X-ray Structures of the Proprotein Convertase Furin Bound with

Biochemistry , 2018, 57 (6), pp 925–934. DOI: 10.1021/acs.biochem.7b01124. Publication Date (Web): January 9, 2018. Copyright © 2018 American Chemi...
1 downloads 5 Views 3MB Size
Subscriber access provided by University of Sydney Library

Article

X-ray structures of the proprotein convertase furin bound with substrate analog inhibitors reveal substrate specificity determinants beyond the S4 pocket. Sven O Dahms, Kornelia Hardes, Torsten Steinmetzer, and Manuel E Than Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.7b01124 • Publication Date (Web): 09 Jan 2018 Downloaded from http://pubs.acs.org on January 10, 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 free 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 accessible to all readers and 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.

Biochemistry 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 36 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

Biochemistry

X-ray structures of the proprotein convertase furin bound with substrate analog inhibitors reveal substrate specificity determinants beyond the S4 pocket.

Sven O. Dahms1,2,*, Kornelia Hardes3, Torsten Steinmetzer3 and Manuel E. Than2,*

1

Department of Molecular Biology, University of Salzburg, Billrothstrasse 11, A-5020

Salzburg, Austria

2

Protein Crystallography Group, Leibniz Institute on Aging - Fritz Lipmann Institute (FLI),

Beutenbergstr. 11, 07745 Jena, Germany

3

Department of Pharmaceutical Chemistry, Philipps University Marburg, Marbacher Weg 6,

D-35032 Marburg, Germany

* To whom correspondence should be addressed: Sven O. Dahms: [email protected], Tel: +43-66280-447227 Manuel E. Than: [email protected], Tel: +49-3641-656170

ACS Paragon Plus Environment

Biochemistry 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

Page 2 of 36

Abstract

The proprotein convertase (PC) furin is a highly specific serine protease modifying and thereby activating proteins in the secretory pathway by proteolytic cleavage. Its substrates are involved in many diseases including cancer and infections caused by bacteria and viruses. Understanding furin’s substrate specificity is of crucial importance for the development of pharmacologically applicable inhibitors. Using protein X-ray crystallography we investigated the extended substrate binding site of furin in complex with three peptide derived inhibitors at up to 1.9 Å resolution. The structure of the protease bound with a hexapeptide inhibitor revealed molecular details of its S6 pocket, which remained completely unknown so far. The arginine residue at P6 induced an unexpected turn-like conformation of the inhibitor backbone, which is stabilized by intra- and intermolecular H-bonds. In addition, we confirmed the binding of arginine to the previously proposed S5 pocket (S51). An alternative S5 site (S52) could be utilized by shorter sidechains as demonstrated for a 4aminomethyl-phenylacetyl residue, which shows steric properties similar to a lysine side chain. Interestingly, we also observed binding of a peptide with citrulline at P4 substituting the highly conserved arginine. The structural data might indicate an unusual protonation state of Asp264 maintaining the interaction with uncharged citrulline. The herein identified molecular interaction sites at P5 and P6 can be utilized to improve next generation furin inhibitors. Our data will also help to predict furin substrates more precisely based on the additional specificity determinants observed for P5 and P6.

2 ACS Paragon Plus Environment

Page 3 of 36 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

Biochemistry

Introduction

Furin belongs to the proprotein convertases (PCs), a subtilisin-related family of serineproteases. The PCs catalyze the post-translational modification of numerous proteins in the secretory pathway by limited proteolysis 1. Therefore, the inhibition of furin and of other PCs could be a promising strategy for the treatment of many pathological conditions, including e.g. hypercholesterolemia, cancer and infectious diseases 1, 2.

The physiological functions of the PCs imply stringent substrate specificity. Most of the family members (furin, PC1, PC2, PC4, PACE4, PC5/6 and PC7) recognize various multibasic substrate sequences and cleave after the general pattern (R/K)Xn(R/K)↓ (n= 0, 2, 4, and 6, where X represents any amino acid and “↓” the scissile peptide bond) 3. Furin, often regarded as the prototypical PC, prefers the core consensus motive R-X-K/R-R↓

4-6

. Substrate

recognition is facilitated solely by the catalytic domain, which harbors the active site and requires the additional P- or HomoB-domain for catalytic activity. The substrate binding cleft of furin is highly negatively charged 5, 7-9. Structurally known enzyme-substrate interactions at the S1-, S2- and S4-pockets of classical furin substrates are mainly facilitated by charge driven interactions as well as hydrogen bond networks. Substrate binding to furin is accompanied by a structural change of the active site cleft to trigger its catalytically active conformation 8. Potent substrate-like inhibitors were developed based on the canonical cleavage motif of furin including inhibitory peptides (e.g.

9-16

) and proteins (e.g.

17-19

).

Because of the high sequence homology of the substrate binding pockets among the PCs, the development of isoenzyme specific inhibitors is a challenging task 20, 21. Targeting binding pockets beyond S4 might be a key strategy to achieve a more specific inhibition of certain PC family members 22. In this region the amino acid conservation and hence the preservation of the interaction surface was shown to be lower. Indeed, P5-extensions of peptide-like furin inhibitors showed comparable lower inhibition of PC7 and PC2

11

. N-terminal multi-leucine

extensions improved the selectivity of peptide inhibitors for PACE4 compared to furin

23, 24

.

Sequence preferences beyond the consensus motif were also suggested by substrate 3 ACS Paragon Plus Environment

Biochemistry 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

Page 4 of 36

profiling approaches performed on known furin substrates or on a genome wide scale

25-28

. In

fact, basic amino acids are overrepresented also at the P5 and P6 positions of furin substrates. In hemagglutinins of highly pathogenic avian influenza viruses, which are activated by furin

29

, insertions of arginine at P5 and P6 increased substrate turnover and

enhanced viral pathogenicity 30, 31.

Poly-arginine peptides were also shown to strongly inhibit furin as well as the processing of furin substrates in living cells

13, 32

. Elongation of tetra-peptides by P5- and P6-arginines

largely improved the Ki values and hence underlines the importance of these positions for protease-ligand interactions 13.

A potential S5-pocket was identified by structural analyses of furin-inhibitor complexes

7-9

. In

these structures 3- or 4-guanidinomethyl-phenylacetyl moieties are present at the P5 position of the inhibitors. Similarly, binding of P5-arginine residues might be sterically possible at this interaction site as suggested by molecular modelling 33. In contrast, lysine side chains are too short to trigger such interactions and probably require a different binding mode. Structural data about protease-substrate interactions at the S6 site also remained completely unknown so far.

Targeting of specific steric properties as well as differences in charge density at the active site clefts of PC family members could be an alternative approach to gain more specificity 22. Several atypical substrates of furin lack basic amino acids usually found at the P2 or P4 positions according to the consensus cleavage motif (e.g. in Protein C 34, Albumin 35, proIGFIA

36

or Hepatitis B virus P22

37

, also reviewed in

28

). So far it is largely unclear how such

atypical substrates interact with furin’s substrate binding pocket. Interestingly, such substrates frequently contain basic amino acids at the P5 or P6 positions. It is of high interest for inhibitor development to gain information about potential alternative protease-substrate interaction modes. Substitutions of charged amino acids by non-charged analogs as found

4 ACS Paragon Plus Environment

Page 5 of 36 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

Biochemistry

for atypical substrates and a consequential reduction in net charge might be used to increase the bioavailability of furin inhibitors.

In this study we analyzed the P4, P5 and P6 interactions of furin with substrate-like peptide inhibitors using X-ray crystallography. The structures provide detailed insights how substrate specificity is mediated at these binding sites.

5 ACS Paragon Plus Environment

Biochemistry 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

Page 6 of 36

Materials and Methods

Expression, Purification and Crystallization of Furin. The detailed procedures for expression and purification of homogenously glycosylated human furin (UNIPROT ID 7, 8

. Shortly, the protein was expressed by transient

P09958) are given in previous work

transfection of HEK293S cells and purified in three chromatography steps including immobilized metal affinity chromatography, immobilized inhibitor affinity chromatography and gel permeation chromatography (GPC). For crystallization of unliganded furin (~9 mg/ml in 10 mM Hepes, pH 7.5, 100 mM NaCl, 2 mM CaCl2) the protease was mixed with equal volumes of crystallization solution (100 mM MES, 200 mM K/NaH2PO4, pH 5.5-6.0 and 3-4 M NaCl and 3% DMSO) and equilibrated against 3-4 M NaCl in the reservoir of vapor diffusion experiments at 20°C as described previously 8.

Synthesis and analysis of inhibitors. The syntheses and the characterization of the inhibitors I1

38

(H-Arg-Arg-Arg-Val-Arg-4-aminomethyl-benzamidine), I2

Val-Arg-4-aminomethyl-benzamidine), aminomethyl-benzamidine), I4

10

I3

9

10

(phenylacetyl-Cit-

(4-aminomethyl-phenylacetyl-Arg-Tle-Arg-4-

(acetyl-Val-Arg-4-aminomethyl-benzamidine), and I5

(phenylacetyl-Arg-Val-Arg-4-aminomethyl-benzamidine)

12

was described previously. The

chemical structures of the inhibitors are given in Fig. 1.

Crystal soaking and structure determination. The crystals were soaked for ~16 h in soaking solution (3.16 M NaCl; 100 mM Mes/NaOH; pH 5.5; 200 mM Na/KH2PO4; 1 mM CaCl2) supplemented with 2 mM I1, 1 mM I2 or 1 mM I3. Prior to flash cooling in liquid N2 the crystals were transferred to soaking solution supplemented with 11% ethylene glycol and 0.8 mM I1, 0.4 mM I2 or 0.4 mM I3 for several seconds. Diffraction data collection was performed at the synchrotron beamline 14.1 of the Helmholtz Zentrum Berlin (BESSY II) The data were processed using XDS

40

(v.01/2016) with XDS-APP

41

39

.

(v1.0) and the CCP4

program suite 42 (v.7.0.019). COOT 43 (v.0.8.3) was used for model building. Refinement was performed in PHENIX

44

(v.1.11.1) using the structure of furin in complex with the peptide 6 ACS Paragon Plus Environment

Page 7 of 36 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

Biochemistry

inhibitor 3-guanidinomethyl-phenylacetyl-Arg-Val-Arg-4-aminomethyl-benzamidine (PDB-ID 5JXH 8) as initial model. The Rfree-set was always transferred from PDB-ID 5JXH. Geometry restraints for refinement of the inhibitors were generated using the PRODRG-server Electron density maps were calculated in PHENIX

44

45

.

(v.1.11.1) performing positional

refinement with simulated annealing of the structures with omitted inhibitors. Molecular graphics and

sequence based

(http://www.pymol.org). COOT

43

structure alignments

were calculated

in PYMOL

(v.0.8.3) was used for manual docking and geometry

optimization of the penta-peptide H-Lys-Arg-Val-Arg-4-aminomethyl-benzamidine. The coordinates and structure factors of the complexes of furin with I1, I2 and I3 have been deposited at the World Wide Protein Data Bank with the PDB IDs 6EQX, 6EQV and 6EQW, respectively.

Substrate sequence analyses. “IceLogos” the FurinDB

27

46

were generated from substrates deposited at

, showing only amino acids with significant differences (P