Laser microdissection of semi-thin sections from plastic-embedded tissue for studying plant–organism developmental processes at single-cell resolution

Abstract

The ability to analyze the gene activity occurring within a single cell has ushered in a new understanding of complex biological processes. Furthermore, this capability has established the prerequisite technologies for the analysis of cells involved in complex pathogenic and/or symbiotic interactions. Collectively, the identification of biological models permitting the analysis of individual cells and improvements in histological technology are allowing for analyses of cells positioned within tissues and involved in complex cellular interactions at unprecedented resolution. Here, a plastic embedding procedure is used for laser microdissection of plant tissues infected with a pathogen. This technique enabled the acquisition of nucleic acids from semi-thin sections that can be used for downstream biological studies of host–pathogen interaction at the single-cell resolution.

Keywords:

• Technovit 9100
• plastic embedding
• laser microdissection
• pathogen
• development
• RNA

Introduction

The identity of a cell is defined by its function. In this regard, determining a cell's activity at the genomics level would aid in understanding cell identity and function. It is well known from histological studies that many cell types are involved vastly in different metabolic processes even from their direct neighbors (Figure 1). Due to the nature of cell integration within tissues, isolating cells from close neighbors is difficult and tends to complicate downstream analyses. The ability to isolate individual cells from their neighboring cells would improve the ability to identify individual cell chemistries and biological processes. When aided by high-throughput reverse genetic capabilities, functional studies of the genes identified from those genomics analyses can be performed without the need for traditional time-consuming genetic screens (Gandotra et al. 2013). This is particularly important when the system under study is not genetically tractable.

Figure 1. Specialization of cell types during plant–nematode interaction. A, a syncytium formed during a compatible interaction between the soybean cyst nematode (H. glycines) and soybean (G. max), three days post infection (dpi), bar = 25 µm; B, 6 dpi, bar = 50 µm; C, 9 dpi, bar = 50 µm, (this image is taken from the same series of sections, but is different than Matsye et al. 2012). The red line surrounding some cells represents the boundary of the syncytium whose development is engaged by the nematode. Black arrow represents the nematode.

The isolation of cells for biological analyses

A number of cell isolation procedures have been developed over the years in order to determine the gene activity of specific cell types. The procedures fall into two main categories: (1) cell culturing and processing procedures that lack the requirement of embedding or infiltration in a support medium to facilitate cell isolation and collection, and (2) processing procedures that require embedding or infiltration in a support medium which facilitates cell isolation and collection. Each type is discussed briefly.

Cell culture and cell sorting processes are strategies that do not require a support medium and allow for a variety of physical separation techniques to be performed. Cell culturing methods permit cells to proliferate by mitosis under defined culture conditions and then can be collected in greater numbers. Cell culturing processes can, but do not always, include tissue maceration to separate cells. The cells then are allowed to grow in culture for defined periods of time. Physical characteristics of the cells and/or marker-assisted tools are then used to separate the cells by fluorescence-activated cell sorting (FACS). Details of FACs and related methods can be obtained from a series of reviews (Fraker et al. 1995; Galbraith et al. 1999; Herzenberg & De Rosa 2000; Miura et al. 2000; Valitutti & Dessing 2000; Maric & Barker 2004; Tung et al. 2007; Zhu & Murthy 2013). An underlying commonality of cells used in these isolation procedures is that they are relatively easy to separate based on their physical or chemical characteristics.

Some cell types, particularly those in plants, are difficult to separate because of structural barriers like a cell wall and/or their position within the tissue or organ. However, FACS has been used successfully in plants (Birnbaum et al. 2003; Bargmann & Birnbaum 2009; Bargmann et al. 2013). In cases where the cells are difficult to purify, cell separation can be accomplished through histology-associated physical procedures such as laser-assisted microdissection (LM; Isenberg et al. 1976; Emmert-Buck et al. 1996, Asano et al. 2002). Unlike the FACS technique, which is limited by the need for the presence of a specific cell marker, LM has the advantage of not requiring a specific cell marker. The LM procedure entails the embedding of tissue in one of several types of tissue support media, followed by microdissection under a microscope (Isenberg et al. 1976; Emmert-Buck et al. 1996). Some of the common support media in which the tissue is embedded or infiltrated are paraffin (Goldsworthy et al. 1999; Tam et al. 1999) or those that allow better preservation of the biological molecules (i.e. RNA) through cryopreservation (Tissue-Tek® optimal cutting temperature [OCT®] Compound, Sakura Finetek, INC., Torrance, CA, USA; Bhattacharya et al. 2003; Tadros et al. 2003; Dalmas et al. 2008). There are advantages to each embedding/infiltration type. For example, paraffin is easy to maintain in the lab, requires little technical experience to use, and allows tissue to be stored and used at ambient temperature. Tissue embedded in paraffin also can potentially be stored for years at ambient temperature or temperatures as low as 4°C at which point the paraffin becomes brittle and cracks. Another major advantage of paraffin is that it preserves anatomical detail well which aids in the ability to identify the cells during LM. However, for studies focused on gene expression, the paraffin-embedding procedure has drawbacks. The major drawback of paraffin is contamination with RNAses, resulting in RNA degradation (Jonsson & Lagerstedt 1957; Nair 1958; Groelz et al. 2013; Staff et al. 2013). Furthermore, some of the early steps in the paraffin embedding procedure, prior to the casting of tissue into molds, involve temperatures >50°C which has negative impacts on the tissue. Another drawback of the paraffin-embedding strategy is that the cells must be sectioned at thicknesses of 4 µm or greater (typically 10 µm) because of the physical limitations of the paraffin support medium with regard to sectioning. In contrast, OCT-based methods rely on cold-stored tissue. The OCT-infiltrated tissue must be cryosectioned, requiring relatively more technical capability which could be a problem since more expensive sectioning devices requiring more expertise are required. Notably, OCT which is composed of water-soluble glycols and resins can compromise the tissue whereby membranes are disrupted and cellular contents are released. The formation of ice crystals is also a major problem, because of how they can puncture membranes and disrupt cellular structure. Furthermore, the fine structure of plants is not preserved well when tissue is infiltrated with OCT (Emrich et al. 2007). This feature of OCT makes the identification of specific cell types in cryopreserved tissues virtually impossible. Therefore, these characteristics of cryopreservation may allow cells to gain false or misleading structural information in a field of sectioned tissue. Thus, the desired cells cannot be collected without further staining or use of molecular markers. This problem presented by OCT is exacerbated if there are no available stains or molecular probes to detect the cells. In many cases, the identification of uniquely expressed genes is the reason why the LM technique is used in the first place. Furthermore, like paraffin embedding, cells in OCT must also be thick-sectioned, increasing the probability that undesired cells are collected.

The collection of semi-thin sections

The LM technique has been a very powerful tool in a variety of biological fields including plant and animal research where cell collection can be problematic (reviewed in Simone et al. 1998, 2000; Emmert-Buck et al. 2000; Gillespie et al. 2001; Espina et al. 2004, 2005, 2009; Klink et al. 2005; Balestrini et al. 2009; Vicidomini et al. 2010; Gandotra et al. 2013). In addition, the LM technology has allowed for the study of DNA mutations, transcriptome, proteome, metabolome, and miRNAs (Meier-Ruge et al. 1976; Isenberg et al. 1976; Hedrum et al. 1994; Emmert-Buck et al. 1996; Bernsen et al. 1998; Schütze & Lahr 1998; Sgroi et al. 1999; Ornstein et al. 2000; Jessani et al. 2002, 2004; Schad et al. 2005; Nonn et al. 2010; Cancer Genome Atlas Research Network 2011; Nishimoto et al. 2012). Unfortunately in some areas of study like plant–organism interactions, isolating cells directly involved in the process is often difficult because relatively few cells are actually involved in the interaction while the remainder of the plant body actually masks the molecular events occurring locally within the infected cells.

The number of localized responses that would be aided by the increased specificity afforded by obtaining quality RNA from tissue sections is vast (Santos et al. 2012; Ghodke et al. 2013; Gupta et al. 2013; Ozawa et al. 2013). For example, plant parasitic nematodes cause as much as 100–150 billion dollars in damage annually, worldwide (Sasser & Freckman 1987; Chitwood 2003). Therefore, with this plant–pathogen interaction alone there exists an urgency to identify the precise spatial gene expression patterns that underlie susceptibility. Since many plant pathogens infect systemically or are interwoven with the plant cells, it is difficult to differentiate between cells that are directly infected versus the cells that are passively damaged. However, LM of tissue supported in paraffin has allowed for understanding the localized gene activity in the cells undergoing infection and has led directly to the identification of resistance genes (Jammes et al. 2005; Klink et al. 2005, 2007, 2009, 2010, 2011; Matsye et al. 2011; 2012; Matthews et al. 2013; Pant et al. in press). Notably, the paraffin-embedding method, followed by LM and gene expression analyses allowed for the identification of the vesicular transport machinery being involved in the defense response of Glycine max (soybean) to Heterodera glycines (soybean cyst nematode) (Matsye et al. 2012). The identified α-SNAP resistance gene which is located at the rhg1 resistance locus is not a classical resistance (R) gene (Matsye et al. 2012). However, it could be that α-SNAP regulates other genes involved in resistance. Important contributions have also been made in plant–fungal studies (Chandran et al. 2009; Hacquard et al. 2010). Even with the ability to perform genomics analyses of collected cell populations, recent advances in single-cell genomics are showing that even cell types that appear similar or identical are undergoing vastly different genetic programs (Guo et al. 2010). Thus, to gain a better understanding of these cells, it is necessary to perform single-cell gene expression analyses to further increase the resolution of the analysis.

Plastic embedding and single-cell gene study

The application of single-cell genomics has permitted the study of various developmental processes (Schütze & Lahr 1998; Chiang & Melton 2003; Guo et al. 2010). Such a goal is complicated in plants due to the problems presented by the cell wall and the limits it places with regard to the isolation of cells. However, cell populations can be collected from complex tissues by FACs methods or LM (Asano et al. 2002; Birnbaum et al. 2003). In these analyses, hundreds if not thousands of cells are required to perform the molecular studies. As can be imagined, it is possible that each cell in those populations may actually have unique gene expression profiles based on their specific function and absolute position in the tissue or organ (Guo et al. 2010). This problem would represent a major challenge for genomics studies involving plant–organism interactions because of the varied position of the pathogen, the limited amount of sample, and technical problems involved in isolating individual cells. However, the number of localized responses that would be aided by the increased specificity afforded by obtaining quality RNA from semi-thin sections is extensive (Santos et al. 2012; Ghodke et al. 2013; Gupta et al. 2013; Ozawa et al. 2013).

One particular type of pathogen, plant parasitic nematodes present a unique opportunity whereby the plant cells that are intimately associated with the pathogen are well defined. This feature of their biology permits the collection and downstream analysis of individual cells at the genomic and molecular levels. For example, some plant parasitic nematodes create a nurse cell called a syncytium that is actually a conglomeration of many cells that fuse together due to cell wall degradation (Figure 1). As already noted, the combination of LM and Illumina® deep sequencing of nurse cells undergoing defense has led to the identification of resistance genes (Matsye et al. 2011, 2012). In contrast, a different type of nurse cell known as a giant cell that develops from a single cell can be formed in plants (e.g. cotton) from nuclear division followed by cell expansion during infection by root knot nematodes (e.g. Meloidogyne incognita), Giant cells are well suited for single-cell genomics analyses. However, giant cells form in many different plants that interact with root knot-forming plant parasitic nematodes.

Materials and methods

The methods used in this analysis follow our previously published procedures (Klink et al. 2005, 2007, 2009, 2013). All glassware and metal utensils were made RNAse-free by baking at 180°C for 8 hr. Solutions were made RNAse-free with Nano-pure® (Barnstead, Thermo Fischer Scientific Inc., Waltham, MA) water with 0.1% diethyl pyrocarbonate (DEPC; Sigma, St. Louis, MO). The DEPC was dissolved in the aqueous solutions and stirred overnight for 12 hr before autoclaving to remove the DEPC. Tissue was fixed in 75% ethanol:25% glacial acetic acid (Klink et al. 2005). The specimens were dehydrated with a graded ethanol series (75%, 90%, and 3× 100%). Tissues were infiltrated and embedded with Technovit (Electron Microscopy Sciences, Hatfield, PA) with xylene as a transitional fluid according to the manufacturer's instructions (http://www.ebsciences.com/histology/methacrylate.htm) following Klink et al. (2013). Tissue pieces were placed into Beem® capsules and polymerized at –8°C for 24 hr (Klink et al. 2013). Semi-thin sections were cut with a diamond knife at 0.8 µm with a Reichert-Jung Ultracut-E ultramicrotome (Leica Microsystems®, Germany). Several drops of DEPC-treated water were placed on PEN MembraneSlides® (Leica Microsystems®) that are set on a slide warmer set at 40°C, allowing for the plastic sections to flatten out and adhere to the slide once the DEPC-treated water evaporates. The slides were used immediately for LM once the water was completely evaporated. LM was performed on an Arcturus® Veritas® microscope (Molecular Devices, Sunnyvale, CA) at the College of Veterinary Medicine, Mississippi State University. No special treatment of the slides was required. Cells were collected on CapSure HS LCM Caps (Applied Biosystems, Foster City, CA). Cells from the LM were washed-off of the HS cap by micropipetting 20 µl of XB buffer (Applied Biosystems) onto the HS cap, and moving the solution to a microfuge tube. RNA samples were isolated with the PicoPure RNA Isolation Kit (Applied Biosystems) according to the manufacturer's instructions (Klink et al. 2005, 2007, 2009) with the addition of a DNAse treatment using DNA-free (Ambion), just before the second column wash (Klink et al. 2005, 2007). RNA quality and yield were tested with the Nanodrop® (Biorad, Hercules, CA) according to the manufacturer's instructions (Klink et al. 2013). In comparative studies, RNA was isolated from G. max roots using the UltraClean® Plant RNA Isolation Kit (Mo Bio Laboratories®, Inc.; Carlsbad, CA) and treated with DNase I to remove genomic DNA.

Results and discussion

The plastic embedding LM (pe-LM) approach, used in concert with LM and RNA isolation methodologies, has recently been demonstrated showing that high-quality RNA can be obtained from cells long understood to be difficult for their isolation (Thibaudeau & Altig 1988; Klink et al. 2013). The protocol focused on using the Technovit 9100 polymer. Technovit 9100 has been applied to the study of tadpole tooth and soft tissues and has proven to be superior in retaining antigenicity as compared to paraffin sections and other plastics (Arnold et al. 1998, 2003; Yang et al. 2003; Brorson & Reinholt 2008; Vertenten et al. 2008; Singhrao et al. 2009; Wittenburg et al. 2009; Steiniger et al. in press). Because all tissue processing steps could be done at low temperatures (−20°C or lower), RNA quality would be expected to maximize preservation (Klink et al. 2013). The developed procedure has allowed for the sectioning of semi-thin sections at a thickness of 0.8 µm (Klink et al. 2013). This thickness is far thinner than the 10 µm typically used for paraffin sectioning and LM (Klink et al. 2005). Certainly obtaining ultrathin sections, while not a focus here, is possible with Technovit 9100. A major advantage of the pe-LM procedure is that the tissue can be sectioned and collected onto slides for LM without having to etch away the plastic (Figure 2). This property is in contrast to tissues that are embedded in paraffin where the paraffin must be dissolved away prior to the LM procedure. In LM procedures using paraffin as a tissue support, the crystalline nature of the paraffin interferes with the visualization of the cells due to the refraction of the light as it passes through the paraffin. Furthermore, loss of biological molecules in the form of RNA happens as the paraffin is dissolved away (Jonsson & Lagerstedt 1958; Nair 1958; Groelz et al. 2013; Staff et al. 2013). As can be imagined, this situation may be particularly problematic for small RNA species like microRNAs (miRNAs) that would be easily liberated from the section. In contrast, the collection of serial, semi-thin sections from plastic-embedded tissues allows for the reconstruction of those section files onto LM slides. Furthermore, light passing through the plastic section does not experience the same interference/diffraction, allowing the cells to be easily visualized, even when the plastic is not removed from the section (Klink et al. 2013). Thus, the desired cells are easily observed, allowing for their collection by LM and all materials in the plastic-embedded cell are captured since no etching is done (Figure 3). With the reduction of these technological limitations presented by small sample volume in transcriptional analyses, it is possible to study the interaction of plants in relation to model plant pathogens at high resolution without the loss of significant cellular information (Figure 4), making analyses of single cells possible (Abad et al. 2008; Guo et al. 2010; Citri et al. 2011; Jang et al. 2011; Byers et al. 2012; Klink et al. 2013; Livak et al. 2013).

Figure 2. Loss of biological information during LM. A, embedded in paraffin: 1, a cell, represented in dark green with a red nucleus is embedded in paraffin as a support medium. The vertical arrow represents a 4 µm limit of sectioning. 2, paraffin is dissolved away with the loss of biological information represented in the cell and nucleus having lighter hues and arrows showing the loss of molecules out of the cell. 3, the cell after LM. B, embedded in pTechnovit 9100: 4, a cell, represented in dark green with a red nucleus is embedded in plastic as a support medium. 5, Plastic does not have to be etched away because both visualization and sectioning can be done effectively without that step, thus, no loss of biological molecules. 6, pe-LM of an ultrathin section containing a cell. The pe-LM procedure vaporizes neighboring cells while allowing for the collection of the desired cell.

Figure 3. pe-LM of a nurse cell (giant cell) formed by the root knot nematode (Meloidogyne incognita) in cotton (Gossypium hirsutum). A, Before LM. B, After LM, inset, the captured cell that was dissected out in B. Bar = 50 µm.

Figure 4. RNA quality. The x-axis represents wavelength and the y-axis represents absorbance. A, Diluted RNA isolated from a whole root sample, A260/A280 = 2.33, A260/230 = 2.37, 10.1 ng/µl. B, RNA isolated from 50 pe-LM obtained cells sectioned at 0.8 µm thickness, A260/A280 = 2.36, A260/230 = 2.10, 10.1 ng/µl.

Conclusion

Genomics-level analyses of complex cellular processes at cellular resolution are difficult to perform, particularly when the cells are buried deeply in tissues. This problem is true in plants for the study of complex interactions of pathogenic or symbiotic organisms. Recent advances in tissue processing permit the collection of semi- and ultrathin sections while minimizing the loss of biologically important molecules (Klink et al. 2013). The combination of this strategy with recently developed genomics tools is revolutionizing the study of single cells undergoing important biological interactions, making the understanding of the intricacies of complex biological processes possible in systems once thought to be non-model and genetically intractable.

Acknowledgments

This publication was made possible by the start-up support of Mississippi State University and the Department of Biological Sciences (DBS). The work is supported jointly between the College of Arts and Sciences (DBS) and the Mississippi Agricultural and Forestry Experimental Station (MAFES). The authors are thankful to Dr Robert Nichols, Cotton Inc., and Dr Larry Heatherly, Mississippi Soybean Promotion Board, for their continued support. The authors thank George Hopper, Reuben Moore, and Wes Burger (MAFES) whose support has made the work possible. Support has also been provided by Amanda Lawrence of the Institute for Imaging and Analytical Technologies, Mississippi State University.