1-(3-Dimethylaminopropyl)-3-ethylcarbodiimide hydro (EDC), N-hydroxy succinimide (NHS), cysteine (Cys), hyaluronic acid (HA), and methacryloyl chloride (MA) were purchased from Aladdin Reagent, Ltd. (Shanghai, China). All chemicals used in this study were of analytical grade, and reagents were used without further purification. Dulbecco’s modified Eagles medium (DMEM) cell culture medium, fetal bovine serum, streptomycin, and penicillin were purchased from Thermo Fisher Scientific Co., Ltd. (China). All aqueous solutions were prepared using ultrapure deionized water (DI water), which was obtained through a Millipore Milli-Q water purification system (Billerica, United States) and had an electric resistance >18.2 MΩ.

The UV–vis absorption spectra were recorded on a UV–vis spectrometer (Lambda 35 UV–vis spectrometer, Perkin-Elmer, United States) at room temperature. The sizes of micelle nanoparticles were measured using ZEN3690 zetasizer (Malvern instruments, Zetasizer Nano-ZS) at room temperature. The confocal fluorescence images of the cell were collected on a Zeiss Laser Scanning Confocal Microscope (ZEISS LSM 780, Germany). In vivo fluorescence images were captured by the Xenogen IVIS Spectrum system (Caliper Life Sciences, Hopkinton, MA).

All animal experiments were approved by the Institutional Animal Ethics Committee at Southern Medical University, and the experiments were performed in compliance with the National Institutes of Health (NIH) guidelines for the Care and Use of Laboratory Animals of Southern Medical University. The granuloma-bearing mouse models were in accordance with the previous model reported according to the literature (Carlsson et al., 2010; Oehlers et al., 2015; Fenaroli et al., 2018; Liao et al., 2020). Briefly, female C57BL/6 (B6) mice (8–12 weeks old) were purchased from Guangdong Medical Laboratory Animal Center (Guangdong, China). To establish the granuloma-bearing mouse models, Mycobacterium marinum (M.m) were first grown in 7H9 broth media for 7–10 days in a bacterial shaking incubator. When the mid-log phase (OD600 = 0.7 ± 0.2) was achieved, the bacteria were collected, washed, and then passed 20 times through a needle to disrupt bacterial aggregates. After that, the supernatants were transferred to new tubes and diluted in sterile PBS to a final concentration of 2 × 108 CFU ml^-1. Two hundred microliters of the M.m suspension was intravenously injected into mice using an insulin syringe. The bacteria infection-induced tail granulomas were successfully formed after 2 weeks.

In order to realize the theranostic effect on tuberculosis, the GSH-responsive nanoagent was rationally designed for targeted imaging and therapy of tuberculosis. RIF@Cy5.5-HA-NG was first synthesized between two types of extensively biocompatible hyaluronic acid (HA) as the host material and near-infrared dye Cy5.5 as a contrast agent via the photo-initiated bioorthogonal reaction and the hydrophilic-hydrophilic interaction. Furthermore, the loaded rifampicin was released through the dissociation of disulfide bonds by the original GSH in granulomas, realizing targeted tuberculosis therapy and providing especially accurate treatment mapping for tuberculosis. Glutathione, as a tripeptide and antioxidant, is synthesized at high levels under intracellular oxidative stress (Aquilano et al., 2014; Zheng et al., 2019a), playing an important role in apoptosis and regulating pathogen-infected-host interaction, including inhibition of M. tuberculosis replication (Seres et al., 2000; Venketaraman et al., 2003; Venketaraman et al., 2005). The resulting system was characterized for GSH-responsive rifampicin release, real-time monitoring, and antibiosis properties. Then, the prolonged retention time of drug-release in vitro and the in vivo was demonstrated using fluorescence imaging techniques.

In this study, HA-Cys-MA and HA-Lys-Tet, which could first form nanocages via UV-induced click reaction, were mixed with rifampicin (RIF) to create a RIF-loaded carrier (RIF@HA-NG). Photo-inducible click chemistry has been widely applied to functionalize and investigate the dynamics and roles of biomolecules in living systems (Le Droumaguet et al., 2010; Zhou et al., 2016; Nainar et al., 2017; Liu et al., 2021). The fluorescent imaging contrast Cy5.5 was then modified on the RIF@HA-NG through amidation in the existence of carbodiimide (EDC) and N-hydroxysuccinimide (NHS) to obtain the aimed nanosystem, which combined the diagnosis and therapy of tuberculosis (Scheme 1). Among them, the synthetic routes of HA-Cys-MA and HA-Lys-Tet are summarized in Supplementary Figures S1, S2, respectively. On the one hand, GSH plays an important role in many diseases, including cancer and tuberculosis (Allen et al., 2015; Zheng et al., 2019b). Moreover, the cysteine (Cys) containing disulfide bond (Yang et al., 2015; Wang et al., 2020) was reasonably chosen to possess the GSH-responsive peculiarity. On the other hand, a polymer pre-monomer containing the photo-click functional groups, including methacryloyl (MA) (Maiti et al., 2020) and tetrazolium (Tet) (Wu et al., 2019; Liu et al., 2021), was designed and synthesized to obtain a controllable nano-delivery system.

The synthesis of RIF@HA-NG was first characterized by the dynamic light scattering (DLS) analyzer. As shown in Figure 1A, the RIF@HA-NG nanoagent had a hydrodynamic diameter of approximately 120 nm, which was slightly larger than that of HA-NG owning to the RIF loading (Figure 1A). To further confirm the successful loading of RIF, the zeta potentials were measured, showing the zeta potential change from −27.5 to −31.3 mV. In other words, the negative RIF obviously decreased the zeta potential of the nanoagent RIF@HA-NG (Figure 1B). Transmission electron microscopy (TEM) revealed that RIF@HA-NG exhibited a uniform morphology and size with a diameter of 100 nm, indicating that no obvious changes were recorded in the size and shape of the nanoagent after loading RIF (Figure 1C). Meanwhile, the drug-release ability of nanocarriers was assessed, so the RIF release analysis was explored (Figure 1D). Obviously, the ∼55% of the total release was observed within 10 h, and the total release reached ∼70% upon GSH treatment for 70 h. In contrast, the total release reached only ∼20% in PBS solution up to 70 h. In vitro GSH-triggered drug-release analysis indicated that a greater amount of RIF can be released in Mycobacterium-infected macrophage cells. Collectively, these results confirmed that the GSH-responsive water-soluble RIF@HA-NG was successfully synthesized.

With the nanoagent in hand, we then aimed to estimate the deliverability of RIF in targeting cells. To determine whether the RIF-loaded nanoagent was endocytosed by the macrophage cell, the RIF@HA-NG was modified by the near-infrared fluorescence dye Cy5.5, simultaneously realizing the imaging of tuberculosis. As shown in Figure 2, the red fluorescence was detected in the RIF@Cy5.5-HA-NG-treated group compared to the Cy5.5-treated group, indicating that RIF@HA-NG can accumulate in the granuloma. That is to say, RIF@HA-NG was successfully modified by Cy5.5. Thus, it can be rationally used to monitor tuberculosis.

Owing to the outstanding drug-release properties of RIF@Cy5.5-HA-NG toward GSH-enriched tuberculosis, we then investigated its antibacterial performance. To investigate the antibacterial activity of RIF and RIF@HA-NG in vitro, M. tb-infected M1 macrophage cells were first incubated with RIF and RIF@Cy5.5-HA-NG at different times. Survival analysis was performed to confirm the antibacterial effects of RIF and RIF@Cy5.5-HA-NG. As displayed in Figure 3A, the survival rate of the group treated with RIF@Cy5.5-HA-NG for 1–3 h was lower than that of the RIF-treated group. In particular, RIF@Cy5.5-HA-NG or RIF was co-incubated with cells for 3 h, and the survival rate of bacteria decreased to 28 and 63%, respectively. When the processing time was extended, the bacterial damage caused RIF@Cy5.5-HA-NG to increase, but the survival rate changed gently, which may have been caused by the phytocytosis of macrophages. Meanwhile, the survival analysis of RIF@Cy5.5-HA-NG and RIF treated with Mycobacterium-infected M2 macrophage cells also showed a similar antibacterial tendency, indicating the antibacterial activity of nanoparticles (NPs) was superior to that of pure RIF (Figure 3B). Taken together, these results performed that RIF@Cy5.5-HA-NG NPs have clipping high antibacterial efficiency against cellular bacteria in vitro and therefore hold potential for tuberculosis treatment.

Having established that RIF@Cy5.5-HA-NG nanoagent could efficiently kill Mycobacterium, our next goal was to validate its potential to monitor the Mycobacterium-infected mice. First, we established the tuberculosis model by injecting Mycobacterium marinum into the tail vein of mice according to our previously reported methods (Liao et al., 2020). The mice were imaged under a 633 nm laser using the fluorescence in vivo imaging system at different times (0, 6, 12, and 24 h). As exhibited in Figure 4A, the fluorescence intensity at 690 nm in the granuloma region was found to reach the maximum at 24 h, indicating that the concentration of RIF@Cy5.5-HA-NG increased gradually over time and was maintained at a relatively high level even at 24 h after injection. To further investigate the distribution of nanoagent in various organs at 2, 4, and 24 h time points, ex vivo fluorescence imaging was also studied. Intense fluorescence signals at 690 nm were observed in the liver (Figure 4B), which indicated that the nanoagent was preferred to selectively accumulate in the liver. The selective accumulation may be due to the reticuloendothelial system (Zheng et al., 2019; Peng et al., 2017), which indicated that these nanoagents could be metabolized through the liver. Furthermore, hematoxylin and eosin (HE) staining of various important organs revealed no pathological changes after RIF@Cy5.5-HA-NG nanoagents injection at different time points (Figure 4C). Taken together, these results firmly demonstrate that RIF@Cy5.5-HA-NG is capable of directly reflecting tuberculosis, revealing the feasibility of our nanoagent for monitoring the mycobacterium in vivo as an excellent biomaterial.