|Home | About | Journals | Submit | Contact Us | Français|
A major goal in cell biology is to understand the molecular mechanisms of the biological process under study, which requires functional information about the roles of individual proteins in the cell. For many non-genetic model organisms researchers have relied on the use of inhibitory reagents, such as antibodies that can be microinjected into cells. More recently, the advent of RNA-mediated interference (RNAi) has allowed scientists to knockdown individual proteins and to examine the consequences of the knockdown. In this chapter we present a comparison between microinjection of inhibitory reagents and RNAi for the analysis of protein function in mammalian tissue culture cells, providing both a description of the techniques as well as a discussion of the benefits and drawbacks of each approach. In addition, we present a strategy to employ RNAi for organisms without a sequenced genome. While the focus of our research is on the organization of the mitotic spindle during cell division and thus the examples utilized are from that system, the approaches described here should be readily applicable to multiple experimental models.
Understanding the molecular mechanism of any biological process requires the complete description of the role of each protein involved in that process. To achieve this goal, it is necessary to have an experimental means to perturb protein function as well as an assay to determine the functional consequences of that perturbation. The choice of approach will be determined by the reagents available to the protein of interest, the equipment available, whether or not the gene sequence is available, as well as the time-course of the experimental process being analyzed.
Microinjection of inhibitory antibodies or dominant-negative reagents has long been a powerful means to inhibit protein function in many cell types (1). However, it has often been criticized because it is questionable whether true loss of protein function is achieved. It is possible that the antibody is exhibiting cross-recognition of other proteins within the cell or that the antibody is simply binding in situ to the protein of interest and causing non-specific blocking of other protein interactions. Often these drawbacks can be overcome by examining the effects of multiple antibodies to the same protein or by complementing antibody injection studies with other inhibition methods.
RNAi has become an extremely useful tool for looking at protein function in many cell types. RNAi has revolutionized how most scientists view protein function studies, and the importance of this discovery is best highlighted by the awarding of a 2006 Nobel Prize to Andrew Z. Fire and Craig C. Mello, the scientists who first described this process (2). To carry out RNAi in vertebrate cells, short dsRNAs are introduced into the cell by transfection (3). This dsRNA then pairs with the endogenous mRNA and induces its degradation by a series of enzymatic activities. Because RNAi knocks out the mRNA, new protein synthesis is inhibited, and the protein levels decrease over the timecourse of the normal turnover of the protein of interest.
In contrast to microinjection, RNAi does not require a purified antibody or dominant-negative reagents, but it does require some information about the individual gene sequence. For organisms in which the genome is sequenced, finding siRNAs to knockout any gene of interest is as easy as searching the website of companies such as Dharmacon or Ambion for their collection of pre-designed RNAs. If a favorite gene is not included in the pre-designed collection, then designing a siRNA only requires entering the accession number of a protein into programs such as Block-IT siRNA Designer (http://rnaidesigner.invitrogen.-com/rnaiexpress/) or Dharmacon siDesign Center (http://www.dharmacon.com/sidesign/default.aspx). In the case of organisms without sequenced genomes, it is still possible to use these siRNA design programs by entering a short amount of sequence obtained by RT-PCR or from a cDNA clone.
Perturbation of protein function by either microinjection of inhibitory antibodies or RNAi should be considered complementary methods of inhibition. Both methodologies have their own strengths and weaknesses that influence their suitability to answer a particular scientific question. For example, microinjection of inhibitory antibodies is quick and will typically display immediate changes in cell behavior and morphology. This allows the experimenter to time the injection relative to the process being analyzed. In contrast, RNAi requires a period of incubation to allow time for the targeted protein to be degraded. With antibody injection the experimenter can inject higher concentrations of the antibody to achieve complete inhibition, whereas with RNAi, sufficient residual protein may remain to carry out all or part of its cellular function. In microinjection, only a small number of cells are often examined, but the exact cell that was injected is known and therefore can be examined phenotypically. In contrast, RNAi is useful to examine a large number of depleted cells. However, since knockdown can vary across a population of cells, it is often difficult to determine if a particular cell shows a phenotypic effect due to depletion unless appropriate antibodies are available. Because of the unique characteristics of each methodology, we use both techniques as complementary approaches to more fully understand the cellular processes we are studying.
Tissue culture cells generally adhere to glass somewhat poorly. Treating glass coverslips with poly-l-lysine helps cells remain attached to the surface keeping them flatter, which improves the ability to follow the intracellular events more clearly. The coverslips are first extensively washed with acid, which etches the glass, and then coated with poly-l-lysine.
All of the subsequent protocols should be performed in a sterile tissue culture hood using sterile techniques and solutions to avoid contamination and its potential spread to other cultures (see Note 7.). For more in-depth information on media, cells, and maintenance techniques, please refer to (9,10).
In the study of cellular processes, the analysis of fixed cells by fluorescence microscopy is usually the first step to assessing the initial phenotypic effects on the cell by either microinjection or RNAi. Specific steps for these procedures are outlined below for the study of PtK2 cells. For more indepth information, please refer to (11).
Microinjection of either antibodies or dominant-negative reagents into vertebrate tissue culture cells is a powerful way to analyze protein function. The cells can be injected during interphase or during mitosis so it is possible to achieve temporal resolution of the experiment. The first part of this section describes the preparation of the injectate (Section 3.4.1.) and the microscope and rose chamber (Section 3.4.2.) followed by an outline of the microinjection steps (Section 3.4.3.). We then present the application of the microinjection for either a fixed time-point experiment (Section 3.4.4.) or for live imaging (Section 3.4.5.).
When first analyzing the function of a protein by microinjection of antibodies, it is easiest to assess the phenotype by injecting the antibody, waiting a set amount of time and then fixing the cells and processing them for immunofluorescence. If it is not known whether the affected process is during interphase or mitosis, it is simpler to start with interphase cells because they are easier to inject and there are more of them on a coverslip. When looking for a mitotic defect, it is easiest to inject only cells at prophase and then allow the cells to incubate at 37°C to progress through mitosis. For interphase cells, we usually start with a 2-h incubation in antibody and have extended that time period up to 24-h post-injection before analysis. For mitotic cells, we usually fix at 30-min post-injection for prometaphase/metaphase defects and 40-min post-injection for anaphase defects. The cellular morphology as well as the cell cycle stage is scored under a fluorescence microscope.
Knocking down protein levels by RNAi is a fairly simple way to address protein function. However it relies on the availability of cDNA sequence to design the siRNA. We describe a fairly straight-forward approach to obtain a sufficient amount of sequence for siRNA design to apply this technology to model systems without a sequenced genome. We then describe a basic protocol for transfecting adherent cells grown in culture. All steps should be performed in a sterile tissue culture hood using sterile techniques and solutions.
The rat kangaroo genome has yet to be sequenced, therefore no databases exist to search for gene sequences of interest. Traditional methods of cloning cDNAs are laborious and time consuming; however, comparison of homologous sequences across different mammalian species often shows a high degree of DNA identity through portions of the coding region if not the whole sequence (6) that can be used to design primers for RT-PCR, which in turn are used to generate siRNAs.
This is a modified version of the Invitrogen protocol “Transfecting siRNA into HeLa Cells Using Oligofectamine” (http://www.invitrogen.com/content/sfs/protocols/sirna_oftsf_proc.pdf) (3,13).
The authors would like to thank Susan Kline for early instruction in microinjection and suggestions on transfections of PtK2 cells. The authors would also like to thank Chantal LeBlanc for editing of the manuscript. Work in the Walczak lab is supported by NIH R01GM059618, an ACS Scholar award RSG CSM-106128, and in part by the Indiana METACyt Initiative of Indiana University, funded in part through a major grant from the Lilly Endowment, Inc. Rania Rizk is supported by a predoctoral fellowship from the American Heart Association.
1HCl is extremely corrosive, inhalation of vapor can cause serious injury, and liquid acids can cause severe damage to skin and eyes. Use appropriate precautions when making and handling the hot acid, and dispose of the acid properly. If a hot plate and glass container are used to acid-wash the coverslips, the temperature and acid level will have to be monitored much more closely to make sure the acid does not overheat, and the glass container should be loosely covered to prevent excessive evaporation. The use of a hot plate requires that this step be carried out under a ventilated hood.
2All tissue culture plates are not created equal. We have experienced problems with unhealthy cells when we have attempted to try plates from different manufacturers. When changing any reagents involved with tissue culture, it is always prudent to test the new batch on a subpopulation of cells before completely switching over.
3The antibodies used will determine the specific fixative needed. To look at microtubule structure with DM1α anti-tubulin antibody (Sigma), we find that fixation with 4% formaldehyde; 0.1% glutaraldhyde in PHEM buffer (60-mM PIPES, 25-mM HEPES, 10-mM EGTA, and 4-mM MgSO4, pH 7) works well. However, many of our other polyclonal antibodies do not work well with glutaraldehyde fixation, and we typically use 2–4% formaldehyde in PHEM buffer without the glutaraldehyde. Cold (−20°C) 100% methanol can also be used to fix cells in which microtubules are to be visualized, and this works for some of our other antibodies as well. It should be noted that methanol fixation can cause some distortion in the condensed chromatin of mitotic cells. Dispose used fixative in accordance to local and state regulations.
4Antibody dilutions can be stored at 4°C; however, the concentrations and the lengths of time these dilutions can be used will vary from antibody to antibody and will have to be empirically determined by each lab.
5For siRNA transfection in PtK2 cells, several popular lipid-based reagents were tested, many of which caused vesicularization in cells, which can interfere with subsequent imaging by microscopy. We found that Oligofectamine (Invitrogen) produced the most consistent results with the least degree of cytotoxic effects compared to other transfection reagents tested.
6siRNAs purchased from Dharmacon have been more successful for us than siRNAs from other companies both in their reliability and efficiency of knockdown. For a negative control, Dharmacon’s non-targeting siRNA #2 designed to Luciferase works well in PtK2 cells, whereas the GFP siRNA (GCAAGCUGACCCUGAAGUUCAU) (14) produced cytotoxic effects in our PtK2 cells.
7It cannot be stressed enough the importance of an ever-present diligence to avoid contamination. Cultures should be routinely examined under the microscope to assess the health of the culture and inadvertent cross-contamination of other cell lines, or contamination by fungi or bacteria. Cells should also routinely be tested for contaminants such as mycoplasma, either by PCR or immunofluorescent assays.
8Many cell lines are sensitive to cell density; plating at densities either too low or too high can have an adverse effect on the cells.
9A 35-mm plate holds 2-ml of media and four 12-mm coverslips; a 60-mm plate holds 4-ml of media and 12 12-mm coverslips or two 22 × 22 mm coverslips and four 12-mm coverslips; and one 100-mm plate holds 10-ml of media and up to 30 12-mm coverslips.
10The number of wells to be plated for RNAi will depend on how cells will be analyzed. We typically plate the cells in 35-mm dishes or six-well plates (each well is equivalent to a 35-mm dish). For live imaging, cells are plated on 22 × 22 mm coverslips in 60-mm dishes. For immunoblots, cells from three wells for each experimental condition are trypsinized, washed, and counted. The number of wells harvested for immunoblots can also be dependent on the antibody used.
11The density at which cells will be plated for RNAi experiments depends on many factors such as the growth rate of the cells, the media in which the cells are grown, and the number of days required to achieve knockdown of the protein as determined by either immunofluorescence, immunoblot, or qRTPCR. For Eg5 RNAi in our PtK2 cells, the effect is seen in as little as 24-h after transfection though we fix cells for immunofluorescence by 48-h. For kinesin-13 MCAK RNAi, cells are processed after 72-h for efficient knockdown (6). Other factors include cell density at processing; overly confluent cells at processing will retard the number of mitotic cells and cause difficulty in imaging. However, the transfection efficiency is at times better if the cells are more confluent versus less, so sometimes a balance needs to be maintained. The optimal plating density will need to be empirically determined.
12The concentration of antibody used must be empirically determined. We often start with a needle concentration of 1–2 mg/ml and have rarely needed to go above 5 mg/ml. There are, however, reports in the literature in which people have used much higher concentrations. The buffer in which the antibody is stored must be at physiological pH, cannot have too high a salt concentration and must not contain sodium azide or other preservatives, which will kill cells.
13While it would be ideal to have a thermoprobe directly monitoring the temperature of the media while imaging, this is not possible due to space restraints and perturbation of the cells. Instead, we attach the thermoprobe to the top-side of the rose chamber. Because the temperature at the top of the chamber will be different than inside the media-filled chamber, it is necessary to predetermine this difference before injections. To do this, place the assembled rose chamber on the stage (Fig. 7.1B). Attach one probe directly on top of the glass coverslip, using tape. Add ~1-ml of media and a thin layer of mineral oil on top of the media. Place a second thermoprobe at the top-side of the rose chamber (Fig. 7.1B). Turn the ASI on and allow the temperature reported by the probe inside the rose chamber to reach the desired temperature and remain constant. This probe reflects the closest estimation to the temperature of the cells during imaging. Now record the temperature of the probe placed at the side of the chamber. This is the temperature that will be used to maintain cells at the actual desired temperature. In our setup, we have determined that the temperature of the coverslip within the rose chamber is typically 1°C lower than that of the probe on the top of the rose chamber.
14As the needle is lowered into the imaging media, it is in a direct flow of the ASI and will heat up to temperatures several degrees warmer than the media. This increased temperature of the glass needle will kill the injected cell, making it imperative that the ASI be turned off right before injections.
15Troubleshooting injections: For a clogged needle, press the clear button, or gently scrape the tip of the needle along a cell-free area of the coverslip to unclog it. Be careful because this latter method can also damage the needle tip. A clogged needle can also be due to a high concentration of the injectate or aggregates in the injectate. To alleviate these problems, try diluting the sample or centrifuging at a higher speed before loading the needle.
16For live imaging of mitotic PtK2 cells under phase contrast microscope we use 100-ms exposures at 30-s intervals for 120-min, which allows us to follow the events of mitosis without damaging the cell. This can be optimized depending on the nature of the experiment.
17Using degenerate primers is not typically recommended in RT-PCR. However we have had some success if we can limit the degree of degeneracy of the primers. Additionally, in some instances our RT-PCR gave multiple products, which required isolating the different bands and sequencing each band. By comparing the sequences back to the original alignments, we identified the band with the most identity to the homologous gene sequences as the intended product. Additionally, if we could obtain at least a partial sequence of the RT-PCR product, we could further amplify the PCR product by doing nested PCR on the RT-PCR product.
18If transfecting cells in multiple wells with the same siRNA, cocktail mixes can be used. In this case, multiply each reagent added by a fraction over the number of wells to be transfected. For example, if transfecting three wells with luciferase siRNA, add 3.2-fold more of each reagent to the appropriate tubes: 9.6-µl of Oligofectamine to 38.4-µl of incomplete media and 32-µl of siRNA to 560-µl of incomplete media. The siRNA/lipid complexes can then be added to 2560-µl of RNAi media in a 5-ml sterile tube instead of the 800-µl of RNAi media being added to the complexes. Aliquots of this cocktail are then used to replace the rinse media.
19The ratio of lipid to siRNA may need to be optimized for each siRNA used to achieve the most efficient knockdown. Using excess lipid or siRNA will increase non-specific cytotoxic effects of the RNAi transfection. Some gene targets may require multiple siRNAs for efficient knockdown or may require that the concentration of oligonucleotide be adjusted to increase the transfection efficiency. Dharmacon has a useful complementary guide book, RNA Interference – Technical Reference & Application Guide (www.dharmacon.com) that has many useful suggestions on how to troubleshoot RNAi.
20The length of time a cell line has been subcultured affects the efficiency of the knockdown. This seems to be particularly true for PtK2 cells, and only those cells subcultured for 8 weeks or less should be used for RNAi.
21The percentage of knockdown that needs to be attained to see a phenotypic effect can vary greatly depending on the gene of interest. For example, we see spindle defects in PtK2 cells when we have reduced MCAK by 70% as shown by western immunoblot. However, CENP-A has to be reduced by >90% to cause mislocalization of other centromere proteins (15).
22The ideal situation is that both the analysis of the phenotypic effects of knockdown and the assessment of knockdown efficiency by immunofluorescence would be performed on the same cell. We can do that for Eg5 RNAi, since both the Eg5 antibody and the anti-tubulin antibodies used to determine spindle defects work in methanol fix. However, this is not always possible because the antibodies used to assess both the knockdown and the phenotypic effects may require different fixation methods. In this case, we plate four coverslips per well so that two coverslips are used to determine knockdown and use the other two coverslips for the phenotypic analysis in order to reduce as many experimental differences as possible in the treatment of these cells.