MeioSeed: a CellProfiler-based program to count fluorescent seeds for crossover frequency analysis in Arabidopsis thaliana
© The Author(s) 2018
Received: 8 December 2017
Accepted: 7 April 2018
Published: 18 April 2018
The formation of crossovers during meiosis is pivotal for the redistribution of traits among the progeny of sexually reproducing organisms. In plants the molecular mechanisms underlying the formation of crossovers have been well established, but relatively little is known about the factors that determine the exact location and the frequency of crossover events in the genome. In the model plant species Arabidopsis, research on these factors has been greatly facilitated by reporter lines containing linked fluorescence marker genes under control of promoters active in seeds or pollen, allowing for the visualization of crossover events by fluorescence microscopy. However, the usefulness of these reporter lines to screen for novel modulators of crossover frequency in a high throughput manner relies on the availability of programs that can accurately count fluorescent seeds. Such a program was previously not available in scientific literature.
Here we present MeioSeed, a novel CellProfiler-based program that accurately counts GFP and RFP fluorescent Arabidopsis seeds with adjustable detection thresholds for fluorescence intensity, making use of a robust seed classifier which was trained by machine learning in Ilastik. Using the previously published reporter line Col3-4/20 as an example, we explain the use of MeioSeed and the steps taken to optimize the thresholding settings of the program to fit the published model for recombination frequency and transgene segregation. The use of MeioSeed is illustrated by investigating salt stress as a novel abiotic trigger for changes in crossover frequency in Col3-4/20 (♂) × Ler-0 (♀) F1 hybrids. Salt stress was found to trigger increases in crossover frequency between the marker genes of up to 70% compared to the control treatment without salt stress. Genotyping of control and salt treated populations revealed that the changes in crossover frequency were not limited to the region between the marker genes, but that fluctuations in crossover frequency are likely to occur genome-wide after treatment with high salt concentrations.
MeioSeed allows for the high throughput recognition and counting of fluorescent Arabidopsis seeds and can facilitate the screening for novel abiotic and biotic modulators of crossover frequency using reporter lines in Arabidopsis.
For sexually reproducing eukaryotic organisms the process of meiosis is pivotal for the formation of haploid gamete cells and for the redistribution of the genetic material among the progeny. Meiosis consists of two consecutive cell divisions , with the first one (meiosis I) resulting in the separation of homologous chromosomes usually accompanied by meiotic crossing-over events between non-sister chromatids. During the second division (meiosis II) the sister chromatids are separated, resulting in four haploid daughter cells each with a unique reassembly of the genetic properties of the diploid progenitor meiocyte.
An essential step in the accurate redistribution of chromosomes during meiosis I is the correct formation of pairs of homologous chromosomes (bivalents), which is mechanically facilitated by the invasion of strands of homologous chromosomal DNA at double stranded break sites generated by the transesterase SPO11 [2, 3], resulting in 5′ end resection of the breaks and the formation of D-loops [1, 4]. The resolution of D-loops can result in the reciprocal exchange of large parts of the chromosome arms, events which are called crossovers [1, 4]. Crossovers are of special interest to plant breeders, as they allow for the redistribution of allelic variants among populations of F2 plants. The pathways for the resolution of D-loops into crossovers or non-crossovers have been well established [1, 4, 5], but relatively little is known about the factors that determine which SPO11 breaks are resolved into crossovers and how the frequency of crossovers along the genome is determined.
Studies on crossover frequency in plants make use of different types of approaches to visualize crossover events. Fundamental biological questions on the formation of bivalents and crossovers are often addressed by cytogenetic studies to count chiasmata or foci for markers of double stand break formation with microscopy [6–10]. The main advantage of these cytogenetic techniques is that the formation of crossovers can be studied in vivo, but they require extensive expertise [11–13]. Perhaps the most widely used approach is PCR-based genotyping (e.g. Kompetitive Allele Specific (KASP) genotyping ) of hybrid F2 plants with primer sets for genetic markers containing single nucleotide polymorphisms (SNPs) among the parental accessions. When genotyping is done for multiple different markers along the chromosome and for a large number of F2 individuals, a genetic map can be constructed which gives an overview of the crossover landscape of the genome. The main advantage of the genotyping approach is that primers can be designed for any polymorphic locus in the genome and that there are ample databases with established SNP markers that have been made publicly available (e.g. ). However, the high throughput application of genotyping requires large numbers of DNA extractions and PCR reactions and all the required equipment. Finally, the research on crossover frequency in the model species Arabidopsis has been greatly facilitated by the availability of fluorescence reporter lines for all five chromosomes [16–18]. These lines contain in cis linked fluorescence reporter constructs under control of seed or pollen specific promoters and are separated by known genetic distances, thereby enabling the visualization of crossover events between the two fluorescence marker constructs in the F2 seeds or in the pollen of F1 plants, respectively. The main advantage of this approach is that it in principle only relies on the availability of a fluorescence microscope.
Here, we present a novel, high throughput program to count fluorescent Arabidopsis seeds, which we have called MeioSeed. The program consists of a batch-wise image processing algorithm in the freeware program CellProfiler, combined with a seed classifier trained by machine learning in the program Ilastik. Using the meiotic reporter line Col3-4/20 as an example, we show that MeioSeed accurately counts red and green fluorescent seeds with adjustable detection thresholds for fluorescence from microscopy images, giving overlay images with seed counts and fluorescence colors and a raw CSV data file as an output. We further illustrate how we optimized the pipeline for Arabidopsis seeds in our setup, which could be adjusted for any other fluorescence microscopy setup. Finally, we illustrate the use of MeioSeed by investigating salt stress as a novel abiotic trigger for changes in crossover frequency in Col3-4/20 (♂) × Ler-0 (♀) F1 hybrids.
Plant material and growth conditions
Homozygous seeds of the meiotic tester line ‘Col3-4/20’  harboring in cis linked napA::GFP (insertion site: 256,516 bp; 4 cM) and napA::RFP (insertion site: 5,361,637 bp; 20 cM) fluorescence reporter constructs at the top of the long arm of chromosome 3  were kindly provided by Dr. Cathy Melamed-Bessudo and Prof. Avraham A. Levy (Department of Plant Sciences, Weizmann Institute of Science, Rehovot, Israel). Col3-4/20 (♂) plants were crossed with the accession Landsberg erecta (Ler-0; ♀) to obtain F1 hybrids, which were selfed to obtain pools of F2 seeds for measurements of crossover frequency and/or genetic mapping. F1 hybrids of which data are presented in Fig. 3 were grown in a climate-controlled growth chamber at 20 °C, 70% relative humidity, a 16 h photoperiod and at ~ 150 μmol m−2 s−1 of photosynthetically active radiation (PAR). The F1 hybrids which received NaCl treatments and control treatments with 0 mM NaCl (data presented in Fig. 4) were grown in a climate-controlled growth chamber at 20 °C, 70% relative humidity, a 12 h photoperiod and at ~ 200 μmol m−2 s−1 PAR. Rapeseeds (Brassica napus cv. Westar) were kindly provided by Arezoo Rahimi (Department of Molecular and Developmental Genetics, Leiden Institute of Biology, Leiden, The Netherlands). Barley grains (Hordeum vulgare) were kindly provided by Fytagoras B.V. (Leiden, The Netherlands).
Collection of seed images
Bright field images, and GFP and RFP fluorescence images of Arabidopsis seeds were collected using a LeicaMZ16FA fluorescence stereomicroscope (Leica, Wetzlar, Germany). F2 seeds were placed on the microscope platform by uniformly distributing them on a sheet of paper in a line-shape with a diameter of ~ 1.5–2 cm (in total approximately 3000–5000 seeds). By moving the platform of the microscope following this line at a zoom drive of 7.17×, ten non-overlapping bright field (exposure: 1.1 s, gain: 1.0, saturation: 1.0, gamma: 1.21), RFP (exposure: 1.1 s, gain: 1.0, saturation: 1.0, gamma: 1.21) and GFP (exposure: 4.0 s, gain: 1.0, saturation: 1.0, gamma: 1.21) images were captured and stored as 2592 × 1944 pixel2 TIFF files with the 1.5 × scaling setting applied. The automatic white balance function was applied on the bright field images and switched off during capturing of the fluorescence images. The same settings were used for all images collected. Bright field images of rapeseeds and barley grains were collected using a fixed digital camera setup (Canon EOS 1100D). Because fluorescence images could not be captured for these species, the bright field images were also used as input for fluorescence classification to allow normal operation of MeioSeed.
Construction of the MeioSeed package and data analysis
The MeioSeed image analysis workflow was based on the combined  use of a CellProfiler (version 2.1.1; ) pipeline with an integrated Ilastik (version 0.5.12; ) classifier, enabling recognition and segmentation of seeds by machine learning. A random forest classifier was used in the interactive machine learning step based on user-drawn image labels on ‘White’ fluorescence stereomicroscopy images. After training the classifier was exported for integration in the CellProfiler pipeline. Further details on the operation of MeioSeed are provided in the results section. The MeioSeed package is available at http://cellprofiler.org/examples/published_pipelines. Smoothly running MeioSeed requires at least a quadcore processor. The TIFF microscopy images used as input for MeioSeed should have a format of 2592 × 1944 pixel2.
Optimization of the MeioSeed thresholding settings
Optimization of the MeioSeed ‘Filter objects’ thresholding settings was performed using fluorescence stereomicroscopy images of F2 seeds from ten independent Col3-4/20 (♂) × Ler-0 (♀) F1 hybrids. This data set was run in the MeioSeed program on a trial-and-error basis with different ‘FilterObjects’ settings for the detection of red and green fluorescence until the output of the program gave the best possible fit with the model for crossover frequency described previously . Some of the different combinations of settings which were tried are presented in Fig. 3. In order for the best fit to be achieved, a filtering step was introduced to remove microscopy images which had a clear bias for RedOnly seeds compared to GreenOnly, or vice versa, as theoretically there is no reason to assume that the RedOnly:GreenOnly ratio should be different from 1:1. To this end, images with a higher than fourfold skewed RedOnly to GreenOnly ratio (or vice versa) were disregarded for the data analysis in subsequent experiments described below. This threshold was chosen arbitrarily.
At 42 ± 2 days after germination all open flowers, open flower buds and siliques were removed from Col3-4/20x Ler-0 F1 hybrids to ensure that only meiotic crossover events after the NaCl treatments were represented among the F2 seeds. Subsequently, the plants were watered from below with fixed volumes of NaCl solutions of the indicated concentrations in demineralized water. This treatment was done once. Photos of the plants were taken every 4 days to document the effect of salt stress on the developing primary inflorescences. After seed set, the complete F2 seed pool of the primary inflorescence of each treated plant was harvested and subjected to fluorescence stereomicroscopy as described above. The distribution of the raw crossover frequency data of the 0 mM NaCl control treatments was visually examined with histograms and Q–Q plots, and was concluded to be normal. The data from the other treatments were then normalized to the control treatment with 0 mM NaCl to combine data sets, and statistically analyzed with the heteroscedastic T test function of Microsoft Excel assuming unequal variance between samples. A p value of 0.05 was used as a threshold for significance.
KASP genotyping was performed as described previously  on genomic DNA extracted from a single leaf (diameter of ~ 1 cm) of 192 F2 plants for 0 mM NaCl and 300 mM NaCl treatments, respectively (therefore n = 384 chromosomes per treatment), using the same set of SNP probes. Genetic maps were constructed using the program JoinMap .
Composition and operation of MeioSeed
Optimization of ‘FilterObjects’ thresholding settings for red and green fluorescence
Example: using MeioSeed to investigate salt stress as a novel trigger for increases in crossover frequency in Arabidopsis
To illustrate the use of MeioSeed we investigated whether it can be used to accurately detect changes in crossover frequency and whether it can be used to find novel treatments to modulate crossover frequency in Arabidopsis. To this end, we investigated whether salt stress as a known abiotic trigger for large scale chromatin remodeling [29, 30] and anther development  of Arabidopsis could be used as a novel abiotic trigger of changes in crossover frequency.
Altogether these data showed that crossover analysis of seed fluorescence reporter lines with MeioSeed could be used to screen for novel treatments or other approaches to modulate crossover frequency between reporter loci, but also at other loci in the genome.
Using MeioSeed for counting other seed species
The focus of this study was on Arabidopsis seeds because fluorescent seed reporter lines with in cis linked reporter constructs are at present exclusively available for Arabidopsis. Should fluorescent seed reporter lines become available for other model plant species or crop species as well, we were interested whether MeioSeed could also be used to count seeds of other plant species. To this end, we generated seed classifier files for rapeseeds (Brassica napus cv. Westar), which are similar to Arabidopsis seeds in shape, but larger and rather heterogeneous in size, and for barley grains, which are much larger and have a more complex morphology, which might be more difficult to recognize consistently. We subsequently captured bright field images of these seeds using a digital camera setup, because stereomicroscopy is not feasible for capturing images of larger seeds, and used these images as an input for seed counting in MeioSeed. As fluorescence images could not be collected for these species, the bright field images were run in triplicate as if they were fluorescence images, to allow for normal operation of MeioSeed as for Arabidopsis seeds. The classifier which was generated for rapeseeds was very robust and allowed for efficient counting of a heterogeneous pool of rapeseeds while excluding flower debris from the analysis (Additional file 3), while otherwise using the same settings as for Arabidopsis seeds. The classifier generated for barley grains was less robust (Additional file 4), likely due to the more complex morphology of the grains and the ridges in the seed coat, which hampered recognition. Regardless, a significant fraction of the barley grains could be counted. Together, these data showed that MeioSeed is in principle compatible with other types of seeds and images, provided that a classifier file is trained by machine learning in Ilastik and that the morphology of the seeds is not too complex.
Here, we have presented MeioSeed, a novel high throughput program to count Arabidopsis seeds from fluorescence microscopy images. Using the meiotic reporter line Col3-4/20, we showed that MeioSeed accurately counts red and green fluorescent seeds using an Ilastik-based seed classifier with adjustable detection thresholds for fluorescence. Furthermore, we demonstrated the optimization of the fluorescence threshold settings to fit the known model for recombination frequency and transgene segregation in the F2 seed progeny of crosses with Col3-4/20. Finally, to illustrate the use of MeioSeed, salt stress was investigated as a novel trigger for changes in crossover frequency, through which we found that salt stress can increase crossover frequency between the fluorescence marker genes in Col3-4/20, and can induce genome-wide fluctuations in crossover frequency.
The combined use of CellProfiler and an Ilastik-based classifier by MeioSeed is fairly new, and has only to a small extent been explored for image analysis . Previously, seed counting and image analysis relied largely on the use of other software packages such as ImageJ [32, 33], or the combined use of Adobe Photoshop and ImageJ . CellProfiler has previously been coined as the best program to reliably count seeds of various crop species , especially because it has a higher accuracy and can be used on images of different sources whereas ImageJ is most suitable for digital camera images . In addition, establishing an automated and high throughput seed counting and classification algorithm in ImageJ requires quite extensive programming experience, whereas CellProfiler allows for user friendly, module-based and linear image analysis pipelines which can easily be fine-tuned and optimized without programming knowledge. As presented here, the use of CellProfiler can also be integrated with Ilastik, allowing for very robust pixel classification and therefore more efficient recognition of seeds. It has to be noted, however, that CellProfiler lacks scripting possibilities to enable more complex workflows (e.g. involving programming loops). When combined with a robust seed classifier, CellProfiler therefore likely allows for the best available recognition and counting of seeds. A disadvantage of CellProfiler, however, is that it can be relatively slow at counting large image files. In order to have MeioSeed fluently running, we experienced that this at least requires a quadcore processor.
In the present study we have described and optimized the settings of MeioSeed for images generated by the LeicaMZ16FA fluorescence stereomicroscope. In principle, MeioSeed is compatible with images generated by any type of fluorescence microscope or digital camera setup, given that the exposure settings are such that seeds are clearly recognizable from the images. In case MeioSeed is unable to count seeds from images generated by another microscope, it is likely that optimization of the exposure settings is required. Here, we have optimized the ‘FilterObjects’ fluorescence thresholding settings (Fig. 3) based on a rather specific set of exposure settings for the LeicaMZ16FA, so it is likely that these thresholds are very different when another microscope with other exposure settings is used. We have noticed that the classifier is robust enough to deal with substantial deviations in the protocol. However, even though the classifier is very robust, it is likely that another microscope setup requires reoptimization of the ‘FilterObjects’ settings to accurately recognize the seeds and for the output to match the theoretical segregation pattern of the transgenes. Furthermore, in our experience Col3-4/20 seeds and Col3-4/20 × Ler-0 hybrids seeds display hardly any or even no autofluorescence at all in both the RFP and the GFP channels, but this might of course be different for other reporter lines or in other hybrid backgrounds. In such cases counting seeds requires assessment of autofluorescence levels of the seeds and optimization of the ‘FilterObjects’ settings to match the relevant model for recombination frequency and transgene segregation. With higher levels of autofluorescence finding a good fit between the ‘FilterObjects’ settings and the model might require more fine-tuning.
The seed classifier of MeioSeed was trained by machine-learning to specifically recognize the morphology and shape of Arabidopsis seeds, and to exclude other objects from the data analysis. This was done in an iterative process of learning from manually labelled ‘White’ images of seeds. Although the final batch processing of MeioSeed was for our purpose based on recognition of Arabidopsis seeds through the classifier, it can in principle be adapted to function for the seeds of any other plant species or for any other type of object (e.g. pollen, vegetative tissue, callus). This would in principle only require training of a classifier in Ilastik with ‘White’ images of those seeds or objects, and subsequently optimizing the ‘FilterObjects’ settings for accurate recognition depending on the levels of fluorescence reporter gene expression and background fluorescence. In fact, the classifier file for Arabidopsis seeds which was generated in this study can already be used for seeds of other plant species or for counting any other objects, but the performance of MeioSeed will drop along with increasing differences in shape and size. It might be very worthwhile to consider the combined use of CellProfiler and Ilastik for seed counting purposes in crop species, as we have shown in the present study for rapeseeds and barley grains (Additional files 3 and 4). Seed count is considered to be an important agronomic parameter and is commonly used as a measure of yield [34, 35]. For applications in meiosis research, similar to our present study, there are to our knowledge no seed reporter lines harboring in cis linked fluorescence reporter genes yet available for crop species. Such lines might very well become available for any given crop species, but it might be technically more challenging than for the model plant Arabidopsis. Obviously, the efficiency of fluorescence detection is likely to depend on the thickness of the seed coat and on the amount of endosperm relative to the size of the embryo; thin-walled Arabidopsis seeds which also contain very little endosperm thus allow for a clear view on the embryos representing the F2 population that express the fluorescence markers. For other species, the situation is likely to be less favorable.
Using MeioSeed we have found that salt stress can act as a novel abiotic modulator of crossover frequency in Arabidopsis. Our data therefore indicate that fluorescent seed or pollen reporter lines are viable options to screen for novel abiotic treatments to enhance crossover frequency in plants. In principle this just requires the exposure of F1 plants (homozygous reporter line × other homozygous accession) to abiotic treatments and fluorescence microscopy analysis of the F2 seed pool or the F1 pollen. It has to be noted, however, that screening for treatments using fluorescence reporter lines will require further investigation of candidate F2 populations by genotyping and/or sequencing to get a complete overview of the genome-wide impact of the treatment, and of the actual positions and the number of crossovers. In addition, whether or not an effect on crossover frequency can be observed in fluorescence reporter lines might depend on the position of the markers in the genome, meaning that it might be worthwhile to screen for treatments using multiple reporter lines in parallel. Previously, temperature stress was reported to enhance crossover frequency in Arabidopsis using a fluorescent pollen reporter line  and fluorescent seed reporter lines , further indicating that other abiotic stresses (e.g. light stress, drought stress and osmotic stress) could also be explored as novel modulators of crossover frequency. This is supported by the fact that abiotic stresses are triggers of chromatin remodeling in plants [29, 30], which is important for the accessibility of a genomic locus to a crossover event . From a more evolutionary and ecological point of view, one could also hypothesize that it would be beneficial for a plant to upregulate the frequency of crossovers upon experiencing abiotic stress to genetically diversify its offspring. Evidence for a similar hypothesis has recently been found in the case of biotic stress, where recombination rate was found to be enhanced at disease resistance loci .
We have shown that MeioSeed allows for the high throughput counting of green and red fluorescent seeds for meiotic crossover analysis, and presented an overview of the optimization steps for our microscopy setup. These settings could be adjusted and optimized for any other fluorescence microscopy setup. In addition, we showed that MeioSeed can also be used for other plant species by retraining the classifier for those seeds, and can thus in principle be used for crossover frequency analysis in other types of seeds as well, provided that reporter lines become available. Using salt stress as an example, we showed that MeioSeed in combination with fluorescence reporter lines can be used to screen for novel treatments affecting crossover frequency in Arabidopsis.
NvT, MR and BvdZ designed the study. PvL wrote the MeioSeed program. MR optimized the MeioSeed thresholding settings. NvT and MR performed the experiments and analyzed the data. NvT drafted the manuscript. BvdZ revised the manuscript. All authors read and approved the final manuscript.
We would like to thank Johan Pinas, Nick Surtel and Jessica van Bokkum for technical assistance, Bastiaan de Snoo and Nathalie Dek (employees of Rijk Zwaan) for help with KASP genotyping and data analysis, and Prof. Dr. Paul Hooykaas for critically reading the manuscript.
PvL is an employee of Rijk Zwaan. BvdZ has received funding from Rijk Zwaan.
Availability of data and materials
The MeioSeed program is publicly available at http://cellprofiler.org/examples/published_pipelines. Raw data are available upon reasonable request and with permission from Rijk Zwaan.
Consent for publication
All authors have given consent for the data to be published.
Ethics approval and consent to participate
This research was supported by a partnership between Rijk Zwaan and the Dutch Technology Foundation STW (partnership Project Number 12427), which is the applied science division of NWO, and the Technology Programme of the Ministry of Economic Affairs, Agriculture and Innovation.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.
- Mercier R, Mezard C, Jenczewski E, Macaisne N, Grelon M. The molecular biology of meiosis in plants. Annu Rev Plant Biol. 2015;66:297–327.View ArticlePubMedGoogle Scholar
- Keeney S, Giroux CN, Kleckner N. Meiosis-specific DNA double-strand breaks are catalyzed by Spo11, a member of a widely conserved protein family. Cell. 1997;88(3):375–84.View ArticlePubMedGoogle Scholar
- Keeney S, Neale MJ. Initiation of meiotic recombination by formation of DNA double-strand breaks: mechanism and regulation. Biochem Soc Trans. 2006;34(Pt 4):523–5.View ArticlePubMedGoogle Scholar
- Osman K, Higgins JD, Sanchez-Moran E, Armstrong SJ, Franklin FC. Pathways to meiotic recombination in Arabidopsis thaliana. New Phytol. 2011;190(3):523–44.View ArticlePubMedGoogle Scholar
- Lawrence EJ, Griffin CH, Henderson IR. Modification of meiotic recombination by natural variation in plants. J Exp Bot. 2017;68(20):5471–83.View ArticlePubMedGoogle Scholar
- Ferdous M, Higgins JD, Osman K, Lambing C, Roitinger E, Mechtler K, Armstrong SJ, Perry R, Pradillo M, Cunado N, et al. Inter-homolog crossing-over and synapsis in Arabidopsis meiosis are dependent on the chromosome axis protein AtASY3. PLoS Genet. 2012;8(2):e1002507.View ArticlePubMedPubMed CentralGoogle Scholar
- Kurzbauer MT, Uanschou C, Chen D, Schlogelhofer P. The recombinases DMC1 and RAD51 are functionally and spatially separated during meiosis in Arabidopsis. Plant Cell. 2012;24(5):2058–70.View ArticlePubMedPubMed CentralGoogle Scholar
- Choi K, Zhao X, Kelly KA, Venn O, Higgins JD, Yelina NE, Hardcastle TJ, Ziolkowski PA, Copenhaver GP, Franklin FC, et al. Arabidopsis meiotic crossover hot spots overlap with H2A.Z nucleosomes at gene promoters. Nat Genet. 2013;45(11):1327–36.View ArticlePubMedGoogle Scholar
- Perrella G, Consiglio MF, Aiese-Cigliano R, Cremona G, Sanchez-Moran E, Barra L, Errico A, Bressan RA, Franklin FC, Conicella C. Histone hyperacetylation affects meiotic recombination and chromosome segregation in Arabidopsis. Plant J. 2010;62(5):796–806.View ArticlePubMedGoogle Scholar
- Sanchez-Moran E, Jones GH, Franklin FC, Santos JL. A puromycin-sensitive aminopeptidase is essential for meiosis in Arabidopsis thaliana. Plant Cell. 2004;16(11):2895–909.View ArticlePubMedPubMed CentralGoogle Scholar
- Koornneef M, Fransz P, de Jong H. Cytogenetic tools for Arabidopsis thaliana. Chromosome Res. 2003;11(3):183–94.View ArticlePubMedGoogle Scholar
- Lysak MA, Fransz PF. From transposon to chromosome and polyploidy. An update on cytogenetics and genomics of Arabidopsis. Chromosome Res. 2014;22(2):99–101.View ArticlePubMedGoogle Scholar
- de Jong H. Visualizing DNA domains and sequences by microscopy: a fifty-year history of molecular cytogenetics. Genome. 2003;46(6):943–6.View ArticlePubMedGoogle Scholar
- He C, Holme J, Anthony J. SNP genotyping: the KASP assay. Methods Mol Biol. 2014;1145:75–86.View ArticlePubMedGoogle Scholar
- Schmid KJ, Sorensen TR, Stracke R, Torjek O, Altmann T, Mitchell-Olds T, Weisshaar B. Large-scale identification and analysis of genome-wide single-nucleotide polymorphisms for mapping in Arabidopsis thaliana. Genome Res. 2003;13(6A):1250–7.View ArticlePubMedPubMed CentralGoogle Scholar
- Melamed-Bessudo C, Yehuda E, Stuitje AR, Levy AA. A new seed-based assay for meiotic recombination in Arabidopsis thaliana. Plant J. 2005;43(3):458–66.View ArticlePubMedGoogle Scholar
- Francis KE, Lam SY, Harrison BD, Bey AL, Berchowitz LE, Copenhaver GP. Pollen tetrad-based visual assay for meiotic recombination in Arabidopsis. Proc Natl Acad Sci USA. 2007;104(10):3913–8.View ArticlePubMedPubMed CentralGoogle Scholar
- Wu G, Rossidivito G, Hu T, Berlyand Y, Poethig RS. Traffic lines: new tools for genetic analysis in Arabidopsis thaliana. Genetics. 2015;200(1):35–45.View ArticlePubMedPubMed CentralGoogle Scholar
- Pecinka A, Fang W, Rehmsmeier M, Levy AA, Mittelsten Scheid O. Polyploidization increases meiotic recombination frequency in Arabidopsis. BMC Biol. 2011;9:24.View ArticlePubMedPubMed CentralGoogle Scholar
- Melamed-Bessudo C, Levy AA. Deficiency in DNA methylation increases meiotic crossover rates in euchromatic but not in heterochromatic regions in Arabidopsis. Proc Natl Acad Sci USA. 2012;109(16):E981–8.View ArticlePubMedPubMed CentralGoogle Scholar
- Yelina NE, Choi K, Chelysheva L, Macaulay M, de Snoo B, Wijnker E, Miller N, Drouaud J, Grelon M, Copenhaver GP, et al. Epigenetic remodeling of meiotic crossover frequency in Arabidopsis thaliana DNA methyltransferase mutants. PLoS Genet. 2012;8(8):e1002844.View ArticlePubMedPubMed CentralGoogle Scholar
- Lario LD, Botta P, Casati P, Spampinato CP. Role of AtMSH7 in UV-B-induced DNA damage recognition and recombination. J Exp Bot. 2015;66(11):3019–26.View ArticlePubMedGoogle Scholar
- Ziolkowski PA, Underwood CJ, Lambing C, Martinez-Garcia M, Lawrence EJ, Ziolkowska L, Griffin C, Choi K, Franklin FC, Martienssen RA, et al. Natural variation and dosage of the HEI10 meiotic E3 ligase control Arabidopsis crossover recombination. Genes Dev. 2017;31(3):306–17.View ArticlePubMedPubMed CentralGoogle Scholar
- Logan DJ, Shan J, Bhatia SN, Carpenter AE. Quantifying co-cultured cell phenotypes in high-throughput using pixel-based classification. Methods. 2016;96:6–11.View ArticlePubMedGoogle Scholar
- Kamentsky L, Jones TR, Fraser A, Bray MA, Logan DJ, Madden KL, Ljosa V, Rueden C, Eliceiri KW, Carpenter AE. Improved structure, function and compatibility for Cell Profiler: modular high-throughput image analysis software. Bioinformatics. 2011;27(8):1179–80.View ArticlePubMedPubMed CentralGoogle Scholar
- Sommer C, Straehle C, Köthe U, Hamprecht FA. ilastik: Interactive Learning and Segmentation Toolkit. In: Eighth IEEE international symposium on biomedical imaging (ISBI) 2011, proceedings. 2011. p. 230–233.Google Scholar
- Wijnker E, van Dun K, de Snoo CB, Lelivelt CL, Keurentjes JJ, Naharudin NS, Ravi M, Chan SW, de Jong H, Dirks R. Reverse breeding in Arabidopsis thaliana generates homozygous parental lines from a heterozygous plant. Nat Genet. 2012;44(4):467–70.View ArticlePubMedGoogle Scholar
- Stam P. Construction of integrated genetic-linkage maps by means of a new computer package—Joinmap. Plant J. 1993;3(5):739–44.View ArticleGoogle Scholar
- Kim JM, Sasaki T, Ueda M, Sako K, Seki M. Chromatin changes in response to drought, salinity, heat, and cold stresses in plants. Front Plant Sci. 2015;6:114.PubMedPubMed CentralGoogle Scholar
- Luo M, Liu X, Singh P, Cui Y, Zimmerli L, Wu K. Chromatin modifications and remodeling in plant abiotic stress responses. Biochim Biophys Acta. 2012;1819(2):129–36.View ArticlePubMedGoogle Scholar
- De Storme N, Geelen D. The impact of environmental stress on male reproductive development in plants: biological processes and molecular mechanisms. Plant Cell Environ. 2014;37(1):1–18.View ArticlePubMedGoogle Scholar
- Joosen RV, Kodde J, Willems LA, Ligterink W, van der Plas LH, Hilhorst HW. GERMINATOR: a software package for high-throughput scoring and curve fitting of Arabidopsis seed germination. Plant J. 2010;62(1):148–59.View ArticlePubMedGoogle Scholar
- Herridge RP, Day RC, Baldwin S, Macknight RC. Rapid analysis of seed size in Arabidopsis for mutant and QTL discovery. Plant Methods. 2011;7(1):3.View ArticlePubMedPubMed CentralGoogle Scholar
- Mussadiq Z, Laszlo B, Helyes L, Gyuricza C. Evaluation and comparison of open source program solutions for automatic seed counting on digital images. Comput Electron Agric. 2015;117:194–9.View ArticleGoogle Scholar
- Bilsborrow P, Cooper J, Tetard-Jones C, Srednicka-Tober D, Baraniski M, Eyre M, Schmidt C, Shotton P, Volakakis N, Cakmak I, et al. The effect of organic and conventional management on the yield and quality of wheat grown in a long-term field trial. Eur J Agron. 2013;51:71–80.View ArticleGoogle Scholar
- Saini R, Singh AK, Dhanapal S, Saeed TH, Hyde GJ, Baskar R. Brief temperature stress during reproductive stages alters meiotic recombination and somatic mutation rates in the progeny of Arabidopsis. BMC Plant Biol. 2017;17(1):103.View ArticlePubMedPubMed CentralGoogle Scholar
- Marand AP, Jansky SH, Zhao H, Leisner CP, Zhu X, Zeng Z, Crisovan E, Newton L, Hamernik AJ, Veilleux RE, et al. Meiotic crossovers are associated with open chromatin and enriched with Stowaway transposons in potato. Genome Biol. 2017;18(1):203.View ArticlePubMedPubMed CentralGoogle Scholar
- Choi K, Reinhard C, Serra H, Ziolkowski PA, Underwood CJ, Zhao X, Hardcastle TJ, Yelina NE, Griffin C, Jackson M, et al. Recombination rate heterogeneity within Arabidopsis disease resistance genes. PLoS Genet. 2016;12(7):e1006179.View ArticlePubMedPubMed CentralGoogle Scholar