Rhenium fac tricarbonyl bisimine complexes: Biologically useful fluorochromes for cell imaging applications

Article abstract of DOI:10.1039/b706657k

A series of lipophilic and hydrophilic fac tricarbonyl rhenium bisimine complexes have been prepared, their membrane-permeabilities explored in liposomes and their potential for application in fluorescence microscopy cell imaging demonstrated in the first

Full text of DOI:10.1039/b706657k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Rhenium fac tricarbonyl bisimine complexes: biologically useful
fluorochromes for cell imaging applications{
Angelo J. Amoroso,a Michael P. Coogan,*a Jennifer E. Dunne,a Vanesa Ferna´ndez-Moreira,a Jacob B. Hess,a
Anthony J. Hayes,c David Lloyd,b Coralie Millet,b Simon J. A. Popea and Craig Williamsa
Received (in Cambridge, UK) 18th May 2007, Accepted 4th June 2007
First published as an Advance Article on the web 20th June 2007
DOI: 10.1039/b706657k
to lanthanides, these seemed appealing targets for new cell staining
systems. The most studied system is probably the ruthenium
trisbipyridyl, and there are studies of the cellular uptake of
ruthenium complexes,5 however the excited state can be localised
on any of the three bipyridine units. Conversely, fac rhenium
tricarbonyl bipyridines4 and related species offer systems in which
the excited state is localised on the single bipyridine unit, making
them ideal for modification to allow a response to their environ-
ment. There are ruthenium complexes in which the excited state is
localised on a single ligand, the bipyridyl tetracyanoruthenates,
however these are highly polar anionic systems which were
thought unlikely to be membrane-permeable and thus not ideal
candidates for cell work.6 The fluorescence properties of these
species in simple solvents, with visible light excitation avoiding
problems of UV damage to tissue, and their relatively lipophilic
and monocationic nature seem suitable for cell imaging. However,
this apparently useful luminescence could be subject to quenching
or photobleaching as a result of interaction with species present in
complex biological media, and the degree and specificity with
which such rhenium complexes interact with organelles are
unknown, as is their toxicity. As heavy metal toxicity is often
associated with metal–DNA interactions these systems seemed
appropriate as they are d6 low spin and kinetically inert to ligand
substitution. Thus, a family of fac [Re(bisim)L(CO)3]+ complexes
have been prepared in which (bisim) represents a bisimine ligand,
i.e. a bipyridine, phenanthroline or derivative, and L is a pyridine
or derivative each of varying charges and lipophilicities, and their
membrane permeabilities and suitability for cell staining tested for
the first time as Re complexes previously studied in cells by fluores-
cence microscopy have relied upon organic fluorophores rather
A series of lipophilic and hydrophilic fac tricarbonyl rhenium
bisimine complexes have been prepared, their membrane-
permeabilities explored in liposomes and their potential for
application in fluorescence microscopy cell imaging demon-
strated in the first application of MLCT-fluorescent rhenium
complexes in cell imaging.
Fluorescence microscopy is a powerful, high resolution technique
in biological imaging and specific fluorescent staining techniques
make it especially useful for diagnostic investigations of tissue
samples. One of the challenges, however, is to differentiate the
endogenous fluorescence of biological species (autofluorescence)
from that derived from a dye which is applied to specifically stain a
target organelle or cell component.1 There is a large literature on
solving the problems associated with autofluorescence, but still it is
often difficult to distinguish clearly between those areas which have
accumulated the dye and those which show only autofluores-
cence.1 Autofluorescence may be filtered out from the desired
signal if there is an appreciable difference in wavelength between
the two modes of fluorescence, and this is particularly likely if the
dye has a large Stokes shift (the difference in wavelength between
the absorbed and emitted light), as the typical Stokes shifts for
the species involved in autofluorescence are small.1 Alternatively,
time-resolved microscopy can provide a method for eliminating
autofluorescence if the fluorescence lifetime of the added dye is
much longer than the short-lived autofluorescence. As both long
fluorescence lifetimes and large Stokes shifts are available in
certain lanthanide complexes, recently systems for cell labelling
using such complexes have been developed.2 Certain transition
metal complexes also offer Stokes shifts and lifetimes which, while
not as great as those of lanthanide complexes, could allow
differentiation from autofluorescence. In particular, there are
ruthenium3 and rhenium4 complexes which have well understood
triplet metal-to-ligand-charge-transfer (3MLCT) luminescence
characterised by useful Stokes shifts. As time resolved microscopy
is not as widely available as standard confocal fluorescence
microscopy, and in view of the much simpler coordination
chemistry of these particular transition metal complexes compared
3
than MLCT Re, the complexes being studied as models for Tc.7
In order to access complexes of the form fac [Re(bisim)L(CO)3]+
in which L is highly lipophilic a series of esters of 3-hydroxy-
methylpyridine, Py-3-CH2O2CR9 (R9 5 octyl, merystyl, steryl)
1a–c were synthesised{ (Scheme 1) and incorporated into the
complexes following an adaptation to the literature routes4 to such
complexes. This involved initial formation of rhenium tricarbonyl
bipyridyl chloride, then activation of the chloride (either exchange
aDepartment of Chemistry, Cardiff University, Cardiff, Cymru/Wales,
UK CF10 3AT. E-mail: cooganmp@cf.ac.uk; Fax: (44)2920874030;
Tel: (44) 2920874066
bCardiff School of Biosciences, Main Building, Museum Avenue,
Cardiff, Cymru/Wales, UK CF10 3TL
cConfocal Microscopy Unit, Cardiff School of Biosciences, Life Sciences
Building, Cardiff, UK CF10 3US
{ Electronic supplementary information (ESI) available: Experimental
details. See DOI: 10.1039/b706657k
Scheme 1
3066 | Chem. Commun., 2007, 3066–3068
This journal is ß The Royal Society of Chemistry 2007

View Article Online
Table 1 Luminescent properties of complexes 4 and 69
Compound
lmax excitation (nm)
lmax emission (nm)
4aa
4ba
4ca
4da
6yb
6zb
a
360
355
358
365
375
378
554
554
552
556
566
563
Scheme 2
b
Solvent: MeCN. Solvent: H2O.
to the triflate, or via halide abstraction to the acetonitrile complex),
then finally displacement with the substituted pyridine (Scheme 2).
The more water soluble anionic systems were accessed via
substitution of the bipyridyl unit with a commercially available
dianionic analogue, bathophenanthroline sulfate.8 This required
a series of counterion exchanges, as it was found that while the
sodium salt of the bathophenanthroline sulfonate reacted with
Re(CO)5Cl to give the required complex, further steps were
hindered by the insolubility of this species in anything but water, in
which substitution of chloride by hydroxide occurred with a half
life of approximately 2 hours at room temperature. No solvent or
phase transfer catalyst was found which allowed chemistry to be
performed on these salts, so an alternative route was developed in
which the bathophenanthroline sulfate sodium salt was first
treated with aqueous tetrabutylammonium hydrogen sulfate, to
precipitate the tetrabutylammonium salt of the ligand as a gel,
which upon drying gave a handleable anhydrous solid. This salt is
soluble in normal organic solvents, and allowed the synthesis of
the monoanionic rhenium complexes as the tetrabutylammonium
salts 6y,z by the normal route (Scheme 3). Ion exchange
chromatography on Amberlite IR 120-H as the ammonium form
gave the final complexes as their water-soluble ammonium salts.
With a variety of complexes in hand, their fluorescence
behaviour and lipophilicities were examined to assay their
suitability for cell microscopy applications. All the complexes 4
and 6 showed the photophysical properties expected from 3MLCT
based emission of rhenium bisimine complexes. The complexes
containing simple bipyridines and lipophilic pyridines showed the
expected photophysical properties for such species, with broad
excitation maxima around 350 nm, and broad emission maxima
around 555 nm. The bathophenanthroline complexes showed the
expected red shifts associated with more conjugated ligands of
around 20 nm for excitation to approximately 370 nm and 10 nm
emission to 565 nm. These higher excitation maxima are
particularly attractive as the broad excitation spectra give good
excitation characteristics in the visible, avoiding problems of UV
tissue damage and poor tissue penetration. These data are
summarised in Table 1. The lipophilicities of the complexes were
assayed by their interactions with liposomes, lipid bilayer vesicles
which are often used as models for cell membranes.10 If a solute is
encapsulated in the vesicle, then after chromatographic separation
from the initial solution, the ability of other species to penetrate
membranes can be studied by their addition to the liposome
preparation followed by examination of the interaction with the
encapsulated solute. Similarly the rate at which an encapsulated
species is able to leave the liposome can be studied by the
interaction with solutions of species which are known to be
membrane impermeable. In this study the membrane permeability
of the complexes 4 and 6 was studied by their interactions with
membrane impermeable fluorescence quenchers. The complexes
bearing lipophilic chains 4a–c appeared to reside mainly in the
lipid membranes, as after exposure to preformed liposomes
followed by GPC isolation, all the fluorescence was localised in
the fractions containing liposomes, indicating that they were
associated with the liposomes. Conversely, the bathophenanthro-
line sulfate derived complexes 6y,z appeared to reside solely in the
aqueous phase and be membrane impermeable. Exposure of
preformed liposomes to an aqueous solution of these complexes
led to no entrapment, whereas liposome generation using buffer
solutions containing complex led to partition of the fluorescence
between the liposomes and the small molecule fractions after GPC,
typical of a non-permeable species.
Addition of water soluble, non-membrane permeable quenchers
to 6y,z encapsulated in liposomes gave no loss of fluorescence,
whereas addition of the same species to 4a–c incorporated into
liposomes gave a significant but not total loss of fluorescence,
interpreted as indicating that the rhenium resides at both the inner
and outer boundaries of the membranes, and the quenchers are
partially able to quench those at the outer membrane.
In order to test the viability of these species as fluorophores in
cell microscopy their ability to act as microscopy stains was tested
in Spironucleus vortens, a parasitic flagellate which is found in
many species of fish.11 These organisms were chosen as they are of
interest in their own right as important pathogens which are little
understood, but also as they are prolific feeders and there was an
expectation that they would actively take up the complexes,
allowing a preliminary investigation of their properties in vivo
without considering cell-specific issues of active transport. Live
cells were harvested by centrifugation, incubated with solutions of
each complex in DMSO for 2 h (typically 10–40 mL of a 0.5 mM
solution of fluorophore incubated with 200 mL of cell suspension)
before being washed in buffer (PBS, pH 7.2), mounted in
methylcellulose to restrict movement and examined by confocal
fluorescence microscopy using a Leica TCS SP2 AOBS spectral
confocal microscope system. The samples were imaged using
Scheme 3
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 3066–3068 | 3067

View Article Online
Fig. 1
excitation at 405 nm and images were recorded using detection
between 520–570 nm, thus eliminating autofluorescence with a
Stokes shift of , 115 nm, and giving sharp fluorescence images.
The images are displayed with artificial colour fluorescence images
overlaying white dispersed light images. The results of these
preliminary experiments indicate that complexes of the type
[Re(bisim)L(CO)3]+ may offer a viable fluorophore for cell imaging.
The highly lipophilic species 4a–c were toxic at high concentrations
(40 mL of a 2 mM solution incubated with 200 mL of cell
suspension), apparently disrupting the membranes and leading to
cell lysis (Fig. 1(A)). The fluorescence was strongly associated with
cell fragments, and at lower concentrations (20 mL of a 2 mM
solution) lysis was avoided and the complexes appeared to be
associated with internal membranes partitioning cell compartments
and with the constituents within organelles (Fig. 1(B)). The simple
bipyridyl complexes 4d are more toxic than their sulfonated
analogues, but again show accumulation in cells. It is worth noting
that the neutral complex [ReCl(phen)(CO)3]12 was not only toxic
but rapidly photobleached. This photobleaching is not observed
in vitro and is not simply a result of photo-assisted hydrolysis, as
the analogous aqua complexes are highly emissive. Significantly,
no photobleaching was observed with any of the other dyes under
the experimental conditions, indicating that this is associated with
the chloride and that other Re complexes are resistant to photo-
bleaching. The more polar sulfonated complexes 6y,z were the
most successful species but showed a difference in toxicity between
themselves. The hydroxymethylpyridine complex 6z seemed to be
non-toxic or at least of low toxicity (cells were alive and mobile
after 2 hours) and it appears that localisation in certain cell
domains, thought to be digestive vacuoles, is occurring by phago-
cytosis (Fig. 1(C)). The simple pyridine analogue 6y was also
accumulated in the cells, but seems to be toxic as all the cells stained
with this species were dead. This series of results implies that there
is no intrinsic problem with the toxicity of rhenium complexes of
the type [Re(bisim)L(CO)3]+ per se, however toxicity may be
associated with certain ligands, and the choice of ligands will be
considered carefully in future studies targeting specific organelles.
In conclusion, fac rhenium tricarbonyl bisimine complexes
[Re(bisim)L(CO)3]+ have been shown to retain their useful
fluorescence in biological systems, and in certain cases have been
shown to have, at the worst, low toxicity, making them promising
candidates for the design of specific cell imaging agents.
We thank the EPSRC National Mass Spectrometry Service
in Swansea for their invaluable assistance, particularly in the
characterisation of the rhenium complexes and in situ generated
derivatives. We thank the EPSRC LSI for financial support to this
project (EP/D080401/1) and for the support which allowed the
purchase of the extruder for liposome generation (EP/C528638).
Notes and references
1 (a) N. Billinton and A. W. Knight, Anal. Biochem., 2001, 291, 175
(b) S. A. Schnell, W. A. Staines, M. W. Wessendorf and W. Martin,
J. Histochem. Cytochem., 1999, 47, 719; (c) M. Neumann and D. Gabel,
J. Histochem. Cytochem., 2002, 50, 437–439; (d) For a comprehensive
and authoritative review of issues relating to fluorescence see
J. Lakowicz, Principles of Fluorescence Spectroscopy, Springer,
New York, 3rd edn, 2006.
2 J. C. Frias, G. Bobba, M. J. Cann, C. J. Hutchison and D. Parker, Org.
Biomol. Chem., 2003, 1, 905; J. Yu, D. Parker, R. Pal, R. A. Poole and
J. Cann Martin, J. Am. Chem. Soc., 2006, 128, 2294; R. Pal and
D. Parker, Chem. Commun., 2007, 474; R. J. Aarons, J. K. Notta,
M. M. Meloni, J. Feng, R. Vidyasagar, J. Narvainen, S. Allan,
N. Spencer, R. A. Kauppinen, J. S. Snaith and S. Faulkner, Chem.
Commun., 2006, 909; C. D. B. Vandevyver, A.-S. Chauvin, S. Comby
and J.-C. G. Bunzli, Chem. Commun., 2007, 1716.
3 A. Juris, V. Balzani, F. Barigellatti, S. Campagna, P. Belser and
A. von Zelewsky, Coord. Chem. Rev., 1998, 84, 85.
4 T. J. Henly, Coord. Chem. Rev., 1989, 93, 269.
5 C. A. Puckett and J. K. Barton, J. Am. Chem. Soc., 2007, 129, 46;
B. Onfelt, L. Gostring, P. Lincoln, B. Norden and A. Onfelt,
Mutagenesis, 2002, 17, 317.
6 M. D. Ward, Coord. Chem. Rev., 2006, 250, 3128.
7 P. Haefliger, N. Agorastos, A. Renard, G. Giambonini-Brugnoli,
C. Marty and R. Alberto, Bioconjugate Chem., 2005, 16, 582;
V. Polyakov, V. Sharma, J. L. Dahlheimer, C. M. Pica, G. D. Luker
and D. Piwnica-Worms, Bioconjugate Chem., 2000, 11, 762.
8 There is already an example of a rhenium carbonyl diphosphine
complex in which bathophenanthroline sulfonate was incorporated
to impart water-solubility: Y. Shen, B. P. Maliwal and J. R. Lakowicz,
J. Fluoresc., 2001, 11, 315.
9 Unusually, these complexes are not particularly sensitive to dissolved
dioxygen with only a small apparent increase in emission intensity upon
degassing.
10 Liposomes were prepared using a LiposoFast Basic with a LiposoFast
Pneumatic Actuator purchased from Avestin Inc according to: M. J.
Hope, M. B. Bally, G. Webb and P. R. Cullis, Biochim. Biophys. Acta,
1980, 812, 55.
11 S. L. Poynton, W. Fraser, R. Francis-Floyd, P. Rutledge, P. Reed and
T. A. Nerad, J. Eukaryot. Microbiol., 1995, 42, 731.
12 M. K. Itokazu, A. S. Polo, D. L. A. de Faria, C. A. Bignozzi and
N. Y. M. Iha, Inorg. Chim. Acta, 2001, 313, 149.
3068 | Chem. Commun., 2007, 3066–3068
This journal is ß The Royal Society of Chemistry 2007