Bioinspired enzymatic compartments constructed by spatiotemporally confined in situ self-assembly of catalytic peptide

Enzymatic compartments, inspired by cell compartmentalization, which bring enzymes and substrates together in confined environments, are of particular interest in ensuring the enhanced catalytic efficiency and increased lifetime of encapsulated enzymes. Herein, we constructed bioinspired enzymatic compartments (TPE-Q18H@GPs) with semi-permeability by spatiotemporally controllable self-assembly of catalytic peptide TPE-Q18H in hollow porous glucan particles (GPs), allowing substrates and products to pass in/out freely, while enzymatic aggregations were retained. Due to the enrichment of substrates and synergistic effect of catalytic nanofibers formed in the confined environment, the enzymatic compartments exhibited stronger substrate binding affinity and over two-fold enhancement of second-order kinetic constant (kcat/Km) compared to TPE-Q18H nanofibers in disperse system. Moreover, GPs enabled the compartments sufficient stability against perturbation conditions, such as high temperature and degradation. This work opens an intriguing avenue to construct enzymatic compartments using porous biomass materials and has fundamental implications for constructing artificial organelles and even artificial cells.

Authors was packaging catalytic peptides in glucan cage particles (GPs) to enhance catalytic activity plus endurance and stability. This reviewer has general questions about this approach. Of course, the protection will be beneficial but the conformation is probably more important. When peptides were assembled in GPs, how do authors know that the peptide assemblies are in the best conformation? Is the confinement really paid off for merely 2-fold enhancement, which I am guessing that it is still not practical as compared to natural enzymes. This issue needs to be addressed otherwise this manuscript could be regarded as a low impact work. In general, this manuscript is sloppy. There are many typos and mis-labeling figures. (e.g., Fig 2-f in page 8 is really Fig. 2-h). The illustration in Fig. 1 is misleading. There is no evidence that peptides were assembled in the fiber form inside GP (I see no structural difference between Fig. 2-e and 2-f) and pore size of GPs is unknown and not investigated (although a dye molecule seems to be in and out). Nanofibers are in the size range of > 1um, which can really fit in the 3 um diameter of GP? Why such as large fibers are not visible in TEM images? How many % of the incubated peptides were taken by GP? What is the uptake kinetics of Rhodamine and releasing kinetics? Protease was used to examine the protective effect via peptide degradation but can protease penetrate inside GPs? Pores are that large on GPs?
Reviewer #2 (Remarks to the Author): This paper describes the entrapment of a self-assembled peptide with hydrolytic catalytic activity in the pores of glucan microparticles. The system design is based on previous literature about related peptides with a pendant histidine unit that becomes catalytic active, hydrolase-like, upon selfassembly into nanofibers. The use of glucan provides some advantages to this system: a) The self-assembled catalyst is supported into an inert matrix, and consequently, its separation from the reaction media by centrifugation is possible b) Notably, the nanofiber catalysts are protected from degradation by temperature or tripsin upon incorporation in the glucan particles. c) The catalytic performance of the entrapped nanofibers is somewhat improved.
Overall, the research shows novelty and provides interesting results. However, some essential experimental details are missing, and some issues should be clarified before publication. 1) The term "Spatio-temporal control" is used in the title and the main text. Clarify the relevance of this issue compared to other cases where such control is not achieved.
2) The text states, "was dissolved in solution (acetonitrile/water=3:7) to get a dispersive solution". What is a dispersive solution? Are the authors talking about dispersion? Is it a suspension or a colloid?
3) The encapsulation efficiency is claimed to be over 90%. This result seems doubtful. The loading is performed using a "dispersion" of the peptide in acetonitrile/water, and the loading efficiency is determined after centrifugation by UV-Vis study of the supernatant. However, the centrifugation probably also removes "dispersed"/precipitated peptides outside the glucan particles. Therefore, a control experiment is required in the absence of glucan. 4) Calculated errors should be provided for the kinetic constants. 5) How can the thermal stability induced by entrapment into glucan be explained? 6) The hydrolysis rate of PNPA in the studied medium without the peptide should be indicated numerically 7) In the experimental section: "The lyophilized powder was resuspended with 1× PBS to obtain TPE-Q18H@GPs" What volume of PBS? 8) "The UV-vis spectra of TPE-Q18H in different concentrations were obtained on a UV-vis spectrophotometer... " What solvent was used?
9) The kinetic measurements seem to be flawed by this ambiguous description: "After 1 minute, the reaction mixture was fast centrifuged" Are the authors reporting kinetic data for a reaction time of one minute that requires centrifugation before measurement of absorbance?! What is "fast" centrifugation? In situ measurements of the hydrolysis product seem mandatory in this case or the use of longer reaction times.
Reviewer #3 (Remarks to the Author): The paper reports spatiotemporally confined encapsulation of self-assembled fibers into hollow glucan particles to construct enzymatic compartments. The resulting biomimetic system shows improved catalytic efficiency and stability against perturbation conditions. The results are well addressed and could be interesting to the broad audience. Overall, this is an interesting concept, and it can be practically useful. This manuscript offers a brand-new perspective inspired by the natural structure of cell compartments, and I recommend acceptance of this manuscript after the following minor issues are addressed. 1. The authors claimed that one advantage of such strategy was that the artificial catalysts could be retained within the cavities of the GPs, while the free biomacromolecules are able to enter or leave the cavities freely, and they used rhodamine dye as an indicator to show this semi-permeability. Yet, the biomacromolecule model was not disclosed in the manuscript. 2. From the sequence design of TPE-Q18H, the KRKR sequence was chosen to increase the solubility of TPE moiety, while, as I know, in another published work the acidic amino acid residues were ever used to make the TPE-hybrid peptide soluble in water. So, did the authors have other considerations about the sequence design to choose such an alternative way. 3. The authors monitored the catalytic activities in several different conditions individually. The blanks were used for these conditions should be referred in detail, separately. 4. The scale bars in Figure 2 should be labelled more clearly, especially for the AFM image.  Based on the results above, we believe that the peptides assembled inside GPs and the structural features of these assemblies were consistent with those of TPE-Q18H in the GP-free state.

Q2.
Is the confinement really paid off for merely 2-fold enhancement, which I am guessing that it is still not practical as compared to natural enzymes?
Response: Thank you very much for your concern about our design. It is still the challenge for most studies of artificial enzymes that the catalytic efficiencies of artificial enzymes are in different order of magnitudes compared to those of natural enzymes. While, here we just wanted to claim a new strategy of constructing bioinspired compartments by immobilizing artificial enzymes into natural porous materials, to improve the activities and especially the stability against perturbation conditions with easier recycling, which avoided the extensive consumption and waste of natural products.
Moreover, based on current strategy, our ongoing work is proposing an approach to fabricate protocells by in situ self-assembly of multiple natural enzymes to carry out cascade reactions within GPs. To explore the universality of our strategy, we have immobilized a modified green fluorescence protein with a designed peptide sequence capable of self-assembly (EGFP-peptide) in GPs. This model EGFP-peptide can self-assembly inside GPs like the way of TPE-Q18H. Confocal laser scanning microscopy (CLSM) was used to observe the green fluorescence of EGFP-peptide inside GPs ( Figure R2a), indicating that protein could be retained within GPs by salt-triggered peptide assembly. Almost no fluorescence was recorded in the supernatant solution and the encapsulation efficiency is 95.51% in the fluorescence measurement ( Figure R2b). It lays a foundation for the subsequent research to immobilize natural enzymes into GPs. Hence, we believe the robust and recyclable properties of our work would pay off, and the spatiotemporally confined in situ self-assembly strategy has significant potential in construction of unnatural organelles.

Response:
We really appreciate reviewer's comments and the typos and mislabeling mistakes have been revised, which has been highlighted in our revised manuscript.

Q4. The pore size of GPs is unknown and not investigated (although a dye molecule seems to be in and out).
Response: Thank you for your valuable suggestion. In our submission, we could check the pore size of GPs under SEM with high resolution in Figure 2c, and furthermore, in our revised manuscript we supplementary measured the surface properties of GPs by nitrogen adsorption isotherms ( Figure R3).
The adsorption-desorption curve of GPs is a type II isotherm, and the BET surface area and average pore diameter of GPs were 12.84 m 2 /g and 4.14 nm respectively, indicating that not only small molecules but also biomacromolecules can pass though the pore easily. This result has been placed and highlighted in our revised manuscript as shown in Figure 2d. About the size issue, we claimed that TPE-Q18H was not at an assembled state when peptide entered into GPs, since we did not add any salt into the system to trigger the assembly process at this time point.

Q7. What is the uptake kinetics of Rhodamine and releasing kinetics?
Response: Thanks for your valuable suggestion. We supplement the uptake and releasing kinetics of Rhodamine in GPs (Supplementary Figure 9 in our revised supporting information). Rhodamine and empty GPs incubated for 10 minutes. One group was filtered and measured leachate's fluorescence to measure the uptake. Another group was centrifuged and resuspended 5 times to measure the releasing of Rhodamine from GPs. Free Rhodamine solution was prepared for calculations.
Calculate the encapsulation and releasing efficiency using the formulas below: Uptake efficiency= (FL intensity of free Rhodamine -FL intensity of leachate) ⁄ fluorescence of free Rhodamine × 100% (FL means fluorescent intensity) Releasing efficiency= FL intensity of supernatant ⁄ (FL intensity of free Rhodamine -FL intensity of leachate) × 100% In particular, the calculation of first centrifugation releasing is the FL intensity of supernatant of the first centrifugation minus the FL intensity of the leachate.
As shown in Figure R4, 50% Rhodamine was released from GPs in the first centrifugation, and all Rhodamine was released after 5 centrifugations. Response: Thank you for your significant reminder. The pore average size was characterized to be around 4.14 nm, so theoretically Trypsin (maximum diameter 4.2 nm) could enter into GPs. To investigate whether the protease Trypsin (24 kDa) can penetrate GPs and TPE-Q18H@GPs, we supplement experiments using a green fluorescent protein (PSP2, 27 kDa) as a model, because trypsin and PSP2 share similar sizes. PSP2 was incubated with GPs or Q13H@GPs (Q13H replaces TPE-Q18H to eliminate the effect of TPE on fluorescence of PSP2 protein) solution for 10 min, and CLSM was used to detect the fluorescent intensity of PSP2 in GPs and Q13H@GPs. The green fluorescence of PSP2 was observed in GPs ( Figure R5a) and Q13H@GPs ( Figure R5b). Moreover, the average intensity in Q13H@GPs was lower than that in GPs because peptide Q13H nanofibers occupied some space ( Figure R5c). With this result, we can conclude that Trypsin can penetrate into GPs, and we speculate that the limited internal space of GP makes the conformation of peptide assemblies relatively fixed, so the structural stability was improved. Response: Thank you for pointing out the inaccurate statement in our manuscript. Peptide TPE-Q18H dissolved very well in solution (acetonitrile/water=3:7), so it is actually a clear solution and the peptide was in the monomer state. The expression "was dissolved in solution (acetonitrile/water=3:7) to get a dispersive solution" has been changed into "was dissolved in solution (acetonitrile/water=3:7) to get a clear solution and the peptide was in the monomer state" in the revised manuscript to avoid this misunderstanding.

Q3.
The encapsulation efficiency is claimed to be over 90%. This result seems doubtful. The loading is performed using a "dispersion" of the peptide in acetonitrile/water, and the loading efficiency is determined after centrifugation by UV-Vis study of the supernatant. However, the centrifugation probably also removes "dispersed"/precipitated peptides outside the glucan particles. Therefore, a control experiment is required in the absence of glucan.
Response: As corrected in the Q2, peptide TPE-Q18H was dissolved very well in solution (acetonitrile/water=3:7) and the peptide monomers entered GPs by capillarity owing to the porous structure of GPs. The assembly of TPE-Q18H occurred in situ in the cavity of GPs by the addition of 1×PBS buffer and the assembled nanofibers within GPs could not escape from GPs because of their micrometer length size, even during centrifugation, while the monomers and the nanofibers out of GPs would be separated into the supernatant

Q4. Calculated errors should be provided for the kinetic constants.
Response: Thanks for your comments. The calculated errors have been added to Figure 5d and the data has been updated in the description part in our revised manuscript.

Q5. How can the thermal stability induced by entrapment into glucan be explained?
Response: Thanks a lot for your concern. Normally, no matter natural enzymes or artificial enzymes lose activities at high temperature mainly due to the change of their best conformation. The limited internal space of GPs made the conformation of peptide assemblies relatively fixed, so the thermal stability was improved.

Q6. The hydrolysis rate of pNPA in the studied medium without the peptide should be indica
Response: The self-hydrolysis rate of pNPA (1 mM) was 14.79 μM/min in 1 PBS. Meanwhile, in the calculations of the initial catalytic rate of TPE-Q18H or TPE-Q18H@GPs, the self-hydrolysis of pNPA was also subtracted to ensure that the activities were purely from TPE-Q18H or TPE-Q18H@GPs.

Q7. In the experimental section: "The lyophilized powder was resuspended with 1× PBS to obtain TPE-Q18H@GPs" What volume of PBS?
Response: The volume of 1× PBS was 1 mL. In the experimental section, we have changed the description "The lyophilized powder was resuspended with 1 mL 1× PBS to obtain a well-dispersed TPE-Q18H@GPs particles solution and stocked at 4 °C. When measuring the enzyme activities, add another 4 mL 1× PBS to get a diluted solution (TPE-Q18H: 33.42 µM; GPs: 1 mg/mL)".

Q8. "The UV-vis spectra of TPE-Q18H in different concentrations were obtained on a UV-vis spectrophotometer... "What solvent was used?
Response: The solvent was 1× PBS. In the experimental section, we have changed the description into "The UV-vis spectra of TPE-Q18H 1× PBS solution in different concentrations was obtained on a UV-vis spectrophotometer from 190 nm to 450 nm to make a standard curve".

Q9.
The kinetic measurements seem to be flawed by this ambiguous description: "After 1 minute, the reaction mixture was fast centrifuged" Are the authors reporting kinetic data for a reaction time of one minute that requires centrifugation before measurement of absorbance? What is "fast" centrifugation?
In situ measurements of the hydrolysis product seem mandatory in this case or the use of longer reaction times.

Response:
We are appreciative of your suggestion. we did not measure the hydrolysis product in situ because the 3-5 μm diameter of GPs will interfere with absorbance, leading to large errors. On the other hand, centrifugation allowed the products to escape from GPs. The fast centrifugation is 5000 rpm using Mini Centrifuge (SCILOGEX S1010E) for 5 seconds. The ambiguous description has been revised in the manuscript.   Response: Thanks for your significant reminder. In the parts of determining catalytic kinetics and thermostability, the blank was the self-hydrolysis rate of pNPA in 1 PBS, and the calculations of initial catalytic rate subtracted the self-hydrolysis rate of pNPA in 1 PBS to ensure that the activities were purely from TPE-Q18H or TPE-Q18H@GPs.

Q1. The authors claimed that one advantage of such strategy was that the artificial catalysts could
In the part of protecting from protease digestion, the blank was the self-hydrolysis rate of pNPA in 1PBS in trypsin untreated group, and the calculations of the initial catalytic rate were described as above. In the trypsin treated groups, for the trypsin could also hydrolyze pNPA, the calculations of the initial catalytic rate of TPE-Q18H with trypsin needed to subtract the hydrolysis rate of pNPA by trypsin to ensure that the activities were purely from TPE-Q18H. The calculations of the initial catalytic rate of TPE-Q18H@GPs with trypsin subtracted the hydrolysis rate of pNPA in trypsin with GPs to ensure that the activities were purely from TPE-Q18H@GPs. We also added the description in detail in the revised manuscript. Figure 2 should be labeled more clearly, especially for the AFM image.

Response:
We thank the reviewer for the careful review and kind suggestion. We have clearly labeled the scale bars in Figure 2, checked and modified the other bars in the revised manuscript and supporting information. There is significant improvement on this manuscript with more data sets. However, there are still a few questions needed to be addressed with new data. 1. Now pore size is identified and release kinetics was also studied and it is very fast. Then, it comes a question that how long the peptide assembly takes and what time point it shows enzymatic activity. These numbers should be compared to justify all of authors' hypothesis. 2. Are Figure R2 and its explanation in revised manuscript? It should be in to support a tremendous claim like these peptide-incorporating GPs are potentially developed as unnatural organelles. Ideally, authors should incorporate another catalytic peptide in GP and simultaneously assemble both and analyze how they can respectively trigger two distinct enzymatic reactions. There are a plenty of papers that used other cage-like polymer or protein shells to incorporate multiple enzymes and thus authors' date could be compared and commented with respect to them.

References
Reviewer #2 (Remarks to the Author): The remarks raised in my first review have been properly answered in this revised version.
Reviewer #3 (Remarks to the Author): The athuors addresssed all issues we concerned.I recommend to publish.

Q1. Now pore size is identified and release kinetics was also studied and it is very fast. Then, it comes a question that how long the peptide assembly takes and what time point it shows enzymatic activity. These numbers should be compared to justify all of authors' hypothesis?
Response: Thank you very much for your concern about our design. To explore how long the peptide assembly takes and what time point it shows enzymatic activity, we first monitored the fluorescence intensity change of Th T that represents the conformation change of peptide TPE-Q18H, and then measured the catalytic activities of TPE-Q18H and TPE-Q18H@GPs before and after adding salt respectively. As shown in Figure RR1 a, for the free peptide TPE-Q18H, the fluorescence intensity of Th   Figure R2 and  Response: Thanks for your comment. We have followed your suggestion about incorporating another catalytic peptide in GPs, and here we designed and constructed bifunctional enzymatic compartments with hydrolase and peroxidase-like activities by immobilizing hemin, the key part of the peroxidase catalytic center, onto the surface of TPE-Q18H nanofibers with a peptide/hemin ratio of 10/1 ( Figure   RR2 a). The hemin bound nanofibers (TPE-Q18H/Hemin10) were characterized using UV-Vis spectroscopy (Figure RR2 b). The free hemin chloride displayed a Soret peak at 384 nm along with a shoulder at 365 nm, indicating the presence of dimeric (μ-oxo bihemin) along with some monomeric hemin hydroxide (haematin). 1 In contrast, hemin bound nanofibers (TPE-Q18H/Hemin10) displayed a broad Soret band at 400 nm, similar to hemin in aqueous micelle solutions and proteins, suggesting monomeric hemin chloride. 2,3 The absorption peak at 330 nm was from TPE. We then explored the peroxidase-like activity of hemin using 3,3',5,5'-tetramethylbenzidine (TMB) and H2O2 as the substrates.

Q2. Are
Peroxidases facilitated the oxidation of a colorless TMB to a blue product with maximum absorbance at 652 nm in the presence of H2O2. The activity was characterized by monitoring the absorbance at 652 nm of reaction supernatants. As shown in Figure RR2 c, TPE-Q18H showed no catalytic activity but TPE-Q18H/Hemin10 exhibited a higher activity than that of free hemin. Similarly, GPs and TPE-Q18H@GPs were proved to have no contribution to the catalytic activity, while TPE-Q18H/Hemin10@GPs carried out the oxidation redox reaction. The absorbance peak of 652 nm was not detected in Hemin@GPs, which may be due to too few products and not completely escaping from the GPs.
In addition, the hydrolytic activities were measured using pNPA as the substrate (Figure RR2 d). GPs, Hemin, Hemin@GPs did not show hydrolytic activity. TPE-Q18H, TPE-Q18H@GPs and TPE-Q18H/Hemin10@GPs exhibited dramatic hydrolytic activity towards pNPA, which was in consistent with previous results.
In summary, we have successfully constructed a dual aritificial enzyme system in porous GPs, which exhibits both hydrolase and peroxidase-like activities. As suggested by editor, we prefer to put Figure   R2, Figure RR2 and the related discussion into the transparent peer review scheme.