Abstract
Albumin-based hydrogels have emerged as promising nanoparticle systems for the effective delivery of hydrophobic anticancer drugs. Anti-cancer drugs often cause some adverse effects, such as toxicity and rapid clearance by mononuclear phagocytic systems. Herein, a new strategy of synthesizing N-hydroxysuccinimide (NHS)-activated linker to form crosslinkable albumin-based hydrogels (CABH) is reported. The CABH favored physiochemical characteristics improvement of doxorubicin (Dox) and drug release. The CABH was constructed depending on the crosslinking reaction between NHS activated glycerol and albumin. The size of CABH was approximately 200 nm examined by dynamic light scattering (DLS) and transmission electron microscopy (TEM). It was found that the particle size and size distribution of the CABH remained stable in neutral PBS for 1 week. Dox loaded CABH would be controllably released in weak acidic environment verified by in vitro release and in vitro cell imaging. The Dox loaded hydrogel results in significant killing in the case of acidic culture medium. Our work provides a crosslinking method to formulate albumin nanoplatform and improve the size, stability, drug loading capacity and controlled release, which throws light on the potential application in drug delivery.

Introduction
The traditional cancer treatment modalities (e. g. chemotherapy, radiotherapy and surgery) achieve limited therapeutic effect due to the heterogeneity of tumor biology [1]. Moreover, a number of anti-cancer drugs, such as Dox, paclitaxel, and camptothecin, on the market are poorly water soluble, which faces major challenges in intravenous administration and bioavailability [2]. In such situation, an increase in dose and frequency are demanded to maintain the desired therapeutic effect. High doses contribute to extra toxicity and lead to poor patient compliance. However, nanomedicine has shown tremendous potential to improve the downside of traditional chemotherapy [3]. In vivo study, nanomedicine typically offers the possibility of enhanced therapeutic efficacy and reduced side effects compared to traditional chemotherapy. For example, the enhanced permeability and retention (EPR) effect has demonstrated that nano drugs are allowed to accumulate and enter into tumor tissue because of its aberrant vascularization [4]. Besides, the abnormal cancer cell proliferation results in insufficient tumor oxygen supply. Myriad evidences have supported that tumor cells gain energy through oxygen-independent glycolysis and produce a large quantity of lactates. The acidic waste tends to accumulate in tumor tissues because of theirs high metabolic activity, which contributes to enhanced acidification (pH ∼6.5–7.0) [5].Therefore, safe and efficient drug delivery systems are eagerly desirable for poorly water-soluble anti-cancer drugs and respond to a specific pH window[6].
Furthermore, we are still limited by the knowledge about the in vivo behavior of nanoparticles in human body, such as high blood circulation stability, excellent aqueous solubility, acceptable toxicity, and biodegradation [7]. Therefore, different materials are exploited as drug carriers, including inorganic materials (quantum dots [8], metal nanostructures [9], and mesoporous silica [10]), organic materials (micelles [11], liposomes, dendrimers, and hydrogels), and biomacromolecules (proteins, cellulose, polysaccharide). In particular, the approval of several commercial nano anti-cancer drugs, such as Abraxane, Daunoxome and Doxil, remain significant clinical effects, which shows great potential clinical application of nano anti-cancer drugs. Particularly, the proteins in organisms play important biological roles as transporters, enzymolysis, and nutrients. The various classes of proteins also play important roles as drug carriers in the nanomedicine. For example, bovine serum albumin (BSA), human serum albumin (HSA), ovalbumin and collagenase are often employed as carriers to load water insoluble chemical drugs, genes and immune cytokines by taking advantage of their high biocompatibility, biodegradability, non-immunogenicity. Especially, BSA comprises of three homologous domains with subdomains I, II and III with the molecular weight of 69 kDa, which provides large hydrophobic regions. There are large amounts of functional groups at the surface of BSA, such as thiol, amine, and carboxyl groups, which offer favorable conjugation sites. In addition, it has high affinity with other species via hydrophobic effect, hydrogen bonding and electrostatic interactions, which are also suitable for the loading of hydrophobic and hydrophilic drugs. The complex of Dox with BSA has gained a lot of attention in drug delivery field. For example, Bordbar reported a novel synthetic strategy for the preparation of doughnut-shaped BSA nanoparticles with narrow size distributions and homogeneity [12]. Dalvi developed albumin microbubbles that were synthesized using a new method of connecting two microfluidic T-mixers in series, which showed a complete dissolution of such bubbles and release of Dox and curcumin [13]. Jiang reported a complex of albumin and DNA synthesized via click chemistry and DNA hybridization reactions carried out to construct DNA-conjugated albumin nanocarriers for targeted therapy [14]. Cong reported a wound microenvironment-responsive hydrogel drug-loading system constructed by cross-linking of the internal electron-deficient polyester and BSA via catalyst-free amino-yne bioconjugation to release the basic fibroblast growth factor (bFGF) [15]. As compared to liposome, cellulose and mesoporous silica, BSA shows a marked increase of the drug permeability and the prolonged circulation time. Meanwhile, it has the specifications, such as non-immunogenicity, nontoxicity, and biodegradability, which make them excellent candidates for drug delivery [16]. Unfortunately, a large proportion of albumin based drug carriers have failed in the development of clinical trials. The factors of inapposite size and complicated preparation process impede their further use.
Moreover, compared with free BSA, BSA composites are more conducive as pharmaceutical excipients because of their appropriate size and larger drug loading capacity. Fatemeh et al. [17] investigated different cross-linking agents, such as tannic acid, ascorbic acid, citric acid, sorbitol, and glucose to optimize BSA nanoparticle. Yang et al. [18] developed pH-sensitive Dox-loaded BSA system via desolvation method, followed by crosslinking through Schiff base bonds. Locatelli et al. [19] extended cellulose nanocrystals containing carbamate linkers to afford releasable cellulose–Dox conjugate. The linker consisted of a spacer arm and carbamate ligation site, which was proved to be stable in presence of basic or neutral pH but not in presence of acidic pH. Goepferich et al. developed cleavable carbamate linkers for controlled protein delivery from hydrogels using aromatic succinimidyl carbonate linkers to cage the function of lysozyme [20]. Clausen et al. [21] reported a prodrug strategy based on ROS-labile 4-methylphenylboronic acid promoieties to link the drugs via carbamate linkage. Langer et al. [22,23] made a systematic study on the preparation process and enzymatic degradation for HSA nanoparticles. Hence, the development of fast, simple and stable BSA composites preparation methods are urgently expected for drug delivery in translational nanomedicines.
In this work, BSAs were crosslinked to construct CABH via activated ester chemistry with the help of crosslinker. This assemblies could endow system with hydrophilic layer in shell and hydrophobic layer in core [24,25]. As illustrated in Scheme 1, NHS activated glycerol was applied to crosslink BSAs through the abundance amino groups of BSAs. The crosslinking reaction of BSAs was available to execution in aqueous solution and avoid the use of organic solvents, which could maintain their conformational stability and biological activity. The CABH could offer controlled particle size, narrow size distribution, and structural stability. The cleavable carbamate linker was employed to control the dissolution of albumin-based hydrogels through pH trigger and the loaded Dox could be released through free diffusion and out of the dissolutive albumin-based hydrogels. Besides, the crosslinker with three NHS-activated arms offers the extended hydrophobic spaces of albumin-based hydrogels, which increases the encapsulation rate of such poorly water-soluble drugs. The therapeutic effect of Dox-loaded CABH against tumor cells was verified using a real-time method. Finally, Dox loaded CABH would be controllably released in weak acidic environment and resulted in significant killing in the case of acidic culture medium. The Dox loaded CABH were anticipated to be a promising anti-tumor drug carrier. Schematic illustration of CABH formulation and drug release.
Materials and methods
Chemical and reagents
Glycerol, triphosgene, NHS, activated charocal, triethylamine, tetrahydrofuran (THF), and Dox were purchased from Energy Chemical (Shanghai, China). Triethylamine and THF were dried with sodium and CaH2, respectively. Ethyl ether, dichloromethane (DCM), methanol, Na2HPO4, KH2PO4 and NaCl were obtained from Greagent (Shanghai, China). BSA was purchased from Shanghai yuanye Bio-Technology Co., Ltd (Shanghai, China)
Characterization
1H NMR spectra were recorded using Bruker AV 500 spectrometer. TEM samples were prepared on 300 mesh ultra-thin carbon film supported copper grids (Zhongjingkeyi Technology Co., Ltd.), and were imaged using Talos F200i (ThermoFisher Scientific, Talos F200i TEM). Fluorescent analyses were performed using fluorometer (Thermo Scientific Fluoroskan Ascent FL). Fluorescent images were captured using Operetta High-Content Imaging System (PerkinElmer).
Synthesis of NHS activated glycerol crosslinker
In a 50 mL round-bottom flask, triphosgene (3.25 g, 11 mmol) was dissolved in dry ether (10 mL) and activated charocal (0.075 mg) was added. The solution was stirred for 1 h at room temperature. Then the solution was cooled to 0°C, and glycerol (0.5 g, 5.4 mmol) in dry THF (5 mL) was added dropwise. The resultant was stirred for 12 h. The residual solvent was filtered and concentrated under vacuum. Then, in a 50 mL round bottom flask, NHS (1.875 g, 16.3 mmol) was dissolved in dry THF (10 mL) under N2 protection. The abovementioned concentrated solution was dissolved in dry THF (5 mL) and was injected into a round bottom flask. Then, dry triethylamine (1.35 mL, 16.5 mmol) was injected into the round bottom flask gradually. The reaction was stirred at 40°C for 12 h. The mixture was filtered to remove precipitates. The residual solvent was concentrated under vacuum. The crude product was purified by silica gel column chromatography using DCM-methanol mixtures of appropriate composition for elution (3.79 g, 70% yield). 1HNMR (CDCl3, 500 MHz):.2.76 (s, 8 H), 2.86 (s, 4 H), 4.36 (m, 2 H), 4.61 (m, 2 H), 5.01 (m, 1 H).
Synthesis of the CABH
The prepared crosslinkers were dissolved in PBS and BSAs were then added to this solution with a crosslinker/BSA ratio of 2:1. The mixture was stirred at room temperature for 12 h. The resultant BSA composites were washed with PBS in an Amicon centrifugal filter (molecular weight cut off = 100 kDa; millipore, USA) to remove unconjugated crosslinker and free BSAs.
Preparation of Dox-loaded CABH and Dox loading capacity measurement of CABH
Typically, the CABH (20 mg) were dissolved in water (1 mL). Then Dox (6.5 mg) was added and the mixtures were stirred at room temperature for 48 h to encapsulate Dox. Dox-loaded albumin-based hydrogels were isolated by centrifugation with Amicon filter (molecular weight cut off = 100 kDa; millipore, USA) and followed by washing with water (2 mL) for three times to remove unloaded Dox molecule. In order to calculate the quantity of Dox loaded into CABH, Dox-loaded albumin-based hydrogels (20 mg) were stirred vigorously in acidic PBS to decompose CABH and extract Dox. The solution was filtered by microfiltration membrane and added into a 96-well opaque plate. The solution was analyzed by fluorometer (Thermo Scientific Fluoroskan Ascent FL) using the 485/510 excitation/emission filter pair. All experiments were carried out in triplicate and each value was reported representing the mean, ±SD The drug loading capacity of Dox loaded CABH was calculated through the following formula: Drug loading capacity (%) = (amount of Dox loaded to CABH/total amount of the Dox loaded CABH) × 100.
The hydrodynamic size and ζ-potential of the CABH
The hydrodynamic diameter and surface charge of CABH were measured using a Malvern ZEN 3600 Instruments equipped with ζ-potential analyzer. The samples (10 mg of free BSA, CABH and Dox loaded CABH, respectively) were diluted in PBS to a total volume of 15 mL for the hydrodynamic size and ζ-potential analysis.
TEM analysis of the CABH
The morphology of CABH was analyzed using TEM. The TEM samples were prepared by way of the freeze-drying method. A diluted aqueous solution of CABH (1 mg/mL) was deposited on super-thin carbon film and the solution was frozen immediately. Then, the frozen sample was dried by lyophilization for TEM measurement.
In vitro release of Dox loaded CABH
Dox loaded CABH were released in different pH PBS solutions (pH = 5.6, 6.5, 7.4 and 9) to simulate the drug release in acidic, neutral and basic conditions. Typically, Dox loaded CABH (10 mg) were dissolved in PBS solution (0.2 mL) and placed into Amicon inner tube. Then, the inner tube was placed into the Amicon outer tube with PBS solution (1 mL) and shocked by a shaker. At defined times, 100 μL of solution was withdrawn and placed into 96-well opaque plates. The Dox concentrations were detected by fluorometer. After detection, the 100 μL of solution was replaced into original Amicon inner tube to continue release process.
Cell viability evaluation of CABH and Dox loaded CABH
Cell viability was evaluated by Cell Counting Kit-8 (CCK-8) (Dojindo, Japan). Skov 3 cells were seeded onto 96-well plates and cultured in RPMI 1640 medium with 10% fetal bovine serum in a 5% CO2 environment for 12 h. Then, the cells were treated with CABH (10 μg/mL, 20 μg/mL, 50 μg/mL, 100 μg/mL, 200 μg/mL and 500 μg/mL of crude CABH) for biocompatibility evaluation. Skov 3 cells were seeded onto 96-well plates and cultured in RPMI 1640 medium with 10% fetal bovine serum in a 5% CO2 environment for 12 h. The culture medium was removed from each well and washed with PBS buffer three times. Then, 100 μL of different pH PBS (pH = 6.5, 7.4, and 9) containing Dox loaded CABH (Dox-equivalent dose of 0.114 μg/mL, 0.228 μg/mL, 0.342 μg/mL, 0.57 μg/mL, 0.798 μg/mL, 1.026 μg/mL, 1.254 μg/mL, 1.482 μg/mL, 1.71 μg/mL and 1.938 μg/mL) were added to the cell culture media respectively for cytotoxicity assessment. Free Dox was used as control group. The absorbance was determined by using a TECAN infinite M200 plate reader (Tecan, Mechelen, Belgium) at 450 nm.
In vitro cell imaging studies
Skov 3 cells were performed to investigate the potential release capability of CABH. The cells were seeded onto six-well plates and cultured in RPMI 1640 medium with 10% fetal bovine serum in a 5% CO2 environment. After 12 h, the culture medium was removed from each well and washed with PBS buffer three times. Then, 2 mL of different pH PBS (pH = 6.5, 7.4, and 9) containing Dox loaded CABH (Dox-equivalent dose of 1.14 μg/mL) were added to the cell culture media respectively and incubated for 1 and 3 h at 37°C in a 5% CO2 environment respectively. After 1 and 3 h, the corresponding culture media were removed from each well and washed with PBS buffer three times. The excitation and emission wavelengths of Dox were 480 nm and 593 nm, respectively. The nuclei were stained with DAPI (Ex: 340 nm, Em: 488 nm). The typical bright field and fifluorescence images were obtained using an operetta high-content imaging system (Olympus, Bh2-RFCA, Japan).
Results and discussion
The nontoxic glycerol containing NHS crosslinker was synthesized by a simple and efficient carbonyl chloride reaction. 1H NMR analysis indicated that the integral of the peak at 2.76 and 2.86 ppm corresponded to three NHS groups, as seen in Figure 1. Note that NHS groups were grafted to glycerol with carbamate linkage successfully. In addition, the aqueous phase reaction between crosslinker and BSAs could lead to the formation of CABH. The NHS contained glycerol crosslinker was also designed to be cleaved in the presence of acidic conditions, followed by release of the loaded cargo. 1HNMR spectra of NHS contained glycerol crosslinker.
The fluorescence change of Dox was monitored during CABH encapsulation process. When mixing different concentrations of free Dox (5.5, 6.5, 7.5 mg/mL) with CABH (20 mg), there was significant reduction in the fluorescence signal, as shown in Figure 2. Typically, the CABH (20 mg) were dissolved in water (1 mL) and divided into three parts. Then Dox (5.5, 6.5 and 7.5 mg) was added to each part and the mixtures were stirred at room temperature for 48 h. Throughout this period, each part was withdrawn and placed into 96-well opaque plates to measure Dox concentration. The high concentration of Dox could lead to more striking fluorescence quenching of Dox during encapsulation process. Therefore, this sort of encapsulation brings about a highly quenched state of the Dox resulting from the close vicinity of the Dox molecules to the hydrophobic region of CABH and the interaction of the Dox with CABH. In addition, the fluorescence intensity generates barely noticeable gradient after 20 h, suggesting that the loading capacity of CABH was already saturated. The final Dox loading capacities were 2.56 mg/20 mg, 2.5 mg/20 mg and 2.66 mg/20 mg CABH based on the standard curve of Dox and the average Dox loading capacity was 2.57 mg Dox per 20 mg CABH. The standard curve presented in Figure 2(b) was calculated to reflect the relationship between the Dox concentration and the fluorescence intensity. Based on the regression equation (y = 76.945x + 0.01 (R2 = 0.9964)), the loading capacity of Dox in CABH was calculated as 11.4 wt%. The drug loading capacity of CABH was similar to the drug loading capacity of BSA based drug carrier and lower than the drug loading capacity of other types of drug carriers (Table 1). This was ascribed to the limited number of loading sites in the CABH. Therefore, the crosslinking was an effective approach to expand hydrophobic space of BSAs and the extra hydrophilic character of CABH could contribute to increasing solubility of Dox. The fluorescence changes of different concentrations of Dox (5.5, 6.5 and 7.5 mg/mL) in the presence of CABH (20 mg) and the standard curve of Dox. The volume of each group was 1 mL. (Ex = 544 nm, Em = 576 nm). Each bar represents the mean ± SD (n = 3). Comparison of drug loading capacity of CABH with other drug carriers.
Particle size, polydispersity index, zeta potential of free BSA and CABH loaded and unloaded with Dox (data represent mean ± SD, n = 3).

The hydrodynamic diameter of free BSAs (a), CABH (b), Dox loaded CABH (c), CABH stored for a week (d) and CABH incubated with acidic solution (e).

Typical TEM of free BSA (a), CABH (b) and Dox loaded CABH (c). Images collected at 200 kV. Scale bar is 50 nm.
The morphology of CABH was also supported by TEM. But the sizes of CABH measured by TEM were correspondingly lower than those obtained by DLS measurements, probably due to the dehydrated state of CABH. Figure 4 showed images of free BSA, the CABH and Dox loaded CABH, which showed the representative process of CABH from crosslinking to drug loading. Figure 4(a) exhibited the formulation of spherical, homogeneous and well-dispersed morphology for free BSA. Figure 4(b) and (c) showed the formulation of a slightly distorted spherical and blurred shadows morphology for CABH and Dox loaded CABH. In contrast to free BSAs, several BSA molecules were gathered together to form crosslinkable structure, as shown in Figure 4(b). But no obvious aggregates were observed, which indicated that CABH were stable in aqueous solution and could be well dispersed as individuals. TEM results indicated that the average size of the CABH was 100–200 nm, which was proportional to the number of the crosslinked BSA molecules. The size of CABH was also somewhat similar to other BSA-based nanoparticles [18]. The particle size was efficiently controlled by crosslinking between BSAs and crosslinker. With the crosslinking of BSA molecules, more hydrophobic areas are provided, which is favorable for the loading of Dox, as shown in Figure 4(c).
The pH induced release of Dox from CABH was examined in different pH PBS solutions. Figure 5 showed the fluorescence intensities of the released Dox at different time-intervals. An interesting distinction can be made between the drug release profiles. The different drug release behaviors were actuated by the pH gradient. Evidently, the fluorescence intensities of the released Dox in pH = 4.5 was the highest, implying that the structure of CABH was unstable in acid medium and resulted in the release of the loaded Dox over time. Consistent with the release behavior of Dox loaded CABH in pH = 4.5, the released Dox in pH = 5.6 and 6.5 also increases. Thus, the release rate progressively increased at the lower pH level. But the fluorescence intensities of the released Dox in pH 7.4 and 9 did not obviously change and reached a saturation value upon further prolongation of the release process. Roughly half of the drugs still remained into the samples after release for 50 h. This may also be associated with hydrophobic segments of CABH that contributes to limiting the diffusion of encapsulated drugs. However, Dox loaded CABH incubated in 4.5 pH led to a continuous increase in Dox release over 50 h. The construct of CABH could be rapidly disassembled at lower pH. Compared with other pH-sensitive BSA nanoparticles, CABH exhibited significant drug release behavior due to the cleavage of carbamate bonds under acidic condition [18]. These results suggested that CABH could release the cargoes under acid conditions and remain relatively stable in neutral and alkaline conditions. Probing the pH triggered release of Dox from CABH by assaying the release responsiveness of CABH in different pH media. p ** for p < 0.01, *** for p < 0.001, p ns for not significant, as compared with pH 7.4 group. Each bar represents the mean ± SD (n = 3).
To understand the biocompatibility of these CABH, the cell viability assay was done with different concentrations on Skov 3. Nearly 90% of cells were survival even when the concentration of CABH was 500 μg/mL, as shown in Figure 6(a). The result suggested that there was no toxicity of the CABH on cells. However, in the case of Dox loaded CABH, the cytotoxicity of CABH was pH-dependent (Figure 6(b)). The half maximal inhibitory concentration values (IC 50) of Dox loaded CABH were 0.45 μg/mL, 0.92 μg/mL and 1.48 μg/mL in acidic, neutral and alkaline culture media, respectively. The IC 50 value of Dox loaded CABH in acidic medium was 2 times lower than that in neutral medium and three times lower than that in alkaline medium, owing to a greater release of Dox in the acidic culture medium. Free Dox showed strong cytotoxicity and an indifferent killing effect in three pH conditions with IC 50 of approximately 0.18 μg/mL (Figure 6(c)). This supports that the CABH can be employed as drug vehicle for controlled and sustained drug delivery. Biocompatibility of CABH (a), cell viability of Skov 3 after treatment with Dox loaded CABH or free Dox in acidic, neutral and alkaline culture media (pp *** for p < 0.001, as compared with Dox loaded CABH in acidic medium) (b) and the IC 50 of Skov 3 after treatment with Dox loaded CABH or free Dox in acidic, neutral and alkaline culture media (pp *** for p < 0.001, as compared with Dox loaded CABH in acidic medium) (c). Each bar represents the mean ± SD (n = 6).
To enable better understanding of drug release behavior, the in vitro cellular uptake and release of Dox loaded CABH was examined by operetta high-content imaging system. Skov 3 cells were treated with different pH PBS solutions to measure the pH-responsive ability of CABH, as shown in Figure 7. The acidic, neutral and alkaline cellular environments were regulated by acidic, neutral and alkaline PBS. Compared to the neutral and alkaline groups, the red fluorescence intensities of acidic group were enhanced when the cells incubated with Dox loaded CABH for 1 h. As time went on, the red fluorescence intensities of acidic group were more distinct than that of the neutral and alkaline group at 3 h. Cell imaging confirmed that Dox were mainly located in the nucleus. While, a high concentration release of Dox was observed as determined in control group using free Dox within the same timeframe, implying free Dox was the absence of triggered release ability and possessed better cell membrane penetration behaviors. Thus, it indicated the Dox anchored on the CBAH was more stable than free Dox and could avoid bulk release and unwanted side effects. This positive data proved that CABH possessed excellent pH-responsive property and could be safely employed as a drug delivery system with increased anticancer efficacy. The bright field and fluorescence images of Skov3 cells after incubation with Dox loaded CABH in acidic, neutral and alkaline environment for 1 and 3 h (a) and free Dox in acidic, neutral and alkaline environment for 1 and 3 h (b). Scale bar: 50 μm. Note: Nuclei were stained by DAPI. Dox was red (Ex: 480 nm, Em: 593 nm) and DAPI was blue (Ex: 340 nm, Em: 488 nm).
Conclusions
Overall, we developed a new strategy to effectively improve the size, stability, controlled release of BSA nanoplatforms by chemical corsslinking. A traceless linking strategy and mild reaction condition were employed to obtain a safe and effective approach for albumin-based hydrogels composite. The NHS activated glycerol linker was utilized as bridge for the albumin-based hydrogels. The CABH could provide hydrophobic area to load hydrophobic anti-cancer drugs. The association and dissociation of the CABH were in response to pH change and controllable in drug release. The degraded albumin-based hydrogels are non-cytotoxic and structural plasticity, which meets more complex requirements of drug delivery application.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Scientific Research Fund of Hunan Provincial Education Department (Grant No. 19C0774), Natural Science Foundation of Hunan Province, China (Grant No. 2020JJ5156) and Student Research and Innovation Program of Hunan University of Science and Technology (Grant No. YZ2143 and Grant No. ZZ2109).
