We present an automated method for three-dimensional reconstruction of the Caenorhabditis elegans germline. Our method determines the number and position of each nucleus within the germline and analyses germline protein distribution and cytoskeletal structure.
The Caenorhabditis elegans (C. elegans) germline is used to study several biologically important processes including stem cell development, apoptosis, and chromosome dynamics. While the germline is an excellent model, the analysis is often two dimensional due to the time and labor required for three-dimensional analysis. Major readouts in such studies are the number/position of nuclei and protein distribution within the germline. Here, we present a method to perform automated analysis of the germline using confocal microscopy and computational approaches to determine the number and position of nuclei in each region of the germline. Our method also analyzes germline protein distribution that enables the three-dimensional examination of protein expression in different genetic backgrounds. Further, our study shows variations in cytoskeletal architecture in distinct regions of the germline that may accommodate specific spatial developmental requirements. Finally, our method enables automated counting of the sperm in the spermatheca of each germline. Taken together, our method enables rapid and reproducible phenotypic analysis of the C. elegans germline.
The conservation of signaling pathways with mammals makes C. elegans an excellent model to study multiple biological processes1,2. In our lab, we use the C. elegans germline to study stem cell development, apoptosis, and gene expression. While the germline is a three-dimensional structure, many studies are two dimensional due to the time-consuming and labor-intensive nature of three-dimensional analysis. It is highly likely that two-dimensional analysis may misrepresent in vivo events in the germline. The C. elegans adult hermaphrodite has two germline arms, each of which houses a somatic distal tip cell (DTC) that maintains distal germ cells in an undifferentiated state3,4. These germ cells begin to differentiate as they move away from the DTC, escaping its influence, and become oocytes and sperm as they reach the proximal end of the germline. During this process, germ cell nuclei undergo mitosis, before transitioning to meiosis5,6. Sperm production is completed by larval stage 4 (L4) of the development, after which oocytes are produced during the adulthood. The sperm are stored in the spermatheca where they fertilize oocytes to generate embryos.
There are multiple genetic and environmental factors that can influence germline development in C. elegans resulting in changes in the number of nuclei, number of apoptotic events, chromosome dynamics, and protein expression and/or localization7,8,9,10,11. The analysis of these events requires the identification of each stage of differentiation based on nuclear morphology and distribution. To accurately analyze these parameters manually with a large sample size is labor-intensive and time-consuming. To circumvent these drawbacks and to enable the consistency of analysis, we developed an automated method for three-dimensional examination of the C. elegans germline for nuclei counting, nuclei distribution, protein expression, and cytoskeletal structure. By combining confocal microscopy with three-dimensional rendering, we generated size and shape parameters for the identification of each stage of germ cell differentiation. Further, this method enables counting of germ cell nuclei and sperm plus scoring of chromosome number in each oocyte.
One crucial structure in the germline is the cytoskeleton, which provides stability to the germline compartment, aids cytoplasmic streaming and protection to germline nuclei12. Using computational rendering, we performed three-dimensional reconstruction of the germline cytoskeleton and identified distinct cytoskeletal features within the germline. Here, we describe a step-by-step protocol to illustrate how computational analysis combined with confocal imaging enables comprehensive analysis of the C. elegans germline.
We propose a rapid method for the three-dimensional analysis of C. elegans germline (Figure 1). Using three-dimensional analysis, it is possible to study the three-dimensional distribution of germline nuclei (Figure 2 and Figure 3), automated counting of cells (Figure 2), reconstruction of the germline cytoskeleton (Figure 3), distribution of proteins (Figure 4), and scoring the number of sperm in the spermatheca and chromosomes in oocytes (Figure 5). The method not only enables easy and accurate quantification of the germline but identifies physiologically relevant phenotypes.
1. Preparation and Worm Husbandry
Note: Refer Table of Materials for all product information.
2. Germline Dissection and Staining6,13
3. Confocal Microscopy
4. Post Imaging Analysis of Nuclei Number and Distribution
Note: Refer Supplementary Figure 1 for screenshots of the software tools and buttons used.
5. Post Imaging Analysis for Scoring Sperm and Chromosome Number
6. Post Imaging Analysis for Cytoskeletal Reconstruction of the Germline
Figure 1 indicates the time required for three-dimensional germline analysis. L4 hermaphrodites incubated at 20 °C were dissected to isolate germlines and stained with DAPI, phalloidin, and antibodies against germline proteins. Germlines are imaged using confocal microscopy. Staining and confocal microscopy requires approximately 24 h. Computational analysis for the complete germline requires 10 – 15 min to count the number and position of nuclei, identify the protein distribution, analyze the cytoskeletal structure and score the number of sperm. Complete manual analysis requires over 2 h per germline and is subject to operator influence, therefore, our automated high-throughput method reduces the analysis time and enhances the reproducibility.
Automated nuclei scoring reveals that a germline arm houses approximately 1,000 – 1,200 nuclei (Figure 2C), corresponding well with previously published manual scoring data1. To confirm the accuracy of scoring, multiple mutants and different conditions are used (Figure 2D-G). For instance, we compared rnp-8(tm435), cpb-3(bt17), and glp-1(e2141) mutant worms to wild type animals. Automated counting reproduced the known reduction of number of nuclei in cpb-3 and glp-1 mutant germlines14,15,16,17. A severe reduction in nuclei number was also observed in the glp-1(e2141) temperature-sensitive mutant when incubated at 25 °C. The distribution of nuclei is dependent on the stage of differentiation. The mitotic region, which contains approximately 250 nuclei in wild type, is tightly-packed with nuclei compared to the rest of the germline (Figure 3)1. The spacing between mitotic nuclei is minimal and nuclei are located throughout the mitotic region. However, as the nuclei move away from distal end, they appear more towards the circumference of the germline (Figure 3). The change in distribution starts at the transition zone and completes as the nuclei enter meiosis/pachytene.
The germline has specific cytoskeletal structures at different stages of differentiation. We visualized F-actin by staining the germline with phalloidin. The distribution of F-actin from the distal to proximal end of the germline was distinct at each region (Figure 3 and Figure 5). In general, there are two separate structures that appear as a 'cylinder within cylinder.' At the mitotic region, inner actin appears as a solid mass with a specific shape (Figure 3). As the germline reaches the transition zone, the inner actin assumes a more cylindrical structure and it becomes a hollow cylinder as it reaches late pachytene. The second layer of actin covers the inner layer giving shape to the germline. This 'cylinder within cylinder' actin structure disappears towards the oocyte region (Figure 5). At the oocyte region, actin appears as thick fibers. The oocytes are separated by an actin rachis, though it appears to be dynamic to allow the movement of oocytes to the spermatheca. Finally, actin in the spermatheca also forms thick bundles of actin fibers. However, the fibers appear to be packed closer than in the oocyte region. The cytoskeletal structure appears to complement the distribution patterns of nuclei (Figure 3). At the mitotic region, the nuclei appear to be placed around the inner actin structure. As they reach meiosis/pachytene, the nuclei organize between the two actin cylinders leaving the middle of the germline mostly devoid of nuclei (Figure 3). This could probably aid uninterrupted cytoplasmic streaming in the germline. Germlines were stained with an antibody against the REC-8 protein and analyzed by confocal microscopy. REC-8 is homogeneously distributed around early nuclei of the germline and is later cleaved and degraded during meiosis18. REC-8 distribution in three-dimensional cross-sectional analysis of the germline shows the distribution of protein in the mitotic region (Figure 4).
Three-dimensional rendering of the proximal end of the germline includes the spermatheca and first three oocytes distal to the spermatheca. Our analysis recognized an average of 151 sperm in each spermatheca (n = 18). This corresponds to the published literature, indicating the reliability of the method (Figure 5F)19. This analysis may be performed together with cytoskeletal or protein distribution studies. The C. elegans genome is distributed among six chromosomes. In fully developed wild type oocytes, these chromosomes are visible and can be counted. This chromosome separation can be used to study chromosome stability or other defects associated with oocyte development. By three-dimensional rendering, the number of chromosomes can be visualized, and if the chromosomes are separated properly, the number can be accurately obtained (Figure 5).
Figure 1. Timeline for automated analysis. Time comparison between manual and computational analysis of the C. elegans germline. Please click here to view a larger version of this figure.
Figure 2. Nuclei distribution in the germline. (A) DAPI stained germline showing mitotic, transition, and pachytene/meiotic regions. (B) Computational three-dimensional model of the germline. (C) Scoring for the total number of nuclei in wild type germlines. (D-F) Scoring for the number of nuclei at mitotic, transition, and meiotic regions of wild type, rnp-8, cpb-3, and glp-1 mutant germlines. (G) Comparison in the total number of nuclei in the glp-1 mutant at 20 °C and 25 °C. Error bar represents standard error. Student's test. ****n <0.0001, ***n <0.001, **n <0.01, *n <0.05. Please click here to view a larger version of this figure.
Figure 3. Cytoskeletal structure of the germline. (A-B) Organization of the inner actin (red) cytoskeleton of the germline at the mitotic region. The inner actin has a diffuse structure compared to outer actin at the mitotic region. (C-D) At the pachytene region, the cytoskeleton assumes a cylindrical structure by forming two cylinders between which the nuclei (blue) are organized. (E-F) Organization of germline nuclei in the mitotic and pachytene regions. The distribution of nuclei is more towards the circumference of the 'germtube' at the pachytene region compared to mitotic region (green) and middle of the germline become devoid of nuclei. Please click here to view a larger version of this figure.
Figure 4. REC-8 expression in the mitotic region. (A) Micrograph showing the distal end of a REC-8-stained germline. (B-C) Three-dimensional rendering of REC-8 staining (shown in blue) and distribution of protein between the mitotic nuclei (green). REC-8 staining is less abundant proximal to the mitotic region (marked by white dots representing nuclei). Please click here to view a larger version of this figure.
Figure 5. Germline structure at the proximal end. (A) DAPI staining of the germline at the proximal end. (B) Three-dimensional rendering of DAPI staining showing sperm in the spermatheca (yellow) and chromosomes in the oocytes (white). (C-E) Three-dimensional cytoskeletal structure at the proximal end of the germline. Red marks the complete proximal end and green marks the spermatheca. Cross sectional analysis shows that the cytoskeleton not only forms a filamentous structure of actin around oocytes, but also separates between oocytes. (F) Graph showing the number of sperm in each spermatheca and average obtained from multiple spermatheca (n = 18). Error bar represents standard error. Please click here to view a larger version of this figure.
Supplementary Figure 1. Software tools description. (A) The spots and surface tools for detecting nuclei and protein staining. (B-C) The selection of three-dimensional region of interest using the surface tool. (D) Defining the XY-diameter, using spots tool, for nuclei detection. Depending on the region of interest and staining, the Z-axis diameter and background correction can be used. (E) Minimum and maximum intensity threshold detection. A similar approach can be used for surface function where instead of defining the diameter, and the surface details can be defined. Please click here to view a larger version of this figure.
The goal of this protocol is to improve the accuracy and reduce the time required for germline analysis. After standard preparation of dissected germlines, a three-dimensional model of germline nuclei is prepared by computational rendering. While allowing the observation of germline nuclei distribution in space, three-dimensional rendering calculates the number of nuclei at specific regions of the germline. The critical aspect of our method is accurate definition of size and shape parameters of nuclei. This depends on clarity of staining and magnification used to image. We confirmed the accuracy of the method by comparing to published manually-analyzed data1. To further confirm the accuracy of our method, we used multiple mutants under different conditions and confirmed the results with published literature. It also confirmed the ability of the method to detect subtle and strong phenotypes – cpb-3 and glp-1 mutants, respectively. In addition, this method shows adaptability to changes in nuclei size and shape within the germline or even between strains. This property also enables the identification and scoring of smaller structures in the germline such as chromosomes and sperm. The method also allows further troubleshooting for size and shape parameters, if required. Our analysis was able to elucidate the distribution of nuclei in the germline and showed that nuclei in mitotic and pachytene regions are organized in different patterns. By applying three-dimensional rendering to REC-8 and DAPI staining, we also showed the distribution of REC-8 within nuclei of the mitotic region.
Three-dimensional rendering of actin staining enabled detailed examination of germline cytoskeletal structures. The germline appears to have two separate actin structures, though they maintain physical contact. While the reconstruction of the cytoskeleton for the entire germline is possible as a single entity, the best results can be obtained if the regions of interest are preselected into mitotic, transition, and meiotic regions. The distal germline cytoskeleton has an undefined shape and evolves into a filamentous structure at the proximal end. Our protocol therefore enables three-dimensional reconstruction of the germline actin cytoskeleton by defining specific parameters to accommodate differences in actin distribution. The most critical aspect of the analysis is the definition of the parameters such as nuclei size and surface detail requirement. Lowering the nuclei size below the actual size will risk the software to pick bright spots within the nuclei as separate objects. Similarly, high surface details values will lead to the loss of accurate three-dimensional structures. The proposed method relies on the precision of dissection and the quality of the germline staining. A damaged germline may lead to the release of nuclei from the germline, causing inaccurate numbering. High background from the staining will also cause the inaccurate nuclei count and improper three-dimensional rendering.
Our method extends previously known automated methods of germline analysis where we are able to analyze the nuclei number and distribution/expression of germline proteins in space with a single protocol20. Secondly, our method enables scoring of smaller structures such as chromosomes within the germline. Chromosome segregation is an important aspect of oocyte development21. Using our method, it is possible to detect the number and position of chromosomes where the distance can be defined by calibrating against wild type oocytes. Any mutation affecting the chromosome separation can be studied using this tool. Also, the method can be extended to study chromosomes in embryos at the 2-cell stage. The computational analysis also provides the number of sperm in the spermatheca, number of chromosomes in an oocyte and germline cytoskeletal structures. Taken together, the method provides comprehensive analysis of the C. elegans germline using a single protocol while reducing analysis time considerably. The method eliminates the requirement of multiple tools for each analysis (nuclei number, protein distribution, and cytoskeletal structure) and removes the coding time required in any specific software. Finally, this method can be effectively extended to other worm studies including the three-dimensional analysis of sedated live worms.
The authors have nothing to disclose.
We thank Monash Microimaging for their technical support. Some strains were provided by the Caenorhabditis Genetics Center, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). This work was supported by a Monash University Biomedicine Discovery Fellowship, NHMRC Project Grant (GNT1105374), NHMRC Senior Research Fellowship (GNT1137645) and veski innovation fellowship: VIF 23 to Roger Pocock.
C. elegans strains: wild type (N2, Bristol), rnp-8(tm2435) I/hT2[bli-4(e937) let-?(q782) qIs48] (I;III), cpb-3(bt17) I, glp-1 (e2141) III | Caenorhabditis Genetics Center (CGC) | ||
OP50 Escherichia coli bacteria | Homemade | ||
Nematode Growth Media (NGM) plates | Homemade | ||
polyclonal rabbit anti-REC-8 | SDIX | 29470002 | |
Alexa 488 conjugated antibody raised in goat | Thermofisher Scientific | A-21236 | |
Cytoskeletal dye phalloidin | Thermofisher Scientific | A-12380 | |
DAPI | Thermofisher Scientific | 62248 | |
Poly-L-lysine | Sigma Aldrich | P5899 | |
Tetramisol | Sigma Aldrich | P5899 | |
MgSO4 | Sigma Aldrich | M7506 | |
1M HEPES buffer, pH 7.4 | Sigma Aldrich | G0887 | |
10X PBS pH 7.4 | Thermofisher Scientific | AM9625 | |
Tween-20 | Sigma Aldrich | P1389 | |
EGTA | Sigma Aldrich | E3889 | |
37% Paraformaldehyde solution | Merck Millipore | 1040031000 | |
Normal goat serum | Sigma Aldrich | G9023 | |
Fluoroshield fixing reagent | Sigma Aldrich | F6182 | |
Ethanol | Millipore | 1009832511 | |
Methanol | Sigma Aldrich | 34860 | |
20°C & 25°CIncubator | Any brand | ||
Light microscope | Any brand | ||
Confocal microscope | Any brand (Leica, Zeiss) | ||
Computer equipped with Imaris suit 8.4.1 or later version, full licence to use the software and Matlab software. | Bitplane | ||
Phospho buffered saline, pH 7.4 | Homemade | ||
Teflon microscope slides | Tekdon | 941-322-8288 |