2.1 MiSp-based NP preparation and characterization

The process of forming NPs by salting out using potassium phosphate is widely regarded as the most biocompatible approach for producing silk protein NPs. Nonetheless, numerous factors, such as protein concentration, phosphate concentration, pH of the potassium phosphate solution, protein purification method, and amino acid sequence, can impact the properties of the resultant protein spheres.^{15, 16, 24, 25} With the goal of identifying appropriate NPs for utilization as potential delivery carriers, the influence of inclusion bodies (IBs) solubilization approaches on NP properties were subsequently explored. SDS-PAGE (sodium dodecyl sulfate polyacrylamide gel electrophoresis) analysis indicated that all three methods generated recombinant NM proteins with good purity and high solubilization efficiency (Figure 1A), resulting in a final yield of approximately 50 mg per liter of LB medium. Circular dichroism (CD) measurements showed that soluble NM proteins from the above approaches shared similar secondary structures (Figure 1B), dominated by α-helix that is probably due to the α-helix bundle conformation of the NT domain.²⁶ Revealed by scanning electron microscopy (SEM), despite being prepared using different approaches, the recombinant NM proteins were all able to form NPs (NM-NPs), which were spherical in shape and exhibited a smooth surface (Figure 1C). However, the mean diameters of the NPs were diverse, 211.2 ± 9.1, 299.8 ± 8.1, and 355.6 ± 7.9 nm for freezing–thawing, heating, and traditional methods, respectively (Figure 1D and Table S1). These diameters are smaller than previously reported spidroin (eADF4)-derived NP size,^{15, 16} which might be due to the different amino acid composition, while in this study the NT domain is included and the M mainly consists of Gly and Ala. Further, the freezing–thawing and heating methods generated more homogeneous NPs than the traditional 8 M urea method, indicated by the polydispersity index (Figure 1E and Table S1). Despite of these, all three types of NPs exhibited similar negative zeta potentials as measured by dynamic light scattering (DLS) (Figure 1F and Table S1). This observation suggests that the negatively charged N-terminal domain (with a pI of 4.1) might be situated on the peripheral side, while the repetitive region M likely forms the core. Based on these observations, it is evident that the various IBs solubilization approaches had discernible effects on the size and homogeneity of the spheres. Taking into account the protein preparation process, sphere characteristics, and the possibility of urea modification, the freezing–thawing method was chosen as the standard method to produce NM proteins for NP preparation for subsequent applications.

2.2 Model protein loading, activity, and releasing behavior of NM-NPs

As NM-NPs are negatively charged at a neutral pH, they could probably form complexes with positively charged molecules via electrostatic interactions. As such, the lysozyme with an isoelectric point of 11.35 and a molecular weight of 14.3 kDa was selected as a model protein. Lysozyme could be efficiently loaded on NM-NPs with loading efficiency nearly 100% and loading at 10% (w/w-ratios) of lysozyme:NM-NPs of 1:10 (Table S2). The SDS-PAGE analysis further confirmed the successful loading of lysozyme on NM-NPs (Lyso-NM-NPs), as indicated by the tiny amount of free lysozyme left in the solution (Figure 2A). SEM observations showed that the Lyso-NM-NPs were uniform and spherical (Figure 2B), but with an increased average size of 510 ± 10.3 nm (Figure 2C) and a decreased negative zeta potential of −9.64 ± 0.5 mV compared with NM-NPs (Figure 2D). The increased size and decreased zeta potential could be probably explained by the lysozyme surface loading on the particles and charge compensation by the opposite charge of lysozyme.

We also studied whether the Lyso-NM-NPs affected the lysozyme activity due to the protein encapsulation. The activity of lysozyme in NM-NPs was evaluated using a common turbidimetric method with Lysozyme (LZM) Assay kit (S1.6, Supporting Information). As shown in Figure 2E, the activity of lysozyme in NM-NPs was not significantly different from that of free lysozyme, with an activity of about 25,000 and 19,000 units/mg, respectively.

To monitor the release kinetics of loaded lysozyme from Lyso-NM-NPs, we tested the in vitro releasing efficiency at different pH levels (Figure 2F). Interestingly, at both pH 2.0 and 4.0, almost 90% of loaded lysozyme proteins were released immediately during the incubation from Lyso-NM-NPs, which is likely due to that the electrostatic interactions were broken up by low pHs. Conversely, Lyso-NM-NPs exhibited sustained lysozyme-release behavior at pH 7.4, with around 22% of preloaded lysozyme released over a period of 27 days, suggesting NM-NPs support sustained release of loaded protein in vitro, which is an important factor for reduced dosage and frequency from the immunization point of view. These results indicate positively changed macromolecular proteins could be loaded on NM-NPs efficiently with minor activity impact, which further present a pH-sensitive releasing profile and supports a sustained release behavior under neutral pH.

Finally, the degradation behavior of NM-NPs was characterized by immersion tests in simulated gastric fluids (SGFs) (S1.7, Supporting Information). The morphology of the NM-NPs was studied with SEM after a static immersion test in a water bath incubator at 37°C for 24 h. As shown in Figure 2G, it is clear that the presence of the organic molecules in the SGFs did not affect the morphology of the particles. The results indicated that the NM-NPs was stable in SGFs.

2.3 Biodistribution and delivery efficiency of NM-NPs in vivo

To comprehensively assess the possibility of NM-NPs as a viable and effective protein carrier, the in vivo biodistribution and the delivery effects of protein drugs were evaluated when administrated through different pathways, including parenteral (subcutaneous [s.c.], intramuscular [i.m.], and intravenous [i.v.] injections) and nonparenteral (per-oral) administration routes. To track the NPs or protein in vivo, different formulations NM-NPs (Cy5.5 labeled), Lyso-NM-NPs (Cy5.5-labeled Lyso), or free lysozyme (Lyso, Cy5.5 labeled) were administrated in Balb/c mice (Table S3), while mice administrated with PBS were as control and analyzed by the noninvasive imaging modality that allows the visualization of biological processes in living organisms using fluorescent Cy5.5, a near-infrared fluorescent dye clinically used in medical diagnostics (Table S4 and Figure S1).

2.3.1 Delivery effect for different administrations

After oral administration of the different formulations, all the mice presented strong fluorescence signals (Figures 3A and B), indicating the mice are successfully administrated. After a 2-h period, only weak fluorescence was detected in the Lyso group, whereas strong signals were still present in the abdominal region for both the NM-NPs and Lyso-NM-NPs groups even until the 4-h mark. For the i.v. route, the mice were administered via tail vein injection. As illustrated in Figures 3C and D, the fluorescent profile showed a rapid distribution from the plasma within the first 6 h, which dropped below the limit of quantification at 24 h for the Lyso group. However, fluorescent signals in the NM-NPs and Lyso-NM-NPs group were still detectable in the abdominal region even at 24 h after administration. The delivery effect of lysozyme loading on NPs over time following s.c. administration is presented in Figures 3E and F. After 6 h of injection, weak fluorescence was detected in the Lyso group, whereas strong fluorescence intensities of Lyso-NM-NPs and NM-NPs were maintained at 12 h postinjection. For i.m. administrations in mice models, the fluorescence intensity of Lyso-NM-NPs was significantly higher and of longer duration at the injection site compared with the Lyso group (Figures 3G and H). While the fluorescence intensity declined at the injection sites at 78 h after injection in Lyso-injected mice, strong fluorescence was still observed for Lyso-NM-NPs after 126 h postinjection. Interestingly, the fluorescence intensity declined more quickly for the NM-NPs group than the other two groups for muscle administration (Figures 3G and H).

2.3.2 Organ biodistribution

To further characterize the tissue distribution of protein and NPs, the mice were sacrificed and organs were collected for ex vivo fluorescence analysis at indicated time point. For oral administration, the fluorescent signals were predominantly observed in the liver and small intestine (Figure 4A). To investigate whether NM-NPs or the loaded protein had reached the Peyer's patch (PP) dome, histological analysis of the PPs from the mice was conducted through frozen sections and observed under a fluorescence confocal microscope. Previous studies have showed that NPs smaller than 10 µm are primarily taken up by M-cells and transported into the PPs.^{27, 28} When compared the differently treated groups, that is, PBS, Lyso, Lyso-NM-NPs, and NM-NPs, the most prominent fluorescence was observed in the subepithelial dome for NM-NPs (Figure 4B). These results demonstrated that NM-NPs were resistant to the degradation of the gastrointestinal tract and efficiently reached the small intestine, where they were taken up by M-cells and transported into the PPs. However, free and NM-NPs delivered lysozymes were not observed in the PPs (Figure 4B), which could be explained by the quick release of the loaded lysozyme from NM-NPs at acidic pHs and its subsequent digestion in the gastrointestinal tract. These findings indicate that NM-NPs are not the suitable carrier for delivering protein drugs or vaccine to the PPs subepithelial dome via the per-oral route.

For parenteral administration routes, the results indicated that the kidney was the primary target of accumulation for both formulations, followed by the liver (Figures 4C–E). Notably, the fluorescence signals of the Lyso-NM-NPs and NM-NPs formulations were significantly higher than that of the free lysozyme group at different organs. Further, the biodistributions of lysozyme and NM-NPs in the heart, spleen, and lungs after i.v. administration were also observed (Figure 4C). The NM-NPs group had the highest fluorescence intensity in the kidney compared with the Lyso-NM-NPs and free lysozyme groups, which may explain the faster signal declining at the injection site after i.m. administration (Figures 3G and 4E).

These results suggest that NM-NPs can effectively prolong the residence time of loaded proteins at injection sites and provide a sustained release effect in vivo, likely due to the depot effect and NP protection. These effects were observed for different administration routes, including i.v., s.c., and i.m. injection, but not for oral delivery, indicating NM-NPs have potential as a protein carrier for various administration routes that could be valuable for drug and antigen delivery.

2.4 Cellular uptake and cross-presentation properties of Lyso-NM-NPs

Considering efficient internalization by APCs is critical for immune activation, we assessed the ability of NM-NPs with antigen loading to be taken up by bone marrow derived cells (BMDCs). We incubated BMDCs with Cy5.5-labeled lysozyme (Lyso) or Cy5.5-labeled lysozyme loaded on NM-NPs (Lyso-NM-NPs), while cells incubated with PBS were as negative control. Confocal microscopy and flow cytometry results showed that the Lyso-NM-NPs group presented ∼3-fold higher fluorescence compared with the Lyso group, suggesting NM-NPs as a delivery carrier can significantly enhance the cellular uptake of loaded protein antigens (Figures 5A and B). We further evaluated the ability of NM-NPs to facilitate cytoplasmic delivery of antigens after internalization into BMDCs using confocal microscopy. After 1 h incubation of BMDCs with Lyso-NM-NPs, yellow fluorescence resulting from the merging of antigen (red) and lysosome (green) was observed, indicating efficient cellular uptake of Lyso-NM-NPs. After 4 h incubation, discrete fluorescence signals for antigen and lysosomes appeared, suggesting successful escape from lysosomes and cytoplasmic delivery of the model antigen lysozyme (Figure 5C). These results together indicate that the loading of lysozyme on NM-NPs enhanced DC internalization, lysosomal escape, and cross-presentation.

2.5 Ability of Lyso-NM-NPs to activate immune responses and the underlying mechanism

Since the solvent-free preparation and the processing conditions can be easily scaled-up, and the high efficiency of loading macromolecular protein as well as the sustained release effect in vivo of different administration routes, the MiSp-based NPs are per se suitable for the loading and delivery of protein drug/vaccine. Hence, we focused on the evaluation of whether NM-NPs could enhance immune responses to load protein vaccine in vivo through s.c. and i.m. administrations that are the primary routes for subunit protein vaccine, and lysozyme was used as a model antigen (Table S5 and Figure S2).

2.5.1 NM-NPs enhance immune responses of loaded antigen through s.c. vaccination

First, mice were subcutaneously vaccinated with soluble antigen alone (Lyso) or antigen loaded on NM-NPs (Lyso-NM-NPs) three times on day 0, 14, and 28, with both groups receiving an equivalent amount of lysozyme, and sacrificed on day 35 and blood was collected for lysozyme-specific antibody measurements by enzyme-linked immunosorbent assay (ELISA). The results revealed that levels of total specific immunoglobulin G (IgG), IgG1, and IgG2a in the Lyso-NM-NPs-treated group were significantly higher compared with the Lyso group (Figures 6A–C). The predominant antibody subtype induced by Lyso-NM-NPs vaccination was IgG1, followed by IgG2a, suggesting a potent Th2 and Th1 immune response. These results indicate that the implementation of NM-NPs as a carrier for protein antigen is able to enhance immune responses in vivo. To further investigate the cytokine profiles, splenocytes were harvested from vaccinated mice and restimulated ex vivo with lysozyme. The levels of endogenous cytokines interferon gamma (IFN-γ) and interleukin-4 (IL-4), indicative of Th1 and Th2 responses, respectively, in the supernatant were analyzed by ELISA. Strikingly, splenocytes from mice immunized with antigen loaded NM-NPs produced much higher levels of IFN-γ and IL-4 cytokines than those injected with the antigen alone (Figures 6D and E). These results suggest that the NM-NPs loading formulation can lead to a stronger mix of Th1 and Th2 immune responses.

2.5.3 Underlying mechanism

To unravel the mechanisms of NM-NPs to enhance immune responses, the draining lymph nodes were collected for ex vivo fluorescence analysis after 24 h injection. Fluorescence images demonstrated that the fluorescence intensity of Lyso-NM-NPs in the lymph nodes was ∼7-fold higher (for s.c. vaccination; Figures 7A and B) or 5-fold higher (for i.m. vaccination; Figures 7C and D) than that of Lyso, which was further confirmed by confocal imaging of lymph node slices (Figure 7E). These findings suggest that Lyso-NM-NPs not only provides sustained antigen exposure but also facilitates efficient antigen uptake by APCs, ultimately leading to enhanced immune responses. As the NP significantly prolongs the residence time of the model antigen at the injection site and significantly affects antigen transport to the draining lymph nodes over time, we next analyzed the extent of DC maturation in the draining lymph nodes by determining the expression of costimulatory molecules CD80 and CD86. The results showed that a significantly higher expression of CD80/86 was observed for Lyso-NM-NPs-treated group compared with free lysozyme group at 48 h after s.c. injection (Figures 7F and G and S3), indicating Lyso-NM-NPs possesses potent immunostimulatory ability.

Collectively, these findings offer compelling evidence that the NM-NPs loading approach significantly augments systemic immune responses, whether administered subcutaneously or intramuscularly. This underscores the remarkable adaptability of NM-NPs as delivery vehicles across diverse administration routes.

2.6 Cytotoxicity and biosafety of NM-NPs

The excellent loading behavior and sustained releasing profile make the NM-NPs as a promising candidate for protein drug/vaccine delivery. With the aim of biomedical applications, we conducted an evaluation of the cytotoxicity of NM-NPs on APCs. Specifically, we exposed BMDCs to different concentrations of NM-NPs (50–500 µg mL⁻¹). After 24 h of exposure, no discernible signs of toxicity upon addition of different concentrations of NM-NPs to the culture media were detected, as demonstrated by the qualitative comparison of cell density between the control group (PBS) and the NM-NPs-treated group (Figure 8A).

Next, we investigated whether these NPs lead to unspecific DC maturation in themselves. BMDCs were incubated with NM-NPs for 24 h before testing the costimulatory factors on the cell surface (CD80 and CD86). The result showed that none of the particle in themselves increased the density of the costimulatory factors, while Lyso and Lyso-NM-NPs significantly induced the BMDCs maturation through upregulation of CD80 and CD86 (Figures 8B and C and S4). Thus, the spider silk NPs derived from NM protein did not show unspecific immune activation in vitro.

To get some insights into the biocompatibility, the biosafety of the NM-NPs in vivo were assessed after multiple i.v. administration to mice. We conducted a histological examination of tissues from the treated mice. Notably, repeated administration of NM-NPs did not induce any signs of inflammation or fibrosis upon microscopic observation of the heart, liver, spleen, lungs, kidney, and intestine (Figure 8D). The excellent biocompatibility and biosafety renders the possibility to implement NM-NPs as a delivery protein drug/vaccine carrier in vivo.

5.1 Protein preparation

A gene fragment encoding the 161-aa N-terminal sequence followed by 261-aa repetitive sequence of A. ventricosus MiSp (designated as NM) was synthesized and inserted into pET-32a plasmid within NdeI and XhoI restriction sites (pET-NM) as described before.⁵¹ For protein preparation, the cells were lysed and the pellets (NM-IBs) were collected for subsequent solubilization study (S1.1, Supporting Information). Three methods of solubilization were applied to obtain NM protein according to the previous study (S1.2, Supporting Information).^{51, 52} Supernatants (10 µL) were analyzed by SDS-PAGE to check the quality of the protein. Protein concentration was estimated by Micro BCA Protein Assay Kit (Thermo Scientific, Rockford, USA). The secondary structures of soluble proteins were measure using CD spectroscopy (S1.3, Supporting information)

5.2 Particles preparation and characterization

The NM-NPs were prepared by salting out with a potassium phosphate solution as described previously.^{15, 16, 51} The particle preparation and characterization was described in S1.4, Supporting Information.

5.3 In vitro protein loading and releasing

5.4 Mice

All experimental protocols and animal analyses were conducted according to the Guide for the Care and Use of Medical Laboratory Animals (Ministry of Health, China, 1998) and with the ethical approval the guide of Soochow University. BALB/c mice were purchased from the Experimental Animal Center of Chinese Academy of Science (Shanghai, China) and used at 6−8 weeks of age.

5.5 Tissue biodistribution in vivo

For different administration routes (oral/i.v./s.c./i.m.), mice were treated with 100 µL of PBS containing 100 µg of Cy5.5-labeled NM-NPs or 100 µg of Cy5.5-labeled free lysozyme (Lyso) or 100 µg of Cy5.5-labeled lysozyme loaded on NM-NPs (Lyso-NM-NPs), and PBS as a negative control (S1.8, Supporting Information). Detailed descriptions of different administration routes was shown in supporting information (S1.9). At the indicated time points, mice were scanned using the PerkinElmer's in vivo imaging system (IVIS Spectrum, Massachusetts, USA) at an excitation of 673 nm and emission of 692 nm. A background scan was taken with administration of PBS, and this provided the thresholding for adjusting the images collected at later time points. Then at indicated time for different administration forms, the heart, kidney, liver, lung, spleen, and small intestines were collected for imaging observation. The fluorescence intensity was quantified by using Living Image 4.0 software.

For immunofluorescence assay, the fresh isolated organs were embedded with OCT (optimal cutting temperature compound). Five-micrometer frozen sections were prepared and dropped Hoechst 33342 (Beyotime, China) onto the slide and cover the tissue adequately. The images were obtained using a confocal microscope (Nikon, Japan).

5.6 Cellular uptake and lysosomal escape of antigens in vitro

BMDCs were generated as previously described (S1.10, Supporting Information).⁵³ For cellular uptake, BMDCs were seeded in 24-well plates at a density of 1 × 10⁵ cells/well and incubated in complete medium for 24 h before use, and then the medium in each well was replaced with medium containing 10 µg/mL (final concentration) of Cy5.5-labeled lysozyme loaded on NM-NPs (Lyso-NM-NPs) or Cy5.5-labeled free lysozyme (Lyso). Cells were incubated for 2 h at 37°C and the cell pellet was washed three times with PBS. Then the cells were observed using confocal fluorescence microscope (Nikon) and analyzed using Canto II & Calibur flow cytometry (Beckman Coulter, USA).

The lysosomal escape ability of antigens was evaluated on BMDCs with a confocal laser scanning microscope (Nikon). BMDCs seeded in confocal dishes were incubated with NM-NPs containing a concentration of 10 µg/mL Cy5.5-labeled lysozyme for 1 h and 4 h at 37°C. Cells were then stained with LysoTracker Green and DAPI to visualize the lysosomes and nucleus.

5.7 Animal immunization and immune responses

As shown in Table S5 and Figure S2, male BALB/c mice aged 6−8 weeks were randomly divided into four groups and immunized with 100 µL of different vaccine formulations containing 25 µg of antigen (lysozyme). For s.c. immunization (six animals per group), mice were immunized three times at 2-week intervals and blood samples were collected for later analysis at 7 days after the third immunization. For i.m. vaccination (five animals per group), mice were immunized with different vaccines on day 0 and a booster immunization on day 60 (50 µL/hind leg). Then, serum was collected on day 14, 28, 42, 60, and 74 for measurement of antigen-specific antibody titers of IgG by ELISA (S1.11, Supporting Information). For determination of cytokine production, IFN-γ and IL-4 were measured by the ELISA kit (eBioscience, USA) according to the manufacturer's instructions (S1.12, Supporting Information).

5.8 Uptake of lysozyme and induction of DC maturation in vivo

The mice were grouped with three mice per group and s.c./i.m. given 100 µL of PBS containing 25 µg Cy5.5-labeled lysozyme in different formulations (Lyso and Lyso-NM-NPs). At 24 h after administration, mice were sacrificed by neck dislocation and isolated the draining lymph nodes. For imaging observation, the isolated lymph nodes were scanned using the in vivo imaging system (IVIS Spectrum) at an excitation of 673 nm and emission of 692 nm. For the induction of DC maturation, cells were collected from draining lymph nodes, and expression of CD80 and CD86 on CD11c⁺ DCs was determined by flow cytometry (Beckman Coulter) (S1.13, Supporting Information).

5.9 NM-NPs biosafety assay

The biosafety assay of NM-NPs including cell cytotoxicity assay (S1.14, Supporting Information), tolerance in mice (S1.15, Supporting Information), and unspecific immune responses assay (S1.16, Supporting Information).

5.10 Statistical analysis

The results presented in this study are expressed as means ± SD. Statistical analyses were conducted using GraphPad Prism 9.0 software (San Diego, CA, USA). To compare two groups, an unpaired, two-tailed Student's t-test was performed, while one-way ANOVA followed by Tukey's multiple comparison test was used for comparisons among multiple groups. Statistical significance was expressed as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ns, no significance.