Uptake and localisation of rhenium fac-tricarbonyl polypyridyls in fluorescent cell imaging experiments

Article abstract of DOI:10.1039/c004610h

The synthesis of a series of rhenium fac tricarbonyl bisimine complexes and their application as lumophores in fluorescence imaging of yeast and human adenocarcinoma cells is reported. A wide range of complexes are synthesised with varying charges and lip

Full text of DOI:10.1039/c004610h

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Uptake and localisation of rhenium fac-tricarbonyl polypyridyls in fluorescent
cell imaging experiments†
Vanesa Ferna´ndez-Moreira,‡a Flora L. Thorp-Greenwood,‡a Angelo J. Amoroso,a Joanne Cable,b
Jonathan B. Court,b Victoria Gray,b Anthony J. Hayes,c Robert L. Jenkins,a Benson M. Kariuki,a David Lloyd,b
Coralie O. Millet,b Catrin Ff. Williamsb and Michael P. Coogan*a
Received 24th March 2010, Accepted 4th June 2010
First published as an Advance Article on the web 30th June 2010
DOI: 10.1039/c004610h
The synthesis of a series of rhenium fac tricarbonyl bisimine complexes and their application as
lumophores in fluorescence imaging of yeast and human adenocarcinoma cells is reported. A wide
range of complexes are synthesised with varying charges and lipophilicities, all of which have
photophysical properties which make them suitable as cell imaging agents. After attempts to apply these
as imaging agents in various strains of yeast which showed limited uptake, an investigation was
undertaken of their applications as imaging agents in mammalian cells. In general the uptake was high
and short-term toxicity and photobleaching appear to be low. The patterns of uptake and localisation
are correlated with structural and electronic features of the complexes in an attempt to establish
ground-rules for the design and application of rhenium complexes in imaging of eukaryotes.
display phosphorescence, and the technique is described as fluo-
rescence microscopy, the term lumophore is used throughout to
Introduction
avoid differentiating between mechanisms of emission).2 Typically,
autofluorescence Stokes shifts are small (tens of nm) and large
Stokes shift lumophores allow autofluorescence to be eliminated
without losing signal intensity. Similarly, luminescence lifetime
can be used to filter out short-lived autofluorescence when a
lumophore has a significantly longer lifetime.2 Photobleaching
usually occurs when the excited state of a lumophore reacts
withoxygen species (sometimes autogenous photoproducts) giving
non-emissive products. Rapid destruction of lumophores upon
intracellular irradiation then leads to loss of signal intensity
and can seriously compromise time-resolved studies of cellular
processes.2 However, while there are now a number of examples
of individual d6 complexes applied in cell imaging, as yet the
basic rules for the uptake and localisation of each family of
complexes are only just beginning to emerge.1 Much of the focus
with Ru complexes has concerned the use of complexes of highly
conjugated aromatic ligands designed to act as intercalators for
DNA/RNA especially for demonstration of nuclear localisation.3
There have been studies of Ir and Re species bearing specific
chemical groups designed to interact with particular biological
entities,4 but as yet there is no general study of the uptake
and localisation of a range of complexes of these metals. In
comparison, the rules relating to the patterns of localisation
of certain families of lanthanide-based lumophores are better
understood and have been more extensively studied.5 Previous
reports of the application of rhenium complexes in cell imaging fall
into two distinctive patterns. This group has previously reported
the application of a small range of rhenium complexes of relatively
simple structure in a parasitic flagellate, Spironucleus vortens, as
a preliminary investigation into their uptake and localisation.6
Other reports, both from this group and others, have involved
the imaging of mammalian cells (usually human carcinoma cells,
either HeLa or MCF-7) and the imaging agents have been highly
Since the advent of metallopharmaceuticals such as cisplatin the
application of synthetic metal complexes, including radioisotopes,
in diagnostic and therapeutic biomedical contexts has developed
dramatically; however, the ground-rules appertaining to the up-
take and transport of such species are far less clear than for
either essential metal ions, or exogenous organic therapeutic and
imaging agents.1a,b Many of the best known organic fluorophores
commonly applied as imaging agents are planar or near planar,2
and it is likely that there will be significant differences between
their interactions with e.g. sheet-like phospholipid bilayers and
those with more complex and sterically demanding geometry.
A fuller understanding of how typical metalloagents such as
octahedral transition metal complexes interact with biological
membranes and behave in cellular environments will be necessary
for many imaging and therapeutic applications. In particular, the
application of d6 metal complexes in fluorescent cell imaging is an
emerging research area which has recently received a great deal
of attention.1c The attractive intrinsic photophysical properties
common to many Ir(III),1d Re(I)1e and Ru(II)1f complexes, such as
large Stokes shifts, long luminescence lifetimes and resistance to
photobleaching, make these excellent candidates for applications
in cell imaging. Large Stokes shifts are useful in fluorescence
imaging as autofluorescence from endogenous fluorophores, e.g.
NAD(P)H, flavin or chlorophyll, can be confused with desired
lumophore emission (NB although rhenium complexes commonly
aSchool of Chemistry, Cardiff University, Cardiff, Wales, UK CF10 3AT.
E-mail: cooganmp@cf.ac.uk; Fax: (+) 4402920874030
bSchool of Biosciences, Cardiff University, Cardiff, Wales, UK CF10 3AT
cConfocal Microscopy Unit, Cardiff School of Biosciences, Life Sciences
Building, Cardiff, Wales, UK CF10 3US
† Electronic supplementary information (ESI) available: X-Ray crystallo-
graphic data for 14 and 15. See DOI: 10.1039/c004610h
‡ Equal contribution from these authors.
3888 | Org. Biomol. Chem., 2010, 8, 3888–3901
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
specialised, and usually very complex. To date, rhenium complexes
conjugated to fMLF (a small peptide used as a targeting agent for
formyl peptide receptor, FPR),7 biotin8 and cobalamine9 have been
applied in cell imaging, as has a chloromethyl rhenium complex
designed to target mitochondria,10 and a dipicolylamine rhenium
derivative which responds to zinc ions.11 There has not, however,
been a systematic investigation of the uptake and localisation
properties intrinsic to the rhenium tricarbonyl skeleton to inform
the design of future imaging agents. Here, we report the results of
a more systematic qualitative study of the patterns of uptake and
localisation of a series of rhenium tricarbonyl bisimine complexes
of varying lipophilicity, charge, size and structural type in yeast
and in human adenocarcenoma MCF-7 cells.
coupling reactions of the complex and synthesis of the conjugated
ligands followed by coordination. Interestingly, it will be seen
that the preferred approach varies with the overall charge of the
complexes.
Cationic complexes
In order to provide mutually comparable results it was decided
that the polypyridyl portion of the complex should be represented
in each cationic case by bipyridine, and the axial pyridine
ligand should be varied through a series of functional group
variations in the 3-position. Thus, rhenium tricarbonyl bipyridine
halides were prepared by reaction of bipyridine with the rhenium
pentacarbonyl halides. Direct reaction of these halo-species with
pyridine and hydroxymethyl pyridine, respectively, in the presence
of silver tetrafluoroborate gave the complexes 112 and 26 in good
yields (Scheme 1). For more complex species the halides were
activated as the triflate 3 or acetonitrile derivative 4 as previously
discussed (Scheme 2).
Results and discussion
Design and synthesis of complexes
The standard methods for the preparation of luminescent rhenium
tricarbonyl species are typically to react the commercially available
rhenium pentacarbonyl halides with a bisimine ligand such as
bipyridine or phenanthroline.10 These reactions typically give
stable rhenium bisimine fac-tricarbonyl halides which require
further activation in order to allow coordination of the axial
pyridine derivative. There are a variety of methods available
for this activation, the three most common being: i) halide
abstraction with silver salts in the presence of a pyridine can lead
directly to the cationic pyridine derivative; ii) halide abstraction
in acetonitrile to give the (usually) cationic and highly labile
acetonitrile complex; iii) conversion of the halide to the triflate by
treatment with triflic acid, being the most commonly encountered
three approaches. In the latter two cases the activated complexes
are then treated with the desired pyridine derivatives (usually in
THF) to give the final species. As the bisimine ligand is the one
involved in the MLCT process12,13 it was decided to use the axial
ligand in order to incorporate groups which tune the solubility,
lipophilicity etc. of the complex without (usually) affecting the
photophysical properties. Conjugation of simple pyridines in
the axial position of the complexes to alkyl chains and rings
was selected to tune the lipophilicity of the agents. While it is
more usual to perform coupling reactions on free ligands, post-
coordination modification of ligands has previously proved useful
in the synthesis of rhenium complexes for use in cell imaging, and
as part of this investigation two approaches to the systems required
were investigated: coordination of simple pyridines followed by
Scheme 1 Synthesis of simple Re(CO)3(bipy)(PyX) complexes
A series of non-commercially available pyridyl ligands of vary-
ing lipophilicities and structures were prepared by simple organic
transformations. As it was of interest to compare lipophilic species
of different structures, alkyl chains and cyclohexyl rings were
incorporated into the ligands. As there are cases in which esters
have been known to break down in the cellular environment14
an amide bearing a lipophilic group was also prepared (derived
from cyclohexylamine). The alkyl chain-containing complexes
were derived from fatty acids conjugated to a hydroxyl-substituted
pyridine, while the cyclohexyl rings were incorporated as the ester
and amide derived from cyclohexanol15 and cyclohexylamine,16
respectively, and nicotinic acid. In all cases two approaches to the
Scheme 2 Activation of rhenium complexes by halide loss.
The Royal Society of Chemistry 2010 Org. Biomol. Chem., 2010, 8, 3888–3901 | 3889
This journal is
©

View Article Online
complexes were attempted, namely pre-formation of the lipophilic
ligand followed by coordination to rhenium, and the conjugation
of a preformed rhenium complex to the lipophilic component.
Coupling of 3-hydroxymethylpyridine with a variety of fatty acids
using EDCI gave the derived esters, 5–76 (Scheme 3) which were
then reacted with 4 to give a series of complexes 8–10 bearing
differing chain length fatty acids (Scheme 4). The analogous
coupling reactions of the pre-coordinated species 2 with the
relevant fatty acids gave the complexes 8–10 but in lower overall
yields than the alternative route in which the pyridine and fatty
acid were pre-conjugated (Scheme 5).
Scheme 4 Synthesis of lipophilic rhenium complexes from pre-coupled
ligands.
approach, giving very poor yields of the required species after
complex isolation procedures (Scheme 7).
These cyclohexyl species were more crystalline than the fatty
acid derivatives 8–10 and crystals for X-ray diffraction19 were
grown of 15; however, the crystals of the cyclohexanol-derived
species 14 were of poor quality and while various diffraction data
sets were obtained which could be solved, a satisfactory refinement
could not be achieved. The analogous amide 15 gave a well resolved
structure in space group P21/c. The structure showed essentially
octahedral geometry and confirmed the expected fac disposition
of the carbonyl ligands (Fig. 1). The bond lengths and angles were
again within the range expected for analogous complexes and
the deviation from ideal octahedral geometry (average N(Py)–Re–
N(bipy) angle = 83.7◦) can be explained by crystal packing forces
(Fig. 2). Crystal packing is dominated by interactions between
Scheme 3 Synthesis of fatty-acid-derived pyridine ligands.
Reaction of nicotinic acid chloride 1117 with cyclohexanol
and cyclohexylamine gave the analogous ester 12 and amide 13.
These cyclohexyl ligands were then reacted with rhenium bipyridyl
acetonitrile complex 4 to give the derived cationic complexes 14
and 15, respectively (Scheme 6). Again, the attempted conjugation
of the nicotinic acid complex 1618 with cyclohexylamine and
cyclohexanol proved less successful than the more traditional
Scheme 5 Synthesis of lipophilic rhenium complexes by post-coordination modification.
Scheme 6 Synthesis, and coordination of cyclohexyl pyridine ligands.
3890 | Org. Biomol. Chem., 2010, 8, 3888–3901 This journal is The Royal Society of Chemistry 2010
©

View Article Online
Scheme 7 Synthesis of cyclohexyl complexes by post-coordination modification.
which were measured for a simple cationic complex 2, log Po/w
-0.37, and the myristryl derived species 9, log Po/w = 0.91.
=
Anionic complexes
As a range of complexes of varying overall charge was desired,
anionic ligands which could compensate for the cationic Re(I)
centre were required. We have previously reported the synthesis of
a small number of rhenium complexes derived from the anionic
ligand bathophenanthroline disulfonic acid, which is commercially
available as the sodium salt. However, in order to synthesise a
larger range of complexes derived from this ligand a new synthetic
method was required to overcome the difficulty of handling these
species in non-aqueous solution, as it was rapidly realised that the
rhenium halides derived from this ligand are highly susceptible to
hydrolysis. Reaction of bathocuprine disulfonic acid sodium salt
with rhenium pentacarbonyl chloride or bromide in acetonitrile
gave the expected complex 17. However, this anionic complex was
of such low solubility in organic solvents that characterisation
was only possible as a solution in water, and it was observed that
upon standing in D2O solution the complex suffered hydrolysis to
the aqua species, 18, which proved to be unreactive towards the
pyridine ligands required. For this reason a counterion exchange
process was performed on the ligand prior to coordination. The
commercial disodium salt was treated with two equivalents of
tetrabutylammonium hydrogen sulfate in water which resulted in
the immediate precipitation of a jelly-like substance, assumed to
be the organic electrolyte 19. After an extraction into DCM and
drying the product, 19 was isolated from the organic phase as
an easily handleable, pale pink amorphous solid (Scheme 9). The
use of the tetrabutylammonium salt 19 allowed the formation
of soluble, anionic rhenium complexes which could be handled
in organic solvents, thus avoiding problems of hydrolysis, which
was problematical for 17 (Scheme 8). Reaction with rhenium
pentacarbonyl chloride gave the expected tricarbonyl complex
20 in good yield (Scheme 10). The activated rhenium complex
21 was synthesised by stirring the anionic complex with triflic
acid in DCM, and rapid substitution of the halide was observed
(Scheme 11). With this activated complex 21 in hand, substitution
of the axial ligand was attempted by heating with pyridine and with
3-hydroxymethylpyridine in polar solvents (Scheme 12). While
in each case the desired products 22 and 23, respectively, were
obtained successfully, prolonged reaction times (5–7 days) were
Fig. 1 ORTEP20 image of asymmetric unit of 15 (50% probability).
Fig. 2 ORTEP20 image of packing of 15 (50% probability).
the cyclohexyl rings and the bipyridines, leading to the <90◦
angle observed between these groups. However, an additional
interaction between the nicotinamide units leads to a close
˚
interaction (<3.5 A) between these p-systems (Fig. 2). Full details
of refinement and structure, and tables of selected bond angles
and lengths are included in the supplementary information.† The
relative lipophilicities of these species are indicated by the log Po/w
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3888–3901 | 3891
©

View Article Online
Scheme 8 Hydrolysis of water-soluble complexes.
the complexes once a pyridine derivative was coordinated, the
hydroxymethyl complex 27 was reacted with myristic acid in
the presence of EDCI and DMAP in DCM, giving a much
more lipophilic complex 28 than those previously synthesised.
Interestingly, in this case precoordination of 3-hydroxymethyl
pyridine followed by post-coordination modification proved the
more effective approach (Scheme 16). After having synthesised a
variety of sulfonated rhenium complexes as tetrabutylammonium
salts, the recovery of the water solubility was then achieved using
ion exchange chromatography with Amberlite IR 120-H, as the
ammonium form. Thus, three monoanionic complexes of varying
lipophilicity have been prepared, each based around the rhenium
fac-tricarbonyl bathophenanthroline disulfonate core, with axial
ligands represented by pyridine, 3-hydroxymethylpyridine and 3-
tetradecanoyloxypyridine, the latter being a lipophilic anionic
species. In addition to the cationic species synthesised above,
and the chloromethyl species 29 prepared from 2 (Scheme 17),
this gave a range of species of varying lipophilicities, charges and
reactivities. The simple anionic species are less lipophilic than the
cations, with the simple pyridyl species 22 having log Po/w = - 0.53.
Scheme 9 Ion exchange to lipophilic sulfonated ligand.
required. It is postulated that these extended reaction times are
due to the presence of sulfonic acids in the reaction mixtures
derived from protonation of the ligand by triflic acid to the
neutral complex 24 or the cationic species 25 (Scheme 13).
Either of these species could protonate the potential ligands and
thus lead to the prolonged reaction times, even in the presence
of an excess of the pyridine, by lowering the nucleophilicity
of the nitrogen by hydrogen bonding. In order to avoid this
complication, activation as the acetonitrile derivative 26 was
attempted by halide abstraction with silver tetrafluoroborate in
acetonitrile (Scheme 14) and the new activated rhenium complex
26 substituted with pyridine ligands (Scheme 15). In an attempt
to demonstrate the possibility of further functionalisation of
Photophysical characteristics of imaging agents
As the polypyridyl ligand involved in the MLCT luminescence
was not varied throughout each of the series, with the cationic
species containing a bipyridyl ligand and the anionic complexes
a bathophenantholine disulfonate, the photophysics of the com-
pounds were essentially identical for each series. The cationic
Scheme 10 Coordination of lipophilic sulfonated ligand.
3892 | Org. Biomol. Chem., 2010, 8, 3888–3901 This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Scheme 11 Activation of anionic rhenium complex as triflate.
Scheme 12 Coordination of pyridines to activated anionic rhenium complex.
Scheme 13 Protonated forms of sulfated rhenium complexes.
Scheme 14 Activation of sulfated rhenium complex as the acetonitrile
Scheme 15 Formation of hydroxymethyl rhenium complex 27.
adduct.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3888–3901 | 3893
©

View Article Online
cations (lmax excitation, lmax emission) are summarised in the
supplementary information.† While the excitation maxima of all
the species lies in the UV region of the electromagnetic spectrum,
1
importantly the nature of the MLCT excitation band, which is
very broad, allows sufficient absorption in the visible region to
allow the use of a 405 nm laser as the excitation source. It is
important for cell imaging purposes to avoid UV excitation as UV
has poor tissue penetration and also causes tissue damage. It is also
convenient that many commercial fluorescence microscopes have
an option for 405 nm excitation in confocal mode, and the poor
excitation of TMRE at this wavelength (Fig. 5) allows differential
excitation of the complexes in co-localisation experiments. The
large Stokes shifts of all the species studied (150–200 nm) allows
easy differentiation between emission emanating from the rhenium
complexes and autofluorescence.
Scheme 16 Post-coordination modification of anionic rhenium com-
plex 27.
Scheme 17 Preparation of 29 according to ref. 10
species (Fig. 3) show broad featureless excitation and emission
maxima at 380 and 550 nm, respectively, and the anionic species
at 380 and 560 nm (Fig. 4). The excitation bands were assigned
Fig. 5 Excitation and emission spectra of TMRE (note minimal excita-
tion at 405 nm).
1
3
as MLCT and the emission bands as MLCT by comparison
with previous studies. As the aim of this work was to investigate
uptake and localisation of the species an in-depth study of the
photophysical properties of the complexes (variable temperature
measurements, time-dependent studies) was not undertaken, but
the characteristics important for fluorescence microscopy appli-
Cellular studies and confocal microscopy
Imaging experiments in yeasts
In order to establish ground-rules relating to the uptake and
localisation of rhenium complexes by eukaryotic cells, an initial
investigation of their uptake into yeasts (Saccharomyces cerevisiae)
was undertaken. The results of these investigations were disap-
pointing, in that across a variety of complexes both cationic and
anionic, and with a range of lipophilicities, only a small proportion
of the cells showed any detectable uptake of the lumophores
(Fig. 6). The best uptake was seen with the cationic lipophilic
species 9, but even with this species only some of the population
showed appreciable rhenium-based luminescence. Some uptake
was also observed even for the anionic species (although less than
in cationic species) and it appeared that the cells which showed
uptake could be those which were in the process of budding,
suggesting that the relatively impermeable cell wall associated with
yeasts was preventing uptake, and it was only those in which the cell
wall-membrane complex, involved in cellular proliferation-related
activities, was imperfect which were taking in the lumophores
(this pattern was also observed, although to a lesser extent,
for the cationic species 9, 29) (Fig. 7). In order to probe the
hypothesis that uptake of rhenium species was associated with
division of yeast cells, further investigations were undertaken with
fission yeast, Schizosaccharomyces pombe.21 In this species, which
typically grows to become a 14 ¥ 3.5 mm cylinder and then divides
transversely upon reaching this length, it follows that in a healthy
growing and dividing sample there should be a correlation between
Fig. 3 Excitation and emission spectra of 14.
Fig. 4 Excitation and emission spectra of 27.
3894 | Org. Biomol. Chem., 2010, 8, 3888–3901
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Fig. 9 Overlaid fluorescence and transmitted light images of 22 in fission
yeast (scale bar = 12 mm).
numbers of the lengthier cells showing intense luminescence, so it
was not possible to unambiguously correlate uptake with length,
and therefore with stage of the division cycle. Again, as for bakers
yeast, however, the majority of the sample of any length showed no
luminescence, and so the level of uptake is low, and the mechanism
of uptake remains ambiguous. While there appeared to be some
interesting patterns of uptake and localisation of rhenium species
in yeasts, it was felt that further study of systems in which such
low levels of uptake was observed was not worthwhile. Therefore,
attention turned to mammalian cells which have less impervious
limiting layers. Thus, each of the species described above was
incubated with◦MCF-7 (human adenocarcinoma) cells at a low
temperature (4 C) at which endocytosis is essentially inoperative.2
This temperature was chosen so that the ability of the lumophores
to permeate cells by passive diffusion could be assessed. It is
important for lumophores not to be entrained in endosomes if
they are to be capable of localising in e.g. mitochondria or nucleoli,
and thus a non-endocytotic uptake mechanism was desirable.
While the imaging experiments themselves were carried out at
ambient temperature, the fluorophore-containing medium had
been washed away by this point so re-emergence of endocytosis
should not have interfered. It has previously been noted8 for related
complexes that both endocytosis and passive mechanisms can be
at play in mammalian cells, so the low temperature was chosen in
order to eliminate entrainment in endosomes which could give a
misleading picture of the preferred sites of localisation.
Fig. 6 Overlaid fluorescence and transmitted light images of 8 in
Saccharomyces cerevisiae showing low uptake (scale bar = 15 mm).
Fig. 7 Overlaid fluorescence and transmitted light images of 23 in
Saccharomyces cerevisiae showing fluorescence concentrated in budding
cells (scale bar = 6 mm).
length of a given cell and the stage of the division cycle. If, as
had been hypothesised from the observations with Saccharomyces
cerevisiae, uptake of lumophore is associated with a weakening of
the cell boundary at a certain stage of division, then the intensity
of luminescence should be a function of the stage of the division
cycle and thus the length. Again, the lipophilic species 9 showed
the best uptake (Fig. 8), but there was significant uptake for all
species tested, including the anionic cases (Fig. 9). Unfortunately
while the most intense luminescence did appear to be in small cells
(i.e. those which had recently divided) there were also significant
Cationic complexes in MCF-7
Uptake of the cationic lipophilic complexes 8–10 bearing alkyl
chains in MCF-7 cells was good to excellent with consistent
patterns of uptake and localisation being observed. Liposome
studies6 had already indicated that these species were highly
membrane permeable and therefore a high degree of cellular
uptake had been expected. There were, however, several positive
features to the observed uptake. Firstly, in no case was significant
toxicity over the lifetime of the experiment observed, with normal
cell structural morphology being maintained as well as it was in
parallel experiments using known fluorophores of low toxicity
such as TMRE; visual inspection indicated healthy cell morphol-
ogy in >90% of cells in each sample. As a semi-quantitative test of
toxicity MCF-7 cells were co-incubated with Trypan blue, which
is only taken up by dead or dying cells, and a small number of
complexes. Cell counts showed that Trypan blue had not been
taken up by a greater number of cells in the samples treated with
anionic species, (normalised to untreated blanks). However, the
Fig. 8 Overlaid fluorescence and transmitted light images of 9 in fission
yeast (scale bar = 8 mm).
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3888–3901 | 3895
©

View Article Online
cationic highly lipophilic species 9 showed relatively high toxicity
at high concentrations, with as low as ca. 25% cell viability at
100mg ml-1 (normalised to untreated blanks). Presumably this is a
result of membrane perturbation as previously noted in parasites.6
Secondly, it was apparent that species 8–10 showed differing
patterns of localisation. This was particularity gratifying as it had
been suspected that the integrity of the ester linkages could be
compromised within the cells by enzymatic hydrolysis. However, as
hydrolysis by esterases would yield 2 in each case and the pattern
of localisation of 8–10 were notably different (and in each case
different from that observed for 2) this was clearly not a significant
problem. Localisation of 8–10 was observed in the cytoplasmic
membrane and cytoplasm, with the cytoplasmic patterns being
non-uniform, indicating localisation at lipophilic sites in organel-
lar membranes, endoplasmic reticulum and the Golgi apparatus.
The longer the chain the more specifically localisation was
observed in the membranes, with 10 being essentially a membrane
stain, and 8 and 9 more general cytoplasm (Fig. 10) and internal
membrane targeting agents. A small degree of nuclear staining
was observed with 9 with a highly specific pattern that appeared
to indicate localisation in the nucleolus (Fig. 11); however, the
small number of cells per sample showing this phenomenon
made co-localisation to confirm this impractical. As the long-
chain appended complexes could in principle act as surfactants
and render membranes permeable, as a semi-quantitative test
of permeability, MCF-7 cells were co-incubated with propidium
iodide, which is only taken up when membranes are damaged
or rendered permeable by surfactant, along with complexes 27
and 9. Cell counts compared to untreated blanks showed that
propidium iodide had not been taken up by a greater number
of cells in the samples treated with anionic complex 27; however,
again complex 9 clearly shows perturbation of membranes with ca.
50% of cells above standards showing uptake. In order to eliminate
the possibility that 8–10 were in fact hydrolysing to give 2, the
pattern of localisation of 2 was examined. Somewhat surprisingly
2 showed a pattern of localisation reminiscent of the mitochondrial
localisation previously reported for the chloromethyl-substituted
complex 29 (Fig. 12 and 13). This pattern of localisation has been
confirmed by co-localisation experiments with TMRE and, in the
case of 29, had been attributed to the cationic lipophilic nature
of the complex leading to a propensity for accumulation in mi-
tochondria; thiol reactivity associated with the chloromethyl unit
‘fixes’ the lumophore in the mitochondria by irreversible reaction
with e.g. GSH.
Fig. 12 Overlaid fluorescence and transmitted light images of 29 in
MCF-7 cells showing mitochondrial localisation (scale bar = 14 mm).
Fig. 13 Overlaid fluorescence and transmitted light images of 2 in MCF-7
cells showing mitochondrial localisation (scale bar = 9 mm).
That 2 shows such specific mitochondrial localisation (while
other cationic lipopophilic complexes do not, vide infra/supra)
suggests that 2 may also show thiol reactivity. While 2 lacks the
chloromethyl group which is assumed to be the thiol-reactive unit
of 29, it is postulated that phosphorylation of the hydroxyl unit
of 2 to give 30 confers thiol reactivity (Scheme 18), although
it should be stressed that no evidence was obtained to support
this hypothesis. A simpler cationic lipophilic complex which lacks
the possibility for phosphorylation of 2 (and thus the ability to
acquire thiol reactivity) is offered by the cyclohexyl ester 14.
The localisation of this species was more comparable to that
observed for the fatty-acid-derived species 8–10, showing non-
uniform localisation in the cytoplasm, with little nuclear uptake
but apparent localisation in organelles in the perinuclear region
(Fig. 14). As it was of interest to compare this species with the
chloromethyl species 29 (which is known to be thiol reactive) and
the hydroxymethyl species which is thought to be phosphorylated
Fig. 10 Overlaid fluorescence and transmitted light images of 8 in MCF-7
cells showing localisation in the perinuclear cytoplasm (scale bar = 6 mm).
Fig. 11 Overlaid fluorescence and transmitted light images of 9 in MCF-7
cells showing nuclear localisation (scale bar = 12 mm).
3896 | Org. Biomol. Chem., 2010, 8, 3888–3901
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
Scheme 18 Proposed activation of 2 by phosphorylation.
Fig. 16 Fluorescence from Re (A) TMRE (B) and transmitted light (C)
images of 14 and TMRE co-administered to MCF-7 cells demonstrating
the more diffuse distribution of 14.
gradient across the mitochondrial membrane, but are also to
be found at high concentrations in other lipophilic localities
of the cell. It is also known that 29 is a thiol-reactive species
which would be expected to be irreversibly concentrated in the
mitochondria. The fact that the pattern of distribution of 2, which
would be expected to be less lipophilic than 14, more resembles
that of the thiol-reactive species 29 than 14 suggests that it is
endowed with some property which ‘fixes’ it in mitochondria,
of which phosphorylation to the thiol-reactive entity seems a
likely possibility. Other than in the mitochondria, the non-uniform
pattern of distribution of 14 throughout the cytoplasm indicated
that the hydrophobic membranes of the endoplasmic reticulum
and the Golgi body were targeted. Remarkably, the apparently
very closely related simple cyclohexylamine-derived amide 15 was
extremely poorly taken up by MCF-7 cells. It was uncertain
from confocal microscopy experiments whether any uptake had
occurred at all. Some extremely faint luminescence was detected
from intact cells centred at shorter wavelength than is typical
for 3MLCT which was assigned to autofluorescence (Fig. 17).
The only intense rhenium-based (i.e. centred around 550 nm)
luminescence from any cells treated with 15 was seen in that small
population of dead or damaged cells found in the sample (Fig. 18).
There was no larger proportion of dead or damaged cells in this
sample than others, indicating that the association of lumophore
followed cell damage, rather than was responsible for it. This
association of rhenium complexes with biological constituents
of dead and damaged cells, which has been observed for a
number of the more lipophilic species, is thought to be a function
of physical or simple chemical interactions (e.g. hydrophobic
interactions) and completely different from (and simpler than)
cellular uptake processes. This dramatic difference in uptake
Fig. 14 Overlaid fluorescence and transmitted light images of 14 in
MCF-7 cells (scale bar = 14 mm).
and also accumulates in mitochondria, similar colocalisation
experiments were performed with 14 and TMRE to assess the
importance of thiol reactivity in the localisation of these species.
The colocalisation experiments (Fig. 15 and 16) demonstrated
that while some of the species had colocalised, indicating that
14 is present in mitochondria, its distribution was far more
widespread and diffuse than that of TMRE. In comparison to
the thiol reactive species 2 and 29, the distribution of 14 in
the cells was again more diffuse, supporting the theory that
thiol reactivity is important to the mitochondrial localisation
of 2 and 29. It is reasonable that cationic, lipophilic species
(such as 14) are taken into mitochondria, due to the potential
Fig. 15 Colocalisation of 14 and TMRE showing significant overlap.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3888–3901 | 3897
©

View Article Online
size or shape why e.g. 1 and 2 should differ so markedly in their
uptake, and so it is suggested that a difference in the interaction
between these species and the glycocalyx layer surrounding the
lipid bilayer prevents these complexes from diffusing into the
cytoplasm. That small differences in the structures of rhenium
complexes have profound effects upon their interactions with
the glycocalyx became apparent in experiments involving anionic
complexes (vide infra).
Anionic complexes
The distribution within cells of anionic complexes 22, 23 and 28
was studied by incubation with MCF-7 cells followed by confocal
microscopy. Interestingly, only the more polar species 22 and 23
showed appreciable association with the cells, while the more
lipophilic species 28 showed no evidence of uptake at all. Of
the two species which showed cell association, the more polar
hydroxylated species 23 showed the better association, while the
pattern of localisation was similar. Both species 22 and 23 showed
uptake in the outer layers surrounding the plasma membrane
region, presumably binding to cationic residues in the peripheral
glycoprotein layers (the ‘glycocalyx’). The diffuse pattern of
localisation can be observed from the Z-stacks of the cells
incubated with 23 as a diffuse halo of luminescence surrounding
the cell in all dimensions (Fig. 19). A 3D representation of the
localisation obtained by recombining the Z-stacks further reveals
the diffuse distribution around the exterior perimeter of the cell
(Fig. 20). Interestingly, localisation within the pellicular layers is
Fig. 17 Overlaid fluorescence and transmitted light images of 15 in
MCF-7 cells showing autofluorescence.
Fig. 18 Overlaid fluorescence and transmitted light images of 15 in
MCF-7 cells showing accumulation in dead cells (scale bar = 8 mm).
between such similar complexes 14 and 15 led to a question as
to whether this difference in uptake could possibly be rationalised
solely in terms of passive diffusion, or whether a more complex
explanation must be invoked. In order to clarify this, the simplest
possible analogue, 1, was incubated with MCF-7 and examined by
confocal microscopy. Again, as for 15, uptake was only observed
for damaged or dead cells and essentially no rhenium-based
luminescence was observed in healthy cells. The apparent uptake
in dead cells simply reflects the moderate lipophilicity of such
species which will associate with organic material in preference
to aqueous buffer if they are able to come into contact with
material such as lipoprotein membranes. It had been noted that
the highly lipophilic species 8–10 were taken up well by dead
or damaged cells, whereas the other complexes were only weakly
associated with these, and good uptake was seen only in apparently
healthy cells. This can be rationalised by proposing that the highly
lipophilic species partition so strongly into e.g. lipoprotein that
they show apparently good uptake whatever the state of the cell,
whereas the less lipophilic species show good uptake only if the
cells are healthy enough to maintain the potential to provide
a healthy cross-membrane gradient which assists the uptake of
these cationic complexes. That certain cationic complexes are not
taken up when apparently they should be of similar permeability
is more surprising. Liposome studies have previously confirmed
that species such as 15 are highly membrane-permeable, and it
would normally be assumed that the hydroxymethyl group of
2 would render it more polar and less permeable than 1. It is
therefore difficult to rationalise on grounds of lipophilicity, charge,
Fig. 19 Overlaid fluorescence and transmitted light images of 23 in
MCF-7 cells showing localisation in the exterior plasma membrane (scale
bar = 12 mm).
Fig. 20 3D reconstruction showing 23 in the plasma membrane of MCF-7
cell showing diffuse distribution.
3898 | Org. Biomol. Chem., 2010, 8, 3888–3901
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
◦
clearly non-uniform, with surface patches showing heterogeneous
distribution across and within the cellular boundary structure.
Both in the image obtained from reconstruction of the Z-scan
series and from individual images it is clear that there are hotspots
of localisation, the nature of which is yet to be identified.
different lumophore, final concentration 100 mg ml-1, at 4 C for
30 min. Cells were finally washed three times in phosphate buffer
saline (PBS, pH 7.2), harvested by centrifugation (5 min, 800 g)
and mounted on a slide for imaging. Preparations were viewed
using a Leica TCS SP2 AOBS confocal laser microscope using
¥63 or ¥100 objective, with excitation at 405 nm and detection at
520–570 nm.
Conclusions
Co-localisation experiments. MCF-7 adenocarcinoma cells
were harvested as indicated above, resuspended in HMEM and
incubated with both TMRE (final concentration 100 nM) and
rhenium complexes (final concentration 20 mg ml-1) for 1.5 h
at room temperature, before being washed three times in PBS,
mounted and viewed as above by confocal microscopy. Images
were recorded in each case with excitation at either 405 or 543 nm
and detection between 500–600 nm.
The localisation and uptake of rhenium lumophores is a complex
area, which will require more study to obtain a complete picture;
however, certain basic principles are becoming apparent. It
appears from this preliminary investigation that cationic species
are taken up well by passive diffusion, except for certain cases
in which it is proposed that the complex is prevented from
entering the lipid bilayer by interaction with the glycocalyx. Highly
lipophilic complexes appended with fatty-acid-derived units are
taken into dead cells, whereas the other species require a healthy
membrane potential to facilitate uptake. Localisation can in many
cases be explained in terms of simple chemical processes e.g.
ionic, hydrophobic, electrostatic hydrogen bonding interactions;
however, in some cases the picture is more complex. Simple
cationic lipophilic species localise in hydrophobic membranes
of the organelles or other cytoplasmic structures, while those
which are electrophilic, or can be rendered electrophilic by phos-
phorylation, localise in the mitochondria. Anionic polar species
associate with particular, as yet unidentified, layers associated
with the outer face of the plasma membrane, while anionic
lipophilic species show little or no uptake. There appears to be
little intrinsic toxicity associated with the rhenium tricarbonyl
bisimine pyridine unit over the duration of the observation as
judged by little deterioration of ultrastructure observable by
optical microscopy. Very little photobleaching was observed under
the conditions employed in mammalian cells whereas some was
observed in yeast cells, albeit still less than expected for an organic
fluorophore.
Chemical synthesis
General considerations. All starting materials, reagents and
solvents were purchased from commercial suppliers and used as
supplied unless otherwise stated. 1H-NMR and 13C-NMR spectra
were recorded at 400 and 100 MHz, respectively, on a Bruker
DPX and referenced to residual solvent peaks unless otherwise
reported. IR spectra were recorded on a Perkin Elmer 1600 FT IR
as thin films or nujol mulls and are reported in wavenumbers.
Mass spectra were recorded on a VG Fisons Platform II or
at the EPSRC national mass spectrometry service in Swansea
(HRMS). Elemental analyses were performed by Warwick An-
alytical Services (University of Warwick). All photophysical data
were obtained on a JobinYvon–Horiba Fluorolog spectrometer
fitted with a JY TBX picosecond photodetection module and a
Hamamatsu R5509-73 detector (cooled to -80 ◦C using a C9940
housing). Excitation and emission maxima are limited in accuracy
6
to the monochromator slit width of 5 nm. ReCl(bipy)(CO)3
[ReCl(bipy)(CO)3(CH3CN)]BF46 and 4, 8, 9, 19, 20, 21, 22 and 23
were prepared according to the literature6,10 methods. Numbering
for rhenium complexes given in Fig. 21.
Experimental
Cell culture and imaging
Yeast cell incubation. Saccharomyces cerevisiae IFO 0233
or Schizosaccharomyces pombe 972h- cells were harvested by
centrifugation (1000 g, 3 min), then washed twice in phosphate
buffer saline (PBS, pH 7.4), before resuspension in PBS, incubated
with lumophores (final concentration 100 mg ml-1) for 30 min in
0.3% yeast extract, 1% peptone and 1% glucose, rinsed in PBS
(¥3) and mounted on a slide. Preparations were viewed using a
Leica TCS SP2 AOBS confocal scanning laser microscope using a
¥63 or ¥100 objective, with excitation at 405 nm and detection at
520–570 nm.
Fig. 21 Numbering for rhenium complexes.
fac-[Re(bipy)(CO)3PyCH2OH][BF4] (2)
Human cell incubation. Human adenocarcinoma cells (MCF-
7), obtained from the European Collection of Cell Cultures,
Porton Down, Wiltshire, UK, were maintained in Hepes modified
minimum essential medium (HMEM) supplemented with 10%
foetal bovine serum, penicillin and streptomycin. Cells were
detached from the plastic flask using trypsin–EDTA solution,
and suspended in an excess volume of growth medium. The
homogenous cell suspension was then distributed into 1 ml
aliquots with each aliquot being subject to incubation with a
Re(bipy)(CO)3CH3CN](BF4)6 (255 mg, 0.4 mmol) and 3-
(hydroxymethyl)pyridine (0.45 ml, 4 mmol) were heated at reflux
in tetrahydrofuran (175 ml) under a nitrogen atmosphere for
1.5 h. After the solution had cooled, petroleum ether was added
dropwise to crystallize the product as a yellow solid (178 mg, 62%
yield), m.p. 200 ◦C. dH (CD3CN) 9.18 (2H, d, J = 5.5 Hz, CH(6)
bipy), 8.30 (2H, d, J = 8.1 Hz, CH(3) bipy), 8.2 (2H, m, CH(4)
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3888–3901 | 3899
©

View Article Online
bipy), 8.12 (1H, s, CH(2) Py), 8.04 (1H, d, J = 5.5 Hz, CH(6) Py),
7.68–7.78 (m, 3H, CH(4) Py, CH(5) bipy), 7.18 (1H, m, CH(5) Py),
4.38 (2H, s, CH2), 3.33 (1H, s, OH). dC (CD3CN) 195.4 (COeq.),
191.3 (COap.), 155.4 (C(3), py), 153.5 (C(6), bipy), 150.0 (C(6),
py), 149.5 (C(2), py), 140.8 (C(3), bipy), 140.6 (C(2), bipy), 137
(C(5), py), 127.8 (C(4), bipy), 125.8 (C(4), py), 124.3 (C(5), bipy),
59.6 (CH2). nmax (Nujol) 2032 s, 1915sb u (CO). m/z (ESI) 536.1
[MH]+, 427 [M–BF4–PyCH2OH]+. Elemental analysis, found C,
36.68; H, 2.30; N, 6.56. ReC19H15O4N3BF4 requires C, 36.65; H,
2.41; N, 6.75.
fac-[Re(1,10-bathophenanthroline
sulfonate)(CO)3PyCH2OCO(CH2)12CH3][H4N] (28)
Mystiric acid (18 mg, 0.08 mmol) was added to a mixture
of EDCI (26 mg, 0.1 mmol) and DMAP (4 mg, 0.03 mmol)
in dry dichloromethane (5 ml) at 0 ◦C. After 15 min stirring
under a nitrogen atmosphere, a solution of complex 236 (50 mg,
0.04 mmol) in dry DCM (3 ml) was added to the reaction mixture,
and stirring was continued at room temperature for 2 h. The
solvent was evaporated to dryness and the residue was washed
with ethyl acetate to remove the excess EDCI and DMAP. The
yellow solid was dissolved in the minimum amount of methanol
and it was passed through an ion exchange column (Amberlite
fac-[Re(CO)3(2,2-bipyridine)(cyclohexylnicotinate)][BF4] (14)
[Re(CO)3(2,2-bipy)NCCH3](BF4)6 (100 mg, 19 mmol) and cyclo-
hexylnicotinate (100 mg, 49 mmol) were dissolved in chloroform
(2 ml) and were heated to 60 ◦C for 12 h, and the product was
precipitated with diethyl ether (10 ml) to afford a yellow powder
(67 mg, 52%). dH (CD3CN) 9.17 (2H, d, J = 5.1 Hz, A), 8.85 (1H, s,
H), 8.76 (2H, d, J = 7.9 Hz, C), 8.38 (4H, m, D+E+G), 7.82 (2H,
t, J = 7.0 Hz, B), 7.59 (1H, td, J = 6.4 Hz,F), 5.02 (1H, m, I) 1.92 -
1.36 (10H, m, J+K+L); dC (CD3CN) 162.2, 155.8, 155.6, 153.4,
151.4, 141.0, 139.7, 129.3, 128.5, 126.6, 124.5, 74.4, 30.6, 24.9, 22.7.
nmax (CH3CN) 2022 (CO), 1917 (CO), 1635 (COOR). Theoretical
isotope pattern 630.12 (58%), 631.12 (17%), 632.12 (100%), 6.33.12
(29%), 634.13 (5%). Observed isotope pattern 630.16 (53%), 631.16
(17%), 632.16 (100%), 633.17 (28%), 634.17 (5%). Crystal data for
14: C26H12.50BF4N3O4Re, M = 703.90, yellow needle, 0.20 ¥ 0.20 ¥
+
IR120 H resin as NH4 form) which had previously been washed
with aqueous ammonia eluting with methanol–water (1 : 1). The
yellow coloured fraction which eluted immediately was collected,
evaporated and dried under vacuum to give 28 as a yellow solid
◦
(27 mg, 62%), m.p. 167 C dH (CD3OD) 9.71 (2H, m CH(2,9)),
8.45 (1H, m, CH(6¢)), 8.32 (1H, s, CH(2¢)), 8.10 - 7.97 (8H,
m, 4 ¥ CH(o), CH(3,5,6,8)) 7.78 (1H, dd, J = 1.5, 7.8 Hz
CH(4¢)), 7.69 - 7.70 (4H, m, 4 ¥ CH(m)) 7.25 (1H, t, J =
7.8, 5.7 Hz, CH(5¢)), 4.83 (2H, s, PyCH2O), 2.03 (2H, t, J =
7.4 Hz, CH(10¢)), 1.25 - 0.85 (22H, m, CH(11¢21¢)), 0.79 (2H, m,
CH3(22)). dC (CD3OD) 197.1 (COeq), 192.9 (COap), 174.9 (CO(9)),
154.1, 152.6, 152.5 (2C), 152.2, 147.0, 145.0, 143.7, 140.4, 138.1,
131.9, 130.9, 130.6, 130.3, 129.8, 129.6, 128.1, 127.9, 127.6 (2C),
127.4 (3C), 127.2, 127.1, 126.8 (2C), 126.6, 126.2, 62.6 (PyCH2O),
33.9 (C(10)), 32.4 (C(11)), 30.0, 29.9 (2C), 29.7 (2C), 29.4, 29.1
(2C), 25.1 (C(20)), 23.0 (C(21)), 13.7 (C(22)). nmax (CH3CN) 2030,
1826, 1738 (CO); m/z (ESI) 1080.3 [MH2+NH4]+, 761.1 [MH2–
PyCH2OCO(CH2)12CH3–NH4]+. Theoretical isotope pattern [M]-
1078.2 (55%), 1079.2 (30%), 1080.2 (100%), 1081.2 (58%), 1082.2
(38%), 1083.2 (6%). Actual isotope pattern [M]- 1078.4 (42%),
1079.4 (45%), 1080.3 (100%), 1081.3 (60%), 1082.4 (21%), 1083.3
(2%) found C 49.07 H 4.94 N 5.08 C45H51N4O11ReS2·H2O requires
C 49.48 H 4.89 N 5.13%.
0.15 mm3, orthorhombic, space group Pnma, a = 16.7420(3), b =
3
˚
˚
12.4980(4), c = 12.0760(4) A, V = 2526.80(12) A , Z = 4, Dc =
-3
˚
1.850 g cm , F000 = 1350, Mo-Ka radiation, l = 0.71073 A, T =
150(2) K, 2qmax = 55.0◦, 9904 reflections collected, 3019 unique
(Rint = 0.0561). Final GooF = 1.341, R1 = 0.0474, wR2 = 0.1363,
R indices based on 3019 reflections with I > 2s(I) (refinement on
F2), 212 parameters, 0 restraints
fac-[Re(CO)3(2,2-bipyridine)(cyclohexylnicotinamide)][OTf] (15)
[Re(CO)3(2,2-bipyridine)(OTf)] (60 mg, 12 mmol) and cyclohexyl-
nicotinamide (24 mg, 12 mmol) were dissolved in acetonitrile
(2 ml). The solution with heated to 60 ◦C for 12 h and the product
was precipitated with diethyl ether (10 ml) to afford a yellow
powder (59 mg, 81%). dH (CD3CN) 9.27 (2H, d, J = 5.1 Hz,
A), 8.51 (1H, d, J = 1.7 Hz, E), 7.84 (2H, td, J = 1.1 Hz, J¢ =
4.5 Hz, B), 7.40 (1H, dd, J = 5.7 Hz, J¢ = 2.2 Hz, F), 7.10 (1H, d,
J = 7.2 Hz, I) 3.75 (1H, m, J), 1.73 (5H, m, K), 1.28 (5H, m, L). dC
(CD3CN) 195.4 (CO), 161.6 (CONH), 155.5, 153.5, 150.8, 140.9,
137.4, 128.5, 125.9, 124.5, 48.9, 31.8, 24.9, 24.5. nmax (CH3CN)
2036 (CO trans to Py), 1933 (CO), 1635 (CONH). Theoretical
isotope pattern 629.13 (61%) 630.14 (22%) 631.14 (100%) 632.14
(32%) 633.14 (11%). Observed isotope pattern 629.14 (36%) 630.15
(11%) 631.13 (100%) 632.14 (27%) 633.13 (81%). Crystal data
for 15: C26H24F3N4O7ReS, M = 779.75, 0.20 ¥ 0.20 ¥ 0.20 mm3,
monoclinic, space group P21/c (No. 14), a = 12.1380(3), b =
fac-[Re(bipy)(CO)3PyCH2Cl][PF6] (29)
A solution of fac-[Re(bipy)(CO)3PyCH2OH](BF4) and 2 (150 mg,
0.2 mmol) in thionyl chloride (3 ml) was stirred under a nitrogen
atmosphere overnight. The mixture was cooled in an ice bath and a
saturated aqueous ammonium hexafluorophosphate solution was
added dropwise over 10 min, with extremely rapid stirring until
a yellow solid precipitate formed. The precipitate was separated
from the mixture of solvents by filtration and washed several times
with water to give 29 as a yellow solid (132 mg, 78% yield), m.p.
180 ◦C. dH (CD3CN) 9.18 (2H, d, J = 5.4 Hz, CH(6) bipy),
8.42 (2H, d, J = 6.4 Hz, CH(3) bipy) 8.32 (3H, m, CH(2) Py,
CH(4) bipy), 8.20 (1H, d, J = 5.5 Hz, CH(6) Py), 7.91 (1H, d,
J = 4.5 Hz, CH(4) Py), 7.82 (2H, m, CH(5) bipy), 7.33 (1H, m,
CH(5) Py), 4.58 (2H, s, CH2). dC (CD3CN) 195.0 (COeq.), 191.2
(COap.) 155.2 (C(2), bipy), 153.4 (C(6), bipy), 151.0, 151.1 (C(2),
Py, C(6), Py), 140.7 (C(4), bipy), 139.0 (C(4), Py), 136.6 (C(3),
Py), 128.3 (C(3), bipy), 126.1 (C(5), Py), 124.2 (2C(5), bipy), 41.1
(CH2). nmax (Nujol): 2027, 1922, 1895 (CO). m/z (ESI) 553.9 [MH]+
and 427.0 [M–PF6–PyCH2Cl]+. Theoretical isotope pattern 552.0
(52%), 553.0 (11%), 554.0 (100%), 550.0 (22%), 556.0 (31%) and
557.0 (8%); observed isotope pattern 551.9 (50%), 552.9 (11%),
◦
˚
12.1560(3), c = 19.4130(5) A, b = 105.8630(10) , V = 2755.30(12)
3
-3
˚
A , Z = 4, Dc = 1.880 g cm , F000 = 1528, Mo-Ka radiation,
l = 0.71073 A, T = 293(2) K, 2qmax = 61.0◦, 11446 reflections
˚
collected, 7278 unique (Rint = 0.0300). Final GooF = 1.071, R1 =
0.0373, wR2 = 0.0698, R indices based on 6007 reflections with
I > 2s(I) (refinement on F2), 379 parameters, 0 restraints. Lp and
absorption corrections applied, m = 4.559 mm-1.
3900 | Org. Biomol. Chem., 2010, 8, 3888–3901
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
553.9 (100%), 554.9 (22%), 556.0 (30%) and 557.0 (5%). HRMS
5 (a) C. P. Montgomery, B. S. Murray, E. J. New, R. Pal and D. Parker,
Acc. Chem. Res., 2009, 42, 925; (b) E. J. New, D. Parker, D. G. Smith
and J. W. Walton, Curr. Opin. Chem. Biol., 2010, 14, 238; (c) G-L. Law,
R. Pal, L. O. Palsson, D. Parker and K-L. Wong, Chem. Commun.,
2009, 7321.
(ESI) calculated [M–PF6]+ = 552.0248; measured [M–PF6]+
=
552.0252. ReC19H14O3N3ClPF6 requires C, 32.65; H, 2.02; N, 6.01;
found C, 32.46; H, 1.92; N, 5.93%.
6 A. J. Amoroso, M. P. Coogan, J. E. Dunne, V. Ferna´ndez-Moreira, J. B.
Hess, A. J. Hayes, D. Lloyd, C. Millet, S. J. A. Pope and C. Williams,
Chem. Commun., 2007, 3066.
7 S. James, K. P. Maresca, J. W. Babich, J. F. Valliant, L. Doering and J.
Zubieta, Bioconjugate Chem., 2006, 17, 590.
8 K. K.-W. Lo, M.-W. Louie, K.-S. Sze and J. S.-Y. Lau, Inorg. Chem.,
2008, 47, 602.
9 N. Viola-Villegas, A. E. Rabideau, M. Bartholom, J. Zubieta and R. P.
Doyle, J. Med. Chem., 2009, 52, 5253.
Acknowledgements
We thank the EPSRC for support (EP/D080401/1;
EP/H501118/1) and the EPSRC National Mass Spectrometry
Service Centre Swansea for assistance with obtaining mass ions
on the labile rhenium complexes.
10 A. J. Amoroso, M. P. Coogan, J. E. Dunne, V. Ferna´ndez-Moreira, J. B.
Hess, A. J. Hayes, D. Lloyd, C. Millet, S. J. A. Pope and C. Williams,
New J. Chem., 2008, 32, 1097.
11 M-W. Louie, H-W. Liu, M. H-C. Lam, T-C. Lau and K. K-W. Lo,
Organometallics, 2009, 28, 4297.
12 D. J. Stufkens, A. Vlcek and Jr, Coord. Chem. Rev., 1998, 177, 127.
13 M. S. Wrighton and D. L. Morse, J. Am. Chem. Soc., 1974, 96, 998.
14 R. Y. Tsien, Nature, 1981, 290, 527.
15 C. O. Badgett and C. F. Woodward, J. Am. Chem. Soc., 1947, 69, 2907.
16 B. Singh, Tetrahedron Lett., 1971, 12, 321.
17 Commercially available, but prepared as in D. Sharma, B. Narasimhan,
P. Kumar and A. Jalbout, Eur. J. Med. Chem., 2009, 44, 1119.
18 I. Costa, M. Montalti, P. Piersandro, A. Perotti, L. Prodi and N.
Zaccheroni, J. Organomet. Chem., 2000, 593, 267.
19 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008,
64, 112.
Notes and references
1 (a) S. Ishida, J. Lee, D. J. Thiele and I. Herskowitz, Proc. Natl.
Acad. Sci. U. S. A., 2002, 99, 14298; (b) A. Younes, J. A. Songadele,
J. Maublant, E. Platts, R. Pickett and A. Veyre, J. Nucl. Cardiol.,
1995, 2, 327; (c) V. Ferna´ndez-Moreira, F. L. Thorp-Greenwood and
M. P. Coogan, Chem. Commun., 2010, 46, 186; (d) M. S. Lowry,
W. R. Hudson, R. A. Pascal and S. Bernhard, J. Am. Chem. Soc.,
2004, 126, 14129; (e) D. J. Stufkens and A. Vlcek, Jr, Coord. Chem.
Rev., 1998, 177, 127; (f) A. Juris, V. Balzani, F. Barigellatti, S.
Campagna, P. Belser and A. von Zelewsky, Coord. Chem. Rev., 1988, 84,
85.
2 Handbook of Biological Confocal Microscopy, ed. J. B. Pawlet, Springer,
New York, 2006.
3 C. A. Puckett and J. K. Barton, J. Am. Chem. Soc., 2007, 129, 46.
4 K. K.-W. Lo, W.-K. Hui, C.-K. Chung, K. H.-K. Tsang, D. C.-
M. Ng, N. Zhu and K.-K. Cheung, Coord. Chem. Rev., 2005, 249,
1434.
20 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
21 M. P. Coogan, J. B. Court, V. L. Gray, A. J. Hayes, S. H. Lloyd, C. O.
Millet, S. J. A. Pope and D. Lloyd, Photochem. Photobiol. Sci., 2010, 9,
103.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 3888–3901 | 3901
©