Concentration Insensitive Supramolecular Polymerization Enabled by Kinetically Interlocking Multiple-Units Strategy

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Aug 2019Concentration Insensitive Supramolecular Polymerization Enabled by Kinetically Interlocking Multiple-Units Strategy Jiezhong Shi†, Haoyang Jia†, Hao Chen, Xi Wang, Jiang-Fei Xu, Weibin Ren, Jiang Zhao, Xin Zhou, Yuanchen Dong and Dongsheng Liu Jiezhong Shi† Key Laboratory of Organic Optoelectronics & Molecular Engineering of the Ministry of Education, Department of Chemistry, Tsinghua University, Beijing 100084 (China) , Haoyang Jia† Key Laboratory of Organic Optoelectronics & Molecular Engineering of the Ministry of Education, Department of Chemistry, Tsinghua University, Beijing 100084 (China) , Hao Chen Key Laboratory of Organic Optoelectronics & Molecular Engineering of the Ministry of Education, Department of Chemistry, Tsinghua University, Beijing 100084 (China) , Xi Wang School of Physical Sciences, University of Chinese Academy of Science, Beijing 100049 (China) , Jiang-Fei Xu Key Laboratory of Organic Optoelectronics & Molecular Engineering of the Ministry of Education, Department of Chemistry, Tsinghua University, Beijing 100084 (China) , Weibin Ren Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 (China) , Jiang Zhao Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 (China) , Xin Zhou School of Physical Sciences, University of Chinese Academy of Science, Beijing 100049 (China) , Yuanchen Dong Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190 (China) and Dongsheng Liu *Corresponding author: [email protected] Key Laboratory of Organic Optoelectronics & Molecular Engineering of the Ministry of Education, Department of Chemistry, Tsinghua University, Beijing 100084 (China) https://doi.org/10.31635/ccschem.019.20190009 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail We proposed and demonstrated a kinetically interlocking multiple-units supramolecular polymerization strategy. Through rationally designed multiple-units monomers, the degree of polymerization (Xw) detected was more than 50 with a polydispersity index of ∼1.4. The prepared polymers were stable when diluted to 20 μM or lower concentrations. By introducing the cooperative effects of multiunit interactions into a traditional condensation polymerization model, we found that the relationship between the concentration and the Xw was heavily influenced by the unit number involved in the kinetic interlocking. Factors that influenced the Xw and their distributions were studied systematically, and all the results matched the theoretical predictions excellently. This strategy provides novel insight into supramolecular polymerization and offers a simple but effective way to improve positive correlation of the reversibility and stability of supramolecular interactions in polymer assembly. Download figure Download PowerPoint Introduction Supramolecular polymers are polymers whose repeating units are connected by noncovalent interactions. In 1990, Jean Marie Lehn proposed the first mesomorphic polymolecular associates based on a triple hydrogen bonds molecular system.1 After that, quadruple,2 and sextuple3 hydrogen bonds, and other supramolecular interactions, including metal–ligand coordination,4–11 host–guest12–17 and π–π interactions18–21 were employed to prepare supramolecular polymers. The reversibility of their noncovalent interactions confers supramolecular polymers and materials unique properties such as responsiveness, self-healing, and thixotropy. In the meantime, such a mechanism also brings high concentration dependency to the polymerization process,22,23 making characterization of the degree of polymerization (Xw) and its distribution under diluted conditions a great challenge. From a kinetic view point, these supramolecular polymerizations are similar to condensation polymerization processes. As illustrated in Scheme 1a, each participating part of the supramolecular interaction could be considered a hard sphere; the reaction between monomers must obey the hard-sphere collision theory of bimolecular elementary reaction model, which is a first-order reaction and concentration-dependent, whereas the depolymerization process is a zeroth-order reaction and concentration independent. Due to the reversible nature of supramolecular interaction, supramolecular polymers are hard to characterize under diluted conditions. Herein, we reported a kinetically interlocking multiple-units (KIMU) strategy, which can overcome the kinetic imbalance and easily acquire supramolecular polymers at diluted conditions. As illustrated in Scheme 1b, multiple supramolecular half units are covalently linked by the distance between each unit and it is carefully designed to ensure that the interaction between two neighbor units does not exceed the range of the hard-sphere collision theory, thereby, interrupting the synchronous and cascade breaking of different units. According to this model, when one of the supramolecular units dissociates, the other units remain connected, and the two breaking unit parts are kept within a collision distance for an immediate reconnection, and no longer relate to the total monomer concentration in the reaction solution. Scheme 1 | (a) Scheme of traditional noncovalent interaction of condensation–polymerization-like supramolecular polymerization. (b) Scheme of supramolecular interaction based on kinetically interlocking multiple-units. (c) Formation of DNA supramolecular polymer. Download figure Download PowerPoint To prove this concept, we used DNA to build a model system based on the following advantages: First, a single-stranded DNA (ssDNA) sequence is composed of different nucleotides covalently connected by phosphodiester bond,24 hence one or several nucleotides could be considered a half unit of supramolecular interaction. Second, DNA sequences could be easily synthesized and optimized. Third, the formed double helix is relatively rigid and highly negatively charged, and thus, avoid cyclization and nonspecific intermolecular aggregation. As shown in Scheme 1c, we designed a 20mer ssDNA (S20, CATCGCGATGGACTGCAGTC), containing two self-complementary segments; under mild conditions, it was proven to be able to generate a linear polymer with Xw larger than 50 in a wide range of monomer concentration, and as low as micromolar levels. The final products were resistant to dilution, and their polymerization kinetics could be verified, using the proposed KIMU strategy. Result and Discussion System design and polymerization In the model system, S20 (5'-CATCGCGATGGACTGCAGTC-3') was designed to contain two different self-complementary sequences, which are covalently connected sequentially. Thus any duplex formation will generate two sticky ends that are self-complementary to themselves. In the meantime, the rigidity of the formed duplex prevents internal cyclization. Furthermore, it has a two-units-in-one-strand, designed to avoid concentration differences and eliminate unwanted termination. In a typical S20 polymerization reaction, 100 μM S20 in 1× TAE buffer containing 12.5 mmol·L−1 MgAc2 was heated to 95 °C for 5 min and then cooled at room temperature for 2 h. Subsequently, the products are analyzed by 10% Native PAGE. As shown in Figure 1a, the DNA self-assemblies reveal a blurred band at several hundred base pairs position. This result shows that the polymerization occurred, as designed, and the assembly was stable enough to withstand the condition of PAGE. The stability of the assemblies was further confirmed by ultraviolet (UV)–visible spectrometry (). The melting point of the assemblies was 55 °C, which is higher than 10 bp duplex (48 °C), but lower than 20 bp duplex (70 °C). From these results, we envisaged that the assembly is a long duplex structure, but had repeating nicks at 10 bp interval (Scheme 1). Figure 1 | (a) 10% Native PAGE (19∶1) analysis of DNA assemblies in TAE buffer containing 12.5 mmol·L−1 MgAc2, (b) AFM image for KIMU polymer, (c) ratio-dependent AsF-FFF elution curves of the KIMU polymer, and (d) mass distribution of the KIMU polymer. PAGE, polyacrylamide gel electrophoresis, TAE buffer, Tris base, acetic acid, and EDTA buffer; AFM, atomic force microscopy; KIMU, kinetically interlocking multiple units; AsF-FFF, asymmetric flow field-flow fractionation. Download figure Download PowerPoint We then performed atomic force microscopy (AFM) to investigate the conformation of the assemblies. As shown in Figure 1b, linear structures spread over the image, whose average length was ∼72.7 nm. The measured height of the structures was ∼2 nm (), which matched the size of a B-helix. This result further supported our assumption that the polymerization had happened as designed. It is noteworthy that the sample was measured at 1 μM concentration, which proved that the polymer chains could form at very low concentration. In addition, fluorescence correlation spectroscopy was employed to observe the in situ polymerization process via diffusion coefficient changes, which is heavily related to the size of the molecules.25–27 At 20 °C, a random sequence with 20 bases is 135.90 μm2·s−1, but the diffusion coefficient of the S20 assemblies is 26.34 μm2·s−1, an indication of the formation of large assemblies. Molecular weight and PDI of KIMU polymers According to the proposed KIMU mechanism, the formed polymers might be stable under very diluted conditions. Thus, we used asymmetric flow field-flow fractionation (AsF-FFF) to characterize its molecular weight and distribution. By applying a lateral field to a continuous fluid, AsF-FFF could separate molecules with different molecular weight without the application of excessive shearing force.28 As shown in Figure 1c, the blue line at 260 nm in the UV absorption is linearly related to the DNA concentration; the black line is the molecular weight detected by multi-angle light scattering. The calculated absorption coefficient of the DNA KIMU polymer is 21.8 mL·(mg·cm)−1 (for detailed calculation, see ). The mass distribution of the molecular weight of the KIMU polymers was summarized in Figure 1d. It is apparent that the population of the polymer increases rapidly with an increase in molecular weight, and then the population begins to decrease slowly, consistent with the condensation mechanism proposed. Based on these data, the weight-average molecular weight (Mw) was calculated to be 3.3 × 105 g·mol−1, the weight-average degree of polymerization (Xw) is ∼ 54, and the polydispersity index (PDI) of the DNA polymers is ∼ 1.4. For a typical condensation polymerization, the PDI and the monomer conversion p should have the relationship: PDI=1+p.29 However, the PDI of our KIMU polymers is only 1.4, even at high monomer conversion. We attribute this unique character to a phenomenon of rigidity and charge repulsion of the DNA units. DNA duplex is highly negatively charged and has a persistence length over 50 nm,30,31 which make DNA assemblies repulse to each other and prevent entanglements. Hence, the active groups can be efficiently exposed for further polymerization, resulting in narrow distribution of the molecular weights. In brief, based on KIMU strategy, the supramolecular polymerization is realized at low concentration, and its molecular weight and distribution could be characterized easily using several methods, including AsF-FFF. The polymerization degree is higher than 50 with a PDI at 1.4. Kinetic studies on KIMU supramolecular polymerization With the supramolecular polymers in hand, we further studied the factors that will influence the polymerization processes. As illustrated in Figure 2a, the concentration dependency of KIMU polymerization was investigated in two different ways: (1) "Dilute before annealing" means S20 was annealed to polymerize at different concentration of 20, 50, 100, 200, and 500 μM. (2) "Dilute after annealing" indicates S20 was annealed to polymerize at 500 μM concentration firstly, and then diluted to 200, 100, 50, and 20 μM and incubated at room temperature overnight. The results show that there is no obvious difference between these two methods. Xw of the polymers decreased very slightly with the decrease of concentration from 500 to 20 μM, which is nearly in the variation of measurement. This result indicates that the KIMU polymer is kinetically stable that can bear with 25 times dilution, and KIMU strategy can enable the polymerization at ultralow monomer concentration. Figure 2 | Relationship between the weight-average degree of polymerization and (a) monomer concentration, (b) spacer length, (c) base number, and (d) terminator ratio. Download figure Download PowerPoint Second, as indicated above, the rigidity of DNA plays a vital role in polymerization. We verified this assumption by inserting a single-stranded spacer (1, 2, 4, 8, or 12 T bases) into the middle of the monomer (for sequence details, see ) to add more flexibility to the assemblies. As shown in Figure 2b, the insertion greatly influenced Xw of systems. A single T base insertion decreased the Xw from 54 to 24, and eight T insertions reduced the Xw to 2.7. This phenomenon can be explained by the ring-chain equilibrium mechanism of supramolecular polymerization32; the increased flexibility allowed the formation of circular assemblies, which inhibited the chain extension. 10% Native PAGE analysis demonstrated that the main products of the polymerization were dimers when the spacer was eight T (). In the KIMU design, the length of each self-complementary sequence in the DNA monomers also played an important role. As shown in Figure 2c, when the length is 6 and the total length of the monomer is 12, no clear polymerization happened. We attributed this result to insufficient stability of formed duplex at room temperature. However, increasing self-complementary length is not always beneficial to the increase of Xw. In this study, we observed that when the total length of monomers increased from 20 to 36 in a stepwise manner, the Xw decreased dramatically from 54 to 8. As all self-complementary sequences have the intrinsic property to form hairpin structures, this result could be explained by the competition between the duplex and hairpin formation: When a monomer is short (S16 and S20), the hairpin structure is less stable than the duplex, so the polymerization is mainly influenced by the stability of the duplex. However, when a monomer becomes longer, the intramolecular hairpin structure acquires more stability, and in the polymerization system, hairpins formation leads to self-termination. To further prove the termination effect, we introduced different ratio terminators into S20 KIMU polymerization system, where the terminator is a 10 bases long ssDNA complementary to half S20 (red part in Scheme 1). As shown in Figure 2d, Xw barely changed when the ratio increased from 0.001 to 0.01, but gradually decreased from 54 to 7 as the ratio increases from 0.01 to 1. This result suggests that the degree of polymerization could be finely tuned by the terminator ratio. Collectively, our results demonstrated that the KIMU strategy enabled supramolecular polymerization at ultralow concentration, and the mechanistic studies revealed that the stability of the prepared polymers is insensitive to dilution. We also found that, in most cases, the KIMU polymerization could be referred to as a condensation polymerization model, for example, in control of the Xw of the final products by the addition of terminators. Mechanistic studies on KIMU polymerization Following the KIMU polymerization strategy, intermediates exist in the polymerization process. Thus we can describe the equation of the reaction, as shown below: [ n ] + [ m ] ⇄ k 1 ′ k 1 [ n · m ] intermediate → k f ( m , n ) [ ( n + m ) ] → k [ n − l ] + [ m + l ] Because the intermediate connection [n·m] is much weaker than that in the polymerized chain [n + m], it can quickly equilibrate with [n] + [m] in comparison with polymerizing to the chain [n + m]. Therefore, when we approximately have the equation, [ n · m ] ≈ k 1 k 1 ′ [ n ] [ m ] , the polymerization from [n·m] to [n + m] is equivalent to the direct polymerization from [n] and [m], as outlined in the equation: k f ( n , m ) [ n · m ] ≈ k f ( n , m ) k 1 k 1 ′ [ n ] [ m ] . Here, we have already proposed that the formation of the polymerized products is determined by the experimental conditions denoted by the formula: f ( n , m ) = 1 n m l p 2 γ 1 2 γ 2 2 . Thus, after this simple derivation, the overall master chemical equation of polymerizing and splitting reaction of these polymer segments in the solution can be presented as: d P n d τ = − ( n − 1 ) P n + 2 ∑ g = n + 1 ∞ P g − 2 K ∑ h = 1 ∞ 1 n h P n P h + K ∑ m = 1 n − 1 1 m ( n − m ) P n − m P m (1)As indicated below, only one parameter is determined by experimental conditions, given by: K = c k k ′ l p 2 γ 1 2 γ 2 2 , (2)and it includes the number of unpaired bases of DNA-polymer ends (k′,γ1), the spacer length inside the polymer (lp), the terminator ratio(γ2), and the monomer concentration of free polymer chains, (c = ∑n[n]) (see Theoretical Section in ). When the experimental condition is determined, it also determines K. For each K parameter, we can calculate the equilibrium distribution from the equation, Pn by d P n d τ = 0 . Due to the nonlinear characteristic of the equation, we can numerically calculate the equilibrium distribution, and then obtain the mass distribution of the degree of polymerization to compare with the experimental result. The weight-average degree of polymerization (Xw) could be calculated using formula (3) below, and the relationship between Xw and K can be visualized in Figure 3a. X w = ∑ n n 2 P n ∑ n n P n (3) Figure 3 | Theoretical results (blue lines are theoretical values, and dots are experimental values). (a) Theoretical weight-average degree of polymerization of DNA assemblies with different lgK values, (b) weight-average degree of polymerization of DNA assemblies with different terminator ratio, and (c) fluctuation (ΔX2) of DNA assemblies with a different weight-average degree of polymerization. Download figure Download PowerPoint To verify the proposed mechanism, we compared the theoretical and experimental results under various experimental conditions. For example, in equation (2), K is shown to be linearly related to γ2, and thus, the relationship between the terminators ratio versus Xw could be calculated, shown as a solid line graph in Figure 3b, and the square symbol denoting the experimental results. The data reveals that the theoretical and the experimental results fit excellently, suggesting the accuracy of the contribution of assumed terminators in the mechanism. Similarly, we have studied the influence of binding constant of monomer (k′), the possibility of the formation of hairpin structure (γ1), and persistence length (lp), and the results all fit the theoretical predictions as well (data are shown in ), demonstrating that the mechanism we proposed has a correct implication in physics. Based on the theoretical distributions, we calculated the fluctuations of weight in the degree of polymerization (ΔX2) as a function of the weight average of the polymer through the parameter K. As shown in Figure 3c, the dotted symbols represent the experimental results under different conditions, and the line graph represents the theoretical relationship between the fluctuations and Xw. It is apparent that the experimental fluctuation is good in consistency with the theoretical one. Thus our results verified that the theoretical model is suitable for the description of our proposed model system. It is noteworthy that the above-mentioned mechanism is applicable for all supramolecular polymerization processes. For a KIMU system, an intermediate such as the dimer, [n·m], exists and could be driven to higher intermediate (or assembly) such as the trimer [n·m·l], the tetramer [n·m·l·q], and so on, depending on the assembling constant and the initial total monomer concentration of polymers, (ct), where ct = ∑n [n]t. Here, [n]t depicts the concentration of n-degree polymers, including the contribution of intermediates. As ct increases, more intermediates could form and the average number of chains in each intermediate increases; thus, the increasing rate of the monomer concentration of the free chains, c, becomes slower. As shown in the Theoretical Section of , there are approximately, [ n ] t ≈ ∑ m [ n · m ] + ∑ m , l [ n · m · l ] + ⋯ , which is proportional to the concentration of the free chain [n] (i.e., [ n ] t ∝ [ n ] ∝ P n , while the distribution of degrees of polymerization remains the same. Therefore, there is approximately c t = ∑ x a x c x , where x is the number of polymers in the intermediate, and the coefficient ax is determined by the equilibrium constant of the assembly. For each specific small range of ct, there is approximately c t ∝ c x , thus from the equation (2) K ∝ c t 1 / x At an ultimate condition, where there are no intermediates,33 all monomeric units participate in the polymerization, c = ct, thus x = 1, as shown by the solid line in Figure 4. From our experimental data, we observed that a traditional supramolecular polymerization system (solid blue dots) by two giant monomers is in between x = 1and x = 2 (dashed line), even when the binding affinity of the system is more than 1012. In the meantime, our KIMU system (solid black squares) is closer to the x = 4 (dotted line). These results represent a verification of the proposed mechanism, and thus, we reasoned that the KIMU effects could slow down the dissociation process and enable more intermediates to form, which presented as an increase in x and a decrease of the slope. Figure 4 | Comparison of the theoretical and the experimental relationship of (lgK) of KIMU and single-unit polymers and monomer concentration. lgK, weight-average degree of polymerization; KIMU, kinetically interlocking multiple units. Download figure Download PowerPoint Conclusion Our study proposed a KIMU supramolecular polymerization strategy and demonstrated its occurrence by experiments. With rationally designed KIMU monomers, the detected degree of polymerization (Xw) could be more than 50, even under very diluted conditions. The theoretical model of KIMU polymerization kinetics has also been built, which could explain the concentration insensitive of KIMU supramolecular polymerization and the concentration dependency of non-KIMU systems. According to the theoretical model, factors that influence the Xw and their distributions have been systematically studied, and the results all matched the theoretical predictions excellently. This strategy provides novel insight into the supramolecular polymerization and offers a simple but effective way to improve the positive correlation of the reversibility and stability of supramolecular interactions. We believe that our proposed scheme will not only the of supramolecular polymers but also the design of the of with structures and beneficial of DNA supramolecular polymers DNA was in 1× buffer containing 12.5 mmol·L−1 the were heated to 95 °C for 5 cooled to and to form supramolecular polymers. AFM DNA polymers were diluted to 1 sample was and with 10 of 20 solution. the was on for After 3 the AFM was using in on the of the degree of polymerization by AsF-FFF experimental conditions for AsF-FFF could be as The monomer concentration was 100 the flow was the flow was and then decreased gradually to and the flow rate was The was of and the was 1× buffer containing 12.5 mmol·L−1 is of The no This was supported by the of and The Xi for their and on the We also for with the AsF-FFF the and most of the experiments. and the results and the model for the polymerization system. with the AsF-FFF experiments. and with the fluorescence correlation spectroscopy experiments. the results and the 1. Lehn of Supramolecular from 2, from Lehn From Through and A 8, 4. and of With and in and and Strategy 8. Molecular Supramolecular and in the Supramolecular Supramolecular of an of Supramolecular From and From of Supramolecular by Liu Supramolecular by Based Molecular Wang Dong Liu and Supramolecular and Supramolecular of in of and of A Supramolecular Supramolecular as a in of on the Formation of Molecular From 2, Zhao on Wang Chen Zhao of in its A by and Ren Zhao the of a DNA 20, Based on a of and 1, in The of and DNA on the of DNA Chen Xu Polymerization From to Chen Xu Polymerization Through Kinetic P and Chen T of the of Chemistry, Xu Dong Chen Xu and Liu DNA Supramolecular with Chemistry, Wang Chen Wang and X Kinetic of a System by Chemistry, and Zhou for Chemistry, and of Molecular on Supramolecular in Polymerization of Chemistry, & Chinese interlocking multiple was supported by the of and The Xi for their and on the We also for with the AsF-FFF the and most of the experiments. and the results and the model for the polymerization system. with the AsF-FFF experiments. and with the fluorescence correlation spectroscopy experiments. the results and the times