This web page was produced as an assignment for an undergraduate course at Davidson College.

Paper Reviewed:

"Programmable ligand-controlled riboregulators of eukaryotic gene expression"

Travis S. Bayer & Christina D. Smolke


Introduction to Antiswitches:

In this paper, Bayer and Smolke explore a synthetic mechanism for controlling gene expression. The regulation is conveyed through molecules called antiswitches, which are RNA molecules designed to turn genes on and off by acting as ligands in an allosteric regulatory mechanism.

This paper examines the mechanism and conditions required for antiswitch action by engineering a plasmid containing a gene encoding Green Fluorescent Protein (GFP) in addition to an ampicillin resistance gene and transforming S. cerevisiae with this plasmid. The target of the synthetic antiswitch used in this experiment is the gfp coding region of the plasmid. If the antiswitch does not disrupt the translation of GFP, the S. cerevisiae should fluoresce as a result of the expression of this gene. If the antiswitch is able to inhibit the expression of GFP, the S. cerevisiae will not be fluoresce green. Upon the basis of this general protocol, the authors tested the efficacy of antiswitches using variables of effector concentration, time, and specificity.

The antiswitches designed have two critical areas: the antisense sequence and the aptamers. The antisense sequence is designed to hybridize to a sequence surrounding the promoter of the target mRNA. The aptamer is a group of nucleotides that allow for ligand binding.


Figure 1

a. This figure is the authors' depiction of the antiswitch mechanism. Highlighted in this part of the figure is the fact that without the presence of the effector, the antiswitch is unable to bind to the desired mRNA. The authors suggest that this is a result of the antisense strand folding back upon itself, leaving no bases free to complement the mRNA. When this happens, GFP production is unaltered because the gfp coding region was undisrupted. However, upon binding with the effector, the antiswitch undergoes a conformational shift that frees the antisense strand of RNA. The antiswitch is then able to bind to the target mRNA and inhibit translation. GFP is no longer produced.

b. The sequence and hypothesized folding of the antiswitch synthesized for the purposes of this experiment is shown here. The authors have displayed critical portions of the molecule in various colors for clarity. The blue bases indicate the aptamer stem sequence. The antisense sequence is shown in red, while green indicates the start codon of the target mRNA. The relative free energy values for the antisense sequence and the aptamer stem sequence are also given. The free energy for the antisense sequence is more negative than that of the aptamer stem sequence; therefore, the former is less energetic and more stable than the latter. This suggests that the default conformation of the antiswitch is with the antisense sequence in a double-stranded complex.

c. Here the authors are looking at relative expression of GFP with changing concentrations of the effector, theophylline. Two negative controls and one positive control are included in this experiment. The authors expose the less reactive aptamer stem construct to theophylline and find, as they predicted, that the expression of GFP was unaltered. The graph of this result is depicted in green. Another negative control is characterized by the synthetic antiswitch, s1, exposed to caffeine as the effector. Again, little inhibition of GFP expression (only ~22% inhibition, relative to the first negative control) was seen, and this control is seen in orange on the graph. As a positive control, the authors used the antisense construct in the presence of theophylline. With this trial, there was roughly 95% inhibition of GFP, relative to the negative control. This result is apparent on the graph provided in red. The results stemming from the experimental conditions are displayed in blue. One can see that the drop-off in expression of GFP is quite abrupt, meaning that the antiswitch acts in an on/off mechanism rather than effecting a gradual change in expression. It seems that .8 mM theophylline is necessary for significant GFP inhibition.

d. The response time for antiswitch inhibition to take effect upon the addition of the effector is plotted in this figure. Yeast cells expressing GFP were cultured in the presence of the antiswitch construct only for slightly more than three hours. GFP expression during this period of time (red) increases nearly exponentially, as is indicated by the increase in relative emission of fluorescence. After three hours had passed, the effector, theophylline, was added in vivo to the yeast cells. The expression of GFP after the addition of the effector declined almost as rapidly as it had accumulated in the first three hours. The time taken for theophylline to effectively inhibit expression of GFP was roughly 4.5 hours. This experiment is well-controlled, also displaying a culture of GFP-expressing cells in blue in the graphical depiction of the experiment. The control cells seemed to level off in their GFP production shortly after the 3.5 hour mark; this expression curve, therefore, takes the form of a standard bacterial growth curve.

e. The authors then performed immunofluorescence with radiolabeled s1. In this assay, the amount of target transcript in solution was kept constant, but the concentration of the effector was altered. These solutions were run on a gel to determine their relative affinity. It seems that with higher concentrations of the effector comes higher affinity for the target mRNA. However, the authors did not provide any sort of molecular weight standard or positive control in this assay. Because of this, one cannot be sure that the same amount of solution was loaded into each well of the gel. Also, the resulting bars on the gel are quite smeared, so it is hard to make any comparative judgments concerning affinity in earnest.


Figure 2

a. Three derivatives of antiswitch s1 were synthesized by the authors, and this figure shows their predicted structures. Each structure has a different thermodynamic stability in its aptamer stem and in its antisense strand. Antiswitch s2 differed from s1 only in one base, indicated in green, which decreases the stability of the molecule when it is not bound to its ligand. In antiswitch s3, five nucleotides have been added to the antisense strand of s1. An aptamer stem also formed with three base pairs of the lower stem, creating stability and decreasing the energy of the molecule. Antiswitch s4 is a further destabilization of s2. There is an additional alteration in the loop sequence that changes a uracil to a cytosine. This makes the molecule more thermodynamically unstable and more energetic. In this figure, all antisense sequences are depicted in red, aptamer stem sequences in blue, and alterations in green. Relative stabilities of these molecules are confirmed by their relative free energies.

b. Because they determined the regulation activity of s1 over various effector concentrations in Figure 1c, they can now compare this with the activity of the new antiswitch constructs with various concentrations of the effector. In the case of s2 and s4 (red and green, respectively), the destabilization of the antisense stem caused the regulation activity to take effect with lower concentrations of theophylline than with s1. On the other hand, the stabilization gained in s3 necessitates a larger concentration of effector before regulatory action can begin. This displays that the thermodynamic activity of the antiswitch plays a role in the ease of the binding of theophylline.

c. Further comparisons of various antiswitch constructs occur in this figure. This graph shows the comparison in regulatory behavior of s5 and s6 constructs with that of s1. The s5 aptamer has a much lower affinity for theophylline than does that of s1. Antiswitch s6 takes the low affinity aptamer of s5 and destabilizes the antisense strain like in s2. The original construct, s1, is labeled in blue. Antiswitch s5 is depicted as green, while s6 is red. Because both new constructs have a very low affinity for theophylline, the authors predicted that a higher concentration of theophylline would be needed to inhibit GFP expression; this is, indeed, supported by the data.

d. This figure investigates whether or not the gene regulation trend observed in s1-s4 is unique to responses to theophylline. In this case, another antiswitch, s7, was synthesized to be sensitive to tetracycline. When exposed to various concentrations of tetracycline, s7 displayed a very similar response curve to that of s1. This indicates that all synthesized antiswitches may have a similar response in the presence of their respective effectors, but more experiments of this sort will need to be done before this conclusion can be validated.


Figure 3

a. In this experiment, the antiswitch construct is redesigned so that its default configuration allowed for an unpaired antisense strand that could hybridize with the target mRNA. This construct, called s8, changes shape in the presence of theophylline to cease its inhibition of GFP gene translation. This figure shows the sequence and predicted structure of s8 in the presence and the absence of theophylline. Critical regions are once again color-coded. The antisense sequence is red; the aptamer stem sequence is blue, and the start codon of the desired mRNA is green.

b. Here, s8 inhibition of GFP is tested over various concentrations as in Figure 2. As would be expected from an antiswitch whose preferred configuration is the inhibition of gene translation, GFP is almost entirely inhibited at low concentrations of theophylline. As the concentration of the effector grows, more inhibition is seen, and the switch takes place around 1 mM theophylline. In high concentrations of theophylline, almost no GFP expression exists. This is contrasted with the regulation behavior of s1, labeled in red. The two graphs switch at roughly the same concentration of effector, providing an interesting link between 'on' switches and 'off' switches.


Figure 4

a. Now that it has been determined that single antiswitches can dictate expression of single genes, the authors sought to discover whether or not two or more antiswitches could have a combined effect on gene expression. In this figure, the authors propose a general mechanism that will govern their hypotheses for part b. Two separate genes have been inserted into a plasmid and transformed in the yeast. One gene encodes GFP, while the other encodes Yellow Fluorescent Protein (YFP). GFP expression is still regulated by the effector, theophylline, but YFP production is regulated by the effector, tetracycline. Through this mechanism, the authors predict that addition of only theophylline will inhibit GFP translation and that the addition of only tetracycline will inhibit YFP translation.

b. This figure is the supporting evidence for the mechanism proposed in part a. The first pair of bars in the graph constitute a negative control. No effectors were added, so no inhibition was expected. When 5 mM of theophylline only was added, GFP expression was inhibited, just as the authors predicted. Upon adding only 5 mM tetracycline, only YFP translation was inhibited. When both effectors were present, both genes were inhibited almost completely. This is a well-presented figure, which adequately demonstrates the hypothesized mechanism of regulation. This figure shows that two antiswitches can have combined effects on multiple genes at one time.


Future Direction of Study

One possible step researchers could take at this point is to design more antiswitches using different effectors and compare the response curves for these different constructs. Data in this experiment suggest that all response curves will act alike, and further investigation in this area could confirm or dispel this belief. The discovery of antiswitches that are more or less sensitive to their respective effectors could present a specificity of action and timing of expression that doesn't exist now.

Another area for investigation would be the efficacy of antiswitches in different organisms. This study confirms that antiswitches can effectively regulate gene expression in bacteria, but such implications may not be contiguous in other species. Antiswitches could be activated in mice as a next step to see if gene regulation is as clean in higher organisms as it is in lower species. Once the presence or absence of gene regulation in mice was established, research could begin pertaining to gene therapy methods. Antiswitches could be engineered to inhibit the translation of cancerous oncogenes. For example, a mouse with liver cancer could be given a hepatitis-encased antiswitch to attempt to turn off an unregulated oncogene in the liver. Such tests would help researchers to discern whether or not antiswitches are good candidates for gene therapy.

Experiments could also be done in developing embryos of a model organism. The ability to prevent the translation of unwanted gene products could have far-reaching implications for dominant diseases like Huntington's disease. Antiswitches could possibly turn off the translation of the dominant, mutant allele of these genes, effectively preventing their onset. If such regulation was imposed on the germ cells, researchers could test to see if the regulatory mechanisms could even be passed on to the next generation.

Nonetheless, the biology of antiswitches opens up new possibilities concerning the specificity of synthetic gene regulation. It will be exciting to see to what lengths this newly found on/off switch will take gene expression studies.


Bayer T S, Smolke C D. 2005. " Programmable ligand-controlled riboregulators of eukaryotic gene expression." Nature Biotechnology 23 (3): 337-343.


Davidson College Biology Home Page

Molecular Biology Home Page

© Copyright 2005 Department of Biology, Davidson College, Davidson, NC 28036

If you have any questions, comments, or suggestions concerning this page, please contact