Hydroxyethyl Starch-Based Functionalization of Gold Nanorods: A Possible Alternative to Polyethylene Glycol as a Surface Modifier

National Engineering Research Center for Nanomedicine, College of Life Science and Technology, Huazhong University of Science and Technology, Wuhan 430074, China Protein & Peptide Pharmaceutical Laboratory, Institute of Biophysics, Chinese Academy of Sciences, Beijing 100101, China University of Chinese Academy of Sciences, Beijing 100049, China Department of Microbiology, Birendra Multiple Campus, Tribhuvan University, Bharatpur, Chitwan 44200, Nepal State Key Laboratory of Environmental Aquatic Chemistry, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China


Introduction
Surface modifiers are extensively added to functionalize desired nanoparticles in order to achieve diagnostic and targeted therapeutic goals in the nanomedicine field [1]. Most of the common surface modifiers are hydrophilic polymers that inhibit the aggregation between nanoparticles in an aqueous solvent [2]. The functionalization of nanoparticles using surfactant polymers began in the 1980s, where surfactant micelles were used and exhibited higher targeting efficiency for neuroleptic action [3]. The appropriate selection of surface modifiers is essential to functionalize the surface of nanoparticles to lower the nontargeted distribution into healthy cells [4]. Polyethylene glycol (PEG), one of the common and traditional polymeric compounds [5] and polysaccharides [6], has gained remarkable attention recently.
PEG is widely used to modify nanoparticles by the process called PEGylation, which involves the functioning of nanoparticles with PEG to prevent nanoparticles from being engulfed by phagocytes and to improve their stability as well as plasma life [5,7]. High stability has made PEG more inclined to targeted delivery as a surface modifier of nanoparticles, which is a "gold standard" approach to alleviate the cytotoxicity in anticancer therapy [8]. PEGylation has been reported to offer the nanoparticles an extended circulation time in vivo and an improved solubility rate to hydrophobic drugs with low nonspecific cytotoxicity [9,10].
Despite PEG's various advantages, its practical application is hindered by several limitations. First of all, PEG causes hypersensitivity reactions in the body [11]. A plethora of clinical as well as animal studies reported that a high amount of anti-PEG antibodies were produced after repetitive parenteral dose, which resulted in a comparatively shorter plasma life in the microvascular system compromising the functions of PEG [12][13][14][15]. Moreover, the nondegradability of PEG in vivo is another issue that leads to severe toxicity due to assemblage inside the body, if given in maximum dose [16]. Likewise, PEG-based chemotherapeutics lose stability upon rehydration due to PEG crystallization upon lyophilization [17]. Furthermore, the structure of PEG has been shown to contain relatively less reactive sites, which do not allow any ligands or conjugates to get attached to the PEG surface in a convenient manner [18]. These limitations of PEG have motivated the search for efficient and reliable methods to functionalize nanoparticles, such as by using varieties of polysaccharides [6,19,20]. Polysaccharides and PEG bear similar characteristics of enhancing plasma life in microvascular circulation as well as resisting the adsorption of proteins [18]. However, polysaccharides offer better biocompatibility and biodegradability compared to PEG along with their significant virtues, such as low cost and low cytotoxicity [6,20].
Hydroxyethyl starch (HES), a polysaccharide, has been proven to possess superior performances as a nanoparticle modifier in a targeted drug delivery system compared to other nondegradable biopolymers such as PEG [18,21]. The application of HES was initiated in the early 2000s and was reported to be highly biodegradable and tunable [22]. HES is naturally found in the form of "waxy starch" and is synthesized by hydroxyethylation of amylopectin [23]. HES has been used as an excellent plasma-volume expander for decades and recently has been proved to enhance the plasma half-life of nanomedicines in vivo [24][25][26][27]. In addition, there are mounting evidences that HES offers biofriendly advantages over PEG, such as low immunogenicity in vivo [26,28]. HES is a highly soluble polymer that contains negligible immunological reactions to cells [29]. Its structure analogy to human glycogen is a key reason for minimal immunogenicity [28]. Moreover, HES is reported to be more stable compared to PEG in the gene delivery system. For example, Noga et al. [21] used HESdecorated polyplex and reported higher stability with minimal aggregation effect compared to PEG-polyplex. In a further study, HES exhibited potential shielding property for the targeted delivery of nucleic acid and increased transfection efficiency compared to PEG [10,17,21].
Several modification strategies have been attempted on HES for the selective delivery of drugs as well as for stabilizing the nanoparticles, such as chemical modification [4,30,31] and biological modification [4,32]. Herein, HES was chemically modified into HES-SH and was used to functionalize AuNRs by Tris-based loading method. Further, it was compared with PEG-SH-coated AuNRs at different aspects. "Seed mediated growth method" was applied to synthesize AuNRs using cetyltrimethylammonium-bromide (CTAB) as a precursor surfactant. The optical property was determined by UV-visible spectrophotometry, and the hydrodynamic size and morphology were characterized by dynamic light scattering (DLS) and transmission electron microscope (TEM), respectively. The structural characterization of the HES-SH structure was determined by FT-IR (Fou-rier transform infrared) spectroscopy and 1 H NMR (nuclear magnetic resonance) spectroscopy. Then, the functionalization of AuNRs by HES-SH at different ratios was carried out. Finally, photothermal analysis of AuNRs, HES-S-AuNRs, and PEG-S-AuNRs was performed to confirm the effect of functionalization on the photothermal conversion property of gold nanorods. Our results confirmed that gold nanorods were functionalized successfully with both PEG-SH and HES-SH. Furthermore, we showed that AuNRs retained their colloidal stability and preliminary spectra after functionalization. In addition, functionalization of AuNRs with HES-SH maintained the stability of photothermal conversion by gold nanorods as PEG-SH. In summary, HES-SH can be used as a promising alternative to PEG for the surface modification of gold nanorods.   [33]. Briefly, 2,2′-dipyridyl disulfide (4.4075 g, 40 mM) dissolved in methanol-glacial acetic acid (20.8 mL) and 2-mercaptoethylamine hydrochloride (1.144 g, 20 mM) dissolved in methanol (8.75 mL) were mixed dropwise under stirring. Then, the reaction mixture was kept under an argon atmosphere for 48 h to lower thiol oxidation and was concentrated under reduced pressure on the following day. The product thus obtained was approximately 5-10 mL of yellow-colored oil-like liquid, which was further precipitated by the addition of cold ether (100 mL) following the purification by redissolving in methanol (20 mL). The same purification process was repeated five times and further dried for approximately 2 h to get the final product and stored at 4°C for further use. The synthetic scheme of PDA-HCl is shown in Figure S1. Journal of Nanomaterials et al. [18] with slight modifications. Briefly, HES (2 g, 130/0.4) was dissolved in water (40 mL) in a round-bottom flask, and sodium hydroxide (1.6 g) dissolved in a suitable amount of water was (approximately 5-10 mL) added to it. Then, the reaction was allowed for 3 h at 70°C using a condenser pipe after the addition of monochloroacetic acid (MCA). The resultant mixture was cooled, added into methanol (approximately 300 mL) under stirring, and waited for a while allowing them to precipitate. Supernatant was discarded, and the product collected was further dissolved in water (15 mL) under stirring. Later, the product was dialyzed against water for two days at least four times. Finally, lyophilization was carried out for 3 days to get the final product.

Materials and Methods
2.4. Synthesis of HES-PDA. The methods to synthesize HES-PDA were adopted from Wu et al. [18]. In brief, all of the three compounds, namely, EDCI (416 mg, 1.09 mM), NHS (125 mg, 0.54 mM), and PDA-HCl (242 mg, 0.54 mM) were alternately added to aqueous HES-COOH (1000 mg, 40 mL) under mild stirring. Then, the reaction was left for 24 h at room temperature, and dialysis was carried for three days. At last, the final dried product (HES-s-s-R) was obtained through lyophilization at -50°C [18].

Synthesis of HES-SH.
HES-SH was synthesized using DTT following the methods by Wu et al. with slight modifications [18]. Briefly, HES-PDA (200 mg) dissolved in DMSO (5 mL) and DTT dissolved in DMSO (1 mL) were mixed. Later, the solution mixture was kept in the argon environment to eliminate the oxygen and left under stirring for 24 h at room temperature following the dialysis for four times and freeze-drying at -50°C. Collected products were stored below -20°C for the structural characterization by 1 H NMR spectroscopy (model-AscendTM, 600 MHz, Bruker) and FT-IR spectroscopy (model-Vertex 70, Bruker).

Synthesis of AuNRs by Seed-Mediated Growth Method.
AuNRs were synthesized by a seed-mediated growth method following the protocol prepared by Ye et al. [35]. Firstly, all the glasswares were well cleaned by aqua regia. For the synthesis of seed solution, HAuCl 4 /Au III (5 mL, 0.5 mM) and CTAB (5 mL, 0.2 M) were mixed under mild stirring. Then, freshly prepared cold NaBH 4 (0.6 mL, 0.01 M) was added into the previous solution under stirring for 2 min at a speed of 1200 rpm. The yellow color was changed to brownish yellow, and the solution was kept for 30 min at room temperature. For the synthesis of growth solution, CTAB (1.8 g, 0.05 M) and 5-bromosalicyalic acid (0.22 g) were dissolved in warm water (50 mL, 50-70°C  Figure 2(b) confirms the successful production of HES-COOH (2). This range represents to the methylene group's protons present in MCA [37]. In the second step, PDA was added to HES-COOH. As a result, there was a formation of an amide bond. Figure 2(c) exhibits the successful synthesis of HES-PDA. The lately emerged three peaks at the range from 7 to 8.5 ppm in Figure 2(c) determine the protons present in the pyridyl group (4,5,6,7). Similarly, the appearance of two new peaks to the right side (range 2.8 to 3.2 ppm) in Figure 2(c) represents to the methylene group's protons present in ethylamino (2, 3) [33]. In the third step, DTT was added to HES-PDA. As DTT is a highly reducing agent, it is responsible for breaking disulfide (s-s) bonds to form 2-mercaptopyridine [18]. In Figure 2(d), almost all the pyridyl proton signals are vanished, which confirms the cleavage of disulfide bonds to produce HES-SH. The FT-IR characterization was performed and analyzed as suggested by Wu et al. [18] using FT-IR spectrometer, and the patterns of FT-IR spectra we observed were similar as they reported. The FT-IR spectra of HES (130/0.4), HES-COOH, and HES-SH are shown in Figure 2(e). While comparing the bands a and b, the remarkable change appeared around the peak 1606 cm -1 . This change determines the -COO¯(carboxylate ion) stretch in the reaction, thus confirming the synthesis of HES-COOH. Similarly, while comparing the bands b and c, the significant difference was shown around the peak 1637 cm -1 , which displays that there is a stretch at -CONH-bond in the amide group of HES-PDA indicating the presence of PDA in the reaction. None of any notable peak within the range from 1513 cm -1 to 1637 cm -1 confirms the absence of a pyridyl group in HES-PDA due to the disulfide bond cleavage by DTT to produce HES-SH [18]. Based on both 1 H NMR and FT-IR spectra analyses, it is confirmed that HES-SH was synthesized successfully.

Synthesis of AuNRs.
AuNRs were synthesized by a simple and effective seed-mediated growth method reported by Ye et al. [35]. The synthetic scheme of AuNRs is displayed in Figure 3(a). Two solutions, namely, "seed solu-tion" and "growth solution" were prepared and used cautiously. The seed solution was prepared reducing Au 3+ (gold III) ion with NaBH 4 (sodium borohydride) and surfactant cetyltrimethylammonium-bromide (CTAB, 0.2 M). The growth solution was prepared with CTAB (0.05 M) and gold III ion solution (HAuCl 4 ) in the presence of additive 5-bromosalicyalic acid, silver nitrate (AgNO 3 ), and ascorbic acid. The hydrophobic benzene ring of 5bromosalicyalic acid favors the penetration into the hydrophobic alkyl tail of CTAB molecules thereby allowing the rod-like transition of CTAB micelles due to the reduced electrostatic repulsion between micellar surface charge and COOions [35]. Ag + ions allow the anisotropic growth of gold nanorods by breaking the symmetry of gold seeds [38]. The addition of ascorbic acid in the CTAB-HAuCl 4 solution led to the disappearance of yellow color due to the reduction of Au 3+ ions [35]. It is crucial to address the multiple parameters, such as temperature, pH of growth solution, concentration of reagents, and amount of Ag + as well as the amount of VC (ascorbic acid) in order to maintain the aspect ratio of gold nanorods [35,36,39]. The injection of a little amount of seed solution in growth solution at the final step acted as a precursor to yield the growth of AuNRs. The resulted red color after 12 h of incubation displayed that there is a production of AuNRs with longitudinal surface plasmon resonance (LSPR) around 800 nm. For the further confirmation of LSPR wavelength, UV-vis spectrophotometry was conducted.

Characterization of Optical Properties of AuNRs by UV-Visible Spectrophotometry.
The "anisotropic" behavior of AuNRs caused the formation of two surface plasmon resonance bands [38]. The "absorption maxima" in longitudinal plasmon resonance of AuNRs was measured 804 nm in UVvisible spectrophotometry, whereas 527 nm of maximum absorption was found in transverse plasmon resonance mode. The gold nanorods prepared at the NIR region have potential applications in photothermal ablation of cancer cells, where they strongly absorb NIR light and generate heat upon excitation thereby resulting in a selective heating of the tumor cells [40][41][42]. The optical image and UV spectra of synthesized AuNRs with 804 nm are shown in Figure S2 and Figure 3(c), respectively. Synthesized AuNRs were centrifuged in falcon tubes to remove excess CTAB ( Figure S3). Then, samples were centrifuged two times at the speed of 9500 rpm for 25 min. The supernatants were discarded in each step and finally redispersed in PBS. These steps removed almost all the excess CTAB present in the solution. The UV-visible spectrophotometry was again carried out in centrifuged samples, and the longitudinal plasmon resonance band was observed at 777 nm (Figure 3(c)). It is because double centrifugation led to the sedimentation of AuNRs with larger volume and smaller aspect ratio, thus resulting in the blue shift of longitudinal surface plasmon resonance band [43]. Our results demonstrated that there were no aggregations between AuNRs (Figure 3(c)). Figure 3(d). The scale bar is 100 nm. The morphology of AuNRs was rod-4 Journal of Nanomaterials shaped having the length 50:67 ± 0:22 nm and width 12:17 ± 0:04 nm. TEM image also showed that little spherical particles were present along with AuNRs ( Figure 3(d)).

HESylation and PEGylation of AuNRs. HESylation and
PEGylation were conducted according to the protocol published by Zhang and Lin [36]. HESylation and PEGylation were conducted simultaneously using the same method with the same concentration of AuNRs in the presence of Tris buffer. The experiment was performed for different ratios of AuNRs and HES-SH and compared. The ratios were as follows: 1AuNRs : 2HES-SH, 1AuNRs : 2PEG-SH, 1AuNRs : 4HES-SH, 1AuNRs : 4PEG-SH, 1AuNRs : 8HES-SH, and 1AuNRs : 8PEG-SH. Tris buffer contains an amine (-NH 2 ) as its functional group, which has a strong affinity toward the gold (Au) surface [36]. Thus, -NH 2 group might have displaced the CTAB-coated cationic bilayers [36]. As a result, the thiol (-SH) group of HES-SH and PEG-SH bounds to the gold surface by strong covalent bond [44]. Another reason for successful functionalization might be due to the neutralization of anions present in HES-SH and PEG-SH molecules, which facilitates HES-SH and PEG-SH molecules to be bonded quickly with the gold surface [36]. The application of 808 nm laser in functionalized gold nanorods resulted in a dissipation of heat without affecting the photothermal conversion stability of gold nanorods. The process of HESylation and PEGylation is shown in Figure 3(b). 5 Journal of Nanomaterials [45]. From the obtained AuNRs spectra, it is demonstrated that no aggregations between the AuNRs were observed after CTAB displacement with both HES-SH and PEG-SH. It means that the functionalization with both polymers restored the colloidal stability and initial spectra of AuNRs, presumably due to the steric hindrance caused by polymer coating [46]. It is reported that HES-SH or PEG-SH has strong-affinity towards water molecules. Surface functionalization of gold nanorods with HES-SH or PEG-SH forms "hydration layer" with water that prevents the aggregation of gold nanorods [2].

Characterization of HESylated AuNRs and PEGylated
TEM characterizations of HES-S-AuNRs and PEG-S-AuNRs are shown in Figures 3(f) and 3(h). TEM images were taken by Hitachi microscope at a voltage 100 kV, and the scale bar is 200 nm. The morphology of AuNRs in both specimens was rod-shaped. The length and width of HES-S-AuNRs were measured 60:5 ± 0:17 nm and 15:5 ± 0:06 nm, respectively. The length and width of PEG-S-AuNRs were measured 60:8 ± 0:11 nm and 14:4 ± 0:05 nm, respectively. Our result showed that the length and width of both HES-S-AuNRs and PEG-S-AuNRs were increased bỹ 10 nm and~3 nm, respectively, compared to centrifuged   Figure 3(i). The result displays that the zeta potential of CTAB-stabilized AuNRs was +38.3 mV. The reason for the positive zeta potential is due to the presence of highly positive cationic CTAB in AuNRs solution [47]. HES (130/0.4) and HES-SH produced -7.29 mV and -13 mV zeta potential, respectively, whereas HES-S-AuNRs exhibited -25.1 mV surface charge. The result revealed that there was a huge reduction in the zeta potential charge (from +38.3 mV to -25.1 mV). This suggests that the CTAB bilayer in AuNRs was replaced and properly coated by HES-SH. On the other hand, the zeta potential of PEG-S-AuNRs was also decreased to -9.08 mV suggesting that the PEGylation approach was also completed while displacing the micelle bilayer formed by CTAB on AuNRs surface.
The hydrodynamic diameters (D H) of AuNRs, HESylated AuNRs, and PEGylated AuNRs are displayed in Figure 3(j). The original hydrodynamic diameter of AuNRs was found 81 nm. After HESylation and PEGylation, the diameter was increased to 188.2 nm and 183.1 nm, respectively. The hydrodynamic diameter is obtained as a result of the spherical approximation of gold nanorods. Thus, it cannot be compared directly to the "effective" size obtained by TEM [48]. It is observed that HESylation provides a narrower size distribution in comparison with PEGylated samples (Figure 3(j)). respectively. The water sample was taken as a control (Figure 4(d)). The images were captured at the highest temperature generated by each sample. The initial temperature was set to 27°C, and the delta of temperature (Δ T) was calculated by subtracting the initial temperature from the maximum temperature generated by each sample. The ideal temperature needed for photothermal therapy in clinical application is around 50°C, and the basal temperature in the human body is 37°C. Hence, the jump in the temperature needed is >13°C [49]. The maximum heat dissipation achieved by AuNRs, HES-S-AuNRs, and PEG-S-AuNRs was 50.8°C, 50.6°C, and 50.3°C, respectively (Figures 4(a)-4(c)), while the temperature in water remained constant ( Figure 4(d)). The Δ T achieved was~23°C in 3.5 min for all the three samples. While comparing the temperaturetime curve by HES-S-AuNRs and PEG-S-AuNRs, the increment rate of temperature is almost similar, and both functionalization techniques did not affect the heat dissipation process by AuNRs, thereby maintaining the photothermal conversion stability of AuNRs (Figure 4(e)).

Conclusion
In conclusion, surface functionalization strategy by polymer surfactants is essential to improve the stability of gold nanorods. HESylation is one of the potential functionalization approaches that can overcome the limitations over 9 Journal of Nanomaterials PEGylation, such as hypersensitivity and nonbiodegradability. In the current study, we reported that the functionalization with both PEG-SH and HES-SH restored the colloidal stability and initial spectra of gold nanorods. In addition, both HESylation and PEGylation did not affect the photothermal effect produced by gold nanorods to exceed the optimal jump of the temperature, thereby maintaining the photothermal conversion stability of AuNRs. Owing to the significant advantages of HES over PEG, especially biocompatibility and biodegradability, HESylation can be a promising alternative for PEGylation. However, further in vitro and in vivo experiments must be performed to confirm the biocompatible and biodegradable behavior of HESylated gold nanorods for the clinical application.

Data Availability
All data are integrated in manuscript.

Conflicts of Interest
The authors declare no conflict of interest.