A Nanobody-Mediated Virus-Targeting Drug Delivery Platform for the Central Nervous System Viral Disease Therapy

ABSTRACT Viral diseases of the central nervous system (CNS) represent a major global health concern. Difficulties in treating these diseases are caused mainly by the biological tissues and barriers, which hinder the transport of drugs into the CNS. To counter this, a nanobody-mediated virus-targeting drug delivery platform (SWCNTs-P-A-Nb) is constructed for CNS viral disease therapy. Viral encephalopathy and retinopathy (VER), caused by nervous necrosis virus (NNV), is employed as a disease model. SWCNTs-P-A-Nb is successfully constructed by employing single-walled carbon nanotubes, amantadine, and NNV-specific nanobody (NNV-Nb) as the nanocarrier, anti-NNV drug, and targeting ligand, respectively. Results showed that SWCNTs-P-A-Nb has a good NNV-targeting ability in vitro and in vivo, improving the specific distribution of amantadine in NNV-infected sites under the guidance of NNV-Nb. SWCNTs-P-F-A-Nb can pass through the muscle and gill and be excreted by the kidney. SWCNTs-P-A-Nb can transport amantadine in a fast manner and prolong the action time, improving the anti-NNV activity of amantadine. Results so far have indicated that the nanobody-mediated NNV-targeting drug delivery platform is an effective method for VER therapy, providing new ideas and technologies for control of the CNS viral diseases. IMPORTANCE CNS viral diseases have resulted in many deadly epidemics throughout history and continue to pose one of the greatest threats to public health. Drug therapy remains challenging due to the complex structure and relative impermeability of the biological tissues and barriers. Therefore, development in the intelligent drug delivery platform is highly desired for CNS viral disease therapy. In the study, a nanobody-mediated virus-targeting drug delivery platform is constructed to explore the potential application of targeted therapy in CNS viral diseases. Our findings hold great promise for the application of targeted drug delivery in CNS viral disease therapy.

Expression and affinity analysis of NNV-Nb. For targeted drug delivery, nanocarriers need to conjugate with targeting ligands, which can specifically recognize and bind to the targets. As one of the most popular targeting ligands, antibody has been widely used owing to its excellent properties (18). However, the application of traditional antibody as targeting ligand is restricted by factors such as large molecular weight, strong immunogenicity, and high production cost. Hamers-Casterman et al. (22) found that a variable domain of heavy-chain antibodies (HCAbs) that are devoid of light chains exists in camelids (camels, llamas, alpacas). The antigen-binding site of HCAbs is composed of one single domain, referred to as Nb, which is considered to be the smallest antibody with complete antigen-binding function (22,31). The characteristics of Nb have just covered the shortages of traditional antibodies, such as small molecular weight (;15 kDa), strong tissue penetration, weak immunogenicity, good solubility, and easy expression in various hosts (20,21), making it an ideal targeting ligand.
Specific Nb can be acquired from nanobody libraries by affinity screening. Nanobody libraries include the naive library and the immunized library. Generally, the naive library is easy to operate without immunization, but the Nb acquired from naive library has a weaker affinity to the target compared with that from immunized library. In addition, more types of Nb with strong affinity can be acquired from the immunized library (21,32). An immunized phage-displayed nanobody library was previously constructed by immunization alpaca with purified NNV. Twenty-two phage clones with strong binding activity to NNV were acquired, and the neutralizing activity of Nbs was checked (unpublished data). In the study, a specific Nb, NNV-Nb, without neutralizing activity against NNV was selected as the targeting ligand.
The deduced amino acid sequence of NNV-Nb is shown in Figure 1A. NNV-Nb is composed of four framework regions (FRs) and three complementarity determining regions (CDRs), which is similar to the variable domain of the heavy chain (VH) of traditional antibodies. However, the CDR1 and CDR3 of Nb are larger than those of VH, not only providing a sufficiently large antigen interacting surface but also forming a variety of paratope structures to recognize special antigenic epitope (21,33). As shown in Figure 1B, NNV-Nb can be detected in supernatant and sediment fractions of the ruptured cells after induction by IPTG (isopropyl-b-D-thiogalactopyranoside). The molecular weight of NNV-Nb is approximately 18 kDa, which is consistent with the theoretical molecular weight and another study (34). NNV-Nb in the supernatant fraction was purified using Ni-chelating affinity chromatography and analyzed by SDS-PAGE (Fig. 1C).
Affinity of the purified NNV-Nb to NNV was checked by indirect enzyme-linked immunosorbent assay (ELISA). As shown in Figure 1D, NNV-Nb exhibited a binding ability to NNV even at 0.035 mg/mL. Previously, an NNV-specific Nb was acquired from a naive library, and results showed that the Nb had no binding ability to NNV below 3.125 mg/ mL (30). The data indicated that Nb acquired from naive library has an obviously weaker affinity compared with that from immunized library.
Construction and characterization of the drug delivery system. As one of the most promising nanocarriers, CNTs have been wildly used owing to their excellent properties, such as strong tissue penetration, high carrying capacity, and needle-like structure (15). CNTs have an intercellular diffusion rate higher than that of globular nanoparticles with similar weight due to their high aspect ratio and needle-like structure (13). CNTs are uniquely equipped to carry drugs and other ligands across biological membranes; particularly, they have shown an intrinsic ability to cross the BBB in vitro and in vivo (14). In addition, CNTs can effectively penetrate the fish epidermis and enter into various tissues (35) and the CNS (30) by immersion. Drugs, proteins, and nucleic acids can be conjugated with CNTs by covalent or noncovalent interaction (17,18). Chemical compounds with extended p -structures can easily be bound to CNTs by p-p interactions (36), such as fluorescein isothiocyanate (FITC).
In the present work, SWCNTs, amantadine, and NNV-Nb were, respectively, selected as the nanocarrier, anti-NNV drug, and targeting ligand to construct a targeted drug delivery platform ( Fig. 2A). To improve the dispersibility and biocompatibility, SWCNTs were first oxidized by a H 2 SO 4 /HNO 3 mixture and then functionalized with polyethylenimine (PEI). In addition, PEI provides a large number of active amino groups for subsequent reactions (37). NNV-Nb was linked on the outermost layer using butanedioic anhydride as a linker to prevent adverse effects on the binding activity. FITC was conjugated by p -p interactions.
As shown in Figure 2B, the P-SWCNTs were fibrous with various lengths. For SWCNTs-P-A-Nb, an obvious layer around SWCNTs surface was observed visually ( Fig. 2C and D), indicating that PEI, amantadine, and NNV-Nb may be conjugated on the surface of SWCNTs. To further verify the conjugation, SWCNTs-P-A-Nb was analyzed by X-ray photoelectron spectroscopy (XPS). As shown in Figure 2E, SWCNTs surface consisted mainly of carbon (C; 284 eV), as well as small amounts of oxygen (O; 532 eV), iron (Fe; 711 [Fe 2p1/2] and 725 [Fe 2p3/2] eV), cobalt (Co; 782 eV), and nickel (Ni; 850 eV). For O-SWCNTs, only carbon and oxygen can be identified from the spectrum. Surface oxygen contents for P-SWCNTs and O-SWCNTs were 3.18 and 14.10 atomic percent, indicating that the oxidation of SWCNTs was sufficient, and the impurities were removed. Nitrogen (399 eV) can be identified from the spectrum of SWCNTs-PEI, indicating that PEI was successfully conjugated with SWCNTs. Surface nitrogen contents for SWCNTs-PEI and SWCNTs-P-A-Nb were 6.81 and 7.23%, indicating that the SWCNTs-P-A-Nb was successfully constructed. The loading efficiencies of amantadine and NNV-Nb were 37.92% and 15.03%, respectively.
As shown in Figure 2F, PEI and amantadine were absolutely degraded at 430 and 220°C, respectively. P-SWCNTs showed a better thermostability than O-SWCNTs because the carboxyl groups and other functional groups of O-SWCNTs are unstable at high temperature. Based on the weight losses of O-SWCNTs (6.61%) and SWCNTs-PEI (19.71%) at 430°C, it can be calculated that the loading efficiency of PEI on SWCNTs was about 13.10%. Previously, PEI was loaded on multiwalled carbon nanotubes (MWCNTs), and the loading efficiency was 9.11% (30). The data indicate that SWCNTs have a higher PEI loading efficiency than MWCNTs, which may be attributed to the larger specific surface area of SWCNTs (36). The loading efficiencies of FITC were 8.34% and 6.41% for SWCNTs-P-F-A and SWCNTs-P-F-A-Nb, respectively.
The particle sizes and zeta potentials of the constructs were measured to check their stability. As shown in Figure 2G and Table 1, the average sizes for P-SWCNTs, O-SWCNTs, SWCNTs-PEI, and SWCNTs-P-A-Nb were 722.86, 136.97, 187.39, and 221.15 nm, respectively. The data indicate that P-SWCNTs were easily aggregated in water due to the hydration and reduction of electrostatic repulsion (38). The dispersibility was significantly improved following oxidization and PEI conjugation, which is consistent with the results reported in a previous study (39). The sizes were gradually increased following amantadine and NNV-Nb conjugation, suggesting the successful conjugation. Zeta potential analysis ( Fig. 2H and Table 1) revealed a negative surface charge (219.06 mV) for P-SWCNTs, which decreased to 249.90 mV following oxidization due to the increase of carboxyl group. After PEI conjugation, the zeta potential was increased to 133.59 mV. The zeta potential was decreased to 236.20 mV for SWCNTs- . Characterization of the constructs using X-ray photoelectron spectroscopy (E), thermogravimetric analysis (F), and nano-particle size and Zeta potential analysis (G: particle size; H: zeta potential). P-A-Nb due to the negative charge of NNV-Nb. The data further proof that the system was constructed successfully.
NNV targeting of the drug delivery system. The specific recognition and binding ability of drug delivery system to the targets is the basis and premise of targeted therapy. Therefore, the NNV targeting of SWCNTs-P-A-Nb was well checked before the anti-NNV activity evaluation. As shown in Figure 3A, both SWCNTs-P-F-A and SWCNTs-P-F-A-Nb were internalized in NNV-infected cells, and a stronger green fluorescence was observed in SWCNTs-P-F-A-Nb group. The internalization was quantitatively measured by flow cytometry analysis. As shown in Figure 3B, the percentage of positively labeling cells was 91.2% in SWCNTs-P-F-A-Nb group, while that was 65.5% for SWCNTs-P-F-A treatment. Previously, the internalization of MWCNTs (uncombined with Nb) in SSN-1 cells was checked, and results showed that the percentage of positively labeling cells was 32.57% (30). The data indicated that SWCNTs have a stronger penetrability than MWCNTs, which is consistent with a previous study (40). To further verify NNV targeting, the specific recognition and binding ability of SWCNTs-P-A-Nb to NNV was checked by transmission electron microscopy (TEM) observation. As shown in Figure 3C, the specific binding of virions with SWCNTs-P-A-Nb can be observed, indicating that the increased internalization of SWCNTs-P-A-Nb in SSN-1 cells is correlated with the specific binding of Nb to NNV.
For NNV targeting in vivo, SWCNTs-P-F-A was distributed mainly in the abdomen, while SWCNTs-P-F-A-Nb could enter into the head in addition to the abdomen (Fig. 4A). Moreover, the fluorescence in abdomen for SWCNTs-P-F-A-Nb groups was weaker than that for SWCNTs-P-F-A groups, indicating that the nonspecific distribution is reduced following NNV-Nb conjugation. The brains were collected and subjected to fluorescence imaging (Fig. 4A) and tissue section observation (Fig. 4B). Stronger fluorescent signals were observed following SWCNTs-P-F-A-Nb treatment than in the  SWCNTs-P-F-A treatment group, indicating that more SWCNTs-P-F-A-Nb can enter into the brain tissue of diseased grouper. All the results indicated that SWCNTs-P-A-Nb has a good NNV-targeting ability in vitro and in vivo and improves the specific distribution in the NNV-infected sites under the guidance of NNV-Nb. Potential ways for ingestion and excretion. As shown in Figure 5, SWCNTs-P-F-A-Nb was distributed in the muscle, gill, intestine, and kidney of the NNV-infected grouper. Distribution in muscle indicates that SWCNTs-P-F-A-Nb can effectively penetrate the epidermis and muscle by immersion. In addition, SWCNTs-P-F-A-Nb may also enter into the circulatory system by crossing the membranes of gill and intestine. Similar phenomena have been reported by previous studies. For example, Zhu et al. (35) checked the distribution of SWCNTs in rare minnow (Gobiocypris rarus) and found that SWCNTs can enter into the  body through the skin and accumulate in the muscle (35). Moreover, another study found that nanoparticles can pass through the membranes of gill/intestine and enter into the circulatory system (41). Distribution in kidney indicates that SWCNTs-P-F-A-Nb can be excreted by kidney. Zhang et al. (23) constructed a targeted SWCNTs-based vaccine delivery system (SWCNTs-MG) for spring viremia of carp therapy. Their results showed that the SWCNTs-MG is distributed in the muscle, intestine, kidney, spleen, and liver of the diseased common carp, which is similar to the present results.
Anti-NNV activity of the drug delivery system. Targeted drug delivery platforms have been widely explored for disease therapy (17,19). However, the most mature area lies in cancer therapy; applications against CNS viral diseases are rarely reported (23). To counter this, a nanobody-mediated NNV-targeting drug delivery platform is constructed to explore the application of targeted therapy in CNS viral diseases.
Toxicity of the drug delivery system was first evaluated, and data showed that both SWCNTs-P-A and SWCNTs-P-A-Nb were safe to SSN-1 cells and grouper juveniles under the concentrations used in the study. As shown in Figure 6A and B, SWCNTs-P-A-Nb showed an anti-NNV file stronger than that of amantadine alone or SWCNTs-P-A group. NNV infection induces host cell death, which generally indicates mitochondria-mediated caspase-dependent cell apoptosis (25). Therefore, apoptosis analysis was performed to check the protective effects on SSN-1 cell from NNV-induced apoptosis. As shown in Figure 6C, compared with those in the negative-control group (NNV-uninfected), the percentages of early (7.3%) and late apoptosis (44.5%) were significantly increased in the positive-control group (NNV-infected). The percentages of normal cells were 77.9%, 86.5%, and 92.3% for amantadine, SWCNTs-P-A, and SWCNTs-P-A-Nb treatments with the same amantadine concentration (12.5 mg/L), respectively. The data indicate that SWCNTs-P-A-Nb has the best protective effect on SSN-1 cells from NNV-induced apoptosis. Similarly, compared with amantadine and SWCNTs-P-A, SWCNTs-P-A-Nb also showed the best anti-NNV activity in vivo (Fig. 7A). As shown in Figure 7B, the survival rate of grouper juveniles was only 4% following infection with NNV for 10 days and increased, respectively, to 27%, 39%, and 51% after amantadine, SWCNTs-P-A, and SWCNTs-P-A-Nb treatments with the same amantadine concentration (40 mg/L).
Metabolism of amantadine in brain was analyzed using a liquid chromatography-mass spectrometer. As shown in Figure 7C, the maximum amantadine contents were reached at 24 h after exposure to amantadine (11.88 mg/kg) or SWCNTs-P-A (19.34 mg/kg), while that was reached at 12 h for SWCNTs-P-A-Nb treatment (34.57 mg/kg). After transferring to seawater without drug, the contents of amantadine were decreased. There was no detectable signal at around 120 h after exposure to amantadine or SWCNTs-P-A, while that lasted to 192 h for SWCNTs-P-A-Nb treatment. These data indicate that SWCNTs-P-A-Nb can transport amantadine to the brain in a fast and efficient manner and prolong the action time of amantadine. A similar result has been reported by previous studies (17,42). For example, Ren et al. (42) have constructed a dual-targeting drug delivery system (O-MWCNTs-PEG-ANG) for brain glioma treatment. They demonstrated that the O-MWCNTs-PEG-ANG can greatly increase the accumulation of doxorubicin in brain (42). Targeted drug delivery system can transport and release drugs to the specific sites and decrease the distribution in normal tissues, thereby reducing toxicity and side effects and prolonging the circulation time, ultimately improving the effects of drug therapy (9, 10).
The above results show that SWCNTs-P-A-Nb has a stronger anti-NNV activity than amantadine or SWCNTs-P-A treatment group with the same amantadine concentration. Therefore, it can be concluded that SWCNTs-P-A-Nb can efficiently transport and release amantadine at the NNV-infected sites under the guidance of NNV-Nb, improving the anti-NNV activity of amantadine.
Conclusions. In summary, the potential application of targeted drug delivery in CNS viral disease therapy was explored using VER as a disease model. A nanobodymediated virus-targeting drug delivery platform was constructed employing SWCNTs, grouper juveniles were collected and subjected to fluorescence imaging and tissue section observation using a fluorescence stereomicroscope (Leica MZFL III, Germany) (30). Healthy grouper juveniles treated without SWCNTs-P-F-A or SWCNTs-P-F-A-Nb were used as the controls.
Ingestion and excretion analysis. Following treatments as described in the section above, the grouper juveniles were killed. The muscle, gill, intestine, and kidney were collected for tissue section observation using the fluorescence stereomicroscope to check the potential ways for ingestion and excretion of SWCNTs-P-F-A-Nb.
Anti-NNV activity of the drug delivery system in vitro. To evaluate the anti-NNV activity of the drug delivery system in vitro, toxicity of SWCNTs-P-A and SWCNTs-P-A-Nb to SSN-1 cells was first checked by trypan blue staining test (44). SSN-1 cells were grown to a monolayer in 12-well plates and infected with 10 2 TCID 50 NNV. Following infection for 2 h, the cells were, respectively, treated with amantadine, SWCNTs-P-A, and SWCNTs-P-A-Nb with the same amantadine concentrations (3.125, 12.5, and 50 mg/L). Following treatment for 48 h, the relative NNV contents in SSN-1 cells were analyzed by reverse transcriptase quantitative PCR (RT-qPCR) and Western blotting. Moreover, apoptosis analysis was carried out based on Annexin V/propidium iodide (PI) staining (Beyotime Biotech, Nantong, China). Briefly, SSN-1 cells were grown to a monolayer in 6-well plates and treated as described above. Approximately 2 Â 10 5 cells were collected and stained with annexin V-FITC (5 mL) and PI (5 mL) following the manufacturer's instructions. After staining, flow cytometry analysis was conducted immediately. FITC fluorescence (FL1) and PI fluorescence (FL2) of each cell were quantitated using the Cell Quest Pro software. The annexin V-FITC 2 /PI 2 , annexin V-FITC 1 /PI 2 , and annexin V-FITC 1 /PI 1 populations were regarded as normal cells, early apoptosis, and late apoptosis, respectively.
Anti-NNV activity of the drug delivery system in vivo. Toxicity of SWCNTs-P-A and SWCNTs-P-A-Nb to grouper juveniles was first evaluated. For anti-NNV activity evaluation, healthy grouper juveniles were infected with 10 5 TCID 50 NNV by intraperitoneal injection for 1 day and then, respectively, treated with amantadine, SWCNTs-P-A, and SWCNTs-P-A-Nb with the same amantadine concentrations (2.5, 10, and 40 mg/L) by immersion. Grouper juveniles infected with NNV and treated without drugs were used as the control. Nine grouper juveniles were randomly collected at 3, 5, and 7 days postinoculation (dpi), and the brain was collected for NNV contents analysis. The same treatment was performed to evaluate the protective effects of amantadine, SWCNTs-P-A, and SWCNTs-P-A-Nb on grouper from NNV infection over a 10-day period.
Amantadine metabolism. Grouper juveniles were infected with 10 5 TCID 50 NNV by intraperitoneal injection for 3 days and then, respectively, treated with amantadine, SWCNTs-P-A, and SWCNTs-P-A-Nb with the same amantadine concentration (20 mg/L) by immersion. Following exposure for 24 h, the grouper juveniles were transferred to clean seawater without drug. The brains were collected at the specific time points and fully ground by a homogenizer. After centrifugation for 30 min, the supernatants were collected. Contents of amantadine in the supernatants were measured using a liquid chromatography-mass spectrometer (Thermo Scientific, USA) according to a previous study (45).
Statistical analysis. Data were presented as mean 6 standard deviation (SD). Statistical comparisons between two groups were performed using the two-tailed unpaired Student's t test, and comparison between more than two groups was analyzed with one-way analysis of variance (ANOVA). Values of *, P # 0.05, **, P # 0.01, and ***, P # 0.001 were applied to annotate statistical significance.

ACKNOWLEDGMENTS
This work was supported by the China Postdoctoral Science Foundation (2020M683222 and 2021T140567), Natural Science Foundation of Chongqing (cstc2021jcyj-msxmX0450), and Fundamental Research Funds for the Central Universities (SWU120034).
All animal experiments in the study were approved by the Animal Ethical and Welfare Committee of Northwest A&F University (Yangling, China) and conducted in accordance with the Ethics Procedures and Guidelines for Animal Use of the People's Republic of China. All authors agree to be published. All data generated or analyzed during this study are included in this published article. Song Zhu and Gao-Xue Wang designed the research. Song Zhu, Fei Luo, Bin Zhu, and Tian-Qiang Liu performed the experiments. Fei Ling and Er-Long Wang provided help in data processing and figure preparing. Song Zhu, Fei Luo, and Gao-Xue Wang wrote the manuscript. All authors read and approved the manuscript.
We declare no conflicts of interest.