Synthesis and In Vitro Biological Evaluation of a Second-Generation Multimodal Water-Soluble Porphyrin-RAPTA Conjugate for the Dual-Therapy of Cancers

In this study we report the synthesis and biological evaluation of a novel cationic porphyrin-[Ru(η 6 -arene)(C 2 O 4 )PTA] (RAPTA) conjugate with potential as a multimodal dual-therapeutic agent. In the absence of high intensity light, relative to untreated cells our conjugate resulted in a 83% decrease in viable human adenocarcinoma cells at a concentration of 10 μM , which is significantly more active than the 57% decrease achieved with the same concentration of the unconjugated RAPTA complex alone. With a light dose of 20 J cm -2 (400 – 1200 nm) a reduction of 98% of viable cells was observed for the same concentration of conjugate. The conjugate is internalized by HT-29 cancer cells as proven by ICP-MS analysis and fluorescence microscopy: the latter result suggesting that the conjugate has applications as a multimodal agent by acting as a fluorophore to obtain in vivo biodistribution data. Furthermore, the conjugate has an excellent relative singlet oxygen quantum yield, and the tetrapyrollic unit was found to be photostable under irradiation by either white light or red light.


INTRODUCTION
Increasingly, novel metal-based compounds are being investigated as cytotoxic agents. Of key significance in this area are ruthenium-based organometallic compounds of which the ruthenium(II) arene 1,3,5-triaza-7-phosphaadamantane (PTA) family, or so-called "RAPTA" complexes, have gained interest in the past two decades as anti-angiogenic or anti-metastatic agents. 1,2 The distinct, but promiscuous, biomacromolecular interactions of many ruthenium-based compounds offers promise in overcoming limitations associated with traditional platinum-based and organic-based chemotherapeutics. 3,4 For example, platinum-based therapies are associated with more than 40 side effects 5 and many cancers become resistant to platinum via mechanisms including: decline in adduct levels, reaction with intracellular reducing agents, binding to transporters, reduced endocytosis and cross-linking repair. 6 The unique chemistry of many ruthenium-based metallodrugs may enable these issues to be circumvented. Like clinically approved platinum-based chemotherapeutics, the majority of ruthenium-based metallodrugs operate via mechanisms of action dominated by the aquation of the metal center followed by metalation of intra-or extracellular biomacromolecular targets. 7 However, the elucidation of the in vivo biodistribution of ruthenium-based metallodrugs, such as the RAPTA family, is challenging 3 and has only been achieved via invasive biopsies followed by elemental analysis such as ICP-MS techniques or by analyzing the radioactivity of resected organs (when a 103 Ru-labelled compound was utilized). 8 Porphyrins are well known for their fluorescent properties, and also an ability to generate cytotoxic reactive oxygen species (ROS) by energy or electron transfer from the excited triplet state, upon interaction with light. 9 Porphyrins have previously been used to generate theranostic agents for combined cancer therapy and imaging. 10 Photodynamic therapy (PDT) is a minimally invasive technique requiring a photosensitizer (PS), light and molecular oxygen. Absorption of light of the correct wavelength promotes the porphyrin into an electronically excited singlet state which undergoes Laporte forbidden intersystem crossing to the triplet state. 11 From there, two main photochemical processes can occur to produce cytotoxic species. The Type I mechanism creates ROS through immediate electron transfer to surrounding substrates, while the Type II mechanism is dominated by interaction with molecular oxygen producing singlet oxygen. 11 An added and well documented, yet controversial, 12 benefit of conjugating to porphyrins is they are able to utilize the enhanced permeation and retention (EPR) effect to passively accumulate and be retained in tumors, this has been proven with many substrates appended to porphyrins which has been demonstrated in both murine models and humans. 13,14,15 It is well understood that cancers are far less resilient when combating two therapies at the same time, and, the combination of PDT with the anti-metastatic/cytotoxic effects of a ruthenium(II) metallodrug represents an underexplored therapeutic combination. While the conjugation of 4 porphyrins with pyridyl groups to Ru-(η 6 -arene) complexes has been carried out previously, four ruthenium moieties were required to obtain IC50 values in the low micro molar region. 16 A different study investigated the photophysical properties of first generation ruthenium-porphyrin conjugates using 2,2'-bipyridine groups on the porphyrin macrocycle to act as ligands to the metal center. 17 Cytotoxicity was observed in the low micromolar range albeit with four ruthenium metal centers appended to the porphyrin. There is clear scope to improve the activity of such conjugates with a concomitant reduction in the number of appended Ru-(η 6 -arene) groups. Furthermore, these Ru-(η 6 -arene)-porphyrin conjugates possessed limited aqueous solubility as a result of the highly lipophilic porphyrin components and lipophilic bifunctional linkers, highlighted by the use of organic solvent to acquire spectral data for these compounds. 16 Clearly, further development of Ru-(η 6 -arene)-porphyrin conjugates is required in order to produce a clinically relevant and economically viable chemotherapeutic possessing a single metal center that is sufficiently hydrophilic. The employment of hydrophilic porphyrins bearing solubilizing methyl pyridinium groups and a hydrophilic bifunctional spacer is obviously advantageous. Furthermore, cationic N-methylpyridyl porphyrins possess a natural ability to passively accumulate in cancer cells and neoplastic tissues. 18 It has been well documented elsewhere that cationic N-methylpyridinium porphyrins are internalized by cancer cells, often localizing to the mitochondria, 19,20 and lysosomes. 21,22 Therefore, our studies have focused on a cationic porphyrin covalently tethered to a single [Ru(η 6 -arene)(C2O4)PTA] complex with an established reactivity profile.

5
As previously discussed, several attempts to synthesise tetranuclear Ru(II)-porphyrin conjugates have been reported, however, in these previous examples the ruthenium ion was coordinated to the porphyrin via N-heterocycle nitrogen atoms. We have investigated the tethering of the [Ru(η 6arene)(C2O4)PTA] complex by the arene ring, thus enabling the known reactivity of the ruthenium complex to be unimpeded by its conjugation to the porphyrin. As discussed, the mechanism of action of these [Ru(η 6 -arene)(C2O4)PTA] complexes is exerted through coordination of biomacromolecular targets to the metal centre in the extra-and intracellular environment.

General information
All reagents were reagent grade and purchased from Fluorochem, Alfa Aesar, and Sigma Aldrich and were used as received, unless stated otherwise. Dry solvents were obtained by drying over activated 4Å or 3Å molecular sieves or anhydrous metal salts as stated for at least 24 h. Bulk solvents were removed at an appropriate temperature using a Buchi rotavapor R-210 at reduced pressure with a Vacuubrand PC101 membrane pump. Solids were dried overnight or until reaching constant mass in a vacuum oven (Buchi glass oven B-585) at 40 o C. Reaction progress and purity were analysed by TLC using Fluka analytical TLC plates (0.2 mm thickness and 10 cm length; silica 60 Å). TLC was visualized using UV irradiation. (Spectroline; ENF-260C/FBE). Products were purified by column chromatography using an appropriate mobile phase as stated using a glass chromatography column and a stationary phase of silica gel obtained from Fluorochem; LC60Å 35-70 μm. Progress of column chromatography was visualized using UV irradiation, tracked against relevant TLC plates. NMR were recorded on a JEOL ECZ 400 spectrometer at 400 MHz 6 for proton NMR and at 100.5 MHz for carbon NMR. All samples contained an internal standard of tetramethylsilane (TMS) in deuterated solvent as stated. NMR spectra splitting patterns were designated as s (singlet), d (doublet), m (multiplet) or br s (broad singlet), dd (doublet of doublets) as appropriate. All chemicals shifts, δ, for proton and carbon NMR spectra were quoted as parts per million, ppm. J values are quoted in Hz unless stated otherwise. Mass spectrometry data was obtained by the EPSRC National Mass Spectrometry Service at Swansea. Combustion elemental analysis was performed with 15 mg of compound analysed in a Carlo Erba EA1108 CHN Fischer instrument with a thermal conductivity detector for determination of CHN; analysis quoted in %.
HPLC analyses were carried out on an Agilent series 1200 HPLC system. Separations were carried out on an ACE-5 C-18 column (4.6 mm x 250 mm). All solvents were HPLC grade containing 0.1% trifluoroacetic acid (TFA) in both eluents. Analytical RP-HPLC (ACE-5, C-18 column, 4.6 mm x 250 mm) with HPLC grade methanol with 0.1% TFA and HPLC grade water with 0.1% TFA as eluents A and B respectively. The flow rate was 1 mL per minute. Initially, an isocratic gradient of 5% rising to 95% (20 mins, isocratic 95% (3 mins), falling to 5% (2 mins) and remaining isocratic 5% A (5 mins). UV-Vis data were obtained from a Varian Cary Bio50 spectrometer with aliquots of standard solutions of known concentration in an optical glass cuvette (path length = 1 cm) of optical clarity 300-900 nm. Data was obtained from solutions of 20 μL additions to 2 mL of solvent from a 10 mL stock solution of concentration 1x10 -2 mol dm -3 . The molar extinction coefficient was calculated at the λ max of the Soret band (400-450 nm) by the application of the Beer -Lambert law. Fluorescence data were obtained from a Varian Cary Eclipse spectrometer with aliquots of standard solutions of known concentration in an optical glass cuvette (path length = 1 cm) of optical clarity 300-900 nm. Data was obtained from solutions of 7 20 μL additions to 2 mL of solvent from a 10 mL stock solution of concentration 1x10 -2 mol dm -3 .

Photostability
Porphyrin photostability experiments were carried out in a quartz fluorescence cuvette (46 x 12.5 x 12.5 mm), H2O (2 mL) was added followed by photosensitizer (100 μL, 150 μM). The cuvette was irradiated at 298 K under continuous irradiation from a 600 mW Paterson Xenon short arc lamp equipped with a band pass filter (617 -651 nm). The intensity of the light from the same distance was measured with a Macam R203 Radiometer (1035.8 W m -2 ). Measurements were taken every 60 seconds with a Varian Cary Bio 50 UV-Vis photospectrometer. Experiments were carried out in triplicate (n=3). RAPTA conjugate photostability experiments were carried out in standard NMR tubes by irradiating samples (5 mg) in D2O (0.7 ml). The NMR tube was irradiated at 298 K under continuous irradiation from an Oriel 1000W QTH white light source (400-1200 nm) in 20 J cm -2 increments followed by 31 P{ 1 H} NMR analysis.

Cell culture
Human colorectal adenocarcinoma (HT-29) cells were cultured in an incubator with a humidified 5% CO2 atmosphere at 37 o C in T25 flasks until reaching 60-75% confluence. RPMI complete media (Life Science Productions) substituted with 1% L-glutamine and 5% foetal bovine serum (Life Science Productions) was used as culture media.

MTT cell viability assay
The cell viability was determined using MTT (3- The samples were diluted into concentrated nitric acid and water (Elga Purelab Flex). ICP-MS was carried out using an Agilent 7500cx instrument. The cool gas was argon (BOC Cryospeed) with a flow rate of 15 L/min, the auxiliary flow rate was 0.2 L/mins and the nebuliser was 0.8 L/min. The detector (Perkin Elmer) was peltier cooled to -43 o C. The nebuliser used was a high solids modified V-groove PEEK unit (Perkin Elmer). The Ru standards used were certified 1000 ppm (Romil, UK), which were diluted with 2% v/v nitric acid (Spa grade, Romil, UK) and water (Purelab).

Statistical analysis
All biological experiments were carried out in triplicate. All data are given as the mean values, error bars are presented as standard deviations (X±SD) of three independent plated experiments performed in triplicate (n=3). Data were plotted on GraphPad Prism 7.0 using recommended posthoc statistical testing.

Complex 2 MS analysis
Accurate mass measurements were performed at the University of Hull using a Bruker Maxis Impact QqTOF MSMS. Before mass measurement the instrument was calibrated against sodium formate over the range 90 to 1550 Da. Resolution used was typically 45000. Samples (as solutions in methanol, 10 -5 M) were injected into a solvent stream from a syringe pump at 3 μl min -1 via a 5 μl loop injector. The data was then internally mass measured against an internal calibrant peak The solvent was eluted to allow the yellow suspension to settle onto the silica gel. The column was then eluted with acetone (200 mL to remove DMSO) then MeOH (300 mL) followed by MeOH/H2O (500 ml, 9:1). The product eluted as a yellow solution -this was filtered then dried    (1678), 651 nm (58 mol -1 dm 3 cm -1 ).      (2-(2-Acetimideethoxy) The collected solid was then dissolved in H2O (1 mL) followed by the addition of acetone (9 mL), the mixture was vigorously shaken then allowed to stand for 30 min, followed by centrifugation to collect the solid. This process was repeated followed by washing of the collected solid in acetone (3 x 10 mL) and collection by centrifugation.

Synthesis of an Amine Reactive RAPTA Complex
The synthesis of the [Ru(η 6 -arene)(C2O4)PTA] complexes was achieved using previously reported methodology. [24] The known complex 1, bearing a pendant carboxylic acid functionality, was chosen as it provides a route by which it can be conjugated to a porphyrin bearing a pendant amine group. Additionally, the use of the oxalate ligand within these complexes confers good aqueous solubility [25] whilst acting as a protecting group in amide-forming reactions. (Scheme 1: 1, 2).

Synthesis of a Carboxylic Acid Reactive Porphyrin
The synthesis of 5 was carried out by adopting the method described by Yap et al. 23

RAPTA-Porphyrin Conjugation
The synthesis of the conjugate was carried out via peptide coupling utilizing DMSO as solvent, and TBTU and DIPEA as coupling reagents (Scheme 3: 9). The conjugate was purified firstly by liquid-liquid extraction and isolated by centrifugation, followed by counter-ion exchange to yield the more water-soluble trichloride salt of 9. The novel RAPTA-porphyrin conjugate (9) was found to be highly soluble in DMSO, water, and biological media. Interestingly, the NMR spectrum of 9 21 recorded in DMSO-d6 revealed signals corresponding to the η 6 -arene protons in the expected region at ~5.8 ppm. However, in D2O these signals are shifted downfield to ~7.3. The origins of these differences are likely to be related to different conformations of the conjugate in these solvents.

UV-Vis and Fluorescence
Ideally, a photodynamic sensitizer should absorb in the red-NIR region for deep tissue penetration, however, the use of optical fibers which can be inserted into the body can negate this requirement somewhat. Conjugation of 2 to the porphyrin did not deleteriously alter the porphyrins optoelectronic properties. The results of the UV-Vis absorption studies for the porphyrins 8 and the conjugate 9 are shown in Figure S1 (supplementary information) and Table 1 recorded at room temperature in water. The conjugate was maintained as the free-base rather than undergoing complexation with a heavy-metal such as Pt 2+ or Pd 2+ in order to preserve the Q-bands in order to profit from the favorable red/NIR absorption band at 650 nm 22 The Soret bands (ca. 420 nm) were found to have the highest molar absorptivity coefficient, while the Q-bands had the lowest. The highest wavelength absorptions were at ca. 650 nm. As expected, the tabulated results are extremely similar for 8 and 9, suggesting no, or minimal, electronic interaction between the porphyrin and RAPTA components. Fluorescence spectra were recorded and, once again, the presence of the RAPTA group did not deleteriously affect the normalized fluorescence spectra, suggesting no significant excited state perturbations occur upon conjugation ( Figure S2; supplementary information).

Relative Singlet Oxygen Quantum Yield
The relative singlet oxygen quantum yield is a measure of the efficiency of a photosensitizer to convert triplet state molecular oxygen into highly cytotoxic singlet oxygen. 9,10-anthracenediylbis(methylene)dimalonic acid, (ABDA), is commonly used as a diagnostic tool in which photooxidation of the central arene ring takes place when in the presence of singlet oxygen yielding an endoperoxide and a reduction in UV-Vis absorption spectra relative to a control. We followed the method according to Senge et al. 26 In this case, the water soluble photosensitizer meso-tetra(Nmethyl-4-pyridyl)porphyrin (TMePyP) was used as a standard during singlet oxygen quantum yield experiments. The results of the irradiation of ABDA (absorption monitored at 380 nm) with time are shown in Figure S3 (supplementary information). Qualitatively, a decrease in the absorption maxima corresponds to the production of singlet oxygen. It can clearly be seen that a decrease in the absorbance signal of ABDA can be seen over time when irradiated with red light (617-651 nm, 1035.8 Wm -2 ). This is an indication that singlet oxygen is being produced.
Irradiation of ABDA in the absence of a photosensitizer did not produce any statistically significant indication singlet oxygen. The relative singlet oxygen quantum yields of 8 and 9 were then calculated relative to that of TMePyP. 8 gave a value of 0.96 (relative to a normalized value of 1 for TMePyP), however 9 gave a value of 1.57, this value could be attributed to steric hindrance of the ruthenium component minimizing porphyrin aggregation (Table 3). It is well understood that even water-soluble cationic porphyrins aggregate and π-π stack in solution which limits their relative singlet oxygen quantum yields, possibly explaining the differential between 8 and 9.

Photostability
Photostability is a major concern for any photosensitizer/fluorophore, as small molecule fluorophores often experience irreversible photobleaching upon continuous illumination with light, therefore, limiting their use in biomedical applications. The photostability of 8 and 9 was investigated with Rose Bengal being utilized as a non-porphyrin control for samples irradiated with white light, and TMePyP being used as a control for samples irradiated with red light. When irradiated continuously for 60 minutes in H2O, 8, 9, and TMePyP were all found to behave similarly ( Figure 1). Previous studies have also demonstrated that the tetrapyrrolic core has been found to be appreciably photostable. 27 None of the samples were found to have a normalized absorbance of less than 95% after irradiation with red light for 60 minutes. When irradiated continuously with white light for 10 minutes in H2O, 8 and 9 were both found to be appreciably photostable with normalized absorbance values of 88.65±0.03% (n=3) and 86.60±0.01% (n=3) respectively, which did not deviate significantly from the normalized absorbance when irradiated continuously for 60 minutes with red light. Meanwhile, the commercially available Rose Bengal control was found to fully photobleach to a normalized absorbance value of 1.58±0.01% (n=3) after continuous irradiation with white light for 10 minutes. Therefore, compared to the commercially available photosensitizer/fluorophore, the porphyrin-RAPTA conjugate can be said to be relatively photostable under these specific conditions. In order to probe the photostability of the [Ru(η 6 -arene)(C2O4)PTA] complex we took a known mass of 9 into D2O (0.7 mL) in an NMR tube and recorded the 31 P{ 1 H} NMR and 1 H NMR spectra at room temperature. The sample was protected from light, then irradiated sequentially in 20 J cm -2 increments of white light (400-1200 nm) and the NMR spectrum was recorded after each irradiation. With regards to the 31 P{ 1 H} NMR spectra, the relative intensity of the peak originating from the bound PTA ligand did not increase or decrease after irradiation with white light, furthermore, no other peaks were observed even at high doses of 60 J cm -2 as seen in Figure S4 (supplementary information). Again, the 1 H NMR remained unchanged. These data suggests that no significant changes to the Ru(II) ligands were occurring in solution after irradiation with white light. Therefore, we are confident that this unique conjugate consists of not only a photostable tetrapyrrole, but, also a stable RAPTA moiety. It is interesting to note that these results contrast with those reported in related systems 28,29 where ligand loss, and therefore activation of the

In Vitro Anticancer Activity
In order to determine the biological anticancer efficacy of 2, 8, and 9, we chose to examine these compounds in the human Caucasian colorectal adenocarcinoma (HT-29) cell line, chosen as a model of colorectal cancers which are the third most common cancers worldwide, where 20% of cases suffer from metastases and 56% of patients with colorectal cancers die from this disease. [30] Cells conditions of this experiment that the ruthenium centre will be activated through oxalato ligand exchange, as reported for an isostructural ruthenium centre in a previous study. 24 The loss of the bidentate oxalato ligand, that acts as a protecting group to prevent ruthenium-centred reactivity, 'activates' the complex to a form able to metallate intracellular biomacromolecules that results in the observed anticancer activity. 24 It is important to note that these data have been obtained after a maximum incubation period of 48 h, while similar values were obtained with four times the amount of ruthenium atoms whilst requiring a 72 h incubation period. 16

Fluorescence Imaging
As well as being known for their photosensitizing properties, porphyrins are in their own right excellent fluorophores which allows monitoring of their in vitro and in vivo biodistribution and cell-uptake by fluorescence imaging. The conjugate, 9, is intrinsically fluorescent, meaning that 29 further synthetic functionalization was not required. Cellular uptake studies were carried out with (9) and control compound (8) according to the methods of Boyle et al to determine whether or not internalization was occuring. 14 Both 10 µM of 8 and 9 were incubated with HT-29 for 24 h. Diffuse fluorescence could clearly be observed emanating from within the sub-cellular environment by fluorescence microscopy, as seen in Figure S5 (supplementary information). While the compounds clearly were being internalized, only diffuse fluorescence from 8 and 9 could be observed suggesting that at the dose of 10 µM the conjugate is not specifically localized within any given cellular organelle. In the case of both 8 and 9, brightfield micrographs also provide qualitative morphological evidence of porphyrin photosensitization indicated by blebbing of cells which is indicative of early onset apoptosis due to activation of the photosensitizer from the microscope light source.

ICP-MS
Inductively coupled plasma mass spectrometry was used as a highly sensitive technique to quantify the relative concentrations of 101 Ru in HT-29 cells that had been incubated with compounds 2 or 9 for 48 h, followed by washing, harvesting, and re-suspending the cells in PBS.
The protocol was developed support whether 2 or 9 had been internalized by HT-29 cells ( Figure   3). We predicted that due to the natural ability of N-methylpyridinium functionalized porphyrin to be internalized by cancer cells that the relative concentrations of 101 Ru in the HT-29 cells should be higher due to the ability of the porphyrin to chaperone the RAPTA metallodrug into the cell through the predominately negatively charged cell membrane. The results unequivocally indicated that the concentration of 101 Ru was significantly higher (p<0.01) for cells that were treated with 10 μM of 9 compared to 10 μM of 2 and compared to the cell only sample (vehicle controlled).
Combined with the fluorescence data described above the porphyrin-RAPTA conjugate is significantly more internalized by the HT-29 cells, relative to unconjugated 2, even at an active dose.

Conclusion
To conclude, we have successfully synthesized a second-generation water-soluble porphyrin-RAPTA conjugate and evaluated its photochemical and biological properties. The photostability of the porphyrin component of the conjugate was found to be good, regardless of whether it was irradiated with white light or red light, and the conjugate retained its ground and excited state properties relative to the non-conjugated porphyrin. Furthermore, it exhibited a relatively higher singlet oxygen quantum yield compared to a known porphyrin standard. Biological evaluation has allowed us to determine that the conjugate is a viable photosensitizer in its own right, but it also operates as an anticancer agent by controlling cell proliferation in the 'dark' at low concentrations (10 μM), while 2 alone had an IC50 >500 μM. In fact, a 2.5-fold decrease in cell viability was observed when compared to 2 alone at the same concentration in the 'dark'. Furthermore, we obtained this data at shorter incubation times (48 h), while previous work on porphyrin-ruthenium conjugates, and RAPTA compounds, have required 72 h incubations in order to achieve acceptable IC values. Fluorescence microscopy and ICP-MS has allowed us to study cellular uptake, from which we attribute the enhanced anticancer efficacy of this conjugate to the natural ability of the cationic porphyrin to increase the effective concentration of the ruthenium metallodrug internalized in the cell through its internalization characteristics. We can therefore envisage such porphyrin-RAPTA conjugates having clinical potential for the combined therapy of solid tumours; with localized light treatment being used to generate cytotoxic reactive oxygen species via the porphyrin, and the RAPTA component inhibiting growth of any tumor cells which escape photodynamic destruction. We believe, that these studies have demonstrated that porphyrin-RAPTA conjugates have to potential to be clinically viable, and that we have laid the foundation for further studies to closer examine the biological efficacy of these conjugates in vivo.

Author Contributions
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

TOC GRAPHIC
Patients are presenting with ever more complicated and challenging healthcare issues. Multimodal theranostic medicines are key to improving therapeutic outcomes by personalizing the treatment and imaging of cancers. A novel conjugate bearing a photosensitizing/fluorescent porphyrin tethered to a cytoactive RAPTA metal-based drug has been synthesized and investigated through in vitro biological evaluation. hν