The discovery of nitric oxide (NO) as a signalling molecule in various physiological and pathological pathways has spurred research in the design of exogenous NO donors as drugs. In recent years, metal nitrosyls (NO complexes of metals) have been investigated as NO-donating agents. Results from our laboratory during the past few years have demonstrated that metal nitrosyls derived from designed ligands can deliver NO under the total control of light of various frequencies. Careful incorporation of these photoactive nitrosyls into polymer matrices has afforded a set of nitrosyl–polymer composites that can be used to make such NO delivery site-specific. The composite materials have shown excellent antineoplastic and antimicrobial actions in several in vitro experiments. This review highlights our key results in the context of recent developments in this area of NO donors that deliver NO on demand.
Nitric oxide (NO) is a vital mediator of a diverse set of biological functions in mammalian physiology with the capacity to function as a signalling molecule, cytoprotectant and cytotoxin . The propensity of NO function in mammalian systems is controlled through the highly regulated activity of NO-producing enzymes (NO synthase) and systemic and local networks of transporting and storing molecules. Together, these players maintain a delicate homeostasis of this highly diffusible and reactive gas . Disruption of NO homeostasis through dysregulated or inadequate NO production has been implicated in numerous human pathologies . Methods for the therapeutic intervention in the homeostasis of NO have therefore been developed, and recovery from acute health crises has been achieved by the administration of an exogenous source of NO . The therapeutic use of exogenous NO was first reported in 1867 by Thomas Lauder Bruton, who noted that insufflation of amyl nitrate caused rapid relief of angina pectoris in ischaemic patients . Although the vasodilatory property of alkyl nitrates was suggested to play a role in the observed effects, the exact pharmacological effect of nitrovasodilators (a misnomer) was identified following two paramount discoveries. The first breakthrough came over a century after Bruton's discovery when, in 1977, Murad and co-workers  reported the generation of free NO from decomposition of several nitrovasodilators under physiological conditions and the nearly identical vasodilatory properties of the NO-donating molecules and NO(g). In the next decade, the work of Luis Ignarro and Robert Furchgott demonstrated that the vasodilatory property of NO resulted from its activation of soluble guanylyl cyclase, which initiates a signal cascade leading to relaxation of the smooth muscles surrounding the vasculature and dilation of the blood vessels [7–9]. The discovery of NO as a signalling molecule stimulated further research that rapidly established the biological effects of NO in blood pressure regulation, neurotransmission, cellular protection, innate immune response to infection and cellular apoptosis . It is now evident that NO is the evolutionary choice for the armamentarium of our body's own defences against infection and cell proliferation. This realization has inspired researchers to emulate such an immune response through targeted delivery of cytotoxic doses of NO under controlled conditions.
The versatile functions of NO in mammalian biology are closely linked to the ability of NO to convert between various oxidation formulations in the biological redox milieu . The enzymatic production of NO requires an aerobic environment owing to the obligatory role of O2 in the NO-generating oxidation of l-arginine by NO synthase. Therefore, newly generated NO may undergo oxidative processing steps that result in formulations involving N atom oxidation states from +1 to +5 before encountering a cellular target. In addition to the high reactivity as a free radical, the conversion of NO to various oxidative congeners confers on NO the potential to invoke a broad range of reversible and irreversible modifications of various cellular targets. These congeners include species with enhanced toxicity (such as nitrogen dioxide, NO2), increased reactivity (dinitrogen trioxide, N2O3), oxidative activity (peroxynitrite, ONOO−) and greater stability (nitrite and nitrate, NO, NO) . The production of these reactive nitrogen species (RNS) has a significant impact on the nature of NO-induced cellular stress owing to their ability to cause irreversible modification of biological molecules and genetic material as opposed to the direct cytotoxic reactions of NO with biological targets which usually involves the reversible inhibitory binding of NO to the metal centres of respiratory and anabolic enzymes . Owing to the high level of reactive oxygen species at sites of infection and neoplasm from the activation of macrophage NADPH oxidase and the high metabolic activity of the replicating cells, delivery of high doses of NO to such sites requires consideration of the effects of the production of RNS on the lifetime and diffusion of exogenously delivered NO and its impact on peripheral host tissues [12–15]. Because of vital differences in cellular structure, prokaryotes exhibit increased sensitivity to NO compared with mammalian cells. The enhanced sensitivity to NO and RNS can be explained by the relatively higher abundance of Fe/S clusters, which can be degraded by reaction with NO and RNS resulting in mobilization of free iron which then binds, oxidizes and fragments (through OH production) microbial DNA in the cytosol [13,14]. By contrast, the same concentration of NO can have the opposite effect on mammalian cells and tissues, where it functions as a cytoprotectant that neutralizes peroxides and as a proliferative agent that induces angiogenesis and the production of extracellular matrix. Although these events promote rapid wound healing during infection , low concentrations of NO can also exacerbate the metastasis of neoplasms . Such observations demonstrate the need to tailor NO-donating therapeutics to the unique pathophysiology and variable environments found at sites of infection and neoplasm [17,18]
2. Nitric oxide delivery from exogenous nitric oxide donors
The desire to modulate NO concentrations at biological targets via delivery of exogenous NO has inspired research in the area of designed molecules that release NO on demand (NO donors). NO donors, such as glycerine trinitrate, nitrosothiols and N-diazeniumdiolates (NONOates; figure 1), afford NO upon exposure to heat, oxidants or thiols and in some cases via enzymatic reactions [19–24]. Although a few of these systemic drugs have found wide use in hospitals, a lack of control on NO release often forbids the use of these NO donors for specific purposes such as selective destruction of infected or malignant sites via higher concentrations of NO. In order to avoid this problem, we decided to synthesize designed photoactive metal nitrosyls (NO complexes of transition metals) to deliver NO at selected sites through the control of light exposure. Our design has been based on specific chemical principles in addition to results of density functional theory (DFT) calculations . In order to deliver the photoreleased NO to a specific location, we have made smart use of material chemistry to isolate nitrosyl–polymer composite materials and the materials have been used to deliver NO at targeted locations via light triggering. The inorganic or polymeric matrix of the composite materials serves to localize the photoreleased NO to the site while retaining the metal nitrosyl and its photoproducts within the host matrix. This step alleviates any potential side effect(s) from the metal complexes. This review highlights the utility of our photoactive metal nitrosyls and material composites in delivering cytotoxic doses of NO to cancer cells and pathogens
3. Photoactive metal nitrosyls derived from designed ligands
Although the photolability of NO from transition metal nitrosyls has been demonstrated previously, the synthesis of photoactive metal nitrosyls in our laboratory has focused on promoting NO release with the aid of low-power light (in the mW range) of selected wavelengths. During the past few years, we have developed a series of designed ligands (figure 2) with peripheral functionalities that could be systematically altered using theoretical predictions based on DFT calculations to tune the photoactivity of the metal nitrosyls and progressively red-shift the photoband (electronic transition that leads to NO release) across the UV, visible and near-infrared (NIR) regions from 350 to 1000 nm . For the sake of brevity, this review will focus on two sets of ligand scaffolds, namely those derived from the pentadentate ligand frame PaPy3H and those derived from the tetradentate ligand frame H2bpb (Hs are dissociable carboxamide protons), that have demonstrated the utility of the light-triggered delivery of NO to biological targets. The pentadentate ligand design is more effective with Fe and Mn metal centres to achieve remarkably high quantum yields and photoactivity in the visible and NIR region up to 1000 nm. The tetradentate ligand design has been used to isolate highly stable Ru nitrosyls that allow for the tight binding of a functional ligand at the open sixth coordinate site transto NO. Thus far, we have used the open sixth coordinate site to attach (i) ligands functionalized with polymerizable groups that copolymerize with the desired polymer allowing tethering of the metal nitrosyl to a polymeric matrix and (ii) fluorescent dyes that act as light-harvesting photosensitizers to increase quantum yields and sensitivity of the dye–nitrosyl conjugate to visible light (better for phototherapy). As described below, the fluorescence to the dye-sensitized metal nitrosyls allows easy tracking of such NO-donating drugs in the cellular matrix.
The pentadentate ligand, PaPy3H (N,N-bis(2-pyridylmethyl)amine-N-ethyl-2-pyridine-2-carboxamide), was the first designed ligand used in our laboratory for the successful synthesis of metal nitrosyls that release NO under the control of low-power visible light. This ligand includes features observed in the metal-chelating locus of the microbial enzyme nitrile hydratase (Fe-NHase), which exhibits rapid photorelease of NO from its Fe(III) centre upon exposure to visible light (sunlight in Nature) to initiate diurnal activity [26,27]. These features include (i) a unique deprotonated carboxamide N-donor group that serves to stabilize the Fe(III) oxidation state observed in the natural enzyme and (ii) a negatively charged donor (thiolate in the case of Fe-NHase) transto NO which could promote photorelease through the trans-labilization effect. Upon reaction with NO(g), the Fe(III) complex derived from the deprotonated PaPy ligand afforded the first nitrosyl [(PaPy3)Fe(NO)]2+(1, figure 3), which upon illumination with low-power visible light released NO (quantum yield at 500 nm, ϕ500=0.185) quite readily [28,29]. While these results demonstrated the capability of the PaPy3H ligand to afford a photoactive metal nitrosyl, stability studies indicated that the iron nitrosyl 1 exhibits limited stability in biological media. By contrast, [(PaPy3)Ru(NO)]2+ (2) with Ru in place of Fe displays excellent stability in biological systems and releases NO upon illumination with UV light (5–20 mW; ). The structural integrity of the Ru nitrosyl makes it a useful NO donor for delivering NO to in vitro biological targets such as proteins and isolated tissues [31–33]. Therapeutic development of this nitrosyl, however, was not continued because of the detrimental effects of UV light on living systems. In the isoelectronic Mn(II) nitrosyl [(PaPy3)Mn(NO)]+ (3), the lower oxidation state of the Mn centre promotes more covalent binding of the NO ligand. As a consequence, 3 is very stable in biological media and yet exhibits electronic transitions in the visible region that lead to NO photolability [34,35]. Further red-shift of the NO-releasing photoband was achieved by increasing the extent of conjugation in the ligand frame through substitution of the pyridine group (Py) with a quinoline ring (Q), as in the Mn(II) nitrosyl [Mn(PaPy2Q)(NO)]+ (4), the first metal nitrosyl to demonstrate biological stability and NIR light activation up to 900 nm . Results of DFT calculations on the [M(PaPy3)(NO)]n+ (M=Fe, Ru, Mn) nitrosyls indicate that, for Fe and Ru analogues, electronic transition(s) from the predominantly dπ(M)-NO(π*) bonding orbital with partial carboxamide character to orbitals with dπ(M)-NO(π*) antibonding character leads to NO photolability . In the case of the Mn analogue 3, additional transitions between the Mn d-orbitals and the Py π-frame in the visible region contribute to sensitivity to low-energy light (500–800 nm) through an indirect mechanism involving intersystem crossing [25,34,35]. This kind of transition allows one to alter the ligand frame judiciously to isolate metal nitrosyls that are sensitive to NIR light. For example, we have now synthesized Schiff base ligands in which the imine-N is conjugated to a Py (SBPy3) or a Q (SBPy2Q) donor. The metal nitrosyls derived from these designed Schiff base ligands, namely [Mn(SBPy3)(NO)]2+ (5) and [Mn(SBPy2Q)(NO)]2+ (6), display enhanced NO photorelease kinetics in comparison with the carboxamide-based nitrosyls 1–4 and sensitivity to low-power (5–20 mW) light up to 1000 nm .
The tetradentate ligand frame, namely H2bpb (1,2-bis(pyridine-2-carboxamido)-benzene) and its congeners (figure 2), allowed us to synthesize stable Ru nitrosyls that could be further attached to a third functional ligand trans to NO. The deprotonated bpb2− coordinates metal centres as a tetradentate planar ligand similar to the naturally occurring porphyrin ligands with four N-donor atoms and two delocalized electrons across a conjugated set of aromatic rings. Ru nitrosyls derived from bpb2− and the related ligands with systematic substitutions on the ligand frame (figure 2) demonstrated that a progressive red-shift of the of the photoband can be achieved through careful selection of the substituents [38,39]. Attachment of electron-donating groups of increasing donor strength (H < Me < OMe) to the phenyl ring of the phenylenedicarboxamide region of the ligand frame causes a red-shift of the π(Ru–NO) →π*(Ru–NO) transition. This transition can be red-shifted further through the extension of electron conjugation in the ligand frame by substituting the pyridine donors with quinoline (Q) and isoquinoline (IQ1). Appropriate combination(s) of substituents lowers the energy of transition for NO release through different contributions of ligand character into the ground state (π(Ru–NO)) and the excited state (π*(Ru–NO)) of these species [38,39]. Although these alterations of the equatorial ligand frames of the Ru nitrosyls brought the of the photobands into the visible region, their quantum yield values remained quite modest (ϕ500=0.01–0.05), and hence the designed nitrosyls were not very effective in delivering high doses of NO to induce cytotoxic effects in malignant cells and pathogens. Additional photosensitization was achieved through direct coordination of dyes (as light-harvesting antennae) to these kinds of Ru nitrosyls in our later work . For example, when the dye resorufin (Resf; ε600=105 000 M−1 cm−1) was attached directly to the Ru centre, the nitrosyls (figure 4) exhibited enhanced NO photorelease with quantum yield values at 500 nm (ϕ500) in the range of 0.20–0.30. Effective intermixing of the dye π→π* transition(s) with transitions leading to NO photorelease leads to this enhancement . The utility of the dye-tethered Ru nitrosyls in combating malignancy has recently been demonstrated by our group. When MDA-MB-231 human breast cancer cells were incubated with the Resf-tethered Ru nitrosyl [(Me2bpb)Ru(NO)(Resf)], uptake of the NO donor was readily evident by its red fluorescence. No loss of viability was observed with these cells in the dark. However, when such cells were exposed to visible light (300 mW) for 1 min, the photoreleased NO promoted apoptosis within 6 h (as indicated by DNA fragmentation and cell wall blebbing; ). Complete loss of viability was noted within 24 h. These results demonstrate that selective NO delivery to malignant sites can be achieved with these designed NO donors under the control of light. Another advantage of the dye-tethered nitrosyls lies in the fact that the NO delivery can be easily tracked in cellular matrices owing to the strong fluorescence of these NO donors. Upon NO photorelease, the diamagnetic Ru nitrosyls are converted into paramagnetic Ru(III) photoproducts and lose their fluorescence. As a consequence, these dye-tethered Ru nitrosyls can be used as ‘NO-donors with a fluorometric on/off light-switch’ .
4. Site-specific nitric oxide delivery using novel nitrosyl–polymer composites
Once the control over NO release from the designed metal nitrosyls was achieved, the focus of our research shifted to ‘targeted delivery’. In order to deliver NO in a site-specific manner, we decided to incorporate the designed photoactive metal nitrosyls into various polymer matrices to produce nitrosyl–polymer composites that can be directly applied to the desired site [44,45]. Optically transparent and NO permeable materials were selected as hosts to maintain the quantum efficiency and NO delivery capacity. Furthermore, to ensure the therapeutic utility of the nitrosyl composites, the host matrices were limited to those with known biocompatibility. In our earlier effort, the highly efficient NO donors 3 and 4 were encapsulated within a transparent and biologically inert silicate matrix under mild conditions using sol–gel chemistry . Upon exposure to visible light (in the case of 3 and 4) and NIR light (in the case of 4), the sol–gel composites 3 • SG and 4 • SG (SG = sol–gel) demonstrated rapid change in colour owing to NO release from the incorporated nitrosyl (figure 5a). Precise spatial control of NO release from the hybrid materials was evident when a photomask was placed in between the light source and the sol–gel patch (figure 5b). The sol–gel encapsulation step led to modest inhibition of the quantum yield of NO release for both nitrosyls. For example, the quantum yield value for 3 of 0.55 (measured in H2O, 10 mW of 532 nm light) was attenuated to 0.25 for 3 • SG. In recent years, sol–gel materials have served as hosts for organic NO donors such as NONOates (figure 1; [46,47]). Schoenfisch and co-workers  have also reported NO release from a silicate-based material constructed through nitrosylation of thiol derivatized silica xerogels. While the quantum yield values of NO release reported for S-nitrosothiols are considerably lower than that of 3 • SG, the S-nitrosylated sol–gel materials have found use as coatings for blood-contacting surgical accessories and agents to inhibit bacterial adhesion/infection at the site of implanted medical devices such as pacemakers and stents. It is important to note that 3 • SG and 4 • SG are the first examples of NO-releasing materials derived from photoactive metal nitrosyls that could be activated with visible or NIR light. To date, a few Ru nitrosyls have also been incorporated into biocompatible polymer matrices and their capacities of NO release under UV illumination have been reported [49,50].
Measurement with an NO-sensitive electrode demonstrated that NO release from both 3 • SG and 4 • SG is proportional to the duration of exposure to light and the photoreleased NO can be readily delivered to reduced myoglobin in physiological buffer. Since NO exhibits strong antimicrobial activity [13,14] and has been used in the treatment of chronic infections , we decided to test the utility of the nitrosyl–polymer composites to eradicate bacterial loads in in vitro models. In our first attempt, a fibre-optic-based catheter was engineered that contains 3 • SG at its tip (figure 6). When light was sent through the fibre-optic line, the tip released high doses of NO and successfully eradicated loads of Staphylococcus aureus, Pseudomonas aeruginosa and Escherichia coli . These results strongly suggest that a device that releases NO via illumination through optical fibres may have clinical applications in combating both Gram-positive and Gram-negative bacteria in remote infected locales (such as infected body cavities in ventilator-related infections).
While sol–gel chemistry is advantageous for molecular encapsulation owing to the mild polymerization conditions, the tendency of the solvent ‘sol’ to evaporate from the sol–gel produces a brittle glass-like ‘xerogel’ that is highly susceptible to fragmentation. Such fragmentation could be detrimental during therapeutic application and limits the shelf life of sol–gel-based materials. To address such concerns, we have recently reported a flexible light-sensitive polyurethane-based film containing dispersed xerogel particles (dia = approx. 100 μm) incorporating up to 3 mol% of 3 within a silicate matrix . The polyurethane films were cast by spin coating the liquid polyurethane with dispersed xerogel particles onto a glass slide and allowing the polyurethane–xerogel–nitrosyl (PUX-NO) composite to cure overnight. When the green flexible film was placed over bacterial colonies (grown on soft-agar layers) and exposed to low-power visible light, rapid eradication of loads of S. aureus, E. coli and Acinetobacter baumannii was observed.
Hydrogels based on poly(2-hydroxyethyl methacrylate) (pHEMA) are widely used in biomedical applications (such as bandages and wound dressings) owing to their swellability, oxygen permeability and biocompatibility . We have recently incorporated the manganese nitrosyl 3 into the pHEMA matrix along with methylene blue (MB; a singlet oxygen generator under light). Upon illumination with visible light, the nitrosyl–polymer–MB composite affords a combination of NO and singlet oxygen, which leads to potent antibiotic activity against P. aeruginosa and E. coli in in vitro experiments . Although patches of this pHEMA-based material are quite convenient as a bandage material, they require a coating of polyurethane to retain the metal nitrosyl and its photoproduct(s) within the gel matrix. This requirement was circumvented in another approach in which we used [(Me2bpb)Ru(NO)(4-vpy)](BF4) (7), a Ru nitrosyl derived from the tetradentate ligand frame Me2bpbH2 (figure 2) with 4-vinylpyridine (4-vpy) as the sixth ligand. During radical-induced copolymerization of HEMA and ethyleneglycol dimethacrylate (a cross-linker), the 4-vpy ligand allowed covalent attachment of the nitrosyl with the polymer backbone . The transparent 7–pHEMA composite readily releases NO upon exposure to low-power UV light (5 mW) while completely retaining the photoproduct bonded to the polymer frame. Since NO is known to accelerate wound healing , the pHEMA–nitrosyl composites could serve as excellent wound dressings with strong wound-healing properties in addition to light-controlled antibiotic effects.
Finally, in our continuing effort to make new composites to deliver NO to selected sites under the control of light, we have recently looked into porous materials as hosts for photoactive nitrosyls. The alarming rate of infection by extensive drug-resistant (XDR) strains of the bacterium A. baumannii has plagued soldiers wounded on the battlefield during recent conflicts and has since spread to civilian communities and hospitals. Current reports indicate that up to 40 per cent of A. baumannii clinical isolates comprise strains showing resistance to all but one or two known antibiotics (referred to as XDR-Ab). To combat the spread of such pathogens, we have developed a readily usable MCM-41-type silicate-based powder with pores that entrap the photoactive metal nitrosyl 3 (figure 7; ). To increase the interaction between the pore walls and the cationic nitrosyl, an aluminosilicate-based material with 3 mol% Al substitution (Al-MCM-41) has also been used to introduce negative point charges throughout the material framework. The loading of 3 within the unidimensional pore system, determined by inductively coupled plasma mass spectroscopy, was noted up to 25 wt%. This composite material rapidly releases NO upon illumination with visible light. In an in vitro model of skin and soft-tissue infection containing approximately 106 colony-forming units per millilitre, use of 50 mg of the powder in conjunction with 1.5 h of illumination completely eradicated an XDR-Ab clinical isolate collected from the sputum of a wounded soldier. The light flux used in this experiment was equivalent to that which one experiences on a sunny day (approx. 100 mW cm−2). These results strongly suggest that powders of this type could serve as the first line of treatment for soldiers wounded on the battlefield.
To date, the use of NO in medicine has been limited to slow release of low doses into the blood stream through decomposition of various exogenous systemic NO donors (figure 1). However, site-specific delivery of this reactive gas in high doses to combat chronic infection and neoplasms still remains a challenge. Although delivery of gaseous NO directly to infected wounds has shown promise [58,59], handling of this noxious and highly diffusible gas in a hospital setting presents a discouraging task in general. In our research, both control on the dosage of NO through light-triggering and delivery to specific sites through entrapment of the photoactive nitrosyls in material hosts of various forms (patch, bandage and powder) have been achieved. Results of these proof-of-the-concept experiments demonstrate that new NO-donating formulations can indeed be isolated for specific use of this unique diatomic molecule in the treatment of various maladies. Our results along with reports from other groups, as compiled in recent reviews [60–62], now firmly uphold the potential of metal nitrosyls and NO precursors in tumour phototherapy and as antimicrobial agents.
Research in the author's laboratory was supported by grants from the National Science Foundation (CHE-0553405, CHE-0957251 and DMR-1105296).
One contribution of 18 to a Discussion Meeting Issue ‘Photoactivatable metal complexes: from theory to applications in biotechnology and medicine’.
- © 2013 The Author(s) Published by the Royal Society. All rights reserved.