Ben-Zhan
Zhu
*ab,
Xi-Juan
Chao
a,
Chun-Hua
Huang
a and
Yan
Li
a
aState Key Laboratory of Environmental Chemistry and Ecotoxicology, Research Center for Eco-Environmental Sciences, The Chinese Academy of Sciences, P.O. Box 2871, Beijing, P. R. China 100085. E-mail: bzhu@rcees.ac.cn; Fax: +86-10-62923563; Tel: +86-10-62849030
bLinus Pauling Institute, Oregon State University, Corvallis, OR 97331, USA
First published on 5th April 2016
The dipyridophenazine (dppz) based ruthenium polypyridyl complexes are known as molecular ‘light-switches’ for DNA. This property is poised to serve in diagnostic and therapeutic applications, but the poor cellular uptake restricts their use in live cells. Herein, we show that the cellular uptake, and more interestingly and surprisingly, the nuclear uptake of cell-impermeable Ru(II)–polypyridyl cationic complexes such as [Ru(bpy)2(dppz)]2+ were remarkably enhanced by three structurally unrelated biochemical agents (pentachlorophenol, carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone and tolfenamic acid), by forming lipophilic and relatively stable ion-pair complexes, via a passive diffusion mechanism. Enantioselective imaging of live-cell nuclear DNA was observed between the two chiral forms of Ru(II) complexes. This represents the first report of an unprecedented new method for delivering the DNA ‘light-switching’ Ru(II) complexes into the nucleus of living cells via ion-pairing, which could serve as a promising general live-cell delivery method for other potentially bio-medically important but cell-impermeable metal complexes.
In the quest for new and better biological imaging agents, transition metal complexes provide a promising avenue for the design of diagnostic and therapeutic agents. Metal complexes able to emit from triplet metal-to-ligand charge transfer (MLCT) states offer many advantages as luminescent probes of DNA structure. Ever since it was discovered that the cationic ruthenium complex [Ru(bpy)2(dppz)]2+ (bpy = 2,2′-bipyridine, dppz = dipyrido[3,2-a:2′,3′-c]phenazine) functions as a molecular ‘light switch’ for DNA,3 great attention has been drawn to the DNA binding properties of polypyridyl complexes of d6 octahedral metal ions, specifically toward the development of highly sensitive and structure-specific DNA probes.4–6
As DNA imaging tools, such complexes offer several advantages over existing systems: MLCT excitations in the visible region, high Stokes shifts along with relatively low toxicity, and chemical and photo-stability. Until recently, study has been largely focused on the development of in vitro probes. The few studies involving direct imaging of DNA in live cells with such systems have had very limited success,7 with poor membrane permeability still being ascribed the major limiting factor.8–11
Barton's group found that the classic and structurally simplest [Ru(bpy)2(dppz)]2+ was unable to permeate into cells due to its poor lipophilicity.8 Strategies using either lipophilic ancillary ligands, such as 4,7-diphenyl-1,10-phenanthroline (DIP)8 or conjugation to protein (or steroids and peptides),10–12 have been employed to increase the membrane permeability of DNA-binding metal complexes. Although this has led to the successful cellular uptake of several metal-based MLCT luminescent systems,8,10–12 the in cellulo DNA binding of such systems has not been successfully demonstrated in live cells, although observed in permeabilized and fixed cells.13,14
During our study on the mechanism of synergistic chemical and biological effects between organic and inorganic compounds (especially metal complexes),15–20 we unexpectedly found that not only the cellular uptake, but more interestingly and importantly, the nuclear uptake of the cell-impermeable Ru(II)–polypyridyl cationic complexes such as [Ru(bpy)2(dppz)]2+ were remarkably enhanced by three structurally unrelated biochemical agents, possibly via forming novel lipophilic and relatively stable ion-pair complexes.
We found that not only the cellular, but more interestingly and surprisingly, the nuclear uptake of the cell-impermeable model Ru(II)–polypyridyl cationic complex, [Ru(bpy)2(dppz)]2+, was remarkably enhanced in the presence of PCP (Fig. 1a and S1a†). PCP is a biochemical agent that has been used primarily in the protection and preservation of wood products worldwide, and for eliminating snails to prevent schistosomiasis in developing countries.20,21 As can be observed by the relative intensities of cellular luminescence at different times, as revealed by CLSM, the cellular uptake of [Ru(bpy)2(dppz)]2+ also gradually increases with time when present together with PCP (Fig. S1b†). Quantitation by line plots indicates nuclear uptake of intense luminescence in the nucleus compared to other regions (Fig. S1c†).
To further confirm that the luminescence is in the nucleus, we co-stained [Ru(bpy)2(dppz)]2+/PCP with the two known membrane-permeable DNA stains (DAPI and Hoechst 33342) and a RNA stain (SYTO 9), and tracked them by both CLSM and the three dimensional structured illumination microscopy (SIM) with higher spatial resolution. The almost complete overlay images clearly show the nuclear DNA stain of [Ru(bpy)2(dppz)]2+ in the presence of PCP (Fig. 1b and c). It should be emphasized that the cells were alive after all these treatments (Fig. S2†) (see Experimental section for details on how to determine the health of the cells).
In addition to these luminescence microscopy studies, which rely on the DNA binding and subsequent activation of the ‘light-switch’ effect to observe the in cellulo location of [Ru(bpy)2(dppz)]2+, we were also able to show the nuclear distribution of [Ru(bpy)2(dppz)]2+ in live cells using transmission electron microscopy (TEM), with better spatial resolution7 (while organic dyes cannot be observed with this method). We found that [Ru(bpy)2(dppz)]2+ was clearly incorporated into the nucleus of cells in the presence of PCP, which is consistent with CLSM and SIM studies (Fig. 2d).
In clear contrast, cells treated with [Ru(bpy)2(dppz)]2+ complex (0.1 mM) alone for a short time (3 h) showed only a minor change in the luminescence profile. At higher concentration (0.5 mM) and even after longer incubation time (24 h), the [Ru(bpy)2(dppz)]2+ complex could be taken up by cells, but only in the cytoplasm, not in the nucleus (Fig. S3a†).
Similar nuclear uptake results were observed when the QSG-7701 cell was substituted with other cell-lines, including HeLa, HepG-2, HL-7702, MCF-7 and PC-12 cells, as well as bacteria such as Staphylococcus aureus (Fig. S3b†).
Taken together, the complementary application of flow cytometry, live-cell CLSM, SIM coupled with co-staining and TEM studies demonstrated unequivocally that [Ru(bpy)2(dppz)]2+ were not only taken up by live eukaryotic and prokaryotic cells in the presence of PCP, but more importantly, they were readily delivered into the nucleus of living cells.
In addition to [Ru(bpy)2(dppz)]2+, other cationic Ru(II) polypyridyl complexes, including the typical [Ru(phen)2(dppz)]2+ (phen = 1,10-phenanthroline), [Ru(phen)3]2+ and [Ru(bpy)3]2+, have also been extensively studied.4,24 We found that their cellular and nuclear uptakes were all enhanced by PCP, FCCP and TA (Fig. 2c, and S4a†).
The abovementioned results suggest that the enhanced cellular and nuclear uptake, with three structurally unrelated hydrophobic weak acids, is a general phenomenon for all cationic Ru(II) polypyridyl complexes.
It is well known that organic cation transporters (OCT) can facilitate the diffusion of endogenous organic cations, as well as a variety of drugs and toxins.5 The possible role of an OCT was then explored. We found that [Ru(bpy)2(dppz)]2+ uptake is not significantly affected in cells co-incubated with the OCT inhibitor cimetidine (Fig. S5d†), indicating that OCT is not responsible for the cellular uptake of [Ru(bpy)2(dppz)]2+.
With all the inhibitors used above, the results suggest that passive diffusion should be the major mechanism of cellular uptake of [Ru(bpy)2(dppz)]2+ in the presence of PCP. Passive diffusion is less cell-type specific, allows greater freedom for modification of the complex than transport via membrane proteins, and does not lead to entrapment in endosomes, as often occurs with endocytosis. As a result, this mechanism of passive diffusion may portend the broad applicability of metal complexes in different cell types for different intracellular functions (see above, Fig. S3b†).
Because lipophilicity has been considered an important factor in the cellular uptake of metal coordination complexes,8,9 we speculate that a lipophilic adduct might be formed between the hydrophilic and cell-impermeable [Ru(bpy)2(dppz)]2+ and PCP, thus readily penetrating through cell cytoplasmic membranes. Indeed, we found that an oily red droplet was formed and precipitated when PCP was added to the yellow [Ru(bpy)2(dppz)]2+ in an aqueous buffer solution (pH, 7.4) (Fig. 3a), suggesting a new adduct may be formed between them. To test whether this new adduct is lipophilic or not, 1-octanol/aqueous partitioning study was conducted, because the 1-octanol/water partitioning system is the common reference system for the determination of lipophilicity, widely employed for structure–activity studies. We found as expected that [Ru(bpy)2(dppz)]2+ alone cannot be partitioned into the organic phase, because it has a low partition coefficient (logP = −2.50)8 (Fig. 3a). However, when PCP was present, [Ru(bpy)2(dppz)]2+ was readily partitioned into the 1-octanol phase (Fig. 3a). These results indicate that a lipophilic adduct was indeed formed between [Ru(bpy)2(dppz)]2+ and PCP. Analogous effects were observed with other Ru(II) polypyridyl complexes and PCP, FCCP and TA (Fig. 3b, and S4b–e†). Furthermore, there arose the question of the binding nature of this lipophilic adduct.
Ion pairing has been considered to be one of the most fundamental atomic interactions in both chemistry and biology. Ion pairs were visualized as associations of two oppositely charged ions, retaining their basic properties when bonded together by coulombic forces, and to a lesser extent, by other interactions.28,29 Chemists have been fascinated by ion pairs for some time, particularly when one or more of the ions is a transition metal coordination compound.29 Such ion pairs have practical importance in catalysis and chromatography, as well as in some types of batteries and solar cells.28–30
The electric neutrality of the ion-pairs makes them non-conducting and sometimes lipophilic; therefore, they can be extracted into organic solvents or penetrate through cytoplasmic membranes.31,32 Ion-pairing reactions have often been utilized by coupling with solvent extraction and HPLC analysis, which have been used mainly for the separation and the concentration of ionic analytes. Although it has been assumed that ion pairing may also play an important role in biological system such as in ionic drug delivery, there have been very few convincing examples to date, especially in the case of hydrophilic metal cationic complexes.31,32
Since [Ru(bpy)2(dppz)]2+ is a coordinatively saturated and substitutionally inert, positively-charged cationic complex, and PCP is a weak hydrophobic acid that could be readily dissociated to produce its negatively-charged pentachlorophenolate anion under physiological pH, we speculate that a neutral and lipophilic ion pair complex might be formed between the two oppositely charged bulky aromatic ions. If this were true, the chemical composition of the ion pair should be [Ru(bpy)2(dppz)]2+[PCP−]2. However, we found that the typical analytical methods used for solutions were not suitable for structure determination of the unknown, mainly because of its poor solubility in both water and organic solvents. We also tried to co-crystallize it, but unfortunately, our efforts were not successful (possibly due to the presence of the phenazine nitrogens of the dppz ligand).
Because PCP was also found to enhance the cellular and nuclear uptake of other Ru(II) complexes such as [Ru(phen)3]2+ (a close analog of [Ru(bpy)2(dppz)]2+), we then extended our study to the interactions between [Ru(phen)3]2+ and [PCP−]. Fortunately, a single crystal could be cultivated after incubation of [Ru(phen)3]2+ and [PCP−], and its solid-state structure was determined using single-crystal X-ray diffraction. The solution of the diffracted data clearly showed that an ion pair was indeed formed between [Ru(phen)3]2+ and [PCP−], with a 1:2 stoichiometry (Fig. 3c; for detailed crystal data, see Tables S1 and S2†). The ion pair [Ru(phen)3]2+[PCP−]2 should belong to the outer sphere ion-pair that represents contact ion-pairs, in which the coordinatively saturated first coordination sphere of the cation is no longer accessible to the anion, and as a consequence, the anion is relegated to the second coordination sphere, interacting with the cation through electrostatic and other weak forces only (aromatic pi stacking, H-bonding, δ–δ, and CH–δ) (Fig. 3d). We assume that a similar ion pair complex should be formed between [Ru(bpy)2(dppz)]2+ and [PCP−], and indeed, we found that the binding affinity between [Ru(bpy)2(dppz)]2+ and the three biochemical agents (PCP, FCCP and TA) are relatively strong, as measured by the fluorescence displacement method using calf thymus DNA (Table S3 and Fig. S6†) (see Experimental sections for details on how to measure binding affinity).
The most interesting and surprising finding of this study is that PCP can enhance not only cellular uptake, but more importantly, it can also deliver the Ru(II) complex directly into the nucleus while maintaining the DNA recognition characteristics of the parent coordination complex (Fig. S6†). The reason for this and the possible underlying mechanism were investigated. We propose that it can be readily explained by the formation of not only neutral and lipophilic ion-pairs, but also relatively strong and stable ion-pair complexes between [Ru(bpy)2(dppz)]2+ and PCP (Scheme 1). The ion pair [Ru(bpy)2(dppz)]2+[PCP−]2 should be strong and stable enough (Table S3†) to compete with various binding components present either in cell culture medium (such as 10% FSB) or in cytoplasm, and lipophilic enough (Fig. 3) to penetrate through cytoplasmic and nuclear membranes. At the same time, it should be weak and labile enough to be readily dissociated, once it reaches nuclear DNA (that has stronger binding affinity for [Ru(bpy)2(dppz)]2+ (K = ∼106),3 and the dissociated [Ru(bpy)2(dppz)]2+ would then bind tightly with nuclear DNA (possibly via intercalation).33 The “freed” PCP may diffuse back to the medium, and then transport more [Ru(bpy)2(dppz)]2+ into the cell, possibly serving as a shuttle carrier.
This is in clear contrast to other studies aimed at effecting membrane permeability by increasing hydrophobicity through coupling to bigger molecules to form conjugates.10–12 While changes in hydrophobicity via conjugation can modulate cellular uptake, they can also affect either cellular localization of the parent complex, which often leads to a decrease in nuclear targeting, or the characteristics of luminescence of the parent complex.10–12 The study outlined herein shows that such an approach is not always required and represents a significant step in the development of these DNA binding systems toward in vivo applications.
In this study, we found that the enhanced cellular uptake, and especially nuclear uptake of the cell-impermeable cationic Ru(II) complexes can be achieved via forming lipophilic and relatively stable ion-pair complexes with three structurally unrelated hydrophobic weak acids. We suggest that the formation of ion pair complexes with similar lipophilic character and stability could be of relevance as a general mechanism for the delivery of other potentially bio-medically important, but cell-impermeable metal complexes into the expected cellular targets.
Compared with two other conventional DNA stains (Hoechst 33342 and DAPI) in living cells, [Ru(bpy)2(dppz)]2+/PCP has the following remarkable advantages as a fluorescence probe (Table 1): [Ru(bpy)2(dppz)]2+ is highly water soluble, photostable (Fig. S7†) and displays a large Stokes shift value with a long lifetime far-red emission, a factor that makes it extremely compatible with other imaging agents such as the heavily used green fluorescent protein (GFP). It is also well tolerated by both eukaryotic and prokaryotic cells, and more importantly, it can be used for both luminescence and TEM studies, which is one of the most distinguished characteristics of this system. Lastly, we found that there is a dramatic difference between the two chiral forms of [Ru(bpy)2(dppz)]2+ and [Ru(phen)2(dppz)]2+, therefore the chirality of the Ru(II) complex is another very important feature as an in cellulo DNA structural probe.
Ru(II)–dppza/PCP | DAPI | Hoechst 33342 | |
---|---|---|---|
MGB: minor-groove binder.a Ru(II)–dppz: [Ru(bpy)2(dppz)]2+ and [Ru(phen)2(dppz)]2+.b Molecular probes.c Ref. 33. | |||
Excitation/emission maximum (nm) | 445/620 | 358/461b | 350/461b |
Stokes shift | >150 | 103 | 111 |
TEM | + | − | − |
Photo-bleaching | Low | High (1–2 min) | High |
UV damage | − | + | + |
Enantioselectivity | + | − | − |
Cell permeable | + | −/+ | + |
Living cell/fixed cell | +/+ | +/+ | +/+ |
Binding mode | Intercalatorc | MGB/AT preferenceb | MGB/AT preferenceb |
Working concentration | 80 μg ml−1 | 0.5–10 μg ml−1 | 0.25–20 μg ml−1 |
Ru(II)–dppz complexes have been used in a wide variety of applications, including DNA detection, DNA topoisomerase and RNA polymerase inhibition, DNA footprinting, amyloid-β-fibrillisation and α-synuclein aggregation detection, cell imaging, photodynamic therapy, anti-proliferative and anticancer effects.4–14,25,26 However, almost all these applications have not yet been realized in living cells. Because there are huge differences between non-cellular and live-cell systems, luminescent Ru(II) complexes capable of passive cell delivery may find a huge number of potential in cellulo or in vivo applications in those areas, which may lead to more interesting new findings.
Footnote |
† Electronic supplementary information (ESI) available. CCDC 1046781. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c5sc03796d |
This journal is © The Royal Society of Chemistry 2016 |