Multimodal cell imaging by ruthenium polypyridyl labelled cell penetrating peptides

Abstract

The capacity of ruthenium polypyridyl complexes as probes for combined confocal luminescence and resonance Raman imaging, enabled by their large Stokes shift, is demonstrated for a novel membrane sensitive Ru(II) polypyridyl peptide. Confocal luminescence and resonance Raman imaging provides complementary information about the membrane phospholipid regions of the cell and the location of the dye within the cell.

---

Advances in fluorescence microscopy, in particular scanning confocal imaging technologies, have revolutionised the study of live cells. Similar advances in Raman microscopy have emerged in recent years, and scanning confocal Raman microscopes are now commercially available, opening the possibility of Raman imaging of cells. Label free Raman mapping of cells has been the focus of a number of studies as it holds the potential for mapping concentrations of key biochemicals within the cell.

However, it is limited by the low sensitivity of conventional Raman which requires long acquisition times; the resulting dwell times of the laser on a single spot is potentially damaging.¹ In addition, in the complex matrix of a cell, it inevitably suffers from background interference. However, if as in fluorescence microscopy, exogenous labels can be incorporated into the cell which can be excited under resonance absorption conditions, the dye’s distribution and, unique to Raman, its structure, can be probed with excellent sensitivity. This may be particularly useful if that structure exhibits environmental sensitivity.

A particularly useful paradigm arises if a single label can be used for both luminescence and resonance Raman imaging of cells. Such multimodal imaging is impossible with the fluorescent organic dyes commonly exploited for live cell imaging since their small Stokes shift means fluorescence overwhelms the weaker Raman response. However, as we demonstrate here, ruthenium polypyridyl complexes which typically exhibit up to 200 nm separation between their absorbance and emission maxima can permit the collection of intense interference free resonance Raman spectra at a wavelength that also excites emission.³ Another key advantage of Ru(II) complexes is the capacity to make their photophysical properties dependent on their environment. For example, the sensitivity of the excited state lifetime and hence the phosphorescent intensity to O₂ is well known for many ruthenium complexes.⁴ In addition, exploitation of ligands such as dppz (dipyrido-[3,2-a:2′,3′-c]-phenazine) induces sensitivity in these parameters to the aqueous environment.^5,6

Examples of ruthenium polypyridyl complexes applied to cell imaging are increasing,² although a major barrier to this application is that they do not generally cross the cell membrane spontaneously. However, we recently described how conjugation of a cell-penetrating polyarginine peptide to a ruthenium polypyridyl complex allows ready passage of the metal complex across the cell membrane.^6,7

In this contribution we illustrate how a peptide linked ruthenium complex can be exploited in multimodal imaging, by demonstrating the application of a novel environmentally sensitive ruthenium polypyridyl peptide conjugate, [Ru(dppz)₂PIC-Arg₈]¹⁰⁺, Scheme 1, as a membrane probe in live SP2 myeloma cells; dppz is dipyrido-[3,2-a:2′,3′-c]-phenazine, PIC is 2-(4-carboxyphenyl)imidazo[4,5-f][1,10]phenanthroline and Arg is arginine. The complex was synthesized from [Ru(dppz)₂Cl₂] and the parent ligand according to commonly reported methods and the peptide modified at its N-terminus with a 6-carbon spacer (6-amino-hexanoic acid) was prepared by Merrifield’s solid phase peptide synthesis.¹¹ The parent complex and peptide conjugates were purified by HPLC and structure was confirmed by NMR and MALDI-TOF mass spectrometry.

Scheme 1 Structure of [Ru(dppz)₂PIC-Arg₈]¹⁰⁺. In the parent complex, the aryl amide on the imidazole ring is replaced with an aryl acid.

The parent complex, [Ru(dppz)₂PIC]²⁺, exhibits a metal-to-ligand charge transfer (MLCT) absorbance centred at approximately 450 nm, and in non-aqueous media, an intense emission centred at 610 nm which is readily excited at 458 nm by an argon ion laser.¹¹

In degassed 9/1 acetonitrile–DMSO the complex exhibits a luminescence lifetime of 0.68 ± 0.017 μs. Following conjugation of the polyarginine peptide to [Ru(dppz)₂PIC]²⁺, no significant change was observed in the absorption or emission λ_max.¹¹ However, the lifetime of the peptide conjugate in degassed 9/1 acetonitrile–DMSO increased to 0.78 ± 0.022 μs. The enhanced luminescence of the conjugate was also reflected in the quantum yields in aerated 9 : 1 acetonitrile–DMSO which were 0.025 and 0.048 for [Ru(dppz)₂PIC]²⁺ and [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ respectively.⁸ The origin of the enhanced luminescence may arise from environmental protection conferred by the large peptide chains on the dppz ligands.

Fig. 1(ii) shows the resonance Raman spectra of [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ and [Ru(dppz)₂PIC]²⁺ in buffer excited at 458 nm. Comparison with [Ru(dppz)₃]²⁺ and similar complexes⁹ confirms the majority of the modes can, in each case, be assigned to the dppz ligand. This is consistent with excitation resonant with a Ru(dπ) to dppz π* MLCT transition. The spectra are similar for dye and conjugate except that bands at ∼1357 cm⁻¹ and ∼1390 cm⁻¹ are shifted to higher energy (∼20 cm⁻¹) for [Ru(dppz)₂PIC-Arg₈]¹⁰⁺. In addition, a shoulder on the band at ∼1592 cm⁻¹ appears in the spectrum of [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ which may originate from the peptide.

Fig. 1 (i) The emission spectra of the free dye in an aerated acetonitrile solution with sequential addition of deionised water. (ii) The resonance Raman spectra of [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ and [Ru(dppz)₂PIC]²⁺ in PBS buffer and from within the cell (from resonance Raman mapping) after excitation at 458 nm.

The dppz ligands confer ‘molecular light-switch’ properties on the ruthenium complex wherein the emission is quenched in an aqueous environment due to hydrogen bonding to the phenazine nitrogens. This effect is illustrated in Fig. 1(i) wherein emission is seen to steadily decrease with addition of increasing volumes of water to a 100% acetonitrile solution of [Ru(dppz)₂PIC]²⁺. Complete extinction of emission occurs at approximately 18% v/v water. Overall, the photophysical properties of the conjugated complex mirror those of the parent, which exhibits the same aqueous light switch property. The ability of a lipid bilayer to switch the luminescence of these complexes on was confirmed by applying the complex to a suspension of large unilamellar vesicles formed from dipalmitoylphosphatidylglycerol (dppg) in PBS buffer at pH 7.4. Both free dye and conjugated dye showed no emission in the buffer alone but after addition to the vesicle suspension, strong emission was observed from the vesicles.¹¹ The capacity of [Ru(dppz)₂PIC]²⁺ and [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ to function as cellular membrane probes was then explored using confocal fluorescence imaging using mouse SP2/0 myeloma cells. Typically, 35 μM of dye or dye conjugate was added to the cell culture medium and incubated with the cells for approximately 48 h. Although the complex shows some toxicity to the cells after 48 h incubation (approx. 30% cell death) the cell culture remains viable. The cells were collected and imaged on a Zeiss Meta confocal microscope, live at 37 °C using an incubator. Fig. 2 shows the confocal luminescence images for the myeloma cells stained with (A) [Ru(dppz)₂PIC-Arg₈]¹⁰⁺, (B) with the parent dye excited at 458 nm.¹⁰

Fig. 2 Confocal luminescence images of live SP2 myeloma cells at 37 °C after incubation for 48 h with (A) [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ and (B) [Ru(dppz)₂PIC]²⁺. λ_ex = 458 nm, λ_em = 620 nm.

Following incubation with [Ru(dppz)₂PIC]²⁺ (B), only the outer membrane of the cells exhibits emission, Fig. 2(B). In contrast, following incubation with [Ru(dppz)₂PIC-Arg₈]¹⁰⁺, the outer and internal membranes associated with the sub-cellular structures emit strongly. Counter-staining indicates that these are the nucleus and cell organelles. In addition, on the cytoplasmic side of the outer cellular membrane, small features can be seen which are thought to be endosomal vesicles that occur when the dye is internalised via endocytosis. The emission spectrum of the internalised dye, collected on the Zeiss Meta, corresponded to the ³MLCT emission of the ruthenium complex. A z-stack was also performed to investigate the lipid structures and dye distribution throughout the cell at 0.2 μm increments over 20 μm; this is presented as a video in supplementary material.†¹¹ Similar results are observed on shorter incubation times but the most intense images require overnight incubation.

Although the emission images indicate that the distribution of the parent and conjugate are different, as only dye associated with membrane emits fluorescence imaging could not be used to assess the true distribution of the dye in the cell or to determine conclusively whether [Ru(dppz)₂PIC]²⁺ had crossed the membrane. Resonance Raman mapping, however, does not rely on the dye emission and was employed for the first time here as a complementary technique to the same cell medium to assess the location of the dye and conjugate. Fig. 3 shows the white light image and resonance Raman map for the SP2 cells incubated with the dye or the dye conjugate as described above for the confocal luminescence imaging; the excitation line was 458 nm. The resonance Raman image was generated by analysing the intensity of the 1598, 1564 and 1455 cm⁻¹ bands at each point in the map. Fig. 1(ii) shows examples of the resonance Raman spectra used to construct the map.

Fig. 3 Resonance Raman intensity map of a live myeloma cell after incubation with [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ (A1) and with the free dye [Ru(dppz)₂PIC]²⁺ (B1) after excitation at 458 nm. A and B represent the corresponding white light images of the live cells.

The plot intensity represents the intensity of these vibrational modes and therefore the relative dye concentration around the cell. Fig. 3 clearly reveals that whereas the parent complex accumulates in the outer membrane without crossing it, [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ crosses the membrane and is distributed throughout the cell. The concentration peaks close to the centre of the cell in the cytoplasm, but its distribution does not appear to reflect pre-concentration into the membranes. For the parent complex, the resonance Raman intensities match the luminescence images, indicating the dye does not transfer across the cell membranes but accumulates in the outer membrane and particularly at junctions between cells.

Fig. 1(ii) compares the resonance Raman spectra of [Ru(dppz)₂PIC]²⁺ and [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ in the myeloma cells and in pH 7.4 buffer. It is important to highlight that the confocal emission and resonance Raman images were collected from the same dye incubated cell medium, i.e. the concentrations of dye used for generating the emission images and resonance Raman images were the same, and that the spectra shown in Fig. 1(ii) are those that were used to generate the Raman image. The high resonance enhancement of the Raman spectra of this dye made it possible to produce high quality resonance Raman maps that reflect the distribution and structure of the dye in situ at an excitation wavelength that can also be used to excite emission. Overall, we have demonstrated Ru(II) probes can be exploited for simultaneous or consecutive imaging by emission and resonance Raman without the requirement for changes to excitation or dye concentration. In the example presented here, the ‘light-switch’ nature of the complex [Ru(dppz)₂PIC-Arg₈]¹⁰⁺ permitted its use as a membrane probe capable of selectively imaging the subcellular structures. Resonance Raman spectroscopy could be used to validate these results and showed the dye distribution within the cell, confirming the parent complex accumulates in the outer cell membrane but does not penetrate the cell. Overall, the Ru(II) complexes are highly versatile molecular probes for cell imaging and we believe their use in multimodal imaging as described here represents a powerful new paradigm in cell imaging.

This material is based upon work supported by Science Foundation Ireland under Grant No. [05/IN.1/B30] and The Higher Education Authority under PRTLI IV.

IRCSET are gratefully acknowledged for postgraduate scholarship funding.

Prof. Richard O’Kennedy, DCU, is sincerely thanked for providing cell culture.

Notes and references

1. (a) N. Uzunbajakava, A. Lenferink, Y. Kraan, E. Volokhina, G. Vrensen, J. Greve and C. Otto, Biophys. J., 2003, 84, 3968; (b) C. Krafft, Anal. Bioanal. Chem., 2004, 378, 60–62; (c) H. J. van Manen, A. Lenferink and C. Otto, Anal. Chem., 2008, 80, 9576.
2. (a) K. Y. Zhang, Inorg. Chem., 2009, 48, 6011–6025; (b) K.-W. Lo, T. K.-M. Lee, J. S.-Y. Lau, W.-L. Poon and S.-H. Cheng, Inorg. Chem., 2008, 47, 200; (c) M. R. Gill, J. Garcia-Lara, S. J. Foster, C. Smythe, G. Battaglia and J. A. Thomas, Nat. Chem., 2009, 1, 662; (d) K. L. Haas and K. J. Franz, Chem. Rev., 2009, 109, 492 (and references therein).
3. (a) G. Piszczek, Arch. Biochem. Biophys., 2006, 453, 54–62; (b) R. Iino and A. Kusumi, J. Fluoresc., 2001, 11, 187–195; (c) J. H. Rao, A. Dragulescu-Andrasi, H. Q. Yao and H. Q. Yao, Curr. Opin. Biotechnol., 2007, 18, 17–25; (d) A. Gomez-Hens and M. P. Aguilar-Caballos, TrAC, Trends Anal. Chem., 2004, 23, 127–136.
4. K. Szacilowski, W. Macyk, A. Drzewiecka-Matuszek, M. Brindell and G. Stochel, Chem. Rev., 2005, 105, 2647–2694.
5. (a) J. Olofsson, B. Onfelt and P. Lincoln, J. Phys. Chem. A, 2004, 108, 4391–4398; (b) C. G. Coates, J. J. McGarvey, P. L. Callaghan, M. Coletti and J. G. Hamilton, J. Phys. Chem. B, 2001, 105, 730–735; (c) M. K. Brennaman, J. H. Alstrum-Acevedo, C. N. Fleming, P. Jang, T. J. Meyer and J. M. Papanikolas, J. Am. Chem. Soc., 2002, 124, 15094–15098.
6. U. Neugebauer, Y. Pellegrin, M. Devocelle, R. J. Forster, W. Signac, N. Morand and T. E. Keyes, Chem. Commun., 2008, 5307–5309.
7. (a) S. M. Fuchs and R. T. Raines, Cell. Mol. Life Sci., 2006, 63, 1819–1822; (b) J. P. Richard, K. Melikov, E. Vives, C. Ramos, B. Verbeure, M. J. Gait, L. V. Chernomordik and B. Lebleu, J. Biol. Chem., 2002, 278, 585–590; (c) F. Duchardt, M. Fotin-Mleczek, H. Schwarz, R. Fischer and R. Brock, Traffic, 2007, 8, 848–866.
8. The presence of DMSO reduces the emission intensity and lifetime of [Ru(dppz)₂PIC]²⁺ compared to 100% acetonitrile solution but was used here to overcome the poor solubility of the ruthenium–peptide conjugate in organic solvent.
9. (a) C. G. Coates, J. Olofsson, M. Coletti, J. J. McGarvey, B. Onfelt, P. Lincoln, B. Norden, E. Tuite, P. Matousek and A. W. Parker, J. Phys. Chem. B, 2001, 105, 12653–12664; (b) W. R. Browne and J. J. McGarvey, Coord. Chem. Rev., 2006, 250, 1696–1709; (c) J. R. Schoonover, W. D. Bates and T. J. Meyer, Inorg. Chem., 1995, 34, 6421.
10. Incubation over shorter periods also worked (down to 40 min) but partitioning into the membranes was lower.
11. See supplementary material.

---

Footnote

† Electronic supplementary information (ESI) available: Detailed synthesis and characterisation data; video. See DOI: 10.1039/b918611e

---