Introduction

Meiosis is a specialized type of cell division in which a prior round of DNA replication is followed by two successive rounds of chromosome segregation. The resulting cells have half of the original chromosomal complement. During fertilization, these cells fuse to reconstitute the original complement thereby completing the sexual cycle. In most eukaryotes, a feature of the first meiotic division (MI) is crossing over, in which homologous chromosomes interact physically, exchanging material between the paternal and maternal copies¹. Crossing over provides a connection between homologs that is required in most organisms for the accurate segregation of chromosomes during MI (ref. 2). Chromosomes that fail to have at least one crossover (CO) often segregate aberrantly, resulting in aneuploidy. In addition, crossing over generates genetic diversity by creating new combinations of paternal and maternal alleles³.

Cells undergoing meiosis regulate recombination at multiple levels. To ensure that each chromosome pair is physically connected, COs are distributed among chromosomes nonrandomly such that each chromosome pair typically undergoes at least one CO even if the total number of COs per chromosome is less^{4, 5}. The distribution of COs along chromosomes is also tightly regulated. In most organisms, COs are distributed such that one CO event inhibits the chances of another nearby event⁶. The term used to describe this phenomenon, first observed by Drosophila melanogaster researchers at the beginning of the twentieth century, is CO interference^{7, 8}.

One of the challenges of studying CO interference is that it is a probabilistic phenomenon of populations. CO interference does not result in a complete lack of closely spaced COs, but instead reduces the proportion of these events in a population compared with what one would expect if the events were randomly positioned⁹. For this reason, making statistically significant claims often requires large data sets. Patterns of CO interference vary widely from organism to organism with regards to strength versus distance relationships and in some organisms there appear to be both interfering and noninterfering COs^{10, 11, 12, 13}. Even within organisms, parameters such as sex and chromosomal location have been shown to have a profound effect on CO interference^{14, 15}.

Using Arabidopsis thaliana as a model, we present here a fluorescent pollen tetrad method for quick and relatively easy production of data sets that can be used to analyze CO interference. Using this method, one can assay almost any chromosomal region, customize interval size and examine the effects of mutants and experimental treatments on CO interference. This method has been used to assay wild-type levels of CO interference on a region of chromosome 5 and also to determine a small but detectable difference in CO interference between wild-type plants and Atmus81 mutant plants^{16, 17}.

Fluorescent tetrad analysis in A. thaliana

Our lab has developed a unique pollen-based visual assay for meiotic recombination in A. thaliana that can be used to generate large data sets amenable to interference analysis quickly^{16, 17}. A. thaliana quartet (qrt) mutants produce pollen tetrads in which the four meiotic products are held together, allowing all products of a single meiotic event to be studied relative to one another^{41, 42}. This assay system is based on a collection of transgenic lines, in the qrt background, each carrying a gene encoding either a red, cyan or yellow fluorescent protein that can be excited by different wavelengths of light. Expression of these markers is directed by a postmeiotic pollen-specific promoter (LAT52)⁴³. Researchers can construct visually assayable genetic intervals by crossing lines that carry linked markers. Lines carrying two or more marker genes on the same chromosome expressing differently colored proteins produce tetrads that segregate the marker genes (and thus the proteins they encode) in the pollen tetrads in patterns that reflect whether or not a recombination event has happened between them. This system, can detect CO events directly in the pollen grains, and through the construction of double intervals delineated by three colors, can be used to measure CO interference throughout the A. thaliana genome.

We have constructed a genome-wide library of single-insertion fluorescent-tagged lines (FTLs) by transforming A. thaliana (Col) with these marker genes¹⁶. All the markers shown in Figure 1 map outside genes. There are currently 113 FTLs available including 35 DsRed2, 41 eYFP and 37 eCFP. Combinations of these three groups can be crossed to create lines with three different markers. Pollen from plants that are heterozygous for these markers is viewed using an epi-fluorescence microscope with three different filters. Such a set of three linked markers on chromosome 5 has been used by our lab to detect subtle differences in interference between wild-type and mus81 mutant plants¹⁷.

Figure 1: Map of fluorescent transgenes.

The chromosomal (green bars) insertion site of the transgene carried by each fluorescent-tagged line (FTL) is indicated by a red (DsRed2), yellow (eYFP) or cyan (eCFP) circle. The identification number of each insertion is given above the circle. The genetic intervals (I1a, I1b, I1c, I2a, I2b, I2c, I3a, I3b, I3c, I3d, I5a, I5b, I5c and I5d) that are available by request from G.P.C. are delineated by brackets. The Arabidopsis Information Resource (TAIR) 'chromosome map tool' (http://www.arabidopsis.org/) was used to place T-DNA insertion points on the map. T-DNA, transferred DNA.

Full size image (97 KB)

Materials

Reagents

• A. thaliana qrt1 (Col) seeds (Arabidopsis Biological Resource Center)
• A. thaliana FTL seeds (available from G.P.C.)
• (Optional) A. thaliana seeds heterozygous for mutant(s) of interest
• Metro-mix 400 (Sun Gro Horticulture Inc)
• Sucrose
• 1 M CaCl₂
• 0.5 M Boric acid
• Triton-X detergent (Fisher Scientific, cat. no. BP151-100)
• PGM screening solution (see REAGENT SETUP)

Equipment

• Fluorescence microscope (Nikon Eclipse E1000; Nikon) (see EQUIPMENT SETUP)
• eCFP, eYFP and DsRed2 filters (Chroma Technologies)
• X-Cite 120 Illumination system (EXFO)
• 75 25 mm², 1-mm thick glass microscope slides (Okando via Genesee Scientific)
• 18 18 mm²-cover glass (Fisher Scientific)
• 5–50-l Finn pipette (Thermo Electron)
• Pipette tips
• Fine point stainless steel tweezers (Biomedical Research Instruments)
• Laboratory counter (Fisher Scientific)
• Soft cotton cloth

Reagent setup

• PGM screening solution To 950 ml H₂O, add 170 g sucrose (17% wt/vol), 2 ml 1 M CaCl₂ (2 mM), 3.25 ml 0.5 M boric acid (1.625 mM) and 100 l Triton-X (0.1% vol/vol).

Equipment setup

• Fluorescence microscope A Nikon Eclipse E1000 epi-fluorescence microscope equipped with a motorized filter set including eCFP, DsRed2 and eYFP filters was used to develop these methods. No special setup of the microscope is necessary. Any epi-fluorescent microscope equipped with these filters will allow the researcher to execute this protocol, but we recommend the use of motorized filters.

Procedure

1. Points from here (point 1) up to and including point 1 are related to Selection of appropriate FTLsFrom the FTL insert library, select the appropriate lines that will define two adjacent intervals (5 cM per Mb is a reasonable estimate for the average genetic-to-physical ratio in A. thaliana).
   Critical step One line for each of the eCFP, DsRed2 and eYFP markers needs to be selected, but they may occur in any order. The sizes of the two intervals do not have to be similar, but it will slightly complicate analysis if either of the intervals is large enough such that an abundance of NPD tetrads in either single interval is observed. Note that for the remainder of the protocol, the line that carries an insert closest to the North end of the chromosome will be referred to as 'A', the next will be 'B' and the line with an insert closest to the South end will be 'C'.
2. Points from here (point 2) up to and including point 6 are related to Construction of first interval (I1)Plant at least 12 seeds each from line A and line B (seeds will usually be heterozygous for the interval). Grow plants on Metro-mix 400 soil in an environmentally controlled growth chamber with a long photoperiod (16 h light at 20 °C/8 h dark at 20 °C).
   Critical step Steps 2–11 can be omitted if the researcher wishes to use interference intervals that our lab has already constructed (see Fig. 1). These lines will be supplied as a segregating population of seeds from an ABC/+++ parent and are available by request from G.P.C.
   Critical step If the study includes analysis of interference in mutant backgrounds, it will save time to cross the mutants into the qrt1 background during the construction of the three-color interval. We recommend maintaining and crossing meiotic mutants in the heterozygous state to avoid genomic abnormalities. Useful plants will have the genotype mutant/+; qrt1/qrt1.
3. As each marker will be segregating 1:2:1 in this generation, it is necessary to determine the marker genotype of each plant by examining the pollen under the epi-fluorescence microscope when plants begin flowering (4–5 weeks). As this step is diagnostic and does not require any counting, use the simplified screening process outlined in Box 1. Plants homozygous (4:0 fluorescent pollen, see Fig. 2) for their respective markers will be used for the next crossing step. Collect seeds from 2:2 plants (A/+ and B/+) and 4:0 plants (A/A and B/B) to save as stocks.

   Figure 2: Single locus segregation patterns in pollen tetrads.

   

   A locus with a transgenic marker construct encoding a fluorescent protein (in this case DsRed2) can be homozygous (A/A) or hemizygous (A/+) for the marker or it can be wild type (+/+). The fluorescence signal in the pollen tetrads will reflect the marker genotype and yield 4:0 (left) 2:2 (middle) or 0:4 (right) pollen tetrads respectively.

   Full size image (20 KB)
   
   Pause Point Store the seeds at room temperature (22–24 °C) in a labeled 1.5-ml polypropylene tubes with a small hole punctured at the top with a needle to prevent molding.
4. To create the first interval, cross a line-A plant that is homozygous for one color insert to a line-B plant homozygous for a different colored insert. Make several crosses to be sure of a viable cross. For details and tips on crossing A. thaliana, see ref. 49. The stigma of the female used for the cross will mature into a silique containing seeds used for the next step.Troubleshooting
5. After the siliques from the crosses (Step 4) have dried (2–3 weeks), harvest the seeds and plant them as described in Step 2. At the same time, plant at least 12 seeds from line C, which will define the second interval.
6. When these plants begin flowering (4–5 weeks), diagnose the fluorescent genotypes of each plant as described in Box 1. Individuals from the A–B cross should yield pollen tetrads with the same fluorescent pattern (2 color-A:2 color-B); discard plants that do not. For line C, select plants homozygous for the C marker. Save seed stocks from the A/B plants and the C/+ and C/C plants as described in Step 3.
7. Points from here (point 7) up to and including point 11 are related to Construction of second interval (I2), and cross into mutant backgroundCross the A/B plants to the C homozygote as in Step 4.
   Critical step This cross needs to be made several times because the useful plants in the next generation will have all the three markers, which can only be achieved when C is fertilized with a recombinant pollen grain harboring both the A and B inserts in a cis configuration. The probability (as a percent) of this event can be calculated by dividing the recombinant frequency between A and B (in percent) by 2. This should give a good idea of how many crosses to complete (depending on how many seeds you usually get from your crosses). As a general rule, perform enough crosses to ensure that you have three times as many seeds as would be needed to produce a single three-colored plant.Troubleshooting
8. After the siliques from these crosses have dried (2–3 weeks), plant all of the seeds as described in Step 2. Also plant at least eight qrt1/qrt1 (nonfluorescent) seeds and/or the mutant/+; qrt1/qrt1 seeds.
9. When these plants begin flowering, diagnose the marker genotype of each plant as described in Box 1. Each of the three fluorescent marker genotypes needs to be determined under the appropriate colored filter in succession. Pollen tetrads from useful plants will fluoresce 2AB:2C, having the genotype AB+/++C. The markers will have had a chance to recombine during the meiosis that produced the tetrads being examined. In useful plants, the tetrads will show many combinations in relation to one another. What is important is that all the three colors show a 2:2 pattern.
10. At this stage, if the study includes interference analysis in mutant backgrounds, cross the three color lines to mutant/+; qrt1/qrt1 lines. Otherwise, cross the three color lines to the Col qrt1 mutants using a similar strategy as in Step 7 with the purpose of producing lines that are ABC/+++; qrt1/qrt1 (and optionally mutant/+).Troubleshooting
11. When the siliques from these crosses (Step 10) have dried (2–3 weeks), plant all of the seeds. Useful plants will have the genotype ABC/+++ and show a majority of pollen tetrads with a 2ABC: 2 nonfluorescent pattern. These plants can be analyzed for interference, but in order to generate large amounts of data from multiple individuals, it is necessary to analyze the appropriate progeny from self-crosses. If crosses were made into mutant backgrounds, useful plants will have the genotype ABC/+++; mutant/+; qrt1/qrt1.
12. Points from here (point 12) up to and including point 21 are related to Scoring three-color tetrads for recombinationPlant at least 50 seeds collected from ABC/+++; (mutant/+); qrt1/qrt1 individuals. These plants need to be distinguished from one another throughout the scoring process, so it is useful to sow the seeds in divided flats. We plant in 6 4 divided flats (24 cells) where the flats measure 26 54 cm² and each cell measures 6.5 9 cm². This provides each plant enough room to grow while it is also easy to distinguish one plant from another. At this stage, the plants need to be grown in a temperature-controlled growth chamber.
    Critical step Columbia FTLs have been extensively tested in our lab. In general, A. thaliana plants are sensitive to environmental changes, so it is important to keep a constant environment as much as possible. Work done in our lab has shown that temperature levels influence recombination, so it is critical to maintain the same levels for all experimental and control populations¹⁶.
13. After the plants begin flowering (4–5 weeks), it is necessary to determine the fluorescent genotype of each individual plant as described in Box 1. Only plants that are ABC/+++ can be scored for recombination in both intervals simultaneously; discard all others. Note that plants of the genotype ABC/ABC are useful for quantifying pollen viability, see Box 2 for details¹⁷.
14. To begin the scoring process, select an ABC/+++ plant for analysis.
15. Rub a microscope slide briskly for 10 s with a soft cotton cloth.
16. Place two 10-l drops of PGM onto the slide 3 cm apart. The drops should form a tight bead on the slide.
17. Using the forceps, remove a single open flower and place it face down into the first drop of PGM, allow to soak for at least 20 s. Note the flower number on the plant by counting up the bolt.
    Critical step It is best to only score flowers from the primary bolt. Work in our lab has shown that recombination rates are altered in secondary axes¹⁶. In addition, we recommend scoring only the fifth to the thirtieth flowers because younger and older flowers tend to have lower pollen counts and elevated lethality. However, we have not observed any relationship between flower number and recombination¹⁶.
18. Use a tapping motion to release the pollen into the PGM. This is best done on a black bench top that will allow the researcher to see the pollen being released into the solution. Remove the flower from the first drop of PGM and repeat the process with the same flower in the second drop. This maximizes the amount of pollen that can be scored from each flower. Gently place a cover slip on top of each pollen-containing drop of PGM.Troubleshooting
19. The fluorescent markers in the ABC/+++ individuals will be segregating in the pollen. In addition, if a mutant allele has been introduced, the plants will be mixture of wild-type, heterozygotes and mutant genotypes—these genotypes will be scored in Step 20. The researcher can score each four-member pollen tetrad, using the 20 objective on the microscope. The researcher will place each tetrad in the appropriate recombinant category based on the pattern of fluorescence in the grains (see Table 1 and Figs. 3 and 4). Note that in the examples (Fig. 3; Tables 1, 2, 3), the markers were in the red–yellow–cyan (North to South) configuration. Three-color intervals with a different configuration (such as yellow–cyan–red) are perfectly acceptable, but the North–South order of the different colored markers will determine how the fluorescent tetrads are categorized. To determine how to classify each type of tetrad fluorescence pattern, it may be useful to redraw Figure 3 using the appropriate configuration of markers (Supplementary Fig. 1 online).

    Figure 3: Classification of tetrad fluorescent patterns.

    

    DsRed2 (red oval), enhanced yellow fluorescent protein (eYFP) (yellow oval) and enhanced cyan fluorescent protein (eCFP) (cyan oval) transgenes define two hypothetical adjacent genetic intervals (I1 and I2). The panels (a–l) correspond to the classes of tetrads described in Table 1. The four chromatids present after DNA replication can experience no crossovers (COs) (a), single COs (b and c), and double COs (DCOs) in the combined interval including two strand DCOs (d), both kinds of three-strand DCOs (e and f), and four strand DCOs (g). A four-strand DCO in either of the single intervals can also be observed as an nonparental ditype (NPD) tetrad either in the absence of an adjacent CO (h and i) or the presence of an adjacent CO (j and k). A four-strand DCO in both single intervals will result in a tetrad that is NPD for each (l). Each of these events can be distinguished by observing the segregation of the fluorescent markers in the pollen tetrads. In each panel the top, second, third and bottom chromatids are shown segregating into the pollen grains at the top, right, bottom and left positions, respectively.

    Full size image (96 KB)
    
    

    Figure 4: Examples of multicolor fluorescent tetrads.

    

    To assess three-color intervals each tetrad must be visualized through each of three different fluorescent filters (red, yellow and cyan) on the epi-fluorescence microscope. These individual images can be merged (right column) into a composite image using graphics software such as Adobe Photoshop. A plant that is heterozygous for three markers in cis configuration can yield nonrecombinant pollen tetrads that have all the three colors in the same two pollen grains (top row). A single crossover (CO) (in I1 in this case) will yield a pollen tetrad that has one grain with all three colors, one grain with two colors, one grain with one color and one with no color (middle row). Double COs (DCOs) will yield pollen tetrads with a variety of segregation patterns (see Fig. 3), in this case a four-strand DCO results in a tetrad with one red, one yellow/blue, one blue and one red/yellow grain (see Fig. 3g).

    Full size image (33 KB)
    
    

    Table 1: Letter codes for classification of tetrads.

    

    
    Full table
    
    

    Table 2: Example of data produced by one individual plant.

    

    
    Full table
    
    

    Table 3: Example of aggregate data comparing a mutant to wild type.

    

    
    Full table
    
    
    Critical step Do not score tetrads in clumps, as this can lead to erroneously classifying tetrads. The Triton-X detergent is added to the PGM to prevent clumping. Only score four-member tetrads.
20. If a mutant allele is segregating in this generation, genotype the mutants after scoring is complete. Blind scoring is a good experimental practice to avoid bias in the counting process.
21. Repeat for as many plants as the analysis demands—this will differ from experiment to experiment.

Timing

Steps 1–6, selection of appropriate FTLs and construction of the first interval: 6–7 weeks
Steps 7–11, construction of the second interval, and cross into mutant background: 12–14 weeks
Steps 12–21, scoring three-color tetrads for recombination: 4–6 weeks

Troubleshooting

Troubleshooting advice can be found in Table 4.

Table 4: Troubleshooting table.

Full table

Anticipated results

The type of results for a single plant generated by three-color FTL interference experiments is shown in Table 2. Compiled results from an experiment designed to test the interference phenotype of a mutant is shown in Table 3. If the object of the analysis is to test a mutant phenotype, and the crosses are conducted in the manner described, then the mutant (assuming recessivity) will be segregating 1:3 in the scoring generation. This is ideal because the blind controls for the experiment are built into the design.

To calculate interference levels from the data that are produced using this method, we recommend using the Malkova et al. method for quantifying interference¹⁹. Using this method, the researcher determines the map distance of one of the intervals without and with a CO in the adjacent interval, and the ratio between the two represents the level of interference. Using letter codes from Table 1, the following formulas are used:

Interference ratio = X _I1 w/o adjacent CO/X _I1 with adjacent CO, where, as above, X refers to the map distance generated from the Perkins equation⁴⁴.

Ratios between sample populations can be compared using Stahl Lab Online Tools: http://molbio.uoregon.edu/~fstahl.