Rational Engineering of a Designed Protein Cage for siRNA Delivery

Aug 9, 2018 - We describe a simple approach to transform a designed porous protein cage into a nucleic acid delivery vehicle. By introducing arginine ...
0 downloads 0 Views 2MB Size
Download PDF
Communication Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

pubs.acs.org/JACS

Rational Engineering of a Designed Protein Cage for siRNA Delivery Thomas G. W. Edwardson, Takahiro Mori, and Donald Hilvert* Laboratory of Organic Chemistry, ETH Zurich, 8093 Zurich, Switzerland

J. Am. Chem. Soc. Downloaded from pubs.acs.org by WASHINGTON UNIV on 08/09/18. For personal use only.

S Supporting Information *

To assess this possibility, we devised a simple approach to transform a nonfunctional protein cage into an efficient oligonucleotide delivery device. As a starting scaffold, we chose O3-33, a nonviral selfassembling protein cage computationally designed by Baker and co-workers.38 This highly stable assembly comprises 24 monomers arranged in octahedral geometry with an external diameter of ∼13 nm, a spherical internal cavity of ∼8 nm, and six pores of ∼3.5 nm (Figure 1a). The complete cage is easily obtained by overexpression in Escherichia coli, and recently, through use as a scaffold for the formation of extracellular

ABSTRACT: Oligonucleotide therapeutics have transformative potential in modern medicine but are poor drug candidates in themselves unless fitted with compensatory carrier systems. We describe a simple approach to transform a designed porous protein cage into a nucleic acid delivery vehicle. By introducing arginine mutations to the lumenal surface, a positively supercharged capsule is created, which can encapsidate oligonucleotides in vitro with high binding affinity. We demonstrate that the siRNA-loaded cage is taken up by mammalian cells and releases its cargo to induce RNAi and knockdown gene expression. These general concepts could also be applied to alternative scaffold designs, expediting the development of artificial protein cages toward delivery applications.

R

ecent advances in computational protein engineering have provided assemblies of numerous forms,1−4 which have proven useful across broad areas of application.5−10 Designs that create well-defined nanoscale compartments are of particular interest.11−16 Similar protein compartments appear in nature and are intimately tied to important cellular processes, serving as enzymatic reaction vessels,17 storage containers,18,19 chaperones for protein (un)folding and degradation,20,21 and even chambers with yet unknown purposes.22,23 The packaging, transport, and delivery of cargo within biological systems is one area where proteinaceous capsules perform particularly well; viruses are a convincing example.24 Although oligonucleotide-based therapeutics25 have wide reaching clinical potential, their suboptimal pharmacokinetic properties have hindered development.26 A major barrier is delivery to the cytosol, where oligonucleotides usually elicit their therapeutic effect.27 Further challenges include susceptibility to nucleases present in biological sera, activation of immune responses, and rapid clearance from the body.28,29 All of these hurdles could be overcome by efficient delivery systems, and a wide range of materials has been exploited to this end.26,30−32 However, effective, nontoxic delivery solutions are still sought and designed protein cages are promising candidates. Protein nanostructures offer uniform size, atomically resolved structure, broad chemical functionality, and inherent biodegradability. Additionally, recombinant protein production is simple, scalable, economical, and allows both genetic and chemical modification. Both viral capsids and vault ribonucleoproteins have been engineered for oligonucleotide delivery.33−37 However, the advent of de novo-designed protein cages offers an alternative to this repurposing approach. Through judicious design, artificial protein cages should be engineerable for specific tasks, while avoiding inherent, potentially undesirable features. © XXXX American Chemical Society

Figure 1. (a) Surface models of the OP cage viewed along 4-fold and 2fold symmetry axes, with one trimer shown as a colored ribbon. (b) Arginine mutations in the monomer and (c) trimer are shown as blue spheres. (d) Negative stain transmission electron micrograph of OP. (e) Superimposed monomers of O3-33 (gray) and OP (orange) from X-ray diffraction data (PDB ID for OP: 6FDB). Received: June 19, 2018

A

DOI: 10.1021/jacs.8b06442 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society vesicles, its suitability for biological applications has been demonstrated.39 To transform this naı̈ve container into a nucleic acid delivery system, we first introduced binding interactions to promote cargo loading. Electrostatic interactions provide a strong driving force for the formation of host−guest complexes40,41 and are well suited for binding highly anionic nucleic acids.42−45 As such, six solvent-exposed residues on the lumenal surface (Thr11, Pro39, Glu66, Trp103, Phe130, and Leu163) were replaced with arginine to create an O3-33 variant, referred to as OP, which has a highly positively charged interior cavity (Figure 1b, Supporting Information (SI) Figure S1). The OP protein can be overexpressed in E. coli cells and isolated by Ni-NTA affinity chromatography through a Cterminal His6 tag (SI Figure S2). Because of the many arginine residues, there was significant contamination from endogenous E. coli RNA. However, these guests were easily removed with high ionic strength buffer to weaken electrostatic interactions and RNase A to digest contaminant RNA. The ability of OP to form the desired cage-like quaternary structure was assessed by size-exclusion chromatography (SEC), dynamic light scattering (DLS), and native agarose gel electrophoresis (AGE) (SI Figures S3−S4). All data were consistent with the formation of an assembly with the same size as the original O3-33 cage. Furthermore, OP and O3-33 were indistinguishable by Transmission electron microscopy (TEM) (Figure 1d, SI Figure S5) and X-ray diffraction confirmed that the addition of 144 arginine residues to the ∼8 nm diameter cavity had negligible effect on monomer fold and cage structure (Figure 1e). Rootmean-square deviations were 0.5−0.7 Å for all Cα atoms. With the structural integrity of the empty protein cage established, we investigated nucleic acid cargo loading in vitro. It was expected that DNA and RNA duplexes, which are 2 nm in diameter, would be able to pass through the ∼3.5 nm diameter pores, allowing encapsulation. Electrophoretic mobility shift assays (EMSAs)46 revealed that the parent O3-33 cage had no effect on the electrophoretic mobility of DNA or RNA (SI Figures S6−S7). In contrast, colocalization of protein and oligonucleotide bands was observed for OP with 21 nt singlestranded (ss) DNA (Figure 2a) and double-stranded (ds) DNA and RNA (SI Figure S8). In all cases, near complete binding of the oligonucleotides was observed upon addition of 0.5 equiv of OP, suggesting two guests per cage. This stoichiometry was confirmed from the A260/A280 ratio of SEC-purified complexes of OP with ssDNA, dsDNA, or dsRNA (Figure 2b, SI Figure S11). We took advantage of encapsulation-induced fluorescence quenching of the DNA-conjugated fluorophore Atto488 to determine binding kinetics (SI Figure S13). The association rate constant, kon = (6.6 ± 0.6) × 105 M−1 s−1 (Figure 2c), and dissociation rate constant, koff = (1.6 ± 0.1) × 10−4 s−1 (Figure 2d), were used to calculate a dissociation constant, Kd = (2.4 ± 0.3) × 10−10 M, for this protein−DNA complex. This Kd value is 2 orders of magnitude lower than those reported for RNA− protein interactions in viruses (10−8−10−7 M)47 and in the range of the tightest binding transcription factors (10−15−10−10 M).48 However, a crucial difference in this artificial system is the lack of sequence specificity, which is advantageous for a delivery device, as it allows packaging of any desired nucleotide sequence (SI Figure S9), and oligonucleotide chemical modifications that retain negative charge.49 An important role of any nucleic acid delivery system is to protect its cargo from nuclease degradation. As such, we conducted digestion assays with three nucleases: DNase I (30

Figure 2. (a) Electrophoretic mobility shift assay of 21 nt Atto488labeled ssDNA with OP, visualized by Atto488 fluorescence. (b) Absorbance spectra of SEC-purified OP:nucleic acid complexes and calculated number of guests per cage (inset). (c) Plot of kobs values, from fluorescence quenching of Atto488-ssDNA cargo upon addition of OP, vs [OP] to determine kon. (d) Plot of fluorescence recovery of Atto488-ssDNA released from OP cages induced by addition of a 100fold excess of unlabeled competitor DNA.

kDa), RNase A (14 kDa), and benzonase50 (60 kDa). On the basis of these molecular weights and the OP pore size, it was expected that only RNase A would be able to enter the cage and digest the cargo. This hypothesis was confirmed by AGE (SI Figure S14), revealing protection of the cargo from the larger enzymes but not RNase A. Although susceptibility to RNase A is undesirable, careful engineering of the pores may afford OP variants that provide better protection.44,51 These data corroborate the kinetic studies, suggesting that guests are tightly bound in the inner cavity with a negligible dissociation rate. With many nucleic acid delivery vectors, such as cationic lipids and polymers, a compromise must be made between transfection efficiency and cytotoxicity.52 We measured in vitro cytotoxicity in HeLa cells after 24 and 48 h using WST-8 to quantify dehydrogenase activity (SI Figure S15) and found that OP has negligible toxicity up to 400 nM cage (193 μg/mL protein). These results compare favorably with the commonly used transfection reagent Lipofectamine 2000, a cationic lipid, which reduced cell viability to 35−45% at 19-fold lower concentrations by mass (10 μg/mL) under our assay conditions. Next, the ability of the cage to enter cells and deliver its nucleic acid cargo was investigated. Seven different cell lines were treated with 200 nM of either free Atto488-ssDNA or Atto488-ssDNA packaged within OP cages, and the cellular fluorescence was measured by flow cytometry. Efficient cellular uptake of OP cages was observed in six out of the seven cell lines tested (SI Figure S17). For example, a plot of median fluorescence intensity (Figure 3a) shows an almost 30-fold increase in cellular uptake of ssDNA over background in HeLa cells. An even larger effect was seen with Vero and HepG2 cells. To confirm, confocal fluorescence microscopy (CFM) was carried out on cells treated with either free Atto488-ssDNA or Atto488-ssDNA-loaded OP cages (SI Figure S18). Cells treated with free Atto488-DNA exhibited no appreciable fluorescence, while cells treated with OP:Atto488-DNA had intracellular B

DOI: 10.1021/jacs.8b06442 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

proton sponge effect55 and the ability of protonated histidine to permeabilize endosomal membranes.56 As each OP monomer has a His6 tag, the complete 24-mer cage presents a total of 144 histidine residues on its exterior. To test whether this affected siRNA delivery, we produced a variant of OP that has no His6 tag and assayed it for RNAi activity (SI Figure S22 and S-IIs.). Under the same experimental conditions we observed only a ∼40% reduction in GFP fluorescence (Figure 4a), approximately half as effective as OP, indicating that the His6 tags indeed contribute to delivery efficiency. Second, the release of cargo molecules in the cytoplasm can be explained in terms of the kinetic parameters shown in Figure 2d. While OP−nucleic acid complexes are stable, encapsulation is reversible in the presence of high concentrations of competing guest molecules. Thus, we expected that high cytosolic concentrations of tRNA could provide a location-specific release mechanism for the siRNA cargo, liberating it for RNAi induced gene knockdown. To validate this hypothesis in vitro, we treated OP cages containing A488-dsDNA with intracellular concentrations of tRNA (26 μM)57 in PBS at 37 °C and observed timedependent cargo release over 6 h (Figure 4b). Conversely, the same experiment carried out with 10% fetal bovine serum showed no release or degradation of cargo (SI Figure S24). For oligonucleotides that act upon mRNA in the cytoplasm, taking advantage of cytosolic physicochemical properties offers a simple solution for cargo release. It should be possible to finetune this mechanism, as the binding affinity could be adjusted by simply altering the number of arginines displayed on the cage interior. In summary, we have established the first steps toward development of nonviral protein containers for nucleic acid delivery. By introducing minimal mutations to generate function, in this case a simple electrostatic encapsulation strategy, a naı̈ve computationally designed scaffold was transformed into an efficient oligonucleotide carrier. While this carrier system has not yet been optimized, the amenability of the protein to both genetic and chemical modification should facilitate further engineering to address issues, such as nuclease susceptibility, targeting and potential immunogenicity. Additionally, the electrostatic-based encapsidation could be applied in alternative scaffold designs. These findings augur well for the further development of artificial, designed protein cages for diverse applications in biology and medicine.

Figure 3. (a) Cellular uptake of Atto488-DNA with or without OP in HeLa cells, determined by flow cytometry (n = 3). (b) Confocal fluorescence microscopy of HeLa cells treated with Atto488-DNA or OP:Atto488-DNA. Blue, Nuclei (Hoechst stained); Green, Atto488.

fluorescence localized in many intense foci, corroborating the high uptake efficiency (Figure 3b). These punctate structures are suggestive of endocytosis, which is often observed for nanoparticles in this size regime.53 With the high cellular uptake of OP cages confirmed, we tested the ability of the cages to deliver siRNA and modulate gene expression using HeLa cells that stably express green fluorescent protein (GFP) as a model to measure protein expression. Using Lipofectamine 2000 as a positive control, we treated GFP-expressing HeLa cells with naked siRNA, OPencapsulated siRNA, empty cage, and OP cages loaded with a scrambled siRNA. To determine gene knockdown efficiency, GFP fluorescence was measured qualitatively, by fluorescence microscopy (SI Figure S19), and quantitatively, by flow cytometry (Figure 4a). After treatment with 20 pmol of OP-

Figure 4. (a) GFP knockdown determined by flow cytometry (n = 4). OP:scramble refers to a scrambled siRNA not targeting the GFP mRNA or other endogenous sequences, and No His6 refers to siRNA packaged in OP cages lacking His6 tags. (b) Native AGE showing tRNA-induced release of dsDNA cargo from OP over time. Visualized by Atto488 fluorescence.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.8b06442.

packaged siRNA and 48 h incubation, we observed a ∼70% reduction in GFP fluorescence, an efficacy within the range of Lipofectamine 2000. Moreover, because of the lower toxicity of OP, cell viability was improved relative to the commercial reagent (SI Figures S15, S19). The lack of effect of OP cages loaded with a scrambled sequence siRNA confirmed that expression levels were not due to induction of the antiviral interferon pathway machinery and were target mRNA specific.54 Successful induction of RNAi requires delivery of siRNA to the cytoplasm, suggesting that OP cages can somehow achieve endosomal escape and release their cargo. Although an in-depth study must follow to probe the exact mechanisms, we can suggest two explanations. First, the ability of histidine-rich peptides to facilitate endosomal escape has been well-documented, both due to the

Materials, methods, and supplementary data (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*[email protected] ORCID

Thomas G. W. Edwardson: 0000-0001-8661-8036 Donald Hilvert: 0000-0002-3941-621X Notes

The authors declare no competing financial interest. C

DOI: 10.1021/jacs.8b06442 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Communication

Journal of the American Chemical Society

■

(25) Lundin, K. E.; Gissberg, O.; Smith, C. I. E. Hum. Gene Ther. 2015, 26, 475. (26) Whitehead, K. A.; Langer, R.; Anderson, D. G. Nat. Rev. Drug Discovery 2009, 8, 129. (27) Dowdy, S. F. Nat. Biotechnol. 2017, 35, 222. (28) Gantier, M. P.; Williams, B. R. Cytokine Growth Factor Rev. 2007, 18, 363. (29) Iversen, F.; Yang, C.; Dagnaes-Hansen, F.; Schaffert, D. H.; Kjems, J.; Gao, S. Theranostics 2013, 3, 201. (30) Ding, Y.; Jiang, Z.; Saha, K.; Kim, C. S.; Kim, S. T.; Landis, R. F.; Rotello, V. M. Mol. Ther. 2014, 22, 1075. (31) Lehto, T.; Ezzat, K.; Wood, M. J. A.; El Andaloussi, S. Adv. Drug Delivery Rev. 2016, 106, 172. (32) Seeman, N. C.; Sleiman, H. F. Nat. Rev. Mater. 2017, 3, 17068. (33) Qazi, S.; Miettinen, H. M.; Wilkinson, R. A.; McCoy, K.; Douglas, T.; Wiedenheft, B. Mol. Pharmaceutics 2016, 13, 1191. (34) Choi, K.-m.; Choi, S.-H.; Jeon, H.; Kim, I.-S.; Ahn, H. J. ACS Nano 2011, 5, 8690. (35) Fang, P. Y.; Gomez Ramos, L. M.; Holguin, S. Y.; Hsiao, C.; Bowman, J. C.; Yang, H. W.; Williams, L. D. Nucleic Acids Res. 2017, 45, 3519. (36) Choi, K.-m.; Kim, K.; Kwon, I. C.; Kim, I.-S.; Ahn, H. J. Mol. Pharmaceutics 2013, 10, 18. (37) Han, M.; Kickhoefer, V. A.; Nemerow, G. R.; Rome, L. H. ACS Nano 2011, 5, 6128. (38) King, N. P.; Sheffler, W.; Sawaya, M. R.; Vollmar, B. S.; Sumida, J. P.; André, I.; Gonen, T.; Yeates, T. O.; Baker, D. Science 2012, 336, 1171. (39) Votteler, J.; Ogohara, C.; Yi, S.; Hsia, Y.; Nattermann, U.; Belnap, D. M.; King, N. P.; Sundquist, W. I. Nature 2016, 540, 292. (40) Zschoche, R.; Hilvert, D. J. Am. Chem. Soc. 2015, 137, 16121. (41) Tetter, S.; Hilvert, D. Angew. Chem., Int. Ed. 2017, 56, 14933. (42) Belyi, V. A.; Muthukumar, M. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 17174. (43) Lilavivat, S.; Sardar, D.; Jana, S.; Thomas, G. C.; Woycechowsky, K. J. J. Am. Chem. Soc. 2012, 134, 13152. (44) Butterfield, G. L.; Lajoie, M. J.; Gustafson, H. H.; Sellers, D. L.; Nattermann, U.; Ellis, D.; Bale, J. B.; Ke, S.; Lenz, G. H.; Yehdego, A.; Ravichandran, R.; Pun, S. H.; King, N. P.; Baker, D. Nature 2017, 552, 415. (45) Azuma, Y.; Edwardson, T. G. W.; Terasaka, N.; Hilvert, D. J. Am. Chem. Soc. 2018, 140, 566. (46) Hellman, L. M.; Fried, M. G. Nat. Protoc. 2007, 2, 1849. (47) Bayer, T. S.; Booth, L. N.; Knudsen, S. M.; Ellington, A. D. RNA 2005, 11, 1848. (48) Slutsky, M.; Mirny, L. A. Biophys. J. 2004, 87, 4021−4035. (49) Deleavey, G. F.; Damha, M. J. Chem. Biol. 2012, 19, 937. (50) Benedik, M. J.; Strych, U. FEMS Microbiol. Lett. 1998, 165, 1. (51) Terasaka, N.; Azuma, Y.; Hilvert, D. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 5432. (52) Lv, H.; Zhang, S.; Wang, B.; Cui, S.; Yan, J. J. Controlled Release 2006, 114, 100. (53) Sahay, G.; Alakhova, D. Y.; Kabanov, A. V. J. Controlled Release 2010, 145, 182. (54) Seth, R. B.; Sun, L.; Chen, Z. J. Cell Res. 2006, 16, 141. (55) Akinc, A.; Thomas, M.; Klibanov, A. M.; Langer, R. J. Gene Med. 2005, 7, 657. (56) Pichon, C.; Gonçalves, C.; Midoux, P. Adv. Drug Delivery Rev. 2001, 53, 75. (57) See SI Section S-IIt for the estimation of intracellular tRNA concentration.

ACKNOWLEDGMENTS We thank B. Blattmann (Protein Crystallization Core Facility, University of Zurich) and the staff at the Swiss Light Source (Paul Scherrer Institute). We also thank the Scientific Center for Optical and Electron Microscopy (ScopeM), ETH Zurich, for help with TEM and CFM experiments and the Flow Cytometry Core Facility, ETH Zurich, for technical support. Additionally, we are grateful to the staff at EMBL Hamburg (DESY), for aid with SAXS measurements. This work was supported by the ETH Zurich and the European Research Council (Advanced ERC grant ERC-AdG-2012-321295 to D.H.). T.E. is very grateful to the Human Frontier Science Program for a long-term fellowship.

■

REFERENCES

(1) Yeates, T. O.; Liu, Y.; Laniado, J. Curr. Opin. Struct. Biol. 2016, 39, 134. (2) Huang, P.-S.; Boyken, S. E.; Baker, D. Nature 2016, 537, 320. (3) Baker, E. G.; Bartlett, G. J.; Porter Goff, K. L.; Woolfson, D. N. Acc. Chem. Res. 2017, 50, 2085. (4) Ljubetič, A.; Gradišar, H.; Jerala, R. Curr. Opin. Chem. Biol. 2017, 40, 65. (5) Liu, Y.; Zhang, X.; Tan, Y. L.; Bhabha, G.; Ekiert, D. C.; Kipnis, Y.; Bjelic, S.; Baker, D.; Kelly, J. W. J. Am. Chem. Soc. 2014, 136, 13102. (6) Kries, H.; Blomberg, R.; Hilvert, D. Curr. Opin. Chem. Biol. 2013, 17, 221. (7) MacPhee, C. E.; Woolfson, D. N. Curr. Opin. Solid State Mater. Sci. 2004, 8, 141. (8) Phippen, S. W.; Stevens, C. A.; Vance, T. D. R.; King, N. P.; Baker, D.; Davies, P. L. Biochemistry 2016, 55, 6811. (9) Strauch, E.-M.; Bernard, S. M.; La, D.; Bohn, A. J.; Lee, P. S.; Anderson, C. E.; Nieusma, T.; Holstein, C. A.; Garcia, N. K.; Hooper, K. A.; Ravichandran, R.; Nelson, J. W.; Sheffler, W.; Bloom, J. D.; Lee, K. K.; Ward, A. B.; Yager, P.; Fuller, D. H.; Wilson, I. A.; Baker, D. Nat. Biotechnol. 2017, 35, 667. (10) Bick, M. J.; Greisen, P. J.; Morey, K. J.; Antunes, M. S.; La, D.; Sankaran, B.; Reymond, L.; Johnsson, K.; Medford, J. I.; Baker, D. eLife 2017, 6, e28909. (11) Padilla, J. E.; Colovos, C.; Yeates, T. O. Proc. Natl. Acad. Sci. U. S. A. 2001, 98, 2217. (12) Fletcher, J. M.; Harniman, R. L.; Barnes, F. R. H.; Boyle, A. L.; Collins, A.; Mantell, J.; Sharp, T. H.; Antognozzi, M.; Booth, P. J.; Linden, N.; Miles, M. J.; Sessions, R. B.; Verkade, P.; Woolfson, D. N. Science 2013, 340, 595. (13) Hsia, Y.; Bale, J. B.; Gonen, S.; Shi, D.; Sheffler, W.; Fong, K. K.; Nattermann, U.; Xu, C.; Huang, P.-S.; Ravichandran, R.; Yi, S.; Davis, T. N.; Gonen, T.; King, N. P.; Baker, D. Nature 2016, 535, 136. (14) Sciore, A.; Su, M.; Koldewey, P.; Eschweiler, J. D.; Diffley, K. A.; Linhares, B. M.; Ruotolo, B. T.; Bardwell, J. C. A.; Skiniotis, G.; Marsh, E. N. G. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 8681. (15) Bale, J. B.; Gonen, S.; Liu, Y.; Sheffler, W.; Ellis, D.; Thomas, C.; Cascio, D.; Yeates, T. O.; Gonen, T.; King, N. P.; Baker, D. Science 2016, 353, 389. (16) Lapenta, F.; Aupic, J.; Strmsek, Z.; Jerala, R. Chem. Soc. Rev. 2018, 47, 3530. (17) Kerfeld, C. A.; Heinhorst, S.; Cannon, G. C. Annu. Rev. Microbiol. 2010, 64, 391. (18) Liu, X.; Theil, E. C. Acc. Chem. Res. 2005, 38, 167. (19) Walsby, A. E. Microbiol. Rev. 1994, 58, 94. (20) Saibil, H. Nat. Rev. Mol. Cell Biol. 2013, 14, 630. (21) Coux, O.; Tanaka, K.; Goldberg, A. L. Annu. Rev. Biochem. 1996, 65, 801. (22) Kedersha, N. L.; Rome, L. H. J. Cell Biol. 1986, 103, 699. (23) Axen, S. D.; Erbilgin, O.; Kerfeld, C. A. PLoS Comput. Biol. 2014, 10, e1003898. (24) Lodish, H. B. A.; Zipursky, S. L., Viruses: Structure, Function, and Uses. In Molecular Cell Biology, 4th ed.; W.H. Freeman: New York, 2000. D

DOI: 10.1021/jacs.8b06442 J. Am. Chem. Soc. XXXX, XXX, XXX−XXX