Introduction to the FRET Test
A fluorophore such as green fluorescent protein (GFP) engages in fluorescence in the visual spectrum. This means it has the ability to absorb a short wavelength photon and emit multiple longer wavelength photons. For example, cyan fluorescent protein (CFP) has an excitation wavelength at 400 nm and emits photons to give a peak at 430 nm; however yellow fluorescent will not be excited at 400 nm and requires an excitation of a longer wavelength which is closer to the emission peak at 530 nm. When two distinct fluorophores are in close proximity (within 10 nm) a resonance energy transfer (FRET) can occur, resulting in a relative change in emission peak magnitudes between fluorophores. For example, excitation at 400 nm of a CFP-YFP pair would give a reduction in the CFP peak and an increase in the YFP peak amplitudes. Due to the strong distance dependence of FRET, an observed energy transfer event is a good indication for the successful association of two proteins.
Methods and Results
With the eventual aim of expressing both CIB1 and CRY2 within a single E. coli cell, we designed a construct for FRET containing both a fusion of CIB1 with CFP and a fusion of CRY2 with YFP. It was hypothesised that after growing the cells in the dark, a fluorescence spectrum with excitation at 400 nm would produce only a CFP peak (as no FRET would occur with the CRY2 and CIB1 domains separate). Then upon induction with blue light, the CRY2 and CIB1 domains would associate, producing a FRET and an observable shift in the peak magnitudes, thus confirming their successful association.
After transforming the construct to BL21 E. coli and culturing with IPTG induction, a very low expression level was observed as determined by the fluorescence of the peaks being comparable to background lysate. Excitation at 400 nm produced a small CFP and YFP peak, and blue light induction for 10 min showed no observable change in the spectrum. Unfortunately the individual fusion proteins could not be separated or purified due to the location of the purification tags.
It was then hypothesised that the expression of both fusions on the single plasmid was resulting in reduced expression levels due to the lengthy recombinant sequences not being supported by the E. coli system. From here we decided to redesign the constructs.
In order to conduct a well-defined FRET assay, we required the fusions to be expressed separately, allowing separate spectra to be acquired for the CFP and YFP as a reference. The purified proteins could then be combined in the dark and their spectra compared before and after blue light induction, allowing measurement of any FRET that may occur. Furthermore, appropriate purification tags were to be included to conduct the assay in the absence of any interfering fluorophores within E. coli. A recent paper discusses the importance of the C-terminus of CRY2 and the N-terminus of CIB1 being free for successful mediation of their associative interaction. Thus to prevent interference from the fused GFP, the fusion protein was designed to be fused to the other terminus of each protein. This gave the following final design for the two separate constructs.
In an attempt to obtain the CFP-CRY2 fusion protein, the transformed cells were then cultured in five different growth media, including autoinduction media and stress media containing ethanol. Negligible levels of protein were observed for all media with the exception of the TB2.0 autoinduction growth media. After cell lysis and centrifugation, strong fluorescence was observed in the pellet and faint fluorescence in the soluble fraction, suggesting that the CFP-CRY2 fusion has mainly formed inclusion bodies or is membrane bound. Purification of the soluble fraction via Nickel his-trap column gave ~1 mg of protein. This was used to conduct the FRET assay.
FAD was added to the reaction mix to enable photoreduction of CRY2. Upon assembling the separate proteins in the dark then in inducing with blue light for 5 min, the fluorescence spectrum at 430 nm excitation produced no measureable shifts between the ratios of CFP to YFP peaks meaning the association under blue light conditions of CRY2 and CIB1 cannot be confirmed. The concentration of FAD was increased sequentially and produced a quenching of the CFP peak, and an increase in a peak around 550 nm. The CFP quenching we significant and confirmed to not be as a result of a dilution effect (as confirmed by a negative control experiment). The increasing peak at 550 nm could be due to a resonance transfer from CFP to YFP, however the YFP fluorophore cannot be distinguished from that of FAD (which lies in the same region of the spectrum). From this experiment it can be inferred that although the observations are consistent with a FRET phenomenon occurring due to energy transfer from CFP to YFP, it may also be due to a FRET induced by the binding of FAD to CRY2. Future experiments could see the use of a different FRET pair, such as a red fluorescent protein. This would allow separation from the FAD chromophore and accurate measurements of fluorophore ratios.
From our efforts to express fluorescent fusion constructs of CRY2 and CIB1 for the aim of conducting a FRET assay, it was discovered that a CRY2 fusion protein has poor expressability in E. coli. Despite the poor expression of the CFP-CRY2 construct relative to the high expression of the CIB1-YFP fusion, the FRET test was performed successfully, and suggested the presence of an energy transfer event when compared to background dilution effects. However, due to the experimental design, we were unable to distinguish the origin of the FRET event as occurring between CFP to YFP or CFP to FAD.
