Activatable NIR-II photoacoustic probes for cancer imaging and imaging-guided therapy

Abstract

Photoacoustic (PA) imaging in the second near-infrared (NIR-II) window offers a compelling approach for deep-tissue tumor imaging. However, most reported NIR-II PA probes are “always on”, resulting in persistent background signals that limit their imaging specificity and signal-to-background ratios. To address these limitations, activatable NIR-II PA probes that remain “off” under normal physiological conditions have been developed. Upon entering tumor microenvironment, they undergo tumor biomarker-mediated activation and turn “on”. Therefore, this stimulus-responsive strategy significantly reduces background interference and enables highly specific cancer imaging. Additionally, the intrinsic photothermal conversion capability of NIR-II PA probes makes them ideal platform for imaging-guided cancer therapy. This review focuses on the recent development in activatable NIR-II PA probes that respond to tumor-related biomarkers, including pH, reactive oxygen species, nitric oxide, glutathione, and enzymes. Then, we highlight recent advances in NIR-II PA imaging-guided cancer therapy based on activatable probes. Finally, current challenges and future perspectives for developing intelligent NIR-II PA theranostic systems are discussed. This review aims to provide useful insights for the design of activatable NIR-II PA probes for precise cancer imaging and therapy.

Graphical abstract

Activatable NIR-II PA probes circumvent the limitations of conventional “always on” probes, such as low specificity and high background interference. They turn “on” only in the presence of tumor biomarkers, thereby significantly improving imaging accuracy. In addition, their intrinsic photothermal conversion capability allows for imaging-guided therapy. Recent progress in this field is reviewed.

Highlights

• •Summarizes the recent development in activatable NIR-II PA probes for tumor-specific imaging and theranostics.
• •Highlights molecular and nanomaterial design strategies for ROS-, pH-, GSH-, enzyme-, and hypoxia-activated NIR-II PA systems.
• •Discusses activation mechanisms, including molecular transformation, assembly enhancement, and ratiometric photoacoustic imaging.
• •Emphasizes the integration of activatable NIR-II PA imaging with photothermal, photodynamic, and synergistic cancer therapies.
• •Provides perspectives on probe sensitivity, biosafety, quantitative imaging, and clinical translation for future bioactive nanotheranostic platforms.

1. Introduction

Cancer remains a major contributor to global mortality [[1], [2], [3], [4]]. Precise tumor detection and treatment monitoring are essential for improving patient survival [5]. As a powerful non-invasive modality for deep-tissue tumor imaging, photoacoustic (PA) imaging has attracted considerable attention [6,7]. During PA imaging, a nanosecond-pulsed laser irradiates PA contrast agents within biological tissues, leading to a rapid local temperature rise. An ultrasound transducer then captures the acoustic waves generated by thermoelastic expansion. Following algorithmic reconstruction, optical absorption maps of the tissue can be obtained [8]. This technique synergizes the high spatiotemporal resolution of ultrasound imaging with the superior sensibility of optical imaging, enabling PA imaging to achieve deeper tissue penetration (approximately 5 cm in vivo) and improved spatial resolution compared with conventional optical imaging modalities [9]. In contrast to fluorescence imaging, which is limited by strong light scattering, and computed tomography or magnetic resonance imaging, which often suffer from lower molecular sensitivity, PA imaging combines optical contrast with ultrasonic resolution, allowing direct visualization of endogenous and exogenous absorbers in deep tissues. Therefore, PA imaging is well-suited for scenarios that require both deep penetration and high sensitivity, such as angiography, organ imaging, and cancer diagnosis [[10], [11], [12]].

Conventional endogenous PA contrast agents, such as collagens, lipids, and oxyhemoglobin, mainly absorb light in the visible region (400–750 nm) [13,14]. In this region, excitation light suffers significant attenuation from reflection and multiple scattering by tissue components with various refractive indices, such as water, lipids, and hemoglobin [15]. These effects significantly limit laser penetration depth and image clarity [16]. Moreover, the small concentration differences of these endogenous agents between diseased and normal tissues result in low signal-to-background ratio (SBR) [17]. To address these limitations, exogenous PA contrast agents that absorb in the first near-infrared (NIR-I, 700–1000 nm) window have been explored, offering reduced tissue absorption and autofluorescence [[18], [19], [20]]. Although NIR-I agents significantly improved imaging depth, they remain insufficient for visualizing many deep-seated tumors [21]. Recent advances in chemistry and materials science have extended absorption of contrast agents further into the second near-infrared (NIR-II, 1000–1700 nm) window [[22], [23], [24], [25], [26]]. This window offers enhanced tissue penetration, reduced photon scattering, and negligible background interference, thus enabling high-contrast imaging of deep tumors [27].

Beyond their imaging function, NIR PA probes possess intrinsic photothermal conversion capability [28]. Upon being excited by NIR light, these probes undergo electron transfer to reach an excited state, followed by nonradiative relaxation back to the ground state. This process results in a release of heat, enabling PA contrast agents to simultaneously serve as photothermal transducers to ablate tumor tissues [29]. Such dual-functional property has propelled PA technology from a diagnostic tool toward integrated theranostic platforms capable of imaging-guided cancer therapy [30].

To date, many NIR-II PA probes have been designed for cancer imaging and therapy [[31], [32], [33]]. However, most of them are “always on” probes that rely on passive tumor accumulation (Fig. 1a) [[34], [35], [36]]. Such probes exhibit persistent background signals in normal tissues, resulting in compromised SBR and non-specific thermal damage during laser irradiation [37]. In contrast, activatable probes remain “off” under physiological conditions and are selectively turned “on” in response to specific cancer biomarkers (Fig. 1b) [38]. This stimulus-responsive strategy enhances imaging specificity, minimizes off-target toxicity, and enables precise spatiotemporal control of therapeutic outcome [39]. Thus, developing activatable NIR-II PA probes has emerged as a critical step toward precise and intelligent cancer theranostics.

Fig. 1.

Schematic illustration of a) conventional NIR-I PA imaging that suffers from low SBR, strong photon scattering, and limited tissue penetration and b) NIR-II PA imaging that enables high SBR, low photon scattering, and improved tissue penetration. Schematic illustration of activatable NIR-II PA probes for c) cancer imaging and d) imaging-guided therapy.

Although several reviews have discussed NIR-II PA imaging probes and contrast agent design, most of them primarily focus on general NIR-II PA materials, imaging platforms, or theranostic applications [23,[40], [41], [42], [43], [44]]. A systematic overview specifically dedicated to activatable NIR-II PA probes remains lacking. Herein, we provide a comprehensive review of activatable NIR-II PA probes for cancer imaging and imaging-guided therapy. First, we highlight various stimuli-responsive NIR-II PA probes for tumor-specific imaging. Second, we summarize the latest progress in NIR-II PA imaging-guided cancer therapy based on activatable probes. Finally, current challenges and future perspectives for the development of intelligent NIR-II PA theranostic systems are discussed.

2. Activatable NIR-II PA probes for cancer imaging

Activatable NIR-II PA probes are typically designed to integrate a biomarker-responsive moiety with an NIR-II fluorophore [45]. The responsive moiety undergoes a specific chemical or structural transformation upon encountering the biomarker of interest in the tumor microenvironment [46]. This transformation subsequently alters the electronic configuration or aggregation behavior of the probe, resulting in a significant shift or enhancement in optical absorption [47]. As a result, an off-on or ratiometric NIR-II PA signal can be finally detected (Fig. 1c) [48]. Compared with “always on” probes, activatable systems significantly reduce background interference and improve the SBR, enabling precise visualization of the spatiotemporal distribution of probes within tumor area [49]. Activatable NIR-II PA probes have been widely investigated for deep-tissue imaging of tumor-associated biomarkers, including pH, reactive oxygen species (ROS), nitric oxide (NO), glutathione (GSH), and enzymes. Table 1 summarizes the representative activatable NIR-II PA probes for cancer imaging.

Table 1.

Summary of the activatable NIR-II PA probes for cancer imaging.

Stimulus	NIR-II PA Probe	Activation	Chromophore	λabs [nm]	Mechanism	Tumor model	Ref.
pH	OctaNPs	off-on	Octaphyrin	1200	pH-triggered protonation for π-conjugation extension	HepG2 tumor-bearing mice	[50]
	PPE	off-on	Conjugated polymer	NIR-II	pH-triggered protonation for aggregation-enhanced PA	PC-3 tumor-bearing mice	[51]
H2O2	SHT	off-on	TMB	1064	HRP-catalyzed oxidation of TMB to form a charge-transfer complex	4T1 tumor-bearing mice	[52]
	AuNNR@MSi-AuNPs and AuNP@MBN/MPA	ratiometric	ABTS and AuNNR	750, 1250	HRP-catalyzed oxidation of ABTS	4T1 tumor-bearing mice	[53]
	1-FCuSA	ratiometric	EM 1 and FCuSA	808, 1064	FCuSA-catalyzed generation of ·OH to oxidize EM 1	HCT116 tumor-bearing mice	[54]
	SPA	off-on	SPA	1065	HRP-catalyzed intratumoral polymerization	4T1 tumor-bearing mice	[55]
	AuNCs@SiO2	off-on	AuNCs	NIR-II	H2O2-triggered transformation of AuNCs into string-like aggregates	4T1 tumor-bearing mice	[56]
NO	APNO-1080	off-on	Cyanine	1080	NO-induced nitrosation	4T1 and A549 tumor-bearing mice	[57]
	BDPNPs	off-on	Semiconducting polymer	1064	NO-triggered conversion to enhance intramolecular donor–acceptor interaction	4T1 tumor-bearing mice	[58]
	SPANO1-2	ratiometric	BODIPY	950, 1040	NO-triggered nitrosation to disrupt intramolecular charge transfer effect	4T1 tumor-bearing mice	[59]
	DPPBT2NH2	off-on	Semiconducting polymer	1064	NO-triggered oxidation under acidic conditions to enhance intramolecular charge transfer	4T1 tumor-bearing mice	[60]
GSH	MC-PSE	ratiometric	Cyanine	900, 980	GSH-triggered nucleophilic substitution to induce disassembly	4T1 tumor-bearing mice	[61]
	Ox-POM@Cu	off-on	polyoxometalate	1065	GSH-triggered reduction of POM	4T1 tumor-bearing mice	[62]
	MnO2-AuNR-Ppa	off-on	AuNR	1250	GSH-triggered MnO2 reduction for AuNR aggregation	MCF-7 tumor-bearing mice	[63]
H2S	Pd@Cu2O	off-on	Cu	NIR-II	H2S-triggered sulfidation to generate localized surface plasmon resonance	HCT-116 tumor-bearing nude mice	[64]
Caspase-3	AuNNP@DEVD-IR1048	off-on	AuNNPs	1250	Caspase-3-mediated cleavage and GSH reduction-induced AuNNPs aggregation	HepG2 and luciferase-transfected MC38 tumor-bearing mice	[65]
	DOX@Gel-DEVD-AuNR	off-on	AuNR	1250	Caspase-3-mediated cleavage and GSH reduction-induced AuNR aggregation	CT26 tumor-bearing mice	[66]
Furin	AuNR@Peptide	off-on	AuNR	1250	Furin-mediated cleavage and GSH reduction-induced AuNR aggregation	HCT116 tumor-bearing mice	[67]
Transglutaminase	AMCGL NPs	off-on	AuNR	1260	Transglutaminase-mediated formation of isopeptide linkages for AuNR aggregation	4T1 tumor-bearing mice	[68]
Nitroreductase	IR1048-MZ	off-on	Cyanine	980	Nitroreductase -mediated nitro reduction to suppress electron transfer	A549 tumor-bearing mice	[69]

Abbreviation: OctaNPs, octaphyrin-based nanoparticles; PPE, 3,4-ethylenedioxythiophene-alt-3,4-ethylenedioxythiophene copolymer with carboxylate side chains; SHT, HRP- and TMB-loaded mesoporous silica nanoplatform; TMB, 3,3′,5,5′tetramethylbenzidine; HRP, horseradish peroxidase; AuNR/AuNNR, gold nanorod; AuNPs/AuNNPs, gold nanoparticles; MSi, mesoporous silica; MBN, 4-mercaptobenzonitrile; MPA, 4-mercaptobenzoboric acid; ABTS, 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid); FCuSA, copper single-atom nanozyme; EM1, diene electrochromic material; SPA, (N-(3-sulfopropyl) p-aminodiphenylamine; AuNCs, gold nanoclusters; NO, nitric oxide; APNO-1080, nitric oxide-activated NIR-II cyanine probe; BDPNPs, semiconducting polymer-intrachain donor–acceptor nanoprobe; SPA_NO1-2, BODIPY-based NIR-II PA system; BODIPY, boron-dipyrromethene dye; DPPBT2NH₂, diketopyrrolopyrrole-based small molecule derivative; GSH, glutathione; MC-PSE, cyanine-based NIR-II PA probe; Ox-POM, oxidized polyoxometalate–copper hybrid system; Ppa, pyropheophorbide-a; DEVD, tetrapeptide Asp-Glu-Val-Asp; DOX, doxorubicin; AMCGL, AuNR-based Janus nanoprobe; IR1048-MZ, nitroreductase-activated IR-1048 probe.

2.1. pH-activated

pH is a critical physiological parameter that influences both intracellular and extracellular environments [[70], [71], [72]]. Disruption of pH homeostasis may cause various pathological conditions [73,74]. Notably, abnormal pH regulation is a hallmark of cancer and is characterized by a “reversed” pH gradient [75,76]. In normal cells, intracellular pH (pHi) is approximately 7.2, which is slightly lower than the extracellular pH (pHe, ∼7.4). In contrast, cancer cells have an elevated pHi (>7.4) while exhibiting an acidic extracellular microenvironment (pHe ≈ 6.7–7.1) [77]. This dysregulated pH facilitates tumor progression by promoting cell proliferation [78], suppressing apoptosis [79], supporting metabolic reprogramming [80], and enhancing migration and invasion [81]. Therefore, in vivo pH detection represents a valuable strategy for tumor diagnosis and monitoring.

Liu and co-workers developed a porphyrin-based pH-activated NIR-II PA probe named OctaNPs for precise tumor imaging (Fig. 2a) [50]. They first synthesized Octaphyrin with extended π-expansion. Under neutral conditions, Octaphyrin remained in a quinoidal form with 32 π-electrons, which showed weak absorption in the NIR-II region. Upon pH decrease, Octaphyrin underwent protonation, accompanied by a two-electron reduction. This process converted the quinoidal Octaphyrin into a 34 π-electron aromatic species. This electronic structure change significantly extended the π-conjugation and shifted the absorption into the NIR-II region, peaking at 1200 nm. To improve stability and biocompatibility, Octaphyrin molecules were further encapsulated into nanoparticles with PEG to yield Octaphyrin nanoparticles (OctaNPs). In vitro experiments indicated that the PA signal of OctaNPs at 1200 nm increased significantly under acidic conditions, confirming the pH responsiveness of OctaNPs. In vivo NIR-II PA imaging showed that OctaNPs generated a significantly higher PA signal (42-fold) in tumor tissues than that in normal tissues. This study indicated that expanded porphyrins enable NIR-II PA tumor imaging. Overall, this system offers a molecular switching mechanism with large spectral shifts into the NIR-II window and high imaging contrast. However, the activation relies on reversible protonation and redox balance, which may be influenced by physiological pH fluctuations. This may result in potential signal instability.

Fig. 2.

pH-activated NIR-II PA probes. a) Chemical structure of Octaphyrin and schematic illustration of pH-activated NIR-II PA imaging with OctaNPs. Reproduced with permission [50]. Copyright 2021, Royal Society of Chemistry. b) Chemical structure of the conjugated polymer PPE during oxygen doping. c) Schematic illustration of pH-triggered aggregation-enhanced NIR-II PA imaging with PPE. Reproduced with permission [51]. Copyright 2021, American Chemical Society.

In addition, pH-triggered aggregation has been explored to activate NIR-II PA signals. Mei and co-workers developed a pH-activatable aggregation-enhanced NIR-II PA probe based on an oxygen-doped conjugated polymer named PPE. The polymer consists of two functional parts: 3,4-ethylenedioxythiophene-alt-3,4-ethylenedioxythiophene backbone and carboxylate side chains that provide pH-responsive properties (Fig. 2b) [51]. Notably, the polymer could be doped by ambient oxygen without additional oxidants. After oxygen doping, PPE absorption red-shifted into the NIR-II region. At physiological pH, the carboxyl groups were deprotonated and remained in carboxylate anions. These negatively charged groups increased the water solubility of PPE. Upon entering tumor microenvironment, the decrease in pH resulted in the protonation of the carboxylate groups to form carboxylic acids. This change reduced the water solubility of PPE and promoted intermolecular hydrophobic interactions, leading to the polymer aggregation (Fig. 2c). This pH-triggered aggregation significantly enhanced the NIR-II PA signal of PPE. Specifically, the PA intensities at pH 5.5 and 6.5 were 5.0- and 3.1-fold higher than that at pH 7.4. In vivo experiments further confirmed the pH responsiveness of PPE. After injection into PC-3 tumor-bearing mice, the PA intensity in tumors was 3.4-fold higher than that in normal tissues. This study provided a simple and effective strategy for developing pH-activatable NIR-II PA probes with aggregation-enhanced signal intensity. This probe benefits from aggregation-induced signal amplification and structural simplicity, providing enhanced NIR-II PA signal under acidic conditions.

pH-activated NIR-II PA probes are designed based on the reversed pH gradient in tumor microenvironments, where acidic extracellular pH and altered protonation states provide a reliable endogenous trigger. Design of pH-activatable NIR-II PA probes mainly focus on protonation-driven electronic reconfiguration, solubility changes, or aggregation-induced PA signal enhancement. These approaches offer the advantages of simple stimulus activation. However, limitations of pH-activatable NIR-II PA probes remain such as partial reversibility of protonation processes and potential interference from heterogeneous pH distributions in vivo. Additionally, quantitative capability of activatable NIR-II PA probes for pH gradient in tumor microenvironment is still lacking. Future development is expected to focus on irreversible activation mechanisms (e.g., chemical or nanoscale transformation) and ratiometric strategies to enhance quantitative capability.

2.2. ROS-activated

ROS can be classified into radical species, including superoxide anion radical (·O₂⁻) [82] and hydroxyl radical (·OH) [83], as well as nonradical oxidants such as hydrogen peroxide (H₂O₂) [84,85] and singlet oxygen (¹O₂) [86]. These species play essential roles in many physiological processes [87]. For instance, ROS commonly serve as toxic metabolic byproducts that damage cellular components [88]. Additionally, they are important signaling molecules that regulate cell proliferation, differentiation, metabolic adaptation, and immune responses [89]. Specifically, elevated ROS level is a hallmark of cancer [90]. Such abnormal ROS accumulation promotes tumor progression, metastasis, and resistance to conventional therapies [91]. Therefore, the development of precise and sensitive probes for in vivo ROS detection is of great importance. H₂O₂ is one of the most common reactive oxygen species in biological systems [92,93]. In tumors, H₂O₂ levels are higher than in normal tissues [94]. This difference makes H₂O₂ an ideal stimulus for activatable NIR-II PA probes.

In 2019, Xing and co-workers developed an H₂O₂-activated nanoplatform called SHT for tumor NIR-II PA imaging. The probe was based on mesoporous silica nanoparticles loaded with horseradish peroxidase (HRP) and 3,3′,5,5′-tetramethylbenzidine (TMB). In addition, the probe surface was modified with folic acid for tumor targeting. In the presence of H₂O₂, HRP catalyzed the oxidation of TMB to form a charge-transfer complex with strong absorption within NIR-II window, consequently turning the NIR-II PA signal “on” (Fig. 3a) [52]. In vitro tests indicated that the PA intensity at 1064 nm increased linearly with H₂O₂ concentration. In 4T1 tumor-bearing mice, the PA signal at the tumor site increased and reached its peak 4 h after intravenous (i.v.) injection of SHT and remained visible at 24 h. In contrast, the PA signal remained “off” in normal tissues, demonstrating its capability for specific deep-tissue tumor imaging. Similarly, Jiang and co-workers developed a functional conjugated polymer probe named SPA that enables in situ biosynthesis at the tumor site based on H₂O₂/HRP catalysis [55]. These nanoprobe offers high sensitivity toward H₂O₂ and excellent tumor selectivity. However, the reliance on exogenous enzyme stability (HRP) and potential diffusion of catalytic components may reduce its spatial precision of activation.

Fig. 3.

H₂O₂-activated NIR-II PA imaging. a) Chemical structures of TMB and schematic illustration of SHT construction, H₂O₂-triggered activation, tumor-targeted NIR-II phototheranostic. Reproduced with permission [52]. Copyright 2019, American Chemical Society. b) Schematic illustration of ratiometric NIR-II PA imaging with the core-satellite nanoprobe. Reproduced with permission [53]. Copyright 2021, Wiley-VCH. c) Construction of 1-FCuSA and its POD-like activity-triggered conversion to 2-FCuSA. d) Schematic illustration of ratiometric NIR-II PA imaging with 1-FCuSA. Reproduced with permission [54]. Copyright 2025, American Chemical Society.

To quantitatively analyze in vivo H₂O₂, ratiometric strategies have been introduced to minimize background interference [95,96]. Song et al. developed a dual-ratiometric core–satellite nanoprobe for quantitative H₂O₂ detection by integrating surface-enhanced Raman scattering (SERS) and NIR-II PA imaging (Fig. 3b) [53]. The nanoprobe was constructed by assembling gold nanoparticles (AuNPs, served as satellites) onto mesoporous silica-coated nanogapped gold nanorod (AuNNR, served as a core) through amide linkage. The AuNPs were functionalized with boronic acid derivatives while HRP and 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) were loaded into the mesoporous silica shell. The nanoprobe showed broad NIR-I/II absorption due to the plasmonic coupling between AuNPs and AuNNRs. On the other hand, two Raman reporters were incorporated into different regions of the nanostructure. Specifically, 4-mercaptobenzonitrile (MBN, with a peak at 2228 cm⁻¹) was anchored on the AuNPs, while 2-naphthalenethiol (NAT, with a peak at 1418 cm⁻¹) was loaded into AuNNR. Upon exposure to H₂O₂, two processes occurred simultaneously. First, the boronic groups on the AuNP surface were oxidized to phenol groups. This chemical transformation weakened the interaction between AuNPs and the core structure, leading to the dissociation of AuNPs from AuNNRs. As a result, the SERS signal of MBN located on the detached AuNPs decreased significantly. In contrast, the Raman signal of NAT molecules in the AuNNRs remained unchanged. In the meantime, ABTS underwent HRP-triggered oxidation into its active form, showing a significant absorption increase with a peak at 750 nm, while the intrinsic absorption of AuNNRs at 1250 nm remained unchanged. Therefore, ratiometric SERS and PA signals could be obtained to quantitatively analyze in vivo ROS. This combination of stable internal references and stimulus-responsive signal changes effectively minimized background interference and improved detection reliability. In vitro experiments demonstrated that the ratiometric SERS and PA responses showed good sensitivity toward H₂O₂. In vivo studies further indicated that this nanoprobe enabled real-time and quantitative imaging of tumor H₂O₂ in 4T1 tumor-bearing mice. This study demonstrated that integrating ratiometric SERS and PA imaging is a powerful strategy to achieve precise and quantitative detection of ROS in deep tissues. This platform provides a dual modality ratiometric strategy, which significantly improves quantitative reliability and reduces background interference. However, the complex nanostructure and multi-component design may limit reproducibility and increase synthetic difficulty, which could hinder large-scale translation.

Nanozymes are nanomaterials with enzyme-like catalytic properties [97]. Owing to their high stability and adjustable activity, they have been widely explored in recent years [98]. Among them, nanozymes showing peroxidase (POD)-like activity can convert H₂O₂ into highly reactive ·OH, offering a promising strategy for ROS-triggered imaging in tumor microenvironments [99]. To monitor catalytic processes of nanozymes, Wu et al. designed a ratiometric NIR-II PA probe named 1-FCuSA. This nanoprobe consists of three functional components: i) a copper single-atom nanozyme (FCuSA) with POD-like activity to convert overexpressed intracellular H₂O₂ into ·OH; ii) an ·OH-responsive chromophore diene electrochromic molecule (EM 1); and iii) DSPE-PEG for encapsulating FCuSA and EM 1 into nanoparticles (Fig. 3c). In vivo activation mechanism of 1-FCuSA is shown in Fig. 3d. Upon entering tumor microenvironment, FCuSA catalyzed the decomposition of endogenous H₂O₂ to yield ·OH under acidic conditions. The generated ·OH then oxidized EM 1 into its dication form (EM 2), which induced a significant red-shift in absorption to the near-infrared region. As a result, the 808 nm PA signal showed a significant increase. In contrast, FCuSA provided an “always-on” 1064 nm PA signal. Thus, ratiometric PA signal (PA808/PA1064) increased with ·OH production and reflected the catalytic activity of the nanozyme. In addition, the generated ·OH further caused cancer cell death, enabling self-prediction of the therapeutic effects of 1-FCuSA. In vitro experiments showed a linear correlation between the PA signal ratio and H₂O₂ concentration, indicating good sensitivity of 1-FCuSA for H₂O₂ detection. In addition, a strong correlation between the PA ratio and nanozyme catalytic activity was observed, demonstrating that the probe could accurately report the POD-like activity of FCuSA. In vivo experiments indicated that, after i.v. Injection of 1-FCuSA, the 808/1064 nm PA signal in the tumor region of mice increased over time, reaching a maximum at 24 h post-injection. Quantitatively, the ratiometric PA intensity increased by 1.7-fold compared to saline-treated negative control group. Collectively, this work extends ROS-activated NIR-II PA probes from HRP-catalyzed systems to nanozyme-based platforms, providing a useful strategy for real-time imaging of catalytic activity of nanozyme in vivo.

ROS-activatable NIR-II PA probes mainly rely on oxidation-driven chemical transformation. A common design strategy is to introduce oxidation-sensitive chromophores or recognition groups, such as phenylboronic acid, TMB, or other redox-active moieties, which can be selectively transformed by ROS into products with altered electron structures. This transformation induces changes in intramolecular charge transfer or aggregation state, resulting in an “off-on” or ratiometric PA response, thereby enabling sensitive and reliable imaging of redox abnormalities in tumors. Recent progress should focus on enhancing stability of probes and developing ratiometric imaging strategies to reduce background interference and improve quantitative analysis of ROS.

2.3. NO-activated

NO is a crucial signaling molecule in regulating vascular tension, neuronal signal transmission, immunological responses, and tumor progression [100,101]. Its biological function is strongly concentration-dependent in cancer, where NO shows both tumor-promoting and tumor-suppressing effects [102]. Therefore, real-time imaging and quantification of NO are essential for clarifying its complex roles.

To this end, Chan and co-workers developed an activatable NIR-II PA probe named APNO-1080 for NO detection. They synthesized APNO-1080 from the commercial NIR-II cyanine dye IR-1048 by replacing the meso-chloro group with 4-methoxyaniline (Fig. 4a) [57]. Upon reacting with NO, APNO-1080 transformed swiftly to its N-nitrosated product. This reaction altered the electronic structure of the cyanine π-conjugation system. The transformation of electronic structures potentially enhanced the intramolecular charge transfer effect or extends the effective conjugation length, consequently resulting in a red-shift of the absorption peak of APNO-1080 from 874 to 1080 nm (Fig. 4b). Furthermore, the NO imaging capability of APNO-1080 was tested in orthotopic 4T1 breast cancer and A549 liver metastasis models. Results showed that the NIR-II PA signals were 1.3-fold in tumor tissue and 1.65-fold in liver metastases of those of control groups, demonstrating the capabilities of APNO-1080 to detect NO in deep-seated tumors (Fig. 4c and d). This probe characterizes with a simple molecular design, enabling direct deep-tissue imaging of NO activity. However, the relatively modest PA enhancement (1.3-fold) and passive tumor accumulation may limit imaging contrast and quantitative accuracy in tumors.

Fig. 4.

NO-activated NIR-II PA imaging. a) Chemical structures of APNO-1080 and its nitrosation upon reacting with NO. b) Normalized absorbance of APNO-1080 with or without NO incubation. c) Cartoon schematic illustration and PA images of tumor-bearing and control mice after probe administration. d) Quantitative analyses of normalized PA intensities in c. Reproduced with permission [57]. Copyright 2021, American Chemical Society. e) Chemical structures of BDPNP and its mechanism of NO-activated NIR-II PA imaging. Reproduced with permission [58]. Copyright 2026, Elsevier.

Based on the conventional “off–on” probes, further efforts focused on minimizing background signals caused by passive intratumoral accumulation to improve imaging accuracy. To achieve this, Huang and co-workers developed a semiconducting polymer-based NO-activated NIR-II PA probe named BDPNP for monitoring NO generation during M2-to-M1 macrophage reprogramming in immunotherapy (Fig. 4e) [58]. BDPNP consists of two functional parts: i) a hydrophobic semiconducting polymer (BDP) backbone containing o-phenylenediamine units that serve as NO-responsive moieties, and ii) an amphiphilic polymer F-127 to form nanoparticles (BDPNPs). BDPNPs showed an ultralow NIR-II absorption, which was nearly same to that of pure phosphate-buffered saline, due to the weak electron-accepting ability of the o-phenylenediamine unit. Upon exposure to NO, the o-phenylenediamine unit underwent a chemical transformation into a benzotriazole derivative. This reaction significantly increased the electron-withdrawing strength of the acceptor unit and enhanced the intramolecular donor–acceptor interaction. Consequently, the π-conjugation and intramolecular charge transfer were strengthened, leading to a significant increase in NIR-II absorption and activation of the PA signal. In vivo experiments showed that, after treatment with the immunostimulant R848, a 21.4-fold increase in NIR-II PA intensity in tumor tissues was observed at 48 h post-injection, indicating rapid NO generation during macrophage activation. These results demonstrated that BDPNPs provided highly enhanced signal “turn-on” performance and enabled high-fidelity of NO dynamics in tumors during M2-to-M1 macrophage reprogramming in immunotherapy and provided a powerful tool for evaluating immunotherapy response. However, its response depends strongly on local immune heterogeneity, which may affect reproducibility and quantitative interpretation across different tumor models.

To achieve quantitative NO imaging, Liu et al. developed ratiometric NIR-II PA probes named SPA_NO1 and SPA_NO2 (Fig. 5a) [59]. They used an ethylene-bridged BODIPY scaffold (BisBDP1) as the core structure. In SPA_NO1, N,N-dimethylaniline was attached to the α-position of BisBDP1 as an electron donor, and N-methyl-4-vinylaniline was linked to the α′-position which served both as an NO-responsive group and an electron-donating group. SPA_NO2 was further synthesized by introducing a PEG chain and a biotin moiety into SPA_NO1 to improve water solubility and tumor targeting, respectively. SPA_NO1 showed a strong absorption peak at 1010 nm due to the push-pull and conjugation effects. Upon reacting with NO, the N-methylaniline moiety underwent nitrosation. This reaction disrupted the intramolecular charge transfer effect, consequently resulting in a blue shift of the absorption peak to 950 nm. As the NO concentration increased, the PA intensity at 1040 nm decreased while the intensity at 950 nm increased, leading to an increase in the PA ratio (PA950/PA1040). Based on this ratiometric response, SPA_NO2 was further applied in 4T1 tumor-bearing mice to evaluate chemotherapy effects using sodium nitroprusside (SNP) or cisplatin (CDDP) (Fig. 5b). During treatment, the ratiometric PA signal in tumors increased (Fig. 5c). Furthermore, the intratumoral NO levels were quantified. Quantitative results showed that both SNP and CDDP treatment groups maintained intratumoral NO concentrations above 80 nmol/cm³, whereas the control group consistently remained below this threshold during the whole treatment (Fig. 5d). Importantly, tumors with NO levels exceeding 80 nmol/cm³ showed suppressed growth, while those below this concentration continued to expand. These findings demonstrated a critical concentration threshold that determined the “double-edged” role of NO in tumor growth. Overall, this work represents the first report on quantitative NIR-II ratiometric PA imaging of NO in vivo, revealing the important role of NO in tumor growth. However, the multi-step synthetic design and relatively complex molecular architecture may limit translational applicability.

Fig. 5.

NO-activated NIR-II PA imaging with SPA_NO1-2. a) Chemical structures of SPA_NO1-2 and NO detection mechanism of SPA_NO2. b) Treatment timeline and ratiometric NIR-II PA images of 4T1 tumor-bearing mice in different groups. c) Quantification analyses of ratiometric PA intensities in a. d) NO concentrations in different groups during treatment calculated from b. Reproduced with permission [59]. Wiley-VCH.

NO-activated NIR-II PA probes are mainly designed based on the high reactivity of NO in biological systems. In most reported strategies, probes are engineered in weakly absorbing or silent states by controlling electron donor-acceptor balance, disrupting intramolecular charge transfer, or masking conjugated chromophores. Upon reaction with NO, key structural moieties such as aromatic amines or o-phenylenediamine derivatives undergo nitrosation, leading to the formation of stronger electron-withdrawing groups or extended conjugation systems. These chemical transformations result in significant changes in absorption wavelength or intensity, subsequent turning the PA signal “on”. Together, NO-activated NIR-II PA probe design follows two key principles: (i) NO-specific nitrosation or redox reactions to achieve high chemical selectivity and (ii) modulation of intramolecular charge transfer or conjugation length to broaden absorption into the NIR-II region. In addition, ratiometric imaging is a powerful strategy to improve quantitative analysis of NO in complex biological environments.

2.4. GSH-activated

GSH is a predominant intracellular biothiol [103]. It is essential for regulating cellular redox homeostasis and mitigating oxidative damage [104]. In cancer cells, GSH levels are significantly elevated to counteract excessive ROS [105]. This distinctive redox feature makes GSH an ideal biomarker for cancer detection and evaluation.

To detect in vivo GSH, Zhao et al. designed a ratiometric NIR-II probe named MC-PSE. The probe was constructed on an anionic cyanine scaffold. The chlorine atom was replaced by a phenylselenophenols group as both a fluorescence quencher and a GSH-responsive unit (Fig. 6a) [61]. In aqueous solution, MC-PSE self-assembled to form stable J-aggregates through π–π stacking interactions. This aggregation resulted in a strong NIR-II absorption peak at 1000 nm. Upon reaction with GSH, a nucleophilic substitution occurred at the phenylselenide site, yielding a water-soluble product. This chemical transformation triggered the disassembly of J-aggregates into monomers (MC-G) and a decrease in the absorption at 1000 nm. Therefore, this GSH-triggered disassembly generated a clear ratiometric PA response. Specifically, the 980 nm PA signal showed a significant decrease after GSH treatment, whereas the 900 nm PA signal showed no significant change. In addition, the PA900/PA980 ratio showed a linear correlation with GSH concentration within a range of 0–120 μM (Fig. 6b). In vivo studies further confirmed the GSH-responsive performance of MC-PSE. After intratumoral injection into 4T1 tumor-bearing mice, MC-PSE showed a significantly higher PA900/PA980 ratio in tumor tissue compared to normal tissue. Additionally, when tumors were pretreated with a GSH scavenger, the ratiometric signal was significantly suppressed, demonstrating the GSH-specific activation of MC-PSE (Fig. 6c and d). Overall, this work introduced a GSH-activated disassembly strategy for NIR-II ratiometric PA imaging, enabling aggregation-to-disassembly conversion to turn “on” ratiometric PA signal sensitive for GSH detection. Besides, Liang and co-workers developed a metal-based GSH-activatable NIR-II PA probe named Ox-POM@Cu. GSH activated the NIR-II PA signal by reducing Mo⁶⁺ to Mo⁵⁺ in Ox-POM@Cu, forming a mixed-valence Mo⁶⁺/Mo⁵⁺ polyoxometalate that exhibits strong NIR-II absorption.

Fig. 6.

GSH-activated NIR-II PA imaging with MC-PSE. a) Chemical structures of MC-PSE and its activation mechanism for GSH detection. b) Linear relationship between ratiometric PA intensities and GSH concentrations. c) Time-dependent in vivo NIR-II PA images of tumor-bearing mice at 900 and 980 nm with GSH scavenger (NEM+) or without GSH scavenger (NEM–). d) Quantitative analysis of NIR-II PA intensities in c. Reproduced with permission [61]. Copyright 2023, American Chemical Society.

GSH-activated NIR-II photoacoustic probes are mainly based on the high intracellular abundance of glutathione in tumor cells, which provides a strong nucleophilic and redox environment for selective chemical transformation. Probes are typically designed in an initially “off” state. Upon reaction with GSH, key electrophilic groups such as halogenated aromatics or chalcogen-containing moieties undergo nucleophilic substitution or redox conversion, leading to structural disassembly or electronic structure rearrangement. Overall, the design of a GSH-activatable NIR-II PA probe relies on the selective recognition of intracellular thiols via electrophile–nucleophile chemistry to ensure tumor specificity, which subsequently triggers the PA signal “On”.

2.5. Enzyme-activated

Development of cancer commonly leads to the upregulation of specific enzymes, such as Alkaline Phosphatase (ALP) [106], Cathepsin B (CTSB) [107], Matrix Metalloproteinases (MMP) [108], Nitroreductase (NTR) [109], and Aminopeptidase N (APN) [110]. These enzymes are important in tumor development [111]. Moreover, they are highly active and significantly overexpressed in the tumor microenvironment while remaining at low levels in normal tissues [112]. This difference makes enzymes reliable tumor biomarkers. Representative enzyme-activated NIR-II PA probes include the furin-responsive AuNR@Peptide [67] and the transglutaminase-responsive Janus AuNR-MnO₂ [68], both of which utilize biomarker-triggered in situ aggregation to amplify PA signals for image-guided cancer therapy.

2.5.1. Caspase-3-activated

Caspase-3 is a key protease in the apoptosis pathway of cancer cells [113]. After activation, caspase-3 initiates the apoptosis process of cancer cells by cleaving various functional proteins [114]. Therefore, in vivo detection of caspase-3 holds great importance in evaluating treatment effect and prognosis, revealing tumor drug resistance mechanisms, and guiding targeted therapy strategies [[115], [116], [117]]. To evaluate radiotherapy effect, Song et al. designed a caspase-3-activated NIR-II PA probe named AuNNP@DEVD-IR1048 (Fig. 7a) [65]. The probe consists of three functional parts: i) an Asp-Glu-Val-Asp (DEVE) peptide backbone which can be specifically cleaved by caspase-3; ii) an NIR-II fluorophore IR-1048 conjugated to the N-terminal of the peptide backbone; and iii) Cys (StBu)-AuNNP-CBT sequence serves as an activatable NIR-II PA contrast component. In addition, the core-shell structured nanogapped gold nanoparticles (AuNNPs) functioned as radiosensitizers to enhance radiotherapy. Initially, the probe remained in an “off” state due to fluorescence resonance energy transfer effect between the energy donor IR-1048 and the acceptor AuNNP. After intracellular caspase-3-mediated cleavage, IR-1048 was released, accompanied by the NIR-II fluorescence signal turning “on”. Importantly, the overexpressed intracellular GSH reduced the StBu protecting group to yield Cys-AuNNP-CBT to trigger a Cys-CBT click reaction [118]. The product dimers further self-assembled into nanoparticles. This aggregation of AuNNPs strengthened the localized surface plasmon resonance effect and significantly enhanced the NIR-II PA signal, peaking at 1250 nm. The intracellular mechanism of AuNNP@DEVD-IR1048 was shown in Fig. 7b. AuNNP@DEVD-IR1048 accumulated in the cancer cells during circulation following i. v. Injection. Upon X-ray irradiation, the cancer cells initiated apoptotic pathway, leading to the upregulation of caspase-3. Then, AuNNP@DEVD-IR1048 underwent caspase-3-mediated cleavage and GSH-mediated reduction, leading to the turn-on of NIR-II fluorescence and PA signals. These optical signals reflected the apoptotic extent of the cancer cells, which was correlated with the radiotherapy efficacy. Studies further verified the responsiveness of AuNNP@DEVD-IR1048 and demonstrated its capability to monitor radiotherapy-induced apoptosis in tumors. This work first applied activatable NIR-II PA imaging on radiotherapy effect evaluation.

Fig. 7.

Enzyme-activated NIR-II PA probes for tumor imaging. a) Chemical structures of AuNNP@DEVD-IR1048 and its products after caspase-3-cleavage. b) Mechanism of caspase-3-triggered activation of AuNNP@DEVD-IR1048 in vivo. Reproduced with permission [65]. Copyright 2022, Wiley-VCH. c) Chemical structure of DOX@Gel-DEVD-AuNR and schematic illustration of its sequential enzyme-responsive activation. Reproduced with permission [66]. Copyright 2025, Springer Nature. d) Chemical structures of IR1048-MZ and its products after NTR-mediated reduction. e) Schematic illustration of NTR-activated NIR-II PA imaging and photothermal therapy in hypoxic tumors with IR1048-MZ. Reproduced with permission [69]. Copyright 2018, Ivyspring International Publisher.

Based on this study, Song and co-workers expanded this system to imaging-guided cancer therapy. They designed a cascade-responsive nanoplatform named DOX@Gel-DEVD-AuNR, that enables caspase-3-activated NIR-II PA imaging to guide synergistic chemo–radiotherapy (Fig. 7c) [66]. Similar to the abovementioned system, the probe consists of four functional domains: i) a DEVE peptide serves as a substrate for caspase-3-specific cleavage; ii) doxorubicin (DOX)-encapsulated gelatin nanoparticles act as apoptosis inducers which are covalently attached to the Cys through a disulfide bond; iii) gold nanorods (AuNRs) serve as both NIR-II PA contrast agents and radiosensitizers, and iv) CBT for Cys-CBT click reaction. After i.v. Injection, the enhanced permeability and retention effect facilitated the passive targeting and accumulation of DOX@Gel-DEVD-AuNR in tumor tissues. Upon entering cancer cells, the overexpressed intracellular trypsin degraded the gelatin nanoparticles, resulting in DOX release. The released DOX induced chemotherapy-mediated apoptosis of cancer cells, which subsequently upregulated intracellular caspase-3 levels. The activated caspase-3 then cleaved the DEVD peptide substrate, yielding Cys-AuNR-CBT moiety which underwent a GSH-triggered CBT-Cys click reaction to form a cyclic dimer. The dimer further self-assembled into AuNR aggregates. The aggregation of AuNRs induced strong plasmonic coupling between neighboring nanorods, resulting in a significant red-shift of absorption and an enhancement of the NIR-II PA signal with a peak at 1250 nm. Notably, the activated PA signal reflected the level of caspase-3 activity and thus indicated the extent of tumor apoptosis induced by chemotherapy. Based on this mechanism, NIR-II PA imaging was utilized to determine the optimal timing for X-ray irradiation, which corresponded to the maximum PA signal intensity. Once the optimal irradiation time was identified, X-ray exposure was performed, and the aggregated AuNRs functioned as effective radiosensitizers to amplify reactive oxygen species generation and DNA damage in tumor cells. According to the in vivo PA imaging, the optimal X-ray irradiation timing was 21 h post-injection. Further therapeutic evaluation demonstrated that this chemotherapy cascade-enhanced radiotherapy with X-ray irradiating at 21 h post-injection significantly inhibited tumor growth compared with control treatments. These results demonstrated that DOX@Gel-DEVD-AuNR enables activatable NIR-II PA imaging-guided radiotherapy, allowing precise control of treatment timing and improving therapeutic efficacy. Together, these studies demonstrated that caspase-3 is a reliable biomarker for tumor imaging and caspase-3-responsive NIR-II PA probes enable real-time visualization of apoptosis of cancer cells and provide guidance for optimizing tumor treatment.

2.5.2. Nitroreductase-activated

Nitroreductase (NTR) is an important enzyme that is highly expressed in hypoxic tumors and is widely used as a biomarker for tumor hypoxia evaluation [119]. Since hypoxia is a well-established driver of both tumor malignancy and resistance to therapy [120], monitoring NTR activity can provide important information for tumor diagnosis and therapy evaluation [121]. To detect tumor hypoxia, Cai and co-workers developed an NTR-activated NIR-II PA probe named IR1048-MZ (Fig. 7d) [69]. IR1048-MZ was constructed by introducing a nitroimidazole derivative 2-(2-nitroimidazolyl)ethylamine (MZ) to an IR-1048 dye, which served as the NIR-II PA contrast agent. Initially, IR1048-MZ remained in an “off” state due to an electron-transfer process that quenched IR-1048 caused by the electron-withdrawing group MZ. Under hypoxic conditions, the overexpressed NTR reduced the nitro group of the MZ moiety to an amine in the presence of NADH. This enzymatic reaction suppressed the electron-transfer process and restored the optical properties of the IR-1048 dye. Consequently, the NIR-II PA signal was turned “on”. In vivo activation mechanism of IR1048-MZ is shown in Fig. 7e. After i. v. Injection, IR1048-MZ circulated throughout the body, but only the hypoxic tumor microenvironment with elevated NTR level could activate the probe. Consequently, the NIR-II PA signal was selectively turned “on” in tumor tissues. This study demonstrated that NTR-responsive probes can selectively visualize tumor hypoxia and enable deep-tissue PA imaging, providing a useful strategy for the hypoxic tumor microenvironment imaging.

Enzyme-activated NIR-II PA probes rely on the overexpression of tumor-associated enzymes to achieve tumor-specific activation of PA signal. They offer high biological specificity and enable real-time visualization of dynamic tumor events and therapy response. These probes are designed in an initial “off” state through fluorescence quenching, suppressed intramolecular charge transfer, or non-aggregated nanostructures. Upon enzymatic cleavage, key functional motifs such as peptide substrates, protecting groups, or electron-withdrawing units are selectively removed or converted, leading to restored conjugation, enhanced donor–acceptor interactions, or triggered self-assembly. These molecular or supramolecular changes subsequently result in increased NIR-II absorption and enhanced PA signals. In brief, enzyme-activatable NIR-II PA probe design can be attributed to two key principles: (i) high substrate specificity toward tumor-overexpressed enzymes and (ii) efficient conversion from an “off” state to an “on” state after enzyme-mediate cleavage. These strategies collectively enable high-contrast imaging and provide a robust platform for monitoring tumor-associated enzymatic activities in vivo. In addition, recent development trends focus on developing multi-enzyme cascade responsiveness strategy to better mimic tumor biochemical microenvironment. Further efforts are also directed toward quantitative ratiometric imaging to improve accuracy in heterogeneous tumor microenvironments.

3. Imaging-guided cancer therapy based on activatable NIR-II probes

Activatable NIR-II PA probes have provided a powerful tool for tumor-specific imaging. Furthermore, they also hold great potential for precise cancer therapy due to their intrinsic photothermal conversion capabilities [28]. In most of the reported studies, NIR-II PA agents are encapsulated within nanocarriers to improve their biostability and tumor accumulation [[122], [123], [124]]. However, these systems are typically “always on” during circulation, thus generating persistent background signals that limit the SBR of PA imaging. Additionally, non-specific uptake by surrounding healthy tissues may induce unintended thermal damage upon laser irradiation. In contrast, activatable probes remain “off” in blood and healthy organs. They turn “on” and become capable of generating localized heat or reactive species, only when the probe encounters tumor-specific stimuli. This selective activation not only minimizes off-target toxicity but also enables precise spatiotemporal control of therapeutic output [38]. Moreover, real-time PA imaging provides direct visualization of probe activation extent [125]. According to the PA imaging, the optimal irradiation timing when the probe was maximally activated for PTT can be determined to avoid insufficient treatment (Fig. 1d) [126,127]. To date, many activatable NIR-II PA probes have been developed for imaging-guided cancer therapy, with activation strategies evolving from “single-key” to “dual-key” that significantly enhance therapeutic precision and safety.

Early studies in activatable NIR-II PA-guided cancer therapy primarily focused on single-stimulus activation. In 2020, Zhang et al. developed an H₂O₂-activatable NIR-II PA probe named AuNCs@SiO₂ (Fig. 8a) [56]. In detail, self-assembled gold nanochains (AuNCs) were encapsulated in a silica shell to form the nanoprobe. Upon exposure to overexpressed H₂O₂ in the tumor microenvironment, the citrate ligands went a redox reaction, resulting in ligand detachment and rapid fusion of adjacent gold nanoparticles within the silica space. This fusion transformed the initial chain-like nanostructure into string-like aggregates, resulting in a red-shift of the localized surface plasmon resonance into the NIR-II region with a peak at 1180 nm, thereby generating strong NIR-II absorption. After H₂O₂-triggered activation, the string-like aggregates showed a photothermal conversion efficiency of 82.2% under 1064 nm irradiation, enabling efficient PTT. Notably, this H₂O₂-triggered activation significantly enhances biosafety, as the probe remains “off” during circulation, thereby minimizing nonspecific heating in normal tissues and reducing potential side effects. In vivo PA imaging indicated that the PA signal peaked at 24 h post injection, indicating it as the optimal irradiation timing for PTT. Subsequent therapeutic studies demonstrated effective tumor ablation with almost complete inhibition of tumor growth under NIR-II laser irradiation at 24 h post injection, positioning AuNCs@SiO₂ as a representative system for precise and noninvasive NIR-II PA imaging-guided PTT.

Fig. 8.

Single-stimulus-activated NIR-II PA probes for tumor imaging-guided therapy. a) Schematic illustration of the H₂O₂-activatable NIR-II plasmonic system. Reproduced with permission [56]. Copyright 2021, Wiley-VCH. b) Design and mechanism of PEGylated Pd@Cu₂O core-shell nanoprobe for H₂S-activated NIR-II PA imaging-guided ferroptosis. Reproduced with permission [64]. Copyright 2025, Elsevier. c) Schematic illustration of the Janus nanoprobe named MnO₂-AuNR-Ppa for GSH-activated NIR-II PA imaging-guided PDT. Reproduced with permission [63]. Copyright 2022, American Chemical Society.

In recent years, single-stimulus activation strategies have been extended to other therapeutic modalities. For instance, Zhao et al. developed a PEGylated Pd@Cu₂O (PCO) core-shell probe for H₂S-activated NIR-II PA imaging-guided ferroptosis in colorectal tumors (Fig. 8b) [64]. Upon being internalized by cancer cells, the Cu₂O shell underwent H₂S-triggered sulfidation to form non-stoichiometric Pd@Cu_2-xS with copper vacancies. This generated localized surface plasmon resonance for efficient PTT and turned NIR-II PA signal “on”. Additionally, the transformation yielded mixed-valence copper (Cu⁺/Cu²⁺). Cu⁺ catalyzes Fenton-like reactions with intracellular H₂O₂ to generate hydroxyl radicals, and Cu²⁺ consumes intracellular GSH to inhibit glutathione peroxidase 4 (GPX4) expression level and promote lipid peroxidation to consequently induce ferroptosis. In vivo experiments indicated that the experimental group showed complete tumor ablation after a 16-day treatment, demonstrating the excellent therapeutic efficacy of PCO. Song et al. reported a Janus nanoprobe (MnO₂-AuNR-Ppa) for GSH-activated NIR-II PA imaging-guided photodynamic therapy (PDT) (Fig. 8c) [63]. The probe is based on gold nanorods (AuNR), with MnO₂ and photosensitizer pyropheophorbide-a (Ppa) integrated at two ends. After being internalized by cancer cells, the MnO₂ component was reduced and etched by the overexpressed intracellular GSH, leading to the release of Mn²⁺. These ions further coordinated with Ppa for the AuNR aggregation. This red-shifted the probe absorption to the NIR-II region, amplifying NIR-II PA signal and prolonging tumor retention. Moreover, the GSH-triggered reduction of MnO₂ is accompanied by the generation of O₂, which significantly alleviates tumor hypoxia, thus enhancing PDT efficacy. In vivo PA imaging showed that the PA intensities in tumor tissues of mice treated with MnO₂-AuNR-Ppa were 2.5-fold higher than those in the control group 24 h post-injection, indicating 24 h as the optimal irradiation timing for PDT when MnO₂-AuNR-Ppa reached its maximum accumulation in tumors. Further therapeutic evaluation at this time point demonstrated the excellent tumor inhibition capability of MnO₂-AuNR-Ppa, achieving a 42% reduction in tumor volume. Collectively, these findings demonstrate that activatable NIR-II PA probes enable sensitive and selective cancer imaging as well as efficient imaging-guided cancer therapy.

While single-stimulus activation strategies have shown promise in tumor-specific imaging-guided therapy, the intrinsic complexity of biological systems often leads to false-positive outcomes [[128], [129], [130]]. To overcome this limitation, dual-key activation has emerged as a natural evolution [[131], [132], [133]]. This strategy requires two tumor-specific stimuli to activate the probe, thereby achieving higher specificity and minimizing off-target effects [134]. To this end, Fan and co-workers developed a semiconducting polymer nanoprobe (DPPBT2NH₂) responsive to both NO and acidity. The probe integrated diketopyrrolopyrrole (DPP) with a weak electron accepter benzo [c] [1,2,5]thiadiazole-5,6-diamine (BT2NH₂). After being internalized by cancer cells, the BT2NH₂ moiety underwent NO-triggered oxidation under acidic conditions to form a strong electron-withdrawing moiety 5H- [[1], [2], [3]]triazolo [4,5-f],3-benzothiadiazole (BT3N) (Fig. 9a) [60]. This transformation significantly enhanced intramolecular charge transfer, leading to an increase in NIR-II absorption for effective PA imaging and PTT. Additionally, they designed an NIR-I control probe DPPB2NH₂ by replacing BT2NH₂ with B2NH₂. In vitro experiments demonstrated that DPPBT2NH₂ showed strong absorption within NIR-II region upon incubation with NO and acidity. To enhance its biocompatibility, DPPBT2NH₂ was further fabricated with PEG to form nanoprobe DPPBT2NH₂NPs (Fig. 9b). In a 4T1-tumor bearing mouse model, DPPBT2NH₂NPs showed a 2.63-fold higher NIR-II PA signal in tumor tissues than that in normal tissues at 1 h post-injection (Fig. 9c). Upon NIR-II laser irradiation, the tumor temperature increased to 46.9 °C, which was 6.8 °C higher than that of normal tissues (40.1 °C) (Fig. 9d). Further therapeutic evaluations confirmed the effective PTT performance of DPPBT2NH₂NPs, with obvious tumor inhibition in the DPPBT2NH₂NPs + NIR-II group (Fig. 9e). This work proposed the dual-key activation strategy for deep-tissue NIR-II tumor PA imaging for the first time, highlighting its potential for precise tumor theranostics.

Fig. 9.

Dual-stimulus-activated NIR-II PA probe for tumor imaging-guided therapy a) Syntheses and chemical structures of DPPBT2NH₂ and control probe DPPB2NH₂ and their products after NO and acidity activation. b) Schematic illustration of dual-key activation mechanism of DPPBT2NH₂NPs. c) NIR-II PA image and its quantitative analysis of tumor and normal tissues at 1 h post-injection. d) Photothermal image of tumor and normal tissues at 1 h post-injection. e) Tumor growth in mice following different treatments. Reproduced with permission [60]. Copyright 2025, Wiley-VCH.

Activatable NIR-II PA probes enable precise tumor visualization, improved signal-to-background ratio, and spatiotemporally controlled photothermal conversion. These advantages significantly enhance their therapeutic accuracy and reduce off-target damage. Design of activatable NIR-II PA imaging-guided therapeutic probes generally rely on three principles: (i) integration of tumor-specific responsive moieties that ensure selective activation in tumor microenvironments; (ii) introduction of signal enhancement mechanisms such as aggregation, plasmonic coupling, or intramolecular charge transfer to improve imaging contrast; and (iii) combination of PA contrast agents with therapeutic agents to enable synchronized diagnosis and treatment. Despite significant progress in activatable NIR-II PA imaging-guided therapy, several limitations remain. For instance, most current systems rely on metal-based or polymer-based nanomaterials, which may raise potential concerns regarding long-term biosafety. In addition, their complex synthetic routes may introduce challenges in reproducibility, pharmacokinetics, and clinical translation. Future directions are expected to focus on multi-stimuli synergistic activation, quantitative PA imaging-guided therapy, and simplified but stable probe designs that balance functionality with translational feasibility.

4. Summary and prospects

In this review, we summarized activatable NIR-II PA probes that respond to tumor-associated biomarkers such as pH, ROS, NO, GSH, and enzymes. Compared with conventional “always on” probes, activatable probes remain “off” under physiological conditions and are activated only in the presence of target tumor biomarkers. This stimulus-responsive strategy significantly reduces background interference and improves the SBR of PA imaging. Additionally, the NIR-II window provides reduced photon scattering, deeper tissue penetration, and improved imaging contrast compared with the NIR-I region. In addition, we highlighted the application of these probes in imaging-guided cancer therapy. In these strategies, the real-time visualization of probe activation enables precise control of therapeutic timing and provides direct feedback on therapeutic outcomes. These advances demonstrated the great potential of activatable NIR-II PA probes for precise cancer imaging and therapy. Despite these encouraging developments, several challenges remain before these systems can be widely translated into clinical applications.

First, the development of NIR-II PA contrast agents faces challenges in material selection, synthetic complexity, and biological safety. To date, both organic and inorganic materials have been widely investigated [135]. Organic probes, such as cyanine dyes, BODIPY derivatives, and semiconducting polymers, commonly show good biocompatibility and facile chemical modification. Nevertheless, many organic probes suffer from limited photostability, low water solubility, and complex synthetic procedures [136]. On the other hand, inorganic contrast agents, such as gold nanomaterials and transition metal compounds, often exhibit good photostability, excellent photothermal conversion efficiency, and superior PA imaging contrast. However, their long-term biosafety, biodegradability, and clearance from the body remain concerns [137]. Thus, the development of simple, stable, and biocompatible NIR-II probes will be essential for future clinical translation. The design of hybrid systems that combine the advantages of both organic molecules and inorganic nanomaterials may represent a promising approach.

Second, improving the specificity and accuracy of the activatable probe is another challenge. Many tumor biomarkers such as ROS or enzymes may also be highly expressed in inflamed tissues or other pathological conditions. This may lead to false-positive signals and reduced imaging specificity. To address this issue, more sophisticated probes are needed to enhance activation specificity and imaging accuracy. For instance, dual-stimulus or multi-stimulus activation strategies can significantly enhance selectivity by requiring multiple tumor-specific biomarkers for probe activation [138]. Ratiometric PA imaging is another effective strategy that enables quantitative analysis by comparing signals at two wavelengths [139]. In addition, the modification of tumor targeting group on the activatable probes can further improve probe accumulation in tumor tissues, thereby enhancing the imaging accuracy [140].

Third, a unique challenge of activatable NIR-II PA probes arises from the intrinsic characteristics of PA signal generation. Unlike fluorescence imaging, where quenching can approach near-zero background signals, PA imaging relies on a non-radiative process. Consequently, the “off” state of PA probes typically corresponds to weak absorption rather than complete signal elimination. This inherent limitation makes it difficult to achieve extremely high signal-to-background ratios in activatable PA systems. In addition, NIR-II PA imaging generally requires high-energy pulsed laser excitation, which needs the high photostability of contrast agents. Under such conditions, certain organic small-molecule PA probes are easy to photobleaching or structural degradation, leading to reduced signal stability during repeated imaging. Therefore, developing activatable PA probes that simultaneously exhibit simple synthetic accessibility, high photostability, and strong activation contrast remains a significant challenge.

Finally, combining PA imaging with other imaging modalities, such as fluorescence imaging (FLI) [141,142], magnetic resonance imaging (MRI) [143], positron emission tomography (PET) [144], or computed tomography (CT) [145], may further improve detection accuracy. Each imaging modality possesses unique advantages. Although PA imaging provides high spatial resolution and optical absorption contrast, it is still constrained by limited penetration depth and fluence-dependent signal variability. In this context, FLI can provide highly sensitive real-time optical readouts that are particularly useful for validating probe activation at the cellular level, complementing the spatially resolved but relatively depth-limited PA signals. PET imaging offers excellent molecular sensitivity and absolute quantitative capability, which can be used to validate probe biodistribution and pharmacokinetics in vivo, thereby addressing the semi-quantitative nature of PA signals. MRI and CT, on the other hand, provide deep-tissue anatomical reference with high spatial stability, enabling precise localization of PA signals within complex biological structures. By integrating these complementary strengths, multimodal imaging systems enable more comprehensive characterization of tumor location, molecular activity, and tissue microenvironment [[146], [147], [148]]. In addition, integrating activatable PA probes with therapeutic agents enables intelligent theranostic platforms that combine diagnosis, therapy, and treatment monitoring within a single system. However, the integration of multimodal imaging systems brings additional regulatory and technical challenges. Multimodal agents must meet multiple sets of performance criteria, including sensitivity, resolution, and quantitative accuracy across different imaging platforms. This increases the complexity of standardization and reproducibility evaluation. Furthermore, regulatory approval processes (e.g., those required by the FDA) demand rigorous evaluations. These include comprehensive toxicological studies, long-term biodistribution analysis, and large-scale manufacturing consistency. Unfortunately, meeting these demands remains difficult for most nanomaterial-based systems. Therefore, future efforts should focus not only on improving probe performance but also on developing clinically compatible evaluation frameworks and scalable production strategies to facilitate translation. With continued advances in probe design and translational research, activatable NIR-II PA probes hold great promise for precise cancer theranostics in the future.

Ethics approval and consent to participate

This article is a review of existing literature and does not contain any original studies with human participants or animals performed by any of the authors. Therefore, ethical approval and consent are not applicable.

CRediT authorship contribution statement

Runqun Tang: Conceptualization, Investigation, Writing – original draft, Writing – review & editing. Ziyi Zhang: Conceptualization, Investigation, Writing – original draft, Writing – review & editing. Shurong Shen: Funding acquisition, Investigation, Supervision. Gaolin Liang: Conceptualization, Funding acquisition, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

All the data of this study can be acquired from the corresponding author upon reasonable request.