Methods for Studying the Evolution of Plant Reproductive Structures: Comparative Gene Expression Techniques

[32]

617

gene expression techniques in plants

Whitfield, C. W., Cziko, A.-M., and Robinson, G. E. (2003). Gene expression profiles in the brain predict behavior in individual honey bees. Science 302, 296–299. Wolfinger, R. D., Gibson, G., Wolfinger, E. D., Bennett, L., Hamadeh, H., Bushel, P., Afshari, C., and Paules, R. S. (2001). Assessing gene significance from cDNA microarray expression data via mixed models. J. Comput. Biol. 8, 625–637. Yang, Y. H., and Speed, T. (2002). Design issues for cDNA microarray experiments. Nat. Rev. 3, 579–588. Zimmer, D. P., Soupene, E., Lee, H. L., Wendisch, V. F., Khodursky, A. B., Peter, B. J., Bender, R. A., and Kustu, S. (2000). Nitrogen regulatory protein C-controlled genes of Escherichia coli: Scavenging as a defense against nitrogen limitation. Proc. Natl. Acad. Sci. USA 97, 14674–14679.

[32] Methods for Studying the Evolution of Plant Reproductive Structures: Comparative Gene Expression Techniques By Elena M. Kramer Abstract

A major component of evolutionary developmental (evo-devo) genetics is the analysis of gene expression patterns in nonmodel species. This comparative approach can take many forms, including reverse-transcriptase polymerase chain reaction, Northern blot hybridization, and in situ hybridization. The choice of technique depends on several issues such as the availability of fresh tissue, as well as the expected expression level and pattern of the candidate gene in question. Although the protocols for these procedures are fairly standard, optimization is often required because of the specific characteristics of the species under analysis. This chapter describes several methods commonly used to determine gene expression patterns in angiosperms, particularly in floral tissues. Suggestions for adapting basic protocols for diverse taxa and troubleshooting are also extensively discussed.

General Considerations for Working with RNA

RNA is, by nature, a less stable molecule than DNA and there are many sources of RNase in the environment, which can lead to its rapid degradation. These facts often lead to trepidation on the part of researchers who are not experienced with RNA work. However, such concerns are largely unnecessary because simple precautions can effectively prevent

METHODS IN ENZYMOLOGY, VOL. 395

Copyright 2005, Elsevier Inc. All rights reserved. 0076-6879/05 $35.00

618

functional analyses

[32]

RNA degradation. First, researchers must always wear sterile gloves when working with RNA. Gloves should be changed if the skin or items generally handled without gloves, such as doorknobs, are touched. Many labs go so far as to have specific RNA-use benches or even rooms, but this is not an absolute requirement as long as the bench space is kept clean. Second, all plasticware (tips, pipets, tubes) should be purchased RNase-free and kept that way by maintaining them as separate lab stocks that are handled only with gloves. For instance, RNase-free supplies can be stored in a separate cabinet that is kept locked to prevent unintentional contamination. Third, glass or metalware should be made RNase-free either by baking at 235
for more than 2 h (careful baking bottle caps, most will melt!) or by treating with 0.1 M NaOH overnight followed by thorough rinsing with sterile water. Before baking, wrap bottle mouths and metal items, such as slide racks or stir bars, in aluminum foil to help keep them RNase-free after baking. Fourth, all solutions should be made with water, chemicals, stir bars, cylinders, bottles, and others, which are all RNase-free. RNase-free water can be prepared using diethylpyrocarbonate (DEPC), which is highly toxic and should be used only under a fume hood. DEPC is typically diluted in double-deionized water (ddH2O) to produce a 0.1% solution. Because DEPC is unstable once the bottle seal is broken, it is best to prepare large batches of DEPC water at one time in order to use the entire volume. Take clean bottles of various sizes (100 ml to 2 L), fill with fixed amounts of ddH2O, and place under a fume hood. Aliquot the appropriate amounts of DEPC into each bottle, tightly secure the bottle caps, and shake. Let the bottles sit in the fume hood for 2–24 h. DEPC has only a 30 min half-life in ddH2O, but autoclaving for 15–30 min will ensure that all DEPC is inactivated. It is important to note that DEPC can be added to many solutions directly, but it is not compatible with Tris or MOPS buffers. Tris and MOPS solutions can be prepared using DEPC water, and RNase-free bottles and chemicals so treatment is not necessary after solution preparation. Also, remember that the bottles that are used to make DEPC water can also be treated as RNase-free themselves after they are emptied. See http://www.ambion.com/techlib/basics/rnasecontrol/ for more information on DEPC water and RNase control methods. Methods Based on RNA Extraction

Collecting and Storing Tissue Given the instability of RNA, the collection and storage of tissue becomes a much greater concern than when the goal is DNA extraction. There are two primary methods for storing tissue for future RNA extraction:

[32]

gene expression techniques in plants

619

freezing at 80
or infiltration with a special RNA preservative such as RNAlater (Ambion). In either case, it is critical to quickly collect the material and complete the selected treatment. Some tissues may be more sensitive to RNA degradation than others, but no more than 15 min at ambient temperature or 1 h on ice is ideal. For best results, tissue should be treated immediately. In cases in which dissection is required, material can be kept on ice during the dissection and the separated tissues treated as soon as the desired amount is obtained (but observing the time frame mentioned earlier). For freezing, it is best to collect tissue in plastic screw-top or snap-top tubes. Plastic or paper bags perform poorly during long-term storage at 80
. Make sure to mark all samples with ink or labels that are compatible with cold storage. Once material is frozen, it is critical that it be maintained at temperatures less than 50
; RNA will not tolerate freeze–thaw cycles of more than 30
. Material can be stored at 80
for very long periods, however, as long as the temperature is controlled. As a general rule, tissue that has been fixed in formaldehyde or similar chemical fixatives is not suitable for RNA extraction. Likewise, silicapreserved tissues will not yield usable RNA. These issues must be taken into consideration when planning any experiment that will use RNA and, unfortunately, often mean that remote field collections are not compatible with RNA-based techniques. Methods of RNA Extraction Many expression analysis techniques, such as Northern blot hybridization and reverse-transcriptase polymerase chain reaction (RT-PCR), start with the extraction of total RNA. Although numerous plants are perfectly amenable to this process, it is not uncommon for the presence of polyphenols, polysaccharides, mucilage, and other compounds to inhibit RNA extraction from plant tissues. A number of specialized protocols have been developed to deal with particularly recalcitrant species or tissue types (Chang et al., 1993; Suzuki et al., 2003; Wang et al., 2000). As a general rule of thumb, younger tissues, such as flower buds or immature leaves, are easier to work with. Many kits and products for RNA extraction are available (see http://www.ambion.com, http://www. qiagen.com, and http://www.invitrogen.com, among others). We prefer to use Concert Plant RNA Reagent by Invitrogen (Carlsbad, CA) because of the simple scalability of the protocol for various amounts of tissue (100 mg to 5 g). This protocol is not reproduced here because we follow the manufacturer’s instructions quite closely, but a number of general considerations are noteworthy.

620

functional analyses

[32]

. It is unnecessary to treat items such as mortars and pestles or spatulas to make them RNase-free. Although they should be clean, there is much more RNase in the tissue than there is on these tools. . The best way to prevent degradation of RNA during the grinding process is to keep all utensils very cold. This will also aid in grinding. Prechill mortars and pestles with liquid nitrogen for 5 min before grinding the tissues in liquid nitrogen. Grind small amounts of tissue in microtubes using disposable micropestles (e.g., VWR KT49521-1590) prechilled in liquid nitrogen. . Best results will be obtained by grinding the frozen tissue to a dustlike consistency. This should be rapidly transferred to a large tube containing the appropriate amount of extraction buffer using a prechilled spatula. For small amounts of tissue, further maceration of material in the extraction buffer can greatly increase yield. . Once an RNA pellet has been obtained, rinse it carefully in 70% ethanol made with diethylpyrocarbonate (DEPC) water. The dry pellet should become translucent and glassy. . Resuspension of the RNA is aided by the use of prewarmed (50–60
) RNase-free water or 2 mM of ethylenediaminetetraacetic acid (EDTA) pH 8.0. Resuspend the pellet in as small a volume as possible to facilitate downstream protocols. Inability to resuspend the pellet may result from overdrying (never dry RNA in a SpeedVac) but is also often indicative of the presence of secondary compounds or starch. This may indicate that alternative preparation protocols will be necessary. . RNA should be stored at 80
. Northern Blot Hybridization Northern blots can be useful to examine the expression of genes in various tissue types or over a range of developmental stages. They can be prepared using either total RNA or poly(A) RNA. Several standard protocols for Northern blot preparation are available, including that of Sambrook and Russel (2001a). General tips for performing Northern blot hybridization are as follows: . Prepare the electrophoresis apparatus, gel casting tray, and gel comb by treatment with 0.1 M NaOH for 12 h, followed by thorough rinsing in water. . The RNA must be suspended at a reasonably high concentration to allow loading of a large amount of RNA (5–10
g) in 1–5
l of sample. Care should be taken to equalize the loading amount of each sample so direct comparisons can be made across different lanes.

[32]

gene expression techniques in plants

621

. Once the RNA transfer to the membrane is complete, the transfer efficiency can be checked by illuminating the blot with a handheld ultraviolet (UV) light source. The nucleic acids must be fixed to the membrane surface, which can be done using commercially available UV cross-linkers or by baking in a vacuum oven for at least 1 hr at 80
. Using both methods increases the stability of the blot and facilitates repeated probing and stripping. . I recommend a hybridization solution for Northerns composed of 50% formamide, 3 SSC (3.0 M NaCl, 0.3 M sodium acetate), 0.5% sodium dodecyl sulfate (SDS), 0.1 mg/ml of herring sperm DNA, 5X Denhardt’s solution (1% Ficoll, 1% polyvinylpyrrolidone, 1% bovine serum albumin [BSA], filter sterilized), and 25 mM of ethylenediaminetetraacetic acid (EDTA) pH 8. This increases the stringency of the hybridization. Blots prepared in this manner can be stripped and hybridized with different probes up to six times. Relative lane loading can also be assessed by hybridizing the blot with a control probe, such as ubiquitin or actin, in the final hybridization. When combined with phospho imaging, this can allow more accurate quantification of relative expression levels. PCR-Based Methods To an increasing degree, PCR-based approaches are replacing the use of Northern blots to assess gene expression. These techniques include ‘‘semiquantitative’’ RT-PCR and quantitative or real-time RT-PCR (RT-qPCR). Before PCR can be conducted, first-strand cDNA must be synthesized from the RNA. Although total RNA can be used as the template for this reaction, I have found that using poly(A) RNA yields the best results. A caveat to this statement, however, is that poly(A) RNA constitutes only about 10% of a total RNA extraction, and typical yields are in the range of 1–5%. This means that extraction of poly(A) from total RNA is worth pursuing only when starting with a fairly large amount of total RNA (at least 50–100
g). Many poly(A) extraction kits are available, including some that allow direct preparation of poly(A) from tissue (see Ambion, Qiagen, Invitrogen, Clontech). I prefer to use magnet-based methods (Dynal, Ambion, Novagen), which allow the elution of poly(A) RNA in very small volumes. Before cDNA synthesis, RNA should be treated with RNase-free DNAse, particularly when working with total RNA. The stop solutions typically used with DNAse, however, contain concentrations of EDTA that may inhibit further enzymatic reactions. This means that treated RNA must be column cleaned using a product such as RNAqueous (Ambion, Austin, TX) or RNeasy (Qiagen, Valencia, CA) and eluted in DEPC water.

622

functional analyses

[32]

RNA can be easily quantified with UV spectrophotometry; high-quality RNA should give an A260/A230 reading of more than 1.8. Electrophoresis in a typical TBE or TAE gel is also useful to assess RNA quality. Keeping a small electrophoresis rig, gel casting tray, and gel comb RNase-free is useful if this is going to be performed frequently. Good quality total RNA will appear as a long, bright smear punctuated by two to four distinct bands of moderate molecular weight, which represent the rRNAs and sometimes rbcL transcripts if the tissue is photosynthetic. Poly(A) RNA does not look like much on a gel, just a bright smear in the range of 400–2000 bp. A faint smear with a bright spot at low molecular weight (<100 bp) indicates that the RNA is degraded. Preparation of first-strand cDNA can be done in a separate reaction or at the same time as the PCR (one-step RT-PCR). Again, many manufacturers supply products for both types of procedure (Ambion, Invitrogen, Stratagene, Clontech, Promega, etc.). Because my laboratory is typically interested in the expression of multiple genes, my colleagues and I prefer to prepare cDNA in a separate reaction using an anchored poly-T primer [see Kramer et al. (1998)]. This type of cDNA is stable at 20
for several months and can be used with any combination of degenerate or specific PCR primers. Following cDNA synthesis, the reaction can be treated with RNase to remove all remaining RNA, and the cDNA can be column purified using any available PCR cleanup kit. The best way to assess the success of the cDNA reaction is to use it as a template in a PCR with primers for control loci such as ubiquitin or actin. Semiquantitative RT-PCR simply involves performing PCR on cDNA using primers that are designed to be specific to the gene of interest. It is advisable to design the primers so that they span at least one intron. This allows the amplification of contaminating genomic DNA to be simply detected because the product derived from genomic DNA will be larger than that derived from cDNA. In addition to the typical concerns for designing PCR primers (annealing temperature, GC content, etc.), primer specificity should be carefully considered so that only the gene of interest is amplified, particularly when working in large gene families. Every effort should be made to use equivalent amounts of cDNA in each reaction, and it is important to run positive control reactions using primers for a gene that is expected to be expressed fairly uniformly, such as actin. Several caveats are important to consider when performing semiquantitative RT-PCR and interpreting the results. The highly sensitive nature of PCR means that even samples that have vanishingly small amounts of target can yield bands if subjected to enough amplification cycles. For this reason, it is advisable to use cycle numbers in the 20–25 range to increase

[32]

gene expression techniques in plants

623

confidence that the detection of cDNA is significant. Of course, some loci are expressed at such low levels that additional cycles are necessary. In this case, it is preferable to show the results for several experiments using both low and high cycle numbers. Real-time or quantitative RT-PCR is rapidly becoming a standard method for more accurate analysis of relative gene expression levels. These methods require specialized equipment, including light-detecting PCR machines. The technical considerations and data analyses involved in this procedure are quite complex and beyond the scope of this chapter. Several excellent web sites on this topic exist, including http:// www.wzw.tum.de/gene-quantification/. For both semiquantitative and RTqPCR, however, several limitations remain. Early stages of development are often difficult to analyze, because of the inability to dissect separate organs for RNA extraction. In addition, important aspects of gene expression, such as spatial patterns of RNA distribution within an organ, are not observable. For these reasons, the best tool for assessing gene expression patterns remains in situ hybridization. In Situ Hybridization

In situ hybridization is, as already mentioned, the best available method for obtaining information on the specific spatial and temporal patterns of gene expression in developing tissues and organs. At the same time, the method is complex and requires considerable preparation and expertise. Commonly used protocols involve hybridization of radioactive or nonradioactive RNA probes to sectioned tissue. Radioactive probes are thought to be more sensitive but have many drawbacks, including poor localization of the signal, instability of the probe, and the requirement to process slides in complete darkness. Several protocols for labeling and hybridization of radioactive probes are available, including that of Weigel and Glazebrook (2002). Nonradioactive in situ hybridization is generally preferable and is described here. Many variations on the protocol exist, but all are derived from Jackson (1991). The following protocol has been successfully used with magnoliid dicot, monocot, and eudicot floral tissue. The procedure is described in six parts: fixation and embedding of tissue, preparation of RNA probes, preparing for in situ hybridization, sectioning and prehybridization, posthybridization, and imaging. Fixation and Embedding of Tissue As with tissue collected for RNA extraction, tissue intended for ultimate use in in situ hybridization must be carefully processed.

624

functional analyses

[32]

1. Dissect or collect tissue and immediately submerge in ice cold, freshly prepared FAA (50% ethanol, 10% formalin, 5% acetic acid). It is critical to use freshly prepared FAA and preferable to use formalin that is not more than a year old. 2. Place the tube containing the FAA and tissue into a beaker filled with ice, and place this into a desiccator. Vacuum infiltrate the tissue for 1.0–1.5 h. Hold the vacuum for 15 min and then release slowly. Repeat. Tissue may sink, and many protocols suggest that this is a necessity. However, it is common for plant materials that are very pubescent (hairy) to remain buoyant in FAA. It is more important not to overfix your tissue than to wait for it to sink. 3. Following vacuum infiltration, refill the tube with fresh FAA as necessary and place it on an oscillating shaker at 4
. Protocols vary as to how long tissue should be incubated in fixative after vacuum infiltration. Very small herbaceous tissue can be immediately removed from the fixative following vacuum infiltration and dehydrated. Larger, dense tissue typically requires 4–12 h of incubation in FAA for proper fixation. Never fix tissue for more than 16 h total. Note: Many people are not aware that tissue can be overfixed. This is a serious problem that must be avoided for optimal in situ hybridization results. Tissue that has been in fix for more than 24 h is unlikely to be useful for in situ hybridization. 4. Dehydrate the samples through the following ethanol series at 4
with agitation for 30-90 min each: 50% ethanol, 70% ethanol, 85% ethanol, 90% ethanol, 100% ethanol. The duration of the incubation depends on the nature of the tissue. Similar to the case with fixation, small herbaceous tissue requires shorter periods than denser tissues. Tissue can be stored overnight or for longer periods at 4
during the 70% ethanol stage. After the 100% ethanol step, exchange the solution for fresh 100% ethanol and leave overnight at 4
. 5. Exchange the solution for fresh 100% ethanol and incubate at room temperature (RT) for 1 h. Follow this by subsequent RT incubations in 50% ethanol/50% Citrisolv (Fisher Scientific 22-143-975) for 2 h, then 100% Citrisolv for 2 h. 6. Transfer tissue to a small glass beaker or scintillation vial and add just enough fresh Citrisolv to cover the tissue. Fill the beaker with Paraplast Plus chips and incubate overnight at 55–60
. For the next 2 days, exchange the Paraplast with fresh molten Paraplast two to three times each day. Material must be maintained at 55–60
. With dense tissue or organs that tend to retain air bubbles (such as spurs), vacuum infiltration of the molten Paraplast may be necessary. This can be accomplished using a vacuum oven set to 60
, applying moderate vacuum for 30–60 min.

[32]

gene expression techniques in plants

625

7. Embed the tissue by filling molds (such as Peel-A-Way or Tissue-Tek) with molten Paraplast and placing individual samples in each. This can be done on a hot plate to allow time to properly orient the tissue. Allow the molds to cool in a RT waterbath. 8. Embedded tissue can be stored for long periods at 4
. Preparation of RNA Probes For nonradioactive probes, antisense and sense RNA is typically labeled using digoxigenin (DIG), which is then detected using an anti-DIG antibody conjugated to alkaline phosphatase. All DIG-labeling supplies mentioned below are available from Roche Applied Science (http://www. roche-applied-science.com). All solutions are prepared with DEPC water and RNase-free chemicals or treated with DEPC and autoclaved after preparation. 1. Chose a region of your gene of interest to serve as the probe template. Generally, a 200-500 bp fragment is optimal. Although in situ hybridization is typically quite stringent in its specificity, it is preferable to use regions that do not show high conservation across members of a gene family, such as DNA-binding domains. However, using solely 30 UTR sequence as template is not advisable because these regions often form secondary structures that can interfere with probe hybridization. 2. Generate a linearized version of the template either by restriction digestion or by PCR. If using PCR, the RNA polymerase binding site can be incorporated into the fragment in one of the oligonucleotide primers. PCR products should be cleaned and concentrated using a spin column. In the case of plasmid linearization, it is important to use a restriction enzyme that does not create a 30 overhang, which can result in nonspecific polymerase initiation. The chosen enzyme should cut at the end of the template insert opposite from the RNA polymerase binding site. Confirm complete digestion by running linearized plasmid on an agarose gel. 3. Extract the linearized DNA with an equal volume of phenol/chloroform, then chloroform. Use RNase-free tubes and tips following the first extraction. Precipitate the DNA by adding 2 volumes 100% ethanol and 0.1 volume of 3 M NaAC, incubating at 20
for 2 h, and centrifuging at high speed for 10 min. Wash the pellet with ice cold 70% ethanol, dry, and resuspend in DEPC water at a concentration of about 1
g/
l. 4. Set up a runoff transcription reaction as follows: 1
l DNA at 1
g/
l 2
l NTP/DIG–UTP mix 2
l 10X transcription buffer

626

functional analyses

[32]

2
l T7, T3, or SP6 RNA polymerase 1
l RNase inhibitor 12
l DEPC water Mix by gently pipetting and incubate at 37
for 2 h. 5. Set aside a 1-
l sample for later use (see step 7). 6. Add 2
l of RNase-free DNAse. Mix and incubate at 37
for 15 min. 7. Take another 1-
l sample. Run this sample side by side with the pre-DNAse sample (step 5) on a small agarose gel (use RNase-free loading buffer to run the samples). Bright RNA bands should be visible in both lanes. RNA probe often runs as multiple bands. DNA template should be visible in the pre-DNAse sample but not in the post-DNAse. If the DNA is not eliminated, add 1
l of DNAse and incubate another 30 min at 37
. Repeat analysis as needed. 8. Once DNAse completion has been confirmed, stop the reaction by adding 4
l 200 mM EDTA pH 8.0. Precipitate the RNA by adding 5
l 4 M LiCl and 150
l ethanol. Incubate at 20
for 2 h before precipitation by centrifugation for 10 min at maximum speed. 9. Wash pellet with ice cold 70% ethanol and allow to air dry. 10. Hydrolize RNA probe to desired length by resuspending pellet in 50
l of 0.1 M NaHCO3 pH 10.2. Incubate at 60
for an amount of time determined by the formula t ¼ (Li  Lf)/K(Li)(Lf), where t is the time in min, K is 0.11 breaks/min, Li is the initial length of the probe in kilobases, and Lf is the final desired length in kilobases. For instance, to hydrolyze an initial probe of 0.5 kb to 0.15 kb (the typically recommended length), t ¼ (0.5–0.15)/(0.11  0.5  0.15) ¼ 42 min. See Trouble-shooting in situ hybridization section for further discussion of probe hydrolysis. 11. Stop the hydrolysis reaction by adding 5
l 5% acetic acid, 5
l 3 M sodium acetate, and 125
l ethanol. Incubate at 20
for two h. Precipitate by centrifugation for 10 min at maximum speed, wash pellet in ice cold 70% ethanol, dry, and resuspend in 20
l deionized formamide. 12. Quantify probe concentration using the Roche protocol, available at http://www.roche-applied-science.com under the title ‘‘Estimating the Yield of DIG-labeled Nucleic Acids.’’ Probe can be stored at 20
for several months. Obtaining a good-quality probe is very important for the success of the hybridization. Reactions that yield poor concentrations are unlikely to perform well. In my experience, T7 or T3 RNA polymerase gives better results than SP6. For probe templates that are particularly GC rich, it may be necessary to prepare a special NTP/DIG–UTP mix. The DIG–RNA labeling mix supplied by Roche has 3.5 mM DIG-11–UTP/6.5 mM dTTP and 10 mM of each remaining NTP. DIG-11–UTP and NTP solutions can

[32]

gene expression techniques in plants

627

be purchased separately in order to prepare alternate concentrations, such as 6.5 mM DIG-11–UTP/3.5 mM TTP. Higher DIG-11–UTP concentrations will result in more DIG incorporation, but it also will reduce the efficiency of the transcription reaction. It is recommended to prepare both sense and anti-sense probes. Anti-sense probe will be produced by transcription from an RNA polymerase binding site at the 30 end of the template fragment, whereas the sense probe is made by transcription from the 50 end. Preparing for In Situ Hybridization Preparing to perform in situ hybridization takes 3–5 days, depending on how many solutions need to be made and your experience with the process. The following basic RNase-free stock solutions are required [refer to Sambrook and Russel (2001b) for details]: 10 PBS 5 NTE 20 SSC 10 PBS with 20 mg/ml glycine (store at 4
) 1 M Tris pH 9.5 1 M pH Tris 8.0 1 M Tris 7.5 0.5 M EDTA pH 8.0 5 M NaCl 1 M MgCl2 Additional required RNase-free solutions include 10 pronase buffer: .5 M Tris 7.5, 50 mM EDTA (see the discussion on pronase, later in this chapter) 10 in situ salts: 3 M NaCl, 100 mM Tris pH 8, 100 mM NaH2PO4 pH 6.8, 50 mM EDTA Hybridization solution (800
l): 100
l 10 in situ salts, 400
l deionized formamide, 200
l 50% dextran sulfate (this will require heating to dissolve), 20
l 50 Denhardt’s solution, 10
l tRNA (100 mg/ml in DEPC water), and 70
l DEPC water Note on the hybridization solution: It is best to prepare a relatively large volume (10–15 ml) and aliquot this into working amounts (1 ml) in RNase-free microtubes. It is absolutely critical that this solution be RNase free, so prepare it with care. In addition to preparing these solutions, all glassware (additional bottles, graduated cylinders, Coplin jars), slide racks, and stir bars must be baked. Other plasticware, such as bottle caps, can be treated with 0.1 M

628

functional analyses

[32]

NaOH as described earlier. I recommend using square polyethylene (VWR 36318-045) and flat Nalgene (VWR 36212-204) boxes for incubation and washing steps. These can be thoroughly washed twice in an automated dishwasher, once with detergent and once without. Dry upside down on paper towels. Glass staining dishes can be similarly treated. Prepare at least 8 L of ddH2O by autoclaving for 30 min in baked bottles. This ‘‘clean’’ water will be used for preparation of the hydration series and all diluted working solutions. Sectioning and Prehybridization Place five to six ProbeOn Plus slides (Fisher 15-188-52) onto a slide warmer set at about 40
and cover the unfrosted surface of each slide with a pool of DEPC water. It is important to use ProbeOn Plus slides to promote proper adhesion of the sections and allow sandwiching of slides (see below). Paraplast Plus embedded tissue should be sectioned at 8-
m thickness on a rotary microtome. Separate the ribbon of sections into strips of two to three sections using a clean razor blade and place sequentially onto the slides. Small paint brushes or wooden applicator sticks can be used to manipulate the sections. The sections must be allowed to flatten on the surface of the DEPC water. After about 15 min, carefully remove excess water using a Pasteur pipette around the edges of each slide. After another period of about 15–30 min, remaining water can be removed by gently tipping the edge of slide against a stack of Kim wipes or paper towels. Repeat this process as needed in order to obtain 20–25 slides. Sectioning can be done the day before the prehybridization, but sectioning on the same day generally yields better results. However, for the sections to adhere properly, it requires at least 4 h of incubation from the time the slides are completely drained. During this period, slides should remain on the slide warmer with the lid closed. If you choose to section the day before, do so late in the afternoon and turn the temperature of the slide warmer down below 30
before leaving the slides overnight. Before starting the prehybridization, several solutions need to be prepared. First, the hydration series should be arranged using square polyethylene boxes, which typically hold 300 ml each. Second, prepare all the diluted solutions, including 1.2 L of 1 PBS, 300 ml of 1 pronase buffer, and 300 ml of 1 PBS with 0.2% glycine. These can be made by dilution of the stocks (see earlier discussion) with the prepared ‘‘clean’’ water. It is also necessary to make a fresh solution of 4% paraformaldehyde in 1 PBS. Several methods are available for preparing this solution, but I prefer the following: Put 300 ml of 1 PBS in an RNase-free flask with a treated stir bar and bring to an active boil in the microwave. Immediately place the flask on a magnetic stir plate under a fume hood

[32]

gene expression techniques in plants

629

and add 12 g of paraformaldehyde to the hot solution. Stir. The powder should go into solution quickly. Allow the flask to cool and cover the mouth with parafilm or foil. Place the flask on ice under the hood to cool completely. This must be done before starting the prehybridization to allow the solution to cool but should not be performed a day in advance. The following is the prehybridization protocol. Unless otherwise noted, the steps are performed in square polyethylene boxes in volumes of 300 ml. 1. Place slides into a metal slide rack, leaving every other slot empty to allow circulation of the solutions. 2. Incubate 10 min in about 300 ml of Citrisolv in a glass staining box. Repeat with fresh Citrisolv. 3. Rehydrate the slides by processing through an ethanol series as follows: 2 min in 100% ethanol, 1 min in 100%, and 1 min each in 95%, 85%, 70%, 50%, and 30%. Follow by 2 min in 150 mM NaCl, and then 2 min in 1 PBS. 4. Incubate the slides for 20 min at 37
in 1 pronase buffer (50 mM Tris pH 7.5, 5 mM EDTA) with 10
g/ml pronase. Note on pronase treatment: Either pronase or proteinase K can be used for digestion, but they use different buffers, so care must be taken to ensure use of the correct buffer. This step requires considerable optimization. First, the activities of both enzymes vary between preparations. When preparing the enzyme stock, follow the manufacturer’s instructions to make a large volume (10–20 ml) at 10 mg/ml in DEPC water. This can be aliquoted into small working volumes in RNase-free microtubes, which can be stored long term at 20
. This will allow the use of the same stock (and hence, the same activity level) throughout repeated experiments. Second, once the stock is prepared, the specific activity must be assessed for the particular tissue being analyzed. The best way to do this is to run test digestions on sectioned tissue using a range of concentrations (5–20
g/ml) or incubation lengths (20–30 min). For simplicity, vary only one component (concentration, time, or temperature) in the experiment. Ideally, the tissue should be digested as much as possible without affecting tissue/cell integrity. It is important to ensure that the cell contents are not affected by the digestion (the cell walls may look perfect even when the contents are gone). The use of stains can aid in the evaluation of the optimal enzyme treatment. Different tissue types or stages may require different treatments. The findings of this process will determine the enzyme concentration that will be used at this step in the actual prehybridization. 5. Following digestion, incubate in 1 PBS with 0.2% glycine for 2 min.

630

functional analyses

[32]

6. Rinse in 1 PBS for 2 min. 7. Incubate in 4% formaldehyde for 20 min under a fume hood. 8. Rinse in 1 PBS for 2 min (dispose of the PBS from this step in the formaldehyde waste). 9. Acetylation: Incubate slides for 10 min in 0.1 M triethanolamine pH 8.0 with 0.5% acetic anhydride. Note: This solution must be prepared fresh but takes a while to complete because the triethanolamine is a thick liquid and is difficult to pipette. It is best to start preparing the 0.1 M of triethanolamine during the incubation period of step 7. Under a fume hood, place a glass staining dish on a magnetic stir plate. Place a stir bar and two pieces of plastic sterile 10-ml pipettes in the bottom. The pipettes should be broken so that they fit in the bottom of the container with room for the stir bar to spin. Fill the dish with 589.8 ml of ‘‘clean’’ water and add 7.8 ml of triethanolamine while stirring. Bring the pH to 8.0 with 2.4 ml of concentrated HCl. Check the pH using a pH strip. Carefully add 3.6 ml of acetic anhydride. Momentarily stop the stir bar and allow the pipette pieces to settle on either side of the stir bar. Place the slide rack into the glass box so that the rack is supported above the stir bar with sufficient room for the bar to move. Restart the stirring and incubate for 10 min. 10. Rinse in 1 PBS for 2 min (do this step in the fume hood and dispose of the PBS in the acetic anhydride waste). 11. Rinse in 150 mM NaCl for 2 min. Follow by processing through the ethanol dehydration series with 1 min of incubation per step: 30%, 50%, 70%, 85%, 95%, and 100%. It is not necessary to make a fresh dehydration series; in fact, the same solutions can be used for several experiments. 12. Incubate in fresh 100% ethanol for 2 min. Carefully remove slides from rack and place them section-side up onto fresh paper towel. Allow to dry completely. 13. Prewarm the necessary number of tubes of hybridization solution in a heat block at 80
(this should correspond to the number of slide pairs, see step 14). Set the hybridization oven to the appropriate temperature (see below). 14. Carefully examine the dry slides and choose those that have the best quality of tissue at the desired stage. Arrange the chosen slides into pairs. Mark the pairs as to which probe will be used on them (each pair will have the same probe). At least one pair should be hybridized with sense probe as a negative control. 15. Probe should be used at 0.5 ng/
l/kb. The amount of probe per slide pair is determined by Lp  100
l  (0.5 ng/
l/kb), where Lp is the length of the probe in kilobases. For instance, a probe hydrolyzed to

[32]

gene expression techniques in plants

631

0.15 kb will require 0.15  100  0.5 ¼ 7.5 ng of probe. The concentration of the probe is determined in step 12 in the section ‘‘Preparation of RNA Probes.’’ The first time that a given probe is hybridized, it should be tried at both the concentration calculated above (this is the 1 concentration) and at 5. 16. For each slide pair, the desired amount of probe should be added to 50% deionized formamide to make the volume up to 40
l. Heat the probe to 80
. 17. Add 200
l of preheated hybridization solution to each tube of heated probe and mix by pipetting. The hybridization solution is very viscous and is difficult to pipette accurately. 18. Apply the probe to each slide pair. The technique of probe application is a matter of taste and requires some practice to determine what works best for you. This is the method I use: Take the first slide of the pair and pipette 100
l of probe mix onto the slide. Carefully use the side of an RNase-free pipette tip to spread the hybridization solution across the entire surface of the slide. Pipette the remaining probe solution onto the second slide along the narrow edge of the slide, opposite the frosted area. Take the spread slide and place its corresponding edge on the narrow edge of the second slide. The wet side of the first slide should be facing with the sections downward toward the second slide. Slowly, lower the spread slide onto the second slide, allowing the adhesion of the probe solution to pull the liquid across the entire surface of the slide. The final product will be a slide ‘‘sandwich’’ with the probe solution in between the surfaces of the two slides. 19. Elevate the slide sandwiches above wet paper towels in a flat plastic box. This can be achieved by breaking plastic sterile 1-ml pipettes into halves and placing them across the bottom of the box, which is lined with wet paper towels. Line the lid with Saran wrap and close tightly. Place the box into the preheated hybridization oven for 14–16 h. Hybridization temperatures generally range from 38 to 45
. Initial hybridizations should be performed at a lower temperature, such as 40
. Make careful notes concerning the probes used and their concentration, the hybridization temperature, and any other variables such as the pronase concentration. Posthybridization At the end of the prehybridization protocol, several solutions for the posthybridization should be prepared 1 day in advance. Prepare 1.5 L of 0.2 SSC and place at 48–55
. The temperature depends on the desired stringency of your wash. A good rule of thumb is to wash at a temperature 10


632

functional analyses

[32]

above the hybridization temperature. Prepare 1.5 L of 1 NTE solution and place at 37
. In addition, you can prepare 300 ml of 1 PBS; 650 ml of 100 mM Tris 7.5, 150 mM NaCl; and 250 ml 100 mM Tris pH 9.5, 100 mM NaCl, 50 mM MgCl2. The morning of the posthybridization, several additional solutions should be prepared using the solutions above: 120 ml 1.0% Roche Blocking Reagent (Roche 1 096 176) in 100 mM Tris 7.5, 150 mM NaCl (this requires warming for complete incorporation, so allow time for cooling before use), and 520 ml 1.0% BSA in 100 mM Tris 7.5, 150 mM NaCl, and 0.3% Triton. The following is the posthybridization protocol. Unless otherwise noted, the incubations are performed in square polyethylene boxes in volumes of 300 ml. 1. Fill a square plastic box with about 300 ml of preheated 0.2 SSC. Immerse each slide sandwich and gently separate by opening, not sliding. Place the slides in a slide rack, leaving an empty slot between each one for circulation. 2. Wash the slides for 1 h in fresh preheated 0.2 SSC with gentle agitation in a hybridization oven set to the chosen wash temperature. 3. Repeat 60-min wash with fresh preheated 0.2 SSC. 4. Incubate the slides in preheated 1 NTE for 5 min at 37
with gentle agitation. 5. Repeat 5-min 1 NTE wash. 6. Treat with 20
g/ml RNase A in preheated 1 NTE for 20 min at 37
. 7. Incubate the slides in preheated 1 NTE for 5 min at 37
with gentle agitation. 8. Repeat 5-min 1 NTE wash. 9. Wash for 60 min in preheated 0.2 SSC at wash temperature with gentle agitation. 10. Incubate 5 min in 1 PBS at RT. 11. Place each slide on the bottom of a flat plastic box, section-side up with no slides overlapping each other. The flat boxes cited above will hold 10 slides. Fill each box with 100 ml of 1.0% Roche Blocking Reagent in 100 mM Tris 7.5, 150 mM NaCl. This should be just enough to cover the slides. Incubate for 45 min on a rocking platform at RT. 12. Replace block solution with 100 ml 1.0% BSA in 100 mM Tris 7.5, 150 mM NaCl, 0.3% Triton. Incubate for 45 min on a rocking platform at RT. 13. Dilute 8
l of alkaline phosphatase-conjugated anti-DIG antibody (Roche 1 093 274) into 10 ml of 1.0% BSA in 100 mM Tris 7.5, 150 mM NaCl, 0.3% Triton. Make a puddle of the solution in a large plastic weigh dish. Place each slide pair in the puddle with their long edges in the

[32]

gene expression techniques in plants

633

solution. Sandwich the slides together, with the sections on the inside, drawing up the antibody solution between them. Drain the slide pair on a stack of Kim wipes and place the edge back into the puddle. Solution will be drawn up by capillary action. Repeat once, avoiding air bubbles. 14. Elevate the slide sandwiches with antibody solution above wet paper towels as described above. Allow to sit at RT for 2 h. 15. Drain slides on Kim wipes and separate carefully. Place on the bottom of a flat plastic box as in step 11. Wash with 100 ml of 1.0% BSA in 100 mM Tris 7.5, 150 mM NaCl, 0.3% Triton for 15 min at RT on a rocking platform. Repeat three times for a total of 4 washes. 16. Replace BSA/Triton solution with 100 ml 100 mM Tris pH 9.5, 100 mM NaCl, 50 mM MgCl2. Wash for 10 min with rocking. 17. Fill a Coplin jar with 100 mM Tris pH 9.5, 100 mM NaCl, and 50 mM MgCl2 and dip each slide into the solution to ensure that all of the detergent is removed. 18. Prepare substrate solution for color detection by adding 22
l NBT (Roche 1 383 213) and 16
l BCIP (Roche 1 383 221) to 10 ml of 100 mM Tris pH 9.5, 100 mM NaCl, 50 mM MgCl2. 19. Apply solution to slide sandwiches as described in step 13. 20. Elevate slide sandwiches above wet paper towels and seal box tightly. Carefully wrap box in aluminum foil and place in a dark drawer to prevent light contamination. Again, make notes on all the variables, such as the wash temperature. Imaging Color development generally takes 12–48 h. Good signal is usually visible to the naked eye, although this depends on the size of the expression domain. Carefully examine the sandwiches under a microscope to assess staining. When adequate signal is detectable, immerse the sandwiches in TE to separate and stop the staining reaction. Place slides into a slide rack and incubate in 1 PBS for 5 min at RT. Counterstaining with calcofluor (also known as Fluorescent Brightener, Sigma F-3397) can significantly improve contrast and visualization. Incubate slides in 0.002% solution of calcofluor in 1 PBS. Follow by rinsing for 5 min in 1 PBS. Dehydration will significantly reduce staining intensity, so mount slides with water, glycerol, or other aqueous medium (I prefer water). Do not let the slides dry out at any step. The slides can now be photographed. To take advantage of the calcofluor counterstaining, sections should be illuminated simultaneously with fluorescent and white light. Dial down the intensity of the white light until the fluorescence of the tissue is just visible. This will appear as a blue-white

634

functional analyses

[32]

glow in the cell walls. Certain tissue absorbs calcofluor better than others and this may be inconsistent between sections. ‘‘Real’’ signal should be light brown or lilac to dark brown. Dark blue staining is generally not real signal (see below). Photodocument every section with informative staining because water-mounted slides will not be stable for more than a week or two (at 4
elevated above wet paper towels). Troubleshooting In Situ Hybridization Results As mentioned earlier, performing in situ hybridization can be a challenging process and the protocol requires optimization of many steps. The following are suggestions for troubleshooting: No or low signal: This may be due to a number of factors including low probe concentration, hybridization, or wash temperatures that are too high, RNase contamination of reagents (particularly hybridization solution), overfixation of tissue, insufficient tissue digestion, or improper probe hydrolization (see below). High background: This typically results from hybridization or wash temperatures that are too low. It is important to note that high background can be general, affecting all tissues, or disconcertingly specific, only affecting certain tissues. Distinguishing nonspecific staining from ‘‘real’’ staining is sometimes difficult. Tissues that are prone to nonspecific staining include stamens and pollen, vascular tissues, and very small dense meristematic cells. Nonspecific signal often takes the form of very dark blue staining that is associated with the cell walls. It is important to remember that in situ hybridization targets mRNA, which should be in the cytoplasm, not the nucleus or cell walls. The caveat to this statement is that in mature plant cells that are highly vacuolated, the cytoplasm may be closely pressed to the cell wall. The most commonly analyzed genes for comparative gene expression, such as MADS-box containing genes or CYCLOIDIA homologues, typically have high levels of expression that should be clearly discernible. Comparisons with sense control slides can sometimes help to distinguish the real signal from the nonspecific staining. Tissue appears to be degraded or cell contents are gone: Tissue has been overdigested. Lower the enzyme concentration. Note on probe hydrolysis: Protocols commonly call for probe to be hydrolyzed to 70–150 bp (Jackson, 1991; Weigel and Glazebrook, 2002). Although this works well for many probes, others perform better at different lengths. Optimization of probe length is a trial-byerror process; there are no good rules of thumb. Although 200 bp

[32]

gene expression techniques in plants

635

may work well for a gene in one species, the ortholog in another taxon may work best at 300 bp. If little signal is observed even at low hybridization temperatures and high probe concentrations, using a shorter probe is often successful. Alternatively, high background (either general or specific) can be eliminated by using longer probe. Several alternatives to the type of in situ hybridization described earlier have been developed. For organs with complex three-dimensional structure, whole-mount in situ hybridization (Zachgo et al., 2000) can yield more informative results. Another alternative that has yet to be broadly used is in situ RT-PCR. One drawback to the DIG-labeled form of this protocol (Johansen, 1997) is that it sometimes results in nonspecific labeling of all nuclei, which can make interpretation of the data difficult. A new method using Oregon green–labeled UTP (Ruiz-Medrano et al., 1999) has yielded beautiful results, however (Kim et al., 2003a,b). Concluding Remarks

The generation of gene expression data is only the first step in the difficult process of assessing gene function. It has been demonstrated for many genes that expression is not necessarily a simple proxy for the spatial extent of gene function, both because of non–cell autonomous effects and posttranscriptional or posttranslational regulation. Therefore, any data resulting from the types of analyses described earlier should be interpreted with care. However, given that direct analyses of gene function using genetic tools are commonly unavailable in nonmodel species, gene expression is often our only comparative tool. Therefore, it is preferable to use a combination of all of the aforementioned techniques to obtain a detailed and clear picture of gene expression patterns. References Chang, S., Puryear, J., and Cairney, J. (1993). A simple and efficient method for isolating RNA from pine trees. Plant Mol. Biol. Rep. 11, 113–116. Jackson, D. (1991). In situ hybridisation in plants. In ‘‘Molecular Plant Pathology: A Practical Approach’’ (D. J. Bowles, S. J. Gurr, and P. McPherson, eds.), pp. 163–174. Oxford University Press, Oxford. Johansen, B. (1997). In situ PCR on plant material with sub-cellular localization. Ann. Bot. 80, 697–700. Kim, M., McCormick, S., Timmermans, M., and Sinha, N. (2003a). The expression domain of PHANTASTICA determines leaflet placement in compound leaves. Nature 424, 438–443. Kim, M., Pham, T., Hamidi, A., McCormick, S., Kuzoff, R. K., and Sinha, N. (2003b). Reduced leaf complexity in tomato wiry mutants suggests a role for PHAN and KNOX genes in generating compound leaves. Development 130, 4405–4415.

636

[33]

functional analyses

Kramer, E. M., Dorit, R. L., and Irish, V. F. (1998). Molecular evolution of genes controlling petal and stamen development: Duplication and divergence within the APETALA3 and PISTILLATA MADS-box gene lineages. Genetics 149, 765–783. Ruiz-Medrano, R., Xoconostle-Cazares, B., and Lucsa, W. (1999). Phloem long-distance transport of CmNACP mRNA: Implications for supracellular regulation in plants. Development 126, 4405–4419. Sambrook, J., and Russel, D. W. (2001a). ‘‘Molecular Cloning,’’ Vol. 2. Cold Spring Harbor Press, Cold Spring Harbor, NY. Sambrook, J., and Russel, D. W. (2001b). ‘‘Molecular Cloning,’’ Vol. 3. Cold Spring Harbor Press, Cold Spring Harbor, NY. Suzuki, Y., Hibino, T., Kawazu, T., Wada, T., Kihara, T., and Koyama, H. (2003). Extraction of total RNA from leaves of Eucalyptus and other woody and herbaceous plants using sodium isoascorbate. Biotechniques 34, 988–993. Wang, S. X., Hunter, W., and Plant, A. (2000). Isolation and purification of functional total RNA form woody branches and needles of Sitka and white spruce. Biotechniques 28, 292–296. Weigel, D., and Glazebrook, J. (2002). ‘‘Arabidopsis: A Laboratory Manual.’’ Cold Spring Harbor Press, Cold Spring Harbor, NY. Zachgo, S., Perbal, M.-C., Saedler, H., and Schwarz-Sommer, Z. (2000). In situ analysis of RNA and protein expression in whole mounts facilitates detection of floral gene expression dynamics. Plant J. 23, 697–702.

[33] Developing Antibodies to Synthetic Peptides Based on Comparative DNA Sequencing of Multigene Families By Roger H. Sawyer, Travis C. Glenn, Jeffrey O. French, and Loren W. Knapp Abstract

Using antisera to analyze the expression of specific gene products is a common procedure. However, in multigene families, such as the -keratins of the avian integument where strong homology exists among the scale (ScK), claw (ClK), feather (FK), and feather-like (FlK) subfamilies, determining the cellular and tissue expression patterns of the subfamilies is difficult because polyclonal antisera produced from any one protein recognize all family members. Traditionally, researchers produced and screened multiple monoclonal antisera produced from the proteins of interest until an antiserum with sufficient specificity could be obtained. Unfortunately, this approach requires a lot of effort, and once obtained, such antisera may have limited applications. Here, we present procedures by which comparative DNA sequences of members from the -keratin multigene family were

METHODS IN ENZYMOLOGY, VOL. 395

Copyright 2005, Elsevier Inc. All rights reserved. 0076-6879/05 $35.00