News & Views | Published:

Use of microRNA sponges to explore tissue-specific microRNA functions in vivo

Nature Methods volume 6, pages 873874 (2009) | Download Citation



MicroRNA depletion by sequestration produces flexible hypomorphs that mimic mutants obtained by microRNA gene targeting.


MicroRNAs (miRNAs) are relatively new on the scene as regulators of gene expression. Their comparative novelty has captured the imagination of biologists and stimulated a flurry of activity in developing tools to study their functions in vivo. In broad terms, the approaches fall into two classes: those that manipulate the genome to compromise miRNA expression and those that seek to interfere with function by sequestering the mature miRNA product. In this issue of Nature Methods, Loya et al.1 report a new twist on miRNA sequestration for use in transgenic Drosophila melanogaster.

The traditional approach to studying gene function in vivo is to produce a mutant whose DNA is altered so as to impair gene function. Perhaps because of their small size, few miRNA genes have been hit in conventional genetic screens2,3. Transposon-based mutagenesis has provided access to some miRNA genes in the fly, but the method of choice for the past few years has been gene disruption using targeted homologous recombination. Gene targeting has the advantage of producing precisely the desired alteration and has been used to make fly and mouse miRNA mutants (for example, refs. 4,5).

The key advantage of targeted mutation strategies is that they remove the DNA encoding the miRNA from the genome. This provides a clean slate, in which the consequences of removing a miRNA can be assessed. The power of the approach is in the absolute lack of the product, but this is also its primary limitation: no product anywhere. The mutant organism will have a clear phenotype for the first time during development when the miRNA function is essential. Potentially interesting roles in other tissues or at other times may be obscured. It would clearly be useful to have the possibility of producing conditional mutants allowing tissue- or stage-specific impairment of gene function.

Conditional mutagenesis is standard in the mouse but is relatively new to the fly community. A new generation of targeting vectors allows the use of recombinase-mediated cassette exchange (RMCE) to facilitate subsequent modification of a targeted locus6,7. Essentially any sequence can be exchanged into the targeted locus8, without the need to go back to the beginning for a new round of targeting. This is an appealing way to make a conditional allele: target the locus to delete the miRNA, replace the RMCE cassette with the miRNA gene, now flanked by inverted loxP and attP sites. This rescues the original mutation and permits genetic validation that any defect observed was due to loss of the miRNA. At the same time it allows for tissue- or stage-specific deletion by conditional expression of a suitable recombinase to excise the miRNA cassette.

But for all its versatility, gene targeting takes time. Antisense RNA calls out the siren song, 'we could find out sooner'.

The logic of antisense-mediated miRNA depletion is simple. miRNAs bind their targets by base-pairing. Providing a template complementary to the miRNA should compete for binding to the 'real' target, sequestering the miRNA and alleviating regulation of endogenous targets. This strategy has been used effectively in cell-based assays and by injection of chemically modified synthetic antisense oligonucleotides in vivo9. However, there have also been cases in which the results obtained by antisense injection did not stand the test of time when compared with mutations removing the miRNAs (reviewed in ref. 2).

As an alternative to injecting synthetic targets, expression of transcripts with multiple miRNA sites provides a means to sequester miRNAs in vivo. The idea of using a transgene to 'soak up' endogenous miRNAs has led to the use of the term miRNA 'sponge' for this approach10. Virus-based expression of sponges has proven to be an effective means of miRNA depletion in cells, but there are limitations to the usefulness of this approach in vivo for conditional or tissue-specific depletion.

Loya et al.1 developed a variant of the miRNA sponge idea for use in transgenic flies that address these limitations. Gal4-regulated gene expression permits tissue-specific expression of miRNA sponges encoded by upstream activator sequence transgenes. The sponges contain 10 miRNA target sites with central bulges, to stably sequester the miRNA. The group placed these target sites in the 3′ untranslated regions of transgenes that express a fluorescent protein to label cells expressing the sponge in vivo.

How well do they work? The authors carefully assessed the phenotypes produced by expressing sponges for miR-7, miR-8 and miR-9a and compare these with existing miRNA mutants11,12,13. In each case the sponge produced an accurate but milder version of the known loss-of-function mutant phenotype. The miR-9a and miR-8 sponges were more effective if one copy of the miRNA gene had been removed. This makes sense, given that a sponge is unlikely to completely sequester its target miRNA.

The real potential of this tool is in tissue-specific miRNA depletion. Loya et al.1 use their sponge to describe a new role for miR-8 in formation of the neuromuscular junction and also observed this defect in the miR-8 deletion mutant. They then used Gal-4–directed expression to demonstrate that miR-8 is required in the muscle cells to support neuromuscular junction formation. My confidence in the conclusions of this study comes in part from the comparison of the effects of the sponges with loss-of-function mutants. If mutants are not available, it seems advisable to confirm, as these authors have done, that any defects observed using sponges are stronger in flies lacking one copy of the miRNA.

This study amply demonstrates the potential of miRNA sponges as a means to explore miRNA functions in vivo with high spatial and temporal resolution. It also holds the promise for depleting multiple members of a miRNA family with similar seed sequences, which may be difficult to achieve by mutation. There seems little doubt that miRNA sponges will be popular tools. Experience will tell whether we can rely on the data they provide without the confirmation provided by miRNA mutants.


  1. 1.

    , , & Nat. Methods 6, 897–903 (2009).

  2. 2.

    & Annu. Rev. Cell Dev. Biol. 23, 175–205 (2007).

  3. 3.

    et al. Nat. Genet. 41, 614–618 (2009).

  4. 4.

    , , & Curr. Biol. 18, 501–506 (2008).

  5. 5.

    et al. Genes Dev. 22, 3242–3254 (2008).

  6. 6.

    et al. Science 324, 54 (2009).

  7. 7.

    , , , & Genetics 183, 399–402 (2009).

  8. 8.

    , & Genetics 173, 769–777 (2006).

  9. 9.

    et al. Nature 438, 685–689 (2005).

  10. 10.

    , & Nat. Methods 4, 721–726 (2007).

  11. 11.

    , , & Genes Dev. 20, 2793–2805 (2006).

  12. 12.

    & Cell 123, 1267–1277 (2005).

  13. 13.

    , , , & Cell 131, 136–145 (2007).

Download references

Author information


  1. Stephen M. Cohen is in the Temasek Life Sciences Laboratory and Department of Biological Sciences, National University of Singapore, Singapore.

    • Stephen M Cohen


  1. Search for Stephen M Cohen in:

Corresponding author

Correspondence to Stephen M Cohen.

About this article

Publication history



Further reading

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing