Top texture: © Laguna Design / Science Source;

CHAPTER 13: Homologous and Site-Specific Recombination

Edited by Hannah L. Klein and Samantha Hoot

Chapter Opener: Laguna Design/Getty Images.

13.1 Introduction

Homologous recombination is an essential cellular process required for generating genetic diversity, ensuring proper chromosome segregation, and repairing certain types of DNA damage. Evolution could not happen efficiently without genetic recombination. If material could not be exchanged between homologous chromosomes, the content of each individual chromosome would be irretrievably fixed in its particular alleles, only changing in the event of a mutation. In the event of a mutation, it would then not be possible to separate favorable from unfavorable changes. The length of the target for mutation damage would effectively be increased from the gene to the chromosome. Ultimately, a chromosome would accumulate so many deleterious mutations that it would fail to function.

By shuffling the genes, recombination allows favorable and unfavorable mutations to be separated and tested as individual units in new assortments. It provides a means of escape and spreading for favorable alleles, as well as a means to eliminate an unfavorable allele without bringing down all the other genes with which this allele is associated. This is the basis for natural selection.

In addition to its role in genetic diversity, homologous recombination is also required in mitosis for repair of lesions at replication forks and for restarting replication that has stalled at these lesions. The importance of mitotic recombination events is highlighted by examples of human diseases that result from defects in recombination repair of DNA damage where altered activity of homologous recombination proteins is seen in some types of cancers. Homologous recombination is also essential for a process known as antigenic switching, which allows disease-causing parasites called trypanosomes to evade the human immune system.

Recombination occurs between precisely corresponding sequences so that not a single base pair is added to or lost from the recombinant chromosomes. Three types of recombination involve the physical exchange of material between duplex DNAs:

  • Recombination involving a reaction between homologous sequences of DNA is called generalized or homologous recombination. In eukaryotes, it occurs at meiosis, usually both in males (during spermatogenesis) and females (during oogenesis). Recombination happens at the “four-strand” stage of meiosis and involves only two nonsister strands of the four strands (see the chapter titled Genes Are DNA and Encode RNAs and Polypeptides).

  • Another type of event sponsors recombination between specific pairs of sequences. This was first characterized in prokaryotes where specialized recombination, also known as site-specific recombination, is responsible for the integration of phage genomes into the bacterial chromosome. The recombination event involves specific sequences of the phage DNA and the bacterial DNA, which include a short stretch of homology. The enzymes involved in this event act in an intermolecular reaction only on the particular pair of target sequences. Some related intramolecular reactions are responsible during bacterial division for regenerating two monomeric circular chromosomes when a dimer has been generated by generalized recombination. This latter class also includes recombination events that invert specific regions of the bacterial chromosome.

  • In special circumstances, gene rearrangement is used to control expression. Rearrangement may create new genes, which are needed for expression in particular circumstances, as in the case of the immunoglobulins. This is an example of somatic recombination, which is discussed in the chapter titled Somatic Recombination and Hypermutation in the Immune System. Recombination events also may be responsible for switching expression from one preexisting gene to another, as in the example of yeast mating type, where the sequence at an active locus can be replaced by a sequence from a silent locus. Rearrangements are also required to control expression of surface antigens in trypanosomes, in which silent alleles of surface antigen genes are duplicated into active expression sites. Some of these types of rearrangement share mechanistic similarities with transposition; in fact, they can be viewed as specially directed cases of transposition.

Let us consider the nature and consequences of the generalized and specialized recombination reactions. FIGURE 13.1 demonstrates that generalized recombination occurs between two homologous DNA duplexes and can occur at any point along their length. The crossover is the point at which each becomes joined to the other. The overall organization of the DNA does not change; the products have the same structure as the parents, and both parents and products are homologous.

FIGURE 13.1 No crossing over between the (a) and (b) genes gives rise to only nonrecombinant gametes. Crossing over between the A and B genes gives rise to the recombinant gametes Ab and aB and the nonrecombinant gametes AB and ab.

Specialized recombination occurs only between specific sites. The results depend on the locations of the two recombining sites. FIGURE 13.2 shows that an intermolecular recombination between a circular DNA and a linear DNA results in the insertion of the circular DNA into the linear DNA. Specialized recombination is often used to make changes such as this in the organization of DNA. The change in organization is a consequence of the locations of the recombining sites. We have a large amount of information about the enzymes that undertake specialized recombination, which are related to the topoisomerases that act to change the supercoiling of DNA in space (see the chapter titled Genes Are DNA and Encode RNAs and Polypeptides).

FIGURE 13.2 Site-specific recombination occurs between the circular and linear DNAs at the boxed region (a). Integration results in an insertion of the A and B sequences between the X and Y sequences (b). The reaction is promoted by integrase enzymes. Reversal of the reaction results in a precise excision of the A and B sequences.

Data from B. Alberts, et al. Molecular Biology of the Cell, Fourth edition. Garland Science, 2002.

13.2 Homologous Recombination Occurs Between Synapsed Chromosomes in Meiosis

Homologous recombination is a reaction between two duplexes of DNA. Its critical feature is that the enzymes responsible can use any pair of homologous sequences as substrates (although some types of sequences may be favored over others). In fact, in most species a crossover event is required for accurate separation of homologs at the first meiotic division; thus there is usually at least one crossover per homologous chromosome pair. The frequency of recombination is not constant throughout the genome, but is influenced by both global and local effects, and both recombination hotspots and coldspots can be identified. The short region of homology between the mammalian X and Y chromosomes (the “pseudoautosomal” region) is the only available region of crossover between the X and Y, and thus is subject to 10 times higher rates of crossover per length than the average for the rest of the genome. The phenomenon of crossover interference refers to the tendency (but not a rule) of a crossover event to reduce the likelihood of another crossover nearby. Crossovers are also rare in or near centromeres, are uncommon near telomeres in some species, and are generally suppressed in heterochromatic regions. Certain histone modifications can also influence recombination positively or negatively. The overall frequency of recombination may be different in oocytes and in sperm; recombination occurs twice as frequently in female as in male humans.

Recombination occurs during the protracted prophase of meiosis. FIGURE 13.3 shows the visible progress of chromosomes through the five stages of meiotic prophase. Studies in yeast have shown that all of the molecular events of homologous recombination are finished by late pachytene.

FIGURE 13.3 Recombination occurs during the first meiotic prophase. The stages of prophase are defined by the appearance of the chromosomes, each of which consists of two replicas (sister chromatids), although the duplicated state becomes visible only at the end.

The beginning of meiosis is marked by the point at which individual chromosomes become visible. Each of these chromosomes has replicated previously and consists of two sister chromatids, each of which contains a duplex DNA. The homologous chromosomes approach one another and begin to pair in one or more regions, forming bivalents. Pairing extends until the entire length of each chromosome is apposed with its homolog. The process is called synapsis or chromosome pairing. When the process is completed, the chromosomes are laterally associated in the form of a synaptonemal complex, which has a characteristic structure in each species, although there is wide variation in the details between species.

Recombination between chromosomes involves a physical exchange of parts (achieved through a double-strand break on one chromatid to initiate recombination), formation of a joint molecule between the chromatids, and resolution to break the joint and form intact chromatids that have new genetic information. When the chromosomes begin to separate, they can be seen to be held together at discrete sites called chiasmata. The number and distribution of chiasmata parallel the features of genetic crossing over. Traditional analysis holds that a chiasma represents the crossing-over event. The chiasmata remain visible when the chromosomes condense and all four chromatids become evident.

What is the molecular basis for these events? Each sister chromatid contains a single DNA duplex, so each bivalent contains four duplex molecules of DNA. Recombination requires a mechanism that allows the duplex DNA of one sister chromatid to interact with the duplex DNA of a sister chromatid from the other chromosome. This reaction must be able to occur between any pair of corresponding sequences in the two molecules in a highly specific manner so that the material can be exchanged with precision at the level of the individual base pair.

We know of only one mechanism for nucleic acids to recognize one another on the basis of sequence: complementarity between single strands. If (at least) one strand displaces the corresponding strand in the other duplex, the two duplex molecules will be specifically connected at corresponding sequences. If the strand exchange is extended, a more extensive connection can occur between the duplexes.

13.3 Double-Strand Breaks Initiate Recombination

Genetic exchange is initiated by a double-strand break (DSB). The double-strand break repair (DSBR) model is illustrated in FIGURE 13.4. Recombination is initiated by an endonuclease that cleaves one of the partner DNA duplexes, the “recipient.” In meiosis this is performed by the Spo11 protein, which is related to DNA topoisomerases (FIGURE 13.5). DNA topoisomerases are enzymes that catalyze changes in the topology of DNA by transiently breaking one or both strands of DNA, passing the unbroken strand(s) through the gap, and then resealing the gap. The ends that are generated by the break are never free, but instead are manipulated exclusively within the confines of the enzyme—in fact, they are covalently linked to the enzyme. Spo11 undergoes a similar covalent attachment when it forms DSBs during meiosis.

FIGURE 13.4 The double-strand break repair (DSBR) model of homologous recombination. Recombination is initiated by a double-strand break. Following nuclease degradation of the ends, called DNA resection, single-strand tails with 3′–OH ends are formed. Strand invasion by one end into homologous sequences forms a D-loop. Extension of the 3′–OH end by DNA synthesis enlarges the D-loop. Once the displaced loop can pair with the other side of the break, the second double-strand break end is captured. DNA synthesis to complete the break repair, followed by ligation, results in the formation of two Holliday junctions. Resolution at the blue arrowheads results in a noncrossover product. Resolution of one Holliday junction at the blue arrowheads and the other Holliday junction at the red arrowheads results in a crossover product.

FIGURE 13.5 Spo11 is covalently joined to the 5′ ends of double-strand breaks.

In mitotic cells DSBs form spontaneously as a result of DNA damage or through the action of specific processes that are programmed to form breaks, such as V(D)J recombination or mating-type switching in yeast. Exonuclease(s), which can work in concert with a DNA helicase, degrade one strand on either side of the break, generating 3′–single-stranded termini; this process is known as 5′-end resection. In earlier models, this included the formation of a significant gap at the site of the DSB, but more recent data suggest that large gaps are not usually present in vivo. One of the free 3′ ends then invades a homologous region in the other (“donor”) duplex. This is called single-strand invasion. The formation of heteroduplex DNA generates a D-loop (displacement loop), in which one strand of the donor duplex is displaced. The point at which an individual strand of DNA crosses from one duplex to the other is called the recombinant joint. An important feature of a recombinant joint is its ability to move along the duplex. Such mobility is called branch migration. The D-loop is extended by repair DNA synthesis, using the free 3′ end as a primer to generate double-stranded DNA. FIGURE 13.6 illustrates the migration of a single strand in a duplex. The branching point can migrate in either direction as one strand is displaced by the other.

FIGURE 13.6 Branch migration can occur in either direction when an unpaired single strand displaces a paired strand.

Branch migration is important for both theoretical and practical reasons. As a matter of principle, it confers a dynamic property on recombining structures. As a practical feature, its existence means that the point of branching cannot be established by examining a molecule in vitro (because the branch may have migrated since the molecule was isolated).

Branch migration can allow the point of crossover in the recombination intermediate to move in either direction. The rate of branch migration is uncertain, but, as seen in vitro, it is probably inadequate to support the formation of extensive regions of heteroduplex DNA in natural conditions. Any extensive branch migration in vivo must therefore be catalyzed by a recombination enzyme.

The second resected single strand subsequently anneals to the donor, forming a second single-end invasion (SEI) and converting the D-loop into two crossed strands or recombinant joints called Holliday junctions. Overall, the resected region has been repaired by two individual rounds of single-strand DNA synthesis. The joints must be resolved by cutting.

If both joints are resolved in the same way, the original noncrossover molecules will be released, each with a region of altered genetic information that is a footprint of the exchange event. If the two joints are resolved in opposite ways, a genetic crossover is produced.

The involvement of DSBs at first seems surprising. Once a break has been made right across a DNA molecule, there is no going back. In the DSBR model, the initial cleavage is immediately followed by loss of information. Any error in retrieving the information could be fatal. However, the very ability to retrieve lost information by resynthesizing it from another duplex provides a major safety net for the cell.

The joint molecule formed by strand exchange must be resolved into two separate duplex molecules. Resolution requires a further pair of nicks. We can most easily visualize the outcome by viewing the joint molecule in one plane as a Holliday junction. This is illustrated in the bottom half of Figure 13.4, which represents the resolution reaction. The outcome of the reaction depends on which pair of strands is nicked.

If the nicks are made in the pair of strands that was not originally nicked (the pair that did not initiate the strand exchange), all four of the original strands have been nicked. This releases crossover recombinant DNA molecules. The duplex of one DNA parent is covalently linked to the duplex of the other DNA parent via a stretch of heteroduplex DNA.

If the same two strands involved in the original nicking are nicked again, the other two strands remain intact. The nicking releases the original parental duplexes, which remain intact, with the exception that each has a residuum of the event in the form of a length of heteroduplex DNA. These are noncrossover products that nonetheless contain sequence from the donor DNA duplex, and as such are considered recombinant. Although this description suggests that the outcome is random, newer evidence suggests that numerous factors influence crossover versus noncrossover outcomes, and the distinction is established as early as the stage of D-loop formation.

What is the minimum length of the region required to establish the connection between the recombining duplexes? Experiments in which short homologous sequences carried by plasmids or phages are introduced into bacteria suggest that the rate of recombination is substantially reduced if the homologous region is less than 75 bp. This distance is appreciably longer than the 10 bp or so required for association between complementary single-stranded regions, which suggests that recombination imposes demands beyond annealing of complements as such.

13.4 Gene Conversion Accounts for Interallelic Recombination

The involvement of heteroduplex DNA explains the characteristics of recombination between alleles; indeed, allelic recombination provided the impetus for the development of a recombination model that invoked heteroduplex DNA as an intermediate. When recombination between alleles was discovered, the natural assumption was that it takes place by the same mechanism of reciprocal recombination that applies to more distant loci. That is to say, both events are initiated in the same manner: A DSB repair event can occur within a locus to generate a reciprocal pair of recombinant chromosomes. In the close quarters of a single gene, however, formation and repair of heteroduplex DNA itself is responsible for the gene-conversion event.

Individual recombination events can be studied in the ascomycete fungi, because the products of a single meiosis are held together in a large cell called the ascus (or, less commonly, the tetrad). Even better is that in some fungi the four haploid nuclei produced by meiosis are arranged in a linear order. (Actually, a mitotic division occurs after the production of these four nuclei, giving a linear series of eight haploid nuclei.) FIGURE 13.7 shows that each of these nuclei effectively represents the genetic character of one of the eight strands of the four chromosomes produced by meiosis.

FIGURE 13.7 Spore formation in ascomycetes allows determination of the genetic constitution of each of the DNA strands involved in meiosis.

Meiosis in a heterozygous diploid should generate four copies of each allele in these fungi. This is seen in the majority of spores. Some spores, however, have abnormal ratios. These spores are explained by the formation and correction of heteroduplex DNA in the region in which the alleles differ. Figure 13.7 illustrates a recombination event in which a length of hybrid DNA occurs on one of the four meiotic chromosomes, a possible outcome of recombination initiated by a DSB.

Suppose that two alleles differ by a single point mutation. When a strand exchange occurs to generate heteroduplex DNA, the two strands of the heteroduplex will be mispaired at the site of mutation. Thus, each strand of DNA carries different genetic information. If no change is made in the sequence, the strands separate at the ensuing replication, each giving rise to a duplex that perpetuates its information. This event is called postmeiotic segregation, because it reflects the separation of DNA strands after meiosis. Its importance is that it demonstrates directly the existence of heteroduplex DNA in recombining alleles.

Another effect is seen when examining recombination between alleles: The proportions of the alleles differ from the initial 4:4 ratio. This effect is called gene conversion. It describes a nonreciprocal transfer of information from one chromatid to another.

Gene conversion results from exchange of strands between DNA molecules, and the change in sequence may have either of two causes at the molecular level, known as gap repair or mismatch repair:

  • Gap repair: As indicated by the DSBR model in Figure 13.4, one DNA duplex may act as a donor of genetic information that directly replaces the corresponding sequences in the recipient duplex by a process of gap generation, strand exchange, and gap filling.

  • Mismatch repair: As part of the exchange process, heteroduplex DNA is generated when a single strand from one duplex pairs with its complement in the other duplex. Repair systems recognize mispaired bases in heteroduplex DNA, and then may excise and replace one of the strands to restore complementarity (see the chapter titled Repair Systems). Such an event converts the strand of DNA representing one allele into the sequence of the other allele.

Gene conversion does not depend on crossing over, but rather is correlated with it. A large proportion of the aberrant asci show genetic recombination between two markers on either side of a site of interallelic gene conversion. This is exactly what would be predicted if the aberrant ratios result from initiation of the recombination process as shown in Figure 13.4, but with an approximately equal probability of resolving the structure with or without recombination. The implication is that fungal chromosomes initiate crossing over about twice as often as would be expected from the measured frequency of recombination between distant genes.

Various biases are seen when recombination is examined at the molecular level. Either direction of gene conversion may be equally likely, or allele-specific effects may create a preference for one direction. Gradients of recombination may fall away from hotspots. We now know that recombination hotspots represent sites at which DSBs are preferentially initiated, and that the gradient is correlated with the extent to which the gap at the hotspot is enlarged and converted to long single-stranded ends (see the section in this chapter titled The Synaptonemal Complex Forms After Double-Strand Breaks).

Some information about the extent of gene conversion is provided by the sequences of members of gene clusters. Usually, the products of a recombination event will separate and become unavailable for analysis at the level of DNA sequence. When a chromosome carries two (nonallelic) genes that are related, though, they may recombine by an “unequal crossing-over” event (see the chapter titled Clusters and Repeats). All we need to note for now is that a heteroduplex may be formed between the two nonallelic genes. Gene conversion effectively converts one of the nonallelic genes to the sequence of the other.

The presence of more than one gene copy on the same chromosome provides a footprint to trace these events. For example, if heteroduplex formation and gene conversion occurred over part of one gene, this part may have a sequence identical with, or very closely related to, the other gene, whereas the remaining part shows more divergence. Available sequences suggest that gene-conversion events may extend for considerable distances, up to a few thousand bases.

13.5 The Synthesis-Dependent Strand-Annealing Model

The DSBR model accounts for meiotic homologous recombination that gives crossover products, but it cannot explain all homologous recombination because mitotic gene conversions are typically not accompanied by crossing over. The synthesis-dependent strand-annealing (SDSA) model serves as a better model for what occurs during mitotic homologous recombination in which DSB repair events and gene conversion are not associated with crossing over. Studies of the DSB that occurs during mating-type switching events in yeast (discussed later in this chapter) led to the development of SDSA as a model for mitotic recombination.

The synthesis-dependent strand-annealing pathway, shown in FIGURE 13.8, is initiated in a mechanism similar to the DSBR model in that DSBs are processed by 5′-end resection. Following strand invasion and DNA synthesis, the second end is not captured as it is in the DSBR model. In the SDSA model, the invading strand, which contains newly synthesized DNA identical in sequence to the strand it displaced, is itself displaced. Following displacement, the invading strand reanneals with the other end of the DSB. This is followed by synthesis and ligation to repair the DSB. In this model, the break is repaired using the homologous sequence as a template, but does not involve crossing over. This feature of the SDSA model makes it suitable for mitotic gene conversions for which there is no associated crossing over. The SDSA pathway is also responsible for recombination without crossover in the first phase of meiosis (discussed in the section in this chapter titled The Synaptonemal Complex Forms After Double-Strand Breaks).

FIGURE 13.8 The synthesis-dependent strand-annealing (SDSA) model of homologous recombination. Recombination is initiated by a double-strand break and is followed by end processing to form single-strand tails with 3′–OH ends. Strand invasion and DNA synthesis repair one strand of the break. Instead of second-strand capture as depicted in Figure 13.4, the strand in the D-loop is displaced. The single strand can anneal with the single strand of the other end. Repair synthesis then completes the double-strand break repair process. No Holliday junction is formed, and the product is always noncrossover.

13.6 The Single-Strand Annealing Mechanism Functions at Some Double-Strand Breaks

Some homologous recombination events to repair double-­strand breaks are not dependent on strand invasion, D-loop formation, or the proteins that promote these processes. In order to account for these recombination events, which typically take place between direct repeats (repeat sequences that are oriented in the same direction), a model has been devised in which homology between single-strand overhangs is used to direct recombination (see FIGURE 13.9). When a DSB occurs between two direct repeats, the ends are resected to give single strands. When resection proceeds to the repeat sequences such that the 3′–single-strand tails are homologous, the single strands can anneal. Processing and ligation of the 3′ ends then seals the DSB. As shown in Figure 13.9, this resection, followed by annealing, eliminates the sequence between the two direct repeats and leaves only one copy of the repeated sequence. Some human diseases arise from the loss of the sequence between the direct repeats, presumably through a single-strand annealing (SSA) mechanism. These diseases include insulin-dependent diabetes, Fabry disease, and α-thalassemia.

FIGURE 13.9 The single-strand annealing model of homologous recombination. A double-strand break occurs between direct repeats, depicted as red arrows. Following end processing to form single-strand tails with 3′–OH ends, the single strands anneal by homology at the red arrows. The single-strand tails are removed by endonucleases that recognize branch structures. The end product is double-strand break repair with a deletion of the sequences between the repeats and loss of one repeat sequence.

13.7 Break-Induced Replication Can Repair Double-Strand Breaks

We saw in the previous section that DSBs between direct repeats can induce the single-strand annealing mechanism. There are other types of repeat sequences at which DSBs induce a repair mechanism known as break-induced replication (BIR). During DNA replication, certain sequences termed fragile sites are particularly susceptible to DSB formation. They often contain repeat sequences related to those found in transposable elements (discussed in the chapter titled Transposable Elements and Retroviruses) and are located throughout the genome. Fragile sites are prone to breakage during DNA replication, creating a DSB at the site of replication. Break-induced replication can initiate repair from these DSBs by using the homologous sequence from a repeat on a nonhomologous chromosome, creating a nonreciprocal translocation, as shown in FIGURE 13.10.

FIGURE 13.10 Break-induced replication can result in nonreciprocal translocations. A DNA break on the red chromosome results in loss of the chromosome end and a break with only one end. The end is repaired by recombination, using a homologous sequence found on a different chromosome, here the blue chromosome. Because there is only one end at the broken chromosome, repair occurs by copying the blue chromosome sequence to the end. This results in a translocation of some of the blue chromosome sequence to the red chromosome.

The mechanism of BIR involves resection of the double-strand break end to leave a 3′–OH single-strand overhang, which can then undergo strand invasion at a homologous sequence, as shown in FIGURE 13.11. The invading strand causes the formation of a D-loop that can be thought of as a replication bubble. The invading strand is then extended using the donor DNA as template for replication. When the invading strand is displaced, it can then act as a single-stranded template on which synthesis can be primed to create double-stranded DNA. The template strand is used until replication reaches the end of the chromosome; as a result, gene conversions from BIR events can be hundreds of kilobases long. Additionally, chromosome translocations can occur from this process if the homology used during strand invasion is a result of repeat sequences present at various sites in the genome. Template switching that occurs during BIR can result in some of the complex chromosomal rearrangements that are seen in tumor cells.

FIGURE 13.11 Possible mechanisms of break-induced replication. Strand invasion into homologous sequences by a single-strand tail with a 3′–OH end forms a D-loop. In (a), synthesis results in a single-strand region that is later converted into duplex DNA. In (b), a single replication fork is formed that moves in one direction to the end of the template sequence. Resolution of the Holliday junction results in newly synthesized DNA on both molecules. In (c), the Holliday junction branch migrates to result in newly synthesized DNA only on the broken strand, as in (a). (d) Shows the final products after resolution.

Data from M. J. McEachern and J. E. Haber, Annu. Rev. Biochem. 75 (2006): 111–135.

13.8 Recombining Meiotic Chromosomes Are Connected by the Synaptonemal Complex

A basic paradox in recombination is that the parental chromosomes never seem to be in close enough contact for recombination of DNA to occur. The chromosomes enter meiosis in the form of replicated (sister chromatid) pairs, which are visible as a mass of chromatin. They pair to form the synaptonemal complex, and it has been assumed for many years that this represents some stage involved with recombination—possibly a necessary preliminary to exchange of DNA. A more recent view is that the synaptonemal complex is a consequence rather than a cause of recombination, but we have yet to define how the structure of the synaptonemal complex relates to molecular contacts between DNA molecules.

Synapsis begins when each chromosome (sister chromatid pair) condenses around a proteinaceous structure called the axial element. The axial elements of corresponding chromosomes then become aligned, and the synaptonemal complex forms as a tripartite structure, in which the axial elements, now called lateral elements, are separated from each other by a central element. FIGURE 13.12 shows an example.

FIGURE 13.12 The synaptonemal complex brings chromosomes into juxtaposition.

Reproduced from D. von Wettstein. Proc. Natl. Acad. Sci. USA 68 (1971): 851–855. Photo courtesy of Diter von Wettstein, Washington State University.

Each chromosome at this stage appears as a mass of chromatin bounded by a lateral element. The two lateral elements are separated from each other by a fine, but dense, central element. The triplet of parallel dense strands lies in a single plane that curves and twists along its axis. The distance between the homologous chromosomes is considerable in molecular terms at more than 200 nm (the diameter of DNA is 2 nm). Thus, a major problem in understanding the role of the complex is that, although it aligns homologous chromosomes, it is far from bringing homologous DNA molecules into contact.

The only visible link between the two sides of the synaptonemal complex is provided by spherical or cylindrical structures observed in fungi and insects. They lie across the complex and are called nodes or recombination nodules; they occur with the same frequency and distribution as the chiasmata. Their name reflects the possibility that they may prove to be the sites of recombination.

From mutations that affect synaptonemal complex formation, we can relate the types of proteins that are involved to its structure. FIGURE 13.13 presents a molecular view of the synaptonemal complex. Its distinctive structural features are due to two groups of proteins:

  • The cohesins form a single linear axis for each pair of sister chromatids from which loops of chromatin extend. This is equivalent to the lateral element of Figure 13.12. (The cohesins belong to a general group of proteins involved in connecting sister chromatids so that they segregate properly at mitosis or meiosis; they are discussed further in the chapter titled Epigenetics II.)

  • The lateral elements are connected by transverse filaments that are equivalent to the central element of Figure 13.12. These are formed from Zip proteins.

FIGURE 13.13 Each pair of sister chromatids has an axis made of cohesins. Loops of chromatin project from the axis. The synaptonemal complex is formed by linking together the axes via Zip proteins.

Mutations in proteins that are needed for lateral elements to form are found in the genes coding for cohesins. The cohesins that are used in meiosis include Smc3 (which is also used in mitosis) and Rec8 (which is specific to meiosis and is related to the mitotic cohesin Scc1). The cohesins appear to bind to specific sites along the chromosomes in both mitosis and meiosis. They are likely to play a structural role in chromosome segregation. At meiosis, the formation of the lateral elements may be necessary for the later stages of recombination, because although these mutations do not prevent the formation of DSBs, they do block formation of recombinants.

The zip1 mutation allows lateral elements to form and to become aligned, but they do not become closely synapsed. The N-terminal domain of the Zip1 protein is localized in the central element, but the C-terminal domain is localized in the lateral elements. Two other proteins, Zip2 and Zip3, are also localized with Zip1. The group of Zip proteins forms transverse filaments that connect the lateral elements of the sister chromatid pairs.

13.9 The Synaptonemal Complex Forms After Double-Strand Breaks

Evidence suggests that DSBs initiate recombination in both homologous and site-specific recombination in yeast. DSBs were initially implicated in the change of mating type, which involves the replacement of one sequence by another (see the section in this chapter titled Unidirectional Gene Conversion Is Initiated by the Recipient MAT Locus). DSBs also occur early in meiosis at sites that provide hotspots for recombination. Their locations are not sequence specific. They tend to occur in promoter regions and to coincide with more accessible regions of chromatin. The frequency of recombination declines in a gradient on one or both sides of the hotspot. The hotspot identifies the site at which recombination is initiated, and the gradient reflects the probability that the recombination events will spread from it.

We may now interpret the role of DSBs in molecular terms. The blunt ends created by the DSB are rapidly converted on both sides into long 3′–single-stranded ends, as shown in the model of Figure 13.4. A yeast mutation (rad50) that blocks the conversion of the blunt end into the single-stranded protrusion is defective in recombination. This suggests that DSBs are necessary for recombination. The gradient is determined by the declining probability that a single-stranded region will be generated as distance increases from the site of the DSB.

In rad50 mutants, the 5′ ends of the DSBs are connected to the protein Spo11, which, as discussed previously, is homologous to the catalytic subunits of a family of type II topoisomerases. Spo11 generates the DSBs. Recall that the model for this reaction, shown in Figure 13.5, suggests that Spo11 interacts reversibly with DNA; the break is converted into a permanent structure by an interaction with another protein that dissociates the Spo11 complex. Removal of Spo11 is then followed by nuclease action. At least nine other proteins are required to process the DSBs. One group of proteins is required to convert the DSBs into protruding 3′–OH single-stranded ends. Another group then enables the single-­stranded ends to invade homologous duplex DNA.

The correlation between recombination and synaptonemal complex formation is well established in most species, and recent work has shown that all mutations that abolish chromosome pairing in Drosophila or in yeast also prevent recombination (a few species appear to lack this strict dependence, however). The system for generating the DSBs that initiate recombination is generally conserved. Spo11 homologs have been identified in several higher eukaryotes, and a mutation in the Drosophila gene blocks all meiotic recombination.

A few systems are available in which it is possible to compare molecular and cytological events at recombination, but recently there has been progress in analyzing meiosis in Saccharomyces cerevisiae. The relative timing of events is summarized in FIGURE 13.14.

FIGURE 13.14 Double-strand breaks appear when axial elements form and disappear during the extension of synaptonemal complexes. Joint molecules appear and persist until DNA recombinants are detected at the end of pachytene.

DSBs appear and then disappear over a 60-minute period. The first joint molecules, which are putative recombination intermediates, appear soon after the DSBs disappear. The sequence of events suggests that DSBs, individual pairing reactions, and formation of recombinant structures occur in succession at the same chromosomal site.

DSBs appear during the period when axial elements form. They disappear during the conversion of the paired chromosomes into synaptonemal complexes. This relative timing of events suggests that formation of the synaptonemal complex results from the initiation of recombination via the introduction of DSBs and their conversion into later intermediates of recombination. This idea is supported by the observation that the rad50 mutant cannot convert axial elements into synaptonemal complexes. This refutes the traditional view of meiosis that the synaptonemal complex represents the need for chromosome pairing to precede the molecular events of recombination.

It has been difficult to determine whether recombination occurs at the stage of synapsis, because recombination is assessed by the appearance of recombinants after the completion of meiosis. By assessing the appearance of recombinants in yeast directly in terms of the production of DNA molecules containing diagnostic restriction sites, though, it has been possible to show that recombinants appear at the end of pachytene. This clearly places the completion of the recombination event after the formation of synaptonemal complexes.

Thus, the synaptonemal complex forms after the DSBs that initiate recombination, and it persists until the formation of recombinant molecules. It does not appear to be necessary for recombination as such, because some mutants that lack a normal synaptonemal complex can generate recombinants. Mutations that abolish recombination, however, also fail to develop a synaptonemal complex. This suggests that the synaptonemal complex forms as a consequence of recombination, following chromosome pairing, and is required for later stages of meiosis.

The DSBR model proposes that resolution of Holliday junctions gives rise to either noncrossover products (with a residual stretch of hybrid DNA) or to crossovers (recombinants), depending on which strands are involved in resolution (see Figure 13.4). Recent measurements of the times of production of noncrossover and crossover molecules, however, suggest that this may not be true. Crossovers do not appear until well after the first appearance of joint molecules, whereas noncrossovers appear almost simultaneously with the joint molecules (see Figure 13.14). The appearance of these two types of products corresponds to what is considered two independent phases of meiotic recombination. In the first phase, DSBs are repaired through a SDSA reaction, leading to noncrossover products, whereas in the second phase the DSBR pathway is predominant and results largely in crossover products. The molecular outcomes of these phases are illustrated in FIGURE 13.15. If both types of product were produced by the same resolution process, however, we would expect them to appear at the same time. The discrepancy in timing suggests that crossovers are produced as previously thought—by resolution of joint molecules—but that other routes, such as SDSA, lead to production of noncrossovers. Current research has uncovered roles for a group of proteins known as ZMMs, which in yeast include the proteins Zip1-4, Msh4 and Msh5 (mismatch repair proteins), Mer3, and Spo16. These proteins are well conserved, include a number of distinct functions, and have roles in crossover determination, synapsis, and other aspects of recombination.

FIGURE 13.15 Model of meiotic homologous recombination. A DNA duplex (a) is cleaved by Spo11 to form a double-strand break with Spo11 covalently attached to the ends (b). After Spo11 is removed the ends are resected by the MRX/N complex to give single-strand tails with 3′–OH ends, which are complexed with Rad51 and Dmc1. Strand exchange occurs by strand invasion (d and g). Second-end capture results in a double Holliday junction, which is resolved to form crossover products (e and f). Most of the double-strand breaks do not engage in a second-end capture mechanism and instead engage in a synthesis-dependent strand-annealing mechanism (h and i), which results in noncrossover products.

Data from M. J. Neale and S. Keeney, Nature 442 (2006): 153–158.

13.10 Pairing and Synaptonemal Complex Formation Are Independent

We can distinguish the processes of pairing and synaptonemal complex formation by the effects of two mutations, each of which blocks one of the processes without affecting the other.

A mutation in the ZMM protein Zip2 allows chromosomes to pair, but they do not form synaptonemal complexes. Thus, recognition between homologs is independent of recombination or synaptonemal complex formation.

The specificity of association between homologous chromosomes is controlled by the gene HOP2 in S. cerevisiae. In hop2 mutants, normal amounts of synaptonemal complex form at meiosis, but the individual complexes contain nonhomologous chromosomes. This suggests that the formation of synaptonemal complexes as such is independent of homology (and therefore cannot be based on any extensive comparison of DNA sequences). The usual role of Hop2 is to prevent nonhomologous chromosomes from interacting.

DSBs form in the mispaired chromosomes in the synaptonemal complexes of hop2 mutants, but they are not repaired. This suggests that, if formation of the synaptonemal complex requires DSBs, it does not require any extensive reaction of these breaks with homologous DNA.

It is not clear what usually happens during pachytene, before DNA recombinants are observed. It may be that this period is occupied by the subsequent steps of recombination, which involve the extension of strand exchange, DNA synthesis, and resolution.

At the next stage of meiosis (diplotene), the chromosomes shed the synaptonemal complex; the chiasmata then become visible as points at which the chromosomes are connected. This has been presumed to indicate the occurrence of a genetic exchange, but the molecular nature of a chiasma is unknown. It is possible that it represents the residuum of a completed exchange, or that it represents a connection between homologous chromosomes where a genetic exchange has not yet been resolved. Later in meiosis, the chiasmata move toward the ends of the chromosomes. This flexibility suggests that they represent some remnant of the recombination event rather than providing the actual intermediate.

Recombination events occur at discrete points on meiotic chromosomes, but it is not yet possible to correlate their occurrences with the discrete structures that have been observed; that is, recombination nodules and chiasmata. Insights into the molecular basis for the formation of discontinuous structures, however, are provided by the identification of proteins involved in yeast recombination that can be localized to discrete sites. These include Msh4 (a mismatch repair protein in the ZMM group) and Dmc1 and Rad51 (which are homologs of the Escherichia coli RecA protein). The exact roles of these proteins in recombination remain to be established.

Recombination events are subject to a general control. Only a minority of interactions actually mature as crossovers, but these are distributed in such a way that, in general, each pair of homologs acquires only one to two crossovers, yet the probability of zero crossovers for a homologous pair is very low (less than 0.1%). This process is probably the result of a single crossover control, because the nonrandomness of crossovers is generally disrupted in certain mutants. Furthermore, the occurrence of recombination is necessary for progress through meiosis, and a “checkpoint” system exists to block meiosis if recombination has not occurred. (The block is lifted when recombination has been successfully completed; this system provides a safeguard to ensure that cells do not try to segregate their chromosomes until recombination has occurred.)

13.11 The Bacterial RecBCD System Is Stimulated by chi Sequences

The nature of the events involved in exchange of sequences between DNA molecules was first described in bacterial systems. Here the recognition reaction is part and parcel of the recombination mechanism and involves restricted regions of DNA molecules rather than intact chromosomes. The general order of molecular events is similar, though: A single strand from a broken molecule interacts with a partner duplex, the region of pairing is extended, and an endonuclease resolves the partner duplexes. Enzymes involved in each stage are known, although they probably represent only some of the components required for recombination.

Bacterial enzymes implicated in recombination have been identified by the occurrence of rec− mutations in their genes. The phenotype of rec− mutants is the inability to undertake generalized recombination. Some 10 to 20 loci have been identified.

Bacteria do not usually exchange large amounts of duplex DNA, but there may be various routes to initiate recombination in prokaryotes. In some cases, DNA may be available with free single-stranded 3′ ends: DNA may be provided in single-stranded form (as in conjugation; see the chapter titled Extrachromosomal Replicons), single-stranded gaps may be generated by irradiation damage, or single-stranded tails may be generated by phage genomes undergoing replication by a rolling circle. In circumstances involving two duplex molecules (as in recombination at meiosis in eukaryotes), however, single-stranded regions and 3′ ends must be generated.

One mechanism for generating suitable ends has been discovered as a result of the existence of certain hotspots that stimulate recombination. These hotspots, which were discovered in phage lambda in the form of mutants called chi, have single base–pair changes that create sequences that stimulate recombination. These sites lead us to the role of other proteins involved in recombination.

These sites share a constant nonsymmetrical sequence of 8 bp:

  1. 5′ GCTGGTGG 3′

  2. 3′ CGACCACC 5′

The chi sequence occurs naturally in E. coli DNA about once every 5 to 10 kb. Its absence from wild-type lambda DNA, and also from other genetic elements, shows that it is not essential for recombination.

A chi sequence stimulates recombination in its general vicinity, within about a distance of up to 10 kb from the site. A chi site can be activated by a DSB made several kilobases away on one particular side (to the right of the sequence shown previously). This dependence on orientation suggests that the recombination apparatus must associate with DNA at a broken end, and then can move along the duplex only in one direction.

chi sites are targets for the action of an enzyme encoded by the genes recBCD. This complex possesses several activities: It is a potent nuclease that degrades DNA (originally identified as the activity exonuclease V); it has helicase activities that can unwind duplex DNA in the presence of a single-strand binding (SSB) protein; and it has ATPase activity. Its role in recombination may be to provide a single-stranded region with a free 3′ end.

FIGURE 13.16 shows how these reactions are coordinated on a substrate DNA that has a chi site. RecBCD binds to DNA at a double-stranded end. Two of its subunits have helicase activities: RecD functions with 5′→3′ polarity, and RecB functions with 3′→5′ polarity. Translocation along DNA and unwinding the double helix is initially driven by the RecD subunit. As RecBCD advances, it degrades the released single strand with the 3′ end. When it reaches the chi site, it recognizes the top strand of the chi site in single-stranded form. This causes the enzyme to pause. It then cleaves the top strand of the DNA at a position between four and six bases to the right of chi. Recognition of the chi site causes the RecD subunit to dissociate or become inactivated, at which point the enzyme loses its nuclease activity. It continues, however, to function as a helicase—now using only the RecB subunit to drive translocation—at about half the previous speed. The overall result of this interaction is to generate single-stranded DNA with a 3′ end at the chi sequence. This is a substrate for recombination.

FIGURE 13.16 RecBCD nuclease approaches a chi sequence from one side, degrading DNA as it proceeds; at the chi site, it makes an endonucleolytic cut, loses RecD, and retains only the helicase activity.

13.12 Strand-Transfer Proteins Catalyze Single-Strand Assimilation

The E. coli protein RecA was the first example of a DNA strand-transfer protein to be discovered. It is the paradigm for a group that includes several other bacterial and archaeal proteins, as well as eukaryotic Rad51 and the meiotic protein Dmc1 (both discussed in detail in the section in this chapter titled Eukaryotic Genes Involved in Homologous Recombination). Analysis of yeast rad51 mutants shows that this class of protein plays a central role in recombination. They accumulate DSBs and fail to form normal synaptonemal complexes. This reinforces the idea that exchange of strands between DNA duplexes is involved in formation of the synaptonemal complex and raises the possibility that chromosome synapsis is related to the bacterial strand assimilation reaction.

RecA in bacteria has two quite different types of activity: It can stimulate protease activity in the SOS response (see the chapter titled Repair Systems), and it can promote base pairing between a single strand of DNA and its complement in a duplex molecule. Both activities are activated by single-stranded DNA in the presence of ATP.

The DNA-handling activity of RecA enables a single strand to displace its homolog in a duplex in a reaction that is called single-strand assimilation (or single-strand invasion). The displacement reaction can occur between DNA molecules in several configurations and has three general conditions:

  • One of the DNA molecules must have a single-stranded region.

  • One of the molecules must have a free 3′ end.

  • The single-stranded region and the 3′ end must be located within a region that is complementary between the molecules.

The reaction is illustrated in FIGURE 13.17. When a linear single strand invades a duplex, it displaces the original partner to its complement. The reaction can be followed most easily by making either the donor or recipient a circular molecule. The reaction proceeds 5′→3′ along the strand whose partner is being displaced and replaced; that is, the reaction involves an exchange in which (at least) one of the exchanging strands has a free 3′ end.

FIGURE 13.17 RecA promotes the assimilation of invading single strands into duplex DNA as long as one of the reacting strands has a free end.

Single-strand assimilation is potentially related to the initiation of recombination. All models call for an intermediate in which one or both single strands cross over from one duplex to the other (see Figure 13.4). RecA could catalyze this stage of the reaction. In the bacterial context, RecA acts on substrates generated by RecBCD. RecBCD-mediated unwinding and cleavage can be used to generate ends that initiate the formation of heteroduplex joints. RecA can take the single strand with the 3′ end that is released when RecBCD cuts at chi, and then use it to react with a homologous duplex sequence, thus creating a joint molecule.

All of the bacterial and archaeal proteins in the RecA family can aggregate into long filaments with single-stranded or duplex DNA. Six RecA monomers are bound to DNA per turn of the RecA-DNA filament, which has a helical structure with a deep groove that contains the DNA. The stoichiometry of binding is three nucleotides (or base pairs) per RecA monomer. The DNA is held in a form that is extended 1.5 times relative to duplex B DNA, making a turn every 18.6 nucleotides (or base pairs). When duplex DNA is bound, it contacts RecA via its minor groove, leaving the major groove accessible for possible reaction with a second DNA molecule.

The interaction between two DNA molecules occurs within these filaments. When a single strand is assimilated into a duplex, the first step is for RecA to bind the single strand into a presynaptic filament. The duplex is then incorporated, probably forming some sort of triple-stranded structure. In this system, synapsis precedes physical exchange of material, because the pairing reaction can take place even in the absence of free ends, when strand exchange is impossible. A free 3′ end is required for strand exchange. The reaction occurs within the filament, and RecA remains bound to the strand that was originally single, so that at the end of the reaction RecA is bound to the duplex molecule.

All of the proteins in this family can promote the basic process of strand exchange without a requirement for energy input. RecA, however, augments this activity by using ATP hydrolysis. Large amounts of ATP are hydrolyzed during the reaction. The ATP may act through an allosteric effect on RecA conformation. When bound to ATP, the DNA-binding site of RecA has a high affinity for DNA; this is needed to bind DNA and for the pairing reaction. Hydrolysis of ATP converts the binding site to low affinity, which is needed to release the heteroduplex DNA.

We can divide the reaction that RecA catalyzes between single-stranded and duplex DNA into three phases:

  • A slow presynaptic phase in which RecA polymerizes on single-stranded DNA

  • A fast pairing reaction between the single-stranded DNA and its complement in the duplex to produce a heteroduplex joint

  • A slow displacement of one strand from the duplex to produce a long region of heteroduplex DNA

The presence of SSB stimulates the reaction by ensuring that the substrate lacks secondary structure. It is not clear yet how SSB and RecA both can act on the same stretch of DNA. Like SSB, RecA is required in stoichiometric amounts, which suggests that its action in strand assimilation involves binding cooperatively to DNA to form a structure related to the filament.

When a single-stranded molecule reacts with a duplex DNA, the duplex molecule becomes unwound in the region of the recombinant joint. The initial region of heteroduplex DNA may not even lie in the conventional double-helical form, but could consist of the two strands associated side by side. A region of this type is called a paranemic joint, as compared with the classical intertwined plectonemic relationship of strands in a double helix, depicted in FIGURE 13.18. A paranemic joint is unstable; further progress of the reaction requires its conversion to the double-helical form. This reaction is equivalent to removing negative supercoils and may require an enzyme that solves the unwinding/rewinding problem by making transient breaks that allow the strands to rotate about each other.

FIGURE 13.18 Formation of paranemic and plectonemic joints. Once homology is found, side-by-side pairing is formed, called paranemic pairing, which then transitions to plectonemic pairing, where the paired DNA strands are in a double-helix configuration. Note that these pairing stages involve strand invasion and D-loop formation.

Data from P. R. Bianco and S. C. Kowalczykowski. Encyclopedia of Life Sciences. John Wiley & Sons, Ltd., 2005.

All of the reactions we have discussed so far represent only a part of the potential recombination event: the invasion of one duplex by a single strand. Two duplex molecules can interact with each other under the sponsorship of RecA, provided that one of them has a single-stranded region of at least 50 bases. The single-stranded region can take the form of a tail on a linear molecule or of a gap in a circular molecule.

The reaction between a partially duplex molecule and an entirely duplex molecule leads to the exchange of strands. An example is illustrated in FIGURE 13.19. Assimilation starts at one end of the linear molecule, where the invading single strand displaces its homolog in the duplex in the customary way. When the reaction reaches the region that is duplex in both molecules, though, the invading strand unpairs from its partner, which then pairs with the other displaced strand.

FIGURE 13.19 RecA-mediated strand exchange between partially duplex and entirely duplex DNA generates a joint molecule with the same structure as a recombination intermediate.

At this stage, the molecule has a structure indistinguishable from the recombinant joint in Figure 13.4. The reaction sponsored in vitro by RecA can generate Holliday junctions, which suggests that the enzyme can mediate reciprocal strand transfer. Less is known about the geometry of the four-strand intermediates bound by RecA, but presumably two duplex molecules can lie side by side in a way consistent with the requirements of the exchange reaction.

The biochemical reactions characterized in vitro leave open many possibilities for the functions of strand-transfer proteins in vivo. Their involvement is triggered by the availability of a single-stranded 3′ end. In bacteria, this is most likely generated when RecBCD processes a DSB to generate a single-stranded end. One of the main circumstances in which this is invoked may be when a replication fork stalls at a site of DNA damage (see the chapter titled Repair Systems). The introduction of DNA during conjugation, when RecA is required for recombination with the host chromosome, is more closely related to conventional recombination. In yeast, DSBs may be generated by DNA damage or as part of the normal process of recombination. In either case, processing of the break to generate a 3′–single-stranded end is followed by loading the single strand into a filament with Rad51, followed by a search for matching duplex sequences. This can be used in both repair and recombination reactions.

13.13 Holliday Junctions Must Be Resolved

One of the most critical steps in recombination is the resolution of the Holliday junction, which determines whether there is a reciprocal recombination or a reversal of the structure that leaves only a short stretch of hybrid DNA (see Figure 13.4). Branch migration from the exchange site (see Figure 13.6) determines the length of the region of hybrid DNA (with or without recombination). The proteins involved in stabilizing and resolving Holliday junctions have been identified as the products of the ruv genes in E. coli. RuvA and RuvB increase the formation of heteroduplex structures. RuvA recognizes the structure of the Holliday junction. RuvA binds to all four strands of DNA at the crossover point and forms two tetramers that sandwich the DNA. RuvB is a hexameric helicase with an ATPase activity that provides the motor for branch migration. Hexameric rings of RuvB bind around each duplex of DNA upstream of the crossover point. A diagram of the complex is shown in FIGURE 13.20.

FIGURE 13.20 RuvAB is an asymmetric complex that promotes branch migration of a Holliday junction.

The RuvAB complex can cause the branch to migrate as fast as 10 to 20 bp per second. A similar activity is provided by another helicase, RecG. RuvAB displaces RecA from DNA during its action. The RuvAB and RecG activities both can act on Holliday junctions, but if both are mutant, E. coli is completely defective in recombination activity.

The third gene, ruvC, encodes an endonuclease that specifically recognizes Holliday junctions. It can cleave the junctions in vitro to resolve recombination intermediates. A common tetranucleotide sequence provides a hotspot for RuvC to resolve the Holliday junction. The tetranucleotide (ATTG) is asymmetric, and thus may direct resolution with regard to which pair of strands is nicked. This determines whether the outcome is patch recombinant formation (no overall recombination) or splice recombinant formation (recombination between flanking markers). Crystal structures of RuvC and other junction-resolving enzymes show that there is little structural similarity among the group, in spite of their common function.

We may now account for the stages of recombination in E. coli in terms of individual proteins. FIGURE 13.21 shows the events that are involved in using recombination to repair a gap in one duplex by retrieving material from the other duplex. The major caveat in applying these conclusions to recombination in eukaryotes is that bacterial recombination generally involves interaction between a fragment of DNA and a whole chromosome. It occurs as a repair reaction that is stimulated by damage to DNA, but this is not entirely equivalent to recombination between genomes at meiosis. Nonetheless, similar molecular activities are involved in manipulating DNA.

FIGURE 13.21 Bacterial enzymes can catalyze all stages of recombination in the repair pathway following the production of suitable substrate DNA molecules.

All of this suggests that recombination uses a “resolvasome” complex that includes enzymes catalyzing branch migration as well as junction-resolving activity. It is possible that mammalian cells contain a similar complex.

Although resolution in eukaryotic cells is less well understood, a number of proteins have been implicated in mitotic and meiotic resolution. S. cerevisiae strains that contain mus81 mutations are defective in recombination. Mus81 is a component of an endonuclease that resolves Holliday junctions into duplex structures. The resolvase is important both in meiosis and for restarting stalled replication forks (see the chapter titled Repair Systems). Other proteins known to be involved in the resolution process are described in the broader context of eukaryotic homologous recombination factors in the following section.

13.14 Eukaryotic Genes Involved in Homologous Recombination

Previously, we briefly mentioned some of the proteins involved in homologous recombination in eukaryotes. In this section, they are discussed in more detail, focusing on the DSBR and SDSA models. (Their roles in repair are also discussed further in the Repair Systems chapter.) Additionally, the steps in the single-strand annealing and break-induced replication mechanisms that overlap with those of DSBR and SDSA proceed by the same enzymatic processes.

Many of the eukaryotic homologous recombination genes are called RAD genes because they were first isolated in screens for mutants with increased sensitivity to X-ray irradiation. X-rays make DSBs in DNA; thus it is not surprising that rad mutants sensitive to X-rays also are defective in mitotic and meiotic recombination. The DSBR model shown in Figure 13.4 indicates at which step the proteins described in the following paragraphs act.

1. End Processing/Presynapsis

In mitotic cells, DSBs are produced by exogenous sources such as irradiation or chemical treatment and from endogenous sources such as topoisomerases and nicks on the template strand. During replication nicks are converted to DSBs. The ends of these breaks are processed by exonucleolytic degradation to have single-strand tails with 3′–OH ends. In meiosis, DSBs are induced by Spo11-dependent cleavage. The first step in end processing entails binding of the broken end by the MRN or MRX complex, in association with the endonuclease Sae2 (CtIP in mammalian cells).

Mre11 works as part of a complex with two other factors, called Rad50 and Xrs2 in yeast and Rad50 and Nbs1 in humans. Xrs2 and Nbs1 have no similarity to each other. Rad50 is thought to help hold DSB ends together via dimers connected at the tips by a hook structure that becomes active in the presence of zinc ion, as shown in FIGURE 13.22. Rad50 and Mre11 are related to the bacterial proteins SbcC and SbcD, which have double-stranded DNA exonuclease and single-stranded endonuclease activities. Xrs2 and Nbs1 have DNA-binding activity. Nbs1 is so named because a mutant allele was first discovered in individuals with Nijmegen breakage syndrome, a rare DNA damage syndrome that is associated with defective DNA damage checkpoint signaling and lymphoid tumors. Rare mutations that produce MRE11 with low activity have been found in humans who have ataxia-telangiectasia-like disorder (ATLD). Patients with this syndrome have not been reported to be cancer prone, but they have developmental problems and show defects in DNA damage checkpoint signaling. Mutations in MRE11, RAD50, or XRS2 render cells sensitive to ionizing radiation and diploids have a poor meiotic outcome. Null mutations of MRE11, RAD50, or NBS1 in mice are lethal.

FIGURE 13.22 Structure of Rad50 and model for the MRX/N complex binding to double-strand breaks. Rad50 has a coiled coil domain similar to SMC (structural maintenance of chromosomes) proteins. The globular end contains two ATP-binding and hydrolysis regions (a and b) and forms a complex with Mre11 and Nbs1 (N) or Xrs2 (X). The other end of the coil binds a zinc cation and forms a dimer with another MRX/N molecule. The globular end binds to chromatin. The complex binds to double-strand breaks and can bring them together in a reaction involving two ends and one MRN/X complex (top right figure) or through an interaction between two MRX/N dimers (bottom right figure).

Data from M. Lichten, Nat. Struct. Mol. Biol. 12 (2005): 392–393.

After MRN/MRX and CtIP/Sae2 have prepared the DSB ends and removed any attached proteins or adduct that would inhibit end resection, the ends are resected by nucleases that act in concert with DNA helicases that unwind the duplex to expose single-strand DNA ends. Recent studies have identified the Exo1 and Dna2 exonucleases and the Sgs1 (in yeast) and BLM (in mammalian cells) helicases as critical factors for end processing.

After the DSBs have been processed to have 3′–OH single-strand tails, the single-strand DNA is bound first by the single-strand DNA-binding protein RPA to remove any secondary structure. Next, with the aid of mediator proteins that help Rad51 displace RPA and bind the single-strand DNA, Rad51 forms a nucleofilament. Rad51 is related to RecA with 30% identity and forms a right-handed helical nucleofilament in an ATP-dependent process, with six Rad51 molecules and 18 nucleotides of single-strand DNA per helical turn. This binding stretches the DNA by approximately 1.5-fold, compared to B-form DNA. Rad51 is required for all homologous recombination processes except single-strand annealing. RAD51 is not an essential gene in yeast, but null mutants are reduced in mitotic recombination and are sensitive to ionizing radiation. DSBs form but become degraded. In mice, RAD51 is essential, and mice that are homozygous for mutant rad51 do not survive past early stages of embryogenesis. This is thought to reflect the fact that, in vertebrates, at least one DSB occurs spontaneously during every replication cycle as a result of unrepaired template strand nicks.

In vitro, the mediators help in the removal of RPA and in the assembly of Rad51 on the single-stranded DNA and promote in vitro strand-exchange reactions. In yeast, the mediators are Rad52 and Rad55/Rad57. Rad55 and Rad57, which form a stable heterodimer, have some homology to Rad51, but have no strand-exchange activity in vitro.

In human cells, the mediators are also related to RAD51, with 20% to 30% sequence identity, and are called RAD51B, RAD51C, RAD51D, XRCC2, and XRCC3, or the “RAD51 paralogs.” (Recall that paralogs are genes that have arisen by duplication within an organism and therefore are related by sequence but have evolved to have different functions.) The human mediator proteins form three complexes: one composed of RAD51B and RAD51C, a second composed of RAD51D and XRCC2, and a third composed of RAD51C and XRCC3. The paralogous genes have been deleted in chicken cell lines and knocked down in mammalian cells. Although the cell lines are viable, they are subject to numerous chromosome breaks and rearrangements and have reduced viability compared to normal cell lines. Mice in which the paralogous genes have been deleted are not viable and undergo early embryonic death.

The human BRCA2 protein, which is mutated in familial breast and ovarian cancers and in the DNA damage syndrome Fanconi anemia, has mediator activity in vitro. Given that BRCA2 interacts physically with RAD51 and can bind to single-stranded DNA, this is not an unexpected activity for BRCA2. Indeed, genetic studies in mouse cells have shown that BRCA2 is required for homologous recombination. The related Brh2 protein of the pathogenic fungus Ustilago maydis binds in a complex to Rad51 and recruits it to single-strand DNA coated with RPA to initiate Rad51 nucleofilament formation.

Yeast mutants deleted for RAD55 or RAD57 show temperature-dependent ionizing radiation sensitivity and are reduced in homologous recombination. Neither mutant undergoes successful meiosis.

Rad52 is not essential for recombination in vivo in mammalian cells and does not appear to have a mediator role in these cells. It is, however, the most critical homologous recombination protein in yeast, as rad52 null mutants are extremely sensitive to ionizing radiation and are defective in all types of homologous recombination assayed. RAD52-deficient cells never complete meiosis.

2. Synapsis

Once the Rad51 filament has formed on single-strand DNA in the DBSR and SDSA processes, a search for homology with another DNA molecule begins and, once found, strand invasion to form a D-loop occurs. Strand invasion requires the Rad54 protein and the related Rdh54/Tid1 protein in yeast, and RAD54B in mammalian cells. Rad54 and Rdh54 are members of the SWI/SNF chromatin remodeling superfamily (see the chapter titled Eukaryotic Transcription Regulation). They possess a double-strand DNA-dependent ATPase activity, can promote chromatin remodeling, and can translocate on double-stranded DNA, inducing superhelical stress in double-stranded DNA. Although Rad54, Rdh54, and RAD54B are not DNA helicases, the translocase activity causes local opening of double strands, which may serve to stimulate D-loop formation. In yeast, RAD54 is required for efficient mitotic recombination and for repair of DSBs, because RAD54-deficient cells are sensitive to ionizing radiation and other DNA-damaging compounds. RDH54-deficient cells have a modest defect in recombination and are slightly DNA-damage sensitive. This sensitivity is enhanced when both RAD54 and RDH54 are deleted. In meiotic cells, rad54 mutants can complete meiosis but have reduced spore viability. The rdh54 mutants are more deficient in meiosis and have a stronger effect on spore viability. The double mutant does not complete meiosis. In chicken cells and mouse cells, RAD54 and RAD54B deletion mutants are viable, in contrast to other homologous recombination gene-deletion mutants. The cells show increased sensitivity to ionizing radiation and other clastogens (agents that cause chromosomal breaks) and have reduced rates of recombination.

3. DNA Heteroduplex Extension and Branch Migration

The proteins involved in this step are not as well defined as those required in the early steps of homologous recombination, yet the homologous DSBR and SDSA recombination pathways both have D-loop extension as an important part of the process. D-loop formation results in Rad51 filament being formed on double-stranded DNA. Rad54 protein has the ability to remove Rad51 from double-stranded DNA. This step might be important for DNA polymerase extension from the 3′ terminus. DNA polymerase delta (δ) is thought to be the polymerase for repair synthesis in DSB-mediated recombination; however, some recent studies have also implicated DNA polymerase h/Rad30 as being able to extend from the strand invasion intermediate terminus.

4. Resolution

The search for eukaryotic resolvase proteins has been a long process. Mutants of the DNA helicases Sgs1 of yeast and BLM in humans result in higher crossover rates. These helicases have thus been proposed to normally prevent crossover formation by promoting noncrossover Holliday junction resolution. This is proposed to occur by branch migration of the double Holliday junctions to convergence, through the DNA helicase action, as shown in FIGURE 13.23. The end structure is suggested to be a hemicatenane, where DNA strands are looped around each other. This structure is then resolved by the action of an associated DNA topoisomerase: Top3 in the case of Sgs1 and hTOPOIIIα in the case of BLM. In vitro, BLM and hTOPOIIIα can dissolve double Holliday junctions into a noncrossover molecule.

FIGURE 13.23 Double Holliday junction dissolution by the action of a DNA helicase and topoisomerase. The two Holliday junctions are pushed toward each other by branch migration using the DNA helicase activity. The resulting structure is a hemicatenane where single strands from two different DNA helices are wound around each other. This is cut by a DNA topoisomerase, unwinding and releasing the two DNA molecules and forming noncrossover products.

While the helicase–topoisomerase complex can resolve Holliday junctions as noncrossover in mitotic cells, the meiotic Holliday junction resolvase that can result in crossovers has not been fully identified. Additional endonuclease activities contained in the Mus81–Mms4 and Slx1–Slx4 complexes in yeast and the MUS81–EME1 and SLX1–SLX4 complexes in mammalian cells can cleave nicked Holliday junction–like structures and branched DNA structures. The relationship of this activity to meiotic crossover formation, however, is not fully defined. Recently, eukaryotic resolvase homologs were identified in humans and S. cerevisiae. The proteins GEN1 in humans and Yen1 in yeast are capable of resolving Holliday structures in vitro. These proteins are not normally essential for resolving recombination intermediates in vivo, but become essential in the absence of Mus81–Mms4.

13.15 Specialized Recombination Involves Specific Sites

Specialized recombination involves a reaction between two specific sites. The lengths of target sites are short and are typically in a range of 14 to 50 bp. In some cases the two sites have the same sequence, but in other cases they are nonhomologous. The reaction is used to insert a free phage DNA into the bacterial chromosome or to excise an integrated phage DNA from the chromosome, and in this case the two recombining sequences are different from one another. It is also used before division to regenerate monomeric circular chromosomes from a dimer that has been created by a generalized recombination event (see the chapter titled Replication Is Connected to the Cell Cycle). In this case the recombining sequences are identical.

The enzymes that catalyze site-specific recombination are generally called recombinases, and more than 100 of them are now known. Those involved in phage integration or related to these enzymes are also known as the integrase family. Prominent members of the integrase family are the prototypical Int from phage lambda, Cre from phage P1, and the yeast FLP enzyme (which catalyzes a chromosomal inversion).

The classic model for site-specific recombination is illustrated by phage lambda. The conversion of lambda DNA between its different life forms involves two types of events. The pattern of gene expression is regulated as described in the chapter titled Phage Strategies. The physical condition of the DNA is different in the lysogenic and lytic states:

  • In the lytic lifestyle, lambda DNA exists as an independent, circular molecule in the infected bacterium.

  • In the lysogenic state, the phage DNA is an integral part of the bacterial chromosome (called the prophage).

Transition between these states involves site-specific recombination:

  • To enter the lysogenic condition, free lambda DNA must be inserted into the host DNA. This is called integration.

  • To be released from lysogeny into the lytic cycle, prophage DNA must be released from the chromosome. This is called excision.

Integration and excision occur by recombination at specific loci on the bacterial and phage DNAs called attachment (att) sites. The attB attachment site on the bacterial chromosome is formally called attλ in bacterial genetics. The locus is defined by mutations that prevent integration of lambda; it is occupied by prophage λ in lysogenic strains. When the attλ site is deleted from the E. coli chromosome, an infecting lambda phage can establish lysogeny by integrating elsewhere, although the efficiency of the reaction is less than 0.1% of the frequency of integration at attλ. This inefficient integration occurs at secondary attachment sites, which resemble the authentic att sequences.

For describing the integration/excision reactions, the bacterial attachment site (attλ) is called attB, consisting of the sequence components BOB′. The attachment site on the phage, attP, consists of the components POP′. FIGURE 13.24 outlines the recombination reaction between these sites. The sequence O is common to attB and attP. It is called the core sequence, and the recombination event occurs within it. The flanking regions B, B′ and P, P′ are referred to as the arms; each is distinct in sequence. The phage DNA is circular, so the recombination event inserts it into the bacterial chromosome as a linear sequence. The prophage is bounded by two new att sites (the products of the recombination) called attL and attR.

FIGURE 13.24 Circular phage DNA is converted to an integrated prophage by a reciprocal recombination between attP and attB; the prophage is excised by reciprocal recombination between attL and attR.

An important consequence of the constitution of the att sites is that the integration and excision reactions do not involve the same pair of reacting sequences. Integration requires recognition between attP and attB, whereas excision requires recognition between attL and attR. The directional character of site-specific recombination is controlled by the identity of the recombining sites.

The recombination event is reversible, but different conditions prevail for each direction of the reaction. This is an important feature in the life of the phage, because it offers a means to ensure that an integration event is not immediately reversed by an excision, and vice versa.

The difference in the pairs of sites reacting at integration and excision is reflected by a difference in the proteins that mediate the two reactions:

  • Integration (attB × attP) requires the product of the phage gene int, which encodes an integrase enzyme, and a bacterial protein called integration host factor (IHF).

  • Excision (attL × attR) requires the product of phage gene xis, in addition to Int and IHF.

Thus, Int and IHF are required for both reactions. Xis plays an important role in controlling the direction; it is required for excision, but inhibits integration.

A similar system, but with somewhat simpler requirements for both sequence and protein components, is found in the bacteriophage P1. The Cre recombinase encoded by the phage catalyzes a recombination between two target sequences. Unlike phage lambda, for which the recombining sequences are different, in phage P1 they are identical. Each consists of a 34-bp-long sequence called loxP. The Cre recombinase is sufficient for the reaction; no accessory proteins are required. As a result of its simplicity and its efficiency, what is now known as the Cre/lox system has been adapted for use in eukaryotic cells, where it has become one of the standard techniques for undertaking site-specific recombination.

13.16 Site-Specific Recombination Involves Breakage and Reunion

The att sites have distinct sequence requirements, and attP is much larger than attB. The function of attP requires a stretch of 240 bp, whereas the function of attB can be exercised by the 23-bp fragment extending from −11 to +11, in which there are only 4 bp on either side of the core. The disparity in their sizes suggests that attP and attB play different roles in the recombination, with attP providing additional information necessary to distinguish it from attB.

Does the reaction proceed by a concerted mechanism in which the strands in attP and attB are cut simultaneously and exchanged? Or, are the strands exchanged one pair at a time, with the first exchange generating a Holliday junction and the second cycle of nicking and ligation occurring to release the structure? The alternatives are depicted in FIGURE 13.25.

FIGURE 13.25 Does recombination between attP and attB proceed by sequential exchange or concerted cutting?

The recombination reaction has been halted at intermediate stages by the use of “suicide substrates,” in which the core sequence is nicked. The presence of the nick interferes with the recombination process. This makes it possible to identify molecules in which recombination has commenced but has not been completed. The structures of these intermediates suggest that exchanges of single strands take place sequentially.

The model illustrated in FIGURE 13.26 shows that if attP and attB sites each suffer the same staggered cleavage, complementary single-stranded ends could be available for crosswise hybridization. The distance between the lambda crossover points is 7 bp, and the reaction generates 3′–phosphate and 5′–OH ends. The reaction is shown for simplicity as generating overlapping single-stranded ends that anneal, but actually occurs by a process akin to the recombination event of Figure 13.4. The corresponding strands on each duplex are cut at the same position, the free 3′ ends exchange between duplexes, the branch migrates for a distance of 7 bp along the region of homology, and then the structure is resolved by cutting the other pair of corresponding strands.

FIGURE 13.26 Staggered cleavages in the common core sequence of attP and attB allow crosswise reunion to generate reciprocal recombinant junctions.

13.17 Site-Specific Recombination Resembles Topoisomerase Activity

Integrases use a mechanism similar to that of type I topoisomerases in which a break is made in one DNA strand at a time. The difference is that a recombinase reconnects the ends crosswise, whereas a topoisomerase makes a break, manipulates the ends, and then rejoins the original ends. The basic principle of the system is that four molecules of the recombinase are required, one to cut each of the four strands of the two duplexes that are recombining.

FIGURE 13.27 shows the nature of the reaction catalyzed by an integrase. The enzyme is a monomeric protein that has an active site capable of cutting and ligating DNA. The reaction involves an attack by a tyrosine on a phosphodiester bond. The 3′ end of the DNA chain is linked through a phosphodiester bond to a tyrosine in the enzyme. This releases a free 5′–OH end.

FIGURE 13.27 Integrases catalyze recombination by a mechanism similar to that of topoisomerases. Staggered cuts are made in DNA and the 3′–phosphate end is covalently linked to a tyrosine in the enzyme. The free hydroxyl group of each strand then attacks the P–Tyr link of the other strand. The first exchange shown in the figure generates a Holliday structure. The structure is resolved by repeating the process with the other pair of strands.

Two enzyme units are bound to each of the recombination sites. At each site, only one of the units attacks the DNA. The symmetry of the system ensures that complementary strands are broken in each recombination site. The free 5′–OH end in each site attacks the 3′–phosphotyrosine link in the other site. This generates a Holliday junction.

The structure is resolved when the other two enzyme units (which had not been involved in the first cycle of breakage and reunion) act on the other pair of complementary strands.

The successive interactions accomplish a conservative strand exchange, in which there are no deletions or additions of nucleotides at the exchange site, and there is no need for input of energy. The transient 3′–phosphotyrosine link between protein and DNA conserves the energy of the cleaved phosphodiester bond.

FIGURE 13.28 shows the reaction intermediate, based on the crystal structure. (Trapping the intermediate was made possible by using a suicide substrate like that described for att recombination, which consists of a synthetic DNA duplex with a missing phosphodiester bond so that the attack by the enzyme does not generate a free 5′–OH end.) The structure of the Cre–lox complex shows two Cre molecules, each of which is bound to a 15-bp length of DNA. The DNA is bent by about 100° at the center of symmetry. Two of these complexes assemble in an antiparallel way to form a tetrameric protein structure bound to two synapsed DNA molecules. Strand exchange takes place in a central cavity of the protein structure that contains the central six bases of the crossover region.

FIGURE 13.28 A synapsed loxA recombination complex has a tetramer of Cre recombinases, with one enzyme monomer bound to each half site. Two of the four active sites are in use, acting on complementary strands of the two DNA sites.

The tyrosine that is responsible for cleaving DNA in any particular half site is provided by the enzyme subunit that is bound to that half site. This is called cis cleavage. This is true also for the Int integrase and XerD recombinase. The FLP recombinase cleaves in trans, however, which involves a mechanism in which the enzyme subunit that provides the tyrosine is not the subunit bound to that half site, but rather is one of the other subunits.

13.18 Lambda Recombination Occurs in an Intasome

Unlike the Cre/lox recombination system, which requires only the enzyme and the two recombining sites, phage lambda recombination occurs in a large structure and has different components for each direction of the reaction (integration versus excision).

The host protein IHF is required for both integration and excision. IHF is a 20-kD protein of two different subunits, which are encoded by the genes himA and himD. IHF is not an essential protein in E. coli and is not required for homologous bacterial recombination. It is one of several proteins with the ability to wrap DNA on a surface. Mutations in the him genes prevent lambda site–specific recombination and can be suppressed by mutations in λint, which suggests that IHF and Int interact. Site-specific recombination can be performed in vitro by Int and IHF.

The in vitro reaction requires supercoiling in attP, but not in attB. When the reaction is performed in vitro between two supercoiled DNA molecules, almost all of the supercoiling is retained by the products. Thus, there cannot be any free intermediates in which strand rotation could occur. This was one of the early hints that the reaction proceeds through a Holliday junction. We now know that the reaction proceeds by the mechanism typical of this class of enzymes, which is related to the topoisomerase I mechanism (see the section in this chapter titled Site-Specific Recombination Resembles Topoisomerase Activity).

Int has two different modes of binding. The C-terminal domain behaves like the Cre recombinase. It binds to inverted sites at the core sequence, positioning itself to make the cleavage and ligation reactions on each strand at the positions illustrated in FIGURE 13.29. The N-terminal domain binds to sites in the arms of attP that have a different consensus sequence. This binding is responsible for the aggregation of subunits into the intasome. The two domains probably bind DNA simultaneously, thus bringing the arms of attP close to the core.

FIGURE 13.29 Int and IHF bind to different sites in attP. The Int recognition sequences in the core region include the sites of cutting.

IHF binds to sequences of about 20 bp in attP. The IHF-binding sites are approximately adjacent to sites where Int binds. Xis binds to two sites located close to one another in attP, so that the protected region extends over 30 to 40 bp. Together, Int, Xis, and IHF cover virtually all of attP. The binding of Xis changes the organization of the DNA so that it becomes inert as a substrate for the integration reaction.

When Int and IHF bind to attP, they generate a complex in which all the binding sites are pulled together on the surface of a protein. Supercoiling of attP is needed for the formation of this intasome. The only binding sites in attB are the two Int sites in the core. Int does not bind directly to attB in the form of free DNA, though. The intasome is the intermediate that “captures” attB, as indicated schematically in FIGURE 13.30.

FIGURE 13.30 Multiple copies of Int protein may organize attP into an intasome, which initiates site-specific recombination by recognizing attB on free DNA.

According to this model, the initial recognition between attP and attB does not depend directly on DNA homology, but instead is determined by the ability of Int proteins to recognize both att sequences. The two att sites then are brought together in an orientation predetermined by the structure of the intasome. Sequence homology becomes important at this stage, when it is required for the strand-exchange reaction.

The asymmetry of the integration and excision reactions is shown by the fact that Int can form a similar complex with attR only if Xis is added. This complex can pair with a condensed complex that Int forms at attL. IHF is not needed for this reaction. A significant difference between lambda integration/excision and the recombination reactions catalyzed by Cre or Flp is that Int-catalyzed reactions bind the regulatory sequences in the arms of the target sites, bending the DNA and allowing interactions between arm and core sites that drive each reaction to its conclusion. This is why each lambda reaction is irreversible, whereas recombination catalyzed by Cre or Flp is reversible. Crystal structures of λ-Int tetramers show that, like other recombinases, the tetramer has two active and two inactive subunits that switch roles during recombination. Allosteric interactions triggered by arm-binding control structural transitions in the tetramer that drive the reaction.

Much of the complexity of site-specific recombination may be caused by the need to regulate the reaction so that integration occurs preferentially when the virus is entering the lysogenic state, whereas excision is preferred when the prophage is entering the lytic cycle. By controlling the amounts of Int and Xis, the appropriate reaction will occur.

13.19 Yeast Can Switch Silent and Active Mating-Type Loci

The yeast S. cerevisiae can propagate in either the haploid or diploid condition. Conversion between these states takes place by mating (fusion of haploid cells to give a diploid) and by sporulation (meiosis of diploids to give haploid spores). The ability to engage in these activities is determined by the mating type of the strain, which can be either a or α. Haploid cells of type a can mate only with haploid cells of type α to generate diploid cells of type a/α. The diploid cells can sporulate to regenerate haploid spores of either type.

Mating behavior is determined by the genetic information present at the MAT locus. Cells that carry the MATa allele at this locus are type a; likewise, cells that carry the MATα allele are type α. Recognition between cells of opposite mating type is accomplished by the secretion of pheromones: α cells secrete the small polypeptide α factor; a cells secrete a factor. A cell of one mating type carries a surface receptor for the pheromone of the opposite type. When an a cell and an α cell encounter one another, their pheromones act on their receptors to arrest the cells in the G1 phase of the cell cycle, and various morphological changes occur (including “schmooing,” in which cells elongate toward each other). In a successful mating, the cell cycle arrest is followed by cell and nuclear fusion to produce an a/α diploid cell.

Mating is a symmetrical process that is initiated by the interaction of pheromone secreted by one cell type with the receptor carried by the other cell type. The only genes that are uniquely required for the response pathway in a particular mating type are those coding for the receptors. Either the a factor–receptor interaction or the α factor–receptor interaction switches on the same response pathway. Mutations that eliminate steps in the common pathway have the same effects in both cell types. The pathway consists of a signal transduction cascade that leads to the synthesis of products that make the necessary changes in cell morphology and gene expression for mating to occur.

Much of the information about the yeast mating-type pathway was deduced from the properties of mutations that eliminate the ability of a and/or α cells to mate. The genes identified by such mutations are called STE (for sterile). Mutations in the genes for the pheromones or receptors are specific for individual mating types, whereas mutations in the other STE genes eliminate mating in both a and α cells. This situation is explained by the fact that the events that follow the interaction of factor with receptor are identical for both types.

Some yeast strains have the remarkable ability to switch their mating types. These strains carry a dominant allele HO and change their mating type frequently—as often as once every generation. Strains with the recessive allele ho have a stable mating type, which is subject to change with a frequency of about 10−6.

The presence of HO causes the genotype of a yeast population to change. Irrespective of the initial mating type, within a very few generations large numbers of cells of both mating types are present, leading to the formation of MATa/MATα diploids that take over the population. The production of stable diploids from a haploid population can be viewed as the raison d’être for switching.

The existence of switching suggests that all cells contain the potential information needed to be either MATa or MATα but express only one type. Where does the information to change mating type come from? Two additional loci are needed for switching. HMLα is needed for switching to give a MATa type; HMRa is needed for switching to give a MATa type. These loci lie on the same chromosome that carries MAT. HML is far to the left and HMR is far to the right.

The mating-type cassette model is illustrated in FIGURE 13.31. It proposes that MAT has an active cassette of either type α or type a. HML and HMR have silent cassettes. In general, HML carries an α cassette, whereas HMR carries an a cassette. All cassettes carry information that encodes mating type, but only the active cassette at MAT is expressed. Mating-type switching occurs when the active cassette is replaced by information from a silent cassette. The newly installed cassette is then expressed.

FIGURE 13.31 Changes of mating type occur when silent cassettes replace active cassettes of the opposite genotype; recombination occurs between cassettes of the same type, and the mating type remains unaltered.

Switching is nonreciprocal; the copy at HML or HMR replaces the allele at MAT. We know this because a mutation at MAT is lost permanently when it is replaced by switching—it does not exchange with the copy that replaces it. This is, in effect, a directed gene-conversion event. The directionality is established by the DSB initiation event, which occurs in the active MAT gene and not in the silent cassettes.

If the silent copy present at HML or HMR is mutated, switching introduces a mutant allele into the MAT locus. The mutant copy at HML or HMR remains there through an indefinite number of switches.

Mating-type switching is a directed event, in which there is only one recipient (MAT), but two potential donors (HML and HMR). Switching usually involves replacement of MATa by the copy at HMLα or replacement of MATα by the copy at HMRa. In 80% to 90% of switches, the MAT allele is replaced by one of the opposite type. This is determined by the phenotype of the cell. Cells of a phenotype preferentially choose HML as donor; cells of α phenotype preferentially choose HMR.

Several groups of genes are involved in establishing and switching mating type. In addition to the genes that directly determine mating type, they include genes needed to repress the silent cassettes, to switch mating type, or to execute the functions involved in mating, and, most important, the homologous recombination factors described earlier in this chapter.

By comparing the sequences of the two silent cassettes (HMLα and HMRa) with the sequences of the two types of active cassettes (MATa and MATα), the sequences that determine mating type can be delineated. The organization of the mating-type loci is summarized in FIGURE 13.32. Each cassette contains common sequences that flank a central region that differs in the a and α types of cassette (called Y a or Yα). On either side of this region, the flanking sequences are virtually identical, although they are shorter at HMR. The active cassette at MAT is transcribed from a promoter within the Y region.

FIGURE 13.32 Silent cassettes have the same sequences as the corresponding active cassettes, except for the absence of the extreme flanking sequences in HMRa. Only the Y region changes between a and α types.

13.20 Unidirectional Gene Conversion Is Initiated by the Recipient MAT Locus

A switch in mating type is accomplished by a gene conversion in which the recipient site (MAT) acquires the sequence of the donor type (HML or HMR). Sites needed for the recombination have been identified by mutations at MAT that prevent switching. The unidirectional nature of the process is indicated by lack of mutations in HML or HMR.

The mutations identify a site at the right boundary of Y at MAT that is crucial for the switching event. The nature of the boundary is shown by analyzing the locations of these point mutations relative to the site of switching (this is done by examining the results of rare switches that occur in spite of the mutation). Some mutations lie within the region that is replaced (and thus disappear from MAT after a switch), whereas others lie just outside the replaced region (and therefore continue to impede switching). Thus, sequences both within and outside the replaced region are needed for the switching event.

Switching is initiated by a DSB close to the Y–Z boundary that coincides with a site that is sensitive to attack by DNase. (This is a common feature of chromosomal sites that are involved in initiating transcription or recombination.) It is recognized by the endonuclease encoded by the HO locus. The HO endonuclease makes a staggered DSB just to the right of the Y boundary. Cleavage generates the single-stranded ends of four bases illustrated in FIGURE 13.33. The nuclease does not attack mutant MAT loci that cannot switch. Deletion analysis shows that most or all of the sequence of 24 bp surrounding the Y junction is required for cleavage in vitro. The recognition site is relatively large for an endonuclease, and it occurs only at the three mating-type cassettes.

FIGURE 13.33 HO endonuclease cleaves MAT just to the right of the Y region, which generates sticky ends with a 4-base overhang.

Only the MAT locus, and not the HML or HMR locus, is a target for the endonuclease. It seems plausible that the same mechanisms that keep the silent cassettes from being transcribed also keep them inaccessible to the HO endonuclease. This inaccessibility ensures that switching is unidirectional.

The reaction triggered by the cleavage is illustrated schematically in FIGURE 13.34 in terms of the general reaction between donor and recipient regions. The recombination occurs through an SDSA mechanism, as described earlier. As expected, the stages following the initial cut require the enzymes involved in general recombination. Mutations in some of these genes prevent switching. In fact, studies of switching at the MAT locus were important in the development of the SDSA model.

FIGURE 13.34 Cassette substitution is initiated by a double-strand break in the recipient (MAT) locus and may involve pairing on either side of the Y region with the donor (HMR or HML) locus.

13.21 Antigenic Variation in Trypanosomes Uses Homologous Recombination

The single-celled parasites known as trypanosomes cause two major types of human disease: African sleeping sickness (human African trypanosomiasis) and Chagas disease. These organisms are able to evade the host immune response through a process known as antigenic variation, in which expression of the major surface antigen is altered in a cyclical pattern in response to immune pressure. The variant surface glycoprotein (VSG) of trypanosomes is the major target of the immune system, but once antibodies are present to a given VSG trypanosomes are able to switch expression to one of the many hundreds of VSG genes in their genomes. The VSG genes are organized into multiple subtelomeric tandem arrays and are also located in telomeric arrays on minichromosomes. Although all the genes in these arrays are silenced, they are either intact genes or pseudogenes. The switch is controlled by a recombination event in which a silent VSG gene is moved to a transcriptionally active, subtelomeric site known as an expression site (ES). This is illustrated in FIGURE 13.35. Twenty subtelomeric expression sites have been identified, but only one of these is actively transcribed at a time. The transcriptionally active ES is thought to be a hotspot for recombination due to the open chromatin in this region. In fact, VSG recombination occurs at a higher frequency than would be expected for random events, leading to a VSG switch rate ranging from 10−2 to 10−3 switch events per cell per generation. Segmental gene-conversion events using different VSGs can create chimeric VSG genes at the active expression site that contain sequences from multiple donor VSG genes.

FIGURE 13.35 Switching mechanisms in trypanosome antigenic variation. Most of the VSG genes are arranged in arrays in subtelomeric locations and consist of silent complete genes and pseudogenes. Gene conversion of the active VSG gene using information from one of the silent genes in the arrays results in a change in the sequence information in the active gene and a change in the surface antigen of the trypanosome. A second mode of variation comes from telomere exchange, to switch an inactive telomeric VSG gene from minichromosomes to the site of the active VSG gene. Both mechanisms use homologous recombination factors, but the precise mechanism of exchange is not known.

Reprinted from Trends Genet., vol. 22, J. E. Taylor and G. Rudenko, Switching trypanosome coats …, pp. 614–620. Copyright 2006, with permission from Elsevier [http://www.sciencedirect.com/science/journal/01689525].

DNA rearrangement through gene conversion, telomere exchange, and other unidentified processes is responsible for replacing an inactive VSG allele for the one in the active ES. The gene-conversion event results in a duplication of the inactive VSG gene at the active ES locus, allowing for expression of the previously inactive VSG. Despite the specificity of the genomic loci involved in the VSG-switching event itself, the process has been shown to depend on general recombination factors.

Trypanosome mutants that do not express Rad51 are greatly impaired in VSG switching, indicating that homologous recombination is essential for this process. Further work has demonstrated a role for the trypanosome homologue of BRCA2 in VSG switching. It is unclear whether enzymes specific to VSG switch recombination are involved in this process as well. Despite the fact that gene conversion is required for VSG switching, defects in mismatch repair pathway genes in trypanosomes do not affect antigenic variation.

13.22 Recombination Pathways Adapted for Experimental Systems

Site-specific recombination not only has important biological roles, as discussed earlier, but has also been exploited to create targeted recombination events in experimental systems. Two classic examples of site-specific recombination have been adapted for experimental use: the Cre/lox and FLP/FRT systems.

The Cre/lox system is derived from bacteriophage P1. The Cre enzyme recognizes and cleaves lox sites. One of the most common uses of the Cre/lox system is in gene targeting in mice, as shown in FIGURE 13.36. Cre/lox can be used to conditionally turn off or turn on a gene in mice. A construct is designed that is flanked by lox sites, with the Cre gene under control of an inducible promoter that can be turned on by temperature, hormones, or in a tissue-specific pattern. Expression of Cre results in production of the Cre protein; the Cre protein then recognizes and cleaves the lox sites and promotes rejoining of the cut lox sites to leave behind a single lox site, with the material between the lox sites having been excised.

FIGURE 13.36 Using Cre/lox to make cell type–specific gene knockouts in mice. loxP sites are inserted into the chromosome to flank exon 2 of the gene X. The second copy of the X gene has been knocked out. The mouse formed with this construct is called the loxP mouse. Another mouse, called the Cre mouse, has the cre gene inserted into the genome. Adjacent to the cre gene is a promoter that directs expression of the cre gene only in certain cell types or in response to certain conditions. This mouse also carries a knockout of one copy of gene X. When the two mice are crossed, progeny that carry the loxP construct, the gene X knockout, and the cre gene are produced. When Cre protein is expressed in cells that activate the promoter, it catalyzes site-specific recombination between the loxP sites, and exon 2 of gene X is deleted. This inactivates the one functional copy of gene X in those cells expressing Cre.

Data from H. Lodish, et al. Molecular Cell Biology, Fifth edition. W. H. Freeman & Company, 2003.

The Cre/lox system can be used to conditionally remove an exon from a mouse gene, resulting in a gene knockout (see the chapter titled Methods in Molecular Biology and Genetic Engineering), or it can fuse the gene of interest to a promoter and thereby control expression of the gene of interest. Expression of a gene in tissues where it is not normally expressed or at a time when the gene is not normally expressed is called ectopic expression. Ectopic expression studies can reveal information about gene redundancy, specificity, and cell autonomy.

Another system that has been adapted for experimental use is derived from the yeast S. cerevisiae. The 2-micron yeast plasmid is an autonomously replicating episome that is present in high copy numbers. The plasmid, which has no apparent benefit to the cell, is amplified through a site-specific recombination reaction that is carried out by a specialized recombinase known as Flp (flip). Flp recognizes inverted repeat sequences known as FRT (Flp recombinase target) sites. During replication, Flp-mediated recombination promotes rolling-circle replication that results in amplification of the 2-micron plasmid. The Flp/FRT system is used in Drosophila to induce site-specific mitotic recombination events that can be used to create homozygous mutations or to make conditional knockouts, as shown in FIGURE 13.37.

FIGURE 13.37 Using Flp/FRT to make homozygous recessive cells by homologous recombination. A fly is heterozygous for a mutant gene and homozygous insertion of the FRT site on the same chromosome. Induction of the Flp gene allows the FLP recombinase protein to be made. Flp recognizes the FRT site and makes a double-strand break, which promotes homologous recombination. Some of the recombination events occur by the double-strand break repair mechanism and result in crossing over. Following chromosome segregation, one daughter cell receives two mutant copies of the gene and the other daughter cell receives two normal copies of the gene. In the example shown, a patch of mutant cells is formed on the wing of a Drosophila. This technique allows assessment of a recessive mutant phenotype at a late stage in development.

Data from B. Alberts, et al. Molecular Biology of the Cell, Fourth edition. Garland Science, 2002.

To use the Flp/FRT system in Drosophila, FLP gene expression is regulated. When Flp is expressed, it cuts the FRT sites, which have been inserted on a chromosome where there is a gene of interest centromere-distal to the FRT site. The cutting of the FRT site, which is not 100% efficient, induces a DSB at the FRT site. The DSBs are repaired by homologous recombination, and some of them will result in crossing over. Depending on how the chromosomes then segregate, some cells will now be homozygous for the mutant gene. In genetic studies, the chromosome is often marked by a gene that affects a pigment, to give a visual readout for the recombination. The mitotic recombination uncovers the recessive pigmentation mutation and the mutant gene of interest, making them homozygous recessive. One use of this system is to see the effects of a lethal recessive mutation: When the zygote is homozygous recessive, the mutation will be lethal. If it is carried in the heterozygous state, though, the organism will be viable. Then the gene is rendered homozygous in clones of cells by induction of Flp, either by temperature or tissue-specific transcription regulation, enabling the investigator to ask about the effects of loss of the gene in specific cells at a specific time during development.

In recent years, Flp/FRT has been further adapted to construct recyclable selectable marker cassettes. In these systems, a selectable marker is placed between two flanking FRT sites. Also contained within the cassette is the FLP gene under the control of a regulatable promoter. Targeted integration of the FLP/FRT cassette is used to replace a locus of interest with the FLP marker cassette. Following integration, induced expression of the Flp recombinase catalyzes recombination between the flanking FRT sites, resulting in excision of the selectable marker cassette. This recyclable marker strategy is advantageous in diploid organisms because it allows for sequential rounds of targeted integration to make homozygous deletions of a gene of interest.

Summary

Recombination is initiated by a double-strand break (DSB) in DNA. The break is enlarged to a gap with a single-stranded end. The free single-stranded end then forms a heteroduplex with the allelic sequence. Correction events may occur at sites that are mismatched within the heteroduplex DNA. The DNA in which the break occurs actually incorporates the sequence of the chromosome that it invades, so the initiating DNA is called the recipient. Gap repair, using the donor genetic information to repair the gap in the recipient DNA molecule, can also result in a gene-conversion event. Hotspots for recombination are sites where DSBs are initiated. A gradient of gene conversion is determined by the likelihood that a sequence near the free end will be converted to a single strand; this decreases with distance from the break. After gap repair, if the invading strain disengages from the recombination intermediate and anneals with the other end of the break, only gene conversion occurs. This is called the synthesis-dependent strand-annealing (SDSA) model. If instead the second end of the break is captured into the recombination intermediate, two Holliday junctions are formed. Resolution of the Holliday junctions can give crossover products if resolved in the appropriate direction. Recombination initiated by a DSB and processed to yield a double Holliday junction intermediate is called double-strand break repair (DSBR).

Meiotic recombination is initiated in yeast by Spo11, a topoisomerase-like enzyme that creates DSBs and becomes linked to the free 5′ ends of DNA. The DSB is then processed by generating single-stranded DNA that can anneal with its complement in the other chromosome. Yeast mutations that block synaptonemal complex formation show that recombination is required for its formation. Formation of the synaptonemal complex may be initiated by DSBs, and it may persist until recombination is completed. Mutations in components of the synaptonemal complex block its formation but do not prevent chromosome pairing, so homolog recognition is independent of recombination and synaptonemal complex formation.

The full set of reactions required for recombination can be undertaken by the Rec and Ruv proteins of E. coli. A single-stranded region with a free end is generated by the RecBCD nuclease. The enzyme binds to DNA on one side of a chi sequence and then moves to the chi sequence, unwinding DNA as it progresses. A single-strand break is made at the chi sequence. chi sequences provide hotspots for recombination. The single strand provides a substrate for RecA, which has the ability to synapse homologous DNA molecules by sponsoring a reaction in which a single strand from one molecule invades a duplex of the other molecule. Heteroduplex DNA is formed by displacing one of the original strands of the duplex. These actions create a recombination junction, which is resolved by the Ruv proteins. RuvA and RuvB act at a heteroduplex, and RuvC cleaves Holliday junctions.

The enzymes involved in site-specific recombination have actions related to those of topoisomerases. Among this general class of recombinases, those concerned with phage integration form the subclass of integrases. The Cre/lox system uses two molecules of Cre to bind to each lox site, so that the recombining complex is a tetramer. This is one of the standard systems for inserting DNA into a foreign genome. Phage lambda integration requires the phage Int protein and host IHF protein and involves a precise breakage and reunion in the absence of any synthesis of DNA. The reaction involves wrapping of the attP sequence of phage DNA into the nucleoprotein structure of the intasome, which contains several copies of Int and IHF; the host attB sequence is then bound and recombination occurs. Reaction in the reverse direction requires the phage protein Xis. Some integrases function by cis-cleavage, where the tyrosine that reacts with DNA in a half site is provided by the enzyme subunit bound to that half site; others function by trans-cleavage, for which a different protein subunit provides the tyrosine.

The yeast S. cerevisiae can propagate in either the haploid or diploid condition. Conversion between these states takes place by mating (fusion of haploid cells to give a diploid) and by sporulation (meiosis of diploids to give haploid spores). The ability to engage in these activities is determined by the mating type of the strain. The mating type is determined by the sequence of the MAT locus and can be changed by a recombination event that substitutes a different sequence at this locus. The recombination event is initiated by a DSB—such as a homologous recombination event—but then the subsequent events ensure a unidirectional replacement of the sequence at the MAT locus.

Replacement is regulated so that MATa is usually replaced by the sequence from HMLα, whereas MATα is usually replaced by the sequence from HMRa. The endonuclease HO triggers the reaction by recognizing a unique target site at MAT. HO is regulated at the level of transcription by a system that ensures its expression in mother cells but not daughter cells, with the consequence that both progeny have the same (new) mating type.

Homologous recombination is also essential for the process of antigenic variation in trypanosomes. Recombination is required to switch inactive VSG genes into active VSG expression sites. The molecular mechanisms behind this phenomenon are not completely understood, but it is clear that it does not involve non-homologous end-joining (NHEJ) or mismatch repair enzymes. Rad51 is essential for this process, indicating the importance of homologous recombination.

Recombination pathways have been exploited as experimental tools for generation of gene knockouts and other recombination-mediated events. Two major examples of these experimental tools include the Cre/lox and Flp/FRT systems. Both tools rely on site-specific recombination to create targeted recombination events in experimental systems.

References

13.2 Homologous Recombination Occurs Between Synapsed Chromosomes in Meiosis

Reviews

  1. Brachet, E., Sommermeyer, V., and Borde, V. (2011). Interplay between modifications of chromatin and meiotic recombination hotspots. Biol. Cell 104, 51–69.

  2. Hunter, N. (2015). Meiotic recombination: the essence of heredity. Cold Spring Harb. Perspect. Biol. 7:a016618.

  3. Phadnis, N., Hyppa, R. W., and Smith, G. R. (2011). New and old ways to control meiotic recombination. Trends Genet. 27, 411–421.

13.3 Double-Strand Breaks Initiate Recombination

Reviews

  1. Lichten, M., and Goldman, A. S. (1995). Meiotic recombination hotspots. Annu. Rev. Genet. 29, 423–444.

  2. Szostak, J. W., Orr-Weaver, T. L., Rothstein, R. J., and Stahl, F. W. (1983). The double-strand-break repair model for recombination. Cell 33, 25–35.

Research

  1. Hunter, N., and Kleckner, N. (2001). The single-end invasion: an asymmetric intermediate at the double-strand break to double-Holliday junction transition of meiotic recombination. Cell 106, 59–70.

13.5 The Synthesis-Dependent Strand-Annealing Model

Review

  1. Paques, F., and Haber, J. E. (1999). Multiple pathways of recombination induced by double-strand breaks in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 63, 349–404.

Research

  1. Ferguson, D. O., and Holloman, W. K. (1996). Recombinational repair of gaps in DNA is asymmetric in Ustilago maydis and can be explained by a migrating D-loop model. Proc. Natl. Acad. Sci. USA 93, 5419–5424.

  2. Keeney, S., and Neale, M. J. (2006). Initiation of meiotic recombination by formation of DNA double-strand breaks: mechanism and regulation. Biochem. Soc. Trans. 34, 523–525.

  3. Nassif, N., Penney, J., Pal, S., Engels, W. R., and Gloor, G. B. (1994). Efficient copying of nonhomologous sequences from ectopic sites via P-element-induced gap repair. Mol. Cell Biol. 14, 1613–1625.

13.6 The Single-Strand Annealing Mechanism Functions at Some Double-Strand Breaks

Research

  1. Ivanov, E. L., Sugawara, N., Fishman-Lobell, J., and Haber, J. E. (1996). Genetic requirements for the single-strand annealing pathway of double-strand break repair in Saccharomyces cerevisiae. Genetics 142, 693–704.

13.7 Break-Induced Replication Can Repair Double-Strand Breaks

Reviews

  1. Kraus, E., Leung, W. Y., and Haber, J. E. (2001). Break-induced replication: a review and an example in budding yeast. Proc. Natl. Acad. Sci. USA 98, 8255–8262.

  2. Llorente, B., Smith, C. E., and Symington, L. S. (2008). Break-induced replication: what is it and what is it for? Cell Cycle 7, 859–864.

13.8 Recombining Meiotic Chromosomes Are Connected by the Synaptonemal Complex

Reviews

  1. Roeder, G. S. (1997). Meiotic chromosomes: it takes two to tango. Genes Dev. 11, 2600–2621.

  2. Zickler, D., and Kleckner, N. (1999). Meiotic chromosomes: integrating structure and function. Annu. Rev. Genet. 33, 603–754.

Research

  1. Blat, Y., and Kleckner, N. (1999). Cohesins bind to preferential sites along yeast chromosome III, with differential regulation along arms versus the central region. Cell 98, 249–259.

  2. Dong, H., and Roeder, G. S. (2000). Organization of the yeast Zip1 protein within the central region of the synaptonemal complex. J. Cell Biol. 148, 417–426.

  3. Klein, F., Mahr, P., Galova, M., Buonomo, S. B., Michaelis, C., Nairz, K., and Nasmyth, K. (1999). A central role for cohesins in sister chromatid cohesion, formation of axial elements, and recombination during yeast meiosis. Cell 98, 91–103.

  4. Sym, M., Engebrecht, J. A., and Roeder, G. S. (1993). ZIP1 is a synaptonemal complex protein required for meiotic chromosome synapsis. Cell 72, 365–378.

13.9 The Synaptonemal Complex Forms After Double-Strand Breaks

Reviews

  1. McKim, K. S., Jang, J. K., and Manheim, E. A. (2002). Meiotic recombination and chromosome segregation in Drosophila females. Annu. Rev. Genet. 36, 205–232.

  2. Petes, T. D. (2001). Meiotic recombination hot spots and cold spots. Nat. Rev. Genet. 2, 360–369.

Research

  1. Allers, T., and Lichten, M. (2001). Differential timing and control of noncrossover and crossover recombination during meiosis. Cell 106, 47–57.

  2. Weiner, B. M., and Kleckner, N. (1994). Chromosome pairing via multiple interstitial interactions before and during meiosis in yeast. Cell 77, 977–991.

13.11 The Bacterial RecBCD System Is Stimulated by chi Sequences

Research

  1. Dillingham, M. S., Spies, M., and Kowalczykowski, S. C. (2003). RecBCD enzyme is a bipolar DNA helicase. Nature 423, 893–897.

  2. Spies, M., Bianco, P. R., Dillingham, M. S., Handa, N., Baskin, R. J., and Kowalczykowski, S. C. (2003). A molecular throttle: the recombination hotspot chi controls DNA translocation by the RecBCD helicase. Cell 114, 647–654.

  3. Taylor, A. F., and Smith, G. R. (2003). RecBCD enzyme is a DNA helicase with fast and slow motors of opposite polarity. Nature 423, 889–893.

13.12 Strand-Transfer Proteins Catalyze Single-Strand Assimilation

Reviews

  1. Kowalczykowski, S. C., Dixon, D. A., Eggleston, A. K., Lauder, S. D., and Rehrauer, W. M. (1994). Biochemistry of homologous recombination in Escherichia coli. Microbiol. Rev. 58, 401–465.

  2. Kowalczykowski, S. C., and Eggleston, A. K. (1994). Homologous pairing and DNA strand-exchange proteins. Annu. Rev. Biochem. 63, 991–1043.

  3. Lusetti, S. L., and Cox, M. M. (2002). The bacterial RecA protein and the recombinational DNA repair of stalled replication forks. Annu. Rev. Biochem. 71, 71–100.

13.13 Holliday Junctions Must Be Resolved

Reviews

  1. Lilley, D. M., and White, M. F. (2001). The junction-resolving enzymes. Nat. Rev. Mol. Cell Biol. 2, 433–443.

  2. West, S. C. (1997). Processing of recombination intermediates by the RuvABC proteins. Annu. Rev. Genet. 31, 213–244.

Research

  1. Boddy, M. N., Gaillard, P. H., McDonald, W. H., Shanahan, P., Yates, J. R., and Russell, P. (2001). Mus81-Eme1 are essential components of a Holliday junction resolvase. Cell 107, 537–548.

  2. Chen, X. B., Melchionna, R., Denis, C. M., Gaillard, P. H., Blasina, A., Van de Weyer, I., Boddy, M. N., Russell, P., Vialard, J., and McGowan, C. H. (2001). Human Mus81-associated endonuclease cleaves Holliday junctions in vitro. Mol. Cell 8, 1117–1127.

  3. Constantinou, A., Davies, A. A., and West, S. C. (2001). Branch migration and Holliday junction resolution catalyzed by activities from mammalian cells. Cell 104, 259–268.

  4. Kaliraman, V., Mullen, J. R., Fricke, W. M., Bastin-Shanower, S. A., and Brill, S. J. (2001). Functional overlap between Sgs1-Top3 and the Mms4-Mus81 endonuclease. Genes Dev. 15, 2730–2740.

13.14 Eukaryotic Genes Involved in Homologous Recombination

Reviews

  1. Kowalczykowski, S. C. (2015). An overview of the molecular mechanisms of recombinational DNA repair. Cold Spring Harb. Perspect. Biol. 7:a016410.

  2. Krogh, B. O., and Symington, L. S. (2004). Recombination proteins in yeast. Annu. Rev. Genet. 38, 233–271.

  3. San Filippo, J., Sung, P., and Klein, H. (2008). Mechanism of eukaryotic homologous recombination. Annu. Rev. Biochem. 77, 229–257.

  4. Sung, P., and Klein, H. (2006). Mechanism of homologous recombination: mediators and helicases take on regulatory functions. Nat. Rev. Mol. Cell Biol. 7, 739–750.

Research

  1. Gravel, S., Chapman, J. R., Magill, C., and Jackson, S. P. (2008). DNA helicases Sgs1 and BLM promote DNA double-strand break resection. Genes Dev. 22, 2767–2772.

  2. Hollingsworth, N. M., and Brill, S. J. (2004). The Mus81 solution to resolution: generating meiotic crossovers without Holliday junctions. Genes Dev. 18, 117–125.

  3. Ip, S. C., Rass, U., Blanco, M. G., Flynn, H. R., Skehel, J. M., and West, S. C. (2008). Identification of Holliday junction resolvases from humans and yeast. Nature 456, 357–361.

  4. Mimitou, E. P., and Symington, L. S. (2008). Sae2, Exo1 and Sgs1 collaborate in DNA double-strand break processing. Nature 455, 770–774.

  5. Zhu, Z., Chung, W. H., Shim, E.Y., Lee, S. E., and Ira, G. (2008). Sgs1 helicase and two nucleases Dna2 and Exo1 resect DNA double-strand break ends. Cell 134, 981–994.

13.15 Specialized Recombination Involves Specific Sites

Review

  1. Craig, N. L. (1988). The mechanism of conservative site-specific recombination. Annu. Rev. Genet. 22, 77–105.

Research

  1. Metzger, D., Clifford, J., Chiba, H., and Chambon, P. (1995). Conditional site-specific recombination in mammalian cells using a ligand-dependent chimeric Cre recombinase. Proc. Natl. Acad. Sci. USA 92, 6991–6995.

  2. Nunes-Duby, S. E., Kwon, H. J., Tirumalai, R. S., Ellenberger, T., and Landy, A. (1998). Similarities and differences among 105 members of the Int family of site-specific recombinases. Nucleic Acids Res. 26, 391–406.

13.17 Site-Specific Recombination Resembles Topoisomerase Activity

Research

  1. Guo, F., Gopaul, D. N., and van Duyne, G. D. (1997). Structure of Cre recombinase complexed with DNA in a site-specific recombination synapse. Nature 389, 40–46.

13.18 Lambda Recombination Occurs in an Intasome

Research

  1. Biswas, T., Aihara, H., Radman-Livaja, M., Filman, D., Landy, A., and Ellenberger, T. (2005). A structural basis for allosteric control of DNA recombination by lambda integrase. Nature 435, 1059–1066.

  2. Wojciak, J. M., Sarkar, D., Landy, A., and Clubb, R. T. (2002). Arm-site binding by lambda integrase: solution structure and functional characterization of its amino-terminal domain. Proc. Natl. Acad. Sci. USA 99, 3434–3439.

13.21 Antigenic Variation in Trypanosomes Uses Homologous Recombination

Review

  1. Taylor, J. E., and Rudenko, G. (2006). Switching trypanosome coats: what’s in the wardrobe? Trends Genet. 22, 614–620.

Research

  1. Machado-Silva, A., Teixeira, S. M., Franco, G. R., Macedo, A. M., Pena, S. D., McCulloch, R., and Machado, C. R. (2008). Mismatch repair in Trypanosoma brucei: heterologous expression of MSH2 from Trypanosoma cruzi provides new insights into the response to oxidative damage. Gene 411, 19–26.

  2. Proudfoot, C., and McCulloch, R. (2005). Distinct roles for two RAD51-related genes in Trypanosoma brucei antigenic variation. Nucleic Acids Res. 33, 6906–6919.

13.22 Recombination Pathways Adapted for Experimental Systems

Research

  1. Egli, D., Hafen, E., and Schaffner, W. (2004). An efficient method to generate chromosomal rearrangements by targeted DNA double-strand breaks in Drosophila melanogaster. Genome Res. 14, 1382–1393.

  2. Le, Y., and Sauer, B. (2001). Conditional gene knockout using Cre recombinase. Mol. Biotechnol. 17, 269–275.