Design of Glyco-Linkers at Multiple Structural Levels to Modulate

Subscriber access provided by AUSTRALIAN NATIONAL UNIV

Biophysical Chemistry, Biomolecules, and Biomaterials; Surfactants and Membranes

Design of Glyco-Linkers at Multiple Structural Levels to Modulate Protein Stability Xudong Feng, Xiaoyan Wang, Beijia Han, Changling Zou, Yuhui Hou, Lina Zhao, and Chun Li J. Phys. Chem. Lett., Just Accepted Manuscript • DOI: 10.1021/acs.jpclett.8b01570 • Publication Date (Web): 30 Jul 2018 Downloaded from http://pubs.acs.org on August 1, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry Letters

Design of Glyco-Linkers at Multiple Structural Levels to Modulate Protein Stability

Xudong Feng1**, Xiaoyan Wang13**, Beijia Han1, Changling Zou1,2, Yuhui Hou1, Lina Zhao2*, Chun Li1*

1

Department of Biochemical Engineering/Institute for Synthetic Biosystem, School of

Chemistry and Chemical Engineering, Beijing Institute of Technology, Beijing 100081, China 2

Key Laboratory for Biomedical Effects of Nanomaterials and Nanosafety, Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049, China

3

Center of Biotechnology, COFCO Nutrition & Health Research Institute, Beijing 102209, China

*Corresponding authors: Lina Zhao ([email protected]), Chun Li ([email protected])

**

These authors contributed equally to this work

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract N-glycosylation has critical roles in regulating protein stability, but the molecular basis is poorly understood. In this study, we integrated experimental and computational techniques to investigate the mechanism by which full-length N-glycans modulate protein stability from quaternary structure perspective. We found the two inherent N-glycans of β-glucuronidase expressed in Pichia pastoris function as “glyco-linkers” that hold spatially proximal motifs together to compact the local protein structure. We further designed and placed glyco-linkers in the unusual form of glyco-bridge and glyco-hairpin at the interfaces between domains and monomers with higher structural level, respectively, which conferred dramatically higher kinetic stability and thermodynamic stability than the inherent N-glycans. Our study not only provides unique insight into the interactions between glycans and proteins from a quaternary structure perspective but also facilitates the rational design of N-glycans as general tools that can enhance protein stability.

2


Page 2 of 25

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table of Contents

3



N-linked glycosylation is one of the most prevalent post-translational modifications in eukaryotes.1-2 This modification consists of attachment of an N-glycan to asparagine in the consensus sequence Asn-X-Thr/Ser, where X denotes any residue except proline.3-4 In many cases, N-linked glycosylation has been proven to be a key determinant in the regulation of protein secretion, folding, function and stability.5-10 Specifically, many studies have shown that N-glycosylation greatly affects protein stability and that the introduction of exogenous N-glycans to native proteins sometimes stabilizes their structure.11-12 N-glycans modulate protein function mainly by inducing the formation of a network of glycan-protein interactions, including hydrogen bonding, hydrophobic interactions and CH-π interactions.13-15 Several successful cases have been reported in which protein stability is enhanced when the protein is expressed heterologously for introducing N-glycosylation in eukaryotic cells such as fungal or mammalian cells; however, in most cases, such natural N-glycosylation processes are rather random and uncontrollable, and N-glycosylation at an incorrect site may even destabilize a protein.16-17 Therefore, enhancing protein stability by rationally designing N-glycosylation sites is quite challenging. Generally speaking, interactions within a multimeric protein tend to be weaker at higher levels of structural organization than at lower levels (i.e., strongest between motifs, less strong between domains, and weakest between monomers); thus, regions between domains and monomers that are insufficiently reinforced needs special attention where the overall structure starts to collapse under harsh conditions. Therefore, we propose that the structure level should be considered in modulating 4


Page 4 of 25

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


protein stability, which is often neglected while designing mutation sites. Typically, N-glycans in eukaryotic cells contain abundant sugar residues (for example, 11-13 sugars in P. pastoris and more than 100 sugars in S. cerevisiae). Herein, we envision that if properly designed at the multiple structural levels (i. e. the interfaces between motifs, domains or monomers), the long and branched chain of N-glycan may establish interaction with different domains or monomers of protein, thus enable it work as a “glyco-linker” to hold these different structures tightly together to compact the local structure, thus improving the overall stability of proteins. To test this hypothesis, we selected β-glucuronidase from Aspergillus oryzae Li-3 (PGUS) as a model protein, of which the crystal structure has been solved recently by our group.18 PGUS in homotetrameric form composed of two asymmetric units, and each monomer contains three conserved domains: a sugar-binding domain, an immunoglobulin-like β-sandwich domain and a catalytic (β/α)8-triosephosphate isomerase (TIM) barrel domain. Two catalytic residues, E414 and E505, act as an acid/base and a nucleophile, respectively.18 PGUS was heterologously expressed in P.

pastoris (PGUS-P), a host that is frequently used to introduce N-glycosylation with high mannose content into proteins,19-20 and four N-glycans N28, N231, N383 and N594 sites were naturally introduced as verified by the N-glycosylation sequence analysis (Fig. S1) and mutagenesis experiment (Fig. S2a). Then we tested the effect of N-glycan on PGUS-P thermostability at 70 °C. As shown in Fig. S2b, ∆PGUS-P (all the glycans were removed by mutating Asn to Gln) was the least thermostable, maintaining less than 30% of its original activity after 20 min of incubation. In sharp 5



contrast, PGUS-P was the most thermostable; its t1/2 of 52.5 min was two times higher than that of ∆PGUS-P. The thermostability of the N383Q and N594Q mutants was significantly lower, with t1/2 values of only 19.4 and 11.9 min, respectively; these half-lives were similar to that of ∆PGUS-P (17.2 min). These results indicate that the N-glycans at N383 and N594 play important roles in PGUS-P thermostability. Then we employed molecular dynamics (MD) simulation to investigate if the glyco-linker effect existed for the N-glycan at N383 and N594. When this research was performed, the structure of Man7GlcNAc2 in the PDB database (PDB ID: 2WAH) was the glycoform most similar to the structure in this study, so Man7GlcNAc2 was chosen as a representative structure for MD simulation. The molecular conformations of PGUS decorated with an N-glycan at N383 and N594 were individually analyzed by performing MD simulations. After 120 ns, the root mean square deviation (RMSD) values of these two PGUS-P mutants became relatively stable (Fig. S3 a, d). Meanwhile, the RMSD of N-glycan versus PGUS protein (Fig. S3 b, e) fluctuated dramatically which corresponded to the dynamic adjustment of N-glycan to form a stable compatible conformation with PGUS. We thus took a snapshot at 120 ns to capture a representative structure for analysis (Fig. S3 c, f). Fig. S4a and b show the interaction network that forms between the N-glycan at N383 and PGUS. First, residue N383 is located at the point where the 3α-helix (residues 383-400) intersected loop 376-382 in the TIM barrel domain. The 3α-helix is connected to the 4α-helix (residues 422-436) by a long loop (401-421), and these two helixes are stabilized by only one salt bridge formed by R386 and E424, resulting in that this subdomain is not 6


Page 6 of 25

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


sufficiently stable. When an N-glycan is introduced at N383, it forms a strong hydrogen bond network with neighboring residues situated in the 3α-helix (N383 and R386), 4α-helix (E424) and linking loop 376-382 (F376 and N382). Notably, sugar residues 4, 6, and 8 of the N-glycan establish six polar interactions with the PGUS protein. Therefore, the N-glycan reinforces interactions among the 3α-helix, 4α-helix and loop 376-382, thus increasing the stability of this subdomain (residues 376-436). As shown in Fig. S4c, the N-glycan at N594 forms abundant hydrogen bonds with three regions of the TIM barrel domain: the C-terminal loop (E598), 9α-helix (R586, T590 and N591) and 1α-helix (H321) (Fig. S4d). Therefore, the N-glycan at N594 enhances the interactions among these three regions, thus decreasing the flexibility of the C-terminal loop and improving the overall stability of the TIM barrel domain. In summary, the above results suggest that N-glycans at N383 and N594 share a common ability to enhance protein stability; that is, the N-glycan establishes extensive interactions with spatially proximal secondary structures (mainly α-helixes and β-strands) and motifs within the same domain, enabling N-glycan to function as a “glyco-linker” holding these structures tightly together, compacting the local domain structure and thus improving the overall stability of the protein. In addition, sugars other than the two GlcNAc residues are deeply involved in interactions between N-glycans and the protein; these interactions are indispensable in the function of the N-glycan as a “glyco-linker”. Then, we used ∆PGUS-P as a starting point to design glyco-linkers on the basis of an in-depth structure analysis. N-glycosylation sites are likely to be located at 7



segments between ordered secondary structures and β-turns;21 thus, we focused on all 40 loops/turns that were present in each monomer of PGUS. Next, the loops/turns that were located at the hydrophobic cave and inside the interface between monomers were considered unsuitable to host the glyco-linker, because the glycan would either disturb the hydrophobic interactions or the intermonomeric interactions between monomers and thus possibly cause the protein to incorrectly fold. On the basis of the above analysis, we picked 24 of the 40 loops/turns situated at the interfaces of domains and monomers to host the glyco-linker, as shown in Table S1. We divided these sites into two groups on the basis of the structural level at which they interacted with the protein: interdomain and intermonomer levels (Fig. 1). We subsequently introduced an N-glycan to these 24 loops/turns by placing four types of consensus sequons, N-X-T, F-N-X-T, F-Y-N-X-T and F-Y-Z-N-X-T (X denote any residues except proline, while Y and Z denote any residues), depending on the length of the target loops/turns in PGUS-P into the protein sequence (Fig. 1, Table S1). These sequences were chosen because the presence of aromatic residues adjacent to the glycosylation site is favorable for increasing N-glycosylation efficiency and N-glycan homogeneity.22-25 We designed 36 N-glycosylation sites for 24 loops/turns (for long loops/turns, two different N-glycosylation sites were introduced), and the mutants were named PGUS-PX, where X denotes the new N-glycosylation site. Then, we established a three-step screening process to rapidly identify mutants with both activity and N-glycan decoration (Fig. S5), and 14 mutants with N-glycans at 12 loops/turns at domain/monomer interfaces were obtained (highlighted in red in Table 8


Page 8 of 25

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


S2). Subsequently, we characterized the thermostability of these 14 mutants (in crude enzyme form) at 70 °C (Fig. S6). Compared with ∆PGUS-P, most of the PGUS-P mutants with the new glyco-linkers showed improved thermostability, exhibiting high residual activity after 180 min of incubation. Two mutants PGUS-P208 and PGUS-P40

had much

higher

thermostability

than

PGUS-P,

whereas

the

thermostability of PGUS-P418 mutant was lower than that of ∆PGUS-P. These three mutants were further purified and systematically characterized as three representative mutants. The combination of SDS-PAGE and PNGase F assays confirmed that the mutants PGUS-P208, PGUS-P40 and PGUS-P418 were indeed N-glycosylated (Fig. 2a). In agreement with the thermostability results measured for the crude enzyme (Fig. S6), the purified PGUS-P208 and PGUS-P40 mutants were dramatically more stable, exhibiting half-lives (t1/2) of 67.3 and 147.5 min, which were 2.7 and 7.1 times higher than that of ∆PGUS-P (18.2 min), respectively. In addition, the t1/2 values of PGUS-P208 and PGUS-P40 mutants were 0.26 and 1.8 times longer, respectively, than that of PGUS-P (53.3 min), indicating that the introduction of one N-glycan at the proper site exceeded the stabilizing effect of the four inherent N-glycans of PGUS-P (Fig. 2b, c). In addition, PGUS-P40 showed a higher kinetic stability than did PGUS-P208, thus indicating that introducing an N-glycan between monomers is more effective than between domains. This conclusion is reasonable, because the interactions between monomers are usually weaker than those between domains; thus, introducing a glyco-linker to reinforce interactions between monomers may yield a 9



more pronounced improvement in stability. Intriguingly, the kinetic stability was significantly lower (10.3 min) in PGUS-P418 than in ∆PGUS-P, thereby indicating that the inappropriate introduction of an N-glycan was detrimental to protein stability (Fig. 2d). Previous study has also demonstrated that weak interactions between monomer interfaces are a limiting factor in multimeric protein stability and that optimizing the interactions in the interfaces between monomers thus significantly improves protein stability.26 To confirm that all the changes in the enzymatic properties of PGUS-P were caused by the presence of an N-glycan, we also generated the unglycosylated counterparts of PGUS-P208, PGUS-P40 and PGUS-P418 in E.

coli, referred to as PGUS-E208, PGUS-E40 and PGUS-E418, respectively, and purified them (Fig. 2b, c, d). The kinetic stability of purified PGUS-E208, PGUS-E40 and PGUS-E418 was similar to that of ∆PGUS-P, indicating that changes in the stabilities were because of the introduced N-glycans and not simply the mutated residues themselves. The dynamic stability of PGUS-P208, PGUS-P40 and PGUS-P418 was further characterized by using denaturation induced by guanidinium (Fig. S7). PGUS-P40 exhibited the highest Gibbs free energy of unfolding (∆G, 44.7 kJ mol-1) among all the PGUS-P mutants; this value was 14.6 and 11.1 kJ mol-1 greater than those of ∆PGUS-P and PGUS-P, respectively, thus indicating that introducing the N-glycan at N40 also conferred greater dynamic stabilization than did the three inherent N-glycans at N231, N383 and N594. PGUS-P208 showed the second highest ∆G of unfolding (38.5 kJ mol-1), which indicated that its folded state was still 4.9 kJ mol-1 10


Page 10 of 25

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


more stable than that of PGUS-P. The △G value of the N418 mutant, at 20.8 kJ mol-1, was the lowest, 9.3 kJ mol-1 lower than that of ∆PGUS-P. These results were in good agreement with those of the kinetic stability described above, thus confirming that the designed glyco-linker at N208 and N40 increased the protein stability over that of PGUS-P. Strikingly, we also found that the introduction of a glyco-linker at N208 and N40 significantly improved PGUS activity (Table S3). PGUS-P40 had the highest specific activity (51.57 U mg-1) and catalytic efficiency kcat/Km (140.2 s-1 µM-1) of all the PGUS-P mutants, and these values were 0.8 and 1.7 times higher than those of ∆PGUS-P, respectively. PGUS-P208 also showed slightly higher specific activity (33.01 U mg-1) and catalytic efficiency kcat/Km (78.11 s-1 µM-1) than ∆PGUS-P. These results indicated that the glyco-linker not only improved protein stability but also enhanced its activity. We then performed 200 ns all-atom MD simulations to determine the conformations of PGUS-P208, PGUS-P40 and PGUS-P418 mutants. The crystal structure of PGUS protein demonstrates that it is a homotetramer containing two asymmetric units and that each asymmetric unit is composed of two identical monomers (e.g., Chains A and B).

18

The RMSD of PGUS-P208 and PGUS-P40

becomes stable after 104 and 120 ns, respectively, indicating that a stable glyco-linker conformation is formed (Fig. 3a and Fig. 3b). The longer time required to stabilize the conformation of PGUS-P40 compared with PGUS-P208 is a result of dramatic fluctuations in RMSD after 100 ns. In contrast, the RMSD of N-glycan in 11



PGUS-P418 fluctuates and increases during the entire simulation until end of the simulation at 200 ns, indicating that the N-glycan is not able to find a position with optimal energy to support a stable conformation of PGUS-P418 (Fig. 3c). Snapshots of the structures at 200 ns were taken to be representative for conformation analysis. The conformation analysis of PGUS-P208 mutant indicated that the N-glycan at N208 is buried deeply inside a cavity formed by the three domains of one monomer (Fig. 3d, Movie S1). Because this N-glycan was designed to interact at the domain level, no interaction with the neighboring monomer is observed. Nonetheless, the introduced N-glycan at N208 forms substantial polar interactions with two domains within the same monomer (Fig. 4a). This N-glycan forms eight hydrogen bonds with the 3α-helix (residues 383-400, TIM barrel domain), one hydrogen bond with the N-terminal domain (sugar-binding domain), and one hydrogen bond with the motif 100-115 (sugar-binding domain). These interactions enable N-glycans to function as “glyco-bridge” to support the neighboring domain and alleviate their structure collapse under heat, enhancing the protein stability at tertiary structural level. In the PGUS-P40 structure, the N-glycan introduced in one monomer (loop 23-42, Chain B) forms substantial polar interactions with a subdomain of the neighboring monomer (Chain A) in the same asymmetric unit (Fig. 3e, Movie S2), whereas the N-glycan does not interact with residues in the same monomer (Chain B). Subsequently, the interactions between the N-glycan attached to the N40 residue in Chain B and the surrounding residues of Chain A were analyzed in detail. As shown in Fig. 4b, the N-glycan forms two hydrogen bonds with the C-terminus, four 12


Page 12 of 25

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


hydrogen bonds with the subdomain 310-330 (composed of the 1α-helix, the 4β-strand and the linking loop323-327), and six hydrogen bonds with the subdomain 231-257 (composed of the 5β-strand, the 6β-strand and the linking loop237-250). We propose that this N-glycan acts as a “glyco-hairpin” to thread neighboring monomers (Chains A and B) to form a more compacted integral fold, thereby enhancing the overall stability of PGUS at quaternary structural level and resulting in the stability being most increased in PGUS-P40. It should be noted that the N-glycan at N40 is located far from the two catalytic residues and has no direct interaction with the critical loops contributing to the active site (Fig. S8). As shown in Movie S1, Movie S2, and Fig. 5, the entire simulation of PGUS-P208 and PGUS-P40 showed that the number of contact atoms and hydrogen bonds between the proteins and their N-glycans began to increase as the glyco-linker formed, and existed all the time until the end of the simulation, indicating once the glyco-linker was formed, the interface was stable as a function of time. In addition, the dramatic increase of the RMSD of residues interacting with the N-glycan of both PGUS-P208 and PGUS-P40 at the initial stage of the simulation was observed, which may be correlated to the interaction with N-glycan, and the RMSD reached stable at 200 ns, indicating a stable conformation has been obtained (Fig. S9). The N-glycan in PGUS-P418 swings severely during the simulation and cannot establish binding interactions with surrounding residues; the N-glycan thus cannot form a stable conformation (Fig. 3f, Movie S3). This motion may explain why the stability of the PGUS-P418 mutant was even poorer than that of ∆PGUS-P. 13



We also used differential scanning calorimetry (DSC) to experimentally verify the effect of the N-glycans at N40 and N208 on PGUS-P protein conformation (Fig. S10). The DSC plots of all PGUS-P mutants showed two peaks, one at approximately 60 °C and one at 74 °C, indicating that the proteins experienced two significant conformational transitions. The second peak corresponded to the complete denaturation of PGUS-P. Notably, the DSC curve of PGUS-P40 showed a third peak, near 72 °C, which indicated that one more dramatic conformational transition occurred in this mutant than in other PGUS-P mutants. This additional transition may be caused by the intermonomer glyco-hairpin enhancing this mutant’s resistance to heat-induced structural collapse. The N-glycans in P. pastoris are unevenly decorated, and the number of sugar groups generally ranges from 11-13.20 To account for the effect of differences in glycoforms on the glyco-hairpin conformation, we added four more mannose moieties to the backbone of Man7GlcNAc2 in the 2WAH crystal structure (based on the typical glycoform found in P. pastoris), thus generating the N-glycan Man11GlcNAc2 (denoted long N-glycan) for MD simulations of PGUS-P40. We performed an MD simulation of homotetrameric PGUS that had a long N-glycan attached to the N40 residue of each monomer (so that a total of four N-glycans were attached). At 200 ns, three of the four long N-glycans were contacting 50-60 atoms with the amino acids of the neighboring monomer (associated with the formation of approximately four hydrogen bonds), thereby indicating formation of three stable glyco-hairpins (Fig.

14


Page 14 of 25

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


S11). Thus, the glycoform of the N-glycan does not affect the formation of glyco-hairpins. In conclusion, we used experiments and all-atom MD simulations to identify a fundamental mechanism by which N-glycans stabilize protein structure; specifically, natural N-glycans function as a “glyco-linker” that holds spatial subdomains together and enhances local stability. Inspired by this result, we successfully designed and introduced various glyco-linkers at interdomain (glyco-bridge) and intermonomer (glyco-hairpin) interfaces where the local interaction was too weak to maintain a stable structure; dramatically, these glyco-linkers improved the protein stability to a much greater extent than the inherent N-glycosylation. Our study not only provides unique insight into the interactions between glycans and proteins from a quaternary structure perspective but also facilitates the rational design of N-glycans as general tools that can enhance protein stability. Supporting Information Experimental methods; Natural N-glycosylation characterization of PGUS-P; MD simulation of PGUS-P decorated with natural N-glycan; Screening process for PGUS-P mutants; Thermostability and dynamic stability of PGUS-P mutants with introduced N-glycan; MD simulation of PGUS-P208, PGUS-P40, and PGUS-P418. DSC thermograms of PGUS-P and mutants; MD simulation of homotetrameric PGUS-P40 with four N-glycans attached. Acknowledgements

15



This research was funded by grants from the National Natural Science Fund for Distinguished Young Scholars (No. 21425624), and the National Natural Science Foundation of China (No. 21506011, No. 31571026, No. 21727817). The authors also acknowledge the Special Program for Applied Research on Super Computation of the NSFC-Guangdong Joint Fund (second phase) under Grant No.U1501501.

References (1) Dalziel, M.; Crispin, M.; Scanlan, C. N.; Zitzmann, N.; Dwek, R. A. Emerging Principles for the Therapeutic Exploitation of Glycosylation. Science 2014, 343, 1235681-1235681. (2) Sun, S.; Shah, P.; Eshghi, S. T.; Yang, W.; Trikannad, N.; Yang, S.; Chen, L.; Aiyetan, P.; Hoti, N.; Zhang, Z.; Chan, D. W.; Zhang, H. Comprehensive Analysis of Protein Glycosylation by Solid-Phase Extraction of N-Linked Glycans and Glycosite-Containing Peptides. Nat. Biotechnol. 2016, 34, 84-88. (3) Nita-Lazar, M.; Wacker, M.; Schegg, B.; Amber, S.; Aebi, M. The N-X-S/T Consensus Sequence Is Required but Not Sufficient for Bacterial N-Linked Protein Glycosylation. Glycobiology 2005, 15, 361-367. (4) Kornfeld, R.; Kornfeld, S. Assembly of Asparagine-Linked Oligosaccharides. Annu. Rev. Biochem. 1985, 54, 631-664. (5) Jiang, J.; Yang, J. F.; Feng, P.; Zuo, B.; Dong, N. Z.; Wu, Q. Y.; He, Y. N-Glycosylation Is Required for Matriptase-2 Autoactivation and Ectodomain Shedding. J. Biol. Chem. 2014, 289, 19500-19507. (6) Shirke, A. N.; Su, A.; Jones, J. A.; Butterfoss, G. L.; Koffas, M. A. G.; Kim, J. R.; Gross, R. A. Comparative Thermal Inactivation Analysis of Aspergillus oryzae and Thiellavia terrestris Cutinase: Role of Glycosylation. Biotechnol. Bioeng. 2017, 114, 63-73. (7) Dotsenko, A. S.; Gusakov, A. V.; Volkov, P. V.; Rozhkova, A. M.; Sinitsyn, A. P. N-Linked Glycosylation of Recombinant Cellobiohydrolase I (Cel7a) from Penicillium verruculosum and Its Effect on the Enzyme Activity. Biotechnol. Bioeng. 2016, 113, 283-91. (8) Ellis, C. R.; Maiti, B.; Noid, W. G. Specific and Nonspecific Effects of Glycosylation. J. Am. Chem. Soc. 2012, 134, 8184-8193. (9) Payne, C. M.; Resch, M. G.; Chen, L. Q.; Crowley, M. F.; Himmel, M. E.; Taylor, L. E.; Sandgren, M.; Stahlberg, J.; Stals, I.; Tan, Z. P.; Beckham, G. T. Glycosylated Linkers in Multimodular Lignocellulose-Degrading Enzymes Dynamically Bind to Cellulose. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 14646-14651. (10) Chen, L. Q.; Drake, M. R.; Resch, M. G.; Greene, E. R.; Himmel, M. E.; Chaffey, P. K.; Beckham, G. T.; Tan, Z. P. Specificity of O-Glycosylation in Enhancing the Stability and 16


Page 16 of 25

Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Cellulose Binding Affinity of Family 1 Carbohydrate-Binding Modules. Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 7612-7617. (11) Culyba, E. K.; Price, J. L.; Hanson, S. R.; Dhar, A.; Wong, C. H.; Gruebele, M.; Powers, E. T.; Kelly, J. W. Protein Native-State Stabilization by Placing Aromatic Side Chains in N-Glycosylated Reverse Turns. Science 2011, 331, 571-575. (12) Price, J. L.; Culyba, E. K.; Chen, W.; Murray, A. N.; Hanson, S. R.; Wong, C.-H.; Powers, E. T.; Kelly, J. W. N-Glycosylation of Enhanced Aromatic Sequons to Increase Glycoprotein Stability. Pept. Sci. 2012, 98, 195-211. (13) Fonseca-Maldonado, R.; Vieira, D. S.; Alponti, J. S.; Bonneil, E.; Thibault, P.; Ward, R. J. Engineering the Pattern of Protein Glycosylation Modulates the Thermostability of a GH11 Xylanase. J. Biol. Chem. 2013, 288, 25522-25534. (14) Hebert, D. N.; Lamriben, L.; Powers, E. T.; Kelly, J. W. The Intrinsic and Extrinsic Effects of N-Linked Glycans on Glycoproteostasis. Nat. Chem. Biol. 2014, 10, 902-910. (15) Wang, X. Y.; Ji, C. G.; Zhang, J. Z. H. Exploring the Molecular Mechanism of Stabilization of the Adhesion Domains of Human Cd2 by N-Glycosylation. J. Phys. Chem. B 2012, 116, 11570-11577. (16) Gavrilov, Y.; Shental-Bechor, D.; Greenblatt, H. M.; Levy, Y. Glycosylation May Reduce Protein Thermodynamic Stability by Inducing a Conformational Distortion. J. Phys. Chem. Lett. 2015, 6, 3572-3577. (17) Price, J. L.; Shental-Bechor, D.; Dhar, A.; Turner, M. J.; Powers, E. T.; Gruebele, M.; Levy, Y.; Kelly, J. W. Context-Dependent Effects of Asparagine Glycosylation on Pin Ww Folding Kinetics and Thermodynamics. J. Am. Chem. Soc. 2010, 132, 15359-15367. (18) Lv, B.; Sun, H.; Huang, S.; Feng, X.; Jiang, T.; Li, C. Structure-Guided Engineering of the Substrate Specificity of a Fungal β-Glucuronidase toward Triterpenoid Saponins. J. Biol. Chem. 2018, 293, 433-443. (19) Yang, M.; Yu, X. W.; Zheng, H. Y.; Sha, C.; Zhao, C. F.; Qian, M. Q.; Xu, Y. Role of N-Linked Glycosylation in the Secretion and Enzymatic Properties of Rhizopus chinensis Lipase Expressed in Pichia pastoris. Microb. Cell. Fact. 2015, 14, 40. (20) Krainer, F. W.; Gmeiner, C.; Neutsch, L.; Windwarder, M.; Pletzenauer, R.; Herwig, C.; Altmann, F.; Glieder, A.; Spadiut, O. Knockout of an Endogenous Mannosyltransferase Increases the Homogeneity of Glycoproteins Produced in Pichia pastoris. Sci. Rep. 2013, 3, 3279. (21) Petrescu, A. J.; Milac, A. L.; Petrescu, S. M.; Dwek, R. A.; Wormald, M. R. Statistical Analysis of the Protein Environment of N-Glycosylation Sites: Implications for Occupancy, Structure, and Folding. Glycobiology 2004, 14, 103-114. (22) Murray, A. N.; Chen, W. T.; Antonopoulos, A.; Hanson, S. R.; Wiseman, R. L.; Dell, A.; Haslam, S. M.; Powers, D. L.; Powers, E. T.; Kelly, J. W. Enhanced Aromatic Sequons Increase Oligosaccharyltransferase Glycosylation Efficiency and Glycan Homogeneity. Chem. Biol. 2015, 22, 1052-1062. (23) Price, J. L.; Powers, D. L.; Powers, E. T.; Kelly, J. W. Glycosylation of the Enhanced Aromatic Sequon Is Similarly Stabilizing in Three Distinct Reverse Turn Contexts. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 14127-14132.

17



(24) Chen, W.; Kong, L.; Connelly, S.; Dendle, J. M.; Liu, Y.; Wilson, I. A.; Powers, E. T.; Kelly, J. W. Stabilizing the CH2 Domain of an Antibody by Engineering in an Enhanced Aromatic Sequon. ACS Chem. Biol. 2016, 11, 1852-1861. (25)Hsu, C. H.; Park, S.; Mortenson, D. E.; Foley, B. L.; Wang, X. C.; Woods, R. J.; Case, D. A.; Powers, E. T.; Wong, C. H.; Dyson, H. J.; Kelly, J. W. The Dependence of Carbohydrate-Aromatic Interaction Strengths on the Structure of the Carbohydrate. J. Am. Chem. Soc. 2016, 138, 7636-7648. (26) Bosshart, A.; Panke, S.; Bechtold, M. Systematic Optimization of Interface Interactions Increases the Thermostability of a Multimeric Enzyme. Angew. Chem.-Int. Edit. 2013, 52, 9673-9676.

18


Page 18 of 25

Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure legend Figure 1. The display of 24 loops/turns chosen for introducing glyco-linker at tertiary and quaternary structure of PGUS. The targeted 24 loops/turns in chain B at the interface of domains and monomers are shown in magenta and red, respectively. Figure 2. Design of a glyco-linkers to modulate the stability of PGUS. (a) SDS-PAGE analysis of PGUS-P40, PGUS-P208, and PGUS-P418 and their counterparts PGUS-E40, PGUS-E208, and PGUS-E418, which were expressed in E.

coli; “-” denotes mutants not treated by PNGase F, “+” denotes mutants treated by PNGase F. (b-d) The kinetic stability of PGUS-P208 (b), PGUS-P40 (c) and PGUS-P418 (d). The results are reported from triplicate measurements, and error bars represent the average ± one standard deviation. Figure 3. (a-c) The RMSD of PGUS-P208 (a), PGUS-P40 (b) and PGUS-P418 (c) from 200 ns MD simulation. (d) A 200 ns snapshot showing the glyco-bridge conformation of the N-glycan of PGUS-P208. (e) A 200 ns snapshot shows the glyco-hairpin conformation of the N-glycan of PGUS-P40. (f) A 200 ns snapshot showing the unstable conformation of the N-glycan of PGUS-P418. The bulk structure of PGUS is shown as ribbon in which the sugar-binding, β-sandwich and TIM barrel domains are shown in magenta, cyan and yellow, respectively, and the N-glycan is shown as purple spheres Figure 4. (a) The interaction network analysis between the N-glycan at N208 and the surrounding residues. (b) The interaction network between the N-glycan at N40 of Chain B and residues of Chain A. The ribbon represents the PGUS structure, and the 19



purple sticks represent the N-glycan. The residues of PGUS-P that interact with the glycan are shown as orange sticks, and the polar interactions are shown as gold dashed lines. Residue N40 in Chain B is shown in red sticks and is located in the loop 23-42 of Chain B (cyan). The right panel shows a schematic diagram of the detailed interaction network between PGUS and the N-glycan. Figure 5. The contact atom number (left panel) and hydrogen bond number (right panel) between PGUS and N-glycan (9 sugar residues) of PGUS-P208 (a) and PGUS-P40 (b) during 200 ns MD simulation.

20


Page 20 of 25

Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1 64x50mm (600 x 600 DPI)



Figure 2 112x84mm (600 x 600 DPI)


Page 22 of 25

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3 94x108mm (600 x 600 DPI)



Figure 4 115x95mm (600 x 600 DPI)


Page 24 of 25

Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5 108x91mm (600 x 600 DPI)


Design of Glyco-Linkers at Multiple Structural Levels to Modulate

Recommend Documents