An E. coli strain carrying BBa_K1758377 in pSB1C3 was induced to express T7 polymerase in medium with different concentrations of either GBL or GHB. As control, medium without GBL nor GHB was used. Induction lead to expression of sfGFP. However, sfGFP coding sequence follows on blc operator sequence. As the strain constitutively expresses BlcR, we expected the fluorescence signal to be higher when GBL or GHB were present in the medium as both analytes interact with BlcR and inhibit its binding to the operator.

Fluorescence signals of strains that had grown in medium with analytes were slightly higher, except for cultures with 1% GHB which showed inhibited growth.

These results indicated that, although a difference could be seen, the device has its limits in vivo. We conducted a CFPS with extract from strain constitutivly expressing BlcR. As reporter plasmid, BBa_K1758376 was used. This plasmid equals our CFPS positive control P_T7-UTR-sfGFP (see CFPS results) except that T7 promoter is followed by the blc-operator.

This however did not stop us from further testing. We took E. coli that constitutively expressed BlcR (BBa_K1758370). In less than a day we cultivated the cells, made cell extract via sonification and performed CFPS with BBa_K1758376 as reporter. In standard extract, fluorescence signals of our positive control plasmid and BBa_K1758376 were similar.

The results of the CFPS reaction surpassed all expectations. In vivo, BlcR reacts on GHB and GBL and thereby dissociates from the blc-operator (Chai et al. 2007). This effect could be observed when 0.3% GBL was present in the reaction, as the flurescence signal was greater when compared to the reaction without GBL. Still, for higher concentrations of GBL, protein synthesis was inhibited.

GHB also negatively affected protein synthesis in BlcR containing extract. Strikingly however, detrimental effects were far smaller than in standard extract! Especially interesting was that for 3% GHB, the fluorescence signal surpassed the 1% GHB signal. We suppose two reasons that together lead to this effect: When BlcR binds to GHB, on the one hand GHB is removed from the reaction and can not act detrimental on the molecular machinery, and on the other hand the polymerase is no longer blocked by BlcR, that means sfGFP can be expressed.

When we normalized the signals from BlcR containing extract on our standard extract in which GHB was strongly inhibiting, the effect of BlcR could not be overlooked. Consistent with findings from Chai et al. 2007, BlcR reaction on GHB is stronger than on GBL.

We therefore demonstrated that we can detect GHB at concentrations of 1% and 3% by normalizing the fluorescence signal to a control reaction. As our CFPS system is very robust even at ethanol concentrations of 5%, we can say that we built a cell-free sensor for GHB that can be used to detect the noxious substance in liquids.

In the final application, the potential of our sensor became evident. In our paper-based CFPS reaction, water that contained 1% (v/v) GHB was used for rehydration. Fluorescence signals were measured, data quickly evaluated and our app demonstrated that the water was contaminated with date rape drugs ingredients. For detail see our final sensor approach.