PlantGeneMapping Techniques

Transcription

1 PlantGeneMapping Techniques David Grant, Iowa State University, Ames, Iowa, USA Randy C Shoemaker, Iowa State University, Ames, Iowa, USA Plant genetic maps are graphic representations of the organization of chromosomes. Maps are developed using a wide range of classical and molecular techniques and population structures.. Introduction Introductory article. Classical Mapping Article Contents. Family Sizes and Structures (Inbreds, F 1 Backcross, RILs, etc.). DNA Markers. Mapping in Aneuploids and Polyploids. Comparative Maps Introduction The development of plant genetic maps has given us a picture of how genes are arranged in chromosomes. The maps are composed of markers, which may be genes controlling visible phenotypic traits (classical markers) or molecular markers whose phenotype is revealed using modern molecular biology techniques (e.g. DNA markers). The markers on the maps are like road signs telling geneticists where they are relative to other markers or to important genes. While in principle genetic maps could be developed using any population of plants that were segregating for a set of markers, the usual method is to minimize the effort needed to do the mapping by mating appropriately chosen parents to develop specialized progeny families. Genetic maps and the markers that define them have many applications. For example, easily assayed genetic markers that map close to genes controlling important plant traits such as yield, seed quality or disease resistance can be used to indirectly select for those traits in breeding programmes. This marker-facilitated selection can greatly speed the development of new and improved plant varieties. Comparison of the organization of genetic maps of different species using markers in common between the maps provides information on the history of these species. These kinds of data are important tools in understanding the evolution of the plant kingdom. Such comparisons can also allow a scientist to apply information developed in one species to solving problems in another. Classical Mapping The driving forces that maintain genetic diversity in a population are mutation and genetic recombination. A mutation is a change in the genetic code. This alteration can be a simple point mutation affecting a single nucleotide, or a larger, structural change such as a deletion or rearrangement. These mutations create genetic variation among individuals. All traits that differ among individuals or populations (tall versus short plant, purple versus white flower, etc.) are the result of changes in the genetic code. These observable traits (or phenotypes) are usually referred to as classical traits. Unless it is known otherwise, each classical trait is assumed to be controlled by a single gene. Each gene defines a genetic locus and each variant of a gene is termed an allele. In many species, progeny of a cross between two plants contain one chromosome from each parent and, because chromosomes are linear arrays of genes, parental loci are contributed to progeny in the exact order in which they were present in each parent. This means that each such pair of homologous chromosomes contains all of the genetic variation originally present in the parents. For any given locus along a chromosome, the progeny of a cross can be homozygous (both parents contributed identical alleles) or heterozygous (the parental alleles were different). When the progeny undergo further sexual reproduction, at one stage of meiosis the chromosomes align and genetic material can be exchanged between homologous chromosomes. This exchange is termed recombination or crossing-over and results in new combinations of alleles on the chromosomes of each gamete. These recombinant gametes can then produce plants that carry novel, nonparental combinations of alleles. The process of recombination is the foundation for the construction of genetic maps. Recombination is a random process; the chance of a recombination event occurring at any place on a chromosome can be assumed to be equal to that at any other place. Therefore, the closer together two genes are on a chromosome, the less likely they will be separated during recombination, and the farther apart they are, the more likely they will be separated during recombination events. The frequency of recombination is used to create genetic maps. For example, imagine that two plants homozygous for each of three genes, but with different alleles at each locus, are crossed to produce a segregating population. This could be diagrammed as: A B C a b c A B C a b c ENCYCLOPEDIA OF LIFE SCIENCES 2001, John Wiley & Sons, Ltd. 1

2 Table 1 Parental and nonparental allele combinations Parental allele combination How often observed Nonparental allele combination A-B or a-b 70% A-b or a-b 30% A-C or a-c 95% A-c or a-c 5% B-C or b-c 75% B-c or b-c 25% How often observed Consider these hypothetical results for all of the parental and nonparental allele combinations as shown in Table 1. The parental alleles for trait A and trait C are observed together 95% of the time, while those for traits A and B are observed together only 70% of the time. At the same time, the parental alleles for traits B and C are observed together 75% of the time. This tells us that the genes conferring traits A, B and C are aligned on the chromosome in the order A - C - B and that A and C are close together, while B is distant from A and C. The frequency with which genes or markers are observed to recombine defines the genetic distance between points on the map. Genetic maps are simply a graphical representation of the relative rates of recombination between adjacent loci along the chromosome. Before the onset of molecular biology, geneticists had developed genetic maps of many plant species using classical traits. The mutations that had naturally occurred in various genotypes resulted in variations in plant traits such as height, flower colour, seed morphology, etc. By making crosses and constructing appropriate families, geneticists were able to create genetic maps using only the segregation of the phenotypic characters. Classical genetic maps are exceptionally useful and informative because each marker on the map represents the position of a gene conferring a plant trait. But, by definition, the number of markers on a classical map is limited to the number of genes for which observable traits can be identified. Classical maps also are difficult to create as often only a limited number of observable genetic variations can segregate in the progeny of any family without causing reproductive or viability problems for individual plants. Thus, classical genetic maps always contain many fewer markers than the estimated number of genes in an organism. Family Sizes and Structures (Inbreds, F 1 Backcross, RILs, etc.) Several different family types and structures are commonly used in plant genetic mapping studies. Although some of these are similar to those used in any diploid organism, the fact that many plant species can be self-pollinated provides a powerful tool not readily available to animal geneticists. Inbred parent Inbred parent Self Self F 1 Inbred parent F 2 F 3 1 F 3 2 F 3 3 Self F 4 1 F 4 2 F 4 3 RIL 1 RIL 2 RIL 3 BC 1 Inbred parent BC 2 Inbred parent Figure 1 shows a generalized scheme for developing some of the most common family types used in plant genetics. Most commonly, highly inbred plant varieties are crossed to generate the F 1 generation. The F 1 plants of this cross are identical to each other and maximally heterozygous, that is they are heterozygous at every locus where the inbred parents differed. F 1 plants can then be crossed to either of the parental inbreds (backcross, BC) or crossed to themselves (self-pollinated or selfed). The selfed, or F 2, family produces what are considered standard Mendelian allele segregation ratios (3:1 or 1:2:1 depending on whether the locus shows dominant/recessive or codominant expression, respectively). Backcrossing can be used when self pollination is impossible owing to selfincompatibility. A family of plants derived from a backcross normally shows a 1:1 segregation ratio of alleles at each locus. Since Mendel s laws predict that each selfing generation will reduce the heterozygosity of the progeny by half, repeated selfing generations eventually results in new inbred lines, sometimes referred to as recombinant inbred BC introgression line Figure 1 Development of mapping populations. Crossing strategies for developing both RILs (recombinant inbred lines) and BCs (backcrosses) are shown. 2

3 lines or RILs. If a series of RILs is developed by single-seed descent from individual F 2 plants, the RILs effectively constitute an immortal mapping family that accurately reflects the genotypes of the original F 2 plants. Backcrossing can be used to transfer one gene or a small number of genes from one inbred parent (donor) to the genetic background of the other (recipient). This is accomplished by choosing a plant at each BC generation that carries the gene(s) being introgressed and backcrossing it to the donor inbred parent. Since each backcross generation reduces the heterozygosity by 50%, it is possible to effectively recover the donor s genotype with only the desired gene substitution. Each family type has strengths and weaknesses and thus the choice for a particular project is usually based on the types of questions being addressed (Table 2). The family sample size and number of markers chosen for a project are usually a compromise between the number of individuals needed for determining precise map positions and the realities of the cost and effort needed to genotype the individuals. In general, the standard error of the computed map position decreases as (1) the number of individuals in the family increases and (2) the number of loci analysed (i.e. the average spacing between the loci) decreases. Around individuals are normally sufficient for generating acceptably precise maps using codominant markers in an F 2 family. DNA Markers In the 1970s several researchers realized that instead of being limited to using whole-plant phenotypes as mappable traits, newly developed molecular biology technologies made it possible to directly measure DNA variation at any chromosomal location. Over the years a number of types of sequence variation have been discovered and many of these have been developed as genetic markers. Restriction fragment length polymorphism Restriction fragment length polymorphisms (RFLPs) were the first method developed to study and map DNA sequence variation. This technique relies on the observation that certain bacterial endonucleases have the ability to recognize and cleave DNA at specific short (4 8 base) sequences. Since these DNA sequences are otherwise unremarkable, they occur randomly many times in any genome. Restriction enzymes show extreme sensitivity to their recognition sequence. The change of even a single base will completely abolish recognition and cleavage of the DNA at that site. This means that whenever the DNA of two individuals differs at one of the bases in such a cleavage site, the restriction enzyme will cut one of the DNAs but not the other. An RFLP analysis is based on DNA hybridization and typically consists of several steps: 1. Generate a mapping family. 2. Isolate genomic DNA from each individual in the family. 3. Digest these DNAs with a restriction enzyme. 4. Separate the resultant fragments by size using gel electrophoresis. 5. Transfer the DNA fragments to a membrane. 6. Hybridize a labelled probe to the DNA fragments on the membrane. 7. Score each plant in the population for its alleles at the locus defined by the probe. Table 2 Family type Informativeness F 2 : Complete classification of linkage phase Highest informativeness (i a ) On average twice as informative as backcross F 2 : Codominant markers 1.0 i 0.5 i increases as p decreases b i not affected by linkage phase F 2 : Dominant markers 0.5 i 0 Markers in coupling: i increases as p decreases Markers in repulsion: i decreases as p decreases Backcross (or doubled haploid) i i not affected by linkage phase RILs 1.0 i i not affected by linkage phase i RIL 4 i BC when p 12.5 cm a i, relative information content per individual. b p, recombination fraction; proportional to interlocus distance. 3

4 RFLPs have another property that makes them very useful in genetic mapping. Unlike most phenotypic traits, which show a dominant/recessive form of expression, RFLPs are codominant, i.e. the presence of one RFLP allele does not affect the ability to simultaneously detect a different allele in the same individual. This means that the individuals in a population can be completely classified as to whether they are homozygous or heterozygous for the alleles contributed by the parents. Several other kinds of DNA sequence variation have been identified since RFLPs were first used. The polymerase chain reaction (PCR) process has allowed the development of these types of DNA sequence as genetic loci. Although these types of genetic markers utilize different methods to detect the sequence variation, they all reveal genetic loci that can be mapped exactly as wholeplant phenotypes or RFLPs are mapped. The choice of which marker type is used is usually based on several factors: the number of individuals to be analysed and the cost of each analysis; the type(s) of variation found; and the kinds of questions that will be studied in the experiment. Genetic analyses using these types of molecular markers are done in the same manner as those described above (family construction, genetics, analysis) but with different biochemical steps appropriate for detection of each kind of DNA variation. Simple sequence repeat Simple sequence repeats (SSRs), or microsatellites, are regions of the chromosome made up of tandem, head-totail repeats of a short DNA sequence, typically 2 6 bases in length. SSR loci are usually assayed by amplifying the repeats using PCR with flanking, locus-specific primers and determining the size(s) of the amplified products by gel electrophoresis. Alleles of SSR loci are defined by the size of the PCR product, which is an indirect measure of the number of repeats in the tandem array. Since the PCR primers are locus-specific, they will amplify any allele. Thus, both alleles in a heterozygous individual will be detected and SSRs behave as codominant markers. All plant genomes studied have SSRs located throughout their chromosomes. SSRs are not usually found within a gene and so there is little or no selection against DNA sequence changes in these regions. This means that many different alleles are often found at each locus. This makes SSRs very useful in mapping studies as it is likely that any two parents will differ from each other at many SSR loci. Random amplified polymorphic DNA Random amplified polymorphic DNA loci (RAPDs) are loci where PCR is done with a single, short oligonucleotide primer. This technique takes advantage of the facts that these short primers will hybridize at many sites in the chromosome and that sometimes the sites will be close enough together and orientated relative to each other such that PCR amplification of the region between them can occur. In the most common implementation of this technique, 10-base oligonucleotides are used as primers and 5 20 bands are seen for each primer. RAPDs are dominant markers as alleles are defined as the presence or absence of a band on an electrophoresis gel. Amplified fragment length polymorphism Amplified fragment length polymorphisms (AFLPs) are a way to detect single base changes that occur within 2 4 bases of a restriction enzyme cut site as well as those that are within a restriction enzyme cut site. Chromosomal DNA is cleaved with two restriction enzymes. Because restriction enzyme cut sites are essentially randomly distributed in the DNA, this will result in three kinds of DNA fragments: those with both ends produced by restriction enzyme 1, those with both ends produced by restriction enzyme 2, and those with one end produced by each restriction enzyme. Universal double-stranded DNA adapters, one complementary to the cohesive ends produced by each of the restriction enzymes, are ligated to the mix of fragments. Next, primers complementary to the two adapters and with 2 4 extra bases at the 3 end are used in a PCR reaction with the ligated fragment mix. The only restriction enzyme fragments that will amplify are those that (1) were produced by the action of both restriction enzymes and thus have different adapters ligated to their ends and (2) have the same 2 4 bases next to the restriction enzyme cut site that were added to the PCR primers. The PCR products are separated by gel electrophoresis. As with RAPDs, AFLPs are scored as dominant markers, with an allele defined by the presence or absence of a particular band. Single nucleotide polymorphism Single nucleotide polymorphisms (SNPs) are single nucleotide differences in the DNA sequences of different plants. As might be expected, they represent the most abundant form of DNA polymorphism. A large number of biochemical methods have been developed to detect SNPs, including direct sequencing and allele-specific amplification or hybridization. Since these detection methods differ in their cost, their ability to be automated and the time it takes to do the procedure, the choice of which detection method is used is almost always based on the number of individuals and loci to be analysed and the timeframe available for doing the analysis. 4

5 Mapping in Aneuploids and Polyploids Polyploidy, in which entire complements of chromosomes exist in more than one dosage (tetraploidy, hexaploidy, etc.), is relatively common in plants. Aneuploids, in which only a single chromosome or part of a chromosome is present in an increased (trisomy) or decreased (monosomy) dosage, are often generated by various experimental methods. Whereas polyploidy can add complexity to genetic mapping, aneuploidy can often be used to facilitate placement of markers on maps. Mapping in aneuploids Researchers have developed efficient methods of utilizing deficiency aneuploids (2n 2 1) for genetic mapping and for assigning markers to chromosomes or chromosome segments. Deficiency aneuploid genetic stocks are often arrayed for analysis so that it is easy to recognize when a gene or locus falls within the chromosome deficiency. Figure 2 shows an example of using aneuploid stocks to quickly map an RFLP locus to a particular chromosome. Probes are sequentially tested against DNA preparations of each aneuploid stock. When the chromosome containing the marker (probe) is absent, the fragment corresponding to it is also absent. By observing whether both parental alleles are present or only the nonaneuploid parent s allele is present, it is easy to determine whether the sequence corresponding to the probe came from the missing chromosome. Mapping stocks can also be developed where, instead of an entire chromosome being lost, only a piece of a chromosome is absent in each aneuploid line. These segmental deletion aneuploids can be used in a similar manner to determine relatively precise locations for a marker on a chromosome or linkage group. The smaller the piece of chromosome missing, the more precise the localization of the marker. This approach works best if the probe detects a polymorphism between homologous loci on homologous chromosomes and it is therefore easy to tell whether there is one or there are two hybridizing fragments. Deletion mapping is still possible in instances where the fragment pattern produced by the probe is monomorphic, although in this case this absence can be detected only by observing a 50% decrease in hybridization intensity of the probe. Similar strategies can be used with trisomics (2n 1 1). In this case the dosage increase of a chromosome or chromosome segment results in a 50% increase in intensity of the observed hybridization signal. Both of these approaches allow researchers to quickly determine to which linkage group or portion of a linkage group a marker belongs. Because these approaches do not depend on developing a mapping population, problems associated with abnormal pairing, sterility or chromosomal rearrangements that often complicate the genetics of aneuploids are avoided. Mapping in polyploids Mapping in polyploids requires many of the same steps as in classical mapping in diploids. First, an acceptable population needs to be identified. This is generally one that is large enough to critically distinguish standard segregation ratios, and one in which adequate polymorphism can be found. Then, the dosage of each polymorphic locus needs to be determined, i.e. single, double, or more. Segregation data can be used to determine the type and level of ploidy. Single-dose polymorphic fragments (Aaaa) are then mapped, followed by placement of the higher-dose fragments onto linkage groups. The mapping of singledose fragments results in the construction of the 2n number of linkage groups. Because analysis of single-dose fragments cannot allow the detection of linkages in repulsion, homologous linkage groups cannot be unambiguously identified by mapping single-dose fragments only and they must be resolved by mapping the multiple-dose markers. A consensus map of the homologous groups is then compiled by combining linkage information from each chromosome map for each homologue. Polyploids present several complicating factors when one is attempting to create genetic maps. Because of the extra number of dosages of each DNA segment, a relatively large number of genotypes is required for analysis. For P 1 P 2 F (a) (b) Figure 2 Mapping with monosomics. Panels (a) and (b) represent an array of DNA samples. P 1,P 2 diploid parents; F 1, progeny of P 1 P 2 ;1,_, 7, progeny of P 1 monosomic derivatives of P 2, each of which was missing a single chromosome. DNAs in each panel were hybridized with an RFLP (restriction fragment length polymorphism) probe with a unknown chromosomal location. The results indicate that the probe used in panel (a) came from chromosome 3 while the one used in panel (b) came from chromosome 6. 5

6 example, at a single locus in a diploid there are two allele possibilities (A and a). Among selfed progeny of this heterozygote there are three possible genotypes: AA, Aa and aa. A similar situation with an autotetraploid, which has the possibility of four alleles at a locus (A, a, A and a ), would yield nine possible genotypes among the progeny. Also, depending on the sizes of the fragments corresponding to each allele, fragments may comigrate on a gel, thus complicating interpretation. Finally, polyploids may exhibit aberrant pairing behaviour and resultant aberrant segregation patterns, thus complicating the determination of inheritance patterns and recombination frequencies. Comparative Maps Once genetic maps had been constructed for a wide range of plant species, researchers realized that the genes that mapped close together in one species were often similarly close together in other species. This observation of synteny or colinearity of chromosome organization between species was greatly expanded by molecular genetic mapping that allowed hundreds or thousands of common molecular markers to be mapped in different species. For example, corn molecular probes that were mapped onto the corn map could be mapped onto the rice genetic map as well. And rice probes that were mapped onto the rice map could be used to map in sorghum or barley, and so forth. This application of common markers in mapping studies has allowed a genome-wide comparison of the organization of genetic maps from different species. The most striking colinearity of genome composition can be found among the grasses: maize, rice, barley, oats, rye, sugarcane, etc. Synteny among these genomes is generally observed even though a tenfold or greater difference in the size of the genomes exists between some members of the grass family. Size differences among the genomes are generally due to different amounts of repetitive DNA elements. Most types of DNA markers are chosen from nonrepetitive regions of the genome. This means that the repetitive DNA manifests itself as large empty regions between genetic markers. Some degree of synteny has also been observed among species of dicots, although the extensive colinearity observed among the monocots is not generally observed among dicot species. The reason for this is unknown. Synteny between species has a number of useful consequences. Simply put, genetic information gained in one species can be transferred to another without being hindered by the experimental limitations of either species. For example, since rice has a small genome and very little repetitive DNA, there is not much physical distance between genes. Wheat, on the other hand, has a very large genome size and much repetitive DNA, so the physical distance between genes is greater. The difficulty of isolating a particular gene for further study is often proportional to the actual physical distance between a gene and a genetically linked marker. This means that if the genetic maps of two organisms are equivalently saturated with markers, gene isolation will be easier in the species with the smaller genome. The observed synteny between rice and maize means that it is often preferable for a maize researcher to isolate a gene of interest from the small rice genome and then use this as a tool to isolate the maize homologue rather than try to isolate the maize gene directly. Further Reading Bennetzen J and Freeling M (1993) Grasses as a single genetic system: genome composition, collinearity and compatibility. Trends in Genetics 9: Caetano-Anolles G and Gresshoff PM (eds) (1997) DNA Markers: Protocols, Applications, and Overviews. New York: Wiley-Liss. Silva J and Sorrells M (1996) Linkage analysis in polyploids using molecular markers. In: Jauhas PP (ed.) Methods of Genome Analysis in Plants, pp New York: CRC Press. 6