A multifunctional nanoaggregate‐based system for detection of rheumatoid arthritis via Optoacoustic/NIR‐II fluorescent imaging and therapy via inhibiting JAK‐STAT/NF‐κB/NLRP3 pathways

Rheumatoid arthritis (RA) is a debilitating autoimmune disease that causes chronic pain and serious complications, presenting a significant challenge to treat. Promising approaches for treating RA involve signaling pathways modulation and targeted therapy. To this end, a multifunctional nanosystem, TPC‐U@HAT, has been designed for RA therapy, featuring multitargeting, dual‐stimuli response, and on‐demand drug release capabilities. TPC‐U@HAT is composed of a probe/prodrug TPC, a JAK1 kinase inhibitor upadacitinib, and the drug carrier HAT. TPC is composed of an aggregation‐induced emission (AIE)‐active NIR‐II chromophore TPY and an NF‐κB/NLRP3 inhibitor caffeic acid phenethyl ester (CAPE), connected via boronic ester bond which serves as the reactive‐oxygen‐species‐responsive linker. The carrier, HAT, is created by grafting bone‐targeting alendronate and hydrophobic tocopheryl succinate onto hyaluronic acid chains, which can encapsulate TPC and upadacitinib to form TPC‐U@HAT. Upon intravenous injection into mice, TPC‐U@HAT accumulates at inflamed lesions of RA through both active and passive targeting, and the overexpressed hyaluronidase and H2O2 therein cleave the hyaluronic acid polymer chains and boronate bonds, respectively. This generates an AIE‐active chromophore for detection and therapeutic evaluation of RA via both optoacoustic imaging and NIR‐II fluorescent imaging and concomitantly releases CAPE and upadacitinib to exert efficacious therapy by inhibiting NF‐κB/NLRP3 and JAK‐STAT pathways.

Kinase inhibitors are a class of molecularly targeted drugs used in the treatment of RA.They function by precisely targeting and modulating specific immune/inflammation signaling pathways closely correlated to this pathological process. [18]Among these signaling pathways, the JAK-STAT pathway (namely the Janus kinase-signal transducer and activator of transcription) plays a crucial part in controlling immune responses during RA, and JAK-STAT inhibitors have emerged as an effective treatment strategy for inflammatory diseases, including RA. [19] Upadacitinib, a novel JAK1 inhibitor, has received approval from the US Food and Drug Administration (FDA) for RA treatment. [20]It can be used as both monotherapy and in combination with csDMARDs, making it a useful option for patients intolerant to or with a contraindication to MTX. [21,22] Additionally, among the inflammatory signaling pathways, the nuclear factor kappa B (NF-κB) and NLRP3 (namely NOD-, LRR-and pyrin domain-containing protein 3) inflammasome, in response to danger signals such as reactive oxygen species (ROS), initiate the secretion of proinflammatory cytokines.[26] The combination of upadacitinib and CAPE is likely to be highly effective in treating RA due to their excellent therapeutic properties.
However, the efficacy of the therapeutic drug is usually affected by factors such as relatively low bioavailability, short blood half-life, poor water solubility, and low permeability.Fortunately, the fast-developing engineered nanomaterials have the potential to address the challenges in RA treatment and overcome the disadvantages of molecular medicines. [27,28]As multifunctional platforms, nanosized systems can effectively target multiple pathogenic factors in RA microenvironment by modifying their structural composition, surface functionalization, and drug loading.They can also enter the inflamed joint through the extravasation via the leaky vasculature and subsequent inflammatory cell-mediated sequestration (ELVIS), [29][30][31] achieving passive targeting.Compared with passive-targeting systems, active-targeting ones can further improve the accuracy of treatment, leading to better therapeutic effects and reduced side effects in RA therapy.Therefore, the rational design and preparation of nanoparticles with multiple targeting properties is of great significance for the treatment of RA.
Activated macrophages are known to be the primary inflammatory cells in RA and play an essential part in the development of joint inflammation and injury. [32,33]Consequently, targeting activated macrophages has been recognized as an effective approach to alleviate RA symptoms. [34][40] The hyaluronic acid-based nanoparticles modified with ALN are therefore expected to be able to effectively deliver drugs to the bones affected by arthritis.The combination of multiple targeting actions is expected to greatly enhance the efficiency of drug delivery to the site of RA disease.
Once the drug-containing nanoparticles reach the foci of RA disease, the on-demand release of active drugs triggered by in-situ pathological stimuli would be ideal for the precise treatment of rheumatoid arthritis.[43][44][45] However, a single-stimulus trigger may cause nonspecific leakage of the drug.In contrast, dual-stimuli response, triggered by in-situ hydrogen peroxide (H 2 O 2 ) and hyaluronidase (HAase), could provide double insurance for on-demand drug release at the RA lesion site.[51] Hence, using in-situ H 2 O 2 and HAase as dual pathological stimuli is an ideal approach to precisely trigger on-demand drug release at the site of RA lesions.
Visualizing RA foci in situ and tracking the treatment outcome in real time are critical for guiding treatment plans.[54][55][56][57] Furthermore, optoacoustic imaging, which uses ultrasound waves as reporting signals, can provide greater imaging depth.[60][61][62] And combining fluorescence imaging with optoacoustic imaging can provide inter-corroborating information.[65][66][67][68][69] However, imaging probes made from NIR-II chromophores often exhibit aggregation-caused fluorescence quenching, which negatively affects imaging performance, and this is due to their hydrophobic planar structure and the aqueous biological environment.72][73][74] Considering the above, we aim to create a nanosystem that is highly efficient in treating rheumatoid arthritis (RA), while also being capable of monitoring therapeutic progress and detecting lesions in real-time.To achieve F I G U R E 1 Illustration of the nanosystem's preparation and actions.Schematic illustrations for (A) fabrication of the nanosystem TPC-U@HAT and (B) the multitargeting, dual-biomarkers-responsive detection/imaging and therapy for rheumatoid arthritis.
this goal, we develop a multifunctional nanosystem TPC-U@HAT, which possesses multiple features, including multi-targeting capabilities for RA foci, on-demand drug release in response to dual pathological stimuli, in-situ biomarker-activated MSOT and NIR-II fluorescence imaging, as well as RA therapy by inhibiting the JAK-STAT pathway and suppressing NF-κB/NLRP3 pathways (as illustrated in Figure 1).The nanosystem TPC-U@HAT comprises a molecular probe/prodrug TPC, a JAK1 inhibitor upadacitinib, and an amphiphilic hyaluronic acid-based polymer HAT acting as the drug carrier.
The nanosystem TPC-U@HAT could target and accumulate in the inflamed joints of RA via active targeting (ALN targets bone and HA targets CD44 receptor) and passive targeting (ELVIS effect).When triggered by the overexpressed HAase in the inflamed joints, the nanosystem disintegrates upon HAT being degraded by HAase, thereby releasing upadacitinib and TPC in situ.The released upadacitinib wields therapeutic effect for RA by inhibiting the JAK-STAT pathway.Meanwhile, the overproduced H 2 O 2 in the inflamed joint cleaves the boronic ester bond in TPC, which generates the aggregation-induced emission (AIE)active chromophore TPY for both MSOT imaging and NIR-II fluorescent imaging and releases the active drug CAPE for therapy by scavenging ROS and hence suppressing the NF-κB/NLRP3 pathways.The nanosystem TPC-U@HAT was applied to the mouse models of zymosan-induced rheumatoid arthritis and collagen-induced rheumatoid arthritis, and the experimental results manifest the nanosystem's multiple targeting capabilities, dual-stimuli response leading to on-demand release of drugs and high treatment efficacy, as well as in-situ biomarker-activated dual-mode imaging of the disease location and monitoring of the therapeutic efficacy.

Design and synthesis of the probe/prodrug TPC
The chromophore-drug dyad TPC, which is designed to act as both the probe and prodrug, was synthesized following a procedure illustrated in Scheme S1.First, an AIE-active NIR-II chromophore TPY was attained by linking an electrondonating xanthene with N-phenyl-4-pyridinamine to further strengthen its electron-donating capability and ensure the AIE feature, followed by the incorporation of an electronwithdrawing tricyanofuran onto the xanthene moiety.The related intermediates and resultant product TPY were characterized and confirmed by high-resolution mass and 1 H NMR spectrometry (Figure S1-S4).To verify the AIE feature of the chromophore TPY, we measured its fluorescence intensity in a mixture of tetrahydrofuran (THF)/water, where THF acted as the good solvent and water as the poor solvent.By adjusting the ratio of water, the solubility of the solvent mixture was fine-tuned, and the degree of aggregation of the chromophore was correspondingly adjusted.As illustrated in Figure 2A,B, as the water fraction increases, the TPY solution demonstrates a significant increase in the fluorescent emission intensity, displaying a typical AIE effect.When the proportion of water reaches 95%, the fluorescence intensity shows about 10.2-fold enhancement compared with that in the THF solution.
Afterwards, an intermediate TPY-BOH was synthesized by introducing 4-bromomethylphenylboronic acid onto the pyridine ring of the TPY framework via pyridinium salt formation; and the biomarker-activatable probe/prodrug TPC was finally obtained by forming a boronic ester bond between the ortho-diphenol moiety on the ROS scavenger CAPE and the boric acid on TPY-BOH.The structures of TPY-BOH and TPC were verified by 1 H NMR and high-resolution mass spectrometry, as presented in Figure S5-S8.
In the structure of TPC, the presence of the electronwithdrawing groups, namely the positively charged pyridinium group and the boronic ester group, significantly weakens the electron-donating capability of the electron donor, causing TPC to exhibit weak fluorescence (Figure S9).Whereas, upon reaction with H 2 O 2 , the boronic ester is cleaved and the positively charged pyridinium is transformed into the neutral pyridine on the donor side, thereby generating the chromophore TPY which exhibits strong NIR-II fluorescence, illustrated in Figure S9.

2.2
Preparation of the drug carrier (HAT) and the nanoparticles (TPC@HT, TPC@HAT and TPC-U@HAT) Hyaluronic acid is a biocompatible and hydrophilic polysaccharide that can be modified to form amphiphilic polymer by incorporating hydrophobic groups onto its polymer backbone as side chains, and this modification facilitates the formation of stable nanoparticles. [36,37]In this study, tocopheryl succinate (TOS) was esterified onto the hyaluronic acid backbone as hydrophobic side chains, resulting in the amphiphilic polymer HT.Next, to obtain the amphiphilic bone-targeting polymer HAT, the bone-targeting ALN sodium was introduced onto the hyaluronic acid backbone through amidation reaction, followed by the grafting of TOS onto the hyaluronic acid backbone via esterification reaction.The synthesis routes for the polymers HT and HAT are presented in Schemes S2 and S3, respectively.The structures of HAT and HT were characterized using 1 H NMR spectrometry (Figure S10-S12).The graft ratios of ALN and TOS in HAT were determined as 18 and 11.5 mol%, respectively, while the graft ratio of TOS in HT was found as 14.25 mol%, as determined by 1 H NMR spectrometry.The appropriate graft ratio ensures that the properties of hyaluronic acid are not substantially changed, while the added hydrophobicity facilitates the formation of stable nanoparticles.Additionally, the FT-IR spectra of the HA (hyaluronic acid), ALN, HA-ALN (hyaluronic acid grafted with ALN), TOS, HT (hyaluronic acid grafted with TOS), and HAT (hyaluronic acid grafted with both ALN and TOS) were measured (Figure S13).The FT-IR spectrum of HAT shows a sharp peak at 1633 cm −1 corresponding to the N-H stretching for amide linkage, whereas the characteristic peaks at the wave number 2866 cm −1 and that at 2925 cm −1 correspond to -CH 2 -stretching vibration peaks of TOS.These results indicate that ALN and TOS have been successfully grafted onto hyaluronic acid as side chains.
Owing to its amphiphilic properties, HT and HAT readily form nanoparticles in aqueous media.To prepare TPC@HAT and TPC@HT nanoparticles, the positively-charged TPC was mixed with the negatively-charged HAT or HT in aqueous medium, this allowed for TPC encapsulation by HT or HAT through electrostatic and hydrophobic interactions.Furthermore, the chromophore-drug dyad TPC, together with the JAK1 kinase inhibitor upadacitinib, were mixed with HAT in aqueous medium to form the nanosystem TPC-U@HAT.As shown in Figure S14, the HAT nanoparticles have an average diameter of approximately 151 nm and display spherical morphology, whereas TPC@HT and TPC@HAT show an average diameter of around 123 and 135 nm, respectively, and spherical structures.As depicted in the inset of Figure 2C, the image obtained from the transmission electron microscopy (TEM) measurement shows that the nanosystem TPC-U@HAT has a comparatively uniform and spherical morphology and the nanoparticles' diameter is approximately 144 nm.The nanosystem TPC-U@HAT's particle size distribution is shown in Figure 2C, which was obtained from the measurement by use of the dynamic light scattering method (DLS).Furthermore, there were no obvious changes in particle size for the nanoaggregate TPC-U@HAT in PBS with varying ion concentrations or in buffer solution with different pH values upon being stored for 48 h, indicating that the nanoaggregate TPC-U@HAT exhibits good stability (Figure S14D-E).The loading capacity of TPC and upadacitinib in TPC-U@HAT nanoparticles was determined as 26.36 and 11.63 wt% respectively via absorption spectrometry.Consequently, the weight percentage of CAPE in TPC-U@HAT was calculated as 7.31 wt%.

MSOT and NIR-II fluorescent responses of TPC-U@HAT toward HAase and H 2 O 2
The nanosystem TPC-U@HAT is capable of responding to dual pathological biomarkers, hyaluronidase (HAase) and hydrogen peroxide (H 2 O 2 ), to generate the reporting signals for both NIR-II fluorescent imaging and MSOT imaging.First, the optical properties of the nanosystem TPC-U@HAT were examined upon incubation with HAase, H 2 O 2 , or both, at 37 • C in PBS (pH 7.4) for varying time periods (as illustrated in Figure 2D-E and Figure S15).The fluorescence intensity remains almost unchanged after incubation with HAase only (Figure S15A), whereas a moderate increase in fluorescence intensity (10.6-fold) occurs in the presence of H 2 O 2 only (Figure S15B).Interestingly, simultaneous incubation with dual stimuli (HAase and H 2 O 2 ) results in a dramatic enhancement (25.4-fold) of the fluorescent intensity at 920 nm (Figure 2D and Figure S15E).In the absorption spectra, no obvious change can be observed in TPC-U@HAT after incubation with HAase only (Figure S15C).Incubation with only H 2 O 2 mildly increases the absorption at 750 and 685 nm (Figure S15D), while an obvious enhancement of the absorption at 750 and 685 nm can be found in the nanosystem after incubation with the two biomarkers (HAase and H 2 O 2 ) (Figure 2E and Figure S15F).The absorption at 808 nm as shown in Figure 2E also increases with the reaction time between the probe and the stimuli, hence the activated probe can be excited by 808 nm laser and be applicable for NIR-II imaging.The optoacoustic intensities for the nanosystem TPC-U@HAT upon incubation with HAase (2 U⋅mL −1 ) and H 2 O 2 (100 μM) for different time periods are shown in Figure 2F.The optoacoustic signals demonstrate a significant increase with increasing incubation time.Furthermore, the fluorescence and absorption spectra's concentration dependency for the nanosystem TPC-U@HAT upon incubation with HAase (2 U⋅mL −1 ) and H 2 O 2 of varied concentrations were recorded in PBS at 37 • C, as displayed in Figure S16.With increasing H 2 O 2 levels, the fluorescence intensity at 920 nm increases as well (Figure S16A,B); while the nanosystem TPC-U@HAT displays rather weak fluorescent emission without the presence of HAase and H 2 O 2 .The absorption spectra for the nanosystem TPC-U@HAT before and after HAase and H 2 O 2 incubation are shown in Figure S16C,D; before responding to HAase and H 2 O 2 , the nanosystem TPC-U@HAT exhibits maximum absorption band at around 635 nm; upon responding to HAase and H 2 O 2 , the red-shifted maximum absorption bands are at approximately 685 and 750 nm.This is because HAase can only disintegrate hyaluronic acid polymer, leading to the disintegration of the nanoparticles; and only H 2 O 2 can trigger the probe/prodrug TPC to release the AIE-active chromophore TPY, which then brings on pronounced MSOT signals and NIR-II fluorescent signals.Moreover, the selectivity of the nanosystem TPC-U@HAT towards HAase, H 2 O 2 , HAase plus H 2 O 2 , and other potential interfering substances, including ions and biomolecules, were evaluated.The results demonstrate that no remarkable increase in fluorescence intensity or relative optoacoustic intensity of the nanosystem TPC-U@HAT could be observed upon incubation with ions, thiols, or the biologically relevant enzymes (Figure S17).These results indicate that the nanosystem TPC-U@HAT can operate as a dual-stimuli responsive system, displaying obvious optoacoustic and NIR-II fluorescence signals as the reporting signals.
It is suggested that the boronic ester group in TPC (the probe/prodrug) can be cleaved upon reacting with H 2 O 2 , resulting in the conversion of TPC to TPY (the activated form of TPC) and the simultaneous release of the active drug CAPE.In order to validate this response mechanism, first, high-performance liquid chromatography (HPLC) was used to assess TPC after incubation with H 2 O 2 .As depicted in Figure S18A, the characteristic peak of TPC at 5.35 min decreases, while the two newly-emerging peaks at 1.77 and 2.11 min are perspicuously visible, corresponding to CAPE and TPY, respectively.In addition, the spectra of absorption and emission for TPY-U@HAT and TPC-U@HAT before and after reaction with HAase plus H 2 O 2 were compared.It can be seen from Figure S18B,C that the absorption and emission spectra of the reaction product are similar to those for the TPY-U@HAT.Furthermore, mass spectrometry was performed on TPC after its incomplete reaction with H 2 O 2 .As shown in Figure 2G and Figure S18D, the m/z peak at 562.2253 [M+H] + and the one at 307.0946 [M+Na] + represent the chromophore TPY and the active drug CAPE, respectively.The weak m/z peak at 944.3624 [M+H] + corresponds to TPC.These results together signify that the probe/prodrug TPC is converted into TPY (namely the chromophore and the activated form of the probe) and the active drug CAPE upon reaction with H 2 O 2 .
Moreover, the in vitro cumulative release rates of CAPE or upadacitinib from TPC-U@HAT nanoparticles triggered by HAase and H 2 O 2 were measured at different time points in PBS solution (pH 7.4).As presented in Figure S19, TPC-U@HAT nanoparticles show increased release rates with increasing HAase and H 2 O 2 levels.When stimulated with HAase (2 U⋅mL −1 ) and H 2 O 2 (200 μM), the drug CAPE is released at a cumulative rate of approximately 68%, while the drug upadacitinib is released at a cumulative rate of around 82%.These results demonstrate that TPC-U@HAT nanoparticles can efficiently release the active drugs CAPE and upadacitinib when triggered by the biomarkers HAase and H 2 O 2 , which could provide on-demand release of the drugs in situ that accordingly produce therapeutic effects.

TPC-U@HAT's in vitro targeting and ROS inhibition capabilities in RAW 264.7 cells
Macrophages play crucial roles in the pathogenesis/etiology of RA by producing various proinflammatory cytokines and chemokines and promoting cartilage and bone destruction. [32,33]Therefore, in this study we used the RAW264.7 cells to investigate the effects of the nanosystem TPC-U@HAT at the cellular level and used L929 cells for comparison.We first evaluated the cytotoxicity of the nanosystem TPC-U@HAT towards both cell lines using MTT assays.As depicted in Figure S20, it is clear that the nanosystem TPC-U@HAT has low cytotoxicity to both cell lines even after being incubated with 200 mg⋅L −1 of the nanosystem, indicating that it is suitable for cell imaging.Next, we employed lipopolysaccharide (LPS) stimulation to induce reactive oxygen species (known as ROS) in the RAW 264.7 cells.The commonly used fluorescent probe for ROS 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) was used as an indicator of ROS, as DCFH-DA responds to ROS and generates bright green fluorescence in live cells. [75]After treating the cells with LPS for different times (3 or 6 h), we incubated them with DCFH-DA and imaged them using a fluorescence microscope.The results are presented in Figure 3A.Brighter fluorescence can be seen in RAW 264.7 cells after 6 h stimulation by LPS compared with the shorter stimulation time (3 h); while no obvious fluorescence is generated in LPS-unstimulated cells.These results demonstrate that RAW 264.7 cells produce ROS when stimulated by LPS, and the longer the induction time, the higher the ROS production, which is consistent with previous literature. [76]Therefore, the above results confirm the utility of the nanosystem TPC-U@HAT in cell imaging and its suitability for further research on ROS-induced inflammatory responses in RA.
Afterward, the in vitro targeting capability of the nanosystem was assessed by incubating RAW264.7 (LPS +/−) cells with different formulations (TPC, TPC@HT, and TPC@HAT).As shown in Figure 3B-D, it is evident that the optoacoustic intensities and NIR-II fluorescence intensities increase when LPS-stimulated RAW264.7 cells are incubated with TPC@HT or TPC@HAT, while no significant signals can be observed in RAW 264.7 cells that have not been stimulated by LPS.[79] To investigate the ROS inhibition capability of TPC-U@HAT nanoparticles in RAW 264.7 cells, different formulations (saline, CAPE, upadacitinib, TPC@HAT, TPC-U@HAT) were incubated with RAW 264.7 cells (LPS +/-), which were then stained by DCFH-DA, as depicted in Figure 3E.Only weak fluorescence could be seen in the CAPE, TPC@HAT, or TPC-U@HAT treatment groups, indicating that these formulations can scavenge ROS effectively.These findings demonstrate that the nanosystem TPC-U@HAT has potential as a therapeutic agent, which can target CD44-receptor-rich disease sites and scavenge ROS.

TPC@HAT's in vivo targeting capability and imaging in zymosan-induced rheumatoid arthritis mouse model
To assess the in vivo targeting capability of TPC@HAT to the RA disease site, we established the RA mouse model on the left ankle of the mice by intra-articular injection of zymosan (25 mg⋅kg −1 , referred to as ZIA group).In mice, zymosan with intra-articular administration leads to the appearance of inflammatory parameters similar to RA, making it a suitable animal model for studying this disease. [80,81]The control group was given an equal volume of saline via intra-articular injection.After 24 h, we intravenously injected the ZIA group of mice with TPC, TPC@HT, and TPC@HAT, respectively; while an intravenous injection of TPC@HAT nanosystem was given to the mice of the control group.And then the NIR-II fluorescence imaging was conducted (Figure 4A).At the left ankle joint area, the ZIA model groups that received TPC@HT or TPC@HAT by i.v.injection exhibit apparently much more evident fluorescence as compared with the control group, and this is likely because of the release of the chromophore TPY in response to the overexpressed H 2 O 2 in the joint inflammation site.In particular, the ZIA group that received the TPC@HAT nanosystem by i.v.injection demonstrates much more intense fluorescence in the left ankle joint site than the ZIA group i.v.injected with TPC@HT, and there is no fluorescence in the ZIA group that received TPC (i.v.injection) or the control group that received TPC@HAT (i.v.injection).This is likely because the TPC@HT nanoparticle can accumulate at the RA joint site due to the fact that hyaluronic acid on the particle surface can target CD44 receptor and the nanoparticle has passive targeting ability (ELVIS effect), whereas the nanosystem TPC@HAT can accumulate at the RA joint site at a much higher level due to both its passive targeting (ELVIS effect) and active targeting (HA targets CD44 receptor and ALN targets the bone).When the nanoparticles reach and accumulate at the RA site, the overexpressed HAase therein disintegrates the nanoparticles, and the probe/prodrug TPC is released.TPC's boronic ester bond is then cleaved by the overexpressed H 2 O 2 at the RA site and, accordingly, the NIR-II chromophore TPY is generated, producing NIR-II fluorescence signals, whereas free TPC cannot accumulate in the RA joint site due to a lack of targeting capabilities.These results confirm that a higher accumulation of the TPC@HAT nanosystem in the RA joint site can be achieved upon intravenous injection due to the nanosystem's passive targeting (ELVIS effect) and active targeting (HA specifically binds to the overexpressed CD44 receptor, and ALN binds calcium in the bone).Furthermore, the average NIR-II fluorescence intensities corresponding to the left ankle joint area (region of interest (ROI)) of mice in the four groups are exhibited in Figure 4B.At about 4 h post i.v.injection of TPC@HT or TPC@HAT, the fluorescence intensity reaches the maximum, after that it gradually grows faint and disappears because of the metabolism.Moreover, the ex-vivo images of hind legs and major organs at 4 h after different formulations i.v.injection confirm the highest accumulation of TPC@HAT in the RA joint (Figure 4C and Figure S21 and S22, Supporting Information).These results verify that the ALN decoration and HA on the nanoparticle's surface significantly improve the in vivo targeting and accumulation of the nanosystem TPC@HAT in the RA joint sites.
Based on the fact that the TPC@HAT nanosystem can effectively target and accumulate at the RA joint site and subsequently respond to the biomarkers (HAase and H 2 O 2 ) at the RA joint site, we utilized the nanosystem TPC@HAT to image the RA site in the ZIA mouse model.It is well known that ROS (such as H 2 O 2 ) and hyaluronidase are overexpressed in joint region of RA, and thus the H 2 O 2 or hyaluronidase can act as endogenous biomarkers for RA [46][47][48][49][50][51] The experimental protocol is illustrated in Figure 5A.To induce rheumatoid arthritis in mice, we administered zymosan intraarticularly at varying doses for 24 hours.The nanosystem TPC@HAT was then intravenously administered into the mice of different groups, and subsequently, both NIR-II fluorescent imaging and MSOT imaging were conducted for them.Figure 5B shows the NIR-II fluorescent images of the control (treated with isovolumetric saline) as well as the groups that were treated with varied doses of zymosan at 0 and 4 h postinjection of TPC@HAT.No fluorescence is observed in any of the mice before injection of the nanosystem TPC@HAT.However, in the ZIA groups, with increasing doses of zymosan, the fluorescent intensities enhance at 4 h upon the nanosystem TPC@HAT's injection.The quantified fluorescent intensities covering the left ankle joint area (ROI) are presented in Figure 5C.
Afterward, 3D MSOT imaging by use of the nanosystem TPC@HAT (equivalent dosage of 2.5 mg⋅kg −1 TPC) was carried out, and the data are shown in Figure 5D.The location of inflammation in the left ankle joint has been definitely revealed by the 3D MSOT images.Obviously, the ZIA mice treated with a bigger dose of zymosan (25 mg⋅kg −1 ) exhibit much more evident optoacoustic signals compared with those treated with a smaller dose of zymosan (10 mg⋅kg −1 ). Figure 5E displays photographs of the dissected left legs from the control and the ZIA mice.In contrast to the control group, the ZIA mice display obvious redness and swelling at the left leg, with the group of mice treated with a bigger dose of Zymosan exhibiting much more severe swelling.Moreover, histological analyses by hematoxylin and eosin (H&E) staining was adopted for analyzing the left ankle joint of the control and the zymosan-induced rheumatoid arthritis model mice, and the results are shown in Figure 5F, it is explicitly clear that as for the mice of the model groups, apparent cartilage destruction and immune cell infiltration can be noticed, as compared with the control group; while the ZIA group of mice treated with a bigger zymosan dose exhibit the conditions with a much more severe degree.These results verify that the nanosystem TPC@HAT is able to detect the inflamed joint site by in-situ responding to the biomarkers hyaluronidase and H 2 O 2 , thereby enabling noninvasive detection of rheumatoid arthritis by means of both the NIR-II fluorescent imaging and MSOT imaging, Furthermore, the 3D MSOT images are capable of identifying and locating the disease site.

TPC-U@HAT's imaging, RA therapy, and outcome monitoring in collagen-induced rheumatoid arthritis mouse model
The nanosystem TPC-U@HAT was employed to treat RA and evaluate its efficacy in monitoring the therapeutic outcome.For this purpose, collagen-induced arthritis (CIA) was used, since CIA is the most commonly used model of rheumatoid arthritis which resembles rheumatoid arthritis in humans.[84] The groups of mice were labeled as G1-G6, to which different treatment formulations were given.G1 represents the healthy mice, to which no treatment was given.G2 represents the CIA model group, to which therapy with saline was given.G3 represents the CIA model group, to which therapy with free CAPE was given.G4 represents the CIA model group, to which therapy with free upadacitinib was given.G5 represents the CIA model group, to which therapy with TPC@HAT was given.G6 represents the CIA model group, to which therapy with the nanosystem TPC-U@HAT was given.The experimental process, which included RA therapy, imaging, and outcome tracking, is presented in Figure 6A.As for the mice of the CIA model, four rounds of treatment were performed (for treatment with TPC-U@HAT: intravenous injection of TPC-U@HAT 13.68 mg⋅kg −1 at day 34, 38, 42, 46, equivalent to 1 mg⋅kg −1 of CAPE and 1.6 mg⋅kg −1 of upadacitinib), and at 4 h after the nanosystem TPC-U@HAT was i.v.injected (equivalent to the dosage of 2.5 mg⋅kg −1 TPC), both the NIR-II fluorescent imaging and 3D MSOT imaging were conducted.The obtained NIR-II fluorescent images and the average NIR-II fluorescent intensities in the mice's ankle regions are depicted in Figure 6B,C.The comparison of NIR-II fluorescent images in the mice's ankle regions before and after treatment (day 34 and day 60) are given in Figure S23 and S24.A mouse's photograph indicating the region for scanning during MSOT imaging experiment is displayed in Figure 6D, and the obtained 3D MSOT images are displayed in Figure 6E.From both the 3D MSOT images and NIR-II fluorescent images, it is explicit that for the CIA model mice receiving treatments with TPC-U@HAT (namely G6) for four rounds of therapeutic injection, the ankle inflammation has mostly been ameliorated, and both the MSOT signals and fluorescent signals in the ankle regions are much weaker in comparison with other treatment groups.
To further evaluate the therapeutic efficacity of different formulations in CIA mouse model, the groups of mice treated with different formulations (G1-G6, represented with different colors as shown in Figure 7A) were employed in the evaluation of clinical scores and hindpaw thickness variation for the mice during the period ranging from day 30 to 60.As shown in Figure 7B,C, it is evident that the group treated with TPC-U@HAT (G6) exhibits the strongest inhibitory effect on the swelling of paws, whereas the saline-treated group (G2) shows paw swelling with the most serious extent.Figure 7C displays the clinical scores, which range from 0 to 4, and reflects the inflammation and swelling in terms of severity degree.At day 36, all groups showed severe inflammation with poor clinical scores; while at day 60, the group treated by the nanosystem TPC-U@HAT (G6) showed the most substantial reduction in arthritis severity compared with the other treatment groups.The group treated with TPC@HAT (G5) also shows notable improvement at day 60, but the free upadacitinib-treated group (G4) shows much poorer improvement, likely owing to the non-targeting and the consequent low accumulation at RA site and the rapid clearance of the drug upadacitinib by the metabolic system.The variation in body weight was also measured for the mice.As presented in Figure 7D, the groups of mice exhibit both slight weight loss or weight gain during the course of therapy.
Patients with advanced RA often experience a loss of flexibility, and regaining even basic mobility can be extremely challenging.In order to evaluate the CIA model mice's mobility upon therapy, we placed every mouse into a separate chamber which was fitted with an electronic monitor and a running roller (Figure 7E), and their running circles were recorded.The data are depicted in Figure 7F, and clearly, the CIA group treated with TPC-U@HAT (G6) exhibits stronger motility than the other treatment groups.
To further validate the effectiveness of TPC-U@HAT in the treatment of RA, on day 60 the mice were sacrificed for their resected ankle joints which then underwent histological analyses, including safranin-O staining and H&E staining analyses.The results are presented in Figure 7G,H.With regard to H&E staining, as compared with the control, the CIA model group which were treated with saline exhibit cartilage erosion and destruction to a severe degree.Treatment with upadacitinib or TPC@HAT partially alleviates the destruction and erosion of cartilage in the ankle joint sections, and treatment with TPC-U@HAT reduces cartilage erosion to a much greater extent (Figure 7G).Moreover, safranin-O staining analyses reveal the degree of cartilage damage in terms of cartilage proteoglycan content, since safranin-O binds to the glycosaminoglycans in cartilage and stains them red.Figure 7H clearly shows that there is a loss of proteoglycan in the CIA model mice treated with saline, indicating severe damage and degeneration to articular cartilage.In comparison, the CAPE-treated group shows slight improvement in cartilage repair, whereas the upadacitinib or TPC@HAT-treated groups show preservation of the cartilage structure.Whereas the TPC-U@HAT-treated group demonstrates the best efficacy, with the cartilage similar to that of the control group.Such serum proinflammatory cytokines as IL-6, TNF-α, and IL-1β levels of the mice were then tested by Elisa kits.From Figure 7I, it is clear that the expression of these proinflammatory cytokines is markedly higher in the saline-treated group (G2).In contrast, the lowest cytokine expression is found in the TPC-U@HAT-treated group (G6) among the groups except the control.These data further confirm the good therapeutic efficacy of the nanosystem TPC-U@HAT.
It is known that CAPE has therapeutic effect on inflammatory diseases through suppressing NLRP3 inflammasome and NF-κB pathway, [23] which are key pathways involved in inflammation by inducing various proinflammatory mediators.On the other hand, the JAK-STAT signaling pathway plays an essential part in controlling immune responses and is crucial for cellular interaction during the pathological process of RA. [19][20][21][22] The inhibition of these pathways is therefore beneficial for treating inflammatory diseases.[22] To further assess the therapeutic effect of the different formulations, the expression levels of the downstream proteins related to the NF-κB, NLPR3, and JAK-STAT pathways were examined for the ankle joint tissues from different groups of mice by Western blotting analyses, and the experimental results are displayed in Figure 7J.As for the control group (G1), the levels of NF-κB p65, NLRP3 inflammasome, and phosphorylated STAT1 (p-STAT1) are rather low, which indicates in healthy mice's ankle joints these pathways generally remain in low activity.As for the saline-treated CIA model mice group (G2), it is clear that all the protein levels including NF-κB p65, NLRP3 inflammasome, and p-STAT1 turn out markedly high, which suggests the rheumatoid arthritis is closely correlated to the activation of NLRP3 inflammasome, NF-κB, and JAK-STAT pathways.Whereas in the TPC-U@HAT-treated group (G6), the levels of NLRP3 inflammasome, NF-κB p65, and p-STAT1 reduce significantly and become comparable to those in group G1.These results clearly reveal that TPC-U@HAT rehabilitates rheumatoid arthritis by inhibiting the NF-κB signaling pathway and JAK-STAT signaling pathway as well as suppressing NLPR3 inflammasome activation.To validate the biosafety of the nanoformulations, we performed H&E staining of major organs of the groups treated by the nanosystems.There are no obvious lesions in major organs, which indicates that the nanosystems have a good biosafety (Figure S25).
Taken together, it is clear that the nanosystem TPC-U@HAT can specifically target the RA joint sites and release two drugs (CAPE and upadacitinib), triggered by the overexpressed in-situ biomarkers hyaluronidase and ROS (H 2 O 2 ).On one hand, the CAPE released from the nanosystem not only suppresses the NF-κB signaling pathway but also inhibits NLRP3 inflammasome, resulting in the decreased production of proinflammatory cytokines.On the other hand, the released upadacitinib inhibits the JAK-STAT pathway by downregulating the expression of phosphorylated STAT1.These actions together ensure an effective therapeutic outcome.

CONCLUSION
In summation, we have developed a nanosystem TPC-U@HAT that can be used for imaging and treatment of rheumatoid arthritis (RA).The nanosystem TPC-U@HAT responds to dual pathological biomarkers of RA, hyaluronidase, and ROS.With the ALN-modified HA polymer as the carrier, the resultant nanosystem TPC-U@HAT has multitargeting capabilities, including active targeting (CD44 receptor targeting and bone targeting) as well as passive targeting, allowing it to effectively accumulate at RA foci.The pathological biomarkers present at the RA site trigger the activation of the AIE-active chromophore TPY for both MSOT and NIR-II fluorescence imaging, enabling disease focus detection and treatment efficacy monitoring, and in the meantime inducing the release of two drugs (CAPE and upadacitinib) for treating RA by suppressing the NF-κB and JAK-STAT signaling pathway as well as inhibiting the NLRP3 inflammasome.Additionally, the obtained 3D MSOT images enable the visualization of the RA foci with 3D information.Overall, the nanosystem herein provides an actionable and practical tool for the imaging and treatment of RA, and the approach may provide insightful and useful information for designing other multifunctional nanosystems for imaging and treating other refractory diseases.

A C K N O W L E D G E M E N T S
This work was supported by NSFC (22274057, 21875069, and 21788102), Guangdong Provincial Basic and Applied Basic Research Fund Regional Joint Fund Project (Youth Fund Project) (2022A1515110842), Chinese Postdoctoral Science Foundation (2022M711194).

C O N F L I C T O F I N T E R E S T S TAT E M E N T
The authors declare no conflict of interest.

E T H I C S S TAT E M E N T
All experiments involving animals were conducted with approval from the Ethics Committee of the Laboratory Animal Center of South China Agricultural University (approval no.: 2022-D082).

D ATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the authors upon reasonable request.

F I G U R E 4
Evaluation of the in vivo targeting ability of TPC@HAT in zymosan-induced rheumatoid arthritis.(A) NIR-II fluorescent images of the ZIA mice (zymosan dose of 25 mg⋅kg −1 ) and the control at various time points following the i.v.injection of TPC, TPC@HT, or TPC@HAT.Mice in prone posture during imaging experiment.Excitation wavelength: 808 nm.(B) Average NIR-II fluorescent intensities of the ankle joint area (ROI, green circle) in the mice of (A).(C) NIR-II fluorescent images for the left legs from different groups at 4 h upon i.v.injection of different formulations.Excitation wavelength: 808 nm.

F
I G U R E 5 TPC@HAT's imaging for zymosan-induced rheumatoid arthritis.(A) Schematic presentation for the experimental outline of establishment and imaging of the zymosan-induced rheumatoid arthritis in mice.(B) Representative bright-field images (namely photographs taken with halogen light) and NIR-II fluorescent images of the ZIA model mice (mice were pretreated with the zymosan dose of 10 or 25 mg⋅kg −1 for 24 h) and the control group at 0 and 4 h after the nanosystem TPC@HAT's i.v.injection.Green circle: ankle joint area.Mice in prone posture during imaging experiment.Excitation wavelength: 808 nm.(C) Average NIR-II fluorescent intensities for the ankle joint areas matching the ROI (indicated by green circle) in (B).(D) Representative orthogonal-view 3D MSOT images of the mice at 0 h and 4 h upon the nanosystem TPC@HAT's i.v.injection (scale bar: 3 mm).1:left leg; 2: tail; 3. right leg.Mice in prone posture during imaging experiment.White dotted rectangular: ankle joint area.Mice's legs were stretched straight during MSOT imaging.(E) Photographs of the left legs from the ZIA model mice with treatment by Zymosan at the dose of 10 or 25 mg⋅kg −1 for 24 h and the control.(F) H&E staining analysis of different groups' left ankle joint sections (scale bar: 200 μm).

F I G U R E 6
TPC-U@HAT's therapy for collagen-induced rheumatoid arthritis and outcome monitoring.(A) Schematic outline showing the treatment course for the collagen-induced rheumatoid arthritis model.(B) NIR-II fluorescence images for the groups G1-G6 at 4 h upon TPC-U@HAT's i.v.injection.The imaging experiment was performed for the mice at day 60.Green circle: ankle joint area.Mice in prone posture during imaging experiment.Excitation wavelength: 808 nm.(C) Average NIR-II fluorescent intensities for the ankle joint areas matching the ROI (indicated by green circle) in (B).(D) A mouse's photograph showing the region for scanning during MSOT imaging experiment.(E) Orthogonal-view 3D MSOT images of the G1-G6 groups at 4 h after the nanosystem TPC-U@HAT's i.v.injection (scale bar: 3 mm).Mice in prone posture during imaging experiment.Mice's legs were stretched straight during MSOT imaging.G1: control; G2: saline; G3: CAPE; G4: Upadacitinib; G5: TPC@HAT; G6: TPC-U@HAT.In the first 3D image on the upper row, coordinate planes represented by each subpanel are indicated by green coordinate axes.

F I G U R E 7
Confirmation of TPC-U@HAT's therapeutic efficacy for collagen-induced rheumatoid arthritis.(A) Notation with colors and marks for each group.(B) Values of relative hind paws' thickness (in mm, relative to the values at day 30) in the day 30-60 period for different groups.(n = 5).(C) Clinical scores of hind paws' in the day 30-60 period for different groups.(n = 5).(D) Body weights measured from day 30 to 60 for G1-G6 (n = 5).(E) A mouse's photograph showing it in the running chamber where the running circles' numbers were recorded.(F) Number of running circles during a 48-h period at day 60 for G1-G6.(G) Analyses by H&E staining of the ankle joint sections at day 60 for different groups (scale bar: 200 μm).(H) Analyses by Safranin-O staining of the ankle joint sections at day 60 for different groups (scale bar: 200 μm).(I) Expression levels of serum IL-6, TNF-α, and IL-1β of different groups at day 60.(J) p-STAT1, NLRP3, and NF-κB levels in ankle joint tissues assessed by Western blotting analyses for G1-G6.Histones H3 as the loading control for nuclear protein, and β-actin as the loading control for cytosolic protein.Data: the mean ± SD.