fragment screen and was found to be able to
stabilize bZiPro in a thermal shift assay (fig. S9).
Changes in the 1H-15N HSQC spectra upon binding
not only confirmed the direct binding of the compound to bZiPro, but also showed that it did not
cause any large-scale conformational changes to
the protein (Fig. 3A). Indeed, EN300 is sandwiched
between the Y161 aromatic ring and A132 by
stacking interactions and forms a hydrogen bond
with Y150 (Fig. 3). It only partially occupies the
S1 pocket and forms no direct contact with NS2B
cofactor. Therefore, the closed conformation of
bZiPro is captured again in this structure, which
is unlikely to be attributable to the binding of
the small compound. The protein-ligand complex
structure serves as a starting point to guide
further chemical modifications for optimizations
in binding potency and inhibition of protease activity inhibition in a targeted manner (movie S1).
We show that the crystal structure of free
unlinked ZIKV NS2B-NS3 protease exists in the
closed conformation. Solution NMR studies con-
firmed that free protease is predominantly in a
closed conformation, with local conformational
dynamics occurring at the NS2B-NS3 binding
interface. Substrate peptide binding does not
induce further substantial conformational changes.
This finding has relevance for the discovery of
antiviral drugs targeting ZIKV protease. The
high-resolution crystal form that we present will
be of practical use because these crystals can be
readily soaked or cocrystalized with inhibitors.
Chemical library screens including fragment-
based screening will have a higher likelihood of
obtaining potent lead compounds and achieving
successful structure-based lead optimization. We
also report the cocrystal structure of a flaviviral
protease bound to a peptide in a C-to-N ori-
entation. The reverse direction of the peptide
bond is unable to form a tetrahedral intermediate
at the protease active site and is therefore not
cleavable. Although this might be a crystal ar-
tifact, the good fit of the peptide within the
protease pocket provides an attractive starting
point for developing novel peptidomimetic in-
hibitors, such as cyclic peptides.
REFERENCES AND NOTES
1. V. M. Cao-Lormeau et al., Lancet 387, 1531–1539
2. L. Schuler-Faccini et al., MMWR Morb. Mortal. Wkly. Rep. 65,
3. T. C. Pierson, M. S. Diamond, in Fields Virology, D. M. Knipe,
P. M. Howley, Eds. (Lippincott-Raven Publishers, ed. 6, 2013),
chap. 26, pp. 747–794.
4. B. D. Lindenbach, C. L. Murray, H.-J. Thiel, C. M. Rice, in Fields
Virology, D. M. Knipe, P. M. Howley, Eds. (Lippincott-Raven
Publishers, ed. 6, 2013), chap. 25, pp. 712–746.
5. S. C. Weaver et al., Antiviral Res. 130, 69–80 (2016).
6. S. Ekins et al., F1000Research 5, 150 (2016).
7. C. Nitsche, S. Holloway, T. Schirmeister, C. D. Klein, Chem. Rev.
114, 11348–11381 (2014).
8. D. Luo, S. G. Vasudevan, J. Lescar, Antiviral Res. 118, 148–158 (2015).
9. B. Falgout, M. Pethel, Y. M. Zhang, C. J. Lai, J. Virol. 65,
10. T. J. Chambers et al., Proc. Natl. Acad. Sci. U.S.A. 87,
11. D. Leung et al., J. Biol. Chem. 276, 45762–45771 (2001).
12. D. Luo et al., J. Virol. 82, 173–183 (2008).
13. P. Erbel et al., Nat. Struct. Mol. Biol. 13, 372–373 (2006).
14. C. G. Noble, C. C. Seh, A. T. Chao, P. Y. Shi, J. Virol. 86,
15. G. Robin et al., J. Mol. Biol. 385, 1568–1577 (2009).
16. S. Chandramouli et al., J. Virol. 84, 3059–3067 (2010).
17. M. Z. Hammamy, C. Haase, M. Hammami, R. Hilgenfeld,
T. Steinmetzer, ChemMedChem 8, 231–241 (2013).
18. A. E. Aleshin, S. A. Shiryaev, A. Y. Strongin, R. C. Liddington,
Protein Sci. 16, 795–806 (2007).
19. R. Assenberg et al., J. Virol. 83, 12895–12906 (2009).
20. J. Lei et al., Science 353, 503–505 (2016).
21. L. de la Cruz, W. N. Chen, B. Graham, G. Otting, FEBS J. 281,
22. X. C. Su et al., PLOS Negl. Trop. Dis. 3, e561 (2009).
23. Y. M. Kim et al., J. Biol. Chem. 288, 12891–12900
24. W. W. Phoo et al., Nat. Commun. 7, 13410 (2016).
25. C. Nitsche, M. A. Behnam, C. Steuer, C. D. Klein, Antiviral Res.
94, 72–79 (2012).
26. C. Kang et al., Antiviral Res. 97, 137–144 (2013).
27. B. A. Johnson, Methods Mol. Biol. 278, 313–352 (2004).
We thank W. Bin, E. Abbas, C. Liew (Nanyang Technological
University), and scientists from the Australian Light Source MX
beamline and the Swiss Light Source X06DA (PXIII) beamline for
their help with diffraction data collection. We thank S. Liu for
technical support for fragment screening. We thank A. Matter and
T. H. Keller for critical reading of this manuscript. The data
presented in this manuscript are tabulated in the main paper and
the supplementary materials. The corresponding coordinates and
structure factors are available from the PDB under accession
codes 5GPI for bZiPro and 5H4I for its complex with EN300.
Assignments of protease in the absence and presence of AcKR
have been deposited in the Biological Magnetic Resonance Data
Bank with accession numbers 26928 and 26927, respectively.
This work was supported by (i) a start-up grant from Lee Kong
Chian School of Medicine, Nanyang Technological University;
(ii) National Medical Research Council grant CBRG15May045 to
D.L.’s laboratory; and (iii) A*STAR Joint Council Office grant
1431AFG102/1331A028 to C.K. The authors declare no
competing financial interests.
Materials and Methods
Figs. S1 to S9
Tables S1 and S2
1 September 2016; accepted 22 November 2016
Published online 8 December 2016
1600 23 DECEMBER 2016 • VOL 354 ISSUE 6319 sciencemag.org SCIENCE
Fig. 3. Structure of bZiPro in complex with a
compound fragment. (A) Overlay of 1H-15N HSQC
spectra of bZiPro in the absence (black) and presence (red) of the compound EN300. (B) The simulated annealing omit map of the bZiPro crystal
structure is contoured at 4s in green mesh. (C) Close-up view of the interactions between NS3 and the
compound. (D) Electron density map at the compound binding site. The 2mFo-DFc electron density
map is contoured at 1s in blue, and the simulated
annealing omit map is contoured at 3s in green.
Clear electron density is observed for the benz-imidazole portion of EN300, forming a p-stacking
interaction with Y161. Weaker electron density observed around the hydroxyl group could be indicative
of partial hydrolysis of the compound.