Two sensors were constructed measure glutamate. One was designed to be soluble, good for measurements of the effectiveness of the isolated sensor. The other was designed to be attached to cell membranes, good for measuring glutamate near neurons.

The soluble form of the sensor was made by tethering together the two fluorescent proteins, ECFPenhanced cyan fluorescent protein, and Citrinea type of yellow fluorescent protein. This put the two proteins in proximity to one another so that the fluorescent emission of ECFPenhanced cyan fluorescent protein might be absorbed and re-emitted by Citrinea type of yellow fluorescent protein. GltIglutamate-binding protein was interposed between the two proteins. This protein from Escherichia coli binds glutamate and changes its shape when it has done so. The resulting protein emits yellow-green fluorescence to a degree dependent on the presence of glutamate, as described in the Introduction.

An artificial gene was made encoding these three proteins without interruption so that a single protein would be synthesized when the gene was expressed in the cell. An additional six amino acids (six histidines) were encoded at the beginning of the gene to facilitate purification of the protein in high amounts. This was accomplished by cloning into a special-purpose plasmid, pRSETB, used to express large amounts of protein in E. coli.

Fig. 3: Map of soluble glutamate sensor protein. Each colored box represents an amino acid sequence taken from an existing protein: ECFPenhanced cyan fluorescent protein, GltIglutamate-binding protein, and Citrinea type of yellow fluorescent protein. These three proteins, along with six copies of the amino acid histidine, were fused together to make a single protein.

The membrane form of the sensor was made in a similar fashion, except that two protein components were added to each end to direct the sensor to the outside of the cell, attached to the cell's membrane. In this way, the protein would sense glutamate levels outside but nearby the cell. This was necesary, because all proteins are made within cells, and if special modifications are not made, the sensor would report on glutamate inside its cell. But the glutamate concentration of interest is what lies outside the cell, near neurotransmitter receptors.

The sensor was modified in two ways to direct it to outside the cell. First, the three-protein fusion protein encoded by an artificial gene was preceded by several amino acids (taken from a rat immunoglobulin protein) that serves as a signal to begin transport of the protein out of the cell. Second, a protein segment that spans a membrane was added to the end of the fusion protein, so that the protein would not completely leave the cell but would be stuck to the membrane. The protein segment used was derived from Platelet-Derived Growth Factor Receptor (PDGFR), but the nature of the protein is not important.

All of this was accomplished by cloning into a special purpose plasmid, pDISPLAY, used to express proteins on the surface of mammalian cells.]

Fig. 4: Map of membrane-bound glutamate sensor protein. The colored boxes are as in Fig. 3, and in addition, these proteins are flanked by two protein segments (IG-K and PDGFR) that direct the fused protein to the cell membrane. These components were fused together to make a single protein.