UPCONVERTING VESICLES AND USES

Field of the invention

The present invention relates to vesicles, such as liposomes and polymerosomes, for use in triplet-triplet annihilation upconversion (TTA-UC). The present invention also relates to the use of such vesicles in a variety of applications including in vivo applications such as phototherapy.

Background to the invention

Light-sensitive compounds such as ruthenium(ll) polypyridyl complexes, have recently been proposed as prodrugs for photoactivatable anticancer therapy (PACT).' ¹¹ As shown in classical photodynamic therapy (PDT), the use of light to treat cancer allows for spatially and temporally controlling the toxicity of an anticancer drug, which lowers side effects for cancer patients. Meanwhile, loading anticancer drugs into drug carriers such as liposomes or polymerosomes helps targeting the compounds to tumor tissues. Sterically hindered liposomes, e.g., those grafted with polyethylene glycol chains, have been recognized as versatile and biocompatible drug carriers for the treatment of various diseases because of their long lifetime in the blood circulation. With such PEGylated liposomes tumor uptake is enhanced due to the so-called enhanced permeability and retention (EPR) effect.' ²¹ In PACT, activation of for example liposomes functionalized with ruthenium polypyridyl complexes could be realised after cell uptake using visible light (see for example S. Bonnet, B. Limburg, J. D. Meeldijk, R. J. M. Klein Gebbink, J. A. Killian, J. Am. Chem. Soc, 201 1 , 133, pp 252- 261 ). However, most ruthenium(ll) polypyridyl compounds require activation with blue light (400-500 nm), i.e., outside the phototherapeutic window (600-1000 nm), in which light permeates mammalian tissues optimally.

The process of photon energy upconversion (UC), wherein photons strongly blue shifted relative to the excitation wavelength are generated, has been studied intensively. Different UC techniques, including two-photon absorption (TPA) and energy transfer UC (ETU), have been used. In these techniques, near-infrared (NIR) or infrared (IR) light sources are used. The importance of such UC optical processes is evident from the remarkable number of examples reported, notably for biological imaging, sensing and photodynamic therapy of cancer. However, these upconversion techniques have typically low quantum efficiencies and require extremely intense light beams to occur, which is sometimes problematic for therapeutic applications in vivo. In contrast, energetically conjoined triplet-triplet annihilation-assisted photon energy UC (TTA-UC) has the advantage that extremely low excitation intensities, such as 10 mW.cm-2, or less are required. Thus, specialised equipment such as pulsed lasers are not generally required.

In TTA-UC, low energy photons are converted into higher energy photons by means of a bimolecular mechanism involving a sensitizer and two annihilator molecules.' ³¹ The sensitizer absorbs the low-energy light, undergoes intersystem crossing (ISC) to a triplet state, and transfers its energy to an annihilator molecule via triplet-triplet energy transfer (TTET). After building up the concentration of triplet annihilator molecules, collision of two of them leads to triplet-triplet annihilation (TTA), whereby one annihilator molecule is promoted to the excited singlet state, whereas the other one falls back to the ground state. The singlet annihilator returns to the ground state by emission of a high-energy photon, thus realizing upconversion. TTA-UC with a range of molecule pairs has been realized in organic solvent,[ ^{[3a, 3b, 3dl} ionic liquid,' ⁴¹ polymers,[ ^{[3a, 3c} ' ⁵¹ and various water-soluble nanoparticles.' ⁶¹

Most up-converting systems based on triplet-triplet annihilation (TTA) have been described in organic solvents, as the two molecules necessary for TTA, i.e., the sensitizer that absorbs the low energy light, and the annihilator that emits the higher energy photon, are usually highly lipophilic.

A few water-compatible up-converting systems have been published, but they are based on lanthanide doped up-converting nanoparticles, micelles, or polymer nanoparticles/nanocapsules with an organic solvent or soybean oil core. The typical upconversion quantum yield of lanthanide-based upconverting nanoparticles is low (0.005-3% in hexane solution), and decreases dramatically in water; in addition, their safety for drug delivery has been questioned. Micelles allow for dissolving in water the molecular pair for TTA upconversion, but they do not allow for combining lipophilic and hydrophilic components in the same drug carrier, and their stability in vivo is problematic.

There is therefore a need to obviate and/or mitigate at least one of the aforementioned disadvantages.

Summary of the invention

The present invention is based upon work conducted by the inventors into developing vesicles which are suitable for in vivo applications, and which may be used in TTA-UC. However, as well as the in vivo applications discussed herein, the present invention may be of use in other applications where TTA-UC may be employed, such as in photovoltaics, photocatalysis and photochemical processes and especially in processes carried out in aqueous solution.

In a first aspect there is provided a system for use in triplet-triplet annihilation upconversion (TTA-UC), the system comprising a water soluble vesicle comprising one or more types of vesicle-forming amphiphilic molecules, the vesicles further comprising a sensitizer molecule and an annihilator.

Unlike in other systems described in the prior art, the vesicles of the present invention may be provided in an aqueous solution and so do not need to be provided within a solid, such as a polymeric or gel like matrix described in the prior art. However, the surface of the vesicles may be coated, such as with an inorganic layer, in order to prevent or minimise ingress or egress of material into or from the vesicles.

The vesicles are artificially prepared vesicles comprising amphiphilic molecules. Suitable vesicle-forming amphiphilic molecules include ionic and non-ionic surfactants, lipids, phospholipids, amphiphilic block copolymers and the like. Many types of vesicle-forming amphiphilic molecules are known to the skilled addressee, as well as methods of their preparation. Suitable examples are described, for example, in [V. P. Torchilin, Nature Rev. Drug Discovery 2005, 4, 145-160 and J. S. Lee, J. Feijen, J. Controlled Release 2012, 161 , 473-483]. The vesicles of the present invention are globular objects composed of a usually spherical, closed amphiphilic membrane with a lipophilic core and two hydrophilic interfaces. The spherical membrane surrounds an aqueous compartment that is to some extent isolated from external influences, such as those found in biological systems. Preferred vesicles are in the form of liposomes where the membrane-forming amphiphiles are lipids, or polymerosomes, where the membrane-forming amphiphiles are amphiphillic block copolymers.

The vesicles may be unilamellar or multilammelar. The membrane of the vesicles may comprise a single type of amphiphilic molecule or a plurality of amphiphilic molecules.

Typically, the lipid membrane of liposomes may be comprised of phospholipid(s). Phospholipids generally contain a diglyceride, a phosphate group and a simple organic molecule such as choline, the exception being sphingomyelin, which is derived from sphingosine instead of glycerol. Examples of phospholipids include phosphatidylethanolamine (cephalin) (PE); phosphatidylcholine (lecithin) (PC); phosphatidylserine (PS); and phosphatidylinositol (PI), as well as sphingomyelin. Many natural as well as synthetic derivatives are also known to the skilled addressee. Other components may be present inside the lipid bilayer of the liposome, such as proteins, glycolipids and/or sterols, such as cholesterol.

The lipid mixture used in the formulation of the bilayer may be varied in order to control the temperature of the phase transition of the bilayer. This is important, because the maximum upconversion efficiency is usually found near the transition temperature of the bilayer. For example, using a 2:3 DMPC-DPPC mixture allows for tuning the transition temperature near 37 deg Celsius, to get maximum upconversion efficiency (for this lipid mixture) at that temperature.

For polymersomes, the membrane is typically comprised of one or several amphiphilic block copolymers. In typical examples of such copolymers the hydrophilic block of the copolymer is polyethylene glycol, poly(methyloxazoline), poly(4-vinyl pyridine), poly(acrylic acid), dextran, or poly(L-glutamic acid), and the hydrophobic block is poly(butadiene), poly(butylene oxide), polyisobutylene, polystyrene, poly(caprolacton), poly(dimethylsiloxane) or poly(lactic acid). Many synthetic derivatives are also known to the skilled addressee. Other component(s) may be present inside the vesicle, ie, in the aqueous compartment surrounded by the membrane, and/or inside the hydrophobic core of the membrane.

The vesicles may also be modified so as to attach reactive groups or other moieties, such as polyethylene glycol (PEG), a marker such as for example an antibody, DNA and/or a molecule probe designed to target a specific cell and/or tissue, or a light-activatable prodrug, to the surface of the vesicle. The vesicles may encapsulate a hydrophilic moiety and/or a hydrophobic moiety may be partitioned within the vesicle membrane or hydrophobic core of the vesicle bilayer.

The vesicles may also be modified as to cover the membrane surface with a thin coating, made from, for example, silica, polymer, silver, or gold. The coating can be in the thickness range of 1 -50 nm, more typically 3-10 nm. The coating may serve to prevent ingress or egress of components into or from the vesicles. For example, the coating may prevent oxygen or other gases from penetrating into the vesicles. The coating may contain additional functionalities such as entities to target the liposomes to specific cell and/or tissue or a light- activatable prodrug. The coating may or may not fall apart under the influence of visible to NIR light, ultrasonic sound, differences in pH, or other means of physical or chemical agitation. The aqueous core of the vesicle may contain one or more (bio)chemical components, such as those that scavenge molecular oxygen. Examples of such components are sodium sulfite (Na ₂ S0 ₃ ) or a mixture of glucose oxidase enzyme (GOx), catalase, and D-glucose.

The vesicles may typically be in the size range of 25 nm - 100 μιη, such as 50 nm - 500 nm, more conveniently 100 nm -250 nm.

The sensitizers and annihilators (often termed emitters and used interchangeably throughout) for use in the present invention are often hydrophobic in nature and as such the sensitizers and annihilator may partition within the hydrophobic bilayer of the vesicles/liposomes. Alternatively they may be amphiphilic and distribute with the hydrophilic portion of the vesicle, at the membrane-water interface (e.g. with the phosphate heads of a phospholipid), and their lipophilic part within the lipophilic bilayer. Any hydrophilic sensitizers and/or emitters could be distributed within the core or cores of the vesicle.

A "sensitizer" is a molecule which is able to absorb light in the photodynamic window, the sensitizer being either an organic dye or a metal-organic complex, preferably with a high populated triplet state.

In accordance with the invention, said sensitizer and said emitter molecule are chosen such that a TTA upconverting process between the sensitizer and emitter is possible.

Typically the sensitizer is capable of absorbing light of a first, higher wavelength (lower energy) and the emitter is capable of emitting light of a shorter wavelength (higher energy) than the first wavelength. In preferred embodiments, the sensitizer is capable of absorbing light in the visible or near IR range (such as 380 nm - 1 .1 μιη range) and the annihilator is capable of emitting light at a shorter wavelength to the absorbed light. In a preferred embodiment the sensitizer is capable of absorbing light which is in the green, red or near infra-red range (typically 500 nm -1 μιη) and the annihilator is capable of emitting light in the near UV, violet, blue, green, yellow, or orange part of the spectrum (typically 350 nm - 600 nm).

A metal-organic complex is per definition a compound containing at least a metal center M surrounded by one or more molecules, the so-called ligands L which are generally bound to the metal ion by one or several coordination bond(s). The ligands are organic molecules, cyclic or acyclic, aromatic or non-aromatic, monodentate or polydentate. In case they are extended aromatic systems they are themselves organic dye sensitizers without being bound to a metal. For better understanding: Both, the Pd-porphyrine (= metal organic complex) but also the metal-free porphyrine (=organic molecule) can be sensitizers.

In one embodiment said organic dye, is selected from the group comprising organic molecules having populated triplet states and especially metal-organic complexes having populated triplet states, for example but not limited to Li, Mg, Al, Sc, Ti, Zr, V Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Mo, Rh, Ru, Pd, Ag, Re, Os, Ir, Pt, Tc, Au, Pb, Sn, U containing organic complexes, or any combination of the foregoing to ensure wavelength control, wherein, preferably, said first organic dye is selected from the group comprising compounds with a populated triplet state, including but not limited to

- porphyrins, including extended porphyrins, texaphyrins, sapphyrins, orangerins, any carbon-bridged pyrrolic system, substituted porphyrins and any of the foregoing containing metals including but not limited to Li, Mg, Al, Sc, Ti, Zr, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Mo, Rh, Ru, Pd, Ag, Re, Os, Ir, Pt, Tc, Au, Pb, Sn, U

- phthalocyanines, including extended phthalocyanines, substituted phthalocyanines, and any of the foregoing phthalocyanins containing metals including but not limited to Li, Mg, Al, Sc, Ti, Zr, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Mo, Rh, Ru, Pd, Ag, Re, Os, Ir, Pt, Tc, Au, Pb, Sn, U

- benzopyridines, benzopyrazines, quinolates and hydroxyquinolates, acetyl-acetonates, substituted benzopyridines, benzopyrazines, quinolates and hydroxyquinolates, acetyl- acetonates, and any of the foregoing derivatives thereof containing metals including but not limited to Li, Mg, Al, Sc, Ti, Zr, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Mo, Rh, Ru, Pd, Ag, Re, Os, Ir, Pt, Tc, Au, Pb, Sn, U .

- methylene blue and other phenothiazinium derivatives, xanthenes dyes, BODIPY dyes, cyanine dyes, perylenediimide dyes, curcumin derivatives, or any dyes used in

photodynamic therapy and described in for example in A. B. Ormond, H. S. Freeman, Materials 2013, 6, 817-840, and any of the foregoing derivatives thereof containing metals including but not limited to Li, Mg, Al, Sc, Ti, Zr, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, Ga, Mo, Rh, Ru, Pd, Ag, Re, Os, Ir, Pt, Tc, Au, Pb, Sn, U .

More preferably, said organic dye (i.e. sensitizer) is a porphyrin, or a metalloporphyrin in particular octaethylporphyrin, or tetraphenyltetrabenzoporphyrin-Palladium (or -Pt or -Cu or - Zn), or a phthalocyanin.

Suitable sensitizers are described, for example, in EP2067839, to which the skilled reader is directed and the contents of which, when directed to sensitizers, are hereby incorporated. Annihilator molecules (or "emitter molecules") include organic molecules with (efficient) emitting singlet states. The emitter molecules may be selected from the group comprising compounds showing light emission in the range of from 350 to 600 nm, e.g. anthracenes, tetracenes, pentacenes, perylenes substituted anthracenes, tetracenes, pentacenes, perylenes phenyl (bi-, tri-phenyl)-bridged anthracenes, tetracenes, pentacenes, perylenes fluorenes, thiophenes, polyfluorenes and oligofluorenes, with or without any side-chain pattern and their copolymers, polyparaphenylenes, including polyparaphenylene vinylene, BODIPY, and polyphenyleneethinylenes. Two particular emitter molecules are diphenylanthrancene and perylene.

Other preferred examples of emitter molecules are described, for example, in EP2067839, to which the skilled reader is directed and the contents of which, when directed to said emitter molecules, is hereby incorporated. In addition, biologically occurring emitters such as NAD(P)H, pterins, flavins, coumarins, isolated nucleotides, RNA, or DNA, fluorescent proteins (such as GFP, chitin, collagen or elastin), or their derivatives, may be used as annihilator as well if their combination with a sensitizer provides a TTA upconverting couple.

In all case, the sensitizer and/or annihilator molecules described can be further modified to comprise one or more organic side groups, such as halide substituents (F, CI, Br, I groups), electron-withdrawing groups such as nitro, esters, amides, nitriles, trifluoromethane, or electron-donating groups such as hydroxo, amine, dialkylamine, or n-alkyl groups (methyl, ethyl, propyl, butyl, pentyl, C6-C18 tails, etc.), in order to provide useful physiological properties, such as enhanced solubility or reduced aggregation in the vesicle membrane, or to tune the electronic properties of the sensitizer and/or annihilator and make their energy levels fitting for upconversion.

Further suitable TTA upconverting couples comprising a sensitizer and an annihilator are described for example in Zhao, J, et al, (201 1 ), RSC Advances, 1 , 937-950.

In a further aspect there is provided a system for use in a method of treatment, the system comprising a) a vesicle comprising a sensitizer molecule and an annihilator; and b) a photosensitive moiety which is substantially incapable or poorly capable of being activated by light of a first wavelength, but is capable of being activated upon action of up-converted light emitted by the annihilator, in order to generate a therapeutically or physiologically active molecule. Definitions in relation to the terms vesicle, sensitizer and annihilator are described above. It is understood that light of a first wavelength is of a longer, lower energy wavelength than the up-converted light. Typically the light of the first wavelength is in the range of 500 nm -1 μιη and the up-converted light is in the range of 350-600 nm.

"Substantially incapable or poorly capable of being activated by light of a first wavelength", is understood to mean and recognise that it is possible under certain circumstances for light of the first wavelength to cause activation of the photosensitive moiety, albeit poorly or slowly. As such the present invention provides a way in which activation of the photosensitive moiety, upon irradiation with light of the first wavelength, is able to occur more efficiently and/or rapidly. This of course can easily be tested by the skilled addressee using a suitable assay for detecting any particular activated photosensitive moiety.

The photosensitive moiety may be a compound, such as described in Schatzschneider U., Eur. J; Inorg. Chem. 2010, 10, 1451 -1467 or any other molecule which is commonly used or may be used in photoactivated chemotherapy (PACT). PACT involves the use of non-toxic light-sensitive compounds, a prodrug, which when exposed to light of a suitable wavelength become toxic to surrounding cells and/or tissue without the need of molecular dioxygen being present locally during light irradiation. Often, the compounds become toxic through a change of its chemical structure, for example, through cleavage of one bond or several bonds within the compound, thereby generating a toxic derivative of the prodrug. Like in photodynamic therapy, PACT may be used to target malignancies, such as cancer cells, but has also been used to kill bacteria, fungi and viruses.

In one embodiment, the physiologically active molecule can be conjugated to a marker such as for example an antibody, DNA and/or a molecule probe designed to target a specific cell and/or tissue. Alternatively, the upconverting vesicle can be targeted to the diseased tissue, for example for imaging or diagnostic purposes, by conjugating them to a marker such as an antibody, DNA and/or a molecule probe designed to target a specific cell and/or tissue.

In a further aspect there is provided a method of treatment, the method comprising providing a system comprising a) a vesicle comprising a sensitizer molecule and an annihilator; and b) a photosensitive moiety which is substantially incapable or poorly capable of being activated by light of a first wavelength, but is capable of being activated upon action of up-converted light emitted by the annihilator, in order to generate a therapeutically or physiologically active molecule and irradiating the vesicles with light in the range of 500 nm -1 μιη and the up- converted light is in the range of 350-600 nm. Preferably the method is used in a method of treating a skin condition, such as cancer. The method may be used to treat a disease, such as cancer or pre-cancerous lesions (e.g. basal cell carcinoma and actinic keratosis (pre-cancerous lesions) which is beneath the skin and/or within a tissue by virtue of the ability of the irradiating light to be able to penetrate further than light of a shorter wavelength. An alternative condition would be the treatment of inflammation and/or acne.

The photosensitive moiety may be provided separately to the upconverting vesicle, or may be a component of the vesicle and hence trapped or otherwise associated with the vesicle. When provided separately, the photosensitive moiety may be provided in solution or by other suitable means known to the skilled addressee. The photosensitive moiety may be provided by a further vesicle, which may be formed from similar or alternative vesicle-forming materials, as described above, to the vesicles comprising the sensitizer and annihilator molecules. The skilled addressee is well versed with methods of administering the vesicles and photosensitive moiety to a subject, including parenterally, orally, optically, nasally, rectally, vaginally etc.

In use, the vesicles and photosensitive moiety may be administered concurrently, sequentially or separately and not necessarily by the same route of administration to a subject. However, at least a portion of the vesicles and photosensitive moiety should be localised to a desired site of action or targeting. In some embodiments, the lipid vesicles and photosensitive moiety will be administered locally to a subject. Light of the first wavelength is then allowed to contact the vesicle, which comprises the sensitizer/annihilator combination. Conveniently, light of the appropriate wavelength is shone or otherwise directed towards the vesicles comprising the sensitizer/emitter couple. Unlike prior art techniques which require the use of high energy light, such as blue light to be applied directly towards the photosensitive/prodrug moiety, the present invention allows light of a lower energy, such as green, red, or NIR light, to be used. Such lower energy light is able to penetrate tissue further and so light be applied to the skin, as well as shone easily in lungs, oesophagus, eye, mouth, nose, throat, stomach, bladder, prostate, and ultimately in internal organs using fiber optics. The lower energy light may activate photosensitive molecules deeper within the tissue than can be reached with the use of blue light, for example. It is also possible to irradiate tissues from outside the tissue, such as the liver and kidney, where any diseased tissue may be present deep within the tissue itself and hence activation of the photosensitive molecule should desirably occur deeper within the tissue than blue light, for example, can penetrate. In accordance with the present invention, the vesicles/liposomes are capable of transforming low-energy photons into photons of higher energy. Typically, green or preferably red or NIR photons may be transformed into blue photons. These blue photons are generated locally, at the vesicle/liposome surface, and can in certain embodiments be directly absorbed by a light-activatable prodrug, for example, encapsulated in a nearby vesicle/liposome, nanoparticle, or drug delivery vector, without having to travel long distances in biological tissues. Alternatively, the photosensitive prodrug can be encapsulated in the same membrane as the photosensitizer and annihilator dyes of the TTA-UC pair, in which case non-radiative energy transfer may occur between the annihilator singlet (high energy) excited state and the excited state of the prodrug that needs to be activated. Non-radiative energy transfer is typically more efficient than radiative energy transfer, which can lead to faster activation of the prodrug by upconversion compared to cases where the upconverting liposomes and the compound to be activated are simply mixed in solution. Typical light- activatable prodrugs that may be encapsulated are metal-based photosensitive compounds activated via ligand photosubstitution or photoreduction processes, such as for example those based on ruthenium(ll), platinum(IV), rhodium(lll), iron(ll), or other transition metals, or organic or inorganic compounds activated via light-induced cleavage of an organic photosensitive linker or photo-labile protecting group. Examples of such molecules are described for example in Schatzschneider U., Eur. J; Inorg. Chem. 2010, 10, 1451 -1467, or Leonidova, A.; Pierroz, V. ; Rubbiani, R. ; Lan, Y. ; Schmitz, A. G. ; Kaech, A.; Sigel, R. K. O.; Ferrari, S.; Gasser, G. Chem Sci 2014, doi:10.1039/C3SC53550A; and Higgins, S. L. H.; Brewer, K. J. Angew Chem Int Edit 2012, 51 , 2-5.

Particular advantages of the present invention include: The up-converted photons are generated locally, near the liposome surface, and can be directly absorbed by a light- activatable drug such as encapsulated in the same or a different liposome. This is a significant advantage as the penetration depth of near-UV or blue light is short in most biological tissues, whereas most light-activatable compounds require high-energy photons to be activated by light. Thus, near-UV or blue light activation is only possible for very superficial phototherapy applications such as those involving skin diseases. Red and near-IR light, on the other-hand, penetrates tissues much further (up to 1 cm) and can treat larger lesions or diseases located deeper in biological tissues. Red photons are often not able to activate light-activatable prodrugs based on transition metals, however, and specific compounds must be developed for which dark stability or biological properties need to be reevaluated and/or re-optimized. Secondly, the presently described vesicles/liposomes represent a new method to obtain up- conversion in an aqueous-based environment. Most up-converting systems based on triplet- triplet annihilation (TTA) have been described in organic solvents, as the two molecules necessary for TTA, the sensitizer that absorbs the red light, and the annihilator that emits the blue photon, are usually highly lipophilic. By encapsulating these two molecules in a lipid bilayer membrane efficient TTA upconversion is obtained without a need to add solubilizing groups on the dyes, which would change their photophysical properties. As compared to prior art systems, the TTA up-conversion systems of the present invention offers several critical advantages: 1 ) up-conversion quantum yields are relatively higher even in water (typically 1 -30%); 2) the molecules involved in TTA-UC have a much higher molar absorptivity than the lanthanoid elements in up-converting nanoparticles of the prior art, hence less light is required for activation and tissue heating is avoided; 3) TTA-UC does not require coherent light (lasers) or high intensity light (pulsed lasers); it already occurs with low light intensities such as that experienced in sunlight or LED devices, and reaches maximum efficiency at 100 mW.cm ^"2 ; 4) inclusion of TTA sensitizer/emitter couple in vesicles/liposomes leads to higher local concentrations compared to other systems, which enhances bimolecular processes such as those necessary for TTA-UC to occur; 5) lipophilic and hydrophilic moiety can be provided/associated with a single vesicle unlike prior art systems; and 6) vesicles/liposomes are a standard, recognized tool in drug delivery and other in vivo applications.

Another advantage is that a unique drug-delivery system is developed that can activate photosensitive compounds of many different types. For example, not only ruthenium-based compounds, but also rhodium-based ones, or platinum-based ones, or even organic molecules, would be able to absorb the near-UV, violet, or blue photons generated by the upconverting vesicles/liposomes.

A further advantage of the technology is that high concentrations of the drug can be put near the up-converting vesicles/liposomes, as ultimately all the up-converted photons may be used for drug activation, and as penetration of the blue light further into tissues is not required. When direct irradiation of compounds is used for light activation using one-photon activation, high concentrations of the compound that need to be activated prevent the light penetrating further in the sample or tissue, and a compromise must be found between the amount of drug to be activated and the light penetration depth. With up-converted vesicles/liposomes this is not an issue, as the red light is not absorbed by the prodrug itself. As well as in the aforementioned in vivo applications, the systems described herein may find application in photovoltaic, photocatalysis and photochemical processes and especially in processes carried out in aqueous solution, where light generated through up-scaling techniques as described herein may be of use.

The present invention may also be of use in place of conventional photodynamic therapy (PDT), or as an adjunct thereto, as PDT does not work when 0 ₂ concentration in tissues is too low, for example in hypoxic tumours. The systems of the present invention may be used in areas of low 0 ₂ concentration, and may be, for example, complementary to PDT when 0 ₂ concentration is a problem. The vesicles of the present invention may employ PDT dyes as sensitizers and such vesicles may kill cells via regular PDT (reactive oxygen species generation) when 0 ₂ concentration is high enough, and via PACT via TTA-UC when 0 ₂ is too low for regular PDT.

Thus, in a further aspect there is provided a method of photodynamic therapy (PDT), or as an adjunct thereto, as PDT does not work when 0 ₂ concentration in tissues is too low, for example in hypoxic tumours, the method comprising providing vesicles of the present invention to undesirably proliferating cells, such as cancer cells and irradiating the cells with light of a suitable wavelength in order to enable PDT or PACT via TTA-UC to occur.

Detailed Description of the invention

The present invention will now be further described by way of example and with reference to the figures which show:

Figure 1 : Shows chemical structures of platinum octaethylporphyrin (PtOEP), 9,10- diphenylanthracene (DPA), palladium tetraphenyltetrabenzoporphyrin (PdTPTBP), and perylene.

Figure 2: Shows digital photographs of UCL-1 and UCL-2 under irradiation at 532 nm and 630 nm, respectively, with 27 mW excitation power (for both systems) in a 2.6 mm diameter beam (intensity: 0.51 Wcm ^"2 ). (a) UCL-1 without filter - green light is visible, (b) UCL-1 with 533 nm notch filter and <575 nm short pass filter - blue light is visible, (c) UCL-2 without filter - red light is visible, (d) UCL-2 with 633 nm notch filter - blue light is visible. Samples were deoxygenated and maintained at 298 K until shortly before photography Figure 3: Shows emission spectra of liposome sample UCL-1 (solid line) and of a toluene solution of couple 1 at the same bulk concentrations (PtOEP 3.5 μΜ and DPA 100 μΜ, dashed line), (b) Emission spectra of the liposome sample UCL-2 (solid line) and of a toluene solution of couple 2 at the same bulk concentrations (PdTPTBP 2.5 μΜ and perylene 50 μΜ, dashed line). Asterisks indicate excitation (532 nm for UCL-1 and 630 nm for UCL-2). The samples were deoxygenated before measurement. All spectra acquired at 298 K. Excitation power for both samples 27 mW, 2.6 mm diameter beam, intensity 0.51 W.cm-2.

Figure 4: Shows (a) Luminescence spectrum of UCL-1 at 288 K (dotted), 293 K (dashed) and at 298 K (solid), (b) Time dependency of the upconversion at 436 nm (solid) and of the phosphorescence at 646 nm (dashed) of UCL-1 during three warming and cooling cycles from 288 to 298 K and from 298 to 288 K. (c) Luminescence spectrum of UCL-2 at 288 K (dotted), 293 K (dashed) and 298 K (solid), (d) Time dependency of the upconversion at 473 nm (solid) and of the phosphorescence at 800 nm (dashed) of UCL-2 during three warming and cooling cycles from 288 to 298 K and from 298 to 288 K. Asterisks indicate excitation wavelengths (532 nm for UCL-1 and 630 nm for UCL-2). Samples were deoxygenated before measurement. Excitation power for both samples: 27 mW, 2.6 mm diameter beam, intensity 0.51 Wcm-2.

Figure 5: Shows chemical structures of [1]2+ and [2]2+ and the photochemical reaction of [1 ]2+ into [2]2+. (b) Cartoon illustrating the TTA-UC process in the lipid bilayer, using a sensitizer (S) and an annihilator (A). Radiative energy transfer from the annihilator to complex [1 ]2+, indicated with a single wavy arrow, triggers light-induced hydrolysis of [1 ]2+ to release [2]2+ in solution.

Figure 6: Shows (a) Absorption spectra, after baseline correction, during red light irradiation (630 nm) of a 1 :1 vol% mixture of liposome samples UCL-2 and L-3 (Table 1 ). Solid black line: spectrum at t=0; dashed black line: spectrum at t=240 min; grey lines: spectra measured every 30 min. (b) Plot of the absorbance at 490 nm during red light irradiation (630 nm) of a 1 :1 vol% mixture of UCL-2 and L-3 (dots), of a 1 :1 vol% mixture of L-2b and L- 3 (diamonds), and absorbance at 490 nm of a 1 :1 vol% mixture of L-2b and L3 left in the dark (triangles). Irradiation conditions: power 120 mW, beam diameter 2.6 mm, intensity 2.3 Wcm-2, T=298 K, sample volume 1 mL.

Figure 7: Shows absorbance (left axes, solid lines) and normalized emission (right axes, dashed lines) spectra of PtOEP, PdTPTBP, DPA, and perylene in toluene solution (red) and in liposome samples (black). Liposome samples were diluted 12 times with respect to the formulation given in Table 1 , to keep absorbance values low enough. Absorbance spectra of liposome samples are uncorrected for scattering. Samples containing PtOEP or PdTPTBP were deoxygenated thoroughly before measurement by bubbling the sample with argon for 30 min with a rate of ~2 bubbles per second. All spectra were taken at room temperature, (a) Sample L-1 a (0.3 μΜ PtOEP) and PtOEP in toluene (7 μΜ). For emission, Aexc = 532 nm. (b) Sample L-1 b (8 μΜ DPA) and DPA in toluene (20 μΜ). For emission Aexc = 378 nm. (c) Sample L-2a (0.2 μΜ PdTPTBP) and PdTPTBP in toluene (5 μΜ). For emission Aexc = 630 nm. (d) Sample L-2b (4 μΜ perylene) and perylene in toluene (20 μΜ). For emission, Aexc = 416 nm.

Figure 8: Shows transmission curve of the 633 nm notch filter used in this work (Thorlabs part no. NF633-25). The low transmission for λ < 450 nm explains the difference between perylene emission in Figure 7d and upconverted perylene emission observed in spectra that were acquired by the upconversion emission spectroscopy setup, using the 633 nm notch filter (Figure 3b).

Figure 9: Shows (a) Emission spectra of couple 1 (PtOEP 3.5 μΜ, DPA 100 μΜ) in toluene at 288 K (black line), 293 K (red line), and 298 K (green line), (b) Emission spectra of couple 2 (PdTPTBP 2.5 μΜ, perylene 50 μΜ) in toluene at 288 K (black line), 293 K (red line), and 298 K (green line). Asterisks indicate excitation wavelength: couple 1 and couple 2 were excited with 532 nm and 630 nm laser light, respectively. Samples were thoroughly deoxygenated before measurement. Excitation power for both samples 27 mW in a 2.6 mm diameter beam (intensity 0.51 W.cm ^"2 ).

Figure 10: Shows setup used for upconversion emission spectroscopy. Legend: (1 ) laser source, (2) optical fibers, (3) filter holder, (4) band pass filter that can be installed or removed, (5) variable neutral density filter that can be installed or removed, (6) temperature- controlled cuvette holder, (7) notch filter, (8) CCD spectrometer.

Figure 11 : Shows setup used for absolute quantum yield measurements. Legend: (1 ) laser source, (2) power meter adjustable in position, (3) integrating sphere with sample tube in the center, (4) filter holder, (5) notch filter that can be installed or removed, (6) variable neutral density filter that can be installed or removed, (7) CCD spectrometer, (8) optical fibers.

Figure 12: Shows setup used for photosubstitution experiments using red light. Legend: (1 ) 630 nm laser source, (2) optical fibers, (3) filter holder, (4) 630 nm band pass filter, (5) variable neutral density filter that can be installed or removed, (6) halogen-deuterium light source for absorption measurements, (7) temperature controlled cuvette holder, (8) CCD spectrometer. Figure 13: Shows example output generated by the beam profiling setup in combination with the Beams software package. Axes represent chip width and height in pixels. Colors represent light intensity in increasing order from blue to red.

Figure 14. Emission spectra at 298 K (left) and 308 K (right) of DLPC-1 (long dashes), DMPC-1 (dash-dot-dot), 2:3 DMPC:DPPC-1 (dots) and DPPC-1 (solid) liposomes containing equal bulk concentrations of lipid (5000 μΜ), PEG-2K (200 μΜ), perylene (12.5 μΜ) and Pd(ll)TPTBP (0.625 μΜ) in Na ₂ S0 ₃ PBS buffer. Excitation with 630 nm laser light (0.51 Wcnr ² , 108 mW).

Figure 15. Left: Evolution of upconversion intensity (at 473 nm) vs. temperature for upconverting PEGylated liposomes based on different lipids (DLPC-1 , DMPC-1 , 2:3 DMPC:DPPC-1 mixture, DPPC-1 ). Lipid formulation: lipid (5000 μΜ), PEG-2K (200 μΜ), perylene (12.5 μΜ) and Pd(ll)TPTBP (0.625 μΜ). Conditions: Na ₂ S0 ₃ PBS buffer, excitation at 630 nm, 0.51 W.cm ^"2 , 108 mW. Right: Evolution of upconversion emission intensity at 473 nm with increasing concentrations of dyes, ie from DMPC-1 (diamonds), DMPC-2 (squares) and DMPC-3 (triangles) liposomes in Na ₂ S0 ₃ PBS with temperature. All formulations contained 5000 μΜ of DMPC and 200 μΜ of PEG-2K. DMPC-1 : 12.5 μΜ perylene, 0.625 μΜ Pd(ll)TPTBP; DMPC-2: 25.0 μΜ perylene, 1 .25 μΜ Pd(ll)TPTBP; DMPC-3: 37.5 μΜ perylene, 1 .875 μΜ Pd(ll)TPTBP. Excitation source: 630 nm (0.51 Wcm ^"2 , 108 mW).

Figure 16. a) Evolution of the absorption spectrum during red light irradiation (630 nm) of liposomes made of 0.5 mM DMPC, 0.02 mM DSPE-PEG-2000, 0.02 mM [1](PF ₆ ) ₂ , 2.5 μΜ perylene, 0.25 μΜ PdTPTBP in PBS. b) Evolution of the emission spectrum of the same liposome sample during red light irradiation. Excitation source is not blocked by filters, explaining the large peak at 630 nm. c) Plot of the absorbance of the liposome sample at 490 nm versus irradiation time, d) Plot of the integrated upconverted emission (400-580 nm) versus irradiation time. Irradiation conditions: power 120 mW, beam diameter 4 mm, intensity 0.95 W.cm ^"2 , T=310 K, sample total volume 1 .5 mL, irradiated volume 0.13 mL (8% of total volume), sample deoxygenated for 30 min before irradiation.

Figure 17. Plot of the absorbance at 490 nm as a function of irradiation time for liposomes made of 0.5 mM DMPC, 0.02 mM DSPE-PEG-2000, 0.02 mM [1 ](PF ₆ ) ₂ , 2.5 μΜ perylene, 0.25 μΜ PdTPTBP in PBS (red circles), and of liposomes in which the sensitizer is omitted, while keeping all other concentrations identical (black squares).

Figure 18: Shows the synthesis scheme of PiB1000-PEG350-Me. Figure 19: Size exclusion chromatograms of PiB1000-SA (dotted), PEG350-Me (dashed), and block copolymer product PiB1000-PEG350-Me (solid line). Flow rate I mL.min ^"1 with CHCI ₃ as the fluid phase.

Figure 20: Shows absorbance (left axes, solid lines) and normalized emission (right axes, dashed lines) spectra of PtOEP, PdTPTBP, DPA, and perylene in toluene solution (red) and in polymerosome samples (black). Polymerosome samples were diluted 5 times with respect to the formulation given in Table 4, to keep absorbance values low enough. Absorbance spectra of polymerosome samples are uncorrected for scattering. Samples containing compound PtOEP or PdTPTBP were deoxygenated thoroughly before measurement by bubbling the sample with argon for 30 min with a rate of ~2 bubbles per second. All spectra were taken at room temperature (293 K). (a) Sample P1 ([PtOEP] = 1 .4 μΜ) and PtOEP in toluene (7 μΜ). For emission, A _exc = 532 nm. (b) Sample P2 ([DPA] = 40 μΜ) and DPA in toluene (20 μΜ). For emission A _exc = 378 nm. (c) Sample P3 ([PdTPTBP] = 0.5 μΜ) and PdTPTBP in toluene (5 μΜ). For emission A _exc = 630 nm. (d) Sample P4 ([perylene] = 10 μΜ) and perylene in toluene (20 μΜ). For emission, A _exc = 416 nm.

Figure 21 : Emission spectra of polymerosome sample P1 -2 (left) and P3-4 (right), diluted x10 in PBS, with the upconversion signal magnified x50 (dashed lines). Asterisks indicate excitation (532 nm for P1 -2 and 630 nm for P3-4). The samples were deoxygenated before measurement. Spectra acquired at 293 K, excitation power for both samples 70 mW, 4 mm diameter beam, intensity 0.56 W.cm ^"2 .

Figure 22. Microscopy images of fixated 3T3 cells treated with upconverting liposomes. Left: brightfield; middle: upconverted emission (450<λ<575 nm) upon red light excitation (630 nm, laser spot size ^« 100-120 μιη, intensity 1 .5 kWcm ^"2 ); right: emission by blue light excitation (405 nm, laser spot size ^« 15-25 μιη). Imaged under anoxic conditions by supplementing the PBS buffer with 0.3 M Na ₂ S0 ₃ . Temperature 298 K.

Example 1

Two well-investigated TTA-UC couples were considered for incorporation in liposomes: platinum octaethylporphyrin and 9,10-diphenylanthracene (couple 1 , see Figure 1 ), and palladium tetraphenyltetrabenzoporphyrin and perylene (couple 2). Obviously, when included in liposomes these highly apolar molecules favour the lipophilic interior of the lipid bilayer. Liposomes made of 1 ,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC) and containing 4 mol% of sodium N-(carbonyl-methoxypolyethylene glycol-2000)-1 ,2-distearoyl- sn-glycero-3-phosphoethanolamine (DSPE-MPEG-2000), the sensitizer, and/or the annihilator of each couple, were prepared by extrusion in DPBS buffer solution (see Table 1 ). The diameter of the liposomes (130-170 nm) were measured by dynamic light scattering. UV-VIS absorption and luminescence spectra of the liposomes containing either the sensitizer or the annihilator, i.e., of samples L1 a, Li b, L2a, and L2b, were comparable to that of the corresponding compounds in toluene solution (see Figure 7). Thus, incorporation of any of the four molecules shown in Figure 1 into the DMPC bilayers did not change their spectroscopic properties.

Table 1

Overview of liposomal formulations used in this work. Concentrations are given as bulk concentration in DPBS buffer.

(mM)

UCL-1 20 0.80 3.5 100 - -

L-1a 20 0.80 3.5 - - -

L-1 b 20 0.80 - 100 - -

UCL-2 20 0.80 - - 2.5 50

L-2a 20 0.80 - - 2.5 -

L-2b 20 0.80 - - - 50

L-3 5.0 0.20 - - - 0.20

Although in UCL-1 and UCL-2 both molecules of each upconverting couple were successfully inserted into the bilayer it was initially uncertain whether their diffusion in the two-dimensions of the bilayer would be sufficient to allow TTA-UC to occur. ^[3a] After deoxygenation these samples were excited at either 532 nm or 630 nm, respectively, corresponding to the absorption of the highest Q-band of PtOEP ^ _max =536 nm) and PdTPTBP nm), respectively. A bright blue luminescence was observed in both cases after suppressing the scattered excitation light using notch and/or short pass filters (Figure 2). Under the same experimental conditions, no blue emission was observed for L- 1 a, L-1 b, L-2a, or L-2b, thus proving that both components of each upconverting couple are necessary for the upconversion to occur. As both green-to-blue and red-to-blue upconversion was obtained using similar formulations, TTA-UC in liposomes seems not to be specific of one particular couple, i.e., liposomes represent a straightforward manner to solubilize TTA-UC couples in aqueous solution.

The luminescence spectra of UCL-1 and UCL-2 were measured at 298 K under argon (Figure 3). Upon excitation at 532 nm, UCL-1 shows a structured upconversion band at 433 nm, corresponding to DPA emission in toluene (see Figure 7b). A second band was present as well; its emission maximum (646 nm) was consistent with the phosphorescence of PtOEP in toluene (Figure 7a). Similarly, for UCL-2 excitation at 630 nm leads to an upconversion band at 473 nm, and a second band was observed at 800 nm (Figure 3b). The upconversion emission corresponds to perylene emission in toluene (Figure 7d), apart from the first peak at 447 nm that was filtered by the 633 nm notch filter used for rejecting the scattered excitation (Figure 8). The peak at 800 nm in the emission spectrum of UCL-2 corresponded to the phosphorescence of PdTPTBP, as observed in toluene (Figure 7c).

The luminescence spectra of both upconverting couples were measured in toluene using the same experimental conditions as for UCL-1 and UCL-2, and notably the same bulk concentrations of the sensitizer and annihilator (Figure 3). The upconversion intensity for couple 1 was found four times weaker in liposomes than in toluene at 298 K, and for couple 2 it was comparable for both sample types. Upon inserting the sensitizer and annihilator in the lipid bilayer two phenomena take place simultaneously. On the one hand, compartmentalization of the lipophilic molecules in the bilayer increases their local concentrations, which increases the probability of intermolecular collision and therefore the rates of TTET and TTA. On the other hand, two-dimensional diffusion in a lipid bilayer is somewhat slower than in a non-viscous isotropic toluene solution, which may decrease TTA- UC efficiency in liposomes. Overall, our data show that the trade-off is excellent and allows efficient TTA-UC to occur in PEGylated DMPC liposomes (at 298 K). Table 2. Quantum yield of upconversion (4> _uc ) at 293 K

[a] 10 mW 532 nm excitation power in 1 .5 mm diameter beam (intensity 0.57 Wcm ^"2 ). [b] [PtOEP] = 3.5 μΜ, [DPA] = 100 μΜ. [c] 10 mW 630 nm excitation power in 2.5 mm diameter beam (intensity 0.20 Wcm ^-2 ). [d] [PdTPTBP] = 2.5 μΜ, [perylene] = 50 μΜ.

Measurements of upconversion quantum yields (Ouc) are usually done by relative actinometry. ^[3a] However, intense scattering in liposome samples would make any comparison with a reference compound in homogeneous solution challenging. For this reason, the upconversion quantum yields of UCL-1 and UCL-2 were measured using an absolute method, i.e., an integrating sphere and a calibrated spectrometer. The setup was similar to that used by Boyer et al. for determining the upconversion quantum yield of lanthanoid-based nanoparticles. ^[7] For UCL-1 , UCL-2, and for their toluene analogues, Ouc was determined upon irradiation using a 10 mW continuous beam. The beam diameter was 1 .5 mm for couple 1 (intensity: 0.57 Wcm-2) and 2.5 mm for couple 2 (intensity: 0.20 Wcm- 2). At 293 K Ouc in PEGylated DMPC liposomes was found roughly half of that in toluene for both couples, with values of 2.3% and 0.5% for UCL-1 and UCL-2, respectively, versus 5.1 % and 1.2% in toluene. Such quantum yields are rather competitive in comparison with lanthanoid-based upconverting nanoparticles in hexanes, which upconvert 980 nm light with efficiencies ranging from 0.005 to 3.0%. ^[7] To the best of our knowledge, it is the first time that the quantum yields of TTA-UC are determined using an absolute method.

The TTA-UC process is diffusion controlled, and therefore depends on temperature and on the viscosity of the medium. For this reason, luminescence spectra were measured for UCL- 1 and UCL-2 at 288, 293, and 298 K (Figure 4a and Figure 4c). Upon warming the sensitizer phosphorescence decreased for both samples, while the upconversion emission increased markedly. In contrast, for toluene samples at the same bulk concentrations both the upconversion and phosphorescence intensities slightly decreased with increasing temperatures (Figure 9) as a result of faster non-radiative decay. The liposome samples were subjected to three warming-cooling cycles while continuously monitoring their luminescence (Figure 4b and 4d). The temperature dependence of the upconversion was found reversible, which advocates for a reversible, physical cause rather than an irreversible chemical evolution (such as aggregation or photoreactions). As the variation of the upconversion vs. phosphorescence efficiency occurs at a temperature that fits the gel-to- fluid phase transition temperature (Tm) of DMPC membranes (296.9 K), we interpret this variation as a consequence of the much increased translational diffusion coefficient (DT) of membrane-embedded molecules above Tm, compared to that at temperatures below Tm . ^[8] With DT being typically two orders of magnitude higher in the liquid phase TTET and TTA are both expected to be much more frequent above Tm , leading to an increase in intermolecular processes (upconversion) at the cost of monomolecular processes (phosphorescence). Similar observations were made for TTA-UC in rubbery polymer matrixes by Sing-Rachford and co-workers. ^[5e]

In order to prove that in situ upconverted blue photons may be used to activate light- activatable prodrugs using red light, for example for PACT, ruthenium-functionalized liposomes were mixed with the upconverting liposomes UCL-2 (Figure 5b). The ruthenium complex [Ru(tpy)(bpy)(SRR'))]2+ ([1 ]2+, see Figure 5a and further information below) was selected because it has a single light-sensitive Ru-S bond. This kind of photoactivatable ruthenium compound shows stability in the dark but hydrolyses to the aqua species [Ru(tpy)(bpy)(H20)]2+ ([2]2+) upon irradiation with blue light into its metal-to-ligand charge transfer state.' ⁹¹ A thioether-cholesterol ligand (SRR') can be used to anchor the complex to lipid bilayers, as has been demonstrated in our group. ^{193, 9c]} PEGylated DMPC liposomes bearing 3.7 mol% of complex [1]2+ were prepared (sample L-3, Table 1 ) and added in 1 :1 volumetric ratio to red-to-blue upconverting liposome sample UCL-2. As both types of liposomes are grafted with sterically hindering polyethylene glycol (PEG) tails, fusion of the liposomes does not occur, and only radiative energy transfer between the upconverting liposomes and the ruthenium-functionalized liposomes should take place (Figure 5b). ^[10]

The liposome mixture was deoxygenated and irradiated at 298 K for 2 hours with a 120 mW 630 nm laser light beam from a clinical grade Diomed PDT laser. The photoreaction was monitored by UV-Vis spectrometry at fixed intervals during irradiation (Figure 6). Although the absorbance of perylene dominates the spectrum the characteristic band of the hydrolysed photoproduct ([2]2+) can be clearly seen rising between 450 and 550 nm as a function of irradiation time. The isosbestic point at 457 nm showed that a single photochemical process was taking place. Monitoring the absorbance at 490 nm allowed for quantitatively measuring the build-up of [2]2+ as a function of irradiation time, which reached a plateau after 3 hours irradiation (Figure 6b). As a control experiment, a mixture of conditions as above. In liposomes L-2b the absence of sensitizer prevents upconversion to occur, and the red photons can only excite the ruthenium complex by direct absorption in the 1 MLCT band. The extinction coefficient of [1]2+ being very low at 630 nm (ε<100 M-1 cm-1 ), even under a strong photon flux the photoconversion to [2]2+ is much slower than in presence of UCL-2 (Figure 6b), i.e., the upconverting liposomes achieve efficient sensitization of the substitution reaction. A second control experiment showed that no photodissociation occurred in absence of light. Overall, these data are the first evidence that blue photons produced in situ by upconversion of PDT-compatible red photons, can be used to enhance the photodissociation rate of polypyridyl ruthenium complexes.

In conclusion, triplet-triplet annihilation upconversion was realised in liposomes and characterized by absolute quantum yield measurement. Red-to-blue upconverting liposomes UCL-2, when mixed with ruthenium-functionalized, PEGylated liposomes L-3 and irradiated with a clinical grade PDT laser at 630 nm, were able to trigger via radiative energy transfer the hydrolysis of the Ru-S bond and to release complex [2]2+. The upconverting liposomes transform two low-energy photons, which penetrate far in biological tissues but are poorly absorbed by the ruthenium complex, into one blue photon that does not need to travel into tissues and can directly promote the complex into its photoreactive excited state. Metal- ligand photodissociation mediated by upconverted light represents exciting perspectives for photoactivatable chemotherapy, for example, in oxygen-poor tissues such as hypoxic tumors. The high quantum yield of TTA-UC in liposomes and the excellent molar absorptivity of porphyrin sensitizers, for example compared to lanthanoid-based upconverting nanoparticles, may offer fascinating applications in biological imaging, photoactivatable chemotherapy, and other biological applications where the in situ generation of blue light is required.

Example 2.

To further show the potential of prodrug activation via upconverted light, a liposomal formulation was prepared containing DMPC, 0.05 mol% PdTPTBP as sensitizer, 0.5 mol% perylene as annihilator, and 3.7 mol% of the ruthenium complex [1 ] ²⁺ . In this formulation, the TTA upconversion dyes and the prodrug to be activated are mixed in the same membrane, which allows for non-radiative energy transfer between perylene and the ruthenium complex [1 ] ²⁺ . In order to minimize activation of the Ru complex by radiative energy transfer, i.e., via emission of a blue photon by the annihilator and re-absorption by the Ru complex, the sample was diluted to low concentrations ([DMPC] _bU i _k = 0.5 mM) so that the Optical Density at the perylene emission wavelength (440<λ<550 nm) was low (OD<0.2 for a 1 cm optical path length).

The sample was deoxygenated, thermostated at 310 K (i.e. biological temperature), and irradiated at 630 nm and 120 mW (0.95 W.cm ^"2 ) using a clinical grade (Diomed) PDT laser. The UV-VIS absorption spectrum of the sample and its emission spectrum were monitored in 15-minute intervals during the irradiation (see Figure 16). In the UV-VIS absorption spectrum the characteristic band of the hydrolysed photoproduct ([2] ²⁺ ) could clearly be seen between 450 and 550 nm during irradiation. The isosbestic point at 457 nm showed that a single photochemical reaction was taking place. Monitoring the absorbance at 490 nm allowed for quantitatively measuring the build-up of [2] ²⁺ as a function of irradiation time, which reached a plateau after 2.5 h of irradiation (Figure 16c). When the sensitizer was omitted from the formulation so that no upconversion took place, the absorbance at 490 nm only slowly increased with irradiation time due to the low extinction coefficient of the ruthenium complex at that wavelength (<100 M ^"1 .cm ^"1 , Figure 17). These results demonstrate that the photodissociation of [1 ] ²⁺ is accelerated by upconvertion.

The emission spectrum shows a clear rise of the upconversion emission between 400 and 600 nm as a function of irradiation time (Figure 16b). Meanwhile the phosphorescence of the porphyrin, centered at 800 nm, decreased slightly, indicating a small amount of photobleaching. A plot of the upconverted intensity vs. irradiation time plateaued after 2.5 h of irradiation (Figure 16d). At this point the upconversion emission had increased tenfold with respect to the starting point. This increase is attributed to the diffusion of the hydrolysed aqua product [2] ²⁺ away from the lipid membrane once the photosensitive Ru-S bond has been broken, so that quenching of the annihilator emission by energy transfer to the Ru complex no longer occurs and radiative emission is observed instead. After 2.5 hours the upconversion started decreasing, which we attribute to photobleaching.

In conclusion, the results clearly show that in a liposomal formulation combining the sensitizer, the annihilator, and a ruthenium prodrug on the same lipid membrane, light- induced hydrolysis of [1 ] ²⁺ is facilitated via triplet-triplet annihilation upconversion. Considering that irradiation is performed with a clinical grade PDT laser and at biological temperature (310 K), the potential of these liposomes for photoactivated chemotherapy is demonstrated.

Example 3.

Diffusion rates of molecules in a bilayer are essential for TTA upconversion. These diffusion rates depend on the phase of the lipid bilayer (gel or liquid), which depend on the temperature and on the transition temperature of the lipid mixture composing the membrane. A comparison of the upconversion emission intensity of PEGylated upconverting liposomes made of DLPC, DMPC, a 2:3 DMPC:DPPC mixture, and DPPC, was made in Na ₂ S0 ₃ PBS buffer with an excitation wavelength of 630 nm, and the perylene/PdTBTPP couple as upconverting couple. For each lipid type three samples were prepared with increasing concentrations of the perylene and porphyrin dyes (see formulations Table 3). DMPC:DPPC lipid mixtures are quasi-perfect, which allows for tuning the transition temperature of the lipid bilayer linearly between 297 K (transition temperature of pure DMPC) and 315 K (transition temperature of pure DPPC). A 2:3 DMPC:DPPC lipid mixture for example has a transition temperature of 309 K, as measured by Differential Scanning Calorimetry, which is very near human body temperature.

Upconversion intensity was measured at 298 K (Figure 14, left) and 308 K (Figure 14, right). In all cases emission is seen near 800 nm, which corresponds to the phosphorescence of Pd(ll)TPTBP, and in the region 450 - 530 nm, which corresponds to the delayed fluorescence of perylene. It can be seen from Figure 14 (left) that the upconversion intensity at 298 K is greatest for the DMPC-1 liposomes, followed closely by the DLPC-1 liposomes. The upconversion of the DMPC:DPPC-1 and the DPPC-1 liposomes is considerably lower than that of the DMPC-1 or DLPC-1 samples while their phosphorescence is considerably higher. This is likely due to more triplet states of the porphyrin undergoing phosphorescence because fewer are being quenched by the annihilators in the TTET processs.

Table 3. Formulations of upconverting PEGylated liposomes prepared for studying the evolution of upconversion intensity with temperature and lipid type.

Label Lipid [Lipid] [PEG-2K] [Perylene] [PdTPTBP]

Bulk Bulk (μΜ) Bulk (μΜ) Bulk (μΜ)

(μΜ)

DLPC-1 DLPC 5000 200 12.5 0.625

DMPC-1 DMPC 5000 200 12.5 0.625

DMPC:DPPC-1 DMPC:DPPC (2:3) 5000 200 12.5 0.625

DPPC-1 DPPC 5000 200 12.5 0.625

DLPC-2 DLPC 5000 200 25.0 1.250

DMPC-2 DMPC 5000 200 25.0 1.250

DMPC:DPPC-2 DMPC:DPPC (2:3) 5000 200 25.0 1.250

DLPC-3 DLPC 5000 200 37.5 1.875

DMPC-3 DMPC 5000 200 37.5 1.875

DMPC:DPPC-3 DMPC:DPPC (2:3) 5000 200 37.5 1.875

At 308 K (Figure 14 right) the features of the spectra remain the same as those seen at 298 K, including the emission at 800 nm and the emission at 450 - 550 nm, however there are changes in the observed intensities. Most importantly the upconversion emission of the DMPC-1 and DLPC -1 liposomes decreases while that of the DMPC:DPPC-1 mixture and the DPPC-1 increases. The porphyrin phosphorescence peak of all four liposome compositions can be seen to have decreased on going to higher temperature. This is predominantly due to increased quenching of the porphyrin excited state at higher temperature. The particularly large decrease in the porphyrin phosphorescence of the DMPC:DPPC-1 sample is likely due to the large increase in upconverted light, i.e. the phosphorescence of the porphyrin decreases due to more of the Pd(ll)TPTBP triplet excited states undergoing triplet-triplet energy transfer to the perylene hence leaving less to undergo phosphorescence.

The global evolution of upconversion intensity as a function of lipid type, temperature, and dye concentration in the membrane, is shown in Figure 15. Basically, when temperature increases diffusion rate in the membrane increases, which leads to improved upconversion intensity. Above the transition temperature quenching of the triplet state of the porphyrin increases, which diminishes the upconversion intensity. In all tested cases a maximum in the upconversion intensity was found at the transition temperature of the lipid bilayer. By increasing the concentration of the annihilator and photosensitizer dyes in the membrane, ie, along the series DMPC-1 , DMPC-2, DMPC-3 for example, the upconversion intensity increases correspondingly. In practical terms, for a given dye concentration the greatest upconversion intensity at biological temperature (308 K) was observed in DMPC liposomes, followed by DLPC liposomes, 2:3 DMPC:DPPC, liposomes, and finally DPPC liposomes. The latter are poor media for upconversion at biological temperature.

Example 4.

Besides liposomes, numerous other nanodevices are currently under development for applications in drug delivery and others. Among them, polymerosomes, i.e. self-assembled vesicles composed of an amphiphilic block copolymer membrane enclosing an aqueous core, receive increasing attention due to their fascinating properties. ¹¹¹¹ A wide choice of block copolymers and a broad range of molecular weights offer ample opportunity to rationally tune the vesicles' critical parameters such as membrane fluidity, thickness, morphology, stability, permeability, encapsulant retention, et cetera. A synthetic block copolymer was designed that is composed of polyisobutylene (PiB) as hydrophobic block and polyethyleneglycol (PEG) as hydrophilic block with a molecular weight ratio of PiB:PEG = 1000:350, which fits the requirement for vesicle formation. ^111c] It was envisioned that polyisobutylene would be easily able to act as a host matrix for the highly hydrophobic molecules involved in TTA-UC. Moreover, the well-known barrier properties of polyisobutylene could prevent molecular oxygen from reaching the TTA-UC molecules and therefore preventing quenching of the upconversion by oxygen. This property has been exploited before by Kim and Kim who designed oxygen-independent upconverting polymer shell microcapsules in aqueous medium with an interior phase consisting of hexadecane and 5 wt % polyisobutylene. ^[6c]

Amphiphilic block-copolymer PiB1000-PEG350-Me was synthesized according to Figure 18. Size exclusion chromatography was performed on the starting materials and the product (see Figure 19). The main peak of the product chromatogram is shifted towards higher molecular weight (smaller elution volume) with respect to the two block polymers, indicating that the product is indeed the combination of the two block polymers.

The resulting amphiphilic block copolymer was used for the assembly of polymerosomes by using the classical hydration-extrusion protocol commonly used for liposome assembly. The block copolymer readily self-assembles into polymerosomes upon hydration of the dried polymer film as can be seen by dynamic light scattering (before extrusion typically z-ave = 136 nm, 0.260 PDI) and they can be further shaped into smaller particles by extrusion (after extrusion typically z-ave = 133 nm, 0.193 PDI).

Table 4: Overview of polymerosomal formulations used in this work. Concentrations are given as bulk concentration in DPBS buffer.

Code [PiB ₁₀ oo-PEG ₃ 5o-Me] [PtOEP] [DPA] (μΜ) [PdTPTBP] [perylene]

(mg/ml) (MM) (MM) (MM)

P0 10 -

P1 10 7 - - -

P2 10 - 200 - -

P1 -2 10 7 200 - -

P3 10 - - 2.5 -

P4 10 - - - 50

P3-4 10 - - 2.5 50

Polymerosomes with sensitizer PtOEP or PdTPTBP, and/or annihilator DPA or perylene were prepared (see). The average size (z-ave) of all samples was 140 ± 5 nm with a polydispersity index (PDI) of 0.226 ± 0.027. Samples P1 , P2, P3, and P4 were investigated by UV-VIS and emission spectroscopy and the presence of the molecular dyes in the sample was confirmed (Figure 20). Samples P1 -2 and P3-4 were investigated for upconversion by luminescence spectroscopy at 293 K under deoxygenated conditions (see Figure 21 ). Upon excitation in the Q-band of PtOEP at 532 nm, P1 -2 shows a structured upconversion band at 433 nm, corresponding to emission of DPA in polymerosome or toluene (see Figure 20). A second band was present as well; its emission maximum (646 nm) was consistent with the phosphorescence of PtOEP in polymerosome or toluene (see Figure 20). Similarly, for P3-4 excitation at 630 nm leads to an upconversion band at 473 nm, and a second band at 800 nm (see Figure 4). The upconversion emission corresponds to emission of perylene in polymerosome or toluene (see Figure 20). The peak at 800 nm in the emission spectrum of P3-4 corresponds to the phosphorescence of PdTPTBP in polymerosome or toluene (see Figure 20). For both P1 -2 and P3-4, the upconversion signals were weak in comparison with the phosphorescence, likely due to the high viscosity of the polyisobutylene membrane which makes the molecular dyes less mobile than in liposomes and thus less efficient in TTA-UC. However, the results clearly indicate that the polymer membrane of polymerosomes constructed from PiB1000-PEG350-Me successfully forms a host lattice that is able to facilitate upconversion via TTA-UC. No upconversion was observed when the samples were not deoxygenated so under these experimental conditions the nanoscale polyisobutylene membrane was not able to act as a barrier for dissolved oxygen.

In conclusion, amphiphilic di-block copolymer PiB1000-PEG350-Me was synthesized and used for the self-assembly of polymerosomes in PBS buffer solution. The combination of the inexpensive starting materials, the facile synthesis of the polymer, the biocompatibility of both of the block polymers, ¹¹²¹ and the ease of polymerosome assembly make these nanoparticles promising for applications such as drug delivery or aqueous nano reactor chemistry. The membrane of the polymerosomes was able to act as a host matrix for highly hydrophobic molecular dyes and, for the first time in polymer vesicles triplet-triplet annihilation upconversion was demonstrated. Future endeavours are devoted to increasing the dye load of the membrane and decreasing the viscosity of the polyisobutylene matrix in order to increase the upconversion intensity.

Example 5

To demonstrate that upconverting liposomes can be taken up by cells, a qualitative cell uptake assay was realized where the upconverted emission of red-to-blue upconverting liposomes in cells is measured by luminescence microscopy. To this end, 3T3 mouse embryo fibroblast cells were incubated with upconverting liposomes containing PEG, perylene, and PdTPTBP, subsequently fixated, and imaged under anoxic conditions i) with bright field illumination, ii) 405 nm excitation (blue light), and iii) 630 nm excitation (red light). Figure 22 shows a typical example of images acquired after 3 h incubation time. The bright field image shows two cells attached to the glass surface. Under 405 nm (blue light) excitation, not only the perylene fluorescence is seen, but also a lot of autofluorescence from the cell structures, which renders the identification of the liposomes impossible. However, under 630 nm (red light) excitation and after notching out all photons above 575 nm or below 450 nm, emission from the perylene annihilator, i.e., upconverted emission, was clearly detected from numerous sites in the cytosol. No upconversion emission was detected in control experiments in which no liposomes were administered to the cells. These results clearly indicate that the upconverting liposomes are being taken up by 3T3 cells and that the upconversion emission could be detected in fixated cells by luminescence microscopy.

Materials & Methods and additional information

Synthesis

General. Platinum octaethylporphyrin (PtOEP) and palladium tetraphenyl tetrabenzoporphyrin (PdTPTBP) were purchased from Frontier Scientific, Inc. (Logan, Utah, USA). Diphenyl anthracene (DPA) was purchased from Alfa Aesar GmbH & Co KG (Karlsruhe, Germany). Sodium N-(carbonyl-methoxypolyethylene glycol-2000)-1 ,2- distearoyl-sn-glycero-3-phosphoethanolamine (DSPE-MPEG-2000) and 1 ,2-dimyristoyl-sn- glycero-3-phosphocholine (DMPC) were purchased from Lipoid GmbH (Ludwigshafen, Germany) and stored at -18 °C. Polyisobutylene succinic anhydride (PiB1000-SA, "Dovermulse H-1000") with a molecular weight of 1000 gmol-1 and hydrolysis number 58.1 mg KOH/g was kindly provided by Dover Chemical Corporation (Dover, OH, USA). Methoxypolyethylene glycol (PEG350-Me) with a molecular weight of 350 gmol-1 was purchased from Santa Cruz Biotechnology, Inc. (Heidelberg, Germany). Dulbecco's phosphate buffered saline (DPBS) was purchased from Sigma Aldrich Chemie BV (Zwijndrecht, The Netherlands) and had a formulation of 8 g.L-1 NaCI, 0.2 g.L-1 KCI, 0.2 g.L- 1 KH2P04, and 1.15 g.L-1 K2HP04 with a pH of 7.1 -7.5. All chemicals were used as received. The syntheses of the thioether-cholesterol conjugate SRR' and [Ru(terpy)(bpy)(CI)](CI) are described elsewhere ^{[9c 13]} .

[1](PF Synthesis of [1 ](PF6)2. [Ru(terpy)(bpy)(CI)](CI) (100 mg, 0.18 mmol), ligand SRR' (1 17 mg, 0.21 mmol), and AgBF4 (73 mg, 0.37 mmol) were dissolved in acetone (30 ml_). The reaction mixture was refluxed for 20 h in the dark. After cooling to room temperature it was filtered hot over Celite, and the solvent was removed by rotary evaporator under reduced pressure. The product was purified by column chromatography on silica gel (acetone/H20/sat. aq. KPF6 100:10:1 .5, Rf=0.35). Acetone was evaporated under vacuum, upon which the product precipitated as an orange solid. [1]2+(PF6)2 was filtered, washed with water and dried under vacuum for 4 h. (124 mg, 52%). 1 H NMR (300 MHz, δ in CDCI3) 9.72 (d, J = 5.3 Hz, 1 H, A6), 8.55 (m, J = 8.2 Hz, 3H, A3 + T3'), 8.41 (d, J = 7.9 Hz, 2H, T3), 8.34 (d, J = 8.0 Hz, 1 H, B3), 8.27 - 8.14 (m, 2H, A4 + T4'), 8.03 - 7.85 (m, 3H, A5 + T4), 7.74 (t, 1 H, B4), 7.68 (d, J = 5.0 Hz, 2H, T6), 7.36 (m, 2H, B5 + B6), 7.16 (m, 2H, T5), 5.30 (d, J = 4.8 Hz, 1 H, 6), 3.75 (t, J = 6.6 Hz, 2H, ζ ), 3.64 - 3.37 (m, 10H, α + β + γ + δ + ε), 3.13 (s, 1 Η, 3), 2.40 - 0.75 (m, 47H), 0.67 (s, 3H). 13C NMR (75 MHz, δ in CDCI3) 157.67 + 157.01 + 156.31 + 156.29 ( B2+ A2 + T2 + T2'), 153.18 (T6), 151 .95 (A6), 149.80 (B6) ,

140.86 (5), 139.09 (T4), 138.56 +138.37 (B4 + A4), 137.56 (Τ4'), 128.91 (T5), 128.35 (A5),

127.87 (B5), 125.16 (T3), 124.85 (A3), 124.48 (Τ3'), 124.03 (Β3), 121 .86 (6), 79.56 (3), 70.88 + 70.35 + 70.30 + 67.52 + 67.30 (a + β + γ + δ + ε), 56.86, 56.28, 50.26, 42.44, 39.88, 39.64, 39.22, 37.28, 36.97, 36.31 , 35.91 , 34.47, 32.06, 32.01 , 29.82, 28.35, 28.13, 24.42, 23.97, 22.95, 22.69, 21 .19, 19.53, 18.85, 15.04, 12.00. UV-Vis: Amax (ε in L.mol-1 . cm-1 ) in CHCI3: 457 nm (6090). ES MS m/z exp. (calc.): 519.7 (519.4, [M-2PF6]2+). Elemental analysis for C59H79F12N503P2RuS: (calc); C, 53.31 ; H, 5.99; N, 5.27; S, 2.41 . (Found); C, 53.34; H, 6.22; N, 5.15; S 2.41.

Synthesis of PiB1000-PEG350-Me. 2.0 g PiB1000-SA (2.1 mmol) and 0.72 g methyl- polyethylene glycol (PEG350-Me, MW=350, 2.1 mmol) were added to toluene (10 ml_). The mixture was heated to 60 - 80 °C, after which air was removed using Schlenk techniques. Then, the mixture was refluxed overnight and allowed to cool to room temperature. The solvent was removed by rotary evaporation at 50 °C and subsequently under high vacuum overnight. The viscously fluid yellow product was used without further purification. 1 H NMR (300 MHz, CDCI3) δ 4.9-4.7 (C=C of PiB), 4.3-4.1 (C(0)-0-CH2-CH2), 3.5-3.8 (0-CH2-CH2- O of PEG), 3.4 (s, 3H, O-Me of PEG), 3.1 -2.9, 2.7-2.3, 2.3-2.1 , 2.1 -0.7 (PiB); ratio of all integrated PEG versus PiB signals is approximately 1 :4. IR-spectroscopy (cm ¹ ): 2950, 2882 (C-H of PiB), 1735 (C=0 succinic acid and ester), 1637 (C=C of PiB), 1470, 1389, 1365, 1230 (PiB skeleton), 1 105 (ether vibration of PEG). All spectral data agree with literature values.' ¹⁴¹ Size exclusion chromatography was done with a double PLgel 3 μιη MIXED-E column from Agilent Technologies (Amstelveen, The Netherlands) with 100 μΙ_ injection volume, and 1 mL.min ^"1 flow rate with CHCI ₃ as the fluid phase. Detection was done by an Optilab DSP interferometric refractometer from Wyatt Technology Europe GmBH (Dernbach, Germany).

Liposome preparation

All liposome formulations were prepared by the classical hydration-extrusion method. As an example, the preparation of UCL-1 is described here. Aliquots of chloroform stock solutions containing the liposome constituents were added together in a flask to obtain a solution with 20 μηιοΙ DMPC, 0.8 μηιοΙ DSPE-MPEG-2000, 100 nmol DPA, and 3.5 nmol PtOEP. The organic solvent was removed by rotary evaporation and subsequently under high vacuum for at least 30 minutes to create a lipid film. 1 .0 ml_ DPBS buffer was added and the lipid film was hydrated by 5 cycles of freezing the flask in liquid nitrogen and thawing in warm water (50 °C). The resulting dispersion was extruded through a Whatman Nuclepore 0.2 μιη polycarbonate filter at 40-50 °C at least 1 1 times using a mini-extruder from Avanti Polar Lipids, Inc. (Alabaster, Alabama, USA). The number of extrusions was always odd to prevent any unextruded material ending up in the final liposome sample. The extrusion filter remained colorless after extrusion, suggesting complete inclusion of the TTA-UC compounds in the lipid bilayer. Liposomes were stored in the dark at 4 °C and used within 7 days. The average liposome size and polydispersity index were measured with a Malvern Instruments Zetasizer Nano-S machine, operating with a wavelength of 632 nm.

Polymerosome preparation

The polymerosome formulations were prepared by a hydration-extrusion method identical to methods used for liposome preparation. As an example, the preparation of P1 -2 is described here. Aliquots of chloroform stock solutions containing the polymerosome constituents were added together in a flask to obtain a solution with 10 mg/mL PiB1000-PEG350-Me, 200 nmol DPA, and 7 nmol PtOEP. The organic solvent was removed by rotary evaporation and subsequently under high vacuum for at least 30 minutes to create a transparent polymer film. 1 .0 mL DPBS buffer was added and the polymer film was hydrated by 3-5 cycles of freezing the flask in liquid nitrogen and thawing in warm water (50 °C). The resulting dispersion was extruded through a Whatman Nuclepore 0.2 μιη polycarbonate filter at room temperature at least 1 1 times using a mini-extruder from Avanti Polar Lipids, Inc. (Alabaster, Alabama, USA). The number of extrusions was always odd to prevent any unextruded material ending up in the final sample. The extrusion filter remained colorless after extrusion, with only weak fluorescence under 365 nm UV light, suggesting near-complete inclusion of the TTA-UC compounds in the polymer membrane. The polymerosomes were used for physical characterizations within 24 h after preparation. The average polymerosome size and polydispersity index were measured with a Malvern Instruments Zetasizer Nano-S machine, operating with a wavelength of 632 nm.

Liposome uptake in 3T3 cells and fluorescence imaging: DMPC liposomes (composition: 95.6 mol% DMPC, 3.8 mol% DSPE-PEG-2000, 0.48 mol% perylene, 0.024 mol% PdTPTBP, 2.5 mM DMPC bulk concentration in PBS) were assembled using the standard hydration- extrusion protocol. 3T3 cells were grown on thin glass microscopy slides and the growth medium was replaced by DMEM medium (-FBS +pen/strep +glutamax) 1 .5 h before liposome incubation. The cells were then incubated for 3 h with a 5:3 v/v mixture of the liposome dispersion and DMEM medium (-FBS +pen/strep +glutamax) and subsequently washed thoroughly to remove liposomes that were not taken up. The cells were fixated with paraformaldehyde (4% in PBS) for 15 min and thoroughly washed with PBS. Before imaging the buffer was replaced by PBS that was supplemented with 0.3 M Na ₂ S0 ₃ to induce anoxic conditions. Imaging was performed with a standard inverted microscope with 405 nm and 630 nm lasers as excitation sources and 100x objective. For imaging of upconversion, 630 nm excitation was cleaned up with a 630 nm band pass filter, and emission was filtered through a 633 nm notch filter and 575 nm short pass filter to selectively allow passage of 450-575 nm light. Temperature was controlled with a mini-incubator from Tokai Hit.

Instrumentation

UV-Vis absorption spectrometry

Regular UV-Vis absorption spectra were taken on a Varian Cary 50 UV-Vis spectrometer or with a customized setup with parts from Avantes (Apeldoorn, The Netherlands), using an AvaLight-DHc as light source and an 2048 StarLine spectrometer as detector, both connected to a CUV-UV/VIS-TC temperature-controlled cuvette holder at a 180° angle using FC-UVxxx-2 (xxx = 200, 400, 600) optical fibers.

Emission spectrometry

Emission spectra with excitation wavelengths 416 and 378 nm were taken on a Shimadzu RF-5301 PC spectrofluorimeter at ambient atmosphere. Emission spectra with excitation wavelengths 532 nm and 630 nm were measured in the same setup as for upconversion emission spectrometry, detailed below, and were always collected from deoxygenated samples that had been thoroughly bubbled with argon (Argon 4.6, LindeGas) for at least 30 minutes with a rate of ~2 bubbles per second. Setup for upconversion emission spectroscopy

The following setup was used for upconversion emission spectroscopy is shown in Figure 10:

Upconversion emission spectra were measured with a custom-built setup. All optical parts were connected with FC-UVxxx-2 (xxx = 200, 400, 600) optical fibers from Avantes (Apeldoorn, The Netherlands), with a diameter of 200-600 μιη, respectively, and that were suitable for the UV-Vis range (200-800 nm). The excitation source was either a continuous wave Aries 150 532 nm portable DPSS laser from LaserGlow (Toronto, ON, Canada), or a clinical grade Diomed 630 nm PDT laser. The 630 nm light was filtered through a FB630-10, 630 nm band pass filter (Thorlabs, Dachau/Munich, Germany) put between the laser and the sample. The excitation power was controlled using a NDL-25C-4 variable neutral density filter (Thorlabs), and measured using either a PM20 optical power meter or a S310C thermal sensor connected to a PM100USB power meter (Thorlabs). Sample deoxygenation was performed in an external ice-cooled pear-shaped flask by bubbling argon for 30 minutes with a rate of 2 bubbles per second, after which the sample was transferred to the cuvette by cannulation under argon. The sample was held under argon in a 104F-QS or 104F-OS semi- micro fluorescence cuvette from Hellma GmbH & Co. KG (Mullheim, Germany) in a CUV- UV/VIS-TC temperature-controlled cuvette holder (Avantes), and was irradiated from the top with a collimated 2.6 mm diameter beam or from the side with a collimated 4 mm diameter beam going through the short side (4 mm optical path length) of the cuvette. Emission measurement was performed by means of a 2048 or 2048L StarLine CCD spectrometer from Avantes under a 90° angle with respect to excitation. The excitation light was rejected using either a NF533-17 533 nm or NF633-25 633 nm notch filter from Thorlabs.

Quantum yield measurement setup

The setup used for determining O _uc is shown in Figure 1 1 For QY measurements with the PtOEP-DPA molecule couple, the excitation source was a 532 nm continuous wave Aries 150 portable DPSS laser from LaserGlow (Toronto, ON, Canada) with a beam diameter of 1 .5 mm. For QY measurements with the PdTPTBP/perylene molecule couple, the excitation source was a clinical grade Diomed 630 nm PDT laser. This laser only operates with an optical fiber connection, and was therefore connected with a FC-UV200-2 optical fiber (Avantes, Apeldoorn, The Netherlands) to a collimating lens, after which the light passed a 630 nm band pass filter and a mechanical iris to produce a ca. 2 mm beam. The excitation power was measured using a S310C thermal sensor connected to a PM100USB power meter (Thorlabs, Dachau/Munich, Germany). An AvaSphere-30-IRRAD integrating sphere, customized with a sample holder and an extra aperture, and an AvaSpec-ULS2048L StarLine CCD spectrometer were purchased from Avantes. The integrating sphere and spectrometer were calibrated by the manufacturer so that the observed intensities are expressed with the dimension of a photon flux (photon. s-1 .m-2, where the surface is that of the aperture in the integrating sphere). The filter holder was fabricated by our own mechanical department, and held a NDL-25C-4 variable neutral density filter (Thorlabs), or a NF533-17 notch filter (Thorlabs) in case of excitation with 532 nm light, or a 575 nm short pass filter (Edmund Optics, York, United Kingdom, part no. 64-603) in case of excitation with 630 nm light. The FC-UVxxx-2 (xxx = 200, 400, 600) optical fibers with 200-600 μιη diameter were purchased from Avantes and were suitable for the UV-Vis range (200-800 nm). Spectra were recorded with Avasoft software from Avantes (The Netherlands) and further processed with Microsoft Office Excel 2010 and Origin 8.5 software.

Quantum yield determination procedure

The quantum yield of upconversion (O _uc ) is defined as number of upconverted photons q _P -em

°uc = 2 x — —— 7 = 2 x - Equation 1

number or low-energy photons absorbed q _P -abs where q _p - _em is the emission photon flux of the singlet annihilator species (in photons. s ^"1 ) and q _p -abs is the photon flux absorbed by the sensitizer species (in photons. s ^"1 ). The factor 2 arises from the fact that upconversion intrinsically has a maximum quantum yield of 50% and thus must be scaled to attain a maximum value of 100%. O _uc can be calculated by f ^A 2

'annihilator( )dA

Equation 2

Qp-abs where LnnihiiatorM is the spectral luminescence intensity (in photons. s ^' Vnm ^"1 ) of the annihilator species, and λ ₂ are the low- and high-wavelength boundaries, respectively, of the upconverted annihilator emission spectrum. q _p . _ab s is determined by subtracting the spectral light intensity of the excitation source that has passed through the sample ( c-sampie, in photons. s ^' Vnm ^"1 ) from the spectral light intensity of the excitation source that has passed through a blank sample (l _exo -biank, in photons. s ^"1 .nm ^"1 ), and by integrating over the excitation wavelength range λ ₃ to λ ₄ , see Equation 3. The blank sample resembled the upconverting sample in all ways, except that it did not contain any sensitizer, and thus did not absorb at the excitation wavelength.

Qp-abs = lexc-blankW ^~ 'exc-sample O^dA Equation 3 Equation 2 can then be expressed as Equation 4:

Φ uc = 2 x Equation 4

WdA

The spectrometer and the integrating sphere were calibrated by the manufacturer so that the observed intensities are directly proportional to the photon flux, i.e. Ι(λ) oc [mol of photons. s ^-1 . nm ^-1 ] . Therefore, integrating these values over the relevant wavelength regions gave directly the flux of photons arriving at the spectrometer.

Because the intensity of the upconverted light is relatively low compared to that of the exciting laser source the absorption and emission of the sample cannot be measured at the same time. In other words, the laser light saturates the spectrometer, which prevents upconversion to be measured. To circumvent this problem, the absorption was measured using a variable neutral density filter with known attenuation (typically F _attn ~0.01 , i.e., ~99% attenuation). This filter was placed between the integrating sphere and the spectrometer to measure the absorbed photon flux, whereas it was replaced for the measurement of the upconverted emission by a notch (533 nm) or short pass filter (<575 nm) to remove the excitation wavelength. Thus, Equation 4 was changed into Equation 5. The attenuation factor Fattn was assumed to be constant over the wavelength range of the laser excitation. Then, although at the first order the notch or short pass filter was assumed to only block the laser signal from reaching the spectrometer, in reality there was a small reduction of transmission for wavelengths situated in the upconversion range as well. This filtering can be corrected when calculating O _uc by dividing the upconversion luminescence intensity by the transmission curve Τ(λ) of the notch or short pass filter in the wavelength range of the upconverted light. The corrected equation for O _uc became:

Φ uc = 2 x Equation 5

The boundary wavelengths that were used for determining O _uc given in the main text, as well as the measured values for q _p . _em and q _p - _a bs, are given in Error! Reference source not found.. Table 4: Measured values used for Φ„ _η determination λ1 A2 A3 A4 q _p . _em q _P abs ue (nm) (nm) (nm) (nm) (nmol (nmol

photons.s ¹ ) photons.s ¹ )

UCL-1 390 525 520 545 0.0384 3.32 0.023

Couple 1

toluene ³ 390 525 520 545 0.124 4.81 0.051

UCL-2 400 575 615 645 0.0306 1 1 .4 0.0054 Couple 2

toluene ^b 400 600 615 645 0.0420 7.13 0.012 ^a [PtOEP] = 3.5 μΜ, [DPA] = 100 μΜ. [Pc!TPTBP] = 2.5 μΜ, [perylene]

For each measurement, two samples were prepared: one blank sample, deprived of sensitizer (L-1 b or L-2b), and one with the upconversion couple (UCL-1 or UCL-2). Since the concentration of the sensitizer is very small compared to the other liposome constituents ( [sensitizer]≤ 0.02 mol%), we assume that removal of the sensitizer from the lipid mixture did not influence the physical properties of the liposomes (membrane fluidity, scattering properties of the sample, or others). The upconverting sample or the blank sample was loaded into specially designed measurement tubes that were made of a quartz EPR-tube bottom (± 7 cm) fused to a NS-14 glass connector (± 2 cm), at the top of which a septum was adapted. The tube fit precisely a hole made in the integrating sphere, and reached the center of the sphere, where it was hit by the excitation laser beam. Deoxygenation of the sample was performed in a separate ice-cooled, pear-shaped flask, by bubbling the sample with argon for at least 30 minutes with a rate of ~2 bubbles per second. The degassed sample was then transferred in the measurement tube by cannulation in the strict absence of oxygen. Degassing in the tube was found to be impossible due to foam formation.

Photosubstitution experiments using red light

The following setup was used for photosubstitution experiments using red light:

1 mL of the liposome mixture, prepared as described in the main text, was deoxygenated by bubbling argon through the sample with a rate of ~2 bubbles per second for at least 30 minutes in an external ice-cooled pear-shaped flask, after which the sample was transferred by means of cannulation under argon to a 104F-QS or 104F-OS semi-micro fluorescence cuvette from Hellma GmbH & Co. KG (Mullheim, Germany) in a CUV-UV/VIS-TC temperature-controlled cuvette holder from Avantes. The sample was held under argon atmosphere at a constant temperature of 298 K and irradiated for 4 hours from the top with a 120 mW 630 nm laser light beam from a clinical grade Diomed 630 nm PDT laser. The laser was collimated to a beam of 2.6 mm diameter to reach an intensity of 2.3 W.cm ^"2 ; in such conditions, a cylinder of approximately 0.13 cm ³ was simultaneously excited by the laser. UV-Vis absorption spectra were measured using an Avalight DH-S-BAL halogen-deuterium lamp (Avantes) as light source and an 2048L StarLine spectrometer (Avantes) as detector, both connected to the cuvette holder at a 180° angle. A UV-Vis absorption spectrum was measured every 15 min; each time the laser was switched off, the halogen-deuterium lamp was turned on, a spectrum was recorded, the halogen-deuterium lamp was switched off, and the laser was switched on again. Each UV-vis measurement took approximately 10 seconds in total. The baseline of each spectrum was corrected for Tyndall and Rayleigh scattering and drift of the halogen-deuterium light source, using Microsoft Excel 2010 and Origin 8.5 software.

Beam profiling

A beam profiler was used for measuring the beam diameters of the laser beams in the aforementioned setups. It consisted of a Trust Webcam Spotlight Pro, of which the front lens was pried off and replaced by NE510A (OD=1 ) and NE520A (OD=2) absorptive neutral density filters (Thorlabs). The laser beam was pointed directly on the photovoltaic chip of the webcam (4.8 mm wide and 3.6 mm high). Then, 1/e ² laser beam diameters in pixels were determined by Beams, an open source beam profiling software downloadable from http://ptomato.name/opensource/beams/beams.html . The beam diameter in millimeters was calculated by dividing the average beam diameter in pixels by the total number of horizontal pixels and multiplying this with the chip width in millimeters. For example, the diameter of the

339 Dx

beam in Fig ³ ure 13 was determined to be— 640— px x 4.8 mm = 2.5 mm.

References

[1 ] a) L. Zayat, C. Calero, P. Albores, L. Baraldo, R. Etchenique, J. Am. Chem. Soc.

2003, 125, 882-883; b) P. J. Bednarski, F. S. Mackay, P. J. Sadler, Anticancer Agents Med Chem 2007, 7, 75-93; c) N. J. Farrer, L. Salassa, P. J. Sadler, Dalton Trans. 2009, 10690-10701 ; d) U. Schatzschneider, Eur. J. Inorg. Chem. 2010, 2010, 1451 -1467; e) S. L. H. Higgins, K. J. Brewer, Angew. Chem., Int. Ed. 2012, n/a-n/a; f) B. S. Howerton, D. K. Heidary, E. C. Glazer, J. Am. Chem. Soc. 2012, 134, 8324- 8327; g) M. A. Sgambellone, A. David, R. N. Garner, K. R. Dunbar, C. Turro, J. Am. Chem. Soc. 2013, 135, 11274-1 1282.

[2] R. R. Sawant, V. P. Torchilin, Soft Matter 2010, 6, 4026-4044.

[3] a) T. N. Singh-Rachford, F. N. Castellano, Coord. Chem. Rev. 2010, 254, 2560-2573; b) J. Zhao, S. Ji, H. Guo, RSC Advances 2011 , 1, 937-950; c) Y. C. Simon, C.

Weder, J. Mater. Chem. 2012, 22, 20817-20830; d) Y. Y. Cheng, B. Fuckel, T.

Khoury, R. I. G. C. R. Clady, N. J. Ekins-Daukes, M. J. Crossley, T. W. Schmidt, J. Phys. Chem. A 2011 , 115, 1047-1053.

[4] Y. Murakami, H. Kikuchi, A. Kawai, J. Phys. Chem. 62013, 117, 2487-2494.

[5] a) C. Wohnhaas, K. Friedemann, D. Busko, K. Landfester, S. Baluschev, D. Crespy, A. Turshatov, ACS Macro Letters 2013, 446-450; b) R. R. Islangulov, J. Lott, C. Weder, F. N. Castellano, J. Am. Chem. Soc. 2007, 129, 12652-12653; c) P. B.

Merkel, J. P. Dinnocenzo, J. Lumin. 2009, 129, 303-306; d) A. Monguzzi, R. Tubino, F. Meinardi, J. Phys. Chem. A 2009, 113, 1171 -1 174; e) T. N. Singh-Rachford, J. Lott, C. Weder, F. N. Castellano, J. Am. Chem. Soc. 2009, 131, 12007-12014.

[6] a) A. Turshatov, D. Busko, S. Baluschev, T. Miteva, K. Landfester, New J. Phys.

2011 , 13, 083035; b) J.-H. Kang, E. Reichmanis, Angew. Chem., Int. Ed. 2012, 51, 1 1841 -1 1844; c) J.-H. Kim, J.-H. Kim, J. Am. Chem. Soc. 2012, 134, 17478-17481 ; d) Q. Liu, B. Yin, T. Yang, Y. Yang, Z. Shen, P. Yao, F. Li, J. Am. Chem. Soc. 2013, 135, 5029-5037; e) K. Tanaka, H. Okada, W. Ohashi, J.-H. Jeon, K. Inafuku, Y.

Chujo, Bioorg. Med. Chem. 2013, 21, 2678-2681 ; f) C. Wohnhaas, V. Mailander, M. Droge, M. A. Filatov, D. Busko, Y. Avlasevich, S. Baluschev, T. Miteva, K.

Landfester, A. Turshatov, Macromol. Biosci. 2013, DOI: 10.1002/mabi.201300149; g) Y. C. Simon, S. Bai, M. K. Sing, H. Dietsch, M. Achermann, C. Weder, Macromol. Rapid Commun. 2012, 33, 498-502.

[7] J.-C. Boyer, F. C. J. M. van Veggel, Nanoscale 2010, 2, 1417-1419.

[8] a) R. N. A. H. Lewis, N. Mak, R. N. McElhaney, Biochemistry 1987 , 26, 61 18-6126; b) C.-H. Chang, H. Takeuchi, T. Ito, K. Machida, S.-i. Ohnishi, J. Biochem. 1981 , 90, 997-1004; c) D. Marsh, Handbook of Lipid Bilayers, 2nd ed., Taylor & Francis Group,

LLC, Boca Raton, FL, USA, 2013.

[9] a) S. Bonnet, B. Limburg, J. D. Meeldijk, R. J. M. K. Gebbink, J. A. Killian, J. Am.

Chem. Soc. 2010, 133, 252-261 ; b) R. E. Goldbach, I. Rodriguez-Garcia, J. H. van

Lenthe, M. A. Siegler, S. Bonnet, Chem.Eur. J. 2011 , 17, 9924-9929; c) A.

Bahreman, B. Limburg, M. A. Siegler, R. Koning, A. J. Koster, S. Bonnet, Chem.Eur.

J. 2012, 18, 10271 -10280.

[10] D. Needham, T. J. Mcintosh, D. D. Lasic, Biochim. Biophys. Acta, Biomembr. 1992,

1108, 40-48.

[1 1 ] a) C. LoPresti, H. Lomas, M. Massignani, T. Smart, G. Battaglia, J. Mater. Chem.

2009, 19, 3576-3590; b) D. E. Discher, A. Eisenberg, Science 2002, 297, 967-973; c)

J. P. Jain, W. Y. Ayen, N. Kumar, Curr. Pharm. Des. 2011 , 17, 65-79.

[12] J. E. Puskas, Y. Chen, Y. Dahman, D. Padavan, Journal of Polymer Science Part A:

Polymer Chemistry 2004, 42, 3091 -3109.

[13] K. J. Takeuchi, M. S. Thompson, D. W. Pipes, T. J. Meyer, Inorg. Chem. 1984, 23,

1845-1851 .

[14] U. G. Karl, DE), Sierakowski, Claudia (Seeheim-Jugenheim, DE), Darijo, Mijolovic (Mannheim, DE), Haberer, Markus (Osnabruck, DE), Hartl, Helmut (Braunschweig, DE), Basf Aktiengessellschaft (Ludwigshafen, DE), United States, 2008.