Controlling the Chain Folding for the Synthesis of Single-Chain Polymer Nanoparticles Using Thermoresponsive Polymers
He Zhang, Liang Zhang, Jichun You, Niboqia Zhang, Linxiuzi Yu, Huanyu Zhao, Hu‐Jun Qian, Zhong‐Yuan Lu
Abstract
Open AccessCCS ChemistryRESEARCH ARTICLE1 Aug 2021Controlling the Chain Folding for the Synthesis of Single-Chain Polymer Nanoparticles Using Thermoresponsive Polymers He Zhang, Liang Zhang, Jichun You, Niboqia Zhang, Linxiuzi Yu, Huanyu Zhao, Hu-Jun Qian* and Zhong-Yuan Lu He Zhang State Key Laboratory of Supramolecular Structure and Materials, Institute of Theoretical Chemistry, Jilin University, Changchun 130012 , Liang Zhang College of Material, Chemistry, and Chemical Engineering, Hangzhou Normal University, Hangzhou 311121 , Jichun You College of Material, Chemistry, and Chemical Engineering, Hangzhou Normal University, Hangzhou 311121 , Niboqia Zhang State Key Laboratory of Supramolecular Structure and Materials, Institute of Theoretical Chemistry, Jilin University, Changchun 130012 , Linxiuzi Yu State Key Laboratory of Supramolecular Structure and Materials, Institute of Theoretical Chemistry, Jilin University, Changchun 130012 , Huanyu Zhao *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Supramolecular Structure and Materials, Institute of Theoretical Chemistry, Jilin University, Changchun 130012 , Hu-Jun Qian* State Key Laboratory of Supramolecular Structure and Materials, Institute of Theoretical Chemistry, Jilin University, Changchun 130012 and Zhong-Yuan Lu State Key Laboratory of Supramolecular Structure and Materials, Institute of Theoretical Chemistry, Jilin University, Changchun 130012 https://doi.org/10.31635/ccschem.020.202000190 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesTrack Citations ShareFacebookTwitterLinked InEmail Synthetic polymer single-chain nanoparticles (SCNPs) are an emerging new class of nanomaterials that possess similar folded structures as natural proteins. However, most SCNPs reported so far are packed loosely in their interior, resembling those of intrinsically disordered proteins (IDPs). Here, we report a facile strategy to synthesize SCNPs with controllable folding in aqueous solution. The precursor was mainly composed of N-isopropylacrylamide (NIPAM), a small percentage of hydrophobic photo-cross-linkers, and their hydrophilic counterparts. Contrary to conventional approaches that started from extended self-avoiding chain conformations, we started from prefolded conformations of thermoresponsive precursors at different temperatures. SCNPs were synthesized in situ by triggering reactions between hydrophobic photo-cross-linkers located in the interior under UV light irradiation. Folding structures ranging from loosely packed IDP-like to fully collapsed globules were obtained at different temperatures. Our experimental findings were also confirmed by using molecular dynamics simulations. Theoretical scaling analysis of the hydrodynamic radius Rh and molecular weight distributions for both precursors and corresponding SCNPs showed that these SCNPs synthesized at high temperatures (>305 K) indeed had fully folded globule structure. Besides, these SCNPs were thermoresponsive in a wide range, and near human body temperature; thus, could be superior candidates for future applications in drug cargo systems. Download figure Download PowerPoint Introduction There has been growing interest in synthesizing single-chain nanoparticles (SCNPs) by folding/collapsing isolated polymer chains at high dilution via intrachain cross-linking.1–8 As a new type of macromolecular architecture, SCNPs have found wide applications in drug delivery,9–11 imaging,12,13 nanoreactor,14,15 and catalysis.16–19 Beyond these applications, society's goal is to produce tightly folded protein-like globule SCNPs accessing the true globule limit of natural proteins. Practically, it is believed to be an easy and efficient way to reproduce nature's way of folding biomacromolecules such as proteins into defined 3D architectures and understand the association of the protein folding process into their natural functions. However, the SCNPs produced by state-of-the-art synthesizing methods are still far away from the well-controlled degree of folding and the corresponding morphology. After summarizing a large library of SCNPs available in the literature, Pomposo et al.20 found that instead of globular conformations, most SCNPs adopted open, sparse morphologies resembling those of intrinsically disordered proteins (IDPs) with locally compact portions of the peptide chain connected by flexible linkers. Conventional synthesis of SCNPs often started from extended self-avoiding conformations of precursors in a good solvent, with reactive functional groups (cross-linkers) randomly attached to the precursor backbone. An intrachain cross-linking reaction had a much higher probability between these cross-linkers with a shorter contour distance along the precursor backbone under such a condition. This resulted in the local spheroidization along the backbone and the formation of a loosely packed IDP-like structure of SCNPs.21,22 It has also been confirmed by recent computer simulations23,24 that intrachain cross-linking reaction probability decreases rapidly at large contour distances. Based on these concerns, many efforts have been devoted to improving the long-distance reaction probability and the subsequent degree of folding of the precursor. For instance, orthogonal or multistep cross-linking schemes have been proposed to enhance the degree of folding of SCNPs.23,25,26 At a given content of the cross-linkers, separating reactive functional groups with different linker types can effectively enlarge the average contour distance between cross-linkers of the same type and, therefore, their reaction probability at large contour distance. Recent simulation and small-angle neutron scattering works27,28 demonstrated that, in the presence of nonreactive precursors acting as purely steric crowders in a concentrated solution, ring precursors adopted crumpled globular conformations that favored the fabrication of SCNPs with a higher folding degree. Although all of these methods could improve the extent of folding, they all have limitations. For the former, intrachain cross-linking reactions still have a higher probability at short contour distances due to the extended self-avoiding conformation of the precursor. For the latter, separation of SCNPs from a concentrated solution of nonreactive crowders could be challenging; especially, there would be an unavoidable concatenation of SCNPs with surrounding nonreaction crowders. Another important issue for SCNP synthesis is to avoid interchain collapsing; this is conventionally achieved by doing synthesis at high dilution. Recent works by Yang group7,29 demonstrated that interchain aggregation could be inhibited substantially by introducing electrostatic interactions on the precursor to enable the intrachain-cross-linking reaction to be performed in concentrated solutions. On the other hand, recent experiments by ter Huurne et al.30,31 indicated that a delicate balance between intra- and interchain self-assembly pathways could be directed by carefully tuning the polymer's hydrophilic–hydrophobic interactions in the synthesis of SCNPs using a supramolecular approach, whereby intrachain folding could be achieved via threefold hydrogen-bonding interactions. They found that by replacing part of the benzene-1,3,5-tricarboxamide (BTA) grafts (hydrogen-bonding agents) with hydrophobic dodecyl, while keeping the hydrophobic content constant, could promote the intrachain folding, thereby favoring the formation of globular SCNPs with a more structured interior. Similarly, Terashima group32,33 reported that amphiphilic random copolymers consisting of both hydrophilic and hydrophobic pendants could undergo reversible single-chain folding in water via intramolecular hydrophobic interactions. A more detailed discussion on the role of hydrophilic/hydrophobic balance in the folding of amphiphilic heterograft copolymers into SCNPs can be found in a recent review by Meijer group,1 who showed that this balance appears to be delicate indeed. Recently, Gormley group34 studied the folding behavior of an extensive combinatorial library (>450) of polymers with varying hydrophobicity and intrachain hydrogen-bonding strength, using a combinatorial and high-throughput approach. They found that only a small group (9/457) of poly(ethylene glycol) (PEG)-functionalized random heteropolymers and block copolymers exhibited compactness similar to that of bovine serum albumin (BSA). These results indicated that, although chain hydrophobicity is a good factor for tuning the intrachain folding, one must be aware that a careful tuning of the chain composition might be necessary for the fabrication of SNCPs with compact interior structure. In the above mentioned systems, single-molecule intrachain folding was usually a cause-and-effect process resulting from intrachain cross-linking via covalent or noncovalent chemistry, where the latter was not only the driving force of the former but also fixed the SCNP configuration in situ. In such strategies, the extended self-avoiding chain conformation of the precursor was the main factor determining the probability of high intrachain cross-linking at short contour distances and the subsequent formation of IDP-like structures of the resultant SCNPs. To improve the degree of folding of the SCNP, one has to improve the probability of the intrachain cross-linking at long contour distances, which is hard to achieve. In this study, we inverted the picture of cause and effect; namely, we triggered the intrachain cross-linking after precursors' precollapse. Such intramolecular cross-linking along precursor backbone at various contour distances was expected to be close to even, which was confirmed in our recent computer simulations,24 where precollapsed conformation of the precursor was fabricated under a co-nonsolvency condition. In this study, we chose poly(N-isopropylacrylamide) (PNIPAM) as the main chain of the precursor, which is a thermoresponsive polymer widely known for its lower critical solution temperature (LCST) phenomenon and a coil-to-globular transition at ∼32 °C in aqueous solution. As demonstrated by Wu group,35,36 the individual PNIPAM chain at high dilution could change from a random coil to a crumpled coil, then to a molten globule, and eventually to a fully collapsed thermodynamically stable single-chain globule as the temperature increased. In this study, photo-cross-linking reactions were used to fix these precursor conformations, thereby obtaining various folding degrees for SCNPs in situ. Here, 4-acryloyloxybenzophenone (ABP) was applied as a photo-cross-linking agent, which reacted efficiently and unselectively with the approaching C−H bonds upon UV irradiation; thus, ensuring high cross-linking efficiency.37 On the other hand, hydrophobic ABP favored the protected SCNP core position, thereby somewhat avoiding the intermolecular reaction. For the tailor-made precursors, hydrophilic poly(ethylene glycol) methyl ether acrylate (PEGMA) segments were used to balance the ABP's hydrophobic feature. Furthermore, we demonstrated that the resulting SCNPs with PNIPAM backbone could exhibit a thermoresponsive behavior at approximately human body temperature. Experimental Section Synthesis of the precursors and SCNPs The precursors (named as PNAE in the following text) are synthesized with N-Isopropyl acrylamide, ABP, and PEGMA by reversible addition-fragmentation chain transfer (RAFT) copolymerization, and the SCNPs were prepared from these precursors under irradiation by UV light in an aqueous solution. Their structures and properties were characterized by proton nuclear magnetic resonance (1H NMR) spectra, dynamic light scattering (DLS), 1H NMR diffusion-ordered spectroscopy (DOSY), size-exclusion chromatography (SEC), 1H nuclear Overhauser effect (NOE) differential spectroscopy, transmission electron microscopy (TEM), and UV–vis absorption spectra. Further details regarding materials and characterizations can be found in the Supporting Information. Exemplary experimental procedure for the synthesis of precursor PNAE-10.5 N-Isopropyl acrylamide (1.92 g, 17 mmol), ABP (252 mg, 1 mmol), PEGMA (0.96 g, 2 mmol), 4-cyano-4-[(dodecylsulfanylthiocarbonyl) sulfanyl] pentanoic acid (32 mg, 0.08 mmol), and 2,2′-azobis(2-methylpropionitrile) (AIBN) (2.62 mg,0.016 mmol) were dissolved in anhydrous 1,4-dioxane (7 mL) in 10 mL Schlenk tube. The mixed solution was degassed by five successive freeze–pump cycles in the Schlenk tube and stirred for 24 h at 343 K. The polymer solution was diluted by 5 mL tetrahydrofuran (THF) and purified via precipitation in n-pentane/ethyl ether (9/1). The PNAE-10 was obtained by filtration as a yellow powder (yield 2.73 g). 1H NMR (400 M, CDCl3): 7.40–7.89 ppm (m, 9H), 4.03 ppm (br, 1H), 3.65 ppm (s, 18H), 3.38 ppm (s, 3H), 1.14–2.99 ppm (br, 3H), 1.19 ppm (br, 6H). The other precursors are synthesized by different feed ratios of monomers. Exemplary experimental procedure for the synthesis of SCNP-10.5-298 PNAE-10.5 (100 mg) is dissolved in ultrapure water (100 mL) in a 500 mL three-neck round-bottom flask with a full jacket. The solution temperature is lowered to 278 K, ultrasonic for 30 min, and then increased to 298 K. The solution is irradiated with 365 nm UV light for 11 h at 298 K. After evaporation to remove water, SCNP-10.5-298 is diluted by 5 mL THF and purified via precipitation in N-pentane/ethyl ether (9/1). The SCNP-10.5-298 is obtained by filtration as a yellow powder (yield 96 mg). The other SCNPs were prepared at varying reaction temperatures from the corresponding precursors. Computational methods In our simulation, we used the coarse-grained molecular dynamics (CGMD) simulation technique. PNAE precursor copolymers and the SCNPs are simulated by the bead-spring polymer model. As depicted in Supporting Information Scheme S1, monomers on the chain are represented by four different bead types: N-isopropylacrylamide (NIPAM) and ABP are represented by A and B, respectively. While C and D represent the backbone and side-chain of PEGMA, respectively. Each precursor has a chain length of N = NA + NB + NC = 100, with NA = 85, NB = 5, and NC = 10. Each PEG tail is attached to C modeled by nine D beads. We use S to represent the solvent water molecule. Scheme 1 | Synthesis of PNIPAM, PEG-PNIPAM, ABP-PNIPAM, and PNAE. Download figure Download PowerPoint Finitely extensible nonlinear elastic (FENE) potential is used to describe the interactions between bonded monomers along the chain contour, U FENE ( r ) = − 1 2 k R 0 2 ln [ 1 − ( r R 0 ) 2 ] + U LJ ( r , r cut = 2 1 / 6 σ ) , where R0=1.5σ, k=30ɛ/σ2. Nonbonded interactions between different components are described by a Lennard–Jones (LJ) potential, U LJ ( r ) = 4 ɛ LJ [ ( σ r ) 12 − ( σ r ) 6 ] . Simulations are performed with the reduced units, with the energy, length, and mass in units of ɛ, σ, and m, respectively; therefore, the time unit is τ = σ m / ε and the temperature unit is ε/kB, where kB is the Boltzmann constant. These nonbonded interactions between monomers are truncated at a cutoff radius, rcut. Further simulation details are supplied in the Supporting Information. Results and Discussion Design and synthesis of precursors A set of random copolymers, named PNAE, is synthesized by RAFT copolymerization of NIPAM, ABP, and PEGMA (Mn = 480 g/mol, with ∼8.9 oxyethylene units on average) with a 4-cyano-4-[(dodecylsulfanylthiocarbonyl) sulfanyl]pentanoic acid as a chain-transfer agent and azobis(isobutyronitrile) (AIBN) as an initiator in 1,4-dioxane, as shown in Scheme 1. For the composition of the precursors, the ABP content was fixed at ∼5 mol %, and the PEGMA content was varied at 5.3, 10.5, and 12.8 mol %. They were labeled as PNAE-5.3, PNAE-10.5, PNAE-12.8, respectively. The molecular weight, degree of polymerization, and composition of the polymer samples were verified by SEC and 1H NMR, as listed in Table 1. For comparison, bicomponent PNIPAM-PEGMA and PNIPAM-ABP random copolymers and PNIPAM homopolymer were also prepared as control samples via RAFT polymerization under the same reaction conditions. Detailed data about our experiments could be found in the Supporting Information. Table 1 | Overview of the Composition (Mol %), Number-Averaged Molecular Weight (Mn), Degree of Polymerization (DP), Molar Mass Dispersity (Đ) of the Precursor Polymers Compositiona (mol %) Sample NIPAM ABP PEGMA Mn, SECb (g/mol) ĐSECb DPSECb Mn NMRa (g/mol) DPNMRa PNIPAM 100 0 0 18,200 1.28 161 14,100 125 ABP-PNIPAM 95.1 4.9 0 23,900 1.39 199 21,200 180 PEG-PNIPAM 94.0 0 6.0 22,500 1.43 167 18,000 134 PNAE-5.3 90.1 4.6 5.3 29,400 1.44 212 21,200 153 PNAE-10.5 84.7 4.8 10.5 27,700 1.42 175 21,300 135 PNAE-12.8 82.1 5.1 12.8 25,600 1.56 153 24,500 147 aDetermined using 1H NMR spectroscopy. bDetermined using SEC in THF, calibrated with polystyrene standard. Phase behavior of PNAE precursors We first characterize the behavior of the precursor copolymers in water upon the change of temperature. Since our synthesized PNAE copolymers had a large content of NIPAM, a similar LCST-type phase behavior was expected for these PNAE polymers. Techniques of DLS, DOSY, and 1H NOE difference spectroscopy are used. The results of the temperature-dependent DLS measurements are shown in Figure 1a, where the hydrodynamic radius (Rh) of the assembly in aqueous solutions was plotted at different temperatures ([polymer] = 1 mg/mL). It shows that all polymers, including PNIAPM, PEG-PNIPAM, and PNAE series, aggregated in water upon heating but at different temperatures, as indicated by a sudden increase in the Rh at a specific temperature. Note that the copolymer of ABP-PNIPAM is insoluble in water due to the hydrophobicity of ABP segments' copolymerization in the chain, indicating a considerable decrease in the aggregation temperature well below the current testing temperature range. Meanwhile, the introduction of hydrophilic PEGMA segments in PNAE copolymers elevated the aggregation temperature, which is in good agreement with a literature report.38 These results indeed show that PNAE copolymers exhibit typical LCST phase behavior, and we could easily manipulate the aggregation temperature by incorporating a small portion of hydrophilic/hydrophobic constituents. Notably, with the hydrophobic ABP segments content set at ∼5%, the aggregation temperature of the PNAEs could be tuned in a wide range by varying the hydrophilic PEGMA content. Specifically, the aggregation temperature of PNAE-5.3 has an LCST temperature of ∼290 K, lower than that of PNIPAM. In comparison, PNAE-10.5 and PNAE-12.8 had LCST temperatures of ∼305 and ∼312 K, respectively, which were close to or higher than that of PNIPAM. Figure 1 | (a) Temperature-dependent DLS analysis of PNIPAM, PNAE-5.3, PNAE-10.5, and PNAE-12.8 in water at 1 mg/mL. (b) Hydrodynamic radius Rh of PNAE-12.8 random copolymers in D2O calculated from 1H DOSY NMR spectra using Stokes–Einstein equation. (c and d) 1H NMR (500 MHz) spectra (top) and 1H NOE differential spectra (middle and bottom) of PNAE-12.8 in CDCl3 (c) and D2O (d), respectively, at 298 K. The molecular structure on the top left corner shows the assignments of 1H NMR and 1H NOE signals. Download figure Download PowerPoint We speculated that PNAE copolymers were in an isolated single-chain state before the aggregation in Figure However, the delicate of the chain before aggregation is hard to due to the of DLS For temperature-dependent DOSY was for aqueous copolymer solution in the temperature range K. the Stokes–Einstein ( in Supporting the Rh at different temperatures is The results for PNAE-12.8 are plotted in Figure It shows that Rh of PNAE-12.8 in aqueous solution decreases from to nm in the range it a at nm between and K. Note that the decrease of Rh at at a lower temperature, with LCST aggregation temperature of ∼312 via DLS It is that DOSY the single-chain behavior, while DLS the is the LCST temperature of PNIPAM, indicating that the PNIPAM backbone of the PNAE precursor intrachain upon and it a fully folded thermodynamically stable single-chain globule at K. The degree of folding of the precursor PNAE-12.8 in water was characterized by using 1H NOE difference spectroscopy. The NOE effect from the magnetic interactions a proton is close in to and it is from In the of the 1H NOE on the distance between the proton and the irradiated proton 1H NMR (500 MHz) spectra were also as a and show the results in and water respectively, both are performed at 278 K. The top show the results for 1H NMR spectra, the and the are for 1H NOE spectra. we that the methyl in the of PEG were irradiated (middle in both CDCl3 and of the NOE for other On the the methyl (c) on the NIPAM monomers were irradiated in a good solvent of CDCl3 in Figure we only NOE for the and and on the first PEG with acrylate backbone on that these and on the had similar distance as the irradiated methyl (c) on the NIPAM therefore, they had a high probability of in close in The of and was in the in Figure PNAE is a random copolymer with different therefore, there were of PEGMA and These results indicated that in a good CDCl3 solvent, the backbone of the PNIPAM, PEGMA chain, and thus, the PNAE-12.8 chain adopted an extended chain For the 1H NOE spectra in D2O solvent in Figure we found that the methyl (c) on the NIPAM monomers were were for the and and on the first PEG with acrylate backbone on monomers and the for and were much than in CDCl3 solvent in Figure Such could be to the intrachain folding of the PNAE-12.8 chain in water, resulting in of of (c) and and NOE were in D2O for on acrylate backbone a and which was for intrachain The results of the experiments using DLS, DOSY, and NOE shown in Figure 1 indicated that PNAE copolymer precursors could into different degrees in aqueous solution. Such intrachain folding is temperature-dependent and could be upon degrees of folding could be achieved and by the temperature, which is a critical for the fabrication of SCNPs Synthesis and of SCNPs For the of the aqueous solutions ([polymer] of the PNAE-10.5 and PNAE-12.8 were prepared and irradiated with UV for 10 h at a of reaction temperatures. to the DLS and DOSY the reaction temperature was set in the range of for PNAE-10.5 and for PNAE-12.8 before the aggregation The SCNPs obtained were labeled as and respectively. To the the stable the copolymer solution was at the set temperature for 1 h before the irradiation. After the solvent was and the SCNPs obtained were via UV–vis absorption spectroscopy, and the UV–vis of synthesized at 278 and with that of their precursor, PNAE-12.8, in aqueous solution, demonstrated at nm for indicating a reaction of the ABP ( Supporting Information Figure we could the degree of cross-linking as the content of the ABP in the precursor, which was in the range of to as listed in Table 1. time and molecular weight of SCNPs were obtained by SEC in THF and with the data of the corresponding precursors. As shown in Figure time in SEC were for both the and the at different reaction temperatures, the of intrachain folding of these SCNPs upon the temperature. Figure and Table 2 the molecular weight data of these of SCNPs. As the reaction temperature the of SCNP indicating a folding of the The difference between the of SCNPs at temperatures could be to the difference of the precursors. However, such a difference at high temperatures where the LCST phenomenon and the precursor a crumpled coil or molten globule structure. a increase in the temperature for a was after K, which was with the behavior of the precursor in DOSY measurements The SCNPs were by Figure shows the of synthesized at on the from the THF solution with It showed small large which for the formation of SCNPs. Figure 2 | SEC of the precursor PNAE-10.5 PNAE 12.8 and their corresponding SCNPs synthesized at different temperatures. (c) molecular weight of SCNPs corresponding to these in (a) and (b) by A of nanoparticles of synthesized at from the THF solution and with Download figure Download PowerPoint Table 2 | Molecular and Molar Mass Dispersity (Đ) of the PNAE-10.5, PNAE-12.8, and Their SCNPs from SEC by Sample Mn Sample Mn PNAE-10.5 / 27,700 1.42 PNAE-12.8 / 25,600 1.56 278 278 298 22,500 21,200 298 Figure shows that SCNPs synthesized at different temperatures have different A of the of such SCNPs is shown in the left part of Scheme On the as depicted in the part of Scheme we also expected the SCNPs to the same thermoresponsive with their precursors, that SCNPs upon and upon For we performed temperature-dependent DOSY measurements in aqueous solution for synthesized at K. The results are plotted in Figure and with those of the corresponding precursor As Rh of both the SCNPs and the precursor upon At temperature Rh of synthesized at was than that of the PNAE-12.8 due to