Determining the Fluorescence Spectrum of a Protein

One of the main values of fluorescence as a technique for probing protein conformation is that it is highly sensitive and very economical of material. This means, however, that small traces of fluorescent impurities in the solvent or on the cell are readily detected and, if care is not taken, can lead to misinterpretation of the spectra. The essential aim in using this technique, therefore, must be to obtain a fluorescence emission spectrum for a protein that is guaranteed free from all too easily generated artifacts. 

The following points are essential to obtaining reproducible spectra for reliable comparison between samples. 

Buffers should have low absorbance and low fluorescence in the regions of excitation and emission. Absorbance will usually be expected to be 0.1 and fluorescence close to zero. This will normally be the case for standard buffers made from analytical-grade reagents or materials of equivalent purity, but they should nevertheless be checked routinely for fluorescence. If a fluorescent component is added to the solution—e.g., as a ligand—it should be checked that the observed fluorescence arises solely from that component. The actual buffer solution used to dissolve or to dialyze the protein should be used for the fluorescence blank. Plastic containers (and stirring bars) may contribute fluorescent agents if these are used, appropriate blanks should be carefully monitored for fluorescence. 

It is essential to monitor clarification if artifacts are to be avoided. There are three indicators of the completeness of clarification, each available from the standard procedures that should be adopted for spectroscopic characterization. 

Determination of protein concentration (unitbu) requires an absorbance spectrum to be recorded on a good quality spectrophotometer from 240 to 350 nm. Aromatic amino acid residues do not absorb above 320 nm, so the spectrum between 320 and 350 nm should be only marginally above baseline. The presence of turbidity will result in finite attenuance in this region that increases toward lower wavelengths. 

---

B3.6 Determining the Fluorescence Spectrum of a Protein B3.6.1 Strategic Planning B3.6.1 Basic Protocol l Recording a Fluorescence Emission Spectrum B3.6.5 Basic Protocol 2 Determination of Fluorescence Quenching B3.6.9 Support Protocol Basic Theory and Interpretation of Fluorescence Spectra B3.6.12 Commentary B3.6.19... 

Native fluorescence of a protein is due largely to the presence of the aromatic amino acids tryptophan and tyrosine. Tryptophan has an excitation maximum at 280 nm and emits at 340 to 350 nm. The amino acid composition of the target protein is one factor that determines if the direct measurement of a protein s native fluorescence is feasible. Another consideration is the protein s conformation, which directly affects its fluorescence spectrum. As the protein changes conformation, the emission maximum shifts to another wavelength. Thus, native fluorescence may be used to monitor protein unfolding or interactions. The conformation-dependent nature of native fluorescence results in measurements specific for the protein in a buffer system or pH. Consequently, protein denatur-ation may be used to generate more reproducible fluorescence measurements. 

Figure 10.6 shows fluorescence emission spectra of lens culinaris agglutinin (LCA) (a) (Lmax = 330 nm), of inaccessible Trp residues (b) (A.max = 324 nm) obtained by extrapolating to [I-] = oo, and of quenched Trp residues (c) obtained by subtracting spectrum (b) from spectrum (a). The emission maximum of accessible Trp residues is located at 345 nm, a characteristic of emission from Trp residues near the protein surface. Thus, both classes of Trp residues contribute to the fluorescence spectrum of LCA (Albani 1996). The presence of five Trp residues makes the analysis by the modified Stern-Volmer equation very approximate nevertheless, a selective quenching method allows the percentage of accessible fluorophores to the quencher to be determined. 

The constant Rq is dependent on several parameters 1) the relative orientation of the transition dipole moments of the two molecules (these dipoles are the spectroscopic transition dipoles), 2) the extent that the fluorescence spectrum of the donor overlaps with the absorption spectrum of the acceptor, and 3) the surrounding index of refraction. We will deal with each of these below (see Equation 8). Because many proteins have diameters less than lOnm, this distance dependence explains the usefulness of ERET for determiiung distances inside proteins as well as between interacting proteins, which is the reason that the name spectroscopic ruler was coined for FRET (20). ERET is a convenient method for determining the distance between two locations on proteins, or for determining whether two proteins interact intimately with each other. Fluorescence instrumentation is available in many laboratories, and a plethora of dyes and a wide variety of fluorescent proteins are now readily available. Therefore, FRET is a viable option for most researchers. With care, FRET can yield valuable information concerning protein-protein interactions, interactions of proteins with other molecules, and protein conformational changes. 

A very common method to put into evidence the fluorescence spectrum of each component is the decay associated spectra. This method allows combining the dynamic time-resolved fluorescence data with the steady-state emission spectrum. The fluorescence decays are globally or individually fitted to the n-exponential function (n being the number of species, tryptophan or tyrosine residues for example), and the decay associated spectra are constructed (Krishna and Periasamy, 1997). Figure 8.27 displays the decay associated spectra of Trp residues of ai-acid glycoprotein performed by Hof et al (1996). The authors found four fluorescence lifetimes instead of the three we obtained. They attributed the peak of 337 nm to Trp-122 with a location between the surface of the protein and its core. The determination of the degree of hydrophobicity around the Trp residues showed that Trp-25 residue is located in a hydrophobic environment and Trp-160 is the most exposed to the solvent. 

The fluorescence intensity of fluorescent proteins is pH dependent and most fluorescent proteins are less fluorescent at lower pH mainly because of a reduction in absorbance. Since the absorbance of the acceptor determines the FRET efficiency, changes in the acceptor absorbance spectrum due to pH variations can be wrongly interpreted as changes in FRET efficiency. Thus, a pKa well below physiological pH is recommended to prevent artifacts due to pH changes inside cells. This is especially challenging if the fluorescent proteins are to be targeted to acid cellular compartments, for example, endosomes, lysosomes, or plant vacuoles. 

The spectroscopic and photochemical properties of the synthetic carotenoid, locked-15,15 -cA-spheroidene, were studied by absorption, fluorescence, CD, fast transient absorption and EPR spectroscopies in solution and after incorporation into the RC of Rb. sphaeroides R-26.1. High performance liquid chromatography (HPLC) purification of the synthetic molecule reveal the presence of several Ai-cis geometric isomers in addition to the mono-c/x isomer of locked-15,15 -c/x-spheroidene. In solution, the absorption spectrum of the purified mono-cA sample was red-shifted and showed a large c/x-peak at 351 nm compared to unlocked all-spheroidene. Spectroscopic studies of the purified locked-15,15 -mono-c/x molecule in solution revealed a more stable manifold of excited states compared to the unlocked spheroidene. Molecular modeling and semi-empirical calculations revealed that geometric isomerization and structural factors affect the room temperature spectra. RCs of Rb. sphaeroides R-26.1 in which the locked-15,15 -c/x-spheroidene was incorporated showed no difference in either the spectroscopic properties or photochemistry compared to RCs in which unlocked spheroidene was incorporated or to Rb. sphaeroides wild type strain 2.4.1 RCs which naturally contain spheroidene. The data indicate that the natural selection of a c/x-isomer of spheroidene for incorporation into native RCs of Rb. sphaeroides wild type strain 2.4.1 was probably more determined by the structure or assembly of the RC protein than by any special quality of the c/x-isomer of the carotenoid that would affect its ability to accept triplet energy from the primary donor or to carry out photoprotection. 

In general, when one wants to determine if global and/or local structural modifications have occurred within a protein, circular dichroism experiments are performed. Also, one can record the fluorescence excitation spectrum of the protein. If perturbations occur within the protein, one should observe excitation spectra that differ from one state to another. One should not forget to correct the recorded spectra for the inner filter effect. 

---