Textbook assignment: Chapter 16, 481 - 492, 496 - 511. Our textbook contains extensive material on alternative molecular models of genetic recombination on pages 492-496. These models are examined in detail in MCDB 3500 and will not be covered in this course beyond a brief description of the basic Holliday model.
Major concepts
Introduction: This lecture and the chapter it is based on examine genetic recombination and the organization of the genome at the level of fine structural details. Four interrelated topics are covered: 1)the use of tetrad analysis for detailed study of all four of the products of a single meiosis; 2) molecular models of recombination that attempt to account for all of the observed details from tetrad analysis; 3) the fine structural organization of genes as revealed by detailed analysis of complementation and recombination over very short map distances; and 4) events related to crossing over that occur in mitotic cells. Please note that we are skipping over some of the detailed analysis of recombinational mechanisms (textbook pages 492-496), which will be covered thoroughly in MCDB 3500.
Tetrad analysis: In the diploid multicellular organisms that have until now received most of our attention, we have only been able to sample single products of meiosis (randomly selected haploid genomes contained in egg, sperm, or pollen cells). For a more detailed analysis of meiosis, it is desirable to sample all four products from the same meiotic event. This is possible in a number of "simple" organisms in which all four products of meiosis form viable spores that can be recovered individually and analyzed genetically.
Ordered spores: The mold Neurospora crassa produces an ordered set of spores that provide highly detailed information about the products of meiosis. Haploid cells of two different mating types fuse to form a diploid cell (Fig.16.1) which immediately undergoes meiosis. The developing haploid spores are enclosed in a membranous structure called an ascus, which has an elongated shape and is narrow enough so that nuclei do not move past each other (figures 19.2 and 1.6g -- also please see the enlarged ascus shown on page 480). A single mitotic division occurs after meiosis is compleded such that of each of the four initial products of meiosis forms two identical spores. This permits the meiotic products to be sampled twice as neighboring spores. Because the spores remain physically ordered as they are generated by the meiotic (and mitotic) divisions, it is possible to determine precisely at which meiotic division genetic segregation is occurring (figures 16.3, 16.4, and 16.5). The products of first meiotic division are arranged 4:4 along the length of the ascus (figure 16.4), and the products of second division are arranged in a 2:2:2:2 relationship (figure 16.5).
Using centromeres as markers: Because of the ordered positions of the spores in the ascus, it is possible to use the centromere of each chromosome as a genetic marker. The centromeres from the homologous parental chromosomes separate at the first division and thus exhibit a 4:4 distribution in the ascus. If a crossover occurs between the centromere and a heterozygous genetic marker, the recombinant chromatids will now have an allele from one of the parental chromosomes associated with the centromere from the other. When the recombinant sister chromatid separates from the parental sister chromatid at the second meiotic division, the two products of that division will remain side by side in one half of the ascus, and will divide once mitotically to generate two parental and two recombinant spores. The same thing will also happen in the other half of the ascus, as shown in figure 16.5.
2:2:2:2 distributions: The formation of two parental and two recombinant spores in each half of the ascus will result in a 2:2:2:2 distribution of possible combinations of centromeres and genetic markers. Although the two centromeres cannot be distinguished in the progeny, the descendants of the two original centromeres remain in the two halves of the ascus in a 4:4 distribution. Thus if a crossover has occurred between non-sister chromatids in the tetrad, there will be a 2:2 distribution of the heterozygous genetic marker in each half of the ascus (figure 16.5). If there is no crossover between the marker and the centromere, each of the markers will stay with its original centromere to generate a 4:4 distribution (figure 16.4)
2:4:2 distributions" Because the orientation of the second meiotic division is random, the order of the pairs of parental and recombinant spores in each half of the ascus is also random, generating four possible sequences (figure 16.5). These four possibilities can be visualized as follows:
where + is wild type at the locus under study and a represents a mutation at that locus. Note that two of these four configurations are seen as a 2:4:2 distribution of ascospores. The critical requirement to verify that recombination has occurred is a 2:2 distribution in each half of the ascus.
Advantages of ordered ascospore analysis: Full analysis of ordered ascospores requires dissecting out the individual spores in order and keeping precise records of the cultures derived from individual spores obtained from specific positions in the ascus (figure 16.3). This is a tedious process, but it is essential for three types of study:
Map units in Neurospora: Since only half of the spores in each ascus are recombinant, a map unit is 100 times one half of the asci that contain recombinant spores, divided by the total number of asci. It is also possible to do three point crosses involving the centromere and two markers as well as conventional two point or three point crosses with genetic markers.
Unordered spores in budding yeast: The the budding yeast, Saccharomyces cerevissiae, undergoes meiosis to produce an ascus containing four haploid spores that represent all four products of meiotic division. However, in this case, the spores are mixed together in the ascus in such a way that the products of the two meiotic divisions cannot be distinguished based on their positione. However, before examining the analysis of unordered spores, it is desirable to clarify the different types of yeast used in laboratory research.
Types of yeast: Two very different types of yeast are widely used in genetic studies. Although they have retained certain shared properties, such as alcoholic fermentation, they have been evolutionarily separated from each other for as long as either of them has been separated from our own ancestors. They are budding yeast, Saccharomyces cerevisiae, which is widely used as a baking and brewing yeast, and fission yeast, Schizosaccharomyces pombe, which is used in the brewing of an African beer. Because references to "yeast" are often made rather casually as if all yeasts were similar, it is important to look for species identification and to be aware that the two types are very different in many aspects of their life cycles.
Yeast life cycles: In brief summary, S. cerevisiae, multiplies by budding, grows preferentially in the diploid state, and is triggered by starvation to undergo meiosis and form spores. When conditions improve, the spores germinate and become haploid vegetative cells, which normally undergo conjugation at the first opportunity, thus regenerating the diploid state. However, they can also be maintained as haploids. S. pombe, on the other hand, multiplies by a fission process that is very similar to mitosis in higher organisms, lives preferentially as a haploid, and is triggered by starvation to undergo conjugation to form a diploid, which then usually immediately undergoes meiosis and forms haploid spores. The spores germinate when conditions improve and proliferate as haploid vegetative cells. Some strains can be maintained in the diploid state through the use of selective media. S. pombe produces asci containing ordered spores, much like those described above for Neurospora, except for absence of the final mitotic division. However, they are small and difficult to work with, and thus seldom studied.
Unordered spores: When a diploid S. cerevisiae (budding yeast) cell undergoes meiosis, all four of the resultant haploid nuclei form viable spores contained in a membranous ascus whose structure is loose enough so they are not held in any particular order (figure 16.6). The individual spores can be dissected out and cultured separately to give rise to haploid strains representing each of the four products of meiosis.
Effects of recombination: The first meiotic division separates centromeres of homologous chromosomes and the second division separates sister chromatids. Thus, if there is no crossing over in a dihybrid test cross, two of the spores will be of one parental genotype and the other two of the other parental genotype (figure 16.7a). This is referred to as a parental ditype (PD) because it contains two types of spores that are both the same as the parental genotypes. If there is a single crossover event, there will be one spore with each of the parental genotypes and the other two will exhibit reciprical recombination (figure 16.7b). This is referred to as a tetratype (TT) because each ascus contains four different types of spores. It is also possible to obtain a non-parental ditype (NPD) if a second crossover involving the other two strands of the meiotic tetrad occurs (figure 16.7c). This is a relatively less frequent event, and is scored as a double crossover. A similar system of nomenclature can also be used to identify the possible products of independent assortment. However, in cases where independent assortment occurs, all four possible combinations of elleles would be expected to occur with equal frequency.
Map units in yeast: A map unit is defined as a recombination
frequency that yields 1% recombinant progeny. Because only half
of the spores in a TT ascus are recombinant, whereas all spores
in an NPD ascus are recombinant, map units are scored as one-half
of the number of TT asci plus the number of NPD asci, divided
by the total number of asci examined and multiplied by 100 to
convert to percentage.
Gene conversion: In a small percentage of cases, an aberrant pattern of spore distribution is observed in Neurosopora and other similar ascomycetes with ordered tetrads (figure 16.8). In most cases, there is an increase in one allele and a decrease in the other. This process is called gene conversion. Through the use of flanking markers, it has been shown that crossing over is involved in about half of the cases of gene conversion. Sevaral relatively complex molecular models have been developed in an attempt to explain how gene conversion can be achieved. These will be examined in detail in MCDB 3500. We will only briefly examine the Holliday model, which was the first to be proposed, and which forms the basis for the more complex variations that have been suggested since.
Holliday model of recombination: If crossing over were a simple breaking and rejoining of two parallel DNA double helices at the same location on each, no gene conversion would be expected. The gene conversion phenomenon suggests that crossing over involves transient formation of heteroduplexes (double helices with one strand from each of the original DNA molecules), followed by mismatch repair. The Holliday model proposes single strand breaks at the same place in the same strands of two DNA double helices that are physically adjacent to each other (shown in figure 16.9a with the helices unwound for clarity). The next step is heteroduplex formation between complementary strands from the two original DNA molecules, together with ligation of the gaps. This binds the two DNA helices together through an X-shaped junction that has come to be known as a "Holliday junction" (figure 16.9b). The junction then migrates, elongating the heteroduplex region (figure 16.9c). This is followed either by a "north-south" cut, which breaks the two remaining strands of the original DNA double helices and results in an exchange of the ends of the DNA molecules (crossing over), or else by an "east-west" cut, which severs the Holliday junction and restores the original parental pattern except for the residual heteroduplex region (figure 16.9d). In either case, if the heteroduplex region spans a point mutation, mismatch repair will replace one of the mis-paired bases. This can result in gene conversion, either with recombination in the case of the north-south cut, or in the absence of recombination in the case of the east-west cut (figure 16.10a). More complex alternative models, which we will not examine in this course, are shown in figures 16.11 and 16.12..
Complementation analysis: We have already studied the basic principles of complementation in several earlier lectures. In lecture 9 (textbook pages 163 - 172), we saw that genomes with mutations blocking different steps in a metabolic pathway had the potential to complement each other, such that between the two of them, they could code for all of the enzymes needed for successful completion of the metabolic pathway. In lecture 10, we saw how the partial genome carried on an F factor in a merozygote could complement a genetic defect on the bacterial chromosome to restore function. We have also seen a similar principle in complementary gene action (lecture 26, pages 395-398), where a dominant allele from each of two different loci must be present to obtain the dominant phenotype in the F2 population, yielding a 9:7 phenotypic distribution. Our textbook has waited until now (pages 498-500) to formally introduce the concept of complementation analysis and the cis/trans test. The notes that follow integrate this material into the study of fine structure mapping in a somewhat different sequence than the textbook, but both cover the same material.
Bacteriophage genetics: The life cycles of virulent and lysogenic bacteriophage were described earlier in the semester (textbook pages 202 - 207) and the molecular mechanisms responsible for switching between the two states have also been described (pages 233 - 237). The size and shape of plaques and the strains of bacteria that will support lytic cycle multiplication of particular types of bacteriophage can be used as markers for genetic studies in bacteriophage. Through use of a high multiplicity of infection (far more bacteirophage than there are bacteria), it is possible to infect each bacterial cell with several bacteriophage particles. Interaction of the bacteriophage genomes within the individual host cells makes it possible to undertake a variety of genetic studies, including recombination and complementation.
Fine structure mapping: After the discovery by several investigators in the late 1940's that recombination could occur when two different strains of bacteriophage invaded the same bacterial cell, Seymour Benzer undertook a series of studies designed to push recombination over short genetic distances to its limit. For these studies, he employed a mutant strain of bacteriophage T4 designated rII, which could form plaques on E. coli strain B, but not on strain K12. His goal was to look for recombination within a single gene, taking advantage of the ability to detect as few as one recombinant phage in 100 million.
Complementation: When certain pairs of rII strains were coinfected into strain K12 bacteria, they were found to form plaques readily without the need for rare recombinational events. Benzer proposed that each strain was providing a function that the other lacked, such that by working together they could replicate under conditions where neither could replicate alone. He called this phenomenon complementation. His research provided evidence for the existence of two separate genetic functions in the rII region, which came to be known as rIIA and rIIB. Subsequent studies have shown that rIIA and rIIB are in fact two separate, but closely-linked genetic loci, each coding for a different protein product. A functional copy of each coding sequence is needed to support full function. There are also some special cases in which two defects in the same protein can complement each other by interacting to form a functional dimer. These relatively rare cases will be discussed later in this lecture.
Complementation in other organisms: The basic concept of complementation or complementary gene action has been known for a long time, and occurs in highly diverse organisms. Mutations are said to be complementary if the genomes (or partial genomes) that contain them support normal function when both are introduced into the same cell. The two separate genomes can be introduced in many different ways, including the maternal and paternal components of a heterozygous diploid genome, or a full haploid genome plus a partial genome in a bacterial merozygote, or two different types of nuclei in a common cytoplasm in heterokaryons of Neurospora, or two different bacteriophage strains infecting the same bacterial cell or any other combination that allows the undamaged part of each genome to compensate for the damage that has occurred in the other one. The critical requirement is that the two mutations must not destroy precisely the same genetic function and that when combined in a diploid or diploid-like situation, each of the damaged genomes must be able to compensate for the losses suffered by the other.
Restricted to one phenotypic variable: If the definitions given above are taken literally, any two recessive mutations in a trans- relationship can be called complementary. However, in practice, the term "complementation" is usually only used to describe the interaction between mutations that influence the same general phenotypic variable. One example of this is the complementary gene action we saw earlier in the semester, in which knocking out either of two enzymes that acted sequentially in the synthesis of purple pigment in sweet peas resulted in a white flower. However, the F1 hybrid between these two mutant strains produced purple flowers, because each of the mutant strains introduced a wild type allele of the gene coding for the other enzyme. This is the essence of complementation: each mutant strain must be deficient in a sufficiently different function so that the two together can achieve full function without the need for genetic recombination to occur.
Limitations on complementation tests: There are several additional limitations to complementation tests. The mutant phenotypes that are being studied need to be recessive to wild type. A dominant gene will exert its effects irrespective of the overall genetic background it is in. In addition, the extent of recombination must be substantially less than the level of complementation that is observed so that it will be clear that the restoration of function occurred because the two mutations destroyed two different units of function, and not because recombination created a normal gene and a doubly-defective gene. Also, the gene products that are involved must be diffusible, such that they can move enough within the cytoplasm to interact in a way that will allow normal function. Thus, strictly cis-acting mutations (those whose effect is on the function of the DNA molecule that they are part of), such as mutations in promoters or operators, would not be expected to be capable of complementation.
Cis-trans test: As used here, the term cis refers to genetic changes that are on the same DNA molecule in simple organisms or in the same haploid genome in cases where there are mutiple chromosomes. The term trans refers to genetic changes that are carried by different genomes that have been introduced into the same cell (in any of the possible ways discussed above). The cis-trans test asks whether two mutations that occur in different genomes have altered the same unit of fuction (usually the coding sequence for a single protein). It is expected that when the two mutations are put together in the same genome (cis) that they will not support normal function, even if they occur in two separate functional units. Such a test is straightforward in haploid organisms. However, to see cis lack of function in a diploid organism, it is necessary to make the organism homozygous for the cis genome. Otherwise there will be a wild type genome opposite it that will support normal function. In most cases, the cis part of the cis-trans is an implied control that could be run, rather than an important part of the actual experimental study.
The trans test, where complementation is expected to occur if two different coding sequences are involved, requires the presence of two complete genomes, or at least two copies of the portions of the genomes where the mutations occur, so that a good copy of each coding unit will be present (in cases where the mutations involve two different coding units). Thus, in haploid organisms, it becomes necessary to find a way to introduce a second genome (merodiploids in bacteria, optional diploid phase in yeast, heterokaryons in Neurospora, mixed infection of bacteriophage into bacterial cells, etc.).
Complementation in diploid organisms: As discussed above, use of the term "complementation" is usually reserved for cases in which the two mutations whose interactions are being studied have a similar phenotypic effect. Numerous examples of such effects can be cited, including the ebony and black mutations in Drosophila, the A and B color loci for corn kernels, or the C and P color loci in sweet peas. In every case, it is necessary to have at least one wild-type allele at both genetic loci in order to obtain a wild-type phenotype (normal body color in Drosophila, purple kernels of corn, purple flowers on the sweet peas). Such systems typically yield a 9:7 phenotypic ratio in the F2 (figures 13.16).
Cistron: Returning now to the studies of genetic fine structure undertaken by Seymour Benzer, he introduced the term cistron to describe the genetic units that were identified in his cis-trans tests as the basic units of complementation in bacteriophage T4. When the original work was done in the early 1950's, genes could only be identified as units of heredity that were known to have a 1:1 relationship to enzymes and other protein products in some cases. (At that time, the genetic code was not yet known and practical methods for sequencing DNA had not been developed). The cis-trans test allowed the rII "gene" in the bacteriophage T4 genome to be subdivided by complementation testing into two subunits, rIIA and rIIB, which Benzer described as separate cistrons.
One cistron-one polypeptide: With the rapid advances in molecular biology that were occurring at the time, it soon became evident that a cistron was, in fact, the coding unit for one polypeptide chain (as opposed to one enzyme, which could be composed of several subunits). For many years, cistron was the preferred term for identifying the coding sequence for a single protein chain. However, as older vague usages have slowly faded away, the word "gene" has gradually regained popularity as the appropriate word to describe a single coding sequence. Nevertheless, use of the term "cistron" has persisted in certain cases, such as the description of the transcripts from operons as "polycistronic" (meaning that they contain the coding sequences for several different proteins in a single large RNA molecule).
Intracistronic complementation: In certain rare cases, complementation can occur within a single coding sequence. This is commonly referred to as intracistronic complementation. The simplest way to visualize this is to think of subunits of a dimeric enzyme. In certain cases identical subunits carrying the same defect may not be able to form a dimer, whereas those that have complementary defects can. Thus, for example, a hydrophobic interaction between two different regions in the monomer might be lost because a charged amino acid has replaced a hydrophobic amino acid. However, if a second mutation places the opposite charge at the interacting site, the positive and negative might interact and allow a stable dimer to be formed, even though neither type of monomer could interact with itself.
Beta-galactosidase as an example: A real life example cited in a previous textbook involves two different mutations in the beta-galactosidase (lacZ) gene in E. coli that can complement each other. In this case, the functional enzyme is a homotetramer. There are two known mutations that each individually prevent formation of an active tetramer. However, the mutations involve different parts of the protein monomer, such that a mixture of the two types of proteins can still form a heterotetramer (containing two copies of each of the mutant proteins) that is catalytically active.
Deletion mapping: Benzer accumulated some 20,000 rII mutations, which were distributed about equally between rIIA and rIIB. Detailed recombinaitonal analysis was greatly facilitated by the use of a number of deletion mutations. Since each deletion mutant had lost a chromosomal region, it could not restore function by recombining with any point mutations that fell within the region covered by the deletion (figure 16.21). This allowed point mutations to be localized to restricted areas within one of the cistrons before precise recombinational mapping with nearby loci was undertaken.
Linear arrays of mutations: Benzer's work demonstrated that the smallest functional genetic unit, the cistron, was, in fact, a linear array of genetic information that was entirely consistent with the growing belief that the sequence of nucleotides in a linear DNA molecule was the carrier of genetic information (figure 16.22). He identified the smallest unit of recombination that he could detect a muton, which is now known to correspond to an individual nucleotide. Other concepts also emerged, such as mutational hot spots, which were discussed earlier in the semester (boxed example 5.2, page 134).
Fine structure mapping in other organisms: Benzer was able to do detailed fine-structure mapping because the bacteriophage assay system allowed unusually large numbers of progeny to be analyzed selectively for rare recombinational events occurring at very short map distances. A few cases of fine structure mapping have also been done in other species such as Drosophila (boxed example 16.5). However, such studies are so labor-intensive that not many have been undertaken. In addition genomic sequencing has now provided an even better method for precise analysis of individual genetic loci and their varient alleles.
Sister chromatid exchange: Although we commonly think of crossing over as a meiotic phenomenon, there are certain situations where it also occurs in mitotic cells. The most common is called sister chromatid exchange. When special labeling techniques are used that permit the two sister chromatids to be distinguished from each other, they are seen to have undergone multiple crossing over events in ordinary mitotic chromosomes (figure 16.24). These studies are usually done in cultured cellsr. The cells are allowed to complete one round of DNA synthesis with 5-bromodeoxyuridine (BrdU) added to the culture medium. The BrdU is a thymidine analog, which is converted to a triphosphate and incorporated into the newly synthesized strand of DNA. The cells are then allowed to go through one more round of DNA synthesis in normal medium. During that second round of semiconservative DNA synthesis, the labeled strand is separated from the unlabeled strand and each synthesizes an unlabeled complementary strand. A special staining technique gives chromatids that contain BrdU a different color than those that do not. Through use of this technique, it can be seen that a typical pair of sister chromatids have undergone recombination at several different locations (figure 16.24). Because the sister chromatids are genetically identical, no genetic consequences of such exchanges are seen.
Mitotic Crossing over: Crossing over between homologous chromosomes does not normally occur in mitotic cells. However, in certain rare cases, it has been observed. The best known example is twin-spotting in Drosophila (figure 16.25). This occurs when a fly is a trans heterozygote for two different recessive markers on the same arm of a chromosome (figure 16.26). If there is temporary homologous pairing and a crossover occurs between the centromere and the two markers, and if the recombinant chromatids then segregate to the poles in the right pattern, one of the daughter cells will become homozygous for one of the mutations and the other will become homozygous for the other mutation. If this phenomenon occurs early enough in development, the descendants of the two homozygous cells will each multiply to generate a patch of tissue with one of the recessive phenotypes. If two body surface markers are involved, the results will be adjacent spots containing the two recessive phenotypes on an otherwise wild-type background. This is illustrated in figure 16.25, but the two phenotypes do not show well in that figure.