RNAi In Drosophila Embryos

RNAi Synthesis

                           While we have not systematically determined the optimum length of a double-strand RNA (dsRNA) for maximum interference activity, some circumstantial evidence suggests that lengths of 700 to 800 bp are most active. However, we have found that dsRNAs as short as 200 bp and as long as 2000 bp have potent interfering activities. The dsRNA can be made from cDNA or genomic DNA templates, as long as most of the dsRNA corresponds to exon regions. dsRNAs with two or more exon regions interrupted by introns will work quite well. One can use dsRNAs corresponding to coding sequence and/or untranslated regions (UTRs). However, the most prudent approach is to use a dsRNA corresponding to UTR sequence (either UTR can mediate interference) or unique coding sequence since a dsRNA to one gene can cross-interfere with another gene's activity if the dsRNA is sufficiently similar in coding sequence to the second gene.

                           There are two choices in preparing dsRNA for interference. One, you can synthesize RNA strands separately from linearized plasmid templates, and subsequently anneal the strands together. Two, you can synthesize both RNA strands simultaneously from a PCR fragment which contains a T7 promoter on each end. Both methods are given below.

A. Plasmid Template Method

1.                     Subclone a fragment of your gene of interest into a vector with T7, T3, and/or SP6 promoters flanking the polylinker, e.g. BlueScript. Linearize plasmid on either side of the insertion site to create two preps of linear DNA template, one to synthesize the sense strand and the other to synthesize the antisense strand. Remove enzymes and restriction buffer by your favorite method, and dissolve DNA in TE buffer.

2.                     Perform RNA synthesis reactions in 50l volume with 1g of DNA template using appropriate phage RNA polymerase. Follow standard protocols as described in general lab manuals.

3.                     Remove DNA template with RNase-free DNAase. Phenol-chloroform extract and precipitate the RNA with NH4OAc and Ethanol.

4.                     Dissolve RNA in 5l annealing buffer (1mM Tris pH 7.5, 1mM EDTA) and measure yield spectrophotometrically. Typical yields of RNA from 1 g DNA template are in the 40 to 50g range.

5.                     To anneal, mix equimolar quantities of sense and antisense RNAs in annealing buffer to a final concentration of 0.45M each.

6.                     Heat small aliquots (11.1l) of the mixture in a 150ml beaker of boiling water for 1 min, at which time remove the beaker from the heat source and allow it to cool to room temperature for 18 hours. IMPORTANT: The 18 hours is critical to produce high yields of dsRNA. Shorter time periods are not effective for some reason. Degradation of RNA is not a problem if care is taken in earlier steps not to contaminate with RNase, i.e. always wear gloves!

7.                     Store RNA aliquots as a NaOAc/ethanol precipitate at -80C until immediately before use.


B. PCR Template Method

1.                     Choose primer sequences that will amplify the region you want to act as the dsRNA template. We use MIT's Primer3 program at http://www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi to choose optimal primer sequences for any given region. Complementary sequences should be 20 to 24 nt in length with a 22 nt optimum and 60C optimum Tm.

2.                     The 5' end of each primer should correspond to a 27 nt T7 promoter sequence (TAATACGACTCACTATAGGGAGACCAC). Thus, each primer will be approximately 47 to 51 nt long (27 + [20 to 24]).

3.                     Perform a 50l PCR reaction with T7-linked primers and suitable template. Cloned plasmids or phage are optimal, but the method will also work on RT-PCR DNA or genomic DNA. The first 10 cycles should have a 40C annealing step, followed by 35 cycles with a 55C annealing step.

4.                     Phenol-chloroform extract and ethanol precipitate in NH4OAc. Dissolve in TE buffer and measure concentration spectrophotometrically.

5.                     Perform an RNA synthesis reaction in 50l volume with 1g of PCR DNA template using T7 RNA polymerase. Follow standard protocols as described in general lab manuals.

6.                     Remove DNA template with RNase-free DNAase. Phenol-chloroform extract and precipitate the RNA with NH4OAc and Ethanol.

7.                     Dissolve the dsRNA in TE buffer and measure yield by spectrometer. Typical yields of RNA from 1 g DNA template are in the 80 to 100g range.

8.                     Aliquot and store dsRNA as a NaOAc/ethanol precipitate at -80C until immediately before use. The RNA becomes double-stranded during the synthesis reaction.


                           No matter how the dsRNA is prepared, you should test its condition by native agarose gel electrophoresis in TBE. Electrophorese 6-10g of RNA and stain with ethidium bromide. Alternatively, trace-label RNA with 32P-ATP during synthesis, and visualize electrophoresed products by autoradiography. Only preparations in which the electrophoretic mobility of most of the RNA is shifted to that expected for dsRNA (very close to duplex DNA mobility) of the appropriate length should be used. Sometimes we will observe a smear of higher-order RNAs migrating slower than the dsRNA species. Although these may constitute over half of the RNA present, we find that the interfering activities of these preparations are similar to more homogeneous preparations.


When preparing to transform Drosophila embryos, several days of preparation are required before injections begin.  Flies that are deficient in the allele that is to mark the transformants, most often ry506 or w1118, must be in abundant supply, both for egg collections and later crosses.  Bottles of flies for egg collection are set up by taping an egg laying plate to the bottom of a plastic beaker with air holes punched in it with a 20 guage or smaller needle.  Bottles  should be moved to a day for night schedule at least two days before collections begin, as this will improve the number of eggs being laid. 1. Dissolve the dsRNA precipitate in injection buffer (Rubin and Spradling, 1982) dissolve in injection buffer solution (0.1mM Na Phosphate pH 7.8, 5mM KCl).  Just before filling the needle, spin injection solution for 10 minutes to avoid any particulate matter that might enter and later jam it to a final concentration of no greater than 5M. If the solution is too viscous or makes alot of bubbles in the needle, try a lower concentration of dsRNA. We do not keep the dsRNA solution longer than a day before discarding. Microfuge the RNA at 25 for 10 min.                    

2. Needles are borosilicate glass capillaries with an internal filament (World Precision Instruments TWF100-4). Pull needles and bevel or break a tip to a diameter of 0.5-2.5m. Wear gloves!

There are a variety of ways to prepare needles.  This is one of the most important steps for injections, as the sharpness of the needle and its ability to stay unclogged will to a large degree determine how many embryos survive the process and thus how many days will have to be spent injecting.  I use a method of beveling that uses a slurry of grinding powder and a regular magnetic stirring set-up (see technical notes by James Powers, <A HREF="gopher://ftp.bio.indiana.edu/00/Flybase/news/dinvol12.txt">DIN Vol. 12</A>)  The slurry is made from silicon carbide powder (Grit 120; Buehler Ltd., Lake Bluff, IL) and ddH2O at a 1:3 ratio.  The grit must be washed several times to remove the smallest particles which will remain suspended after the bulk has settled out.  The slurry should be autoclaved before use and every week to prevent any bacterial contamination.  A shallow beaker should be used.  For injections we use a 1mm capillary pipet, pulled to form a needle and then back filled with the DNA solution before sharpening.  With the slurry being stirred at a relatively fast speed, but not so fast as to throw grit everywhere, the tip of the pipet should be inserted at 135˚ angle with respect to the direction of flow of the slurry.  By holding the needle steady for 4-5 minutes, the tip of the needle will be beveled to a sharp point.  This may provide a wide enough opening for injecting, but if the DNA does not flow freely from the tip when injecting begins, gently touch the tip of the needle to the edge of a slide under a microscope while applying gentle positive pressure with your syringe.  One side of the tip will be weakened from the beveling and  should break easily, providing a wide but sharp tip for injections.

3. Back- or front-fill the needle with dsRNA solution.

4. Eggs must be collected every hour and injected within the hour, in order to introduce the DNA before cellularization takes place.  Before the first round of injections, a fresh egg laying plate with yeast paste should be put on the bottle for half an hour to induce the flies to lay any over developed eggs they may have been carrying.  The eggs are then collected every hour by exchanging a fresh plate with yeast paste for the old plate, which is then wetted slightly and brushed lightly to release the eggs from the agar.  The eggs are then strained out and washed in ddH2O or PBS The chorion, the "shell" on the embryos, must be removed before injecting.  Although some use bleach for this step, this drastically reduces survival rates.  We manually dechorionate by brushing the embryos onto a piece of double-stick tape.  A relatively sharp tipped instrument (a push-pin stuck on a pencil will do) is used to gently rub the side of each embryo till it rolls over and pops out of its chorion.   The embryo is then picked up and placed on the edge of a thin strip of agar cut from an egg laying plate.  The eggs should all face in the same direction, with their anterior end facing out from the agar.  You can recognize the anterior end by the dorsal appendages, which are two string like parts that come off the egg with the chorion.  Also located on the anterior end is the micropile, a small nipple-like extension from the egg that should be visable after removing the chorion.  When 40-50 embryos have been lined up on the agar strip, a coverslip with double stick tape on it is gently lowered onto the eggs.  The eggs should end with one quarter to one fifth of their posterior ends over the edge of the double stick tape.  It may be a good idea to use 1/4" 3M, type 415 tape, as this tape has supposedly been tested non-toxic, whereas other double-stick tape has not (see technical notes by Mary Whiteley and Judith A. Kassis, <A HREF="gopher://ftp.bio.indiana.edu/00/Flybase/news/dinvol11.txt">DIN Vol. 11</A>)  The coverslip should then be placed in a container with Drierite for 5-15 minutes.  This time should be adjusted until the embryos are loose enough to inject without leaking, but not so loose as to appear "wrinkled" or bag-like.  After desiccation, the eggs are covered by a thin layer of oil.  We use Halocarbon Oil Series 700, CAS# 9002-83-9, Halocarbon Products Corporation. Collect embryos over a 60 min period at 25C and dechorionate in 50% bleach for 2 min. After rinsing embryos with water, attach them to a coverslip coated with either rubber cement (when embryos are to be later analyzed by cuticle morphology) or Tape Glue (when embryos are to be later analyzed by histochemistry).

5. Dessicate embryos and cover in 700 Halocarbon Oil.

6. Injection location is typically on the ventral side extending from 30-75% egg length. However, any location seems to work since varied locations give similar RNAi effects on the same target gene.  The dsRNA diffuses uniformly throughout the syncitial embryos within minutes after injection.  The average injection volume we use is 85pL but ranges from 65 - 110pL, as determined by measuring the diameter of droplets injected into halocarbon oil. We use a pneumatic pump (PicoPump, WPI) for injections. This is obviously the point at which most of the embryos get killed.  The needle should be inserted quickly in the center of the posterior end and the injection should place as much of the solution as close to that end as possible, as this is where the germ cells we want to transform are located.  The needle should then be pulled out quickly to avoid any leakage.  There are several problems that will tend to arise during injections, each with its own solution.<P>


<LI>If embryos leak immediately upon being punctured, they have been under-desiccated.  Do the best you can with this slide and increase the desiccation time by a few minutes next time.<P>

<LI>If many of the embryos are "empty shells", reduce the desiccation time.<P>

<LI>If the eggs are not easily punctured by the needle and leak often when the needle is pulled out, the needle is not sharp enough.  Try re-breaking the tip on the edge of the slide, or make a new needle.<P>

<LI>If the embryo leaks while you are injecting, you are trying to inject too much DNA solution.  Use a lighter touch or a smaller syringe.<P>

<LI>If the needle jams frequently, there is either an excess of particulate matter in the injection solution or the needle's opening is too small.  Remember to spin down the solution before back-filling your needle.

7. Typically, we try to deliver approximately 0.2 fmoles of dsRNA per embryo. This is a dose that has produced interference effects for all of the genes that we have tested thus far.

Eggs should be placed in a humidified environment at 18˚C.  They should be checked every twelve hours in order to verify that they are properly humidified and covered with oil.  Add as much oil as is nescesary to keep the eggs covered, although not too much.  Larva will begin hatching after two days and should be collected at least twice a day using the same type of instrument that was used for dechorionation.

Cuticle Analysis

1. Incubate embryos at 18C under oil for two days.

2. Dissect embryos that have secreted cuticle from their vitelline membranes.

3. Wash embryos in Rinse Buffer (PBS + 0.5% Triton X-100) to remove oil.

4. Wash briefly in Glycerol/Acetic Acid (1:4) and fix overnight at 60C.

5. Mount in 3:1 (v/v) CMCP-10 Mounting Medium/Lactic Acid and bake overnight at 70C. CMCP-10 can be purchased (without a DEA license) from Masters Chemical Company (Bensenville, IL).

Immunohistochemical Analysis

1. Incubate embryos at 25C under oil. Embryos should be attached to coverslips with Tape Glue (extract of 3 cm length of Scotch double stick tape per ml heptane).

2. At the appropriate stage, collect the embryos for staining. With a razor blade, scrape excess oil from coverslip taking care not remove embryos. Remove as much oil as possible.

3. Take coverslip off of slide and hold coverslip over a 60 mm dish of heptane. With a pasteur pipet, wash the embryos on the slide with heptane until the oil is washed away (about 6 sprays). Continue washing until all of the embryos are washed into the dish.

4. Remove the heptane and replace with fixative in the depression well. The fixative is 10:3:7 (v/v) n-heptane/37% formaldehyde/PBS that has been previously vortexed to saturate both phases. Ensure that the well has both phases present. Cover with a slide and incubate at room temperature for 30 min.

5. With a tip-cut P1000, remove the embryos from the fix interface and pipet onto the outer surface of a fine mesh basket stuffed with Kimwipes. Let the solvent blot through the mesh leaving the embryos on top. It may be necessary to blot them with a dry Kimwipe from the inside.

6. Immediately pick up the embryos from the mesh with a strip of double stick tape.

7. Immediately apply the strip to the bottom of a 60 mm plastic dish (embryo side up) and submerge under PBS.

8. Manually devitellinize embryos with the tip of a 21 gauge needle and forceps. Popping them with a nudge from the posterior is often sufficient to release them.

9. With a tip-cut P1000, transfer the embryos to a glass depression well filled with PBS + 0.1% Triton-X100 (PBT).

10. Carry out standard antibody incubations and washes with all steps done in the depression well.

Beta-Galactosidase Detection

1. Perform steps 1 - 3 as in Immunohistochemical Analysis.

2. Add glutaraldehye-fix to the depression well. Cover with a slide and incubate at room temperature for 7 min. Fix is prepared by mixing together 20:1:19 (v/v) n-heptane/50% glutaraldehyde/PBS and vortexing.

3. Perform steps 6 - 10 as in Immunohistochemical Analysis.

4. Wash embryos a couple of times with PBT.

5. Incubate embryos at 37C in XGal Stain (XS) Buffer and stop reaction by washing embryos with PBT.

Final Comments

                           We have observed interference of gene activity with as little as 30 molecules of dsRNA per cell, well below the transcript abundance of most genes. Thus, dsRNA at sub-stoichiometric levels is sufficient to interfere with gene activity. That said, however, we also observe variability in the interference activities of different dsRNAs. While some dsRNAs systemically generate null phenotypes with high efficiency, others generate localized or patchy null effects at the same doses. Several factors may play a role in this variability. First, some phenotypes may be less sensitive to gene activity than others. Second, differences in size or sequence composition may affect interference activity, stability, or transport of dsRNA within an embryo. Third, different genes may be unequally sensitive to interference based on relative accessibility to dsRNA. On the other hand, some dsRNAs at high doses (0.2-0.4 fmoles) produce phenotypes that are greater than the known null phenotypes of their corresponding genes. We suspect that these dsRNAs are interfering with several genes sharing a common sequence domain since these dsRNAs correspond in part to those domains.

                           Therefore, it is prudent to initially perform a dose titration experiment with a dsRNA to determine the range of phenotypes that are produced. A phenotype that progressively becomes more severe or systemic is a likely true representation of a gene's loss-of-function phenotype. This can be confirmed by interfering with a different dsRNA corresponding to the same gene.

                           This basic protocol has worked well in our hands. We have tried RNAi on ten genes and successfully generated phenotypes in all ten. The protocol is an adaptation of the one used in our paper: [Kennerdell, J.R. and Carthew, R.W. (1998) Use of dsRNA-mediated genetic interference to demonstrate that frizzled and frizzled 2 act in the Wingless pathway. Cell 95, 1017-1026.] References to specific manufacturers and products are just meant as suggestions. Good luck and if you have any questions, please contact us.

Rich Carthew                                                                                         Jason Kennerdell
(412) 648-7687                                                                                      (412) 648-7688
carthew@pop.pitt.edu                                                             jasonk@pitt.edu

Department of Biological Sciences
Clapp Hall Room 217
4249 Fifth Avenue
University of Pittsburgh
Pittsburgh, PA 15260