A sulfonated mesoporous silica nanoparticle for enzyme protection against denaturants and controlled release under reducing conditions

Journal Pre-proofs A Sulfonated Mesoporous Silica Nanoparticle for Enzyme Protection against Denaturants and Controlled Release under Reducing Conditions Hui Li, Yanxiong Pan, Jasmin Farmakes, Feng Xiao, Guodong Liu, Bingcan Chen, Xiao Zhu, Jiajia Rao, Zhongyu Yang PII: DOI: Reference:

S0021-9797(19)30964-6 https://doi.org/10.1016/j.jcis.2019.08.063 YJCIS 25319

To appear in:

Journal of Colloid and Interface Science

Received Date: Revised Date: Accepted Date:

26 June 2019 15 August 2019 16 August 2019

Please cite this article as: H. Li, Y. Pan, J. Farmakes, F. Xiao, G. Liu, B. Chen, X. Zhu, J. Rao, Z. Yang, A Sulfonated Mesoporous Silica Nanoparticle for Enzyme Protection against Denaturants and Controlled Release under Reducing Conditions, Journal of Colloid and Interface Science (2019), doi: https://doi.org/10.1016/j.jcis.2019.08.063

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier Inc.

A Sulfonated Mesoporous Silica Nanoparticle for Enzyme

Protection

against

Denaturants

and

Controlled Release under Reducing Conditions Hui Li,+,1,3 Yanxiong Pan,+,1 Jasmin Farmakes,1 Feng Xiao,2 Guodong Liu,1 Bingcan Chen,*,3 Xiao Zhu,4 Jiajia Rao*,3 and Zhongyu Yang*,1 + These authors contribute equally * Correspondence: [email protected]; [email protected] [email protected] 1. Department of Chemistry and Biochemistry, North Dakota State University, Fargo, North Dakota 58102, United States 2.

Department

of

Civil

Engineering, University

of

North

Dakota, Grand

Forks, North

Dakota 58202, United States 3. Department of Plant Sciences, North Dakota State University, Fargo, North Dakota 58108, United States 4. Research Computing, Information Technology at Purdue (ITaP), Purdue University, West Lafayette, IN, 47907, United States Key words: sulfonated mesoporous silica-nanoparticles, enzyme protection, enzyme controlled release, enzyme orientation, degree of enzyme freedom 1

Abstract Mesoporous silica nanoparticles (MSiNPs) are attractive enzyme hosts, but current MSiNPs are limited by leaching, poor enzyme stabilization/protection, and difficulty in controlling enzyme release. Sulfonated MSiNPs are promising alternatives, but are challenged by narrow channels, lack of control over enzyme adsorption to particle surfaces, and controlled release of enzyme. By introducing amines on particle surfaces and sulfonate groups into the channels via disulfide bonds, we developed a unique sulfonated MSiNP to selectively encapsulate enzymes to the channels with enhanced enzyme stabilization under denaturing conditions. Via pore-expansion, the channel diameter was increased which allows for encapsulating nm-sized enzymes. This new concept/strategy to immobilize and deliver enzymes or other biomacromolecules were demonstrated using two model enzymes. Furthermore, we combine site-directed spin labeling with Electron Paramagnetic Resonance to confirm the enhanced enzyme-host interaction and reveal the preferred enzyme orientation in the channels. Lastly, the presence of disulfides allows for enzyme release under reducing conditions, a great potential for cancer treatments. To the best of our knowledge, this is the first report of sulfonated MSiNPs that simultaneously offer enhanced stability against denaturants and controlled release of enzymes under reducing conditions, with enzyme orientation resolved in the channels.

2

1. Introduction Mesoporous silica nanoparticles (MSiNPs) as advanced porous platforms have improved chemical catalysis, molecular delivery, and separation science [1-6]. Of particular interest is their bioapplications such as enzyme storage and delivery, biocatalysis, biosensing, and biomedicine; MSiNPs are attractive for these applications due to their low toxicity, high biocompatibility, low cost, tunable pore size, and ease of preparation [7-12]. Meanwhile, a few limitations of MSiNPs as enzyme carriers preventing further expansion of MSiNP applications are: leaching [13], lack of enzyme protection against harsh or denaturing conditions [10, 14-17], and difficulty in controlling enzyme release [18]. In principle, these limitations originate from the relatively weak and non-specific enzyme-pore interactions; the difficulty in solely restricting enzymes in pores/channels (instead of on particle surface) could also lead to leaching and poor protection [19]. These drawbacks are especially severe for enzymes with sizes much smaller than channel size, which limit many important applications requiring small-enzyme delivery such as prodrug activation, antioxidant and antimicrobial enhancement, and nutrition delivery [20-26]. A potential solution is to enhance the enzyme-channel interaction by coating the channels with sulfonate groups. Sulfonated MSiNPs have been applied to chemical catalysis and drug delivery [27-30] but are not directly applicable to enzyme loading due to the narrow channel diameters and cannot provide controlled enzyme release due to lack of stimuli-responsive groups (i.e. disulfide bonds). In this work, we combine surface modification, disulfide sulfonation and pore expansion to develop sulfonated MSiNPs that are suitable for enzyme hosting. By introducing amines on the outer surface of the particle and coating channels with sulfonates (via disulfides), a particle that allows for restricting enzymes selectively in the channels was formed. This selectivity is achieved by taking advantage of the significantly different pKa values of the two functional groups (Figure 1). Pore expansion was used to load enzymes with the desired size wherein 1,3,5-trimethylbenzene (TMB) was used as the swelling agent to increase the size of the pore-templates formed by cetyl trimethyl ammonium bromide (CTAB) 3

surfactant [31, 32]. These features allow for enhancement of enzyme protection, leaching prevention (by enhancing enzyme-channel interactions and restricting enzymes in channels), and enzyme release under reducing conditions (due to disulfides; Figure 1). For proof-of-concept, we demonstrated these capabilities by encapsulating 2 representative small enzymes with different electrostatic points (pIs) in the channels. We found significantly enhanced protection for both enzymes against 6 M urea in our sulfonated MSiNPs, as well as high enzyme stability under acidic pHs and high salt concentrations. We then probed the molecular level origins of the enhanced interaction by using Site-Directed Spin Labeling (SDSL)-Electron Paramagnetic Resonance (EPR) spectroscopy to reveal enzyme behavior within the channels of sulfonated and regular MSiNPs [33, 34]. This method probes site-specific enzyme backbone dynamics and local environment regardless of enzyme surroundings and location (on surface or in channels) [33-44]. SDSL-EPR revealed an enhanced degree of enzyme restriction in the sulfonated MSiNPs compared to regular MSiNPs in the absence and presence of urea. Based on this data, a preferred enzyme orientation within the sulfonate MSiNP channels was proposed. This information is of fundamental importance in understanding and characterizing enzyme behavior under porous environments [45, 46]. Additionally, we demonstrated enzyme release under reducing conditions via cleavage of the disulfide bonds.

4

Figure 1. Schematic illustration of sulfonated MSiNP design to simultaneously achieve three goals listed on the right. Enzyme immobilization position can be controlled by adjusting the loading pH as shown in the insets (pI: isoelectric point). To the best of our knowledge, this is the first report on selectively encapsulating and restricting enzymes in the channels of sulfonated MSiNPs via disulfide bonds (to enable the controlled release of enzymes); the enhanced enzyme-channel interactions directly lead to significantly enhanced enzyme protection against denaturing and harsh conditions. Therefore, our sulfonated MSiNPs deliver new functions and duties by simultaneously allowing for biocatalysis in inhospitable environments as well as enzyme release under reducing conditions (such as cancerous cellular environments). Our platforms can be applied to enzymes with various sizes by adjusting pore expansion scales. This work is also the first report comparing enzyme microenvironments, at the residue level, in different porous silica

5

environments, which is meaningful for guiding the rational design of silica and other hierarchical porous platforms for enzyme encapsulation. The key to selectively restricting enzymes in the channels of MSiNPs is the low pKa of the sulfonate groups inside the channel and the high pKa of the amines on the particle surface. Specifically, for high pI enzymes (>7) under neutral pH, both the particle surface and the enzyme are positively charged while the channel is negatively charged, driving the enzymes to preferentially adsorb to the channels (Figure 1). For low pI enzymes (<7) under neutral pH, both the channels and the enzyme are negatively charged while the surface is positively charged; enzymes only adsorb to the surface (Figure 1). However, under more acidic pHs (ca. 3.5), both the particle surface and the enzyme are positive while the sulfonate channels are negative (pKa of ~2.5); enzymes only adsorb to the channels (Figure 1). Once adsorbed, altering pH does not affect encapsulation, as shown below for the two model enzymes; this is likely due to the steric hindrance. 2. Experimental Section The synthesis of precursors to prepare the Ms-NH2 as well as the preparation of both Ms-NH2 and M-NH2 were described in the main text with details in the Electronic Supplementary Materials (ESM). TEM images were obtained using a JEOL JEM-2100 LaB6 transmission electron microscope (JEOL USA, Peabody, Massachusetts) running at 200 kV. For TEM test, a drop of sample was placed on a 300-mesh formvar-carbon coated with copper TEM grid (Electron Microscopy Sciences, Hatfield, Pennsylvania, USA) for 30 seconds and wicked off with a filter paper. Protein-containing samples were stained by adding 1% phosphotungstic acid, adjusted to pH 7-8, placed on the grid for 2 minutes, then wicked off and allowed to air dry. All Zeta potential measurements were carried out with a Nano ZS Zetasizer (Malvern Instrument Ltd.). FTIR was conducted with an FTIR spectrometer (Thermo Scientific Nicolet is-10) equipped with Attenuated Total Reflection (ATR) element of Smart iTX AR Diamond and an Omnic 5.1 software. The porosity and pore size distribution of particles were measured by means of 6

N2 porosimetry at 77 K (Autosorb-iQ, Quantachrome, Boynton Beach, FL). EDX imaging and analysis were performed on a JEOL JEM-2100 LaB6 transmission electron microscope (JEOL USA, Peabody MA) operating at 200 kV. The expression, purification, spin labeling, and characterization of involved lysozyme mutants were conducted following standard procedures as described in our recent work [47, 48] and the ESM. To immobilize enzymes on MSiNPs, ~ 1 nmol of enzyme and MSiNPs were dispersed in 400 µL PBS buffer (50 mM, pH = 7.4) and incubated at ambient temperature for 16 hr under vibration. Each incubated sample was washed with the PBS buffer at least three times to remove any free enzyme. All activity measurements were repeated for three times to obtain the uncertainties. For each EPR spectrum, ~ 20 µL (~ 50-100 µM) the sample was loaded into a borosilicate capillary tube (0.70 mm i.d./1.25 mm o.d.; VitroGlass, Inc.), which was mounted in a Varian E-109 spectrometer fitted with a cavity resonator. All continuous wave (CW) EPR spectra were obtained with an observed power of 200 µW under a modulation frequency of 100 kHz and a modulation amplitude of 1.0 G. 3. Results and Discussions 3.1. Mesoporous Silica Nanoparticle (MSiNP) preparation and characterization To simplify discussion, through this work we define M-OH, M-NH2, and Ms-NH2 as the regular MSiNPs, the regular MSiNPs with amines on the surface, and the sulfonated MSiNPs with amines on the surface, respectively. The procedures of the one-pot synthesis of M-NH2, and Ms-NH2 are shown in Scheme 1. In brief, M-NH2 was synthesized according to the co-condensation method [31, 49] wherein cetyl-trimethyl-ammonium bromide (CTAB) was imbedded as templates in the standard silica nanoparticles

prepared

with

tetraethyl

orthosilicate

(TEOS)

in

the

presence

of

3-

aminopropyltriethoxysilane (APTES, which is to introduce –NH2 to the particle, including the surface, walls, and pores of the particle), followed by pore expansion for 4 days under autoclave and CTAB removal, sequentially (Scheme 1A). A 4-day pore-expansion was selected because this can create a proper pore size to load the model enzymes (see below). As introduced above and in the ESM, ~11.3 % 7

APTES was added to TEOS/APTES mixture, which means the surface, walls, and pores of the resultant particles have ~11.3% chance to contain –NH2. Still, the dominant functional group in the channel is– OH. To introduce sulfonate groups into the channels of Ms-NH2, a precursor, 2-[3-(trimethoxysilyl)propyldisulfanyl]-ethanesulfonic acid sodium salt (SDSP-TMS), was prepared first [49, 50]. Herein, 2,2′dithiodipyridine and 2-mercaptoethanesulfonic acid were reacted under glacial acetic acid to generate SDSP, followed by reaction with (3-mercaptopropyl) trimethoxysilane (MPTMS) to yield SDSP-TMS (Schemes S1&S2) [51]. Then, SDSP-TMS was reacted with CTAB as the new template, which was later included in the silica nanoparticle prepared with TEOS in the presence of APTES. Lastly, the product was treated with pore expansion and CTAB removal (Scheme 1B) as described above. Details see the ESM. Although there is still ~11.3 % chance for the resultant particle to contain –NH2, the majority of – NH2 is expected to exist in the surface and walls of the particle, because the addition of APTES/TEOS was on the opposite site of the interface of SDSP-TMS and CTAB, meaning the chance for APTES to enter the channel is low (if not zero). Therefore, the channels are dominantly coated with the sulfonate groups.

8

Scheme 1. Schematic illustration of the synthetic pathways of the M-NH2 (A) and Ms-NH2 (B). Pink outer surface = -NH2; brown channel surface in (B) = sulfonate; blue = silicate bulk. (Inset) Structures of TEOS and APTES. Details see the main text and ESM. Note that ordered pores are drawn to illustrate the chemical modification to the pores for clarity. The real pores/channels are in random directions (see below TEM images). The TEM images of the M-NH2 and Ms-NH2 are shown in Figures 2A and 2B, respectively, with particle sizes of ~120 nm and 270 nm. Particle size was measured using ImageJ software. The Ms-NH2 obtained from one-pot synthesis seems to be less structured than the spherical M-NH2 particles. Upon loading of a model enzyme, the TEM images appear to have similar morphologies as compared to the particle with no enzyme (Figure S1). This indicates that enzymes were most likely encapsulated in the channels and did not cause any surface changes.

Figure 2. The TEM images of the M-NH2 (A) and Ms-NH2 (B) prepared in this work. 9

FTIR data (Figure 3A) of the M-OH showed absorption at 3400, 1000-1200 and 797 cm-1, which were attributed to the O-H stretching, the asymmetric stretching vibration of Si-O-Si, and O-H bending vibrations, respectively, typical for MSiNPs. The disappearance of peaks at 2930 and 2850 cm-1 in MOH demonstrates the template (CTAB; Figure 3A black versus red) was effectively removed via reflux. Compared to M-OH, the appearance of the new peaks of M-NH2 and Ms-NH2 at 2930 and 2850 cm-1 indicate additional –CH2– stretching, likely caused by the addition of –CH2-NH2 of APTES [52]. For Ms-NH2, the presence of new peaks at 1409, 1340, and 1304 cm-1 was attributed to the stretching vibrations of S=O and -SO3-, demonstrating the presence of the sulfonate group [53, 54]. Energydispersive X-ray (EDX) analysis also confirmed the presence of sulfur as shown in Figure S2. To further characterize the pore sizes of Ms-NH2, nitrogen absorption experiments were carried out (Figure 3B). As shown in Figure 3C, the pore size distribution in the absence of enzyme for M-NH2 and Ms-NH2 were ~ 2.5-5.0 nm with surface areas of 443.5 and 343.1 m2/g, respectively. Upon enzyme loading, the surface areas were slightly decreased to 318.6 and 159.6 m2/g, respectively, consistent with tapped enzymes occupying pore spaces. Specifically, the enzyme seems to occupy most of the larger channels (4-5 nm) while leaving the smaller channels (2-3 nm) empty. This is reasonable considering the size of the model enzymes. Lastly, ζ-potential experiments showed that the surface charge of the M-OH shifted from -37.1 +/- 6.6 mV to 39.1 +/- 1.0 mV upon surface modification by amines (M-NH2), confirming the success of particle surface modification. A similar positive surface potential was also observed for Ms-NH2. Upon loading either a high or a low pI enzyme (see below), the ζ-potential was almost unchanged, indicating the loaded enzymes mostly entered the channels.

10

Figure 3. (A) The FTIR spectra of the MSiNPs-CTAB, MSiNPs, M-NH2, and Ms-NH2. (Inset) Amplified spectra to highlight the peaks indicating the addition of –NH2 and presence of sulfonate. (B) Nitrogen absorption experiments of the M-NH2 and Ms-NH2 in the absence (open black and red) and upon loading of enzyme (black squares and red dots). (C) The pore size distribution of the M-NH2 and Ms-NH2 (with 11

amine coated surfaces) in the absence (open black and blue) and presence of enzyme (black squares and blue dots). 3.2. Enzyme stability under denaturing conditions in M-NH2 and Ms-NH2 We selected two model enzymes, hen lysozyme and bovine erythrocytes Cu/Zn superoxide dismutase (SOD1) to prove principles of our design. SOD1 and lysozyme are selected because they have vastly different pIs (5.9 and 11.2, respectively) and relatively small size (2 x 3 x 5 and 2 x 3 x 6 nm, respectively), which make them good models for this work. Lysozyme cleaves the 1,4-glocosidic bond of polysaccharides. The substrate could be as small as oligosaccharides or as large as bacterial cell walls, the latter of which represents lysozyme native function. Both reactions can be quantified using established methods [55, 56]. SOD1 degrades the superoxide ion radicals (O2-•) in nature and its activity can also be quantified using established approaches as well [57], wherein the extent of suppression of O2-• freshly generated in a reaction indicates the presence of active SOD1. First, lysozyme was loaded into both M-NH2 and Ms-NH2 by incubating ~1.0 mg of lysozyme with 1.0 mg of each particle in 1.0 ml 50 mM PBS buffer (pH 7.4) at room temperature under constant nutation for at least 4 hr. The adsorption capacity of M-NH2 and Ms-NH2 is ~0.4 mg enzyme per 1 mg particle for both particles according to UV-Vis spectroscopy. Unattached protein was removed via centrifugation. The lysozyme activity test (see ESM) indicated comparable efficiency in degrading oligosaccharides for both M-NH2 and Ms-NH2 (97.4+/-1.9 % and 97.9+/-3.0 %; Figure 4A-1 vs 4A-4). Upon incubation with 6 M urea, the activity was significantly reduced to 45.5+/-4.8 % in M-NH2, likely due to enzyme unfolding, but in Ms-NH2 activity was retained by ~75 % (Figure 4A2 vs 4A5). This indicates Ms-NH2 channels can effectively prevent enzyme unfolding. Upon washing off the urea, lysozyme activity in M-NH2 was further reduced (to ~25 %) but enhanced in Ms-NH2 (to ~ 90%), indicating that in M-NH2 the trapped enzyme could not refold properly when the urea was removed. It is

12

also possible that some unfolded enzyme was removed from the channel of M-NH2 upon treatment of urea. In contrast, for Ms-NH2, the enzyme had better refolding success upon urea removal. Next, we loaded SOD1 into both M-NH2 and Ms-NH2 by incubating SOD1 with each particle in 1.0 ml 50 mM acetate buffer (pH 3.5 to ensure that SOD1 only loads to the channel) at room temperature under constant nutation for at least 4 hr. Unadsorbed SOD1 were removed via thorough washing with acetate buffer and high salt buffer (to ensure removal of surface adsorbed enzymes). The adsorption capacity of SOD1 in M-NH2 and Ms-NH2 is comparable to that of lysozyme (~ 0.4 mg per 1 mg particle) according to UV-Vis spectroscopy. As shown in Figure 4B, upon generation of O2 -• by xanthine oxidase, the working solution becomes yellow with absorption at 450 nm. Pure SOD1 at ~1.0 μM concentration completely suppresses the generation of 450 nm absorption, while buffer did not suppress 450 nm absorption (see Figure 4B blue and green triangles, respectively). Without urea, SOD1 loaded to both MNH2 and Ms-NH2 showed strong suppression of the 450 nm absorption (Figure 4B red dots and black squares). Upon incubation with 6 M urea, the M-NH2 showed > 70 % loss of activity while the Ms-NH2 showed < 30 % loss, confirming that Ms-NH2 was successful in protecting SOD1 against urea induced unfolding. Comparable activity for both model enzymes on Ms-NH2 were observed under low pHs and 0.5 M NaCl (see Figures S7&S8). These findings indicate that the sulfonated MSiNPs, Ms-NH2, are able to enhance stability for loaded enzymes against harsh conditions.

13

Figure 4. (A) The relative activity of lysozyme loaded in M-NH2 and Ms-NH2 (A1&A4), upon mixing with urea (A2&A5) and after removal of urea (A3&A6). The activity of lysozyme after treatment under reducing conditions (5 mM DTT) in the supernatant (A7) and particle channels (A8) was also plotted. Error bars were obtained via three independent activity tests. (B) The activity assay of SOD1 upon loading into M-NH2 (black squares) and Ms-NH2 (red dots) and after treatment with and removal of urea (open black squares and open red circles, respectively). Negative control with a buffer or empty MSiNPs is shown in green triangles and positive control of pure SOD1 enzyme with a similar concentration is shown in blue triangles. No SOD1 activity can be measured in the presence of urea. Error bars are about the size of the symbols for each sample. 3.3. Molecular level understanding of the enhanced enzyme-channel interaction 14

To obtain molecular level understanding of the enhanced interaction between enzyme and MNH2 versus enzyme and Ms-NH2, we conducted SDSL-EPR studies using a model lysozyme as a pilot system, T4 phage lysozyme (T4L). This lysozyme has been extensively studied using SDSL-EPR with well-established structure-function relationships [58-60]. We selected six representative residues which cover most of the enzyme surface regions to attach our spin labels, one at a time (Figure 5A). The resultant spectra show three regions, the low-, mid-, and high-field regions due to hyperfine splitting. Within each region, there are two spectral components observed, one originates from the label under high motional restriction (labeled as the “im” domain to indicate “immobile” motion) and that under less restriction (labeled as “m” to indicate “mobile”) [33]. Details of principles and spectral analysis presented in our recent work and ESM [47, 61]. Spectral simulations using first principal theory confirmed such assignment (example simulations see our recent work) [47]. Furthermore, spectral analysis suggested enhanced restriction of enzyme backbone dynamics in the channels of Ms-NH2 compared to those in the channels of M-NH2. For example, 65R1 in the M-NH2 channels shows enhanced peak intensity of the “m” domain but a less intense “im” peak (Figure 5B black curve), compared to that in the Ms-NH2 channels (Figure 5C black curve). Upon incubation with urea, the same labeled site shows three very sharp lines in M-NH2 channels (Figure 5B red curve), indicating significantly enhanced backbone dynamics motion [47], consistent with enzyme being unfolded. In contrast, the spectrum of Ms-NH2 (Figure 5C red) was mostly unchanged, indicating that urea did not influence enzyme backbone dynamics. Sharp lines upon addition of urea would indicate unfolding of enzymes adsorbed to the surface, which was not observed, thus indicating there were no enzymes adsorbed to the surface. Upon removal of urea the spectrum of 65R1 in Ms-NH2 showed no change (Figure 5C blue curve). However, in M-NH2 the spectrum was different (Figure 5B blue curve) from that before urea treatment. This indicates that although removal of urea may refold the enzyme, the local environment of 65R1 was not the same as that before urea treatment, which explains why the lysozyme show less activity even upon removal of 15

urea. Other labeled sites show similar trends. This explains why the enzyme lost activity in M-NH2 but not in the Ms-NH2 (Figure 4A).

Figure 5. (A) Six representative sites selected in this work and the scheme of site-directed spin labeling. (B) The EPR spectra of a representative site, 65R1, upon loading into M-NH2, addition of urea, and removal of urea are shown in black, red, and blue respectively. The immobile and mobile spectral components are highlighted by grey and yellow shades as well as the “im” and “m” markers, respectively. Only the low-field regions were highlighted because of the high resolution compared to other spectral regions. (C) The EPR spectra of a representative site, 65R1, upon loading into Ms-NH2, addition of urea, and removal of urea are shown in black, red, and blue, respectively. (D) The EPR spectra of all labeled sites upon loading into Ms-NH2. Stars represent sites with enhanced population in the mobile component. Furthermore, a careful look at all T4L mutants in Ms-NH2 channels (Figure 5D) indicates that 3 out of the 6 labeled sites, 44, 65, and 118, show a slightly stronger “m” peak (Figure 5D stars). Spectral 16

simulations also confirmed these three sites show slightly more than 50 % relative population of the “m” component (see Table S1 blue highlights). This finding indicates that these three sites, 44, 65, and 118 have a slightly higher chance to “hang out” in the channels without contact as compared to other sites which slightly prefer to contact the channel. Based on the relative chance of contacting the channel, a model of enzyme preferred orientation within the channel is proposed and shown in Figure 6A. This can be explained by the surface charge distribution of T4L shown in Figure S10 and the steric hindrance of the channel. Overall, all labeled residues are surrounded by both positive and negative residues (see Figure S10; red=positive; blue= negative), in line with the fact that all labeled sites have a chance to contact with or be “repelled” by the channels. Residues 44 and 65 seem to have slightly more negative positive surroundings than others, which may explain why these two sites have more “m” component. Furthermore, the relatively smaller ordering parameters (see ESM) of the “m” component from fitting also indicates that these two sites seem to have slightly higher degree of freedom (Table S1 green highlights) compared to other sites. Residue 118 shows a relatively higher order parameter but more mobile population (see Table S1 purple). A closer look at the 118 site indicates that it is partially buried which explains the high ordering parameter. However, due to the geometry of the enzyme (2 x 3 x 5 nm) and the channel diameter (2.5-5 nm), it is more likely for the enzyme to “lay down” along the channel norm as shown in Figure 6B, which reduces the chance for 118 to contact the channel walls. It has to be noted that a relatively higher degree of freedom at the N-terminus (represented by 44 and 65) is important for enzyme activity since the active site (as indicated by the yellow stars in Figure 6) is wrapped by the N- and C-terminus.

17

Figure 6. Preferred orientation of model enzyme lysozyme in Ms-NH2 channels (A) and that from the sideview (B). Relatively more restricted residues due to stronger contact with the channel are shown in blue, while residues with slightly less chance to contact the channel are shown in orange. (B) Residue 118 shows more restriction in motion possibly due to the relatively buried position in the protein. Still, due to geometry, 118 has a lower chance to contact the channel. Star = active site. 3.4. Enzyme release under reducing conditions from sulfonated Mesoporous Silica Nanoparticles (MSiNPs)

18

To show the feasibility of enzyme release under reducing conditions, we incubated the lysozyme and SOD1 loaded Ms-NH2 in a 5 mM dithiothreitol (DTT) reducing solution for ~12 hr under gentle nutation. The supernatants and particles were then separated via centrifugation. The amount and activity of released enzymes in the supernatant of each enzyme-particle composites were measured using UVVis spectroscopy and the standard activity assays discussed above, respectively. The activity of unreleased enzymes was also measured for each composite. For lysozyme and SOD1, we observed 93.4+/-2.5 and 97.8+/-1.3 % (~0.31 and 0.39 mg per 1 mg nanoparticle) release of protein as judged by UV absorption. The released enzymes were mostly functional (Figure 4A-7&8), indicating the potential of our Ms-NH2 in enzyme delivery applications. In contrast, there was no enzyme release from the MNH2 upon DTT treatment. 4. Conclusions We discovered a new mesoporous silica nanoparticle (MSiNP) which can simultaneously enhance enzyme protection against denaturing environments and enable enzyme controlled release under reducing conditions. Our key innovations are to introduce sulfonates (via disulfides) to the channel of MSiNPs and amines to the outer surface which can encapsulate enzymes with various pIs in the channels. Our hypothesis is that the encapsulated enzymes have enhanced interactions with the channels which can prevent the enzyme from denaturing and harsh conditions and leaching, especially for small enzymes. Meanwhile, the disulfide bonds allow for enzyme release under reducing conditions. While most existing sulfonated MSiNPs were focused on chemical catalysis and drug delivery applications [27-30], our unique approach in introducing sulfonates allows for enhanced enzyme encapsulation and controlled release, comparing to regular MSiNPs. We demonstrate the concept using two small enzymes, lysozyme and SOD1. The loaded enzymes in Ms-NH2 were active and stable in the presence of denaturants (6 M urea) as well as under low pHs and high salt concentrations, likely due to the enhanced stability that the sulfonate offers. Using SDSL-EPR, we confirmed the enhanced enzyme-channel interaction at the 19

residue level in Ms-NH2 compared to M-NH2. Furthermore, we revealed preferred orientations of a model enzyme in Ms-NH2 channels based on differences in backbone dynamics of different enzyme regions. Lastly, we demonstrated the feasibility of enzyme release under reducing conditions. Our MsNH2 is applicable to other enzymes with various sizes because pore sizes can be adjusted by tuning the pore expansion time. Enzyme orientation in silica channels is also meaningful for guiding future porous materials design for enzyme confinement so that optimal enzyme orientation (substrate access) and catalytic efficiency are achieved for the loaded enzymes. As future directions, we plan to improve the extent of sulfonate modification in order to enhance the loading capacity. Also, we will control the kinetics of enzyme release under various reducing conditions by tuning the degree of sulfonate modification. Our Ms-NH2 may be a bit large for possible biomedical applications. However, one of the main purposes of this manuscript is to demonstrate the principles of using sulfonated mesoporous silica nanoparticles for enzyme protection and controlled release. Future work is definitely needed to optimize the particle properties for biomedical applications.

Acknowledgements This work is supported by the New Faculty Startup Funds from the North Dakota State University (Z.Y.) as well as the 2018 ND EPSCoR Seed Award (FAR0030347). We appreciate Prof. Hubbell for generously providing the EPR spectral simulation package. The TEM work is supported by the National Science Foundation under Grant 0821655. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

20

Conflict of interest There is no conflict of interest. References [1] Liang, J.; Liang, Z.; Zou, R.; Zhao, Y., Heterogeneous Catalysis in Zeolites, Mesoporous Silica, and Metal-Organic Frameworks. Adv. Mater. 2017, 29, 1701139. [2] Aznar, E.; Oroval, M.; Pascual, L.; Murguía, J. R.; Martínez-Máńez, R.; Sancenón, F., Gated Materials for On-Command Release of Guest Molecules. Chem. Rev. 2016, 116, 561-718. [3] Song, N.; Yang, Y.-W., Molecular and Supramolecular Switches on Mesoporous Silica Nanoparticles. Chem. Soc. Rev. 2015, 44, 3474-3504. [4] Li, W.; Liu, J.; Zhao, D., Mesoporous Materials for Energy Conversion and Storage Devices. Nat. Rev. Mater. 2016, 1, 16023. [5] Park, S. S.; Santha Moorthy, M.; Ha, C.-S., Periodic Mesoporous Organosilicas for Advanced Applications. Npg Asia Mater. 2014, 6, e96. [6] Park, S. S.; Ha, C.-S., Hollow Mesoporous Functional Hybrid Materials: Fascinating Platforms for Advanced Applications. Adv. Funct. Mater. 2018, 28, 1703814. [7] Díaz, J. F.; Balkus, K. J., Enzyme Immobilization in MCM-41 Molecular Sieve. J. Mol. Catal. B Enzym. 1996, 2, 115-126. [8] Fried, D. I.; Brieler, F. J.; Fröba, M., Designing Inorganic Porous Materials for Enzyme Adsorption and Applications in Biocatalysis. ChemCatChem 2013, 5, 862-884. [9] Hartmann, M.; Kostrov, X., Immobilization of Enzymes on Porous Silicas - Benefits and Challenges. Chem. Soc. Rev. 2013, 42, 6277-6289. [10] Zhou, Z.; Hartmann, M., Progress in Enzyme Immobilization in Ordered Mesoporous Materials and Related Applications. Chem. Soc. Rev. 2013, 42, 3894-3912. [11] Magner, E., Immobilisation of Enzymes on Mesoporous Silicate Materials. Chem. Soc. Rev. 2013, 42, 6213-6222. [12] Popat, A.; Hartono, S. B.; Stahr, F.; Liu, J.; Qiao, S. Z.; Lu, G. Q., Mesoporous Silica Nanoparticles for Bioadsorption, Enzyme Immobilisation, and Delivery Carriers. Nanoscale 2011, 3, 2801-2818. [13] Verma, M. L.; Naebe, M.; Barrow, C. J.; Puri, M., Enzyme Immobilisation on AminoFunctionalised Multi-Walled Carbon Nanotubes: Structural and Biocatalytic Characterisation. PLoS ONE 2013, 8, e73642. [14] Lykourinou, V.; Chen, Y.; Wang, X. S.; Meng, L.; Hoang, T.; Ming, L. J.; Musselman, R. L.; Ma, S., Immobilization of MP-11 into a Mesoporous Metal-Organic Framework, MP-11@mesoMOF: A New Platform for Enzymatic Catalysis. J. Am. Chem. Soc. 2011, 133, 10382-10385. [15] DiCosimo, R.; McAuliffe, J.; Poulose, A. J.; Bohlmann, G., Industrial Use of Immobilized Enzymes. Chem. Soc. Rev. 2013, 42, 6437-6474. [16] Sheldon, R. A., Enzyme Immobilization: the Quest for Optimum Performance. Adv. Syn. Catal. 2007, 349, 1289-1307. [17] Hanefeld, U.; Gardossi, L.; Magner, E., Understanding Enzyme Immobilisation. Chem. Soc. Rev. 2009, 38, 453-468. [18] Yu, A.; Wang, Y.; Barlow, E.; Caruso, F., Mesoporous Silica Particles as Templates for Preparing Enzyme-Loaded Biocompatible Microcapsules. Adv. Mater. 2005, 17, 1737-1741. [19] Küchler, A.; Yoshimoto, M.; Luginbühl, S.; Mavelli, F.; Walde, P., Enzymatic Reactions in Confined Environments. Nat. Nanotechnol. 2016, 11, 409. 21

[20] Ortac, I.; Simberg, D.; Yeh, Y.-S.; Yang, J.; Messmer, B.; Trogler, W. C.; Tsien, R. Y.; Esener, S., Dual-Porosity Hollow Nanoparticles for the Immunoprotection and Delivery of Nonhuman Enzymes. Nano Lett. 2014, 14, 3023-3032. [21] Chen, Y.-P.; Chen, C.-T.; Hung, Y.; Chou, C.-M.; Liu, T.-P.; Liang, M.-R.; Chen, C.-T.; Mou, C.Y., A New Strategy for Intracellular Delivery of Enzyme Using Mesoporous Silica Nanoparticles: Superoxide Dismutase. J. Am. Chem. Soc. 2013, 135, 1516-1523. [22] Wang, Y.; Caruso, F., Mesoporous Silica Spheres as Supports for Enzyme Immobilization and Encapsulation. Chem. Mater. 2005, 17, 953-961. [23] Chang, F.-P.; Chen, Y.-P.; Mou, C.-Y., Intracellular Implantation of Enzymes in Hollow Silica Nanospheres for Protein Therapy: Cascade System of Superoxide Dismutase and Catalase. Small 2014, 10, 4785-4795. [24] Guardado-Alvarez, T. M.; Sudha Devi, L.; Russell, M. M.; Schwartz, B. J.; Zink, J. I., Activation of Snap-Top Capped Mesoporous Silica Nanocontainers Using Two Near-Infrared Photons. J. Am. Chem. Soc. 2013, 135, 14000-14003. [25] Hanefeld, U.; Cao, L.; Magner, E., Enzyme Immobilisation: Fundamentals and Application. Chem. Soc. Rev. 2013, 42, 6211-6212. [26] Hung, B.-Y.; Kuthati, Y.; Kankala, K. R.; Kankala, S.; Deng, J.-P.; Liu, C.-L.; Lee, C.-H., Utilization of Enzyme-Immobilized Mesoporous Silica Nanocontainers (IBN-4) in ProdrugActivated Cancer Theranostics. Nanomater. 2015, 5, 2169-2191. [27] Yambem, S. D.; Timm, J.; Weiss, M.; Pandey, A. K.; Marschall, R., Sulfonated Mesoporous Silica as Proton Exchanging Layer in Solid-State Organic Transistor. Adv. Electron. Mater. 2017, 3, 1700316. [28] Sharghi, H.; Shiri, P.; Aberi, M., An Overview on Recent Advances in the Synthesis of Sulfonated Organic Materials, Sulfonated Silica Materials, and Sulfonated Carbon Materials and Their Catalytic Applications in Chemical Processes. Beilstein J. Org. Chem. 2018, 14, 2745-2770. [29] Trewyn, B. G.; Slowing, I. I.; Giri, S.; Chen, H.-T.; Lin, V. S. Y., Synthesis and Functionalization of a Mesoporous Silica Nanoparticle Based on the Sol-Gel Process and Applications in Controlled Release. Acc. Chem. Res. 2007, 40, 846-853. [30] Wang, X.; Liu, R.; Waje, M. M.; Chen, Z.; Yan, Y.; Bozhilov, K. N.; Feng, P., Sulfonated Ordered Mesoporous Carbon as a Stable and Highly Active Protonic Acid Catalyst. Chem. Mater. 2007, 19, 2395-2397. [31] Kim, M.-H.; Na, H.-K.; Kim, Y.-K.; Ryoo, S.-R.; Cho, H.S.; Lee, K.E.; Jeon, H.; Ryoo, R.; Min, D.-H., Facile Synthesis of Monodispersed Mesoporous Silica Nanoparticles with Ultralarge Pores and Their Application in Gene Delivery. ACS Nano 2011, 5, 3568-3576. [32] Blin, J.L.; Impéror-Clerc, M., Mechanism of Self-assembly in the Synthesis of Silica Mesoporous Materials: In Situ Studies by X-ray and Neutron Scattering. Chem. Soc. Rev. 2013, 42, 4071-4082. [33] Hubbell, W. L.; López, C. J.; Altenbach, C.; Yang, Z., Technological Advances in Site-Directed Spin Labeling of Proteins. Curr. Opin. Struct. Biol.2013, 23, 725-733. [34] Altenbach, C.; López, C. J.; Hideg, K.; Hubbell, W. L.; Peter, Z. Q.; Kurt, W., Chapter Three Exploring Structure, Dynamics, and Topology of Nitroxide Spin-Labeled Proteins Using Continuous-Wave Electron Paramagnetic Resonance Spectroscopy. In Methods in Enzymology, Academic Press, 2015; Vol. 564, pp 59-100. [35] McHaourab, H. S.; Steed, P. R.; Kazmier, K., Toward the Fourth Dimension of Membrane Protein Structure: Insight into Dynamics from Spin-Labeling EPR Spectroscopy. Structure 2011, 19, 1549-1561. [36] Cafiso, D. S., Identifying and Quantitating Conformational Exchange in Membrane Proteins Using Site-Directed Spin Labeling. Acc. Chem. Res. 2014, 47, 3102-3109. 22

[37] Fanucci, G. E.; Cafiso, D. S., Recent Advances and Applications of Site-Directed Spin Labeling. Curr. Opin. Struct. Biol. 2006, 16, 644-653. [38] Borbat, P.; Freed, J., Pulse Dipolar ESR: Distance Measurements. In Structural Information from Spin-Labels and Intrinsic Paramagnetic Centers in the Biosciences. Structure and Bonding., Harmer, J.; Timmel, C., Eds. Springer, Heidelberg, Germany: New York, USA, 2014; Vol. 152, pp 1-82. [39] DeBerg, H. A.; Brzovic, P. S.; Flynn, G. E.; Zagotta, W. N.; Stoll, S., Structure and Energetics of Allosteric Regulation of HCN2 Ion Channels by Cyclic Nucleotides. J. Biol. Chem. 2016, 291, 371-381. [40] Ji, M.; Ruthstein, S.; Saxena, S., Paramagnetic Metal Ions in Pulsed ESR Distance Distribution Measurements. Acc. Chem. Res. 2014, 47, 688-695. [41] Huang, Y.-W.; Lai, Y.-C.; Tsai, C.-J.; Chiang, Y.-W., Mesopores Provide an Amorphous State Suitable for Studying Biomolecular Structures at Cryogenic Temperatures. Proc. Natl. Acad. Sci. 2011, 108, 14145 -14150. [42] Ruthstein, S.; Schmidt, J.; Kesselman, E.; Popovitz-Biro, R.; Omer, L.; Frydman, V.; Talmon, Y.; Goldfarb, D., Molecular Level Processes and Nanostructure Evolution During the Formation of the Cubic Mesoporous Material KIT-6. Chem. Mater. 2008, 20, 2779-2792. [43] Ruthstein, S.; Frydman, V.; Goldfarb, D., Study of the Initial Formation Stages of the Mesoporous Material SBA-15 Using Spin-Labeled Block Co-polymer Templates. J. Phys. Chem. B 2004, 108, 9016-9022. [44] Hanson, S. M.; Francis, D. J.; Vishnivetskiy, S. A.; Kolobova, E. A.; Hubbell, W. L.; Klug, C. S.; Gurevich, V. V., Differential Interaction of Spin-Labeled Arrestin with Inactive and Active Phosphorhodopsin. Proc. Natl. Acad. Sci. 2006, 103, 4900-4905. [45] Nabavi Zadeh, P. S.; Åkerman, B., Immobilization of Enzymes in Mesoporous Silica Particles: Protein Concentration and Rotational Mobility in the Pores. J. Phys. Chem. B 2017, 121, 25752583. [46] Göβl, D.; Singer, H.; Chiu, H.-Y.; Schmidt, A.; Lichtnecker, M.; Engelke, H.; Bein, T., Highly Active Enzymes Immobilized in Large Pore Colloidal Mesoporous Silica Nanoparticles. New J. Chem. 2019, 43, 1671-1680. [47] Pan, Y.; Li, H.; Farmakes, J.; Xiao, F.; Chen, B.; Ma, S.; Yang, Z., How Do Enzymes Orient on Metal-Organic Framework (MOF) Surfaces? J. Am. Chem. Soc. 2018, 140, 16032-16036. [48] Pan, Y.; Neupane, S.; Farmakes, J.; Bridges, M.; Froberg, J.; Rao, J.; Qian, S. Y.; Liu, G.; Choi, Y.; Yang, Z., Probing the Structural Basis and Adsorption Mechanism of an Enzyme on Nanosized Protein Carriers. Nanoscale 2017, 9, 3512-3523. [49] Huh, S.; Wiench, J. W.; Yoo, J.-C.; Pruski, M.; Lin, V. S. Y., Organic Functionalization and Morphology Control of Mesoporous Silicas via a Co-Condensation Synthesis Method. Chem. Mater. 2003, 15, 4247-4256. [50] Lin, V. S. Y.; Lai, C.-Y.; Huang, J.; Song, S.-A.; Xu, S., Molecular Recognition Inside of Multifunctionalized Mesoporous Silicas: Toward Selective Fluorescence Detection of Dopamine and Glucosamine. J. Am. Chem. Soc. 2001, 123, 11510-11511. [51] Radu, D. R.; Lai, C.-Y.; Huang, J.; Shu, X.; Lin, V. S. Y., Fine-Tuning the Degree of Organic Functionalization of Mesoporous Silica Nanosphere Materials via an Interfacially Designed CoCondensation Method. Chem. Commun. 2005, 1264-1266. [52] Bao, Y.; Wang, T.; Kang, Q.; Shi, C.; Ma, J., Micelle-Template Synthesis of Hollow Silica Spheres for Improving Water Vapor Permeability of Waterborne Polyurethane Membrane. Sci. Rep. 2017, 7: 46638.

23

[53] Buzzoni, R.; Bordiga, S.; Ricchiardi, G.; Spoto, G.; Zecchina, A., Interaction of H2O, CH3OH, (CH3)2O, CH3CN, and Pyridine with the Superacid Perfluorosulfonic Membrane Nafion: an IR and Raman Study. J. Phys. Chem. 1995, 99, 11937-11951. [54] Peña, L.; Xu, F.; Hohn, K.; Li, J.; Wang, D., Propyl-Sulfonic Acid Functionalized Nanoparticles as Catalyst for Pretreatment of Corn Stover. J. Biomater. Nanobiotechnol. 2014, 5, 8-16. [55] Van der Straeten, A.; Bratek-Skicki, A.; Germain, L.; D'Haese, C.; Eloy, P.; Fustin, C. A.; Dupont-Gillain, C., Protein-Polyelectrolyte Complexes to Improve the Biological Activity of Proteins in Layer-By-Layer Assemblies. Nanoscale 2017, 9, 17186-17192. [56] Bower, C. K.; Xu, Q.; McGuire, J., Activity Losses among T4 Lysozyme Variants after Adsorption to Colloidal Silica. Biotechnol. Bioengin. 1998, 58, 658-662. [57] de Alteriis, E.; Falanga, A.; Galdiero, S.; Guida, M.; Maselli, V.; Galdiero, E., Genotoxicity of Gold Nanoparticles Functionalized with Indolicidin towards Saccharomyces Cerevisiae. J. Environ. Sci. 2018, 66, 138-145. [58] López, C.J.; Fleissner, M.R.; Guo, Z.; Kusnetzow, A.K.; Hubbell, W.L., Osmolyte Perturbation Reveals Conformational Equilibria in Spin-Labeled Proteins. Prot. Sci. 2009, 18, 1637-1652. [59] López, C.J., Oga, S., Hubbell, W.L., Mapping Molecular Flexibility of Proteins with Site-Directed Spin Labeling: a Case Study of Myoglobin. Biochemistry 2012, 51, 6568-6583. [60] Lerch, M.T.; López, C. J.; Yang, Z.; Kreitman, M.J.; Horwitz, J.; Hubbell, W.L.; StructureRelaxation Mechanism for the Response of T4 Lysozyme Cavity Mutants to Hydrostatic Pressure. Proc. Natl. Acad. Sci. 2015, 112, E2437-E2446. [61] Budil, D. E.; Lee, S.; Saxena, S.; Freed, J. H., Nonlinear-Least-Squares Analysis of Slow-Motion EPR Spectra in One and Two Dimensions Using a Modified Levenberg-Marquardt Algorithm. J. Magn. Reson. A 1996, 120, 155-189.

Graphic Abstract

24