Introduction

Luminescent transition metal complexes constitute an important class of compounds that have found increasing applications in inorganic photochemistry1,2,3,4,5,6,7,8,9, phosphorescent materials for optoelectronics10,11 and luminescent sensing applications2,3,12,13,14,15,16. In particular, ruthenium(II)-diimine complexes such as [Ru(bpy)2(dppz)]2+ (bpy = 2,2′-bipyridine; dppz = dihydro[3,2-a:2′,3′-c]phenazine), have received particular attention due to their “molecular light switch effect” when bound to DNA17,18,19,20. Over the last two decades, the topic of luminescent transition metal complex–nucleic acid interactions has been firmly established as an important field of research21,22,23,24,25,26,27,28,29,30,31,32.

By comparison, reports on the development of luminescent transition metal complexes for protein detection and protein staining have been considerably scarcer. SYPRO Ruby dye, a ruthenium(II) complex of undisclosed formulation, is commercially available as a reagent for protein gel staining33,34. However, its unknown composition hinders investigation and optimization of the staining protocol. We have previously developed cyclometallated iridium(III) complexes bearing 2-phenylpyridine (ppy) ligands for protein detection and staining35. As part of our continuing efforts, cyclometallated platinum(II) complexes with 2-phenyl-1,10-phenanthroline ligands have also been reported for protein gel staining and cellular imaging36. On the other hand, we have recently reported the photophysics of a series of ruthenium(II) and osmium(II) complexes with the general formula [MII(N∧N)(X)3(L)]n+ where (X)3 are facial or meridional tridentate ligands and N∧N are bpy-like aromatic diimines8. Envisioning that the [MII(N∧N)(X)3]n+ core could serve as a luminescent switch-on/off probe when it binds with biomolecules, we designed a water-soluble ruthenium(II) complex based on the [RuII(N∧N)(X)3]n+ core as a reagent for biomolecule detection and protein staining. We report herein a new ruthenium(II) complex bearing 1,4,7-trithiacyclononane ([9]aneS3) and 4,4′-dimethoxycarbonyl-2,2′-bipyridine (dcmb) (complex 1′, Fig. 1). Complex 1′ is highly selective for histidine over other natural amino acids and also functions as a switch-on luminescent probe for protein in solution and for protein staining in a polyacrylamide gel.

Figure 1
figure 1

Chemical structure of ruthenium(II) complexes 1′, formed from the in situ hydrolysis of complex 1.

Full size image

Results

Synthesis

Complex 1 was synthesized by refluxing [Ru([9]aneS3)(dmso)Cl2] with dcmb in methanol under argon for 90 min; recrystallization by slow diffusion of diethyl ether into an acetonitrile solution of 1 gave the acetonitrile-ligated complex [Ru(dcmb)([9]aneS3)(CH3CN)]2+ (1′) in 72% yield. The structures of the complexes were confirmed by 1H and 13C NMR spectroscopy and mass spectrometry. These complexes were sufficiently stable to be handled in air under ambient conditions in solution and solid forms. For example, complex 1′ was stable in aqueous solution for 72 h at room temperature, as revealed by UV-Vis spectroscopy. The absorption spectrum of 1′ in acetonitrile displays two absorption peaks [λ/nm (εmax/dm3 mol−1 cm−1)]: 367 nm (6870) and 493 nm (7120) (Fig. S2†). Excitation of 1′ (31 μM) at 450 nm in acetonitrile resulted in an emission with λmax at 692 nm, a quantum yield of 2.334 × 10−3 and a lifetime of 0.272 μs. This emission was assigned as the triplet dπ(RuII) → π*(N∧N) charge-transfer in nature based on comparison with analogous complexes reported previously7,37,38,39,40,41,42,43,44,45,46,47. The replacement of the bpy ligand with dcmb has been previously reported to red-shift the emission maximum of ruthenium(II) complexes48. The long-wavelength emission of 1′ is especially attractive for luminescent sensing applications, as interference from endogenous fluorophores (autofluorescence) can be reduced.

Binding analysis

We first examined the binding of complex 1′ to the amino acid histidine. Encouragingly, we found that the emission intensity of complex 1′ in 5% CH3CN/95% H2O solution at λmax = 630 nm was enhanced by up to 6.9-fold at [His]/[1′] = 48 (Fig. 2). As a control, we tested the response of complex 1′ to the amino acid glycine. The addition of glycine gave less than 0.2-fold increase in the emission intensity of 1′ even up to [Gly]/[1′] = 78 (Fig. S4†). The addition of groups of all the amino acids (each at 2.4 mM or ca. 50-times excess of the metal complex) also did not give a significant luminescent response (Fig. S5†). To investigate the binding property of complex 1′, electrospray ionization positive-ion mass spectrometry (ESI-MS) was adopted for the binding between complex 1′ and histidine. No covalent attachment peak was observed for complex 1′ upon incubation with histidine for 5 h at 20°C (Fig. S6†). Compared with the complex [Ir(ppy)2(solv)2]+ (where solv = H2O or CH3CN) we previously reported as a covalently binder to histidine, there was an additional peak center at m/z 656.1 which corresponds to the covalent attachment23 indicating that complex 1′ may not covalently bind to histidine. These results demonstrate that complex 1′ might have another interaction mode and high selectivity towards histidine over other natural amino acids, as only the addition of histidine could give a significant luminescent response.

Figure 2
figure 2

Phosphorescence emission spectra of 1′ (50 μM) in 5% CH3CN/95% H2O with increasing concentration of [His]/[1′] (0–48) at 20°C. Inset: phosphorescence emission intensity at 630 nm vs. His concentration.

Full size image

Luminescence response

To study the luminescence response of 1′ towards proteins, we chose the common protein standard bovine albumin serum (BSA) as the test analyte. We observed that the emission intensity of 1′ at λmax = 630 nm was greatly enhanced upon addition of BSA (Fig. 3). An 18-fold increase in the emission intensity of 1′ was registered at [BSA]/[1′] = 1.2. We hypothesize that the binding of 1′ to the histidine residues of the protein protects the aromatic diimine moiety from the aqueous environment, thereby suppressing non-radiative decay of the excited state and promoting 3CT emission.

Figure 3
figure 3

Phosphorescence emission spectra of 1′ (50 μM) in H2O with increasing concentration of [BSA]/[1′] (0–1.2) at 20°C. Inset: phosphorescence emission intensity at 630 nm vs. BSA concentration.

Full size image

Absorption titration

An absorption titration experiment was performed to further investigate the binding of 1′ to BSA. Isosbestic points were observed at 437 nm and 513 nm (Fig. S7†). Using the Scatchard equation, the binding constant K at 20 °C was determined to be 1.70 × 105 mol−1 dm3D (Fig. S7 Inset†)49.

Gel electrophoresis

Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS) is one of the most important techniques in biochemistry and molecular biology to detect and quantitate protein levels. Conventional protein-staining methods include colloidal-silver staining, Coomassie Brilliant Blue staining and Ponceau S staining. However, most involve time-consuming procedures and multiple reagents. The popular and commercially available Coomassie Brilliant Blue (CBB)50 stain requires a long destaining time for optimal performance. We were thus interested to see if we could apply ruthenium(II) complex 1′ to the staining of protein bands in an SDS-PAGE gel.

Fig. 4 shows an emissive image of a gel containing BSA after staining with 1′ (2.6 mg/20 mL) for 30 min. The lowest quantity of the protein mixture detected after staining with 1′ was 0.625 μg of protein (Fig. 4, left). The sensitivity of this system is at least comparable to Coomassie Brilliant Blue staining (Fig. 4, right). Note that in Fig. 4, complex 1′ was applied to the gel for only 30 min and required no destaining step, whereas Coomassie Brilliant Blue was applied for 60 min and was destained over a period of one to two days. This result demonstrates the simplicity and convenience of this protein staining protein utilizing ruthenium(II) complex 1′. Whereas the SYPRO Ruby dye staining solution contains 7% acetic acid, which is a mild irritant, the method described here utilises only 5% acetonitrile, a relatively more benign reagent. We also observed that an increase in the staining time enhanced the sensitivity of the system (Fig. S8).

Figure 4
figure 4

Left: Emissive SDS-PAGE analysis of BSA with complex 1′ (2.6 mg/20 mL) as the detecting agent (staining time: 30 min); Right: Control SDS-PAGE analysis of BSA with Comassie Brilliant Blue staining (staining: 60 min; destaining: 24–48 h).

Full size image

Complex 1′ is postulated to bind histidine residues. Consequently, 1′ should be able to readily detect histidine-rich or histidine-tagged proteins in a protein gel (Fig. 5). Gratifyingly, we observed that 1′ was able to detect 6×His-tagged NF-κB p50 within 30 min, whereas the sensitivity to NF-κB p50 without histidine-tag is relatively reduced (Fig. S9). This property may be useful for the analysis of His-tagged proteins during purification, for example during nickel-column chromatography.

Figure 5
figure 5

Left: Emissive SDS-PAGE analysis of 6×His-tagged NF-κB p50 with complex 1′ (2.3 mg/20 mL) as the detecting agent (staining time: 30 min); Right: Control SDS-PAGE analysis of BSA with Comassie Brilliant Blue staining (staining: 60 min; destaining: 24–48 h).

Full size image

Discussion

In conclusion, we have developed a new ruthenium(II) complex supported by trithiacyclononane and an aromatic diimine as a versatile luminescent switch-on probe for amino acid and protein detection and for protein staining. This complex can be used to selectively sense histidine over all other natural amino acids and is postulated to bind protein through coordination to histidine residues. There are a wide range of alternate stains such as the traditional methods: Coomassie Brilliant Blue, colloidal-silver stain and SYPRO. However, the Coomassie Brilliant Blue requires a long destaining time for optimal performance, colloidal-silver stain involves multiple steps and reagents, as a result the process is relatively time consuming. The untreated silver stain sample is incompatible for mass spectrometry whereas the undisclosed SYPRO Ruby dye limits the investigation and optimization of the staining protocol.

Recently, some metal complex protein staining dyes have been reported via covalent and non-covalent binding which were reviewed by Lo and co-workers51. Generally, luminescent transition metal complexes derivatized with the reactive functional groups have been designed as covalent protein labels. Whereas, the noncovalent dyes generally equip with hydrophobic ligands for non-specific binding or append with a biological substrate that shows specific binding to its receptor protein.

In this study, complex 1′ shows the selectivity towards histidine. However, unlike our previous report iridium(III) complex shows that it covalently and selectively binds to histidine, complex 1′ might display another type of interaction mode since the ESI-MS doesn't show the covalent attachment peak upon addition of histidine. Therefore, it indicates that the interaction mode may not be covalent binding and it is rare to see the noncovalent protein staining dye display the selectivity. We have applied this complex to SDS-PAGE protein gel staining to detect BSA and His-tagged proteins. Importantly, staining can be complete in less than 30 min and no destaining step is required, with a sensitivity comparable to the highly popular, but time-consuming Coomassie Brilliant Blue dye. This complex also possess a favourable emission maximum approaching the red region, which is attractive for sensing applications as the effects of autofluorescence from endogenous fluorophores can be avoided. Potentially, these complexes could be developed into probes for protein levels in cellular imaging. As the [dπ(RuII) + L] → π*(N∧N) 3CT emission of these ruthenium(II) complexes is sensitive to the auxiliary ligands, we envisage that further optimization of the emission maximum, sensitivity and selectivity of the system is possible.

Methods

General Procedures

All chemicals and solvents (AR grade) for syntheses were used as received. Bovine serum albumin (BSA, product no. A8531) was purchased from Sigma Chemical Co. Ltd. and used without further purification. Broad-range molecular weight protein standards (Catalog no. 161-0317) and all chemicals for SDS-PAGE were purchased from Bio-Rad. UV-Visible spectra were recorded on Cary UV-300 (for UV-Visible absorption titration) and a Shimadzu UV-1700 spectrophotometer. Emission spectra were recorded on PTI QM-4/2005 Spectrophotometer. 1H NMR spectra were recorded on Bruker 400 DRX FT-NMR spectrometers. Peak positions were calibrated with solvent residue peaks as internal standard. Electrospray mass spectrometry was performed on a PE-SCIEX API 3000 triple quadrupole mass spectrometer.

Synthesis. [([9]aneS3(dcmb)Ru(Cl)](PF6), complex 1

[Ru([9]aneS3)(dmso)Cl2] (0.1 g, 0.2 mmol) was added to a solution of 4,4′-dimethoxycarbonyl-2,2′-bipyridine dcmb (0.4 mmol) in methanol (20 mL) and the reaction mixture was refluxed under argon for 90 min. Upon cooling to room temperature, the resultant mixture was filtered and the filtrate was added to a saturated NH4PF6 solution to afford a red precipitate. The precipitate was washed with diethyl ether and dried under vacuum. The solid was recrystallized by slow diffusion of Et2O into an acetonitrile solution to give deep red crystals. Yield: 0.098 g, 72%. 1H NMR (400 MHz, CD3CN): δ 2.45–2.78, 2.84–2.95, 2.98–3.10 (m, 12H, [9]aneS3); 4.03 (s, 6H, COOCH3), 8.02 (dd, 2H, J = 5.6, 1.6 Hz, dcmb); 8.94 (s, 2H, dcmb); 9.22 (d, 2H, J = 6 Hz, dcmb). ESI-MS: m/z 589 [M+].

Absorption Titration

UV-Visible spectra were recorded on Cary UV300. A solution of the 1′ (23 μM) was prepared in H2O. Aliquots of a millimolar BSA stock solution were added. Absorption spectra were recorded in the 320–600 nm range, after equilibrium at 25°C for 20 min. The intrinsic binding constant K was determined from the plot if D/Δεap vs D:

where D is the concentration of BSA, Δεap = |εa-εF| and Δε = |εB-εF|. The apparent extinction coefficient, εa, is obtained by calculation Aobs/[metal complex]. εB and εF correspond to the extinction coefficient of BSA-metal complex adduct and the extinction coefficient of free metal complex, respectively.

Emission Titration

Emission measurements were carried out by using PTI QM-4/2005 Spectrophotometer. All measurements were made with 15 nm entrance and exit slit. A stock solution of the Ru(II) complex (50 μM) was prepared in acetonitrile. Aliquots of a millimolar BSA stock solution or various natural amino acids were added. Emission spectra were recorded in the 500–850 nm range, after equilibration at 25°C for 20 min.

Sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE)

SDS-PAGE analysis was conducted according to Bio-Rad Mini-Protean® 3 cell Instruction Manual.

Protein detection

After electrophoresis, the gels were stained with 1′ at room temperature. 1′ (2–3 mg) was first dissolved in 1 mL acetonitrile followed by addition of H2O (19 mL). Approximately 20 mL of staining solution was used for a typical mini-gel (8 cm × 7 cm × 1.5 mm). The gel was placed into water for 10 min and then transferred into the staining solution. The container was covered with aluminum foil to protect the dye from light. The gel was gently agitated for 30 min at room temperature using an orbital shaker (50 RPM). After staining, the gel was viewed using a 300 nm UV transilluminator directly. Detection and imaging of the gels were conducted using AlphaImager™ 2200 imaging system. For the control gel, Coomassie Brilliant Blue stain was used to visualize the protein sample. The gels were stained for 1 h followed by destaining in deionised H2O for 24–48 h to reduce the background staining.