MicroRNAs and Muscle Cell Differentiation miRCURY LNA™ microRNA Knockdown Libraries
Dr. Annick Harel-Bellan
Dr. Anna Polesskaya
Dr. Annick Harel-Bellan (AHB) is Directeur de Recherche at the Institut Andre Lwoff in Paris. She heads a group working on epigenetics and cancer (Laboratoire Epigenetique et Cancer). Dr. Anna Polesskaya (AP) is a senior scientist and longstanding member of this group.
1. What is the main focus of the research conducted in your lab?
(AHB-AP) We are interested in understanding the mechanisms of epigenetic control of cell fate and the underlying regulatory mechanisms that govern cell differentiation.
2. What made you interested in microRNA?
(AHB) We have been interested for many years in exploring the mechanisms of cell fate control during the terminal step of muscle cell differentiation. In particular we are are studying the role of microRNAs and other small ncRNAs in the regulation of gene expression during myogenesis. We have discovered that miR-181 is differentially expressed during differentiation of our murine cell culture model. In vivo
, mir-181 is poorly expressed in adult muscle tissue, but strongly induced in regenerating muscle. In order to explore miR-181 function we designed a loss-of-function assay based on antisense inhibitors containing LNA (we were the first to do that).The experiments showed that mir-181 plays a crucial role in switching on the early programme of gene expression during myogenesis by allowing expression of MyoD, the key transcription factor during muscle cell differentiation . mir-181 excerts this effect by attenuating translation of hoxA11 mRNA, that encodes a transcription factor that inhibits myoD transcription. Mir-181 is therefore essential to the establishment of terminally differentiated muscle cells, but is probably not required for the maintenance of this cell type (Naguibneva et al, Nature Cell Biol., 2006).
3. What is the advantage of using LNA in microRNA inhibitors?
(AHB) As we showed in Naguibneva et al. (An LNA-based loss-of-function assay for micro-RNAs. Biomed Pharmacother. 2006 Aug 28;), LNA microRNA antisense inhibitors are efficient and specific with long lasting effects.
4. You are currently engaged in a large project aimed at identifying novel microRNA involved in muscle cell differentiation. Could you describe the experimental approach chosen for this project?
(AP) We have chosen a novel human skeletal myoblast cell line as our model system. These cells were transfected in 96 well plates with the Exiqon miRCURY LNA™ KD library of microRNA inhibitors. Differentiation was induced 24 h later, and the efficiency of differentiation was evaluated by immunofluorescent staining at 7-day time point. In a parallel approach, the expression of miRNAs during terminal differentiation of these cells was profiled by Exiqon array services.
5. What did the expression profiling tell you?
(AHB) The data shows surprisingly dramatic changes over time. A large number of microRNAs are differentially regulated and they fall in different categories of expression kinetics: generally we observe up- and downregulated microRNAs, but a few microRNAs appear to be transiently differentially regulated during the process. Interestingly, many of the differentially regulated microRNAs are related – either belonging to the same family or closely associated on the chromosome, and possibly deriving from the same primary transcript.
6. How do you monitor the effect of the KD probes on differentiation?
(AP) During myogenesis, individual myoblast cells fuse to form long polynucleate cells. To enable high throughput automated detection of these myotubes, we stain them with Hoechst 33242 (nuclear stain) and antibodies against a late marker of terminal muscle differentiation, MHC (Myosin Heavy Chain). See images below. The images are acquired by an automated microscope and sophisticated image analysis software is used to detect and quantify phenotypic changes such as decreased/increased efficiency of differentiation, increased proliferation or cell death, and other phenomena caused by the KD of specific miRNAs in differentating myoblasts.
7. Setting up large scale parallel transfections and automated image analysis seems like a major challenge. What type of expertise and equipment was required?
(AP) The RNAi group in our laboratory helped us use their pipetting-diluting robot MLStar, (Hamilton) for plating and transfection of the human myoblasts. To acquire and analyze the images, we have set up collaboration with the I-Stem Institute in Evry, France. The specialists of I-Stem have helped us to develop the techniques for the automated image analysis using the ArrayScan (Cellomics).
8. How did you set up the transfection conditions?
(AP) To determine the reagents and conditions to efficiently transfect human myoblasts, we have used the AllStars hs Cell Death control siRNA from Qiagen. Afterwards, we have used the LNA inhibitors of miRNAs which are known to be important for terminal muscle differentiation (miR-181, miR-206, miR-133) to set up the working conditions for the KD screen.
9. How did you setup the image analysis?
(AP) Here again, we have used the LNA inhibitors against miR-181, miR-206, and miR-133 in increasing concentrations, to obtain different levels of inhibition of terminal differentiation. The multinuclear, MHC-positive myotubes were recognized by the image analysis software, and the level of inhibition of terminal differentiation was quantified, and compared to visual observation, and Western Blot analysis of MHC expression.
10. With a project of this scale surely you must have encountered a series of problems?
(AP) We have had problems with edge effects, likely due to the evaporation of the media during the long incubation of the plates (7 days). Increasing the volume of the differentiation media from 100 to 200 m l
resolved this problem. Using robots for transfection meant that we could not plate the cells first and transfect them afterwards, because of mechanical damage to the cells in the middle of wells caused by ejection of transfection mixture from the pipette tips. Therefore, we had to set up a reverse transfection protocol adding the transfection reagents first and the cell suspension afterwards. But it worked out surprisingly well. Unfortunately, the long incubation times combined with the use of robots also increased the risk of contamination, which was important in our original screen (3%). In the ongoing secondary screen, done to confirm the original “hits”, we have paid special attention to the problem of contamination.
11. The primary screen was recently performed and the preliminary results look exciting. Can you describe the general nature of the results and how you plan to go from here?
(AP) We were very (pleasantly) surprized by the large number of “hits”. As many as 120 LNA inhibitors have been included in the secondary screen, based on the variety of strong phenotypical changes observed in the cells. 55 of the “hits” have been detected in our miRNA profiling of myoblasts, as being differentially regulated between myoblasts and myotubes. The vast majority of these “hits” are novel candidates for a role in skeletal myogenesis, though we also identified some previously-described “myoMIRs”, such as miR-133 and others. In order to weed out false positive hits we will analyze the results of a secondary, confirmation screen, and we will then re-synthesize the LNA inhibitors for the most promising miRNA “hits” to test the kinetics and the dose-dependence of the candidate miRNAs in terminal differentiation. In a parallel approach we will try to ascertain whether the KD phenotype can be rescued with synthetic (mimic) miRNAs. We have previously used this approach with success (Naguibneva et al, Nature Cell Biol., 2006). The confirmed and validated novel miRNAs will be analyzed in silico to predict their specific targets with a role in terminal muscle differentiation.
12. What is the value of performing a genome wide screen with microRNA inhibitors in parallel with microRNA expression profiling?
(AHB) The power of the functional screen is that it leads you directly to hits that are functionally important. However, essential microRNA regulation might go unnoticed in a screen due to functional redundancies of individual microRNA- ie. microRNA belonging to families or coregulated microRNA that together play an important regulatory function. Expression profiling generates a different but complimentary set of data that is a useful cross-reference when evaluating screening data. In addition, it is useful to pinpoint microRNA families and transcriptional clusters that merit closer investigation even if they did not show up in the KD screen. We plan to do this using experimental 3rd generation inhibitors specially designed by Exiqon.
13. Why is screening with a complete library important – couldn’t you have performed a restricted screen against differentially expressed microRNA detected in your expression profiling data?
(AP) Many of the LNA inhibitors that significantly changed the phenotype of human myoblasts in our KD screen target microRNAs that appear not to be differentially expressed. This observation suggests that constitutively expressed miRNAs can nevertheless be essential for terminal myogenic differentiation.The truth is that a large proportion of the hits were not obvious candidates based on the expression profiling data.
14. Based on your experience do you have any helpful advice for colleagues contemplating using KD libraries?
(AP) KD libraries represent a very powerful tool for the discovery of novel miRNA-dependent pathways. It is important to apply the KD screening to an efficient and robust cell culture model with a quantifiable read-out that can be adapted to high-throughput conditions.
15. What is the current limitation with functional analysis tools and what type of novel tools would you like to see developed?
(AP) The profiling of miRNA expression during terminal differentiation shows that the members of a number of miRNA families are regulated in the same manner. The LNA inhibitors of individual miRNAs belonging to these families did not induce any phenotypic changes in our KD screen, which can be a result of functional redundancy between the miRNAs of the same family. It will be very useful to include the inhibitors of the whole families of miRNAs (such as let-7, miR-29, miR-30 and others) in KD libraries in the future.
Automated image analysis based on MHC and multinucleated cells
| Hoechst 33342 (note the multinuclear myotubes )
Myosin Heavy Chain
Recognition of MHC-positive multinuclear myotubes by ArrayScan
|Inhibition of differentiation with a-miR-206 |
| Hoechst 33342 (no multinuclear cells)
Myosin Heavy Chain (not detected)
No MHC-positive multinuclear myotubes detected by ArrayScan