Live Imaging Characterization of Centromere Movements During Male Meiotic Prophase in Arabidopsis thaliana

Summary

Chromosome movement during meiotic prophase is a poorly understood phenomenon that is fundamental to spore formation. This protocol describes a method to acquire, analyze, and quantify centromere movements during meiotic prophase in Arabidopsis thaliana male meiocytes.

Abstract

Meiosis is a cellular division, essential for sexual reproduction. During meiotic prophase, homologous chromosomes of maternal and paternal origin recognize each other and associate, a process that is facilitated by their rapid movements in the nucleoplasm. Meiotic prophase movements rely on the attachment of the chromosomes to the nuclear envelope and on cytoplasmic forces transmitted to the chromosomes through the nuclear envelope. In plants, the limited accessibility and fragility of male meiocytes inside the anthers make live imaging approaches and chromosome movement analysis technically challenging. Indeed, these movements are very rapid, and analyzing them requires finding the optimal acquisition conditions. Here, an efficient method to capture rapid chromosome movements during male meiotic prophase in Arabidopsis thaliana is described, in time-lapse movies, by following the movements of fluorescent-tagged centromeres. It is based on anther dissection and image acquisition using confocal laser scanning microscopy (CLSM), in 3D and over time. Acquiring meiotic z-stacks every 7 s allows sufficient resolution to reconstruct and analyze the trajectories of tagged centromeres. The extracted data is subsequently processed, enabling the quantitative analysis of these chromosomal movements. The development of models from this data is essential for understanding the mechanism of rapid chromosome movement, including the nature and intensity of the forces driving this process.

Introduction

Meiosis is a fundamental biological process that ensures the faithful transmission of genetic material from one generation to the next. The process of meiosis is divided into two successive rounds of cell division, meiosis I and meiosis II. Meiosis I is preceded by a long prophase during which homologous chromosomes recognize each other and pair. During this extended prophase, numerous biological processes occur. Besides the pairing of homologous chromosomes, these chromosomes also undergo a process called synapsis. Synapsis involves the formation of a protein complex known as the synaptonemal complex, which connects the homologous chromosomes along their length. The synaptonemal complex promotes homologous recombination, followed by the exchange of genetic material between homologous chromosomes and the formation of crossovers. These establish a physical connection between homologous chromosomes, facilitating their proper alignment and balanced segregation^1,2.

During homologous chromosome recognition, Rapid Prophase Chromosome Movements (RPMs) are known to play a crucial role. Studies conducted in several model species have shown that during leptotene, chromosomes attach to the nuclear envelope (NE). They are connected to the cytoplasmic cytoskeleton through the NE, via the LINC (Linker of Nucleoskeleton and Cytoskeleton) protein complex³. The forces generated in the cytoplasm are transmitted to the chromosomes through the LINC complex, enabling chromosome movement.

The study of RPMs is essential for understanding the fundamental mechanisms of meiosis, particularly those involved in the recognition of homologous chromosomes. However, studying these movements requires specific live imaging experiments. Most live-imaging studies in A. thaliana focus on somatic tissues and vegetative organs. These studies have been conducted to determine how cells give rise to organs⁴. Such studies have been conducted on meristems⁵androots⁶. Recently, a few teams have developed techniques for observing floral development, including meiosis, in real-time in A. thaliana over periods of several days^7,8. However, these protocols can also be demanding and difficult to set up for large numbers of samples.

This article presents a detailed protocol for live-cell imaging and analysis of chromosome dynamics during meiosis in A. thaliana. Through the use of confocal microscopy, a simple technique for sample preparation, in conjunction with fluorescent labeling of chromosomes, chromosome movement was captured with high spatial and temporal resolution over short time periods. The protocol outlines the preparation of plant tissue for imaging, the optimization of imaging parameters to minimize phototoxicity and maximize image quality, as well as the image processing and data analysis. This method enables precise quantification of centromere dynamics during meiosis, which is crucial for understanding chromosome behavior underlying successful gamete formation. This knowledge has direct implications for agriculture, particularly in improving crop breeding programs by facilitating the study of chromosome pairing and recombination processes that affect fertility and genetic diversity.

Protocol

This protocol uses transgenic A. thaliana plants expressing GFP-CENH3 and NUP54-RFP transgenes^9,10. The CENH3 marker allows tracking the movement of the centromeres. The NUP54 marker labels the nuclear pores and thus helps to distinguish the nucleus. Other markers can be used, but these must be resistant to photobleaching as the sample is continuously illuminated. To ensure that the marker does not undergo photobleaching, perform the experiment under the same laser conditions and acquisition parameters, and check that the signal remains stable over the duration of image acquisition. The reagents and the equipment used are listed in the Table of Materials.

1. Plant growth and culture

1. Sow seeds directly in pots in the greenhouse.
2. Place the pots in a greenhouse under the following conditions: 16 h day/8 h night, 20 °C, 70% humidity.
3. Allow plants to develop in the greenhouse for approximately 4 weeks. Plants must reach a stage where they have 5/6 flower stems (see Figure 1A).

2. Dissecting anthers and slide preparation

1. Prepare a glass slide, place a spacer (double-sided adhesive with a 9 mm diameter perforation, 0.12 mm deep) at its center, taking care to remove the protective films on each side. This spacer will hold the sample and create space between the slide and coverslip.
2. Deposit 8 µL of tap water into the center of the spacer.
3. Using forceps, pick an inflorescence from one of the main stems (see Figure 1B) (do not use small inflorescences from the emerging stems).
4. Place the inflorescence on a new slide under a stereomicroscope fitted with a ruler. Select flower buds approximately 0.5 mm in length (Figure 1C).
5. Using sharp needles, gently open the buds, pulling apart the sepals and petals to keep only the anthers. Take care not to damage the anthers nor to let them dry.
6. Transfer delicately the anther bunches into the water in the spacer cavity. To avoid dehydration, do this immediately after removing the sepals and petals.
7. Deposit several anther bunches inside the spacer cavity.
8. Cover the slide with a coverslip, avoiding air bubbles. To do that, add 8 µL of water onto a coverslip and carefully turn the coverslip over with the drop of water on the spacer cavity. Make sure the anther bunches are immersed in a total of 16 µL of water and the coverslip is well glued to the spacer (see Figure 1D; Figure 2).

3. Confocal microscopy time-lapse observation – Imaging setup

NOTE: The way to set up the imaging parameters depends on the confocal system used. For more consideration of the imaging parameters, see the Discussion section. Images were acquired using CLSM and a 20x water immersion lens. Using a water immersion objective provides a better refractive index match, as the biological sample is mounted in water.

1. Set up the microscope by exciting GFP at 488 nm and detecting it with a hybrid detector between 494 nm and 547 nm.
   NOTE: The hybrid detector offset gain was set to 100. RFP is excited at 561 nm and detected with a hybrid detector between 584 and 629 nm. The hybrid detector offset gain was set to 130. The scan speed is 400 Hz, and the scan is bidirectional. A line average of 2 is applied. The two channels were acquired simultaneously to minimize acquisition time. The laser power was set to 15% for the 488 nm laser and 30% for the 561 nm laser.
2. Localize the anther using brightfield illumination.
3. Using the brightfield channel, determine the meiotic stage based on the shape of the cells.
   NOTE: Zygotene or pachytene stages are required to observe RPMs. Variation of the focus along the Z-axis enables estimation of the cell’s 3D shape. At the leptotene stage, meiocytes exhibit a distinctly square shape, and the nucleolus is positioned at the center of the nucleus, making them easily identifiable. During the zygotene and pachytene stages, meiocytes begin to adopt a trapezoid shape, and the nucleolus migrates toward the nuclear envelope. By the diplotene stage, meiocytes display a fully rounded morphology¹⁰.
4. Visualize the anther on the software. The RFP signal is also a good indicator of the meiotic stage of the cells.
   NOTE: Meiocytes at the zygotene/pachytene stage exhibit NUP54 labeling in a crescent shape, corresponding to a clustering of nuclear pores in a restricted territory of the NE at these stages (Figure 3A). If necessary, rotate the anther so that it is horizontally located (see Figure 3).
5. Define an image size of 400 px by 150 px, with a digital zoom of 4, centering the anther in the image.
6. Define the thickness of the sample to be imaged in order to obtain as many intact meiocytes as possible. However, it is important to limit the number of z-slices to minimize acquisition time. For a thickness of 15 µm, 16 z-slices are made. Z-stacks were acquired with a step size of 1.04 µm.
7. In order to perform continuous acquisitions, set the time interval to zero in the time section. With the microscope and the parameters described above, the acquisition time for 16 slices is about 6–8 s. The minimum acquisition duration is 2 min, but it can go up to 20 min.
8. Start the acquisition.

4. Image analysis – Centromere movement tracking

NOTE: Reconstruction of the centromere movements requires image analysis software that is able to track 3D motion. The software will assign an object to a fluorescent signal, based on the parameters defined, at each acquisition time. Then, the tracking tool automatically analyzes moving objects and creates a lineage plot.

1. Image pre-processing
   1. To facilitate the analysis of movements, perform post-acquisition image processing. For example, an adaptive information extraction process (deconvolution algorithm) can be used to reveal fine structures and details and improve the image quality.
2. Global centromere tracking
   1. Track all the centromeres of the image: Use a spot module to automatically track spots (over time). 
   2. Choose the source channel, green in this case. Indicate the XY diameter, 1 µm, for the GFP-CENH3-marked centromere. Object detection identifies all objects in the image (Figure 3B).
   3. Modify the quality of the detection based on the signal intensity using the graph at the bottom of the control window.
   4. Use an auto-regressive motion tracking algorithm to trace the trajectories (Figure 3C). Visualizing all the trajectories makes it easy to identify meiocytes from somatic tissues (Video 1).
3. Tracking centromeres in a single meiocyte
   NOTE: Using a 3D visualization software, check the nuclei and the trajectories, and analyze only nuclei that are intact and have not been cropped at any point during the acquisition. Analysis of a trimmed meiocyte results in erroneous data.
   1. Automatic centromere tracking
      NOTE: When analyzing a single nucleus, the process is restarted from the beginning.
      1. Use the spot module to track spots (over time) on the selected nucleus, defining a corresponding Region of Interest (ROI) using the option Segment only a Region of Interest and track spots (over time). 
      2. Select an ROI that properly includes the nucleus in all three dimensions and over time. (Figure 4A).
      3. Use the same parameters as for the whole anther for centromere detection (Figure 4B) and tracking (Figure 4C).
      4. In a lineage plot (Figure 4F), all trajectories are indicated. The trajectories outside the analyzed meiocyte can be removed (Figure 4C). Select the tracks and delete them using the Delete key on the keyboard. These can be easily identified using the software's 3D viewer.
   2. Manual centromere tracking correction
      NOTE: In most cases, the trajectories need to be manually corrected because tracking software cannot differentiate between centromeres that are too close to each other.
      1. Use the Edit mode to delete or add an object. Make sure that the object is added in the correct plane. 
      2. Manually connect the new objects in Edit Tracks mode using the lineage plot, after a careful and thorough personal analysis of the cell in 3D. Figure 4E,G show the corrected trajectories on the image of the meiocyte and on the linear plot, respectively. Video 2 shows the trajectories of one meiocyte after correction.
      3. Exporting data
         1. Export the tracking data into a spreadsheet program format to facilitate further data processing and analysis. These files contain measurement tables resulting from the tracking performed with the software.
         2. Make sure that the exported spreadsheet file contains the following variables: Time, which records the time point of each measurement; track ID, which uniquely identifies each track; Displacement Delta X, Y, Z, which represents the change in position in three dimensions at each time point; and Position X, Y, Z, which stores the absolute position of each object in space.

5. Data analysis

NOTE: This part describes the data analysis workflow using R¹¹ scripts executed in RStudio¹². They can be retrieved from the Data INRAE repository at: 10.57745/V1NNFI. All figures are generated using ggplot2¹³ to visualize the computed metrics. The scripts sequentially perform the following steps: The meiomove-sub-config.R script is first executed to initialize the R environment, load the required libraries, and define global variables. The link to the script is: https://doi.org/10.57745/V1NNFI

1. Reading and selecting data
   NOTE: The script to use is stored as: meiomove-sub-loader.R."
   1. Gather all individual Excel files containing tracking data and merge them into a single consolidated file.
   2. Save the merged file in TSV format (tab-separated values) to ensure compatibility with further data processing steps.
   3. Compute the x, y, z positions by accumulating the displacement values over time, ensuring that the spatial trajectory of each tracked object is reconstructed accurately. This was done to compensate for a bug in the software that exported incorrect XYZ positions.
2. Data cleaning
   NOTE: The script to use is stored as: meiomove-sub-clean.R."
   1. Remove short tracks, defined as those with a minimum duration of 50 s. These tracks do not contain enough data points for meaningful analysis, as they may introduce noise or inaccuracies.
   2. Identify and delete tracks where the position (X, Y, Z) remains constant over time, as these indicate either artifacts or static structures that are not relevant for the analysis.
   3. Filter out missing displacement values (NA entries) to ensure data integrity and avoid computational errors in later steps.
   4. Detect and delete tracks where the displacement values obtained from image analysis software differ significantly from those recalculated, ensuring consistency and reliability in tracking data.
3. Correcting meiotic cell drift
   NOTE: The script to use is also stored as: meiomove-sub-clean.R."
   1. Account for possible drift during acquisition by distinguishing between meiotic and somatic centromeres. This step ensures that observed movements correspond to biological processes rather than experimental artifacts.
   2. Apply drift correction by adjusting the positions of meiotic cells relative to the somatic cells, which serve as a stable reference.
4. Output generation
   NOTE: The script to use is stored as: meiomove-sub-metrics.R."
   1. Compute the instantaneous speed at each time step to track variations in movement velocity over time.
   2. Calculate the turning angle, which represents the change in direction of movement between consecutive time steps, providing insights into trajectory patterns. The result is stored as: results/dataframes/metrics-by-step.tsv (Output 1).
   3. Compute the outreach ratio using celltrackR¹⁴, which quantifies how far a tracked object moves relative to its starting position, giving an indication of exploratory behavior. The result is stored as: results/dataframes/metrics-celltrackr.tsv (Output 2).
   4. Compute the average speed for each individual track to summarize movement characteristics over the entire observation period. Calculate the average turning angle per track to analyze overall movement directionality and behavioral patterns. The result is stored as: results/dataframes/metrics-by-id.tsv (Output 3).
   5. Determine the centroid position at each timestep for each cell track. The centroid is the average of centromere positions in a given meiocyte.
   6. Compute the centroid size at each timestep, which provides a measure of the spatial dispersion of centromeres around their average position. Variations of centroid size over time give information on global consistency across centromere tracks. The result is stored as: results/dataframes/metrics-by-cell.tsv (Output 4).
   7. Compute speed cross-correlation to analyze relationships between the movement dynamics of different tracked objects, revealing potential coordinated behaviors. The result is stored as: results/dataframes/metrics-correlations.tsv (Output 5).
   8. Compute the mean squared displacement (MSD) for each track, a key metric that quantifies the average squared distance travelled by centromeres over given time intervals (time lags)  to characterize motion type.
      NOTE: By examining how the MSD evolves with increasing time lag, one can infer whether the motion is predominantly diffusive or confined, thus providing insight into the underlying dynamic behavior. The result is stored as: results/dataframes/metrics-msd.tsv (Output 6).

Results

A rapid protocol was developed to visualize chromosome movements during the prophase of meiosis in Arabidopsis thaliana, allowing multiple time-lapse acquisitions and, thus, the analysis of a large dataset. Meiocytes expressing a GFP marker, allowing for the visualization of centromeres, and an RFP marker, allowing for the visualization of the nucleus, are imaged for a duration of 4 min following the described protocol.

Discussion

The current protocol enables in vivo imaging and analysis of centromere movement during meiotic prophase in Arabidopsis thaliana. This protocol can be implemented with either an upright or an inverted confocal microscope, making it accessible to most laboratories equipped with confocal technology. Although confocal systems vary in scanning speed, sensitivity, detectors, and the ability to separate wavelengths, it is possible to adjust the acquisition parameters to obtain images with good temporal and sp...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Breeding ground	TREF (Jiffy)	101820/02227861
Confocal microscope	Leica	TCS SP8 AOBS (Acousto-Optical Beam Splitter)
Confocal software	Leica	LAS X 3.5.0.18371 software
Coverslips	Paul Marienfeld GmbH	107052	22x22 No1,5H
Forceps	Ideal-Tek	4.SA.0	High precision tweezers
GFP:CENH3		Ravi et al. 2010	Arabidopsis thaliana seeds used here
ggplot2		ggplot2: Elegant Graphics for Data Analysis  Reference13	https://ggplot2.tidyverse.org
Handle for needles	Hammacher	Handle for loop, length 170 mm
Imaris	Oxford instruments	Imaris 9.7.2	Image analysis - "Spot" module
Lens	Leica	HC PL APO CS2 20×/0.75 IMM
Lightning	Leica	Adaptive Deconvolution	Automated intelligent information extraction from confocal data
Lights	VEGELED	Apollo LL252-Gen2
Needles	Watkins&Doncaster	E3 pins .0124"(.31mm) x 25mm
NUP54:RFP		Cromer et al. 2024	Arabidopsis thaliana seeds used here
Pots	TEKU (POPPELMANN)	VDF 7x7x6,5 cm
R	RCore Team	R: A language and environment for statistical   computing. R Foundation for Statistical Computing, Vienna, Austria.  Reference11	https://www.R-project.org/
Rstudio	Posit Team	RStudio: Integrated Development Environment for R. Posit Software, PBC, Boston, MA. Reference12	http://www.posit.co/
Slides	KNITTEL GLASS	100003	26x76x1,0mm
Spacers	Thermo-fisher	Secure-Seal Spacer, 8 wells, 9 mm diameter, 0.12 mm deep	cut each spacer individually
Stereomicroscopes	Nikon	SMZ800