In Vitro Protein Stability.

Preparation of therapeutic proteins is often prone to instability during storage and use, and enhanced structural stability is a desirable attribute of modification technologies. Thermal stress tests at elevated temperatures are a facile approach to evaluate differences in protein stabilities in a relatively short time (25–27). A protein conjugate or nanocapsule stable at higher temperatures is likely to be stable at ambient temperatures for a longer period. To test the stability of modified proteins to environmental stressors, their enzymatic activities were assayed after incubation at 65 °C for different time periods and presented as the percentage of preheating activity retained. As shown in Fig. 1E, uricase-PCBX has exceptional thermal stability and retained 100% activity after 2 h of incubation. In contrast, uricase-PEG only exhibited a modest stabilizing effect.

The present result, that uricase-PCBX remains stable at 65 °C, suggests that it would not require low-temperature storage or formulation with stabilizing excipients as is the case for the native protein. This high stability is important when considering situations in which low-temperature storage or transportation is unavailable, and may help reduce the high cost of therapeutic protein production and use. In a previous study, PCB was shown to impart a superior protein-stabilizing effect and increase the substrate affinity to protein binding sites when similar protein conjugation structures were tested (26). The CB monomer unit is a derivative of glycine betaine, a natural molecule found in many living systems as a stabilizing agent. Encapsulating a protein in a PCB gel network is analogous to placing it into a highly concentrated protein stabilizing solution, equipping it with a protective microenvironment that accompanies the protein both in vitro and in vivo. Furthermore, the nanocapsule structure has been reported to stabilize enzymes through multiple covalent attachments between the enzyme core and polymer shell, which effectively hinder conformational changes in adverse environments (27, 28). Combining these two distinct mechanisms, it is unsurprising that PCB gel encapsulation affords superior stability to protein cargos.

PK Study.

Since the 1970s, PEGylation has been the gold standard method to alter PK properties of both proteins and a variety of nanoparticle platforms for drug delivery (3, 29). Surface attachment of PEG significantly prolongs in vivo circulation times by increasing hydrodynamic size to avoid rapid renal clearance (for small particles) and reducing interactions with both blood components (opsonization) and immune cells. The PK profile of ZPNE-modified uricase was compared with this gold standard in healthy Sprague–Dawley rats through repeated i.v. injections. Fig. 2 shows the circulation profiles of native and modified uricase after each injection. All concentration-time curves fit a two-compartment model. PK parameters are listed in Table 1. On the initial injection, uricase-PEG exhibited an elimination half-life (t_1/2β) of 38.2 h, threefold higher than the 13.9 h t_1/2β of native uricase. The area under curve (AUC_∞) of PEGylated uricase is five times that of the native enzyme, roughly in agreement with values reported in the literature (30). In comparison, uricase-PCBX showed exceptionally prolonged circulation behavior, with a t_1/2β of 134.9 h, a 3.6-fold improvement over uricase-PEG, and 10-fold higher than the native enzyme. Moreover, the AUC_∞ of uricase-PCBX was 4.2 times that of uricase-PEG, indicating its significantly higher systemic availability.

Open in a separate window

Fig. 2.

Blood circulation profiles of modified uricase vs. native uricase. Circulation curves after the (A) first i.v. administration, (B) second administration, and (C) third administration.

Uricase is a large tetrameric protein with a molecular weight of ~130 kDa. The rapid clearance of native uricase is considered to be associated with an irreversible uptake process, likely by the reticuloendothelial system (RES), rather than renal filtration (31). Conjugating a total of 20 large PEG molecules (each with a molecular weight of 10 kDa) to uricase can effectively cover a large portion of its surface, minimizing nonspecific uptake clearance. This phenomenon is closely related to material hydration and nonfouling character. As a well-known class of ultra-low fouling materials, zwitterionic polymers have indeed been used to extend the circulation lifetime of nanoparticles (32–34). Apart from material properties, recent research has shown surface coverage density to be another important factor affecting particle uptake by immune cells and in vivo circulation behavior (35). A high surface grafting density of PEG molecules, exceeding the minimum for a brush conformation, is required for nanoparticles to evade immune cell uptake and clearance. However, in the case of protein modification, there is a certain limit to conjugation density due to the restricted number of available functional groups on the surface of a protein. A cross-linked hydrogel “mesh,” which does not require a high number of conjugation sites, can effectively overcome this problem. A thin hydrogel layer can ideally mask the entire protein surface, making it “stealth” to surroundings and slowing down RES clearance. The results of this study confirm our hypothesis that greatly extended circulation profiles of protein therapeutics can be realized by PCB nanocapsule encapsulation. Patients could benefit greatly from this advance by enjoying a reduced administration frequency compared with PEGylated products.

Uricase is administered chronically in clinical practice to treat refractory chronic gout, so repeated injections were performed to evaluate the long-term performance of ZPNE-modified and PEGylated proteins. The second and third injections were given at days 7 and 14 after the initial injection. It should be noted that severe infusion reactions were observed 3–5 min after the second and third injections of both native uricase and uricase-PEG, with symptoms including hypopiesia and dyspnea. In contrast, no adverse effects were observed following the uricase-PCBX injections. Although the calculated t_1/2β of native uricase appears comparable after each injection, the sampling frequency was likely inadequate for its rapid elimination; its accelerated elimination after repeated injections is more clearly reflected in the significantly reduced AUC_∞. Following the second and third injections of uricase-PEG, an obvious accelerated blood clearance (ABC) phenomenon was noted, with t_1/2β decreasing to 30.6 and 24.8 h, and AUC_∞ values dropping 2.8-fold and 5.9-fold, respectively. This accelerated elimination process is most likely due to the induction of anti-PEG antibodies, which will be discussed in the following section. Unlike PEGylated uricase, the PK profiles of uricase-PCBX in this study did not show any significant change following three injections, indicating that uricase-PCBX is likely to maintain its superior PK performance in repeated applications.

Detection of Antibodies.

Serum collected at day 35 after the first injection was tested for both anti-uricase and anti-polymer antibodies using direct ELISA. As shown in Fig. 3 A and B, IgG was the major isotype of anti-uricase observed after repeated injections. The measured antibody titers are listed in Table S2. Due to the strong immunogenicity of the fungal enzyme, the native uricase group developed extremely high IgG antibody titers along with relatively high titers of IgM. In the group treated with PEGylated uricase, the titers of anti-protein antibodies were significantly lower than the native enzyme but still clearly elevated. In contrast, neither anti-uricase IgG nor IgM were detected in the rat group treated with uricase-PCBX. Our results here are consistent with clinical findings, showing that PEGylation cannot completely prevent the generation of anti-protein antibodies, at least for highly immunogenic uricase (9). This finding supports our hypothesis that comprehensive surface coverage of a protein with a stealth polymer network can effectively render the protein “invisible” to the immune system and thus mitigate antibody induction.

Open in a separate window

Fig. 3.

Detection of antibodies by direct ELISAs. (A) Anti-uricase IgM, (B) anti-uricase IgG, (C) anti-polymer IgM, and (D) anti-polymer IgG. Serum from each rat is plotted separately.

The severe infusion reactions and accelerated drug elimination seen in the native uricase group is apparently related to strong production of anti-uricase antibodies. However, for PEGylated protein therapeutics, clinical trial evidence suggests anti-PEG antibodies rather than anti-protein antibodies are primarily responsible for infusion reactions and efficacy loss over chronic injections (8, 9). Accordingly, strategies to suppress the production of anti-polymer antibodies remain a critical challenge in the development of polymer-protein conjugates. Uricase-PEG and uricase-PCBX groups were tested for anti-PEG and anti-PCB anti-polymer antibodies, respectively, and the results are shown in Fig. 3 C and D. Notably, at day 35, the major anti-PEG response comes from IgM (titers 1:6,400) rather than IgG (titers 1:800), suggesting differences in adaptive immunity pathways for the protein and polymer components in the conjugates. Compared with the strong response to PEGylated uricase, no IgG or IgM were detected against the polymeric component of uricase-PCBX.

As repeated injections are commonly needed for therapeutic proteins, the antibody response issue should not be overlooked and merits careful evaluation. It should be noted that in the earlier work establishing PEGylation as a method to alter protein immunogenicity (6), neither anti-BSA nor anti-PEG antibodies were detected after injecting BSA-PEG conjugates into a rabbit model. Three reasons may account for their observations. First, BSA shares a large degree of similarity with rabbit serum albumin (RSA) (36), which makes it less immunogenic as a model protein. In a later study, the generation of antibodies against PEGylated proteins was shown to be strongly related to the protein immunogenicity (7). This result explains why PEGylated nonhuman enzymes, e.g., uricase and asparaginase, have been reported to present the most severe issues in clinical trials resulting from anti-PEG antibodies (8–10). Second, a high degree of modification can be easily attained with BSA. The relationship between modification degree and immunogenicity of conjugates was also clearly demonstrated in the later study (7). Serum albumin is a class of protein with extraordinary ligand binding capacity (36), but this is not the case for most therapeutic proteins. Finally, immunological assays of the time, e.g., immunodiffusion and complement fixation tests, were not capable of detecting low-titer antibodies.

Because there is no standard method to evaluate anti-polymer antibodies, and especially as no positive control is available for the detection of anti-PCB antibodies, we validated our results using a surface plasmonic resonance (SPR) sensor. SPR is an ultra-sensitive technique used to study specific/nonspecific protein adsorption on surfaces with nanogram-level detection limits (21, 37). Compared with ELISA, SPR analysis is more straightforward, as the procedures are simpler, and each step can be carefully controlled and monitored. To detect anti-PEG or anti-PCB antibodies, we modified the surface of gold chips with poly(ethylene glycol) methyl ether methacrylate (PEGMA) or PCB polymer brush layers. Both PEGMA and PCB are nonfouling materials that resist nonspecific protein adsorption from diluted serum. When serum samples flow over the modified chip surface, only the polymer-specific antibodies can bind to the surface and generate an SPR signal through specific adsorption (Fig. 4A). The results are shown in Fig. 4. Sera obtained preinjections were used as a negative control. When sera from uricase-PEG–treated rats were flowed over a PEGMA modified chip, adsorption curves showed typical specific binding kinetics with adsorbed mass increasing linearly with time. Protein binding reached 40–50 ng/cm² in 15 min under experimental conditions. No nonspecific binding was detected on the same chip from either control serum or uricase-PCBX serum, indicating all adsorption came from anti-PEG antibodies. In contrast, none of the sera showed detectable adsorption on a PCB-coated chip surface, suggesting that there is no antibody in the sera that can recognize and bind to PCB polymers.

Open in a separate window

Fig. 4.

Detection of antibodies by an SPR sensor. (A) Scheme showing the detection setup. The gold chip surface was modified with either PEGMA or PCB polymer brushes. The polymer brush serves dual functions, providing a nonfouling background while serving as the investigated antigen. (B) Summary of the SPR detection results. (C) Sensor signal-time curves on PEGMA-coated chips, each curve representing one serum sample. No uricase-PCBX serum showed any adsorption onto a PEGMA surface, though only one typical curve is shown. (D) Sensor signal-time curves on PCB-coated chips.

PK and Biodistribution Study.

The PK of native and modified uricase is studied using Spraque–Dawley rats (male, body weight 74–100 g) as the animal model. Each sample has three duplicates to generate statistical significance. All animal experiments adhered to federal guidelines and were approved by the University of Washington Institutional Animal Care and Use Committee. For PK studies, each protein sample was administered into the rat via tail vein injection at the dose of 25 U/kg body weight. Blood samples were collected from the tail vein at 5 min, 6 h, 24 h, 48 h, and 72 h after the injection. The blood samples were put in heparinized vials and centrifuged, and the enzyme content in plasma was estimated by enzyme activity. The i.v. injections and bleeding procedure were repeated three times with 1 wk as time interval between each injection. Five weeks after the first injection, 5 mL blood was drawn by using cardiac punch, and serum was prepared for antibody detections. All PK parameters were calculated using PKSolver following the instructions.

For the biodistribution study, all rats were given unlabeled samples for the first and second dose, followed by FITC-labeled samples for the third dose. At 72 h after injection, all rats were killed, and the heart, liver, spleen, lung, kidney, and blood were collected for further analysis. The collected organ tissues were homogenized using a tissue ruptor, followed by centrifugation at 3,200 × g for 30 min at room temperature. The fluorescence of particles in the tissues was measured using a microplate reader and compared with a standard curve generated using FITC-labeled samples added to untreated homogenized tissues.

Synthesis of CBAAX.

A solution of acryloyl chloride (5.3 mL, 65.1 mmol) in 20 mL dichloromethane (DCM) was added dropwise to a solution of N-methyl-2,2′-diaminodiethylamine (4 mL, 31.0 mmol) and N,N-Diisopropylethylamine (DIPEA) (11.9 mL, 68.3 mmol) in 50 mL DCM at 0 °C for a period of 30 min. The reaction mixture was allowed to warm to room temperature and stirred for 2 h. The reaction mixture was then washed with (2 × 25 mL) water. The aqueous layer was re-extracted with DCM (3 × 75 mL). The combined organic layers were dried using sodium sulfate, filtered, and concentrated in vacuo to leave a residue that was further purified by column chromatography to give compound 2 (5.2 g) in 75.3% yield. ¹H NMR (300 MHz, D₂O) δ 6.02–5.97 (m, 4H), 5.56–5.51 (m, 2H), 3.14 (t, J = 6.7 Hz, 4H), 2.37 (t, J = 6.7 Hz, 4H), 2.07 (s, 3H).

Compound 2 (3.2 g, 10.5 mmol) was dissolved in 20 mL acetonitrile, and tert-butyl bromoacetate (4.2 mL, 31.6 mmol) was added to it. The reaction contents were stirred for 18 h at 65 °C until the TLC showed complete consumption of starting material. The reaction mixture was concentrated to dryness in vacuo and further purified by column chromatography to give compound 3 (4.3 g) in 72.6% yield. ¹H NMR (300 MHz, DMSO-d₆) δ 6.31–6.08 (m, 4H), 5.67 (dd, J₁ = J₂ = 2.4 Hz, 2H), 4.48 (s, 2H), 3.75–3.54 (m, 8H), 3.29 (s, 3H), 1.45 (s, 9H).

Compound 3 (2.0 g, 4.7 mmol) was dissolved in 15 mL DCM, and 15 mL TFA was added to it. The reaction contents were stirred overnight at room temperature. After complete hydrolysis, the reaction mixture was concentrated in vacuo and coevaporated with DCM (3 × 15 mL). The resulting residual mixture was further dissolved in 15 mL methanol, and 4 g IRN-78 resin was added to it for complete neutralization. The reaction mixture was then filtered over celite and concentrated to dryness in vacuo. The residue was dissolved in water and lyophilized on a freeze dryer to give compound 4 in 84.1% yield. ¹H NMR (300 MHz, DMSO-d₆) δ 6.31–6.05 (m, 4H), 5.63 (dd, J₁ = J₂ = 2.4 Hz, 2H), 3.70 (s, 2H), 3.69–3.50 (m, 8H), 3.21 (s, 3H).

Measurement of Thermal Stability.

Modified protein samples (protein concentration 40 μg/mL in 0.1 M Tris buffer, pH 8.0) were incubated at 65 °C in water bath. At different time points, each sample was taken out and quenched in ice bath until measurement. After the final samples were taken out at 2 h, activities of all samples were measured using the uricase activity kit following the manufacturer’s protocol. Values were recorded as percentage of each sample activity before heating.

ELISA.

The antigens used in direct ELISAs consisted of native uricase (for detection of anti-uricase antibody), BSA-PEG conjugates (for detection of anti-PEG antibody), and BSA-PCB conjugates (for detection of anti-PCB antibody). BSA-PEG conjugates were made following the same procedure as uricase-PEG samples. BSA-PCB conjugates were made following similar procedure as making uricase-PCBX but with no cross-linkers during polymerization. For ELISA experiments, 100 μL antigen solution (10 μg/mL of protein concentration) prepared in 0.1 M sodium carbonate buffer, pH 10.5, was used to coat each well of the 96-well plates. During coating procedure, plates were incubated at 4 °C overnight. After removing antigen solutions, the plates were washed five times using PBS (PBS 7.4) and then filled with blocking buffer (1% BSA solution in 0.1 M Tris buffer, pH 8.0). It is important to avoid using any buffer that contains PEG-like detergents, e.g., Tween 20 and Tween 80. After incubation at room temperature for 1 h, blocking buffer was removed, and all wells were washed by PBS for another five times. Serial dilutions of rat sera in PBS containing 1% BSA were added to the plates (100 μL/well), which were incubated for 1 h at 37 °C. The plates were then washed five times with PBS. Goat anti-rat IgM or IgG conjugated to HRP (Bethyl Laboratories) was used as the secondary antibody for detection of IgM and IgG. After adding the secondary antibody, plates were incubated at room temperature for 1 h and then washed five times using PBS before the addition of 100 μL/well HRP substrate 3,3′,5,5′-tetramethylbenzidine (TMB; Bethyl Laboratories). The plates were shaken for 15 min, and 100 μL stop solution (0.2 M H₂SO₄) was added to each well. Absorbance at 450 (signal) and 570 nm (background) was recorded by a microplate reader. Prebleeding sera were used as negative control for all ELISA detections. Commercially available rat anti-PEG antibodies were used as positive control for anti-PEG detections. The positive signal was defined as absorbance significantly larger than corresponding negative control.

This work was supported by the Defense Threat Reduction Agency (HDTRA1-13-1-0044) and National Institutes of Health (1R21EB020781-01).