter surface residues that are often involved in protein-protein interactions. These predictions have proven to be correct frequently enough to make charged-cluster to alanine mutagenesis a valuable tool and widely accepted as a method for genetically dissecting the different protein-protein interactions of a particular multifunctional protein or protein component of a multiprotein complex. The types of interaction that can be detected by this approach are interactions with substrates, activating subunits, inhibitory subunits, targeting subunits, and other components of complexes. For proteins that function in several processes, the charged-cluster to alanine scanning approach is able to dissect these different functions because the clustered alterations may affect only one of the several functions.
One scans the sequence of a protein usually using an overlapping window of five residues looking for the presence of two or more charged residues in the window. All the charged residues are then changed to alanine using in vitro mutagenesis techniques. The mutant allele is then tested for phenotype in a null mutant strain as described above in the section on reverse genetics.
Bass, S.H., M.G. Mulkerrin, & J.A. Wells (1991) A systematic mutational analysis of hormone-binding determinants in the human growth hormone receptor. Proc. Natl Acad. Sci. USA 88: 4498-4502.
Bennett, W.F., N.F. Paoni, B.A. Keyt, D. Botstein, A.J. Jones, L. Presta, F.M. Wurm, & M.J. Zoller (1991) High resolution analysis of functional determinants on human tissue-type plasminogen activator. J. Biol. Chem. 266: 5191-5201.
Cunningham, B.C. & J.A. Wells (1989) High-resolution epitope mapping of hGH-receptor interactions by alanine-scanning mutagenesis. Science 244: 1081-1085.
Herskowitz, I. (1987) Functional inactivation of genes by dominant negative mutations.
Nature 329: 219-222. Ko, L.J. & C. Prives (1996) p53: Puzzle and paradigm. Genes Dev. 10: 1054-1074. Labbe, S. & D.J. Thiele (1999) Copper ion inducible and repressible promoter systems in yeast. Meth. Enzymol. 306: 145-153.
Genetic Techniques for Biological Research Corinne A. Michels Copyright © 2002 John Wiley & Sons, Ltd ISBNs: 0-471-89921-6 (Hardback); 0-470-84662-3 (Electronic)
Was this article helpful?