Immediate visualization of recombination events and chromosome segregation defects in fission yeast meiosis

Schizosaccharomyces pombe, also known as fission yeast, is an established model for studying chromosome biological processes. Over the years, research employing fission yeast has made important contributions to our knowledge about chromosome segregation during meiosis, as well as meiotic recombination and its regulation. Quantification of meiotic recombination frequency is not a straightforward undertaking, either requiring viable progeny for a genetic plating assay, or relying on laborious Southern blot analysis of recombination intermediates. Neither of these methods lends itself to high-throughput screens to identify novel meiotic factors. Here, we establish visual assays novel to Sz. pombe for characterizing chromosome segregation and meiotic recombination phenotypes. Genes expressing red, yellow, and/or cyan fluorophores from spore-autonomous promoters have been integrated into the fission yeast genomes, either close to the centromere of chromosome 1 to monitor chromosome segregation, or on the arm of chromosome 3 to form a genetic interval at which recombination frequency can be determined. The visual recombination assay allows straightforward and immediate assessment of the genetic outcome of a single meiosis by epi-fluorescence microscopy without requiring tetrad dissection. We also demonstrate that the recombination frequency analysis can be automatized by utilizing imaging flow cytometry to enable high-throughput screens. These assays have several advantages over traditional methods for analyzing meiotic phenotypes. Electronic supplementary material The online version of this article (10.1007/s00412-019-00691-y) contains supplementary material, which is available to authorized users.


Introduction
Meiosis is a highly conserved process that produces haploid sex cells (gametes) as an integral part of sexual reproduction (Hunter 2015). During meiosis, chromosomes are deliberately broken to initiate homologous (meiotic) recombination that physically connects the equivalent maternal and paternal (homologous) chromosomes; this is absolutely essential for correct chromosome segregation (Petronczki et al. 2003;Lam and Keeney 2015). Only if these connections (chiasmata) are achieved accurately, healthy gametes containing a single chromosome complement will result from the two meiotic cell divisions. In the process, homologous chromosomes are reshuffled and genes are re-assorted; this provides the genetic diversity that makes individuals unique. Failure to perform meiosis correctly has been shown to cause infertility, miscarriages, and hereditary disorders in mammals (Hassold and Hunt 2001); meiosis is thus fundamental to sexual reproduction.
Meiotic recombination is initiated by Spo11, a topoisomerase VI-like transesterase, creating meiotic double-stranded DNA breaks (DSBs) . These DSBs are subsequently repaired by homology-directed repair mechanisms driven by the RecA-family recombinases Rad51 and Dmc1. Rad51 and its meiosis-specific paralogue Dmc1 are This article is part of a Special Issue on Recent advances in meiosis from DNA replication to chromosome segregation Bedited by Valérie Borde and Francesca Cole, co-edited by Paula Cohen and Scott Keeney^.
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00412-019-00691-y) contains supplementary material, which is available to authorized users. supported by a host of ancillary factors through loading Rad51 and/or Dmc1 onto a processed DSB site and stabilizing them as multimeric nucleoprotein filaments. These ancillary factors include Rad51 paralogues (Gasior et al. 1998;Grishchuk and Kohli 2003;Bleuyard et al. 2005;Sasanuma et al. 2013;Brown and Bishop 2014;Lorenz et al. 2014;Abreu et al. 2018), and factors evolutionarily unrelated to RecA, such as Rad52, Swi5-Sfr1, and Hop2-Mnd1 (Gasior et al. 1998;Chen et al. 2004;Ellermeier et al. 2004;Zierhut et al. 2004;Petukhova et al. 2005;Kerzendorfer et al. 2006;Vignard et al. 2007;Octobre et al. 2008). In Sz. Pombe, the Hop2-Mnd1 orthologues are called Meu13-Mcp7, and similar to the situation in other eukaryotes, meiotic recombination is strongly reduced in their absence (Nabeshima et al. 2001;Saito et al. 2004). Homology-directed repair can follow several pathways, and ultimately results in crossover (CO) and non-crossover recombination outcomes (Phadnis et al. 2011;Hunter 2015). Only COs between homologous chromosomes support the formation of chiasmata, which together with sister chromatid cohesion are needed for proper chromosome segregation (Marston 2014). Cohesion is achieved by the cohesin complex which physically entraps the sister chromatids right after their replication during S phase (Nasmyth and Haering 2009). Cohesin holds sister chromatids together until all chromosomes are properly attached to microtubules in metaphase, at which point the kleisin subunit of cohesin is destroyed and anaphase ensues (Nasmyth and Haering 2009;Marston 2014). To reduce the diploid chromosome complement to a haploid one, meiosis consists of two cell divisions following a single round of DNA replication; special modifications to sister chromatid cohesion have to be in place to enable this. During meiosis I, homologous chromosomes are segregated from each other, and cohesins are only removed from the chromosome arms, whereas cohesins at centromeres remain protected for the second meiotic division. During meiosis II, centromeric cohesin protection is removed to allow sister chromatids to be segregated from each other (Petronczki et al. 2003;Marston 2014). A key centromeric protector is the Mei-S332 homolog Shugoshin, Sgo1 (Katis et al. 2004;Kitajima et al. 2004;Marston et al. 2004;Rabitsch et al. 2004), and the absence of Sgo1 and chiasmata, indeed, generates a strong chromosome segregation defect during meiosis (Hirose et al. 2011).
Here, we establish and characterize visual assays to quantify chromosome segregation defects and meiotic recombination frequency which are new to Sz. pombe. Visual assays for determining meiotic recombination frequencies were originally established in Arabidopsis, and more recently adapted for budding yeast (Francis et al. 2007;Thacker et al. 2011). These visual recombination assays utilize genes encoding red, yellow, and cyan fluorophores driven by gamete-specific promoters, and are integrated at specific loci on a given chromosome to form genetic intervals. The four products (gametes) of a single meiosis will fluoresce in a color corresponding to the fluorophore gene(s) they receive. In Arabidopsis, the fluorophores are expressed from the pollen-specific post-meiotic LAT52 promoter, and various genetic intervals (fluorescent-tagged lines, FTLs) have been generated and adopted widely (e.g., Yelina et al. 2013;Séguéla-Arnaud et al. 2017;Kurzbauer et al. 2018). Also, the budding yeast version of the visual recombination assay starts to enjoy popularity and several recent studies used spore-autonomous fluorophore expression to determine meiotic recombination frequency (e.g., Vincenten et al. 2015;Arter et al. 2018;González-Arranz et al. 2018;Raffoux et al. 2018;Rogers et al. 2018). In yeasts, this kind of setup allows assessment of the frequency of exchange of flanking markers (COs) and has advantages over traditional methods for studying meiotic recombination-such as using nutritional markers (White and Petes 1994;Smith 2009) or Southern blotting of DNA from meiotic yeast cells (Hyppa and Smith 2009;Oh et al. 2009): (I) spores can be assessed regardless of their viability (ability to form a visible yeast colony), (II) the simplicity of this method will allow its use for high-throughput genetic screens, and (III) achieving large sample sizes is straightforward when using imaging flow cytometry. Additionally, this can also be used as a tool for monitoring chromosome segregation defects, when different fluorophore markers are inserted close to a centromere (Thacker et al. 2011;this study).
These visual assays represent a novel, powerful, and easyto-use experimental tool for fission yeast allowing straightforward analysis of chromosome segregation and homologous recombination defects during meiosis. They also enable the identification and characterization of complex phenotypes (single and double CO formation) in high-throughput screens via imaging flow cytometry.

Molecular and microbiological techniques
Plasmids and details of construction are given in Table S1. DNA-modifying enzymes (high-fidelity DNA polymerase Q5, Taq DNA polymerase, T4 DNA ligase, restriction endonucleases) and the NEBuilder HiFi DNA Assembly Master Mix were obtained from New England BioLabs (NEB), Inc. (Ipswich, MA, USA), and the In-fusion HD Cloning kit from Takara Bio, Inc. (Mountain View, CA, USA). Oligonucleotides (Table S2) were supplied by Sigma-Aldrich Co. (St. Louis, MO, USA). All relevant regions of plasmids were verified by DNA sequencing (Source BioScience plc, Nottingham, UK). Plasmid sequences are available as supporting online material (Lorenz 2018).
Escherichia coli was grown in LB and SOC media, when appropriate media contained 100 μg/ml ampicillin (Sambrook and Russell 2000). Competent E. coli XL1-blue cells (Agilent Technologies, Santa Clara, CA, USA) were transformed following the protocol provided by the manufacturer.
All sequence details and positional information about Sz. pombe genomic loci have been extracted from https://www. pombase.org (Wood et al. 2002).
Spore viability by random spore analysis and meiotic recombination assays have been performed as previously described (Osman et al. 2003;Smith 2009;Sabatinos and Forsburg 2010;Lorenz et al. 2012).

Microscopy
For microscopy cells from sporulating cultures were suspended in sterile demineralized water, and spotted onto microscopic slides. After placing a cover slip over the cell suspension, cells were immobilized by squashing the slide in a filter paper block, and afterwards the cover slip was sealed with clear nail varnish. Microscopic analysis was done using a Zeiss Axio Imager.M2 (Carl Zeiss AG, Oberkochen, Germany) epi-fluorescence microscope equipped with the appropriate filter sets to detect red, yellow, and cyan fluorescence. A 63× objective (Plan-Apochromat, aperture 1.4) was used for taking black-and-white images with a Zeiss AxioCam MRm CCD camera controlled by AxioVision 40 software v4.8.2.0. For chromosome segregation experiments 9-20 and for recombination assays 20-25 randomly selected fields of view were photographed and evaluated. Images were pseudo-colored and overlaid using Adobe Photoshop CC (Adobe Systems Inc., San José, CA, USA). Images of mature four-spored asci were evaluated manually; data was collected and analyzed in Microsoft Excel 2016 MSO (version 16.0.4738.1000, 32-bit).

Imaging flow cytometry
The ImageStreamX Mark II (Merck KGaA, Darmstadt, Germany) is an imaging flow cytometer, where an image of each individual cell is acquired as it flows through the cytometer. It measures hundreds of thousands of individual cells in minutes, combining the high-throughput capabilities of conventional flow cytometry with single-cell imaging. The ImageStream measures not only total fluorescence intensities but also the spatial image of the fluorescence plus bright-field and dark-field images of each cell in a population.
For a more extensive overview of data acquisition and analysis in ImageStreamX, see Basiji (2016). Briefly, the INSPIRE acquisition software generates raw image data (.rif file) without compensation which can then be directly loaded into IDEAS for further analysis. Using the IDEAS software, the .rif files will then be converted into compensated image files (.cif) by applying the compensation matrix (.ctm) generated from single fluorescence controls during the acquisition. The file resulting from analysis is stored as .daf (data analysis file), which is used for plotting features derived from the bright-field, dark-field, and fluorescence single cell images in the form of histograms or bivariate scatter plots. Subpopulations are generated using these plots and saved as analysis template for further datasets.
For imaging flow cytometry, cellular material containing asci was suspended in 1× PBS, pH 7.5 (8 g/l NaCl, 2 g/l KCl, 1.15 g/l Na 2 HPO 4 ·7H 2 O, 2 g/l anhydrous KH 2 PO 4 ), harvested by centrifugation (6000×g, 30 s), and re-suspended in 1× PBS, pH 7.5. Data was acquired on the ImageStreamX Mark II using INSPIRE acquisition software (Merck kGaA). Cellular parameters were measured in Channel 1 (Brightfield, BF), Channel 2 (GFP*, a yellow-shifted version of green fluorescent protein, using a 485 nm laser), Channel 4 (RFP, red fluorescent protein, 561 nm), Channel 7 (CFP, cyan fluorescent protein, 405 nm), and Channel 12 (side scatter, 785 nm) with magnification set to 60×. Briefly, objects of interest (asci) with a BF Barea^of 50 to 200 μm 2 and an Baspect ratio^(ratio of minor axis to major axis) lower than 0.5 (Bdoublet area^) were selected. Focused cells were identified by a Bgradient RMS^feature value of 50 or higher. A typical file contained about 25,000 focused yeast cells.
Data evaluation for identification of asci and spore phenotyping were performed using IDEAS software (version 6.2; Merck). A focused population of asci were identified within the Bdoublet area^and based on the features BModulation^for fluorescent channels (the Modulation texture feature measures the intensity range of an image, normalized between 0 and 1) and BIntensity^for side scatter (SSC) using the custom masks BMorphology^and BObject(right),^respectively. Further refinement was performed each on RFP, GFP*, and CFP fluorescence via BIntensity.^Following analysis of the merged triple fluorescent population using BLength^and BElongatedness^(ratio of the height over width of the object's bounding mask) features (custom BF mask BAdaptiveErode, M01, Ch01, 75^) resulted in identification of asci of interest. Finally, spore phenotype analysis was conducted by evaluating the fluorescent area using custom masks for each fluorescent intensity (GFP* intensity 200-4095, RFP intensity 75-4095, and CFP intensity 150-4095) and by applying Boolean algebra to identify particular combinations of fluorescent colors. Asci with a mask area larger than 3 μm 2 were considered positive for a particular spore phenotype.

Results and discussion
Identifying spore-autonomous promoters in Schizosaccharomyces pombe To test whether a particular upstream regulatory sequence is a spore-autonomous promoter (Thacker et al. 2011), we cloned a 700-931-bp region upstream of the start codon of the Sz. pombe eis1, pil2, eng2, agn2, and mde10 genes in front of a cyan (mCerulean) or red (tdTomato) fluorophore gene inserted in pDUAL, a vector restoring leu1 + by integrating at the leu1-32 mutant locus (Matsuyama et al. 2004). Fluorophore genes were terminated by Saccharomyces spp. PGK1 downstream regulatory sequence: T PGK1 from S. bayanus for mCerulean, and T PGK1 from S. kudriavzevii for tdTomato (Thacker et al. 2011). The candidate promoters were selected on the basis of its corresponding gene being upregulated during late meiosis or sporulation (Mata et al. 2002): eng2, agn2, and mde10 code for proteins involved in spore wall formation, eis1 encodes an eisosome assembly protein, and pil2 a component of the eisosome. The promoter of S. cerevisiae YKL050c has previously been shown to support spore-autonomous expression of fluorophores in budding yeast (Thacker et al. 2011); Sz. pombe eis1 is the single homolog of the S. cerevisiae paralogue pair EIS1 and YKL050c. The resulting plasmids (pALo139, pALo140, pALo141, pALo142, pALo175; Table S1) were digested with ApaI to release the leu1 + integration cassettes containing the constructs; these were transformed into h + and h − fission yeast strains (ALP729 and FO652) carrying the leu1-32 mutation. Two leu1 + strains of different mating types carrying differently colored fluorophore constructs were crossed to each other, and presence or absence of spore-specific fluorescence was recorded on an epi-fluorescence microscope. P eng2 , P agn2 , and P mde10 failed to produce fluorescence levels visible under the microscope (data not shown). P eis1 and P pil2 were strong spore-autonomous promoters yielding clear red or cyan fluorescence in spores of mature asci (data not shown).
To avoid ectopic recombination events between the P eis1 and P pil2 constructs and the upstream regions of endogenous eis1 and pil2, we decided to follow a similar strategy as Keeney and co-workers (Thacker et al. 2011), and investigated whether P eis1 and P pil2 from Schizosaccharomyces species other than Sz. pombe can be used as spore-autonomous promoters in Sz. pombe. Indeed, the upstream sequences of the Sz. japonicus eis1 and pil2 homologs SJAG_04227 and SJAG_02707 , as well as the regions upstream of Sz. cryophilus and Sz. octosporus pil2 homologs SPOG_00147 and SOCG_04642, cloned in front of fluorophores produced strong fluorescence in spores of Sz. pombe asci (Fig. 1a). P SJAG_04227 , P SPOG_00147 , and P SOCG_04642 were selected to drive tdTomato (red fluorescence protein, from now called RFP), GFP* (yellow-shifted green fluorescence protein, terminated by T PGK1 from S. mikatae) (Griesbeck et al. 2001;Thacker et al. 2011), and mCerulean (cyan fluorescence protein, from now on called CFP) expression in all experimental constructs (Fig. 1).

Monitoring meiosis chromosome segregation defects
Markers inserted next to the centromere can be used to monitor meiotic chromosome segregation defects. Previously, this has been exploited in genetic screens by introducing bacterial operator repeats (lacO or tetO) close to centromeres in budding and fission yeast, to identify chromosome segregation mutants via the distribution of LacI-or TetR-GFP fusions binding to their respective operators, thus becoming visible as small foci (Straight et al. 1996;Michaelis et al. 1997;Nabeshima et al. 1998;Katis et al. 2004;Rabitsch et al. 2004;Gregan et al. 2005). Introducing spore-autonomously expressed fluorophore markers with different colors at the centromere (Figs. 1b and 2a) has the advantages of (I) enabling distinction of meiosis I and meiosis II segregation defects in a single assay (Fig. 2) rather than requiring homozygous and heterozygous setups of lacO or tetO repeats integrated close to a centromere, and (II) likely not interfering with chromosome behavior as strongly as lacO or tetO repeats Sofueva et al. 2011). Fission yeast asci are ordered; due to the physical constraints of the zygotic cell size and shape, microtubular spindles can orientate only along the longitudinal axis of the zygote, which means that the neighboring nuclei/spores in one half of the zygote are the sister products generated in meiosis II (Fig. 2b). This makes the evaluation of chromosome mis-segregation a comparatively straightforward undertaking in Sz. pombe.
We tested the functionality of our assay carrying fluorophore genes under the control of the sporeautonomous SPOG_00147-promoter integrated close to CEN1 (Fig. 2a) with a set of mutants defective in meiotic recombination (meu13, spo11) and/or kinetochore function (sgo1) (Keeney et al. 1997;Nabeshima et al. 2001;Sharif et al. 2002;Rabitsch et al. 2004). For this, we only evaluated four-spored asci, and ignored asci with spore counts of 1, 2, or 3, to exclude incidences of clear nuclear division failures in meiosis I or II. As expected, in wild-type and meu13Δ crosses, chromosome 1 is correctly segregated, in almost all cases (Fig. 2c). We did observe a low frequency (3.3%) of CO recombination between the fluorophore marker and the physical centromere in wild type, leading to red-cyan pairs of sister nuclei, rather than red-red and cyan-cyan pairs (Fig. 2c). In meu13Δ, which strongly reduces meiotic recombination (Nabeshima et al. 2001), no COs were observed, but two  Fig. 1 Spore-autonomous expression of fluorophores. a Schematic and examples of main constructs, P SOCG_04642 -GFP*-T PGK1(mik) from strain UoA694, P SPOG_00147 -mCerulean-T PGK1(bay) from strain UoA727, P SPOG_00147 -tdTomato-T PGK1(kud) from strain UoA726, and P SJAG_ 04227 -tdTomato-T PGK1(kud) from strain UoA694; scale bar in example images represents 10 μm. b Plasmid maps of CEN1targeting (CEN1t) constructs using the Sz. octosporus SPOG_ 00147 (pil2) promoter to drive RFP (tdTomato) and CFP (mCerulean) expression. c Plasmid maps of constructs usable for generating genetic intervals (see main text for details); RFP is driven by the Sz. japonicus SJAG_04227 (eis1) promoter in pALo148 and by Sz. octosporus SPOG_00147 (pil2) promoter in pALo181, CFP by the Sz. octosporus SPOG_00147 (pil2) promoter in pALo168 & pALo182, and the yellow-shifted GFP* by the Sz. cryophilus SOCG_04642 (pil2) promoter in pALo179 & pALo186 incidences of chromosome mis-segregation could be recorded (Fig. 2c). As an obvious example for meiotic chromosome mis-segregation, we employed double mutants of sgo1Δ with meu13Δ or spo11Δ. A sgo1Δ single mutant does not produce a strong mis-segregation phenotype ), but in combination with the absence of recombination factors, a meiotic non-disjunction phenotype can be observed (Hirose et al. 2011). Indeed, massive chromosome segregation defects are obvious in asci of meu13Δ sgo1Δ and spo11Δ sgo1Δ double mutants (Fig. 2c). In spo11Δ sgo1Δ, the percentage meiotic non-disjunction is slightly higher than in meu13Δ sgo1Δ, and there are also more meiosis I chromosome missegregation events in spo11Δ sgo1Δ. In meu13Δ, chromosome segregation can presumably be supported to some degree, because a small number of chiasmata is still being produced, whereas in spo11Δ meiotic DSB formation is completely abrogated and thus no chiasmata are formed.
Creating a genetic interval with fluorophore markers to assess meiotic recombination frequency To explore whether fluorophore markers inserted at defined genomic sites on a single chromosome to create a genetic interval that can be used to determine meiotic recombination frequencies, we transformed constructs integrating on chromosome 3 forming a genetic interval of~45 kb around the ade6 locus (Fig. 3a) (Osman et al. 2003;Lorenz et al. 2012). To target the ura4 + -marked P SJAG_04227 -tdTomato construct to the same locus as ura4 + -aim2 on chromosome 3 (at position 1,291,583,~26.5 kb upstream of ade6), a ura4 + -P SJAG_04227 -tdTomato-T PGK1(Skud) cassette was amplified by PCR from pALo148 adding~80 bp of homologous flanking sequences (oligonucleotides oUA113 and oUA114, Table S2) (Bähler et al. 1998), and transformed into strain FO652. All strains harboring ura4 + -aim2-P SJAG_04227 -tdTomato have been derived from the resulting transformant, UoA523 (Table S3), by crossing. A similar approach failed to deliver the his3 + -P SPOG_00147 -mCerulean to the same site as his3 + -aim on chromosome 3 (at position 1,337,447,~19.5 kb downstream of ade6). Therefore, we cloned larger homologous flanking sequence up-and downstream of his3 + -P SP OG_00147 -mCerulean-T PGK1(Sbay) into the NotI sites of pALo182 (Table S1). The backbone and insert (his3 + -P SPOG_00147 -mCerulean-T PGK1(Sbay) ) of pALo182 after a NotI digest were merged with a 436-bp upstream (oligonucleotides oUA189 and oUA190) and a 646-bp downstream flanking sequence (oligonucleotides oUA191 and oUA192, Table S2) amplified by PCR from Sz. pombe genomic DNA (strain MCW1196, Table S3) in a single NEBuilder assembly reaction (in the process, the NotI sites flanking the whole construct were replaced by SmaI sites). The whole cassette was excised from the resulting plasmid (pALo168,  Fig. 2 Chromosome segregation assay using spore-autonomous expression of fluorophores. a Schematic of assay, RFP and CFP are expressed from the Sz. octosporus SPOG_00147 (pil2) promoter integrated at position 3,751,911 on chromosome 1 downstream of the per1 (SPAP7G5.06) locus close to its centromere (CEN1). b Meiotic nuclear divisions generate an ordered tetrad with sister nuclei from meiosis II (MII) ending up next to one another. c Chromosome segregation phenotypes in fourspored wild-type (WT; UoA726 × UoA727, n = 274), meu13Δ (UoA752 × UoA755, n = 101), meu13Δ sgo1Δ (UoA756 × UoA759, n = 53), and spo11Δ sgo1Δ (UoA760 × UoA763, n = 20 asci) asci. A low frequency of crossover (CO) events (3.3%) between the fluorophore genes and CEN1 has been observed in WT and transformed into strain ALP729. This generated strain UoA676 (Table S3), from which all strains carrying his3 +aim-P SPOG_00147 -mCerulean were derived by crossing. Finally, the yellow spore marker (P SOCG_04642 -GFP*-T PGK1(Smik) ) on pALo179 was generated by an NEBuilder assembly of P SOCG_04642 (PCR on genomic DNA of Sz. octosporus yFS286, oligonucleotides oUA201 and oUA202) and GFP*-T PGK1(Smik) (PCR on pSK726, oligonucleotides oUA204 and oUA138) between the ade6-targeting sequences on pALo159 (linearized by a BamHI and BglII digest) (Tables S1 and S2). The ade6 + ::P SOCG_04642 -GFP* strain UoA666 (Table S3) was created by transforming the ade6targeting cassette from pALo179 (amplified by PCR, oligonucleotides oUA142 and oUA143; Table S2)  Genetic interval constructed with spore-autonomously expressed fluorescent markers can be analyzed by imaging flow cytometry. a Schematic of the genetic interval constructed: RFP expressed from the Sz. japonicus SJAG_04227 (eis1) promoter together with a ura4 + marker is integrated on chromosome 3 at position 1,291,583, the same site as ura4 + -aim2 (Osman et al. 2003); CFP expressed from the Sz. octosporus SPOG_00147 (pil2) promoter together with a his3 + marker is integrated on chromosome 3 at position 1,337,447, the same site as his3 + -aim (Osman et al. 2003); GFP* expression driven from the Sz. cryophilus SOCG_04642 (pil2) promoter, the construct is integrated between positions 1,318,042 and 1,318,115 on chromosome 3 (downstream of ade6 at its endogenous locus). Only outcomes of single crossovers (COs) between the three markers are shown; double COs are rare (see Figs. S2 and S3 for double COs observed in this kind of assay). Please note that order of spore colors is not fixed, but can rotate perpendicular to the meiotic spindle axis. b Outline of the workflow to identify asci based on particular cellular features on the Amnis ImageStreamX Mark II. Modulation measures the intensity range of an image normalized between 0 and 1 by calculating Modulation = (Max Pixel − Min Pixel)/(Max Pixel + Min Pixel). c Examples of ascus phenotypes from a cross of wild-type strains (UoA694 × UoA676) as shown in a; Boolean algebra mask equations used to discriminate between the different ascus types as presented in Ch01 BF1 ade6-3′Δ::natMX6 strain UoA570. This transformation restored ade6-3′Δ to ade6 + and removed the natMX6 cassette, all ade6 + ::P SOCG_04642 -GFP* strains were derived from UoA666 by crossing. As this assay visualizes recombination events, we evaluated it using standard epi-fluorescence microscopy, and also tested whether single-cell imaging flow cytometry (Basiji 2016) could be exploited to perform high-throughput screens with the spore-autonomously expressed fluorophore recombination assay. We established a workflow on the Amnis ImageStreamX Mark II imaging flow cytometer to select for mature asci displaying fluorescence from a mixed population of cells in a standard cross (mature fluorescing asci, immature non-fluorescing asci, zygotes, vegetative cells), and subsequently applied customizations in the software to identify spore color phenotypes unique for the recombination outcomes we expected to occur in this assay (Fig. 3b, c).
The important first steps are identifying focused cells by using a measure of the Bgradient RMS^feature of the brightfield image to define the focus quality. Single and multiple cells were determined by plotting the cell mask area versus cell mask aspect ratio, whereby the asci were located in the doublet area. Once focused subspecies are identified via gating, they were used as the starting point to analyze recombination products by determining the spore phenotype, which is only feasible by utilizing the fluorescent markers GFP*, RFP, and CFP.
For this purpose, mainly the BModulation (texture)^feature was applied to objectively discriminate the bright fluorescence pattern of GFP*, RFP, and CFP associated asci. We first gated GFP*-positive objects on the basis of the appropriate BModulation (texture)^against darkfield using the side scatter (SSC) parameter in a bivariate plot. In the next stage, the gated GFP* population was subanalyzed for the modulation of RFPand CFP-containing spores (Fig. 3b).
Employing the ability of the IDEAS software for creating Boolean logic, masks with good determination of spore borders in each fluorescent channel were selected, and these advanced combined masks determined spores with particular fluorescent phenotypes (Fig. 3c). For example, spores with all three fluorescent proteins are only possible, if recombination happened between GFP* and CFP, whereas spores containing RFP and CFP are only possible, if recombination happened between RFP and GFP*. Finally, asci were quantified within the triple merged combined fluorescent populations by using the newly created shape features BLength^versus BElongatedness^in brightfield. Thereby, asci with recombination products were identified within a BLength^< 20 and an BElongatedness^> 2 (Fig. 3).
If a particular experimental setup requires a distinction between four-spored asci and asci with irregular spore numbers (1, 2, 3, > 4), masks using information from the brightfield channel can be programmed to accommodate this.
Because the fluorophore markers were inserted at the same positions as the nutritional markers of an established recombination assay (Figs. 4a, b and S1a, b) (Osman et al. 2003;Lorenz et al. 2012Lorenz et al. , 2014, we could directly compare the outcomes of the different assays assessed by various methods. We used two slightly different recombination assays utilizing nutritional markers: one contained a point mutation at ade6 (ade6-704, a T645A substitution mutation; Park et al. 2007), the other one a dominant drug resistance marker inserted at the 3′ end of ade6 creating a partial deletion (ade6-3′Δ::natMX6). We used the latter to test whether a drastically different recombination frequency is caused by introducing a heterologous piece of DNA into the genetic interval. The natMX6 cassette is~1.25 kb in size and removes 848 bp of genomic DNA at ade6 (429 bp of which are ade6 coding sequence); in comparison, the spore-autonomously expressed GFP* cassette is~2.1 kb in size and inserted just downstream of the ade6 open reading frame (removing 73 bp just downstream of the 3′-untranslated region of ade6).
Despite all these differences between the markers, the recombination frequencies within the genetic intervals were remarkably similar (Figs. 4c and S1c). The genetic intervals with the nutritional markers produced 11.88% (ade6-704) and 13.33% (ade6-3′Δ) COs, respectively (Figs. 4c, Table S4). The interval with the fluorophore markers measured 9.41% COs on the epi-fluorescence microscope and 14.57% COs on the imaging flow cytometer (Fig. 4c, Table S5). The results were comparable, when the ade6-or GFP*-markers were initially linked with his3 + -aim or CFP, respectively (10.63% CO for ade6-704, 8.33% CO for ade6-3′Δ, 7.68% CO for fluorophore markers evaluated by epi-fluorescence microscopy; Fig. S1, Tables S4 and S5). In all types of assays, we could also detect a few rare double CO events (Figs. 4 and S1, Tables S4 and S5). Because asci can be evaluated as an ordered tetrad in the fluorophore-based assay (Figs. 2b and 3a), information about the involvement of two, three, or all four chromatids in the double CO can be extracted. Within the four double CO events over the two slightly different genetic intervals evaluated on the epi-fluorescence microscope (Figs. 4b and S1b), examples for participation of two, three, or four chromatids could be found (Figs. S2 and S3). The observed frequency of double CO in any of the genetic assays is equal with or slightly higher than expected from the frequency in neighboring intervals (Tables S4 and S5), in line with Sz. pombe not displaying CO interference (Munz 1994).
In a meu13 mutant meiotic intra-and intergenic recombination is strongly decreased (Nabeshima et al. 2001). When running the fluorophore-based assay in a meu13Δ background, as expected, a 3.6-to 5.7-fold reduction in CO formation could be observed (Fig. 4c). No double COs were detected in the meu13Δ crosses. This demonstrates that in Sz. pombe a genetic interval consisting of spore-autonomously expressed fluorescent markers behaves very similarly to a genetic interval built from nutritional markers.

Conclusion
Here, we established assays employing spore-autonomously expressed fluorescent proteins to determine meiotic chromosome mis-segregation and meiotic recombination frequencies in the fission yeast, Sz. pombe. We generated a series of plasmids containing selectable markers in addition to the sporespecific fluorophores (Fig. 1, Table S1); this makes the whole system portable enabling the creation of genetic intervals at virtually any position within the Sz. pombe genome.
Ectopic spore-autonomous promoters from Sz. japonicus work in Sz. pombe; this raises the possibility that expression from this type of regulatory elements is conserved, and could be used to develop a similar system in Sz. japonicus. This is of interest, because Sz. japonicus produces 8-spored asci (an additional mitosis following the two meiotic divisions) (Klar Fig. 4 Comparison of genetic intervals generated by nutritional markers and spore-autonomously expressed fluorescent markers. a Schematic of genetic recombination assay using nutritional markers and plating of colonies. In UoA112, the ade6 marker is a point mutation (ade6-704) without hotspot activity; in UoA736, it is a partial deletion of ade6 by integrating a natMX6 cassette (ade6-3′Δ). In both instances, ade6 is at its endogenous locus on chromosome 3, and the position for the coding sequence is 1,316,337-1,317,995. The flanking markers ura4 + and his3 + are the artificially introduced markers (aim) and his3 + -aim, which have been previously described (Osman et al. 2003); ura4 + -aim2 is integrated on chromosome 3 at position 1,291,583, and his3 + -aim at position 1,337,447. b Schematic of spore-autonomously expressed fluorophore recombination assay (see also Fig. 3a), the RFP gene is at the same position as ura4 + -aim2 in a, the CFP gene at the same position as his3 + -aim in a, and the GFP* gene is inserted downstream of ade6 + . c Results from recombination assays in a and b: crossover (CO) recombinant frequencies were determined in wild-type (WT) crosses by random spore analysis for the plating assay (a), using data from n = 3 independent crosses with 160 progeny each. CO recombinant frequencies were determined in WT and meu13Δ crosses either by counting manually on an epifluorescence microscope (UoA694 × UoA676 n = 356 asci, UoA742 × UoA743 n = 305 asci) or by high-throughput single cell assessment on an imaging flow cytometer (ImageStream) (UoA694 × UoA676 n = 916 asci, UoA742 × UoA743 n = 370 asci). Please note that ImageStream can only identify one out of two double CO classes 2013) enabling an even better resolution of genetic events. We validated our system by comparison to an established recombination assay (Osman et al. 2003;Lorenz et al. 2012Lorenz et al. , 2014 utilizing nutritional markers (Fig. 4), and demonstrated that imaging flow cytometry can be used to run genetic highthroughput screens for recombination phenotypes (Figs. 3  and 4). Due to its portability and advantages over existing assays, our fluorophore-based system represents a novel addition to the ever-growing genetic toolkit for probing the cell biology of fission yeast.