ATP-Responsive Drug Delivery Systems

1. Introduction

Advances in material science brought about a wide range of drug delivery systems, from traditional bulk controlled release formulations to modern nanotechnology based carriers, for improved drug efficacy. A wide collection of nanoparticle frameworks, like metallic nanoparticles, liposomes, polymeric nanogels, protein or DNA nanoassemblies, were employed in constructing nanocarriers. Basic nanoparticle framework can be viewed as the chassis of a motor vehicle, upon which various functional modules could be installed, such as targeting ligands for vehicle “steering”, drug binding and environment-sensing moieties for cargo loading and unloading. Uncontrolled drug release from nanocarriers along their way to diseased areas significantly compromises the therapeutic effects while increasing the risk of side effects. The concept of “on-demand” therapeutics would be more promising when the nanocarriers are “smart” to perceive their surrounding environment and react correspondingly. To generate these “smart” formulations for precise delivery, numerous types of stimuli-responsive materials were developed and incorporated into the nanocarriers [1].

Our group has been devoted to the development of “smart” nanocarriers responsive to a variety of triggers for cancer and diabetes treatment. For example, external physical triggers, like focused ultrasound [2] or tensile strains [3] were applied as hints to control drug dosing. Internal physiological triggers, like overexpressed enzymes (such as furin [4] or hyaluronidase [5]), acidic environment [6], reducing gradients [7], hypoxia [8] or elevated blood glucose levels [9] were also harnessed for controlling drug release.

In this editorial, we discuss an emerging strategy utilizing the “molecular unit of currency” in biological energy transfer - ATP - as a trigger for therapeutic delivery. The sharp concentration contrast between extracellular (<0.4 mM) and intracellular (1-10 mM) environments makes ATP a feasible cue for regulating drug release [5].

2. ATP responsive drug delivery systems

To make a nanocarrier responsive to ATP, functional modules capable of differentiating ATP from other cytosolic components need to be incorporated. By far, two types of modules were adopted based on the specific recognition of ATP by biomacromolecules: 1) single stranded DNA (ssDNA) aptamers that specifically bind ATP [10]; or 2) enzymes that consume ATP as a source of energy [11]. The ssDNA aptamer with high binding affinity toward adenosine/ATP was selected in vitro from a large pool of random ssDNA sequences. It has become a popular ATP-responsive module adapted into diverse nano-formulations due to its relatively short sequence (~ 30 bases), easy functionalization and specific response.

In a polymeric nanogel based nanocarrier (Figure 1a) [5], the ssDNA aptamer was hybridized with its complementary nucleotides to form a DNA duplex, which contains a “GC” pair for loading the small molecule anticancer drug Doxorubicin (DOX). In the absence of ATP, the DNA duplex was rather stable to hold the DOX payload; while high concentrations of ATP will compete with the complementary DNA to bind the ATP aptamer, dissociating the duplex and releasing the loaded drug. To neutralize the strong negative charge of the DNA duplex, thus condensing its size in solution for packing into a nanoparitcle, the positively charge peptide protamine was complexed with the DNA duplex. The peptide/DNA complex loaded with DOX formed the core of the particle, upon which a negatively charged polymer hyaluronic acid was coated. The hyaluronic acid coating served three main purposes: 1) it protected the peptide/DNA core from premature degradation while in circulation; 2) the hyaluronic acid is an active targeting ligand that binds receptor like CD44 and RHAMM on cancer cell membrane; and 3) the enzyme hyaluronidase overexpressed in tumor microenvironment could function as an extracellular trigger in addition to intracellular ATP for prompting drug release. In an in vivo xenograft tumor model, systemic administration of the nanogel based particle exhibited higher accumulation at the tumor (~ 4 fold more DOX) owing to passive as well as active targeting effects as compared with non-gel coated particles. After exposing the peptide/DNA duplex per hyaluronidase degradation, the positively charged protamine stimulated endosome escape, escorting the DNA duplex loaded with DOX into the cytosol where the high levels of ATP triggered DOX release. The released DOX gradually accumulated in the nucleus by diffusion. Formulations with the ATP aptamer showed significantly higher tumor growth inhibition effect than non-ATP responsive control groups.

Figure 1.

Representative ATP responsive drug delivery systems. a) Nanogel using ATP responsive DNA aptamer for intracellular delivery of anticancer drug DOX. b) Protein nano-assembly as ATP responsive system based on an ATP consuming protein. Reproduced with permission from [5] and [11].

In addition to directly using innate ATP from intracellular compartment for drug release, a co-delivery system that applied extrinsically supplemented ATP was also developed using liposomes in case the intrinsic ATP level in targeted cells is not sufficient to trigger drug release [12]. In this strategy, the DOX loading DNA duplex and ATP were encapsulated in two liposomes separately with endo-lysosomal acidity as a spatiotemporal cue for liposome fusion. The protamine peptide/DNA duplex complex loaded with DOX was encapsulated into to a fusogenic liposome, which is integrated with the fusogenic lipid dioleoylphosphatidylethanolamine (DOPE). With the assistance of a cell-penetrating peptide, the acid sensitive DOPE could induce the fusion of the DOX loaded liposome with the ATP providing liposome. Further fusions of the liposomes with the endosomal membrane lead to the release of the loaded DOX into the cytosol.

Furthermore, the ssDNA aptamer can serve as a “cage” group toward nanomaterials for loading and releasing different therapeutics with a relatively high loading capacity. For example, the homogeneous pores (2-10 nm) of mesoporous silica nanoparticles (MSN) could be used as a general cavity for loading hydrophilic as well as hydrophobic drugs. When the ATP responsive aptamer hybridized with complementary DNA oligos were grafted onto MSN surface and block the pores, the ON/OFF state of the nano-channels was controlled by the concentration of ATP [13]. Besides MSN, graphene oxide was also demonstrated as a suitable nanocarrier for ATP responsive drug delivery where the drug could be loaded by π-π stacking [14]. The ATP aptamer worked as bridge linking two DNA oligo modified graphene oxide sheets and locked the drugs in between. The presence of ATP would dissociate the assembly and release the drug. A recent study also reported the application of DNA to form a stabilizing shell to coat CaCO₃ microparticles via a lay-by-lay deposition method where the ATP aptamer served as crosslinker of the DNA shell [15]. This method allowed ATP responsive delivery of several different theranostic agents simultaneously.

Aside from the popular ATP aptamer based delivery systems, protein capable of utilizing ATP as substrate, namely exhibiting ATPase activity, was also demonstrated as a module in designing ATP responsive drug delivery systems. As demonstrated by Aida and coworkers (Figure 1b) [11], a mutant form of the molecular chaperone GroEL, which is a extra-large protein (~800 KDa) with a large cavity to capture denatured proteins, was assembled into a tube like structure via merocyanine (MC) modification and Mg²⁺ based coordination. Drugs could be loaded after conjugation onto a guest protein that fit the cavity. Hydrolysis of ATP into ADP caused a conformational change to the chaperone, generating an internal mechanical force to disrupt the Mg²⁺ based coordination, leading to tube disassembly and release of the guest molecules. Further modification of the chaperone with boronic acid rendered cell membrane permeable to the protein assembly, making the protein nanoparticle a promising nanocarrier for intracellular ATP responsive drug delivery.

3. Expert Opinion

ATP is a new member of physiological triggers to achieve “on-demand” therapeutic delivery with several merits: 1) high intracellular ATP concentration and sharp concentration contrast between intracellular and extracellular environment make ATP a robust trigger signal to reduce premature drug release before cellular uptake and enhance intracellular accumulation of drugs; 2) as a common “currency” for energy transfer, ATP is involved in many biological processes, making ATP-based response a general stimuli-responsive strategy that it does not need any special equipment, such as ultrasound or light, to trigger drug release; 3) the ATP level can also be further regulated by other metabolic elements, such as glucose, which makes the response regulation mechanisms broad; 4) a diverse range of interactions between drug and ATP responsive nanocarrier can be used for drug loading, such as intercalation into DNA strands[5, 12], conjugation into a gust protein of molecular chaperones[11], diffusion into the nano-cavities of MSN[13], adsorption on to grapheme oxide sheet via π-π stacking [14] or co-precipitation into CaCO3 particles[15], making ATP responsive drug delivery system a general strategy for delivering therapeutics ranging from small molecules to macromolecular proteins. Though the concept is sound, current systems using ATP as a drug delivery trigger are mostly in the proof-of-principle stages. To make these systems practical for clinical applications, further investigations are needed. First, fundamental study of the cellular-level ATP concentrations in different organelles is essential for further designing more precise and efficient delivery carriers utilizing ATP trigger. Second, current systems are mostly based on the ATP aptamer, which does not have the capability to differentiate ATP from ADP. Potential optimization of its sequence for enhanced ATP selectivity would further improve the specificity of the system. Third, to better use the ATP signal, more types of ATP sensors with diverse range of ATP sensitivity need to be developed to better differentiate diseased cells from normal ones. The abundant collection of proteins that need ATP as a cofactor would be a valuable pool for identifying suitable candidates for ATP mediated drug delivery. Last but not least, Since the ATP binding modules are basically bio-macromolecules like DNA or protein, potential concerns for immunogenicity from the components need to be addressed before clinical translations.