Recent advances in drug release monitoring

2.1 Prodrug-based drug release monitoring

Over the past decades, prodrugs were developed as an effective approach for cancer therapy. This strategy provides great possibilities to overcome various drawbacks of drug molecules, such as poor solubility, systemic instability, and lack of cancer specificity. Recently, theranostic prodrugs, which exhibit the dual purpose of diagnosis and therapy, are emerging as a promising technology in biomedicine. The theranostic prodrugs can covalently bind with fluorophores as fluorescent reporters to monitor the drug release process, and the fluorescence signals could be of great benefit to evaluate their therapeutic effect, pharmacodynamics, and activation mechanism. These prodrugs enable researchers to know when, where, and how the active drug molecules are delivered, and can even semi-quantitatively detect the release of active drugs in a non-invasive manner.

By coupling latent fluorophores to the drug molecules with an activatable linker, real-time information about the drug release process can be obtained by using non-invasive fluorescence detection techniques. For instance, Weinstain et al. [39] reported a DDS for real-time monitoring drug release in vitro by using a coumarin-based linker. In this system, the drug release event occurring in the cytoplasm of human umbilical vein endothelial cells was observed directly with the aid of confocal microscopy, and the amount of released drug was calculated by quantifying the emitted fluorescence, allowing the prediction of therapeutic effect and potential side effects. Similarly, Bazylevich et al. [40] developed a fluorescein-based prodrug by conjugating free phenolic hydroxyl of fluorescein chemosensor to the hydroxyl group of the drugs for monitoring drug release. Hu and Zeng [41] designed a theranostic prodrug, composed of an anti-cancer drug camptothecin (CPT), a cleavable linker based on disulfide bonds, and a fluorophore naphthalimide derivative, and detected the drug release via the fluorescence resonance energy transfer (FRET) mechanism. The FRET occurring between CPT (energy donor) and naphthalimide (energy receptor) was interrupted by the high concentration of intracellular glutathione (GSH), thus achieving dual fluorescence response and specific release of CPT. Liu et al. [42] reported a ratiometric theranostic prodrug consisting of boron dipyrromethene, therapeutic drug CPT, and a long flexible disulfide linker. Once treated with dithiothreitol, the disulfide bond was cleaved and CPT was released, resulting in enhanced blue and green fluorescence emissions. The changes of the fluorescence ratio over time provided the opportunity for the real-time monitoring of drug release at cellular levels. DT-diaphorase, an enzyme overexpressed in some malignant tumors, can be exploited to catalyze the reduction of quinine propionic acid moiety incorporated into prodrug, which could be utilized for concomitant drug release imaging and cancer chemotherapy [43], [44]. In order to monitor drug release more accurately and understand pharmacological mechanisms better, Kong et al. [45] developed a dual turn-on fluorescence based controlled release system (CDox), in which doxorubicin (Dox) and fluorescent dye (CH) are conjugated by a pH-responsive cleavable hydrazone linker. The C=N isomerization and N-N free rotation efficiently quenched the fluorescence of CH and Dox in the CDox complex. When activated under acidic conditions, CDox was hydrolyzed to release Dox and CH, resulting in dual turn-on signals with emission peaks at 595 and 488 nm, respectively. Using this system, two-photon fluorescence imaging of CDox in living tumor tissues was successfully performed by high-definition three-dimensional (3D) imaging.

NIR fluorophores are often used as fluorescent reporters for in vivo drug release monitoring because NIR photons can deeply penetrate the skin and underlying tissue with minimal damage to the biological samples and with low background interference. Zhang et al. [46] reported a real-time drug-reporting conjugate (CPT-SS-CyN), composed of a NIR fluorescent cyanine-amine dye (CyN), a disulfide linker, and a model therapeutic drug (CPT) (Figure 1A). Likewise, treatment with dithiothreitol induces cleavage of the disulfide bond, resulting in the release of NIR dye and the drug CPT. The released CyN has an emission wavelength (760 nm) that is significantly different from that of CPT-SS-CyN (820 nm), enabling easy detection and monitoring of drug release. A linear relationship of the fluorescence intensity at 760 nm versus the amount of released CPT was observed, enabling quantitative monitoring of drug release in live cells and semi-quantitative monitoring in live animals in a real-time and non-invasive manner. Additionally, a NIR fluorophore, dicyanomethylene-4H-pyran derivative (DCM), was conjugated with anti-cancer CPT through a disulfide linker (DCM-S-CPT), which can be cleaved by the overexpressed intracellular GSH, making the drug activation and fluorescence turn on concomitantly. Furthermore, DCM-S-CPT has been successfully utilized for assessing drug release and cancer therapeutic efficacy by analyzing NIR fluorescence in vivo and in situ (Figure 1B) [47].

Figure 1:

Monitoring drug release by fluorescent prodrugs.

(A) Schematic illustration of real-time tracking of active drug release in vivo and proposed mechanism for concurrent NIR/CPT release from CPT-SS-CyN triggered by DTT/GSH. (Adapted from Ref. [46].) (B) Chemical structures of activatable prodrug DCM-S-CPT and non-activatable control DCM-C-CPT. Also, in vivo imaging of tumor-bearing mice at various times (0.15, 1, 2, 6, and 24 h) after intravenous injection and fluorescence images of the internal organs after anatomy. The color bars correspond to the detected fluorescence intensity. (Adapted from Ref. [47].) (C) Proposed mechanism in CPT activation and fluorescence variation of the prodrug Cy-S-CPT by the treatment of GSH. Also, in vivo and ex vivo bioimaging of tumor-bearing mice at various times (0.25, 0.5, 1, 2, 4, 8, and 24 h) after intravenous injection of the different prodrugs. Note that the green signal represents the fluorescence from the intact prodrugs (λ_ex=750 nm, λ_em=825 nm) for tracking the prodrug biodistribution, whereas red signal represents the fluorescence from CyA-K (λ_ex=530 nm, λ_em=650 nm) for tracking the location of drug activation. Yellow represents the coexistence of prodrug and drug release. (Adapted from Ref. [48].)

The fluorescent signals of single-channel fluorescent theranostic prodrugs were usually unobservable before drug release; therefore, it is difficult to simultaneously monitor their biodistribution and drug release in spatiotemporal mode. The metabolism kinetics of prodrugs in a certain organ or tissue becomes a blind spot, which is a critical obstacle in the precise diagnosis and chemotherapy. The development of two-color imaging systems with two-channel fluorescence that can distinguish the signals before and after drug release has opened a window to overcome this hurdle. Liu et al. [49] reported a novel light-activated cleavable prodrug comprising a photoremovable coumarinyl, the anti-cancer drug CPT, a cleavable linker, and the NIR fluorescent dye DCM. In this prodrug, owing to FRET, the fluorescence of coumarinyl and CPT is completely quenched while the DCM moiety shows red emission, which can be used to visualize the drug localization before irradiation. Upon one- or two-photon irradiation, the prodrug releases active CPT, and the release process is monitored via the change in the FRET process. Ye et al. [48] presented a dual-channel NIR activatable theranostic prodrug, in which the CPT was linked to a NIR cyanine dye via a disulfide linkage (Cy-S-CPT) (Figure 1C). The cleavage of the disulfide bond by endogenous GSH activated the anti-cancer drug and induced a remarkable fluorescence shift from 825 to 650 nm, thereby providing dual fluorescent channels to real-time track the prodrug biodistribution and activation in vivo. As demonstrated, the prodrug was activated in all organs, particularly in the liver after an intravenous injection, and achieved predominant accumulation and activation in tumors at 24 h after injection.

In addition, such a dual-channel fluorescence strategy can also be applied to monitor drug release and evaluate therapeutic efficacy simultaneously. As is known, the activity of intracellular apoptotic proteases such as caspase-3 can reflect the degree of cell apoptosis, which is highly related to the therapeutic efficacy. Based on these, Li et al. [50] developed a versatile prodrug for real-time drug release monitoring and in situ therapeutic efficacy evaluation. In this system, the fluorescence of both 5(6)-carboxylfluorescein and Dox was quenched by 4-(dimethylaminoazo)benzene-4-carboxylic acid via FRET. On the one hand, the acid-mediated Dox release from the prodrug was real-time monitored according to the increase of liberated Dox fluorescence. On the other hand, Dox-induced cell apoptosis was in situ assessed by the fluorescence recovery of 5(6)-carboxylfluorescein, resulting from the caspase-3-mediated cleavage of the Asp-Glu-Val-Asp peptide. Through this cascaded imaging strategy, the drug release and consequent cell apoptosis were real-time monitored, enabling the in situ detection of pharmacokinetics and treatment response.

2.2 Macromolecule-based drug release monitoring

As driven by the exponentially increasing demand for “intelligent” materials, responsive polymers, polymeric nanoparticles, micelles, and other macromolecules that are capable of adapting to changes in local environment have attracted tremendous interest in recent years. These materials with autofluorescence or labeled fluorescence have been exploited for controlled drug release with real-time tracking ability due to their low cytotoxicity and good biocompatibility.

Most of the anti-cancer drugs have little aqueous solubility and permeability. One common way to deliver hydrophobic drugs to the target site is to construct a prodrug that can create a synthetic analogue with higher hydrophilicity by modification. However, even with this approach, it is still difficult to extend the circulation time of drugs in the blood and to suppress the fast renal excretion, thus lowering the curative effect. One way to solve this problem is incorporating the drug molecules with polymer nanoparticles. Wu et al. [47] and Ye et al. [48] demonstrated that loading the prodrug into polymeric nanoparticles could significantly improve the therapeutic efficacy and lower the side effects owing to the EPR effect. Another example is perylene, which can act as an efficient phototrigger for carboxylic acids and alcohols in aqueous media under visible light irradiation. The perylene molecules can form nanoparticles with different sizes by the re-precipitation method, and the perylene nanoparticles show different spectral properties compared with the free perylene molecules. Jana et al. [51], [52] reported chlorambucil-conjugated perylene organic nanoparticles as single-component photoresponsive nanocarriers for controlling and monitoring drug release. In these systems, perylene nanoparticles performed four roles: (i) nanocarriers for drug delivery, (ii) phototriggers for drug release, (iii) fluorescent chromophores for cell imaging, and (iv) detectors for real time-monitoring of drug release.

Drug molecules can also conjugate with polymer nanocarriers to improve their efficacy and toxicity. The conjugated drugs are typically inactive in the polymer nanocarriers, and the triggered release will recover their therapeutic activity. To monitor the drug release in the polymer-drug conjugates, Shiran et al. [53] designed two N-(2-hydroxypropyl) methacrylamide co-polymer-based systems that were complementary to one theranostic nanomedicine consisting of self-quenched Cy5 as a reporter probe and paclitaxel as an anti-cancer agent. The established co-polymer systems enabled site-specific release upon enzymatic degradation in cathepsin B-overexpressed breast cancer cells, and the release of paclitaxel occurred concomitantly with the activation of Cy5 to the turn-on state. Apart from conjugating additional moieties for nanoparticle assembly, some drug molecules can self-assemble into nanoparticles for delivery. Zhang et al. [54] developed a self-carried curcumin nanodrug for highly effective cancer therapy in vitro and in vivo with real-time monitoring of drug release. Curcumin has different fluorescence characteristics in its molecular (“on” state, strong green fluorescence) and solid (“off” state, no emission because of intermolecular aggregation) forms, which was exploited in this study for monitoring the release of curcumin molecules (“on”) from drug nanoparticles (“off”) upon cell internalization.

Polymer nanoassemblies, such as polymer micelles, polymer vesicles, and polyionic complexes, are commonly created from self-assembling or cross-linked co-polymers. They are typically engineered to have a hydrophobic core to entrap drugs with low solubility and an enveloped hydrophilic shell to protect the drugs from precipitation, degradation, or non-specific binding to plasma proteins. By virtue of physically encapsulated or covalently linked fluorophores, polymeric nanoassemblies have been extensively explored for tracing the biodistribution of nanocarriers and monitoring the drug release. For example, a series of fluorescent polymeric micelles have been constructed by self-assembly of two amphiphilic co-polymers with fluorescent moieties based on different fluorine derivatives. The drug release from these polymeric micelles could be monitored by the on/off of FRET from micelles to drugs [55]. Specifically, the modulating environment of some polymer nanoparticle cores, such as hydrophobicity or pH sensitivity, can control their fluorescence property. By tethering hydrophilic polymer chains (poly(ethylene glycol), PEG) onto a single polymer backbone (branched poly[ethylene imine], PEI) while modifying the core of the nanoassemblies with lipophilic pendant (palmitate, PAL) to modulate fluorochromism of Nile blue (NB), two types of polymer nanoassemblies were synthesized: PEG-PEI-NB (2PN) and PEG-PEI-PAL-NB (3PN) [56]. 2PN has ionizable moieties in the core that can attract water and quench NB fluorescence, while 3PN has a hydrophobic core modified with PAL to increase entrapment of hydrophobic anti-cancer drugs. Therefore, fluorescence intensity increased upon addition of hydrophobic drugs or exposure to acidic environments. These two polymer nanoassemblies were used to monitor the amount of drug remaining in the hydrophobic core in real time and detect metastatic tumors in a tissue pH-dependent manner. Aibani et al. [57] exploited differences in the hydrophobic/hydrophilic balance between two isomers, spiropyran and merocyanine, to mediate drug release from a polymer micelle. The stable spiropyran with closed hydrophobic four-ring system preferentially favors a non-polar environment, whereas light irradiation can convert to its zwittrionic merocyanine counterpart with open ring, which prefers a more hydrophilic environment. In addition, the drug was quantified using a FRET pair between spiropyran-merocyanine photochromic dyad and co-encapsulated boron dipyrromethene fluorophore. Therefore, this DDS based on polymer micelles not only is capable of releasing cargo under light, but also provides a real-time analysis of unreleased cargo.

The aggregation-induced emission (AIE) molecules, such as tetraphenylethene (TPE), triphenylethene, and siloles, commonly show no emission in dispersed solution due to the decaying of the excited energy via intramolecular rotations, while they exhibit high emission at the aggregated state because of the restricted intramolecular rotations. Such AIE phenomena can also be observed when the AIE molecules are aggregated in body fluids or in solid state, which favors cell imaging, chemical sensors, and investigating drug delivery mechanisms. For instance, AIE micelles prepared by conjugation of TPE with PEG were used as both nanocarriers to deliver anti-cancer drug and intracellular imaging probe to track their locations. Liow et al. [58] constructed a thermogelling co-polymer, poly(PEG/PPG/TPE urethane), consisting of hydrophilic PEG segment, hydrophobic poly(propylene glycol) (PPG) segment, and TPE units with AIE character (Figure 2A). As the AIE of the AIE thermogel is highly dependent on the Dox concentration and temperature, the drug release status and the erosion condition of the matrix can be revealed. Similarly, the PNIPAM thermogel can combine with LaF₃:Eu³⁺ nanocrystals by surface-initiated living radical polymerization [61]. The fluorescence properties and drug release behaviors in the core-shell LaF₃:Eu³⁺ nanocrystals/PNIPAM nanogels are highly related to the temperature. The fluorescence intensity of the nanogels increases gradually with the drug release.

Figure 2:

Monitoring drug release by fluorescent macromolecules.

(A) Schematic drawing of AIE micelles, micelle aggregation, and in situ gelling DDS. (Adapted from Ref. [58].) (B) Schematic illustration of assembly, lactose-mediated endocytosis, NIR fluorescence on, and drug release of Lac-SS-DCM. (Adapted from Ref. [59].) (C) Sequence-activated AND logic dual-channel NIR fluorescent probe. Also, dual-channel fluorescence tracking of the programmable drug release in vivo (yellow-red represented at 830 nm; rainbow represented at 650 nm). (Adapted from Ref. [60].)

Further, coordination polymers comprising organic ligands and metal ions can also be used for efficient monitoring of drug release in tumors. Fe³⁺ and polyphenol gallic acid can form a polygonal nanoscale coordination polymer with extremely high Dox loading efficacy (up to 48.3%) in aqueous solution, which exhibits potent anti-tumor effect [62]. In vitro studies demonstrated that the Dox fluorescence was efficiently quenched once loaded on the coordination polymer. The acidic microenvironment in lysosome could trigger Dox release and fluorescence recovery simultaneously, leading to real-time monitoring of drug release in tumor cells. The biocompatible dipeptide molecules (tryptophan-phenylalanine) could also coordinate with Zn²⁺ to form dipeptide nanoparticles, and the intrinsic fluorescence signal of the peptides shifted from the ultraviolet (UV) to the visible range due to π-π stacking interaction of aromatic groups and the structure rigidification by Zn²⁺ [63]. After being functionalized with MUC1 aptamer and Dox, the dipeptide nanoparticles could target cancer cells and image drug release in real time.

Polymeric nanovesicles obtained via self-assembly of amphiphilic small molecules hold a significant advantage in the ability to obtain a wide range of desired functions via the incorporation of hydrophobic/hydrophilic components with rational selection. A glycol nanovesicle (Lac-SS-DCM) is self-assembled by a rationally designed amphiphilic lactose derivative [59], in which a DCM moiety (used as the hydrophobic part and the fluorescence off-on probe) was pre-quenched and conjugated with a lactose derivative (used as the hydrophilic part and the targeting ligand to galectin-overexpressing tumor cells via enhanced carbohydrate-protein interactions) via a disulfide linkage (the GSH-responsive unit) (Figure 2B). The targeting Lac-SS-DCM nanovesicles were disassembled by GSH, and simultaneously activated the dormant NIR fluorescence of DCM, allowing the drug release triggered by the high GSH concentration in cancer cells and real-time monitoring of the release process.

For precisely controlling the drug release in vivo, DDSs with programmable release behaviors have been reported, which respond to multiple stimuli. Yan et al. [60] incorporated sequence-activated fluorescence responses into DDS to monitor multi-stimuli-triggered programmable drug release (Figure 2C). The smart AND logic dual-channel NIR fluorescent probe P(Cy-S-CPT) had a hierarchical structure composed of an ionizable tertiary amine-containing diblock co-polymer for rendering an ultrasensitive response to pH changes and a dual-channel NIR fluorescence component Cy-S-CPT for tracking the biothiol-triggered drug release. The programmable drug release of this hierarchical probe was conducted in a multi-stage tumor microenvironment, acidic endocytic organelles followed by abnormal GSH-overexpressing cell cytosol, which was synchronous with dual-channel NIR fluorescence output (830 and 650 nm, respectively). Impressively, in vivo fluorescence and 3D imaging demonstrated that P(Cy-S-CPT) could precisely target tumor tissue and perform controllable programmable drug release.

2.3 Mesoporous silica nanoparticle (MSN)-based drug release monitoring

To date, MSNs have become attractive delivery vehicles because of their unique structural features that enable high drug loading, ease in structure control, and versatility in functionalization. The ordered pore network within these MSNs allows drug entrapping by diffusion. Meanwhile, the pores can be gated with various valves such as molecular motifs [64], nanoparticles [65], [66], [67], [68], [69], polymer multi-layers [70], DNA [71], [72], and proteins [73], which were designed to trigger the release of the entrapped drugs in the presence of multifarious external or internal stimuli. In light of numerous reports on the design and development of MSN-based stimuli-responsive DDSs, the development of strategies for real-time monitoring of drug release from MSNs is desired. Typically, optical agents were modified on the surfaces of MSNs or doped in the skeleton to track drug delivery and release in cells or tissues.

Lai et al. [64] designed a delicate molecular valve for the pores of MSN to control and monitor redox-responsive drug release (Figure 3A). The system comprised four components: (i) coumarin (FRET donor)-labeled-cysteine tethered MSNs as the drug carrier, (ii) fluorescein isothiocyanate (FITC, FRET acceptor)-attached β-cyclodextrin (β-CD) as the molecular cap to block the pores, (iii) disulfide linkage as the redox-responsive trigger to release the entrapped drug molecules, and (iv) FRET donor-acceptor for monitoring drug release in real time. Specifically, owing to linkage by the intact disulfide bond, the close proximity between coumarin and FITC resulted in FRET. In the presence of redox stimuli like GSH, the disulfide bond was cleaved, which led to the removal of the molecular cap (FITC-β-CD), thus triggering drug release and eliminating FRET. The release of the drugs from the MSN nanocarrier was monitored in real time following the changes in the FRET signal. It was also demonstrated that any exogenous or endogenous change in the GSH concentration would result in a change in the amount of released drug as well as concurrent change in the FRET signal, suggesting that the FRET-based MSNs could be applied to monitor the release of different drugs in real time.

Figure 3:

Monitoring drug release by fluorescent MSNs.

(A) Schematic representation of the redox-responsive FRET-MSNs. (Adapted from Ref. [64].) (B) Microcarrier fate tracking and drug release monitoring in the NIR-II window using an InGaAs charge coupled device (CCD) camera. (Adapted from Ref. [74].) (C) Fluorescence lifetime imaging for drug release monitoring. (Adapted from Ref. [75].)

By using SiO₂ as template, Chen et al. [76] synthesized rare-earth (Tb, Eu)-doped Gd₂O₂S-based radioluminescent capsules with an X-ray-excited optical luminescence that changed during the release of the absorbing chemotherapy drug, Dox. In vitro experiments have demonstrated the pH-triggered release process of Dox from nanocapsules coated with a pH-responsive polyelectrolyte multi-layer. The Dox was loaded over 5% by weight and released from the capsule with a time constant of ~36 days at pH 7.4 and 21 h at pH 5.0, respectively. Other nanomaterials with specific functions are also combined with mesoporous silica to precisely control drug release. For example, metallic nanoparticles, such as gold and silver nanoparticles, have been widely explored for application in cancer diagnosis and therapies due to their fascinating properties. Gold nanorods (AuNRs) exhibit strong absorption and photothermal effect in the NIR region owing to their longitudinal localized surface plasmon resonance. Lee et al. [77] developed a NIR-responsive DDS based on mesoporous silica-coated AuNRs (AuNR@mSiO₂) and phase change material (PCM). The AuNR core in AuNR@mSiO₂ provided excellent NIR sensitivity and could efficiently quench the fluorescence of Dox, while the mesoporous silica shell provided an increased reservoir of the loaded drug (Dox). The PCM was applied as both a thermosensitive gatekeeper and a medium for loading hydrophobic anti-cancer drugs. The photothermal effect of the AuNRs under NIR irradiation could heat the PCM above its phase transition temperature and facilitate a rapid release of Dox and concurrently fluorescence recovery. An enhanced anti-cancer effect was observed against various cell lines, and the intracellular release of the drug was easily monitored by live cell imaging.

Shan et al. [78] covalently modified the mesoporous channel with the ligand thenoyltrifluoroacetone firstly to in situ immobilize Yb³⁺ ions, thus forming a Yb(III) complex. Then, the two-photon ligand 4,4′-bis[p-(diethylamino)-α-styryl]-2,2′-bipyridine was used to coordinate with the central Yb³⁺ ions to make the Yb(III) complex possess the NIR excitation (850 nm) and NIR emission (980 nm) property. Dox was then embedded into the mesoporous silica channel by coordination and π-π interaction with the Yb(III) complex, resulting in the decrease of photoluminescence intensity of the Yb(III) complex. Upon the drug release, the photoluminescence intensity of the Yb(III) complex became stronger because the coordinated Dox diffused out and the coordination sphere of the central Yb³⁺ ions changed, thus achieving real-time drug release monitoring.

Compared to fluorescence imaging in visible range (400–750 nm) and the first NIR (NIR-I) window (750–900 nm) with low tissue penetration depth of several micrometers to millimeters because of severe absorption and scattering of photons by tissues, the second NIR (NIR-II) window (1000–1400 nm) features deeper penetration depths (1–2 cm) and higher resolution (sub-10 μm) with reduced scattering and minimized autofluorescence. Wang et al. [74] reported a novel microcarrier by integrating lanthanide-based nanoparticles (NaGdF₄: 5%Nd@NaGdF₄) emitting in the NIR-II window with MSNs (Figure 3B). The microcarrier realized in vivo semi-quantitative detection of the released drug in real time by measuring the fluorescence signals in the NIR-II window of lanthanide-based nanoparticles with an absorption competition-induced emission bioimaging system. The authors demonstrated that the microcarriers showed a prolonged residence time in the gastrointestinal tract up to 72 h and released up to 62% of the loaded drug. Moreover, minimal deposition of the microcarriers was found in non-target organs, such as the liver, spleen, and kidney. These findings provide novel insights for the development of therapeutic and bioimaging strategies.

Except for the fluorescence imaging relying on the intensity or spectral shift, fluorescence lifetime imaging is also a very attractive imaging tool for biomedical application, because it can reflect instantaneous local changes and allow distinguishing them from autofluorescence of the biological environment [79], [80], [81]. In view of these appealing features of fluorescence lifetime imaging, Valetti et al. [75] prepared label-free fluorescent mesoporous silica micro- and nanoparticles through in situ generation of carbon dots within the pores. By controlling the thermal decomposition of carboamino groups linked on the surface, three different types of mesoporous silica particles were synthesized: NLAB3, NLAB8, and NLAB14. The particles were endowed with emission in the visible region and had fluorescence lifetimes of up to 9.0 ns, which could be easily discriminated from intrinsic biological autofluorescence. The drug release was monitored successfully in biological environments by means of fluorescence lifetime imaging (Figure 3C).

2.4 Quantum dot (QD)-based drug release monitoring

QDs with the size of few nanometers are appealing candidates for nanodelivery and as optical probes in biomedical applications due to their excellent photostability, metabolic stability, and photoluminescence. Recently, several QDs such as carbon dots and silicon dots have been used to monitor drug release. Usually, the QDs were embedded into or capped onto nanocarriers to track the DDS and monitor drug release.

Carbon nanodots (CDots), owing to their high photostability, good water solubility, excellent biocompatibility, and multi-photon absorption, have emerged as a new class of fluorescence probes [82] and have also been extensively investigated for numerous applications, such as in medical imaging, in disease diagnosis, and as a print ink [83], [84], [85]. Tang et al. [86] reported a FRET-based CDot DDS (FRET-CDots) via directly coupling PEG onto the CDot surface followed by trapping the guest drug molecules within the PEG network and attaching to the CDot surface by π-π stacking, enabling real-time monitoring of drug release on the basis of the FRET signal (Figure 4A). In this case, the close proximity between CDots (serving as both excellent drug carriers and the donors) and Dox (the acceptor) made them form a FRET pair (FRET on). When the CDots were excited at 405 nm, emissions of both 498 and 595 nm were displayed, which correlated with the intrinsic emission of CDots and energy transfer from CDots to Dox, respectively. When Dox molecules are released from the CDot surface in the acidic environment, the FRET between CDots and Dox is terminated (FRET off), thus displaying an enhanced emission at 498 nm. This change in FRET signal could easily be regulated by the release of drug molecules from the CDot surface, thus allowing for convenient cell imaging and direct monitoring of the drug release process in real time. Moreover, the 3D two-photon imaging of the FRET-CDot at tumor tissues of 65–300 μm under the excitation of 810 nm suggested that the CDots had the potential to monitor drug release in deep tumor tissues.

Figure 4:

Monitoring drug release by QDs.

(A) Schematic representation of the FRET-CDot-DDS for drug delivery. Also, 3D two-photon confocal fluorescence images (under 810-nm excitation) of CDot-FA-Dox conjugate-incubated glomerular tissues, accumulated along the z-direction at a depth of 65–300 μm, incubated under pH of a) 7.4, b) 6.5, and c) 5.5. (Adapted from Ref. [86].) (B) Schematic representation of controlled release and real-time monitoring of the anticancer drug chlorambucil from ONBCbl-SiQDs. (Adapted from Ref. [87].) (C) Preparation and application of ATP-responsive GQDs-based nanocarriers. (Adapted from Ref. [88].)

Photoluminescent silicon QDs (SiQDs) have great potential in the area of fluorescence imaging owing to their unique properties, such as tunable emission, high natural abundance, low toxicity, and excellent surface tailorability [89], driving them be well explored as in vitro [90], [91], [92] and in vivo [93], [94] imaging agents. Recently, SiQD-based DDSs, which are responsive to pH changes, have been reported [95], [96], [97]. Paul et al. [87] used SiQDs as biocompatible fluorescent nanocarriers and o-nitrobenzyl (ONB) derivative as a phototrigger to construct a photoresponsive SiQD-based DDS (ONBCbl-SiQDs) for the controlled release of the anti-cancer drug chlorambucil (Figure 4B). The ONB moiety with the caged anti-cancer drug chlorambucil quenched the fluorescence of SiQDs due to the photoinduced electron transfer process, which was termed as the fluorescence “off” state. Upon irradiation with one photon or two photons, uncaging of the drug occurred and led to the fluorescence recovery of SiQDs, considered as the fluorescence “on” state. This photoswitchable fluorescence signal was used to monitor drug release in real time.

Luminescent gold nanoclusters (AuNCs) consisting of several to tens of gold atoms have recently attracted significant scientific interest in fluorescence imaging. Importantly, the latest research showed that aggregated thiolate-AuNCs exhibited enhanced luminescence owing to the mechanism of AIE, and this phenomenon was closely connected to the restriction of intramolecular motion. Cao et al. [98] encapsulated aggregated AuNCs (aAuNCs) into metal-organic frameworks (MOFs, ZIF-8), which restricted the free movements of AuNCs in aggregated states, and thus enhanced their luminescence with an improved quantum yield of up to 7.74%. Once the acidic pH-induced dissociation of the ZIF-8 framework occurred, the aAuNCs were liberated from the MOF and dispersed in the form of individual nanoclusters with weaker luminescence. By encapsulating the hydrophobic drugs into the luminescent MOF simultaneously, the aAuNC-MOFs can achieve controlled release under acidic conditions and real-time imaging of pH-dependent drug release.

Graphene QDs (GQDs) maintaining the intrinsic layered structural motif of graphene have emerged as promising nanomaterials for fluorescence imaging and biosensing [99], [100], [101], [102] due to their small lateral size, fine optical property, and superior biocompatibility. Our group has fabricated a smart drug nanocarrier based on FRET by capping GQDs (the acceptor) onto fluorescent MSNs (the donor) via ATP aptamer for real-time monitoring of ATP-triggered drug release [88] (Figure 4C). In this work, GQDs were used as both gatekeeper and quencher. By virtue of the sensitive interaction between ATP and ATP aptamer, any exogenous or endogenous fluctuation in the ATP concentration resulted in a change in drug release and concurrent variation in the FRET signal, which allowed real-time monitoring the release of drug from the pores.

2.5 Upconverting nanoparticle (UCNP)-based ratiometric luminescence for drug release monitoring

UCNPs are promising for bioimaging and biosensing owing to their minimized photodamage, suppressed autofluorescence, and high tissue penetration depth [103], [104], [105], [106], [107], [108]. The unique ability of upconverting low-energy NIR light to high-energy UV, visible, and NIR light makes UCNP an excellent nanotransducer for controlled drug release and an outstanding tracer for monitoring the drug release.

Wang et al. [109] coated a mesoporous SiO₂/Ca layer onto UCNP to form UCNPs@mSiO₂-Ca core-shell nanocomposites, which were able to deliver anti-cancer drugs and monitor drug release. Generally, the NaYF₄:Yb/Er UCNPs exhibited strong green emission (around 550 nm) and weak dark red emission (around 660 nm). Here, the incorporation of Ca into the nanocomposites induced phase transformation from a pure hexagonal phase to a cubic phase, and facilitated the occurrence of red emission, which strongly overlapped with the maximum absorbance of the anti-cancer drug zinc phthalocyanine to form a FRET pair. Consequently, drug release could be quantified by monitoring the luminescence recovery of UCNPs.

The simultaneous multiplex emission of UCNP in the UV to NIR range is important for the design of multi-functional UCNP-based DDS: (i) the individual sharp emission peaks in the UV-visible region can serve as an energy donor for luminescence resonance energy transfer (LRET) to the entrapped drugs that typically absorb in UV-visible range, thereby allowing for real-time monitoring of the release of different drugs, whereas (ii) the NIR emission remains constant irrespective of the drug molecule and hence can be used for imaging or quantifying the drug release via a ratiometric signal. The conventional method to monitor the drug release using a single signal may be affected by an unknown local concentration of nanocarriers and uncontrolled microenvironment change in biological samples. Li et al. [110] presented a rational design for ratiometric monitoring of drug release kinetics in living cells by virtue of the multi-color emission feature of UCNPs (Figure 5A). In the system of mesoporous silica-coated NaYF₄:Yb,Er nanocomposites loaded with Dox (NaYF₄:Yb,Er@mSiO₂-Dox), the green upconversion luminescence (UCL_G) of the NaYF₄:Yb,Er nanoparticles was selectively transferred to the drug Dox through the LRET mechanism, while the red upconversion luminescence (UCL_R) was essentially unchanged and used as an internal standard. This ratiometric design employed red emission as the internal standard, and thus offered an accurate and reliable platform for localized drug delivery monitoring in living cells, avoiding the interference of the internal microenvironment and dosing errors. A linear relationship between the green-to-red emission intensity ratio (UCL_G/UCL_R) and the percentage of the released Dox was established.

Figure 5:

Ratiometric monitoring of drug release by UCNPs.

(A) Schematic illustration of LRET-based upconversion nanocarriers for Dox delivery and the linear relationship between the Dox release percentage and UCL_G/UCL_R ratio. UCL_G and UCL_R denote green and red upconversion luminescence, respectively. (Adapted from Ref. [110].) (B) Schematic representation of the real-time monitoring of ATP-responsive drug release from polypeptide-wrapped TDPA-Zn²⁺-UCNP@MSNs. Also, the intracellular drug release was monitored by UCL imaging. (Adapted from Ref. [111].)

Almost simultaneously, Lai et al. [111] utilized the multi-color UCNP with a core-shell structure of β-NaYF₄:Yb³⁺/Tm³⁺@NaYF₄:Yb³⁺/Er³⁺ to construct a ratiometric probe for monitoring drug release (Figure 5B). Such a core-shell structure of UCNP minimized the cross relaxation of activators, endowing the UCNP with bright emissions in the UV to NIR region (360, 475, 542, and 656 nm) when excited with a 980-nm NIR laser. This multi-color upconversion nanoparticle was firstly coated by mesoporous silica to form a UCNP@MSN composite nanoparticle. After being functionalized with zinc dipicolylamine analogue (TDPA-Zn²⁺), a compact branched polypeptide, poly(Asp-Lys)-b-Asp, was wrapped on its exterior surface through multi-valent interactions to entrap the drugs in interior mesopores. The UCNP emission in the UV-visible range was quenched due to LRET from the UCNPs to the entrapped drugs, while retaining their strong NIR emission. ATP had higher affinity to TDPA-Zn²⁺ and competitively displaced the surface bound polypeptide, which led to the release of the entrapped drugs and subsequent collapse of LRET. The intracellular drug release was accompanied with the enhancement in the UV-visible emission of UCNP, which allowed for real-time monitoring of the drug release via a ratiometric signal with the NIR emission of UCNP as an internal reference.

3.1 Gd-based MRI agents for drug release monitoring

As commercial MRI contrast agents, gadolinium chelates that can produce positive contrast enhancement in T₁-weighted MRI are beneficial for tumor diagnosis. Our group reported intelligent biodegradable DDS-based gadolinium-Dox-loaded nanoscale coordination polymers (Gd NCPs) for targeted drug delivery and programmed drug release [121]. As shown in Figure 6A, the gadolinium ions not only acted as metallic nodes to conjugate the building blocks of 2-amino-1,4-benzenedicarboxylic acids (NH₂-BDC) to form uniform nanoparticles, but also as anchors to chelate Dox to form Gd³⁺-Dox complexes. Meanwhile, the gadolinium atoms in the core of the Gd NCPs were shielded well from water, making no contribution to T₁-weighted MRI contrast. By further coating a hyaluronic acid shell, the designed multi-functional nanocarriers were taken up especially by CD44-positive cancer cells via receptor-mediated endocytosis, and trapped within the lysosomes and endosomes. The acidic environment of lysosomes and endosomes caused the decomposition of the Gd NCPs into small fragments, along with the release of Dox. Therefore, water molecules could access the gadolinium atoms, manifesting as a “turned-on” T₁ signal for drug release monitoring. In vitro experiments showed that the Gd NCPs exhibited enhanced T₁ signal under acidic condition, suggesting their potential for localizing the tumor and monitoring drug release.

Figure 6:

Monitoring drug release by Gd/Mn based MRI contrast agents.

(A) Schematic illustration of an acid-degradable gadolinium-based nanoscale coordination polymer. r₁ relaxivity curves and T₁-weighted MR images of Gd NCPs with various Gd³⁺ concentrations at pH 5.5 and 7.4, respectively. (Adapted from Ref. [121].) (B) Schematic illustration of the synthetic procedure of HMPB-Mn, and schematic representation of HMPB-Mn as a smart multifunctional theranostic for pH-triggered T₁-weighted MRI monitoring of drug release and tumor chemothermal therapy. The correlation between Dox release and Mn²⁺ release of HMPB-Mn at pH 7.4 and pH 5.0; comparison of the calculated Mn²⁺ concentrations by MRI intensity measurement and the release Mn²⁺ concentrations measured by inductively coupled plasma optical emission spectrometry (ICP-OES) at pH 7.4 and pH 5.0. (Adapted from Ref. [122].)

Kaittanis et al. [123] constructed a polymeric nanobeacon by chelating glucose-based polymer dextran with 89-zirconium and gadolinium. The positron emitter 89-zirconium could not merely image the tumor by positron emission tomography, but also guide the surgical resection of sentinel lymph nodes by their inherent Cerenkov luminescence. The MRI contrast agent gadolinium allowed monitoring the drug release with MRI through the quenching of the gadolinium signal by the co-loaded drug. Kim et al. [124] developed short-chain elastin-like polypeptide-incorporating thermosensitive liposomes (STLs) to co-encapsulate gadobenate dimeglumine (Gd-BOPTA) and Dox (Gd-Dox-STL). Gd-Dox-STL was highly stable under physiological conditions (37°C) and capable of rapidly releasing the encapsulated drug with mild hyperthermia (42°C) induced by high-intensity focused ultrasound (HIFU). Compared with lysolipid-based temperature-sensitive liposome loaded with gadolinium-based MRI contrast agent and Dox (Gd-Dox-LTSL), the Gd-Dox-STLs have higher stability and larger changes in T₁ relaxation time between 37°C and 42°C. In vivo MRI experiments demonstrated the co-release of Dox and Gd-BOPTA from STL under HIFU-induced mild hyperthermia, and thus the effectiveness of drug delivery and heating-triggered drug release could be assessed by detecting the MRI signals.

3.2 Mn-based MRI agent for drug release monitoring

The Mn-based agent is one kind of the earliest reported examples that can be used for enhancing T₁-weighted MRI owing to its efficient positive contrast enhancement. Cai et al. [122] constructed a smart Prussian blue-based theranostic nanoplatform with a core-shell hollow structure (namely HMPB-Mn) by coating an Mn-containing Prussian blue analogue (MnPBA) onto the outer surface and the inner channels of hollow mesoporous Prussian blue nanoparticles (HMPBs) (Figure 6B). The HMPB-Mn was highly sensitive to acidic pH and could release Mn²⁺ at the tumor site, making HMPB-Mn a smart T₁-weighted MRI contrast agent with ultrahigh relaxivity for tumor diagnosis. Meanwhile, the hollow mesoporous structure and high surface area endowed HMPB-Mn with high drug (Dox) loading capacity for tumor chemotherapy. More important, both Mn²⁺ ions and Dox molecules were released concurrently from the prepared HMPB-Mn in a pH-triggering manner, and the Dox release rate was real-time monitored through MRI. Combining the high-efficiency chemotherapy and excellent photothermal conversion property, HMPB-Mn also showed a great potential in synergistic chemothermal therapy.

Dong et al. [125] doped Mn²⁺-chelated chlorine e6 [Ce6(Mn), a photosensitizer for photodynamic therapy] into mono-dispersed PEG-modified mesoporous CaCO₃ nanoparticles by co-precipitation. The as-synthesized mesoporous structure of Ce6(Mn)@CaCO₃-PEG nanoparticles could enable efficient loading of chemotherapeutic agent, Dox. With the degradation of the nanoparticles under weak acidic solutions, Ce6(Mn) and Dox were released attendant with the significantly enhanced T₁-contrast under MRI, making it possible to monitor the pH-dependent drug release in real time. Additionally, as guided by MRI and fluorescence imaging, it was found that the intravenously injected CaCO₃@Ce6(Mn)/Dox-PEG could gradually accumulate in the tumor and contribute to a synergistic anti-tumor effect in the combination of photodynamic therapy and chemotherapy.

3.3 Fe-based MRI agent for drug release monitoring

Although iron oxide nanoparticles as T₂ contrast agents have been widely used in MRI, they are rarely applied for drug release monitoring. Kaittanis et al. [126] reported that the clinical formulation of polymer-coated iron oxide nanoparticles (ferumoxytol) could serve as a DDS in cancer chemotherapy (Figure 7). Specifically, the dextran coating retained diverse therapeutic payloads via weak electrostatic interactions. Once perturbed by mild acidification, the particles could rapidly release their cargo, which was utilized for the treatment of solid tumors. In addition to the improved therapy efficacy, the spatiotemporal drug release was monitored via MRI as the cargo release restored the spin-spin (T₂) and spin-lattice (T₁) relaxation signals of ferumoxytol, while leaving the T₂^* signature unchanged. In such a self-reporting multi-functional drug delivery platform, neither the particle nor the drug is subject to any chemical modification, which could lead to a faster administrative approval.

Figure 7:

Monitoring drug release by Fe based MRI contrast agents.

(A) Schematic representation of the proposed model suggesting that the presence of cargo within the coating of iron oxide nanoparticles hinders the diffusion of water molecules, concomitantly affecting the ability of nanoparticles to efficiently diphase the protons of water. (B) Non-invasive monitoring of drug release in vivo with MRI. (Adapted from Ref. [126].)

5.1 Fluorescence- and MRI-based multi-mode drug release monitoring system

Liu et al. [136] reported a novel concept of real-time monitoring of NIR-triggered drug release in vitro and in vivo simultaneously by using upconverted luminescence (UCL) and MRI. Such a monitoring strategy featured the characteristics of high-sensitivity of UCL as well as high resolution and tissue depth independence of MRI. Concretely, multi-functional NaYF₄:Yb/Tm@NaGdF₄ core/hollow mesoporous silica shell nanoparticles (Gd-UCNP@hmSiO₂) were synthesized and applied in the dual-mode imaging nanosensor. Figure 9A illustrates the prototype nanosensor containing three essential components: (i) modified “photomechanical” azobenzene that acted as a “stirrer” on the mesopore surface of Gd-UCNP@hmSiO₂ for NIR-triggered drug release; (ii) LRET donor-acceptor pair of the UCNP and Dox for monitoring NIR-controlled drug release by UCL signals; and (iii) anti-cancer drug residing in the hollow cavity for monitoring NIR-controlled drug release by MR signals due to the diffusion of water molecules from outside to the Gd-UCNP core. In vitro and in vivo experiments demonstrated that the Gd-UCNP cores could serve as robust nanosensors to visualize and quantify drug release in HeLa cells, zebrafish embryos, and Walker-256 tumor-bearing Sprague-Dawley rats. Shen et al. [139] proposed a dual-mode imaging and real-time monitoring strategy with photoluminescence and MRI. Two-photon-sensitized multi-functional nanospheres were developed via stepwise modification of hollow Gd₂O₃:Eu³⁺ nanoparticles with P(NIPAm-co-MAA), 4,4,4-trifluoro-1-(9-pentylcarbazole-3-yl)-1,3-butanedione anion (THA), and amine-functionalized cRGD (designated as TPNPs@cRGD). In the TPNPs@cRGD nanospheres, the photoluminescence signals of Eu³⁺ and the MR signals of Gd³⁺ were stabilized and enhanced, respectively, by the loaded drug via a coordination effect, while displaying excellent linear decreases upon drug release.

Figure 9:

Multi-mode imaging for drug release monitoring.

(A) UCL/T₁-MRI dual-mode nanosensor for drug release monitoring: confocal laser scanning microscopy images of zebrafish, in which UCNP signals are green and Dox signals are red; T₁-MR images of a Walker-256 tumor-bearing Sprague-Dawley rat injected with nanosensor subcutaneously before and after NIR exposure for different time durations. (Adapted from Ref. [136].) (B) Schematic diagram of pH-triggered drug release tracking in acidic conditions by monitoring the SERS and fluorescence signal from the Dox molecules. (Adapted from Ref. [137].) (C) Mechanism of dual-mode monitoring of drug release and synergistic photothermal chemotherapy under a single NIR light. (Adapted from Ref. [138].)

5.2 Fluorescence- and Raman-based multi-mode drug release monitoring system

The combination of fluorescence and Raman imaging has also been used for drug release monitoring. Kang et al. [137] presented a strategy based on plasmonic-tunable Raman/fluorescence imaging spectroscopy to track the release and delivery of an anti-cancer drug (Dox) from Au NP carriers at the single living cell level (Figure 9B). The pH-responsive drug release profile was attained through the conjugation of Dox to the nanoparticle surface via a pH-sensitive hydrazone linkage. When Dox was bound to the surface of Au NPs, its SERS signal was enhanced but its fluorescence was quenched. In turn, once Dox was released from Au NPs in acidic lysosomes, its Raman enhancement was greatly reduced and the Raman spectrum was changed; meanwhile, the fluorescence signal was recovered. The plasmonic-tunable Raman/fluorescence properties enabled the tracking of Dox release from the Au NP surface in living cells, and this system showed great potential in the investigation of the molecular mechanisms of drug delivery and release, as well as the cellular response to drug action.

Cao et al. [140] utilized a Janus particle with opposing mesoporous silica (MS) and gold faces for monitoring intracellular dual-drug (Dox and 6MP) release in real time based on FRET and SERS. The FRET acceptor Dox was attached to the donor 7-hydroxycoumarin-3-carboxylate (CMR)-conjugated MS with a pH-responsive linker hydrazone, and 6MP was conjugated to the gold surface through the gold-thiol interaction. The pH-triggered Dox release and GSH-responsive 6MP release were monitored according to the change of the CMR fluorescence signal and the decrease of 6MP SERS signal, respectively. This dual-drug system provided new perspectives for the real-time study of multi-drug delivery and release.

5.3 Fluorescence- and photothermal-based multi-mode drug release monitoring system

By utilizing drug-coordinated Eu³⁺ as a blocker to inhibit LRET from the Eu³⁺ complex to gold nanotriangles, multiple functions were integrated into one nanocomposite, such as operating under single NIR light, dual-mode monitoring of light-triggered drug release by photothermal imaging and two-photon luminescence imaging, and achieving synergistic photothermal chemotherapy [138] (Figure 9C). In this nanocomposite, Zhang et al. chose a two-photon-sensitized Eu³⁺ complex [Eu(THA)₃(phen)] as the LRET donor, gold nanotriangles as the acceptor, and a model drug (Dox) as the LRET blocker. The NIR light-mediated LRET enabled the dual-mode real-time monitoring of drug release as well as synergistic activation of multiple therapeutic potencies.