Visualizing siRNA in Mammalian Cells:
Fluorescence Analysis of the RNAi Effect

Mike Byrom, Vince Pallotta, David Brown*, and Lance Ford*
Ambion, Inc.
2130 Woodward Street
Austin, TX 78744-1832
* Corresponding authors dbrown@ambion.com, lford@ambion.com

Abstract
Labeling GAPDH and c-myc siRNA with either fluorescein or Cy3 was demonstrated to have no effect on silencing efficiency, but allowed visualization of the distribution patterns of these molecules after transfection. Using fluorescence microscopy we have analyzed the dynamics of siRNA localization. Our studies revealed that the majority of both GAPDH and c-myc siRNA was localized to the nuclear periphery. In addition, in vivo siRNA strand separation was observed, supporting similar observations from earlier in vitro studies (Zamore et al., 2000).

Using fluorescently labeled siRNA, the time course of gene silencing was examined at both the mRNA and protein level. GAPDH protein and mRNA levels were moderately reduced by 4 hours post transfection, but intensified to maximum levels by 24 to 48 hours post transfection. Both mRNA and protein levels remained suppressed as long as 10 days post transfection. The initiation of the silencing effect was found to correlate well with the release of siRNA from the transfection agent in the cytoplasm of transfected cells and the occurrence of strand separation during this period of time.

While still a new and rapidly evolving technology, siRNA induced silencing is already regarded as a technique that has great promise not only for mammalian functional genomics, but as a therapeutic agent. In order to realize this potential, the mechanism of action in living cells must be fully elucidated. To further the understanding of siRNA distribution and metabolism, the development of methodology to track siRNA in cells is required. Fluorescent labels are commonly used to follow the movement of small molecules in and between cells. However, siRNAs must interact with cellular components and it was not known whether the addition of bulky dyes would negatively affect the capacity of siRNAs to induce gene silencing. We prepared fluorescently labeled double-stranded siRNAs with the Silencer siRNA Labeling Kit (Ambion, Inc.), and compared the labeled siRNAs to their unlabeled counterparts. We confirmed that labeled siRNAs remained fully functional and tracked their movements after transfection. Double labeled siRNAs were used to track siRNA strand separation in intact mammalian cells, supporting previous in vitro observations (Nykanen et al., 2001).

Results
Fluorescently Labeled siRNAs Retain Functionality. Two commonly used fluorescent labels, Cy3, and 5 carboxy-fluorescein (FAM) were coupled to siRNA using the Silencer siRNA Labeling Kit and evaluated for potential interference with siRNA silencing. Cells were transfected with an siRNA to the 3' UTR of c-myc (Jarvis and Ford, 2001; Demeterco et al., 2002) that was either unlabeled, labeled with Cy3 alone (on the antisense stand), or double labeled with Cy3 on the antisense strand and FAM on the sense strand. It has been shown that down-regulation of c-myc can cause a decrease in cell proliferation (Kimura et al., 1995). The ability of unlabeled and labeled (on one or both strands) siRNA to affect cell proliferation rate was evaluated in HeLa S3 cells. Labeled and unlabeled siRNA performed equivalently. Growth rates of cells transfected with the labeled or unlabeled c-myc siRNA were reduced to an equivalent level as compared to the scrambled control siRNA, which remained equivalent to that of non-transfected cells (Figure 1B). To confirm that the siRNA-induced cell proliferation defect was due to c-myc silencing, we examined protein expression 48 hr post-transfection using immunofluorescence microscopy (Figure 1A). The results clearly demonstrate that the labeled and unlabeled siRNAs were equally capable of reducing protein expression, whereas the samples transfected with labeled and unlabeled scrambled control siRNAs showed no silencing.

Figure 1. Labeled siRNA is Functional. A siRNA against the 3' UTR of c-myc was labeled using the Silencer siRNA Labeling Kit (Ambion, Inc.) with Cy3 and FAM on one or both strands as described in Materials and Methods. A. Following labeling the siRNA was transfected into HeLa S3 cells grown on cover slips using siPORT Lipid Transfection Agent (Ambion, Inc.). 48 hr after transfection HeLa S3 cells were analyzed for c-myc protein expression using immunofluorescence as described in Materials and Methods. c-myc protein is shown in green. B. Labeled or unlabeled siRNA targeting c-myc was transfected into HeLa S3 cells. At 72 hours post-transfection HeLa S3 cells that were transfected with siRNA containing one, two, or no label were counted using a hemocytometer. The cell number was calculated relative to a non-treated cell population and graphed.

Figure 2. siRNA Silencing of GAPDH Gene Expression. siRNA to human GAPDH and a scrambled control were fluorescently labeled with Cy3 using Ambion's Silencer siRNA Labeling Kit, transfected into HeLa S3 cells, and analyzed by fluorescence microscopy with an anti-GAPDH antibody. Red: Cy3 labeled siRNA; Blue: DAPI stained nuclei; Green: fluorescein labeled antibody to GAPDH. A. siRNA silencing of GAPDH expression in HeLa S3 cells. B. Scrambled control siRNA has no effect on GAPDH protein levels.

To extend these observations to other genes, fluorescently labeled GAPDH and ß-actin siRNAs were tested (Figure 2) and found to be just as effective at silencing cognate protein expression as their unlabeled controls. Both synthetic and in vitro transcribed siRNAs were capable of silencing, whether labeled on the sense and/or the antisense strand, and labeled with either Cy3 or FAM (data not shown). There appears to be no cytotoxic effect associated with the fluorescent moiety as we observed no additional cell death using the trypan blue exclusion assay (data not shown).

Fluorescently Labeled siRNA Can Be Tracked in Cells. Fluorescently labeled siRNAs to c-myc were generated by labeling the sense strand RNA with FAM and the antisense strand with Cy3. The Cy3-FAM double labeled siRNAs were transfected into cells and visualized by fluorescence microscopy to determine their subcellular distribution pattern. 48 hr post transfection the siRNAs were typically found localized to discrete foci on the cytoplasmic side of the nuclear membrane (Figure 3).

Figure 3. siRNA Strand Separation. A. siRNA against the 3' UTR of c-myc labeled on the sense strand with FAM and on the antisense strand with Cy3. 48 hr following transfection, HeLa S3 cells were fixed and analyzed using an Olympus fluorescence microscope. The two views represent different magnifications. Arrows point to red and green foci (individual siRNA sense and antisense strands). B. siRNA against the 3' UTR of c-myc labeled on the sense strand with Blackhole quencher and on the antisense strand with Cy3 were transfected into HeLa S3 cells. After 4, 24 and 48 hr following transfection cells were fixed and then analyzed using an Olympus fluorescence microscope.

In dividing cells, fluorescence was focused at the midline of division, suggesting that siRNA might divide equally between daughter cells. Although the majority of transfected Cy3-labeled antisense strand (red) and FAM labeled sense strand (green) siRNA was found in large yellow foci on the cytoplasmic side of the nuclear membrane, a small amount of red and green signal was also observed (Figure 4). Isolated red or green signal is the result of fluorescence from individual siRNA strands. This apparent strand separation is consistent with previous in vitro studies (Nykanen et al., 2001). To confirm and extend these observations, we used fluorescence resonance energy transfer (FRET) analysis to assess siRNAs in transfected cells. In this experiment, the sense strand of c-myc was labeled with Blackhole quencher 1 (IDT) and hybridized to the complementary Cy3 labeled antisense strand. In the double strand form, the Cy3 signal is quenched by the Blackhole quencher and almost no signal is observed during fluorescence microscopy. The quenched, double stranded siRNA was transfected into cells and examined for fluorescence at 4 , 24, 48, and 72 hr post transfection. As in the previously described dual labeling experiment, a minor amount of siRNA fluorescence was detected at 4 hr (possibly due to incomplete quenching of Cy3). At later time points, a significant increase in signal was detected (Figure 3B), showing that the siRNA sense and antisense strands separate in intact mammalian cells.

Figure 4. In Vivo Distribution of siRNA. Panel A. Cy3 labeled siRNA was transfected into HeLa S3 cells grown on cover slips using an NBD-labeled carrier cationic lipid. At the time indicated, the samples were analyzed using a fluorescence microscopy for the relative distribution of the siRNA (red) and the lipid (green). The siRNA was localized to regions in close proximity to the nuclear membrane following release from the carrier lipid. Panel B. HeLa S3 cells grown on glass slides were transfected with siRNA against c-myc labeled with FAM on the sense strand or Cy3 on the antisense strand. 48 h post transfection, the cells were processed and analyzed using a Leica confocal microscope. Arrows point to individual sense and antisense siRNA strands (denoted by the red and green foci, respectively).

Figure 5. Induction and Duration of RNAi Induced by siRNA. A. HeLa S3 cells grown in a 24-well dish were transfected with siRNA to the 3' UTR of GAPDH. After 4 hours, 24 hours, 3, 6, 10 and 12 days following transfection, the cells were harvested and analyzed by Northern blot analysis for both GAPDH mRNA and 28S rRNA levels. B. A graph of Northern data in Panel A showing relative GAPDH mRNA levels. C. HeLa S3 cells grown on cover slips in a 24-well dish were transfected with siRNA to the 3' UTR of GAPDH. After 4 hours, 24 hours, 3, 6, 10 and 12 days following transfection, the cells were harvested and the GAPDH protein expression was analyzed using immunofluorescence. DAPI (blue) stains cell nuclei; an antibody to GAPDH (green) binds to expressed protein.

Figure 6. Distribution of siRNA in Dividing HeLa S3 Cells. HeLa S3 cells grown on cover slips were transfected with Cy3 labeled-siRNA (red) and analyzed using immunofluorescence. Dividing cells that contained siRNA were detected in a population of cells on cover slips by observing chromosome condensation and nuclear separation/reformation. DNA was analyzed using DAPI. In all cases where the cell were dividing or appeared to have just undergone a division event the siRNA was localized to the central region of the cell.

Figure 7. siRNA Associated with Nuclei of Dividing Cells. HeLa-S3 cells grown on cover slips were transfected with FAM labeled siRNA to GAPDH. The cells were harvested and fixed using 4% Paraformaldehyde at 48 hours post transfection. The cells were mounted on microscope slides using VectaShield with DAPI and examined using the appropriate fluorescence filters. The siRNA (green) can be seen associated with each nuclei (blue) of a dividing cell.

The localization of transfected siRNA into discrete foci raised the question of whether the siRNA was being concentrated in lipid vesicles derived from the transfection carrier. To address this question, we repeated our transfections using an NBD-labeled carrier lipid (green) to transfect a Cy3 labeled siRNA (red). We examined the distribution of red and green fluorescence at multiple time points over a period of 72 hr after transfection. At as little as 4 hr post transfection, a small amount of siRNA "freed" from lipid vessels was detected while the majority of lipid and siRNA co-localized within foci in the cells, yielding yellow fluorescence. An increased amount of "free" siRNA was observed 24 and 48 hr post transfection (Figure 4). While protein silencing was first detected at 4 hr post transfection, we observed more significant suppression of protein expression at 24 hr and maximum protein reduction at 48 hr post transfection (Figure 5). The period of greatest silencing at both the mRNA and protein level correlated with the time of maximum release of siRNA from the carrier lipid and at the time of greatest increase in the amount of strand separation (data not shown).

HeLa S3 cells were transfected with siRNA targeting GAPDH to examine the induction and duration of silencing. GAPDH had reduced mRNA and protein levels for up to 10 days, during which time the cells should have undergone 10-12 doublings. This data suggests that the siRNA was passed on to daughter cells. If the siRNA was not passed to daughter cells, the percentage of cells with siRNA would be vastly reduced in only 2 or 3 days. Supporting immunofluorescence data showed that the siRNA was centrally localized in dividing cells (Figure 6) and appeared to enter daughter cells (Figure 7).

Discussion
Our observation that fluorescently labeled siRNAs can effectively enter the RNAi pathway and elicit gene silencing enabled us to follow siRNA within the cells and to draw conclusions about its mode of action and metabolism.

1. Localization
Using labeled siRNA, we found that the dsRNA accumulated in the cytoplasm near the nucleus. Four hours post-transfection, the majority of transfected siRNA remains in lipid vessels. However by 48 hr, the siRNA has dissociated from its lipid carrier and appears to be localized in the cytoplasm in close proximity to the nuclear membrane (Figure 4). This localization could represent sites of siRNA processing or sites where the RISC resides. Previous research on RNAi suggests that long dsRNA degrades mature (e.g. cytoplasmic) RNA (Montgomery et al., 1998). While it is unclear whether siRNAs in mammalian cells operate by the same mechanism, the accumulation of siRNA in the cytoplasm seen in these experiments is consistent with this scenario. Another possible explanation for the perinuclear localization is that the siRNA is clustered around nuclear pores where it can "scan" mRNA being transported to the cytoplasm. When a complementary sequence is detected, that message is then targeted for cleavage leading to gene silencing.

2. Strand Separation
Although dissociated single stranded siRNA was detected within the cell, the majority of the transfected siRNA appears to remain in a double stranded state. Perhaps the extent of strand separation is proportional to the molar amount of siRNA required to direct cleavage of the target gene.

3. Duration
Data presented here suggests siRNA is maintained in cells up to 10 days and that siRNA is transferred to daughter cells. If this is true and siRNA amplification does not occur in mammalian cells, then the duration of the RNAi effect should be directly proportional to the cellular concentration of the siRNA and the number of cellular divisions that occur. This suggests that a major experimental variable in siRNA experiments is the effectiveness of siRNA delivery and the initial cellular siRNA concentration.

In a situation in which an effective siRNA is transfected with low efficiency, labeled siRNA can be used to identify individual cells that have been successfully transfected with siRNA. Cells lacking labeled siRNA can thus serve as internal negative controls. Labeled siRNAs can also be used for metabolic studies, especially if performing experiments geared towards identifying clinically important molecules.

Materials and Methods
Labeling siRNA
Chemically synthesized siRNA (Ambion) was labeled using the Silencer Cy3 and FAM siRNA Labeling Kits (Ambion). Briefly, 5 µg of a chemically synthesized siRNA to c-myc and its scrambled control were labeled using 7.5 µl of labeling reagent. After labeling, the single strands were purified, resuspended and hybridized prior to transfection as per the manufacturer's protocol.

Transfections
HeLa S3 cells were plated at 50,000 cells/well in a 24 well tissue culture plate containing glass cover slips (11 mm). Transfections were performed 24 hr after plating using siPORT Lipid Transfection Agent (Ambion) according to protocol with the siRNA at a final concentration of 100 nM in the tissue culture media.

Immunofluorescence
Cells were harvested 48 hours post transfection, washed, and fixed for 5 min in 4% paraformaldehyde prior to immunofluorescent staining for c-myc. Cells were washed with PBS, permeabilized in 0.1% Triton X-100 for 5 min, washed again, and blocked in 3% BSA for 1 hr. The primary antibody was added for 1 hr, the cells were washed, and the secondary antibody was added for 1 hr. Cells were then washed and mounted using VectaShield with DAPI (Vector Labs). The primary antibody c-myc Ab-5 (NeoMarkers, Freemont, CA; Cat# MS-1054) was used at a 1:200 dilution; the primary antibody against GAPDH (RDI; Cat# TRK5G4-6C5) was used at 1:2000 dilution. The secondary antibody, FITC labeled Donkey anti-Mouse IgG (Jackson ImmunoResearch), was also used at a 1:200 dilution. All washes, dilutions, and incubations were done in 1X PBS at room temperature with agitation. Cells were examined using an Olympus BX60 microscope with the appropriate fluorescence filters. Photomicroscopy was performed using a Hitachi KP-C571 camera.

RNA Isolation and Analysis
Total RNA from cell cultures was extracted using the RNAqueous™-4PCR Kit (Ambion) and quantitated by spectrophotometer. The expression level of both target and control genes was determined for each experimental sample using the NorthernMax™-Gly Kit (Ambion).

Cy3 is a trademark of Amersham BioSciences.
The Silencer siRNA Labeling Kits contain reagents manufactured for Ambion by MIRUS.

References

1. Brown D, Jarvis R, Pallotta V, Byrom M, Ford L (2002) RNA Interference in Mammalian Cell Culture: Design, Execution and Analysis of the siRNA effect. Ambion TechNotes 9(1): 3-5.
   
2. Demeterco C, Itkin-Ansari P, Tyrberg B, Ford LP, Jarvis RA, Levine F (2002) c-Myc controls proliferation versus differentiation in human pancreatic endocrine cells. JCEM (in press).
   
3. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, and Tuschl T (2001) Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411: 494-498.
   
4. Hammond SM, Boettcher S, Caudy AA, Kobayashi R, and Hannon GJ (2001) Argonaute2, a link between genetic and biochemical analyses of RNAi. Science 293: 1146-1150.
   
5. Jarvis R and Ford LP, (2001) The siRNA target site is an important parameter for inducing RNAi in human cells. Ambion TechNotes 8(5): 3-5.
   
6. Kimura S, Maekawa T, Murakami A and Abe T (1995) Alterations of c-myc expression by antisense oligodeoxynucleotides enhance the induction of apoptosis in HL-60 cells. Cancer Research 55:1379-1384.
   
7. Montgomery MK, Xu S, Fire A (1998) RNA as a target of double stranded RNA-mediated genetic interference in C. elegans. Proc. Natl. Acad. Sci. USA 95: 15502-15507.
   
8. Nykanen A, Haley B, Zamore PD (2001) ATP Requirements and Small Interfearing RNA Structure in the RNA Interference Pathway. Cell 107: 309-321.
   
9. Tuschl T, Zamore PD, Lehmann R, Bartel DP, and Sharp PA (1999) Targeted mRNA degradation by double-stranded RNA in vitro. Genes Dev. 13: 3191-3197.
   
10. Zamore PD, Tuschl T, Sharp PA, and Bartel DP (2000) RNAi: double-stranded RNA directs the ATP-dependent cleavage of mRNA at 21 to 23 nucleotide intervals. Cell 101: 25-33.

Related Articles

Overview of RNA Interference

siRNA Target Finder

siRNA-Mediated Gene Silencing in Mammalian Cells

RNA Interference Resource

Top Ten Tips for a Successful siRNA Experiment