Activation Strategies in Image-Guided Nanotherapeutic Delivery

: Therapeutic nanomaterials serve as an important platform for drug delivery under image guidance. Despite signiﬁcant growth and broad applications, their design speciﬁcs remain a subject of continued interest primarily due to multifunctional factors involved, ranging from nanomaterial properties, imaging modalities, and therapeutic agents to activation strategies. This review article summarizes key ﬁndings on their design characteristics with a particular interest in strategies developed for therapeutic activation (release). First, their activation can be controlled using either an endogenous factor including low pH and glutathione or an external stimulation by light, ultrasound, or electromagnetic ﬁeld. The former is passively controlled from a spatiotemporal aspect compared to the latter, which is otherwise actively controlled through drug linker photolysis, nanomaterial disassembly, or gate opening. Second, light stimulation serves a most notable strategy due to its essential role in controlled drug release, photothermal activation (hyperthermia), and photodynamic production of reactive oxygen species (ROS). Third, some of those activation strategies that rely on ultrasound, photothermal, photoacoustic, magnetic ﬁeld, or X-ray radiation are dually functional due to their role in imaging modalities. In summary, this review article presents recent advances and new insights that pertain to nanotherapeutic delivery systems. It also addresses their technical limitations associated with tissue penetration (light), spatial resolution (ultrasound, hyperthermia), and occurrence of cellular resistance (ROS).


Introduction
Over recent decades, we have seen a rapid growth in and significant contribution by nanoscale therapeutic systems in image-guided delivery [1]. These systems utilize an integrative approach that combines both imaging and delivery functions in a single nanoscale entity. Their design is composed of three complementary modules that include an imaging probe, a therapeutic component, and a nanomaterial platform. Their imaging function plays a fundamental role in detecting and tracking the nanotherapeutic system following its administration ( Figure 1). This helps to identify whether it is well distributed at its targeted site from a pharmacokinetic perspective and thus to determine an optimal timing for its therapeutic activation. Its therapeutic module is achieved with many strategies collectively by drug release [2][3][4][5], photothermal activation [6], photodynamic activation [7,8], or the combination of these. Nanomaterial platforms are more variable in the range of their structure and function as illustrated in dendrimer polymers [9][10][11], upconversion nanocrystals (UCN) [12][13][14], metal-organic frameworks [15], magnetic nanoparticles [16], and hollow nanostructures [17]. This modular approach employed in nanotherapeutic design offers a combinatorial convenience and broad applicability in various therapeutic areas and personalized medicine [1]. In this article, therapeutic strategies are presented with regard to their activation mechanisms. Activation strategies which are covered here involve stimulations by either endogenous factors (pH, glutathione) [18,19] or external stimuli including light [2][3][4][5], ultrasound (US) [20][21][22], alternating magnetic field [4,16], or electric field [23]. These stimuli provide various mechanisms applicable for therapeutic activation including drug release that occurs through linker cleavage, nanomaterial disassembly, or gate opening. These strategies are described along with design factors involved in the selection and integration of imaging probes, payloads, and linkers to nanomaterials. Of particular focus is their individual role in therapeutic activation that occurs via stimulus-controlled linker cleavage, nanomaterial degradation, or induction of porosity.
Briefly, nanoscale systems are engineered for image-guided delivery through modular assembly. They are variable in modular principles, design characteristics, and activation strategies. Identifying a strategy for most optimal therapeutic efficacy remains an objective of significant interest [39]. This article offers new insights learned from recent advances in image-guided systems.

Low pH
Cells provide several stimulating conditions applicable for controlled drug release. Some of these relate to low pH conditions in subcellular endosomes (pH 5.0-6.0) [40] and lysosomes [41,42] where nanomaterials initially reside after intracellular uptake through receptor-mediated endocytosis [43,44]. Similarly, acid-mediated release is reported to occur at extracellular matrices in tumors because they are more acidic than normal tissues [18]. This release strategy controlled by lowered pH has been used frequently in delivery systems. Some of its mechanistic basis is attributed In this article, therapeutic strategies are presented with regard to their activation mechanisms. Activation strategies which are covered here involve stimulations by either endogenous factors (pH, glutathione) [18,19] or external stimuli including light [2][3][4][5], ultrasound (US) [20][21][22], alternating magnetic field [4,16], or electric field [23]. These stimuli provide various mechanisms applicable for therapeutic activation including drug release that occurs through linker cleavage, nanomaterial disassembly, or gate opening. These strategies are described along with design factors involved in the selection and integration of imaging probes, payloads, and linkers to nanomaterials. Of particular focus is their individual role in therapeutic activation that occurs via stimulus-controlled linker cleavage, nanomaterial degradation, or induction of porosity.
Briefly, nanoscale systems are engineered for image-guided delivery through modular assembly. They are variable in modular principles, design characteristics, and activation strategies. Identifying a strategy for most optimal therapeutic efficacy remains an objective of significant interest [39]. This article offers new insights learned from recent advances in image-guided systems.

Low pH
Cells provide several stimulating conditions applicable for controlled drug release. Some of these relate to low pH conditions in subcellular endosomes (pH 5.0-6.0) [40] and lysosomes [41,42] where nanomaterials initially reside after intracellular uptake through receptor-mediated endocytosis [43,44]. Similarly, acid-mediated release is reported to occur at extracellular matrices in tumors because they are more acidic than normal tissues [18]. This release strategy controlled by lowered pH has been used frequently in delivery systems. Some of its mechanistic basis is attributed to acid-catalyzed linker hydrolysis [3] or degradation of acid-susceptible nanomaterials that include metal-organic frameworks [45], nanogels [35], and mesoporous silica oxide (mSiO 2 ) [46]. Figure 2 shows representative delivery systems based on pH-mediated therapeutic release. These include single-walled carbon nanotubes (SWNT) [43] designed for endosomal release of its cisplatin (IV) prodrug [3] (Figure 2A) and bortezomib-loaded polydopamine nanoparticulates [47] in which bortezomib linked through a boronic acid-catechol bonding is demonstrated for faster release at pH 5.4 than 7.4 ( Figure 2B). The schematic in Figure 2D illustrates polymer micelles designed for extracellular activation catalyzed by acidic pH or proteinases [18]. This endogenous control is likewise applied in the release of doxorubicin loaded in acid-labile nanogels [35] or on the surface of iron oxide nanoparticles (IONP; Fe 3 O 4 ) [48]. Delivery systems based on such magnetic nanomaterials as IONP allow MRI-guided drug delivery [35,48]. Another system controlled by low pH is porous mSiO 2 applied in the delivery of camptothecin, an anticancer drug that inhibits topoisomerase I in the nucleus [46]. Here, the drug molecule loaded in the porous shell layer of mSiO 2 is released when the silica layer is degraded in acidic subcellular compartments.
J. Nanotheranostics 2019, 2, x FOR PEER REVIEW  3 of 26 to acid-catalyzed linker hydrolysis [3] or degradation of acid-susceptible nanomaterials that include metal-organic frameworks [45], nanogels [35], and mesoporous silica oxide (mSiO2) [46]. Figure 2 shows representative delivery systems based on pH-mediated therapeutic release. These include single-walled carbon nanotubes (SWNT) [43] designed for endosomal release of its cisplatin (IV) prodrug [3] (Figure 2A) and bortezomib-loaded polydopamine nanoparticulates [47] in which bortezomib linked through a boronic acid-catechol bonding is demonstrated for faster release at pH 5.4 than 7.4 ( Figure 2B). The schematic in Figure 2D illustrates polymer micelles designed for extracellular activation catalyzed by acidic pH or proteinases [18]. This endogenous control is likewise applied in the release of doxorubicin loaded in acid-labile nanogels [35] or on the surface of iron oxide nanoparticles (IONP; Fe3O4) [48]. Delivery systems based on such magnetic nanomaterials as IONP allow MRI-guided drug delivery [35,48]. Another system controlled by low pH is porous mSiO2 applied in the delivery of camptothecin, an anticancer drug that inhibits topoisomerase I in the nucleus [46]. Here, the drug molecule loaded in the porous shell layer of mSiO2 is released when the silica layer is degraded in acidic subcellular compartments. disulfide building blocks. Figure 2A,C each illustrate such release that includes a cisplatin (IV) prodrug loaded on SWNT which undergoes reductive activation to a platinum (II) species [43] and docetaxel release from degradable polymer micelles made of GSH-cleavable poly(ethylene glycol) (PEG) [49]. The latter micelle system is of relevance to image-guided drug delivery due to its ability for induction of fluorescence by a boron dipyrromethene (BODIPY) derivative that serves as a fluorescent probe when it is co-released with docetaxel.
Other pathophysiological conditions play a significant role in therapeutic activation as well. These are related with upregulated redox enzymes [52,53] and tumor hypoxia (low oxygen) [54]. Each of these is exploitable for prodrug activation or release as reported in prodrugs based on SN38 [53], naloxone [52], and Pt [43,55]. Another condition involves matrix metalloproteinase [56][57][58], a hydrolytic enzyme upregulated at tumors which is applicable for catalytic linker cleavage. It plays a role in the tumor-targeted delivery of methotrexate carried by dendrimer [56,57] or hyaluronic acid polymer [58]. In short, though passively controlled, drug release that occurs under endogenous conditions constitutes one of principal strategies explored in delivery systems.
Selecting light wavelengths that are optimal for linker photolysis is dependent on specific photophysical properties defined by linkers. This is illustrated with UV-cleavable ONB [11,81,82], visible light-cleavable coumarin [51,83,84], and near IR (NIR)-responsive cyanine [113,114]. However, linker integration with UCN opens an indirect route for NIR-triggered linker photolysis due to UCN luminescent emission at the UV and visible region upon irradiation at 980 nm [9,14,90]. This concept is illustrated in Figure 3 with UCN-based delivery systems designed for doxorubicin ( Figure 3C) [14] or hydrogen sulfide ( Figure 3D) [111], each linked through ONB to the UCN shell layer. Thus, NIR irradiation at 980 nm results in doxorubicin-induced cytotoxicity in folate receptor (FAR)-positive KB cancer cells as potent as UV irradiation ( Figure 3C), whereas it accounts for more effective H 2 S release in a pork skin model due to its deep tissue irradiation than UV ( Figure 3D). This NIR control is equally demonstrated in a range of payloads from doxorubicin [9,14,90] to macromolecular CRISPR-Cas9 [104]. Further benefits conferred by UCN relate to irradiation at NIR, which shows deeper tissue penetration compared to UV or visible light. Additionally, its luminescent bands contribute to UCN imaging and tracking as well [14,60,104,115].
J. Nanotheranostics 2019, 2, x FOR PEER REVIEW 6 of 26 decrease in relaxivity (mM −1 s −1 ) in a consistent manner with a decrease in its ONB absorbance at 365 nm.

Gating
Gating refers to a process in which a payload is retained in a carrier until it is released upon gate opening. One approach involves a mechanism that induces photolytic alteration at a gate entrance by which its gate is opened ( Figure 5). As shown in Figure 5A, this often occurs in porous materials like mesoporous silica nanoparticle (MSN), which is temporarily blocked at its gate entrance with a photocleavable valve [69,142,143] or polymer brush [68,115,144]. This system is effective for NIR (974 nm)-controlled doxorubicin delivery in HeLa cells as evidenced with the induction of potent cytotoxicity [69]. Another approach involves valve isomerization at the entrance as illustrated in Figure 5B [143]. Its delivery application is consistent with UV-controlled release of doxorubicin

Gating
Gating refers to a process in which a payload is retained in a carrier until it is released upon gate opening. One approach involves a mechanism that induces photolytic alteration at a gate entrance by which its gate is opened ( Figure 5). As shown in Figure 5A, this often occurs in porous materials like mesoporous silica nanoparticle (MSN), which is temporarily blocked at its gate entrance with a photocleavable valve [69,142,143] or polymer brush [68,115,144]. This system is effective for NIR (974 nm)-controlled doxorubicin delivery in HeLa cells as evidenced with the induction of potent cytotoxicity [69]. Another approach involves valve isomerization at the entrance as illustrated in Figure 5B [143]. Its delivery application is consistent with UV-controlled release of doxorubicin which results in the induction of cytotoxicity in macrophages. Molecular valves commonly utilized in this isomerization approach comprise azobenzne [71,143,145,146], coumarate [147], fumaramide [148], spiropyran [149], and donor-acceptor Stenhouse adduct [150]. Their role in gated release is well established by applications in various payloads including doxorubicin [71,143,145,146,150], camptothecin [150], naproxen [147], curcumin [151], DNA [152], and dye molecules including rhodamine B [148] and 4 ,6-diamidino-2-phenylindole [149]. which results in the induction of cytotoxicity in macrophages. Molecular valves commonly utilized in this isomerization approach comprise azobenzne [71,143,145,146], coumarate [147], fumaramide [148], spiropyran [149], and donor-acceptor Stenhouse adduct [150]. Their role in gated release is well established by applications in various payloads including doxorubicin [71,143,145,146,150], camptothecin [150], naproxen [147], curcumin [151], DNA [152], and dye molecules including rhodamine B [148] and 4′,6-diamidino-2-phenylindole [149]. Whereas most gating systems are designed for activation by UV or visible light, they can be functionalized for opening by NIR. This is achieved by integration of valve-attached mSiO2 with nano gold [68] or UCN [153], each serving as an NIR-responsive nanomaterial. Of note is a TiO2/Au@UCN nanocomposite ( Figure 5C), a dual action system designed for NIR-controlled ampicillin release and ROS production by irradiation at 980 nm. This shows potent antibacterial activities against E. coli and methicillin-resistant Staphylococcus aureus (MRSA) in vitro [153]. Their gate opening occurs through valve alteration or isomerization in response to photothermal heating Whereas most gating systems are designed for activation by UV or visible light, they can be functionalized for opening by NIR. This is achieved by integration of valve-attached mSiO 2 with nano gold [68] or UCN [153], each serving as an NIR-responsive nanomaterial. Of note is a TiO 2 /Au@UCN nanocomposite ( Figure 5C), a dual action system designed for NIR-controlled ampicillin release and ROS production by irradiation at 980 nm. This shows potent antibacterial activities against E. coli and methicillin-resistant Staphylococcus aureus (MRSA) in vitro [153]. Their gate opening occurs through valve alteration or isomerization in response to photothermal heating by nano Au or UV-VIS luminescence emitted from NIR-excited UCN. This NIR control has been successfully applied to several payloads including doxorubicin [68,69,143], fluorescein [115,144], and ruthenium-bipyridyl [142].
Briefly, light control constitutes one of principal strategies employed in the activation of nanotherapeutic systems. Its mechanisms for therapeutic release occur via linker photolysis, disassembly, or pore gating.

Reactive Oxygen Species (ROS)
Photodynamic activation involves the production of reactive oxygen species (ROS) which comprise singlet oxygen ( 1 O 2 ) and oxygen-based radical species [154,155]. This occurs through photosensitizer (PS) excitation as applicable to zinc phthalocyanine (ZnPc) [156,157], chlorin e6 (Ce6) [158], rose bengal [12,13], and NIR-responsive indocyanine dye [159]. Alternatively, ROS are produced by other types of PS based on photoactive polymers including polydopamine ( Figure 2B) [47] and photoactive nanomaterials [76,160] including nanoTiO 2 [161]. ROS production by most PS systems occurs by light absorption at visible wavelengths. However, UCN-integrated PS systems [12,13,153,157,162,163] are able to produce ROS by NIR irradiation at 980 nm due to UCN luminescence at UV-VIS bands. This NIR-based PS activation is more beneficial in therapeutic applications in vivo than in vitro due to its deeper penetration [12,163].

Photodynamic Delivery Systems
A typical system designed for photodynamic delivery consists of PS molecules encapsulated or conjugated in nanomaterials like mesoporous nanoparticles (NPs) [156], liposomes [164][165][166], polymers [47,167,168], polymer micelles [65,67,159,169,170], and lipid NPs [171]. Due to high chemical reactivity, ROS produced from these systems are able to make a broad range of non-selective cellular damages from lipid oxidation responsible for membrane ruptures to oxidative degradation in proteins and nucleic acids [76,160]. This accounts for its potent cytotoxicity induced in mammalian cells and bacterial pathogens [154,172], and it serves as a therapeutic basis for photodynamic therapy (PDT) [155,[173][174][175].
In short, photodynamic activation serves as an important route for a dual therapy due to ROS-mediated cytotoxicity and the ability to participate in controlled drug release. PS systems also offer an optical imaging capability due to their fluorescence emission from PS or photoactive nanomaterials themselves, as frequently applied in image-guided delivery [170,174,177]. Their fluorescence detection is coupled with other imaging modes. This is illustrated with a Ce6-based system co-loaded with an Mn 2+ chelate, which is designed for photodynamic therapy under dual image guidance by fluorescence and MRI [170,174,177].

Challenges in Photodynamic Systems
PS systems are faced with certain challenges that are attributable to unfavorable physicochemical properties and cellular occurrence of ROS resistance. Most PS molecules show poor aqueous solubility and high aggregation tendency because of their hydrophobic aromatic rings that engage in intermolecular π-π stacking interaction. As a consequence, these are responsible for lowered efficiency in ROS production, fluorescence quenching, or suboptimal imaging resolution. This issue can be overcome in part by ROS production using either photoactive nanomaterials or polymer particulates that display aggregation induced emission (AIE) [178,179], including integrin-targeting polymer nanodots [179].
Another challenge relates to occurrence of cellular ROS resistance, a known mechanism responsible for reduction in PDT efficacy. This occurs through reductive inactivation of ROS by intracellular antioxidant thiols, primarily, GSH. A few approaches are reported to address this problem including co-delivery of an Mn 2+ contrast agent [177] or MnO 2 , which releases thiol-oxidizing manganese ions [170]. Their oxidative role contributes to a decrease in GSH level via thiol oxidation. An alternative approach involves facilitating the rate of ROS production by increasing the intracellular concentration of O 2 , a precursor molecule for ROS. This involves co-delivery of catalase, an oxidative enzyme that engages in catalyzing endogenous hydrogen peroxide to O 2 [170,180]. The latter approach is of most relevance to applications in tumors where their oxygen level remains relatively lower (hypoxia) than normal tissues. Potent antitumor efficacy is achieved by catalase co-delivery systems made of Ce6-modified chitosan micelles [170] or indocyanine green-loaded albumin complexes [180].
In short, photodynamic activation produces ROS for application in photodynamic therapy [155,158] as well as dual therapy by engagement in controlled drug release [176]. Photodynamic systems are applied for image-guided delivery through their fluorescence alone or in combination with MRI [170,177], photoacoustic imaging (PAI) [180], and/or thermal imaging [170].

Photothermal Activation
Certain classes of nanomaterials are able to produce localized heat through plasmonic photoactivation [181]. These comprise nano Au (AuNP, gold nanorod (AuNR) [182], hollow Au nanosphere (AuNS) [183]), black phosphorus [36], CuS [184], and MoS 2 nanosheet [185]. Their heat production results in the induction of cytotoxic hyperthermia, and this constitutes a therapeutic basis in photothermal ablation therapy (PTT) [186,187]. The heat is further exploited in release mechanisms for dual therapy. Additionally, it serves as an imaging modality applicable in photothermal imaging, which is widely applied in image-guided systems.

Photothermal Drug Release
Despite its cytotoxicity, it is possible that hyperthermia alone is not potent enough for complete tumor ablation, leaving a marginal area of tumor for potential recurrence. This insufficiency is improved by combination with drug release. Delivery systems ideal for photothermal release are designed through nanomaterial integration that consists of a drug-loadable porous nanomaterial and, separately, a photothermally-active nanomaterial. Their integration is illustrated with various combinations of nano Au and mSiO 2 including smaller AuNR-loaded mSiO 2 [188], Au-coated mSiO 2 [74], mSiO 2 -coated AuNR [72], and UCN coated with carbon dot-loaded mSiO 2 [73].

Photothermal Imaging Delivery
Heat production itself serves as an imaging modality in photothermal systems. This is illustrated with AuNR-coated mSiO 2 developed for photothermal image-guided dual therapy via hyperthermia and 5-fluorouracil release [182]. A few other systems have been developed similarly that include UCN (Gd 2 O 3 :Yb,Er)-coated nano Au [201], black phosphorus-loaded hydrogel [36], polydopamine-coated IONP [202], or extracellular vesicles loaded with Au-IONP [203]. Each of these is effectively applied in the photothermal delivery of doxorubicin [36] or anti-miR-21 aptamer [203] under imaging guidance. Yet, unlike purely imaging modalities, such as MRI scans, photothermal imaging has a risk of inducing premature cytotoxicity or drug activation by photothermal heat produced during its scanning process. Therefore, its application is limited on photothermal verification or optimization purposes at target tissues instead of pre-activation imaging guidance, which is achieved by supplementary modalities such as MRI [201], fluorescence [36,182], and PAI [182,202].

Photoacoustic Imaging (PAI) Delivery
The technique of PAI is based on detecting a sound wave (photoacoustic signal) which is generated upon photoexcitation at certain functional systems. It occurs via photothermal expansion (thermoelastic), vaporization (perfluoro molecule), and nanomaterial breakdown [37]. Photoacoustic systems are designed with chromophore molecules or photoactive nanomaterials including polydopamine [202], AuNR [204], CuS [139], and MoS 2 [205]. These are integrated in various ways by nanomaterial coating [202], encapsulation in nanobubbles [204], or loading in a nanogel [139]. These include a paclitaxel-loaded multifunctional nanobubble composed of AuNR and perfluorohexane [204]. Its photoacoustic activation involves perfluorohexane vaporization that occurs in response to temperature elevation induced by photoactivated AuNR. This leads to shockwave production that triggers bubble implosion, serving as a mechanism for drug release and PAI [204]. Systems loaded with CuS [139] and MoS 2 [205] are reported as well for PAI-guided photothermal therapy as illustrated with a FAR-targeted polyethylenimine (PEI)-based nanogel loaded with CuS and Gd chelates ( Figure 4B) [139]. This nanogel system allows dual image (PAI, MRI)-guided PTT in FAR-overexpressing tumors by laser irradiation at 1064 nm. In a similar way, MoS 2 nanosheet modified with IONP and 64 Cu on its shell surface shows a multifunctional capability that allows PAI-guided PTT in combination with two complementary imaging modes, MRI and PET [205].

Computed Tomography (CT) Imaging Delivery
CT-guided photothermal systems rely on nanomaterials with ability for X-ray absorption. These comprise nano Au [206], CuS [184], CuS-Pt [207], Bi 2 Se 3 nanosponge [208], and MoS 2 nanosheet [185], and many of these exhibit strong absorbance at NIR wavelengths. Therefore, CT systems allow additional imaging capabilities including thermal imaging and PAI [208]. Examples of CT-guided delivery include doxorubicin-loaded MoS 2 nanosheet [185] and CuS-encapsulated hyaluronic acid polymer [184]. Liposomes loaded with IONP@Au are effectively applicable for dual CT and magnetic particle imaging as demonstrated in the delivery of tenofovir disoproxil, an antiviral drug, in HIV-infected microglia cells [206]. Using Bimetallic CuS-Pt is dually beneficial for CT guidance as well as Pt chemotherapy in which cytotoxic Pt (II) is released inside the cell following its photothermal activation at 808 nm [207].

Positron Emission Tomography (PET) Imaging Delivery
The basis of PET imaging relies on detecting γ-rays generated in a tissue when tissue electrons react with positrons emitted from radionucleotide tracers carried in a delivery system [209]. PET tracers used in clinical agents include 11 C, 13 N, 15 O, and 18 F, each incorporated as an isotope in a ligand or drug molecule. Post-transition metal elements constitute another class of PET tracers that includes 89 Zr [45], 64 Cu [205,210,211], and Au [212]. PET systems developed for drug delivery are designed in a number of ways. First, PET tracers are incorporated as a part of the carrier as illustrated with 89 Zr incorporated in a metal-organic framework [45], a system developed for doxorubicin delivery under PET guidance. Second, some of these tracers are available as chelated species as seen in 64 Cu chelated in polymer micelles [211] and 64 Cu-NOTA conjugated to mSiO 2 [213]. Third, PET tracers allow direct integration through chemoadsorption or coating as evident in 64 Cu-adsorbed MoS 2 nanosheets [205], Au-coated PdCu nanotripods [212], and 64 Cu-coated AuNR [210]. Overall, CT imaging systems designed in these ways are applied for therapeutic delivery. This includes doxorubicin-loaded mSiO 2 @ 64 Cu-NOTA demonstrated for tumor vasculature-targeted delivery under PET imaging [213]. Nanodots composed of 64 CuS display a similar capability for PET-guided photothermal ablation as demonstrated in tumor-bearing mice [214].

Single Photon-Emission Computerized Tomography (SPECT) Imaging Delivery
SPECT is a dual imaging modality composed of CT and radiation imaging [209]. Like PET, radionuclide elements are required that include 111 In, 123 I, 131 I, 99m Tc, and 188 Re, and they are often incorporated in aromatic compounds through 123 I/ 131 I labeling [139] and in pharmaceuticals through 99m Tc-chelation [209]. Radionucleotide tracers used in SPECT have longer half-lives (6-68 h) on average than PET tracers. This property is highly beneficial during their radiolabeling, nanomaterial incorporation, and uses because it allows more time until full decay.
Several systems have been reported for SPECT-guided therapy based on nano Au [183], dendrimer [139], and liposomes [215]. These include 111 In-labeled hollow AuNS designed for epidermal growth factor receptor (EGFR)-targeted photothermal ablation in head and neck cancer [183]. 131 I-labeled dendrimer is effectively used as a tumor-targeted radionuclide therapy after ligand functionalization with a tumor targeting the LyP-1 peptide [139]. Micelle serves as another platform for SPECT-guided delivery as illustrated with 188 Re-labeled micelles loaded with IR-780 dye applied for PTT in tumors [215].

Mechanism of Ultrasound Activation
Low-intensity ultrasound (US) is commonly used in diagnostic ultrasonography [38], however high-intensity US can serve as an external stimulus strong enough for inducing therapeutic release. This US activation occurs in nanomaterial systems composed of gas-filled cavities or US-responsive molecules including perfluorocarbon [140,216], fluorous tag [20], and polymer [141]. Like photoacoustic activation described above, the physical basis of US activation involves cavitation which occurs upon vaporization of molecules loaded in the cavity. This leads to induction of bubble ruptures or open pores, both contributing to drug release as illustrated in Figure 4C,D [20]. This is evident in A549 cells treated with GFP-loaded fluorous nanoemulsions that show a significant increase in intracellular fluorescence following US application compared to free GFP as a control ( Figure 4C). Similarly, stimulating microcapsule-treated MCF-7 cells by therapeutic US leads to induction of cytotoxicity, which does not occur without microcapsule treatment or by diagnostic weaker US ( Figure 4D).
High-intensity focused ultrasound (HIFU) is also able to induce hyperthermia, which serves as another mechanism applicable for controlled drug release [22]. Similar to photothermal activation, this occurs in thermo-sensitive systems [22] including doxorubicin-loaded liposomes, which show controlled drug release in response to high-intensity US [217,218]. In spite of its potential and recent advances in transducer designs, the efficacy of HIFU activation is limited by its focal length that can reduce spatial resolution, its slow onset of activation which takes hours, and its challenges in certain tissue areas that require crossing a bone or intact skull [219].

Ultrasound Delivery Systems
Delivery systems that are responsive to US activation comprise microbubbles [20,21,220], nanobubbles [216], probubbles [221], nanoemulsions [140], nanodroplets [222], and microcapsules [141]. Microbubbles are popularly used in US applications due to their large shell or cavity space available for loading supplementary imaging probes and therapeutic agents. These include IONP (Fe 3 O 4 )-coated microbubbles developed for dual MRI and US-guided delivery of tPA [21]. Multifunctional nanobubbles that are loaded with 5-fluorouracil, a Gd chelate and IR-780 dye, are comparably used in US-controlled drug release under guidance by multiple imaging modalities [216]. US activation is also applied in the controlled release of gas molecules as illustrated with liposomal probubbles loaded with a H 2 S precursor for application in tumor ablation therapy [221]. Non-bubble US systems are made of nanoemulsions ( Figure 4C) [140], microcapsules ( Figure 4D) [141], and nanodroplets [222], each applied in the delivery of doxorubicin [141], green fluorescent protein and antibodies [140], and paclitaxel [222], respectively. Porous nanomaterials are useful as well for US activation. This is illustrated with mSiO 2 loaded with an NO precursor and superparamagnetic iron oxide nanoparticle (SPION) for MRI-guided, US-controlled NO release in tumor [223].
One of the limitations in US systems relates to bubble sizes. Most microbubble systems are excellent in contrast imaging capability, although they are considerably too large for efficient tissue penetration and cellular uptake. On the other hand, nanometer-sized systems are reversed in the trend with greater tissue infiltration at lower imaging contrast. This conflicting unbalance has been recently addressed using size-variable polymersome (~200 nm), which is able to grow in size under acidic conditions [224]. Thus, once taken up in tumor tissues and subcellular compartments, its signal intensity in US imaging is enhanced as a result of its enlarged size.

Electric and Magnetic Field Activation
Stimulation under an electrical field is potentially applicable in controlled delivery systems. A notable approach involves electroporation, a technique that applies an electrical pulse to induce temporary pores across cell membranes. This technique has played an important role in enhancing cell permeability as often applied to small drug molecules, biomolecules, and genes [225,226]. However, it is rarely used in nanomaterial delivery though it can be beneficial as illustrated with doxorubicin-loaded SPION demonstrated for tumor uptake through electroporation [227]. Unlike normal electroporation, this system offers an imaging guidance through magnetic particle imaging that helps precise positioning of electrodes at its uptake area.
Applying an alternating magnetic field (AMF) on magnetic nanomaterials is more often used in delivery systems. It is able to produce a localized heat applicable for controlled drug release [3]. This magnetic hyperthermia is observed in IONP systems loaded with DNA [228] or doxorubicin linked through a thermally unstable azo bond [229]. SPION is equally applicable for magnetically induced hyperthermia demonstrated in photothermal therapy [230]. Its use also plays a critical role in MPI-guided spatial and temporal control of magnetic hyperthermia, drug delivery, or the combination of these [230][231][232]. This approach helps to attain precise localization and positioning of magnetic nanocarriers within a target region prior to drug release through AMF-triggered hyperthermia. This is demonstrated with SPION-encapsulated poly(lactic-co-glycolic acid) (PLGA) [232] and doxorubicin-loaded liposomes [231]. Furthermore, applying a magnetic field serves as a physical force that enables the control of particle movement. Therefore, drug-loaded magnetic nanomaterials are guided and transported by a magnet for accumulation in a targeted tissue, which has been effectively applied in a few systems. These include a magnetic nanospear designed for gene transfection in glioblastoma cells [233], SPION-loaded microbubbles for doxorubicin delivery in brain tumors [217], and IONP-coated microbubbles for accumulation at specific blood vessels [21]. Overall, delivery systems developed for magnetic control offer several benefits including magnetically inducible hyperthermia, magnetic particle imaging, and magnetic particle guidance.
It is however worth noting current limitations associated with activation strategies. The degree of spatial resolution conferred by each activation strategy is variable as it is defined by the perimeter of its stimulus. Some strategies characterized by US, electromagnetic field, or thermal stimulation show relatively lower precision in spatiotemporal control compared to light activation. On the other hand, light shows a lower level of tissue penetration compared to US or electromagnetic stimulation [80]. This light limitation is currently worked out using an optical technique that allows tissue bypassing in which laser irradiation is delivered through a catheter inserted in a needle injected in tissue [201].
Another critical aspect which is of broad interest involves how to achieve specific nanomaterial uptake and localization in targeted cells only. Many systems discussed here are designed for tumor uptake via enhanced permeation and retention (EPR), a passive targeting strategy that facilitates particle infiltration through leaky vessels in tumors (Figure 1) [234]. However, this passive targeting is not applicable for distinguishing specific tumor biomarkers or targeted binding and uptake at specific biomarker-overexpressing tumor cells. This lack of specificity is achieved otherwise by an active targeting strategy [3,235,236] in which a drug-loaded nanomaterial is functionalized through multivalent conjugation with a target-specific ligand or antibody. This active targeting has been applied to a few tumor biomarkers that include FAR [46,204,216], α v β 3 integrin [179,210,237,238], IGF1 receptor [48], EGFR [183], CD105 [213], CCR5 [212], and nucloelin [45]. It would be equally applicable to other promising but less explored biomarkers that include prostate-specific membrane antigen (PSMA) receptors [239], Her2 [240], riboflavin receptors [241,242], and transferrin receptors [243] for brain delivery. In brief, multivalent ligand conjugation serves as an important strategy in the design of actively targeted systems.
Finally, clinical translation of nanotherapeutic agents is associated with significant challenges due to their multifunctional design, difficulty in synthetic scalability, and paucity of efficient clinical devices needed for their optimal activation. Nevertheless, they show a growing potential as evident with numerous types of nanotherapeutic agents either approved or advanced to clinical studies [244]. Of those, release control by endogenous factors is most actively engaged as shown in PLGA NPs encapsulated with leuprolide (Eligard ® ) [245], albumin NPs bound with paclitaxel (Abraxane ® ) [246], and liposomes encapsulated with doxorubicin (Doxil ® ) [247], mifamurtide (Mepact ® ) [248], vincristine (Marqibo ® ) [249], or irinotecan (Onivyde ® ) [250]. Thermal control of drug release is also successfully applied in heat-sensitive liposomes loaded with doxorubicin (ThermoDox ® ) [251]. Clinical development of photoactivated nanotherapeutics has been relatively slower but was already demonstrated by PDT-based verteporfin liposome (Visudyne ® ), which is approved for age-related macular degeneration and is currently being investigated for locally advanced pancreatic cancer [252,253]. This strategy is also applicable in topical and superficial treatments as illustrated with PDT nanoemulsion (BF-200) [254], a topical agent investigated for treating actinic keratosis [255]. In summary, nanotherapeutic activation strategies are currently being evaluated for their clinical translation [39,244].
Funding: This research received no external funding.