Schmitz et al., (2014). Cooperative gene regulation by microRNA pairs and identification using a computational workflow.

July 2014, model of the month by Florent Yvon
Original model: BIOMD0000000530


microRNAs (miRNAs) are small non-coding RNA molecules that contain about 22 nucleotides. They are involved in the regulation of gene expression at the post-transcriptional level. By sequence complementarity, they bind to mRNAs, which are then silenced as they cannot be translated any more. They can also induce mRNA degradation.

Human genome can encode about 1800 miRNAs [1]. They are involved in multiple diseases [2], such as:

Recently, it has been shown that miRNAs can cooperate to repress target gene expression more efficiently [3]. Two distinct miRNAs can mutually target a single mRNA, leading to the formation of a RNA triplex. The regulatory mechanism occurring with pairs of miRNAs is illustrated in Figure 1.

Schmitz et al. (2014) [4, BIOMD0000000530] propose a workflow for the identification and the analysis of novel RNA triplexes in human.

Figure 2

Figure 2SBGN representation of the model. Figure taken from [4] (supplementary material).


The authors have shown an effective workflow for the identification of regulatory miRNA pairs. They identified 15062 genes in human that are likely to be targeted by such a process, as a single miRNA pair can target multiple different genes. To be efficient, they showed that this mechanism must rely on a RNA triplex that meets the following criteria:

  • Stable local structure along the binding sites
  • Strong binding affinity between the three RNAs
  • Strong thermodynamical stability
  • Low dissociation and degradation rates

The modellers highlighted the fact that miRNA cooperation is a common mechanism in human and could be extended to other species. It can increase the efficiency of regulation of gene expression at the mRNA level, as single miRNA regulation tends to be only moderate. In the same way, this process shows a higher specificity. These two statements could encourage the design of new therapeutic strategies in human diseases that involve miRNA regulation processes.

Figure 1

Figure 1[On the left] Illustration of the regulation of a mRNA by two cooperative miRNAs compared to the regulation by one miRNA. [On the right] Illustration of the repressive effect on a target mRNA by single miRNAs (blue and red lines) and by cooperating miRNAs (green dashed line). This shows an effective repression effect in the case of cooperation. Figure taken from [4].

First, they identified possible RNA triplexes by predicting miRNA target sites on the 3'-UTR of mRNAs by considering a convenient distance between the two binding sites. Second, they predicted the secondary and tertiary structures of the resulting RNA triplex, and calculated the thermodynamics free energy, i.e., the stability of these complexes, as well as the binding affinity of the three components. Finally, they constructed a kinetic model of the cooperative regulation of mRNAs by two miRNAs, and performed simulations to measure the effectiveness of this repression process on gene expression.


The SBGN representation of the model in shown in Figure 2. Using ODEs, it takes into account the mRNA and miRNAs syntheses, the formation of duplexes by each miRNA with the mRNA, the formation of triplexes with both miRNAs and the mRNA and the dissociation of all these different complexes. The dissociation rate constants are depending on the thermodynamics stability of the corresponding complex that was calculated during the 3D modelisation. This model is a refinement of the one of their previous work [3]. Some other models already exist for describing mRNA regulation by single miRNA, but it was the first to describe such a process with two cooperative miRNAs.


The results are shown on Figure 3. They display the repression gain according to the expression rate of the miRNAs. The simulations were performed following three scenarios:

  • Highly up-regulation of miRNA1
  • Highly up-regulation of miRNA2
  • Moderate up-regulation of both miRNAs

Taken together, the results show that the cooperation is only effective for thermodynamically stable triplexes (Figure 3B lower panel). Moreover, they also show that the synergy only occurs for triplexes that are significantly more stable than their corresponding duplexes (Figure 3C lower panel). In this particular case, one can see that the gene repression is not only the result of the cumulative effect of the two single miRNAs but also the result of a significant synergistic effect from the cooperation of both miRNAs that increase the repression rate. In the cases where the triplex is not more stable than the corresponding duplexes (Figure 3C upper panel), the results show that the repression is due to only one of the two miRNAs, even if they are able to form a triplex.

Figure 3

Figure 3Simulation of gene repression by cooperative miRNA pairs. (A) Illustrative plot with scale. (B) Upper panel: results for unstable triplexes (high free energy). Lower panel: results for stable triplexes (low free energy). (C) Upper panel: results for insignificantly more stable triplexes than duplexes. Lower panel: results for significantly more stable triplexes than duplexes. Figure taken from [4].


  1. Human miRNAs in the miRBase
  2. Jiang et al. miR2Disease: a manually curated database for microRNA deregulation in human disease. Nucleic Acids Res. (2009); 37, D98-104.
  3. Lai et al. Computational analysis of target hub gene repression regulated by multiple and cooperative miRNAs. Nucleic Acids Res. (2012); 40, 8818-8834.
  4. Schmitz et al. Cooperative gene regulation by microRNA pairs and identification using a computational workflow. Nucleic Acids Res. (2014) May 29; pii: gku465. [Epub ahead of print]