Role of Thermolysin in Catalytic-Controlled Self-Assembly of Fmoc-Dipeptides

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Aug 2020Role of Thermolysin in Catalytic-Controlled Self-Assembly of Fmoc-Dipeptides Meiyue Wang†, Qiansen Zhang†, Honglei Jian, Shijie Liu, Jieling Li, Anhe Wang, Qianqian Dong, Peng Ren, Xin Li and Shuo Bai Meiyue Wang† State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190 School of Chemical Engineering, University of Chinese Academy of Sciences, Beijing 100049 †M. Wang and Q. Zhang authors contributed equally.Google Scholar More articles by this author , Qiansen Zhang† Shanghai Key Laboratory of Regulatory Biology, Institute of Biomedical Sciences and School of Life Sciences, East China Normal University, Shanghai 200241. †M. Wang and Q. Zhang authors contributed equally.Google Scholar More articles by this author , Honglei Jian State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190 Google Scholar More articles by this author , Shijie Liu Shanghai Key Laboratory of Regulatory Biology, Institute of Biomedical Sciences and School of Life Sciences, East China Normal University, Shanghai 200241. Google Scholar More articles by this author , Jieling Li State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190 Google Scholar More articles by this author , Anhe Wang State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190 Google Scholar More articles by this author , Qianqian Dong State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190 Google Scholar More articles by this author , Peng Ren State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190 School of Chemical Engineering, University of Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author , Xin Li State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190 School of Chemical Engineering, University of Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author and Shuo Bai *Corresponding author: E-mail Address: [email protected] State Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190 School of Chemical Engineering, University of Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.020.201900116 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail In recent years, short peptide self-assembled materials, prepared under the control of the thermolysin catalyst, have been investigated extensively and shown to acquire various morphologies and functions as building blocks for a wide range of biomaterials and device applications. However, the role played by thermolysin in this enzymatically triggered peptide self-assembly is still ambiguous. Herein, we designed a series of Fmoc-dipeptide amphiphiles to explore the catalytic role of thermolysin. The results from our experiments and computational simulations showed that hydrophobicity and amino acid sequences of substrates have a significant correlation with thermolysin actions, including the binding capacity and catalytic efficiency. Specifically, thermolysin favors a specific substrate pattern with a hydrophilic amino acid in the first residue and hydrophobic amino acid in the second residue. Moreover, thermolysin catalyzed reactions are bidirectional and could move toward hydrolysis or condensation based on the design of its diverse substrates (peptides). However, the specificity of the enzyme action lies in the major site of its cleavage, which is the terminal hydrophobic or bulky amino side chains. We designed a two-step reaction, taking advantage of the bidirectional catalytic actions of thermolysin, to modify the sequence of Nα-fluorenylmethoxycarbonyl (Fmoc)-dipeptide from Fmoc-YL–COOH to Fmoc-YY–NH2 and treated with thermolysin, which resulted in the enzyme-catalyzed gel–sol–gel transition. This work has an instructive significance in the regulation of peptide sequences, secondary amino acid structures, morphology, and the mechanical property of self-assembled hydrogels with precise design and control at the molecular level via thermolysin catalysis. Download figure Download PowerPoint Introduction The fabrication of supramolecular hydrogels based on self-assembly of short peptides containing 2–3 amino acids are emerging as a research hot spot because of their versatile biofunctions, including flexible and well-controlled self-assembly, designed architectures, inherent biocompatibility, and biodegradability.1–4 These properties make them useful as materials for sensors, muscle-type actuators, biological membranes, and others. Nα-fluorenylmethoxycarbonyl (Fmoc) could be utilized as a protective group during synthesis to modify short peptides and then used them to drive the self-assembly of supermolecular structures via aromatic stacking interactions.1,5 Moreover, Fmoc-protected amino acids possessed anti-inflammatory properties.6 By utilizing the “bottom-up” hierarchical nanofabricated supramolecular design principle, hydrogelators could be fabricated with various chemical structures and functionalities, which could be applied widely in biomedical, bio-imaging, bio-templating, and tissue engineering.7–12 Two main strategies could be used to generate hydrogelators: (1) chemical synthesis and (2) enzymatic catalysis. Through the chemical synthesis approach, functional moieties could be incorporated into short peptide molecules to exert influence on the self-assembly and bio-functions of supramolecular systems. In the case of the enzyme-triggered catalytic reaction, an optional biological stimulus has been developed as an effective approach in peptide synthesis and self-assembly.13–16 Also, enzymes could kinetically induce the molecular ordering of peptides in a controlled and reproducible manner leading to a high level of control over the assembly process, thereby, enabling access to structurally diverse self-assembled materials that are inaccessible by conventional self-assembly strategies.17–19 The most used enzymes for peptide synthesis and triggering of self-assembly include subtilisin, phosphatase, chymotrypsin, and thermolysin. Subtilisin is an esterase isolated from Bacillus licheniformis, which could hydrolyze methyl esters from peptide precursors to trigger self-assembly.19,20 Phosphatases exist widely in biological systems, could hydrolyze phosphomonoester bonds, and cleave phosphate groups from the corresponding location of a wide selection of substrates with high efficiency.21,22 Chymotrypsin could convert dipeptide esters into oligopeptides23 or make full use of chemical energy preserved in peptide methyl ester, developed into a nonequilibrium biocatalytic assembly system, with the formation of short peptide scaffold.24 Thermolysin, extracted from thermophile Bacillus thermoproteolyticus, is a thermostable metal endopeptidase with a molecular weight of 34,600 Da.25 The conformation of thermolysin is bilobal with zinc ion in the center of the active site that is essential for catalysis. His142, His146, Glu166 (H142, H146, and E166), and a water molecule are ligands of the central zinc cation required to maintain a tetrahedron conformation to drive the catalysis. Four calcium ions are attached to thermolysin to maintain the thermostability of the enzyme, three (Ca 1, 2, and 4) bind near the C-terminal domain and one (Ca 3) to the N-terminal domain, which is critical for its stability.26 In previous research, the structure and catalytic mechanism of thermolysin were studied using numerous enzyme inhibitors and molecular simulations, and the results demonstrated that thermolysin preferred to catalyze substrates containing hydrophobic side chains.25,27 Advances in the development of short peptide self-assemblies have led to the widespread use of thermolysin in peptide chemistry, and there are increasing numbers of investigations that utilize the enzyme to investigate the self-assembly behavior of short peptides, and also for the preparation of a series of short peptides with different morphologies and functions.5,28,29 For example, Ulijn’s and co-workers5 have utilized thermolysin in the catalysis of condensation reactions of amino acid derivatives. Fmoc-triphenylalanine (Fmoc-FFF) synthesized from Fmoc-F and FF via catalytic action of thermolysin could self-assemble to form a fiber-entangled hydrogel. Apart from synthesis, thermolysin could also be employed to realize phase transition from hydrogel to solution through the hydrolysis of peptide bonds of Fmoc–tyrosine–leucine (Fmoc-YL).29 Nonetheless, up to now, there is no relevant study on the effect of a peptide sequence on the role of thermolysin due to its complexity of binding to different peptides. Hence, exploring the binding capacity and catalytic action of thermolysin for different peptide sequences would promote its development in the field of short peptide self-assembly. Herein, we designed a series of amphiphilic dipeptide derivatives through thermolysin catalysis to explore its role in the process. We anticipate that elucidating the mechanism of thermolysin action would define the design principle of peptide materials and their functionalities significantly. Experiments Methods Materials and chemicals Fmoc-dipeptides (YL, YA, YD, TL, LL, where Y is tyrosine, L is leucine, A is alanine, D is aspartic acid, and T is threonine) were synthesized by GL Biochem Company, Shanghai, China. Fmoc-protected amino acids (L, Y, T, S, where S is serine) and L–OMe/L–NH2/Y–NH2 were purchased from Bachem, Tübinge, Germany. Amino acid (L–COOH, A–COOH, and D–COOH) and lyophilized thermolysin were purchased from Sigma, Mainland, China. A Milli-Q Plus water purification system (resistivity >18.2 MΩ·cm), MilliporeSigma, Shanghai, China, was used to produce H2O used in the research. NaH2PO4·H2O (94 mg, Sigma) and Na2HPO4·7H2O (2.5 mg; Sigma) were dissolved in 100 mL H2O to prepare 100 mM sodium phosphate buffer saline (PBS, pH 8.0). The mobile phase solution was composed of acetonitrile (J&K Chemicals, Shanghai, China) water (50∶50) with the addition of 0.1% trifluoroacetic acid (J&K Chemicals). Sample preparation For the enzyme-triggered hydrolysis reaction, Fmoc-dipeptides (20 mM) and lyophilized thermolysin (1 mg mL−1) were dispersed in 100 mM PBS in a glass vial. Then, the mixture was vortex-mixed for 20 s and sonicated for 1 min to ensure complete dissolution. After that, the reaction of the solution mixture was allowed to proceed for up to 120 h at room temperature. In the case of the enzymic-catalyzed condensation, the precursors of each Fmoc-X (20 mM) and L–OMe/L–NH2/Y–NH2 (80 mM) at molar ratios of 1∶1/1∶2/1∶4 were suspended in 100 mM PBS in a glass vial. The mixture was vortex-mixed for 20 s and sonicated for 1 min to ensure dissolution. Then, 1 mg mL−1 lyophilized thermolysin powder was added to the solution. Samples were incubated at room temperature for up to 120 h. High-performance liquid chromatography An Agilent 1200 high-performance liquid chromatography (HPLC) system (Beijing, China) equipped with an Agilent Eclipse XDB-C18 column of 150 mm length, 4.6 mm internal diameter, and 5 µm particle sizes were used to quantify the percentage conversion of the enzyme-triggered condensation reaction at a flow rate of 0.8 mL min−1. For the sample preparation for HPLC, 10 µL of the reacted mixture was diluted into a 1 mL sample by the mobile phase solution. The conversion percentage of each sample was calculated from integrated peak areas at 265 nm UV detection. The experiment was repeated at least three times, and the average data was presented. Molecule docking Molecular docking studies between the Fmoc-dipeptide fragments (Fmoc-YL, Fmoc-YA, Fmoc-YD, Fmoc-TL, and Fmoc-LL) and thermolysin were performed using Autodock4.2.6.30 Initial structures of the Fmoc-dipeptide fragments were modeled with the aid of GaussView 6.0 ( https://wafiapps.net/gaussview). The crystal structure of thermolysin binding with amyloid-β peptide fragments (PDB entry: 5ONR)31 was used as the receptor structure and prepared for docking. The cocrystallized amyloid-β peptide fragment (Ile31-Gly33 fragment) was used to define the active sites for docking. A grid of 60, 60, and 60 points in x, y, and z direction with a grid spacing of 0.375 Å was created at the binding site for the docking between target dipeptide and thermolysin, respectively. The Lamarckian genetic algorithm was used to generate the binding poses between the dipeptides and thermolysin.32 Exactly 200 docking runs for every dipeptide bound to thermolysin were calculated with the rotation of all non-ring torsion angles. Then, the docking conformations with the lowest binding energy were selected for analysis, respectively. Molecular dynamic simulation Molecular dynamic (MD) simulations of the enzyme-peptide docked complexes were carried out using GROMACS5.1.4.33 Charmm36 force field34 was used for the simulations. The CGenff server ( https://cgenff.umaryland.edu/) was used to generate topology files for Fmoc-dipeptides. The systems were solvated in a cubic box with TIP3P water model, and periodic boundary conditions were used. Exactly 0.15 M NaCl was added further in order to neutralize the systems. The bond lengths were constrained using the Linear constraint solver simulations (LINCS) algorithm,35 and the Particle-mesh Ewald (PME) method36 was utilized to compute long-range electrostatics. Energy minimization of the docked complexes was carried out using the steepest descent algorithm. After energy minimization, the canonical ensemble, particle number, volume, temperature (NVT), and the isothermal–isobaric ensemble, particle number, pressure, temperature (NPT) equilibrations were performed on each system until they reached room temperature (300 K) and water density. Finally, the production run MD was carried out for 100 ns with a time-step of 2 fs. Coordinates of the docked complexes were saved every 10 ps and used further for binding free energy analysis. The molecular mechanics Poisson–Boltzmann surface area (MM-PBSA)37 approach was employed to estimate the binding free energy of protein–dipeptides interaction. For this purpose, “g_mmpbsa” tool38 was used, which calculated potential energy and the free energy of solvation for the molecular mechanics, excluding the entropy calculations. The MM-PBSA calculations were performed by taking 2000 snapshots from the last 20 ns of trajectories of each simulation system. FT-IR spectroscopy Each of the peptide-based gel samples generated from the thermolysin condensation reaction was dissolved in 100 mM PBS to a final concentration of 20 mM. Then 10 μL of the samples were added to CaF2 windows (thickness 1.5 mm) and dried into a film using a vacuum pump. The FT-IR spectra were obtained through Bruker Vertex spectrometer (80/80v FT-IR; Beijing, China) with a resolution of 1 cm−1, and corrected for absorption from a PBS blank sample. Circular dichroism To measure the secondary structure of the different peptides-based hydrogels films, 20 mM samples obtained by initial preparation of stock solutions in 100 mM PBS, and taking 10 μL aliquots to dissolve in 1 mL D2O to the desired final concentration, ignoring the role of salt. Subsequently, the samples were pipetted into a 0.2 mm path-length quartz cell, and the CD spectra were measured on a J-810 circular dichroism (CD) spectrometer (JASCO, Tokyo, Japan) with wavelength (λ) scanning from 260 to 190 nm in the conditions of a scanning speed of 100 nm/min, 4 s integrations, a 1 nm bandwidth, and a data pitch of 0.1 nm due to the dynamic nature of the system. Atomic force microscopy The morphologies of the hydrogel samples were displayed by atomic force microscopy (AFM; FastScan Bio, Bruker, Beijing, China); the samples were added on clear mica sheets surface and air-dried in a dust-free vacuum at room temperature. The samples were slightly washed by ultrapure water to remove the salt and then dried in a vacuum. The images of samples were obtained by scanning the mica sheet surface in the air under ambient conditions using AFM in tapping mode. The specifications of the AFM images obtained were as follows: 512 × 512 pixels resolution and produced topographic morphology in which the brightness of features increased as a function of height; images were first-order flattened using the Nanoscope Analysis software (version 1.5; Bruker). Scanning electron microscopy The hydrogel samples (20 μL) were dropped on a silicon wafer and dried in a vacuum at room temperature. Then, they were washed 3 times to eliminate the influence of salt. Before image acquisition with an S-4800 (10 kV voltage; HITACHI, Tokyo, Japan) instrument, the samples were placed on a silicon wafer and sputtered with platinum to increase conductivity. Results and Discussion Thermolysin-triggered hydrolysis of Fmoc-dipeptide derivatives We were inspired from our previous work in which we showed that Fmoc-YL could be hydrolyzed into Fmoc-Y and L–COOH in the presence of thermolysin.29 Thus, we sought to continue investigating the actions of thermolysin during catalysis by designing a series of Fmoc-dipeptide substrates, changing the peptide sequences by varying the hydrophobicity of the amino acid compositions and the carboxyl terminus. Interestingly, thermolysin behaved significantly different under the designed substrates, including its catalytic actions on hydrolysis or condensation, as well as its catalytic efficiency. The hydrophobicity of the substrates is one of the major driving forces that determined substrate binding with an active site of the enzyme. Therefore, at a molecular level, we designed the terminal amino acid of peptide sequences to obtain varied substrate groups, including Fmoc-YL, Fmoc-YA, and Fmoc-YD, with decreasing hydrophobicity to investigate the enzymic binding and catalytic efficiency (Figure 1a). The reaction systems, each comprising the Fmoc-dipeptides (20 mM) was prepared in vials, and thermolysin (1 mg mL−1) was added to each vial.17 Then, the enzymatic hydrolysis reaction of each sample was allowed to proceed for 120 h. A visible gel–sol transition of Fmoc-YL was apparent, as shown in Figure 1b, and the mechanical strength of the Fmoc-YL reaction showed dramatic difference ( Supporting Information Figure S1a), revealed by an increase in flexibility and stability when compared with the other Fmoc-dipeptide fabrications. On the contrary, the phase transition of Fmoc-YA was not observable; however, its mechanical properties showed a noticeable change in the presence of thermolysin (Figure 1b and Supporting Information Figure S1b). Fmoc-YD remained as a viscous liquid the whole time, and the rheological results showed little differences (Figure 1b and Supporting Information Figure S1c). The conversion efficiency of Fmoc-dipeptide was determined by reversed-phase high-pressure liquid chromatography (RP-HPLC). As shown in Figure 1c and d, Fmoc-YL, Fmoc-YA, and Fmoc-YD were hydrolyzed into Fmoc-Y with a conversion percentage of 97.7%, 25.2%, and 2.7%, respectively. The hydrolysis of Fmoc-YL was significantly faster and more efficient than Fmoc-YA and Fmoc-YD, which was probably due to its higher hydrophobicity. Figure 1 | Thermolysin-triggered hydrolysis of Fmoc-dipeptide derivatives with different terminal amino acid residues. (a) Schematic representation of substrate molecular design of Fmoc-dipeptide. (b) Photographs of corresponding Fmoc-dipeptide materials before and after the addition of thermolysin. (c) HPLC traces after addition of thermolysin for 120 h. (d) Time course of the hydrolysis of the Fmoc-dipeptide to the monopeptide, as shown by percentage conversion by HPLC analysis. (e) Binding mode and binding energy of each Fmoc-dipeptide at the thermolysin catalytic site, calculated by computational simulation. Conditions: [each Fmoc-YX] = 20 mM and [thermolysin] = 1 mg mL−1. Download figure Download PowerPoint We performed a computational simulation to gain an understanding of the binding mode of thermolysin and the Fmoc-dipeptides, using PDB entry: 5ONR (catalytic site with Zn2+ binding to peptide sequences), and conducted through molecular docking strategies.26 As shown in Figure 1e, the Fmoc-dipeptides with different amino acid terminus (YL, YA, and YD) revealed highly conserved overall binding mode at the active site of thermolysin. In these docking structure models, the fluoren-group of Fmoc-dipeptide was positioned in an outward orientation toward the water-accessible hydrophilic environment. The succeeding attached carbonyl group of Fmoc-dipeptides was hydrogen-bonded to the backbone carbonyl oxygen of F114. Meanwhile, the first assigned tyrosine residue (YX) displayed a strong interaction by forming hydrogen bonds to Y110, Y157 and H231, and π–π interaction with F114. As for the second residues of Fmoc-dipeptides, the main-chain atoms, viz, leucine (L), alanine (A), or aspartic acid (D) hydrogen-bonded to N112 and R203 in these modes (Figure 1e). However, there were significant differences observed, such as chemically deviating side chains and the surrounding interaction pattern. For example, for the binding structure of thermolysin/Fmoc-YL, which displayed the highest decomposition rate, the isopropyl of Leucine extended into a deep and entirely hydrophobic cavity (V139, V192, L133, I188, and L202) and formed multiple hydrophobic van der Waals interactions with the cavity-lining residues (Figure 1e–left). However, when this group was replaced by a more hydrophilic methyl group of Fmoc-YA, the decomposition rate declined dramatically, which might be attributable to a decrease in hydrophobicity of the side-chain functional group to induce the decline in the binding affinity (Figure 1e–middle). This finding was supported further by an observation that when the more hydrophilic polar group was introduced to the hydrophobic pocket as with the carboxylic group of Fmoc-YD, thermolysin lost most of its catalytic efficiency for this modified substrate, likely due to the incompatibility between the hydrophilic carboxylic group and the hydrophobic pocket environment (Figure 1e–right). We calculated further the docked complexes of Fmoc-YL, Fmoc-YA, and Fmoc-YD with thermolysin using MD simulation. For all three systems listed, the root-mean-square deviation (RMSD) of the thermolysin Cα atoms were calculated during 100 ns to evaluate the stability of the MD simulations ( Supporting Information Figure S2a). The average RMSD values for thermolysin were about 2 Å, and at the last 50 ns, all the RMSD complexes values reached an equilibrium state, indicating the simulation systems were stable. Also, by comparing the initial and the final conformations of the MD simulations, we found that the orientations of Fmoc-YL, Fmoc-YA, and Fmoc-YD were well conserved in the complexes ( Supporting Information Figure S2b–d). Further, to study the interaction between Fmoc-dipeptides and thermolysin, the binding free energy calculations were carried out with the MM-PBSA program. For each complex, van der Waals (Evdw), electrostatic (Eelec), polar solvation (Gpolar), nonpolar solvation energy (Gnonpolar), and binding free energy (ΔGbinding) were calculated ( Supporting Information Table S1). The estimated binding energy of Fmoc-YL, Fmoc-YA, and Fmoc-YD to thermolysin was −91.515, −30.126, and −20.090 KJ/mol, respectively. It could be interpreted that hydrophobicity of terminal residues would affect the binding capability of peptide sequences at the catalytic site of thermolysin, and then enhanced the catalytic efficiency.30,39 The analysis of the correlation between hydrophobicity and binding energy was also highly consistent with the hydrolysis results from the HPLC evaluation. For comparison, we designed the second group of substrates by changing the first amino acid residues linked to the Fmoc group to generate Fmoc-dipeptides, including Fmoc-YL, Fmoc-LL, and Fmoc-TL (where T is threonine) (Figure 2a). After 120 h hydrolysis reactions, gel–sol transition was observable with Fmoc-YL and Fmoc-LL (Figure 2b), and their mechanical strength displayed a dramatic decrease ( Supporting Information Figure S1a and d). However, Fmoc-TL remained in the solution state, and its condition did not change in the presence of thermolysin. Also, our rheological data revealed that the Fmoc-TL was a liquid with low viscosity due to the viscous modulus being higher than the elastic modulus (G″ > G′), as shown in Supporting Information Figure S1e. The HPLC profiles in Figure 2c display thermolysin-cleaved peptide bonds of Fmoc-dipeptides to Fmoc-X and L–COOH. Fmoc-YL and Fmoc-TL showed 97.7% and 97.2% of conversion at 120 h, respectively, whereas Fmoc-LL had a lower conversion rate of 44.6%/120 h (Figure 2d). The binding modes of Fmoc-dipeptides (Fmoc-XL) and thermolysin calculated from the computational simulation shown in Figure 2e. We demonstrated that both tyrosine (Y) and threonine (T) formed hydrogen bonds to adjacent residues of thermolysin, which could stabilize the binding conformation at the active site (Figure 2e–left and middle). Tyrosine residue of Fmoc-YL formed π–π conjugation with F114 and hydrogen bonds with Y110 to stabilize the binding conformation. Threonine residue of Fmoc-TL could form hydrogen bonds with N112. However, there was no apparent beneficial interaction observed between leucine residue of Fmoc-LL and surrounding hydrophilic residues, which could be the reason for their lower decomposition rate compared with Fmoc-YL and Fmoc-TL. The MD simulation results show that these docked complexes are stable during 100 ns MD simulations ( Supporting Information Figure S2a, b, e, and f). The trajectories obtained from MD simulations were used for computing the binding free energy of the Fmoc-dipeptides in the docked complexes. The binding energy of Fmoc-YL, Fmoc-TL, and Fmoc-LL to thermolysin was −91.515, −95.597, and −43.517 KJ/mol, respectively ( Supporting Information Table S1), consistent with the efficiency of the decomposition catalysis. Consequently, we demonstrated that thermolysin preferred a specific substrate pattern with a hydrophilic amino acid in the first residue and hydrophobic amino acid in the second residue. Figure 2 | Thermolysin-triggered hydrolysis of Fmoc-dipeptide derivatives with different amino acid residues linked to the Fmoc group. (a) Schematic representation of the molecular design of Fmoc-dipeptide substrates. (b) Photographs of corresponding Fmoc-dipeptide materials before and after the addition of thermolysin. (c) HPLC traces of hydrolysis products obtained after the addition of thermolysin to the Fmoc-dipeptides for 120 h. (d) Time course of the rate of conversion (%/120 h) of Fmoc-dipeptides to the monopeptides by HPLC. (e) The binding energy of each Fmoc-dipeptide to the catalytic site of thermolysin calculated form computational simulation. Conditions: [each Fmoc-XL] = 20 mM and [thermolysin] = 1 mg mL−1. Download figure Download PowerPoint Formation of Fmoc-dipeptide amphiphiles by thermolysin-catalyzed condensation reaction On the contrary, by using Fmoc-Y with X–COOH (where “X” denotes L, A, and D), or using Fmoc-X (where “X” denotes Y, L, and T), when L–COOH was presented as a substrate for thermolysin, the enzyme was unable to catalyze the condensation reaction to produce Fmoc-dipeptide derivatives ( Supporting Information Figure S3a and b). According to recent reports, an amino acid amide derivative exhibited a more hydrophobic nature,40 and thermolysin could couple a carboxy component to an amino acid amide at about one order of magnitude faster than to the corresponding amino acid.41 Therefore, we designed