Y66W mutant generates Cyan Fluorescent Protein with lowered fluorescent intensity
Using site directed mutagenesis we produced an Y66W mutant of the Aequorea green- fluorescent protein (cyan fluorescent protein) so as to study its emissive spectra and fluorescence intensity and compare it to the wild type. The His-tagged mutant in pET28c was overproduced in E.coli BL21 (DE3) using auto-induction media, fractionated and purified by nickel chromatography and molecular mass was determined by mass spectroscopy. Fluorescence intensity was 32 times lower than the wild type indicating the crucial role of Tyrosine66 in chromophore structure and function. More mutations are required to make this mutant viable for widespread use.
Key words: GFP; CFP; Y66W; Site directed mutagenesis; fluorescence intensity
The Green Fluorescent Protein (GFP) has taken the scientific world by storm. The 238 amino acid molecule isolated from Aequorea Victoria has a chromophore made of para-hydroxybenzylidene-imidazolinone derived from Ser65, Tyr66 and Gly67 housed in a cylindrical 11 strand ß barrel. A sequence of cyclization, oxidation and dehydration events involving the Tyr66 results in chromophore maturation and excitation at 396 and 475nm leads to fluorescing of green light at 503-508 nm. GFP has a number of attractive features: expression in many organisms like E.coli , non-invasiveness and dynamic nature of its observation, inertness and diffusivity . It has been used extensively in Forster resonance energy transfer (FRET) and photo bleaching experiments and more recently used to map neural network activities in the human brain . To improve the existing properties of GFP and to introduce other useful traits, extensive mutational studies are carried out on the fluorophore. As a result of certain mutations by random or site directed mutagenesis , mutants are produced that could be well expressed in mammals, possess simpler excitation spectra, show marked improvement in fluorescing intensity, improved folding kinetics, are of various colours, more tolerant to pH conditions and that are not prone to dimerization .
GFPuv (GFP with mutations F99S, M153T, V163A; henceforth referred wild type "GFPwt") matures better at 37°C and fluoresces 45 times more intensely than the native GFP . The aim of the experiment is to purposefully induce a mutation by adopting a modified variation of Stratagene's QuikChange site directed mutagenesis and assess the impact on the covalently altered GFPwt fluorophore. The tyrosine at the 66th position on the GFPwt is mutated to tryptophan (Y66W) resulting in the formation of a Cyan Fluorescent Protein (CFP). Mass spectrometry confirms the presence of the CFP and subsequent fluorimetric analysis indicates that the mutation alters the excitation and emission spectra as well as shows reduced fluorescence intensity with respect to GFPwt. The study establishes that the Y66W is not a useful mutation by itself and further mutations must be done in order to compensate for the lost fluorescence intensity.
2. Materials and Methods
2.1 Site directed mutagenesis on GFPuv cloned into vector pET28c
The His-tagged GFPuv ("wild type"), is excised from pET23GFPuv (provided by Prof.McPherson, University of Leeds) using enzymes NdeI (20U/µL) and HindIII (10U/µL) and ligated into pET28c(+) (Novagen) under control of a T7 promoter. A modified protocol based on Stratagene's Quik-Change® is adopted where hot start KOD polymerase (Cat No 71086 Stratagene) is used instead of Pfu Turbo DNA polymerase for reasons of improved fidelity and processivity . A 50µL PCR mix comprising 20ng of GFPuv -Pet28c template DNA , 2 mM dNTPs , 10x PCR buffer , 25 mM MgSO4 , 1U hot start KOD polymerase , distilled water and 125ng each of the Forward primer (F; 5'-CACTTGTCACTACTTTCTCTTGGGGTGTTCAATGCTTTTCC-3') and Reverse primer (R;5'-GGAAAAGCATTGAACACCCCAAGAGAAAGTAGTGACAAGTG-3') was run at 940C for 30 sec for 1 cycle, 940C for 30sec, 550C for 1 min and 680C for 4 min 20 sec for 24 cycles and at 680C for 10 min for 1cycle.
The PCR products are digested with Dpn1 (10U) and incubated at 37°C for 1 hr and E.coli XL 1 blue supercompetent cells (Stratagene) is transformed with digest products by heat shocking. After incubation in NZY+ broth for 1 hr at 37°C, it is plated on LB-Kanamycin plates and incubated overnight at 37°C.
2.2Transformation of site-directed products in pET28c to E.coli BL21 (DE3) cells
The BL21 (DE3) cells (Cat No 69450 Novagen) are cultured in SOB media (1:40) at 37°C till an OD600 value of 0.3 is reached. The cells are made competent (refer manufacturer's instructions) and the Y66W mutant DNA (purified by QIAprep Miniprep, Qiagen) is transformed into the cells by heat shock. Plating of transformed cells and competent cells (control) on LB-Kanamycin plates is done and incubated overnight at 37°C.
2.3Expression of Y66W GFP (CFP) from E.coli cells
The transformed BL21 (DE3) cells are grown in LB-1D media (constituting non-induced control NIC) and SB-5052 (Auto-Induction media) is added and is incubated (shaking at 400rpm) for 20 hrs at 28°C and then cooled (Total induced control TIC). TIC and NIC are pelleted and resuspended in SDS-PAGE buffer and placed on ice. The cooled culture is fractionated using BUGBUSTER® (Novagen) and soluble samples and insoluble samples (pellet resuspended in binding buffer pH 7.9) are collected. All samples are run on a polyacrylamide gel at 150V for 1 hr followed by staining by Coomassie Brilliant dye. To confirm the presence of CFP, Western blotting of the fractionated samples is done and the blot is probed with Hisprobe®-HRP (Novagen) which binds to His-tagged CFP. After washing the blot with TBST, it is analyzed by a chemiluminescent kit comprising Peroxide buffer (1M Tris/HCl pH 8.5, 30% H2O2 and distilled water) and Luminol enhancer (1M Tris/HCl pH 8.5, 475mM Luminol, and 90mM coumaric acid, distilled water) and the film is developed in an X-O graph machine.
2.4 Purification and analysis of CFP
Histidine bind resin(Qiagen) is equilibrated by washes with deionised water, with 1x charge buffer (50mM NiSo4) and with 1x binding buffer (0.5M NaCl, 20mM Tris HCl, 5mM imidazole pH 7.9). 10µL of the soluble fraction is saved and labeled Total soluble fraction, while the rest is spun down with binding buffer (labeled unbound sample), wash buffer (0.5mM Nacl,60mM imidazole, 20mM Tris HCl pH 7.9, labeled wash sample) and elution buffer (1M imidazole, 0.5M NaCl 20mM Tris HCl pH 7.9, labeled Elution samples). The samples are run on a gel and Coomassie stained. Fluorimetric analysis is done with Flwinlab software (Perkin Elmer, MA, USA) with the parameters for wild type GFP (Excitation 450nm, Emission 490-550nm, Slit Widths 4 and 4, Accumulation 5) and for CFP (Excitation 440nm, Emission 460-550nm, Slit Widths and Accumulation same as Wild type). Mass spectrometry of first elution (treated with methanol and 1% formic acid, exposed to 3.5kV with nitrogen as nebulising and drying gas and calibrated with horse heart myoglobin) is also done to confirm presence of CFP.
3.1 Quik-Change™ Site Directed Mutagenesis (Stratagene), sequence verification and alignment
A modified protocol was employed for site directed mutagenesis at the Tyr66 using PCR and the plasmid DNA extracted by QIAprep kit (Qiagen) was diluted to 50ng/mL and sequenced. The sequence (Group 20, ) was queried with the wild type GFP sequence using online sequence alignment tool CLUSTALW2. The translated wild type and sequence data (using Translate tool at the ExPASy proteomics server) was also queried. Results proved that the Y66W mutagenesis was successfully carried out and that the DNA obtained was indeed CFP (Fig.1. ).
Fig. 1. Sequence alignment of wild type and mutant DNA (A) and that of translated amino acid sequence (B) using CLUSTALW2 indicates the Y66W mutation (in red).
3.2 Expression of mutant proteins from E.coli BL21 (DE3) cells using SB-5052 auto-induction media
The BL21 (DE3) cells fractionated by BUGBUSTER™ (Novagen) and DNase 1 were spun down at 13,000rpm at 4°C for 20 minutes (soluble sample) and the resuspended pellet constituted the insoluble sample. SDS-PAGE on a 12% polyacrylamide gel (Fig.2. ) shows the presence of the putative CFP protein around the 27kDa mark. The non induced control has no putative CFP band, while the total induced control shows a thick band. This indicates that the controls are correct and that the auto-induction was successful. The putative CFP band is seen strongly in the soluble fraction indicating good expression of the protein. Due to overexpression, a small band is seen in the insoluble fraction as well.
3.3 Confirming presence of CFP, purification and concentration estimation
A Western Blot of the above samples probed with His-Probe (Novagen) reveals the presence of a band at the 27kDa mark in the induced, soluble and a faint band in the insoluble sample. As the CFP is his-tagged, its presence is confirmed (Fig.3A). Elution samples purified by Ni-NTA chromatography were assayed by Bradford assay (not shown) and a high concentration (2.5mg/mL) was found in the first elution and also reflected on a SDS gel after staining (Fig 3B).
3.4 Analysis of Wild type GFP and CFP by Mass spectroscopy and fluorimetry.
The purified GFP elution sample and CFP analyzed by mass spectroscopy (Provided by Dr. Alison Ashcroft, Leeds,) revealed a molecular mass of 29,583 Da and 29,597 Da respectively (Fig. 4. ).
As fluorescence depends on the concentration of the sample, 1µg/mL of the wild type GFP and CFP were used and were excited at 450nm and 440nm respectively. The fluorescence values contributed by the imidazole buffer due to excitation at 440 and 450nm was accounted for. The wild type's maximum emission peaked at 504nm and the CFP's emission peaked at 481nm. There was a 32 fold difference between peak emission fluorescence of wild type as compared to CFP (Fig.5. ). Data obtained from .
The study generated a mutant CFP protein by site directed mutagenesis, replacing the tyrosine residue at the 66th position with tryptophan. The CFP was confirmed using DNA sequencing, purified by nickel chromatography and its molecular weight verified by mass spectroscopy. Fluorimetry data indicated that the excitation and emission spectra of CFP was 440nm and 460-550nm respectively, with an emission peak at 481nm. Wild type GFP was excited at 450nm and exhibited an emission peak at 504nm.
The GFPs are classified into seven different classes based on the chromophore component that constitutes it. The wild type GFP has a mixture of neutral phenol and anionic phenolates which causes its characteristic double humped excitation spectra and also explains why the emission maxima is dependent on the excitation wavelength. The site-directed mutation Y66W replaces the phenolate or the neutral phenol species with a bulky indole group and the resulting chromophore causes decreased fluorescence as compared to the wild type .
The mutated sequence is compared to the wild type using CLUSTALW2 and the results indicate that the Y66W mutation is successful. There is a mutation at position 231 of the nucleotide sequence but it is a silent mutation due to codon degeneracy and does not alter the amino acid sequence.
The wild type is his-tagged and so, is the resulting CFP. A Western Blot probed with His-Probe confirms CFP's presence on the blot. The blot also indicates the molecular mass of the CFP as approximately 27kDa in accordance with Kajihara . An SDS-PAGE of the components derived from nickel chromatography indicates that the CFP is found primarily in the eluted sample and due to overexpression, there is still a some protein left in the second elute. This is as a result of enhanced stable chromophore maturation at 37°C .
The elution samples analyzed by mass spectroscopy reveal the mass of wild type GFP and CFP around 29 kDa as opposed to the literature value of 27 kDa. The wild type GFP sequence from pET23 provided contains an additional 10 amino acids introduced at the C terminus during PCR amplification when purchased initially from Clontech. This coupled with the His-tag sequence MGSSHHHHHHSSGLVPRGSH could account for the difference of 2 kDa observed.
Fluorimetry analysis shows that the wild type GFP excited at 450nm has a peak emission of 504 nm in exact accordance to published values . The CFP excited at 440nm emits at 481nm in close following with Heim et al's value of 485nm . The fluorescence is normalized by arbitrarily assigning the maximum fluorescence of the respective GFP species as one. This allows a better visualization of the emissive spectral characteristics. The CFP shows the characteristic double humped spectral curve with the shoulder at 508nm instead of at 500nm. The double hump unique to the CFPs is caused by two distinct conformations formed by the residues Tyr145 and His148 . The fluorescent readings indicate that the wild type is 32 fold more fluorescing than the CFP, but there isn't literature that provides relative fluorescence intensities of CFP to wild type at 37°C . But there is sufficient evidence to prove that mutants that are substituted with aromatic sidechains at the Tyr66 residue are considerably less fluorescent compared to the wild type .
Mutations like N146I, M153T, V163A and H148D result in enhanced CFP called Cerulean that is used along with yellow fluorescent proteins (YFP) in FRET studies .The lowered fluorescence coupled with lack of sufficient literature on the CFP Y66W mutant signifies that it is a mutant that cannot be implemented practically and that it requires further mutations before it could be used.
Acknowledgments: We thank Dr. Aysha Divan, University of Leeds for her unwavering help and support which made this study possible. Thanks are extended to Dr. Chi Trinh and Ms. Sophie for their help as lab demonstrators and also to Prof.McPherson, University of Leeds for providing the initial pET23GFP construct and Ms. Alison Ashcroft for the mass spectrometry data.