DECIPHERING THE PROTEOLYTIC MECHANISM OF THE ATP-DEPENDENT
PROTEASE LON USING FLUORESCENT PEPTIDES
by
JESSICA WARD
Submitted in partial fulfillment of the requirements
For the degree of Doctor of Philosophy
Thesis Adviser: Dr. Irene Lee
Department of Chemistry
CASE WESTERN RESERVE UNIVERSITY
January, 2008
CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the thesis/dissertation of
______
candidate for the ______degree *.
(signed)______(chair of the committee)
______
______
______
______
______
(date) ______
*We also certify that written approval has been obtained for any proprietary material contained therein. TABLE OF CONTENTS
Title page…………………………………………………………………………………..i
Committee Sign-off sheet…………………………………………………………...... ii
Table of Contents…………………………………………………………………………iii
List of Tables…………………………………………………………………………….vii
List of Figures…………………………………………………………………………...viii
Acknowledgements………………………………………………………………………xii
List of Abbreviations……………………………………………………………………xiii
Abstract………………………………………………………………………………….xvi
CHAPTER 1 Introduction to Lon protease……………..…………………………………1
CHAPTER 2 Evaluating the contributions of nucleotide binding and hydrolysis towards
the proteolytic activity of E. coli Lon protease…………………………………..11
2.1 Introduction…………………………………………………………………..12
2.2 Methods and Materials……………………………………………………….15
2.2.1 Peptidase activity with various nucleotides………………………..15
2.2.2 Peptidase activity with limiting ATP……………………………....16
2.3 Results and Discussion………………………………………………………17
2.3.1 Peptidase activity with various nucleotides………………………..17
2.3.2 Peptidase activity with limiting ATP………………………………22
2.4 Conclusions…………………………………………………………………..24
CHAPTER 3 Pre-steady-state kinetic characterization of the peptidase activity of E. coli
Lon protease……………………………………………………………………...28
3.1 Introduction…………………………………………………………………..29
iii 3.2 Methods and Materials……………………………………………………….32
3.2.1 Measuring peptide hydrolysis using a discontinuous acid-quench
assay……………………………………………………………...32
3.2.2 Pre-steady-state kinetic analysis of peptide hydrolysis……………32
3.2.3 ADP inhibition of the lag phase in peptide hydrolysis…………….34
3.3 Results and Discussion………………………………………………………36
3.3.1 Measuring peptide hydrolysis using a discontinuous acid-quench
assay……………………………………………………………..36
3.3.2 Pre-steady-state kinetic analysis of peptide hydrolysis……………37
3.3.3 ADP inhibition of the lag phase in peptide hydrolysis…………….43
3.4 Conclusions…………………………………………………………………..46
CHAPTER 4 Investigating the mechanism of E. coli Lon protease using proteolytically
inactive Lon mutants……………………………………………………………..50
4.1 Introduction…………………………………………………………………..51
4.2 Methods and Materials……………………………………………………….53
4.2.1 Generation and characterization of Lon mutants…………………..53
4.2.2 Fluorescence emission scans……………………………………….54
4.2.3 Measuring peptide binding using fluorescence anisotropy………...54
4.2.4 Monitoring Lon-peptide interactions using pre-steady-state kinetic
techniques………………………………………………………..55
4.2.5 ADP inhibition of the Lon-peptide interaction…………………….58
4.3 Results and Discussion………………………………………………………60
4.3.1 Developing an assay to monitor Lon-peptide interactions………...60
iv 4.3.2 Monitoring the Lon-peptide interaction with S679A using pre-
steady-state kinetic techniques…………………………………69
4.3.3 Measuring peptide binding to Lon using fluorescence anisotropy.72
4.3.4 Examining events prior to peptide hydrolysis using S679W
and pre-steady-state kinetic techniques…………………………73
4.3.5 ADP inhibition of the Lon-peptide interaction……………………84
4.4 Conclusions………………………………………………………………….86
CHAPTER 5 Exploring the substrate specificity of E. coli Lon protease using peptide
substrates and λN protein deletion mutants……………………………………..94
5.1 Introduction………………………………………………………………….95
5.2 Materials and Methods………………………………………………………98
5.2.1 Peptide design and peptidase assays………………………………98
5.2.2 λN purification and generation of λN deletion mutants………….100
5.2.3 λN degradation assay……………………………………………..102
5.3 Results and Discussion……………………………………………………..103
5.3.1 Examining Lon substrate specificity using λN peptides…………103
5.3.2 Alanine scan of λN89-98 peptide………………………………...109
5.3.3 Importance of the C-S cleavage site in the λN89-98 peptide……114
5.3.4 Degradation of a 2-site peptide…………………………………..116
5.3.5 Wild type λN and λN deletion mutant protein purification and
degradation……………………………………………………..118
5.4 Conclusions…………………………………………………………………124
CHAPTER 6 Conclusions and future directions……………………………………….127
v Appendix: Jessica Ward’s publications………………………………………………..137
Bibliography……………………………………………………………………………183
vi LIST OF TABLES
CHAPTER 2
2.1. Steady-state kinetic parameters of NTP-dependent peptidase activity……………20
2.2 Peptidase activity rate constants at limiting, stoichiometric and saturating nucleotide
concentrations…………………………………………………………………..23
2.3 Steady-state kinetic parameters of intrinsic and peptide stimulated
NTPase activity…………………………………………………………………25
CHAPTER 3
3.1 Kinetic parameters associated with the pre-steady-state characterization of E. coli
Lon protease……………………………………………………………………..41
CHAPTER 4
4.1 Primers used in site-directed mutagenesis…………………………………………..53
4.2 Kinetic parameters for ATP binding and hydrolysis by wildtype and Lon mutants...62
4.3 Kinetic constants for λN89-98 dansyl peptide interacting with S679A and S679W..72
4.4 Experimental and theoretical rate constants for the E. coli Lon mechanism………...93
CHAPTER 5
5.1 Steady-state kinetic parameters for peptidase activity with λN peptides…………..107
5.2 Steady-state kinetic parameters for ATPase activity with λN peptides……………109
5.3 Steady-state kinetic parameters for peptidase activity with λN89-98 alanine
scan peptides……………………………………………………………………113
vii LIST OF FIGURES
CHAPTER 1
1.1 The domain organization of E. coli Lon protease……………………………………..4
1.2 Proposed Ser-Lys dyad mechanism for peptide bond hydrolysis……………………..5
1.3 Proposed mechanism for E. coli Lon protease………………………………………...6
1.4 E. coli Lon protease mechanism proposed by our lab………………………………...7
1.5 Structures of the fluorescent and non-fluorescent model peptide substrate
λN89-98…………………………………………………………………………..9
1.6 Peptidase activity assay………………………………………………………………10
CHAPTER 2
2.1 Chemical structures of nucleotides…………………………………………………..14
2.2 Scheme for ATP and peptide binding to Lon………………………………………..16
2.3 Peptidase activity at varying concentrations of peptide in the presence of
saturating ATP, CTP, GTP and UTP…………………………………………….18
2.4 Peptidase activity with saturating peptide at varying concentrations of
ATP, CTP, GTP and UTP………………………………………………………..19
2.5 Structures of adenine and cytidine…………………………………………………...21
2.6 Peptidase activity with limiting amounts of ATP or AMPPNP……………………...23
2.7 Limited tryptic digestion of Lon with various nucleotides…………………………..26
2.8 Proposed mechanism for peptide hydrolysis………………………………………...27
CHAPTER 3
3.1 Diagram of stopped-flow and rapid chemical quench……………………………….31
3.2 Monitoring the peptidase reaction using a discontinuous acid-quench assay……….37
viii 3.3 Stopped-flow time courses of peptide hydrolysis by E. coli Lon……………………38
3.4 Steady-state kinetics of peptide cleavage……………………………………………40
3.5 Substrate dependency of the pre-steady-state lag phase in the peptidase reaction…..42
3.6 ADP lengthens the pre-steady-state lag phase of the peptidase reaction…………….44
3.7 ADP inhibits the pre-steady-state of peptide hydrolysis……………………………..45
3.8 Pre-steady-state time courses for ATP hydrolysis and λN89-98 degradation
under identical reaction conditions………………………………………………47
3.9 Proposed mechanism for peptide hydrolysis by Lon………………………………...48
CHAPTER 4
4.1 ATPase activity of Lon mutants……………………………………………………..61
4.2 Peptidase activity of Lon mutants……………………………………………………62
4.3 MANT-ATP binding to Lon…………………………………………………………63
4.4 Limited tryptic digestion analysis of S679A, S679W and S679W, W297F,
W303F, W603F Lon mutants…………………………………………………...64
4.5 Emission scan of S679A Lon with the λN89-98 NF peptide……………………….65
4.6 Emission scan of S679A Lon with λN89-98 dansyl peptide………………………..66
4.7 Peptide binding to S679A can be monitored using the λN89-98 dansyl peptide
and monitoring dansyl fluorescence…………………………………………….68
4.8 Peptide binding to S679A can be monitored using the λN89-98 dansyl peptide and
monitoring tryptophan fluorescence…………………………………………….68
4.9 λN89-98 dansyl peptide binding to S679A is dependent on peptide……………….70
4.10 Scheme for peptide binding to Lon………………………………………………..70
4.11 λN89-98 dansyl peptide binding to S679A is dependent on ATP…………………71
ix 4.12 Equilibrium λN89-98 dansyl binding to Lon can be monitored using fluorescence
anisotropy………………………………………………………………………..73
4.13 Intrinsic tryptophan fluorescence can be used to measure a conformational change
in S679W dependent on ATP and AMPPNP…………………………………….75
4.14 Representative time courses for S679W interacting with the λN89-98 dansyl
peptide at varying concentrations of ATP……………………………………….76
4.15 Representative time courses for S679W interacting with the λN89-98 dansyl
peptide at varying concentrations of λN89-98 dansyl peptide…………………..78
4.16 The first phase of the S679W reaction is dependent on ATP………………………79
4.17 The first phase of the S679W reaction is not dependent on peptide………………..80
4.18 The second phase of the S679W reaction is dependent on ATP…………………...81
4.19 The second phase of the S679W reaction is dependent on peptide………………...82
4.20 AMPPNP supports the peptide-Lon interaction at a reduced rate………………….83
4.21 ADP inhibits the first phase of the S679W reaction………………………………..84
4.22 ADP inhibits the second phase of the S679W reaction…………………………….85
4.23 Proposed mechanism for the first round of peptide hydrolysis by E. coli
Lon protease……………………………………………………………………..88
4.24 Simulation of the proposed peptide hydrolysis mechanism using KinFitSim……...92
CHAPTER 5
5.1 λN protein sequence and cleavage profile…………………………………………...97
5.2 Amino acid sequences of the fluorescent λN peptides………………………………99
5.3 Peptidase activity with λN11-21……………………………………………………104
5.4 Peptidase activity with λN55-65……………………………………………………105
x 5.5 Steady-state kinetics of hydrolysis of the λN peptides……………………………..107
5.6 Calibration curve of peptide digested with trypsin…………………………………111
5.7 Relative kinetic parameters for λN89-98 alanine scan peptides……………………112
5.8 The P3, P1 and P3’ positions are crucial for determining substrate specificity……114
5.9 The P1 position is more important than the P1’ position…………………………..115
5.10 Sequence of the “2-site” peptides…………………………………………………117
5.11 Peptidase activity with the “2-site” peptides……………………………………...118
5.12 SDS-PAGE analysis of purified λN protein………………………………………119
5.13 Wildtype λN protein degradation with and without ATP…………………………120
5.14 λN protein deletion mutants……………………………………………………….122
5.15 λN protein deletion mutant degradation…………………………………………..122
5.16 The C-terminus of the λN protein is important for recognition by Lon…………..123
CHAPTER 6
6.1 Kinetic mechanism for E. coli Lon protease………………………………………..129
6.1 Scheme for a double mixing experiment on the stopped-flow instrument…………132
6.2 First double mixing experiment…………………………………………………….134
6.3 Second double mixing experiment…………………………………………………134
xi ACKNOWLEDGEMENTS
First, I would like to thank my advisor, Irene Lee, for being a great teacher and
mentor over the last five years. Thank you for challenging me and always being accessible and supportive. I want to thank my labmates, Jen Fishovitz, Jason Hudak,
Hilary Frase, Xuemei Zhang, Ed Motea, Jon Huang and James Becker for making the lab an enjoyable environment to work and learn. Thank you to Tony Berdis for all of his scientific help and advice. I also want to thank my best friend, Diana Barko who kept me going throughout this entire journey with scientific discussions, encouragement and
“telling me a story”. Most of all, thank you to my husband, Kevin, my mom and dad and the rest of my family for their unconditional love and support while I pursue my educational goals.
xii LIST OF ABBREVIATIONS
λN 89-98 F a fluorescent model peptide substrate consisting of residues 89-98
of the λN protein and containing the fluorescence donor Abz and
the fluorescence quencher Y-NO2
λN 89-98 NF a nonfluorescent analog of λN89-98 F in which the Abz moiety is
replaced with Bz and the Y-NO2 moiety is replaced with tyrosine.
AAA+ ATPases associated with various cellular activites
Abz anthranilamide, the fluorescent quencher in the fluorescent
peptides
ADP adenosine diphosphate
AMPPNP adenylyl 5- imidodiphospate, a nonhydrolyzable ATP analog
Bz benzoic acid, used in the nonfluorescent peptides in place of Abz
CTP cytidine triphosphate
Clp an ATP-dependent protease belonging to the same family as Lon
DMSO dimethyl sulfoxide
DTT dithiothreitol
E. coli Escherichia coli
FRET fluorescence resonance energy transfer
GTP guanosine triphosphate
HCl hydrochloric acid
HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
xiii HPLC high performance liquid chromatography
HSLU the ATPase subunit of an ATP-dependent protease belonging to
the same family as Lon
HSLV the protease subunit of an ATP-dependent protease belonging to
the same family as Lon
IC50 the concentration of inhibitor required to inhibit the reaction by
50%
IPTG isopropyl β-D-1-thiogalactopyranoside, used to induce gene
expression
kburst the pre-steady-state rate constant for ATPase activity in Lon
kcat the maximum first order rate constant for Lon catalysis
kcat/Km substrate specificity constant, a lower limit for the second-order
rate constant of substrate binding, used to compare the utilization
of different substrates for an enzyme
Kd an equilibrium dissociation constant
klag the pre-steady-state rate constant for peptidase activity in Lon
Km Michaelis Menten constant, the concentration of substrate required
for the enzyme reaction to reach half maximum velocity
kobs an observed rate constant
MANT -ATP 2’- (or 3’-) O-(N-methylanthraniloyl) andenosine triphosphate, a
fluorescent ATP analog used to measure ATP binding to Lon
xiv MG262 a peptidyl boronate and known inhibitor of the proteasome and
Lon n the Hill coefficient, n > 1 suggests positive cooperativity in an
enzyme, n < 1 suggests negative cooperativity in an enzyme
NaOH sodium hydroxide
Ni-NTA nickel nitriliotriacetic acid, an affinity matrix used for purifying 6x
His tagged proteins
PCR polymerase chain reaction
Pd1 the hydrolyzed product of λN89-98 containing the 5 C-terminal
amino acids
PMT photomultiplier tube
SB superbroth, culture media used to grow E. coli consisting of
bactotryptone and yeast extract
SBTI soybean trypsin inhibitor
SDS-PAGE sodium dodecyl sulfate polyacrylamide gel electrophoresis
SSD substrate sensor and discriminatory domain
UTP uridine triphosphate
Y-NO2 nitrotyrosine, the fluorescence quencher used in the fluorescent
peptides
xv Deciphering the Proteolytic Mechanism of the ATP-Dependent Protease Lon Using Fluorescent Peptides
Abstract
by
JESSICA WARD
Lon is an ATP-dependent serine protease that degrades damaged, misfolded and
certain regulatory proteins in the cell. The enzyme exists as a homo-oligomer with one
ATPase domain and one protease domain on each subunit. Using our fluorescent model
peptide substrate, λN89-98, as a tool, we have employed steady-state and pre-steady-state
kinetic techniques to evaluate the proteolytic mechanism of Lon protease. Steady-state
kinetic analysis of the peptidase activity of Lon with nucleotide triphosphates reveals that while nucleotide binding alone is enough to support peptide hydrolysis, the contribution of nucleotide hydrolysis towards the proteolytic activity is significant. Pre-steady-state experiments reveal a lag phase in the peptidase reaction and that ATP hydrolysis occurs before the peptide cleavage event. We propose that the slow step constitues a peptide translocation event that contains two binding events in which the peptide is delivered from the initial peptide binding site to the protease active site. I was able to measure this slow step using proteolytic inactive Lon mutants and a dansylated version of the λN89-98 peptide as well. Taken together, I propose a mechanism for the first round of peptide hydrolysis. Following initial nucleotide and peptide binding, a nucleotide dependent conformational change occurs. Next, upon ATP hydrolysis, the proteolytic site is activated and peptide is delivered to the protease domain where peptide cleavage occurs.
xvi In addition to characterizing the kinetic mechanism of Lon protease, I have also utilized multiple peptide substrates to demonstrate the importance of the P3, P1 and P3’ positions in determining substrate specificity and provide evidence that the C-terminus of the λN protein is important for recognition by Lon.
xvii
CHAPTER 1
INTRODUCTION TO LON PROTEASE
1 Lon (protease La) is an ATP-dependent serine protease found ubiquitously in
nature. In prokaryotic cells the enzyme is found in the cytosol, while in eukaryotes it is
nuclearly encoded and localized to the mitochondria (1-7). The enzyme acquired its
name from the phenotype of E. coli (Escherichia coli) bacteria lacking the Lon gene,
which appear longer than the wild-type cells (8). Lon, like other ATP dependent
proteases, belongs to the AAA+ (ATPases Associated with various cellular Activities) family of proteins. These proteins contain an ATP hydrolysis domain with Walker A and
B motifs and are involved in important cellular functions such as DNA replication, transcriptional regulation, membrane fusion and proteolysis (9-11). Lon represents one of the simplest of these proteases with both the ATPase and protease domains on the same protein subunit (12-14).
Lon is responsible for maintaining proper cellular function by degrading abnormal and short-lived regulatory proteins in the cell (12, 13, 15-18). In bacteria, Lon is not only one of the major contributors to the elimination of damaged proteins but it also regulates many important cellular functions. For example, it regulates activities such as the SOS response by degrading the SulA protein that inhibits cell division and the production of capsular polysaccharide by degrading the transcriptional activator RcsA (19, 20). Lon function is also critical for maintaining the structure and integrity of mitochondria (21).
Studies have indicated that mitochondrial Lon can selectively degrade oxidized proteins
(22). Oxidized proteins accumulate in the mitochondria due to free radicals from oxidative stress and aging. Therefore, Lon is implicated in age related mitochondrial dysfunction (22, 23). In certain pathogenic bacteria such as Salmonella enterica
Typhimurium and Brucella abortus, Lon has been shown to be necessary for systemic
2 infection. Lon proteolysis is required to regulate the correct timing for expression of virulence genes (24-27). Identification of specific bacterial Lon inhibitors could be used as novel therapeutics in treating these bacterial infections (28, 29). Yeast cells lacking
Lon protease have large deletions in their mitochondrial genome and are respiration- deficient (30-32). Lon protease is extremely important for many biological activities, and many details about how the enzyme functions are not known. Therefore, our lab is focused on designing experiments that will provide information on the mechanism of
Lon.
The crystal structure of the E. coli Lon protease domain indicates the enzyme exists as a hexamer (33). Early investigators using gel filtration and glycerol density gradient centrifugation methods indicated the enzyme is a tetramer or octamer (2, 3), while unpublished data from our lab suggests the oligomeric state is dependent on the concentration of enzyme (Lee and Burke, unpublished result). Nevertheless, unlike most other ATP-dependent proteases in which the ATPase and protease domains are separate and oligomerize to form a functional enzyme, Lon is a homo-oligomer with both the
ATPase and protease domains on the same monomeric unit.
With a molecular weight of approximately 88 kDa, each monomer of E. coli Lon consists of four domains; the N-terminal domain, the ATPase domain, the SSD (substrate sensor and discriminatory) domain and the protease domain (Figure 1.1) (3, 7, 34). The
N-terminal domain is poorly conserved amongst the different Lon homologues and while a crystal structure for the E. coli enzyme has been published, its function is not clear at this time (35). The ATPase domain (sometimes called the α/β domain) is highly conserved amongst the members of the AAA+ family (9) and contains the nucleotide
3 binding motifs (Walker A and B) (34). The crystal structure for a portion of the ATPase
domain has been reported, however this structure does not include the Walker A and B
motifs (36). The SSD domain is relatively small and consists of about 100 amino acids.
It has been expressed and purified and shown to fold into a stable structure resistant to proteolytic digestion. Binding assays with the isolated domain indicate it can discriminate between different protein substrates (37). A crystal structure of the protease domain has recently been published and suggests Lon utilizes a Ser679-Lys722 catalytic dyad for peptide bond cleavage (see Figure 1.2 for proposed mechanism) (33, 38). The protease domain crystallographic data also reveals Lon oligomerizes into a barrel shaped structure.
Figure 1.1. The domain organization of E. coli Lon protease. Lon exists as a homo-oligomer. Each monomer consists of 4 domains: an N-terminal domain with an unknown function, an ATPase domain containing the ATP binding motifs, the SSD domain which is known to be able to bind to different protein substrates, and the protease domain containing the catalytic serine residue.
N-terminal ATPase Substrate sensor Protease + and - H3N CO2 domain domain discriminator domain domain
4 Figure 1.2. Proposed Ser-Lys dyad mechanism for peptide bond hydrolysis. 1) Nucleophilic attack 2) Formation of acyl-enzyme 3) Acyl-enzyme hydrolysis 4) Regeneration of active site
P1'
NH
H2N NH - NH O O 1 P1 O O O 722 P1 HN ' P Lys Ser679 P1 2 1 ' H HN P1 O O NH2 HO H N+ H H 2 Ser679 679 O Ser 722 Lys722 Lys O Acyl-enzyme Tetrahedral NH2 intermediate intermediate
NH P1 4 NH 3
O P1 - OH O
OH H O + 679 H2N Ser
Lys722
Many of the initial mechanistic characterizations of Lon protease were performed by Goldberg and coworkers and Maurizi and coworkers. Based on their data they proposed a mechanism for Lon depicted in Figure 1.3 (3, 39). First, ATP binds and activates the proteolytic site. The peptide bond is then cleaved followed by ATP hydrolysis, which leads to inactivation due to ADP being bound. Binding of additional protein substrate promotes the release of ADP (ADP release is the rate limiting step) allowing another ATP molecule to bind and the cycle to repeat.
5 Figure 1.3. Proposed mechanism for E. coli Lon protease. ATP binding activates the proteolytic site and the peptide bond is cleaved. Upon ATP hydrolysis, the enzyme is inactivated due to ADP being bound. Binding of additional protein substrate promotes the release of ADP allowing another ATP molecule to bind and the cycle to repeat. Adapted from Goldberg (3, 39).
Lon:ADP Protein “inactive” substrates
Peptide Products ADP
Pi ATP Lon:ATP “active”
Extensive product inhibition studies by our lab led to a revised proposed kinetic
mechanism of Lon shown in Figure 1.4 (40). The intrinsic ATPase activity (no peptide
or protein substrate) is shown in steps 1-3. Lon binds and hydrolyzes ATP and the
product ADP binds to the enzyme tighter than ATP. When peptide/protein is present, the
peptide is delivered accompanied by ATP binding and hydrolysis (steps 1` and 2`).
When peptide and ATP are in excess, peptide binds to an allosteric site on Lon to promote ATP/ADP exchange (steps 3`-5`). An additional molecule of ATP binds allowing the peptide to be hydrolyzed (step 6`). Lon then isomerizes into a post-catalytic
“F” form upon peptide cleavage (step 7`). The specifics of the isomerization and how
enzyme turnover occurs (steps a-c) are very poorly understood and will need to be addressed in the future.
6 Figure 1.4. E. coli Lon protease mechanism proposed by our lab. In the absence of peptide/protein substrate, Lon has an intrinsic ATPase activity (steps 1- 3). When peptide/protein is present, the peptide is delivered accompanied by ATP binding and hydrolysis (steps 1` and 2`). Additional peptide then binds to the allosteric site on Lon to promote ATP/ADP exchange (steps 3`-5`). Another molecule of ATP binds allowing the peptide to be hydrolyzed (step 6`). Lon then isomerizes into a post- catalytic “F” form upon peptide cleavage (step 7`). Enzyme turnover is not well understood (steps a-c). Adapted from Thomas-Wohlever and Lee (40).
2 3 E:ATP E:ADP E + ADP 1 Pi
a E + ATP c b ATP 1' peptide E ATP F F Pd1 + Pd2 peptide ATP 7' E peptide E ATP 2' Pi F Pd1 + Pd2 ADP E* peptide* 6' peptide 3' ATP E*peptide* ADP peptide E* ADP peptide* ATP peptide 5' 4' E* peptide*
7 We know that optimum peptide bond cleavage requires ATP hydrolysis, however the mechanism of how ATP binding and hydrolysis mediate proteolytic cleavage is not well understood. To aid in the study of the coordination of these two events, our lab has created a fluorescent model peptide substrate, λN89-98 (Figure 1.5) (41). λN89-98 consists of residues 89-98 of the native E. coli Lon substrate, the λN protein. Until the development of λN89-98, most mechanistic characterization of Lon was performed using full length proteins or short fluorogenic tetrapeptides (3, 15, 42-44). Ideally, protein substrates would be used to study Lon because they are more physiologically relevant.
However, they are not good to use because they have multiple Lon cleavage sites. With several cleavage sites in one substrate, it is difficult to correlate a single peptide cleavage event with ATP binding and hydrolysis. The short fluorogenic tetrapeptide substrates require ATP to be cleaved by Lon; however they fail to stimulate the ATPase activity like protein substrates. Our substrate, λN89-98, is superior to and much more physiologically relevant than the tetrapeptide substrates. It is unique in that it stimulates the ATPase activity of Lon, has a substrate specificity constant (kcat/Km) similar to the full length λN protein, and contains a single Lon cleavage site (40, 41).
8 Figure 1.5. Structures of the fluorescent and non-fluorescent model peptide substrate λN89-98. λN89-98 consists of residues 89-98 of the λN protein. The fluorescent analog contains the fluorescent donor Abz and the fluorescent quencher Y-NO2. In the non-fluorescent analog the Abz is replaced with Bz and the Y-NO2 is replaced with Y. Both peptides are degraded identically by Lon. The black arrow indicates the Lon cleavage site.
λN89-98 λN89-98 fluorescent non-fluorescent
+ - + - H3N-Y(NO2)-RGITCSGRQ-K(Abz)-CO2 H3N-Y-RGITCSGRQ-K(Bz)-CO2
NO2 NH NH OH OH O O Quencher NH2
Donor
As mentioned above, λN89-98 is a fluorescent peptide containing residues 89-98 of the λN protein with the fluorescent donor, anthranilamide (Abz), on the C-terminal lysine residue and the fluorescent quencher, 3-nitrotyrosine (Y-NO2), on the N-terminus
Figure 1.5) (41). The Lon cleavage site is between the Cys93 and Ser94 residues. Upon
cleavage of the peptide, the Abz and Y-NO2 separate resulting in an increase in
fluorescence over time (Figure 1.6). Using this model peptide we are able to
continuously monitor the peptide hydrolysis activity of Lon. One limitation of this peptide is the reduced fluorescence changes at high concentrations of peptide due to the inner filter effect (45). The inner filter effect becomes a problem when intermolecular quenching rather than intramolecular quenching occurs at high concentrations of Y-NO2.
To overcome this obstacle we have developed a non-fluorescent version of λN89-98,
9 where Abz is replaced with benzoic acid amide (Bz) and Y-NO2 is replaced with
tyrosine, and showed that the kinetics of the degradation of the two peptides are identical.
By using a mixed peptide substrate (consisting of 90% non-fluorescent and 10%
fluorescent), we are able to continually monitor the kinetics of peptide cleavage at high
concentrations of peptide.
Figure 1.6. Peptidase activity assay. In the presence of Lon and ATP, the peptide λN89-98 is hydrolyzed. Peptide cleavage is monitored by an increase in fluorescence over time as the fluorescence donor (Abz) moves away from the fluorescence quencher (Y-NO2). The rate of the reaction is determined from the slope of the linear portion of the time course.
Y(NO2)-RGITCSGRQ-K(Abz)
1.4 10 5
5 Lon 1.2 10
+ 1 10 5 ATP
8 10 4
6 10 4
4 Y(NO2)-RGITC + SGRQ-K(Abz) 4relative fluorescence 10 0 50 100 150time 200 250 300 350
Given the lack of a full, intact crystal structure or details concerning the molecular
mechanism of Lon protease, our lab has chosen to primarily take a kinetic approach in
characterizing the enzyme. I have concentrated on elucidating the protease mechanism of
E. coli Lon. Specifically, I have utilized pre-steady-state kinetic techniques to determine the kinetic mechanism of the proteolytic activity, explored the role of ATP binding and hydrolysis in peptide cleavage and looked at how Lon recognizes and degrades substrates. This work has resulted in three publications, one manuscript accepted for publication and an additional manuscript in the final stages of preparation.
10
CHAPTER 2
EVALUATING THE CONTRIBUTIONS OF NUCLEOTIDE BINDING AND
HYDROLYSIS TOWARDS THE PROTEOLYTIC ACTIVITY OF E. COLI LON
PROTEASE
11 2.1 INTRODUCTION
Lon is a homo-oligomeric serine protease that is activated by ATP (2, 3). The
-1 enzyme has an intrinsic ATPase activity (kcat, intrinsic ~ 0.2 s ) that is stimulated by the
-1 presence of peptide and protein substrates (kcat, stimulated ~ 1 s ) (41, 43, 46). As a member
of the AAA+ (ATPases Associated with various cellular Activities) family of proteases,
Lon shares many structural similarities with other ATP-dependent proteases such as
HslU/V (a bacterial homologue of the proteasome) (9, 47). HslV is an oligomeric
protease subunit and HslU is an oligomeric ATPase subunit. The two subunits assemble
together to form a functional protease (48). A complete crystal structure of Lon is not
available at this time, however based on the structural similarities and sequence
homology between HslU/V and Lon, HslU/V is a good structural model for studying
Lon.
HslU functions to unfold polypeptide substrates and translocate them through the
central cavity where they can be degraded by the protease subunit HslV (49, 50). By
analogy, our lab proposed that the ATPase activity in Lon is used to translocate
polypeptide substrates into the protease domain for degradation (40). Kinetic evidence
obtained in our lab also supports this proposal. Using our model peptide substrate λN89-
98, we showed that the degradation of this short, single cleavage site peptide is
maximized by ATP hydrolysis (40, 41). AMPPNP (a nonhydrolyzable ATP analog)
supports peptide cleavage at a rate that is ~ 7 fold slower than with ATP. As λN89-98
lacks any defined secondary structure, ATP hydrolysis is not necessary for substrate
unfolding; rather it could be used to support a peptide translocation event.
12 Initial studies with Lon indicated that other nucleotides, cytidine triphosphate
(CTP), guanosine triphosphate (GTP) and uridine triphosphate (UTP) (Figure 2.1) can
support proteolysis of the protein casein at a reduced rate as compared to ATP and that
these nucleotides are hydrolyzed by Lon (15). Based on these results and the kinetic data
generated in our lab, I propose that the differences in proteolytic activity with the other
nucleotides are due to differences in the nucleobase structure. CTP, GTP and UTP will
bind to Lon with a lower affinity, but will support peptide hydrolysis to a greater extent
than AMPPNP because of their ability to be hydrolyzed. The kcat values for the
nucleotide-dependent peptidase reactions should be at least equal to if not higher than the
kcat value for the AMPPNP-dependent reaction if nucleotide hydrolysis provides a greater catalytic advantage than nucleotide binding. To test this hypothesis and determine how
ATP binding and hydrolysis contribute to the proteolytic activity of Lon; I performed an in-depth kinetic study of the peptidase activity of Lon in the presence of ATP, CTP, GTP and UTP.
13 Figure 2.1. Chemical structures of nucleotides. ATP, CTP, GTP and UTP support the peptidase activity of Lon and are hydrolyzed by the enzyme. AMPPNP is a non-hydrolyzable analog of ATP that supports peptide cleavage.
NH2 ATP N NH2 Adenosine Triphophate N AMPPNP N N Adenylyl 5-imidodiphosphate N O O O N
-O P PO OPO N O O O O N
O- O- O- H H H -O P N P OPO O H H OH OH O- O- O- H H H H OH OH O NH GTP 2 N Guanosine Triphosphate NH CTP Cytidine Triphophate N N O O O N NH 2 O N O O O -O P PO OPO O - H H O P OPO OP O- O- O- O H H O- O- O- H H OH OH
O H H OH OH
UTP HN Uridine Triphosphate
O N O O O
-O P OPO OP O O- O- O- H H
H H OH OH
As stated above, Lon functions as a homo-oligomeric enzyme with one ATPase
and one protease domain on each monomeric subunit. However, the ATPase sites in Lon
are unique in that they exhibit two different affinities for ATP (51, 52). The high-affinity
sites bind ATP tightly (Kd < 1 μM) and the low-affinity sites bind ATP relatively weakly
(Kd ~ 10 μM). Using various experimental techniques, our lab demonstrated that the
-1 high-affinity sites hydrolyze ATP very slowly ( kobs ~ 0.01 s ) and the low-affinity sites
-1 hydrolyze ATP rapidly (kobs ~ 11 s ) (52). In addition to examining the contributions of
ATP binding and hydrolysis, I also performed experiments to elucidate how the high- affinity and low-affinity ATPase sites contribute to peptidase activity. The work presented in this chapter has resulted in two publications (46, 52).
14 2.2 MATERIALS AND METHODS
2.2.1 Peptidase activity with various nucleotides
Peptide synthesis and protein purification procedures were performed as
described previously (40). Steady-state velocity data were collected on a Fluoromax 3
spectrofluorimeter (Horiba Group) equipped with a temperature regulated cell holder set
to 37 °C. All reactions were performed in a 3 mm path length cuvette from Hellma. All
assays contained 50 mM Tris-HCl (pH 8.1), 5 mM magnesium acetate, 5 mM DTT
(dithiothreitol), and 125 nM E. coli Lon except in experiments using GTP where 200 nM
E. coli Lon was used, with varying concentrations of peptide substrate (25 to 500 μM).
Each reaction was initiated with the addition of varying concentrations (0-1 mM) of
nucleotide (ATP, CTP, GTP or UTP). The peptide substrate used in the assays consisted
of 10% λN89-98 F (fluorescent) and 90% λN89-98 NF (nonfluorescent) to avoid
complications from the inner filter effect (40, 45). Peptide hydrolysis was monitored by
an increase in fluorescence upon excitation at 320 nm and emission at 420 nm. The
amount of peptide hydrolyzed was calibrated by measuring the fluorescence generated
per micromolar of peptide after complete digestion of known amounts of peptide by
trypsin under identical reaction conditions.
The steady-state rate of each reaction was determined from the slope of the linear
portion of the reaction time course. The observed rate constant for each reaction (kobs) was determined by dividing the observed steady-state rate by the concentration of enzyme used in the reaction. All reactions were performed in triplicate and the averaged rates for each set of nucleotide and peptide concentrations was fit to a general bireactant equation described by equation 2.1
15 n n n kobs = kcat [A] [B] / (KibK’a + K’a[B] + Kb[A] + [A] [B]) (2.1)
where kobs is the observed rate constant, kcat is the maximal rate constant, A is the peptide substrate, B is the nucleotide, n is the Hill coefficient, Ka is the Michaelis constant for A
(peptide), Kib is the intrinsic dissociation constant for B (nucleotide) and Kb is the
Michaelis constant for B (nucleotide) (Figure 2.2). The Michaelis constant for peptide hydrolysis, Ka, was calculated from the relationship log K’a = n log Ka. Global fitting
was performed using Enzfitter (Biosoft) (40, 53).
Figure 2.2. Scheme for ATP and peptide binding to Lon. Binding scheme for a general bireactant enzyme. Kia is the intrinsic dissociation constant for peptide, Ka is the Michaelis constant for peptide, Kib is the intrinsic dissociation constant for nucleotide, Kb is the Michaelis constant for nucleotide.
Kib Lon + ATP Lon:ATP Ka +peptide Lon:ATP:peptide
+ATP K Lon + peptide Lon:peptide b
Kia
2.2.2 Peptidase activity with limiting ATP
All data were collected on a Fluoromax 3 Spectrofluorimeter as described above.
Each reaction contained 50 mM HEPES (pH 8.0), 75 mM potassium acetate, 5 mM DTT,
5 mM magnesium acetate, 1 mM peptide substrate (10% λN89-98 F and 90% λN 89-98
NF) , 5 or 6 μM E. coli Lon, and 0.5 μM, 5 μM or 100 μM ATP or AMPPNP (adenylyl
5- imidodiphospate). The reaction time courses were calibrated and the kobs for each reaction was determined as described above.
16 2.3 RESULTS & DISCUSSION
2.3.1 Peptidase activity with various nucleotides
Previously, our lab examined the kinetics of ATP and AMPPNP (a non-
hydrolysable ATP analog) mediated peptide hydrolysis by Lon and found that AMPPNP
supports peptidase activity at a reduced rate (~ 7 fold slower as compared to ATP) (40).
This result suggested that while ATP binding to the enzyme is sufficient, the energy from
ATP hydrolysis is required for optimum peptidase activity. To further explore how ATP
binding and hydrolysis are coupled with the peptidase activity of Lon, the steady-state
kinetics of peptide cleavage with various nucleotides was determined.
The steady-state rate constants of peptide hydrolysis were measured under conditions of saturating nucleotide (ATP, CTP, GTP or UTP) and varying concentrations
of peptide substrate. The predominant enzyme form is Lon:nucleotide under these
conditions, therefore the differences in peptidase activity can be attributed to differences in nucleotide hydrolysis and not binding. The observed rate constants plotted as a function of [peptide] yields a sigmoidal curve (Figure 2.3) indicating some degree of cooperativity and the need for a Hill coefficient (n) to be associated with the peptide concentration term in equation 2.1. Cooperativity was observed previously by our lab.
As Lon functions as a homo-oligomer, the apparent cooperativity is likely attributed to interaction or communication between the enzyme subunits during peptide hydrolysis
(40). Peptidase activity was also measured under conditions of saturating peptide and varying concentrations of nucleotide. Plotting the observed rate constants as a function of nucleotide concentration yields a hyperbolic curve (Figure 2.4).
17 Figure 2.3. Peptidase activity at varying concentrations of peptide in the presence of saturating ATP, CTP, GTP and UTP. Lon was preincubated with 25, 50, 100, 150, 200 and 500 μM λN89-98 and the reaction was initiated with 1 mM ATP (red), CTP (blue), GTP (black), or 1.6 mM UTP (green). The reaction was monitored by an increase in fluorescence at excitation 320 nm and emission 420. The observed rate constants (kobs) were determined from the linear portions of the time courses and plotted as a function of peptide concentration.
10
8
6 ) -1 (s obs k 4
2
0 0 100 200 300 400 500 600
[peptide] (μM)
18 Figure 2.4. Peptidase activity with saturating peptide at varying concentrations of ATP, CTP, GTP and UTP. Lon was preincubated with 500 μM λN89-98 and the reaction was initiated with 5, 10, 50, 75, 150, 250, and 1000 μM ATP (red), 30, 75, 150, 200, 500, 800 and 1000 μM CTP (blue), 50, 75, 150, 250, 500, 1000, 1300 μM GTP (black), or 100, 200, 400, 600, 800, 1600 mM UTP (green). The reaction was monitored by an increase in fluorescence at excitation 320 nm and emission 420. The observed rate constants (kobs) were determined from the linear portions of the time courses and plotted as a function of nucleotide concentration.
12
10
8 ) -1
(s 6 obs obs k
4
2
0 0 200 400 600 800 1000 1200 1400 1600
[NTP] (μM)
The kinetic constants resulting from globally fitting the data with equation 2.1 are shown in Table 2.1. The Hill coefficient (n) and the Michaelis constant for peptide (Ka)
are not affected by the binding and hydrolysis of nucleotide as these values vary only
slightly amongst the different nucleotide dependent reactions. On the contrary, there are
large variations in the kcat, Kb (Michaelis constant for nucleotide) and Kib (intrinsic
dissociation constant for nucleotide) values. The Kib and Kb values for each nucleotide
are very similar indicating that the Michaelis constant reflects the intrinsic binding
19 affinity of the nucleotide for the enzyme. According to the kcat values, the nucleotides are able to activate the peptidase activity of Lon in the following order: ATP > CTP > UTP
~ GTP > AMPPNP. Interestingly, with the exception of AMPPNP, this order correlates well with decreasing binding affinity (Kb or Kib) of the nucleotides to Lon: ATP ~
AMPPNP > CTP > GTP > UTP.
Table 2.1. Steady-state kinetic parameters of NTP-dependent peptidase activity. The AMPPNP data was taken from (40). ND: not determined.
-1 kcat (s )Ka (μM) Kb (μM) Kib (μM) n
ATP 9.0 + 0.5 102 + 30 7 + 17.4 + 2 1.6
CTP 4.2 + 0.1 151 + 43 100 + 20 73 + 14 1.52
UTP 1.9 + 0.2 99 + 32 350 + 125 389 + 112 1.43
GTP 1.7 + 0.2 219 + 43 200 + 78 250 + 92 1.6
AMPPNP 1.0 + 0.1 77 + 7 ND 10 + 1 1.6
CTP binds to Lon with a weaker affinity than ATP or AMPPNP, but has a 4.2 fold higher kcat than AMPPNP. Comparing the structures of all the nucleotides (Figure
2.1) suggests which structural features are important for nucleotide binding. Figure 2.5 shows the structures of ATP and CTP, which bind with the highest affinity, overlaid on top of one another. The figure illustrates that ATP and CTP share similarities in the N6
(ATP) and N4 (CTP) amino groups in the pyrimidine ring. Pate et al. used constrained energy minimization to show that the N6 amino group of ATP and the N4 amino group of
20 CTP are only 0.7 Å apart (54). Therefore, it is likely that this amino group interacts with the enzyme and is responsible for activating the peptidase activity of Lon. However, this binding interaction alone is not enough. ATP and AMPPNP bind with comparable affinities; however there is a large difference in their kcat values. AMPPNP is nonhydrolyzable and is a poor activator. This result illustrates the importance of nucleotide hydrolysis in the mechanism of Lon. The UTP and GTP data emphasize this point as well. These nucleotides bind very poorly to Lon; however their ability to be hydrolyzed by the enzyme offers a distinct catalytic advantage over AMPPNP. While nucleotide binding is important and can sustain Lon proteolysis, nucleotide hydrolysis is necessary for optimal enzyme activity.
Figure 2.5. Structures of adenine and cytidine. A side and front view of the adenine and cytidine structures overlaid shows that the N6 amino group in ATP and the N4 amino group in CTP are in very close spatial proximity. Adenine is in blue and cytidine is in red.
Side Front
21 2.3.2 Peptidase activity with limiting ATP
The differences in ATP binding affinity were exploited to examine the functional
nonequivalency of the two ATPase sites and how that relates to peptide hydrolysis.
Peptidase activity was monitored under single turnover conditions (0.5 μM ATP, 5 μM
Lon, 1 mM peptide substrate) where the predominant enzyme form is homo-oligomeric
Lon with ATP bound only at the high affinity ATPase sites. As shown in Figure 2.6, this
-1 enzyme form can undergo multiple rounds of peptide hydrolysis (kobs = 0.32 s , Table
2.2). This experiment illustrates that ATP and peptide hydrolysis are not stoichiometrically linked as multiple peptide bonds are hydrolyzed for each molecule of
ATP hydrolyzed. The same experiment was also performed with limiting amounts of
AMPPNP. As shown in Figure 2.6, no peptidase activity is detected. Therefore, we can conclude that under this condition at least one molecule of ATP must be hydrolyzed for peptide cleavage to occur.
Peptide hydrolysis was also measured under stoichiometric Lon/ATP conditions
(5 μM ATP, 5 μM E. coli Lon, 1 mM peptide substrate) where the predominant enzyme form is the high-affinity ATPase sites saturated with ATP and under conditions of saturating ATP (100 μM ATP, 5 μM E. coli Lon, 1 mM peptide substrate) where both the high- and low-affinity ATPase sites are occupied. The results are shown in Table 2.2.
-1 The rate at 100 μM ATP (kobs, 100 μM ATP = 2.69 s ) is faster than at 5 μM ATP (kobs, 5 μM
-1 ATP = 1.52 s ) therefore both ATPase sites are contributing to the peptidase activity of
Lon. Optimal peptide hydrolysis only occurs when both ATPase sites are occupied by
ATP.
22 Figure 2.6. Peptidase activity with limiting amounts of ATP or AMPPNP. 5 μM Lon was incubated with 1 mM λN89-98 and the reaction was initiated with 0 (black), 0.5 (blue) or 100 (red) μM ATP or 0.5 (orange) or 100 (green) μM AMPPNP. Peptide hydrolysis was monitored by an increase in fluorescence at excitation 320 nm emission 420 nm. The rate of the reaction was determined from the linear portion of each time course.
1000
800 M) μ 600
400
peptide hydrolyzed ( hydrolyzed peptide 200
0
0 50 100 150 200 250 300
time (s)
Table 2.2. Peptidase activity rate constants at limiting, stoichiometric and saturating nucleotide concentrations.
[nucleotide] kobs (μM) (s-1) limiting ATP 0.5 0.32 ± 0.07 (high affinity sites occupied) stoichiometric ATP 5 1.52 ± 0.05 (high affinity sites saturated) saturating ATP 100 2.69 ± 0.30 (all sites saturated) saturating AMPPNP 100 0.96 ± 0.08 (all sites saturated)
23 2.4 CONCLUSIONS
Lon is a serine protease with proteolytic and ATPase activities that are tightly
coupled (2, 3). We hypothesized that the dependency on ATP-hydrolysis is a peptide
translocation step similar to those found in other ATP-dependent proteases based on
initial kinetic characterization of the ATP and AMPPNP- dependent peptidase reactions
and structural similarities between Lon and other ATP-dependent proteases. To test this
hypothesis, I examined the steady-state kinetic parameters of the peptidase activity of
Lon with various nucleotides (ATP, CTP, GTP and UTP).
Initial studies indicated Lon hydrolyzes ATP as well as CTP, GTP and UTP,
however the kinetic parameters associated with this activity were not reported (15).
Extensive kinetic studies on the nucleotide hydrolysis reaction have been published by our lab and the results are shown in Table 2.3 (46). All nucleotides are intrinsically hydrolyzed by Lon, and the NTPase activity is stimulated in the presence of λN 89-98 NF peptide. The Km values are very similar amongst the different nucleotides; however there
are large differences in the kcat values. The kcat/Km values in the presence of peptide
indicate that ATP is the best activator of Lon activity, which coincides with the peptidase
activity data (Table 2.1).
24 Table 2.3. Steady-state kinetic parameters of intrinsic and peptide stimulated NTPase activity (46).
intrinsic peptide-stimulated k /K k k /K kcat Km cat m cat Km cat m NTPase (s-1) (μM) (X 103 M-1s-1) (s-1) (μM) (X 103 M-1s-1) enhancement
ATP 0.26 + 0.02 47 + 10 5.5 1.0 + 0.1 49 + 520 3.8
CTP 0.14 + 0.02 60 + 10 2.3 0.28 + 0.02 69 + 20 4.1 2 UTP 0.50 + 0.02 100 + 4 5.0 1.1 + 0.1 132 + 30 8.3 2.2 GTP 0.09 + 0.02 57 + 5 1.8 0.15 + 0.02 42 + 11 3.6 1.7
Our lab also performed limited tryptic digestion analysis and was able to probe a
nucleotide dependent conformational change in Lon (46). Lon was incubated with
various nucleotides and trypsin for a defined period followed by quenching with SBTI
(soybean trypsin inhibitor) and SDS-PAGE analysis. As evidenced by the 67 kDa protein band (Figure 2.7), Lon adopts a more compact conformation in the presence of adenine
nucleotides (ATP, ADP, AMPPNP). Protein sequencing data indicated that this band
corresponds to the ATPase, SSD, and peptidase domains in Lon. The relative stability of the 67 kDa fragment in the presence of the various nucleotides is the same as the ability of the nucleotides to support peptide hydrolysis (ATP > CTP > UTP ~ GTP) suggesting that the adenine induced conformational change is important for activating peptide
hydrolysis.
25 Figure 2.7. Limited tryptic digestion of Lon with various nucleotides. Lon was incubated with λN89-98 and various nucleotides and digested with trypsin. The reactions were quenched at 0, 15 and 30 minutes with SBTI (soybean trypsin inhibitor) and analyzed by SDS-PAGE. The adenine induced conformational change is monitored by the stability of the 67 kDa protein band corresponding to the ATPase, SSD and protease domains. The relative stability of the 67 kDa fragment in the presence of the various nucleotides is the same as the ability of the nucleotides to support peptide hydrolysis (ATP > CTP > UTP ~ GTP).
no ATP ADP AMPPNP ATP GTP CTP UTP nucleotide kDa kDa
89 89 67 67 45 45 35 35 26 26 23 23 SBTI SBTI time time 01530 0 15 30 0 15 30 0 15 30 (min) (min) 01530 0 15 30 0 15 30 0 15 30
Taken together, the data supports a mechanism for the proteolytic activity of Lon,
illustrated in Figure 2.8, in which ATP binding induces a series of conformational
changes in the enzyme thereby activating the protease activity. The Lon:ATP enzyme
form (form II, Figure 2.8) adopts a more compact conformation and the pore leading to the proteolytic chamber is “opened” allowing peptide/protein substrates access to the
protease domain. The energy derived from ATP hydrolysis facilitates the translocation of
substrate to the protease active site (form III, Figure 2.8). Because AMPPNP supports peptide hydrolysis, albeit at a slower rate as compared to ATP, I propose that the adenine
induced conformational change (form II, Figure 2.8) is enough to allow peptide delivery
to occur, however the energy from ATP hydrolysis is required for optimal activity. Once
26 delivered, the substrate is degraded by Lon and the hydrolyzed peptide products and ADP
are released.
Figure 2.8. Proposed mechanism for peptide hydrolysis. Free enzyme is depicted by form I with the ATPase and SSD domains in green and the protease domain in blue. The enzyme is shown as a dimer instead of a hexamer for simplicity. ATP binds causing a conformational change (form II) in the enzyme allowing the peptide/protein substrate access to the protease domain. Upon ATP hydrolysis, the substrate is translocated into the proteolytic chamber where hydrolysis can occur (form III). Since AMPPNP supports peptide hydrolysis albeit at a slower rate as compared to ATP, we propose that the adenine induced conformational change (form II) is sufficient to allow peptide delivery to occur, however the energy from ATP hydrolysis is required for optimal activity.
peptide
A A P T P D AT P AD P peptide ATP ATP binding hydrolysis (I) (II) (III)
The mechanism proposed in Figure 2.8 can be further tested using pre-steady-
state (transient) kinetic techniques. Pre-steady-state kinetic techniques provide information about individual rate constants along the reaction pathway. By comparing the pre-steady-state rates at which Lon hydrolyses ATP and peptide, we can determine which event occurs first along the reaction pathway. In the model, ATP is hydrolyzed before peptide bond cleavage. Therefore, if the proposed kinetic mechanism is accurate, we anticipate the pre-steady-state rate constant for ATP hydrolysis to be faster than the pre-steady-state rate constant for peptide hydrolysis. These proposed transient kinetic experiments are described in detail in Chapter 3.
27
CHAPTER 3
PRE-STEADY-STATE KINETIC CHARACTERIZATION OF THE
PEPTIDASE ACTIVITY OF E. COLI LON PROTEASE
28 3.1 INTRODUCTION
To fully understand the mechanism of Lon protease and how it processes protein
substrates, it is important to determine how the ATPase and peptidase activities are
coupled. Lon has an intrinsic ATPase activity that is stimulated ~ 4 fold in the presence
of peptide and protein substrates (43, 46). Minimally, the peptidase activity requires the
binding of ATP. However, substrate degradation is optimized when ATP is hydrolyzed
(15, 40, 41, 46). Based on structural similarities between Lon and other ATP dependent
proteases as well as data from our lab indicating that the nonhydrolyzable ATP analog,
AMPPNP, supports peptide hydrolysis at a reduced rate as compared to ATP, it is
proposed that the energy from ATP hydrolysis is used to deliver peptide and protein
substrates to the protease domain for degradation (40, 46). The energy from ATP
hydrolysis is also used to unfold protein substrates in other ATP dependent proteases (55-
57). This could also be true for Lon, however our model peptide (λN 89-98) is a short
peptide with no defined structure, therefore the ATPase activity must have an additional
purpose. If the proposed mechanism is correct (Figure 2.8), one would expect that ATP is hydrolyzed before the peptide cleavage event. The experiments described in this chapter are aimed at determining the timing of ATP hydrolysis and peptide cleavage in the mechanism of Lon.
Important aspects of the kinetic mechanism of an enzyme are obtained from steady-state kinetic analysis. From these studies we can determine the order of binding and release of reactants and products as well as parameters such as kcat, Km and kcat/Km
(58). These parameters are complex functions of all the rate constants along the reaction pathway and provide information such as the lower limit for all the first-order rate
29 constants and the apparent second-order rate constant for substrate binding. Used in conjunction with steady-state studies, pre-steady-state (or transient) kinetic techniques are invaluable for elucidating the mechanism of an enzyme. Pre-steady-state methods use stopped-flow or chemical quench flow instruments to rapidly mix the reactants and monitor the reaction on the millisecond timescale (the time for a single enzyme turnover)
(Figure 3.1) (58-60). Using these instruments we are able to examine the formation of individual enzyme intermediates rather than the catalytic turnover of the enzyme. When properly designed, transient kinetic experiments provide information about the rate limiting step of the reaction as well as the rate constants of individual events. A lag phase in a pre-steady-state time course indicates a step before the step being monitored is rate limiting while a burst phase reveals a step after the step being monitored is rate limiting.
Used together, steady-state and pre-steady-state methods are powerful tools for understanding how enzymes work.
30 Figure 3.1. Diagram of stopped-flow and rapid chemical quench. The reagents to be mixed are loaded into syringes A and B. As the motor driven drive plate is lowered, the contents of the two syringes are rapidly mixed. In the stopped-flow instrument, the reagents enter the observation cell where the reaction is monitored continuously by fluorescence. In the rapid chemical quench, the reagents are mixed for a defined period of time followed by chemical quenching. The reaction aliquots at various time points are collected and analyzed by the appropriate technique. A time course with a lag phase indicates a step before the step being monitored is rate limiting. A time course with a burst phase indicates a step after the step being monitored is rate limiting.
Stopped-Flow Rapid Chemical Lag kinetics AB Quench AB
Step prior to chemistry is rate limiting
time
quench observation Burst kinetics cell Δt
Monitor changes in Step after fluorescence chemistry is rate limiting
time
Our lab previously performed extensive steady-state kinetic analysis on Lon
protease, which led to the proposed mechanism summarized in chapter 2 (Figure 2.8) (40,
46). Initial stopped-flow experiments have also been completed and shown that the
peptidase activity of Lon displays lag kinetics. A lag in this pre-steady-state time course
indicates the rate limiting step in the reaction occurs before peptide bond cleavage. Here
we present the results of an extensive study using pre-steady-state kinetic methods to
characterize both the peptidase and ATPase activities of Lon. In support of the proposed mechanism, it was discovered that ATP is hydrolyzed before the peptide cleavage event.
The work presented in this chapter has resulted in one publication (61).
31 3.2 MATERIALS & METHODS
3.2.1 Measuring peptide hydrolysis using a discontinuous acid-quench assay
Peptide synthesis and protein purification procedures were performed as described previously (40). Each reaction was performed at 37°C and contained 50 mM
Tris-HCl (pH 8.1), 5 mM magnesium acetate, 5 mM DTT, 800 μM λN89-98 peptide substrate and 1 μM E. coli Lon. The λN 89-98 peptide substrate was a mixed substrate containing 50% λN89-98 F and 50% λN89-98 NF. The reaction was initiated with 500
μM ATP. Ten microliter aliquots were quenched with 58 μL of 0.5 N HCl at 0, 0.5, 1 and 2 minutes. Lon was removed by trichloroacetic acid precipitation and the reaction was neutralized to pH 8.0 with 1 M Tris/2 N NaOH. The fluorescence of each aliquot was measured (excitation 320 nm emission 420 nm) using a Fluoromax 3 spectrofluorimeter (Horiba Group).
3.2.2 Pre-steady-state kinetic analysis of peptide hydrolysis
All pre-steady-state peptidase reactions were performed on a KinTek stopped
Flow controlled by the data collection software Stop Flow version 7.50 β. The PMT
(photomultiplier tube) was equipped with a 400 nm longpass filter to detect fluorescence emission > 400 nm. The sample syringes were maintained at 37 °C by a circulating water bath. Syringe A contained 5 μM Lon monomer, with variable concentrations of
λN89-98 peptide substrate (5 – 500 μM), 5 mM magnesium acetate, 50 mM Tris-HCl
(pH 8.0), and 5 mM DTT. The λN89-98 peptide substrate was a mixture of fluorescent and nonfluorescent peptides (10% λN89-98 F and 90% λN89-98 NF) to avoid problems from the inner filter effect (40, 45). Syringe B contained varying concentrations of ATP
(1 – 500 μM). The syringe contents were rapidly mixed in the sample cell and peptide
32 hydrolysis was detected by an increase in fluorescence (excitation 320 nm emission 420
nm). The time courses shown are a result of averaging at least four traces and each time
course was normalized to zero. The concentration of hydrolyzed peptide was calibrated
by completely digesting known amounts of peptide with trypsin and determining the
maximum fluorescence generated per micromolar peptide under identical reaction
conditions in the stopped flow apparatus. The averaged and calibrated time courses were
fit to equation 3.1
–k t Y = A exp lag + νsst + C (3.1)
where Y is the concentration of hydrolyzed peptide in micromolar, A is the amplitude of
the lag phase of the reaction, klag is the pre-steady-state rate constant in per seconds, t is
time in seconds, νss is the steady-state rate in micromolar per second, and C is the endpoint. The first order rate constant (kss in per seconds) is obtained by dividing νss by the enzyme concentration (62). Equation 3.1 describes a biphasic time course. When Y
= 0 at t = 0, C = - A and equation 3.1 becomes
–k t Y = - A + A exp lag + νsst
such that
-k t Y = - A(1-exp lag ) + νsst when A is νss – νi/klag, where νi is the rate corresponding to the initial phase of the reaction, equation 3.1 is equal to the equation defining hysteresis (40, 63, 64).
33 All of the rate constants (klag and kss) were plotted as a function of substrate
concentration. Hyperbolic plots were fit with equation 3.2
k = (kmax[S]) / (KS + [S]) (3.2)
where k is the observed rate constant (klag or kss), kmax is the maximum rate constant
(referred to as kcat for kss data and klag,max for klag data), S is the variable substrate, KS is the Michaelis constant for the respective substrate, the concentration of substrate required to reach 50% of the maximal rate constant (referred to as Km,λN89-98 and Km,ATP).
Sigmoidal plots were fit with equation 3.3
n n k = (kmax[S] ) / (K’S + [S] ) (3.3)
where k is the observed rate constant (klag or kss), kmax is the maximum rate constant
(referred to as kcat for kss data and klag,max for klag data), S is the variable substrate, n is the
Hill coefficient. The Michaelis constant, KS, is calculated from the relationship log K’S =
n log KS (see equation 3.2 for description).
3.2.3 ADP inhibition of the lag phase in peptide hydrolysis
Pre-steady state peptide hydrolysis was measured as previously described above in section 3.2.2. Syringe A contained 50 mM HEPES pH 8.0, 75 mM potassium acetate,
5 mM magnesium acetate, 5 mM DTT, 5 μM Lon, 500 μM λN89-98 peptide (10% fluorescent, 90% nonfluorescent mixture) and varying ADP (0-50 μM). Syringe B contained 50 mM HEPES pH 8.0, 75 mM potassium acetate, 5 mM DTT and 500 μM
ATP. Peptide cleavage was detected by an increase in fluorescence (excitation 320 nm and emission 420 nm) upon rapid mixing of the syringe contents. The data shown are a result of averaging at least four traces. The concentration of peptide hydrolyzed was calibrated by determining the fluorescence generated per micromolar peptide after
34 digestion with trypsin under identical reaction conditions in the stopped-flow apparatus.
Time courses were fit with equation 3.1. The observed rate constants were plotted as a
function of ADP and fit with equation 3.4
klag/klag,max = 1 / (1+( [ADP] / IC50)) (3.4)
where klag is the observed pre-steady state rate constant from equation 3.1, klag,,max is the observed rate constant with no ADP in per seconds and IC50 is the concentration of ADP
required to inhibit the reaction rate by 50% in micromolar units.
35 3.3 RESULTS & DISCUSSION
3.3.1 Measuring peptide hydrolysis using a discontinuous acid-quench assay
Previously, our lab used a continuous pre-steady-state fluorescence assay to show that the peptidase activity of Lon displays lag kinetics in the presence of ATP and the length of the lag phase is increased in the presence of AMPPNP (40). The presence of a lag in the pre-steady-state time course suggests that the rate determining step in the reaction occurs before peptide bond cleavage. However, as the fluorescence signal is a result of the fluorescence quencher (Y-NO2 on the N-terminal peptide product fragment)
separating from the fluorescence donor (Abz on the C-terminal peptide product fragment)
the lag phase could also be a result of slow dissociation of the peptide products from the
active site of Lon.
I used a discontinuous acid-quench assay to monitor the peptidase activity of Lon
and determine if the lag phase in the peptidase time course can be attributed to slow
dissociation of the peptide products rather than a slow step in the reaction before
chemistry. The reaction was initiated with ATP and aliquots were quenched with HCl to
denature Lon and release the hydrolyzed peptide products. The fluorescence of each
reaction aliquot was measured and plotted as a function of time. A line drawn tangent to
the linear (steady-state) portion of the time course intercepts the y – axis (t = 0) at a point
below y = 0 indicating a lag still exists in the peptidase time course (Figure 3.2) (58-60).
The experiment qualitatively supports that the lag kinetics observed in the continuous assay are a result of a slow phase in the reaction before peptide bond cleavage and not slow release of peptide products.
36 Figure 3.2. Monitoring the peptidase reaction using a discontinuous acid-quench assay. Lon was incubated with 800 μM λN89-98 and the reaction was initiated by the addition of 500 μM ATP. Reaction aliquots were quenched at various time points with 0.5 N HCl. The fluorescence signal associated with each aliquot was measured and plotted at the corresponding time point. This experiment demonstrates that the lag phase is not a result of a slow dissociation of the peptide products from the enzyme active site.
20
16
12
8 relative fluorescence 4
0 00.511.522.5
time (min)
3.3.2 Pre-steady-state kinetic analysis of peptide hydrolysis
Initial analysis of the peptide hydrolysis reaction showed lag kinetics in the pre-
steady-state indicating a step before chemistry is rate limiting (40). Based on the
mechanisms of other ATP-dependent proteases, which use ATP-hydrolysis to deliver
substrates to the proteolytic active site, we proposed that the slow step in Lon is an ATP-
dependent substrate translocation event (56, 65, 66). To test this hypothesis, I performed
additional pre-steady-state experiments to fully characterize the lag phase and determine
the kinetic parameters associated with it.
37 I monitored λN 89-98 peptide hydrolysis using a stopped-flow apparatus under conditions of saturating ATP (500 μM) and varying peptide (5-500 μM) as well as saturating peptide (500 μM) and varying ATP (1-500 μM). As shown in Figure 3.3, a lag followed by a steady-state is observed under all ATP and peptide concentrations. Each time course was calibrated to show the amount of peptide hydrolyzed. Known amounts of peptide were digested by trypsin in the stopped flow apparatus under identical reaction conditions as the Lon reactions. A calibration curve was generated by determining the maximum fluorescence change associated with each peptide concentration.
Figure 3.3. Stopped-flow time courses of peptide hydrolysis by E. coli Lon. Five micromolar Lon was preincubated with 5 (grey), 10 (pink), 25 (orange), 50 (black), 100 (green), 200 (blue) and 500 (red) μM λN89-98 and rapidly mixed with 500 μM ATP or 5 μM Lon was preincubated with 500 μM λN89-98 and rapidly mixed with 1 (grey), 5 (pink), 10 (orange), 50 (black), 100 (green), 200 (blue) and 500 (pink) μM ATP. Peptide hydrolysis was monitored by an increase in fluorescence (excitation 320 nm, emission 420 nm). The y-axis was converted to the amount of peptide hydrolyzed using a calibration curve generated by determining the maximum fluorescence reading after complete digestion of known amounts of peptide by trypsin.
80 80 80 80 70 5 μM Lon 70 M) 5 μM Lon μ M) 60 vary ATP 60 500 μMATP μ 60 60 500 μM λN89-98 vary λN89-98 50 50
40 40 40 40 30 peptide hydrolyzed ( hydrolyzed peptide 30 peptide hydrolyzed peptide hydrolyzed ( 20 20 20 20
10 10
0 012345 0 time (s) 012345 time (s)
38 The calibrated time courses were fit with equation 3.1 describing a lag followed
by a steady-state to yield the pre-steady-state rate constant (klag) and the rate constant
describing the steady-state of the reaction (kss). Figures 3.4 A and B show the steady
state rate (kss) as a function of peptide and ATP concentration, respectively. Figure 3.4 A
displays sigmoidal behavior and was fit with equation 3.3 describing the Hill equation
while Figure 3.4 B is hyperbolic and was fit with equation 3.2 describing the Michaelis
Menten equation. Fitting the data in this manner yielded the maximum rate constant for
peptide cleavage (kcat), the Michaelis constants for peptide and ATP (Km,λN89-98 and
Km,ATP) as well as the Hill coefficient (n). All the kinetic values are summarized in Table
3.1. According to a previous steady-state kinetic study performed by our lab with 125
-1 nM Lon, kcat = 7.7 s , Km,λN89-98 = 85 μM and Km,ATP = 7.2 μM (40). Because the results
presented in Table 3.1 with 5 μM Lon are in close agreement with these values, we can assume that the concentration of enzyme does not affect the steady-state kinetic parameters.
39 Figure 3.4. Steady-state kinetics of peptide cleavage. The stopped-flow time courses were fit with equation 3.1 and the steady-state rate constants (kss) for the reactions were plotted as a function of the corresponding peptide (A) or ATP (B) concentration. The data in A were fit with equation 3.3 describing a -1 sigmoidal curve and kinetic parameters obtained were kcat = 5.5 ± 0.8 s , Km,λN89-98 = 188 ± 148 μM, and n = 1.2 ± 0.2. The data in B were fit with equation 3.2 describing a -1 hyperbolic curve and the kinetic parameters obtained were kcat = 4.2 ± 0.1 s , Km,ATP = 9.7 ± 0.9 μM. All kinetic parameters are summarized in Table 3.1. 5 A 4
) 3 -1 (s ss k 2
1
0 0 100 200 300 400 500
[peptide] (μM)
5 B
4
) 3 -1 (s ss k 2
1
0 0 100 200 300 400 500
[ATP] (μM)
40 Table 3.1 Kinetic parameters associated with the pre-steady-state characterization of E. coli Lon protease. The values not in parentheses were determined using the steady-state rate constants (kss) while the values in parentheses were determined using the pre-steady-state lag rate constant (klag).
500 μMATP 500 μM λN89-98 vary [λN89-98] vary [ATP]
-1 -1 kcat 5.5 + 0.8 s 4.2 + 0.1 s
Km,λN89-98 188 + 148 μM (67 +34 μM) NA
n 1.2 + 0.2 μM (1.7+ 0.2μM) NA
Km,ATP NA 9.7 + 0.9 μM (7.2 + 1.9 μM)
k -1 -1 lag 0.88 + 0.07 s 1.14 + 0.06 s
The pre-steady-state rate constant (klag) is shown as a function of peptide and ATP
in Figure 3.5 A and B, respectively. Note that klag is dependent on the concentration of
ATP as well as the concentration of peptide. As with the kss data, the varying peptide
data was sigmoidal and fit with equation 3.3 while the varying ATP data was hyperbolic
and fit with equation 3.2. From these fits come the kinetic parameters shown in Table 3.1.
-1 -1 The maximum lag rate constant (klag,max) is 0.88 s and 1.14 s at varying peptide and
varying ATP concentrations, respectively. Under conditions of saturating peptide and
ATP, as is true for the reaction described by klag,max, substrate binding is no longer rate
limiting as can be the case with low concentrations of peptide and/or ATP. Therefore,
klag,max represents the rate constant for the buildup of at least one reaction intermediate before the peptide cleavage event. In addition, the formation of this intermediate(s) is
41 dependent on ATP and peptide, and previous experiments suggest klag is also dependent
on ATP hydrolysis (klag is ~ 7 fold slower with AMPPNP as opposed to ATP) (40).
Figure 3.5. Substrate dependency of the pre-steady-state lag phase in the peptidase reaction. The stopped-flow time courses were fit with equation 3.1 and the pre-steady-state lag rate constants (klag) for the reactions were plotted as a function of the corresponding peptide (A) or ATP (B) concentration. The data in A were fit with equation 3.3 describing a −1 sigmoidal curve and kinetic parameters obtained were klag,max = 0.88 ± 0.07 s , Κm,λN89-98 = 67 ± 34 μM, and n = 1.7 ± 0.2. The data in B were fit with equation 3.2 describing a -1 hyperbolic curve and the kinetic parameters obtained were klag,max = 1.14 ± 0.06 s , Km,ATP = 7.2 ± 1.9 μM. All kinetic parameters are summarized in Table 3.1.
1 A
0.8 )
-1 0.6 (s lag lag k
0.4
0.2
0 0 100 200 300 400 500 600
[peptide] (μM)
42 1.2 B
1 )
-1 0.8 (s lag k 0.6
0.4
0.2
0 0 100 200 300 400 500 [ATP] (μM)
3.3.3 ADP inhibition of the lag phase in peptide hydrolysis
Previous work by our lab and others has shown that ADP is a potent inhibitor of
Lon activity (15, 40, 44). I performed experiments to explore further the mechanism by
which ADP inhibits the peptidase activity of Lon by investigating the effect of ADP
toward the pre-steady-state lag phase in peptide hydrolysis In these ADP inhibition
experiments, Lon was pre-incubated with a 500 μM λN89-98 peptide (a mixed substrate
of 10% fluorescent and 90% nonfluorescent) and varying concentrations of ADP before
mixing with saturating ATP (500μM). The time courses for peptide hydrolysis (500 μM)
in the presence of 500 μM ATP and various concentrations of ADP (0 to 50 μM) are
shown in Figure 3.6. It is observed that the length of the lag phase increases with ADP.
In Figure 3.7, klag is shown plotted as a function of ADP concentration. Fitting these data
43 with equation 3.4 yields an IC50 of 1.56 ± 0.06 μM, which is the concentration of ADP
required to decrease enzyme activity by 50%.
Figure 3.6. ADP lengthens the pre-steady-state lag phase of the peptidase reaction. Five micromolar Lon was incubated with 500 μM λN89-98 and 0 (red), 0.5 (blue), 1 (green), 5 (black), 10 (orange) and 50 (pink) μM ADP and rapidly mixed with 500 μM ATP. Peptidase activity was monitored by an increase in fluorescence (excitation 320 nm, emission 420 nm). The y-axis was converted to the amount of peptide hydrolyzed using a calibration curve generated by determining the maximum fluorescence after complete digestion of known amounts of peptide by trypsin.
300
250 M)
μ 200
150
100 peptide hydrolyzed ( hydrolyzed peptide 50
0 0246810
time (s)
44 Figure 3.7. ADP inhibits the pre-steady-state of peptide hydrolysis. The time courses in Figure 3.6 were fit with equation 3.1 and the pre-steady-state rate constant (klag) was plotted as a function of ADP concentration. The data were fit with equation 3.4 to determine the amount of ADP required to inhibit the reaction by half. IC50 = 1.56 ± 0.06 μM.
1
0.8
0.6 lag,max / k lag k 0.4
0.2
0 0 102030405060
[ADP] (μM)
45 3.4 CONCLUSIONS
Steady-state and pre-steady-state kinetic techniques are valuable tools for
elucidating the mechanism of an enzyme. Here I have used the results from our previous
steady-state analysis to design pre-steady-state experiments that have yielded an
abundance of information about the mechanism of Lon. Lon is a serine protease that
requires ATP to degrade peptide and protein substrates. I have investigated the timing of
ATP hydrolysis and peptide cleavage in the overall reaction scheme.
The current hypothesis is that ATP and peptide hydrolysis are coordinated and
ATP hydrolysis is used to translocate peptide/protein substrates to the proteolytic active
site where they can be degraded (40, 46). Monitoring the pre-steady-state of the
peptidase reaction revealed that peptide hydrolysis requires the buildup of at least one
reaction intermediate before bond cleavage can occur. The pre-steady-state of the
ATPase activity was also monitored by our lab and has been published (61). In these
experiments, the hydrolysis of the gamma phosphate of ATP is being monitored. The
ATPase activity displays burst kinetics, indicating the rate limiting step is after hydrolysis
of the gamma phosphate, possibly ADP release. The pre-steady-state rate constant for
-1 the ATPase activity (kburst) is ~ 11 s . Figure 3.8 compares the pre-steady-state ATPase and peptidase time courses under identical reaction conditions and shows that ATP hydrolysis occurs before peptide cleavage and provides support for the hypothesis. ATP hydrolysis should occur before peptide cleavage if the energy gained in the ATP hydrolysis reaction is used to translocate peptide substrate. Another interesting note is that the steady-state peptide stimulated ATPase reaction occurs at ~ 0.7 s-1, which is in
46 -1 close agreement with the lag rate constant (klag,max = 0.88 s ) offering additional evidence
that the ATPase activity is coupled to the peptidase activity.
Figure 3.8. Pre-steady-state time courses for ATP hydrolysis and λN89-98 degradation under identical reaction conditions. Five micromolar Lon was incubated with 100 μM ATP and 500 μM λN89-98. Peptide hydrolysis (blue) was monitored using stopped-flow fluorescence spectroscopy. ATP hydrolysis (red) was monitored using the rapid chemical quench instrument and [α−32P]ATP.
10 10
8 M) 8 μ
6 6
4 4
2
peptide hydrolyzed or ADP ( ADP or hydrolyzed peptide 2
0 00.511.52
time (s)
Other than demonstrating that the rate limiting step in the ATPase mechanism
occurs after hydrolysis of the gamma phosphate, the pre-steady-state ATPase
experiments also revealed half-site reactivity amongst the ATPase subunits. Half of the
-1 ATPase sites hydrolyze ATP before peptide cleavage (kburst ~ 11 s ) and the other half
-1 hydrolyze ATP relatively slowly (kobs ~ 0.01 s ) (52, 61). Lon has been shown to have
two different affinities for ATP so the functional nonequivalency has been attributed to
the two different types of ATPase sites (51, 61, 67). The low-affinity sites (Kd ~ 10 μM)
47 are responsible for the burst in ADP production while the high-affinity sites (Kd < 1 μM) hydrolyze ATP much slower.
Given the new information gained from the pre-steady-state experiments I have revised our previously proposed mechanism for the first round of peptide hydrolysis by
Lon protease. While I have still not directly shown that the role of ATP hydrolysis in
Lon is to translocate peptide/protein substrates through the proteolytic chamber, all of the results of the pre-steady-state experiments support this mechanism. Figure 3.9 illustrates the revised reaction scheme. Free enzyme is shown in form I with the ATPase and SSD domains in green and the protease domain in blue. The enzyme is shown as a dimer instead of a hexamer for simplicity. The peptide is depicted as a solid black line. First,
ATP and peptide bind (form II). The binding of nucleotide induces a conformational change in the enzyme (form III). Next, ATP is hydrolyzed at the low affinity sites (form
IV), thereby allowing peptide to be delivered to the protease domain (form V) where it is hydrolyzed (form VI).
48 Figure 3.9. Proposed mechanism for peptide hydrolysis by Lon. Free enzyme is depicted by form I with the ATPase and SSD domains in green and the protease domain in blue. The enzyme is shown as a dimer instead of a hexamer for simplicity. The peptide is depicted as a solid black line. First, ATP and peptide bind (form II). The binding of nucleotide induces a conformational change in the enzyme (form III). Next, ATP is hydrolyzed at the low affinity sites (form IV), thereby allowing peptide to be delivered to the protease domain (form V) where it is hydrolyzed (form VI).
ATP ATP ADP ATP ATP ATP ATP ADP ATP ADP ATP
(I) (II) (III) (IV) (V) (VI)
49
CHAPTER 4
INVESTIGATING THE MECHANISM OF E. COLI LON PROTEASE USING
PROTEOLYTICALLY INACTIVE LON MUTANTS
50 4.1 INTRODUCTION
Our lab has conducted multiple mechanistic studies on Lon protease to understand
the coupling between the ATPase and protease activities. Many other ATP-dependent
proteases such as HslU/V and Clp use the energy from ATP hydrolysis to unfold and
translocate protein and peptide substrates into the proteolytic core where they are
degraded (49, 50, 66). Much of our data suggests that Lon adopts a mechanism similar to
these other enzymes and as such a mechanism has been proposed for the first round of
peptide hydrolysis in Lon (see Chapter 3, Figure 3.9). ATP binding is enough to support
cleavage of peptide substrates; however ATP hydrolysis allows the process to occur
optimally. The steady-state peptidase data indicates that peptide hydrolysis occurs at a
rate ~ 7 fold slower in the presence of AMPPNP as compared to ATP (40). Pre-steady-
state kinetic experiments show there is a lag in the peptidase reaction and ATP is
hydrolyzed before peptide bond cleavage. Here I describe how I designed and utilized a
system to examine the microscopic events leading up to peptide hydrolysis.
Previously, Epps and coworkers as well as Fattori and coworkers independently
exploited the FRET (fluorescence resonance energy transfer) property between native
tryptophans in proteins and dansylated peptides to study protein-substrate interactions
(68, 69). In this system, the tryptophan residues are excited and transfer energy to the dansyl moiety on the peptide when the two are close to one another. The interaction between the tryptophan residues and the dansylated peptide can be monitored by exciting the tryptophan residues and monitoring the emission of the dansyl group. I utilized this approach to investigate the microscopic events leading up to peptide hydrolysis in Lon.
To this end, I synthesized the λN89-98 peptide with a dansyl fluorophore on the C-
51 terminal Lys residue (YRGITCSGRQK(dansyl)). I also generated two proteolytically
inactive Lon mutants, S679A and S679W. Both mutants contain three Trp residues that
are intrinsic to the wild-type enzyme: Trp297 and Trp303 located at the proximity of the
ATPase domain (based on the primary amino acid sequence (34)) and Trp603 at the
vicinity of the active site Ser (based on the truncated crystal structure of the protease
domain (33)). Previously, it has been shown that Ser679 is the catalytic residue that may
act as a nucleophile to attack the scissile peptide bond (7). Replacement of Ser679 with an Ala abolishes the proteolytic but not the intrinsic or protein-stimulated ATPase activity (70). Therefore, the S679A mutant should allow the detection of enzyme intermediates generated from Lon interacting with peptide upon binding and hydrolysis of ATP, but the subsequent peptide cleavage event is eliminated.
Because Trp is commonly used as an intrinsic fluorescent probe for monitoring conformational changes within proteins (71-73), the replacement of the active site Ser with Trp in S679W allows the introduction of an additional fluorescent probe to monitor conformational change(s) occurring at the proteolytic site as well as the interaction with peptide substrate. As a complete crystal structure of Lon is unavailable at this time, I
exploited the changes in the fluorescence resonance energy transfer properties of the
intrinsic Trp residues in the two proteolytically inactive mutants upon binding with the
λN89-98 dansyl peptide to detect enzyme intermediate(s) generated after nucleotide
binding and before peptide cleavage. This work has resulted in one manuscript that has
been accepted for publication in Biochemistry.
52 4.2 MATERIALS AND METHODS
4.2.1 Generation and characterization of Lon mutants
The S679A and S679W Lon mutants were generated using the wild type E. coli
Lon protease plasmid (pSG11, a generous gift from Alfred Goldberg, Harvard Medical
School) and the QuikChange Site-Directed Mutagenesis Kit from Stratagene. The forward primer and reverse primers used to generate both mutants are shown in Table
4.1. An additional “no intrinsic Trp” Lon mutant was also generated in which the three intrinsic tryptophan residues were mutated to phenylalanine residues and the catalytic serine residue was mutated to a tryptophan (S679W, W297F, W303F, W603F mutant
Lon). The “no intrinsic Trp” Lon mutant was made by sequentially mutating each residue using the primers and plasmids in Table 4.1. The sequences of all the mutant plasmids were verified by DNA sequencing. All plasmids were transformed into the BL21 DE3 E. coli cell strain.
Table 4.1. Primers used in site-directed mutagenesis. Template Plasmid plasmid Primers
pJW028 pSG11 oJW062 GCCGAAAGATGGTCCGGCTGCCGGTATTGC S679A oJW063 GCAATACCGGCAGCCGGACCATCTTTCGGC
LonS679WFor AAAGATGGTCCATGGGCCGGTATTGCT S679W pJW015 pSG11 LonS679WRev AGCAATACCGGCCCATGGACCATCTTT
pJW016 pJW015 oJW023 GTGCGTGGTTATATCGACTTCATGGTACAGGTGCCG S679W, W297F oJW024 CGGCACCTGTACCATGAAGTCGATATAACCACGCAC
S679W, W297F, pJW017 pJW016 oJW047 GGTACAGGTGCCGTTCAATGCGCGTAGCAAGG W303F oJW048 CCTTGCTACGCGCATTGAACGGCACCTGTACC
S679W, W297F, pJW018 pJW017 oJW027 GTAACCGGTCTGGCGTTCACGGAAGTGGGCGGTG W303F, W603F oJW028 CACCGCCCACTTCCGTGAACGCCAGACCGGTTAC
53 The S679A, S679W, and “no intrinsic Trp” mutant Lon proteins were expressed and purified to homogeneity using the protocols described previously for the wild type
enzyme (41). The intrinsic and peptide stimulated ATPase activity of all three mutants
was measured using radiolabeled ATP as described previously (46, 62). MANT-ATP
(2’- (or 3’-) O-(N-methylanthraniloyl) binding to S679A and S679W was measured using
stopped-flow techniques as described previously (74). Peptidase activity of S679A and
S679W was measured using the continuous λN 89-98 fluorescence assay described
previously (40, 41). The adenine-induced conformational change was monitored using
the limited tryptic digest assay described previously (46).
4.2.2 Fluorescence emission scans
Emission spectra were collected on a Fluoromax 3 spectrofluorimeter (Horiba
Group) with excitation at 290 nm. All assays were performed in microcuvettes (Hellma)
with a 3-mm path length. Reactions contained 50 mM HEPES pH 8.0, 5 mM magnesium
acetate, 5 mM DTT, 100 μM ATP, 5 μM S679A or S679W Lon, and varying
concentrations of λN89-98 dansyl peptide or λN89-98 nonfluorescent peptide (0, 25, 100
μM). The data shown are a result of subtracting the emission spectra of the λN89-98
dansyl peptide alone in the presence of ATP from the emission spectra of the λN89-98
dansyl peptide incubated with Lon in the presence of ATP.
4.2.3. Measuring peptide binding using fluorescence anisotropy
λN89-98 dansyl peptide binding to the S679A Lon mutant was detected by
changes in fluorescence anisotropy (excitation 340 nm emission 520 nm) at 37 °C on a
54 Fluoromax-3 spectrofluorimeter (Horiba Group). Each binding reaction contained 50
mM HEPES pH 8.0, 5 mM magnesium acetate, 5 mM DTT, and 20 μM λN89-98 dansyl
peptide in the presence and absence of 1 mM AMPPNP in a quartz microcuvette
(Hellma) with a 3 mm path length. S679A, S679W or wild-type Lon was titrated into the
cuvette (0-192 μM) and the reaction was incubated at 37 °C for 3 minutes to allow
equilibrium to be reached. Higher Lon concentrations were not used due to problems with
solubility of the protein. The anisotropy measurements were plotted as a function of Lon
concentration and the data were fit with equation 4.1
n n B = Bmax [S] / (Kd’ + [S] ) + C (4.1)
where B is the observed anisotropy, Bmax is the maximum anisotropy, S is the concentration of Lon added, n is the Hill coefficient, C is the endpoint, the equilibrium binding constant Kd is calculated from the relationship log Kd’ = n log Kd.
4.2.4 Monitoring Lon-peptide interactions using pre-steady-state kinetic techniques
Experiments to monitor interactions between peptide substrate with Lon were performed on a KinTek Stopped Flow controlled by the data collection software Stop
Flow version 7.50 β with a 0.5 cm path length. The sample syringes were maintained at
37 ºC by a circulating water bath. Syringe A contained 5 μM S679A or S679W Lon monomer with variable concentrations of λN89-98 dansyl peptide (10-500 μM), 50 mM
HEPES pH 8.0, 75 mM potassium acetate, 5 mM magnesium acetate and 5 mM DTT.
Syringe B contained 100 μM ATP, 50 mM HEPES pH 8.0, 75 mM potassium acetate and
5 mM DTT. λN89-98 dansyl peptide binding to S679A or S679W was monitored by an increase in fluorescence (excitation 290 nm emission 450 nm long-pass filter) upon rapid
55 mixing of the syringe contents. In addition to monitoring with excitation 290 nm and
emission with a 450 nm longpass filter, experiments were performed with excitation 290
nm emission with a 340 nm band-pass filter to monitor changes in Trp fluorescence. To
define the pre-steady-state period of the reaction, each time course was fit over a period
equal to four half-lives of the pre-steady-state rate constant (klag) obtained from the ATP
dependent peptidase reaction. The lag rate constant klag was previously determined for
each reaction condition used (61). The data shown are a result of averaging at least four
traces. Each reaction condition was performed in triplicate (a total of > 12 traces obtained for each data point). The averaged time courses of S679A were fit with equation 4.2 describing a single exponential
(-k t) F = A exp S679A + C (4.2)
where F is relative fluorescence, A is amplitude in relative fluorescence units, t is time in
seconds, C is the endpoint, kS679A is the first order rate constant associated with peptide binding to S679A Lon in per seconds. Note that the PMT (photomultiplier tube) sensitivity was automatically adjusted by the instrument to optimize signal to noise during acquisition of the time course data for the various concentrations of peptide used in the reactions. A higher sensitivity setting was used for the lower [λN89-98 dansyl] and a lower sensitivity setting was used for higher [λN89-98 dansyl]. As a result, the
relative amplitudes of the time courses do not reflect the stoichiometries of the enzyme
intermediates monitored by the signals. At high concentrations of λN89-98 dansyl (>
100 μM), the high absorbance of the dansyl moiety also obscures the amplitude of the
fluorescence signal detected; but the first order rate constants of the reactions do not
change because the dansyl absorbance in each reaction remains constant.
56 The time course of S679W with AMPPNP without peptide was fit with equation
4.3 describing a single exponential equation followed by a steady state rate
(-k t) F = A1exp 1,S679W + νsst + C (4.3)
where F is relative fluorescence, A1 is amplitude in relative fluorescence units, t is time in
seconds, k1,S679W is the first order rate constant associated with the first phase of the reaction in per seconds, νss is the steady state rate in fluorescence units per second, and C
is the endpoint.
The averaged time courses of S679W with ATP and λN89-98 dansyl peptide were
fit with equation 4.4 describing a double exponential
(-k t) (-k t) F = A1 exp 1,S679W + A2 exp 2,S679W + C (4.4)
where F is relative fluorescence, A1 and A2 are amplitudes in relative fluorescence units, t
is time in seconds, C is the endpoint, k1,S679W is the first order rate constant associated
with the first phase of the reaction in per seconds, and k2,S679W is the first order rate
constant associated with the second phase of the reaction in per seconds. Time courses of
S679W with saturating concentrations of ATP and low concentrations of λN89-98 dansyl peptide (25 – 125 μM) were fit with equation 4.5 describing a triple exponential
(-k t) (-k t) (-k t) F = A1 exp 1,S679W + A2* exp 2*,S679W + A2 exp 2,S679W + C (4.5)
where F is relative fluorescence, A1, A2* and A2 are amplitudes in relative fluorescence
units, t is time in seconds, C is the endpoint, k1,S679W is the first order rate constant
associated with the first phase of the reaction in per seconds, k2,S679W is the first order rate
constant associated with the second phase of the reaction in per seconds and k2*,S679W is
the first order rate constant for a phase in the reaction that is only visible under conditions
of low peptide concentrations.
57 The first order rate constants were plotted as a function of substrate concentration
(ATP or peptide). Hyperbolic plots were fit using the data fitting program Kaleidagraph
by Synergy with equation 4.6 describing a 2-step binding event
k = kmax [S] / (Kd + [S]) + krev (4.6)
where k is the observed rate constant from equation 4.2, 4.4 or 4.5 in per seconds, kmax is
the maximum forward rate constant for the second step in the binding event in per
seconds, krev is the reverse rate constant for the second step in the binding event in per
seconds, S is the concentration of ATP or λN89-98 dansyl peptide, and Kd is the
equilibrium binding constant for the substrate in micromolar units. Sigmoidal plots were fit using the data fitting program Enzfitter by Biosoft with equation 4.7
n n k = kmax [S] / (Kd’ + [S] ) + krev (4.7)
where k is the observed rate constant from eq 4.2, 4.4 or 4.5 in per seconds, kmax is the
maximum forward rate constant for the second step in the binding event in per seconds,
krev is the reverse rate constant for the second step in the binding event in per seconds, S
is the concentration of ATP or λN89-98 dansyl peptide, n is the Hill coefficient, the
equilibrium binding constant Kd is calculated from the relationship log Kd’ = n log Kd.
4.2.5 ADP inhibition of the Lon-peptide interaction
ADP inhibition was measured on a KinTek Stopped Flow using the data collection software Stop Flow 7.50β. Syringe A contained 5 μM S679W Lon, 500 μM
λN89-98 dansyl and varying concentrations of ADP in 50 mM HEPES pH 8.0, 75 mM potassium acetate, 5 mM magnesium acetate, 5 mM DTT. Syringe B contained 50 mM
HEPES pH 8.0, 75 mM potassium acetate, 5 mM DTT and 500 μM ATP. Upon rapid
58 mixing of the syringe contents, peptide delivery was monitored by an increase in
fluorescence (excitation 290 nm, emission 450 nm long pass filter). The data shown are
an average of at least 4 traces. The time courses were fit with equation 4.4. The rate
constants, k1,S679W and k2,S679W were plotted versus [ADP] and fit with equation 4.8
k/kmax = 1 / (1+( [ADP] / IC50)) (4.8)
where k is the observed rate constant from eq 4.4 (k1,S679W or k2,S679W), k,max is the observed rate constant with no ADP in per seconds and IC50 is the concentration of ADP
required to inhibit the reaction rate by 50% in micromolar units.
59 4.3 RESULTS & DISCUSSION
4.3.1 Developing an assay to monitor Lon-peptide interactions
In this study, I generated three proteolytically inactive mutants of E. coli Lon for studying the microscopic events occurring along the enzymatic pathway before peptide cleavage. I chose to work with proteolytically inactive mutants to simplify the system and decouple proteolysis from substrate binding. In the S679A and S679W mutants the catalytic Ser679 residue, which is responsible for attacking the scissile peptide bond in a substrate, is replaced with an alanine and tryptophan residue, respectively. In the third mutant, all intrinsic tryptophan residues were mutated to phenylalanine residues and the active site serine was replaced with a tryptophan (S679W, W297F, W303F, W603F Lon).
With only a single tryptophan residue in the enzyme located in the protease active site, any FRET signal can be more easily assigned. This mutant should allow direct detection of a peptide translocation event.
Previously, Starkova et al. showed that the ATPase activity and the proteolytic activity of Lon can be decoupled (70). All three Lon mutants hydrolyze ATP with an intrinsic ATPase activity that is stimulated in the presence of peptide like the wild type enzyme (Figure 4.1, Table 4.2). The ~2-5 fold increase in ATPase activity is comparable to those typically observed in the presence of protein substrates and the λN89-98 peptide
(43, 46). As expected, the λN89-98 F peptide is not cleaved by S679A or S679W in the presence of ATP (Figure 4.2). Figure 4.3 and Table 4.2 shows that the fluorescent ATP analog MANT-ATP binds to S679A and S679W the same as the wild-type enzyme indicating that the mutation in the active site of the protease domain does not affect ATP binding (74).
60 Figure 4.1. ATPase activity of Lon mutants. The intrinsic (●) and peptide stimulated (■) ATPase activity of wildtype (red), S679A (blue), S679W (green) and S679W, W297F, W303F, W603F (black) Lon was measured using an [α32P]ATP assay with 150 nM Lon, in the presence and absence of λN89-98 peptide with varying concentrations of ATP (25-1000 μM). The observed rate constants (kobs) at various concentrations of ATP are shown plotted as a function of ATP. The kinetic parameters obtained as a result of fitting the data to the Michaelis Menten equation are reported in Table 4.2.
1.5
1 -1 , s obs k
0.5
0 0 200 400 600 800 1000
[ATP] (μM)
61 Table 4.2 Kinetic parameters for ATP binding and hydrolysis by wildtype and Lon mutants. a values were taken from (74) ND: not determined
ATPase activity MANT-ATP Binding
-1 5 -1 -1 -1 k (s-1) kcat (s ) Km (μM) kon (10 M s )koff (s ) on,2
a a a WT Lon 0.26 + 0.01 46 + 6 6.8 11 4.1
a a 10 a WT Lon + λN89-98 1.43 + 0.05 82 + 10 6.8 3.7
S679A 0.31 + 0.03 147 + 41 2.95 + 0.70 8.50 + 1.14 3.85 + 0.96
S679A + λN89-98 1.57 + 0.08 131+ 21 ND ND ND
S679W 0.64 + 0.03 139 + 24 4.26 + 1.98 8.59 + 1.51 3.96 + 1.80
S679W + λN89-98 1.07 + 0.06 79 + 16 ND ND ND
S679W, W297F, W303F, W603F 0.58 + 0.07 333 + 34 ND ND ND
S679W, W297F, W303F, W603F + λN89-98 1.06 + 0.11 123 + 37 ND ND ND
Figure 4.2. Peptidase activity of Lon mutants. The peptidase activity of wildtype (red), S679A (blue) and S679W (green) Lon in the presence (●) and absence (■) of ATP was monitored using the λN89-98 peptide and the continuous fluorescence based assay. 125 nM Lon was incubated with 500 μM λN89-98 peptide and the reaction was initiated by the addition of 500 μM ATP. No peptidase activity is detected without ATP or in the S679A or S679W mutants.
250
200 M) μ 150
100
peptide hydrolyzed ( hydrolyzed peptide 50
0
0 50 100 150 200 250 300 350 time (s)
62 Figure 4.3. MANT-ATP binding to Lon. MANT-ATP is a fluorescently labeled ATP analog that allows for the detection of ATP binding to Lon. Five micromolar S679A (red) or S679W (blue) was rapidly mixed with varying concentrations (1-100 μM) of MANT-ATP using a stopped flow apparatus. An increase in fluorescence is observed (excitation 360 nm, emission 400 nm) upon MANT- ATP binding to Lon. The timecourses were fit with a double exponential equation and the resulting rate constants, k1 (●) and k2 (■) were plotted as a function of MANT-ATP concentration. The slope of the line yields the on rate for ATP (kon) and the y-intercept approximates the off rate for ATP (koff). The second rate constant (k2) is independent of ATP and likely represents an ATP binding dependent conformational change. All rate constants are reported in Table 4.2.
35
30
25 ) -1 20 (s 2 or k or
1 15 k
10
5
0 02 10-5 4 10-5 6 10-5 8 10-5 0.0001 0.00012
[Mant-ATP] (M)
Using limited tryptic digestion analyses, we previously demonstrated that wild
type E. coli Lon binds to adenine nucleotides to form a more compact conformation that
is resistant to digestion by trypsin (46). To evaluate whether Ser679 is needed for forming such a conformational change, I subjected S679A and S679W to limited tryptic digestion. As shown in Figure 4.4, both mutants exhibit the same susceptibility toward
limited tryptic digestion as the wild type enzyme, indicating that the formation of the
“adenine-induced” conformational change is independent of the active site serine residue.
63 The S679W, W297F, W303F, W603F Lon mutant was also subjected to limited tryptic digestion and behaves very differently than wild type Lon (Figure 4.4). For this reason, I did not use this mutant to examine the steps leading up to peptide hydrolysis. Only the
S679A and S679W Lon mutants were used. One or more of the native tryptophan residues in Lon are required for the adenine induced conformational change. However, this mutant exhibits wild type-like ATPase activity. Therefore, I conclude that the adenine induced conformational change is not required for ATPase activity.
Figure 4.4. Limited tryptic digestion analysis of S679A, S679W and S679W, W297F, W303F, W603F Lon mutants. Wildtype, S679A, S679W and S679W, W297F, W303F, W603F Lon was incubated with λN89-98 in the presence and absence of ATP and digested with trypsin. The reactions were quenched at various time points with SBTI (soybean trypsin inhibitor) and analyzed by SDS-PAGE. The adenine induced conformational change is monitored by the stability of the 67 kDa protein band corresponding to the ATPase, SSD and protease domains. The 67 kDa band is stable in wildtype, S679A and S679W but not in S679W, W297F, W303F, W603F.
WT S679A S679W 1mM no 1mM no 1mM no kDa ATP ATP ATP ATP ATP ATP 89 67
45 35 26 23
SBTI time 0 30 0000300 30 30 30 30 (min)
WT S679W, W297F, W303F, W603F no 1 mM no 1 mM kDa ATP ATP ATP ATP 89 67
45
35 26 23
SBTI
time 0 10 30 0 10 30 0 10 30 0 10 30 (min)
64 Although E. coli Lon is a homo-hexamer containing three intrinsic Trp residues in each subunit, I could not detect any changes in the fluorescence emission scan of these
Trp residues in S679A Lon incubated with 100 μM ATP and the λN89-98 NF (non- fluorescent) peptide (Figure 4.5) when exciting the sample at 290 nm. However, incubating S679A with 100 μM ATP at increasing concentrations of the λN89-98 dansyl peptide causes a reduction in the fluorescence of Trp with a concomitant increase in the fluorescence of the dansyl moiety when excited at 290 nm (Figure 4.6). This result indicates that the interaction between S679A and the dansyl peptide in the presence of nucleotide can be quantitatively characterized by monitoring the increase in dansyl fluorescence using stopped flow spectroscopy.
Figure 4.5. Emission scan of S679A Lon with the λN89-98 NF peptide. Five micromolar S679A Lon was incubated with 100 μM ATP and 0 (red), 25 (blue) or 100 (green) μM λN89-98 NF (nonfluorescent) peptide. The sample was excited and 290 nm and the fluorescence emission spectrum was monitored. No changes in tryptophan fluorescence are detected.
1
0.8
0.6
0.4 relative fluorescence relative
0.2
0 300 350 400 450 500 550
wavelength (nm)
65 Figure 4.6. Emission scan of S679A Lon with λN89-98 dansyl peptide. Five micromolar S679A Lon was incubated with 100 μM ATP and 0 (red), 25 (blue) or 100 (green) μM λN89-98 dansyl peptide. The sample was excited and 290 nm and the fluorescence emission spectrum was monitored. A decrease in tryptophan fluorescence is detected at 350 nm with a concomitant increase in dansyl fluorescence at 520 nm. 1
0.8
0.6
0.4 relative fluorescence relative
0.2
0 300 350 400 450 500 550
wavelength (nm)
Figure 4.7 shows the stopped-flow time courses for the increase in dansyl
fluorescence generated from pre-incubation of 5 μM S679A with 100 μM λN89-98 dansyl peptide and rapidly mixing with 100 μM ATP. The sample was excited at 290 nm and the fluorescence signal was detected using a cutoff filter at 450 nm, which monitored the fluorescence signal generated from the λN89-98 dansyl peptide. Identical time
courses were obtained when 5 μM S679A was rapidly mixed with 100 μM λN89-98
dansyl peptide pre-mixed with 100 μM ATP. According to Figure 4.7, negligible changes in the dansyl fluorescence were detected when ATP was omitted. Figure 4.8 reports the changes in Trp fluorescence in the same reactions described for obtaining the data shown
66 in Figure 4.7. The increase in dansyl fluorescence (Figure 4.7) mirrors the decrease in
Trp fluorescence (Figure 4.8), which was detected using a 340 nm bandpass filter. Taken together, the increase in the dansylated peptide fluorescence time course (Figure 4.7)
along with the concomitant decrease in the Trp fluorescence time course (Figure 4.8)
suggests that peptide binding to Lon can be monitored by fluorescence resonance energy
transfer (FRET) between the intrinsic Trp residues in S679A and the dansyl moiety in the
peptide. The increase in dansyl fluorescence could be a result of conformational changes
in the enzyme in which the Trp residues are more accessible to interact with the dansyl moiety. Alternatively, the signal can be reflecting the dansyl peptide approaching one or
more of the Trp residues in Lon or a combination of both phenomena. In all cases, the
signal detected in the FRET time courses is interpreted as a result of the λN89-98 dansyl
peptide interacting with Lon and is used as a reporter for monitoring the kinetics of the
binding reactions.
67 Figure 4.7. Peptide binding to S679A can be monitored using the λN89-98 dansyl peptide and monitoring dansyl fluorescence. Five micromolar S679A was incubated with 100 μM λN89-98 dansyl peptide and rapidly mixed with 100 μM ATP. The reaction was excited at 290 nm and monitored using a 450 nm longpass filter to measure dansyl fluorescence (red). No changes in fluorescence were observed in the absence of ATP (blue).
0.16
0.14
0.12
0.1
0.08
0.06 relative fluorescence 0.04
0.02
0 012345 time (s)
Figure 4.8. Peptide binding to S679A can be monitored using the λN89-98 dansyl peptide and monitoring tryptophan fluorescence. Five micromolar S679A was incubated with 100 μM λN89-98 dansyl peptide and rapidly mixed with 100 μM ATP. The reaction was excited at 290 nm and monitored using a 340 nm bandpass filter to measure tryptophan fluorescence (red). No changes in fluorescence were observed in the absence of ATP (blue). 0.1
0
-0.1
-0.2
-0.3 relative fluorescence
-0.4
-0.5 012345 time (s)
68 4.3.2 Monitoring the Lon-peptide interaction with S679A using pre-steady-state kinetic
techniques
To quantitatively evaluate the binding interaction between S679A and the λN89-
98 dansyl peptide, I measured the stopped flow time courses of dansyl fluorescence emission (450 nm long pass filter) by exciting Trp at 290 nm at various dansyl peptide concentrations (10 to 500 μM). The time courses were best fit with equation 4.2 to yield the rate constants of the reactions (kS679A). Plotting the observed rate constant at the
corresponding peptide concentration yields Figure 4.9, which is indicative of a two step
peptide binding mechanism (Figure 4.10). Because the fluorescence signal is generated
from the S679A:dansyl peptide complex, the first step, whose rate constant depends on
peptide concentration, represents the formation of an enzyme-peptide complex. The
second step is therefore an isomerization or conformational change occurring within the
enzyme:peptide complex after the initial binding event (59). As illustrated in Figure 4.9,
at low peptide concentration, the observed rate constants are affected by the initial
binding. Increasing peptide concentration saturates the peptide binding site in S679A
such that at > Kd peptide concentration, the observed rate constant is dominated by the
forward and reverse rate constants of the second step (Figure 4.10). Fitting the data with
equation 4.7 describing a sigmoidal curve yields the forward and reverse rate constants
-1 -1 (0.74 ± 0.10 s and 0.19 ± 0.01 s , respectively (Table 4.3)). The Kd for peptide and the
Hill coefficient (n) are also determined from this fit (Table 4.3) and are very similar to
the Km and n values determined previously for the λN89-98 peptide (40).
69 Figure 4.9. λN89-98 dansyl peptide binding to S679A is dependent on peptide. Five micromolar S679A and varying concentrations of λN89-98 dansyl peptide (10, 15, 25, 35, 50, 75, 100, 150, 250, 300, 450 or 500 μM) was rapidly mixed with 100 μM ATP. The time courses were fit with equation 4.2 and the resulting rate constants are plotted as a function of the corresponding peptide concentration. The data were fit with equation -1 4.7 to yield the kinetic parameters kS679A = 0.74 ± 0.10 s , Kd = 164 ± 35 μM, krev = 0.19 ± 0.01 s-1, n = 1.3 ± 0.2 (Table 4.3).
0.8
0.6 ) -1 (s
S679A 0.4 k
0.2
0 0 100 200 300 400 500 600
[peptide] (μM)
Figure 4.10. Scheme for peptide binding to Lon. Step 1 is an initial peptide binding event independent of nucleotide. This step is measured using equilibrium fluorescence anisotropy. Step 2 is a conformational change following initial binding that is dependent on nucleotide. This step is measured using stopped-flow kinetic techniques. Note that ATP binds to Lon independent of peptide. The Lon*:peptide complex contains ATP.
12
kon kS679A Lon + peptide Lon:peptide Lon*:peptide k k off S679A, rev
70 The stopped flow time courses reflecting the FRET signal between the intrinsic
Trp in 5 μM S679A and 500 μM of the λN89-98 dansyl peptide (5 x Km in wild type
Lon) is also dependent on the concentration of ATP. The time courses were best fit with equation 4.2 to yield the rate constants for the reactions (kS679A). Plotting the observed rate constants at the corresponding concentration of ATP (1 μM to 50 μM) yields a hyperbolic plot, which upon fitting with equation 4.6, yields a Kd of 7.4 ± 2.5 μM for
-1 ATP, a forward rate constant (kS679A,max) of 0.54 ± 0.04 s and a reverse rate constant
-1 (krev,S679A) of 0.19 ± 0.03 s , respectively (Figure 4.11, Table 4.3).
Figure 4.11. λN89-98 dansyl peptide binding to S679A is dependent on ATP. Five micromolar S679A and 500 μM λN89-98 dansyl peptide was rapidly mixed with varying concentrations of ATP (0.5, 1, 3, 5, 10, 25 or 50 μM). The time courses were fit with equation 4.2 and the resulting rate constants are plotted as a function of the corresponding ATP concentration. The data were fit with equation 4.6 to yield a -1 -1 maximum kS679A = 0.54 ± 0.04 s , Kd = 7.4 ± 2.5 μM, krev = 0.19 ± 0.03 s (Table 4.3).
0.8
0.7
0.6
0.5 ) -1
(s 0.4 S679A k 0.3
0.2
0.1
0 0 102030405060
[ATP] (μM)
71 Table 4.3 Kinetic constants for λN89-98 dansyl peptide interacting with S679A and S679W NA: not applicable ND: not determined
S679A S679W k1 k2 Vary peptide Vary ATP Vary peptide Vary ATP Vary peptide Vary ATP
kmax 0.74 +/- 0.10 0.54 +/- 0.04 7.6 +/- 1.0 5.3 +/- 0.6 0.57 +/- 0.10 1.5 +/- 0.4 (s-1)
Kd 164 +/- 35 7.4 +/- 2.5 NA 4.3 +/- 1.9 157 +/- 8 9.3 +/- 9.8 (μM)
krev 0.19 +/- 0.01 0.19 +/- 0.03 NA 2.1 +/- 0.5 0.10 +/- 0.01 ND (s-1)
n 1.3 +/- 0.2 NA NA NA 1.9+/- 0.1 NA
4.3.3 Measuring peptide binding to Lon using fluorescence anisotropy
We used fluorescence anisotropy techniques as another method for measuring
λN89-98 dansyl peptide binding to Lon. Figure 4.12 shows the changes in anisotropy signal from adding S679A (0 – 192 μM) to a cuvette containing 20 μM λN89-98 dansyl peptide and 100 μM AMPPNP. The binding isotherm was fit with equation 4.1 describing a sigmoidal curve to yield Kd = 35.2 ± 18.6 μM and n = 1.5 ± 0.1. Similar results were obtained with no nucleotide as well as with wild-type Lon (Kd = 49.2 ± 28.1, n = 1.5 ± 0.1 and Kd = 38.3 ± 6.0, n = 1.5 ± 0.1, respectively). These results indicate that neither the binding constant nor the Hill coefficient are affected by nucleotide and supports the previous result that Lon has a random ordered Bi Bi kinetic mechanism (40,
53, 60) .
72 Figure 4.12. Equilibrium λN89-98 dansyl binding to Lon can be monitored using fluorescence anisotropy. Twenty micromolar λN89-98 dansyl peptide was incubated with 1 mM AMPPNP and changes in anisotropy were measured (excitation 340 nm emission 520 nm) by titrating in S679A Lon (0 μM - 192 μM). The data were fit with equation 4.1 resulting in Kd = 35.2 ± 18.6 μM with a Hill coefficient (n) of 1.5 ± 0.1. Similar results were obtained in S679A and in the wild-type enzyme when nucleotide was omitted.
0.16
0.12
0.08 anisotropy
0.04
0 050100150200
[S679A Lon] (μM)
4.3.4 Examining events prior to peptide hydrolysis using S679W and pre-steady-state kinetic techniques.
Previously our lab demonstrated that the peptide boronate Z-Leu-Leu-Leu-
B(OH)2 (MG262) inhibits Salmonella Lon, which shares 99% sequence identity with the
E. coli lon homolog, only in the presence of ATP or AMPPNP (28, 29). In MG262, the boronate moiety is anticipated to function as an electrophilic center that sequesters the nucleophilic side chain of Ser679 in the proteolytic site of Lon. The observed dependency in inhibition toward ATP or AMPPNP suggests that nucleotide binding to
73 Lon at the ATPase site allosterically activates the proteolytic site such that Ser679 reacts with MG262 (29). To evaluate the existence of an allosteric communication between the
ATPase site and the protease site in Lon, I generated the S679W mutant. As Trp fluorescence is sensitive to changes in its local environment, the replacement of Ser679 with a Trp residue may allow the detection of a conformational change in the proteolytic site in Lon resulting from ATP or AMPPNP binding. Figure 4.13 shows the stopped flow time course monitoring the changes in Trp fluorescence of S679W (excitation 290 nm, emission 340 bandpass filter) in the absence and presence of ATP or AMPPNP or
ADP with no peptide. A decrease in the fluorescence of the Trp residues in S679W
(three intrinsic and a fourth one at 679W) is detected in the presence of ATP or AMPPNP but not ADP, suggesting the observed changes in Trp fluorescence are associated with
Lon interacting with the gamma phosphate moiety in ATP and AMPPNP. The ATP time
-1 course was fit with equation 4.2 describing a single exponential, k = 9.35 ± 0.34 s . The
AMPPNP time course was best fit with a single exponential followed by a steady state to yield an observed rate constant of 0.94 ± 0.02 s-1 and a steady-state rate of 0.02 fluorescence units per second.
74 Figure 4.13. Intrinsic tryptophan fluorescence can be used to measure a conformational change in S679W dependent on ATP and AMPPNP. Five micromolar S679W was rapidly mixed with buffer (red), 100 μM ATP (blue), 100 μM AMPPNP (green) or 100 μM ADP (orange) in a stopped-flow instrument. The reactions were excited at 290 nm and emission was detected using a 340 nm bandpass filter to detect Trp fluorescence. No significant changes are observed with buffer or ADP. The experimental time course is shown in color and the fitted curve is shown in black.
0.1
0.05
0
-0.05
-0.1
-0.15 relative fluorescence -0.2
-0.25
-0.3 012345
time (s)
Experiments similar to those performed with the S679A Lon mutant were used to characterize λN89-98 dansyl peptide binding to S679W. S679W was preincubated with
500 μM ( ~ 3 x Kd) peptide and rapidly mixed with varying concentrations of ATP (0.5 –
50 μM). As shown in Figure 4.14, at concentrations of ATP < 5 μM, there is an initial increase in fluorescence followed by a slower decrease while at concentrations > 5 μM there is only an increase in fluorescence. These results can be explained when the kinetic results of ATPase activity under similar conditions are considered. At concentrations of
ATP less than 5 μM, our lab has shown that ATP hydrolysis activity is slow and nearly negligible and only 50% of the enzyme is saturated at 10 μM ATP (52, 61, 74). The
75 change in the direction of the fluorescence signal (decrease and then increase) at increasing concentrations of ATP could be attributed to different conformational states of the proteolytic site due to different states of ATP occupancy and/or ATP hydrolysis activity. This data emphasizes the importance of ATP binding and hydrolysis for optimal proteolytic activity.
Figure 4.14. Representative time courses for S679W interacting with the λN89-98 dansyl peptide at varying concentrations of ATP. Five micromolar S679W pre-incubated with 500 μM λN89-98 dansyl peptide was rapidly mixed with 0.5 (red), 1 (blue), 5 (green), and 10 (orange) μM ATP in a stopped flow apparatus. All the time courses were best fit with equation 4.4 describing a double exponential. The experimental time courses are shown in color and the fitted curve is shown in black. At < 5 μM ATP, the second phase of the time courses display negative changes in fluorescence whereas in the time courses measured at > 5 μM ATP, the overall changes in fluorescence were positive.
0.15
0.1
0.05 relative fluorescence relative
0
0 5 10 15 20 time (s)
76 I also examined λN89-98 dansyl peptide binding to S679W with saturating ATP and varying concentrations of peptide (10 – 500 μM). As shown in the representative time courses in Figure 4.15, at sub Kd levels of peptide (< 100 μM) time courses are triphasic. There is an initial increase in fluorescence followed by a transient decrease then another increase. At the low concentrations of peptide, the time courses fit best to equation 4.5 describing a triple exponential equation. The first increasing phase is assigned k1,S679W, the second decreasing phase is assigned k2*,S679W and the third increasing phase is assigned k2,S679W. The k2*,S679W transient phase is independent of
-1 peptide concentration (k2*S679W = 0.95 ± 0.30 s ). The time courses for > 100 μM peptide fit best to a double exponential equation and do not show the transient phase. I propose that the transient phase “disappears” at the higher peptide concentrations because the rate constant k2,S679W approaches the rate constant for k2*S679W. At this point it is not clear what mechanistic event is described by the transient phase (k2*S679W). I do know that under these reaction conditions, ATP is hydrolyzed by Lon generating ADP with a pre- steady-state burst rate of ~ 10 s-1 and a steady-state rate constant of ~ 0.8 s-1. It is possible the transient phase represents an ADP/ATP exchange event (rate constant for
ADP release ~ 0.5 s-1 (74)) but further experimentation is required to fully define this phase.
77 Figure 4.15. Representative time courses for S679W interacting with the λN89-98 dansyl peptide at varying concentrations of λN89-98 dansyl peptide. Five micromolar S679W pre-incubated with 25 (red) 50 (blue) 100 (green) and 500 (orange) μM λN89-98 dansyl peptide were rapidly mixed with 100 μM ATP in a stopped-flow apparatus. The experimental time courses are shown in color and the fitted curve is shown in black. The sub Kd levels of λN89-98 dansyl peptide (25, and 50 μM) were fit with equation 4.5 (triple exponential). The 100 and 500 μM λN89-98 dansyl peptide time courses were fit with equation 4.4 (double exponential equation).
0.35
0.3
0.25
0.2
0.15
relative fluorescence relative 0.1
0.05
0 0246810121416
time (s)
All of the S679W time courses were fit to the appropriate equation (equation 4.4 for a double exponential or equation 4.5 for a triple exponential) and the rate constants were plotted as a function of the respective substrate concentration. As shown in Figures
4.16 and 4.17, k1,S679W is dependent on ATP but independent of peptide. The apparent dependence on peptide shown in Figure 4.17 is probably due to uncertainty in determining k1,S679W from the double exponential equation because I was able to measure k1,S679W without peptide by monitoring the decrease in tryptophan fluorescence (k = 9.35
78 + 0.34 s-1, Figure 4.13). Figure 4.16 was fit with equation 4.6 describing a hyperbolic function to yield the kinetic parameters summarized in Table 4.3.
Figure 4.16. The first phase of the S679W reaction is dependent on ATP. Five micromolar S679W Lon was incubated with 500 μM λN89-98 dansyl peptide and rapidly mixed with varying concentrations of ATP (0.5, 1, 2, 4, 7, 10, 25 or 50 μM). The time courses were fit with equation 4.4 and the resulting rate constants (k1,S679W) are shown as a function of ATP concentration. The data were fit with equation 4.6 to yield -1 the kinetic parameters summarized in Table 4.3. k1,S679W = 5.3 ± 0.6 s is dependent on -1 ATP concentration, Kd = 4.3 ± 1.9 μM, krev = 2.1 ± 0.5 s
8
6 ) -1 (s
4 1,S679W k
2
0 0 102030405060
[ATP] (μM)
79 Figure 4.17. The first phase of the S679W reaction is not dependent on peptide. Five micromolar S679W Lon was incubated with varying concentrations of λN89-98 dansyl peptide (25, 50, 75, 100, 125, 200, 350 or 500 μM) and rapidly mixed with 100 μM ATP. The time courses were fit with equation 4.4 or 4.5 and the resulting rate constants (k1,S679W) are shown as a function of peptide concentration. k1,S679W = 7.6 ± 1.0 s-1 and is independent of peptide. The 0 μM λN89-98 dansyl peptide data point (■) was obtained by fitting the time course shown in Figure 4.13 (S679W + ATP) to equation 4.2 -1 to yield k1,S679W = 9.35 ± 0.34 s .
10
8 )
-1 6 (s 1, S679W 1,
k 4
2
0 0 100 200 300 400 500 600
[peptide] (μM)
The final phase described by k2,S679W is dependent on peptide and ATP as shown in Figures 4.18 and 4.19. Figure 4.18 was fit with equation 4.6 describing a hyperbolic function and Figure 4.19 was fit with equation 4.7 describing a sigmoidal function. All of the kinetic parameters are summarized in Table 4.3. The error in the reverse rate
- constant for k2,S679W at varying concentrations of ATP was rather high (krev = 0.08 ± 0.5 s
1) therefore I have reported it as not determined (ND). The reverse rate constant is the y- intercept of the plot. Visually the y-intercept is approaching zero, therefore I assume krev
80 is nearly negligible in comparison to k2,S679W. From Table 4.3 we can see that both the
S679A data and the second phase S679W data are dependent on ATP and peptide and the rate constants for these steps are very similar. For these reasons, I propose that the same event in the peptidase reaction is being measured in the two different mutants.
Figure 4.18. The second phase of the S679W reaction is dependent on ATP. Five micromolar S679W Lon was incubated with 500 μM λN89-98 dansyl peptide and rapidly mixed with varying concentrations of ATP (4, 7, 10, 25 or 50 μM). The time courses were fit with equation 4.4 and the resulting rate constants (k2,S679W) are shown as a function of ATP concentration. The data were fit with equation 4.6 to yield the kinetic -1 parameters summarized in Table 4.3. k2,S679W = 1.5 ± 0.4 s , Kd = 9.3 ± 9.8 μM.
1.6
1.2 ) -1 (s
0.8 2,S679W k
0.4
0 0 102030405060
[ATP] (μM)
Figure 4.19. The second phase of the S679W reaction is dependent on peptide.
81 Five micromolar S679W Lon was incubated with varying concentrations of λN89-98 dansyl peptide (25, 50, 75, 100, 125, 200, 350 or 500 μM) and rapidly mixed with 100 μM ATP. The time courses were fit with equation 4.4 or 4.5 and the resulting rate constants (k2,S679W) are shown as a function of peptide concentration. The data were fit with equation 4.7 to yield the kinetic parameters summarized in Table 4.3. k2,S679W = 0.57 -1 ± 0.10 s , Kd = 157 ± 8 μM, n = 1.9 ± 0.1.
0.8
0.7
0.6
0.5 ) -1 (s 0.4 2,S679W k 0.3
0.2
0.1
0 0 100 200 300 400 500 600
[peptide] (μM)
Previous peptidase experiments with AMPPNP have revealed that AMPPNP supports peptide cleavage at a rate that is ~ 7 fold reduced as compared to ATP. I monitored the S679W - λN89-98 dansyl peptide interaction in the presence of AMPPNP to determine how nucleotide binding alone affects the events leading up to peptide hydrolysis. Figure 4.20 shows that AMPPNP supports the peptide-Lon interaction while
ADP does not. Fitting the AMPPNP time course with equation 4.4 describing a double
-1 -1 exponential (k1,S679W = 0.50 ± 0.01 s and k2,S679W = 0.04 ± 0.01 s ) shows that both
82 phases of the reaction are ~ 10 fold slower in the presence of AMPPNP as compared to
ATP.
Figure 4.20. AMPPNP supports the peptide-Lon interaction at a reduced rate. Five micromolar S679W was preincubated with 500 μM λN89-98 dansyl peptide and rapidly mixed with 100 μM AMPPNP (red). The reaction was excited at 290 nm and monitored using a 450 nm longpass filter to measure dansyl fluorescence. The time course with AMPPNP was fit with equation 4.4. The fitted curve is shown in black. The -1 -1 k1,S679W and k2,S679W are 0.50 s and 0.04 s , respectively. No changes in fluorescence were observed with ADP (blue).
0.12
0.1
0.08
0.06
0.04
relative fluorescence relative 0.02
0
-0.02 0 5 10 15 20 25 30 35
time (s)
83 4.3.5 ADP inhibition of the Lon-peptide interaction
Previously, our lab and others have shown that ADP is a potent inhibitor of Lon activity (3, 40). As discussed in Chapter 3, ADP inhibits the lag phase in the pre-steady- state peptidase reaction (IC50 = 1.56 ± 0.06 μM). I also examined the effect of ADP on the kinetics of λN89-98 dansyl peptide interacting with S679W. In these experiments,
λN89-98 dansyl and S679W were preincubated with 0-25 μM ADP followed by the addition of saturating amounts of ATP (500 μM). Figures 4.21 and 4.22 show the rate constants (k1,S679W and k2,S679W) as a function of the concentration of ADP. Both steps are inhibited by ADP with an IC50 of 2.35 ± 0.33 μM and 3.65 ± 0.54 μM, respectively.
Figure 4.21. ADP inhibits the first phase of the S679W reaction. Five micromolar S679W was incubated with 500 μM λN89-98 dansyl peptide and varying concentrations of ADP (0-25 μM) and rapidly mixed with 500 μM ATP using a stopped flow apparatus. The time courses were fit with equation 4.4. The observed rate constants (k1,S679W) at various ADP concentrations were divided by k1,S679W with no ADP and plotted as a function of ADP concentration. The data were fit with equation 4.8 to yield IC50 = 2.35 ± 0.33 μM. 1
0.8
0.6 1,S679W max / k 0.4 1,S679W k
0.2
0 0 5 10 15 20 25 30
[ADP] (μM)
84 Figure 4.22. ADP inhibits the second phase of the S679W reaction. Five micromolar S679W was incubated with 500 μM λN89-98 dansyl peptide and varying concentrations of ADP (0-25 μM) and rapidly mixed with 500 μM ATP using a stopped flow apparatus. The time courses were fit with equation 4.4. The observed rate constants (k2,S679W) at various ADP concentrations were divided by k2,S679W with no ADP and plotted as a function of ADP concentration. The data were fit with equation 4.8 to yield IC50 = 3.65 ± 0.55 μM.
1
0.8
0.6 2,S679W max / k 0.4 2,S679W k
0.2
0 0 5 10 15 20 25 30
[ADP] (μM)
85 4.4 CONCLUSIONS
Lon is a serine protease whose activity is regulated by the binding and hydrolysis of ATP. In this study, I demonstrated that the microscopic events associated with peptide interacting with Lon before its cleavage can be monitored by the fluorescence signal generated between dansylated peptide substrate (λN89-98 dansyl) and tryptophan residues in Lon. The two proteolytically inactive Lon mutants S679A and S679W have been used for characterizing the kinetics of peptide interacting with Lon in the presence of ATP. The data presented here supports a two-step peptide binding mechanism. The first step involves the formation of a Lon:ATP:peptide complex, and the second step is a conformational change in the complex, occurring after ATP is hydrolyzed (Figure 4.10).
The kinetic progression of the second step is detected by the increase in dansyl fluorescence in the λN89-98 dansyl peptide upon excitation of the Trp residues in the
Lon mutants. The proposal of a two-step peptide binding mechanism is supported by the results presented in this study. At low concentrations of peptide, the binding of peptide to Lon in the presence of ATP limits the observed rate constant. Increasing the concentration of peptide leads to enzyme saturation such that the observed rate constants reach a finite value, which is defined by the asymptote of the plots shown in Figures 4.9 and 4.19.
Through previous kinetic analyses (described in previous chapters), I have constructed a kinetic model to account for the first turnover of the ATP-dependent peptide bond cleavage activity of E. coli Lon. The data presented in this study is consistent with the proposed mechanism. The proposed mechanism is shown in Figure
4.23. Free enzyme is represented by form I. For simplicity, the enzyme is shown as a
86 dimer with the ATPase and SSD domain in green and the protease domain in blue. The active site serine is shown in red. First, ATP and peptide bind to the enzyme in a random mechanism (step 1, form II), followed by a conformational change that is induced by nucleotide binding (step 2, form III). Previously, our lab utilized MANT-ATP and
MANT-AMPPNP to study the binding of ATP to Lon (74). This study showed that ATP and AMPPNP bind to Lon in an identical manner wherein ATP or AMPPNP binds to
Lon with a rate constant of 0.7 μM-1 s-1 (step 1, Figure 4.23) followed by a nucleotide
-1 binding induced conformational change (kNTP = 5 s , step 2, Figure 4.23). Equilibrium peptide binding experiments performed in this study indicate that peptide initially binds to Lon independent of nucleotide (Figure 4.12, Figure 4.10, step 1). This result corroborates our previous steady-state product inhibition studies that indicated peptide and ATP likely bind to Lon in a random order to form the Lon:ATP:peptide complex
(40).
87 Figure 4.23. Proposed mechanism for the first round of peptide hydrolysis by E. coli Lon protease. The enzyme is shown as a dimer instead of a hexamer for simplicity. The ATPase and SSD domains are shown in green, the protease domain is shown in blue and the active site serine is shown in red. Free enzyme is represented by form I. First, ATP and peptide bind in a random order (form II, step 1). Next, a conformational change occurs as a result of nucleotide binding (form III, step 2). An allosteric activation of the proteolytic site is accompanied by ATP hydrolysis (IV, step 3), followed by a slow peptide delivery/translocation event (formV, step 4),) peptide hydrolysis and product release (formVI, step 5). The rate constants in bold were measured by experiments described in this chapter.
ATP ATP ADP ATP ATP ATP ATP ADP ATP ADP ATP k1 k2 k3 k4 k5, k6
k -1 k-5 , k-6 k -2 k -3 k -4
(I) (II) (III) (IV) (V) (VI)
k1 = kon,ATP k-1 = koff,ATP
k2 = kNTP k-2 = reverse of NTP conformational change
k3 = k1,S679W ~ kburst,ATP k-3 = krev
k4 = kS679A ~ k2,S679W ~ klag k-4 = krev
k5 = peptide hydrolysis k-5 = reverse of peptide hydrolysis
k6 = product release k-6 = reverse of product release
Experiments performed with S679W reveal that the next step in the mechanism is defined by k1,S679W (Figure 4.23, step 3, form IV). The rate constant for this step (k1,S679W
= 7.6 ± 1.0 s-1) is comparable to the rate constant determined for the conformational
-1 change occurring immediately after nucleotide binding (kNTP = 5 s ); however experiments with AMPPNP reveal the two steps in the mechanism are distinct.
According to the data obtained with S697W and AMMPNP (Figure 4.20), k1, S679W is ~10 fold slower with AMPPNP as compared to ATP while kNTP is the same with ATP or
AMPPNP (74). Therefore, the step represented by k1,S679W is the key step that
88 distinguishes the ATP versus AMPPNP bound enzyme form in Lon. A further understanding of the molecular nature of this step should provide insights into how ATP and AMPPNP differ in their ability to activate the proteolytic activity of Lon.
As the formation of enzyme form IV (step 3) can only be measured with the
S679W Lon mutant and it is only dependent on nucleotide (completely independent of peptide), I propose this step involves an allosteric activation of the catalytic serine residue within the protease domain wherein the active site undergoes a gross conformational change. Our lab has shown that inhibition of Lon’s protease activity by the peptide boronate MG262 depends on ATP binding and presumably the boronate moiety sequesters the nucleophilic side chain of Ser679 (28, 29). This mechanism of inhibition provides further evidence for an allosteric activation of the active site in Lon. Under ideal reaction conditions when ATP is present, this step is accompanied by ATP
-1 hydrolysis (kburst,ATP = 11 s (61)), however ATP hydrolysis is not necessary as the same enzyme form is generated with AMPPNP at a reduced rate.
I propose herein that the slow phase leading up to peptide hydrolysis and responsible for the lag kinetics is the formation of enzyme form V (step 4, Figure 4.23). I am able to measure this step using both the S679A and S679W Lon mutants. Work by
Ishikawa et al. on a related ATP dependent protease, ClpAP, showed that substrates are translocated from an initial binding site into the proteolytic chamber (65). ClpAP and
Lon are both members of the AAA+ (ATPases associated with a variety of cellular activities) family of proteases and as a slow peptide translocation step has been shown in
ClpAP, we propose the slow step we observe before peptide hydrolysis in Lon (k2,S679W and kS679A; form V, Figure 4.23) is also a peptide translocation event. As such, the slow
89 phase could be a peptide delivery event whereby the peptide moves from the initial peptide binding site at the top of the “barrel” near the ATPase domain, down to the protease domain where it is degraded. In form VI (step 5), peptide is hydrolyzed followed by product release. The idea of the dansylated peptide “moving” toward the proteolytic site in Lon has been proposed based upon the discovery of a peptide translocation event occurring in ATP-dependent proteases that are related to Lon (56, 65,
66). Recently, Reid and coworkers labeled the protease active site of ClpAP, a heterosubunit ATP-dependent protease belonging to the same family as Lon, with
EDANS (5-(2-(acetamido)ethylamino)naphthalene-1-sulfonic acid) and polypeptide substrates with fluorescein. They used FRET between the two fluorophores to demonstrate translocation of the substrate into the proteolytic active site (66). Unlike the
ClpAP experiments, our system has multiple Trp residues that may interact with the dansyl moiety in the peptide. Therefore, my work cannot conclusively evaluate whether the signal is solely attributed to peptide translocation. However, because the signal is detected only when Trp and dansyl are nearby one another (Förster distance (R0) = 21 Å), perhaps the signal could be reflecting peptide translocation, which may be accompanied by a conformational change(s) within the enzyme (75).
In an attempt to prove that there is a peptide translocation event in Lon, I created a Lon mutant (S679W, W297F, W303F, W603F)in which the three native Trp residues were mutated to Phe residues and the active site serine was mutated to a Trp (as in the
S679W mutant). With this mutant, any signal resulting from exciting the Trp residue at
290 nm and detecting the dansyl emission would provide strong evidence that the peptide is approaching the active site serine. Signal should only be generated when the Trp and
90 dansyl moieties are close to one another. The mutant was expressed and purified, however limited tryptic digest analysis revealed differences between the mutant and wild- type enzyme. As shown in Figure 4.4, the mutant does not have the stable 67 kDa fragment that is typically used to characterize the adenine induced conformational change
(46). The Trp to Phe mutations seemed to greatly disrupt the structure/activity of Lon and therefore the mutant was unable to be used in the current study. Perhaps the Trp residues can be mutated to an amino acid other than Phe in the future resulting in a more wild-type-like enzyme that can be used to demonstrate the peptide translocation event.
To further test the kinetic model, I utilized the program KinFitSim (76) to simulate the pre-steady-state lag phase of the peptidase reaction (Chapter 3) as well as the interaction of S679W and the dansylated peptide. The simulated time courses were obtained by incorporating kinetic constants measured for the indicated steps in the proposed mechanism (Figure 4.23, Table 4.3). As shown in Figure 4.24, representative stopped-flow time courses of peptide hydrolysis with wild-type Lon and λN89-98 dansyl peptide interacting with S679W Lon obtained experimentally under conditions of saturating ATP and peptide (Figure 4.24, solid line) overlays well with the simulated data generated from the proposed mechanism (Figure 4.24, dashed line). The experimental data was then fit to the proposed mechanism to yield the theoretical rate constants. The experimental rate constants relating to this mechanism from previous work in our lab (74) as well as this study are shown in Table 4.4 alongside the theoretical rate constants generated in the KinFitSim program. The close correlation between the experimental data and the simulated time course as well as the similarity in the rate constants provides strong support for our proposed kinetic mechanism for Lon.
91 Figure 4.24. Simulation of the proposed peptide hydrolysis mechanism using KinFitSim. The solid color lines represent experimental time courses for peptide hydrolysis by wildtype Lon (red) and S679W interacting with λN89-98 dansyl peptide (blue). The dashed lines are simulated time courses from the simulation program KinFitSim using the mechanism shown. The numbers correspond to the enzyme forms shown in Figure 4.23. The rate constants, both simulated and experimental, are shown in Table 4.4. As shown in the figure, the raw data overlays well with the simulated time course.
k1 k2 k3 k4 k5 k6 E+A+B EAB FAB GPB HPB HPQ H+P+Q
(I) k-1 (II) k-2 (III) k-3 (IV) k-4 (V)k-5 (VI) k-6 A ATP P ADP B peptide Q peptide products E, F, G, H Lon enzyme forms
1.2 S679W with λN89-98 dansyl 1
0.8
0.6
0.4 relative fluorescence WT Lon peptide hydrolysis 0.2
0
00.20.40.60.81
time (s)
92 Table 4.4 Experimental and theoretical rate constants for the E. coli Lon mechanism. a values obtained from (74) b values obtained from (61) ND: values not yet determined experimentally WT Peptide S679W with Experimental Value Hydrolysis λN89-98 dansyl KinFitSim value KinFitSim value -1 -1 a k1 (μM s ) 0.7 0.23 kon,ATP = 0.7 -1 a k-1 (s ) 7 6.9 koff,ATP = 7 -1 a k2 (s ) 5 8.9 kNTPconf = 5 -1 k-2 (s ) 0.0002 .0002 ND -1 b k3 (s ) 5 25 k1,S679W = 7; kburst,ATP = 11 -1 k-3 (s ) 0.001 .001 krev = 2 -1 k4 (s ) 0.45 0.6 kS679A = 0.6; k2,S679W = 0.6; b klag = 1 -1 -5 k-4 (s ) 0.001 2.1 x 10 0.1 -1 k5 (s ) 15 ND -1 -5 k-5 (s ) 2.3 x 10 ND -1 k6 (s ) 59 ND -1 -1 -7 k-6 (μM s ) 9 x 10 ND Χ2 0.35 0.16
In conclusion, I have utilized stopped flow fluorescence spectroscopy to evaluate the steps leading up to peptide hydrolysis in Lon. I have provided evidence for three distinct conformational change steps along the pathway. One of these steps (step 4, form
V, Figure 4.23) is relatively slow and I propose it is a peptide translocation/delivery event. Another step (step 3, form IV, Figure 4.23) is particularly valuable in understanding the catalytic advantage offered by ATP over AMPPNP as this is the first time differences between ATP and AMPPNP have been assigned to a specific step.
93
CHAPTER 5
EXPLORING THE SUBSTRATE SPECIFICITY OF E. COLI LON PROTEASE
USING PEPTIDE SUBSTRATES AND λN PROTEIN DELETION MUTANTS
94 5.1 INTRODUCTION
Lon protease is responsible for degrading regulatory and damaged or misfolded proteins in the cell into short peptides from 5-15 amino acids in length (3, 13, 16, 17, 39,
42). Many in vivo substrates of bacterial Lon have been identified (SulA, RcsA, SoxS,
λN) by observing the phenotype of E. coli cells lacking the lon gene (12, 16, 18, 77).
Additional work has shown that the half-life of these proteins is greatly increased in lon- cells (cells lacking the gene for Lon protease) (12, 16, 17). We know these proteins are
Lon substrates; however the mechanism by which the enzyme recognizes and initiates degradation is not well understood. Knowledge of the mechanism by which Lon processes regulatory proteins will be beneficial for understanding how Lon controls regulatory pathways in the cell such as RNA transcription.
There are thousands of proteins in E. coli and other bacteria and only a small portion of these are degraded by intracellular proteases. As such, there must be a mechanism by which proteases discriminate amongst these proteins and select their substrates. Many AAA+ proteases degrade substrates containing certain recognition tags at either the N-or C- terminus of the protein (78-84). For example, the Clp protease requires an 11-residue ssrA tag. Fusion of this short tag sequence to virtually any protein results in degradation of that protein by the protease (78). At this time, no known universal substrate recognition tag for Lon exists. Studies have shown that the 8 C- terminal amino acids in the cell division inihibitor SulA protein are required for the protein to be recognized by Lon (81, 85). Mutants lacking these eight residues are not degraded and fail to bind to Lon. Similarly, Shah and Wolf have demonstrated that the
N-terminal amino acids of the transcription activator SoxS are recognized by Lon (82).
95 Currently, the accepted paradigm is that loosely folded or unstructured regions on a protein interact or bind to the SSD (substrate sensor and discriminator) domain of Lon followed by unfolding and translocation into the proteolytic chamber where the protein substrate is degraded (37, 55, 56, 66, 86, 87).
Not only is substrate recognition important, but we are also interested in understanding what determines where Lon cleaves its substrates. Unlike certain proteases such as trypsin, which always cleave after certain amino acid residues, it is not possible at this time to predict Lon cleavage sites in a protein. Analysis of Lon cleavage profiles suggests that Lon prefers to cleave after hydrophobic or aliphatic amino acids, but not after every hydrophobic or aliphatic amino acid (42, 81). For example, in the λN protein, Lon hydrolyzes the bond between Ala16 – Gln17 but not Ala3 – Gln4 (Figure
5.1) (42). The cleavage specificity of Lon is probably directed by many factors, including residues at the cleavage site and residues distal from the cleavage site.
Knowledge of the factors determining substrate specificity could prove important for developing peptide based therapeutic molecules that regulate Lon activity.
96 Figure 5.1. λN protein sequence and cleavage profile. The λN protein has seven major Lon cleavage sites (red arrows). The amino acids that make up the λN89-98 peptide are highlighted in orange.
1M D A Q T R R R E 11R R A Q K Q A Q W K A 21A N P L L V G V S A 31K P V N R P
I L S L 41N R K P K S R V E S 51A L N P I D L T V L A E Y H K Q I E S N 71L Q R I
E R K N Q R 81T W Y S K P G E R G 91I T C S G R Q K I K 101G K S I P L I
This chapter describes my work to examine the substrate specificity of Lon protease as well as the mechanism by which Lon degrades the unfolded λN protein.
Using model peptides and our continuous fluorescence based peptidase assay, I have identified several amino acids close to the cleavage site that affect substrate recognition.
I have also generated several deletion mutants of the λN protein, which provide insight into how Lon initiates the degradation of a full length unstructured protein.
97 5.2 MATERIALS AND METHODS
5.2.1 Peptide design and peptidase assays
All peptides were synthesized using standard Fmoc solid-phase synthesis protocols as described previously(88-90). Fluorescent peptides corresponding to various regions of the λN protein are shown in Figure 5.2. With the exception of λN11-21 all peptides contain the Y(NO2) - Abz FRET pair used in the λN89-98 peptide previously designed for the kinetic characterization of Lon. Nonfluorescent analogs of λN25-35,
λN35-45, λN79-89, and λN89-103, in which a tyrosine residue is used in place of the
Y(NO2) and Bz is used in place of the Abz, were also synthesized (similar to the λN89-
98 peptide (40)). With the exception of λN11-21, peptidase activity was detected by an increase in fluorescence upon excitation at 320 nm and emission at 420 nm. Unlike the other peptides, the λN11-21 peptide consists of an Abz – tryptophan FRET pair (75).
Peptide cleavage is detected by a decrease in fluorescence upon excitation of the Trp at
290 nm and emission of the Abz at 420 nm.
98 Figure 5.2. Amino acid sequences of the fluorescent λN peptides. Peptides were synthesized corresponding to various regions of the λN protein cleaved by Lon. The arrows indicate the seven major cleavage sites. With the exception of λN11-21, nonfluorescent analogs were also synthesized (Y(NO2) replaced with Y and Abz replaced with benzoic acid amide).
λN11-21
+ - H3N-K(Abz)AEKQAQWKAA-CO2 λN25-35
+ - λN35-45 H3N-Y(NO2)VGVSAKPVNK(Abz)-CO2
+ - H3N-Y(NO2)PILSLNRKPK(Abz) –CO2 λN55-65
+ - λN79-89 H3N-Y(NO2)DLTVLAEYHK(Abz) –CO2
- (Abz)QRTWYSKPGY(NO2) –CO2 λN89-103
+ - H3N-Y(NO2)RGITCSGRQK(Abz)IKGKS –CO2
Peptidase activity was measured using a Fluoromax 3 spectrofluorimeter (Horiba
Group) with a circulating water bath to maintain reactions at 37 °C and 3 mm path length microcuvettes from Hellma. Reactions contained 50 mM HEPES pH 8.0, 75 mM potassium acetate, 5 mM magnesium acetate, 5 mM DTT, 125 nM to 1 μM wild-type E. coli Lon (for good peptide substrates less Lon was used while for bad substrates more
Lon was needed to have good signal) and varying concentrations of a mixed peptide substrate (10% fluorescent and 90% nonfluorescent or 1% fluorescent and 99% nonfluorescent) to avoid complications from the inner filter effect. After equilibrating at
37°C for 1 minute, the reaction was initiated with 500 μM ATP. The amount of hydrolyzed peptide was measured by determining the maximum fluorescence generated per micromolar peptide after complete digestion by trypsin, chymotrypsin or elastase.
99 The steady-state rate of the reaction was determined from the tangent of the linear portion of the time course. This rate was converted to an observed rate constant (kobs) by dividing the rate by the enzyme concentration used in the assay. All of the rate constants were plotted as a function of peptide concentration and fit with equation 5.1
n n kobs = (kmax[S] ) / (K’m + [S] ) (5.1) where kobs is the observed rate constant, kmax is the maximum rate constant, S is the concentration of peptide, n is the Hill coefficient. Km, the concentration of peptide required to reach 50% kmax, is calculated from the relationship log Km’ = n log Km. All kinetic data shown is a result of averaging at least 3 separate experiments.
Peptidase activity using the λN11-21 peptide was also examined using HPLC.
Reactions were performed at 37 °C and contained 50 mM HEPES pH 8.0, 75 mM potassium acetate, 5 mM magnesium acetate, 5 mM DTT, 1 μM λN11-21, 1 μM E. coli
Lon, and the reaction was initiated with 500 μM ATP. Reaction aliquots were quenched at 0, 5 and 30 minutes with 0.02% TFA and analyzed on an analytical C-18 HPLC column (Vydac) with a Waters 616 HPLC system. The solvent system included water with 0.05% TFA and acetonitrile with 0.05% TFA and peptide products were separated using a linear gradient of 5 – 25 % acetonitrile with 0.05% TFA in 20 minutes.
Absorbance was monitored at 220 nm. As a control, the peptide was digested by trypsin and the products were analyzed by HPLC to ensure that peptide products were visible under these conditions.
5.2.2 λN purification and generation of λN deletion mutants
The plasmid pHF012 containing an N-terminal 6x His tag was transformed into the BL21 DE3 E. coli Lon cell strain and the cells grown to OD600 = 0.6 in 30 μg/mL
100 kanamycin in SB (superbroth) and induced for 2 hours with 1 mM IPTG (isopropyl-beta-
D-thiogalactopyranoside). The cells were pelleted by centrifugation and lysed by adding
5 mL of lysis buffer per gram of cells (lysis buffer: 100 mM sodium phosphate pH 8.0,
10 mM Tris, 1 mM beta-mercaptoethanol, 8 M urea) and stirring for 45 minutes. All cell debris was pelleted by centrifugation. The cleared lysate was loaded onto a Ni-NTA agarose column (Qiagen) equilibrated with lysis buffer. The column was then washed with lysis buffer to remove all excess proteins. The λN protein was eluted from the column with elution buffer (elution buffer: 100 mM sodium phosphate pH 4.5, 10 mM
Tris, 1 mM beta-mercaptoethanol, 8 M urea). The eluted protein was extensively dialyzed into λN storage buffer to remove urea (λN storage buffer: 20 mM Tris pH 7.6,
50 mM sodium chloride, 1 mM beta-mercaptoethanol, 20% glycerol). A white precipitate formed during dialysis and was removed by centrifugation. The λN protein in λN storage buffer was analyzed by SDS-PAGE on a 12.5% acrylamide gel and stained with coomassie to assess purity. The protein was quantitated using the extinction coefficient
-1 -1 at A280 (ε = 14,060 M cm ).
λN deletion mutants (λNΔ89-98, λNΔ89-107 and λNΔ99-107) were generated using basic cloning techniques. The forward primer used for all the mutants was oHF014. The reverse primers used were oJW103 for λNΔ89-98, oJW102 for λNΔ89-107 and λNΔ99-107rev for λNΔ99-107. PCR was used to amplify the λN gene with the desired deletions from plasmid pHF012. The PCR products were gel purified using a
0.8% agarose gel. The pCOLA-Duet vector (Novagen) and the purified PCR products
(insert) were digested with Bam HI and Hind III restriction enzymes. The digested vector was gel purified and the digested inserts were purified by phenol chloroform
101 extraction. The vector and insert were ligated using the Ligate-IT Rapid Ligation kit
(USB) and transformed into XL1 Blue competent cells (Stratagene). The λN protein deletion mutant plasmids were verified by DNA sequencing analysis and transformed into the BL21 DE3 E. coli cell strain.
5.2.3 λN degradation assay
λN degradation assays contained 50 mM HEPES pH 8.0, 15 mM magnesium acetate, 5 mM DTT, 10 μM wild-type or mutant λN protein, 1 μM E. coli Lon and the reaction was initiated with 5 mM ATP. At various time points, reaction aliquots were quenched with 5 x SDS-PAGE loading dye and incubated at 90°C for 1 minute. The aliquots were loaded and run on a 12.5% SDS-PAGE gel and stained with coomassie to detect the protein. The intensity of each band was quantitated using the Biorad Gel Doc
2000 and Quantity One quantitation software. The amount of λN protein degraded was corrected for loading errors and normalized to 1 by dividing the λN intensity by the Lon intensity at each time point and then dividing each time point by the quantity at time zero.
102 5.3 RESULTS AND DISCUSSION
5.3.1 Examining Lon substrate specificity using λN peptides
The λN protein is a bacteriophage transcription regulatory factor known to be an in vivo substrate of E. coli Lon. The λN89-98 peptide (containing residues 89-98 of the
λN protein) is a simplified, single cleavage site model peptide developed in our lab for use in the mechanistic characterization of Lon. Our peptide is superior to other synthetic peptide substrates because like protein substrates, λN89-98 stimulates the ATPase activity of Lon (41). The degradation profile of the λN protein has been reported showing that Lon cleaves the protein at 7 major cleavage sites (42). To examine the substrate specificity of Lon, I generated a library of fluorescent λN peptides based on the
λN protein cleavage profile. The peptide sequences are shown in Figure 5.2. With the exception of λN11-21, nonfluorescent analogs of all the peptides were synthesized. In all kinetic assays, a mixed substrate was utilized (10% fluorescent and 90% nonfluorescent or 1% fluorescent and 99% nonfluorescent) to avoid problems from the inner filter effect as was done previously with the λN89-98 peptide (40, 45). The inner filter effect becomes a problem when intermolecular quenching rather than intramolecular quenching occurs at high concentrations of Y-NO2. Previously, our lab demonstrated that the fluorescent and nonfluorescent analogs of λN89-98 were degraded by Lon in an identical manner (40). Therefore, for this study we assumed both analogs of the other λN peptides were processed the same by Lon.
With the exception of λN11-21 all peptides were cleaved by Lon. As shown in
Figure 5.3, no peptide cleavage is detected over the control without ATP. To ensure that the apparent lack of peptidase activity was not a result of a poorly designed FRET pair,
103 HPLC analysis was performed and verified that we are unable to detect any cleavage of the λN11-21 peptide by Lon. A control in which λN11-21 was digested with trypsin yielded multiple peaks on the HPLC indicating that the method for detecting λN11-21 peptide products by HPLC is good.
Figure 5.3. Peptidase activity with λN11-21. Three micromolar Lon was incubated with 500 μM λN11-21 and 500 μM ATP was added to initiate the reaction (blue). The reaction was excited at 290 nm and emission at 420 nm was monitored. No fluorescence changes are observed over the no ATP control (red). 8.5
8 ) 4 (x 10
7.5 relative fluorescence fluorescence relative
7 0 200 400 600 800 1000 time (s)
A complete kinetic study of the λN55-65 peptide was not completed due to peptide solubility issues. The peptide was dissolved in DMSO due to its poor solubility in water. Peptidase assays were performed at a low λN55-65 concentration (25 μM) and indicated the peptide is hydrolyzed by Lon; however it is a poorer substrate than the
λN89-98 substrate (Figure 5.4). It was not possible to measure peptide hydrolysis
104 activity at higher λN55-65 concentrations because the peptide precipitated out of solution.
Figure 5.4. Peptidase activity with λN55-65. Lon (125 nM) was incubated with 25 μM λN55-65 and the reaction was initiated with 1 mM ATP (red). A no ATP control was also performed (black). In the inset graph, hydrolysis of λN55-65 (red) is compared to λN89-98 (green) under identical reaction conditions. λN55-65 is hydrolyzed by Lon but is not as good of a Lon substrate as λN89- 98.
4 20
15 M) 3 μ 10 M) μ
peptide hydrolyzed ( hydrolyzed peptide 5 2 0
0 50 100 150 200 250 300
time (s)
1 peptide hydrolyzed ( hydrolyzed peptide
0
-1 0 50 100 150 200 250 300
time (s)
Complete steady-state kinetic analysis of Lon degrading λN25-35, λN35-45,
λN79-89 and λN89-103 was performed using the continuous fluorescence based assay.
The steady-state rate (kobs) of peptide hydrolysis was determined under conditions of saturating ATP and varying concentrations of peptide. A 1% fluorescent and 99% nonfluorescent peptide mixture was used for the λN25-35 peptide due to the very high
Km value and the need to go to very high peptide concentrations (25 μM – 2 mM). For all
105 other peptides a 10% fluorescent mixture was sufficient (25 μM – 1.5 mM). The kobs values were plotted as a function of peptide concentration and yielded sigmoidal plots
(Figure 5.5). The data were fit with equation 5.1 and the results are shown in Table 5.1.
The sigmoidal kinetics with n ~ 1.6 (observed with the λN89-98 peptide as well) likely reflects a certain degree of cooperativity amongst the subunits of Lon or an allosteric activation by peptide binding. Sigmoidal kinetics can also indicate a preferred pathway for substrate binding. For example, the formation of Lon:peptide then Lon:peptide:ATP is preferred over Lon:ATP then Lon:ATP:peptide. If this were the case, we would expect the rate of peptide hydrolysis to be inhibited at high concentrations of ATP and low concentrations of peptide (53, 60). I do not observe inhibition of peptidase activity under these conditions so I assume the sigmoidal behavior is cooperativity or allosteric activation by peptide.
106 Figure 5.5. Steady-state kinetics of hydrolysis of the λN peptides. The steady-state rate constants (kobs) of peptide hydrolysis with varying concentrations of λN25-35 (red), λN35-45 (blue), λN79-89 (green), λN89-103 (orange) and λN89-98 (black) were determined using the continuous fluorescence based peptidase assay. The kobs values were plotted as a function of peptide concentration and the data were fit with equation 5.1. The kinetic parameters are summarized in Table 5.1.
10
8
6 ) -1 (s obs k 4
2
0 0 500 1000 1500 2000
[peptide] (μM)
Table 5.1. Steady-state kinetic parameters for peptidase activity with λN peptides. NC: no cleavage detected, ND: not determined
kcat Km kcat/Km (s-1) (μM) (x 104 M-1s-1) λN 11-21 NC NC NC λN 25-35 11.3 ± 0.7 836 ± 290 1.3 λN 35-45 9.5 ± 0.2 461 ± 92 2.0 λN 55-65 ND ND ND λN 79-89 3.0 ± 0.1 479 ± 87 0.6 λN 89-98 9.0 ± 0.5 102 ± 30 8.8 λN89-103 3.7 ± 0.4 481 ± 237 0.7
107 Comparing the substrate specificity constants (kcat/Km) for the various peptides it becomes obvious that λN89-98 is by far the best Lon substrate (it has the highest kcat/Km value). There is a greater difference in the Km values than the kcat values between the
λN89-98 and the other peptides. The Km for λN89-98 is more than 4 fold lower than any of the other peptides, including λN89-103. The reason for large differences in the kinetic constants between λN89-98 and λN89-103 is not clear at this time considering λN89-103 is the same as λN89-98 except λN89-103 has 5 additional residues at the C-terminus.
In addition to measuring the peptidase activity, all of the λN peptides (with the exception of λN55-65 due to problems with solubility) were tested for their ability to stimulate the ATPase activity of Lon. Steady-state ATPase activity was measured using the [α-32P] ATP assay described previously, with 150 nM Lon, saturating peptide (1mM), and saturating ATP (1 mM). The observed rate constants for the various λN peptides as well as the enhancement over the intrinsic ATPase activity without peptide are reported in Table 5.2. All the λN peptides stimulate the ATPase activity of Lon to approximately the same extent as the λN89-98 peptide. Interestingly, the λN11-21 peptide stimulates
ATPase activity even though it is not cleaved by Lon. The mechanism by which peptides and proteins stimulate ATPase activity is not clear. However, the λN11-21 peptide data demonstrates that a peptide or protein does not need to be degraded by Lon to stimulate the ATPase activity.
108 Table 5.2. Steady-state kinetic parameters for ATPase activity with λN peptides. ND: not determined
kobs (s-1) stimulation intrinsic 0.34 ± 0.02 - λN 11-21 1.00 ± 0.25 2.9 λN 25-35 1.35 ± 0.29 4.0 λN 35-45 1.47 ± 0.39 4.3 λN 55-65 ND - λN 79-89 1.08 ± 0.02 3.2 λN 89-98 1.21 ± 0.06 3.6 λN 89-103 1.45 ± 0.13 4.3
5.3.2 Alanine scan of λN89-98 peptide
After scanning the λN peptides, clearly λN89-98 is a superior substrate for Lon. I performed an alanine scan of λN89-98 to examine the specific residues important for recognition by the enzyme. Each amino acid in λN89-98 was systematically changed to an alanine residue. The ATPase activity of Lon in the presence of the alanine scan peptides was measured using the radiolabeled ATPase assay previously described (46).
All the peptides stimulated the ATPase activity of Lon to approximately the same extent as the λN89-98 peptide.
The steady-state rates of peptide hydrolysis for each peptide with saturating ATP and varying concentrations of peptide (25 – 200 μM) were measured using the continuous fluorescence based assay used previously for λN89-98 (41). Unlike the λN peptides, only the fluorescent analogs were synthesized and used in the assays. No nonfluorescent analogs were used. Calibration curves for each peptide were created by completely digesting known amounts of peptide with trypsin and measuring the
109 maximum amount of fluorescence generated. At peptide concentrations greater than ~
150 μM the linear change in fluorescence is decreased due to the inner filter effect (45). I corrected for the inner filter effect at peptide concentrations greater than 150 μM by determining the percentage the fluorescence is reduced relative to the linear calibration curve. The rates of peptide hydrolysis were corrected by dividing the rate at a specific peptide concentration by the percentage of fluorescence reduction at the same peptide concentration (Figure 5.6) (41). All of the rates were converted to observed rate constants (kobs) by dividing by the enzyme concentration. The observed rate constants were plotted as a function of peptide concentration and the data were fit with equation 5.1 describing a sigmoidal curve. All of the kinetic parameters are summarized in Table 5.3 and shown graphically in Figure 5.7.
110 Figure 5.6. Calibration curve of peptide digested with trypsin. Various concentrations of peptide were digested with trypsin and the maximum changes in fluorescence are plotted as function of the corresponding peptide concentration. At concentrations of fluorescent peptide higher than ~ 150 μM, the fluorescence becomes reduced due to the inner filter effect. We corrected the kobs values for the inner filter effect at peptide concentrations greater than 150 μM by determining the percentage the fluorescence is reduced relative to the linear calibration curve. The rates of peptide hydrolysis were corrected by dividing the rate at a specific peptide concentration by the percentage of fluorescence reduction at the same peptide concentration.
7
6
5 ) 5 4 F (x 10 F (x
Δ 3
2
1
0 0 50 100 150 200 250 300
[peptide] (μM)
111 Figure 5.7. Relative kinetic parameters for λN89-98 alanine scan peptides. The steady-state kinetic parameters for peptidase activity with the λN89-98 alanine scan peptides were determined using the continuous fluorescence based assay. The relative kcat (red), Km (blue) and kcat/Km (green) values are shown. The kcat and kcat/Km values are reduced for the P3, P1 and P3’ positions. The actual kinetic parameters are summarized in Table 5.3.
Y R G I T C S G R Q K P5 P4 P3 P2 P1 P1’ P2’ P3’ P4’
1.6
1.4 k 1.2 cat Km 1 kcat / Km 0.8
relative values relative 0.6
0.4
0.2
0 P51111111111 P4 P3 P2 P1 P1’ P2’ P3’ P4’ λN 89-98 position replaced with alanine
112 Table 5.3. Steady-state kinetic parameters for peptidase activity with λN89-98 alanine scan peptides.
kcat Km kcat/Km (s-1) (μM) (x 104 M-1 s-1)
λN 89-98 R89A 8.5 ± 0.4 165 ± 38 5.1 λN 89-98 G90A 7.2 ± 0.6 104 ± 36 6.9 λN 89-98 I91A 2.1 ± 0.3 168 ± 69 1.2 λN 89-98 T92A 6.2 ± 0.5 121 ± 41 5.1 λN 89-98 C93A 3.3 ± 0.3 146 ± 47 2.2 λN 89-98 S94A 6.7 ± 0.8 80 ± 40 8.4 λN 89-98 G95A 8.7 ± 0.6 123 ± 38 7.0 λN 89-98 R96A 2.5 ± 0.2 117 ± 39 2.1 λN 89-98 Q97A 5.6 ± 0.8 99 ± 51 5.6
λN 89-98 I91R R96I 0.33 ± 0.03 156 ± 58 0.21 λN 89-98 I91A R96A 0.24 ± 0.03 199 ± 83 0.12 λN 89-98 I91A C93A R96A 0.16 ± 0.05 325 ± 200 0.04 λN 89-98 All Ala except I91 R96 6.7 ± 1 133 ± 75 5.0 λN 89-98 All Ala except I91 C93 R96 4.7 ± 0.2 56 ± 15 8.4
λN 89-98 C93G 0.031 ± 0.01 206 ± 98 0.01 λN 89-98 C93S 0.051 ± 0.01 208 ± 92 0.02 λN 89-98 S94G 3.67 ± 0.57 196.4 ± 92 1.9 λN 89-98 C93A S94Q 3.67 ± 0.32 199 ± 94 1.8
λN89-98 I91A, λN89-98 C93A, and λN89-98 R96A have a reduced kcat and kcat/Km compared to all the other peptides. When the location of the residues in the peptide are switched (λN89-98 I91R and R96I), peptidase activity is nearly abolished.
Peptidase activity is also nearly negligible when two of the important residues (λN89-98
I91A R96A) or all three important residues (λN89-98 I91A C93A R96A) are mutated to alanine residues (Table 5.3, Figure 5.8). On the contrary, when only these two or three residues are present in the peptide, (λN89-98 all Ala except I91 R96 or λN89-98 all Ala
113 except I91 C93 R96) peptidase activity approaches that of the normal λN89-98 peptide
(Table 5.3, Figure 5.8). Taken together, these results indicate that the P3, P1 and P3’ positions are crucial for determining the substrate specificity of Lon protease.
Figure 5.8. The P3, P1 and P3’ positions are crucial for determining substrate specificity. The steady-state kinetic parameters for peptidase activity were determined using the continuous fluorescence based assay. The relative kcat (red), Km (blue) and kcat/Km (green) values are shown. The actual kinetic parameters are summarized in Table 5.3.
3.5
3
2.5 kcat 2 Km
kcat / Km 1.5 relative values relative
1
0.5
0 I91R111111I91A I91A All Ala All Ala λN R96I R96A C93A except except 89-98 R96A I91 I91 R96 C93 R96
5.3.3 Importance of the C-S cleavage site in the λN89-98 peptide
The λN89-98 peptide is cleaved by Lon at a single site between the Cys93 and
Ser94 residues (C-S) (40). The results from the alanine scan indicated that changing the
Cys residue to an Ala reduced the rate of peptide cleavage by approximately 3 fold, but changing the Ser residue to an Ala residue had no affect. To further explore the importance of the Cys residue in the P1 position, I synthesized λN89-98 C93G (G-S) and
λN89-98 C93S (S-S). As shown in Figure 5.9 and Table 5.3, the Km of these peptides is
114 similar to the Km of λN89-98; however the rate of peptide hydrolysis is greatly reduced.
On the contrary, changing the Ser94 (at the P1’ position) to a Gly residue (λN89-98
S94G, C-G) does not affect Km, and the kcat is only reduced approximately 3 fold. I also replaced the C-S cleavage site in λN89-98 with the A-Q cleavage site in λN11-21, which was not hydrolyzed by Lon (λN89-98 C93A S94Q). The kinetic parameters of this peptide are comparable to the λN89-98 C93A peptide from the alanine scan.
Figure 5.9. The P1 position is more important than the P1’ position. The steady-state kinetic parameters for peptidase activity were determined using the continuous fluorescence based assay. The relative kcat (red), Km (blue) and kcat/Km (green) values are shown. The actual kinetic parameters are summarized in Table 5.3.
2
1.5 kcat
Km
1 kcat / Km relative values relative
0.5
0 C93G11111 C93S S94G C93A λN89-98 S94Q
In summary, the P1 position is very important in determining substrate specificity and the P1’ position plays only a small role. The Cys mutation results can be explained by the hydrophobicity of the substituted amino acid. Black and Mould calculated the logarithm of the partition coefficient (log P) for transfer from a polar to nonpolar phase (a measure of hydrophobicity) (91). Using the log P values they assigned a value from 0
(most polar) to 1.0 (most hydrophobic) to all 20 amino acids. The peptides with the most
115 hydrophobic amino acid at the P1 position are the best substrates for Lon: Cys (0.680) >
Ala (0.616) > Gly (0.501) > Ser (0.359). The kcat values for λN89-98, λN89-98 C93A,
λN89-98 C93G, and λN89-98 C93S have the same trend. It has previously been suggested that Lon prefers to cleave after hydrophobic and nonpolar residues, although it does not cleave after every hydrophobic residue (81). My results agree with this conclusion. In the future it would be interesting to test this hypothesis further by mutating the Cys in the P1 position to a Phe, the most hydrophobic amino acid with a value of 1.0 (91).
Amino acids are chiral molecules and in nature all proteins are made of the l- enantiomer. To evaluate the importance of chirality for Lon recognition, we synthesized
λN89-98 with d-Cys, λN89-98 with d-Ser and λN89-98 with d-Cys and d-Ser. I failed to detect any cleavage in the peptides with d-amino acids. Chirality is very important in
Lon substrate recognition as peptides with d-amino acids are not cleaved.
5.3.4 Degradation of a 2-site peptide
All of the model peptides used in our lab are short (10-11 amino acids) and typically contain one primary Lon cleavage site. Ultimately, it is of interest to study how
Lon processes a full length protein substrate. Toward this end, I developed a set of “2- site” peptides containing residues 79-98 of the λN protein (λN79-98 79F and λN79-98
89F, Figure 5.10). λN79-98 79F is used to monitor the Ser-Lys cleavage site and λN79-
98 89F is used to monitor the Cys-Ser cleavage site. By separately monitoring the kinetics of each site being cleaved and comparing them to the kinetics of the single site
116 peptides corresponding to the same region of λN, I will be able to determine if peptide hydrolysis at one site helps or impairs peptide hydrolysis at the other site.
Figure 5.10. Sequence of the “2-site” peptides. In the λN79-98 79F peptide, the fluorescent donor and quencher are positioned so the S- K cleavage site can be monitored. In the λN79-98 89F peptide, the fluorescent donor and quencher are positioned so the C-S cleavage site can be monitored.
λN79-98 79F
- Abz-QRTWYSKPGY(NO2)RGITCSGRQK-CO2 λN79-98 89F
+ - H3N-QRTWYSKPGY(NO2)RGITCSGRQK(Abz)-CO2
The steady-state rates of peptide cleavage were measured using the continuous fluorescence based assay with 125 or 300 nM Lon, 1 mM ATP and varying concentrations of peptide (25-250 μM). Only fluorescent peptides were used (not a mixed substrate of fluorescent and nonfluorescent), therefore a calibration curve was generated up to 250 μM peptide and the rates for peptide hydrolysis at concentrations above 150 μM were corrected for the inner filter effect as described previously. Figure
5.11 compares the steady-state rate constants for peptide hydrolysis of the Ser-Lys and the Cys-Ser cleavage sites in the “2-site” and single site peptides. The kinetics of peptide hydrolysis are similar for the 2-site and single site peptides, therefore I conclude that peptide hydrolysis at one site does not facilitate hydrolysis at a second site.
117 Figure 5.11. Peptidase activity with the “2-site” peptides. The steady-state rate constants at various concentrations of λN79-98 79F (blue ■) and λN79-98 89F (red ■) are shown with the steady-state rate constants of λN79-89 (blue ●) and λN89-98 (red ●). The steady-state rate constants are similar for the “1-site” and “2- site” peptides. 7
6
5
) 4 -1 (s obs
k 3
2
1
0 0 50 100 150 200 250 300
[peptide] (μM)
5.3.5 Wild type λN and λN deletion mutant protein purification and degradation
The peptide substrates have been very valuable for determining certain features of the mechanism of Lon such as how the ATPase and peptidase activities are coupled.
However, ultimately we would like to evaluate how Lon processes full length proteins, and not just peptides. Toward this goal, I expressed and purified the λN protein, an intrinsically unstructured native substrate for E. coli Lon. Bacteriophage λN is a transcription regulatory protein. It forms a complex with RNA polymerase, the E. coli transcription elongation factor NusA and boxB RNA to inhibit termination signals causing expression of the phage genes (92-96).
118 The λN protein was expressed as a His-tag protein and purified to near homogeneity under denaturing conditions using a Ni-NTA agarose column. The eluted protein fractions were analyzed by SDS-PAGE and the fractions with λN were pooled and dialyzed into λN storage buffer remove the urea. A white precipitate formed during dialysis and was removed by centrifugation. A small portion of the precipitate was solubilized in a small amount of lysis buffer (with 8 M urea) and run on a 12.5% polyacrylamide gel. The precipitate contained mostly λN protein, however, a significant amount of protein was still in solution (Figure 5.12). Therefore, the precipitated protein was discarded and only the soluble λN was used.
Figure 5.12. SDS-PAGE analysis of purified λN protein. Lane 1 is the soluble purified protein. Lane 2 is the precipitate that formed upon dialysis into λN storage buffer.
12
λN protein
To monitor λN degradation, wild type λN protein and E. coli Lon were incubated at 37°C and the reaction was started by the addition of ATP. Reaction aliquots were quenched in 5x SDS-PAGE loading dye at various time points. A control was also
119 performed with no ATP. All quenched reaction aliquots were run on a 12.5% polyacrylamide gel and stained with coomassie blue stain to detect intact λN protein.
Figure 5.13 illustrates that the λN protein is degraded by Lon, but only in the presence of
ATP.
Figure 5.13. Wildtype λN protein degradation with and without ATP. One micromolar Lon was incubated with 10 μM wildtype λN protein and the reaction was initiated with the addition of ATP. A no ATP control is also shown. Reaction aliquots were quenched at various time points, loaded onto a 12.5% polyacrylamide gel and analyzed by SDS-PAGE.
ATP no ATP
time 051020 (min) 0251020 Lon
λN protein
Lon cleaves the λN protein at seven major cleavage sites. I synthesized and tested model peptides corresponding to these cleavage sites and found that the region spanning residues 89-98 of the λN protein is kinetically favored compared to the other peptides. As the substrates of other ATP-dependent proteases such as Clp contain a
120 “recognition tag sequence” at the N- or C- terminus, I set out to explore if residues 89-98 of λN or other C-terminal residues were important for recognition of λN protein by Lon.
Three λN mutants (λNΔ89-98, λNΔ89-107, and λNΔ99-107) were generated as described in materials and methods and purified using the same procedure as for the wild- type enzyme. λNΔ89-98 lacks residues 89-98, λNΔ89-107 lacks the entire C-terminus, including residues 89-98 and λNΔ99-107 lacks just the C-terminus and still contains residues 89-98 (Figure 5.14). Mutant λN degradation was monitored like wild type λN using the discontinuous protease assay. Pictures of the SDS-PAGE gels are shown in
Figure 5.15. The intensity of the protein bands were quantitated and indicated that
λNΔ89-98 is degraded with kinetics comparable to the wild type protein, while λNΔ89-
107 and λNΔ99-107 are degraded about 2-fold slower (Figure 5.16). These results indicate that residues 89-98 are not important, rather residues 99-107 are required for optimal λN protein degradation by Lon. The λN mutant study suggests that the C- terminus of the λN protein contains some sort of Lon recognition sequence. However, the C-terminus alone does not determine if λN will be degraded by Lon. The λNΔ89-107 and λNΔ99-107 mutants are still degraded by Lon, but at a reduced rate as compared to wild type protein.
121 Figure 5.14. λN protein deletion mutants. Three λN protein deletion mutants were generated. λNΔ89-98 lacks residues 89-98, λNΔ89-107 lacks residues 89-107 and λNΔ99-107 lacks residues 99-107.
wildtype λN 1-88 89-98 99-107
λNΔ 89-98 1-88 99-107
λNΔ 89-107 1-88
λNΔ 99-107 1-88 89-98
Figure 5.15. λN protein deletion mutant degradation. One micromolar Lon was incubated with 10 μM λN protein and the reaction was initiated with the addition of ATP. Reaction aliquots were quenched at various time points, loaded onto a 12.5% polyacrylamide gel and analyzed by SDS-PAGE.
wildtype λN λNΔ89-98 λNΔ89-107 λNΔ99-107
Lon
λN
time 0 5 10 20 0 5 10 20 0 5 10 20 0 5 10 20 (min)
122 Figure 5.16. The C-terminus of the λN protein is important for recognition by Lon. The protein bands in the gels in Figure 5.15 were quantitated and plotted as a function of time. All three mutants are degraded by Lon, but the wildtype and λNΔ89-98 proteins are hydrolyzed faster than the λNΔ89-107 and λNΔ99-107 proteins.
1
0.8
0.6
0.4 N protein protein N remaining λ
relative amount of amount of intact relative
0.2
0 0 5 10 15 20 25
time (min)
123 5.4 CONCULSIONS
In this study, I have utilized model peptide substrates and a full length non- structured protein to examine the substrate specificity of Lon (where the enzyme cleaves along a protein substrate) and the mechanism by which Lon initiates the degradation of a full length protein. This knowledge should provide a better understanding of how Lon regulates important cellular functions such as RNA transcription. Using our model peptide λN89-98, I conclude that the P3, P1 and P3’ positions are substrate determining factors. A mutation at any one of these three positions greatly affects the ability of Lon to degrade the peptide substrate. In addition, I have shown that the C-terminal residues of the λN protein are important for optimal degradation of the protein by E. coli Lon protease.
The bacteriophage λN protein is a transcriptional antitermination protein that forms a ribonucleoprotein complex and allows for transcription of phage proteins.
Although λN is a relatively small protein lacking any defined structure, it has been shown that in the ribonucleoprotein complex, the amino terminus (residues 1-22) binds to boxB
RNA and forms an α-helical structure (92). NMR data indicates that residues 23-107 of the protein remain disordered. Residues 34-47 of the λN protein bind to the E. coli host protein factor, NusA, and the C-terminal residues (73-107) bind to RNA polymerase (92-
96). It has been suggested that the binding of λN to boxB RNA and forming an α-helix while the rest of λN remains disordered is part of a regulatory mechanism. The disordered C-terminus of the λN protein can be recognized and degraded if the NusA protein is limiting or antitermination is restricted. If λN is partially degraded to where the amino terminus is still in an α-helical structure bound to boxB RNA, other λN
124 molecules are unable to bind to the boxB RNA thereby inhibiting antitermination (92). I have found that the C-terminal residues of the λN protein are important for recognition by Lon protease. Therefore, the results of our λN deletion mutant study support this proposed regulatory model.
All of the peptides synthesized and tested in this study stimulate the ATPase activity of Lon protease. Most interesting is the λN11-21 peptide, which is not degraded by Lon but it increases the rate of ATP hydrolysis by ~ 3 fold. This result indicates that peptides or proteins that are not substrates have the ability to bind to Lon and affect ATP hydrolysis. More experimentation is required to identify the mechanism of ATPase stimulation. The intricacies of this process are not understood. However, one possible mechanism is that the peptide/protein initially binds to the enzyme and stimulates ATP hydrolysis. If the potential substrate contains the correct substrate determinants it is then delivered to the proteolytic domain where the protein is degraded.
Recently, studies have indicated that Lon protease is required for systemic infection of the bacterial pathogen Salmonella enterica Typhimurium causing gastroenteritis in humans (26, 27). As such, development of an inhibitor of the bacterial
Lon homolog represents a possible antibiotic target. Our lab has demonstrated there are differences in the substrate specificities of the human and bacterial enzymes and shown that peptidyl boronates are potent inhibitors of Lon activity (28, 29). Perhaps one could design a potent and specific inhibitor by tethering a boronate moiety to a peptide specific for bacterial Lon to avoid cross-reactivity with the human homolog and other serine proteases. The work presented here provides the initial knowledge needed to achieve this ultimate therapeutic goal. Not only do we now have a large peptide library with which
125 we can scan against other proteases to find a bacterial Lon specific peptide sequence to limit cross-reactivity, I have also begun to identify substrate determinate factors for bacterial Lon.
126
CHAPTER 6
CONCLUSIONS AND FUTURE DIRECTIONS
127 Using many different kinetic techniques I have learned a lot about the proteolytic mechanism of E. coli Lon protease. Steady-state kinetic experiments have shown that optimal peptide cleavage depends on nucleotide binding and hydrolysis. λN89-98 is a superior substrate for Lon and the P3, P1 and P3’ positions are crucial for determining substrate specificity. Poorer degradation of the λN protein deletion mutants suggests that the C-terminus of the λN protein contains some sort of recognition element that targets it for degradation by Lon. Extensive pre-steady-state experiments have led to a proposed kinetic model for the first round of peptide hydrolysis by Lon. After initial substrate binding, ATP hydrolysis is accompanied by an activation of the proteolytic active site, a slow peptide delivery event and finally peptide hydrolysis.
All of the kinetic data regarding the proteolytic activity must be considered along with the ATPase data to fully understand the mechanism of E. coli Lon protease. A kinetic scheme combining this data is presented in Figure 6.1 (40, 46, 52, 61, 74). The intrinsic ATPase activity is depicted in enzyme forms I, ii, iii and iv. Enzyme forms I to
VI represent the intermediates we have identified using various techniques. While this work has provided many answers regarding how Lon functions, several questions regarding the mechanism still exist. In the remaining discussion, I will provide an overview of some of the mechanistic aspects of Lon that will require further evaluation.
128 Figure 6.1. Kinetic mechanism for E. coli Lon protease The enzyme is shown as a dimer instead of a hexamer for simplicity. The ATPase and SSD domains are shown in green and the protease domain is shown in blue. The active site serine is shown in red. The steps highlighted in red were determined from experiments described in this thesis. In the absence of peptide Lon has an intrinsic ATPase activity. (I) free enzyme (ii) ATP binds (iii) nucleotide dependent conformational change (iv) ATP is hydrolyzed at the low-affinity site. In the presence of peptide substrate (Ia) ATP binds and (Ib) peptide binds in a random order (II) ATP and peptide are both bound (III) nucleotide dependent conformational change (IIIa) ATP is hydrolyzed at the low-affinity sites (IV) a conformational change that activates the proteolytic site (V) slow peptide delivery (VI) peptide bond cleavage. A post-catalytic form of Lon is generated that conducts multiple turnovers. It is not known how the post- catalytic form of Lon returns to free enzyme. Lon also hydrolyzes ATP very slowly at the high-affinity ATPase site and that is not included in this kinetic scheme.
ATP ATP ATP ATP kNTP = k = ADP ATP -1 burst 5 s -1 kon,ATP = 11 s 6.8 x 106 M-1s-1 ATP (iii) (ii) koff,ADP = (iv) 0.5 s-1
ATP ATP ATP Km ~ 100 μM peptide kon,ATP = 6.8 x 106 M-1s-1 (I) ATP ATP ATP ATP k = ADP ATP (Ia) kNTP = burst -1 11 s-1 ATP 5 s
peptide kon,ATP = (II) 6.8 x 106 M-1s-1 (III) (IIIa) k1,S679W = 7 s-1 Kd ~ 100 μM (Ib) Pd1 ADP + + ATP ADP Pd2 Pi ADP ATP ATP A post- ADP ATP (I) catalytic form Pd1 of Lon Pd2 k , k , ? conducting ? 2,S679W S679A koff,ADP = k = 1 s-1 multiple 0.5 s-1 lag turnovers (VI) (V) (IV)
129 Although our experiments suggest that the slow step in the peptide hydrolysis reaction involves peptide translocation, this result requires further clarification.
Therefore, a future direction of this project could be to generate other Lon mutants to provide additional evidence. In chapter 4, I described an initial, unsuccessful attempt to make an intrinsic tryptophan-free Lon mutant by mutating the three tryptophan residues in Lon into phenylalanine residues. The mutant had wild-type-like ATPase activity, but failed to form the nucleotide dependent conformational change that we detect with limited tryptic digestion. Obviously, one or more of the tryptophan residues (W297,
W303, or W603) is important for maintaining the structural integrity of the enzyme. It would be interesting to individually mutate the tryptophan residues to identify the indispensable position.
To prove the existence of a peptide delivery event I could try once again to make a tryptophan-free mutant by mutating the intrinsic tryptophan residues to tyrosine residues, perhaps a more conservative mutation. According to the Black and Mould hydrophobicity scale, tryptophan (0.878) and tyrosine (0.880) are more similar and less hydrophobic than phenylalanine (1.000) (91). If, with these mutations (W297Y, W303Y,
W603Y), Lon behaves like the wild-type enzyme we can mutate the active site serine to a tryptophan residue (W297Y, W303Y, W603Y, S679W) and use the dansyl peptide to monitor the events leading up to peptide hydrolysis. If the slow step (the lag phase) in the peptidase reaction is a peptide delivery event, we would expect to be able to measure an event with a rate constant of ~ 1 s-1 that is dependent on the concentration of peptide and ATP. With fluorescent signal coming only from the single tryptophan residue in the
130 protease domain, we would be able to assign this signal as a specific interaction between the peptide and the proteolytic active site of the enzyme.
Another mechanistic aspect of Lon that requires further evaluation is the generation of a post catalytic enzyme form. Extensive steady-state and pre-steady state kinetic experiments have led to a model for the first round of peptide hydrolysis by E. coli Lon protease (Figure 4.23). However, we have very little information about the specific kinetic steps after the initial peptide cleavage event and how enzyme turnover occurs. More specifically, steady-state product inhibition studies performed by our lab have indicated that Lon isomerizes into a post catalytic enzyme form and at this point we have very little information regarding this step (Figure 1.4) (40). One of the peptide products (Pd1) was used as a product inhibitor and its effect on peptide cleavage at varying concentrations of ATP and peptide was measured. Pd1 is a noncompetitive inhibitor against peptide and ATP indicating that Pd1 binds to two different Lon enzyme forms. Increasing the concentration of peptide or ATP did not alleviate the inhibition by
Pd1. Taken together, a mechanism was proposed in which Lon isomerizes into a post catalytic form that binds peptide products and ATP and full enzyme turnover requires the conversion of Lon back into a pre-catalytic form (40).
Results of the pre-steady state ATPase experiments also suggest that after the initial round of ATP and peptide hydrolysis Lon adopts a different form. None of the rate constants for the first round of ATP hydrolysis are affected by the presence of peptide; however the steady-state rate is stimulated in the presence of peptide or protein substrates
(52, 61, 74). Perhaps this more active enzyme form that is affected by peptide and protein substrates is the post catalytic form of Lon.
131 Considering all the kinetic evidence we have about the mechanism of Lon, I hypothesize that in the first round of peptide hydrolysis there is a slow (~1 s-1) peptide translocation event dependent on ATP binding and hydrolysis followed by peptide cleavage. After the initial peptide hydrolysis event, Lon isomerizes into the post-catalytic form and hydrolyzes peptide at the “steady-state rate” of ~ 9 s-1. To begin to explore the post catalytic form of Lon, I have performed some preliminary double mixing experiments on the stopped flow instrument using wildtype Lon and the fluorescent
λN89-98 peptide. A scheme describing a double mixing experiment is shown in Figure
6.2. Two reagents are rapidly mixed for a defined period of time, followed by the addition of a third reagent and monitoring the changes in fluorescence. When properly designed, these types of experiments could be useful in monitoring the events after the initial ATP and peptide hydrolysis events.
Figure 6.2. Scheme for a double mixing experiment on the stopped-flow instrument. In a double mixing experiment, the contents of syringes A and B are rapidly mixed for a defined period followed by the addition of a third reagent in syringe C. The reaction is then monitored by changes in fluorescence in the observation cell.
ABC
Δt
Observation cell
132 In the first experiment, wild type E. coli Lon and ATP were rapidly mixed for 3 seconds followed by the addition of fluorescent λN89-98 peptide and monitoring at excitation 320 nm and emission > 400 nm (Figure 6.3). A lag in peptide hydrolysis activity is observed. In the second experiment, Lon was pre-incubated with nonfluorescent λN89-98 and rapidly mixed with ATP for 1.5 seconds followed by the addition of fluorescent λN89-98 peptide and monitoring at excitation 320 nm and emission > 400 nm (Figure 6.4). No lag is observed in the peptidase time course. In the first experiment, ATP is hydrolyzed by Lon before the addition of peptide and there is still a slow phase before attainment of the steady-state rate. On the other hand, as expected in the second experiment, when ATP and peptide are hydrolyzed before the addition of the fluorescent peptide there is no slow phase and only a steady-state turnover of peptide. ATP hydrolysis alone is not enough for fast peptide hydrolysis by the enzyme. Both ATP hydrolysis and peptide hydrolysis are required.
133 Figure 6.3. First double mixing experiment. Five micromolar wildtype Lon was rapidly mixed with 100 μM ATP for 3 seconds followed by the addition of 500 μM fluorescent λN89-98 peptide (red). There is a lag phase in the reaction similar to the single mixing control time course where Lon was preincubated with λN89-98 and rapidly mixed with ATP (blue).
0.8 Lon ATP λN89-98F
0.6
0.4 Δt relative fluorescence relative
0.2
0 0246810
time (s)
Figure 6.4. Second double mixing experiment. Five micromolar wildtype Lon was preincubated with 50 μM nonfluoresecnt λN89-98 peptide and rapidly mixed with 100 μM ATP for 1.5 seconds followed by the addition of 50 μM fluorescent λN89-98 (red). There is no lag phase in the reaction unlike the single mixing control time course where Lon was preincubated with fluorescent λN89-98 and rapidly mixed with ATP (blue).
0.06 Lon + 0.05 λN89-98NF ATP λN89-98F
0.04
0.03 Δt
0.02 relative fluorescence relative
0.01
0 024681012
time (s)
134 These preliminary experiments simply demonstrate that in our lab we have the ability to design and perform experiments to explore the post catalytic form of Lon.
More specifically, we can evaluate what happens after the first round of peptide and ATP hydrolysis. For example, double mixing experiments and a proteolytically active tryptophan Lon mutant could be used to measure peptide delivery after the initial peptide cleavage event. We could substitute one of the hydrophobic residues near the active site serine with a tryptophan residue (I664W, L666W, V668W, V716W, I719W or I724W) and ideally be able to monitor dansyl peptide entering and dansyl peptide products leaving the proteolytic site of Lon. The Lon mutant preincubated with nonfluorescent
λN89-98 could be rapidly mixed with ATP for a defined period followed by the addition of the dansyl peptide. If peptide delivery is the rate determining step during steady-state peptide hydrolysis, we should detect a fluorescent signal resulting from the dansyl peptide approaching the tryptophan residue in the active site of the enzyme with a rate constant of ~ 9 s-1.
Lon protease has been shown to be important for systemic infection in certain pathogenic bacteria such as Salmonella enterica Typhimurium and Brucella abortus (25-
27). After creating multiple peptides to examine the substrate specificity of Lon, we have a large peptide library that could be used in the development of a potent and specific inhibitor of bacterial Lon protease. Previously, our lab has demonstrated that even though human Lon and bacterial Lon share ~ 40% sequence identity they have differences in their substrate specificities (28). We have also shown that peptidyl boronates are potent inhibitors of Lon activity (28, 29). I propose we can exploit these properties and develop a Lon inhibitor in which a boronate moiety provides potency
135 while the peptide moiety provides specificity to the bacterial Lon homolog. We can use our peptide library to screen other proteases such as human Lon, the proteasome and Clp to find a sequence specific for bacterial Lon and use this peptide as the inhibitor scaffold.
136
APPENDIX
137 7432 Biochemistry 2004, 43, 7432-7442
Correlation of an Adenine-Specific Conformational Change with the ATP-Dependent Peptidase Activity of Escherichia coli Lon† Jessica Patterson,‡,§ Diana Vineyard,‡,§ Jennifer Thomas-Wohlever,‡ Ramona Behshad,‡ Morris Burke,| and Irene Lee*,‡ Department of Chemistry and Department of Biology, Case Western ReserVe UniVersity, CleVeland, Ohio 44106 ReceiVed December 2, 2003; ReVised Manuscript ReceiVed March 12, 2004
ABSTRACT: Escherichia coli Lon, also known as protease La, is a serine protease that is activated by ATP and other purine or pyrimidine triphosphates. In this study, we examined the catalytic efficiency of peptide cleavage as well as intrinsic and peptide-stimulated nucleotide hydrolysis in the presence of hydrolyzable nucleoside triphosphates ATP, CTP, UTP, and GTP. We observed that the kcat of peptide cleavage decreases with the reduction in the nucleotide binding affinity of Lon in the following order: ATP > CTP > GTP ∼ UTP. Compared to those of the other hydrolyzable nucleotide triphosphates, the ATPase activity of Lon is also the most sensitive to peptide stimulation. Collectively, our kinetic as well as tryptic digestion data suggest that both nucleotide binding and hydrolysis contribute to the peptidase turnover of Lon. The kinetic data that were obtained were further put into the context of the structural organization of Lon protease by probing the conformational change in Lon bound to the different nucleotides. Both adenine- containing nucleotides and CTP protect a 67 kDa fragment of Lon from tryptic digestion. Since this 67 kDa fragment contains the ATP binding pocket (also known as the R/â domain), the substrate sensor and discriminatory (SSD) domain (also known as the R-helical domain), and the protease domain of Lon, we propose that the binding of ATP induces a conformational change in Lon that facilitates the coupling of nucleotide hydrolysis with peptide substrate delivery to the peptidase active site.
Lon, also known as protease La, is an oligomeric ATP- Lon could be superimposed on the HslUV structure (12), dependent protease which functions to degrade abnormal and thereby validating the utilization of HslU as a structural certain short-lived regulatory proteins in the cell (1-9). Lon model in studying the ATP-dependent peptidase reaction of represents one of the simplest forms of ATP-dependent Lon. proteases because both the ATPase and the protease domain The heterosubunit ATP-dependent protease HslU/HslV is are located within each monomeric subunit (10, 11). The a bacterial homologue of the proteasome (14). It contains functional form of Escherichia coli Lon has been shown to an oligomeric ATPase subunit and an oligomeric protease exist as a tetramer or an octamer, depending on the buffer subunit which assemble to form a functional enzyme (15- composition (3). Recently, the crystal structure of an inactive 17). The HslU subunit is an ATPase that functions to unfold mutant of the Lon protease domain has been reported. This and translocate polypeptide substrates through the central hexameric structure lacks the ATPase domain and contains cavities of the oligomeric enzyme by a threading mechanism an active site Ser to Ala mutation. The structure of this - protein displays a central cavity, which is commonly found (18 20). In light of the structural similarities shared by Lon in the ATPase and protease subunits of other ATP-dependent and HslU/HslV, as well as the high degree of sequence proteases (12). In addition to the protease domain, the homology in the ATPase domains of the two enzymes (21), structure of the R-subdomain of the ATPase subunit of Lon it is conceivable that Lon also couples ATP hydrolysis with has been determined. This subdomain constitutes the last 25% substrate unfolding and translocation. Furthermore, due to + of the carboxyl region of the intact ATPase domain of Lon; their sequence homology with many AAA (ATPase- however, it lacks the conserved Walker motifs found in ATP associated cellular activities) proteins, both Lon and HslU binding proteins (12, 13). Nevertheless, a structural model are classified as AAA+ proteins (21). The AAA+ motif consisting of the R-subdomain and the protease domain of typically contains an ATP binding site in the R/â-domain and a substrate sensor and discriminatory (SSD) site in the R-helical domain. It is believed that the movements of these † This work was supported in part by NIH Grant GM067172 and the PRI award sponsored by the Ohio Board of Regents. R.B. is the two domains initiate chemical energy transduction in enzyme recipient of the SPUR fellowship, a program sponsored by the Howard catalysis (21, 22). Hughes Medical Institute. * To whom correspondence should be addressed. Phone: (216) 368- The structural data on Lon cannot conclusively demon- 6001. Fax: (216) 368-3006. E-mail: [email protected]. strate that it possesses an ATP-dependent peptide translo- ‡ Department of Chemistry. cation step. However, we have obtained kinetic data to § These authors contributed equally to the preparation of the manuscript. support the existence of such a mechanism. In our earlier | Department of Biology. studies, we reported that the kcat of the degradation of a 10.1021/bi036165c CCC: $27.50 © 2004 American Chemical Society Published on Web 05/15/2004 Conformational Change in Lon and Enzyme Activities Biochemistry, Vol. 43, No. 23, 2004 7433 defined peptide (S3)1 by Lon is maximized by ATP hydroly- could be readily compared with our previous observations. sis. The defined model peptide substrate (S3) and its The kinetic data that were obtained were further put into nonfluorescent analogue (S2) were previously used as the context of the structural organization of Lon protease simplified mimics of the E. coli Lon substrate, the lambda by probing the conformational change in Lon bound to N protein, to demonstrate that catalytic turnover of peptide different nucleotides using limited tryptic digestion analyses. cleavage exhibits a dependence on ATP hydrolysis (23, 24). Combining the kinetic data and the limited tryptic digestion Since these model peptides lack any defined secondary data allowed us to evaluate the functional role of a nucle- structure (24), the observed ATPase dependency could reflect otide-specific conformational change in Lon as well as to a peptide translocation step (18-20, 25, 26). Using product quantify the functional relationship between NTP hydrolysis inhibition studies, we further demonstrated that the inhibitory and peptidase activation. effect of ADP toward Lon could be alleviated by increasing peptide substrate concentrations. This result allowed for the EXPERIMENTAL PROCEDURES construction of a minimal kinetic model to account for the Materials. ATP (lot A-7699) was purchased from Sigma, ATPase-coupled peptide translocation step prior to substrate while CTP (lot 2077F), GTP (lot 9311C), and UTP (lot cleavage (23). According to our model, ATP hydrolysis could 6688F) were purchased from ICN. The [R-32P]ATP, [R-32P]- be used to thread the unfolded peptide substrate through the CTP, [R-32P]GTP, and [R-32P]UTP were purchased from central cavity of the oligomeric enzyme, thereby delivering Perkin-Elmer. Fmoc-protected amino acids, Boc-Abz, Fmoc- the substrate to the proteolytic site. The nonhydrolyzable protected Lys Wang resin, and HBTU were purchased from ATP analogue, AMPPCP, cannot support peptide cleavage Advanced ChemTech and NovaBiochem. HATU was pur- because of the lack of energy transduction. Another nonhy- chased from PerSeptive Biosystem. Tris buffer, PEI-cellulose drolyzable analogue, AMPPNP, supports peptide cleavage TLC plates, ammonium molybdate, sodium citrate, malachite with an efficiency reduced compared to that of ATP. Peptide green oxalate salt, cell culture media, IPTG, SBTI, TPCK- - cleavage occurs because the structure of the AMPPNP treated trypsin, PMSF, and chromatography media were magnesium complex may have mimicked a transition state purchased form Fisher, Sigma, and ACROS Organic. in ATP hydrolysis, thus inducing a conformational change General Methods. Peptide synthesis and protein purifica- in Lon allowing for the delivery of the peptide (27). tion procedures were performed as described previously (23). However, because AMPPNP is not hydrolyzed by Lon, the All enzyme concentrations were reported as Lon monomer substrate translocation step is less efficient due to a lack of concentrations. energy transduction. NTP-Dependent Peptidase Assays. Steady-state velocity As NTP-activated casein degradation is accompanied by data were collected on a Fluoromax 3 spectrofluorimeter hydrolysis of the respective nucleotide (at 0.5 mM nucleo- (Horiba Group) as described previously (23). Assays con- tide), it is conceivable that the observed differences in the tained 50 mM Tris-HCl (pH 8.1), 5 mM magnesium acetate, respective NTPase-dependent reactions are attributed to the 5 mM DTT, and 125 nM E. coli Lon except in experiments differences in the nucleotide base structure. Therefore, under using GTP, where 200 nM E. coli Lon was used, with saturating NTP conditions, Lon will exist primarily as the varying concentrations (from 0 to 1 mM) of nucleotide (ATP, - Lon NTP form, and the functional relationship between the CTP, GTP, and UTP) or peptide substrate S3 (from 50 to kcat of NTP hydrolysis and peptidase activation could be 500 µM). All assays were performed at least in triplicate, quantitatively measured. Our proposed reaction model pre- and the averaged value of the rates determined for each set dicts that at saturating NTP concentrations, the hydrolyzable of nucleotide and peptide concentrations was fit to eq 1 as nucleotides that bind with weaker affinity to Lon than ATP described previously (23) using the nonlinear program will support peptide degradation with accompanying hy- EnzFitter (Biosoft). drolysis of the nucleotide triphosphate backbone. Further- more, the kcat for S3 cleavage with the hydrolyzable V [A]n[B] nucleotides should be at least comparable to or higher than V) max (1) ′ + ′ + n + n that with the nonhydrolyzable ATP analogue-mediated (KibK a K a[B] Kb[A] [A] [B]) reaction if energy transduction affects peptide translocation. To test our hypothesis, we measured the steady-state kinetic where V is the observed rate, Vmax is the maximal rate, A is parameters for S3 cleavage in the presence of ATP, GTP, the peptide substrate, B is the nucleotide, Ka is the Michaelis CTP, or UTP, as well as the NTPase activity of Lon in the constant for A, Kib is the intrinsic dissociation constant for absence and presence of S3. Since the kinetic parameters of B, and Kb is the Michaelis constant for B. The kcat value ATP- and AMPPNP-mediated S3 degradation have been was determined by dividing Vmax with the concentration of previously determined (23), results obtained from this study the Lon monomer. The Km values for peptide hydrolysis were calculated from the relationship log K′a ) n log Km, where 1 Abbreviations: AMPPNP, adenylyl 5-imidodiphosphate; AMPPCP, n is the Hill coefficient. adenylyl (â,γ-methylene)diphosphonate; DTT, dithiothreitol; Abz, Radiolabeled NTPase Assays. Steady-state velocity data anthranilamide; Bz, benzoic acid amide; NO2, nitro; Tris, 2-amino-2- for nucleotide hydrolysis were measured as described (hydroxymethyl)-1,3-propanediol; amp, ampicillin; KPi, potassium phosphate; λ N (lambda N protein), lambda phage protein that allows elsewhere (28), and all reactions were performed at least in E. coli RNA polymerase to transcribe through termination signals in triplicate. Briefly, for the NTPase measurements, each the early operons of the phage; SE, standard error; S2, nonfluorescent reaction mixture (50 µL) contained 50 mM Tris-HCl (pH analogue of S3 that is cleaved by Lon in the same manner as S3 8.1), 5 mM magnesium acetate, 2 mM DTT, and 150 nM [YRGITCSGRQK(Bz)]; S3, mixed peptide substrate containing 10% of the fluorescent peptide Y(NO2)RGITCSGRQK(Abz) and 90% S2; Lon monomer for ATP or UTP and 600 nM Lon monomer NTP, nucleotide triphosphate. for CTP or GTP. For the peptide-stimulated NTPase reac- 7434 Biochemistry, Vol. 43, No. 23, 2004 Patterson et al. tions, 500 µM peptide substrate (S2) was added to each S2 peptide, and either 1 mM ATP, ADP, or AMPPNP, 2 reaction mixture, and the reactions were initiated by the mM CTP or GTP, or 3 mM UTP were started by the addition addition of NTP. Subsequently, 5 µL aliquots were quenched of 1/50 (w/w) TPCK (N-p-tosyl-L-phenylalanine chlorom- in 10 µL of 0.5 N formic acid at seven time points (from 0 ethyl ketone)-treated trypsin with respect to Lon. At 15 and to 12 min). A 3 µL aliquot (ATPase) or 2 µL aliquot 30 min, a 3 µL reaction aliquot was quenched in 3 µgof (CTPase, GTPase, and UTPase) of the reaction was spotted soybean trypsin inhibitor (SBTI) followed by boiling. The directly onto a PEI-cellulose TLC plate (10 cm × 20 cm) quenched reactions were then resolved by 12.5% SDS- and the plate developed in 0.3 M potassium phosphate buffer PAGE analysis and visualized with Coomassie brilliant blue. (pH 3.4). Radiolabeled NDP nucleotide was then quantified Identification of Tryptic Digestion Sites in Lon by Peptide using the Packard Cyclone storage phosphor screen Phosphor Sequencing. The trypsin cleavage sites in Lon were identified imager purchased from Perkin-Elmer Life Science. To by sequencing the Lon fragments generated by limited tryptic compensate for slight variations in spotting volume, the digestion in the presence of 1 mM ATP or GTP. After 45 concentration of the NDP product obtained at each time point min, the reactions were quenched with PMSF (phenyl- was corrected using an internal reference as shown in eq 2. methanesulfonyl fluoride) followed by boiling, and then resolved on a denaturing 4 to 15% gradient gel. The Lon NDP ) dlu fragments were electroeluted onto a PVDF membrane and [NDP] ( + )[NTP] (2) NTPdlu NDPdlu sequenced by Edman degradation performed by the Molec- ular Biotechnology Core at the Lerner Research Institute of The kinetic parameters were determined by fitting the kobs the Cleveland Clinic (Cleveland, OH). The first five amino data with eq 3 using the nonlinear regression program acids at the amino termini of each Lon fragment were KaleidaGraph (Synergy). determined and then matched against the primary amino acid sequence of E. coli Lon [GenBank accession number P08177 k [B] ) obs,max (3)] to identify the respective trypsin cleavage sites. kobs + (3) Km [B] RESULTS where k is the observed rate constant, k is the maximal obs obs,max V rate, B is the nucleotide, and Km is the Michaelis-Menten Steady-State Kinetic Analysis of Peptide Clea age. Using constant. the fluorescent peptidase assay previously employed to Malachite Green NTPase Assays. Steady-state velocity compare the kinetics of ATP- versus AMPPNP-mediated data for ATP or CTP hydrolysis were also measured using peptide degradation (23), we determined the kinetic constants a modified assay to detect inorganic phosphate (Pi) release of ATP-, CTP-, UTP-, and GTP-mediated S3 hydrolysis by (28, 29). Solutions containing 0.045% (w/v) malachite green E. coli Lon. The NTP-dependent S3 hydrolysis reactions oxalate (MG) in deionized water, 4.2% (w/v) ammonium were monitored by the increase in fluorescence, which molybdate (AM) in 4 N HCl, 2% (v/v) Triton X-100 in reflects the amount of peptide hydrolyzed over time. The deionized water, and 34% (w/v) sodium citrate‚2H2Oin steady-state observed rate constants of S3 cleavage (kobs,S3) deionized water were prepared. Prior to each NTPase assay, were determined at varying concentrations of S3 and a 3:1 mixture of MG and AM was made, stirred for at least saturating concentrations of the respective NTP. Under these 20 min, and filtered through filter paper. The Triton X-100 conditions, Lon exists predominantly in the Lon-NTP form, solution was then added to this MG/AM solution in the and thus, the observed differences in the NTP-dependent amount of 100 µL per 5 mL of MG/AM solution. A solution peptidase activities reflect the effect of nucleotide hydrolysis of NaHPO4 and NaH2PO4 (pH 8.1) was used for the rather than binding. Plotting kobs,S3 as a function of S3 calibration standard. For the NTPase measurements, we used concentrations yields sigmoidal plots as shown in Figure 1A. a 210 µL reaction mixture containing 50 mM Tris-HCl (pH On the other hand, plotting kobs of S3 hydrolysis at saturating 8.1), 5 mM magnesium acetate, 2 mM DTT, various S3 and increasing NTP concentrations yields hyperbolic plots concentrations of Lon protease (125 nM for ATP), and 500 (Figure 1B). The kobs,S3 values were also measured at varying µM peptide substrate (S2, the nonfluorescent analogue of concentrations of NTP at several fixed peptide concentra- S3) for the peptide-stimulated NTPase assays. For all assays, tions, and the full set of velocity data was fitted with eq 1 to the NTPase reaction was initiated by the addition of the yield the kinetic constants summarized in Table 1. The nucleotide to the reaction mixture. At eight time points (from detection of sigmoidal kinetics at fixed nucleotide but varying 0 to 15 min), a 25 µL aliquot was thoroughly mixed with peptide concentrations indicates that the peptide concentration 400 µL of an MG/AM/Triton X-100 solution. After 30 s, 50 term shown in eq 1 is associated with a Hill coefficient (23). µL of 34% sodium citrate was added for color development. Since Lon functions as an oligomer, the detection of a Hill The absorbance was then recorded at 660 nm on a UV-vis coefficient of >1 at increasing peptide concentrations sug- spectrophotometer for each time point. The amount of Pi gests that the enzyme subunits communicate with each other formed at each time point was determined by comparing the during peptide cleavage, and the further binding of peptide absorbance of the sample to a Pi calibration curve. Initial substrate stimulates peptide hydrolysis. velocities were determined from plots of the amount of Pi As summarized in Table 1, both the Km,S3 and n values released versus time. The kinetic parameters were determined vary only slightly among the different NTP-mediated pep- by fitting the averaged rate data with eq 3. tidase reactions, indicating these parameters are not affected Tryptic Digestions. Tryptic digest reactions in mixtures by nucleotide binding or hydrolysis. The Kib and Kb values containing 1.5 µM Lon, 50 mM Tris (pH 8.1), 5 mM of the different nucleotides represent the intrinsic dissociation magnesium acetate, 2 mM DTT, with or without 800 µM constant and Michaelis constant of the individual nucleotide Conformational Change in Lon and Enzyme Activities Biochemistry, Vol. 43, No. 23, 2004 7435
for the different NTPs also varies with the chemical structure of the nucleotide that is used (Table 1). Comparing the kcat,S3 values of the peptidase reactions in Table 1 reveals the order of the ability of different nucleotides to activate S3 cleav- age: ATP > CTP > UTP ∼ GTP > AMPPNP. With the exception of AMPPNP, the decreasing order in kcat,S3 mediated by different NTPs correlates with the decreasing affinities of Lon for the respective nucleotides that are represented by Kib and Kb (Table 1): ATP ∼ AMPPNP > CTP > GTP > UTP. Although both ATP and AMPPNP bind to Lon with comparable affinity, they exhibit different stimulatory effects toward S3 degradation. On the other hand, CTP, which has a weaker affinity for Lon than AMPPNP, exhibits a 4.2-fold higher value of kcat,S3. Comparing the structures of the nucleotide bases reveals that both ATP and CTP share similarities in the N6 (in ATP) and N4 (in CTP) amino groups in the pyrimidine ring. Therefore, it is likely that the binding interaction between Lon and the amino group in the nucleotides is responsible for activation of peptide hydrolysis. Although both GTP and UTP are less effective than ATP or CTP in activating S3 cleavage, the kcat values of these two NTP-mediated S3 cleavages are at least comparable if not slightly higher than that of AMPPNP- mediated S3 cleavage (23, 24). Collectively, these data suggest that despite the reduction in binding affinity between Lon and CTP, GTP, or UTP, the hydrolysis of the triphos- phate in these nucleotides confers a catalytic advantage over the tight binding AMPPNP. This result indicates that the kcat,S3 is not solely dependent on the affinity of Lon for the nucleotide. The similarity in the Km,S3 values is consistent with our previous report that the peptide substrate binds to Lon independent of the nucleotide (23). This result is also corroborated by the observation that the limited tryptic FIGURE 1: Steady-state peptidase activity of Lon in the presence digestion patterns of Lon are identical regardless of the of different NTPs. (A) Lon was preincubated with 25, 50, 100, 150, 200, and 500 µM S3 prior to the addition of 1 mM ATP (b), presence or absence of the S3 peptide substrate and further CTP (9), and GTP ([) and 1.6 mM UTP (×). All assays were indicate that binding of the S3 peptide does not induce any performed in triplicate, and the averaged kobs values were plotted detectable conformational change in Lon. against the corresponding S3 concentration. The data were best fit Although Lon contains an identical peptide sequence for ) n n with the Hill equation kobs (kobs,max[S] )/([S] Km), where kobs is the ATPase site in each monomer, this oligomeric enzyme the observed steady-state rate constant, kobs,max is the maximum observed steady-state rate constant, [S] is the concentration of the has at least two different affinities for ATP. The Kd values peptide, and n is the Hill coefficient (30). (B) Lon was preincubated for the high- and low-affinity sites are 1 and 10 µM, with 500 µM S3 prior to addition of 5, 10, 50, 75, 150, and 250 respectively (31, 32). Since the Kib and Kb for ATP obtained µM and 1 mM ATP (b), 30, 75, 150, 200, 500, and 800 µM and in this study approximate 7 µM (Table 1), we have concluded 9 1 mM CTP ( ), 50, 75, 150, 250, and 500 µMand1mMGTP that the low-affinity ATP-binding site of Lon is reported in ([), and 100, 200, 400, 600, and 800 µM and 1.6 mM UTP (×). Table 1. On the basis of this observation, we further utilized All assays were performed in triplicate, and the average kobs values were plotted against the corresponding NTP concentration. The data the Kib values summarized in Table 1 to assess the relative were best fit with eq 3. affinity of Lon for the different nucleotides. It should be noted that Kib values only measure the effect of nucleotide Table 1: Steady-State Kinetic Parameters of NTP-Dependent binding on activation of the peptidase activity; therefore, Kib Cleavage of S3 by Lon alone cannot reveal whether Lon binds to CTP, GTP, or UTP kcat,S3 ( SE Km,S3 ( SE Kb ( SE Kib ( SE using the high- or low-affinity nucleotide-binding site. - nucleotide (s 1) (M) (M) (M) n Steady-State Characterization of the NTPase ActiVities of ATP 9.0 ( 0.5 102 ( 30 7 ( 1 7.4 ( 2 1.6 Lon. In addition to being a protease, Lon also possesses CTP 4.2 ( 0.1 151 ( 43 100 ( 20 73 ( 14 1.52 intrinsic ATPase activity that is increased in the presence of UTP 1.9 ( 0.2 99 ( 32 350 ( 125 389 ( 112 1.43 GTP 1.7 ( 0.2 219 ( 43 200 ( 78 250 ( 92 1.6 the peptide (24) or protein substrate (3, 33). Although Lon AMPPNPa 1.0 ( 0.1 77 ( 7NDb 10 ( 1 1.6 can hydrolyze any of the NTPs used in this study in the a These values were obtained from ref 23. b Not determined. presence and absence of casein (34), the kinetic parameters associated with these NTPases have not been reported. To (30), respectively. These values vary for the different NTPs. quantitatively characterize the intrinsic and peptide-stimulated According to Table 1, the Kib of each NTP approximates its NTPase activitiy of Lon, we determined the steady-state Kb value, indicating that the Michaelis constant reflects the kinetic parameters of the respective nucleotide hydrolysis intrinsic binding affinity of the nucleotide for Lon. The kcat,S3 in the absence and presence of the S2 peptide, the nonfluo- 7436 Biochemistry, Vol. 43, No. 23, 2004 Patterson et al.
FIGURE 2: Steady-state NTPase activity of Lon in the absence and presence of 800 µM S3 peptide. The kobs values for NTP hydrolysis were determined in the absence (b, 9, 2) and presence (+) of S3 as described in Experimental Procedures. The concentrations of ATP that were used were 25, 50, 100, 250, and 500 µM and 1 mM (A). The concentrations of CTP or GTP that were used were 25, 50, 100, 250, and 500 µM. The concentrations of UTP that were used were 50, 100, 200, and 600 µM and 1 and 2 mM.
Table 2: Steady-State Kinetic Parameters of NTP Hydrolysis by Lon intrinsic S2-stimulated
kcat,NTP ( SE Km,NTP ( SE kcat/Km(NTP) kcat,NTP ( SE Km,NTP ( SE kcat/Km(NTP) nucleotide (s-1) (µM) (×103 M-1 s-1) (s-1) (µM) (×103 M-1 s-1) NTPase enhancement ATP 0.26 ( 0.02 47 ( 10 5.5 1.0 ( 0.1 49 ( 5 20 3.8 CTP 0.14 ( 0.02 60 ( 10 2.3 0.28 ( 0.02 69 ( 20 4.1 2 UTP 0.50 ( 0.02 100 ( 4 5.0 1.1 ( 0.1 132 ( 30 8.3 2.2 GTP 0.09 ( 0.02 57 ( 5 1.8 0.15 ( 0.02 42 ( 11 3.6 1.7 rescent analogue of S3 (23). Two different discontinuous stimulated kinetic data generated by the radiolabeled nucleo- assays were used to measure the steady-state rate constants tide hydrolysis assay were highly scattered. Therefore, the of NTP hydrolysis (kobs,NTP): a malachite green colorimetric malachite green assay was used to determine the CTPase assay (28, 29) that monitors the release of inorganic kinetic parameters. The plots of kobs,NTP versus nucleotide 32 phosphate (Pi) and an R- P-labeled ATPase assay that concentration in both the absence and presence of S2 are 32 measures [R- P]ADP production (28). Plots of kobs,NTP as a shown in Figure 2. All kinetic assays were performed at least function of nucleotide concentrations in all the NTPase in triplicate, and the averaged values at each time point were experiments yielded hyperbolic plots (Figure 2). Using ATP reported in each plot. In all cases, the observed rate constant as a reference, we have observed that the kinetic parameters data were fitted with eq 3 to yield the kinetic constants obtained by the malachite green assay are comparable to summarized in Table 2. those obtained by the R-32P-labeled ATPase assay. Therefore, As discerned in Figure 2, the hydrolysis of NTPs displays with the exception of CTPase, we used the radiolabeled Michaelis-Menten kinetics, which agrees with the detection NTPase assay to determine the kinetic parameters for NTP of hyperbolic plots in the dependency of kobs,S3 values as a hydrolysis because less reagent was required. With regard function of NTP concentrations (Figure 1B). Table 2 shows to CTP hydrolysis, the intrinsic CTPase but not the peptide- that while the Km values for both the intrinsic and S2- Conformational Change in Lon and Enzyme Activities Biochemistry, Vol. 43, No. 23, 2004 7437 stimulated NTPase reactions are comparable, the kcat values vary considerably. The NTPase enhancement value is the ratio of the S2-stimulated kcat to the intrinsic kcat of the respective NTP, and yields an assessment of the effect of peptide substrate on NTPase stimulation (Table 2). As summarized in Table 2, the kcat of the intrinsic NTPase activity is in the following order: UTP > ATP > CTP > GTP. On the other hand, the kcat of the S2-stimulated NTPase activity is in the following order: ATP ∼ UTP > CTP > GTP. Comparing the order of intrinsic versus peptide- stimulated kcat values for the respective NTPase activities reveals that the S2 peptide exhibits a modest stimulatory effect on the ATPase activity (3.8; Table 2) compared to the CTP, GTP, and UTPase activities (1.7-2.2; Table 2). Furthermore, the kcat/Km value for ATP in the presence of the peptide substrate is higher than those of the other NTPs, thus indicating that ATP is the preferred activator. Probing the Structural Changes in the AAA+ Motif of Lon. To correlate the kinetic data given in Tables 1 and 2 with the function(s) of the AAA+ chaperone motif in Lon (21, 22), we utilized limited tryptic digestion to probe the conformational change of Lon in the absence and presence of the S2 peptide as well as different NTPs. The domain organization of E. coli Lon has been examined by comparing the digestion pattern of the enzyme by overnight digestion with V8 protease (35). Both ATP- and ADP-bound Lon are more resistant to V8 digestion than unbound Lon, suggesting that the enzyme undergoes a conformational change when binding to these nucleotides. The functional role of the ATP- or ADP-induced conformational change of Lon is not known. FIGURE 3: Limited tryptic digestion of Lon in the presence of To address this issue, we examined the susceptibility of Lon nucleotides. Lon was digested with a limiting amount of trypsin to limited tryptic digestion under the reaction conditions used and quenched at the indicated times with soybean trypsin inhibitor in our kinetic studies. Using trypsin rather than V8 protease (SBTI) as described in Experimental Procedures. The first lane as a probe, we were able to detect an adenine-specific shows the molecular markers in kilodaltons (from top to bottom): conformational change in Lon within 30 min of digestion. 172, 110, 79, 62, 48, 37, 24, 19, 13, and 5. The limited tryptic digestion patterns of Lon incubated in Table 3: Identification of the Trypsin Digestion Sites in Lon the absence and presence of 800 µM S2 peptide (8Km,S3) and with a saturating amount of nucleotides are shown in observed molecular sequence domains Figure 3A. Since the NTPs are also hydrolyzed under these mass (kDa) identifieda cleavage site included condition reaction conditions, both the Lon-NDP and Lon-NTP 67 AIQKE and A237/E241-K783 ATPase, SSD, b forms are anticipated to exist during the tryptic digestion ELGEM peptidase conditions. Identical Lon fragmentation patterns were de- 45 ELGEM A237-R587 ATPase, SSD c - tected in the absence or presence of the nucleotide when S2 35 LSGYT L490 K783 SSD, peptidase c 26 MNPER and M1/S6-K236/K240 amino terminus b was omitted (data not shown), indicating that the peptide SERIE did not induce any conformational change in Lon that could 23 ADNEN A588-K783 peptidase c be detected by tryptic digestion. Figure 3A shows that Lon 7 LSGYT L490-R587 SSD c is more resistant to tryptic digestion when incubated with a The first five-amino acid sequence of each Lon fragment was ATP than with the other NTPs, suggesting that the Lon- identified by Edman degradation as described in Experimental Proce- - dures. b Detected in the absence or presence of NTPs. c Detected in ADP or Lon ATP form adopts a more compact conforma- the absence of adenine-containing nucleotides. tion. Since all the limited tryptic digestion results shown in Figure 3A most likely reflect the resistance of the Lon- NDP complex to tryptic digestion, we performed a control Lon might have a less compact conformation than ADP- to further compare the stability of Lon bound to ADP or bound Lon. AMPPNP and without nucleotide (Figure 3B). Since AMP- When tryptic digestion was performed on Lon incubated PNP is not hydrolyzed by Lon and it supports S3 cleavage with ATP, ADP, AMPPNP, or CTP, two prominent frag- (with reduced efficiency compared to ATP), it is used to ments with apparent molecular masses of 67 and 26 kDa mimic the effect of ATP bound to Lon while resisting tryptic (Table 3) were detected within 15 min. In contrast, nucle- digestion. According to Figure 3B, the fragmentation patterns otide-free Lon, as well as the GTP- and UTP-incubated Lon, of Lon incubated with ATP, ADP, and AMPPNP are was mostly digested by trypsin to yield fragments with comparable. However, in the digestion reaction of Lon apparent molecular masses of 67, 45, 35, 26, and 7 kDa, incubated with AMPPNP, a slightly higher level of the 35 respectively (Table 3). Increasing the incubation time of the kDa fragment is detected, suggesting that AMPPNP-bound limited tryptic digestion reactions did not significantly alter 7438 Biochemistry, Vol. 43, No. 23, 2004 Patterson et al. similarity shared by Lon and the ATPase HslU (12), we have constructed a model for the nucleotide-binding site of Lon based upon the crystal structure of HslU bound to dADP (PDB entry 1HT2). We utilized the Swiss PDB viewer program to construct the nucleotide binding site for one of the monomers in HslU (the E chain) bound to dADP. This region contains Ile17-Ile66 of the E chain, which encom- passes the conserved Walker A motif found in the AAA+ protein family (Figure 5A,B). Sequence alignment of E. coli Lon with the nucleotide binding site of HslU reveals that the residues in the Walker A motif of Lon and HslU are highly homologous (Figure 5C). This level of sequence homology validates the fitting of the Lon primary amino acid sequences flanking Ala316-Leu365 to the structure of HslU (consisting of Ile17-Ile66) to yield a model for the nucle- otide binding site in Lon (Figure 5C,D). Figure 5B depicts
FIGURE 4: Fragmentation of Lon resulting from limited tryptic the crystal structure of the region in HslU that interacts with digestion. The peptide fragments generated from limited tryptic the N6 amino group in adenine, whereas Figure 5D illustrates digestion were acquired as described in Experimental Procedures. a model (based upon the sequence alignment shown in Figure The sizes of the Lon fragments were estimated on the basis of their 5C) showing that Lon appears to interact with the N6 amino position on the SDS gel compared to the molecular mass markers. group in ADP via the same mechanism as HslU. In HslU, The relative positions of the fragments compared to the intact Lon monomer were deduced from the sequencing data given in Table the N6 amino group functions as a hydrogen bond donor 3. that could simultaneously interact with the carbonyl oxygen in Ile18 and Val61 along the amide backbone of the enzyme. the digestion patterns of Lon incubated with ATP or ADP. According to Figure 5C, Val61 is located within the Walker However, the 67 and 45 kDa fragments of the reaction A motif and is conserved in Lon. Therefore, it is conceivable mixtures containing GTP, CTP, or UTP were further digested that the carbonyl oxygen of Val360 in Lon makes the same by trypsin to yield more of the 35, 23, and 7 kDa fragments contact with the N6 amino group in adenine. Through (Figure 3A). sequence alignment, we have also identified that the carbonyl The trypsin cleavage sites identified by peptide sequencing oxygen in the amide backbone of Gln317 in Lon (Figure experiments are summarized in Table 3, and the accessible 5D) could adopt the same function as Ile18 in HslU (Figure trypsin cleavage sites of Lon are illustrated in Figure 4. The 5B). Taken together, our model indicates that Lon binds to sizes of the tryptic digestion fragments determined by their the adenine nucleotide via two hydrogen bond interactions relative mobility on the 4 to 15% gradient gel with respect with the N6 amino group in adenine, and disruption of these to the molecular markers agreed well with the calculated interactions may affect the affinity of Lon for the nucleotide. molecular mass of the Lon fragments based upon the identities of the trypsin cleavage sites (Table 3). In the DISCUSSION presence of ATP or ADP, trypsin cleaves Lon primarily at Lon is one of the simplest ATP-dependent proteases within K236 or K240 to yield a 26 kDa fragment corresponding to the AAA+ protease family (3, 21, 22). On the basis of the amino terminus of Lon and a 67 kDa fragment corre- sequence and structural similarities shared by Lon and other sponding to the AAA+ chaperone motif and the protease ATP-dependent proteases, it is proposed that Lon also domain of Lon [Figure 4 and Table 3 (3, 21, 22, 35-37)]. couples ATP binding and hydrolysis to unfold and translocate The AAA+ chaperone motif is further comprised of the peptide substrates into its protease chamber. The unfolded Walker A and B motifs of the ATPase domain and SSD polypeptide is threaded through the central protease chamber domain found in many ATPases (21, 22). When intact Lon which is formed by oligomerization of the enzyme subunits. was treated with trypsin in the presence or absence of non- In previous studies, we reported that Lon exhibited ATPase adenine nucleotide triphosphates, the 67 kDa fragment was dependency in hydrolyzing an unstructured peptide contain- further degraded into smaller fragments (Figure 4 and Table ing one Lon cleavage site (24). Through kinetic characteriza- 3). The 45 and 23 kDa fragments correspond to tryptic tion of the ATP- and AMPPNP-dependent peptidase reac- cleavage at R587 which separates the AAA+ chaperone tions, we assigned the observed ATPase dependency to an domain from the peptidase domain in Lon. Tryptic cleavage energy-coupled peptide translocation step similar to ones of the 67 kDa fragment at R489 gave a 35 kDa fragment found in other ATP-dependent proteases (23). To further test which contained the SSD domain and peptidase domain. The this hypothesis, we compared the steady-state kinetic pa- 35 kDa fragment was further cleaved by trypsin to separate rameters of NTP-dependent S3 cleavage and NTP hydrolysis the SSD domain (7 kDa) and protease domain (23 kDa). No to evaluate the functional relationship between nucleotide fragment corresponding to the ATPase domain alone was binding and hydrolysis with peptide cleavage. Since CTP, detected. This could be attributed to rapid degradation of GTP, and UTP also activate the proteolytic activity of Lon the ATPase domain by trypsin due to its relatively open and are hydrolyzed during the reaction (34), they were used conformation as observed in the non-nucleotide-bound HslU as probes to study the mechanism by which Lon transduces (27, 38-40). chemical energy generated from ATP hydrolysis into the Modeling the Nucleotide Base Binding Site in Lon Using movement of the peptide substrate into the protease chamber. HslU as a Study Model. Because of the architectural At saturating nucleotide concentrations, the predominant Conformational Change in Lon and Enzyme Activities Biochemistry, Vol. 43, No. 23, 2004 7439
FIGURE 5: Modeling the adenine binding site of Lon based upon the structure of HslU bound to ADP (PDB entry 1HT2). (A) The structure of a monomeric HslU containing residues 16-66 is shown in blue. The structure of ADP is shown bound to the top of the cleft. (B) The N6 amino group in adenine functions as a hydrogen bond donor that can interact with the carbonyl oxygen of Ile18 and Val61 in HslU. (C) The sequences corresponding to the Walker A motif of Lon and HslU are highly conserved. This region was used as a reference point to align Ala316-Leu365 of Lon with Ile17-Ile66 of HslU. The residues that are anticipated to interact with the N6 amino group of adenine are highlighted in green and yellow. (D) On the basis of the sequence alignment, the carbonyl oxygens of Val360 and Gln317 in Lon are proposed to form hydrogen bonds with the N6 amino group in adenine. enzyme form is the Lon-NTP form, and thus, the kinetic is 10 µM, which is similar to the Kib for ATP (7.4 µM; Table parameters associated with S3 cleavage should be related to 1) but is lower than those obtained for CTP, GTP, and UTP the effect of nucleotide binding and/or hydrolysis. In the (73, 250, and 389 µM, respectively; Table 1). The limited absence of detailed structural information revealing the tryptic digestion patterns of ATP- and AMPPNP-protected different functional states of Lon, we utilized steady-state Lon also show that both nucleotides induce a more compact enzyme kinetic techniques to demonstrate that at least one conformation in the enzyme than CTP, GTP, and UTP enzyme form of Lon functions to transduce the chemical (Figure 3A,B). Collectively, these data indicate that while energy liberated from ATP hydrolysis to stimulate the AMPPNP is not hydrolyzed by Lon, it still binds to Lon in catalytic turnover of peptide cleavage. This provides evidence a manner similar to that of ATP. Since AMPPNP induces for the existence of a peptide translocation step in Lon. the same conformational change in Lon as ATP or ADP, as The kinetic parameters summarized in Tables 1 and 2 judged by the limited tryptic digestion study (Figure 3B), collectively indicate that the kcat of peptide hydrolysis is the lower kcat for the AMPPNP-mediated S3 hydrolysis partially coupled with energy transduction, as ATP, being reaction is most likely due to the lack of energy transduction the most effective peptidase activator, is the most sensitive associated with nucleotide hydrolysis rather than the absence to peptide-stimulated hydrolysis. As seen in Table 1, the kcat of an adenine-specific conformational change. This result, of S3 cleavage is 9 and 4.2 s-1 for ATP and CTP, in conjunction with our previous observation that AMPPCP respectively, whereas for GTP and UTP, the kcat of S3 failed to support peptide cleavage (24), provides quantitative cleavage is within the range of 1.7-1.9 s-1. Despite the evidence of the fact that at least one enzyme form of Lon variation in kcat,S3 found among the different NTPs, the kcat,S3 couples NTP hydrolysis with peptide cleavage. values obtained for these nucleotides are higher than that Although all the NTPs generate a kcat value for S3 cleavage obtained for AMPPNP (Table 1) (23). Since AMPPNP is a that is higher than that of the nonhydrolyzable analogue nonhydrolyzable analogue of ATP, it is used to evaluate the AMPPNP, the kcat,S3 values for the NTPs differ from one functional role of ATP binding in Lon catalysis. Comparing another. This difference in kcat,S3 loosely follows the ranking the relative affinity of ATP with that of AMPPNP in Table of the relative affinities of Lon for the nucleotides (Kib and 1 reveals that Lon binds to both adenine nucleotides with Kb; Table 1), the magnitude of NTPase enhancement (Table comparable affinities. The Kib of AMPPNP for S3 cleavage 2), and the effectiveness of the nucleotide in protecting Lon 7440 Biochemistry, Vol. 43, No. 23, 2004 Patterson et al. process. Collectively, these data indicate a functional rela- tionship between peptidase activation, enhanced peptide- stimulated ATPase activities, and the nucleotide-induced conformational change in the enzyme. The effective com- munication among these three factors is needed to optimize the peptidase activity of Lon. Using limited tryptic digestion studies, we have demon- strated that a compact structure forms between the three functional domains of Lon, the R/â ATPase domain, the SSD domain (also known as the R-helical domain), and the peptidase domain, in the presence of ATP, ADP, or AMP- PNP, that resists tryptic digestion. These results agree well with the report on the domain organization of nucleotide- bound Lon (35) and HslU/HslV (27). In addition to binding the polypeptide substrate, the SSD domain has been proposed FIGURE 6: Comparison of the structures of adenine and cytidine. to participate in ATP hydrolysis (21, 22, 37). The observed An overlaid view of the structures of adenine and cytidine reveals that the amino groups in the two nucleotide bases are in the spatial 2-fold higher peptide-stimulated ATPase enhancement (3.8; proximity of one another. Adenine is shown in blue and cytidine Table 2) compared to the other peptide-stimulated NTPase in red. enhancement (∼2; Table 2) could therefore be attributed to the weakened interaction between the SSD and the γ-phos- from tryptic digestion (Figure 3A). The order of peptidase phate of the NTP (22, 37). Although our tryptic digestion activation is as follows: ATP > CTP > GTP > UTP. This results show that adenine nucleotides induce a closed order parallels the ability of the respective NTPs to protect conformation in Lon, the adenine-specific conformational the 67 kDa fragment of Lon from tryptic digestion: ATP > change alone cannot support peptide cleavage. For example, CTP > GTP ∼ UTP. According to Table 1, peptide AMPPNP is bound to Lon with an affinity comparable to hydrolysis by Lon is more efficient in the presence of ATP that of ATP and induces a compact conformation. The kcat,S3 and CTP, which taken together with the limited tryptic of AMPPNP, however, is lower than the values obtained for digestion data, suggests that there is an enzyme form in Lon the hydrolyzable nucleotides that bind to Lon with lower that exhibits higher selectivity in transducing the energy affinities (Table 1), which suggests that nucleotide hydrolysis liberated from ATP hydrolysis into the activation of peptidase also plays a role in activating peptide hydrolysis. This activity. argument is further supported by the tryptic digestion results To explain why CTP is a better peptidase activator than showing that ADP induces a compact conformation in Lon GTP and UTP, we constructed a structural model for the that is slightly distinguishable from the AMPPNP-bound nucleotide base binding site of Lon based upon the crystal enzyme form (Figure 3B). structure of HslU bound to dADP [Figure 5 (27, 39, 40)]. Using the structural changes associated with binding of Figure 5D shows that the N6 amino group in adenine serves HslU to an adenine nucleotide as a template, we incorporated as a hydrogen bond donor that could potentially interact with the data obtained in this and previous studies to discuss the the carbonyl oxygen of Gln317 and Val360 in Lon. Accord- role of ATP binding and hydrolysis in mediating energy ing to this model, nucleotides capable of forming similar transduction and peptide hydrolysis in Lon. This model contacts with Lon should mimic the functions of the N6 features an ATPase-dependent peptide translocation and amino group in ATP, which in this case is the induction of protease mechanism similar to that proposed for HslUV. The a closed enzyme form compared to unbound Lon. Figure 6 binding and hydrolysis of ATP induce a series of confor- overlays the structure of adenosine and cytosine, which mational changes in the enzyme that render peptide substrate shows that both amino groups are in the spatial proximity access to the central protease cavity. Figure 7 depicts a of one another. Constrained energy minimization of ATP simplified model of oligomeric Lon containing only two of and CTP performed by Pate et al. showed that the two amino the subunits which face one another to form a central cavity. groups were 0.7 Å apart (41). Given the spatial similarity of It should be noted that the other monomers are omitted for the two amino groups in the respective nucleotides, it is clarity. Furthermore, our current model assumes that each conceivable that CTP could make contacts with Lon similar monomeric subunit is functional as an ATP-dependent to those of adenine and induce a relatively closed confor- protease, and the interaction among the subunits may invoke mational change in Lon compared to GTP and UTP. CTP, positive cooperativity in enzyme catalysis that is character- however, is not a perfect mimic of ATP, as its kcat,S3 is 2-fold ized by a Hill coefficient of >1. However, further experi- lower than that obtained for ATP and it is less effective than ments should be conducted to further evaluate this cooper- any of the adenine nucleotides at protecting Lon from tryptic ativity. digestion. The increase in Kb and Kib for CTP (Table 1) is In the absence of nucleotide, Lon adopts a relatively loose probably caused by the loss of interaction between Lon and conformation in which the ATPase, SSD, and peptidase other parts of the heterocyclic base in the nucleotide. These domains are susceptible to tryptic digestion and yield the results suggest that the kcat,S3 value is related to the compact- fragments shown in Table 3 (I in Figure 7). These loose ness of the nucleotide-induced conformation in the enzyme. conformations among the subunits form a narrow opening CTP, GTP, and UTP have lower kcat,S3 values than ATP to the central cavity in Lon, thereby hindering the access of because they induce a less compact conformation in Lon, the peptide substrate to the protease chamber. The binding which reduces the effectiveness of the energy transduction of ATP to the AAA+ motif in Lon induces a more compact Conformational Change in Lon and Enzyme Activities Biochemistry, Vol. 43, No. 23, 2004 7441
FIGURE 7: Model proposed for the different enzyme forms in Lon that couple ATP binding and hydrolysis in activating peptide hydrolysis. This model is proposed on the basis of the structural similarities shared by Lon and HslU. An ATPase-dependent peptide translocation is proposed in this model (see the Discussion). The R/â-subdomain and the R-helical subdomain of the AAA+ motif are in blue and green, respectively. The protease domain is in white. The protein domains and subdomains are connected by flexible linkers. conformation within the enzyme, which is more resistant to position should be poor activators of Lon. The former tryptic digestion than form I (II in Figure 7). This structural prediction could be readily tested by comparing the pre- rearrangement opens the pore leading into the central cavity, steady-state kinetics of peptide and ATP hydrolysis. One which renders the peptide access to the protease chamber anticipates that the rate constant for ATP hydrolysis will be and facilitates the transduction of chemical energy generated higher than that obtained for peptide cleavage. This endeavor from ATP hydrolysis to the active site of the protease where is currently being examined in our laboratory using the same peptide cleavage occurs (form III in Figure 7). Although this defined peptide substrate. Elucidating the timing of ATP step is speculative in Lon, the coupling of ATP binding and binding and hydrolysis should expand our mechanistic hydrolysis to energy transduction through conformational understanding of Lon. With regard to evaluating the func- changes in enzymes has been reported in molecular motors tional role of the hydrogen bond interaction between Lon such as myosin (42) and HslUV (27). In our studies, the and the nucleotide base in adenine, we propose a detailed existence of this step is supported by the observed correlation investigation utilizing non-natural nucleotides. These nucleo- between peptide-stimulated NTPase activity and the higher tides lacking any hydrogen bond donating properties in the kcat,S3 obtained for the hydrolyzable nucleotides compared heterocyclic bases will be used as probes to evaluate the to those of nonhydrolyzable ATP analogues. The mechanism functional role of nucleotide binding in Lon. The design and by which Lon returns to the free enzyme form is not clear. synthesis of a series of non-natural nucleotides is currently However, on the basis of the detection of a noncompetitive underway to further address this issue. inhibition pattern for one of the hydrolyzed peptide products against the S3 substrate, we propose that Lon isomerizes to ACKNOWLEDGMENT another form during peptide cleavage (23). Since the hydrolyzed peptide product is a truncated version of the S3 We thank Dr. Anthony Berdis, Hilary Frase, and Jonathon substrate, it is anticipated to act as a competitive inhibitor Ipsaro for their assistance in the preparation of the manu- against S3. The noncompetitive inhibition pattern seems to script. suggest that Lon adopts a different form upon peptide REFERENCES cleavage and the postcatalytic enzyme form is specific only for the hydrolyzed peptide product but not for S3. This 1. Chung, C. H., and Goldberg, A. L. (1981) The product of the lon enzyme isomerization step could account for the lack of (capR) gene in Escherichia coli is the ATP-dependent protease, - additional S3 degradation in form IV (Figure 7). protease La, Proc. Natl. Acad. Sci. U.S.A. 78, 4931 4935. The model proposed in Figure 7 predicts that ATP 2. Gottesman, S., Gottesman, M. E., Shaw, J. E., and Pearson, M. L. (1981) Protein degradation in E. coli: the lon mutation and hydrolysis should occur before peptide cleavage, and adenine bacteriophage lambda N and cll protein stability, Cell. 24, 225- nucleotides lacking hydrogen donating properties at the C6 233. 7442 Biochemistry, Vol. 43, No. 23, 2004 Patterson et al.
3. Goldberg, A. L., Moerschell, R. P., Chung, C. H., and Maurizi, 23. Thomas-Wohlever, J., and Lee, I. (2002) Kinetic characterization M. R. (1994) ATP-dependent protease La (lon) from Escherichia of the peptidase activity of Escherichia coli Lon reveals the coli, Methods Enzymol. 244, 350-375. mechanistic similarities in ATP-dependent hydrolysis of peptide 4. Gottesman, S., Maurizi, M. R. (1992) Regulation by Proteolysis: and protein substrates, Biochemistry 41, 9418-9425. Enegry-Dependent Proteases and Their Targets, Microbiol. ReV. 24. Lee, I., and Berdis, A. J. (2001) Adenosine triphosphate-dependent 56, 592-621. degradation of a fluorescent lambda N substrate mimic by Lon 5. Gottesman, S. (1996) Proteases and their targets in Escherichia protease, Anal. Biochem. 291,74-83. V - coli, Annu. Re . Genet. 30, 465 506. 25. Hoskins, J. R., Pak, M., Maurizi, M. R., and Wickner, S. (1998) 6. Maurizi, M. R. (1992) Proteases and protein degradation in - The role of the ClpA chaperone in proteolysis by ClpAP, Proc. Escherichia coli, Experientia 48, 178 201. Natl. Acad. Sci. U.S.A. 95, 12135-12140. 7. Charrette, M., Henderson, G. W., and Markovitz, A. (1981) ATP hydrolysis-dependent activity of the lon (capR) protein of E. coli 26. Reid, B. G., Fenton, W. A., Horwich, A. L., Weber-Ban, E. U. K12, Proc. Natl. Acad. Sci. U.S.A. 78, 4728-4732. (2001) ClpA mediates directional translocation of substrate 8. Schoemaker, J. M., Gayda, R. C., Markovitz, A. (1984) Regulation proteins into the ClpP protease, Proc. Natl. Acad. Sci. U.S.A. 98, - of cell division in Escherichia coli: SOS induction and cellular 3768 3772. location of the sulA protein, a key to lon-associated filamentation 27. Wang, J., Song, J. J., Seong, I. S., Franklin, M. C., Kamtekar, S., and death, J. Bacteriol. 158, 551-561. Eom, S. H., and Chung, C. H. (2001) Nucleotide-dependent 9. Goff, S. A., and Goldberg, A. L. (1985) Production of abnormal conformational changes in a protease-associated ATPase HsIU, proteins in E. coli stimulates transcription of lon and other heat Structure 9, 1107-1116. shock genes, Cell 41, 587-595. 28. Gilbert, S. P., and Mackey, A. T. (2000) Kinetics: a tool to study 10. Chin, D. T., Goff, S. A., Webster, T., Smith, T., and Goldberg, molecular motors, Methods 22, 337-354. A. L. (1988) Sequence of the lon gene in Escherichia coli.A 29. Lanzetta, P. A., Alvarez, L. J., Reinach, P. S., and Candia, O. A. heat-shock gene which encodes the ATP-dependent protease La, - (1979) An improved assay for nanomole amounts of inorganic J. Biol. Chem. 263, 11718 11728. phosphate, Anal. Biochem. 100,95-97. 11. Amerik, A., Chistiakov, L. G., Ostroumova, N. I., Gurevich, A. I., and Antonov, V. K. (1988) [Cloning, expression and structure 30. Copeland, R. A. (1996) Enzymes. A practical introduction to of the functionally active shortened lon gene in Escherichia coli], structure, mechanism and data analysis, VCH Publishers, New Bioorg. Khim. 14, 408-411. York. 12. Botos, I., Melnikov, E. E., Cherry, S., Tropea, J. E., Khalatova, 31. Menon, A. S., and Goldberg, A. L. (1987) Protein substrates A. G., Rasulova, F., Dauter, Z., Maurizi, M. R., Rotanova, T. V., activate the ATP-dependent protease La by promoting nucleotide Wlodawer, A., and Gustchina, A. (2004) The catalytic domain of binding and release of bound ADP, J. Biol. Chem. 262, 14929- Escherichia coli Lon protease has a unique fold and a Ser-Lys 14934. dyad in the active site, J. Biol. Chem. 279, 8140-8148. 32. Menon, A. S., and Goldberg, A. L. (1987) Binding of nucleotides 13. Botos, I., Melnikov, E. E., Cherry, S., Tropea, J. E., Khalatova, to the ATP-dependent protease La from Escherichia coli, J. Biol. A. G., Rasulova, F., Dauter, Z., Maurizi, M. R., Rotanova, T. V., Chem. 262, 14921-14928. - Wlodawer, A., and Gustchina, A. (2004) J. Struct. Biol. 146, 113 33. Waxman, L., and Goldberg, A. L. (1986) Selectivity of intracellular 122. proteolysis: protein substrates activate the ATP-dependent pro- 14. Rohrwild, M., Coux, O., Huang, H. C., Moerschell, R. P., Yoo, tease (La), Science 232, 500-503. S. J., Seol, J. H., Chung, C. H., and Goldberg, A. L. (1996) HslV- 34. Waxman, L., and Goldberg, A. L. (1982) Protease La from HslU: A novel ATP-dependent protease complex in Escherichia Escherichia coli hydrolyzes ATP and proteins in a linked fashion, coli related to the eukaryotic proteasome, Proc. Natl. Acad. Sci. Proc. Natl. Acad. Sci. U.S.A. 79, 4883-4887. U.S.A. 93, 5808-5813. 15. Rohrwild, M., Pfeifer, G., Santarius, U., Muller, S. A., Huang, H. 35. Vasilyeva, O. V., Kolygo, K. B., Leonova, Y. F., Potapenko, N. C., Engel, A., Baumeister, W., and Goldberg, A. L. (1997) The A., and Ovchinnikova, T. V. (2002) Domain structure and ATP- ATP-dependent HslVU protease from Escherichia coli is a four- induced conformational changes in Escherichia coli protease Lon - ring structure resembling the proteasome, Nat. Struct. Biol. 4, revealed by limited proteolysis and autolysis, FEBS Lett. 526,66 133-139. 70. 16. Goldberg, A. L., Akopian, T. N., Kisselev, A. F., Lee, D. H., 36. Smith, C. K., Baker, T. A., and Sauer, R. T. (1999) Lon and Clp Rohrwild, M. (1997) New insights into the mechanisms and family proteases and chaperones share homologous substrate- importance of the proteasome in intracellular protein degradation, recognition domains, Proc. Natl. Acad. Sci. U.S.A. 96, 6678- Biol. Chem. 378, 131-140. 6682. 17. Yoo, S. J., Seol, J. H., Shin, D. H., Rohrwild, M., Kang, M. S., 37. Wickner, S., and Maurizi, M. R. (1999) Here’s the hook: similar Tanaka, K., Goldberg, A. L., and Chung, C. H. (1996) Purification substrate binding sites in the chaperone domains of Clp and Lon, and characterization of the heat shock proteins HslV and HslU Proc. Natl. Acad. Sci. U.S.A. 96, 8318-8320. that form a new ATP-dependent protease in Escherichia coli, J. - 38. Bochtler, M., Hartmann, C., Song, H. K., Bourenkov, G. P., Biol. Chem. 271, 14035 14040. Bartunik, H. D., and Huber, R. (2000) The structures of HslU 18. Lee, C., Schwartz, M. P., Prakash, S., Iwakura, M., and Matous- and the ATP-dependent protease HslU-HslV, Nature 403, 800- chek, A. (2001) ATP-dependent proteases degrade their substrates 805. by processively unraveling them from the degradation signal, Mol. Cell 7, 627-637. 39. Wang, J., Song, J. J., Franklin, M. C., Kamtekar, S., Im, Y. J., 19. Reid, B. G., Fenton, W. A., Horwich, A. L., and Weber-Ban, E. Rho, S. H., Seong, I. S., Lee, C. S., Chung, C. H., and Eom, S. U. (2001) ClpA mediates directional translocation of substrate H. (2001) Crystal structures of the HslVU peptidase-ATPase complex reveal an ATP-dependent proteolysis mechanism, proteins into the ClpP protease, Proc. Natl. Acad. Sci. U.S.A. 98, - 3768-3772. Structure 9, 177 184. 20. Ishikawa, T., Beuron, F., Kessel, M., Wickner, S., Maurizi, M. 40. Sousa, M. C., Trame, C. B., Tsuruta, H., Wilbanks, S. M., Reddy, R., and Steven, A. C. (2001) Translocation pathway of protein V. S., and McKay, D. B. (2000) Crystal and solution structures substrates in ClpAP protease, Proc. Natl. Acad. Sci. U.S.A. 98, of an HslUV protease-chaperone complex, Cell 103, 633-643. 4328-4333. 41. Pate, E., Franks-Skiba, K., White, H., and Cooke, R. (1993) The 21. Neuwald, A. F., Aravind, L., Spouge, J. L., and Koonin, E. V. use of differing nucleotides to investigate cross-bridge kinetics, (1999) AAA+: A class of chaperone-like ATPases associated J. Biol. Chem. 268, 10046-10053. with the assembly, operation, and disassembly of protein com- 42. Park, S., Ajtai, K., and Burghardt, T. P. (1997) Mechanism for plexes, Genome Res. 9,27-43. + Coupling Free Energy in ATPase to the Myosin Active Site, 22. Ogura, T., and Wilkinson, A. J. (2001) AAA superfamily Biochemistry 36, 3368-3372. ATPases: common structuresdiverse function, Genes Cells 6, 575-597. BI036165C Biochemistry 2005, 44, 1671-1682 1671
Monitoring the Timing of ATP Hydrolysis with Activation of Peptide Cleavage in Escherichia coli Lon by Transient Kinetics† Diana Vineyard,‡ Jessica Patterson-Ward,‡ Anthony J. Berdis,§ and Irene Lee*,‡ Department of Chemistry and Department of Pharmacology, Case Western ReserVe UniVersity, CleVeland, Ohio 44106 ReceiVed June 30, 2004; ReVised Manuscript ReceiVed October 14, 2004
ABSTRACT: Escherichia coli Lon, also known as protease La, is an oligomeric ATP-dependent protease, which functions to degrade damaged and certain short-lived regulatory proteins in the cell. To investigate the kinetic mechanism of E. coli Lon protease, we performed the first pre-steady-state kinetic characterization of the ATPase and peptidase activities of this enzyme. Using rapid quench-flow and fluorescence stopped-flow spectroscopy techniques, we demonstrated that ATP hydrolysis occurs before peptide cleavage, with the former reaction displaying a burst and the latter displaying a lag in product production. The detection of burst kinetics in ATP hydrolysis is indicative of a step after nucleotide hydrolysis being rate-limiting in ATPase turnover. At saturating substrate concentrations, the lag rate constant for peptide cleavage is comparable to the kcat of ATPase, indicating that two hydrolytic processes are coordinated during the first enzyme turnover. The involvement of subunit interaction during enzyme catalysis was detected as positive cooperativity in the binding and hydrolysis of substrates, as well as apparent asymmetry in the ATPase activity in Lon. When our data are taken together, they are consistent with a reaction model in which ATP hydrolysis is used to generate an active enzyme form that hydrolyzes peptide.
Escherichia coli Lon, also known as protease La, is an The coupling of ATP binding and hydrolysis to peptide oligomeric ATP-dependent protease, which functions to cleavage is a key question in mechanistic studies of Lon degrade damaged and certain short-lived regulatory proteins protease. On the basis that ADP inhibits the ATPase and in the cell (1-10). In Vitro this enzyme is capable of the protease activities and nonhydrolyzable ATP analogues degrading polypeptides in the presence of certain nonhy- support peptide cleavage, an ADP/ATP exchange mechanism drolyzable ATP analogues. However, optimal protein deg- has been proposed (3, 6, 18-21). However, it is not known radation does require ATP hydrolysis (3, 6, 11). Lon whether peptide cleavage occurs before or after ATP represents one of the simplest forms of ATP-dependent hydrolysis. To evaluate the functional relationship between proteases because both the ATPase and the protease domains ATP hydrolysis and peptide cleavage, we developed a are located within a single monomeric subunit (12, 13). fluorescent peptide substrate whose degradation by Lon Although a complete crystal structure of active Lon has exhibits ATPase dependency as observed in the endogenous yet to be determined, some structural clues to the mechanism protein substrates of Lon (11, 22). This defined model peptide of Lon are available. The crystal structure of a proteolytically substrate containing 10% of the fluorescent peptide Y(NO2)- inactive Lon mutant lacking an ATPase domain reveals that RGITCSGRQK(Abz) and 90% S2 (S3)1 and its nonfluores- Lon is hexameric and contains a central cavity commonly cent analogue (S2) were previously used as simplified mimics found in the ATPase and protease subunits of other ATP- of the E. coli Lon substrate, the λ N protein, to demonstrate dependent proteases (14). Additionally, the structure of the that kcat but not Km of the peptide cleavage is dependent on R subdomain of the ATPase subunit of Lon has been ATP hydrolysis (11). In addition, the kcat of peptide cleavage determined (15). This monomeric subdomain constitutes the is higher when hydrolyzable nucleotides rather than nonhy- last 25% of the carboxyl-terminal region of the ATPase drolyzable ATP analogues are used as activators. During domain and lacks the conserved Walker motifs found in peptide cleavage, the nucleotide hydrolysis activity of Lon ATP-binding proteins. On the basis of these reports, Lon appears to be structurally similar to other ATP-dependent 1 Abbreviations: AMPPNP, adenylyl 5-imidodiphosphate; DTT, proteases such as the bacterial homologue of the proteasome, dithiothreitol; Abz, anthranilamide; Bz, benzoic acid amide; NO2, nitro; HBTU, O-benzotriazole-N,N,N′,N′-tetramethyluroniumhexafluorophos- HslUV (16, 17). Whether and how ATP binding and phate; Tris, 2-amino-2-(hydroxymethyl)-1,3-propanediol; HEPES, N-2- hydrolysis are coupled to peptide cleavage are key inquiries hydroxyethylpiperazine-N′-ethanesulphonic acid; Mg(OAc)2, magne- in understanding the mechanism of the enzyme. sium acetate; KOAc, potassium acetate; PEI-cellulose, polyethylene- imine-cellulose; λ N, also known as the λ N protein, a λ phage protein that allows E. coli RNA polymerase to transcribe through termination † This work was support by the NIH Grant GM067172. signals in the early operons of the phage; S2, a nonfluorescent analogue * To whom correspondence should be addressed. Telephone: 216- of S3 that is cleaved by Lon in the same manner as S3:YRGITCS- 368-6001. Fax: 216-368-3006. E-mail: [email protected]. GRQK(benzoic acid amide) (Bz); S3, a mixed peptide substrate ‡ Department of Chemistry. containing 10% of the fluorescent peptide Y(NO2)RGITCSGRQK(Abz) § Department of Pharmacology. and 90% S2. 10.1021/bi048618z CCC: $30.25 © 2005 American Chemical Society Published on Web 01/11/2005 1672 Biochemistry, Vol. 44, No. 5, 2005 Vineyard et al. is also elevated, suggesting a coupling between the two performed on a KinTek Stopped Flow controlled by the data hydrolytic activities (23). Because both peptides were collection software Stop Flow version 7.50 â. The sample degraded by E. coli Lon in an identical manner, the S3 and syringes were maintained at 37 °C by a circulating water S2 peptide could be used as substrates to monitor the bath. Syringe A contained 5 µM Lon monomer, with variable peptidase and ATPase activity of Lon, respectively (11). concentrations of the peptide substrate S3 (from 5 to 500 While initial velocity and product inhibition studies µM), 5 mM Mg(OAc)2, 50 mM Tris-HCl (pH 8.1), and 5 identified which enzyme forms predominate under certain mM DTT. Syringe B contained varying ATP (1-500 µM). reaction conditions, they could not evaluate whether ATP Peptide cleavage was detected by an increase in fluorescence hydrolysis precedes or follows peptide cleavage (11). This (excitation of 320 nm and emission of 420 nm) following question, however, can be answered by measuring the rates rapid mixing of syringe contents in the sample cell. The of ATP and peptide hydrolysis in the pre-steady state. If ATP baseline fluorescence was normalized to zero, and the data hydrolysis occurs before peptide cleavage, its rate constant shown are a result of averaging at least four traces. The should be higher than that for peptide cleavage. In this study, concentration of hydrolyzed peptide was calibrated by we utilized rapid chemical-quench-flow and fluorescence determining the maximum fluorescence generated per mi- stopped-flow techniques to monitor pre-steady-state time cromolar peptide because of complete digestion by trypsin courses of the two reactions to correlate the timing of ATP under identical reaction conditions in the stopped-flow hydrolysis with peptide cleavage, thereby establishing the apparatus. The averaged time courses were fit with eq 1 sequence of events occurring along the reaction pathway. - ) klagt +V + In addition, despite its existence as a homooligomer, Lon Y A exp sst C (1) exhibits two different affinities for ATP (Kd ) 10 µM and Kd < 1 µM) (19); the mechanism by which these two kinds where t is time in seconds, Y is the concentration of of ATPase sites affect proteolysis are not known. Therefore, hydrolyzed peptide S3 in micromolar, A is the amplitude of by determining the pre-steady-state kinetics of peptide and the reaction, klag is the pre-steady-state rate constant in per V ATP hydrolysis at low (<10 µM) versus high (100 µM) ATP seconds, ss is the steady-state rate in units of micromolar V concentrations, we can evaluate how the binding and/or product per second, and C is the endpoint. The ss value can hydrolysis of the nucleotide at the respective ATPase site be converted to a first-order rate constant (kss in the unit of coordinates with the first turnover of peptide cleavage. per seconds) by division with the enzyme concentration (24). Equation 1 is the general function that quantifies a biphasic MATERIALS AND METHODS time course. When Y ) 0att ) 0, C )-A and eq 1 becomes
- Materials. ATP was purchased from Sigma, whereas )- + klagt +V Y A A exp sst [R-32P]ATP was purchased from Perkin-Elmer or ICN Bio- medical. Fmoc-protected amino acids, Boc-anthranilamide such that (Abz), Fmoc-protected Lys Wang resin, and O-benzotriazole- - ′ ′ )- - klagt +V N,N,N ,N -tetramethyluroniumhexafluorophosphate (HBTU) Y A(1 exp ) sst were purchased from Advanced ChemTech and NovaBio- chem. 2-Amino-2-(hydroxymethyl)-1,3-propanediol (Tris), When A is defined as Vss -Vi/klag, eq 1 becomes equivalent N-2-hydroxyethylpiperazine-N′-ethanesulphonic acid (HEPES), to the equation defining hysteresis (25, 26), where Vi is the and polyethyleneimine-cellulose (PEI-cellulose) TLC plates rate corresponding to the initial phase of the time course. were purchased from Fisher. Data Processing. Plots displaying sigmoidal behavior were General Methods. Peptide synthesis and protein purifica- fit with eq 2 where k is the observed rate constant of the tion procedures were performed as described previously (22). n All enzyme concentrations are reported as Lon monomer k [S] k ) max (2) concentrations. All reagents are reported as final concentra- K′ + [S]n tions.
Acid-Quench Peptidase Assay. Assays were performed at reaction being monitored, kmax is the maximum rate constant 37 °C and contained 50 mM Tris-HCl (pH 8.1), 5 mM (referred to as kcat,S3 for kss,S3 data, klag,S3 for klag data, and magnesium acetate (Mg(OAc)2), 5 mM dithiothreitol (DTT), kcat,ATP for kss,ATP data), [S] is the variable substrate, K′ is the 800 µM S3 peptide substrate (containing 50% of the Michaelis constant for S, and n is the Hill coefficient. The fluorescent peptide, S1, and 50% S2, the nonfluorescent Ks is calculated from the relationship log K′ ) n log Ks, analogue of S1 that is degraded by Lon in an identical where Ks is the [S] required to obtain 50% of the maximal manner as S1), and 1 µM E. coli Lon monomer. The reaction rate constant of the reaction. Plots displaying hyperbolic was initiated by the addition of 500 µM ATP, and 10 µL behavior, i.e., when the Hill coefficient in eq 2 becomes 1, aliquots were quenched with 58 µL of 0.5 N HCl at 0, 0.5, were fit using eq 3 where k is the observed rate constant, 1, 2, 3, and 5 min. After trichloroacetic acid precipitation, k [S] which removed Lon, the reaction mixtures were neutralized ) max with 1 M Tris/2 N NaOH to pH 8 and the fluorescence k + (3) Ks [S] (excitation of 320 nm and emission of 420 nm) of the solution was measured using a Fluoromax 3 spectrofluorim- kmax is the maximum rate constant (referred to as kcat,S3 for eter (Horiba Group). kss,S3 data, klag,S3 for klag data, and kcat,ATP for kss,ATP data), [S] Pre-steady-State Time Courses of S3 CleaVage by Fluo- is the variable substrate, and Ks is the [S] required to obtain rescent Stopped Flow. Pre-steady-state experiments were 50% of the maximal rate constant of the reaction. The Ks Kinetic Mechanism of ATP Hydrolysis and Peptide Cleavage Biochemistry, Vol. 44, No. 5, 2005 1673 value was referred to as Km,S3 when S3 hydrolysis was monitored at varying S3 concentrations. Alternatively, the Ks value was referred to as Km,ATP when ATP hydrolysis was monitored at varying ATP concentrations. Chemical-Quench ATPase ActiVity Assays. The pre-steady- state time courses for ATP hydrolysis were measured using a rapid chemical-quench-flow instrument from KinTek Corporation. All solutions were made in 50 mM HEPES buffer at pH 8.1, 5 mM DTT, 5 mM Mg(OAc)2, and 75 mM potassium acetate (KOAc). A 15 µL buffered solution of 5 µM Lon monomer, with and without 500 µMS2or10µM casein, was rapidly mixed with a 15 µL buffered solution of ATP containing 0.01% of [R-32P]ATP at 37 °C for varying FIGURE 1: Detection of ATP-dependent S3 cleavage by a discon- times (0-3 s) before quenching with 0.5 N formic acid and tinuous acid-quench assay. Lon (1 µM monomer) was incubated then extracted with 200 µL of phenol/chloroform/isoamyl with 800 µM S3, and reaction aliquots were quenched at the alcohol at pH 6.7 (25:24:1). A 3 µL aliquot of the aqueous indicated times. The fluorescence signals associated with peptide solution was spotted directly onto a PEI-cellulose TLC plate cleavage were measured and plotted against their corresponding × reaction time points. The experiment was performed 2 times, and (10 20 cm), and the plate was developed in 0.75 M the averaged data were plotted. The data were fit with eq 1 to yield potassium phosphate buffer (pH 3.4) to separate ADP from a lag rate constant of 1.4 s-1. ATP. The relative amount of radiolabeled ADP and ATP at each time point was quantified by a Cyclone Phosphor as described by Gilbert and Mackey (24). Each spot was imager (Perkin-Elmer Life Science). To compensate for the washed with 10 µL of cold buffer and dried under vacuum slight variations in spotting volume, the concentration of the for 30 min. In the absence of vacuum, the nitrocellulose was ADP product obtained at each time point was corrected for spotted with 2 µL of each reaction and then air-dried. The using an internal reference as shown in eq 4. All assays were radioactive counts at each spot were quantified by Phospho- rImaging. ADP ) dlu [ADP] ( + )[ATP] (4) ATPdlu ADPdlu RESULTS Pre-Steady-State Kinetic Analysis of S3 CleaVage. Previ- performed at least 3 times, and the average of those traces ously, we demonstrated that the pre-steady-state time course was used for data analysis. The burst amplitudes and burst of S3 cleavage could be monitored by stopped-flow fluo- rates were determined by fitting the kobs data from 0 to 400 rescence spectroscopy (11). Using ATP as the activator, we ms with eq 5. where t is time in seconds, Y is [ADP] in detected lag kinetics in S3 cleavage. The lag phase is -k t lengthened by 7-fold when adenylyl 5-imidodiphosphate Y ) A exp burst + C (5) (AMPPNP), a non-hydrolyzable ATP analogue, was used micromolar, A is the burst amplitude in micromolar, kburst is as an activator. Because the cleavage of S3 peptide generates the burst rate constant in per seconds, and C is the end point. a fluorescent signal as a result of the separation of the The observed steady-state rate constants (kss,ATP) were fluorescent quencher, 3-nitrotyrosine, from the fluorescent determined by fitting the data from 600 ms to 1.8 s with the donor, anthranylamide, the lag in S3 cleavage could be linear function, Y ) mX + C, where X is time, Y is [ADP]/ attributed to a slow step prior to peptide bond cleavage or [E], m is the observed steady-state rate constant in per to the slow dissociation of the donor from the quencher seconds, and C is the y intercept. Data fitting was ac- because both hydrolyzed peptides remain bound to the active complished using the nonlinear regression program Kaleida- site of the enzyme. To determine if slow dissociation of Graph (Synergy). peptide products caused the lag, we monitored the time Pulse-Chase ATPase ActiVity Assays. The pre-steady-state course of 800 µM S3 cleavage by 1 µM monomeric Lon time courses of ATP hydrolysis were also measured using a using a discontinuous acid-quench assay. An aliquot of the pulse-chase experiment on the rapid quench. Lon ((0.5 mM reaction was quenched with HCl at the times indicated in S2) was rapidly mixed with radiolabeled ATP at 37 °C for Figure 1, and the fluorescence intensity at each time point 0-1.8 s, followed by a 10 mM unlabeled ATP chase for 60 was measured to yield the time course for S3 cleavage. s before quenching with 0.5 N formic acid. The amount of Despite acid denaturation that released hydrolyzed peptides ADP produced at each time point was quantified as described from Lon, the lag phase remained in the fluorescence time in the chemical-quench assay (see above). The burst ampli- course of S3 cleavage. This result indicates that the separation tude (A) and burst rate constant (kburst) were determined from of the hydrolyzed peptide from Lon does not contribute to the time courses by fitting the data from 0 to 400 ms with the observed lag phase of the reaction. Furthermore, a lag eq 5. rate constant of 1.4 ( 0.6 s-1 was obtained by fitting the Filter Binding Assay. A total of 2-5 µLofa35µM stock data with eq 1, which is a general function used to determine of Lon was incubated with 10 µM[R-32P]ATP in 30 µLof the rate constant (k) associated with a single-exponential 50 mM HEPES at pH 8.1, 5 mM Mg(OAc)2, 75 mM KOAc, phase followed by a steady-state phase, and is related to the and 2 mM DTT at 37 °C for 20 min to convert all ATP to hysteresis equation as described in the Materials and ADP. The reactions were then chilled on ice, and 3 µLof Methods. This observed value is 2-fold higher than that the reactions (performed 3 times) was spotted onto a piece obtained previously using stopped-flow spectroscopy (11) of nitrocellulose mounted onto a dot-blot apparatus (BioRad) and may be attributed to a difference in the detection methods 1674 Biochemistry, Vol. 44, No. 5, 2005 Vineyard et al.
FIGURE 3: Steady-state kinetics of ATP-dependent S3 cleavage by 5 µM E. coli Lon monomer. The kss,S3 values were obtained by dividing the steady-state rates of the reactions by [Lon] as described in the Materials and Methods. (A) Steady-state rates of S3 cleavage (Vss) were obtained by fitting the stopped-flow time courses of peptide cleavage at varying [S3] with eq 1 (Materials and Methods). The data presented in this plot were best-fit with eq 2, and the -1 kinetic parameters obtained were kcat,S3 ) 5.5 ( 0.8 s , Km,S3 ) 188 ( 148 µM, and n ) 1.23 ( 0.2 as summarized in Table 1. (B) Steady-state rates of S3 cleavage (Vss) were obtained by fitting the stopped flow-time courses of peptide cleavage at varying [ATP] FIGURE 2: Stopped-flow analysis of ATP-dependent S3 cleavage with eq 1 (Materials and Methods). The data presented in this plot by E. coli Lon. (A) ATP (500 µM) was incubated with 5 µM were best-fit with eq 3, and the kinetic data obtained from the fit ) ( -1 ) ( monomeric Lon in the presence of 5 (1), 10 (O), 25 (]), 50 (×), were kcat,S3 4.2 0.1 s and KATP 9.7 0.9 µMas 100 (+), 200 (4), and 500 (b) µM S3. The inset zooms in A to summarized in Table 1. show the lag for one peptide hydrolyzed per active site. The fluorescence changes associated with peptide cleavage were concentration of S3 (500 µM) and several fixed levels of converted to product concentrations as described in the Materials ATP (1-500 µM, Figure 2B). As shown in parts A and B and Methods. Each time course shown is an average of 4 traces. of Figure 2, a lag in peptide hydrolysis was observed in all The values on the right y axis represent the concentration of peptide hydrolyzed, whereas the values on the left y axis represent the mole of the time courses. The lag rate constants (klag) and the equivalent of peptide digested by each Lon monomer. (B) S3 (500 observed steady-state rate constants of peptide cleavage µM) was digested by 5 µM Lon monomer in the presence of 1 (kss,S3) were obtained by fitting each time course with eq 1. (b),5(0), 10 (]), 50 (×), 100 (+), 200 (4), and 500 (O) µM The plots of k as a function of S3 and ATP concentrations ATP. The inset zooms in B to show the lag for one peptide ss,S3 hydrolyzed per active site. Each time course shown is an average are shown in parts A and B of Figure 3, and the data were of at least 4 traces. The right y axis shows the amount of peptide fit with eqs 2 and 3, respectively, to yield the kcat,S3 (the hydrolyzed, whereas the left x axis shows the mole equivalent of maximum rate constant for peptide cleavage), KmS3, KATP, peptide cleavage by each Lon monomer. and Hill coefficients (n) as summarized in Table 1. The steady-state kinetic parameters determined from this study (a continuous versus discontinuous assay). Despite the slight (using 5 µM enzyme monomer) agree well with those variation in the lag rate constants, both the stopped-flow and determined previously at 125 nM Lon monomer (kcat,S3 ) -1 the acid-quenched reactions displayed lag kinetics, indicating 7.7 s ; KmS3 ) 85 µM, and KATP ) 7.2 µM) (11), indicating that the dissociation of the peptide products from Lon did that these kinetic parameters are independent of the enzyme not contribute to the lag phase. concentration under the conditions examined. Because pre- To determine the stoichiometry of S3 cleavage, we steady-state lag kinetics could be attributed to the binding generated a calibration curve by measuring the fluorescence of substrates at low concentrations being rate-limiting, we changes associated with the complete degradation of known measured the klag of S3 cleavage at increasing peptide or concentrations of S3 by trypsin in a stopped-flow apparatus. ATP concentrations. As shown in parts A and B of Figure This technique allows us to accurately define the amount of 4, the dependence of klag toward S3 and ATP concentrations peptide cleaved during the time courses for S3 cleavage. reaches saturation with a maximum klag,S3 value. The data Experiments were performed at 500 µM ATP with several were best-fit with eqs 2 and 3, respectively, to yield the fixed levels of S3 (5-500 µM, Figure 2A) as well as a fixed kinetic parameters klag,S3, KS3, and KATP as summarized in Kinetic Mechanism of ATP Hydrolysis and Peptide Cleavage Biochemistry, Vol. 44, No. 5, 2005 1675
-1 Table 1: Parameters Obtained from Pre-steady-State Kinetic of 0.76 s determined previously by fitting the fluorescent Characterization of ATP-Dependent S3 Cleavage by E. coli Lon time courses with the hysteresis equation (11), which is a 0.5 mM ATP 0.5 mM S3 specific form of eq 1 (see the Materials and Methods) used vary [S3] vary [ATP] to characterize enzymes responding slowly to a change in -1 a -1 b the ligand concentration (25, 26). At high S3 or ATP kcat,S3 5.5 ( 0.8 s 4.2 ( 0.1 s a Km,S3 188 ( 148 µM NA concentrations, substrate binding no longer limits the rate a c n 1.23 ( 0.2 (1.7 ( 0.2) NA of the first peptidase turnover, and thus, the maximum klag,S3 ( -1 c ( -1 d klag,S3 0.88 0.07 s 1.14 0.06 s value directly reflects the rate constant for the build-up of ( c KS3 67 34 µM NA at least one reaction intermediate that leads to peptide K NA 9.7 ( 0.9b µM (7.2 ( 1.9 µM)c ATP hydrolysis at the active site of the enzyme. Furthermore, the a These values were obtained by fitting the steady-state kinetic data fit of the k dependence on S3 or ATP concentrations to shown in Figure 3A with eq 2. b These values were obtained by fitting lag,S3 the steady-state kinetic data shown in Figure 3B with eq 3. c These eq 2 (for S3) or eq 3 (for ATP) provided the apparent Kd values were obtained by fitting the lag rate constants obtained from values of the respective substrate (KS3 and KATP in Table 1). the stopped-flow time courses of S3 cleavage shown in Figure 4A with The KS3 value (67 ( 34 µM) is comparable to the Km of S3 eq 2. d NA ) not available. cleavage (188 ( 148 µM), whereas KATP (9.7 ( 0.9 µM) is in close agreement with the weak affinity ATPase site in E. coli Lon, which is 10 µM(17). Chemical-Quench Analysis of ATP Hydrolysis. Because the kcat of S3 cleavage is enhanced by nucleotide hydrolysis (11, 21) and the kcat of ATP hydrolysis is stimulated by the peptide or protein substrate (3, 21, 23), it is very likely that the two hydrolytic processes are coupled through a common enzyme intermediate. To begin identifying such an interme- diate, we measured the time courses of the first turnover of ATP hydrolysis in the absence and presence of the peptide or protein substrate using the rapid chemical-quench tech- nique. Lon (5 µM monomer) was preincubated with 500 µM S2, the nonfluorescent analogue of S3 that was degraded by Lon identically as S3 (5× KmS3, Table 1, see ref 11), and rapidly mixed with 100 µM ATP (10× KATP, Table 1) prior to quenching with formic acid at times between 0 and 3 s. The time courses for ATP hydrolysis measured in the absence and presence of casein (protein substrate) were also deter- mined (Figure 5A) for comparison. All three time courses showed an identical burst in ADP production within the first 200 ms of the reaction, followed by at least one slower phase in product formation. The detection of a pre-steady state burst in the acid-quench experiments is indicative of the rate- limiting step occurring after ATP hydrolysis. Although the S2 peptide is smaller than casein and contains only one Lon cleavage site, it stimulates ATP hydrolysis like the protein substrate, thereby indicating that both substrates share FIGURE 4: Substrate dependency of the lag rate constants of S3 cleavage. (A) Lag rate constant for S3 peptide degradation by 5 identical mechanisms in ATPase stimulation. The data in µM monomeric Lon at varying S3 was determined by fitting the Figure 5A were initially fit with eq 1 (with klag becoming stopped-flow time courses of peptide cleavage as shown in Figure kburst), which characterizes a single-exponential burst phase 2A with eq 1. The data were collectively fit with eq 3 to yield a followed by a linear steady-state turnover rate of the reaction. ) -1 ) ) ( maximal klag,S3 0.88 s , KS3 67 µM, and n 1.7 0.2. The As shown in Figure 5B and the inset, the data obtained during error bars represent the experimental deviations among the different trials. Each data point was obtained from the average of three the first 400 ms of the time courses showed a poor fit with independent experiments, with each experiment containing at least eq 1 and the burst amplitudes were also significantly four stopped-flow traces. (B) Lag rate constant for S3 peptide underestimated. To better evaluate the burst amplitudes as degradation by 5 µM monomeric Lon at varying ATP was well as the burst rates of the reactions, we fit the data within determined by fitting the stopped-flow time courses of peptide the first 400 ms of the time courses, which consisted of a cleavage as shown in Figure 2B with eq 1. The data were -1 burst and a relatively constant transition phase in ADP collectively fit with eq 2 to yield a maximal klag,S3 ) 1.14 s and a KATP ) 7.2 ( 1.9 µM. The error bars represent the experimental formation, with eq 5 to yield the values summarized in Table deviations among the different trials. Each data point was obtained 2. As shown in Table 2 and parts A and C of Figure 5, the from the average of three independent experiments, with each burst amplitudes as well as the burst rate constants for the experiment containing at least four stopped-flow traces. three time courses are comparable but the steady-state rates of the intrinsic versus stimulated ATPase reactions differ. Table 1. The maximum lag rate constant for S3 cleavage The steady-state phase of the time courses (from 600 ms to (klag,S3), determined at varying S3 and varying ATP concen- 1.8 s) were fit with a linear function to yield the steady- trations, was 0.88 ( 0.07 and 1.14 ( 0.06 s-1, respectively. state turnover numbers of 0.23 ( 0.02, 0.69 ( 0.01, and These values are in close agreement with the lag rate constant 0.67 ( 0.03 s-1, corresponding to the intrinsic, S2-stimulated, 1676 Biochemistry, Vol. 44, No. 5, 2005 Vineyard et al.
and casein-stimulated ATPase reactions, respectively. Despite the inclusion of 5 µM monomeric Lon in the reactions however, only ∼2 µM burst of ADP production was detected in all three time courses, suggesting that only 40% of the Lon monomer hydrolyzes ATP during the first enzyme turnover. Comparison of ATPase and S3 CleaVage. To assess how ATP hydrolysis and peptide cleavage are kinetically coor- dinated, we compared the time courses of peptide cleavage and peptide-stimulated ATP hydrolysis under identical reac- tion conditions as shown in Figure 6 (5 µM Lon monomer, 100 µM ATP, and 500 µM peptide). The pre-steady-state phase of ATP hydrolysis consists of the burst of ADP -1 production (kburst,ATP ) 6.5 ( 1.4 s ) and the transition phase, which together spans the first 400 ms of the reaction. Steady- state turnover of ATP hydrolysis, which reflects mostly the rate-limiting step of the reaction, occurs thereafter with a turnover number of 0.69 ( 0.01 s-1. Under identical reaction conditions, the lag in S3 cleavage persists for approximately 1 s with a lag rate constant of 0.94 ( 0.08 s-1 prior to the attainment of steady-state turnover in S3 cleavage at 3.7 ( 0.2 s-1. The close agreement between the lag rate constant for peptide hydrolysis and the kcat of peptide-stimulated ATPase suggests that the first turnover of peptide cleavage may be coupled with the rate-limiting step of peptide- stimulated ATPase activity. As indicated in the above experiments, despite the inclu- sion of 5 µM Lon monomer in the ATPase reactions, only 2 µM ADP formation were detected in the burst phases of the ATPase time course. Because the acid-quench experi- ments measured ADP formation at the active site of the enzyme, the detection of a substoichiometric burst could be attributed to 40% of the enzyme saturated with ATP because of the low nucleotide concentration. If this explanation is correct, the burst amplitude, which reflects the active ATPase concentration, should increase with the nucleotide concentra- tion. To test this hypothesis, we monitored the ATPase reaction in the presence of 5 µM Lon monomer, 500 µM S2, and varying [ATP] for 1.8 s. Figure 7 shows that the time courses display triphasic behavior as discussed above in Figure 5A. The steady-state rates of ATP hydrolysis were obtained by fitting the data from 600 ms to 1.8 s with a FIGURE 5: Pre-steady time courses of ATP hydrolysis by E. coli linear function. Plotting the observed steady-state rate Lon. (A) [R-32P]ATP (100 µM) was incubated with 5 µM constants of ATPase (k ) versus the ATP concentration + ss,ATP monomeric Lon in the absence ( ) and presence of 500 µMS2 yields Figure 8A, which was best fit with eq 2. The k (0)or10µM casein (]), and the reactions were quenched with cat,ATP ( -1 acid at the indicated times. The concentrations of [R-32P]ADP value determined from this analysis was 1.4 0.1 s , and generated in the reactions were determined by TLC followed by the Km,ATP was 76 ( 33 µM, both of which closely agree PhosphorImaging as described in the Materials and Methods. The with that of 1 s-1 and 49 µM determined previously at 150 values on the y axis were obtained by dividing [ADP] produced nM monomeric Lon (23). The Hill coefficient obtained from by 5 µM Lon, which reflects the mole equivalent of ADP produced ( per Lon monomer. (B) [R-32P]ATP (200 µM) was incubated with the fit of these data was 1.7 0.2. Although E. coli Lon 5 µM monomeric Lon in the presence of 500 µM S2. The time exists predominantly as a homooligomer, it exhibits at least points were obtained by quenching the reactions with acid at the two different affinities for ATP (Kd < 1 µM and Kd ) 10 indicated times, and the resulting time course for ADP production µM) (19). Because each monomeric subunit contains only ) - +V + was fit with the equation Y A exp( kobst) sst C, where A one ATP-binding site, it is conceivable that the ATPase sites is the burst amplitude, kobs is the observed burst rate constant, Vss is the steady-state rate, and C is the endpoint. The inset shows the in oligomeric Lon are asymmetrical in their functions. fit of the data spanning 0-400 ms. (C) Time points from 0 to 400 Therefore, the detection of a Hill coefficient close to 2 could ms were fit with eq 5 (Materials and Methods) to yield burst be a measurement of the communication between two amplitudes for intrinsic, casein-stimulated, and S2-stimulated time functionally nonequivalent subunits during enzyme catalysis. courses of 1.92 ( 0.12, 1.8 ( 0.2, and 2.06 ( 0.15 µM, respectively, as summarized in Table 2. The burst rate constants This speculation however, will require further evaluation by were 7.6 ( 1.3, 12.2 ( 3.2, and 6.5 ( 1.4 s-1, respectively, which biophysical approaches that relate the oligomeric state of Lon are also summarized in Table 2. with enzymatic activities. Kinetic Mechanism of ATP Hydrolysis and Peptide Cleavage Biochemistry, Vol. 44, No. 5, 2005 1677
Table 2: Parameters Obtained from Pre-steady-State Kinetic Characterization of the ATPase Activity of E. coli Lon intrinsic ATPase +20 µM casein +500 µMS2 -1 a kcat,ATP NA NA 1.4 ( 0.1 s a Km,ATP NA NA 76 ( 33 µM n NA NA 1.7 ( 0.2a (1.4 ( 0.1)b b KATP NA NA 22 ( 16 µM -1 c -1 c c a -1 d kburst 7.6 ( 1.3 s 12.2 ( 3.2 s 6.5 ( 1.4 (11.3 ( 3.3) (18 ( 2) sec burst amplitude 1.92 ( 0.12 µMc 1.8 ( 0.2 µMc 2.06 ( 0.15c (2.2 ( 0.1)b (4.1 ( 0.1) µMd a These values were obtained by fitting the data in Figure 8A with eq 2. b These values were obtained by fitting the data in Figure 8B with eq 2. c These values were obtained by fitting the acid-quenched time courses (0-400 ms) of ATP hydrolysis that are shown in Figure 5C with eq 5. d These values were obtained by fitting the pulse-chase data in Figure 10 with eq 5.
the burst amplitudes for ATP hydrolysis exhibit a dependency on the ATP concentration and a maximum burst amplitude of 2.2 ( 0.1 µM (44% of the concentration of Lon monomer present), a KATP of 22 ( 16 µM, and a Hill coefficient of 1.4 ( 0.10 were obtained through fitting the data with eq 2. According to Figure 8C, the burst rates of ATP hydrolysis are independent of the nucleotide concentrations (5-200 µM) and the averaged burst rate constant (kburst) is 11.3 ( 3.3 s-1.At5µM ATP, it is anticipated that the tight but not the weak nucleotide-binding sites of Lon will be saturated with ATP. Therefore, the apparent independence of kburst toward the indicated ATP concentrations is likely attributed to ATP hydrolysis occurring at the high-affinity sites. When our data FIGURE 6: Pre-steady-state time courses for ATP hydrolysis and S3 degradation at identical reaction conditions. Monomeric Lon are taken together with the observed 44% burst amplitude, (5 µM) was incubated with 100 µM ATP and 500 µM peptide they suggest that the pre-steady-state burst in ATP hydrolysis substrate. The time course for peptide hydrolysis (O) was deter- occurs at the high-affinity sites (∼50% of the total monomer mined by fluorescence stopped-flow spectroscopy using 100 µM concentration) but the hydrolysis of ATP is coordinated with nonradiolabeled ATP and 500 µM S3 as the substrates. The time course for ATP hydrolysis ([) was determined by rapid acid ATP binding at the low-affinity sites. quenching of a reaction containing Lon, 100 µM[R-32P]ATP, and Characterization of ATP Binding. The substoichiometric 500 µM S2 as described in the Materials and Methods. The values burst in ADP production could be caused by either inefficient on the left y axis were obtained by dividing the concentrations of the peptide or ATP hydrolyzed by 5 µM Lon, whereas the right y nucleotide binding or reflects the coordinated ATP binding axis reports the concentrations of products formed. and hydrolysis between the nonequivalent ATPase sites in Lon. To further evaluate these potential mechanisms, we measured the concentration of functional nucleotide-binding and hydrolysis sites using a filter binding assay and a pulse- chase experiment, respectively. In the filter binding assay, several aliquots of an enzyme stock of Lon, whose concentration was predetermined by the Bradford assay (final enzyme concentration e 5 µM), were incubated with 10 µM[R-32P]ATP at 37 °C to yield [R-32P]ADP that remained tightly bound to Lon. Because ADP competes with ATP for the same binding site and its Ki is 0.3 µM(11), it is anticipated that the concentration of active Lon could be defined by the population of enzyme that bound [R-32P]ADP and was retained as radiolabeled FIGURE 7: Pre-steady-state time courses at varying [ATP] and constant S2. Monomeric Lon (5 µM) was incubated with 500 µM protein onto the nitrocellulose. As such, the concentration S2 and 5 (4), 10 (b), 25 (]), 50 (0), 100 (×), and 200 (O) µM of active Lon could be determined from the radioactive [R-32P]ATP. The reactions were rapidly quenched with acid at the counts of standards containing known ATP concentrations. indicated times, and the amount of [R-32P]ADP formed was A plot of [32P]ADP versus the volume of Lon stock added determined by PhosphorImaging. Each time course was repeated in the binding reaction yielded a linear plot (Figure 9), at least 3 times, and the averaged data are reported. indicating that the concentration of Lon/[32P]ADP formed The burst amplitudes and burst rate constants were was directly proportional to the amount of enzyme present obtained by fitting the data spanning 0-400 ms in Figure 7 in the reaction. The concentration of active Lon was with eq 5 and plotting the respective parameters against their calculated from the slope of the fit, which corresponds to specific ATP concentration (parts B and C of Figure 8). 36 µM monomeric enzyme. This value agrees closely with Because the burst amplitude directly reflects the amount of the concentration of the enzyme stock determined by the ADP production at the active site of the enzyme, the fit of Bradford assay (35 µM) and indicates that all of the the data in Figure 8B with eq 2 provided a measurement of monomeric Lon used in the ATPase reactions could bind the apparent Kd for ATP (KATP). As shown in Figure 8B, the nucleotide. 1678 Biochemistry, Vol. 44, No. 5, 2005 Vineyard et al.
FIGURE 9: Determining the concentration of active Lon by filter binding assay. Total of 2, 3, 4, and 5 µL of a Lon stock (35 µM monomer as determined by the Bradford assay) was incubated with 10 µM[R-32P]ATP at 37 °C for 20 min to convert ATP to ADP. The amount of Lon/[R-32P]ADP formed in each reaction was determined by the filter binding assay described in the Materials and Methods. Plotting the amount of [R-32P]ADP bound to Lon against the volume of enzyme stock added yields a straight line that was fit with the linear equation y ) mx + c, where m equals 36 pmol of ADP/µL of Lon (36 µM Lon momomer).
experiment. In this experiment, 5 µM Lon monomer was incubated with 500 µM S2 and rapidly mixed with 100 µM [R-32P]ATP as in the chemical-quench-flow experiment, except that the mixed reaction was chased with 10 mM unlabeled Mg‚ATP for 60 s before quenching with acid and denaturation with phenol/chloroform. During ATP hydroly- sis, [R-32P]ATP bound to Lon could either be hydrolyzed to yield [R-32P]ADP or dissociate from the enzyme. In the presence of 10 mM unlabeled ATP (100-fold excess over labeled ATP), any enzyme not bound to [R-32P]ATP would be sequestered by the unlabeled ATP and not be detected in the assay. In the acid-quench experiment, any [R-32P]ATP that bound to Lon but was not hydrolyzed remained as unreacted ATP. In contrast, the 60 s delay in the pulse-chase experiment allowed time for the enzyme to hydrolyze any [R-32P]ATP that was bound during the various incubation times, and the presence of 10 mM unlabeled ATP (100-fold excess over labeled ATP) allowed for the complete exchange of [R-32P]ADP for unlabeled ATP. FIGURE 8: Fitting the pre-steady-state kinetic parameters of ATP hydrolysis by Lon. (A) Data from 600 ms to 1.8 s in Figure 7 were Figure 10 compares the time courses of ATP hydrolysis fit with a linear function to provide the steady-state rates of ATP by Lon under the acid-quench and pulse-chase conditions. hydrolysis at varying [ATP]. Each time point was performed at least 3 times, and the averaged data were reported. The error bars The data spanning the first 400 ms were fit with eq 5 to represent the error of the fit for each data point. The kss,ATP values yield the burst amplitudes and kburst of the respective were obtained by dividing the steady-state rates by [Lon]. Plotting reactions. In the presence of ATP chase, the burst amplitude the kss,ATP values against its specific [ATP] yields a sigmoidal plot was 4.1 ( 0.1 µM, which was approximately 80% of the that is best fit with eq 2. The kinetic parameters obtained from the -1 total monomeric Lon present in the reaction. The burst rate fit were kcat,ATP ) 1.4 ( 0.1 s , Km,ATP ) 76 ( 33 µM, and n ) 1.7 ( 0.2 and are summarized in Table 2. (B) Burst amplitudes of the chased reaction was 18 ( 2s-1, which was 2-fold were determined by fitting the data from 0 to 400 ms in Figure 7 higher than that determined for the chemical-quench condi- with eq 5. As summarized in Table 2, the maximum burst amplitude tion under identical labeled ATP conditions. The pulse-chase obtained from the fit was 2.2 µM, which corresponds to ∼44% of the enzyme present in the reaction, a KATP ) 22 ( 16 µM, and n result showed that at least 80% of the Lon monomer could ) 1.4 ( 0.1. The time points reported here are averaged values of bind and hydrolyze ATP. Therefore, the detection of a three different trials. The error bars represent the standard error of reduced burst in ADP production in the chemical-quench- the fit for evaluating the respective burst amplitude values at the specific [ATP]. (C) The burst rates of ATP hydrolysis were flow time course was not due to the presence of inactive determined by the same manner as in B, and the average kburst ) enzyme in the reaction. Moreover, the ATPase sites of Lon 11.3 ( 3.3 s-1. appeared to be functionally nonequivalent, because ∼50% of the sites (tight affinity) hydrolyzed ATP within the first To determine if all of the monomeric Lon prepared for 400 ms of the reaction, whereas the remaining sites (weak this study could hydrolyze ATP, we performed a pulse-chase affinity) hydrolyzed ATP at a much slower rate. Kinetic Mechanism of ATP Hydrolysis and Peptide Cleavage Biochemistry, Vol. 44, No. 5, 2005 1679
Scheme 1a
a E and F are different catalytic forms of Lon along the reaction pathway. FIGURE 10: Monomeric Lon (5 µM) was incubated with 500 µΜ S2 and 100 µM[R-32P]ATP for the indicated times and then burst in ATP hydrolysis further indicates that the rate-limiting O ‚ quenched with acid ( ) or chased with 10 mM Mg ATP for 60 s step of the reaction occurs after nucleotide hydrolysis; prior to quenching with acid (b). The data from 0 to 400 ms were best-fit with eq 5 to yield a burst amplitude of 2.06 ( 0.15 µM therefore, it could be ADP product release. The detection of -1 and a kburst of 6.2 ( 1.4 s for the acid-quench experiment. the rate-limiting step following ATP hydrolysis is consistent Similarly, a burst amplitude of 4.1 ( 0.1 µM and a kburst of 18 ( with the proposal that ADP release limits ATPase turnover - 2s 1 were yielded for the pulse-chase experiment. The amount of (3, 19) and is supported by our observation that both casein [R-32P]ADP generated at each time point was reported on the right y axis. The values on the left y axis were obtained by dividing the and S2 peptide stimulate only the steady-state turnover rate concentrations of ADP formed by 5 µM Lon. of ATP hydrolysis (Figure 5A). Collectively, these results support a reaction model in which activation of peptide DISCUSSION hydrolysis is driven by ATP hydrolysis and the first turnover of peptide cleavage is coupled with the rate-limiting step in Lon is an ATP-dependent protease functioning to degrade ATP hydrolysis. damaged and certain regulatory proteins in ViVo. In this study, This model indicates that ATP hydrolysis precedes peptide we demonstrated that in E. coli Lon, peptide cleavage is cleavage and is in discord with the earlier reaction model driven by ATP hydrolysis, which is coordinated with ATP proposed for Lon, which suggests that peptide hydrolysis binding to the low-affinity sites in the oligomeric enzyme. occurs before ATP consumption (3). The discrepancy Our findings are in discord with the previous proposal that between the two models could be attributed to the choice of peptide cleavage occurs before ATP hydrolysis in Lon peptide substrates used in evaluating the ATPase dependency catalysis (3, 4, 6, 21). In the previous studies, on the basis of the reaction. Some of the earlier studies utilized peptides that proteins but not tetrapeptide substrates stimulate the lacking ATPase stimulation abilities as substrates, which ATPase activity of Lon, it was proposed that proteins bind might have led to an underestimation of the contribution of to an allosteric site in Lon to promote ADP release, thereby ATP hydrolysis toward peptide-cleavage efficiency. Although facilitating enzyme turnover (3, 19). Although it is known S2 is significantly smaller than casein and contains only one that ATP is the most effective activator of Lon protease, Lon cleavage site, it exhibits the same ATPase stimulation mechanistic details concerning how ATP hydrolysis is profile as the latter (23). Together with the previous data coupled with peptide bond cleavage is not available. To indicating that the kcat for peptide degradation is 7-fold higher investigate the molecular mechanism of the ATPase-depend- in the presence of ATP versus AMPPNP, we conclude that ent peptidase activity in E. coli Lon, we performed the first this peptide is more suitable for evaluating the ATPase- pre-steady-state kinetic analysis on this enzyme by measuring dependent protease mechanism of Lon (11). the kinetics of the first turnover of ATP and peptide Pre-steady state kinetic analysis of the ATPase activity of hydrolysis of a synthetic peptide substrate whose degradation Lon further reveals functional nonequivalency in the subunits by Lon exhibits the same ATP dependency as protein of the enzyme, because only 50% of the ATP bound to Lon substrates (22). Because this peptide contains only one Lon is hydrolyzed before peptide cleavage. The observed asym- cleavage site and it stimulates ATP hydrolysis, the data metry in the ATPase activity could be attributed to the two obtained from this study can be directly attributed to the different classes of ATP-binding sites found in E. coli Lon ATPase-dependent peptidase reaction rather than polypeptide as reported by Menon and Goldberg [Kd < 1 µM and Kd ∼ unfolding or processive peptide cleavage. 10 µM(19, 20)], who also observed that optimal protein As presented in this study, E. coli Lon exhibits lag kinetics degradation requires occupancy of ATP at each site. The in the degradation of the model peptide but burst kinetics in molecular basis for such requirement, however, was not clear. ATP hydrolysis. During the lag in peptide cleavage, ∼50% On the basis of the ATPase data obtained in this study, we of the Lon monomer hydrolyzes ATP with a burst rate propose a sequential ATP hydrolysis reaction model that constant of 11.3 ( 3.3 s-1. According to the ATPase kinetic could account for the aforementioned observations (Scheme -1 data, the kcat for peptide-stimulated ATPase is 1.4 s (Table 1). Assuming that ∼50% of the Lon subunits exhibit high 2), which approximates the klag,S3 of S3 cleavage as sum- affinity for ATP, these sites will be saturated at 5 µM ATP, marized in Table 1 (0.88-1.14 s-1). These results show that which is the lowest concentration of ATP used in this study. the first turnover of peptide hydrolysis requires the build up Under this condition, Lon binds but does not hydrolyze ATP. of a reaction intermediate, which coincides with the rapid As the concentration of ATP increases, the low-affinity sites hydrolysis of ATP during the first enzyme turnover. The become occupied with the nucleotide, which upon full 1680 Biochemistry, Vol. 44, No. 5, 2005 Vineyard et al. yielded the steady-state turnover number for ATP hydrolysis of 0.8 sec-1. These values are in accordance with the -1 maximum kburst of 11.3 ( 3.3 s and kcat,ATP of 1.4 ( 0.1 s-1 determined by analyzing the pre-steady-state and steady- state time course of ATP hydrolysis separately as described above (Table 2). While the ATPase time courses presented in this study could be consistent with a sequential reaction mechanism of Lon, these experiments alone cannot conclu- sively determine this mechanism. Therefore, additional pre- steady-state kinetic characterizations are currently underway to further investigate the ATPase reaction mechanism of Lon in more detail. The ATPase mechanism proposed in this study predicts FIGURE 11: Collective fit of acid-quench ATPase data using FitSim. Simulation of the ATPase mechanism outlined in Scheme 1 was that communication between the two classes of ATPase sites performed using FitSim. The resulting solid lines yielded a kburst ) contribute significantly to peptide cleavage. This proposal ( -1 ) -1 8 2.4 s and kss,ATP 0.8 s and were overlayed with the is supported by the observation that both the ATPase and experimental data from Figure 7, demonstrating consistency with peptidase sites display the same level of positive cooperat- the proposed sequential mechanism. ivity in the binding and hydrolysis of substrates. We observe that the klag as well as the observed steady-state rate constant occupancy activates ATP hydrolysis at the high-affinity sites. (kss,S3) values of peptide cleavage exhibit sigmoidal depen- Hydrolysis of ATP at the low-affinity sites occurs thereafter. dency on S3 concentrations with a Hill coefficient (n) ranging On the basis of this model, the first turnover of ATP from 1.22 to 1.7 (Table 1). Moreover, both the steady-state hydrolysis, which occurs at the high-affinity sites, is de- turnover number as well as the burst amplitude of ATP pendent on the binding of ATP to the low-affinity sites and hydrolysis determined at saturating S2 and varying ATP a 50% burst in ATP hydrolysis is anticipated in the burst concentrations also exhibit sigmoidal kinetics. Assuming that phase of the acid-quench experiment as shown in Figures E. coli Lon is a hexamer (14), the maximum n value would 8B and 10. However, when chased with unlabeled ATP, the be 6, and thus, an n value of ∼1.6 measured here either radiolabeled ATP bound to the low-affinity sites will suggests slight positive cooperativity among the six Lon eventually hydrolyze ATP to yield a full burst in ADP subunits or the communication between the two classes of formation (Figure 10). Because initial ATP hydrolysis occurs ATPase sites with a maximum of n ) 2. While the extent at the high-affinity sites, the burst amplitudes but not the of subunit communication cannot be defined in this study, burst rates are dependent on ATP binding to the low-affinity our results clearly reveal that interactions among the subunits sites (parts B and C of Figure 8). Because peptide hydrolysis in Lon affect the ATPase and peptidase activities of the is coupled with a step after ATP hydrolysis, the binding of enzyme. ATP to the low-affinity sites therefore indirectly activates Because ATP hydrolysis does occur before S3 cleavage, the peptidase activity by stimulating ATP hydrolysis. The we propose that nucleotide hydrolysis generates an enzyme triphasic ATPase time courses together with the half-site form that subsequently cleaves S3. Lon exhibits high reactivity in ATP hydrolysis are similar to the ATPase sequence and structural homology with the heterosubunit activity of yeast topoisomerase in which its sequential ATP-dependent proteases, such as HslUV, which utilizes ATPase mechanism contains a burst phase and two slow ATP hydrolysis to deliver unfolded polypeptide substrates steps of comparable magnitudes (27). It is possible that Lon to the proteolytic site (14, 15, 17, 32-34). This similarity adopts a similar mechanism in coupling ATPase activity to suggests that ATP consumption serves a similar function in activate peptide cleavage. To further evaluate this hypothesis, Lon. Previously, we employed steady-state kinetics and we fit the averaged time courses of the acid-quenched product inhibition studies to construct a kinetic model to ATPase data (from 5 to 100 µM ATP) collectively to the account for the observed “ATPase-dependent” S3 cleavage kinetic mechanism shown in Scheme 1 by regression analysis reaction by Lon (11). Because the identities of different using FitSim (28-30) or KinFitSim (31). This minimal enzyme intermediates existing along the peptidase reaction mechanism features the sequential ATPase mechanism pathway are not defined, we cannot fit the time courses of deduced based upon the data analyses presented above and the peptidase reactions to a defined kinetic mechanism as in assumes that one of the product-release steps limits enzyme the case for the ATPase reaction. However, a simplified turnover. In addition, the concentration of enzyme used for version of this model showing the microscopic events the fitting process was varied at the respective ATP associated with S3 cleavage is provided in Scheme 2. This concentration based upon the data reported in Figure 8B. model proposes that ATP hydrolysis facilitates the delivery The two kon and koff for ATP binding are estimated from the of S3 to the peptidase site of Lon and the exchange of ADP two ATP affinities of Lon [<1 and 10 µM, respectively (19)]. with ATP in step 3 is rate-limiting. In the presence of excess Because the kinetics of product release are yet to be S3, the peptide binds to an allosteric site in Lon to promote determined, the rate constants associated with events after the exchange of ADP with ATP, thereby increasing the steady-state turnover are at present hypothetical. Fitting the catalytic turnover of ATP hydrolysis. The data obtained in data to this mechanism yielded a pre-steady-state burst rate this study support this proposed model by revealing that ATP constant for ATP hydrolysis of 8 ( 2.4 sec-1 and a steady- hydrolysis precedes peptide bond cleavage. The hydrolysis state turnover of 4 ( 0.34 µM/sec (Figure 11), which, upon of ATP (step 2) occurs with an apparent rate constant of division by the 5 µM Lon monomer present in the reaction, ∼11 s-1 (Table 2) to generate an active enzyme form “F” Kinetic Mechanism of ATP Hydrolysis and Peptide Cleavage Biochemistry, Vol. 44, No. 5, 2005 1681
Scheme 2 3. Goldberg, A. L., Moerschell, R. P., Chung, C. H., and Maurizi, M. R. (1994) ATP-dependent protease La (lon) from Escherichia coli, Methods Enzymol. 244, 350-375. 4. Gottesman, S., and Maurizi, M. (1992) Regulation by proteoly- sis: Enegry-dependent proteases and their targets, Microbiol. ReV. 56, 592-621. 5. Gottesman, S. (1996) Proteases and their targets in Escherichia coli, Annu. ReV. Genet. 30, 465-506. 6. Maurizi, M. R. (1992) Proteases and protein degradation in Escherichia coli, Experientia 48, 178-201. 7. Charette, M. F., Henderson, G. W., Doane, L. L., and Markovitz, A. (1984) DNA-stimulated ATPase activity on the lon (CapR) protein, J. Bacteriol. 158, 195-201. 8. Schoemaker, J. M., Gayda, R. C., and Markovitz, A. (1984) Regulation of cell division in Escherichia coli: SOS induction that, upon the exchange of ADP with ATP, cleaves S3 (step and cellular location of the sulA protein, a key to lon-associated 3). We tentatively assign step 3 as the slow step because filamentation and death, J. Bacteriol. 158, 551-561. ADP inhibits S3 cleavage and its slow release from Lon 9. Goldberg, A. L., and Waxman, L. (1985) The role of ATP would be consistent with the detection of burst kinetics in hydrolysis in the breakdown of proteins and peptides by protease La from Escherichia coli, J. Biol. Chem. 260, 12029-12034. ATP hydrolysis (Figure 5). If F/ATP/S3 is the enzyme form 10. Goff, S. A., and Goldberg, A. L. (1985) Production of abnormal that catalyzes peptide bond cleavage and subsequent steps proteins in E. coli stimulates transcription of lon and other heat associated with S3 cleavage are fast, then the kinetics of S3 shock genes, Cell 41, 587-595. 11. Thomas-Wohlever, J., and Lee, I. (2002) Kinetic characterization cleavage will exhibit a lag whose duration depends on the of the peptidase activity of Escherichia coli Lon reveals the kinetics of ADP and/or Pi dissociation. Because the lag rate mechanistic similarities in ATP-dependent hydrolysis of peptide -1 - constant for S3 cleavage is ∼1s (Table 1) and the kcat of and protein substrates, Biochemistry 41, 9418 9425. ATP hydrolysis is ∼1.2 s-1 (Table 2), it is conceivable that 12. Chin, D. T., Goff, S. A., Webster, T., Smith, T., and Goldberg, A. L. (1988) Sequence of the lon gene in Escherichia coli.A such coordination between the two hydrolytic events occurs heat-shock gene which encodes the ATP-dependent protease La, in the first turnover of S3 cleavage. However, it is also J. Biol. Chem. 263, 11718-11728. possible that the kcat of ATP hydrolysis and the klag of S3 13. Amerik, A., Chistiakov, L. G., Ostroumova, N. I., Gurevich, A. I., and Antonov, V. K. (1988) Cloning, expression, and structure cleavage are coordinated successively in the reaction pathway of the functionally active shortened lon gene in Escherichia coli, of Lon. In this case, both steps contribute sequentially to Bioorg. Khim. 14, 408-411. limit the turnover of peptide hydrolysis. The data presented 14. Botos, I., Melnikov, E. E., Cherry, S., Tropea, J. E., Khalatova, in this study cannot distinguish between these two possibili- A. G., Rasulova, F., Dauter, Z., Maurizi, M. R., Rotanova, T. V., Wlodawer, A., and Gustchina, A. (2004) The catalytic domain of ties, and further kinetic characterization is needed to further Escherichia coli Lon protease has a unique fold and a Ser-Lys delineate this mechanism. dyad in the active site, J. Biol. Chem. 279, 8140-8148. In addition to peptide translocation, the ATPase activity 15. Botos, I., Melnikov, E. E., Cherry, S., Tropea, J. E., Khalatova, A. G., Rasulova, F., Dauter, Z., Maurizi, M. R., Rotanova, T. V., of Lon could also be used to facilitate other kinetic events Wlodawer, A., and Gustchina, A. (2004) Crystal structure of the prior to peptide cleavage. For example, because Lon func- AAA+Rdomain of E. coli Lon protease at 1.9 Å resolution, J. tions as an oligomer, ATP hydrolysis may be used to promote Struct. Biol. 146, 113-122. 16. Sousa, M. C., Trame, C. B., Tsuruta, H., Wilbanks, S. M., Reddy, enzyme oligomerization or to induce a conformational change V. S., and McKay, D. B. (2000) Crystal and solution structures in the oligomer that facilitates peptide cleavage. The kinetic of an HslUV protease-chaperone complex, Cell 103, 633-643. data presented in this study cannot disprove these possibilities 17. Wang, J., Song, J. J., Seong, I. S., Franklin, M. C., Kamtekar, S., because these events occur before S3 cleavage. However, Eom, S. H., and Chung, C. H. (2001) Nucleotide-dependent conformational changes in a protease-associated ATPase HsIU, studies performed on the oligomerization of Lon homologues Structure 9, 1107-1116. (3, 35, 36) thus far have revealed that the enzyme oligo- 18. Waxman, L., and Goldberg, A. L. (1982) Protease La from merization is independent of nucleotide, thereby excluding Escherichia coli hydrolyzes ATP and proteins in a linked fashion, Proc. Natl. Acad. Sci. U.S.A. 79, 4883-4887. the possibility that ATP consumption is coupled to enzyme 19. Menon, A. S., and Goldberg, A. L. (1987) Binding of nucleotides oligomerization. However, unpublished data (Lee and Burke) to the ATP-dependent protease La from Escherichia coli, J. Biol. suggest that the oligomerization of Lon is a reversible process Chem. 262, 14921-14928. that is sensitive to reaction conditions. As such, studies are 20. Menon, A. S., Waxman, L., and Goldberg, A. L. (1987) The energy utilized in protein breakdown by the ATP-dependent protease (La) currently being conducted to evaluate the effect of ATP on from Escherichia coli, J. Biol. Chem. 262, 722-726. the oligomeric state of Lon under the reaction conditions 21. Menon, A. S., and Goldberg, A. L. (1987) Protein substrates used in these studies. activate the ATP-dependent protease La by promoting nucleotide binding and release of bound ADP, J. Biol. Chem. 262, 14929- 14934. ACKNOWLEDGMENT 22. Lee, I., and Berdis, A. J. (2001) Adenosine triphosphate-dependent degradation of a fluorescent λ N substrate mimic by Lon protease, We thank Hilary Frase and Joyce Jentoft for their Anal. Biochem. 291,74-83. assistance in preparing this manuscript. 23. Patterson, J., Vineyard, D., Thomas-Wohlever, J., Behshad, R., Burke, M., and Lee, I. (2004) Correlation of an adenine-specific conformational change with the ATP-dependent peptidase activity REFERENCES of Escherichia coli Lon, Biochemistry 43, 7432-7442. 24. Gilbert, S. P., and Mackey, A. T. (2000) Kinetics: A tool to study 1. Chung, C. H., and Goldberg, A. L. (1981) The product of the lon molecular motors, Methods 22, 337-354. (capR) gene in Escherichia coli is the ATP-dependent protease, 25. Frieden, C. (1979) Slow transitions and hysteretic behavior in - protease La, Proc. Natl. Acad. Sci. U.S.A. 78, 4931 4935. enzymes, Annu. ReV. Biochem. 48, 471-489. 2. Gottesman, S., Gottesman, M. E., Shaw, J. E., and Pearson, M. 26. Frieden, C. (1970) Kinetic aspects of regulation of metabolic L. (1981) Protein degradation by protease La from Escherichia processes. The hysteretic enzyme concept, J. Biol. Chem. 245, coli, J. Cell. Biochem. 32. 5788-5799. 1682 Biochemistry, Vol. 44, No. 5, 2005 Vineyard et al.
27. Harkins, T. T., and Lindsley, J. E. (1998) Pre-steady-state analysis 33. Ogura, T., and Wilkinson, A. J. (2001) AAA+ superfamily of ATP hydrolysis by Saccharomyces cereVisiae DNA topoi- ATPases: Common structure-diverse function, Genes Cells 6, somerase II. 2. Kinetic mechanism for the sequential hydrolysis 575-597. of two ATP, Biochemistry 37, 7292-7298. 34. Wang, J., Song, J. J., Franklin, M. C., Kamtekar, S., Im, Y. J., 28. Zimmerle, C. T., and Frieden, C. (1989) Analysis of progress Rho, S. H., Seong, I. S., Lee, C. S., Chung, C. H., and Eom, S. curves by simulations generated by numerical integration, Bio- H. (2001) Crystal structures of the HslVU peptidase-ATPase chem. J. 258, 381-387. complex reveal an ATP-dependent proteolysis mechanism, Struc- 29. Frieden, C. (1994) Analysis of kinetic data: Practical applications ture 9, 177-184. of computer simulation and fitting programs, Methods Enzymol. 35. Stahlberg, H., Kutejova, E., Suda, K., Wolpensinger, B., Lustig, 240, 311-322. A., Schatz, G., Engel, A., and Suzuki, C. K. (1999) Mitochondrial 30. Dang, Q., and Frieden, C. (1997) New PC versions of the kinetic- Lon of Saccharomyces cereVisiae is a ring-shaped protease with simulation and fitting programs, KINSIM and FITSIM, Trends seven flexible subunits, Proc. Natl. Acad. Sci. U.S.A. 96, 6787- Biochem. Sci. 22, 317. 6790. 31. Svir, I. B., Klymenko, A. V., and Platz, M. S. (2002) KinFitSims 36. Rudyak, S. G., Brenowitz, M., and Shrader, T. E. (2001) Mg2+- A software to fit kinetic data to a user selected mechanism, linked oligomerization modulates the catalytic activity of the Lon Comput. Chem. 26, 379-386. (La) protease from Mycobacterium smegmatis, Biochemistry 40, 32. Neuwald, A. F., Aravind, L., Spouge, J. L., and Koonin, E. V. 9317-9323. (1999) AAA+: A class of chaperone-like ATPases associated with the assembly, operation, and disassembly of protein com- plexes, Genome Res. 9,27-43. BI048618Z 4602 Biochemistry 2006, 45, 4602-4610
Single-Turnover Kinetic Experiments Confirm the Existence of High- and Low-Affinity ATPase Sites in Escherichia coli Lon Protease† Diana Vineyard, Jessica Patterson-Ward, and Irene Lee* Department of Chemistry, Case Western ReserVe UniVersity, CleVeland, Ohio 44106 ReceiVed NoVember 21, 2005; ReVised Manuscript ReceiVed January 31, 2006
ABSTRACT: Lon is an ATP-dependent serine protease that degrades damaged and certain regulatory proteins in vivo. Lon exists as a homo-oligomer and represents one of the simplest ATP-dependent proteases because both the protease and ATPase domains are located within each monomeric subunit. Previous pre-steady-state kinetic studies revealed functional nonequivalency in the ATPase activity of the enzyme [Vineyard, D., et al. (2005) Biochemistry 44, 1671-1682]. Both a high- and low-affinity ATPase site has been previously reported for Lon [Menon, A. S., and Goldberg, A. L. (1987) J. Biol. Chem. 262, 14921- 14928]. Because of the differing affinities for ATP, we were able to monitor the activities of the sites separately and determine that they were noninteracting. The high-affinity sites hydrolyze ATP very slowly -1 (kobs ) 0.019 ( 0.002 s ), while the low-affinity sites hydrolyze ATP quickly at a rate of 17.2 ( 0.09 s-1, which is comparable to the previously observed burst rate. Although the high-affinity sites hydrolyze ATP slowly, they support multiple rounds of peptide hydrolysis, indicating that ATP and peptide hydrolysis are not stoichiometrically linked. However, ATP binding and hydrolysis at both the high- and low-affinity sites are necessary for optimal peptide cleavage and the stabilization of the conformational change associated with nucleotide binding.
Lon is an ATP-dependent serine protease functioning to account for the effect of the peptide (23). This 10 amino degrade damaged and certain regulatory proteins in vivo (1- acid long (S3 and S2) peptide sequence contains only one 10). Lon belongs to the ATPases associated with a variety cleavage site and comes from the λN protein, which is a of cellular activities (AAA+) superfamily, whose members physiological substrate of Escherichia coli Lon (24). Because include ClpAP, ClpXP, ClpCP, and HslUV (11, 12). They our model peptide (S3 and S2) contains only one Lon share a conserved Walker A (or P loop) and Walker B motif, cleavage site and it stimulates ATP hydrolysis, its kinetics which is associated with nucleotide binding and hydrolysis of degradation can be directly attributed to the ATP- (13). Lon represents one of the simplest of the ATP- dependent peptidase reaction rather than polypeptide unfold- dependent proteases because both the protease and ATPase ing or processive peptide cleavage (22, 24). domains are located within each monomeric subunit (14, 15). We have utilized this peptide substrate in steady-state Crystal structures of portions of the enzyme have been kinetic and product inhibition studies to establish a minimal recently reported and include an inactive mutant of the Lon kinetic mechanism for E. coli Lon protease (22). This protease domain (16-18). This structure shows Lon as a mechanism proposed a sequential mechanism for ATP hexamer organized in a ring with a central cavity, which is binding and hydrolysis, which mediated peptide cleavage commonly found in other ATP-dependent proteases (11, 13, presumably through a predicted ATP-dependent translocation 17). Although it is known that ATP modulates the protease step. The resulting Lon/ATP-bound enzyme form (F) was activity of Lon (4, 5, 7, 19), mechanistic details concerning distinct from the precatalytic Lon (E). Because the steady- how the binding and hydrolysis of ATP are coordinated with state methods and predicted kinetic model could not address peptide bond cleavage is not known. However, it has been the microscopic details along the reaction pathway, we shown that ATP binding and hydrolysis do not affect the utilized pre-steady-state kinetic techniques to determine the oligomeric state of the enzyme (20, 21). timing of events as well as individual rate constants. We have previously developed a continuous fluorescent Previously, we were able to elucidate the timing of ATP peptidase assay to monitor the kinetics of peptide cleavage.
Because the inner-filter effect of fluorescence interferes at 1 1 Abbreviations: AMPPNP, adenylyl 5-imidodiphosphate; DTT, high concentrations of 100% fluorescent peptide, we use S3, dithiothreitol; Abz, anthranilamide; Bz, benzoic acid amide; HBTU, a 10% mixture of fluorescently labeled peptide with its O-benzotriazole-N,N,N′,N′-tetramethyl-uronium-hexafluoro-phos- ′ nonfluorescent analogue (S2) (22). No optical signal from phate; HEPES, N-2-hydroxyethylpiperazine-N -ethanesulfonic acid; KPi, the peptide is needed when monitoring ATPase activity; potassium phosphate; Mg(OAc)2, magnesium acetate; KOAc, potassium acetate; PEI-cellulose, polyethyleneimine-cellulose; S2, a nonfluo- therefore, only the nonfluorescent analogue (S2) is used to rescent analogue of S3 that is degraded by Lon identically as S3 and is used in the ATPase reactions to conserve the fluorescent peptide † This work was support by the NIH Grant GM067172. (S3): YRGITCSGRQK(benzoic acid amide) (Bz); S3, a mixed peptide * To whom correspondence should be addressed. Telephone: 216- substrate containing 10% of the fluorescent peptide Y(NO2)RGITCSGRQ- 368-6001. Fax: 216-368-3006. E-mail: [email protected]. K(Abz) and 90% S2. 10.1021/bi052377t CCC: $33.50 © 2006 American Chemical Society Published on Web 03/21/2006 High- and Low-Affinity ATPase Sites in Lon Biochemistry, Vol. 45, No. 14, 2006 4603
Scheme 1: Enzyme Forms Associated with Various Concentrations of ATPa
a Form A is a free enzyme containing two different sets of ATPase sites that are represented by gray squares and circles. Form B is formed under single-turnover conditions when only 500 nM ATP is present. The occupancy of ATP to an enzyme subunit is illustrated by the change in color from gray to yellow. Form C represents the enzyme form where only the tight sites are occupied by ATP. Form D represents an enzyme form where both the tight and weak sites are saturated with ATP. hydrolysis and peptide cleavage in Lon (25), by demonstrat- or protein substrates and is independent of nucleotide binding ing that ATP hydrolysis was occurring before peptide at the low-affinity ATPase sites. Collectively, the data cleavage during the first turnover of Lon. This study found obtained in this study reveal that peptide cleavage is not that E. coli Lon exhibits lag kinetics in the degradation of stoichiometrically linked to ATP hydrolysis because multiple S3 but burst kinetics in ATP hydrolysis. Furthermore, the rounds of peptide hydrolysis occur under conditions of ATPase activity of Lon reveals functional nonequivalency limiting ATP, where only the high-affinity sites are occupied. in the subunits of the enzyme, because only 50% of the ATP We have also shown that the two ATPase sites hydrolyze bound to Lon is hydrolyzed before peptide cleavage (25). ATP at drastically different rates, which are seemingly The observed asymmetry in the ATPase activity could be unaffected by ATP hydrolysis at the other site. The previ- attributed to the two different classes of ATP-binding sites ously proposed reaction model (25) is therefore revised found in E. coli Lon as reported by Menon and Goldberg accordingly to account for our currently observed data. (Kd < 1 µM and Kd ∼ 10 µM) (26). Thus far, kinetic data on the ATPase-dependent degrada- MATERIALS AND METHODS tion of the model peptide S3 by Lon implicates a reaction model by which the nonequivalent ATPase sites function Materials. ATP and casein was purchased from Sigma, R 32 - cooperatively to modulate the efficiency of peptide cleavage whereas [ - P]ATP was purchased from Perkin Elmer or (25). This model predicts that ATP hydrolysis will occur ICN Biomedical. Fmoc-protected amino acids, Boc-anthra- ∼ -1 nilamide (Abz), Fmoc-protected Lys Wang resin, and O- with a burst rate constant of 12 s at the tight sites of ′ ′ Lon only when the low-affinity sites are occupied by ATP. benzotriazole-N,N,N ,N -tetramethyl-uronium-hexafluoro- phosphate (HBTU) were purchased from Advanced ChemTech Furthermore, optimal peptide hydrolysis is attained through ′ the coordinated ATP binding to the low-affinity sites and and NovaBiochem. Tris, N-2-hydroxyethylpiperazine-N - ethanesulfonic acid (HEPES), and polyethyleneimine-cel- hydrolysis at the high-affinity sites. However, this model is - constructed on the basis of kinetic data obtained under pseuo- lulose (PEI cellulose) thin-layer chromatography (TLC) first-order conditions, where the concentration of ATP is in plates were purchased from Fisher. excess over Lon. As such, the functional roles of the high- General Methods. Peptide synthesis and protein purifica- and low-affinity ATPase sites could not be independently tion procedures were performed as described previously (24). examined, and the validity of the proposed model could not All enzyme concentrations were reported as Lon monomer be rigorously tested. To further investigate the cooperative concentrations. All reagents are reported as final concentra- function of the two kinds of ATPase sites in Lon and their tions. respective impact on the kinetics of peptide cleavage, we Double-Filter-Binding Assay. For the high-affinity ATP monitored the ATPase and peptidase activities under limiting site binding experiment, 50 nM [R-32P]ATP was mixed with nucleotide concentrations, where only the high-affinity ATP- 0.005-6 µM Lon (27)in15µL of 50 mM HEPES at pH binding sites of Lon are occupied (Scheme 1). Under these 8.0, 5 mM magnesium acetate [Mg(OAc)2], 75 mM potas- conditions, although multiple rounds of peptide hydrolysis sium acetate (KOAc), and 2 mM dithiothreitol (DTT). A total occur, the rate constant is 10-fold lower than that obtained of 3 µL of the reactions (performed in triplicate) was spotted when both ATP-binding sites are occupied under saturating onto a piece of nitrocellulose mounted onto a dot-blot levels of ATP. Dependent upon the level of ATP saturation, apparatus (BioRad) with a piece of Immobilon Ny+ below Lon exhibits two distinct kinetic behaviors in its ATPase as described elsewhere (28, 29). All reactions were performed sites, with optimal peptide hydrolysis occurring upon full at least in triplicate. Each spot was washed with 10 µLof occupancy of ATP at both of the sites. Unexpectedly, ATP cold reaction buffer 2 times. The radioactive counts at each hydrolysis at the high-affinity sites is stimulated by peptide spot were quantified by PhosphorImaging using the Packard 4604 Biochemistry, Vol. 45, No. 14, 2006 Vineyard et al.
Cyclone storage phosphor system. The concentration of Chemical-Quench ATPase ActiVity Assays. The acid- bound was determined according to eq 1 quenched time courses for ATP hydrolysis were measured using a rapid-chemical-quench-flow instrument from KinTek NC [bound] ) dlu [[R-32P]ATP] (1) Corporation as described by Vineyard et al. (25). All ( + + ) solutions were made in 50 mM HEPES buffer at pH 8.0, 5 NCdlu NYdlu mM DTT, 5 mM Mg(OAc)2, and 75 mM KOAc. A 15 µL where NCdlu is the radioactive count on the nitrocellulose buffered solution of 6 µM Lon monomer or 6 µM Lon + preincubated with 6 µM ATP, with and without 500 µMS2 membrane and NYdlu is the radioactive count on the Immobilon Ny+ membrane. The binding parameters were or 20 µM casein, was rapidly mixed with a 15 µL buffered determined by fitting the data with eq 2 using the nonlinear solution of 100 µM ATP containing 0.01% of [R-32P]ATP regression analysis program Prism (GraphPad) software at 37 °C for varying times (0-3 s). The reactions were version 4 quenched with 0.5 N formic acid and then extracted with 200 µL of phenol/chloroform/isoamyl alcohol at pH 6.7 [RL] ) (25:24:1). A 3 µL aliquot of the aqueous solution was spotted directly onto a PEI-cellulose TLC plate and treated as above. ([R] + [L] + K ) - x([R] + [L] + K )2 - 4[R][L] d d (2) All assays were performed at least in triplicate, and the 2[L] average of those traces was used for data analysis. The burst amplitudes and burst rates were determined by fitting the R 32 where [L] is the concentration of [ - P]ATP, [R] is the k data from 0 to 400 ms with eq 5 concentration of Lon, [RL] is the concentration of [R-32P]- obs - ATP bound to Lon, and Kd is the equilibrium dissociation Y ) A exp kburstt + C (5) constant for ATP bound at the high-affinity site. Single-TurnoVer ATPase Assays. Single-turnover data for where t is time in seconds, Y is [ADP] in micromolar, A is ATP hydrolysis were measured as described elsewhere (23), the burst amplitude in micromolar, kburst is the burst rate and all reactions were performed at least in triplicate. Briefly, constant in s-1, and C is the end point. The observed steady- for the ATPase measurements, each reaction mixture (70 µL) state rate constants (kss,ATP) were determined by fitting the contained 50 mM HEPES (pH 8.0), 75 mM KOAc, 5 mM data from 600 ms to 3 s with the linear function, Y ) mX + Mg(OAc)2, 5 mM DTT, and 5 or 6 µM Lon monomer. For C, where X is time, Y is [ADP]/[Lon], m is the observed the peptide-stimulated ATPase reactions, 500 µM peptide steady-state rate constant in s-1, and C is the y intercept. substrate (S2) was added to each reaction mixture and the Data fitting was accomplished using the nonlinear regression reactions were initiated by the addition of [R-32P]ATP. program Prism (GraphPad) software version 4. Subsequently, 5 µL aliquots were quenched in 10 µLof0.5 Tryptic Digestions. Tryptic digest reactions in mixtures N formic acid at 12 time points (from 0 to 15 min). A 3 µL containing 6 µM Lon, 50 mM HEPES (pH 8.0), 5 mM aliquot of the reaction was spotted directly onto a PEI- magnesium acetate, 2 mM DTT, (500 µM S2 peptide, and cellulose TLC plate (10 × 20 cm), and the plate developed either 1 mM ATP, 6 µM ATP, or 500 nM ATP were started in 0.75 M potassium phosphate (KPi) buffer (pH 3.4). by the addition of 1/50 (w/w) TPCK (N-p-tosyl-L-phenyl- Radiolabeled ADP was then quantified using the Packard alanine chloromethyl ketone)-treated trypsin with respect to Cyclone storage phosphor screen Phosphor imager purchased Lon. At 0, 2, 4, 20, and 40 min, a 3 µL reaction aliquot was from Perkin-Elmer Life Science. To compensate for slight quenched in 3 µg of soybean trypsin inhibitor (SBTI) variations in spotting volume, the concentration of the ADP followed by boiling. The quenched reactions were then product obtained at each time point was corrected using an resoved by 12.5% SDS-PAGE analysis and visualized with internal reference as shown in eq 3 Coomassie brilliant blue. ADP ) dlu × RESULTS [ADP] ( + ) [ATP] (3) ATPdlu ADPdlu Examining Binding of the ATPase Sites in Lon. Although E. coli Lon contains one ATP-binding domain in each of its All assays were performed at least in triplicate, and the monomeric subunits, the existence of both a high- and low- kinetic parameters were determined by fitting the time-course affinity ATP-binding site is evident in its oligomeric form data with a single-exponential eq 4 using the nonlinear (25, 26). To verify the existence of two different ATPase regression program Prism (Graphpad) software version 4 sites in Lon under our reaction conditions, we measured the - R 32 Y ) A exp kobst + C (4) affinities of Lon for [ - P]ATP using a filter-binding assay adapted from the protocols of Jia et al. and Gilbert and where t is time in seconds, Y is [ADP] in micromolar, A is Mackey (27, 28) and Wong and Lohman (29). The half-life the amplitude in micromolar, kobs is the observed rate constant of the complex, where ATP is bound at the low-affinity sites, in s-1, and C is the end point. can be calculated using the off rate of ATP (Vineyard, D., Peptidase Methods. Peptidase activity was monitored on and Lee, I., manuscript in preparation). Because the half- a Fluoromax 3 spectrofluorimeter (Horiba Group) as de- life is on a millisecond time scale, the filter-binding assay scribed previously (22). Assays contained 50 mM HEPES is not an appropriate method for detecting the affinity of at pH 8.0, 75 mM KOAc, 5 mM DTT, 5 mM Mg(OAc)2,1 ATP to the low-affinity site. However, the binding to this mM S3 peptide (excitation at 320 nm and emission at 420 site has previously been determined under our reaction nm),5or6µM Lon, and either ATP or adenylyl 5-imido- conditions using steady-state kinetic methods (22), and the diphosphate (AMPPNP) (0-100 µM). resulting affinity agreed with the published value of 10 µM High- and Low-Affinity ATPase Sites in Lon Biochemistry, Vol. 45, No. 14, 2006 4605
FIGURE 2: Pre-steady-state time courses of ATPase activity of E. FIGURE 1: Determining the Kd for the high-affinity ATPase site in E. coli Lon using an adapted filter-binding assay. To monitor the coli Lon under single-turnover conditions. The time courses for binding of [R-32P]ATP to only the high-affinity site, various ATP hydrolysis at the high-affinity sites were determined by R 32 9 concentrations of Lon (0.005-6 µM) were incubated with 50 nM incubating 5 µM Lon with 500 nM [ - P]ATP in the absence ( ) 2 [R-32P]ATP at 4 °C. The amount of the Lon/[R-32P]ATP complex and presence ( )of500µM S2 peptide. The reactions were formed was quantified by PhosphorImaging of the nitrocellulose quenched with acid at the indicated times, and the concentrations R 32 membrane, and the free [R-32P]ATP was quantified by Phospho- of [ - P]ADP were determined by TLC followed by Phospho- rImaging of the positively charged Immobilon Ny+ membrane. The rImaging. The kobs values were determined by fitting the time amount of (bound) complex (2) was calculated as described in the courses using a single-exponential equation as described in the Materials and Methods, and the generated data were fit using a Materials and Methods, yielding observed rate constants of 0.006 ( 0.004 and 0.007 ( 0.003 s-1 in the absence and presence of the binding isotherm (eq 2). The resulting Kd value was 0.52 ( 0.096 µM for the high-affinity site. S2 peptide, respectively. The inset shows time courses for 500 nM [R-32P]ATP hydrolysis at the high-affinity sites in the presence of 500 µM S2 peptide at increasing concentrations of Lon: 5 µM by Menon and Goldberg (26). The equilibrium dissociation (2), 7 µM([), and 10 µM(0). The kobs values were determined constant for the high-affinity site was never specifically by fitting the time courses using a single-exponential equation as determined because of the limit of detection of the previous described in the Materials and Methods, yielding observed rate < constants of 0.007 ( 0.001, 0.007 ( 0.001, and 0.006 ( 0.001 assays (Kd 1µM) (26). Therefore, we utilized the filter- -1 binding assay to better define the binding affinity of Lon to s , respectively. the high-affinity ATPase site. occupied by the nucleotide (Kd ) 0.52 µM, see above). To probe the high-affinity site, 50 nM [R-32P]ATP was Enzyme forms A-D (Scheme 1) represent different ATP- incubated with varying amounts of Lon and the resulting bound states of Lon under the various reaction conditions Lon/[R-32P]ATP complex was immobilized onto a nitro- used in this study. The proposed enzyme form under limiting cellulose membrane, whereas unbound [R-32P]ATP was ATP conditions is shown as the enzyme form B in Scheme trapped by a positively charged nylon membrane placed 1(17); thus, we can selectively monitor ATP hydrolysis at directly below the nitrocellulose filter. The concentration of only the high-affinity sites. The inset of Figure 2 shows the Lon was varied rather than [R-32P]ATP to eliminate the high hydrolysis of ATP at the high-affinity sites as Lon is background generated when the concentration of [R-32P]ATP increased (5, 7, and 10 µM). The observed rate constants is increased because the concentration Lon is held constant range from 0.006 to 0.007 s-1 with a standard deviation of in the nanomolar range. A binding isotherm of the high- less than 0.8%. Because the rate constants are identical, the affinity ATP site in Lon was generated by quantifying the binding of ATP is not rate-limiting under the single-turnover amount of 32P immobilized onto the nitrocellulose versus reaction conditions employed in the experiment. Furthermore, the nylon membrane as described in the Materials and as shown in Figure 2, the presence of the S2 peptide does Methods. As shown in Figure 1, the binding of ATP at the not influence ATP hydrolysis at the high-affinity sites high-affinity site was detected with a Kd value of 0.52 ( because the rate constant for the reaction is 0.006 ( 0.0004 0.096 µM. Control experiments were performed to ensure and 0.007 ( 0.0003 s-1 in the absence and presence of a that no [R-32P]ATP hydrolysis was occurring under the saturating amount of S2 (500 µM), respectively. This is reaction conditions (data not shown). These results are consistent with our previous observation that the burst rate consistent with the earlier observation reported by Menon constant associated with ATP hydrolysis under pseudo-first- and Goldberg (26) that Lon contains two affinities for ATP, order conditions was not affected by the presence of the S2 which differ from one another by approximately 10-fold. peptide (25). The rate constant for S2-stimulated ATP Examining ActiVity of the High-Affinity ATPase Sites of hydrolysis at the high-affinity sites under single-turnover -1 Lon. Kinetic analyses performed under pseudo-first-order conditions (kobs ) 0.007 ( 0.0003 s ), however, is consider- conditions, where ATP is in excess over the enzyme ably slower than the burst rate obtained when ATP was in concentration, have revealed an apparent functional asym- excess over the enzyme concentration (kburst ) 11.3 ( 3.3 metry in the ATPase sites (25). We examined the ATPase s-1)(25). These two experiments differ only by the oc- activity of the high- and low-affinity sites independently of cupancy of ATP at the low-affinity sites. This implies that, one another by manipulating the concentration of nucleotide although one ATP-binding site exists per monomer, two such that either just the high-affinity sites or both sets of functionally distinct ATPase sites are evident in the homo- sites were occupied. To monitor the ATPase activity at the oligomeric form of Lon. high-affinity sites of Lon, single-turnover experiments were Because the hydrolysis at the high-affinity ATPase sites employed. When the concentrations of the reactants are is minimal and ATP hydrolysis is required for optimal adjusted such that the Lon concentration (5 µM) is in excess peptide cleavage (23), we questioned whether the catalytic over limiting ATP (500 nM), only the high-affinity sites are efficiency of S3 cleavage is coupled with ATP hydrolysis 4606 Biochemistry, Vol. 45, No. 14, 2006 Vineyard et al.
FIGURE 3: S3 hydrolysis by E. coli Lon under limiting nucleotide FIGURE 4: Representative E. coli Lon time courses of ATP conditions. The 5 µM Lon monomer was incubated with 1 mM S3 hydrolysis at the high-affinity sites. [R-32P]ATP (6 µM) was peptide in the presence of 0, 0.5, and 100 µM ATP and 0.5 and incubated with 6 µM monomeric Lon in the absence (9) or presence 100 µM AMPPNP. The fluorescence changes associated with (2)of500µM S2 peptide and quenched with acid at varying time peptide cleavage were converted to product concentrations as points. To see the effect of nucleotide occupation at the low-affinity described in the Materials and Methods. The kobs values associated sites on the high-affinity site ATP hydrolysis, 100 µM ATP was with each trace are summarized in Table 1. added at 1 half-life into the reaction 60 s in both the absence (1) and presence ([)of500µM S2 peptide to saturate the low-affinity 1 Table 1: Rate Constants Associated with Peptidase Activity ATPase sites. The time of addition of the 100 µM ATP in traces and [ is indicated by the arrow. As described in the Materials and -1 [nucleotide] (µM) kobs,S3 (s ) Methods, the time courses were fit using the equation Y ) A - ( kobs,ATPt) + limiting ATP 0.5 0.32 ( 0.07 exp C, where A is the amplitude, kobs,ATP is the observed stoichiometric ATP 5 1.52 ( 0.05 rate constant, and C is the endpoint. The resulting rate constants saturating ATP 100 2.69 ( 0.30 are summarized in Table 2. The time points reported are an average saturating AMPPNP 100 0.96 ( 0.08 of at least three different trials. at the high-affinity sites. To address this issue, we monitored Table 2: Rate Constants Associated with High-Affinity Site ATPase the kinetics of S3 cleavage under single-turnover conditions, Activity where ATP is limiting (5 µM Lon, 500 nM ATP, and 1 mM intrinsic S2 stimulated -1 -1 S3). Under these conditions, the predominant enzyme form kobs,ATP (s ) kobs,ATP (s ) is homo-oligomeric Lon, with ATP bound only at the high- stoichiometric ATP 0.011 ( 0.001 0.019 ( 0.002 affinity sites (enzyme form B in Scheme 1). As shown in 100 µM ATP chase 0.012 ( 0.001 0.017 ( 0.001 100 µM AMPPNP chase 0.010 ( 0.001 0.014 ( 0.001 Figure 3, although the rate constant for S3 cleavage at ( ( -1 100 µM ADP chase 0.015 0.001 0.017 0.001 limiting ATP concentrations (Table 1, kobs,S3 ) 0.32 s )is slower than with saturating ATP, enzyme form D in Scheme -1 R 32 1 (Table 1, kobs,S3 ) 2.69 s ), Lon is undergoing multiple the concentration of [ - P]ATP was raised to approximately rounds of peptide cleavage with only limiting amounts of 10-fold excess of the Kd (6 µM), which is stoichiometric to ATP. Because no peptide hydrolysis occurs in the presence the amount of Lon in the reaction. Shown as enzyme form of limiting amounts of the nonhydrolyzable ATP analogue, C in Scheme 1, under these conditions, it is assumed that AMPPNP, (Figure 3) at least one molecule of ATP must be the high-affinity sites are saturated and the low-affinity sites hydrolyzed for peptide cleavage to occur under these are left unoccupied. conditions. Although, nonstoichiometric processing of ATP The experiment shown in Figure 4 shows the effect on and the S3 peptide is observed under saturating (100 µM) ATP hydrolysis at the high-affinity sites when the low- ATP conditions, the single-turnover data much more clearly affinity sites were subsequently occupied with unlabeled demonstrate that ATP and peptide hydrolysis are not nucleotide. To accomplish this, ATP hydrolysis was mea- stoichiometric. As previously reported and shown here in sured at the high-affinity sites under stoichiometric [R-32P]- Figure 3 for a comparison to the limiting nucleotide ATP/Lon conditions (enzyme form C in Scheme 1), while conditions, saturating amounts of AMPPNP (100 µM) saturating (100 µM) unlabeled ATP was subsequently added support S3 hydrolysis at a lower rate than under saturating or chased to occupy the weak-affinity sites 1 min into the ATP conditions (100 µM; Table 1) (22). In this case, reaction (enzyme form D in Scheme 1). When the experiment saturating AMPPNP most likely supports slow peptide is performed in this manner, the hydrolysis at the high- hydrolysis because its binding at the low-affinity sites in affinity sites can be monitored for the first half-life of the addition to not generating ADP from the lack of hydrolysis reaction and then the effect of nucleotide occupation at the is sufficient to lock Lon into an active conformation (22). low-affinity sites on this hydrolysis can subsequently be seen. Binding and Hydrolysis of ATP at the Tight Sites Are To ensure that 60 s was an appropriate time to add saturating Independent of Nucleotide Binding and Hydrolysis at the nucleotide, a control experiment was done where the Weak Sites. To further probe the functional nonequivalency saturating nucleotide was added at 10 s and no difference of the two ATPase sites, the activity of each site was was noted (data not shown). The rate constants associated monitored independently of the other using radiolabeled ATP with the hydrolysis of [R-32P]ATP at the high-affinity sites as a selective probe. Because the Kd for binding ATP at the alone as well as after subsequent occupation of the low- high-affinity sites was 0.52 ( 0.096 µM, the single-turnover affinity sites with nucleotide are summarized in Table 2. experimental conditions employed above were not sufficient These experiments were performed in both the presence (S2 to saturate those sites. To detect the full effect of ATP stimulated) and absence (intrinsic) of the S2 peptide substrate hydrolysis when the high-affinity sites were fully occupied, to account for effects of any interaction between the ATPase High- and Low-Affinity ATPase Sites in Lon Biochemistry, Vol. 45, No. 14, 2006 4607 when the high-affinity sites are occupied by unlabeled ADP (2). Hydrolysis of [R-32P]ATP at the low-affinity sites exhibited an initial burst in [R-32P]ADP production, which when fit using eq 6 from the Materials and Methods, yielded an observed burst rate constant of 17.2 ( 0.09 s-1 (2), which is comparable to the value of 15.9 ( 0.07 s-1 (9), where both the high- and low-affinity sites are contributing. As expected, the rate constant obtained here, 15.9 ( 0.07 s-1 (9), agrees well with the previously determined value of 11.3 ( 3.3 s-1 (25). The pre-steady-state time course showing the hydrolysis of ATP at only the low-affinity ATPase sites FIGURE 5: E. coli Lon pre-steady-state chemical-quenched time 2 R 32 ( ) mirrored the pre-steady-state time course reflecting courses of ATP hydrolysis at the low-affinity sites. [ - P]ATP 9 (100 µM) was incubated with 6 µM monomeric Lon (9)or6µM activity at both sites ( ). This would imply that the pre- monomeric Lon preincubated with 6 µM ATP (2) in the presence steady-state burst in ADP production is coming from only of 500 µM S2 peptide as described in the Materials and Methods. the low-affinity ATPase sites. As noted in the previous The preincubation of 6 µM Lon with 6 µM ATP presumably publication, the ATPase burst activity of Lon is unusual resulted in 6 µM Lon/6 µM ADP, where the high-affinity ATPase because it shows half-burst amplitude and the time course sites were saturated. The reactions were quenched with acid at the indicated times, and the concentrations of [R-32P]ADP generated is triphasic (25). The triphasic time course showed a burst in the reactions were determined by TLC followed by Phospho- in ADP production followed by an intermediate slow phase, rImaging. The time courses from 0 to 400 ms were fit with the and then steady-state ATP turnover occurs. The time courses, - equation Y ) A exp( kburstt) + C, where A is the burst amplitude, therefore, cannot be fit using the classical burst equation kburst is the observed burst rate constant, and C is the endpoint. The -1 because of the intermediate slow phase. As described in the resulting kburst for Lon (9) was 15.9 ( 0.07 s , and the resulting -1 Materials and Methods, the time courses are instead split kburst for 1 Lon/1 ADP (2) was 17.2 ( 0.09 s . The kss,ATP values were obtained by fitting the time courses from 600 ms to 3 s with into the pre-steady-state burst phase, which is fit using a a linear function and dividing the slope by the [Lon] in the reaction single-exponential equation, and a linear steady-state phase. ( -1 9 ( -1 and were 0.40 0.021 s for Lon ( ) and 0.11 0.014 s for The kinetics of each site has been monitored in the 1 Lon/1 ADP (2). The time points reported here are an average of at least three different trials. presence and absence of nucleotide occupation at the other site, and the occupancy of ADP at the high-affinity sites does and peptidase activities. As illustrated in Table 2, the not affect ATP hydrolysis at the low-affinity sites. Therefore, observed rate constant for ATP hydrolysis at the high-affinity we have now demonstrated that the ATPase sites hydrolyze sites was unaffected by the occupation of the low-affinity ATP independently of one another and the rate of ATP sites 60 s into the reaction with ATP (Figure 4), AMPPNP hydrolysis at the high-affinity sites is much slower. This (figure not shown), or ADP (figure not shown). Because both behavior explains the triphasic time courses observed in the a nonhydrolyzable ATP analogue (AMPPNP) and a product pseudo-first-order experiments as well as the independence inhibitor (ADP) were also tested, these data indicate that ATP of the burst rate constant on the concentration of ATP, which hydrolysis at the high-affinity sites is independent of ATP have been previously noted (25). Taken together with the binding and/or hydrolysis at the low-affinity sites. The rate pre-steady-state characterization of peptide cleavage deter- constant for ATP hydrolysis at the high-affinity sites was mined previously, we conclude that ATP hydrolysis occurs slightly faster in the presence of the S2 peptide (Table 2) at the low-affinity sites prior to S3 cleavage. and casein (data not shown), thus suggesting that com- Interestingly, there is an observed 4-fold stimulation in munication occurs as a result of the peptide or protein the steady-state rate in the presence of S2 peptide (Figure interacting with the high-affinity ATPase site of Lon. 5) or the unstructured protein substrate casein (30, 31; data Peptide and Protein Substrates Stimulate ATP Hydrolysis not shown). This is consistent with the stimulation observed at the High-Affinity Sites. In the previous experiment, which in the kcat for ATPase activity (23), but it is only observed was depicted in Figure 4 for ATP, 100 µM AMPPNP, ATP, when the high-affinity sites are not occupied by ADP (Figure 32 -1 -1 and ADP did not compete out the [R- P]ATP bound at the 4, kss,ATP ) 0.40 ( 0.021 s , 9; kss,ATP ) 0.11 ( 0.014 s , high-affinity sites because the amplitudes of the time courses 2). This supports the implication that, although the hydrolysis were unaffected. The converse experiment could then be at the high-affinity sites is unaffected by nucleotide occupa- employed to monitor ATP hydrolysis at the low-affinity sites. tion at the low-affinity sites, it is affected by the presence In this experiment, Lon (6 µM) was preincubated with a of the peptide or protein substrate. stoichiometric amount of unlabeled ATP (6 µM; enzyme Optimal Peptidase ActiVity Requires ATP Binding and form C in Scheme 1). During this preincubation, only the Hydrolysis at Both Sites. Communication occurring between high-affinity sites are occupied with unlabeled ATP, which the high-affinity ATPase sites and the S2 peptide or protein was hydrolyzed to ADP that remains bound at the high- substrate has been suggested by the experiments performed affinity sites. As such, the subsequent hydrolysis of 100 µM above. Therefore, the catalytic efficiency of S3 cleavage was [R-32P]ATP by Lon would directly reflect the ATPase activity examined under stoichiometric Lon/ATP conditions as well at the low-affinity sites. To ensure that the ATP hydrolysis (enzyme form C in Scheme 1). The kinetics of S3 cleavage at the high-affinity sites, which occurred during the prein- were monitored under conditions where the high-affinity sites cubation, was not necessary, the experiment was also were saturated with ATP (5 µM Lon, 5 µM ATP, and 1 mM performed where Lon (6 µM) was preincubated with ADP S3), and the proposed enzyme form is shown in Scheme 1. (6 µM) and the traces were identical (data not shown). Figure The rate constants for peptide hydrolysis under these 5 shows the hydrolysis of [R-32P]ATP at the low-affinity sites conditions are summarized in Table 1. Because the rate at 4608 Biochemistry, Vol. 45, No. 14, 2006 Vineyard et al.
FIGURE 6: Limited tryptic digestion of Lon in the presence of varying amounts of ATP. Lon in the presence of 500 µM S2 peptide was digested with a limiting amount of trypsin and quenched at the indicated times with SBTI as described in the Materials and Methods. The first lane shows the molecular markers in kilodaltons (from top to bottom): 172, 110, 79, 62, 48, 37, 24, and 19. 100 µM ATP (enzyme form D in Scheme 1) is faster than ATP (lanes 10-12), and saturating (1 mM) ATP (lanes 13- at 5 µM ATP (enzyme form C in Scheme 1), both ATPase 15). The digest pattern was identical in the absence of the sites are contributing to the peptidase activity. Optimal S2 peptide (data not shown). In accordance with what was peptide degradation is consequently recovered by the low- previously observed, limited tryptic digestion of Lon in the affinity site ATPase activity. This means that the maximal presence of ATP yields fragments varying from 23 to 67 rate of peptide degradation is attained only when both the kDa. As indicated by these data, the 67 kDa fragment is high- and low-affinity sites are saturated with ATP. Because substantially stabilized only by full occupation of the low- peptidase activity is slower in the presence of the non- affinity ATP sites. Because binding and hydrolysis of ATP hydrolyzable analogue, AMPPNP, ATP binding as well as at all ATPase sites are also necessary for optimal peptidase hydrolysis at the low-affinity sites are a necessity for optimal activity in Lon, a correlation between this adenine-specific peptidase activity. conformational change and accessibility for peptide cleavage Tryptic Digest Probes the Conformational Change As- could exist. sociated with ATP Binding. Previously (23), we utilized limited tryptic digestion to probe the functional role of DISCUSSION nucleotide binding to Lon. We detected an adenine-specific In this study, we utilized kinetic techniques to better conformational change at saturating amounts of nucleotide. understand the function of the two nonequivalent ATPase Although the digestion yielded a pattern of bands ranging sites in E. coli Lon protease and their coordination with the in size from 7 to 67 kDa, the adenine-specific conformational protease activity of the enzyme. Despite being a homo- change was monitored primarily by the detection of a stable oligomer, with one ATP-binding site per monomer, Lon 67 kDa fragment. When sequenced, the 67 kDa fragment contains two types of ATPase sites that are functionally or included all domains of Lon (ATPase, SSD, and peptidase), kinetically distinct from one another. As such, the kinetic except the amino-terminal region. Of all of the nucleotides characterization of the activities of the ATPase sites is tested in that study, ATP was shown to be the best activator imperative because they are structurally indistinguishable. of the peptidase activity of Lon. Therefore, it was concluded The experiments represented in Figure 4 demonstrate that that the adenine-specific conformational change contributed ATP hydrolysis at the high-affinity site is unaffected by the to maintaining the optimal catalytic efficiency of S3 cleavage. occupation of nucleotide at the low-affinity site, while the In light of the detection of functional nonequivalency in experiments exemplified in Figure 5 show that hydrolysis the ATP binding and hydrolysis activity in Lon, in this study, of ATP at the low-affinity site is unaffected by the occupation we questioned whether the previously observed conforma- of nucleotide at the high-affinity site. Therefore, the kinetic tional change upon nucleotide binding could be assigned to data support the conclusion that the two ATPase sites are a specific interaction between ATP with either the high- or functioning independent of the other. Our kinetic experiments low-affinity ATPase sites in Lon. To address this issue, we have also allowed us to examine the coupling of the ATPase subjected 6 µM Lon to limited tryptic digestion in the and peptidase activities of Lon protease, which are not yet presence of limiting (500 nM; enzyme form B in Scheme fully understood. Mechanistic studies of other enzymes 1), stoichiometric (6 µM; enzyme form C in Scheme 1), and including Rho protein (32, 33), multidrug-resistance-associ- excess (1 mM) amounts of ATP (enzyme form D in Scheme ated protein (MRP1) (34), P-glycoprotein (PGP) (35), F1- 1) to see the effect on the stability of the adenine-specific ATPase (36), Na+/K+ ATPase (37), topoisomerase II (38, conformational change when one or both ATPase sites were 39), and Rep helicase (40) have also revealed the functional occupied. On the basis of the two Kd values of ATP, we roles of multiple ATPase sites. anticipated that only the tight sites of Lon were occupied by Previously, this lab has constructed a minimal kinetic ATP in the first two cases. Figure 6 shows the Lon fragments model to account for the ATP-dependent S3 cleavage by Lon generated over increasing time in the presence of the S2 using steady-state kinetics (22). Although this model was peptide under conditions of no nucleotide (lanes 2-5), constructed assuming monomeric Lon subunits had equiva- limiting (500 nM) ATP (lanes 6-9), stoichiometric (6 µM) lent ATPase activity, the order of events has been confirmed High- and Low-Affinity ATPase Sites in Lon Biochemistry, Vol. 45, No. 14, 2006 4609 by our current pre-steady-state kinetic studies. This minimal ACKNOWLEDGMENT kinetic model predicted ATP hydrolysis to occur prior to We thank Dr. Anthony Berdis for his assistance and careful peptide cleavage. Current as well as previous pre-steady- reading of this manuscript. state kinetic studies (25) are consistent with this model because E. coli Lon exhibits lag kinetics in the degradation REFERENCES of the model peptide (S3) and burst kinetics in ATP hydrolysis. The burst kinetics indicate a rapid buildup of a 1. Charette, M. F., Henderson, G. W., and Markovitz, A. (1981) ATP hydrolysis-dependent protease activity of the lon (capR) protein reaction intermediate with a rate-limiting step following of Escherichia coli K-12, Proc. Natl. Acad. Sci. U.S.A. 78, 4728- chemistry in the reaction pathway, while the lag kinetics for 4732. S3 peptide degradation are consistent with a need for an 2. Chung, C. H., and Goldberg, A. L. (1981) The product of the lon accumulation of a reaction intermediate prior to peptide (capR) gene in Escherichia coli is the ATP-dependent protease, protease La, Proc. Natl. Acad. Sci. U.S.A. 78, 4931-4935. hydrolysis. The burst kinetics for ATP hydrolysis demon- 3. Goff, S. A., and Goldberg, A. L. (1985) Production of abnormal strated triphasic behavior, where only 50% of the Lon proteins in E. coli stimulates transcription of lon and other heat monomer was hydrolyzed in the duration of the lag in shock genes, Cell 41, 587-595. peptidase activity. Although this unusual behavior suggested 4. Goldberg, A. L., Moerschell, R. P., Chung, C. H., and Maurizi, M. R. (1994) ATP-dependent protease La (lon) from Escherichia the contribution of two ATPase activities, it could not coli, Methods Enzymol. 244, 350-375. separate the two. 5. Goldberg, A. L., and Waxman, L. (1985) The role of ATP The present study has confirmed the existence of two hydrolysis in the breakdown of proteins and peptides by protease - ATPase sites in Lon and that their activities are independent La from Escherichia coli, J. Biol. Chem. 260, 12029 12034. 6. Gottesman, S. (1996) Proteases and their targets in Escherichia of one another. We have now demonstrated with additional coli, Annu. ReV. Genet. 30, 465-506. pre-steady-state experiments that the ATP bound to the low- 7. Gottesman, S., Gottesman, M., Shaw, J. E., and Pearson, M. L. affinity sites is hydrolyzed during the lag in peptide cleavage (1981) Protein degradation in E. coli: The Ion mutation and - ( -1 bacteriophage λ N and cll protein stability, Cell 24, 225 233. with a burst rate constant of 17.2 0.09 s . The peptide is 8. Gottesman, S., and Maurizi, M. R. (1992) Regulation by pro- concomitantly cleaved with a rate constant of 2.69 ( 0.30 teolysis: Energy-dependent proteases and their targets, Microbiol. s-1, while the ATP bound at the high-affinity sites is slowly ReV.56, 592-621. ( 9. Maurizi, M. R. (1992) Proteases and protein degradation in being hydrolyzed with an observed rate constant of 0.019 - -1 Escherichia coli, Experientia 48, 178 201. 0.002 s . The previously determined minimal kinetic model 10. Schoemaker, J. M., Gayda, R. C., and Markovitz, A. (1984) also predicted a Lon/ATP-bound “F” form of the enzyme Regulation of cell division in Escherichia coli: SOS induction following peptide cleavage, which could undergo multiple and cellular location of the sulA protein, a key to lon-associated rounds of peptide hydrolysis before reverting back to the filamentation and death, J. Bacteriol. 158, 551-561. 11. Ogura, T., and Wilkinson, A. J. (2001) AAA+ superfamily free enzyme (22). This is consistent with our data in Figure ATPases: Common structuresDiverse function, Genes Cells 6, 3, which demostrate that peptide and ATP hydrolysis are 575-597. not stoichiometrically linked. It is still not clear how the 12. Kunau, W. H., Beyer, A., Franken, T., Gotte, K., Marzioch, M., enzyme turns over once it reaches the “F” form. Our revised Saidowsky, J., Skaletz-Rorowski, A., and Wiebel, F. F. (1993) Two complementary approaches to study peroxisome biogenesis mechanism for the ATPase activity in Lon predicts that the in Saccharomyces cereVisiae: Forward and reversed genetics, two ATPase sites are hydrolyzing ATP sequentially but are Biochimie 75, 209-224. not communicating. However, binding and hydrolysis of ATP 13. Neuwald, A. F., Aravind, L., Spouge, J. L., and Koonin, E. V. (1999) AAA+: A class of chaperone-like ATPases associated at both the high- and low-affinity ATPase sites are necessary with the assembly, operation, and disassembly of protein com- for optimal peptide or protein degradation. The energy plexes, Genome Res. 9,27-43. generated from ATP hydrolysis at the low-affinity sites could 14. Amerik, A., Chistiakov, L. G., Ostroumova, N. I., Gurevich, A. be used to translocate the protein substrate or may be used I., and Antonov, V. K. (1988) Cloning, expression and structure of the functionally active shortened lon gene in Escherichia coli, to induce a conformational change in the oligomer that Bioorg. Khim. 14, 408-411. facilitates protein cleavage. Therefore, the development of 15. Chin, D. T., Goff, S. A., Webster, T., Smith, T., and Goldberg, an effective continuous assay to measure protein degradation A. L. (1988) Sequence of the lon gene in Escherichia coli.A is currently underway to determine how the high- and low- heat-shock gene which encodes the ATP-dependent protease La, J. Biol. Chem. 263, 11718-11728. affinity ATPase sites are contributing to protease activity. 16. Botos, I., Melnikov, E. E., Cherry, S., Khalatova, A. G., Rasulova, Because the protease activity is ultimately limited by F. S., Tropea, J. E., Maurizi, M. R., Rotanova, T. V., Gustchina, turnover of ATP hydrolysis, further kinetic characterization A., and Wlodawer, A. (2004) Crystal structure of the AAA+R domain of E. coli Lon protease at 1.9 Å resolution, J. Struct. Biol. needs to be performed to delineate the ATPase mechanism. 146, 113-122. We and others (4) have previously suggested that ADP 17. Botos, I., Melnikov, E. E., Cherry, S., Tropea, J. E., Khalatova, release was the rate-limiting step along the reaction pathway A. G., Rasulova, F., Dauter, Z., Maurizi, M. R., Rotanova, T. V., Wlodawer, A., and Gustchina, A. (2004) The catalytic domain of because it is a potent inhibitor of peptidase activity (Ki,ADP ) Escherichia coli Lon protease has a unique fold and a Ser-Lys 0.3 µM) (22). The burst kinetics associated with ATP dyad in the active site, J. Biol. Chem. 279, 8140-8148. hydrolysis would be consistent with this, because they 18. Li, M., Rasulova, F., Melnikov, E. E., Rotanova, T. V., Gustchina, indicated that a step following chemistry is rate-limiting. A., Maurizi, M. R., and Wlodawer, A. (2005) Crystal structure However, the rapid-quench experiment shown in Figure 5 of the N-terminal domain of E. coli Lon protease, Protein Sci. 14, 2895-2900. showed that ATP did not compete out ADP bound to the 19. Menon, A. S., Waxman, L., and Goldberg, A. L. (1987) The energy high-affinity sites of Lon, suggesting that ADP is perhaps utilized in protein breakdown by the ATP-dependent protease (La) only inhibiting at the high-affinity sites. Experiments to from Escherichia coli, J. Biol. Chem. 262, 722-726. 2+ determine other microscopic rate constants along the ATPase 20. Rudyak, S. G., Brenowitz, M., and Shrader, T. E. (2001) Mg - linked oligomerization modulates the catalytic activity of the Lon reaction pathway including ADP release are currently (La) protease from Mycobacterium smegmatis, Biochemistry 40, underway to determine the rate-limiting step of the reaction. 9317-9323. 4610 Biochemistry, Vol. 45, No. 14, 2006 Vineyard et al.
21. van Dijl, J. M., Kutejova, E., Suda, K., Perecko, D., Schatz, G., 32. Kim, D. E., and Patel, S. S. (1999) The mechanism of ATP and Suzuki, C. K. (1998) The ATPase and protease domains of hydrolysis at the noncatalytic sites of the transcription termination yeast mitochondrial Lon: Roles in proteolysis and respiration- factor Rho, J. Biol. Chem. 274, 32667-32671. dependent growth, Proc. Natl. Acad. Sci. U.S.A. 95, 10584-10589. 33. Stitt, B. L., and Xu, Y. (1998) Sequential hydrolysis of ATP 22. Thomas-Wohlever, J., and Lee, I. (2002) Kinetic characterization molecules bound in interacting catalytic sites of Escherichia coli of the peptidase activity of Escherichia coli Lon reveals the transcription termination protein Rho, J. Biol. Chem. 273, 26477- mechanistic similarities in ATP-dependent hydrolysis of peptide 26486. and protein substrates, Biochemistry 41, 9418-9425. 34. Yang, R., Cui, L., Hou, Y. X., Riordan, J. R., and Chang, X. B. 23. Patterson, J., Vineyard, D., Thomas-Wohlever, J., Behshad, R., (2003) ATP binding to the first nucleotide binding domain of Burke, M., and Lee, I. (2004) Correlation of an adenine-specific multidrug resistance-associated protein plays a regulatory role at conformational change with the ATP-dependent peptidase activity low nucleotide concentration, whereas ATP hydrolysis at the of Escherichia coli Lon, Biochemistry 43, 7432-7442. second plays a dominant role in ATP-dependent leukotriene C4 24. Lee, I., and Berdis, A. J. (2001) Adenosine triphosphate-dependent transport, J. Biol. Chem. 278, 30764-30771. degradation of a fluorescent λ N substrate mimic by Lon protease, 35. Sauna, Z. E., and Ambudkar, S. V. (2000) Evidence for a Anal. Biochem. 291,74-83. requirement for ATP hydrolysis at two distinct steps during a 25. Vineyard, D., Patterson-Ward, J., Berdis, A. J., and Lee, I. (2005) single turnover of the catalytic cycle of human P-glycoprotein, Monitoring the timing of ATP hydrolysis with activation of peptide Proc. Natl. Acad. Sci. U.S.A. 97, 2515-2520. cleavage in Escherichia coli Lon by transient kinetics, Biochem- 36. Weber, J., and Senior, A. E. (2001) Bi-site catalysis in F1- istry 44, 1671-1682. ATPase: Does it exist? J. Biol. Chem. 276, 35422-35428. 26. Menon, A. S., and Goldberg, A. L. (1987) Binding of nucleotides 37. Linnertz, H., Urbanova, P., and Amler, E. (1997) Quenching of to the ATP-dependent protease La from Escherichia coli, J. Biol. 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole-modified Na+/K+-ATPase Chem. 262, 14921-14928. reveals a higher accessibility of the low-affinity ATP-binding site, 27. Jia, Y., Kumar, A., and Patel, S. S. (1996) Equilibrium and FEBS Lett. 419, 227-230. stopped-flow kinetic studies of interaction between T7 RNA 38. Harkins, T. T., Lewis, T. J., and Lindsley, J. E. (1998) Pre-steady- polymerase and its promoters measured by protein and 2-amino- state analysis of ATP hydrolysis by Saccharomyces cereVisiae purine fluorescence changes, J. Biol. Chem. 271, 30451-30458. DNA topoisomerase II. 2. Kinetic mechanism for the sequential 28. Gilbert, S. P., and Mackey, A. T. (2000) Kinetics: A tool to study hydrolysis of two ATP, Biochemistry 37, 7299-7312. - molecular motors, Methods 22, 337 354. 39. Harkins, T. T., and Lindsley, J. E. (1998) Pre-steady-state analysis 29. Wong, I., and Lohman, T. M. (1993) A double-filter method for of ATP hydrolysis by Saccharomyces cereVisiae DNA topo- - nitrocellulose-filter binding: Application to protein nucleic acid isomerase II. 1. A DNA-dependent burst in ATP hydrolysis, - interactions, Proc. Natl. Acad. Sci. U.S.A. 90, 5428 5432. Biochemistry 37, 7292-7298. 30. Bhattacharyya, J., and Das, K. P. (1999) Molecular chaperone- 40. Wong, I., and Lohman, T. M. (1997) A two-site mechanism for like properties of an unfolded protein, R(s)-casein, J. Biol. Chem. - ATP hydrolysis by the asymmetric Rep dimer P2S as revealed 274, 15505 15509. by site-specific inhibition with ADP-A1F4, Biochemistry 36, 31. Creamer, L. K., Richardson, T., and Parry, D. A. (1981) Secondary 3115-3125. structure of bovine R s1- and â-casein in solution, Arch. Biochem. Biophys. 211, 689-696. BI052377T Detection and Characterization of Two ATP-Dependent Conformational Changes in Proteolytically Inactive Escherichia coli Lon Mutants by Stopped Flow Kinetic Techniques† Jessica Patterson-Ward, Jon Huang, and Irene Lee* Department of Chemistry, Case Western ReserVe UniVersity, CleVeland, Ohio 44106 ReceiVed August 15, 2007; ReVised Manuscript ReceiVed September 21, 2007
ABSTRACT: Lon is an ATP dependent serine protease responsible for degrading denatured, oxidatively damaged and certain regulatory proteins in the cell. In this study we exploited the fluorescence properties of a dansylated peptide substrate (S4) and the intrinsic Trp residues in Lon to monitor peptide interacting with the enzyme. We generated two proteolytically inactive Lon mutants, S679A and S679W, where the active site serine is mutated to an Ala and Trp residue, respectively. Stopped-flow fluorescence spectroscopy was used to identify key enzyme intermediates generated along the reaction pathway prior to peptide hydrolysis. A two-step peptide binding event is detected in both mutants, where a conformational change occurs after a rapid equilibrium peptide binding step. The Kd for the initial peptide binding step determined by kinetic and equilibrium binding techniques is approximately 164 micromolar and 38 micromolar, respectively. The rate constants for the conformational change detected in the S679A and S679W Lon mutants are 0.74 ( 0.10 s-1 and 0.57 ( 0.10 s-1, respectively. These values are comparable to the lag -1 rate constant determined for peptide hydrolysis (klag ∼ 1s ) [Vineyard, D., et al. (2005) Biochemistry 45, 4602-4610]. Replacement of the active site Ser with Trp (S679W) allows for the detection of an ATP-dependent conformational change within the proteolytic site. The rate constant for this conformational change is 7.6 ( 1.0 s-1, and is essentially identical to the burst rate constant determined for ATP hydrolysis under comparable reaction conditions. Collectively, these kinetic data support a mechanism by which the binding of ATP to an allosteric site on Lon activates the proteolytic site. In this model, the energy derived from the binding of ATP minimally supports peptide cleavage by allowing peptide substrate access to the proteolytic site. However, the kinetics of peptide cleavage are enhanced by the hydrolysis of ATP.
Lon, also known as protease La, is a homo-oligomeric (10, 11) as well as a stress-response protease necessary for ATP1 dependent serine protease localized in the cytosol of virulence in Brucella abortus (12). In mammalian cells, Lon prokaryotes and the mitochondria of eukaryotes (3-9). As is found in the mitochondria and downregulation of Lon a member of the AAA+ (ATPases associated with a variety disrupts mitochondrial function and causes cell death (13). of cellular activities) family of proteases, Lon’s primary Each enzyme subunit of Lon consists of an N-terminal function is to degrade denatured, oxidatively damaged and domain with unknown function, an ATPase domain, an SSD certain regulatory proteins in the cell. Lon has been shown (substrate sensor and discriminatory) domain and a protease to be important for Salmonella enterica infection in mice domain (14, 15). According to the crystal structure of the protease domain (16) and electron microscopy studies (17), † This work was supported by the NIH Grant GM067172. Escherichia coli (E. coli) Lon is a hexamer consisting of * Corresponding author. Phone: 216-368-6001. E-mail: Irene.lee@ case.edu. Fax: 216-368-3006. six identical subunits that adopt a self-compartmentalized 1 Abbreviations: ATP, adenosine triphosphate; ADP, adenosine structure found in many ATP dependent proteases. Mecha- diphosphate; AMPPNP, adenylyl 5-imidodiphosphate; DTT, dithio- nistic characterization of Lon from E. coli reveals that the threitol; Abz, anthranilamide; Bz, benzoic acid; MANT, 2′(or 3′)-O- (N-methylanthraniloyl); HEPES, N-2-hydroxyethylpiperazine-N′-ethane- protease utilizes a Ser-Lys dyad (S679 and K722) to catalyze sulfonic acid; KPi, potassium phosphate; Mg(OAc)2, magnesium acetate; peptide bond hydrolysis (16). Despite the presence of this KOAc, potassium acetate; PEI-cellulose, polyethyleneimine-cellulose; catalytic dyad, the hydrolytic activity of Lon remains dormant dlu, density light unit; TPCK, N-p-tosyl-L-phenylalanine chloromethyl ketone; SBTI, soybean trypsin inhibitor; λN, also known as the lambda until the enzyme binds and hydrolyzes ATP at a site distal N protein, a lambda phage protein that allows E. coli RNA polymerase from the protease domain. In fact, optimal peptide bond to transcribe through termination signals in the early operons of the cleavage is accompanied by ATP hydrolysis. However, the phage; FRET, fluorescence resonance energy transfer; S1, a fluorescent ATPase and peptidase activities of Lon are not stoichiomet- peptide substrate of Lon developed by our lab based on the λN protein (3-NO2)YRGITCSGRQK(Abz), (1)); S2, a nonfluorescent analogue of rically linked, as certain nonhydrolyzable ATP analogues S1 (YRGITCSGRQK(Bz), (2)); S3, 10% S1 + 90% S2 (this substrate such as AMPPNP can activate the degradation of unstruc- mixture was used to correct for the inner filter effect otherwise observed tured protein and peptide substrates (1, 2, 18). Furthermore, at high peptide substrate concentration; (2)); S4 is a peptide with the same sequence as S2 with a dansyl fluorophore (YRGITCSGRQK- substitution of the proteolytic site Ser with Ala affects only (dansyl)) and was used in this study. the protease activity without influencing the intrinsic or the 10.1021/bi701649b CCC: $37.00 © xxxx American Chemical Society Published on Web 11/02/2007 PAGE EST: 12.6 B Patterson-Ward et al. Biochemistry protein-stimulated ATPase activity of the enzyme (19). site of Lon, and facilitates translocation of peptide substrates Other ATP-dependent proteases that belong to the same by inducing at least two distinct conformational changes family as Lon also utilize ATP hydrolysis to promote the within the enzyme complex. Furthermore, ATP hydrolysis translocation of unfolded polypeptide substrates to the offers an additional catalytic advantage in mediating peptide proteolytic active site. It is generally believed that the bond cleavage by facilitating the formation of the two translocation of polypeptides constitutes the rate-limiting step conformational changes detected in this study. of the peptide hydrolysis reaction (20-22). Although it is plausible that Lon utilizes a similar mechanism in mediating MATERIALS AND METHODS peptide bond cleavage, this has not been unambiguously Materials. ATP (adenosine triphosphate), ADP (adenosine demonstrated. The detection of a rate-limiting step in the diphosphate), AMPPNP (adenylyl 5-imidodiphosphate), dan- binding interaction between Lon and the peptide substrate syl chloride, SBTI (soybean trypsin inhibitor), TPCK-treated in the presence of ATP will provide support for the existence trypsin, N-2-hydroxyethylpiperazine-N′-ethanesulfonic acid of the proposed peptide translocation step. (HEPES) and polyethyleneimine-cellulose (PEI-cellulose) To elucidate the timing of ATP hydrolysis with activation TLC plates were purchased from Sigma and Fisher. All of peptide bond cleavage by E. coli Lon, we have previously Fmoc-protected amino acids, Fmoc-protected Lys Wang generated a synthetic peptide (designated S1) which contains resin, Boc-anthranilamide (Abz), and O-benzotriazole- residues 89-98 of the λN protein which is an endogenous N,N,N′,N′-tetramethyluroniumhexafluorophosphate (HBTU) protein substrate of E. coli Lon (1, 2). We demonstrated that were purchased from Advanced ChemTech and EMD the degradation of S1 exhibits the same ATP-dependency Biosciences. [R-32P]ATP was purchased from Perkin-Elmer and cleavage specificity as the full-length λN(1, 2). Using Life Science. QuikChange Site-Directed Mutagenesis Kit was pre-steady-state kinetic techniques, we also demonstrated that purchased from Statagene. Primers were purchased from ATP hydrolysis occurs prior to peptide bond cleavage since Integrated DNA Technologies. a burst in ATPase activity is coupled with a lag in proteolysis (23). Comparing the rate-constants of the two hydrolytic General Methods. Peptide synthesis and protein purifica- reactions suggests that the lag phase of peptide cleavage is tion procedures were performed as described previously (1). dependent on the buildup of an enzyme intermediate gener- All enzyme concentrations are reported as Lon monomer ° ated during and/or after ATP hydrolysis, which could be concentrations. All experiments were performed at 37 C. attributed to the existence of the proposed ATP-dependent All reagents are reported as final concentrations. peptide translocation step. Generation and Characterization of Lon Mutants. Two In this study, we probed the existence of an ATP- Lon mutants, S679A and S679W, were generated using the dependent peptide translocation step using a similar technique wild type E. coli Lon protease plasmid (pSG11, a generous used by Epps et al. and Fattori et al. to monitor the kinetics gift from Alfred Goldberg, Harvard Medical School) and the of peptide-protein interactions (24, 25). We employed pre- QuikChange Site-Directed Mutagenesis Kit from Stratagene. steady-state kinetic techniques to monitor the dynamics of a The forward primer used to generate the S679A mutant dansylated peptide (designated S4) interacting with two (pJW028 plasmid) was 5′-GCCGAAAGATGGTCCGGCT- proteolytically inactive E. coli Lon mutants, S679A and GCCGGTATTGC-3′, and the reverse primer was 5′- S679W. Upon binding to the target protein, the fluorescence GCAATACCGGCAGCCGGACCATCTTTCGGC-3′. The signal associated with a dansylated peptide is indirectly forward primer used to generate the S679W mutant (pJW015 enhanced by excitation of the intrinsic Trp residues located plasmid) was 5′-AAAGATGGTCCATGGGCCGGTAT- within the protease. These authors (Epps et al. and Fattori TGCT-3′, and the reverse primer was 5′-AGCAATAGGGGC- et al.) refer to this fluorescence technique as FRET (fluo- CCATGGACCATCTTT-3′. The sequence of pJW028 and rescence resonance energy transfer) (24, 25). Analysis of the pJW015 was verified by DNA sequencing. Both plasmids fluorescence time courses allowed us to construct a kinetic were transformed into the BL21 E. coli cell strain. S679A model revealing the enzyme intermediates generated from and S679W Lon were expressed and purified to homogeneity the binding and hydrolysis of ATP in the presence of the S4 using the protocols described previously for the wild type peptide. The proteolytically inactive mutant S679A contains enzyme (1). ATPase activity was measured using radiola- the three intrinsic Trp residues found in wild type Lon (15) beled ATP as described previously (26, 27). MANT-ATP while S679W contains an additional Trp residue in place of binding was measured using stopped-flow techniques as the active site Ser. The kinetics of the S4 peptide binding to described previously (28). either Lon mutant in the presence of ATP was monitored Fluorescence Emission Scans. Emission spectra were by exciting the Trp residues at 290 nm and detecting the collected on a Fluoromax 3 spectrofluorimeter (Horiba rates of dansyl fluorescence increase at >450 nm. Applying Group) with excitation of the Trp residues at 290 nm. All stopped flow kinetic techniques, we were able to detect at assays were performed in microcuvettes (Hellma) with a least one distinct conformational change in either E. coli Lon 3-mm path length. Reactions contained 50 mM HEPES pH mutant that are dependent on the presence of the S4 peptide 8.0, 5 mM Mg(OAc)2, 5 mM DTT, 100 µM ATP, 5 µM and nucleotide binding (ATP or AMPPNP). The rate constant S679A or S679W Lon, and varying concentrations of S4 of this step agrees well with the lag rate constant of peptide dansyl peptide or S2 nonfluorescent peptide (0-100 µM). hydrolysis that was determined previously (23). However, The emission spectra for the dansyl peptide S4 interacting an additional conformational change is detected in S679W with Lon were generated by subtracting the emission spectra that is dependent only upon the presence of nucleotide. of the S4 dansyl peptide alone in the presence of ATP from Collectively, these data support a mechanism by which ATP the emission spectra of the same concentration of the S4 binding and hydrolysis allosterically activates the proteolytic peptide incubated with Lon in the presence of ATP. The Biochemistry Mechanism of Peptide Binding to Lon Protease C concomitant changes in the emission spectra of tryptophan The averaged time courses of S679W with ATP and S4 were recorded by subtracting the emission spectra of the S4 peptide were fitted with eq 3 describing a double exponential peptide incubated with Lon and ATP from the emission ) - + - + spectra of the same concentration of dansylated glutamic acid F A1 exp( k1,S679Wt) A2 exp( k2,S679Wt) C incubated with Lon and ATP. (3) Monitoring Peptide-Lon Interactions. Experiments to where F is relative fluorescence, A1 and A2 are amplitudes monitor interactions between peptide substrate with Lon were in relative fluorescence units, t is time in seconds, C is the performed on a KinTek Stopped Flow controlled by the data endpoint, k1,S679W is the first-order rate constant associated collection software Stop Flow version 7.50 â with a 0.5 cm with the first phase of the reaction in per seconds, and k ° 2,S679W path length. The sample syringes were maintained at 37 C is the first-order rate constant associated with the second by a circulating water bath. Syringe A contained 5 µM phase of the reaction in per seconds. Time courses of S679W S679A or S679W Lon monomer with variable concentrations with saturating concentrations of ATP and low concentrations - of S4 (10 500 µM), 50 mM HEPES pH 8.0, 75 mM KOAc, of S4 dansyl peptide (25-125 µM) were fitted with eq 4 5 mM Mg(OAc)2 and 5 mM DTT. Syringe B contained 100 describing a triple exponential µM ATP, 50 mM HEPES pH 8.0, 75 mM KOAc and 5 mM ) - + - + DTT. S4 binding to S679A was monitored by an increase F A1 exp( k1,S679Wt) A2* exp( k2*,S679Wt) in fluorescence (excitation 290 nm emission 450 nm long- - + pass filter) upon rapid mixing of the syringe contents. In A2 exp( k2,S679Wt) C (4) addition to monitoring with excitation 290 nm and emission where F is relative fluorescence, A , A and A are with a 450 nm long-pass filter, experiments were performed 1 2* 2 amplitudes in relative fluorescence units, t is time in seconds, with excitation 290 nm emission with a 340 nm band-pass C is the endpoint, k is the first-order rate constant filter to monitor changes in Trp fluorescence. To define the 1,S679W associated with the first phase of the reaction in per seconds, pre-steady-state period of the reaction, each time course was k is the first-order rate constant associated with the fitted over a period of time equal to four half-lives of the 2,S679W second phase of the reaction in per seconds and k is pre-steady-state rate constant (k ) obtained from the ATP 2*,S679W lag the first-order rate constant for a phase in the reaction that dependent peptidase reaction. The lag rate constant k was lag is only visible under conditions of low peptide concentra- previously determined for each reaction condition used (23). tions. The data shown are a result of averaging at least four traces. The first-order rate constants were plotted as a function Each reaction condition was performed in triplicate (a total of substrate concentration (ATP or S4). Hyperbolic plots of g12 traces obtained for each data point). The averaged were fitted using the data fitting program Kaleidagraph by time courses of S679A were fitted with eq 1 describing a Synergy with eq 5 describing a 2-step binding event single exponential ) + + ) - + k kmax[S]/(Kd [S]) krev (5) F A exp( kS679At) C (1) where F is relative fluorescence, A is amplitude in relative where k is the observed rate constant from eq 1 or eq 3 in fluorescence units, t is time in seconds, C is the endpoint, per seconds, kmax is the maximum forward rate constant for the second step in the binding event in per seconds, krev is and kS679A is the first-order rate constant associated with peptide binding to S679A Lon in per seconds. It should be the reverse rate constant for the second step in the binding noted that the PMT (photomultiplier tube) sensitivity was event in per seconds, [S] is the concentration of ATP or S4 automatically adjusted by the instrument to optimize signal- peptide, and Kd is the equilibrium binding constant for the to-noise during acquisition of the time course data for the substrate in micromolar units. Sigmoidal plots were fitted various concentrations of peptide used in the reactions. A using the data fitting program Enzfitter by Biosoft with eq higher sensitivity setting was used for the lower [S4] and a 6 lower sensitivity setting was used for higher [S4]. As a result ) n ′ + n + the relative amplitudes of the time courses do not reflect the k kmax[S] /(Kd [S] ) krev (6) stoichiometries of the enzyme intermediates monitored by where k is the observed rate constant from eq 1, 3 or 4 in the signals. At high concentrations of S4 (>100 µM), the per seconds, k is the maximum forward rate constant for high absorbance of the dansyl moiety also obscures the max the second step in the binding event in per seconds, k is amplitude of the fluorescence signal detected; but the first- rev the reverse rate constant for the second step in the binding order rate constants of the reactions do not change because event in per seconds, [S] is the concentration of ATP or S4 the dansyl absorbance in each reaction remains constant. peptide, n is the Hill coefficient (n ) 1.6, previously The time course of S679W with AMPPNP without peptide determined for Lon and our model peptide substrate (27)), was fitted with eq 2 describing a single-exponential equation the equilibrium binding constant K is calculated from the followed by a steady-state rate d relationship log Kd′ ) n log Kd. ) - + + Measuring Peptide Binding Using Fluorescence Anisot- F A1 exp( k1,S679Wt) νsst C (2) ropy. S4 dansyl peptide binding to the S679A Lon mutant where F is relative fluorescence, A1 is amplitude in relative was detected by changes in fluorescence anisotropy (excita- fluorescence units, t is time in seconds, k1,S679W is the first- tion 340 nm emission 520 nm) at 37 °C on a Fluoromax-3 order rate constant associated with the first phase of the spectrofluorimeter (Horiba Group). Each binding reaction reaction in per seconds, νss is the steady-state rate in contained 50 mM HEPES pH 8.0, 5 mM Mg(OAc)2,5mM fluorescence units per second, and C is the endpoint. DTT, and 20 µM S4 dansyl peptide in the presence and D Patterson-Ward et al. Biochemistry
Table 1: Kinetic Parameters for ATP Binding and Hydrolysis by Wild-Type and Lon Mutants ATPase activity MANT-ATP binding -1 5 -1 -1 -1 -1 kcat (s ) Km (µM) kon (10 M s ) koff (s ) kon,2 (s ) WT Lon 0.26 ( 0.01 46 ( 6 6.8a 11a 4.1a WT Lon + S2 1.43 ( 0.05 82 ( 10 6.8a 10a 3.7a S679A 0.31 ( 0.03 147 ( 41 2.95 ( 0.70 8.50 ( 1.14 3.85 ( 0.96 S679A + S3 1.57 ( 0.08 131 ( 21 NDb ND ND S679W 0.64 ( 0.03 139 ( 24 4.26 ( 1.98 8.59 ( 1.51 3.96 ( 1.80 S679W + S3 1.07 ( 0.06 79 ( 15 ND ND ND a These values are from ref 28. b ND: not determined. absence of 1 mM AMPPNP in a quartz microcuvette fluorescence of the intrinsic Trp residues of S679A are (Hellma) witha3mmpath length. S679A, S679W or wild- detected upon excitation at 290 nm in a series of reactions type Lon was titrated into the cuvette (0-192 µM) and the containing 100 µM ATP and increasing concentrations of reaction was incubated at 37 °C for 3 min to allow the nonfluorescent peptide S2 (YRGITCSGRQK(Bz), Figure equilibrium to be reached. Higher Lon concentrations were 1A). The low signal generated from S2 binding to the Lon not used due to problems with solubility of the protein. The mutants cannot be used to monitor the kinetics of peptide anisotropy measurements were plotted as a function of Lon interacting with the proteins. In contrast, the fluorescence concentration and the data were fitted with eq 7 of Trp residues in S679A is significantly reduced when the dansylated S4 peptide (YRGITCSGRQK(dansyl)) is used ) n ′ + n + B Bmax[S] /(Kd [S] ) C (7) (Figure 1B and 1C). When excited at 290 nm, the reduction in the fluorescence of Trp is accompanied by an increase in where B is the observed anisotropy, Bmax is the maximum the fluorescence of the dansyl moiety in the S4 peptide anisotropy, [S] is the concentration of Lon added, n is the (Figure 1D). An additional control was performed to correct Hill coefficient, C is the endpoint, and the equilibrium for the decrease in trypotphan fluorescence caused by binding constant Kd is calculated from the relationship log nonspecific absorption of the dansyl moiety. Although a Kd′ ) n log Kd. decrease in tryptophan fluorescence is observed with dan- sylated glutamic acid, a larger decrease in the fluorescence RESULTS signal is observed with the S4 dansyl peptide (Figure 1B). The tryptophan spectra shown in Figure 1C were generated Characterization of the E. coli Lon Mutants S679A and by subtracting the emission spectra of the S4 peptide S679W. We generated two proteolytically inactive mutants incubated with Lon and ATP from the emission spectra of of E. coli Lon to study the microscopic kinetic events the same concentration of dansylated glutamic acid incubated occurring along the enzymatic pathway prior to peptide with Lon and ATP. Collectively, Figures 1A to 1D reveal cleavage. Both mutants replace Ser 679, which is responsible that the reduction in the fluorescence of the tryptophan for attacking the scissile peptide bond in a substrate, with either an Ala (S679A) or Trp (S679W). While both mutants residues in Lon is attributed to the specific interaction display wild-type-like intrinsic and peptide-stimulated between Lon and the S4 peptide. Thus the interaction ATPase activities that are identical to wild-type Lon, both between S679A and the dansylated peptide S4 in the presence are proteolytically inactive (see Supporting Information of nucleotide can be quantitatively characterized by monitor- Figures A and C). As summarized in Table 1, the k and ing the increase in dansyl fluorescence using stopped flow cat spectroscopy. Figure 2A provides the time course for the Km values for the ATPase activity are comparable to those obtained for wild-type Lon and suggest that the mutations change in dansyl fluorescence generated from rapidly mixing do not alter ATP hydrolysis (28). In addition, we used 100 µM ATP versus a preincubated solution containing 5 stopped-flow spectroscopy to validate that MANT-ATP binds µM S679A with 100 µM S4 dansyl. The order of peptide or to both mutants identically compared to the wild type enzyme ATP mixing does not affect the kinetics of the reactions since (Table 1 and Supporting Information Figure B). Taken identical time courses were obtained when S679A was together, our data demonstrates that only the degradation of rapidly mixed with 100 µM S4 dansyl peptide (data not peptide is inactivated by the mutations. Furthermore, limited shown). In addition, negligible changes in the dansyl tryptic digestion analysis reveals that both mutants form a fluorescence were detected when ATP was omitted. compact conformation resistant to digestion by trypsin upon In addition to monitoring changes in dansyl fluorescence, binding to ATP (Supporting Information Figure D), thereby these experiments could also monitor changes in Trp revealing that the mutations do not affect the adenine- fluorescence (Figure 2B). The increase in dansyl fluorescence dependent conformational change that is found in the wild- (Figure 2A) mirrors the decrease in Trp fluorescence (Figure type enzyme (27). The discovery of wild-type-like ATPase 2B) which was detected using a 340 nm bandpass filter. activities in both mutants is consistent with the report of Taken together, the increase in the dansylated peptide Starkova et al. demonstrating that the ATPase activity can fluorescence time course (Figure 2A) along with the con- be decoupled from the proteolytic activity of Lon (19). comitant decrease in the Trp fluorescence time course (Figure Characterizing the Binding Interaction of S4 Peptide with 2B) suggests that peptide binding to Lon can be monitored S679A by Fluorescence Spectroscopy. E. coli Lon is a homo- by detecting the dansyl fluorescence time courses through hexamer containing three intrinsic Trp residues in each excitation of the Trp residues in the Lon mutants. The subunit. Surprisingly though, only small changes in the increase in dansyl fluorescence most likely results from Biochemistry Mechanism of Peptide Binding to Lon Protease E
FIGURE 1: (A) Fluorescent emission scans of S679A with nonfluorescent S2 peptide. Five micromolar S679A was incubated with 100 µM ATP and (9)0µM, (b)25µM, ([) 100 µM of the nonfluorescent S2 peptide. The sample was excited at 290 nm, and emission was monitored from 300 nm to 570 nm. No changes in tryptophan fluorescence are detected. Similar results were obtained with the S679W Lon mutant. (B) Fluorescent emission scan of S679A with S4 dansylated peptide and dansylated glutamic acid. Five micromolar S679A was incubated with 100 µM ATP and 50 µM of the S4 dansylated peptide or 50 µM of dansylated glutamic acid. The sample was excited at 290 nm and emission was monitored from 300 nm to 400 nm. A decrease in tryptophan fluorescence at 350 nm is observed with dansylated glutamic acid, however there is a greater decrease with the S4 dansyl peptide. Similar results were obtained with the S679W Lon mutant. (C) Corrected S679A tryptophan emission spectra. Five micromolar S679A and 100 µM ATP was incubated with (9)0µM, (b)25µM, (2)50µM, ([) 100 µM S4 dansyl peptide or dansylated glutamic acid. The samples were excited at 290 nm, and emission was detected from 300 nm to 400 nm. The spectra shown were corrected by subtracting the emission spectra of S4 dansyl peptide with Lon and ATP from a control reaction containing dansylated glutamic acid, Lon and ATP. (D) Corrected emission spectra for the dansyl moiety on the S4 peptide. Five micromolar S679A was incubated with 100 µM ATP and (9)0µM, (b)25µM, (2)50µM, ([) 100 µM of the dansyl peptide. The sample was excited at 290 nm, and emission was monitored from 450 nm to 570 nm. These spectra were corrected by subtracting the emission scan of S4 dansyl peptide with ATP and no Lon from the emission spectra of S4 dansyl peptide with Lon and ATP. An increase in dansyl fluorescence at 520 nm is observed with increasing concentrations of S4 peptide. Similar results were obtained with the S679W Lon mutant. conformational changes in the enzyme in which the Trp 20 µM of the dansylated peptide S4 and 1 mM AMPPNP, a residues are more accessible to interact with the dansyl nonhydrolyzable analogue of ATP. The binding isotherm in moiety. However, alternative possibilities exist. For example, Figure 3 was fitted to the Hill equation (eq 7) to yield a Kd the signal could reflect the dansyl peptide approaching one ) 35.2 ( 18.6 µM and a Hill coefficient (n)of1.5( 0.1. or more of the Trp residues in Lon or a combination of both Experiments performed using wild-type or S679A Lon in phenomena. Regardless, the detected signal is a result of the the absence of AMPPNP yielded identical binding parameters S4 dansyl peptide interacting with Lon, and is used as a (Kd ) 38.3 ( 6.0 µM, n ) 1.5 ( 0.1; and Kd ) 49.2 ( 28.1 reporter for monitoring the kinetics of the binding reactions. µM, n ) 1.5 ( 0.1, respectively, data not shown). Compa- Measuring S4 Dansyl Peptide Binding to Lon Using rable Kd values were also obtained in the S4 peptide binding Fluorescence Anisotropy. Fluorescence anisotropy experi- to S679W in the presence of AMPPNP (data not shown). ments were used to further characterize the equilibrium Since neither the Kd nor the Hill coefficients are affected by binding of S4 dansyl peptide to S679A Lon. As shown in the presence of nucleotide, we conclude that peptide binding Figure 3, S4 peptide binding to S679A Lon can be measured to Lon occurs via a rapid equilibrium process. This conclu- by titrating S679A Lon (0-192 µM) into a cuvette containing sion is supported by our previous steady-state kinetic studies F Patterson-Ward et al. Biochemistry
FIGURE 3: Equilibrium binding of S4 dansyl peptide to S679A Lon can be measured using fluorescent anisotropy. Twenty micromolar S4 dansyl peptide was incubated with 1 mM AMPPNP (10 x Kd, determined previously and reported in (36)), and changes in anisotropy were measured (excitation 340 nm emission 520 nm) by titrating in S679A Lon (0 µMto192µM). The data were fitted with eq 7 resulting in Kd ) 35.2 ( 18.6 µM with a Hill coefficient (n)of1.5( 0.1. The data shown are an average of 3 trials. Similar results were obtained in S679A and in the wild-type enzyme when nucleotide was omitted (see Results).
from a conformational change in the enzyme making the Trp residues more accessible. The time courses were best fitted to a single-exponential equation (eq 1) to yield the observed rate constants of the reactions, kS679A. Plotting kS679A as a function of peptide concentration yields a sigmoidal plot shown in Figure 4A, which is indicative of a two step peptide binding mechanism (Scheme 1) (30). The assignment of a two-step binding FIGURE 2: Peptide binding to S679A can be monitored using the mechanism can be explained in terms of the signal detected S4 dansyl peptide. The experimental time courses are shown in from S4 is generated from the S679A:dansyl peptide gray, and the fitted curve is shown in black. The time courses with complex. At low peptide concentrations, the formation of ATP were fitted with eq 1 describing a single exponential. (A) Five enzyme:peptide contributes to the observed k (step 1 in micromolar S679A was incubated with 100 µM S4 dansyl peptide S679A and rapidly mixed with 100 µM ATP. The reaction was excited at Scheme 1). As such, the kS679A values increase as the 290 nm and monitored using a 450 nm long-pass filter to measure concentration of peptide increases until the enzyme is dansyl fluorescence. No changes in fluorescence were observed in saturated with peptide. The horizontal asymptote of the plot the absence of ATP. (B) The reactions described in Figure 2A were shown in Figure 4A provides the maximum rate constant of also monitored for changes in Trp fluorescence by excitation at S4 interacting with the enzyme at saturating peptide con- 290 nm and detecting emission with a 340 nm bandpass filter. No changes in fluorescence were observed in the absence of ATP. centrations. This apparent zero-order dependency of a rate- Identical reaction time courses were obtained when S679A was constant toward peptide concentration is consistent with the rapidly mixed with S4 peptide premixed with ATP. existence of an isomerization or conformational change occurring within the enzyme:peptide complex after the initial ligand binding event (Scheme 1) which has been detected revealing that Lon adopts a random ordered Bi Bi kinetic in many enzymes (30, 31). Fitting the data shown in Figure mechanism in catalyzing the ATP-dependent cleavage of a 4A with eq 6 yields the K of S4 for S679A, which is 164 peptide bond (2). d ( 35 µM, n ) 1.3 ( 0.2 (Table 2). Increasing the peptide The Interaction between S679A and Dansyl Peptide concentration saturates the peptide binding site in S679A Exhibits Dependency toward Peptide and ATP Concentra- such that the observed rate constant becomes dominated by tions. To evaluate the kinetics of S679A interacting with S4, the forward and reverse rate constants of the second step at we measured the stopped flow time courses of dansyl peptide concentrations greater than the Kd. Under these fluorescence emission (450 nm long-pass filter) by exciting conditions, the forward and reverse rate constants reach a Trp at 290 nm at various S4 dansyl peptide concentrations maximum, which is 0.74 ( 0.10 s-1 and 0.19 ( 0.01 s-1, (10 to 500 µM). The dansyl fluorescence signal measured is respectively (Table 2). a result of the S4 dansyl peptide interacting with the Trp The interaction between 5 µM of S679A and 500 µMS4 residues in Lon. The Trp residues are excited at 290 nm, dansyl peptide (∼3 × Kd of S679A, see Table 2) is also and the emission from the Trp residues excites the dansyl dependent on the concentration of ATP as illustrated in moiety on S4 either from S4 approaching a Trp residue or Figure 4B. The time courses (excitation 290 nm, emission Biochemistry Mechanism of Peptide Binding to Lon Protease G
Scheme 1: Peptide Binding Steps Measured Using the S679A Lon Mutanta
a (1) An initial peptide binding event independent of nucleotide. This step is measured using equilibrium fluorescence anisotropy. (2) A conformational change step following initial binding that requires nucleotide. This step is measured using stopped-flow kinetic techniques. It should be noted that ATP binds to Lon independent of peptide. The Lon*:peptide complex contains ATP.
allosteric communication between the ATPase site and the protease site in Lon. As Trp fluorescence is sensitive to changes in its local environment, the replacement of the 679S with a Trp residue allows the detection of a conformational change in the proteolytic site in Lon resulting from ATP or AMPPNP binding (32). Figure 5 provides representative stopped flow time courses monitoring the changes in Trp fluorescence of S679W in the absence and presence of ATP as well as in the presence of AMPPNP or ADP without the presence of peptide. A decrease in the Trp fluorescence is detected only in the presence of ATP or AMPPNP but not ADP, suggesting that the gamma phosphate moiety present in ATP and AMPPNP is needed to induce the changes in Trp fluorescence. The time course which contains ATP is best fitted with the single-exponential function (eq 1) to yield an observed rate constant of 9.35 ( 0.34 s-1 (observed k1,S679W). The time course which contains the nonhydrolyzable analogue AMPPNP is best fitted with the single-exponential function followed by a steady-state rate shown in eq 2 to yield an observed rate constant of 0.94 ( 0.02 s-1 and a steady-state rate of 0.02 fluorescence units per second to account for the decrease in the Trp fluorescence. As the observed rate constant for the change in Trp fluorescence is ∼10-fold faster in the presence of ATP than AMPPNP, it is FIGURE 4: S4 dansyl peptide binding to S679A is dependent on likely that the formation of this enzyme form is facilitated peptide and ATP. (A) Five micromolar S679A and varying by the hydrolysis of ATP. Furthermore, the presence of concentrations of S4 dansyl peptide (10, 15, 25, 35, 50, 75, 100, 150, 250, 300, 450 or 500 µM) were rapidly mixed with 100 µM peptide substrate is not needed for this conformational ATP. The time courses were fitted with eq 1, and the resulting rate change. constants are plotted as a function of the corresponding peptide Compared to ATP, AMPPNP Supports the Peptide-Lon concentration. The data in (A) were fitted with eq 6 to yield a Interaction at a Reduced Rate. An increase in dansyl -1 maximum kS679A ) 0.74 ( 0.10 s , Kd ) 164 ( 35 µM, krev ) 0.19 ( 0.01 s-1, n ) 1.3 ( 0.2 (Table 2). (B) Five micromolar fluorescence was detected when 5 µM S679W preincubated with 100 µM S4 was mixed with 100 µM ATP (Figure 6A). S679A and 500 µM(3× Km) S4 dansyl peptide was rapidly mixed with varying concentrations of ATP (0.5, 1, 3, 5, 10, 25 or 50 µM). The samples were excited at 290 nm and the fluorescence The time courses were fitted with eq 1, and the resulting rate signals were detected using a 450 nm long-pass filter, which constants are plotted as a function of the corresponding ATP monitored the fluorescence signal generated from the S4 concentration. The data in (B) were fitted with eq 5 to yield a -1 dansyl peptide. We have previously shown that AMPPNP maximum kS679A ) 0.54 ( 0.04 s , Kd ) 7.4 ( 2.5 µM, krev ) - ∼ 0.19 ( 0.03 s 1 (Table 2). supports peptide hydrolysis by Lon with a kcat that is 7 fold lower and a lag phase that is ∼10-fold longer than those 450 nm) were best fitted to a single-exponential curve (eq detected for ATP (1, 2). To explore how AMPPNP effects 1) to yield the rate constants, kS679A, for the reactions in which the peptide-Lon interaction we used 5 µM S679W with 500 the concentration of ATP is varied. Plotting the observed µM S4 dansyl peptide and rapidly mixed it with 100 µM rate constants at the corresponding concentration of ATP (0.5 AMPPNP. The resulting time course is shown in Figure 6B. µMto50µM) yields a hyperbolic plot, which upon fitting The time course was fitted to a double exponential equation with eq 5 yields a Kd of 7.4 ( 2.5 µM for ATP, a forward (eq 3), and the resulting rate constants, k1,S679W and k2,S679W, -1 rate constant (kS679A,max) of 0.54 ( 0.04 s and a reverse were found to be approximately 20- to 10-fold slower than -1 -1 -1 rate constant (krev,S679A) of 0.19 ( 0.03 s (Figure 4B, Table with ATP (0.50 ( 0.01 s and 0.04 ( 0.01 s , respectively) 2). compared to 7.6 ( 1.0 s-1 and 0.57 ( 0.10 s-1 in the The S679W Mutant Shows an ATP-Dependent Conforma- presence of ATP under identical reaction conditions (see tional Change in Trp That Is Independent of Peptide. The Table 2). The same experiment was also performed in the S679W mutant was generated to probe the existence of an presence of ADP instead of AMPPNP. As shown in Figure H Patterson-Ward et al. Biochemistry
Table 2: Kinetic Constants for S4 Peptide Interacting with S679A and S679W S679A S679W
k1 k2 vary S4 vary ATP vary S3 vary ATP vary S4 vary ATP -1 kmax (s ) 0.74 ( 0.10 0.54 ( 0.04 7.6 ( 1.0 5.3 ( 0.6 0.57 ( 0.10 1.5 ( 0.4 a kd (µM) 164 ( 35 7.4 ( 2.5 NA 4.3 ( 1.9 157 ( 8 9.3 ( 9.8 -1 b krev (s ) 0.19 ( 0.01 0.19 ( 0.03 NA 2.1 ( 0.5 0.10 ( 0.01 ND n 1.3 ( 0.2 NA NA NA 1.9 ( 0.1 NA a NA: not applicable. b ND: not determined.
FIGURE 5: Intrinsic tryptophan fluorescence can be used to measure a conformational change in S679W dependent on ATP and AMPPNP. No significant changes are observed with buffer or ADP. Five micromolar S679W was rapidly mixed with buffer, 100 µM ATP, 100 µM AMPPNP or 100 µM ADP in a stopped-flow instrument. The reactions were excited at 290 nm, and emission was detected using a 340 bandpass filter to detect Trp fluorescence. The reaction with ATP was fitted with eq 1 describing a single exponential, and the reaction with AMPPNP was fitted with eq 2 describing a single exponential followed by a steady-state. The experimental time course is shown in gray, and the fitted curve is shown in black.
6B, no change in fluorescence is detected in the reaction containing ADP, indicating that the gamma-phosphate posi- tion of ATP is needed for generating the conformational changes detected in this study. Fluorescent Analysis of the Interaction of S679W and the Dansylated Peptide. The interaction between S679W and the FIGURE 6: Peptide binding to S679W can be monitored using the S4 peptide was examined using stopped flow techniques S4 dansyl peptide. The experimental time courses are shown in described above for characterizing the S679A mutant. Figure gray, and the fitted curve is shown in black. (A) Five micromolar 7A shows selected time courses reflecting the changes in S679W was preincubated with 100 µM S4 dansyl peptide and the dansyl fluorescence of S4 following excitation of the Trp rapidly mixed with 100 µM ATP. The reaction was excited at 290 - nm and monitored using a 450 nm long-pass filter to measure dansyl residues in S679W at varying concentrations of ATP (0.5 fluorescence. The time course with ATP was fitted with eq 3 50 µM). According to Figure 7A, the time courses obtained describing a double exponential. No changes in fluorescence were at concentrations of ATP at <5 µM and at 500 µMS4(∼3 observed in the absence of ATP. (B) Five micromolar S679W was × Kd of S4) display an increase followed by a slower preincubated with 100 µM S4 dansyl peptide and rapidly mixed decrease in dansyl fluorescence. At g5 µM ATP, however, with 100 µM AMPPNP. The reaction was excited at 290 nm and monitored using a 450 nm long-pass filter to measure dansyl the time courses only show an increase in dansyl fluores- fluorescence. The time course with AMPPNP was fitted with eq 3 cence. Concurrent decreases in the Trp fluorescence time describing a double exponential. AMPPNP does support binding, courses which mirror the observed dansyl fluorescence data however at a rate slower than ATP. The k1,S679W and k2,S679W are - - were also detected (see Supporting Information Figure E). 0.50 s 1 and 0.04 s 1, respectively. No changes in fluorescence In an earlier study, we demonstrated that the ATPase activity were observed with ADP. of Lon was negligible at <5 µM ATP, and the enzyme was ATP likely reflect the conformational states of S679W only 50% saturated with ATP at 10 µM of the nucleotide accompanied by ATP hydrolysis. Comparing the data (23, 28, 33). Therefore the time courses detected at g5 µM obtained at <5 µM versus at g5 µM ATP reveals that the Biochemistry Mechanism of Peptide Binding to Lon Protease I the dansyl fluorescence of S4 upon incubating with S679W at a saturating level of ATP. The time courses obtained at sub Kd levels of S4 (see Table 2 for Kd of S4 peptide) are triphasic in which there is an initial increase followed by a transient decrease culminating in a final phase showing an increase in fluorescence signal (time courses a and b). At the low peptide concentrations (25-100 µM), the time courses fit best to a triple exponential equation (eq 4). We assign the first phase as k1,S679W, the second phase as k2*,S679W and the final phase as k2,S679W. The transient phase, which is represented by k2*,S679W, is independent of [S4] (k2*,S679W ) 0.95 ( 0.30 s-1). As the concentration of peptide is increased, the time courses become biphasic as the intermediate phase diminishes in intensity, and the data were fitted best to the double exponential function (eq 3). We propose that the second transient phase “disappears” at higher S4 concentra- tions since the rate constant k2,S679W, which exhibits depen- dency toward peptide concentration, becomes comparable to the values of k2*S697W. As a result, only two apparent rate constants that are represented by k1,S679W and k2,S679W are detected when the concentration of S4 is higher than the Kd of peptide. The mechanistic feature that is responsible for generating the transient phase is not clear. However, these reactions were conducted at 100 µM ATP, where Lon catalyzes the hydrolysis of ATP with a pre-steady-state burst rate of ∼10 s-1 and a steady-state turnover rate constant of ∼0.8 s-1 (23). Therefore it is possible that the transient phase detected at low concentrations of S4 is attributed to the exchange of ADP generated from the first turnover of nucleotide hydrolysis with the exchange of ATP in the enzyme to maintain protein interaction with S4. Additional experimentation will be required to resolve this issue. Collectively the biphasic time course data obtained at 500 FIGURE 7: Representative time courses for S679W interacting with µM S4 and varying [ATP] or at varying [S4] and saturating the S4 dansyl peptide. All the reactions were excited at 290 nm, ATP were best fitted with the double exponential function and emission signals from the dansyl moiety in S4 were detected defined by eq 3 to yield two observed rate constants k1,S679W using a 450 nm long-pass filter. (A) Five micromolar S679W and k . In regard to the data obtained at 500 µMS4 preincubated with 500 µM S4 was rapidly mixed with (a) 0.5; (b) 2,S679W 1; (c) 5 and (d) 10 µM ATP in a stopped flow apparatus. All the and at varying [ATP], the observed k2,S679W values were time courses were best fitted with eq 3 describing a double determined by fitting the data at >5 µM ATP with eq 3 to exponential. The experimental time courses are shown in gray, and yield the respective k1,S679W and k2,S679W. Plotting the observed < the fitted curve is shown in black. At 5 µM ATP, the second k1,S679W and k2,S679W as a function of [ATP] and [S4] (Figures phases of the time courses display negative changes in fluorescence, 8A to 8D) allows the extrapolation of the maximal rate whereas in the time courses measured at g5 µM ATP, the overall changes in fluorescence were positive. (B) Five micromolar S679W constants for the respective phase of the reactions. Figures preincubated with (a) 25; (b) 50; (c) 100 and (d) 500 µM S4 peptide 8A and 8B collectively show that k1,S679W exhibits depen- was rapidly mixed with 100 µM ATP in a stopped-flow apparatus. dency toward [ATP] but not [peptide] as k1,S679W at 0 µM The experimental time courses are shown in gray, and the fitted S4 dansyl peptide is 9.35 ( 0.34 s-1 (determined by curve is shown in black. The sub Kd levels of S4 (25 and 50 µM) monitoring decreases in Trp fluorescence with 5 µM S679W were fitted with eq 4 (triple exponential). The 100 and 500 µMS4 time courses were fitted with eq 3 (double exponential equation). and 100 µM ATP, Figure 5). Increasing the concentration of S4 peptide does not significantly change the value of direction of the fluorescence changes in the second phase is k1,S679W. The apparent increase in the k1,S679W values discerned reversed as the concentration of ATP increases. As the in Figure 8A may be due to uncertainty involved in proteolytic residue 679S is replaced with a Trp, the observed determining k1,S679W while fitting the time courses. These changes in the direction of the fluorescence time courses at results indicate that the formation of the first conformational increasing concentrations of ATP could be attributed to the change in S679W, which is defined by k1,S679W, is dependent conformation of the proteolytic site of Lon varying with the only on the presence of ATP. The maximal k1,S679W value occupancy of nucleotide and/or the ATPase activity of the was extrapolated from Figure 8A by averaging the values -1 enzyme. As such, optimal peptide cleavage by Lon requires (k1,S679W ) 7.6 ( 1.0 s ) and from Figure 8B by fitting the -1 the hydrolysis of ATP to fully activate the proteolytic site. data to eq 5 (k1,S679W ) 5.3 ( 0.6 s ). All kinetic parameters Identical experiments were performed varying the con- are reported in Table 2. centration of S4 (25-500 µM) using a fixed concentration The final exponential phase, which is defined by k2,S679W, of 100 µM ATP (10 × Kd of the ATPase site (33)). Figure exhibits dependency on the concentration of ATP and 7B shows selected time courses reflecting the changes in peptide. Figures 8C and 8D show the plots relating the J Patterson-Ward et al. Biochemistry
FIGURE 8: In S679W Lon the first phase is dependent on nucleotide only and the second phase is dependent on ATP and peptide. In a stopped-flow apparatus, 5 µM S679W Lon was incubated with varying concentrations of S4 dansyl peptide (25, 50, 75, 100, 125, 200, 350 or 500 µM) and rapidly mixed with 100 µM ATP or 5 µM S679W Lon was incubated with 500 µM S4 dansyl peptide and rapidly mixed with varying concentrations of ATP (0.5, 1, 2, 4, 7, 10, 25 or 50 µM). The resulting time courses were fitted with eq 3 or eq 4, and the -1 resulting rate constants (k1,S679W and k2,S679W) are shown as a function of substrate concentration. (A) k1,S679W ) 7.6 ( 1.0 s and is independent of peptide concentration. The 0 µM S4 peptide data point (9) was obtained by fitting the time course shown in Figure 5 -1 -1 (S679W + ATP) to eq 1 to yield k1,S679W ) 9.35 ( 0.34 s . (B) k1,S679W ) 5.3 ( 0.6 s is dependent on ATP concentration, Kd ) 4.3 -1 -1 ( 1.9 µM, krev ) 2.1 ( 0.5 s . The data was fitted with eq 5. (C) k2,S679W ) 0.57 ( 0.10 s and is dependent on peptide concentration, -1 -1 Kd ) 157 ( 8 µM, krev ) 0.10 ( 0.01 s , n ) 1.9 ( 0.1. The data was fitted with eq 6. (D) k2,S679W ) 1.5 ( 0.4 s is dependent on ATP concentration, Kd ) 9.3 ( 9.8 µM. The data was fitted with eq 5. All the kinetic parameters are summarized in Table 2. observed rate constants to the indicated ATP or peptide dansylated peptide substrate (S4) and tryptophan residues concentrations. The Kd values for ATP and peptide, as well in Lon. We employed pre-steady-state kinetic techniques to as the maximal rate constant (k2,S679W,max) for each fit, are monitor the dynamics of a dansylated peptide (designated summarized in Table 2. We report the krev rate constant for S4) interacting with two proteolytically inactive E. coli Lon varying concentrations of ATP as “not determined” in Table mutants, S679A and S679W. Both mutants contain three Trp 2 as the standard error for this number was rather high (0.08 residues that are intrinsic to the wild-type enzyme: 297W -1 ( 0.5 s ). The krev rate constant is the y-intercept in the and 303W located at the proximity of the ATPase domain plot, and visually it is approaching zero. Despite the (based on the primary amino acid sequence (15)) and 603W uncertainty involved, the data reveals that the reverse rate at the vicinity of the active site Ser (based on the truncated constant is slow and nearly negligible in comparison to the crystal structure of the protease domain (16)). The S679W forward rate constants. mutant contains an additional Trp residue that replaces the proteolytic active site Ser at position 679. Upon binding to DISCUSSION the target protein, the fluorescence signal associated with a Lon is a serine protease whose activity is regulated by dansylated peptide is indirectly enhanced by excitation of the binding and hydrolysis of ATP. In this study we the intrinsic Trp residues located within the protease. demonstrated that the microscopic events associated with Therefore we can determine the kinetic mechanism of peptide peptide interacting with Lon prior to its cleavage can be interacting with S679A and S679W by analyzing the monitored by the fluorescence signal generated between fluorescence time courses obtained in this study. Biochemistry Mechanism of Peptide Binding to Lon Protease K
FIGURE 9: Proposed mechanism for peptide hydrolysis. The enzyme is shown as a dimer instead of a hexamer for simplicity. The ATPase and SSD (sustrate sensor and discriminatory) domains are shown in green, the protease domain is shown in blue and the active site serine is shown in red. (I) Free enzyme. (II, step 1) ATP and peptide bind in a random order. (III, step 2) A conformational change resulting from nucleotide binding. (IV, step 3) Allosteric activation of the proteolytic site accompanied by ATP hydrolysis. (V, step 4). A slow peptide delivery/translocation event. (VI, step 5) Peptide hydrolysis and product release. The rate constants in boldface are measured in this study. All other rate constants have been published previously by our lab.
The detection of a kS679A sharing a comparable value as bound enzyme form in Lon. Previously we demonstrated that k2,S679W as well as dependency toward [ATP] and [S4] fluorescently labeled ATP and AMPPNP bind to E. coli Lon suggests that the same kinetic step is being examined in the with an identical kinetic mechanism but peptide bond respective mutants. A two-step S4 binding mechanism is cleavage is 7- to 10-fold faster in the presence of ATP proposed. The first step involves the formation of a Lon: compared to AMPPNP (2, 28). As such the step represented ATP:S4 complex, and the second step is a conformational by k1,S679W, which is 10-fold higher in the presence of ATP change in the complex, occurring after ATP is hydrolyzed. than AMPPNP, contributes to the observed difference in the The kinetic progression of the second step is detected by ATP-versus AMPPNP-dependent peptidase activity of Lon the increase in dansyl fluorescence in the S4 peptide upon and must occur after nucleotide binding but before the step -1 excitation of the Trp residues in the Lon mutants. At low representing k2,S679W (k2,S679W ) 0.57 ( 0.10 s , Table 2). [peptide], the binding of S4 to Lon in the presence of ATP The ATP-dependent step is detected only in the S679W limits the observed rate constant. Increasing the concentration but not the S679A mutant. Therefore, we propose that this of peptide leads to enzyme saturation such that the observed step monitors the changes in the local environment surround- rate constants reach a finite value, which is defined by the ing residue 679 in Lon. In the wild type enzyme, such a asymptote of the plots shown in Figures 4A and 8C. change, which is induced by the binding and hydrolysis of Previously we reported the degradation of a fluorogenic ATP, may be needed to activate the proteolytic residue 679S. derivative of the S4 peptide exhibiting sigmodial kinetics, This proposal is made on the basis of a previous observation yielding a Hill coefficient of 1.6 (27). The detection of a made in our lab showing that inhibition of the protease comparable Hill coefficient in the data shown in Figures 3, activity of Lon by the peptide boronate MG262 requires the 4A and 8C is consistent with our previous finding. Likewise, presence of ATP. As the boronate moiety in MG262 the hyperbolic plots shown in Figures 4B and 8B reveal that functions to sequester the nucleophilic side chain of 679S the fluorescence signal detected in this study reflects a (34, 35), the observed dependency toward ATP in protease conformational change in the enzyme after the initial inhibition provides support for an allosteric activation of the formation of a Lon:ATP:S4 complex. active site in Lon. Although both S679A and S679W show the same kinetic Through previous kinetic analyses, we have constructed behavior toward the binding and hydrolysis of ATP, as well a kinetic model to account for the first turnover of the ATP- as their interactions with S4, S679W displays an additional dependent peptide bond cleavage activity of E. coli Lon. The fluorescent signal that is dependent on the concentration of data presented in this study is consistent with the proposed nucleotide but not peptide. Experiments performed with mechanism that is summarized in Figure 9. Free enzyme is S679W reveal a step defined by k1,S679W. The rate constant represented by form I. For simplicity, the enzyme is shown for this step is 7.6 ( 1.0 s-1 in the presence of ATP and is as a dimer with the ATPase and SSD domain in green and 0.50 ( 0.01 s-1, which is ∼10-fold lower, in the presence the protease domain in blue. The active site serine is shown of AMPPNP. Therefore the step represented by k1,S679W is in red. First, ATP and peptide bind to the enzyme in a the key step that distinguishes the ATP versus AMPPNP random mechanism (step 1, form II), followed by a confor- L Patterson-Ward et al. Biochemistry mational change that is induced by nucleotide binding (step providing further support for our proposed mechanism. The 2, form III). Previously, our lab utilized MANT-ATP and detailed simulation data is provided in the Supporting MANT-AMPPNP to study the binding of ATP to Lon (28). Information. This study showed that ATP and AMPPNP bind to Lon in In conclusion, we have utilized stopped flow fluorescence an identical manner wherein ATP or AMPPNP binds to Lon spectroscopy to evaluate the steps leading up to peptide with a rate constant of 0.7 µM-1 s-1 (step 1, Figure 9) hydrolysis in Lon. We have provided evidence for two followed by a nucleotide binding induced conformational distinct conformational change steps along the pathway. One -1 change (kNTP ) 5s , step 2, Figure 9). Equilibrium peptide of these steps (step 4, form V, Figure 9) is relatively slow, binding experiments performed in this study indicate that and we propose it is a peptide translocation/delivery event. peptide initially binds to Lon independent of nucleotide Another step (step 3, form IV, Figure 9) is particularly (Figure 3, Scheme 1, step 1). This result corroborates our valuable in understanding the catalytic advantage offered by previous steady-state product inhibition studies which indi- ATP over AMPPNP as this is the first time differences cated peptide and ATP likely bind to Lon in a random order between ATP and AMPPNP have been assigned to a specific to form the Lon:ATP:peptide complex (2). As the formation step. Additional experimentation is currently underway using of enzyme form IV (step 3) can only be measured with the the techniques described here and other nucleotides to further S679W Lon mutant and it is only dependent on nucleotide study the role of ATP binding and hydrolysis in activating (completely independent of peptide) we propose this step Lon activity. Furthermore, the mechanism described herein involves an allosteric activation of the catalytic serine residue is a valuable tool and will be used as a template in elucidating within the protease domain wherein the active site undergoes the mechanism of how Lon processes multiple sites in a full a gross conformational change. length protein such as lambda N. We propose herein that the slow phase leading up to peptide hydrolysis and responsible for the lag kinetics is the ACKNOWLEDGMENT formation of enzyme form V (step 4, Figure 1) (2, 23). We We would like to thank Gordon Hammes, Anthony Berdis, are able to measure this step using both the S679A and Diana Barko, Greg Tochtrop, and Mary Barkley for helpful S679W Lon mutants. Work by Ishikawa et al. on a related discussions and suggestions during the preparation of this ATP dependent protease, ClpAP, showed that substrates are manuscript. translocated from an initial binding site into the proteolytic chamber (22). ClpAP and Lon are both members of the SUPPORTING INFORMATION AVAILABLE AAA+ (ATPases associated with a variety of cellular activities) family of proteases, and as a slow peptide Kinetic data for the characterization of both the S679A translocation step has been shown in ClpAP, we propose and S679W Lon mutants as well as additional kinetic the slow step we observe prior to peptide hydrolysis in Lon simulation data. This material is available free of charge via the Internet at http://pubs.acs.org. (k2,S679W and kS679A; form V, Figure 1) is also a peptide translocation event. As such, the slow phase could be a REFERENCES peptide delivery event whereby the peptide moves from the initial peptide binding site at the top of the self-compart- 1. Lee, I., and Berdis, A. J. (2001) Adenosine triphosphate-dependent mentalized structure near the ATPase domain, down to the degradation of a fluorescent lambda N substrate mimic by Lon protease, Anal. Biochem. 291,74-83. protease domain where it is degraded. In form VI (step 5) 2. Thomas-Wohlever, J., and Lee, I. (2002) Kinetic characterization peptide is hydrolyzed followed by product release. The idea of the peptidase activity of Escherichia coli Lon reveals the of the dansylated peptide “moving” toward the proteolytic mechanistic similarities in ATP-dependent hydrolysis of peptide - site in Lon has been proposed based upon the discovery of and protein substrates, Biochemistry 41, 9418 9425. 3. Charette, M. F., Henderson, G. W., and Markovitz, A. (1981) ATP a peptide translocation event occurring in ATP-dependent hydrolysis-dependent protease activity of the lon (capR) protein proteases that are related to Lon (20-22). Recently, Reid of Escherichia coli K-12, Proc. Natl. Acad. Sci. U.S.A. 78, 4728- and co-workers labeled the protease active site of ClpAP, a 4732. 4. Chung, C. H., and Goldberg, A. L. (1981) The product of the lon heterosubunit ATP-dependent protease belonging to the same (capR) gene in Escherichia coli is the ATP-dependent protease, family as Lon, with EDANS (5-(2-(acetamido)ethylamino)- protease La, Proc. Natl. Acad. Sci. U.S.A. 78, 4931-4935. naphthalene-1-sulfonic acid) and polypeptide substrates with 5. Goff, S. A., and Goldberg, A. L. (1985) Production of abnormal fluorescein. They used FRET between the two fluorophores proteins in E. coli stimulates transcription of lon and other heat shock genes, Cell 41, 587-595. to demonstrate translocation of the substrate into the pro- 6. Goldberg, A. L., Moerschell, R. P., Chung, C. H., and Maurizi, teolytic active site (21). Unlike the ClpAP experiments, our M. R. (1994) ATP-dependent protease La (lon) from Escherichia system has multiple Trp residues that may interact with the coli, Methods Enzymol. 244, 350-375. dansyl moiety in S4. Therefore, our work cannot conclusively 7. Goldberg, A. L., and Waxman, L. (1985) The role of ATP hydrolysis in the breakdown of proteins and peptides by protease evaluate whether the signal is solely attributed to peptide La from Escherichia coli, J. Biol. Chem. 260, 12029-12034. translocation. However, because the signal is detected only 8. Gottesman, S., Gottesman, M., Shaw, J. E., and Pearson, M. L. when Trp and dansyl are nearby one another, it is possible (1981) Protein degradation in E. coli: the Ion mutation and - that the signal could be reflecting peptide translocation, which bacteriophage lambda N and cll protein stability, Cell 24, 225 233. may be accompanied by a conformational change(s) within 9. Maurizi, M. R. (1992) Proteases and protein degradation in the enzyme. To further test our kinetic model, we utilized Escherichia coli, Experientia 48, 178-201. the program KinFitSim (29) to simulate the pre-steady-state 10. Takaya, A., Suzuki, M., Matsui, H., Tomoyasu, T., Sashinami, lag phase of the peptidase reaction as well as the interaction H., Nakane, A., and Yamamoto, T. (2003) Lon, a stress-induced ATP-dependent protease, is critically important for systemic of S679W and the dansylated peptide. A close correlation Salmonella enterica serovar typhimurium infection of mice, Infect. was found between the experimental and simulated data Immun. 71, 690-696. Biochemistry PAGE EST: 12.6 Mechanism of Peptide Binding to Lon Protease M
11. Takaya, A., Tomoyasu, T., Tokumitsu, A., Morioka, M., and Peptide Cleavage in Escherichia coli Lon by Transient Kinetics, Yamamoto, T. (2002) The ATP-dependent lon protease of Biochemistry 44, 1671-1682. Salmonella enterica serovar Typhimurium regulates invasion and 24. Epps, D., Schostarez, H., Argoudelis, C. V., Poorman, R., expression of genes carried on Salmonella pathogenicity island Hinzmann, J., Sawyer, T. K., and Mandel, F. (1989) An 1, J. Bacteriol. 184, 224-232. Experimental Method for the Determination of Enzyme-Competi- 12. Robertson, G. T., Kovach, M. E., Allen, C. A., Ficht, T. A., and tive Inhibitor Dissociation Constants from Displacement Curves: Roop, R. M., 2nd. (2000) The Brucella abortus Lon functions as Application to Human Renin Using Fluorescence Energy Transfer a generalized stress response protease and is required for wild- to a Synthetic Dansylated Inhibitor Peptide, Anal. Biochem. 181, type virulence in BALB/c mice, Mol. Microbiol. 35, 577-588. 172-181. 13. Bota, D. A., Ngo, J. K., and Davies, K. J. (2005) Downregulation 25. Fattori, D., Urbani, A., Brunetti, M., Ingenito, R., Pessi, A., of the human Lon protease impairs mitochondrial structure and - Prendergast, K., Narjes, F., Matassa, V. G., De Francesco, R., function and causes cell death, Free Radical Biol. Med. 38, 665 and Steinkuhler, C. (2000) Probing the active site of the hepatitis 677. C virus serine protease by fluorescence resonance energy transfer, 14. Amerik, A., Chistiakov, L. G., Ostroumova, N. I., Gurevich, A. J. Biol. Chem. 275, 15106-15113. I., and Antonov, V. K. (1988) [Cloning, expression and structure 26. Gilbert, S. P., and Mackey, A. T. (2000) Kinetics: a tool to study of the functionally active shortened lon gene in Escherichia coli], - Bioorg. Khim. 14, 408-411. molecular motors, Methods 22, 337 354. 15. Chin, D. T., Goff, S. A., Webster, T., Smith, T., and Goldberg, 27. Patterson, J., Vineyard, D., Thomas-Wohlever, J., Behshad, R., A. L. (1988) Sequence of the lon gene in Escherichia coli. A heat- Burke, M., and Lee, I. (2004) Correlation of an adenine-specific shock gene which encodes the ATP-dependent protease La, J. Biol. conformational change with the ATP-dependent peptidase activity - Chem. 263, 11718-11728. of Escherichia coli Lon, Biochemistry 43, 7432 7442. 16. Botos, I., Melnikov, E. E., Cherry, S., Tropea, J. E., Khalatova, 28. Vineyard, D., Zhang, X., and Lee, I. (2006) Transient kinetic A. G., Rasulova, F., Dauter, Z., Maurizi, M. R., Rotanova, T. V., experiments demonstrate the existence of a unique catalytic Wlodawer, A., and Gustchina, A. (2004) The catalytic domain of enzyme form in the peptide-stimulated ATPase mechanism of Escherichia coli Lon protease has a unique fold and a Ser-Lys Escherichia coli Lon protease, Biochemistry 45, 11432-11443. dyad in the active site, J. Biol. Chem. 279, 8140-8148. 29. Svir, I. B., Klymenko, O. V., and Platz, M. S. (2002) ‘KINFITSIM’s 17. Park, S. C., Jia, B., Yang, J. K., Van, D. L., Shao, Y. G., Han, S. a software to fit kinetic data to a user selected mechanism, Comput. W., Jeon, Y. J., Chung, C. H., and Cheong, G. W. (2006) Chem. 26, 379-386. Oligomeric structure of the ATP-dependent protease La (Lon) of 30. Johnson, K. A. (2003) in Practical Approach Series, Oxford Escherichia coli, Mol. Cells 21, 129-134. University Press, Oxford. 18. Maurizi, M. R. (1987) Degradation in vitro of bacteriophage 31. Cook, P., and Cleland, W. W. (2007) Enzyme Kinetics and lambda N protein by Lon protease from Escherichia coli, J. Biol. Mechanism, Garland Science, New York. Chem. 262, 2696-2703. 19. Starkova, N. N., Koroleva, E. P., Rumsh, L. D., Ginodman, L. 32. Lakowicz, J. R. (1999) Principles of Fluorescence Spectroscopy, M., and Rotanova, T. V. (1998) Mutations in the proteolytic 2nd ed., Plenum, New York. domain of Escherichia coli protease Lon impair the ATPase 33. Vineyard, D., Patterson-Ward, J., and Lee, I. (2006) Single- activity of the enzyme, FEBS Lett. 422, 218-220. Turnover Kinetic Experiments Confirm the Existence of High- 20. Lee, C., Schwartz, M. P., Prakash, S., Iwakura, M., and Matous- and Low-Affinity ATPase Sites in Escherichia coli Lon Protease, chek, A. (2001) ATP-dependent proteases degrade their substrates Biochemistry 45, 4602-4610. by processively unraveling them from the degradation signal, Mol. 34. Frase, H., and Lee, I. (2007) Peptidyl boronates inhibit Salmonella Cell 7, 627-637. enterica serovar Typhimurium Lon protease by a competitive ATP- 21. Reid, B. G., Fenton, W. A., Horwich, A. L., and Weber-Ban, E. dependent mechanism, Biochemistry 46, 6647-6657. U. (2001) ClpA mediates directional translocation of substrate 35. Frase, H., Hudak, J., and Lee, I. (2006) Identification of the proteins into the ClpP protease, Proc. Natl. Acad. Sci. U.S.A. 98, proteasome inhibitor MG262 as a potent ATP-dependent inhibitor 3768-3772. of the Salmonella enterica serovar Typhimurium Lon protease, 22. Ishikawa, T., Beuron, F., Kessel, M., Wickner, S., Maurizi, M. Biochemistry 45, 8264-8274. R., and Steven, A. C. (2001) Translocation pathway of protein 36. Menon, A. S., and Goldberg, A. L. (1987) Binding of nucleotides substrates in ClpAP protease, Proc. Natl. Acad. Sci. U.S.A. 98, - to the ATP-dependent protease La from Escherichia coli, J. Biol. 4328 4333. Chem. 262, 14921-14928. 23. Vineyard, D., Patterson-Ward, J., Berdis, A. J., and Lee, I. (2005) Monitoring the Timing of ATP Hydrolysis with Activation of BI701649B BIBLIOGRAPHY
(1) Charette, M. F., Henderson, G. W., and Markovitz, A. (1981) ATP hydrolysis-
dependent protease activity of the lon (capR) protein of Escherichia coli K-12.
Proc Natl Acad Sci U S A 78, 4728-32.
(2) Chung, C. H., and Goldberg, A. L. (1981) The product of the lon (capR) gene in
Escherichia coli is the ATP-dependent protease, protease La. Proc Natl Acad Sci
U S A 78, 4931-5.
(3) Goldberg, A. L., Moerschell, R. P., Chung, C. H., and Maurizi, M. R. (1994)
ATP-dependent protease La (lon) from Escherichia coli. Methods Enzymol 244,
350-75.
(4) Wang, N., Gottesman, S., Willingham, M. C., Gottesman, M. M., and Maurizi, M.
R. (1993) A human mitochondrial ATP-dependent protease that is highly
homologous to bacterial Lon protease. Proc Natl Acad Sci U S A 90, 11247-51.
(5) Wang, N., Maurizi, M. R., Emmert-Buck, L., and Gottesman, M. M. (1994)
Synthesis, processing, and localization of human Lon protease. J Biol Chem 269,
29308-13.
(6) Suzuki, C. K., Kutejova, E., and Suda, K. (1995) Analysis and purification of
ATP-dependent mitochondrial lon protease of Saccharomyces cerevisiae.
Methods Enzymol 260, 486-94.
(7) Amerik, A., Antonov, V. K., Gorbalenya, A. E., Kotova, S. A., Rotanova, T. V.,
and Shimbarevich, E. V. (1991) Site-directed mutagenesis of La protease. A
catalytically active serine residue. FEBS Lett 287, 211-4.
183 (8) Howard-Flanders, P., Simson, E., and Theriot, L. (1964) A Locus That Controls
Filament Formation And Sensitivity To Radiation In Escherichia Coli K-12.
Genetics 49, 237-46.
(9) Neuwald, A. F., Aravind, L., Spouge, J. L., and Koonin, E. V. (1999) AAA+: A
class of chaperone-like ATPases associated with the assembly, operation, and
disassembly of protein complexes. Genome Res 9, 27-43.
(10) Ogura, T., and Wilkinson, A. J. (2001) AAA+ superfamily ATPases: common
structure--diverse function. Genes Cells 6, 575-97.
(11) Ogura, T., Whiteheart, S. W., and Wilkinson, A. J. (2004) Conserved arginine
residues implicated in ATP hydrolysis, nucleotide-sensing, and inter-subunit
interactions in AAA and AAA+ ATPases. J Struct Biol 146, 106-12.
(12) Gottesman, S., and Maurizi, M. R. (1992) Regulation by proteolysis: energy-
dependent proteases and their targets. Microbiol Rev 56, 592-621.
(13) Maurizi, M. R. (1992) Proteases and protein degradation in Escherichia coli.
Experientia 48, 178-201.
(14) Rep, M., and Grivell, L. A. (1996) The role of protein degradation in
mitochondrial function and biogenesis. Curr Genet 30, 367-80.
(15) Goldberg, A. L., and Waxman, L. (1985) The role of ATP hydrolysis in the
breakdown of proteins and peptides by protease La from Escherichia coli. J Biol
Chem 260, 12029-34.
(16) Gottesman, S. (1996) Proteases and their targets in Escherichia coli. Annu Rev
Genet 30, 465-506.
184 (17) Gottesman, S., Gottesman, M., Shaw, J. E., and Pearson, M. L. (1981) Protein
degradation in E. coli: the Ion mutation and bacteriophage lambda N and cll
protein stability. Cell 24, 225-33.
(18) Gottesman, S., and Zipser, D. (1978) Deg phenotype of Escherichia coli lon
mutants. J Bacteriol 133, 844-51.
(19) Schoemaker, J. M., Gayda, R. C., and Markovitz, A. (1984) Regulation of cell
division in Escherichia coli: SOS induction and cellular location of the sulA
protein, a key to lon-associated filamentation and death. J Bacteriol 158, 551-61.
(20) Torres-Cabassa, A. S., and Gottesman, S. (1987) Capsule synthesis in Escherichia
coli K-12 is regulated by proteolysis. J Bacteriol 169, 981-9.
(21) Bota, D. A., Ngo, J. K., and Davies, K. J. (2005) Downregulation of the human
Lon protease impairs mitochondrial structure and function and causes cell death.
Free Radic Biol Med 38, 665-77.
(22) Bota, D. A., and Davies, K. J. (2002) Lon protease preferentially degrades
oxidized mitochondrial aconitase by an ATP-stimulated mechanism. Nat Cell Biol
4, 674-80.
(23) Bota, D. A., and Davies, K. J. (2001) Protein degradation in mitochondria:
implications for oxidative stress, aging and disease: a novel etiological
classification of mitochondrial proteolytic disorders. Mitochondrion 1, 33-49.
(24) Matsui, H., Suzuki, M., Isshiki, Y., Kodama, C., Eguchi, M., Kikuchi, Y.,
Motokawa, K., Takaya, A., Tomoyasu, T., and Yamamoto, T. (2003) Oral
immunization with ATP-dependent protease-deficient mutants protects mice
185 against subsequent oral challenge with virulent Salmonella enterica serovar
typhimurium. Infect Immun 71, 30-9.
(25) Robertson, G. T., Kovach, M. E., Allen, C. A., Ficht, T. A., and Roop, R. M., 2nd.
(2000) The Brucella abortus Lon functions as a generalized stress response
protease and is required for wild-type virulence in BALB/c mice. Mol Microbiol
35, 577-88.
(26) Takaya, A., Suzuki, M., Matsui, H., Tomoyasu, T., Sashinami, H., Nakane, A.,
and Yamamoto, T. (2003) Lon, a stress-induced ATP-dependent protease, is
critically important for systemic Salmonella enterica serovar typhimurium
infection of mice. Infect Immun 71, 690-6.
(27) Takaya, A., Tomoyasu, T., Tokumitsu, A., Morioka, M., and Yamamoto, T.
(2002) The ATP-dependent lon protease of Salmonella enterica serovar
Typhimurium regulates invasion and expression of genes carried on Salmonella
pathogenicity island 1. J Bacteriol 184, 224-32.
(28) Frase, H., Hudak, J., and Lee, I. (2006) Identification of the proteasome inhibitor
MG262 as a potent ATP-dependent inhibitor of the Salmonella enterica serovar
Typhimurium Lon protease. Biochemistry 45, 8264-74.
(29) Frase, H., and Lee, I. (2007) Peptidyl boronates inhibit Salmonella enterica
serovar Typhimurium Lon protease by a competitive ATP-dependent mechanism.
Biochemistry 46, 6647-57.
(30) Suzuki, C. K., Suda, K., Wang, N., and Schatz, G. (1994) Requirement for the
yeast gene LON in intramitochondrial proteolysis and maintenance of respiration.
Science 264, 891.
186 (31) van Dyck, L., Neupert, W., and Langer, T. (1998) The ATP-dependent PIM1
protease is required for the expression of intron-containing genes in mitochondria.
Genes Dev 12, 1515-24.
(32) Van Dyck, L., Pearce, D. A., and Sherman, F. (1994) PIM1 encodes a
mitochondrial ATP-dependent protease that is required for mitochondrial function
in the yeast Saccharomyces cerevisiae. J Biol Chem 269, 238-42.
(33) Botos, I., Melnikov, E. E., Cherry, S., Tropea, J. E., Khalatova, A. G., Rasulova,
F., Dauter, Z., Maurizi, M. R., Rotanova, T. V., Wlodawer, A., and Gustchina, A.
(2004) The catalytic domain of Escherichia coli Lon protease has a unique fold
and a Ser-Lys dyad in the active site. J Biol Chem 279, 8140-8.
(34) Chin, D. T., Goff, S. A., Webster, T., Smith, T., and Goldberg, A. L. (1988)
Sequence of the lon gene in Escherichia coli. A heat-shock gene which encodes
the ATP-dependent protease La. J Biol Chem 263, 11718-28.
(35) Li, M., Rasulova, F., Melnikov, E. E., Rotanova, T. V., Gustchina, A., Maurizi,
M. R., and Wlodawer, A. (2005) Crystal structure of the N-terminal domain of E.
coli Lon protease. Protein Sci 14, 2895-900.
(36) Botos, I., Melnikov, E. E., Cherry, S., Khalatova, A. G., Rasulova, F. S., Tropea,
J. E., Maurizi, M. R., Rotanova, T. V., Gustchina, A., and Wlodawer, A. (2004)
Crystal structure of the AAA+ alpha domain of E. coli Lon protease at 1.9A
resolution. J Struct Biol 146, 113-22.
(37) Smith, C. K., Baker, T. A., and Sauer, R. T. (1999) Lon and Clp family proteases
and chaperones share homologous substrate-recognition domains. Proc Natl Acad
Sci U S A 96, 6678-82.
187 (38) Kisselev, A. F., Songyang, Z., and Goldberg, A. L. (2000) Why does threonine,
and not serine, function as the active site nucleophile in proteasomes? J Biol
Chem 275, 14831-7.
(39) Goldberg, A. L. (1992) The mechanism and functions of ATP-dependent
proteases in bacterial and animal cells. Eur J Biochem 203, 9-23.
(40) Thomas-Wohlever, J., and Lee, I. (2002) Kinetic characterization of the peptidase
activity of Escherichia coli Lon reveals the mechanistic similarities in ATP-
dependent hydrolysis of peptide and protein substrates. Biochemistry 41, 9418-25.
(41) Lee, I., and Berdis, A. J. (2001) Adenosine triphosphate-dependent degradation of
a fluorescent lambda N substrate mimic by Lon protease. Anal Biochem 291, 74-
83.
(42) Maurizi, M. R. (1987) Degradation in vitro of bacteriophage lambda N protein by
Lon protease from Escherichia coli. J Biol Chem 262, 2696-703.
(43) Waxman, L., and Goldberg, A. L. (1982) Protease La from Escherichia coli
hydrolyzes ATP and proteins in a linked fashion. Proc Natl Acad Sci U S A 79,
4883-7.
(44) Waxman, L., and Goldberg, A. L. (1985) Protease La, the lon gene product,
cleaves specific fluorogenic peptides in an ATP-dependent reaction. J Biol Chem
260, 12022-8.
(45) Lakowicz, J. R. (1999) Principles of Fluorescence Spectroscopy, second ed.,
Plenum, New York.
(46) Patterson, J., Vineyard, D., Thomas-Wohlever, J., Behshad, R., Burke, M., and
Lee, I. (2004) Correlation of an adenine-specific conformational change with the
188 ATP-dependent peptidase activity of Escherichia coli Lon. Biochemistry 43,
7432-42.
(47) Rohrwild, M., Coux, O., Huang, H. C., Moerschell, R. P., Yoo, S. J., Seol, J. H.,
Chung, C. H., and Goldberg, A. L. (1996) HslV-HslU: A novel ATP-dependent
protease complex in Escherichia coli related to the eukaryotic proteasome. Proc
Natl Acad Sci U S A 93, 5808-13.
(48) Rohrwild, M., Pfeifer, G., Santarius, U., Muller, S. A., Huang, H. C., Engel, A.,
Baumeister, W., and Goldberg, A. L. (1997) The ATP-dependent HslVU protease
from Escherichia coli is a four-ring structure resembling the proteasome. Nat
Struct Biol 4, 133-9.
(49) Park, E., Rho, Y. M., Koh, O. J., Ahn, S. W., Seong, I. S., Song, J. J., Bang, O.,
Seol, J. H., Wang, J., Eom, S. H., and Chung, C. H. (2005) Role of the GYVG
pore motif of HslU ATPase in protein unfolding and translocation for degradation
by HslV peptidase. J Biol Chem 280, 22892-8.
(50) Wang, J., Song, J. J., Franklin, M. C., Kamtekar, S., Im, Y. J., Rho, S. H., Seong,
I. S., Lee, C. S., Chung, C. H., and Eom, S. H. (2001) Crystal structures of the
HslVU peptidase-ATPase complex reveal an ATP-dependent proteolysis
mechanism. Structure 9, 177-84.
(51) Menon, A. S., and Goldberg, A. L. (1987) Binding of nucleotides to the ATP-
dependent protease La from Escherichia coli. J Biol Chem 262, 14921-8.
(52) Vineyard, D., Patterson-Ward, J., and Lee, I. (2006) Single-Turnover Kinetic
Experiments Confirm the Existence of High- and Low-Affinity ATPase Sites in
Escherichia coli Lon Protease. Biochemistry 45, 4602-10.
189 (53) Segel, I. H. (1993) Enzyme Kinetics Behavior and Analysis of Rapid Equilibrium
and Steady-State Enzyme Systems, John Wiley and Sons, New York.
(54) Pate, E., Franks-Skiba, K., White, H., and Cooke, R. (1993) The use of differing
nucleotides to investigate cross-bridge kinetics. J Biol Chem 268, 10046-53.
(55) Kenniston, J. A., Baker, T. A., Fernandez, J. M., and Sauer, R. T. (2003) Linkage
between ATP consumption and mechanical unfolding during the protein
processing reactions of an AAA+ degradation machine. Cell 114, 511-20.
(56) Lee, C., Schwartz, M. P., Prakash, S., Iwakura, M., and Matouschek, A. (2001)
ATP-dependent proteases degrade their substrates by processively unraveling
them from the degradation signal. Mol Cell 7, 627-37.
(57) Weber-Ban, E. U., Reid, B. G., Miranker, A. D., and Horwich, A. L. (1999)
Global unfolding of a substrate protein by the Hsp100 chaperone ClpA. Nature
401, 90-3.
(58) Johnson, K. A. (1992) Transient-state kinetic analysis of enzyme reaction
pathways, in The Enzymes (Sigman, D. S., Ed.), Academic Press.
(59) Johnson, K. A. (2003) in Practical Approach Series, Oxford University Press.
(60) Copeland, R. A. (2000) Enzymes: A Practical Introduction to Structure,
Mechanism, and Data Analysis, 2nd ed., John Wiley & Sons, Inc., New York.
(61) Vineyard, D., Patterson-Ward, J., Berdis, A. J., and Lee, I. (2005) Monitoring the
Timing of ATP Hydrolysis with Activation of Peptide Cleavage in Escherichia
coli Lon by Transient Kinetics. Biochemistry 44, 1671-82.
(62) Gilbert, S. P., and Mackey, A. T. (2000) Kinetics: a tool to study molecular
motors. Methods 22, 337-54.
190 (63) Frieden, C. (1970) Kinetic aspects of regulation of metabolic processes. The
hysteretic enzyme concept. J Biol Chem 245, 5788-99.
(64) Frieden, C. (1979) Slow transitions and hysteretic behavior in enzymes. Annu Rev
Biochem 48, 471-89.
(65) Ishikawa, T., Beuron, F., Kessel, M., Wickner, S., Maurizi, M. R., and Steven, A.
C. (2001) Translocation pathway of protein substrates in ClpAP protease. Proc
Natl Acad Sci U S A 98, 4328-33.
(66) Reid, B. G., Fenton, W. A., Horwich, A. L., and Weber-Ban, E. U. (2001) ClpA
mediates directional translocation of substrate proteins into the ClpP protease.
Proc Natl Acad Sci U S A 98, 3768-72.
(67) Menon, A. S., and Goldberg, A. L. (1987) Protein substrates activate the ATP-
dependent protease La by promoting nucleotide binding and release of bound
ADP. J Biol Chem 262, 14929-34.
(68) Epps, D., Schostarez, H., Argoudelis, C.V., Poorman, R., Hinzmann, J., Sawyer,
T.K., Mandel, F. (1989) An Experimental Method for the Determination of
Enzyme-Competitive Inhibitor Dissociation Constants from Displacement
Curves: Application to Human Renin Using Fluorescence Energy Transfer to a
Synthetic Dansylated Inhibitor Peptide. Analytical Biochemistry 181, 172-181.
(69) Fattori, D., Urbani, A., Brunetti, M., Ingenito, R., Pessi, A., Prendergast, K.,
Narjes, F., Matassa, V. G., De Francesco, R., and Steinkuhler, C. (2000) Probing
the active site of the hepatitis C virus serine protease by fluorescence resonance
energy transfer. J Biol Chem 275, 15106-13.
191 (70) Starkova, N. N., Koroleva, E. P., Rumsh, L. D., Ginodman, L. M., and Rotanova,
T. V. (1998) Mutations in the proteolytic domain of Escherichia coli protease Lon
impair the ATPase activity of the enzyme. FEBS Lett 422, 218-20.
(71) Hansen, S. B., Radic, Z., Talley, T. T., Molles, B. E., Deerinck, T., Tsigelny, I.,
and Taylor, P. (2002) Tryptophan fluorescence reveals conformational changes in
the acetylcholine binding protein. J Biol Chem 277, 41299-302.
(72) Flowers, S., Biswas, E. E., and Biswas, S. B. (2003) Conformational dynamics of
DnaB helicase upon DNA and nucleotide binding: analysis by intrinsic tryptophan
fluorescence quenching. Biochemistry 42, 1910-21.
(73) Zhu, W., and Becker, D. F. (2005) Exploring the proline-dependent
conformational change in the multifunctional PutA flavoprotein by tryptophan
fluorescence spectroscopy. Biochemistry 44, 12297-306.
(74) Vineyard, D., Zhang, X., and Lee, I. (2006) Transient kinetic experiments
demonstrate the existence of a unique catalytic enzyme form in the peptide-
stimulated ATPase mechanism of Escherichia coli Lon protease. Biochemistry 45,
11432-43.
(75) Wu, P., and Brand, L. (1994) Resonance energy transfer: methods and
applications. Anal Biochem 218, 1-13.
(76) Svir, I. B., Klymenko, O. V., and Platz, M. S. (2002) 'KINFITSIM'--a software to
fit kinetic data to a user selected mechanism. Comput Chem 26, 379-86.
(77) Mizusawa, S., and Gottesman, S. (1983) Protein degradation in Escherichia coli:
the lon gene controls the stability of sulA protein. Proc Natl Acad Sci U S A 80,
358-62.
192 (78) Levchenko, I., Seidel, M., Sauer, R. T., and Baker, T. A. (2000) A specificity-
enhancing factor for the ClpXP degradation machine. Science 289, 2354-6.
(79) Ishii, Y., and Amano, F. (2001) Regulation of SulA cleavage by Lon protease by
the C-terminal amino acid of SulA, histidine. Biochem J 358, 473-80.
(80) Ishii, Y., Sonezaki, S., Iwasaki, Y., Miyata, Y., Akita, K., Kato, Y., and Amano,
F. (2000) Regulatory role of C-terminal residues of SulA in its degradation by
Lon protease in Escherichia coli. J Biochem (Tokyo) 127, 837-44.
(81) Nishii, W., Maruyama, T., Matsuoka, R., Muramatsu, T., and Takahashi, K.
(2002) The unique sites in SulA protein preferentially cleaved by ATP-dependent
Lon protease from Escherichia coli. Eur J Biochem 269, 451-7.
(82) Shah, I. M., and Wolf, R. E., Jr. (2006) Sequence requirements for Lon-dependent
degradation of the Escherichia coli transcription activator SoxS: identification of
the SoxS residues critical to proteolysis and specific inhibition of in vitro
degradation by a peptide comprised of the N-terminal 21 amino acid residues. J
Mol Biol 357, 718-31.
(83) Van Melderen, L., Thi, M. H., Lecchi, P., Gottesman, S., Couturier, M., and
Maurizi, M. R. (1996) ATP-dependent degradation of CcdA by Lon protease.
Effects of secondary structure and heterologous subunit interactions. J Biol Chem
271, 27730-8.
(84) Hersch, G. L., Baker, T. A., and Sauer, R. T. (2004) SspB delivery of substrates
for ClpXP proteolysis probed by the design of improved degradation tags. Proc
Natl Acad Sci U S A 101, 12136-41.
193 (85) Higashitani, A., Ishii, Y., Kato, Y., and Koriuchi, K. (1997) Functional dissection
of a cell-division inhibitor, SulA, of Escherichia coli and its negative regulation
by Lon. Mol Gen Genet 254, 351-7.
(86) Navon, A., and Goldberg, A. L. (2001) Proteins are unfolded on the surface of the
ATPase ring before transport into the proteasome. Mol Cell 8, 1339-49.
(87) Sauer, R. T., Bolon, D. N., Burton, B. M., Burton, R. E., Flynn, J. M., Grant, R.
A., Hersch, G. L., Joshi, S. A., Kenniston, J. A., Levchenko, I., Neher, S. B.,
Oakes, E. S., Siddiqui, S. M., Wah, D. A., and Baker, T. A. (2004) Sculpting the
proteome with AAA(+) proteases and disassembly machines. Cell 119, 9-18.
(88) Wellings, D. A., and Atherton, E. (1997) Standard Fmoc protocols. Methods
Enzymol 289, 44-67.
(89) Bodanszky, M. (1993) Peptide Chemistry A Practical Textbook 2nd Edition,
Springer-Verlag, New York.
(90) Chan, W. C., and White, P. D. (2003) Fmoc Solid Phase Peptide Synthesis,
Oxford University Press, New York.
(91) Black, S. D., and Mould, D. R. (1991) Development of hydrophobicity parameters
to analyze proteins which bear post- or cotranslational modifications. Anal
Biochem 193, 72-82.
(92) Mogridge, J., Legault, P., Li, J., Van Oene, M. D., Kay, L. E., and Greenblatt, J.
(1998) Independent ligand-induced folding of the RNA-binding domain and two
functionally distinct antitermination regions in the phage lambda N protein. Mol
Cell 1, 265-75.
194 (93) Prasch, S., Schwarz, S., Eisenmann, A., Wohrl, B. M., Schweimer, K., and Rosch,
P. (2006) Interaction of the intrinsically unstructured phage lambda N Protein
with Escherichia coli NusA. Biochemistry 45, 4542-9.
(94) Rees, W. A., Weitzel, S. E., Yager, T. D., Das, A., and von Hippel, P. H. (1996)
Bacteriophage lambda N protein alone can induce transcription antitermination in
vitro. Proc Natl Acad Sci U S A 93, 342-6.
(95) Van Gilst, M. R., Rees, W. A., Das, A., and von Hippel, P. H. (1997) Complexes
of N antitermination protein of phage lambda with specific and nonspecific RNA
target sites on the nascent transcript. Biochemistry 36, 1514-24.
(96) Van Gilst, M. R., and von Hippel, P. H. (1997) Assembly of the N-dependent
antitermination complex of phage lambda: NusA and RNA bind independently to
different unfolded domains of the N protein. J Mol Biol 274, 160-73.
195