The main experiments we performed to develop our sensor. For our daily activities and experiments, visit our Notebook. For a list of our protocols, check out our Protocols page.

Cell-Free System

Linear versus Plasmid DNA

We obtained promising data from adding plasmid DNA to our cell-free system. However, because the Collins paper specified the use of linear DNA in their experiments, we compared the signal from eGFP in linear and plasmid form in a cell-free reaction. The linear construct does not produce a stronger signal than the plasmid construct.

Reaction Volume Reduction

We decreased the cell-free reaction volume from the 25 µL given by the PURExpress protocol. To make their sensor, the Collins group only freeze-dried 1.8 µL of reaction onto paper discs, so our system would have to produce a signal with a much smaller reaction volume. In addition, performing experiments with a smaller reaction volume would make PURExpress last longer. PURExpress was one of the most expensive reagents we used, yet it was also central to our project. We set up reactions with a range of volumes from 1 to 25 µL and measure the fluorescence from eGFP. The reaction produces an appreciable signal when the volume is as low as 5 µL.


To obtain a good reading from the plate reader, we had to add at least 10 µL of liquid to the plate wells. For endpoint assays, we diluted the completed cell-free reactions by two before reading, which gave us good results. We wanted to try diluting the reaction before incubation to run a time course without increasing the amount of PURExpress used. However, cell-free reactions to not proceed nearly as quickly when diluted.

Back to Top


We performed our DNAzyme assays in two ways. We used polyacrylamide gel electrophoresis (PAGE); however, the gels rarely provided clear, informative results. Thus, we also put the DNAzymes in cell-free reactions with the toehold switch. If the toehold switch was not triggered, the substrate strand of the DNAzyme remained sequestered by the catalytic strand. If the toehold switch was triggered, the substrate strand must have activated the switch, either because cleavage had occurred or because the substrate strand was not properly sequestered.

Substrate Sequestration

Because the substrate strand of the DNAzyme duplex contains the trigger for the toehold switch, it is essential that the strand is annealed to the catalytic strand. The catalytic strand sequesters the substrate strand and prevents it from activating the toehold switch. To look for sequestration, we annealed the substrate and catalytic strands of each DNAzyme together in ratios that ranged from 1:1 to 1:1000. The resulting duplexes were analyzed on a native polyacrylamide gel, which suggested that the substrate strand is never fully sequestered. However, in cell-free reactions, the duplexes did not activate the toehold switch.

Hairpin versus Duplex

To circumvent the presence of unsequestered substrate strand, we synthesized hairpin DNAzymes that connect the two strands into one DNA strand that forms a hairpin loop. Based on experiments in cell-free extract, the hairpin produces minimal activation of the toehold switch in its folded state, and it produces much greater activation of the switch than the duplex when the substrate is cleaved.


To analyze the activity of the DNAzymes, we added metal ion and observed the completed reactions on denaturing polyacrylamide gels. The gels do not show cleavage product for either the lead or thallium DNAzymes, but adding metal ion to the cell-free reactions with the DNAzyme and toehold switch results in expression of LacZ, suggesting that cleavage does, in fact, occur.

Concentration Optimization

To determine the concentration of DNAzyme that would produce the most activation of the toehold switch in reaction without inducing background expression of LacZ, we tried a range of concentrations of DNAzyme in the reaction. Preliminary results suggest that a concentration of 100 nM is too high, but we did not have time to pursue these results.

Lead Titration

To determine the detection limit of our system and to see if lead interfered with our circuit, we tried adding a range of concentrations of lead to the reaction. Preliminary results suggest that concentrations as low as 50 nM can be detected (the EPA MCL is 72.4 nM), but we did not have time to pursue these results.

Back to Top

Toehold Switch

Our initial experiments were performed with the D and G LacZ switches from the paper from the Collins group, "Paper-Based Synthetic Gene Networks". However, because G switch activity was neater in some experiments, we switched to using exclusively G.

RNA Trigger

Using plasmids from the Collins group that contained the switch and trigger for the two LacZ switches, we expressed lacZ in cell-free reactions. The two triggers were switch-specific, and the switch did not produce LacZ in the absence of a trigger.

DNA Trigger

Because our genetic circuit relies on the activation of the toehold switch by the cleavage product of a DNAzyme, we needed to make sure a DNA oligonucleotide could also trigger the toehold switch. When we added the DNA oligo trigger to a cell-free reaction with the switch, lacZ was expressed. DNA triggers toehold switches.

Sequestered DNA Trigger

The DNA oligo should not activate the toehold switch if it is bound to another DNA strand. In our system, the trigger should be sequestered by the catalytic DNAzyme strand as part of the DNAzyme's substrate strand. We added the hybridized DNAzyme duplex in a cell-free reaction with the toehold switch. When hybridized to another DNA strand, the trigger does not activate the switch.

Back to Top



To integrate amilCP into our genetic circuit, we assembled the BioBricks provided by iGEM to form a promoter-gene-terminator construct. Following the scheme illustrated below, we assembled a plasmid to express amilCP regulated by T7. "Prom" refers to the reporter (pT7 with RBS in this case), "report" refers to the reporter (amilCP), and "term" refers to the terminator.



Because wild-type lacZ contains an EcoRI site, it cannot be BioBricked. We were unable to grow a mutated version of lacZ from the bacterial stabs sent to us by iGEM, so we mutagenized the Collins toehold switch to obtain a lacZ sequence with a mutation in the EcoRI site.

Back to Top


To construct our amplifier, we followed the same assembly scheme as illustrated for amilCP. The T3 constructs were synthesized from IDT, then added into the plasmid backbone to assemble.

Back to Top