- Email: [email protected]
NMR Mapping of Disordered Segments from a Viral Scaffolding Protein Enclosed in a 23 MDa Procapsid
Biophysical Letter
NMR Mapping of Disordered Segments from a Viral Scaffolding Protein Enclosed in a 23 MDa Procapsid Richard D. Whitehead III,1 Carolyn M. Teschke,1,2,* and Andrei T. Alexandrescu1,* 1
Department of Molecular and Cell Biology and 2Department of Chemistry, University of Connecticut, Storrs, Connecticut
ABSTRACT Scaffolding proteins (SPs) are required for the capsid shell assembly of many tailed double-stranded DNA bacteriophages, some archaeal viruses, herpesviruses, and adenoviruses. Despite their importance, only one high-resolution structure is available for SPs within procapsids. Here, we use the inherent size limit of NMR to identify mobile segments of the 303-residue phage P22 SP free in solution and when incorporated into a 23 MDa procapsid complex. Free SP gives NMR signals from its acidic N-terminus (residues 1–40) and basic C-terminus (residues 264–303), whereas NMR signals from the middle segment (residues 41–263) are missing because of intermediate conformational exchange on the NMR chemical shift timescale. When SP is incorporated into P22 procapsids, NMR signals from the C-terminal helix-turn-helix domain disappear because of binding to the procapsid interior. Signals from the N-terminal domain persist, indicating that this segment retains flexibility when bound to procapsids. The unstructured character of the N-terminus, coupled with its high content of negative charges, is likely important for dissociation and release of SP during the double-stranded DNA genome packaging step accompanying phage maturation.
SIGNIFICANCE Scaffolding protein (SP) nucleates the assembly of phage P22 coat proteins into an icosahedral capsid structure that encapsidates the viral genome. NMR spectra of free SP show signals from the N-terminus and a helix-turnhelix domain at the C-terminus. When SP is incorporated into phage P22 empty procapsid shells to form a 23 MDa complex, NMR signals from the N-terminal 40 residues persist indicating this segment is disordered. The unfolded nature of the N-terminus coupled with its negatively charged character is important for the functional requirement of SP to exit the capsid as it becomes packaged with its genome.
INTRODUCTION Viruses infect organisms in all kingdoms of life; those that infect bacteria are called bacteriophages or phages. In many viruses and phages, scaffolding proteins (SPs) are required to ensure the correct organization of coat proteins (CPs) and other minor capsid proteins into a precursor structure, called a procapsid (1,2). Although SPs are critical for viral assembly and therefore potential therapeutic targets, their structural properties, with few exceptions (3,4), are poorly understood. Here, we investigate the structure of full-length SP from bacteriophage P22. Phage P22 is a well-characterized model for the class of tailed bacterial viruses with double-stranded DNA (dsDNA) genomes. Between 60 and 300 copies of a 303-residue SP (33.6 kDa) catalyze the assembly of 415 copies of CP (46.8 kDa) to Submitted May 29, 2019, and accepted for publication August 30, 2019. *Correspondence: [email protected] or andrei.alexandrescu@ uconn.edu Editor: David Eliezer. https://doi.org/10.1016/j.bpj.2019.08.038
form the icosahedral T ¼ 7 procapsid of phage P22 (23 MDa) (5). The SPs are also required for incorporation of the dodecameric portal protein complex at one vertex of the icosahedron, as well as other minor internal proteins (6–8). SPs are not found in mature virions because they are released without proteolysis through holes in the procapsid during genome packaging. The SP is then recycled for subsequent use in multiple rounds of procapsid assembly (9). Without SP, phage P22 CP fails to assemble into normal T ¼ 7 procapsids, yielding instead aberrant ‘‘petite’’ T ¼ 4 and T ¼ 7 particles or spiral structures, all of which are noninfectious (10). Lack of proper assembly in the absence of SP is a general feature of these viruses. Although the functional domains of SP are well characterized (11–14), its structure has proven elusive because of the unusual dynamic properties of the protein (15). The amino acid sequence of SP has a high content of charged residues (31%) and a low content of hydrophobic residues, which classify it as an intrinsically disordered protein (16). The distribution of charged amino acids in SP is asymmetric, with a highly acidic N-terminus (pI ¼ 4.4 for residues
Ó 2019 Biophysical Society.
Biophysical Journal 117, 1387–1392, October 15, 2019 1387
Whitehead et al.
1–130) and basic C-terminus (pI ¼ 9.2 for residues 131–303). This charge distribution suggests that SP belongs to the group of ‘‘electrostatic hairpin’’ intrinsically disordered protein structures (16). Biochemical data are consistent with the disordered nature of SP predicted from its sequence. The protein exhibits a noncooperative unfolding transition characteristic of molten globule conformations, which have a secondary structure but a liquid-like fluctuating tertiary structure (12,15). SP also exhibits rapid amide proton hydrogen exchange typical of marginally stable secondary structure (15). These properties are not unique for the SP of P22; conformational flexibility is a hallmark of viral SPs that appears to be required for their functionality (1,2,17). Nevertheless, the SP of P22, as well as those from phages l, f29, SPP1, and T4, have a high content of predicted and/or measured a-helical secondary structure (17–21). Moreover, in the SP of phage P22, the C-terminal domain, which is critical for CP binding during procapsid assembly, adopts an independently folded helix-turn-helix (HTH) domain structure, as determined by NMR of a 238–303 fragment of SP (3). In this fragment, residues 238–267 are disordered, and the last 35 residues 268–303 fold into the HTH domain (3). Thus, although SP lacks a fixed global hydrophobic core, it is organized into distinct functional domains that are used to fulfill the protein’s multiple roles in the procapsid assembly process (14). In previous biochemical cross-linking studies, we saw evidence that P22 SP undergoes conformational changes during encapsulation into procapsids (22). Here, we use the intrinsic size limit of NMR to define the regions of SP that remain unfolded after assembly.
The CD spectrum shifts to one typical of an unfolded protein as the protein becomes acid denatured at low pH. Consistently, the 1H-15N heteronuclear single quantum coherence (HSQC) NMR spectra of SP at low pH under acid-denaturing conditions are typical of an unfolded protein. About 300 1H-15N HSQC crosspeaks are observed for the 303-residue SP protein at pH 2.0, with all of the 1HN chemical shifts falling in the random coil region between 8.7 and 7.7 ppm (Fig. 2). At pH 6.0, at which CD indicates that SP has substantial a-helix structure (Fig. 1), the chemical shifts remain close to the random coil region, but 2/3 of the 1H-15N HSQC correlations expected from the 303-residue sequence are lost. Sequence-specific NMR assignments for the phage
Results Circular dichroism (CD) spectra show that free SP has a markedly a-helical character at physiological pH (Fig. 1).
FIGURE 1 Far-ultraviolet CD of the SP from P22 (3 mM) as a function of pH. The data were obtained at 30 C. To see this figure in color, go online.
1388 Biophysical Journal 117, 1387–1392, October 15, 2019
FIGURE 2 Representative 600 MHz 1H-15N HSQC spectra from a pH titration of 50 mM P22 SP at a temperature of 35 C. At lower pH, more correlations are seen because of unfolding of residues 41–263 in the central part of SP (see Fig. 1).
Biophysical Letter
P22 SP (Fig. S1; Table S1) indicate the surviving 100 1H-15N HSQC peaks at pH 6.0 are mostly from the N-terminal residues 1–40 that have chemical shifts consistent with unfolded random coil conformations and the C-terminal residues 264–303 that correspond to the folded HTH domain of SP responsible for binding CP (3). Many of the 1HN NMR signals from the C-terminal HTH segment are shifted upfield, with backbone carbon chemical shifts (Table S1) that are also consistent with the a-helical NMR structure of the HTH domain (3). The NMR structure of the HTH domain was determined at pH 4.4 (3). The ellipticity at 222 nm, characteristic of a-helical structure, nearly doubles between pH 4.0 and pH 7.4 (Fig. 1). The transition to a-helical structure with increasing pH therefore cannot be due to the HTH domain alone. Rather, a substantial amount of a-helical structure must be formed in the central 41–263 segment, as SP becomes increasingly folded toward neutral pH. The majority of signals from the central portion of SP, between residues 41–263, were not observed in the 1H-15N HSQC spectra near neutral pH. Biophysical studies indicate that this central portion of SP consists of loosely organized domains of a-helical structure, connected by turns and random coil segments (12,15,23). Furthermore, SP exists in a rapidly reversible monomer-dimer-tetramer dynamic equilibrium, at which the dimers are thought to be the catalytically active species in procapsid assembly (24,25). Broadening of NMR resonances from the 41–263 segment of SP, accompanying the folding transition to a-helix structure at neutral pH by CD (Fig. 1), could be due to a combination of the following mechanisms: 1) the large size of the protein (33.6 kDa for the monomer) and of its oligomeric complexes, 2) intermediate exchange on the NMR timescale between the monomer, dimer, and tetramer oligomerization states (24), and 3) exchange broadening due to microsecond-millisecond motion in the SP monomer, if the 41–263 segment adopts a ‘‘molten-globule’’-like conformation with a-helical secondary structure but no fixed tertiary structure (15,23). To distinguish between these possibilities, we compared 1 H-15N HSQC spectra over a range of protein concentrations between 500 mM and 15 mM (Fig. S2 A). The dissociation constants of SP are 91 and 43 mM for the monomer/ dimer and dimer/tetramer equilibria, respectively (24). At 500 mM protein concentration, >95% of SP is present as dimers and tetramers. At 50 and 15 mM concentrations, below the monomer/dimer Kd, the fraction of monomeric SP is 70% and 90%, respectively (24). Yet few, if any, new peaks are seen in the 1H-15N HSQC spectrum because the protein is diluted below its Kd, at which it exists primarily as a monomer. Thus, the missing NMR signals for the 41–263 segment of SP cannot be due to the increase in molecular weight accompanying oligomerization. The 41–263 segment of SP corresponds to a 222-residue domain, and as such, an SP monomer should not exceed the size limit of
NMR. The most likely explanation for the absence of signals from the central 41–263 segment of SP is that it is undergoing microsecond-millisecond conformational exchange. This type of intermediate exchange broadening on the NMR timescale is often associated with proteins that have a molten globule conformation rather than a fixed tertiary structure (26,27). A second source of intermediate exchange NMR signal broadening for the visible peaks of SP is the rapid monomer-dimer-tetramer equilibrium. Thus, crosspeaks in 1H-15N HSQC broaden with increasing protein concentrations above the Kd of SP (Fig. S2 A). Consistent with this, a 500 mM SP sample that predominantly contains the dimeric and tetrameric forms shows increased NMR broadening at a field strength of 800 MHz compared to 500 MHz (Fig. S2 B). The strongest NMR broadening effects are observed for the amide protons, with the largest chemical shift differences from the ‘‘random coil’’ region of the spectrum, such as those from the HTH domain. Because NMR exchange broadening is dependent on both the exchange rate and magnetic field strength (28), the observations are most consistent with the rapid monomer-dimer-tetramer equilibrium moving from the fast to the intermediate chemical shift timescale as the magnetic field strength is increased from 500 to 800 MHz. In summary, free SP near physiological conditions lacks NMR signals from the central segment 41–263. This central segment has an a-helical molten globule conformation, subject to microsecond-millisecond conformational exchange. NMR signals in free SP are seen from the N-terminal residues 1–40 that are disordered and the C-terminal residues that were previously shown (3) to correspond to an independently folded domain with an HTH NMR structure. We next used NMR to investigate the properties of SP when it is encapsulated into phage P22 procapsids (Fig. 3). 15N-labeled SP was incorporated into unlabeled empty procapsid shells as previously described (29). A molar ratio of 60 SP to one empty procapsid shell was used to allow tight rebinding of SP to the interior of the procapsids (30). Higher ratios of up to 300:1, result in a mixture of strongly and weakly bound species (29). The encapsulated SP was separated from any unbound protein by ultracentrifugation. This yielded a capsid-SP complex with a molecular mass between 22 and 24 MDa, depending on the amount of incorporated SP (5). Sedimented procapsids with encapsulated 15N-SP were resuspended in 20 mM phosphate buffer containing 50 mM NaCl for NMR data collection. All 1H-15N HSQC crosspeaks observed from the procapsid-encapsulated SP (Fig. 3 B, shown in red) could be assigned to residues 4–40 from the acidic N-terminal domain of SP, illustrated in red in Fig. 3 A. In the spectrum of free SP under identical conditions (Fig. 3 B, blue contours), most of the assigned correlations are from the disordered N-terminal domain (residues 4–40) also seen in the encapsulated SP, as well as the basic folded HTH domain (residues 264–303) illustrated in blue in Fig. 3 A.
Biophysical Journal 117, 1387–1392, October 15, 2019 1389
Whitehead et al.
FIGURE 3 NMR spectra at 600 MHz of SP incorporated into P22 procapsids. (A) A domain diagram of the P22 SP protein is shown. (B) Superposition of the 1H-15N HSQC spectra of free SP (blue), with SP in the 23 MDa procapsid complex (red) is shown. NMR data were acquired on samples in 20 mM sodium phosphate, 50 mM NaCl (pH 7.0) at a temperature of 35 C. Horizontal lines in the upper right corner indicate unassigned crosspeaks from asparagine and glutamine side-chain NH2 groups. For NMR assignments, see Figs. S3 and S4 and Table S1. To see this figure in color, go online.
NMR signals from the C-terminal HTH domain broaden beyond detection in the encapsulated SP because of its tight binding to the capsid. The central portion of SP (residues 41–263) corresponding to the a-helical oligomerization domain (Fig. 3 A, gray) shows no crosspeaks in the procapsid-encapsulated SP and only 10–20 crosspeaks in the free protein. Similar results to those presented in Fig. 3 were obtained with a variant of SP that contained an N-terminal His6 tag to allow higher protein purification yield (Figs. S5 and S6). Control experiments in which 15N-labeled SP was pelleted with the procapsids and a supernatant spectrum was taken, as well as demonstration of comigration of SP with shells using a sucrose gradient, verified that the observed NMR signals were from encapsulated SP rather than the free protein (Fig. 4; Fig. S7). Discussion NMR signals from the first 40 amino acids of phage P22 SP are retained when it is encapsulated into phage P22 procapsids. Signals from the basic C-terminal domain of SP disappear because it is tightly bound to CP in a procapsid. Persistence of NMR signals from the N-terminal domain indicates this segment is unfolded in the SP-procapsid complex. The intrinsic disorder of the N-terminal segment is likely to be important in allowing SP to exit through channels in the procapsid during packaging of the viral genome (14). During maturation of the P22 procapsid to a virion, there are large structural rearrangements of the N-arm and A-domain of CP, which switch the capsid inner surface
1390 Biophysical Journal 117, 1387–1392, October 15, 2019
FIGURE 4 Control experiments used to demonstrate encapsulation of wild-type SP into capsids. (A) SP encapsulated into P22 capsids is shown. (B) The spectrum of the supernatant obtained after sedimenting the SP-capsid complex at 104,813 g for 20 min is shown. (C) The spectrum of the sedimented SP-capsid complex, after resuspension in 20 mM sodium phosphate (pH 7.0) by shaking at 180 oscillations per min for 20 h at 4 C is shown. The slightly weaker sensitivity of the spectrum in (C) is likely due to the loss of some sample during the sedimentation and resuspension procedure.
(facing the nucleic acid) from a negative to predominantly positive character (31–33). This switch in the electrostatic properties of the capsid inner surface likely triggers exit of SP because Coulombic interactions between SP and CP present in the procapsid are broken upon maturation. Previous deletion mutagenesis studies indicate that the N-terminal domain of SP is necessary for its exit from procapsids because its absence yields virions with an incomplete dsDNA genome (14). We hypothesize that the highly negative N-terminal domain of SP could compete for interactions between the negatively charged surface of CP and the positively charged HTH domain, thereby facilitating exit of SP. Our results show that the well-known size limitation of NMR can be used to an advantage as a filter to identify disordered segments even in very large supramolecular complexes of proteins. In this way, NMR can provide a unique perspective on dynamic and disordered elements of macromolecules not accessible by other techniques. The procapsid encapsulation experiments described herein are conceptually analogous to ‘‘in-cell NMR experiments’’ (34–36) in which signals from small proteins, or flexible segments of proteins, can be observed when they are incorporated inside living cells, as long as the isotope-labeled proteins of interest do not interact strongly with other large cellular components (34–36). We believe the ‘‘in-virus’’ NMR strategy described here (Fig. S8) could be more generally applicable to the study of the dynamic properties of macromolecules encapsidated into virus particles, including cargo molecules encased in viral capsids for nanotechnology applications. Additionally, such studies could assess the level of interaction of cargo molecules with the virus and probe the release properties of cargo.
Biophysical Letter
SUPPORTING MATERIAL Supporting Material can be found online at https://doi.org/10.1016/j.bpj. 2019.08.038.
16. van der Lee, R., M. Buljan, ., M. M. Babu. 2014. Classification of intrinsically disordered regions and proteins. Chem. Rev. 114:6589– 6631.
AUTHOR CONTRIBUTIONS
17. Medina, E. M., B. T. Andrews, ., C. E. Catalano. 2011. The bacteriophage lambda gpNu3 scaffolding protein is an intrinsically disordered and biologically functional procapsid assembly catalyst. J. Mol. Biol. 412:723–736.
C.M.T. and A.T.A. conceived the study. R.D.W. purified proteins, performed CD and NMR experiments, and analyzed data with A.T.A. A.T.A., R.D.W., and C.M.T. wrote the manuscript.
18. Lee, C. S., and P. Guo. 1995. Sequential interactions of structural proteins in phage phi 29 procapsid assembly. J. Virol. 69:5024– 5032.
ACKNOWLEDGMENTS This work was supported by National Institutes of Health grant R01 GM076661 to C.M.T. and A.T.A.
SUPPORTING CITATIONS References (37–40) appear in the Supporting Material.
REFERENCES 1. Dokland, T. 1999. Scaffolding proteins and their role in viral assembly. Cell. Mol. Life Sci. 56:580–603. 2. Fane, B. A., and P. E. Prevelige, Jr. 2003. Mechanism of scaffolding-assisted viral assembly. Adv. Protein Chem. 64:259–299. 3. Sun, Y., M. H. Parker, ., N. R. Krishna. 2000. Structure of the coat protein-binding domain of the scaffolding protein from a doublestranded DNA virus. J. Mol. Biol. 297:1195–1202. 4. Xu, J., D. Wang, ., Y. Xiang. 2019. Structural assembly of the tailed bacteriophage f29. Nat. Commun. 10:2366. 5. Keifer, D. Z., E. E. Pierson, ., M. F. Jarrold. 2014. Charge detection mass spectrometry of bacteriophage P22 procapsid distributions above 20 MDa. Rapid Commun. Mass Spectrom. 28:483–488. 6. Israel, V. 1977. E proteins of bacteriophage P22. I. Identification and ejection from wild-type and defective particles. J. Virol. 23:91–97. 7. Motwani, T., R. K. Lokareddy, ., C. M. Teschke. 2017. A viral scaffolding protein triggers portal ring oligomerization and incorporation during procapsid assembly. Sci. Adv. 3:e1700423. 8. Motwani, T., and C. M. Teschke. 2019. Architect of virus assembly: the portal protein nucleates procapsid assembly in bacteriophage P22. J. Virol. 93. e00187-19. 9. King, J., and S. Casjens. 1974. Catalytic head assembling protein in virus morphogenesis. Nature. 251:112–119. 10. Earnshaw, W. C., and S. R. Casjens. 1980. DNA packaging by the double-stranded DNA bacteriophages. Cell. 21:319–331. 11. Cortines, J. R., P. R. Weigele, ., C. M. Teschke. 2011. Decoding bacteriophage P22 assembly: identification of two charged residues in scaffolding protein responsible for coat protein interaction. Virology. 421:1–11. 12. Greene, B., and J. King. 1999. In vitro unfolding/refolding of wild type phage P22 scaffolding protein reveals capsid-binding domain. J. Biol. Chem. 274:16135–16140. 13. Padilla-Meier, G. P., E. B. Gilcrease, ., S. R. Casjens. 2012. Unraveling the role of the C-terminal helix turn helix of the coat-binding domain of bacteriophage P22 scaffolding protein. J. Biol. Chem. 287:33766–33780. 14. Weigele, P. R., L. Sampson, ., S. R. Casjens. 2005. Molecular genetics of bacteriophage P22 scaffolding protein’s functional domains. J. Mol. Biol. 348:831–844. 15. Tuma, R., P. E. Prevelige, Jr., and G. J. Thomas, Jr. 1996. Structural transitions in the scaffolding and coat proteins of P22 virus during assembly and disassembly. Biochemistry. 35:4619–4627.
19. Mesyanzhinov, V. V., B. N. Sobolev, ., V. P. Efimov. 1990. A proposed structure of bacteriophage T4 gene product 22–a major prohead scaffolding core protein. J. Struct. Biol. 104:24–31. 20. Morais, M. C., S. Kanamaru, ., M. G. Rossmann. 2003. Bacteriophage phi29 scaffolding protein gp7 before and after prohead assembly. Nat. Struct. Biol. 10:572–576. 21. Poh, S. L., F. el Khadali, ., P. Tavares. 2008. Oligomerization of the SPP1 scaffolding protein. J. Mol. Biol. 378:551–564. 22. Padilla-Meier, G. P., and C. M. Teschke. 2011. Conformational changes in bacteriophage P22 scaffolding protein induced by interaction with coat protein. J. Mol. Biol. 410:226–240. 23. Tuma, R., M. H. Parker, ., P. E. Prevelige, Jr. 1998. A helical coat protein recognition domain of the bacteriophage P22 scaffolding protein. J. Mol. Biol. 281:81–94. 24. Parker, M. H., W. F. Stafford, III, and P. E. Prevelige, Jr. 1997. Bacteriophage P22 scaffolding protein forms oligomers in solution. J. Mol. Biol. 268:655–665. 25. Tuma, R., H. Tsuruta, ., P. E. Prevelige. 2008. Detection of intermediates and kinetic control during assembly of bacteriophage P22 procapsid. J. Mol. Biol. 381:1395–1406. 26. Alexandrescu, A. T., P. A. Evans, ., C. M. Dobson. 1993. Structure and dynamics of the acid-denatured molten globule state of alpha-lactalbumin: a two-dimensional NMR study. Biochemistry. 32:1707–1718. 27. Redfield, C. 2004. Using nuclear magnetic resonance spectroscopy to study molten globule states of proteins. Methods. 34:121–132. 28. Palmer, A. G., III 2014. Chemical exchange in biomacromolecules: past, present, and future. J. Magn. Reson. 241:3–17. 29. Parker, M. H., C. G. Brouillette, and P. E. Prevelige, Jr. 2001. Kinetic and calorimetric evidence for two distinct scaffolding protein binding populations within the bacteriophage P22 procapsid. Biochemistry. 40:8962–8970. 30. Teschke, C. M., J. King, and P. E. Prevelige, Jr. 1993. Inhibition of viral capsid assembly by 1,10 -bi(4-anilinonaphthalene-5-sulfonic acid). Biochemistry. 32:10658–10665. 31. Teschke, C. M., and K. N. Parent. 2010. ‘Let the phage do the work’: using the phage P22 coat protein structures as a framework to understand its folding and assembly mutants. Virology. 401:119–130. 32. Parent, K. N., R. Khayat, ., T. S. Baker. 2010. P22 coat protein structures reveal a novel mechanism for capsid maturation: stability without auxiliary proteins or chemical crosslinks. Structure. 18:390–401. 33. Cortines, J. R., T. Motwani, ., C. M. Teschke. 2014. Highly specific salt bridges govern bacteriophage P22 icosahedral capsid assembly: identification of the site in coat protein responsible for interaction with scaffolding protein. J. Virol. 88:5287–5297. 34. Croke, R. L., C. O. Sallum, ., A. T. Alexandrescu. 2008. Hydrogen exchange of monomeric alpha-synuclein shows unfolded structure persists at physiological temperature and is independent of molecular crowding in Escherichia coli. Protein Sci. 17:1434–1445. 35. Serber, Z., P. Selenko, ., V. Do¨tsch. 2006. Investigating macromolecules inside cultured and injected cells by in-cell NMR spectroscopy. Nat. Protoc. 1:2701–2709.
Biophysical Journal 117, 1387–1392, October 15, 2019 1391
Whitehead et al. 36. Theillet, F. X., A. Binolfi, ., P. Selenko. 2014. Physicochemical properties of cells and their effects on intrinsically disordered proteins (IDPs). Chem. Rev. 114:6661–6714. 37. Greene, B., and J. King. 1994. Binding of scaffolding subunits within the P22 procapsid lattice. Virology. 205:188–197. 38. Maciejewski, M. W., A. D. Schuyler, ., J. C. Hoch. 2017. NMRbox: a resource for biomolecular NMR computation. Biophys. J. 112:1529– 1534.
1392 Biophysical Journal 117, 1387–1392, October 15, 2019
39. Prevelige, P. E., Jr., D. Thomas, and J. King. 1988. Scaffolding protein regulates the polymerization of P22 coat subunits into icosahedral shells in vitro. J. Mol. Biol. 202:743– 757. 40. Suhanovsky, M. M., and C. M. Teschke. 2011. Bacteriophage P22 capsid size determination: roles for the coat protein telokin-like domain and the scaffolding protein amino-terminus. Virology. 417:418–429.
NMR Mapping of Disordered Segments from a Viral Scaffolding Protein Enclosed in a 23 MDa Procapsid Richard D. Whitehead III,1 Carolyn M. Teschke,1,2,* and Andrei T. Alexandrescu1,* 1
Department of Molecular and Cell Biology and 2Department of Chemistry, University of Connecticut, Storrs, Connecticut
ABSTRACT Scaffolding proteins (SPs) are required for the capsid shell assembly of many tailed double-stranded DNA bacteriophages, some archaeal viruses, herpesviruses, and adenoviruses. Despite their importance, only one high-resolution structure is available for SPs within procapsids. Here, we use the inherent size limit of NMR to identify mobile segments of the 303-residue phage P22 SP free in solution and when incorporated into a 23 MDa procapsid complex. Free SP gives NMR signals from its acidic N-terminus (residues 1–40) and basic C-terminus (residues 264–303), whereas NMR signals from the middle segment (residues 41–263) are missing because of intermediate conformational exchange on the NMR chemical shift timescale. When SP is incorporated into P22 procapsids, NMR signals from the C-terminal helix-turn-helix domain disappear because of binding to the procapsid interior. Signals from the N-terminal domain persist, indicating that this segment retains flexibility when bound to procapsids. The unstructured character of the N-terminus, coupled with its high content of negative charges, is likely important for dissociation and release of SP during the double-stranded DNA genome packaging step accompanying phage maturation.
SIGNIFICANCE Scaffolding protein (SP) nucleates the assembly of phage P22 coat proteins into an icosahedral capsid structure that encapsidates the viral genome. NMR spectra of free SP show signals from the N-terminus and a helix-turnhelix domain at the C-terminus. When SP is incorporated into phage P22 empty procapsid shells to form a 23 MDa complex, NMR signals from the N-terminal 40 residues persist indicating this segment is disordered. The unfolded nature of the N-terminus coupled with its negatively charged character is important for the functional requirement of SP to exit the capsid as it becomes packaged with its genome.
INTRODUCTION Viruses infect organisms in all kingdoms of life; those that infect bacteria are called bacteriophages or phages. In many viruses and phages, scaffolding proteins (SPs) are required to ensure the correct organization of coat proteins (CPs) and other minor capsid proteins into a precursor structure, called a procapsid (1,2). Although SPs are critical for viral assembly and therefore potential therapeutic targets, their structural properties, with few exceptions (3,4), are poorly understood. Here, we investigate the structure of full-length SP from bacteriophage P22. Phage P22 is a well-characterized model for the class of tailed bacterial viruses with double-stranded DNA (dsDNA) genomes. Between 60 and 300 copies of a 303-residue SP (33.6 kDa) catalyze the assembly of 415 copies of CP (46.8 kDa) to Submitted May 29, 2019, and accepted for publication August 30, 2019. *Correspondence: [email protected] or andrei.alexandrescu@ uconn.edu Editor: David Eliezer. https://doi.org/10.1016/j.bpj.2019.08.038
form the icosahedral T ¼ 7 procapsid of phage P22 (23 MDa) (5). The SPs are also required for incorporation of the dodecameric portal protein complex at one vertex of the icosahedron, as well as other minor internal proteins (6–8). SPs are not found in mature virions because they are released without proteolysis through holes in the procapsid during genome packaging. The SP is then recycled for subsequent use in multiple rounds of procapsid assembly (9). Without SP, phage P22 CP fails to assemble into normal T ¼ 7 procapsids, yielding instead aberrant ‘‘petite’’ T ¼ 4 and T ¼ 7 particles or spiral structures, all of which are noninfectious (10). Lack of proper assembly in the absence of SP is a general feature of these viruses. Although the functional domains of SP are well characterized (11–14), its structure has proven elusive because of the unusual dynamic properties of the protein (15). The amino acid sequence of SP has a high content of charged residues (31%) and a low content of hydrophobic residues, which classify it as an intrinsically disordered protein (16). The distribution of charged amino acids in SP is asymmetric, with a highly acidic N-terminus (pI ¼ 4.4 for residues
Ó 2019 Biophysical Society.
Biophysical Journal 117, 1387–1392, October 15, 2019 1387
Whitehead et al.
1–130) and basic C-terminus (pI ¼ 9.2 for residues 131–303). This charge distribution suggests that SP belongs to the group of ‘‘electrostatic hairpin’’ intrinsically disordered protein structures (16). Biochemical data are consistent with the disordered nature of SP predicted from its sequence. The protein exhibits a noncooperative unfolding transition characteristic of molten globule conformations, which have a secondary structure but a liquid-like fluctuating tertiary structure (12,15). SP also exhibits rapid amide proton hydrogen exchange typical of marginally stable secondary structure (15). These properties are not unique for the SP of P22; conformational flexibility is a hallmark of viral SPs that appears to be required for their functionality (1,2,17). Nevertheless, the SP of P22, as well as those from phages l, f29, SPP1, and T4, have a high content of predicted and/or measured a-helical secondary structure (17–21). Moreover, in the SP of phage P22, the C-terminal domain, which is critical for CP binding during procapsid assembly, adopts an independently folded helix-turn-helix (HTH) domain structure, as determined by NMR of a 238–303 fragment of SP (3). In this fragment, residues 238–267 are disordered, and the last 35 residues 268–303 fold into the HTH domain (3). Thus, although SP lacks a fixed global hydrophobic core, it is organized into distinct functional domains that are used to fulfill the protein’s multiple roles in the procapsid assembly process (14). In previous biochemical cross-linking studies, we saw evidence that P22 SP undergoes conformational changes during encapsulation into procapsids (22). Here, we use the intrinsic size limit of NMR to define the regions of SP that remain unfolded after assembly.
The CD spectrum shifts to one typical of an unfolded protein as the protein becomes acid denatured at low pH. Consistently, the 1H-15N heteronuclear single quantum coherence (HSQC) NMR spectra of SP at low pH under acid-denaturing conditions are typical of an unfolded protein. About 300 1H-15N HSQC crosspeaks are observed for the 303-residue SP protein at pH 2.0, with all of the 1HN chemical shifts falling in the random coil region between 8.7 and 7.7 ppm (Fig. 2). At pH 6.0, at which CD indicates that SP has substantial a-helix structure (Fig. 1), the chemical shifts remain close to the random coil region, but 2/3 of the 1H-15N HSQC correlations expected from the 303-residue sequence are lost. Sequence-specific NMR assignments for the phage
Results Circular dichroism (CD) spectra show that free SP has a markedly a-helical character at physiological pH (Fig. 1).
FIGURE 1 Far-ultraviolet CD of the SP from P22 (3 mM) as a function of pH. The data were obtained at 30 C. To see this figure in color, go online.
1388 Biophysical Journal 117, 1387–1392, October 15, 2019
FIGURE 2 Representative 600 MHz 1H-15N HSQC spectra from a pH titration of 50 mM P22 SP at a temperature of 35 C. At lower pH, more correlations are seen because of unfolding of residues 41–263 in the central part of SP (see Fig. 1).
Biophysical Letter
P22 SP (Fig. S1; Table S1) indicate the surviving 100 1H-15N HSQC peaks at pH 6.0 are mostly from the N-terminal residues 1–40 that have chemical shifts consistent with unfolded random coil conformations and the C-terminal residues 264–303 that correspond to the folded HTH domain of SP responsible for binding CP (3). Many of the 1HN NMR signals from the C-terminal HTH segment are shifted upfield, with backbone carbon chemical shifts (Table S1) that are also consistent with the a-helical NMR structure of the HTH domain (3). The NMR structure of the HTH domain was determined at pH 4.4 (3). The ellipticity at 222 nm, characteristic of a-helical structure, nearly doubles between pH 4.0 and pH 7.4 (Fig. 1). The transition to a-helical structure with increasing pH therefore cannot be due to the HTH domain alone. Rather, a substantial amount of a-helical structure must be formed in the central 41–263 segment, as SP becomes increasingly folded toward neutral pH. The majority of signals from the central portion of SP, between residues 41–263, were not observed in the 1H-15N HSQC spectra near neutral pH. Biophysical studies indicate that this central portion of SP consists of loosely organized domains of a-helical structure, connected by turns and random coil segments (12,15,23). Furthermore, SP exists in a rapidly reversible monomer-dimer-tetramer dynamic equilibrium, at which the dimers are thought to be the catalytically active species in procapsid assembly (24,25). Broadening of NMR resonances from the 41–263 segment of SP, accompanying the folding transition to a-helix structure at neutral pH by CD (Fig. 1), could be due to a combination of the following mechanisms: 1) the large size of the protein (33.6 kDa for the monomer) and of its oligomeric complexes, 2) intermediate exchange on the NMR timescale between the monomer, dimer, and tetramer oligomerization states (24), and 3) exchange broadening due to microsecond-millisecond motion in the SP monomer, if the 41–263 segment adopts a ‘‘molten-globule’’-like conformation with a-helical secondary structure but no fixed tertiary structure (15,23). To distinguish between these possibilities, we compared 1 H-15N HSQC spectra over a range of protein concentrations between 500 mM and 15 mM (Fig. S2 A). The dissociation constants of SP are 91 and 43 mM for the monomer/ dimer and dimer/tetramer equilibria, respectively (24). At 500 mM protein concentration, >95% of SP is present as dimers and tetramers. At 50 and 15 mM concentrations, below the monomer/dimer Kd, the fraction of monomeric SP is 70% and 90%, respectively (24). Yet few, if any, new peaks are seen in the 1H-15N HSQC spectrum because the protein is diluted below its Kd, at which it exists primarily as a monomer. Thus, the missing NMR signals for the 41–263 segment of SP cannot be due to the increase in molecular weight accompanying oligomerization. The 41–263 segment of SP corresponds to a 222-residue domain, and as such, an SP monomer should not exceed the size limit of
NMR. The most likely explanation for the absence of signals from the central 41–263 segment of SP is that it is undergoing microsecond-millisecond conformational exchange. This type of intermediate exchange broadening on the NMR timescale is often associated with proteins that have a molten globule conformation rather than a fixed tertiary structure (26,27). A second source of intermediate exchange NMR signal broadening for the visible peaks of SP is the rapid monomer-dimer-tetramer equilibrium. Thus, crosspeaks in 1H-15N HSQC broaden with increasing protein concentrations above the Kd of SP (Fig. S2 A). Consistent with this, a 500 mM SP sample that predominantly contains the dimeric and tetrameric forms shows increased NMR broadening at a field strength of 800 MHz compared to 500 MHz (Fig. S2 B). The strongest NMR broadening effects are observed for the amide protons, with the largest chemical shift differences from the ‘‘random coil’’ region of the spectrum, such as those from the HTH domain. Because NMR exchange broadening is dependent on both the exchange rate and magnetic field strength (28), the observations are most consistent with the rapid monomer-dimer-tetramer equilibrium moving from the fast to the intermediate chemical shift timescale as the magnetic field strength is increased from 500 to 800 MHz. In summary, free SP near physiological conditions lacks NMR signals from the central segment 41–263. This central segment has an a-helical molten globule conformation, subject to microsecond-millisecond conformational exchange. NMR signals in free SP are seen from the N-terminal residues 1–40 that are disordered and the C-terminal residues that were previously shown (3) to correspond to an independently folded domain with an HTH NMR structure. We next used NMR to investigate the properties of SP when it is encapsulated into phage P22 procapsids (Fig. 3). 15N-labeled SP was incorporated into unlabeled empty procapsid shells as previously described (29). A molar ratio of 60 SP to one empty procapsid shell was used to allow tight rebinding of SP to the interior of the procapsids (30). Higher ratios of up to 300:1, result in a mixture of strongly and weakly bound species (29). The encapsulated SP was separated from any unbound protein by ultracentrifugation. This yielded a capsid-SP complex with a molecular mass between 22 and 24 MDa, depending on the amount of incorporated SP (5). Sedimented procapsids with encapsulated 15N-SP were resuspended in 20 mM phosphate buffer containing 50 mM NaCl for NMR data collection. All 1H-15N HSQC crosspeaks observed from the procapsid-encapsulated SP (Fig. 3 B, shown in red) could be assigned to residues 4–40 from the acidic N-terminal domain of SP, illustrated in red in Fig. 3 A. In the spectrum of free SP under identical conditions (Fig. 3 B, blue contours), most of the assigned correlations are from the disordered N-terminal domain (residues 4–40) also seen in the encapsulated SP, as well as the basic folded HTH domain (residues 264–303) illustrated in blue in Fig. 3 A.
Biophysical Journal 117, 1387–1392, October 15, 2019 1389
Whitehead et al.
FIGURE 3 NMR spectra at 600 MHz of SP incorporated into P22 procapsids. (A) A domain diagram of the P22 SP protein is shown. (B) Superposition of the 1H-15N HSQC spectra of free SP (blue), with SP in the 23 MDa procapsid complex (red) is shown. NMR data were acquired on samples in 20 mM sodium phosphate, 50 mM NaCl (pH 7.0) at a temperature of 35 C. Horizontal lines in the upper right corner indicate unassigned crosspeaks from asparagine and glutamine side-chain NH2 groups. For NMR assignments, see Figs. S3 and S4 and Table S1. To see this figure in color, go online.
NMR signals from the C-terminal HTH domain broaden beyond detection in the encapsulated SP because of its tight binding to the capsid. The central portion of SP (residues 41–263) corresponding to the a-helical oligomerization domain (Fig. 3 A, gray) shows no crosspeaks in the procapsid-encapsulated SP and only 10–20 crosspeaks in the free protein. Similar results to those presented in Fig. 3 were obtained with a variant of SP that contained an N-terminal His6 tag to allow higher protein purification yield (Figs. S5 and S6). Control experiments in which 15N-labeled SP was pelleted with the procapsids and a supernatant spectrum was taken, as well as demonstration of comigration of SP with shells using a sucrose gradient, verified that the observed NMR signals were from encapsulated SP rather than the free protein (Fig. 4; Fig. S7). Discussion NMR signals from the first 40 amino acids of phage P22 SP are retained when it is encapsulated into phage P22 procapsids. Signals from the basic C-terminal domain of SP disappear because it is tightly bound to CP in a procapsid. Persistence of NMR signals from the N-terminal domain indicates this segment is unfolded in the SP-procapsid complex. The intrinsic disorder of the N-terminal segment is likely to be important in allowing SP to exit through channels in the procapsid during packaging of the viral genome (14). During maturation of the P22 procapsid to a virion, there are large structural rearrangements of the N-arm and A-domain of CP, which switch the capsid inner surface
1390 Biophysical Journal 117, 1387–1392, October 15, 2019
FIGURE 4 Control experiments used to demonstrate encapsulation of wild-type SP into capsids. (A) SP encapsulated into P22 capsids is shown. (B) The spectrum of the supernatant obtained after sedimenting the SP-capsid complex at 104,813 g for 20 min is shown. (C) The spectrum of the sedimented SP-capsid complex, after resuspension in 20 mM sodium phosphate (pH 7.0) by shaking at 180 oscillations per min for 20 h at 4 C is shown. The slightly weaker sensitivity of the spectrum in (C) is likely due to the loss of some sample during the sedimentation and resuspension procedure.
(facing the nucleic acid) from a negative to predominantly positive character (31–33). This switch in the electrostatic properties of the capsid inner surface likely triggers exit of SP because Coulombic interactions between SP and CP present in the procapsid are broken upon maturation. Previous deletion mutagenesis studies indicate that the N-terminal domain of SP is necessary for its exit from procapsids because its absence yields virions with an incomplete dsDNA genome (14). We hypothesize that the highly negative N-terminal domain of SP could compete for interactions between the negatively charged surface of CP and the positively charged HTH domain, thereby facilitating exit of SP. Our results show that the well-known size limitation of NMR can be used to an advantage as a filter to identify disordered segments even in very large supramolecular complexes of proteins. In this way, NMR can provide a unique perspective on dynamic and disordered elements of macromolecules not accessible by other techniques. The procapsid encapsulation experiments described herein are conceptually analogous to ‘‘in-cell NMR experiments’’ (34–36) in which signals from small proteins, or flexible segments of proteins, can be observed when they are incorporated inside living cells, as long as the isotope-labeled proteins of interest do not interact strongly with other large cellular components (34–36). We believe the ‘‘in-virus’’ NMR strategy described here (Fig. S8) could be more generally applicable to the study of the dynamic properties of macromolecules encapsidated into virus particles, including cargo molecules encased in viral capsids for nanotechnology applications. Additionally, such studies could assess the level of interaction of cargo molecules with the virus and probe the release properties of cargo.
Biophysical Letter
SUPPORTING MATERIAL Supporting Material can be found online at https://doi.org/10.1016/j.bpj. 2019.08.038.
16. van der Lee, R., M. Buljan, ., M. M. Babu. 2014. Classification of intrinsically disordered regions and proteins. Chem. Rev. 114:6589– 6631.
AUTHOR CONTRIBUTIONS
17. Medina, E. M., B. T. Andrews, ., C. E. Catalano. 2011. The bacteriophage lambda gpNu3 scaffolding protein is an intrinsically disordered and biologically functional procapsid assembly catalyst. J. Mol. Biol. 412:723–736.
C.M.T. and A.T.A. conceived the study. R.D.W. purified proteins, performed CD and NMR experiments, and analyzed data with A.T.A. A.T.A., R.D.W., and C.M.T. wrote the manuscript.
18. Lee, C. S., and P. Guo. 1995. Sequential interactions of structural proteins in phage phi 29 procapsid assembly. J. Virol. 69:5024– 5032.
ACKNOWLEDGMENTS This work was supported by National Institutes of Health grant R01 GM076661 to C.M.T. and A.T.A.
SUPPORTING CITATIONS References (37–40) appear in the Supporting Material.
REFERENCES 1. Dokland, T. 1999. Scaffolding proteins and their role in viral assembly. Cell. Mol. Life Sci. 56:580–603. 2. Fane, B. A., and P. E. Prevelige, Jr. 2003. Mechanism of scaffolding-assisted viral assembly. Adv. Protein Chem. 64:259–299. 3. Sun, Y., M. H. Parker, ., N. R. Krishna. 2000. Structure of the coat protein-binding domain of the scaffolding protein from a doublestranded DNA virus. J. Mol. Biol. 297:1195–1202. 4. Xu, J., D. Wang, ., Y. Xiang. 2019. Structural assembly of the tailed bacteriophage f29. Nat. Commun. 10:2366. 5. Keifer, D. Z., E. E. Pierson, ., M. F. Jarrold. 2014. Charge detection mass spectrometry of bacteriophage P22 procapsid distributions above 20 MDa. Rapid Commun. Mass Spectrom. 28:483–488. 6. Israel, V. 1977. E proteins of bacteriophage P22. I. Identification and ejection from wild-type and defective particles. J. Virol. 23:91–97. 7. Motwani, T., R. K. Lokareddy, ., C. M. Teschke. 2017. A viral scaffolding protein triggers portal ring oligomerization and incorporation during procapsid assembly. Sci. Adv. 3:e1700423. 8. Motwani, T., and C. M. Teschke. 2019. Architect of virus assembly: the portal protein nucleates procapsid assembly in bacteriophage P22. J. Virol. 93. e00187-19. 9. King, J., and S. Casjens. 1974. Catalytic head assembling protein in virus morphogenesis. Nature. 251:112–119. 10. Earnshaw, W. C., and S. R. Casjens. 1980. DNA packaging by the double-stranded DNA bacteriophages. Cell. 21:319–331. 11. Cortines, J. R., P. R. Weigele, ., C. M. Teschke. 2011. Decoding bacteriophage P22 assembly: identification of two charged residues in scaffolding protein responsible for coat protein interaction. Virology. 421:1–11. 12. Greene, B., and J. King. 1999. In vitro unfolding/refolding of wild type phage P22 scaffolding protein reveals capsid-binding domain. J. Biol. Chem. 274:16135–16140. 13. Padilla-Meier, G. P., E. B. Gilcrease, ., S. R. Casjens. 2012. Unraveling the role of the C-terminal helix turn helix of the coat-binding domain of bacteriophage P22 scaffolding protein. J. Biol. Chem. 287:33766–33780. 14. Weigele, P. R., L. Sampson, ., S. R. Casjens. 2005. Molecular genetics of bacteriophage P22 scaffolding protein’s functional domains. J. Mol. Biol. 348:831–844. 15. Tuma, R., P. E. Prevelige, Jr., and G. J. Thomas, Jr. 1996. Structural transitions in the scaffolding and coat proteins of P22 virus during assembly and disassembly. Biochemistry. 35:4619–4627.
19. Mesyanzhinov, V. V., B. N. Sobolev, ., V. P. Efimov. 1990. A proposed structure of bacteriophage T4 gene product 22–a major prohead scaffolding core protein. J. Struct. Biol. 104:24–31. 20. Morais, M. C., S. Kanamaru, ., M. G. Rossmann. 2003. Bacteriophage phi29 scaffolding protein gp7 before and after prohead assembly. Nat. Struct. Biol. 10:572–576. 21. Poh, S. L., F. el Khadali, ., P. Tavares. 2008. Oligomerization of the SPP1 scaffolding protein. J. Mol. Biol. 378:551–564. 22. Padilla-Meier, G. P., and C. M. Teschke. 2011. Conformational changes in bacteriophage P22 scaffolding protein induced by interaction with coat protein. J. Mol. Biol. 410:226–240. 23. Tuma, R., M. H. Parker, ., P. E. Prevelige, Jr. 1998. A helical coat protein recognition domain of the bacteriophage P22 scaffolding protein. J. Mol. Biol. 281:81–94. 24. Parker, M. H., W. F. Stafford, III, and P. E. Prevelige, Jr. 1997. Bacteriophage P22 scaffolding protein forms oligomers in solution. J. Mol. Biol. 268:655–665. 25. Tuma, R., H. Tsuruta, ., P. E. Prevelige. 2008. Detection of intermediates and kinetic control during assembly of bacteriophage P22 procapsid. J. Mol. Biol. 381:1395–1406. 26. Alexandrescu, A. T., P. A. Evans, ., C. M. Dobson. 1993. Structure and dynamics of the acid-denatured molten globule state of alpha-lactalbumin: a two-dimensional NMR study. Biochemistry. 32:1707–1718. 27. Redfield, C. 2004. Using nuclear magnetic resonance spectroscopy to study molten globule states of proteins. Methods. 34:121–132. 28. Palmer, A. G., III 2014. Chemical exchange in biomacromolecules: past, present, and future. J. Magn. Reson. 241:3–17. 29. Parker, M. H., C. G. Brouillette, and P. E. Prevelige, Jr. 2001. Kinetic and calorimetric evidence for two distinct scaffolding protein binding populations within the bacteriophage P22 procapsid. Biochemistry. 40:8962–8970. 30. Teschke, C. M., J. King, and P. E. Prevelige, Jr. 1993. Inhibition of viral capsid assembly by 1,10 -bi(4-anilinonaphthalene-5-sulfonic acid). Biochemistry. 32:10658–10665. 31. Teschke, C. M., and K. N. Parent. 2010. ‘Let the phage do the work’: using the phage P22 coat protein structures as a framework to understand its folding and assembly mutants. Virology. 401:119–130. 32. Parent, K. N., R. Khayat, ., T. S. Baker. 2010. P22 coat protein structures reveal a novel mechanism for capsid maturation: stability without auxiliary proteins or chemical crosslinks. Structure. 18:390–401. 33. Cortines, J. R., T. Motwani, ., C. M. Teschke. 2014. Highly specific salt bridges govern bacteriophage P22 icosahedral capsid assembly: identification of the site in coat protein responsible for interaction with scaffolding protein. J. Virol. 88:5287–5297. 34. Croke, R. L., C. O. Sallum, ., A. T. Alexandrescu. 2008. Hydrogen exchange of monomeric alpha-synuclein shows unfolded structure persists at physiological temperature and is independent of molecular crowding in Escherichia coli. Protein Sci. 17:1434–1445. 35. Serber, Z., P. Selenko, ., V. Do¨tsch. 2006. Investigating macromolecules inside cultured and injected cells by in-cell NMR spectroscopy. Nat. Protoc. 1:2701–2709.
Biophysical Journal 117, 1387–1392, October 15, 2019 1391
Whitehead et al. 36. Theillet, F. X., A. Binolfi, ., P. Selenko. 2014. Physicochemical properties of cells and their effects on intrinsically disordered proteins (IDPs). Chem. Rev. 114:6661–6714. 37. Greene, B., and J. King. 1994. Binding of scaffolding subunits within the P22 procapsid lattice. Virology. 205:188–197. 38. Maciejewski, M. W., A. D. Schuyler, ., J. C. Hoch. 2017. NMRbox: a resource for biomolecular NMR computation. Biophys. J. 112:1529– 1534.
1392 Biophysical Journal 117, 1387–1392, October 15, 2019
39. Prevelige, P. E., Jr., D. Thomas, and J. King. 1988. Scaffolding protein regulates the polymerization of P22 coat subunits into icosahedral shells in vitro. J. Mol. Biol. 202:743– 757. 40. Suhanovsky, M. M., and C. M. Teschke. 2011. Bacteriophage P22 capsid size determination: roles for the coat protein telokin-like domain and the scaffolding protein amino-terminus. Virology. 417:418–429.