INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter ^ce, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event tfiat the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right In equal sections with small overlaps.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.

Bell & Howell Information and Leaming 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600 UMI'

SUBSTRATE RECOGNITION THROUGH MODULAR DOMAINS; PROTEIN

TYROSINE PHOSPHATASE SHP-1 AND TAIL-SPECIFIC PROTEASE (TSP)

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy

in the Graduate School of the Ohio State University

By

Kirk Beebe

*****

The Ohio State University

2000

Dissertation Committee:

Professor Dehua Pei. Adviser

Professor Charles Brooks Approved by

Professor Richard Swenson

Professor James Cowan

Adviser

The Ohio State Biochemistry Program UMI Number: 9994840

UMI

UMI Microform 9994840 Copyright 2001 by Bell & Howell Information and Leaming Company. All rights reserved. This microform edition Is protected against unauthorized copying under Title 17, United States Code.

Bell & Howell Information and Leaming Company 300 North Zeeb Road P.O. 00x1346 Ann Arbor, Ml 48106-1346 ABSTRACT

The nature of two domain regulated enzymes Tail-specific protease (Tsp) from E. coli and SHP-1 protein tyrosine phosphatase from humans has been examined. Tail specific protease (Tsp) is a periplasmic enzyme which selectively degrades proteins bearing a nonpolar C-terminus. To facilitate the study of this enzyme, a continuous tluorimetric assay using a fluorescence donor/quencher system, in which 5-[(2- aminoethyl)amino]naphthalene-l-sulfonic acid (ED ANS) and 4-(4- dimethylaminophenylazo)benzoic acid (DABCYL) are attached to the N-terminus and the lysyl side chain of peptide AARAAK-( 6-aminocaproyl) 2-ENYALAA. respectively was developed. Using this assay along with photoaffinity labeling, further analysis of the protease was carried out in which it was demonstrated that a PDZ domain mediates substrate recognition of Tsp. The use of a separate substrate recognition domain such as a

PDZ domain may be a general mechanism for achieving selective protein degradation. In a similar manner. SH2 domains mediate the activity of SHP-1 and the specificity of theses domains holds the key to the functional activity of the enzyme. To closely analyze these domains and how they may direct the activity of SHP-1. a synthetic phosphotvTosyl peptide library was constructed and screened for binding to both the SH2(N) and SH2(C) domains. The N-terminal SH2 domain is most selective for a leucine at the -2 position: at the C-terminal side of the pY residue, it can recognize two distinct classes of peptides with consensus sequences o f LXpY(M/F)X(F/M) and LXpYAXL (X = any amino acid), respectively. The C-terminal SH2 domain exhibits almost exclusive selectivitv' for peptides of the consensus sequence, (V/I/L)XpYAX(L/V). Several representative sequences selected from the library were individually synthesized and tested for binding to the SH2 domains by surface plasmon resonance and for their ability to stimulate the catalytic activity of SHP-1.

Ill DEDICATED

TO

MY

MOTHER

IV ACKNOWLEDGMENT

I would like to express my appreciation to my adviser. Dehua Pei. for all of the hard work and effort that he put into the projects 1 worked on in the lab. 1 thank him for having a genuine interest in the students' welfare while being able to maintain a highly productive lab. I want to thank my lab associates for assisting me and giving me advice throughout my graduate career. Particularly 1 would like to thanks Jinu Shin. Peng

Wang, and Gulnur Arabaci for their contribution to the projects in this dissertation. I also thank my committee members for their time throughout my graduate career.

Most of all I would like to thank and dedicate this degree to three individuals and my God. The person most responsible for who 1 have become is my Mother. It is my

Mother who first exemplified how to work hard, demonstrated to me how to put others needs first and it is she who has always offered encouragement through anything I have ever done. 1 cannot thank her enough for all she has done for me and 1 love her for it.

With respect to encouragement. 1 am so grateful that I have a brother who encouraged me to tackle things that I never felt 1 had the aptitude to accomplish, while serving as an example of someone who could do these things. I could not ask for a better friend. This thesis is also dedicated to the precious jewel that I discovered in graduate school, my wife

Melanie. Melanie is what gives meaning to this degree and without a doubt the most significant accomplishment I made in graduate school was committing to loving her for the rest of my life. Lastly I would like to dedicate this work to my loving savior Jesus

Christ, who purchased my soul with His blood and without Him. my time on earth graduate school, or life in general, would be utterly futile.

VI VITA

September 24. 1969 Bom - Pontiac. Michigan

1995...... B.S. Central Michigan University

1995 - present ...... Graduate Teaching and Research Associate The Ohio State University

PUBLICATIONS

Research Publications

1. Beebe. K... D.. Wang, P., Arabaci. G.. & Pei. D. "Determination of the Binding Specificity of the SH2 Domains of Protein Tyrosine Phosphatase SHP-1 through the Screening of a Combinatorial PhosphotvTosyl Peptide Librarv.” Biochemistn,' 39 13251- 13260.(2000).

2. Wang, P.. Byeon. I-J.. L.. Liao. H., Beebe. K., D.. Yongkiettrakul. S.. Pei. D.. & Tsai. M-D. "Structure And Specificity of the Interaction between the FHA2 Domain of Rad 53 and Phosphotyrosyl Peptides.” Journal o f Molecular Biology 302(4). 927-940 (2000).

3. Beebe. K.. D.. Shin. J.. Peng, J.. Chaudhury. C.. Khera. J.. & Pei. D.. "Substrate Recognition through a PDZ Domain in Tail-Specific Protease.” Biochemistrv 39. 3149- 3155(2000).

4. Arabaci. G.. Guo. X-C.. Beebe, K., D.. Coggeshall. K.. M.. & Pei. D.. "a- Haloacetophenone Derivatives as photoreversible Covalent Inhibitors of Protein Tyrosine Phosphatases.” Journal o f the American Chemical Society 121. 5085-86 (1999).

5. Beebe. K.. D.. & Pei. D.. "A Continuous Fluorimetric Assay for Tail-Specific Protease." Analytical Biochemistry 263. 51-56 (1998).

FIELDS OF STUDY

Major Field: Biochemistry

VII TABLE OF CONTENTS

Page

Abstract ...... ü

Dedication ...... iv

Acknowledgments ...... v

Vita...... vii

List of Tables ...... xii

List of Figures ...... xiü

List of abbreviations ...... xv

Chapters:

1. General Introduction ...... l

1.1 Src homology phosphatase-1. SHP-1 ...... 2

1.1.1 SH2 domain function ...... 4

l. 1.2 SH2 domain structure ...... 4

1.1.3 Research objectives...... 7

1.2 Tail-specific protease ...... 8

1.2.1 PDZ domain function ...... 10

l .2.2 PDZ domain structure ...... 11

l .2.3 Research objectives ...... 12

2. Development of a continuous assay for Tsp ...... 15

2.1 Introduction ...... 15 viii 2.2 Experimental procedures ...... 17

2.2.1 Materials and Methods ...... 17

2.2.2 Purification of Tail-specific protease ...... 17

2.2.3 Analytical Methods ...... 18

2.2.4 Substrate and inhibitor preparation ...... 19

2.2.5 fsp assay ...... 21

2.2.6 HPLC and mass spectrometry analysis of Tsp products....22

2.2.7 Inhibition of Tsp ...... 22

2.3 Results...... 22

2.3.1 Design and synthesis of Tsp substrates ...... 22

2.3.2 Tsp assays ...... 25

2.3.3 .Analysis of cleavage products ...... 27

2.3.4 Inhibition of Tsp ...... 28

2.4 Conclusions ...... 29

3. Substrate recognition through a PDZ domain in Tail-specific protease ...... 36

3.1 Intoduction ...... 38

3.2 Experimental Procedures ...... 38

3.2.1 Materials...... 38

3.2.2 Synthesis of photoaffinity labels ...... 38

3.2.3 Construction of Tsp mutants ...... 39

3.2.4 Purification of wild type and mutant Tsp ...... 40

3.2.5 Tsp assay ...... 41

3.2.6 Photoaffinity labeling ...... 41

i.x 3.2.7 V 8 proteolysis of Tsp ...... 43

3.2.8 Binding assays ...... 43

3.3 Results...... 44

3.3.1 V8 proteolysis of Tsp ...... 44

3.3.2 PhotoafFinity labeling ...... 45

3.3.3 Direct binding of peptide 1 to Tsp and PDZ domains 49

3.3.4 Catalytic properties of Tsp mutants ...... 50

3.3 Discussion ...... 52

4. Development of combinatorial method for determining the ...... 66 binding specificity of the SH2 domains of SHP-1

4.1 Introduction ...... 66

4.2 Experimental procedures ...... 69

4.2.1 Materials and methods ...... 69

4.2.2 Library construction ...... 70

4.2.3 Library screening and MALDl peptide sequencing ...... 71

4.2.4 SH2-pY peptide binding assays ...... 73

4.2.5 Analysis of novel consensus sequence ...... 74

3.3 Results...... 74

4.3.1 Library design, construction, and screening ...... 74

4.3.2 Analysis of selected sequences ...... 78

4.3.3 Affinity measurment of selected peptides ...... 83

4.3.4 Stimulation of SHP-1 avtivity by pY peptides ...... 85

3.4 Discussion ...... 88

X Appendix ...... 108

I. Proteolytic targeting by Tsp ...... 108

II. PTP inhibitor search ...... 112

List of References ...... 116

XI LIST OF TABLES

Table Paae

3.1 Kinetic constants...... 51

4.1 SH2(N) selected sequences(Ac-LKpYXXXX) ...... 79

4.2 SH2(N) selected sequences(Ac-DEXXpYXXX) ...... 80

4.3 SH2(C) selected sequences!Ac-DEXXp YXXXl ...... 81

4.4 Dissociation constants ...... 86

4.5 Reported SHP-1 binding sites ...... 92

4.5 Pattern match of novel consensus ...... 97

.XII LIST OF FIGURES

Figure Page

1.1 Crystal structure of a SH2 domain ...... 13

1.2 Crystal structure of a PDZ domain ...... 14

2.1 SDS-PAGE gel of Tsp purification ...... 30

2.2 Synthesis of Tsp substrate ...... 31

2.3 Reaction course ...... 32

2.4 Initial reaction rates and Tsp assay ...... 33

2.5 HPLC analysis of Tsp digest of peptide 1 ...... 34

2.6 Inhibition of Tsp ...... 35

3.1 Structure of photoaffmity labels ...... 56

3.2 Tsp mutants ...... 57

3.3 Tsp cleavage by V 8 ...... 58

3.4 V8 proteolyzed and denatured TspH& ...... 59

3.5 Tsp crosslinking to peptide 1 ...... 61

3.6 Competition of crosslinking by peptide 2 ...... 62

3.7 Fluorescence resonance energy transfer ...... 63

3.8 Proposed model of Tsp substrate recognition ...... 64

3.9 Crystal structure of 01 protease ...... 65

4.1 MALDI-mass spectrometry ...... 102

xiii 4.2 Specificity of SH2(N) by Ac-LKpYXXXX ...... 103

4.3 Specificity of SH2(N) and SH2(C) by Ac-EDXXpYXXX ...... 104

4.4 BIAcore analysis ...... 105

4.5 Activation of SHP-1 by pY peptides ...... 106

4.6 Known receptor motifs ...... 107

A. 1 Inhibitors...... 115

XIV ABREVIATIONS

Nie. norleucine

BCIP. 5-bromo-4-chloro-3-indolyl-phosphate

SH2. Src homology 2

MALDI MS. matrix assisted laser desorption ionization mass spectrometry

MBP. maltose-binding protein

GST. glutathione-S-transferase.

Tsp. tail specific protease

EDANS. 5-[{2-aminoethyl)amino]naphthalene-l-sulfonic acid

DABCYL. 4-(4-dimethylaminophenylazo)benzoic acid

HOBT. 1-hydroxybenzotriazole

HBTU. 2-( 1 H-benzotriazol-l -yl)-1.1.3.3-tetramethyluronium hexafluorophosphate:

Fmoc. 9-fluorenylmethyloxycarbonyl

SDS-PAGE. sodium dodecylsulfate-polyacrylamide gel electrophoresis

TOP. time of flight

SP. sulfopropyl

Q. quaternary amine

ESI MS, electrospray ionization mass spectrometry

PY. phosphotyrosine

PTP, protein tyrosine phosphatase

XV CHAPTER 1

GENERAL INTRODUCTION

In vivo regulation of various signaling processes often involves the localization or arrangement of a protein or enzyme within a cell such that its activity is targeted to the appropriate site of action. One of the major means by which this occurs is due to the recruitment and or arrangement via small modular protein domains that include PH. WW.

PTB. SH3. PDZ. and SH2 (Kuriyan and Cowbum. 1997). Regulation, recruitment, and signal amplification (Pawson. 1995; Kuriyan and Cowbum. 1997) occur as a result of these domains. Proteases, kinases, phosphatases and many other enzymes are governed or assisted by these recombinately attached small modular domains. Many of these domains bind small peptide motifs within a specific partner protein resulting in the association of proteins and subsequent change in signal transduction. Some of these domains recognize posttranslationally modified amino acid residues such as pSer. pThr. or pTyr. Change in phosphorylation state leads to propagation, abrogation, or amplification of signals, or localization to a particular compartment. Many times this phosphorylation results in the binding of SH2 domains to a particular site of action. In many cases the SH2 domain recruits an enzymatic body to this site resulting in a message being sent along a complex signal transduction pathway. Other modular domains bind to specific peptide motifs without the requirement of any posttranslational modification.

The PDZ domain is known to bind to apolar C-terminal peptides with a tree carboxy- terminus, mediating protein-protein interaction.

Two enzymes of interest in maintaining cellular viability and determined to utilize the above-mentioned modular domains are the SH2 domain containing phosphatase.

SHP-1 and tail specific protease. Tsp. Two levels of studies on these two enzymes are undertaken in this work. Further refinement of a well defined enzyme SHP-1 was carried out by precisely determining the binding specificity of its SH2 domains. On another level, more basic characterization of a less clearly defined enzyme was carried out in the determination of how a PDZ domain serves to regulate tail specific protease (Tsp).

l.l Src homology phosphatase-1 (SHP-1)

Src homology phosphatase-1 (SHP-1) is one of only two known dual SH2 domain-containing phosphatases in humans and was first discovered in 1991 (Shen et uL.

1991; Plutzky et al.. 1992: Yi et a i. 1992). The enzyme was cloned, overexpressed, and characterized in a series of papers in the mid-1990s (Shen et al.. 1991: Pei et al.. 1993:

Pei et al.. 1994: Pei et al.. 1996). Sequence homology revealed the presence of two N- terminal SH2 domains one more N-terminal designated SH2(N). and one more C- terminal designated SH2(C). Initially, it was discovered that SHP-1 exhibited non­ saturation kinetics toward both small molecule and pY peptide substrates, displaying linear activity versus substrate concentration plots (Pei et al.. 1994). The explanation for this was revealed in the fact that the SH2(N) intramolecularly inhibits the catalytic domain of the phosphatase while the SH2(C) primarily served a role in recruitment of the enzyme to an appropriate pY substrate (Pei et al.. 1996). Indeed, deletion of the SH2(N). resulted in typical Michaelis-Menten kinetic behavior (Pei et al.. 1994). Crystallographic data of the 60% homologous, only other dual SH2 domain containing phosphatase in humans, S HP-2, supported the model derived from biochemical studies and yielded a specific explanation for the inhibition (Hof et al.. 1998). Structural data demonstrated that a loop region in the SH2(N) (DE loop) served as a substrate mimic while simultaneously preventing the closure of a loop within the catalytic domain that was composed of a key general acid/base aspartic acid (Barford and Neel. 1998: Hof et a i.

1998). Additionally upon SH2(N) binding to the catalytic domain, the SH2 pY-peptide binding site is perturbed such that its affinity for pY-peptides is dramatically decreased.

The importance of this lies in the fact that the SH2(C) is not involved in the autoinhibition and its pY-peptide binding site is entirely competent for binding, validating the biochemical data that proposes a role for the SH2(C) in the recruitment of the enzyme to an appropriate site of action. It is also noted that dual phosphor\ lated peptides separated by - 40 Â (Eck et al.. 1996) or 23-25 amino acids serve to most effectively bind SHP-1 (Ottinger et al., 1998). It can be envisioned that the SH2(C)

recruits the enzyme to an appropriate bis-phosphorylated receptor thus enhancing the

avidity of the second SH2(N) site, which has weakened affinity for pY peptides in the

PTP-bound state, for switching to the pY-peptide bound state from the SH2(N)-PTP

bound state. Upon occupation of the SH2(N) to this second pY binding site, inhibition is

relieved and catalysis proceeds efficiently. Physiologically, SHP-1 is implicated in a host

of signaling pathways but is generally thought to abrogate signals that serve to down regulate processes (Tonks. 1996; Neel and Tonks, 1997). SHP-1 is particularly important in immune cell signaling, specifically, B and T-cells (Siminovitch and Neel. 1998; Tamir et al., 20 0 0 ).

1.1.1 SH2 domain function

The SH2 domains of SHP-1 are structurally homologous to all known SH2 domains. SH2 domains typically serve to localize or dock proteins involved in a specific signal transduction cascade to the proper site of action (Pawson and Gish. 1992: Pawson.

1995: Pawson and Scott, 1997) however there are several examples of direct autoregulation of the enzyme by which they are attached. Certain Src family kinases also employ a type of negative regulation, as does SHP-1. The means by which this occurs is different in each case but the end result is that of an enzyme that is autoinhibited by its own SH2 domain when the proper signal is absent. The signal in the case of SHP-1 is that of an appropriate pY peptide that binds to the most N-terminal SH2 of SHP-1. Due to their role in regulating this critical immune cell signaling enzyme, either by recruiting it to a specific site of action or directly inhibiting the enzyme, the SH2 domains of SHP-1 are of great interest.

1.1.2 SH2 domain structure

SH2 domains (Sadowski et al„ 1986) are -100 amino acids in length and are primarily (3-sheet, with minimal helical content. The SH2 domain is characterized by having one large central (3-sheet (typically; (3A-PG) flanked by two a-helices (oA and aB) (Eck et al., 1993; Lee et al., 1994; Mulhem et al., 1997; S legal et al.. 1998). The

4 typical SH2 ligand binds perpendicularly across the central (3-sheet with most of the interactions being in a pocket composed of oA and the PD strand (Figure 1.1) The pY binds in an extended conformation interacting through a salt bridge with a conserved

.Arg(PB5). Data produced from microcalorimetry studies calculates that 50% of the binding energy of an SH2 domain to a pY-ligand can be attributed to the pY pocket with the remainder of the amino acid residues in the peptide making up the other 50% (Grucza el al.. 1999). Typically SH2 domain specificity is largely driven by the residues C- terminal to the pY (Songyang et al., 1993; Kuriyan and Cowbum. 1997: Sawyer. 1998). but there is significant debate about how much specificity exists. Structurally, several key regions determine binding affinity and specificity. As discussed above, the invariant .Arg at PB5 forms a bidentate hydrogen bond interaction with the terminal phosphate and replacement of this residue even conservatively with a Lys. results in abolition of binding

(Mayer et al.. 1992: Sugimoto et al.. 1994: Grucza et al.. 1999). .Another key element that determines specificity is the residue at P05 which is largely predictive of the binding specificity of the residues C-terminal to the pY (+ positions) (Songyang et al.. 1993).

Several years ago SH2 domains were ordered into groups based largely on the residue at the POS position, which was determined to play a significant role in their binding specificity (Songyang et al., 1993). Group 1 SH2 domains bind sequences such as pY- hydrophilic-hydrophilic-hydrophobic and have a Tyr or Phe at the P05 position.

Conversely. Type III SH2 domains have He or Cys at the PBS position and select sequences of the general nature of hydrophobic-X-hydrophobic (Songyang et al.. 1993).

Additionally, residues beyond the +3 position provide little to no additional binding affinity to group I SH2 domains while residues beyond the +3 position do confer greater 5 binding affinity to many SH2 domains of group III (Lee el al.. 1994). Crystallographic evidence has demonstrated why, with the group 1 SH2 domains having a well defined -^3 pocket in which the hydrophobic +3 residue interacts and the group 111 SH2 domains having a less well defined pocket exhibiting what more typically would be categorized as a groove (Eck et al.. 1993; Lee et al.. 1994). The + 3 interactions are controlled largely by the BG and EF loops. In most of the group 1 SH2 domains, these two elements have features that lead to them forming a clamp of which the +3 residue is deeply ensnared.

Compared to the group 111 SH2 domains, the BG loop is shorter and contains a Gly that allows a type I’-tum that gives compactness to this region (a hydrogen bond forms between BGl and BG4) (Mulhem et al.. 1997). Additionally, an amide nitrogen in the

BG loop is allowed to form a hydrogen bond to the hydroxyl group on the P05 residue.

These features, absent in group 111 SH2 domains lead to a closed pocket in group 1 SH2 domains that will not accommodate residues beyond the +3 position and a less well defined region described as a groove accommodating residues beyond the plus three. The precise reason hydrophilic residues are selected at the -i-1 position in group 1 SH2 domains where as in group 111 SH2 domains hydrophobic residues are selected can not be readily determined by structural data. In the early experiments it was thought that SH2 domains did not have much preference or requirement for residues at the positions N- terminal to the pY (-positions) however there appear to be exceptions and this early idea would seem to have been an oversimplification for many SH2 domains (Burshtyn et ai.

1997). The SFI2 domains of SHP-1. SHP-2, and SAP have all exhibited significant preference and contribution to binding from residues at the N-terminal positions in biochemical studies and in work contained in this dissertation (Huyer ei al.. 1995;

Burshtyn et al.. 1997; Poy et al.. 1999).

1.1.3 Research objectives

SH2 domains have long been the target of inhibitors both for pharmacological reasons and cell biological reasons. However, progress to this end has been slow due to several reasons including apparent lack of specificity, lack of cell permeability to designed inhibitors, and dearth of specificity information for individual SH2 domains.

Cantley et al. went a long way in identifying the general trends in SH2 domain pY binding specificity but as discussed in the following sections the information excludes many SH2 domains and is limited to the three C-terminal residues (Songyang et al..

1993: Songyang et al.. 1994). Development of a more general method for specificity studies would greatly assist in systematically determining the specificity, of not only SH2 domains but. in all phosphopeptide binding modules.

To further the understanding of how SHP-1 may be regulated but more importantly develop a general methodology to study all pY-peptide binding modules we embarked on the development of a general methodology to study the binding specificity of the SH2 domains of SHP-1. As discussed above, information as to the specificity determinants of SH2 domains is reasonably well established but despite this, systematic studies of these critical pY-peptide binding modules have not been developed. Since

SH2-pY binding interactions perform critical roles in signal transduction and some are in pharmacological targets, the development of a general methodology to study the

7 specificity of these modules would greatly assist in both of the above pursuits. With the enormously important role that this enzyme plays in signal transduction as well as SH2

domains in general, the specificity of these domains would yield important information as

to what the in vivo function may be through the identification of potential partner

proteins. .‘Additionally, and more importantly, the development of a general methodology

to determine the binding specificity of all pY-binding modules would greatly facilitate

inhibitor design and defining their in vivo roles.

Other processes that assist in maintaining cellular function do not require

posttranslational modification as in the case o f phospho-dependent SH2 domain binding.

One of these processes is that of proteolysis for purposes of regulation and maintenance.

Cells utilize proteolysis to activate dormant zymogens, regulate the concentration of

effecter proteins, and for eliminating potentially harmful proteins. The housekeeping

function of eliminating potentially harmful proteins in bacteria is carried out by a number

of ATP dependant and independent proteases (Gottesman. 1996). The identification of

these proteases and subsequent studies has revealed unique means by which this

proteolysis occurs.

1.2 Tail-specific protease

In 1992 an Escherichia coli periplasmic protease was identified that degraded

recombinately produced proteins containing the C-terminal sequence of WVAAA and

was subsequently named Tail-specific protease (Silber et al.. 1992). It was demonstrated

that without this type of C-terminal sequence, Tsp exhibited negligible activity towards

protein or peptide substrates (Keiler et al.. 1995). Tsp function was further detailed when

8 it was demonstrated that Tsp was involved in proteolyzing proteins that had been tagged during translation with AANDENYALAA (Keiler et al., 1996). This co-translational tagging mechanism occurs in the presence of transcripts that fail to have a stop codon due to a transcriptional error or nuclease activity. Proteins produced from such transcripts become tagged with the above sequence and this sequence ultimately targets the protein for destruction. Since Tsp is a periplasmic protease, it most likely degrades only proteins that enter the periplasm with such C-terminal sequences, with a cytosolic counterpart degrading proteins remaining in the cytoplasm. Biochemical characterization of this enzyme revealed that it had a broad specificity as to the site(s) of cleavage in a protein substrate, with sites existing distally to the C-terminus at external loop regions and regions in the hydrophobic core (Keiler et ai.. 1995). Importantly Tsp has a stringent requirement for the C-terminal sequence, with amidation of or change of the most C- terminal residue to glutamate, abolishing activity (Keiler et al.. 1995). Although Tsp has a stringent requirement for the C-terminal residues it normally in most cases performs catalysis at quite distal locations and does not utilize the last 3-5 C-terminal residues for catalysis (Keiler et al.. 1995). The closest cleavage site to the C-terminus on peptide substrates was known to be some 13 amino acids away (Keiler et al.. 1995). These factors led to the proposal that Tsp may function using independently operating regions or domains with one serving to bind the protein substrate via its C-terminal non-polar tail and one providing the machinery for catalysis. Much of the work that was done in our lab was to refine the mechanism of Tsp and definitively demonstrate that the substrate binding site in Tsp is a PDZ domain. 1.2.1 PDZ domain function

PDZ domains were discovered later than SH2 domains (Cho et al.. 1992: Itoh et al.. 1993: Woods and Bryant. 1993: Ponting and Phillips, 1995). and are well established as protein-protein interaction domains in the formation of oligomeric protein structures

(Ranganathan and Ross. 1997; Craven and Bredt. 1998). Named after the first three proteins in which they were discovered: PSD-95. Drosophila discs-large tumor suppressor protein (Dig), and the mammalian tight-junction protein zona occludens-1

(ZO-1) (Woods and Bryant. 1991. Cho et al.. 1992: Itoh et al.. 1993). In contrast to SH2 domains, which bind pY containing peptides. PDZ domains are known to bind C-terminal peptides, typically nonpolar in nature (Gee et al.. 1998). First identified to bind S/T-X-V. but since been shown to bind many other types of sequences as well as internal P-hairpin structure (Komau et al.. 1995: Songyang et al.. 1997). PDZ domains mediate protein- protein interactions primarily by recognition of C-terminal peptides such as S/T-X-V-

COOH as in PSD-95 and the N-methyl-D-aspartate receptor (Komau et al.. 1995). This protein-protein recognition serves to formulate complex assemblies of proteins typically in terminal neuronal synapses. Regions homologous to known PDZ domains were later identified in bacterial proteases, such as HtrA and Tsp. and proposed to play a role in substrate recognition rather than protein protein interaction (Pallen and Wren. 1997:

Ponting. 1997).

The general roles of PDZ domains, particularly in eukaryotic organisms, have been determined largely to be that of a protein-protein interaction domain (Ranganathan and Ross. 1997; Craven and Bredt. 1998; Gamer et al.. 2000). The PDZ domain of one protein serves to bind a C-terminal peptide of another protein thus leading to a protein-

10 protein complex. This was classically identified in synapses and ion channels (Hata ei al„ 1996; Gamer el a i, 2000). During this time of PDZ domain discovery and characterization Sauer et al discovered an E. coli protease that was responsible for the degradation of recombinately produced proteins containing a non-polar C-terminal tail such as WVAAA (Parsell et al.. 1990; Silber et al.. 1992; Keiler et al.. 1995; Keiler and

Sauer, iWbi. Fhis protease was further implicated in the degradation of potentially aberrant proteins (Keiler et al.. 1996). Through biochemical analysis, a distinct peptide recognition site and a catalytic protease domain were proposed to modulate Tsp. .Among the evidence for this, was the fact that even though Tsp had an absolute requirement for a particular C-terminal sequence but yet in most analyses, this sequence was not among the sites of proteolysis (Keiler et al.. 1995). Despite the PDZ domains primarily being identified in eukaryotic proteins, sequence analysis of Tsp as well as other bacterial proteins revealed regions of homology to Tsp (Ponting. 1997).

1.2.2 PDZ domain structure

The PDZ is arranged as a P-sandwhich composed of six P-strands and two a- helices (Figure 1.2). The interaction between PDZ domain and peptide ligand is complimentary in that the peptide ligand augments the PDZ structure by providing an additional P-strand to the PDZ P-sheet. This P-sheet augmentation has been compared to that of p-clamping observed in viral capsid proteins (Doyle et al.. 1996). Specifically key interactions are in the highly conserved GLGF carboxylate binding loop in which the carboxylate from the ligand hydrogen bonds to the peptide amides o f the two glycines. and phenylalanine amides as well as through a water molecule to a conserved arginine

11 (Doyle et al.. 1996). The two glycine residues provide the flexibility required to allow

for the hydrogen bonding arrangement of the ligand. Crystal structure evidence suggests that the peptide has its strongest interaction at the C-terminal residue with the side chain of this amino acid projecting directly down into a hydrophobic pocket and the

interactions of the terminal carboxylate with the GLGF motif (Doyle et al.. 1996).

Based on known PDZ domain structural data and known tunctional roles in peptide

binding, combined with the biochemical data on Tsp. it was not difficult to envision a

PDZ or PDZ-like domain serving a functional role in Tsp.

1.2.3 Research Objectives

The ultimate goal of the Tsp project involved exploratory work based on

antibiotic design and protein engineering projects. In order to pursue these projects, a

sensitive convenient assay needed to be developed and basic biochemical characterization

had to be carried out. The short term goals of this project were to develop an assay and

characterize the essential features of this enzvme.

12 C-SH2 N-SH2

oB,

BG

oB C 104 D'E

Figure 1.1. Crystal structure of the SH2(N) and SH2(C) of SHP-2 complexed with phosphopeptide. Reproduced from Eck et al.. Nature 1996. /

[P»w»l Cm

p«oMeN4#mwu»

Figure 1.2. Illustrations based on crystal structure of PDZ domain of PSD-95 complexed with a substrate peptide (reproduced from Doyle et al.. Cell. 1996). .\. General topology of the PDZ domain complexed with a substrate peptide (checkerboard). B. Peptide KQTSV complexed with GLG loop in the PDZ domain.

14 CHAPTER 2

DEVELOPMENT OF A CONTINUOUS FLUOROMETRIC

ASSAY FOR TAIL-SPECIFIC PROTEASE

2.1 Introduction

Intracellular proteolysis occurs in all organisms, providing a variety of functions such as the elimination of abnormal proteins, the maintenance of amino acid pools during star\'ation. the generation of protein fragments that act as hormones, antigens, and other effectors, and the regulation of the intracellular concentrations of certain proteins. It has been recognized that the sequence identity at the N-terminal or C-terminal end of a protein serves as a determinant of proteolytic degradation (Bachmair et al.. 1986; Parsell et al.. 1990). The effect of the N-terminal residue on the metabolic stabilit}' of a protein. the N-end rule, has been relatively well studied (Varshavsky. 1996). The mechanism by which the C-terminal sequence of a protein determines its proteolytic susceptibility has only recently started to be elucidated. A major advance in this area was the identification and cloning of a tail-specific protease (Tsp) in Escherichia coli periplasm, which selectively degrades proteins with nonpolar C-terminal sequences (Silber et ai. 1992:

Keiler et al.. 1995; Keiler and Sauer, 1996). Tsp and its cytoplasmic counterpart (Silber and Sauer, 1994; Gottesman et al.. 1998) have also been implicated in the

15 elimination of aberrant polypeptides expressed from defective mRNAs using a peptide tagging system (Keiler et al.. 1996), although Tsp can only participate in a periplasmic housekeeping role.

Tsp represents a new class of protease. It has no sequence homology to any well characterized proteases and is not inhibited by several common protease inhibitors (Silber et al., 1992). Its substrate specificity suggests that it contains a separate binding site, for the hydrophobic tails of its substrate proteins, that is distinct from the catalytic site

(Keiler et al.. 1995). The molecular mechanism for substrate recognition and degradation has. however, not been well understood. Further biochemical characterization of Tsp has been hampered by the lack o f a convenient assay of Tsp activity. The existing assays monitored the degradation o f protein substrates by SDS-PAGE gels or the cleavage of short peptides by HPLC (Silber et al., 1992; Keiler et al., 1995: Keiler and Sauer. 1996).

These assays are relatively time-consuming, lack sensitivity, and are not well suited for quantitative analysis of Tsp activity. Detailed in this chapter is the report of a continuous, highly sensitive fluorimetric assay for Tsp. based on a quenched fluorogenic peptide. Tsp cleavage o f the peptide separates a fluorescent donor. 5-[(2- aminoethyl)amino]naphthaIene-l-sulfonic acid (EDANS), from a quenching acceptor. 4-

(4-dimethylaminophenylazo)benzoic acid (DABCYL). resulting in an increase in fluorescence yield. This assay is highly sensitive and convenient, and can be applied to various types of mechanistic investigations.

16 2.2 Experimental Procedures

2.2.1 Materials

All amino acids were purchased from SynPep Corp. (Dublin. CA). HMP resin

was from Applied Biosystems. EDANS and DABCYL were from Sigma Chemical

Company (St. Louis, MO). N-Fmoc-6-aminocaproic acid was purchased from Penninsula

Laboratories (Belmont. CA). All other chemicals were from Aldrich Chemical Company

(Milwaukee. WI).

2.2.2 Purification of Tail Specific Protease

Overexpression of Tsp in E. coli was carried out as described by Keiler et al.

(Keiler et a i. 1995). £ coli X90 cells harboring plasmid pKS6-lw were kindly provided

by Dr. Robert Sauer (Department of Biology. Massachusetts Institute of Technology).

Typically. 6-10 liters of cells were grown at 37 °C in LB media containing 75 mg/liter

ampicillin to an ODeoo of 0.7 - 0.8. and induced by the addition of isopropyl P-D-

thiogalactopyranoside (IPTG) to a final concentration of 300 pM. The temperature was

then decreased to 28 “C and cells were harvested 12-16 hours later. Cells were harvested

and lysed using a procedure described previously (Rajagopalan et al.. 1997) by addition

of lysozyme followed by sonication in lysis buffer A (50 mM Hepes. pH 7.4. 1% triton

X-100. 0.5 mM 1.10-phenanthroline. 20 mM NaCl. and 5 mM EDTA). After the lysis

suspension was cleared by centrifugation at 15.000 rpm (SS-34 rotor), the supernatant

was decanted into a clean container, diluted 1.5-2 fold in buffer B (30 mM Hepes. pH

7.4. 10 mM NaCl, 1 mM EDTA), and loaded onto an SP-Sepharose column (8 x 2.5 cm)

17 equilibrated with buffer B. Bound protein was eluted with a linear gradient of 10 mM - 1

M NaCl at 2.0 ml/min (total 400 ml). Fractions containing highest Tsp activity were concentrated using a Centriprep-30 concentrator. The concentrated protein sample was desalted on a Pharmacia fast desalting column and loaded onto a Q-Sepharose column

(7.5 X 2.5 cm), which had been preequilibrated with buffer C (20 mM Tris-HCl. pH 8.0. 5 mM NaCl. ImM ED 1 A). .After washing with 1-2 column volumes of buffer C. the protein was eluted with a 5-100 mM NaCl linear gradient (200 ml at 4 ml/min). Fractions containing active Tsp were pooled and were adjusted to 1.5 M with (NH^lzSO^ by slowly adding solid (NlT^liSO; while stirring at 4 ^ C. The resulting solution was then loaded onto a Pharmacia phenyl-Sepharose column (HiLoad 16/10) and Tsp was eluted using a reverse linear gradient of 1.7-0 M (NFl 4)2S0 4 in 20 mM phosphate buffer. pH 7.4 (total

300 ml at 4.0 ml/min). Fractions with greater than 95% purity (as judged by SDS-P.AGE.

Figure 2.1) were pooled, concentrated using a Centriprep-30 concentrator, and quickly frozen in liquid nitrogen after the addition of 30% glycerol. Typical yield was 1.5 mg Tsp per liter of cells. Aliquots of the protein were stored at -80 “^C until use.

2.2.3 Analytical Methods

Protein concentration was determined by Bradford assay (Bradford. 1976). using bovine serum albumin as the standard. Substrate concentrations were spectrophotometrically determined by using the experimentally derived extinction coefficient of DABCYL-Glycine in 20% DMF ( e = 13.850 M'‘ cm'') at 468 nm.

Absorbance readings to determine substrate concentrations were always taken in 20%

1 8 DMF. consistent with the solvent composition used during the determination of the extinction coefficient. Mass spectrometric analysis was performed at the Biological Mass

Spectrometry Facility of The Ohio State University.

2.2.4 Substrate and Inhibitor Preparation

0-Ethoxycarbonyi DABCYL.Ethyl chloroformate (55 mg, U.5 mmol) was added to a suspension of sodium 4-(4-dimethylaminophenylazo)benzoate (146 mg. 0.5 mmol) in 10 ml of methylene chloride. The mixture was vigorously stirred at room temperature for 30 min and then filtered to remove the white solid formed. Evaporation in vacuo afforded 145 mg of a red solid (85% yield): 'H NMR (300 MHz. CDCl;): 6

8.17 (d. J = 8.7 Hz. 2H). 7.91 (d. J = 9.2 Hz. 2H). 7.88 (d. J = 8.7 Hz. 2H). 6.77 (d. J =

9.2 Hz. 2H). 4.42 (quartet. J = 7.2 Hz. 2H). 3.12 (s. 6H). 1.00 (t. J = 7.2 Hz. 3H).

2-N-[(l-Sulfono-5-naphthalene)aminoethyllamino-succinyl-.AARAAKO-(6- aminocaproyOa-ENYALAA. Peptide synthesis was carried out with standard Fmoc chemistry on HMP resin (0.2 mmol scale) on a home-made manual peptide synthesis apparatus. Coupling of each amino acid was monitored with ninhydrin and if judged to be incomplete, was repeated once or twice until complete coupling was achieved. After the last amino acid (alanine) was added onto the resin, the peptide was treated with 20% piperidine in DMF to remove the N-terminal Fmoc group. Succinic anhydride (150 mg.

1.5 mmol) and dimethylaminopyridine (12.2 mg) in 2 ml of dichloromethane were added to the resin and the reaction was allowed to proceed for 2 hours at room temperature. Ax. this point, ninhydrin test showed complete coupling. The resin was exhaustively washed with dichloromethane and DMF and then treated with a DMF solution containing

19 EDANS (200 mg, 0.75 mmol), HOST (100 mg, 0.74 mmol), and HBTU (160 mg. 0.42 mmol) for 5 hr with mixing. The coupling reaction was repeated once to insure complete derivatization by EDANS. The resin was exhaustively washed with DMF to remove any unreacted EDANS. Cleavage and deprotection were effected with a cocktail containing

4.5 ml trifluoroacetic acid. 0.15 ml ethanedithiol. 0.1 ml anisole. 375 mg crystalline phenol, and 0.25 ml thioanisole and gentle agitation at room temperature for 1 hour. The solvents were evaporated with a gentle flow of nitrogen and the resulting semisolid was triturated with diethyl ether (3 x 20 ml). The crude peptide was utilized directly in the solution phase coupling to DABCYL.

2-N-[(l-SuIfono-5-naphthaiene)aminoethyl]amino-succinyi-AARAA-[N‘^-(4-

(4-dimethylaminophenyla2o)benzoyl)Iysyll-(6-aminocaproyl)2-ENYALAA. The crude peptide (5 mg) from above was dissolved in 0.3 ml of a DMF/water/triethylamine mixture (10:1.6:1 v/v) and 0-ethoxycarbonyl DABCYL (2.5 mg. 3 equivalents) was added. The reaction was incubated at room temperature for 10-20 minutes and the derivatized peptide was purified on the reversed-phase HPLC. The peak corresponding to the desired product was collected and lyophilized to an amorphous powder. Electrospray ionization mass spectrometric analysis of the peptide produced a molecular mass of

2146.0 Da (calculated MW: 2144.68). A control peptide with a glutamate residue instead of an alanyl residue at the C-terminus was synthesized in the same manner and produced a molecular mass of 2204.0 Da (calculated MW: 2202.68).

20 GRGYALAA. This peptide was synthesized using standard solid-phase Fmoc chemistry and purified by reversed-phase HPLC. Matrix assisted laser desorption ionization mass spectrometric (MALDI-TOF) analysis gave a peak at m/z 779.4 (M+H)

(calculated MW: 778.48).

2.2.5 Tsp Assay

.All assays were performed at room temperature. Stock solutions of peptide substrates were prepared in anhydrous DMF typically to a concentration of 3-5 mM and stored at -80‘’C. Prior to use, they were diluted to a working stock of 500 pM in water.

DMF concentration in the assay reaction was kept at <0.5%. a concentration that had no detectable effect on enzyme activity. Assay reactions were carried out on a Perkin-Elmer

LS-5 Fluorescence Spectrophotometer at an excitation wavelength of 340 nm and emission wavelength of 520 nm. The slit width was set at 10 nm for both excitation and emission bands. .A total reaction volume of 1.5 ml was used in 50 mM Tris-HCl and 250

mM NaCl at pH 8.0. Substrate concentrations were typically 0-20 pM and 5.0 pg of enzyme was added to initiate the reaction. The increase in fluorescent yield was

monitored over time (typically 3-5 minutes). Substrate conversion to product was always

kept below 15%. A standard curve was developed by hydrolyzing known amounts of

peptide substrate to completion and then measuring the fluorescence yield. Using this

standard curve, the arbitrary fluorescent units could be used to quantify the amount of

product generated during the time course of the assay. As substrate concentrations were

increased (>10 pM), a small amount of quenching was observed. This quenching effect

was corrected by measuring the fluorescent yields of solutions that contained a fixed

21 concentration of the fluorescent product but that contained increasing concentrations of the added substrate. A mild correction to the signal could be readily employed by utilizing this quenching standard.

2.2.6 HPLC and Mass Spectrometric Analysis of Tsp Products

To determine the site(s) of Tsp cleavage. Tsp (5.6 pg) was incubated with 15 pM substrate under the assay conditions described above. At various time points (0-60 min).

300-pl aliquots of the reaction were withdrawn and injected into an HPLC equipped w ith a Ci8 column. A linear gradient of 0-50% acetonitrile (with 0.05 % trifluoroacetic acid) in water was used to separate the mixture. Peaks corresponding to the products were collected, lyophilized, and analyzed by electrospray mass spectrometry.

2.2.7 Inhibition of Tsp.

Assays were performed as described above, except for the presence of 0-150 pM inhibitor peptide (GRGYALAA). The inhibitor peptide was shown to have no effect on the fluorescent yield of either the substrate or the cleavage product. HPLC analysis was also performed to ensure that the inhibitor peptide was not a substrate of Tsp. This was carried out by duplicating the assay reactions and taking portions of the reactions for reverse-phased HPLC analysis.

1 9 2.3 Results

2.3.1 Design and Synthesis of Tsp Substrates

Tail specific protease selectively degrades proteins that contain a hydrophobic C- terminus (Silber e( al.. 1992; Keiler et al.. 1995: Keiler and Sauer. 1996: Keiler et al..

1996). Together with its cytoplasmic counterpart. Tsp also assists in the elimination of potentially toxic peptides translated from damaged mRNAs. which have been tagged with a hydrophobic sequence. AANDENYALAA. at their C-termini (Keiler et al.. 1996).

Small peptides containing nonpolar C-terminal sequences have also been shown to be excellent substrates of Tsp (Keiler et al.. 1995). It appears that Tsp possesses a recognition domain that binds to the hydrophobic C-terminus of a potential substrate and then engages its catalytic site, which is presumably distinct from the recognition site, to a remote location on the substrate for peptide bond cleavage (Keiler et al.. 1995). Previous experiments seem to suggest that while Tsp is highly selective for the nature of the C- terminal sequence of a substrate, it has rather broad specificity for the cleavage site

(Keiler er a/.. 1995).

Based on the above observations, we have designed peptide 1 as a potential Tsp substrate (Figure 2.2). Peptide 1 contains a hydrophobic C-terminal tail. ENY.A.L.AA. derived from sequence information of the in vivo tagging system (Keiler et al.. 1996).

This sequence should serve as the recognition signal for the Tsp enzyme. The N-terminus

contains AARAAK for the folio wring reasons. First, despite the lack of selectivity for its

active site. Tsp does have some preference for cleaving between Ala-Arg residues (Keiler

et al., 1995). Second, the presence o f an arginine residue significantly increases the solubility of this otherwise hydrophobic peptide. Third, inclusion of a lysine residue allows for the attachment of the fluorescent quencher. DABCYL. to its e-amino group.

The fluorescent donor, EDANS, is attached to the N-terminus via a succinyl linker. The spatial proximity of the DABCYL and EDANS groups would result in efficient quenching of any fluorescent emission by the EDANS group (Matayoshi ei al.. 1990).

However. Tsp cleavage in the AARAAK region would separate the donor/quencher pair, causing an increase in fluorescence emission, which can easily be monitored on a spectrofluorimeter and provide a sensitive readout for the Tsp reaction. An unnatural amino acid. 6-aminocaproic acid, was used as a linker to connect the recognition and cleavage sites in order to minimize any cleavage in the linker region, as cleavage in this region would not change the fluorescent yield and thus complicate the quantitative analysis. Additionally since most cleavage sites in known peptide substrates are typically at least 13 amino acids from the C-terminus. it was thought that sufficient distance between the cleavage and recruitment sites may be required. Previous experiments indicate that the closest cleavage site on peptide substrates is some 13 amino acid residues away from the C-terminus (Keiler et al.. 1995). Inclusion of two 6-aminocaproyl units should direct Tsp cleavage to the AARAAK region. A final design consideration was to maximize the synthetic steps that can be carried out on the solid phase, which greatly simplifies product isolation. As discussed in Materials and Methods, peptide 1 was synthesized almost entirely on the HMP resin, except for the last step, the addition of the DABCYL group.

24 2.3.2 Tsp Assays

As expected, peptide 1 gives little fluorescence emission when excited at 340 nm and incubation in the absence of Tsp under the assay conditions did not increase its fluorescence emission over extended periods of time. When active Tsp was added to the reaction, however, a rapid, linear increase in fluorescence emission at 520 nm with time was observed (Figure 2.3). Extended incubation resulted in complete cleavage of the substrate and a >50-fold increase in the fluorescent yield. This increase is not observed upon Tsp addition to peptide 2. demonstrating the Tsp specificity of our substrate system.

.As a positive control, peptide 1 was incubated with a small amount of trypsin, which selectively cleaves the peptide after the arginyl residue, the same increase in fluorescence emission was observed. Additionally, the stability o f the substrate is quite good as no increase in fluorescence is observed across >30 minutes. This demonstrates that peptide

1 is indeed a substrate of Tsp and the strategy described above is viable for monitoring

Tsp activity. It should also be noted that we first tested a similarly designed and synthesized substrate that was shorter than that detailed here (Y.ALAA vs ENYALAA).

Tsp exhibited much less activity towards this shorter version of the substrate and it was hypothesized that, based on previously characterized peptide substrates, that the cleavage site may have been too close to the C-terminal recognition element. The effect of having these two elements too close together is presumably that of efficient binding to the putative C-terminal peptide binding site while spatially segregating the cleavage site from a distant catalytic site.

25 A typical assay reaction is carried out by the addition of 5 pg of Tsp to 1.5 ml of reaction mixture containing 0-20 pM substrate. The increase in fluorescence emission is monitored on a spectrofluorimeter. with excitation and emission wavelengths set at 340 nm and 520 nm. respectively. Under these conditions, fluorescence intensity increases linearly with time for a period as long as 30 min. Initial reaction rate is calculated from the slope of the linear line within the first 3-5 min (Figure 2.4A). At this point, substrate to product conversion is usually -15%. Comparison with a standard curve generated with known concentrations of cleavage products converts the initial rates in fluorescence intensity to product concentration.

Tsp exhibits Michaelis-Menten kinetics toward peptide 1 (figure 2.4B). The catalytic constants were determined by varying the substrate concentrations while keeping the enzyme concentration constant. .A plot of initial rates for product formation against substrate concentration and data fitting to the Michaelis-Menten equation produce a k,at of 0.086 ± 0.002 s'*. Km of 4.0 ± 0.3 pM. and a k«a/KM of 2.2 x lO^* M 's '. As a comparison, the best substrate reported by Keiler et al. has a kcat of 3.7 s'', a Km of 35 pM. and a k^ai/Kw o f 1.1 x 10^ M ''s ' (Keiler et a i. 1995). This suggests that peptide 1 is a fairly potent substrate of Tsp.

It has been determined that Tsp has a very high specificity towards the sequence at the C-terminus of a peptide or protein substrate, especially requiring a small nonpolar residue as the last residue (Keiler et al.. 1995). Upon prior analysis of Tsp. substitution of the C-terminal Ala to Glu resulted in a dramatic reduction in activity (Keiler et a i. 1995).

To ensure that the activity observed is due to Tsp reaction, a control substrate (peptide 2) was prepared with a C-terminal glutamate instead of alanine. As expected, this simple

26 substitution at the C-terminus rendered the peptide (2) a much poorer substrate. No saturation was observed up to 20 |iM peptide and the kcat/K,vi value is at least 380 fold lower than that of peptide 1 (kcat/K.M = 58 M‘‘s‘').

2.3.3 Analysis of Tsp Cleavage Products

Since Tsp is known to be ratlier non-specific and the presence of multiple cleavage sites would complicate data analysis due to nonproductive cleavages or product inhibition, the products were monitored to determine the site(s) of cleavage. Peptide 1 was incubated with Tsp under typical assay conditions as used above and the reaction mixtures were analyzed at various time points by HPLC (Figure 2.5). It can be seen that under the typical assay conditions. Tsp cleavage produces two major product peaks with retention times of 13.3 and 23.6 min. respectively. With incubation times of up to 60 min. no other product was generated at significant levels. Longer incubation times (>6 hours) did yield additional cleavage species (data not shown). This Indicates a single site of cleavage on and beyond the time scale used in the assay, which was typically monitored over the first three minutes. The peak with retention time o f 13.3 min was collected and analyzed by electrospray ionization mass spectrometry and was found to have a molecular mass of 508.3 Da. This species is apparently the N-terminal fragment as it has strong fluorescence emission at 520 nm. The molecular weight is consistent with the sequence of EDANS-Ala-Ala (calculated MW: 508.45). In contrast to the highly fluorescent EDANS peak, the second fragment (retention time of 23.6 min) had no fluorescence emission but was intensely red in color, consistent with being the C-terminal

27 fragment with the attached DABCYL group, but quality mass spectral data could not be achieved with this peak. Therefore, we conclude that Tsp resulted in a single cleavage between the second alanyl and the arginyl residues.

2.3.4 Inhibition ofTsp

If binding to a hydrophobic C-terminus is a prerequisite for cleavage, as the data suggest, a peptide that contains a hydrophobic C-terminal sequence but lacking a cleavage site would be expected to act as a competitive inhibitor of Tsp. To test this notion and to demonstrate the utility of our assay in inhibition analysis, we synthesized peptide GRGYALAA. This peptide is indeed an inhibitor of Tsp. Lineweaver-Burk plots

( 1/V vs 1/[S]) at different inhibitor concentrations produce a series of straight lines with a common intercept on the y-axis. indicating that the observed inhibition is competitive in nature (Figure 2.6A). Data fitting to a competitive inhibition rate equation gives an inhibition constant (Ki) of 31 pM (Figure 2.6B). The pattern of competitive inhibition suggests that Tsp is being successfully occupied by the peptide, and HPLC analysis of the peptide after incubation with Tsp showed no sign of cleavage. Together with the finding that peptide 2 is a poor substrate, this result indicates that proper binding of Tsp to the C- terminus of a substrate is essential for efficient cleavage by its catalytic domain.

However, the fact that this Ki value is 7-fold higher than the Km value obtained for peptide 1 suggests that the length or type of sequence N-terminal to the hydrophobic C- terminus could also be important for optimal binding and cleavage to Tsp.

2 8 2.4 Conclusions

A continuous fluorimetric assay has been developed for tail specific protease.

This assay is highly sensitive and capable of detecting submicromolar concentrations of products. It will be particularly useful in detecting minute amounts of Tsp activity or Tsp mutants that have attenuated catalytic activities. With peptide 2 as a control, this assay can also be applied to quantitate fsp activity in a crude cell lysate which may contain other proteases. While Tsp is 380-fold more active to peptide 1 than peptide 2. contaminating endoproteases should have similar activities toward both peptides. Thus, a subtraction of signal for 2 from that of I would give the signal for Tsp. This assay is highly reproducible and convenient. Each assay reaction can be carried out in 3-5 min. a feature useful for screening a large number of effectors against Tsp. Finally, this assay is a direct assay, which does not have any complications derived from the use of coupling enzymes or other reagents. The only point that may require a user's caution is that any reagents added to the assay must be checked for their effect on the fluorescent yield of the EDANS group.

29 Figure 2.1. 8% SDS-PAGE gel of fractions from different stages of Tsp purification. Lanes 1-3. SP-sepharose fractions. Lanes 4-6. Q-Sepharose fractions. Lanes 7-9. Phenyl-Sepharose fractions.

30 u

HjN-A-A-R-A-A-K-(aminocaproyl)j-E-N-Y-A-L-A-A"^o"^®®'" Sucanic anhydnde. DMAP 0 W no^^^/'v„^^^^,J^^A-A-R-A-A-K-(aminocaproyl)2"E”N-Y-A-L-A-A Q^Rssin I T Pi ,NK HOBT/HBTU L j L J DMF, RT

SOjH

HN' -A-A-R-A-A-K-(aminocaproyl), I - " r ^ t

TFA

SO,H 0 V HN '^'||-''^v.,^pj--A-A-R-A-A-l<-(amjnocaproyl)2-E-N-Y-A-L-A-A^OH

SO,H 0 0 NH,

0 V

HN' '^Vj^.^'^v^^^^A-A-R-A-A-K-(aminocaproyl)2-E-N-Y-A-L-A-Ar'^OH

SOjH

Figure 2.2. Synthesis of Tsp substrate (peptide 1). All steps were accomplished on solid phase with the exception of the addition of the DABCYL group during the last step. Peptide 2 is identical to peptide 1 with the exception of having a C-terminal Glu rather than Ala.

31 240

200

Q) • OU C <%) o 120 (U u(O 2 80 o 3 40

0

0 150 300 450 600 Time, seconds

Figure 2.3. Analysis of reaction course of peptides 1 and 2 upon addition of Tsp. Solid line. Tsp plus peptide 1; dotted line, Tsp plus peptide 2; dashed line peptide 1 in absence of Tsp. 1.5 uM

T im e, m m B

0.0504

I 0.0378 O) I 0.0252 3.

0.0126

0 • 0 8 12 16 [S], mM

Figure 2.4. (A) Initial reaction rates upon addition of 5 pg Tsp to increasing substrate concentrations. (B) Plot of Tsp reaction rates obtained from A vs. substrate concentration. Closed circles. Tsp plus peptide I; closed squares. Tsp plus peptide 2.

33 Ill

Retention Time, min

Figure 2.5 HPLC analysis of Tsp digest of peptide I after various times of incubation (monitored at 214 nm). (a). 0 minutes; (b), 10 minutes; and (c). 60 minutes. Peak 1. EDANS-Ala-Ala; peak II. the DABCYL-containing fragment; peak III. peptide 1.

34 100

80

60 W

40

0 0.08 0.16 0.24-0.08 1 /[S ]

28

21

Km’ 14

7

0

0 50 100 150

[I], ^iM

Figure 2.6. (A) Inhibition of Tsp by peptide GRGYALAA. Data presented are the Lineweaver-Burk plots of initial rates vs substrate concentration at different inhibitor concentrations. Solid circles, 25 pM inhibitor; triangles, 50 pM; squares: 75 pM; and diamonds. 150 uM. (B) Secondary plot of apparent K.m derived from the double reciprocal plot against [1].

35 CHAPTER 3

SUBSTRATE RECOGNITION THROUGH A PDZ DOMAIN IN TAIL-SPECIFIC PROTEASE

3.1 Introduction

Tsp represents a new class of protease that was first Identified as an activity that degrades a À repressor variant with a nonpolar C-terminal sequence. WV.AAA.. but not the wild-type repressor which has a polar C-terminal sequence. RSEYE (Silber ei al..

1992). Upon cloning of the tsp gene and subsequent analysis of its sequence it was apparent that Tsp had no sequence homology to any of the well characterized members of other protease families and was not inhibited by some common protease inhibitors (Silber et ai. 1992). Site-directed mutagenesis study has established that serine 430 and lysine

455. both in the C-terminal portion of the protein, are essential for catalytic activity

(Keiler and Sauer. 1995). Based on this observation, it has been proposed that Tsp belongs to the family of proteases that use a serine-lysine dyad for catalysis; the members of this family include LexA. type 1 signal peptidase, and the class A (3-lactamases

(Paetzel and Dalbey, 1997). The unusual substrate specificity of Tsp suggests that it may contain a separate binding site for the hydrophobic tails of its substrate proteins that is distinct firom the catalytic site (Keiler et al., 1995; Beebe and Pei. 1998). Amino acid sequence analysis has revealed an internal region with some sequence similarity to

36 known PDZ domains (Ponting. 1997). PDZ domains are -100 aa modules initially found in a number of eukaryotic cell-junction-associated proteins, where they serve to localize ion channel proteins within a particular region of the membrane by binding to the C- termini of the channel proteins (Cho et al.. 1992; Itoh et al., 1993; Woods and Brvani.

1993; Ponting and Phillips. 1995). PDZ domains were subsequently found in many other proteins such as phosphatases and kinases (Sato et al.. 1995; Hata et ai. 1996; Komau et al.. 1997; Ponting, 1997). Crystal structural studies of several PDZ domains have revealed that the C-termini of their partner proteins bearing specific sequences bind to a surface cleft in the PDZ domain, with the C-terminal carboxylate group bound by the highly conserved GLGF motif of the PDZ domain through hydrogen bonds to the backbone amides (Doyle et al.. 1996; Morais Cabral et al.. 1996; Schultz et al.. 1998). In addition to the carboxylate group, the C-terminal three or four residues are also involved in intricate interactions with the PDZ domain (Kim et a i. 1995; Komau et al.. 1995:

Schepens et al.. 1997; Songyang et al.. 1997; Strieker et al.. 1997). The substrate specificity of Tsp is consistent with having a PDZ domain as the separate substrate binding site, recognizing proteins with the proper C-terminal sequences for degradation.

In this work, we show that Tsp indeed contains a PDZ domain as the substrate recognition domain through a combination of partial proteolysis, photoaffinity labeling, and site-directed mutagenesis studies. This work continues to reveal how this unique protease performs its function and will aid in future protein engineering work by its description of the general architecture of Tsp.

37 3,2 Experimental Procedures

3.2.1 Materials

All amino acids, Wang resin, and other peptide synthesis reagents were purchased from Advanced ChemTech (Louisville, K.Y). V8 protease, 5-[(2-aminoethyl) amino] napthalene-1-sulfonic acid (EDANS), and biotin were from Sigma Chemical Company.

Fmoc-p-benzoyl-L-phenylalanine (Bpa) was purchased from Peninsula Laboratories.

Oligonucleotide primers were ordered from Integrated DN.A Technologies (Coralville. lA).

3.2.1Synthesis of Photoaffinity Labels

Peptides were synthesized on Wang resin on 0.2-mmol scale using a home-made peptide synthesis apparatus and fluorenylmethoxycarbonyl (Fmoc)-protected amino acids

[28]. Three equivalents of amino acids were used during coupling reactions and double coupling was performed if necessary (as judged by ninhydrin test). Coupling of the biotin moiety was accomplished by the addition of 3 equivalents of biotin, .\- hydroxybenzotriazole monohydrate (HOBt), O-benzotriazole-.V.A'.V'.A'- tetramethyluronium hexafluorophosphate (FIBTU), and Wmethylmorpholine. The

ED ANS group was added as previously described in chapter 2. Cleavage and deprotection of peptides were performed for 1 h at room temperature in a cocktail containing 5 mL of trifluoroacetic acid (TEA). 0.13 mL ethanedithiol. 0.25 mL thioanisole, 0.10 mL of anisole, 0.375 g crystalline phenol and 0.25 mL of water. The solvent was evaporated and the crude peptides were triturated with diethyl ether (3x15

38 mL). The crude peptides were purified by reversed-phase HPLC (C-18 column) and analyzed by matrix-assisted laser desorption ionization (MALDI) mass spectrometr>'.

Peptide 1 (Figure 3.1): calculated MW 1377.0. found 1379.3; peptide 2: calculated MW

1453.1. found 1455.4 (Figure 3.1).

3.2.3 Construction of Tsp Mutants

Site-directed mutants were constructed using the Quick-Change mutagenesis method (Stratagene). Plasmid pKKlOl (Keiler el al.. 1995). which contains the entire isp gene plus six histidine residues at the C-terminus, was used as the template. The primers used were 5*-GGAAGGTATTGGCGCCGAGCTGCAAATGGATG-.A.TG-3' (primer I) and 5 -GGAAGGTATTGGCGCACAGCTGCAAATGGATGATG-3' (primer 2). and their complements for the V229E and V229Q mutations, respectively. .A, new PvuU restriction site was introduced into the plasmid DNA by the V229Q mutation, whereas an

.Marl site was created by the V229E mutation, providing convenient markers for screening for mutants.

Deletion mutants were generated by the polymerase chain reaction (PCR) using plasmid pKKlOl (Keiler et al., 1995) as template. The following primers were used to construct the various mutants: primer 3. 5 ' -GGG AATT CTT ACTT G AC GGGAGCGG-

GTTG-3': primer 4. 5'-GGAAGGCCATATGAGCTATCTTTCCCCGC- GT-3': primer 5. 5 -GGGAATTCTTAACGGGTCTTGGTCCCT-3'; primer 6. 5 -CATGCC.AT-

GGCATATGAGCGAAGATGTTTTC-TCGCT-3' : primer 7. 5 -GGGAATTCTTATT-

TCTCTTTACCGACGGTCTT-3' ; and primer 8, 5’-CATGCCATGGCATATGGTAGA-

AGATATCACGCGTG-3’). Primers 3 and 4 were used to construct the N-terminal

39 deletion mutant, ANT Tsp (amino acids 206-660). The PCR product was digested with restriction endonucleases Ndel and EcoRI and inserted into the corresponding sites of plasmid pET22b (Novagen). The resulting plasmid was digested with Ndel and &ol. and the Tsp coding fragment was ligated to the Ndel-Scal linearized plasmid pET15b

(Novagen). This procedure added an N-terminal histidine tag to ANT Tsp. Primers 6 and

7 were employed to isolate the PDZ domain. PDZ-a (amino acids 185-334). Primers 5 and 6 were used to construct PDZ-b (amino acids 185-307). Primers 4 and 7 were used to construct PDZ-c (amino acids 206-334). Primers 4 and 5 were used to construct PDZ-d

(amino acids 206-307). Primers 7 and 8 were used to construct the C-terminal truncation mutant. ACT Tsp (amino acid 1-334). The PCR products from all above reactions were digested with Ndel and EcoRl and ligated with NdeUEcoKl linearized plasmid pET-28a

(Novagen). This procedure resulted in the addition of a six-histidine tag to the N-terminus of each construct. Amino acid numbering of all mutants used in this study and positions referred to are based on that used in previous work (Keiler and Sauer. 1995) (Figure 3.2).

3.2.4 Purification of Wild-type and Mutant Tsp

Escherichia coli BL21 (DE3) cells carrying the appropriate expression plasmid were grown in 3 L of LB media containing 75 mg/L ampicillin at 37 '’C. When OD 600 reached 0.7. the cells were induced by the addition of isopropyl p-D- thiogalactopyranoside to a final concentration of 140 pM and grown for an additional 3-6 hours at 28 “C. The cells were harvested by centrifugation at 4 "C (6000 rpm for 10 min). resuspended in 60 mL of ice cold column binding buffer (5 mM imidazole. 0.5 M NaCl.

20 mM Tris-HCl, pH 7.9), and lysed by sonication. The crude cell lysate was centrifuged 40 for 15 min at 17.000 rpm (Sorvall SS-34 rotor) and the clear supernatant (65 mL) was loaded onto a 10-mL His'Bind column (Novagen) (Hichuli et al.. 1987) that had been equilibrated with the binding buffer. The bound protein was eluted with 60 mL of an elution buffer (1 M imidazole, 0.5 M NaCl. 20 mM Tris-HCL, pH 7.9) and the fractions containing homogeneous Tsp were pooled. After being concentrated in a Centriprep 10 concentrator (Amicon), the protein sample was mixed with glycerol (J3% t'lnal) and stored at -80“C. Protein concentrations were determined by the Bradford assay using bovine serum albumin as standard (Bradford. 1976). Wild-type Tsp (without histidine tag) was purified as previously described in chapter 2.

3.2.5 Tsp Assay

All assays were performed using 2-W-[( 1 -sulfbno-5-napthalenelaminoethyI] amino- succinyl-.Ala-Ala-Arg-Ala-Ala-[/V^-(4-(4-dimethylaminophenylazo)benzoyl)lysyl]-( 6- aminocaproyl) 2-Glu-Asn-Tyr-Ala-Leu-Ala-Ala as substrate (Beebe and Pei. 1998). .As detailed in chapter 2. cleavage of the substrate at the Ala-Arg peptide bond by Tsp causes an increase in fluorescence yield of the EDANS group, which was monitored on a

Perkin-Elmer LS-5 spectrofluorometer. A typical reaction (1.5 mL) contained 50 mM

Tris-HCl. pH 8.0. 250 mM NaCl. 0-20 pM substrate, and 5.0 pg of Tsp enzyme.

3.2.6 Photoafnnity Labeling

All labeling reactions were carried out at 4 "C in a Rayonet photochemical reactor equipped with two RPR-3500 lamps (350 nm) (Southern New England Ultraviolet Co.).

A reaction mixture (total volume of 30 pL) containing 25 mM Tris*HCl (pH 8.0). 100

41 mM NaCl. wild-type or mutant Tsp protein (typically 50 pg), and 100 pM peptide 1 was irradiated for 20 min under the 350 nm lamps and then analyzed by electrophoresis on

SDS-PAGE gels. The labeled protein bands in the gel were visualized on a Fisher

Biotech 302 nm transluminator (Model FBTI 614) and photographed using a Polaroid camera. In competition experiments, a reaction mixture containing 25 mM Tris'FlCl (pH

8.0). 100 mM NaCl. 50 pg of Tsp. 100 pM fluorescently labeled peptide 1. and 0-1200 pM unlabeled competing peptide (peptide 2) was incubated under the 350 nm lamps for

10 min at 4 '^C. followed by SDS-PAGE analysis.

To identify the site of photolabeling. PDZ-a domain (3 mg. 187 nmoles of PDZ domain) was irradiated in the presence of 40 pM peptide 1 for 3 h at 4 "C (total volume of 10 mL). To remove the excess photoaffinity labels, the sample was loaded onto a

His'Bind metal affinity column (Novagen). After extensive washing, the bound protein was eluted with 20 mM Tris*HCl. pH 8.0. 500 mM NaCl. and 1 M imidazole and exhaustively dialyzed against 100 mM ammonium bicarbonate buffer (pH 7.9). The resulting protein was digested with 1:25 (w/w) trypsin at 37 °C for 3 days. The peptides generated were separated on a reversed-phase HPLC equipped with an analytical C|* column. A linear gradient of 0-50% acetonitrile in water plus 0.05% trifluoroacetic acid was carried out in 1 h at a flow rate of 1.0 mL/min. Individual peaks were collected, lyophilized to dryness, and analyzed on a Perkin-Elmer electrospray ionization mass spectrometer (model API-300) in the positive ion mode.

42 3.2.7 V8 Proteolysis of Tsp

Tsp (6 jig) was digested with 0.3 pg of V8 protease in 0.1 M Tris-HCl. pH 7.4 for varying length of time (0-60 min) at room temperature (total reaction volume of 12 pL). The reactions were quenched by the addition of 12 pL of 2x SDS-PAGE gel loading buffer and heating for 10 minutes at 95 '^C and analyzed on a 10% SDS-PAGE gel. For mass spectrometric analysis, the reaction was similarly carried out (60 pg Tsp in a 120- pL reaction) and quenched by the addition of TEA to 0.5% final concentration and stored at -80 '^C until use. The resulting sample was passed through a C-8 reversed-phase HPLC column (eluted with acetonitrile plus 0.1% trifluoroacetic acid) before being injected into a Perkin-Elmer electrospray ionization mass spectrometer (model .API-300). Further confirmation of the site of cleavage was achieved by digestion of a Tsp variant containing a C-terminal 6-histidine tag (Keiler et al., 1995). dénaturation with 6 M urea, and passing the sample through a Ni’"^ metal affinity column (Novagen) (Hichuli ei al.. 1987). .After washing with 6 column volumes of the binding buffer (20 mM Tris'HCl. pH 7.9. 500 mM NaCl. 5 mM imidazole, and 6 M urea), the column was eluted with 20 mM

Tris'HCl. pH 7.9. 500 mM NaCl. 1 M imidazole, and 6 M urea. The eluted proteins ('2 mL) were precipitated with 4 volumes of cold acetone and analyzed on an 8% SDS-

PAGE gel.

3.2.8 Binding Assay

The dissociation constants (ATd) of peptide 1 to native Tsp and the isolated PDZ-c domain were determined by measuring the radiationless resonance energy transfer from the tryptophan(s) in the protein to the EDANS group on the peptide (Selvin. 1995). A

43 1.5-mL solution containing 50 mM Tris-HCl (pH 8.0), 200 mM NaCl. 250 nM of peptide

I. and 0.1-8.0 pM wild-type Tsp or PDZ-c in a quartz cuvette was placed in a Perkin-

Elmer spectrofluorometer. The excitation wavelength was set at 290 nm while the emission at 520 nm was monitored. The observed fluorescence yield at 520 nm. F. was fitted to the equation F = F^ax • [P]/(A^o +[?]) where [P] is the concentration of Tsp or

PDZ-c added and Fmax is the maximum fluorescence intensity at 520 nm. Inner filtering effects were minimal over the concentration range of 0.1-8.0 pM PDZ-c; in the absence of peptide I. the fluorescence yield of PDZ-c at 370 nm increased linearly with protein concentration. Experiments were carried out in triplicates.

3.3 Results

3.3.1 V8 Proteolysis of Tsp

Partial proteolysis of Tsp by VS protease resulted in two fragments of estimated molecular masses of 55 kDa and 25 kDa (Figure 3.3). accompanied by the loss of Tsp activity. Since the intact Tsp has an apparent molecular mass of 75-80 kDa on an SDS-

PAGE gel. this indicates selective cleavage of Tsp at a single site by V8 protease. To determine whether the cleavage occurred near the N-terminus or C-terminus of Tsp. the

V8 digestion was performed on a Tsp variant that carries a C-terminal histidine tag. After digestion to completion, the resulting firagments were denatured by 6 M urea and loaded onto a Ni“^ affinity column (Hichuli et al.. 1987), upon which the histidine-tagged C- terminal firagment should bind, whereas the N-terminal fragment would not (assuming a single cleavage site). SDS-PAGE analysis revealed that a single bound fragment with an

44 apparent molecular mass of 55 kDa was eluted from the column (Figure 3.4). Thus, the single V8 cleavage occurred closer to the N-terminus in the Tsp sequence. To locate the precise site of cleavage, the reaction products were analyzed by electrospray ionization mass spectrometry (ESI MS). The molecular masses of intact Tsp (wild-type) and the two cleavage fragments were found to be 75168 Da, 49491 Da. and 25689 Da. respectively.

This is most consistent with cieavage of the peptide bond between N211 and T212

(calculated masses of 49506 Da and 25708 Da for the two fragments, despite the absence of a Glu or Asp that V8 typically cleaves after. Surprisingly, all of our attempts to separate the two fragments under nondenaturing conditions (gel-filtration and ion exchange chromatography) failed, suggesting that the two fragments remain tightly associated even after proteolysis but demonstrated no significant activity. We also noted that prolonged storage of the protein at 4°C produced a similar cleavage pattern, suggesting a common region that is sensitive to proteolysis.

3.3.2 Photoaffînity Labeling.

Previous studies with both small peptide and large protein substrates suggest that

Tsp contains a separate substrate-binding site that is responsible for recognizing the hydrophobic C-terminus of a substrate (Keiler et al.. 1995; Beebe and Pei. 1998). To test for the presence of such a substrate recognition site, we synthesized a fluorescent photoaffinity label (peptide 1, Figure 3.1), in which an EDANS group and a photolabile benzophenone group were attached to the N-terminus of the octapeptide GRGYAL.A.A. a sequence known to act as a competitive inhibitor of Tsp, while not acting as a substrate

(Beebe and Pei, 1998). Binding of peptide 1 to Tsp would bring the benzophenone group

45 to the vicinity of the substrate recognition site; upon exposure to 350 nm light, the fluorescent peptide would be covalently attached to the protein via the benzophenone group (Dorman and Prestwich, 1994) and the protein would become fluorescent. ,\s expected, when Tsp (50 pg) was incubated with peptide 1 (100 pM) under 350 nm light for 25 min. the peptide became covalently attached to Tsp. as evidenced by the appearance of a fluorescent band at -75 kDa on SDS-PAGE gels (Figure 3.5. panels .A and B. lane 2). When a control experiment was carried out with phosphatase SHP-1 (Pei et al.. 1993). a 68-kDa protein readily available in this laboratory, no labeling was observed under the same set of conditions (Figure 3.5. panels A and B. lane 5).

.Additionally, no labeling was observed in the presence of the catalytic portion lacking the putative PDZ domain (data not shown). Further, successful labeling required the presence of all three components: Tsp protein, the peptide, and UV light. HPLC analysis of peptide I demonstrated that it was not cleaved by Tsp under the assay conditions despite the high enzyme concentration. This result demonstrates that Tsp indeed contains a recognition site for the nonpolar C-terminus of a substrate, which may or may not be the same as the catalytic site.

To locate the binding site for peptide 1. photoaffinity labeling experiments were carried out with various deletion mutants of Tsp. Primary sequence analysis has previously identified a region (amino acids 217-301) that exhibits some sequence similarity to known PDZ domains (Ponting, 1997). Since PDZ domains are known to bind to the C-termini of their partner proteins, the putative PDZ domain in Tsp may be the substrate recognition site. Consistent with this notion, when photoaffinity labeling experiments were carried out with ANT Tsp, a tnmcation mutant lacking the sequence N-

46 terminal to the putative PDZ domain (amino acids 1-205), efficient labeling was observed (Figure 3.5, panels C and D, lane 3). Similarly, a C-terminal truncation mutant.

ACT Tsp (amino acids 1-334), which lacks all of the sequences C-terminal to the putative PDZ domain, was also efficiently labeled by peptide 1 (Figure 3.5. panels C and

D. lane 4). These results suggest that the substrate recognition site is located between amino acids 205 and 334, the region where the putative PDZ domain is located (Figure

3.2). Therefore, the putative PDZ domain is the likely substrate recognition site of Tsp.

In order to isolate the PDZ domain, several truncations were performed by the

PCR method (PDZ-a-d. Figure 3.2). We found that the PDZ domain core (PDZ-d. amino acids 206-307) as defined by sequence homology (Ponting. 1997) failed to produce any soluble protein. Extension of the PDZ core by 27 amino acids at the C-terminus but not at the N-terminus afforded a soluble protein (PDZ-c. Figure 3.2). When PDZ-c (124 pM) was incubated with peptide 1 (100 pM) in the presence of 350 nm UV light, the PDZ domain was efficiently labeled with the fluorescent probe (Figure 3.5. panels C and D. lane 5). This labeling is both UV- and peptide-dependent. PDZ-a (amino acids 185-334) was also purified as soluble protein and was efficiently labeled by peptide 1 (data not shown). PDZ-b ( amino acids 185-307) was purified only in very small quantities as the majority was present in inclusion bodies, however data suggested that PDZ-b crosslinks to peptide 1 as well.

To further rule out the possibility that the observed labeling may be due to nonspecific modification at multiple sites, we performed competition experiments, in which photoaffinity labeling by peptide 1 was carried out in the presence of a competing peptide (peptide 2, Figure 3.1). Peptide 2 was obtained as a by-product from the synthesis

47 of peptide 1 and was formed by the addition of one molecule of ethylenedithiol. presumably to the benzophenone group, during the peptide deprotection step. Peptide 2 binds Tsp and inhibits Tsp activity but does not covalently crosslink to Tsp due to the loss of the chromophore. As shown in Figure 3.6B. when increasing amounts of peptide 2 were added to the reaction between Tsp and peptide 1. the amount of fluorescently labeled Tsp decreased. The SDS-PAGE gel stained with Coomassie blue displays that there were equal amounts of Tsp in all reactions (Figure 3.6, panel A). Significant competition is observed at concentrations of competitor in only slight excess of peptide 2

(300 pM) despite irreversible labeling of peptide 1. suggesting that both peptides compete for binding to a specific site instead of a series of nonspecific sites. These results taken together demonstrate that PDZ is indeed the domain through which Tsp recognizes and binds its substrates.

To locate the site of covalent attachment on the PDZ domain. PDZ-a was

incubated with excess peptide 1 in the presence of 350 nm light and the unreacted

peptides were removed by dialysis. The covalently modified protein was digested with

trypsin and the resulting peptides were separated by reversed-phase HPLC. AW major

peaks were collected and checked for the presence of fluorescence. Out of the twenty

peaks collected, peaks 11 and 15 showed strong absorbance at 214 nm and were intensely

fluorescent. These two fractions were further analyzed by electrospray ionization mass

spectrometry (ESI MS). Peak 11 gave a peak at m/z = 840.7. Although this closely

matches the calculated molecular mass of an unmodified peptide from the PDZ domain.

L295EILPAGK302 (calculated MET = 840.52), this peptide should not be fluorescent.

Therefore, we assume that the peak at m/z = 840.7 carried two positive charges, derived

48 from a single peptide that had a molecular mass of 840.7 x 2 - 2 = 1679.4 Da. .A.fter trypsin digestion at the arginyl residue, the N-terminal fragment of peptide 2. which contained both the EDANS group and the benzophenone moiety, had a molecular mass of

830.4 Da as determined by ESI MS under the same conditions. Thus, the unknown peptide fragment derived from the PDZ domain should have a molecular mass of 1679.4

- 830.4 = 849.0. This value matches closely the mass of peptide M 3:

(calcd M = 848.5). which Is located at the extreme C-terminus of the PDZ-a protein. No other peaks matched the observed molecular mass (-10 Da deviation). Peak 15 produced a peak at m/z = 1072.0. which did not match either a fragment from the unmodified PDZ- a or any fragment plus the fluorescent probe. One possibility is that a modified peptide might have undergone other structural changes during the experimental process.

Experiments with an N-terminally biotinylated peptide (which has similar sequence to peptide 1) also resulted in labeling of residues in the C-terminus of PDZ-a.

3.3.3 Direct Binding of Peptide 1 to Tsp and PDZ Domain

Binding of peptide I to Tsp was also demonstrated by the presence of fluorescence resonance energy transfer from the tryptophan residue(s) in Tsp (excitation wavelength = 290 nm) to the EDANS group in the peptide (emission wavelength = 520 nm). By varying the Tsp concentration (0.1-8.0 pM) and examining the fluorescence emission from the EDANS group at 520 nm (peptide concentration fixed at 250 nM). a dissociation constant {Kq) of 1.8 ± 0.3 pM was determined (Figure 3.7). The binding

49 affinity of peptide 1 to the PDZ domain (PDZ-c) was similarly determined, with a A.'d value of 1.9 ± 0.3 p,M. The similarity in apparent binding affinities further demonstrates that there is one site of binding and not multiple substrate binding sites.

3.3.4 Catalytic Property of Tsp M utants

All PDZ domains contain a ü L û F loop motif or its homologue (Doyle ei al..

1996: Ponting, 1997). This sequence is critical for recognizing the C-terminus of its partner protein by forming a series of hydrogen bonds between the PDZ backbone amides and the carboxylate of the partner protein (Doyle ei al.. 1996). In Tsp. the corresponding sequence is GIGA and valine 229 is located immediately C-terminal to the GIG.A motif

(Silber et al.. 1992). Using the PDZ domain of PSD-95 as a model (Doyle et al.. 1996).

Val-229 of Tsp should be solvent exposed and is likely involved in hydrophobic interactions with a substrate. With mutation of this residue it was expected that the binding affinity of Tsp to a hydrophobic peptide would change but major changes in the global structure of Tsp would not occur. To this end. valine 229 was mutated to a glutamine and glutamate, respectively, with the expectation that the hydrophilic side chains of Gin and Glu would lower the affinity of Tsp to substrates containing hydrophobic C-termini. Both V229Q and V229E mutants produced soluble proteins. which were efficiently labeled by peptide 1 upon UV irradiation (data not shown) and were catalytically active (Table 3.1). While these mutations had no detectable effect on the value (0.05 s '), they caused two to three-fold increase in the K\\ value when assayed against a peptide substrate (Beebe and Pei, 1998). This is consistent with the notion that changes in the PDZ domain only affect the overall binding affinity of a

50 substrate but do not affect the catalytic power of the active site, which is located at the C- terminal region of Tsp and is physically remote from the PDZ domain.

The deletion mutants were also evaluated for catalytic activity. Although ANT

Tsp contains both the PDZ domain and all of the catalytic residues as identified by mutagenesis studies (Keiler and Sauer. 1995), it was catalytically inactive. The mutant protein was somewhat less stable than the intact Tsp. undergoing dénaturation when the ionic strength of the solution was low. The C-terminal truncation mutant. ACT Tsp. was also catalvticallv inactive.

A-'m Koat kcat^ K\| Enzyme (X 10" M-‘ s'') (pM) (s ')

wild-type 4.4 ± 0.6 0.050 ± 0.000 11

V229Q 9.5 ± 1.3 0.050 ±0.010 5.3

V229E 11.1 ± 1.4 0.050 ± 0.002 4.5

“Assay reactions were performed in 50 mM Tris-HCl. pH 8.0. 250 mM NaCl.

and 5 pg Tsp in the presence of 0-20 pM substrate. Data reported are the mean

± SD of three independent experiments.

Table 3.1. Kinetic Constants for Wild-Type and Mutant Tsp Enzymes^

51 3.4 Discussion

Tsp is different from the other known proteases in that it recognizes both the C- terminus and the sequences flanking the scissile bond of a substrate. The substrate specificity of Tsp is primarily determined by the sequence at the C-terminus. with protein containing small hydrophobic C-terminal residues being the preferred substrates (Keiler et al.. 1995). On the other hand, the sequences flanking the known cleavage sites in several peptide as well as protein substrates are quite diverse (Keiler ei al.. 1995). This unusual specificity allows Tsp to selectively degrade any proteins with nonpolar C- termini but not those proteins which are needed for proper cellular functions and which necessarily do not contain nonpolar C-termini. One of the functions of Tsp (or its homologue ClpXP or ClpAP) is to degrade polypeptides synthesized from damaged mRNAs. which are potentially toxic to the cell and are targeted for destruction even before they are released from the ribosome by tagging a nonpolar tail to their C-termini

(Keiler and Sauer. 1996). Because the cleavage site in a substrate and its C-terminus have been found to be distinct and typically separated by significant distance (Keiler ei al..

1995). it is unlikely that a single active site can interact with both motifs. Therefore, a separate binding domain has been proposed to be responsible for recognizing the C- terminal sequence of a potential substrate (Keiler et al.. 1995). In this work, we have shown that this C-terminal binding site is mediated by a PDZ domain, based on several lines of evidence. First, sequence alignment shows that amino acids 217-301 share significant sequence homology to several well-known PDZ domains (Ponting and

Phillips. 1995). Second, proteolysis by V8 protease, which cleaves after aspartvl and glutamyl residues, resulted in cleavage at a single site immediately before the PDZ

52 domain. This suggests the presence of individual domains in Tsp. Third, deletion mutagenesis locate the C-terminal binding site to amino acids 206-307. the same region where the PDZ domain resides. Fourth, The Tsp fragment containing amino acids 206-

334 forms a soluble protein and is capable of binding to a peptide with a nonpolar C- terminus. Finally, mutagenesis in the PDZ domain, which is remote from the catalytic site, decreased the catalytic activity of Tsp by reducing its affinity to the substrate.

The above data leads us to propose that Tsp contains a minimum of three separate domains: an N-terminal domain (amino acids 1-205). a PDZ domain (amino acids 206-307), and a C-terminal catalytic domain (amino acids 308-660) (Figure 3.8).

We speculate that the PDZ domain first recognizes a cognate substrate by binding to its nonpolar C-terminus. By doing so. it recruits a substrate to the catalytic site, which subsequently cleaves the substrate at multiple sites. This would predict that the cataKiic site is a relatively poor enzyme that has very low affinity towards its substrates so that it cannot form a productive E«S complex without the help from the PDZ domain. Indeed. addition of peptide 1 or other peptides with nonpolar C-termini strongly inhibited the catalytic activity of Tsp. An alternative mechanism is that binding of the C-terminus of a substrate to the PDZ domain causes a conformational change that activates the catalytic domain for peptide cleavage. This latter mechanism is less likely to be the case, as it would predict that peptide 1 or other nonpolar peptides act as activators instead of

inhibitors of Tsp. The function of the N-terminal domain remains unclear. It is clearly

important for catalysis, despite containing none of the residues identified as critical for catalysis (Keiler and Sauer, 1995), as its removal resulted in a protein that is capable of

binding to a substrate but has no catalytic activity. Even a single cleavage in the apparent

53 linker region between the N-terminal domain and the PDZ domain inactivated the enzyme. It is possible that residues from both the N-terminal and C-terminal domains form the active site. Equally likely is the possibility that the lack of catalytic activity from the N-terminal deletion mutant could be entirely due to a structural perturbation in the catalytic domain because of the loss of a N-terminal domain that normally serv es to maintain structure. The explanation may also exist in a more finite mechanistic structural role in which the N-terminal domain assists in the orientation of a loop, or helix that contains a critical catalytic residue. The proposed general topology model (Figure 3.8) would be consistent with any of these scenarios. Note tliat even after V8 proteolysis, the two fragments remained tightly associated, suggesting a strong noncovalent protein- protein interaction existing in Tsp. Indeed, in September of 2000. the crystal structure of a plant homologue of Tsp. Dl protease, was solved. As seen in the crystal structure representation in Figure 3.9. the general topology is that of what we predicted in Figure

3.8 with three distinct domains being present and the NT domain bridging the PDZ covalently and the catalytic domain non-covalently. The precise role of the NT domain remains unclear but it is unlikely that it donates a key catalytic residue as mutagenesis through the NT domain of Tsp has been extensive and the key catalytic residues have been shown to be in the designated catalytic domain. The authors suggest a role in substrate binding, but only one residue from the NT domain is referenced as making up this hydrophobic binding site. As can be seen from the crystal structure representation, a

(3-sheet and loop from the catalytic domain extend up into and actually form part of the

NT domain. The critical lysine general base is located at the base of the ascending (3- strand and therefore it can be seen that the orientation of this residue, based on the

54 position of the strand, may be determined by the interaction with the NT domain. Clearly the catalytic activity takes place at the junction of the three domains but the precise role of each component is still undefined in Tsp.

Since the identification of Tsp in 1992. several Tsp homologues have subsequently been discovered in both prokaryotic and eukaryotic organisms (Keiler and

Sauer. 1995: Oelmuller et a i. 1996). Sequence analysis shows that members of the Htr.A protease family, which are heat shock-induced serine proteases found in bacteria, yeast, and human, also contain one or multiple putative PDZ domains (Ponting and Phillips.

1995; Palien and Wren. 1997). Furthermore, the two-component. ATP-dependent

proteases ClpAP. ClpXP. and ClpYQ as well as the Lon protease have recently been shown to contain a 100-amino-acid sensor and substrate discrimination (SSD) domain in their .ATPase component, which recognizes specific sequences within a substrate protein

(Smith et al., 1999: Van Melderen and Gottesman. 1999). Therefore, it appears that the

use of a separate substrate recognition domain (e.g.. PDZ or SSD domains) is a common

mechanism by which specific proteases discriminate their target protein(s) from other

nonspecific proteins.

55 so,H

OH

HO H,N NH.

SO,H 1

HN, OH

HO H,N NH.

Figure 3.1. Structure of photoaffinity labels.

56 WT NTD PDZ CD

ANT[206-660] PDZ CD

ACT [1-334] NTD PDZ

PDZ-a [185-334] PDZ

PDZ-b [185-307] PDZ

PDZ-c [206-334] PDZ

PDZ-d [206-307] PDZ

Figure 3.2. Schematic representation of the stmctures of wild-type and mutant Tsp used in this work. The three domains in the wild type protein (WT) are pictured. All domain constructs are present with N-terminal six histidine tags. Solid black lines represent regions outside of the PDZ homology region produced with the particular domain constructs. The N-terminal region, and catalytic region are designated as NTD and CD respectively. Amino acids that each domain represents are in brackets to the right of each domain name. Amino acid numbering is based on (Keiler and Sauer. 1995).

57 Time (min) 0 2 4 8 16 20 22 24 26 28 30 40 50

97.2- 66.4 - 55.6- 42.7-

26.6-

20.1 -

Figure 3.3. 10% SDS-PAGE gel showing the cleavage o f Tsp by V8 protease. .At the indicated time points, the reaction was quenched by the addition of equal volume of 2x SDS-PAGE loading buffer and heating at 95°C for 5 min.

58 kDa

9 7 -

4 5 -

Figure 3.4. 8% SDS-PAGE gel of V8 proteolyzed and denatured TspHô Lane 1. molecular weight marker: lane 2. proteolyzed sample that was retained and eluted from Ni"' affinity column.

59 Figure 3.5. Photoaffinity labeling of wild-type Tsp and its deletion mutants. (A) Coomassie blue stained SDS-PAGE gel showing wild-type Tsp labeled with peptide 1. Lane 1. molecular weight marker; lane 2. positive reaction with all components present: lane 3. no peptide; lane 4. no UV light; and lane 5. negative control with SHP-1 protein (no Tsp). (B) Fluorescent image of the same gel shown in panel A. (C) Coomassie blue stained SDS-PAGE gel showing Tsp deletion mutants that had been labeled by peptide 1. Lane 1. molecular weight marker; lane 2. wild-type Tsp; lane 3. ANT Tsp; lane 4. ACT Tsp; and lane 5, PDZ-c domain. (D) Fluorescent image of the same gel shown in panel C.

60 Peptide 1 + - + 4- Tsp + + + - UV A kDa 97. 4 - 66 -

4 5- -

31 - —

21.5 -

kDa 97.4 -

6 1 97.4-

P e p tid e l 2

Figure 3.6. Competition between photoaffinity label (peptide 1) and peptide 2 for binding to Tsp. (A) Coomassie blue stained SDS-PAGE gel showing the labeled Tsp in the presence of 0-1200 pM peptide 2 (lanes 2-6). Lane 1 is molecular weight markers. (B) Fluorescent image of the same gel shown in panel A.

62 15

10 •

#- F

5 #

#

0 012345678

(iM PDZ-c

Figure 3.7 Fluorescence resonance energy transfer. 200 nM peptide 1 titrated with increasing concentration (|iM) of PDZ-c.

63 Figure 3.8. Proposed model of substrate recognition and catalysis by Tsp. Solid arrow, our initially proposed model. Broken arrow, proposed active site location based on crystallographic data from the D1 protease.

64 N-Terminal Domain

N

PDZ Domain

Catalytic Domain

Figure 3.9. Crystal structure of DIP reproduced ffom Nature Structural Biology, v. 7(9), 749-53 (2000)

65 CHAPTER 4

DETERMINATION OF THE BINDING SPECIFICITY OF THE

SH2 DOMAINS OF PROTEIN TYROSINE TYROSINE PHOSPHATASE

SHP-1

4.1 Introduction

Signal transduction, regulation of gene expression, and other cellular processes are all mediated by two types of events: chemical modification (e.g.. phosphorylation and dephosphorylation) and physical association of the proteins involved (i.e.. formation of protein-protein complexes). A major advance in the latter area was the realization in

1990*s that many protein-protein interactions are mediated by small protein modules. which recognize linear peptide motifs in their partner proteins (Bork ef a i. 1997). The

Src homology-2 (SH2) domain was among the first such modules discovered (Sadowski et al.. 1986). It consists of -100 amino acids and is an independently folded functional module found in a wide variety of signaling proteins (Pawson and Gish. 1992; Pawson.

1995). SH2 domains bind to their interacting proteins by recognition of linear phosphotyrosine (pY)-containing sequence motifs (Songyang et al.. 1993: Songyang et al.. 1994). Structural studies of SH2-pY peptide complexes reveal that a key interaction. which is common to all SH2 domains, is the insertion o f the pY side chain into a deep pocket in the SH2 domain, where an invariant arginine residue forms a bidentate

66 interaction with the pY phosphate group (Waksman et a i, 1992; Eck et al.. 1993; Lee el al.. 1994). Additional binding energy is provided by interactions between amino acids adjacent to pY. particularly the three residues immediately C-terminal to pY. and the less conserved surface of the SH2 domain. This latter interaction also governs the selectivity of a given SH2 domain in binding to a specific pY partner. Presumably, the binding specificity of different SH2 domains directs the proteins that contain them to different pY receptors, thereby transducing specific signals to downstream proteins. .A.n important task in signal transduction research is therefore to determine the binding specificities of these

SH2 domains as well as other protein modules.

Several combinatorial library methods have been devised to systematically determine the binding specificity of SH2 domains. The first method employed an SH2 affinity column to enrich the SH2-binding sequences from a pY peptide library.

Sequencing of this enriched pool generated a consensus sequence!s) based on preferentially selected amino acid(s) at a given position (Songyang et a i. 1993;

Songyang et a i. 1994). The second method involved screening support-bound libraries against a fluorescently labeled SH2 domain (Muller et a i. 1996). The positive beads with the bound SH2 were removed ffom the library by fluorescence-activated bead sorting and sequenced as a pool. Both methods provide information on the consensus sequence but do not give individual sequences. As several amino acids are often enriched at a given position, there is still a great deal of uncertainty as to which particular amino acid combination will produce a high-affinity binder. It is possible that a peptide featuring the

most preferred residues at each position actually does not have the highest affinity.

Furthermore, because these methods select for both affinity and abundance of certain

67 types of sequences, a high-affmity peptide that is present only in minute amoimts in the library may be entirely missed. In the third method, bacteriophage bearing short random peptide sequences on their surfaces were phosphorylated by a kinase cocktail and selected against an immobilized SH2 domain (Gram et al.. 1997). The sequences of the

SH2-binding pY peptides were determined by amplifying the bound phage and sequencing their DNA. The limitation of this method is that the starting library is grossly biased by the specificities of the kinases (only 2-3% phage were phosphorv'lated).

Finally, a method has been developed to combine library affinity selection with mass

spectrometry to identify SH2-binding sequences (Kelly et al.. 1996). However, this

method has only been demonstrated with very small libraries (361 members).

We have developed another method to identify specific binding motifs for SH2

domains or other protein modules from a combinatorial peptide library. Libraries ( up to

2.5 X 10^ individual sequences) were synthesized on beads (-90 pm in diameter) using

the split-pool method (Furka et al.. 1991; Lam et al.. 1991), with each bead carrying a

unique peptide sequence. The libraries were screened for SH2 binding using an enzyme-

linked assay and the positive beads were manually removed from the library and

sequenced by matrix-assisted laser desorption ionization (MALDl) mass spectrometry.

This method provides both individual sequences and a consensus sequence(s) and should

be readily applicable to other protein modules. The power of this method has been

demonstrated by the rapid identification of high-affinity pY peptide ligands for the SH2

domains of protein tyrosine phosphatase SHP-1 (Shen et al.. 1991; Matthews et al.. 1992;

Plutzky et al., 1992; Yi et al., 1992).

68 4.2 Experimental Procedures

4.2.1 Materials and General Methods

5-bromo-4-chloro-mdolyl-phosphate (BCIP), biotin, and streptavidin-aikaline

phosphatase conjugate were purchased from Sigma Chemical Company. Gelatin was

purchased from Bio Rad. All peptide synthesis reagents and resins were purchased from

.\dvanced ChemTech (Louisville. fCYj. Soluble peptides including biotinylated peptides

were synthesized on Wang resin using standard Fmoc chemistry on 0.015-0.2 mmole

scale as detailed previously (Beebe et al.. 2000). Phosphotyrosyl (pY) peptides were

synthesized using unprotected N-Fmoc-pY during the coupling reactions (Ottinger et al..

1993). Crude peptides of >80% purity (as judged by analytical HPLC) were used directly.

whereas less pure samples were purified by reversed-phase HPLC on a semi-preparative

Cis column prior to use. The identity of all peptides was confirmed by MALDl mass

spectrometric analysis. SHP-1. its isolated N- and C-terminal SH2 domains, and their

maltose-binding protein (MBP) or glutathione-S-transferase (GST) fusion proteins were

expressed and purified as previously described (Pei et al.. 1993; Pei et al.. 1994: Pei et

al.. 1996) using standard procedures for these fusion proteins. Biotinylation of proteins

was carried out by treating MBP-SH2(N), MBP, or the free SH2(C) protein (180-800 pM

in 50 mM sodium phosphate, pH 7.0, 150 mM NaCl) with 1.8—2.5 molar equivalents of

N-hydroxysuccinimidobiotin for 30 min at room temperature. Dry N-

hydroxysuccinimidobiotin was dissolved first in DMF and immediately before addition to

the protein was diluted into the above phosphate buffer. The samples were passed

through a Pharmacia G-25 Fast Desalting column equilibrated in 20 mM Tris-HCL pH

69 7.5, ISO mM NaCl, 10 mM p-mercaptoethanol. and 1 mM EOTA to remove any free biotin and to quench any unreacted ester with the Tris-based buffer. The protein was collected, concentrated in a Centriprep-10 concentrator, and quickly frozen in the presence of 33% glycerol. To serve as a qualitative check for biotinylation. samples of biotinylated protein were run over a streptavidin-agarose column in parallel with a unbiotinylated sample in a separate column. After washing, the presence of immobilized protein was detected with Bradford reagent with the presence of immobilized protein being detected by a strong increase in blue color compared to the control.

4.2.2 Library Construction

The pY peptide library. Ac-EDXXpYXXXIBBRM or Ac-LKpYXXXXBBRM (X

= Nle or any natural amino acid except for Met and Cys). was synthesized on 5 g or 1 g of TentaGel S NHi resin (80-100 pm. 0.3 mmol/g) as previously described (Hu et til..

1999). Briefly, the diversity of the library was generated during synthesis by the split and mix protocol. Each amino acid is coupled in a separate reaction vessel and after synthesis, all of the resin is combined and mixed. After thorough mixing, the resin is equally split into separate portions for a second coupling of a single amino acid to each resin portion. This process of coupling, combining and mixing, followed by equal splitting of the resin for another coupling is repeated imtil all of the random positions have been generated. Using this split-pool synthesis method (Furka et al.. 1991; Lam et al.. 1991), the coupling reactions were carried out with 5 equivalents of reagents for 3-4 h and repeated once. To generate chain-termination products for peptide ladder sequencing by MALDl mass spectrometry, a small amount of capping reagents were

70 added to the coupling reactions (along with the individual Fmoc-amino acids) during the synthesis of the randomized region (Youngquist el al., 1995). iV-Acetylalanine ( 10%) was used for Nle; a mixture of Y-acetylalanine and iV-acetylglycine (5% each) was used for

Gin and lie: and yV-acetylglycine (10%) was used for all other amino acids. Side-chain deprotection of the resin-bound peptides was performed as previously described (Hu et al.. 1999), using a typical cleavage cocktail (5 mL of trifluoroacetic acid (TF.A). 0.13 mL ethanedithiol. 0.25 mL thioanisole, 0.10 mL of anisole, 0.375 g crystalline phenol and

0.25 mL of water) with mixing for 1 hour.

4.2.3 Library Screening and MALDl Peptide Sequencing

Library beads were exhaustively washed with CHiCL. DMF. HiO. and TBS buffer (20 mM Tris-HCL pH 7.4. 150 mM NaCl. 0.1% Tween 20) containing 0.1% (w v) gelatin. The library beads were incubated for 1 h in the above buffer to block any nonspecific protein binding sites. The beads were next suspended in fresh TBS buffer containing 1.0 pM biotinylated MBP-SH2(N) or 1.5-3.5 pM biotinylated SH2(C) and the

mixture was incubated for 6-12 h at 25 X with constant, gentle end over end mixing.

After initial incubation with SH2 protein, the mixture was supplemented with 10-14.3

nM streptavidin-aikaline phosphatase conjugate and 25 mM sodium phosphate. The

mixture was shaken at room temperature for 15 min and quickly washed under vacuum

with PBS buffer (20 mM sodium phosphate. pH 7.4, 150 mM NaCl) containing 0.1 %

Tween 20 (2 x 3 mL). 3 mL of PBS buffer, and 3 mL of TBS buffer (20 mM Tris-HCL

pH 8.5.150 mM NaCl). The resin beads were suspended in 12 mL of the TBS buffer (pH

8.5) and incubated in the presence of 1.4 mM 5-bromo-4-chloro-3-indolyl-phosphate

71 (BCIP) in a Petri dish while undergoing gentle rocking. The placement of the beads in a

Petri dish allows for convenient periodic monitoring of the staining reaction under a microscope. Typically the staining reaction was allowed to proceed for 30-60 min at room temperatiu-e before the beads were collected in a disposable column and washed several times with ddHiO and quenched by incubation in 6 M guanidine hydrochloride for 15-20 min. The beads were washed exhaustively (15-30 times 1-3 mL) with deionized water prior to sequence analysis. Positive beads were readily identified by their intense turquoise color and were manually removed from the library with a micropipette under a low-power microscope. A control screening was performed, with biotinylated

MBP (no SH2) under the same conditions, but resulted in no colored beads. Individually selected beads with all water removed by evaporation were each incubated with 20 pL of

20 mg/mL CNBr in 70% formic acid for 20-24 hours in the dark. The solvent and excess

CNBr were removed under vacuum and the released peptides from each bead were dissolved in 3 pL of 0.1% TFA in deionized water. Sequence analysis was carried out on a Kratos Kompact MALDl-111 mass spectrometer in the positive ion mode as described previously (Hu et al.. 1999). Briefly. 1 pL of dissolved peptide was added to 2 pL of aCCA (freshly prepared as a saturated solution in 50% acetonitrile containing 0.1%

TFA). mixed quickly followed by spotting 1 pL of the mixture on the MALDl sample slide.

72 4.2.4 SH2-pY Peptide Binding Assays

Two different assays were performed to detect binding of SH2 domains to pY peptides selected ffom the library. In method A. pY peptides were assayed for their ability to stimulate SHP-1 activity (Pei ei a i. 1994; Pei et a i. 1996). The assay reaction

(total volume of 50 pL) contained SHP-1 (73 nM). 10 mM p-nitrophenyl phosphate

(pNPP). 100 mVI Hepes. pH 7.4. 1.0 mM P-mercaptoethanol. 1.0 mM EDTA. 150 mM

NaCl. and 0-600 pM pY peptide. The reaction was allowed to proceed for 31 min at room temperature before being quenched with 950 pL of 1 M NaOH. and the absorbance at 405 nm was measured on a UV-VIS spectrophotometer. Peptide concentrations were determined by hydrolysis of pY to completion with alkaline phosphatase followed by the malachite green assay of the released inorganic phosphate (Lanzetta et al.. 1979). .-Ml reactions were carried out in triplicates.

In method B. the Pharmacia BlAcore was employed (Malmqvist. 1993). pY peptides were biotinylated at their N-termini via solid-phase peptide synthesis and immobilized onto streptavidin-coated sensor chips by passing a pulse of -0.1 pM peptide solution in HBS buffer (10 mM Hepes, pH 7.4. 150 mM NaCl. and 3.4 mM EDT.A) over the surface of a chip. For most peptides. 1-5 pL at a flow rate of 10 pL/min was sufficient to result in 150-350 response units (RU). The amount of immobilized pY

peptide was determined by saturating the surface with GST-SH2(N+C) and measuring

the maximal response units. This low loading density of pY peptides drastically reduces

the avidity effects due to GST dimerization (Ladbury et al., 1995). Data for secondary

plot analysis was generated by passing increasing concentrations (0-50 pM) of an SH2

protein in HBS buffer over the sensor chip at 10 pL/min for 3 min. A sham flow cell with

73 no attached peptide was used to account for any signal generated due to the solvent bulk or any other effect not specific to the SH2-peptide interaction. In between runs, regeneration of the chip surface was facilitated by using the running buffer supplemented with 0.025% SDS and 0.019% NP-40. Data analysis was carried out by plotting the response at equilibrium, obtained by subtracting the response of the sham flow cell from that of the pY peptide flow cell, against the SH2 domain concentration. The data were fit to the equation

RU = RUma.x[SH2]/(/fo+[SH2]) where RU is the measured response and RUmax is the maximum response.

4.2.5 Analysis of novel consensus sequence

The class I SH2(N) consensus sequence was searched against the PIR protein database for pattern matches. The sequence L/L'WY-X-Y-M/F-X-M/F was searched against the

PIR pattern match. All non-human sequences were eliminated from the querry.

4.3 Results

4.3.1 Library Design, Construction, and Screening

The majority of SH2 domains recognize primarily the pY residue and the three residues immediately C-terminal to pY (Songyang et al., 1993; Songyang et al.. 1994). although it has been reported that, for a few SH2 domains including those of SHP-1. the -

2 position (relative to pY) of a pY peptide is also important for high-affinity interaction

(Burshtyn et al.. 1997; Huyer and Ramachandran, 1998). Thus, we initially designed a

74 pY library. acetyi-LKpYXXXXBBïlM-resin in which the four C-terminal residues were randomized followed later by the construction of acetyl-DEXXpYXXXIBBRM-resin. in which the two residues immediately N-terminal to the pY and the three C-terminal residues are randomized [X = norleucine (Nle) or any of the 18 natural amino acids except for Met and Cys; B = P-alanine]. The C-terminal methionine permits the peptides to be released from the resin by CNBr treatment prior to sequencing, whereas the C- terminal arginine provides a fixed positive charge that will improve the solubility of the peptides and. most importantly, increase the sensitivity for sequencing by mass spectrometry (Youngquist et al.. 1995). The tetrapeptide library was originally synthesized to test the methodology and the two N-terminal positions were fixed as LK because a well known natural peptide ligand for the SH2(N) was known to have that specific sequence (Pei et ai. 1994). In the pentapeptide library two acidic residues. .Asp and Glu. were added at the N-terminus to ensure reasonable solubility of all peptides. The two P-alanines add some flexibility to the peptides, making them more accessible to a protein target (Yu et al.. 1997). An invariant lie was placed at the +4 position in the pentapeptide library because after the completion of screening of the tetrapeptide library

(Ac-LKpYXXXXBBRM-resin) by the SH2(N) it was apparent that the selectivity at this position was weak but overall a preference for longer chain aliphatic residues was observed. Methionine is excluded from the randomized region to avoid internal cleavage during CNBr treatment, and is replaced by the isosteric norleucine residue. In all of our studies so far (Rajagopalan et al., 1997; Hu et al., 1999) we have found no significant effect on either protein binding or enzyme catalysis when Nle is substituted for Met. The library was synthesized on TentaGel S NHi resin (2.86 x 10^ beads/g) using the split-pool

75 method (Furka et al., 1991; Lam et al., 1991), with each bead carrying a unique sequence

(-100 pmol peptide on each bead). This method ensures equal representation of all possible sequences in the library. Partial chain termination was effected by the addition of

10% /V-acetylglycine and/or iV-acetylalanine to the individual coupling reactions during the construction of the randomized region (see Experimental Procedures). The addition of the 10% capping reagent during each coupling step creates a mixture of peptides on the bead surface that result in a peptide mass ladder in which the mass difference between an adjacent set of peaks is equal to the residue molecular weight of a particular amino acid (Figure 4.1)

The library (theoretical diversity = 19'*= 130,321 for the tetrapeptide library or 19' =

2.5 X 10^ for the pentapeptide library) was screened for binding to both SH2 domains of

SHP-1. For the N-terminal SH2 domain, a maltose binding protein fusion. MBP-SH2(N). was employed, whereas the free C-terminal SH2 domain was used. Both SH2 domains were biotinylated on a surface lysine residue(s) by treatment with N-

hydroxysuccinimidobiotin. Screening was based on the assumption that binding of the

biotinylated SH2 domain to a bead that carries a high-affinity pY peptide for the SH2

domain should recruit a streptavidin-aikaline phosphatase conjugate to the surface of that

bead. Upon the addition of 5-bromo-4-chloro-3-indolyl phosphate (BCIP). the bound

alkaline phosphatase would cleave BCIP into an indole, which should instantaneously

dimerize in air into indigo, resulting in a turquoise precipitate deposited on the bead

surface. As a result of this reaction cascade, a bead carrying a specific SH2 ligand would

become colored. Before screening was carried out on 1 or more copies of the library, it

was performed on small portions of resin to determine the appropriate quantity of

75 reagents, incubation times, and washing steps. Several important factors should be pointed out that are critical in order to achieve successful results. The first is that although BCIP/NBT is often employed in assays in concert to speed the color development, its use often results in uncontrolled staining and high background, especially when using it at the high pH buffer (pH 9.0) that it is typically supplied with.

It is likely that this high background and extremely rapid color development results, at least partially, from the reduction of NBT from trace quantities of thiol containing compounds left over from the deprotection. Another important factor, especially with complexes that have moderate affinity for their respective ligands is to minimize the wash steps for removal of unbound species. The appropriate amount of washing needs to be determined during pilot or small scale screenings but typically copious amounts of washing, carried out slowly, will strip most of the protein from the bead surface as a new equilibrium is continually being established. Inherent in the screening process is a that

SH2-ligand interactions with slow off rates will be favored in the selection as they will be

least effected by the wash steps. In most cases this should not provide a problem as only those SH2 domains with slow on and fast off rates will be effected as the ones w ith fast on and fast off should have the opportunity, even after the brief washing steps to reengage an appropriate peptide ligand. A further note is that color that takes longer than

2 hours to develop should be regarded with caution, as beads selected after long staining

times, in our experience, were false positives.

77 4.3.2 Analysis of Selected Sequences

Screening was initially carried out on the tetrapeptide library (Ac-

LKpYXXXXBBRM) using 3 complete copies of the library. From this screening 158 total beads were collected, cleaved from the resin and subjected to MALDl-MS analysis.

Out of 158 sequenced. 136 yielded unambiguous full sequence information (86%) (Table

4.1). Screening of 1.0 g of resin representing the pentapeptide library (-2.86 x 10'’ beads) each against the N- and C-terminal SH2 domains of SHP-1 resulted in -150 and -300 colored beads, respectively. Peptide ladder sequencing (Youngquist ei ai, 1995: Hu ei al.. 1999) was performed for all 150 beads selected against MBP-SH2(N) to give 97 unambiguous sequences (Table 4.2.1). The rest of the beads (total 53) produced mass spectra that missed one or a few peaks in the peptide ladder and. therefore, complete sequence assignment was not possible. Out of the 300 beads selected against the SH2(C ) domain. -175 were individually sequenced to give 112 unambiguous sequences (Table

4.2.2). Overall the success rate of full sequence determination for positive bead sequencing from the pentapeptide library was - 65%.

Table 4.1 displays the sequences selected for the SH2(N) binding to the tetrapeptide library and figure 4.2 demonstrates an overall consensus sequence of LKpYM/F-X-M/F-

X. Screening was not thoroughly carried out on the tetrapeptide library with the SH2(C) due to early small scale screening failures while the methodology was being developed and the desire to examine its selectivity for peptides from the pentapeptide libran,.

78 LKpYYRMW LKpYFHMW LKpYFMYS LkpYMEFR LKpYYMAR LKpYMQFY LKpYFEYR LkpYMEFK LKpYYRHH LKpYMQFF LKpYFMFK LkpYMKFA LKpYYRHW LKpYMQMF(2) LKpYFMFS LkpYMKFF LKpYLQFR LKpYMQMM LKpYFMRN LkpYMKYD LKpYLRFY LKpYMQLL LKpYFWRM LKpYMKYW T Vr>VT OMIT T a LKpYLMMM LKpYMQLR LKpYFMLR LKpYMKMM LKpYLRFM LKpYMQKF LKpYFMMM LKpYMHFM LKpYVVMW LKpYMQRM LKpYFLFR LKpYMHFW LKpYLRMF LKpYMQRL LKpYFLFK LKpYMHFH

LKpYFRRR LKpYMNFY LKpYFAFY LKpYMHMM( 2 LKpYFRFK LKpYMNFK LKpYFAFN LKpYMHLR LKpYFRFS LKpYMNFR(2) LKpYFDMF LKpYMHLM LKpYFRMM(;2) LKpYMNFN(2) LKpYFEFM LKpYMRFS(2 LKpYFRMF LKpYMNYM LKpYMLFL LKpYMRFF LKpYFRMS LKpYMNMA LKpYMLFA LKpYMRFH LKpYFRML LKpYMNMF LKpYMLWG LKpYMRFQ LKpYFRMM LKpYMNMH LKpYMLMF(2) LKpYMRFA LKpYFRYM LKpYFSMR LKpYMAFF(2) LKpYMRIG LKpYFRYFI:2) LKpYFSWY LKpYMAMK LKpYMRFM LKpYFRMG LKpYMSAF LKpYMALF LKpYMRYM LKpYFRAW LKpYMSFN LKpYMFMK LKpYMRYL LKpYFQMRI:2) LKpYAYMF LKpYMVMR LKpYMRYA LKpYFQMW LKpYAMMF LKpYMWLH LKpYMRYG LKpYFQAY LKpYAMFR LKpYMMFM LKpYMRLF LKpYFHAY LKpYAQLR LKpYMFMR LKpYMRLL LKpYFHAF LKpYARFF LKpYMMLF LKpYMRLM LKpYFHWY LKpYAKFF LKpYMYMK LKpYMRLW LKpYFHRF LKpYFMYK(2) LKpYMFMM LKpYMRLA LKpYFHMR LKpYFMYR LKpYMMLF LKpYMRML LKpYMELF LKpYMRWW

Sequences appearing more than once are designated by (#). 136 sequences were obtained from the sequencing of 158 positive beads. ______

Table 4.1 SH2(N) selected sequences against the Ac-LKpYXXXX

79 Class I LHpYMFM LHpYMFM MHpYMLF ClassIX LYpYFAF LHpYMFA LQpYMTF IRpYFSF LNpYAFL LLpYFMY LHpYMAF MQpYMLF IHpYMYM LNpYAMF LYpYFNY LHpYMVF LNpYMEL LHpYMVL LNpYAFL LLpYMFM LHpYMVL LNpYMLF IHpYMVF lYpYANL' LYpYMNM LHpYMLF LNpYMYF LWpYLNM LYpYANL* LYpYMDM LHpYYYF LNpYMAM MLpYLAL LYpYAAL LYpYMQL LHpYFLM LNpYMNF TDpYNWV LYpYADL lYpYMQM LHpYLQM LNpYMIF KApYYPF LHpYAIF lYpYMNL LHpYLMM LEpYMMF NHpYFMR LApYAWL LYpYMNV PRpYMAF IQpYMFM SYpYYLR LWpYAQL LYpYMRY LKpYMRF LNpYVLM NHpYADL LYpYMSL LRpYMRM VQpYLYF MHpYVLL LYpYMFP LKpYMAF LNpYFAF LYpYMRF* LKpYMSF LQpYFMM LYpYMAY VKpYMLF LSpYFFM LYpYMND MKpYMRF ISpYFLF LYpYMLE MRpYFRL NYpYMMQ LYpYMFQ LKpYFFW MHpYFMY LYpYLAP LKpYFMH MHpYFVY LYpYLFF LRpYYMF MHpYFMF PYpYMRM VKpYFMF MHpYFMF LHpYMDF LKpYLMF MHpYMYY LHpYMYY LRpYIVM MHpYLYF LHpYMLM LQpYMLF MHpYMLM

Peptides selected for testing. M. norleucine

Table 4.2 Sequences of Peptides Selected from AcDEXXpYXXX Library bv the SH2(N)

80 Class I IVpYAQM ClassII HGpYYMK FVpYYMK TTpYARL MQpYAMI MWpYYAR RWpYYMK IQpYYMK* VHpYARL* YQpYAYL LWpYYMQ KNpYYMK PRpYYMR lApYARL HHpYAVL AWpYYMQ HKpYYMK KRpYYMR HRpYARL SHpYAVL YKpYYMQ HSpYYMK DTpYYMR VMpYARL YQpYAIV YPpYYMQ NHpYYMK GFpYYMR YRpYARL YNpYAIV NGpYYMQ SSpYYMK YYoYYMP QGpYARL FRpYAIL GSpYYMQ AEpYYMR KYpYYMR TRpYARL YNpYALL FWpYYML TEpYYMR NYpYYMR RWpYARL FHpYAIL QQpYYML MApYYMR WGpYYMR VTpYARN LWpYALL LGpYYML RHpYFMR TGpYYMR YRpYARI HHpYALL SDpYFMR KHpYFMR QNpYYMR LNpYARV VKpYALL TNpYYMK YHpYFMK QIpYYMR IHpYAKV LYpYALL VNpYYMK NHpYFMR VGpYYMR IVpYAKL PVpYALL DIpYYMK TWpYFMR LPpYYMR LYpYAKL HFpYAAV NEpYYMK KTpYFMR VSpYYMR RRpYAKV VYpYAAL QQpYYMK YApYYMK ETpYYMR PYpYAKV LWpYSLV EDpYYMK PApYYMK lEpYYMR LYpYAHI VHpYMKL GDpYYMK VSpYYMK QWpYFMR lYpYAEL IQpYMRL REpYYMK GSpYYMK AWpYFMR IQpYAEL TEpYVKV EHpYYMK PSpYYMK VSpYFMR* VEpYAEL RKpYYMN DGpYYMK VSpYFMR VIpYADL RApYYMK SWpYYMK DTpYFMK VIpYANL

Peptides selected For testing. M. norleucine

Table 4.3 Sequences of Peptides Selected from AcDEXXpYXXX Librarv bv the SH2(C)

81 •Ajialysis of the selected sequences against the pentapeptide library reveals that each SH2 domain exhibits two different consensus sequences and the two SH2 domains have overlapping but distinctive specificities (Table 4.2 and Figure 4.3). For the N-terminal

SH2 domain, the majority of selected sequences (85 out of 97) belong to class I. which has a consensus sequence of L(Y/H)pY(M/F)X(F/M). This major consensus sequence is in excellent agreement with our earlier screening of the tetrapeptide library (Ac-

LKpYXXXXBBRM) yielding essentially the same consensus, demonstrating the reproducibility of the library method. A striking feature of this SH2 domain is that the strongest selectivity is at the -2 position, with a leucine being the most frequently observed residue; even norleucine and isoleucine, which are structurally similar to leucine, are only found sparingly at this position. The SH2 domain also exhibits strong selectivity for residues C-terminal to the pY. At position +1. norleucine is most preferred, followed by phenylalanine and leucine; whereas at position +3. phenylalanine is the most preferred residue followed by norleucine. leucine, and tyrosine (Figure 4.3). Some specificity is also observed at position -I. where tyrosine and histidine are frequently selected. There is little selectivity at the +2 position. A small number of the selected sequences (12 out of 97) clearly belong to a different class, with a consensus sequence of

LXpYAXL (class H) (Table 4.2.1). Like in class I peptides, a leucine is the most frequently selected residue at the -2 position. On the C-terminal side, alanine and leucine are almost invariant at the +1 and -t-3 positions, respectively.

The C-terminal SH2 domain also selected two distinct classes of sequences (Table

4.2.2). Class I peptides (total 44) exhibit a consensus of (V/I/L)XpYAX(L/V). in which alanine is almost exclusively selected at the +1 position, whereas leucine is the most

82 preferred amino acid at the +3 position (Figure 4.3). This SH2 domain also shows selectivity at other positions. It prefers a hydrophobic residue (e.g.. valine, isoleucine, leucine, or tyrosine) at position -2, a hydrophilic residue (e.g., histidine, arginine, or glutamine) at position -1. and a positively charged residue (e.g.. arginine or lysine) at position +2. The class II peptides (total 68) have a consensus sequence of

.XXpYYM(K/R) (Table 4.2.2). In contrast to class 1 sequences, the most selective position is actually at position +2. where norleucine is exclusively selected. .A. tyrosine is strongly preferred at position +1. with phenylalanine being the only acceptable substitution. .A positively charged residue, lysine or arginine, is highly preferred at position -i-3. Essentially all of the selected sequences have the consensus sequence at the

C-terminal side: they only differ by their N-terminal sequences. Note that the class 1 sequences of the C-terminal SH2 domain closely resemble the class II sequences of the

N-terminal SH2 domain.

4.3.3 Affinity Measurements of Selected Sequences

Several representative peptides selected from the two libraries were re-synthesized individually and tested for binding to the SH2 domains of SHP-1 using the surface plasmon resonance technique (BIAcore) (Table 4.3). Immobilization of the pY peptides onto streptavidin-coated sensor chips was effected by the addition of a biotin to the N- termini of these peptides during solid-phase synthesis. To maximize the accessibility of the surface bound peptides to an incoming SH2 domain, a flexible linker of Ala-(3-Ala-|3-

Ala-P'Ala (ABBB) is added in between the biotin and the N-terminus of a pY peptide

(e.g., biotin-ABBBIYpYANLI). Initially, we attempted to titrate the surface directly with

83 MBP-SH2(N) as carried out in previous work (Pei et al., 1994; Pei et al.. 1996).

Although we obtained data for some of the pY peptides (not shown), nonspecific binding of MBP to the chip surface became a problem, particularly at high concentrations (e.g..

50 (iM). Therefore, we performed all of the binding experiments with glutathione-S- transferase fusion proteins (GST-SH2). which had much less nonspecific binding to the chip surface under our assay conditions. However. GST is known to dimerize resulting in the measurement of avidity instead of affinity (Ladbury et al.. 1995). To minimize the avidity effect, we used the streptavidin-coated sensor chip that avoids overloading the surface with pY peptides. We also used two known ligands of SHP-1 SH2 domains for comparison: peptide LKpYLYLV of erythropoietin receptor (EpoR) known to bind to the

N-terminal SH2 domain of SHP-1 {Ko = 1.8 pM for Ac-PHLBCpYLYLVVSDK) (Pei et a i. 1994: Pei el al., 1996) and peptide ITpYSLLK of B cell Fc receptor known to bind to the C-terminal SH2 domain (Ko ~ 2.8 pM for EAENTITpYSLLKH) (D'.Ambrosio et ai.

1996: Pei et ai. 1996). Our data in Table 4.3 demonstrates that the selected peptides are comparable to the peptides derived from the natural receptor sequences, thus validating the selection.

Figure 4.5A shows the BIAcore sensograms for the binding of GST-SH2(N) to immobilized peptide lYpYANLI, a class II peptide selected against the N-terminal SH2 domain; flow of increasing concentrations of the SH2 protein (0.4-25 pM) over the chip resulted in increasing and eventually saturating equilibrium response units ( RU^q = response at 12 s after the end o f injection). Plot of the RUeq values against SH2 concentration clearly exhibited a pattern of saturation and data fitting gave an equilibrium dissociation constant (Æo) of 0.60 pM (Figure 4.5B). This peptide also binds to the C-

84 terminal SH2 domain, although with slightly lower affinity (Ko = 1.4 gM) (Table 4.3). consistent with the observed overlapping specificity of the two SH2 domains. Likewise, another class II peptide selected for the N-terminal SH2 domain. LYpY.A.NLl. binds to the N- and C-terminal SH2 domains with Kq values o f 0.65 and 2.0 pM. respectively.

Peptide VHpYARLI. which was selected for the C-terminal SH2 domain, also binds both

SH2 domains but with slightly higher affinity for the C-terminal SH2 domain (Kd = 4.9 and 1.8 pM for the N- and C-terminal SH2 domains, respectively). .A. class I peptide of the N-terminal SH2 domain, LKpYMQMF (which was selected from the initial tetrapeptide library), binds to the N-terminal SH2 domain with a A.'d value o f 2.4 p M .

Thus, all of the selected peptides we have tested bind to SHP-1 SH2 domains with similar affinity to the pY peptides derived from known SHP-1 partner proteins. By all means employed, binding of the C-terminal SH2 domain to the class II peptides (e.g..

IQpYYMKI. IQpYY(Nle)KI. VSpYYMRI) failed to be observed.

4.3.4 Stimulation of SHP-1 Activity by pY Peptides

To further characterize the selected pY sequences, we have examined their ability to stimulate the catalytic activity of SHP-1. It has previously been established that the N- terminal SH2 domain of SHP-1 autoinhibits its phosphatase domain by forming a noncovalent intramolecular SH2-PTP complex (Pei et a i. 1994). Binding of a pY peptide to the N-terminal SH2 domain disengages the intramolecular complex and stimulates the enzymatic activity by -30 fold (Pei et al., 1994; Pei et al.. 1996). There is a general correlation between the binding affinity of a pY peptide to the N-terminal SH2 domain and its ability to stimulate the enzymatic activity, thus providing a simple method to

85 screen pY peptides for binding to the N-terminal SH2 domain. For fair comparison, we synthesized a shorter version of EpoR pY429 peptide. LKpYLYLV. as our benchmark.

As shown in Figure 4.5A, all of the peptides selected against the N-terminal SH2 domain

Peptide^ GST-SH2(C) GST-SH2(N)

LKpYMQMF ND 2 .4 ± 0 .4

LKpYLYLV ND 2.7 ± 0 .4

VHpYARLI 1.8 ± 0.2 4 .9 ± 0 . "

ITpYSLLK 1.2 ± 0.2 ND

LYpYANLI 2.0 ± 0.1 0.65 ± 0.05

lYpYANLI 1.4 t 0.1 0.60 ± 0.06

VYpYANLI 1 .9 r 0 .1 1 .1 z 0 .1 AYpYANLI >50 >50 “Biotin-ABBB (B= (3-Ala) was attached to the N- terminus of the above peptides. ND. not determined.

Table 4.4 Dissociation Constants (K d , M-M) of Selected Peptides

exhibited potent stimulation of SHP-1 in a concentration dependent manner. In tact, the selected peptides are more potent than the EpoR peptide. These results again suggest that the selected pY peptides are capable of binding to the N-terminal SH2 domain with relevant potencies. We have avoided making any quantitative comparison in the binding affinities of these peptides based on the stimulation data, the magnitude of which is

8 6 complicated by the differential ability of these peptides to inhibit the phosphatase activity during the stimulation assays (through competition with the assay substrate p-nitrophenyl phosphate).

Since the N-terminal SH2 domain strongly selects for a leucine at the -2 position, we examined the importance of the -2 residue in binding to the SH2 domain by replacing the leucine in LYpYANLI with an iso leucine, valine, or alanine. .Among the resulting peptides. lYpYANLl (which is one of the selected class II peptides) has essentially the same binding affinity to the parent peptide {Ko = 0.60 vs 0.65 pM) (Table 4.3).

Substitution of a valine for the leucine residue resulted in a 2-fcld reduction in binding affinity (Ko = 1.1 pM for VYpYANLI). However, substitution of an alanine at the -2 position drastically reduced the binding affinity {Kq > 50 pM for AYpY.ANLl).

Consistent with the BIAcore results, peptides LYpYANLI. lYpYANLI. and VYpY.ANLI exhibited similar potencies in stimulating SHP-1. whereas peptide .AYpYANLI was > 10- fold less active (Figure 4.5B). The effect of a Leu->Ala mutation at the -2 position was even greater for the class I peptides. For example, peptide AYpYMRFl was nearly two orders of magnitude less potent in stimulating SHP-1 than LYpYMRFl (Figure 4.5B).

These results demonstrate the importance of the -2 residue in binding to the N-terminal

SH2 domain and validate the strong selection for a leucine (or norleucine. isoleucine) during library screening. It should be noted that a control peptide (not selected from the library), LKpYDHHR, resulted in no stimulation of SHP-1 at concentrations up to 500 pM (Figure 4.5B). This result suggests that the pY residue and a leucine at position -2 are

necessary but not sufficient for high-affinity binding to the N-terminal SH2 domain of

SHP-1.

87 The contribution of the -2 residue for binding to the C-terminal SH2 domain was also examined. While peptides VHpYARLI, LYpYANLI, lYpYANLL and VYpYANLI all bind the SH2 domain with very similar Ku values (~2 pM), peptide AYpYANLI showed much weaker affinity (/fo >50 pM) (Table 4.3). This demonstrates that the C- terminal SH2 domain also requires a hydrophobic residue at the -2 position for high- affinity binding, although the precise nature of the side chain is less critical than for the

N-terminal SH2 domain. Note that 80% of the class 1 sequences selected against the C- terminal SH2 domain contained valine, isoleucine, leucine, tyrosine, threonine, proline, or phenylalanine at the -2 position, whereas none of the sequences contained alanine, aspartic acid, asparagine, glutamic acid, glycine, or lysine at this position (Table 4.2.2 and Figure 4.2).

We have noted that the residues beyond the +4 position may also contribute to the overall binding affinity, at least for the binding of some peptides to the N-terminal SH2 domain of SHP-1. There was a ~5-fold reduction in the stimulatory activity when the

EpoR pY429 peptide Ac-LKpYLYLVVSR was shortened to Ac-LKpYLYLV (Figure

4.5C).

4.4 Discussion

The dual SH2 domain-containing phosphatase SHP-1 and its close relative SHP-2 belong to an intriguing subfamily of non-transmembrane protein tyrosine phosphatases.

Both SHP-1 and SHP-2 have been shown to actively participate in many signaling pathways, acting both as positive and negative regulators (reviewed in (Neel and Tonks.

1997)). Their SH2 domains play essential roles in these signaling events by binding to

8 8 and thereby recruiting the catalytic activity to different tyrosyl phosphorylated receptors.

Although the pY motifs on the receptors that are responsible for SHP-1 and SHP-2 binding have been determined in many cases, the precise mode of interaction (e.g.. which

SH2 is responsible for binding?) is less clear. In this work, we have carried out a systematic evaluation of the binding specificity for both SH2 domains of SHP-1. Such information will greatly facilitate the identification of the cognate binding site(s) on the known receptors for each SH2 domain of SHP-1 as well as new target proteins of SHP-1.

A surprising finding of this work is that, for the N-terminal SH2 domain, the -2 position is most critical for high-affinity binding. Out of the 97 sequences determined. 65

(67%) contained a leucine at position -2; the rest of the sequences contained the highly homologous norleucine. isoleucine, or valine at this position (Figure 4.3). Substitution of an alanine for the -2 leucine resulted in very poor binding sequences (ATp >50 pM for

.AYpY^lRFI and AYpYANLI) (Figure 4.4B and Table 4.3). On the basis of the screening results (Figure 4.3), the BIAcore binding data (Table 4.3). and literature data (vide infra). we propose the following order of preference at the -2 position by the N-terminal SH2 domain of SHP-l; leucine, iso leucine, norleucine > valine » other amino acids. The C- terminal SH2 domain also exhibits strong preference for a hydrophobic residue at the -2 position, although the precise nature of the side chain is less critical than the N-terminal

SH2 domain. Data from this work as well as literature reports suggest the following order of preference: valine, isoleucine, leucine > norleucine. tyrosine, threonine, serine. histidine, proline, phenylalanine » alanine or hydrophilic amino acids. The importance of the -2 residue for binding to the SH2 domains of SHP-1 and SHP-2 has previously been noted by others, through comparison of their recognition pY motifs on various

89 receptors and mutagenesis studies (Burshtyn el al., 1997; Huyer and Ramachandran.

1998; Burshtyn et al., 1999). Our current results are in excellent agreement with these literature data. Since most SH2 domains of known specificity show either no or little selectivity on the N-terminal side of pY (Songyang et al., 1993; Songyang et al.. 1994). the SH2 domains of SHP-1 and SHP-2 represent a novel class of SH2 domains whose binding specificity is governed by both N- and C-terminal residues.

.Another interesting finding is that the N-terminal SH2 domain of SHP-1 is capable of recognizing peptides of two distinct consensus sequences on the C-terminal side of pY [LXpY(M/F)X(F/M) vs LXpYAXL]. The selected peptides trom both classes bind to the N-terminal SH2 domain with similar affinity and effectively stimulate the catalytic activity of SHP-1. Presumably, the two types of sequences bind to the N- terminal SH2 domain in different modes. It is worth noting that among the class 1 sequences, norleucine and phenylalanine are interchangeable at the -+-1 and +3 positions

(i.e.. pYMXF. pYNIXM. pYFXM. and pYFXF were all selected by the SH2 domain), although there is slight preference for norleucine at the +1 position and phenylalanine at the -^3 position. We have previously found through library screening that for peptide deformylase. whose physiological substrates are N-formylmethionyl peptides. N- formylnorleucyl and N-formylphenylalanyl peptides are the only alternative substrates of significant deformylation reaction (Hu et al., 1999). The X-ray crystal structure of a deformylase-inhibitor complex demonstrated that the n-butyl side chain of norleucine adopts a bent conformation that effectively mimics the phenyl ring of a phenylalanine

{Hao et al., 1999).

90 The C-terminal SH2 domain also selected peptides of two distinct consensus sequences. (V/I/L)XpYAX(L/V) (class 1) and XXpY(Y/F)M(K/R) (class 11).

Interestingly, the class I consensus sequence, (V/I/L)XpYAX(L/V). overlaps with the class II consensus sequence of the N-terminal SH2 domain (LXpYAXL). However, due to its narrower specificity at positions -2 and +3. the N-terminal SH2 domain will likely bind only a subset of the SH2(C)-binding sequences (see below for more discussion). The reason for the selection of class II sequences for SH2(C) is not yet clear. However, they are unlikely a result of nonspecific binding, as they did not show up during library selection against the N-terminal SH2 domain or control selections (no SH2 domain).

Since free SH2(C) was used in selection, these sequences may have been selected against a population of biotinylated SH2 domain whose pY peptide-binding site was perturbed by biotin modification.

91 Motif Protein R e f . M o t i f Protein Ref.

VTpYAQL PIRB ( B l é r y e c V T p Y A Q L I L T 3 S V p Y A T L al., 1 9 9 8 ) S V p Y . A T L a l . , 1 9 9 - ; S L p Y A S V V T p Y A K V

S I p Y S T L * XAEA (Philosof- IVpY.AQV gp4 9B1 (Kurciwa ec

Oppenheimer VTpY.AQL a 2., 1993

e c a i . , 2 0 0 0 )

IVpYASL'- = P l L R a -T p iSi-i.*" EC/RIIB (D’.Ambrosio TLpYSVL' ec al., e c al., 1 9 9 6 )

V T p Y . A Q L

I H p Y S E L ' C D 2 2 ( L a w e c I T p Y A A V V S p Y A I L al., 1 9 9 6 ; V D p Y V T L Blasioli ec al., 1999) ITpY.ADL L T p Y . A D L

; Q p Ï T t. / P E C A i M (Pumphrey TEpY.ASI r V p Y S E V e c al., S E p Y . A S V 1 9 9 9 ;

VApYTVL Hucer e z l E p Y L C L ' : l - 3 r (Bone ec TVpYSEV -2 - . , . T" o' al., 199"'.

L K p Y L Y L E p o R 2 3 3 3 , T a y - o r e c r -r “ i ^ , T E p Y S E V al., 1 9 9 9 )

V T p Y A Q L I L T 2 (Colonna ec VTpYAQL'" = XIR S I p Y ' A T L al., 1 9 9 7 ) * C i •

A E p Y L R V EGER ( X e i l h a c . k e c al., 1 9 9 8 ) I T p Y . A D L .“.Cl 3. c r. c? z

V T p Y S T V L y - 4 9 ( O l c e s e e c al., 1 9 9 7 )

I V p Y A S L E D E 0 3 I L p Y A E L 5 s c 2 — R (Liu ec TLpYSVL e c a i . , al., 2 0 0 0 )

'Sequences reported to bind the SH2(C) domain. Sequences reported to bind to the SH2(N) domain

Table 4.5 Reported SHP-1 Binding Sites

92 Our library results are in agreement with the literature data on the recognition motifs of SHP-l. Table 4.4 lists some of the pY motifs found in various receptors that have been implicated in SHP-1 binding. These motifs clearly show a consensus sequence of (V/I/T/L)XpYAX(L/V) (Figure 4.6). At the +1 position, alanine is most frequently observed; the sterically similar serine and, much less frequently threonine and valine, are the only allowed substitutions at this position, in excellent agreement with the library selection results (Figure 4.3). For the majority of these pY motifs, it is not yet clear which of the SH2 domains of SHP-l is responsible for binding. On the basis of our library data,

we predict that all of the peptides in Table 4.4 should be capable of binding the C-

terminal SH2 domain, with the exception of the pY motifs from EpoR (LKpYLYL)

(Klingmuller et a/.. 1995) and IL-3 receptor (LEpYLCL) (Bone et al.. 1997). which

belong to the class 1 sequences of the N-terminal SH2 domain and are indeed known to

bind to the N-terminal SH2 domain (Pei et al.. 1994; Klingmuller et al.. 1995). We

further predict that the subset of sequences in Table 4.4 that contain a leucine or

isoleucine at the -2 position (e.g.. ITpYSLL of Fr/RIIB. IHpYSEL of CD22. LHpY.ASL

of CD33. IVpYASL of PlLRa. IVpYAQV o f gp49Bl. ITpYAAV of LAIR-1, and

(LT)TpYADL of SHPS-l) and, perhaps with lower affinity, the motifs with a valine at

position -2 (e.g.. VTpYAQL) should also be capable of binding the N-terminal SH2

domain. Our predictions have turned out to be correct in all cases for which binding data

are available. For example, peptides ITpYSLL and IVpYASL. the only peptides whose

Ko values have been determined, bind effectively to both SH2 domains of SHP-l (Pei et

al.. 1994; D'Ambrosio et al.. 1996; Mousseau et al.. 2000). Peptide LHpYASL of CD33

has also been reported to bind to both SH2 domains of SHP-1 (Taylor et al.. 1999).

93 Peptide IHpYSEL o f CD22 is known to bind the N-terminal SH2 and stimulate SHP-1 activity (no data yet available for its interaction with the C-terminal SH2 domain) (Law ei al.. 1996; Burshtyn el ai.. 1997; Blasioli el ai.. 1999). The killer cell inhibitor receptor

(KiR) motif. VTpYAQL, showed detectable binding with the isolated SH2(C) domain but not with the SH2fN) domain under similar conditions (Burshtyn ei ai.. 1996).

However, this peptide exhibits potent stimulation of SHP-1 activity in a concentration dependent manner (Burshtyn et ai.. 1997). indicating that it is capable of binding to the

N-terminal SH2 domain at higher peptide concentrations. This is in keeping with our prediction that pYAXL peptides containing a valine at position -2 will bind the C- terminal SH2 domain more effectively than the N-terminal SH2 domain of SHP-1.

Peptide TLpYSVL derived from PlLRa. which does not have Leu. lie. or Val at position

-2. binds only the C-terminal but not the N-terminal SH2 domain of SHP-1 (Mousseau ei ai.. 2000). as we have predicted. The excellent agreement between our library results and the literature data clearly demonstrates the validity of our library method.

Two other notable points have been borne out of this work. First, some of the class 1 peptides selected against the N-terminal SH2 domain (e.g. the pYMXM subset) bear close resemblance to the recognition motifs for the SH2 domains in the p85 subunit of phosphoinositide-3-kinase (Piccione et ai.. 1993). However, upon examination of the pYMXM motifs on PDGF receptor (DGGpY^‘'°MDMSK) (Piccione et ai.. 1993:

Shoelson et ai.. 1993; Carpenter et ai.. 1998), insulin receptor substrate-1

(NGDpY^-^MPMSP, PNGpY^'^MMMSP, TGDpY^^MNMSP. and SEEpY^'^MNMDL)

(Felder et ai., 1993; Piccione et ai., 1993; Shoelson et ai., 1993), and mT antigens

94 (ENEpY“^*MPMAP from hamster and EEEpY^'^MPMED from mouse) (Shoelson ei al..

1993; Carpenter et a i, 1998), which are known to bind to p85 SH2 domains but not to

SHP-1. we found that none of these motifs had a leucine or isoleucine at the -2 position.

Presumably, these proteins have specifically avoided these hydrophobic residues at position -2 to prevent recruitment of SHP-1. which would prematurely terminate the signaling cascade. It remains to be seen whether the (L/I)XpYMXM motif exists in any proteins, which would be predicted to recruit both SHP-1 and phosphoinositide-3-kinase but upon performing a pattern match search for this motif, several potential candidates emerged (Table 4.6). From the FIR database. 2.437 sequences were identified which matched this query out of which, 64 were human proteins. Listed in table 4.6 are 42 of the 64 proteins and their sequences obtained in the search. Excluded in the table are transcription factors, hypothetical proteins, mitochondrial proteins and extracellular proteins. Whether any of these are actual binding sites for SHP-1 is yet to be determined but many are possess features one would expect in a partner protein of being transmembrane (TSPAN-5). hematopoietic in origin (HEM-1), and a bona fide receptor

(KAI1[CD82]). Second, our results have convincingly shown that the most frequently selected sequences do not necessarily have the highest binding affinity. For example, selection against the N-terminal SH2 domain resulted in 7 times more class 1 peptides

(total 85) than class II peptides (total 12) and yet. our limited binding studies seem to suggest that the class II peptides are slightly tighter binders than class I peptides (Table

4.3). Similarly, while out of the 12 class II peptides for the N-terminal SH2 domain, nine contain a leucine and only one has isoleucine at position -2, peptides lYpYANLI and

95 LYpYANLl bind to the SH2 domain with equal affinity. This could be due to one or a combination o f factors with one potentially resulting from the inherent bias of this library strategy.

96 Protein Sequence Reference

H E M - 1 I M p Y M N F (Baumgartner et ai., 1555;

S y k I N p Y F K F (Rowley ec al., 1 9 9 5 ;

Z23~: I K p Y F I F (Classon ec al., 1 9 9 C :

Histamine HI-R I P p Y F I F (Akaishi ec al., 2 G 0 C ;

PTP-3AS VSpYFRM (Cuppen ec al., 1 9 9 ~ ;

Che.mckine ?. IVpYFLF (Samson ec al., 1 9 9 6 ;

Thicredcxin-iike I 3 Y F T F (M iranda-Vicuete ec al., crstein - human 1 9 9 8 )

Calcium-binaing protein LHYFKM ( D e k a e c al., 1 9 S S :

Ma/taurochclace

tot ranspcrting LLYMIF (Hagenbuch and M eier, 19 94 polypeptide

Lamin 3 receptor IIYFTM (Ye and Wcrm.an, 1994

1 M etastasis suppressor IFYFMM ' W h i t e e c al., 1 9 9 5 ' r * ! . " . I L

-iacylgiycerol kinase LWYFEF (Schaap ec a i . , 1 9 9 G '

Vasoactive intestinal VHYMVF (Svoboda ec a i . , 1 9 9 4 ' peptide receptor 2

Cholesteryi ester MLYFV/F transfer protein (D rayna ec ai . , 1 9 5 "

97 Table 4.6 (continued)

cGM P-gaced cation LKYMAF ( C h e n e c a i . , 1 5 9 3 ,

c h a n n e l VKYMNF

S c e r c l 0 - VPYFLF ( C h a n g e c a i . , 1 9 5 3 ; acyitransferase

.-.cyloxyacyi hydrolase L F Y M D F ( H a g e n e c a i . , 1 9 9 1 '

?Tr-nunreceptor type 13 V S Y F R H 1 B anville ec a i . , 1 9 9 4

. - . r ç i n a s e I K Y F S M (Banville ec a i . , 1 9 9 4

Tryctophan-cRNA ligase L D Y M G M (Fleckner ec a i . , 1 9 9 1

Tra.nsforming protein MVYFEM (Nomura ec a l . , 1 9 9 9

s n c - . A

iytckine R P-chain VDYFSF (.nayasnica et

1 9 9 0 !

' M ultidrug R protein 1 LRYMVF C h e n e c a l . , i ! ■.'c it a ce c ec e n d en t Ca IRYFEM (W illiams ec

: cnannei a lE-1

O r p h a n j D P - 7 L Y F D F (Jin ec a i., cicc'crcnosyl- VLYFEF

transferase

( K o i k e e c a i . , 1 9 9 9 ' 3tn LYYFGM

Prccacle transmembrane ( L i s s y e c a i . , 1 9 9 6 ' protein TMC

5',5'-cyclic nucleotide IMYMIF (Horton ec a i . , K phosphodi-esterase- H S P D E 4 A 7

I.ncsitol-polyphoshate VSYFVF ;Laxm inarayar 3-cchsDhate 1 9 9 4 )

Table 4.6 (continued)

98 Table 4.6 (continued)

"ihyarc-pyri.m idase LVYMAF (Hamai ima et al., 1 à Ac relared proceia 2

■Jbiquicin protein LSYETF {Morrione ec a i . , 1 A 9 A ' l i g a s e

.-.cetyl-CoA carboxylase VRYMVM ( H a e c a i . , 1 9 9 4 )

Ar.gioter.sin II R VIYFYM ( M a u z y e c ai., 19 92'

Ir.Gsitol triphosphate R LIYFGF '(M arante, 1 9 9 4 t y p e 1

(Beeson ec a i . , 1 9 5 9 Nicotinic acetylcholine VTYFPF R b e t a

Taurine transport IPYFIF 'Ramiamocrthy ec a 1. , p r c t i e n 1 9 9 4 :

R e t a i n e ' T . A 5 . A t r a n s p o r t r" t ?* .L r (Rascia ec a i . , 1 9 9 :

I n t e g r i n a - E c h a i n LQYFGM ( S h a w e c a l . , 1 9 9 4

Ilutamate transporter VKYFSF iShashidharan and Plaita:

Excitatory ami.no acid IKYFSF (Fairm an ec a i . , 1 9 9 : transporter 4

•Amino acid transporter LIYFVF (Shafqat ec a 1., 1 9 9 : S.ATT IVYMLF

■Adenosine R 1 ( R e n a n d Stiles, 1994 MVYFMF

Table 4.6 Proteins identified through database search. I/M/L/V-Y-M/F-X-M/F was searched against the FIR protein database by using a pattern match search. Abbreviations: Ptn (protein), R (receptor).

99 During the library encoding by partial chain termination, the given coupling efficiency for a given residue to a specific sequence combination relative to the coupling efficiency o f the capping reagent will be different in most cases. The end result of this differential reactivity is that some beads will contain more chain termination product than others while some will have more of the full length peptide. Although it is true that ever\ peptide in the library will be represented by this technique, they may not exist in equimolar quantities. For instance the greater representation of Leu compared to lie could be due to the fact that lie containing peptides were present during the screening at slightly lower concentrations due to greater relative coupling efficiency of the capping reagent. The other reason that certain sequences may be selected more frequently, despite not being the highest affinity binders is due to a tight-binding class having a more rigid sequence requirement across all five positions. Indeed in comparing class I to class

II SH2(N) ligands, class 1 ligands are represented most frequently but they have much more variability at the -2. +1, and +2 positions. The class II sequences have a much more rigid requirement at these positions. Assuming 3 of the 5 positions are fixed in one case that leaves 19‘ or 361 possible sequences. If as in the case of class I peptides only one of the positions is rigid, that leaves many more peptides to potentially select.

Missing these higher affinity underrepresented classes is another strong argument for obtaining individual sequence information when possible rather than subjecting selected libraries to pool sequencing approaches.

In summary, we have developed a new combinatorial library methodolog} for rapid and systematic evaluation of the binding specificity of protein modules such as SH2 domains. This method offers several advantages over the previously reported methods. too First, this method gives individual sequences as well as a consensus sequence(s). Second, because selection is based exclusively on afFinity, any high-affinity sequence is

identified, even if it is present in the library at only minute amounts. This is not true for the methods that employ pool sequencing (Songyang et al., 1993; Muller et at.. 1996).

Previous work has predicted a consensus of pYFXF for the N-terminal SH2 domain of

SHP-1. which resembles the class I consensus from this work, but failed to reveal the

minor class II sequences (Songyang et a i. 1994). Finally, our method is rapid and economical. Microsequencing of a large number of beads was not possible in the past due

to the prohibitive cost in resources and time associated with the conventional Edman

method. By using MALDI-TOF mass spectrometry, however, one can now routinely sequence the peptide from a single resin bead in less than one minute and at a cost of less

than five US dollars. This method should be readily applicable to other protein modules

and the specificity data obtained will greatly facilitate the identification of the

physiological partner proteins of these modular domains.

101 MRBBIPALpYYLED-Ac

3000 MRBBcap MRBBXicap MRBBXiXacap 25 0 0 -, MRBBX-iXzXscap MRBBXiXzXapYcap MRBBXiXaXapYXAcap 2000 MRBBXiX2X3pYX4Xscap MRBBX1X2X3PYX4X5E-AC MRBBX1X2X3PYX4XSED-AC 1500

1000

A L pY Y L + E D 500 - e— V < ------► ■<------■<-----

600 800 1000 1200 1400 1600 M/Z

Figure 4. 1 . MALDI-mass spectrometry of typical peptide mixture derived from selected TentaGel S NHi bead carrying library Ac-DEXXpYXXXIBBRM (B=P- .A.ia).

102 80 60 +1 40 20 0 LARDNEQHGIKMFPSTWYV 80 60 +2 40 8 C 20 2 0 8 LARDNEQHGIKMFPSTWYV CLa co 80 o +3 c 60 40 20 0 I.. Il .. LARDNEQHGIKMFPSTWYV 80 60 +4 40 20 ..I ill M ! ■ LARDNEQHGI KMFPSTWYV

Figure 4.2. Specificity, as determined by the screening of Ac-LKpYXXXXBBRM library, of the N-terminal SH2 domains of SHP-1. Displayed are the amino acids identified at each position from +1 to +4 relative to pY (position 0). Abundance on the y- axis represents the number of selected sequences that contain a particular amino acid at a certain position [out of a total of 136].

103 SH2(N) SH2(C) 70 50 60 -2 40 -2 50 40 30 30 20 20 10 10 ■ _ L 0 - 0 J__ LARDN EQHG I KMFPSTWYV LARDNEQHGIKMFPSTWYV

70 - 50 60 • 40 50 ■ 30 40 • 30 20 20 10 10 0 0 LARDNEQHGIKMFPSTWYV

30 '

20 ■

10 '

0 ■ LARDNEQHGIKMFPSTWYV LARDNEQHGIKMFPSTWYV

50 - 50 40 ■ +2 40 30 ■ 30 20 ■ 20

10 ■ 10

0 - JL 0 - J .1 ______LARDNEQHGIKMFPSTWYV LARDNEQHGIKMFPSTWYV

♦3

■___ ■_BB—M——bb LARDNEQHGIKMFPSTWYV LARDNEQHGIKMFPSTWYV

Figure 4.3. Specificity of the N- and C-terminal SH2 domains of SHP-1 as determined by screening Ac-EDXXpYXXXIBBRM. Displayed are the amino acids identified at each position from -2 to +3 relative to pY (position 0). Abundance on the y axis represents the number of selected sequences that contain a particular amino acid at a certain position [out of a total of 97 for the N-SH2 domain and 44 (class I) for the C-SH2 domain].

104 4 0 0 0 R U eq

3 0 0 0 D a: 2000

1000

0

-50 5 0 100 1 5 0 2 0 0 T im e (s) 2 5 0 0

2000

Sr 1 5 0 0

1000

5 0 0

10 15 2 0 25 [GST-SH2(N)], nlVI

Figure 4.4. BIAcore analysis of the binding of GST-SH2(N) to peptide lYpYANLI. (A) Overlaid sensograms at increasing concentrations (0.39. 0.78, 1.56, 3.12. 6.25. 12.5. and 25 pM) of GST-SH2(N). (B) Plot of resonance signal under the equilibrium binding conditions (12 s after the end of injection) against SH2 concentration. Data were tmed to equation; RU^q = RUn,a.x* [SH2]/(/Cd + [SH2]).

105 1.0

LYpYANLI LYpYANLI 0.8 lYpYANLI lYpYAN LI VHpYARLI Vi'pYANLI LKpYMRFF AYpYANLI LKpYLRMF 8 LYpYMRFl Z- - - - LYpYMRFl AYpYMRFl I LKpYMOMF LKpYDHHR LKpYLYLV

0.1 10 100 1000 0.1 1 10 100 1000 [pY Pepdde], pM [pY Pepddel, pM

0.6

0.5 LKpYLYLVVSR LKpYLYLV ^ 0.4

§; 0-3 I I 0.2

0.0

0.1 10 100 1CC0 [pY Peptide]. pM

Figure 4.5. Activation o f SHP-l by various pY peptides. (A) Comparison of several selected peptides with the EpoR pY429 peptide (LKpYLYLV). Peptides not shown in Table I were selected from the tetrapeptide library LKpYXXXX. (B) Effect of the -2 residue on SH2(N) binding. (C) Effect of residues beyond +4 position on SH2(N) binding.

1 0 6 30 - 25 -

20 • 15 -

10 • 5 - 0 - I Il 1 LARDNEQHGIKMFPSTWYV 30 - 25 -1 20 15 10 5 1 LARDNEQHGIKMFPSTWYV

30

25 -I 20 15 10 5 1 LARDNEQHGIKMFPSTWYV

30 25 20 15 10 5 ■ l l u . 1 LARDNEQHGIKMFPSTWYV 30 25 20 15 10 5 0 1 LARDNEQHGIKMFPSTWYV

Figure 4.6. Analysis of the pY motifs from receptors that are known to bind SHP-1 from table 4.5. Displayed are the amino acids identified at each position from -2 to -i-3 relative to pY (position 0). Number of appearance on the y-axis represents the number of known and nimiitativp SHP-1 hindino çf»nn«»nrp<: that rnntain a nartirniar amtnn arid at a rertain

107 APPENDIX

Part I. Protein degradation by peptide targeting to Tsp

Introduction

Tsp protease has broad specificity (proteoiytically) for a peptide substrate and even, to some degree, for the C-terminal residues. Despite high stringency for apolar residues and a free carboxy terminus. Tsp does not exhibit a strict preference for a

specific combination of residues (Keiler and Sauer. 1996). Although Tsp is not essential

for bacterial survival and therefore not an antibiotic target, we hypothesized that the

activity of Tsp and its cytoplasmic counterpart. ClpXP (Gottesman et a i. 1998). could be

directed against the cell by targeting small nonpolar recognition signals in the tbrm of the

C-terminal peptide sequences to essential bacterial enzymes or proteins. .Although

classically an essential enzyme or protein is targeted directly by conventional inhibitor

design, targeting these proteins indirectly by using the cells own machinery may provide

another possible means by which to destroy a bacterial cell. Since Tsp appears to only

have a substrate requirement for apolar or hydrophobic residues and a free carboxy-

terminus. it is conceivable that it may accept certain non-peptide molecules that mimic

this sequence chemically and therefore still serve to recruit Tsp to a given site.

Additionally, Tsp has broad specificity in terms of its proteolytic targets seeming to

108 require only the C-terminal PDZ recognition motif. This is evidenced by studies on protein substrates that demonstrate a broad range of cleavage sites both spatially and sequentially in a number of protein substrates (Keiler et al.. 1995) and also the ability of

Tsp to bypass unnatural elements in the form of the substrate detailed in chapter 2 to initiate proteolysis. With this evidence in mind, it was thought that small C-terminal apolar peptide-like sequences bearing a N-terminal specificity element for an essential protein target could serve to recruit Tsp and the cytoplasmic counterpart ClpXP to the protein target resulting in proteolysis, destruction of the protein, and cell death. To set up a model and demonstrate the feasibility of this approach, we designed means by which to covalently and noncovalently attached C-terminal peptides suitable for recruiting Tsp. to

target proteins, in an attempt to initiate the directed proteolysis of a protein target by Tsp.

Strategy and methods

We synthesized a variety of peptides containing the Tsp recognition element at

the C-terminus and an N-terminal element that would allow for attachment or

derivitization of a target protein. Ideally, the peptides should serve to recruit Tsp to a

target protein while not serving as a substrate for the catalytic portion of the enzyme. It

was with this objective that several short peptides were synthesized for the purpose of

localizing Tsp to a target while themselves not serving as a substrate. In an attempt to

target streptavidin. biotin-GRGWVAAA was synthesized. In an attempt to derivatize

cysteine containing proteins, maleimide activated GGDENYALAA or thiopyridyl-

cysteiny 1-R.GWVAAA were synthesized. All peptides were synthesized by standard

Fmoc chemistry as described in the previous three chapters. All proteins were purified as

109 described in chapters 2-4. Steptavidin was purchased from sigma. The strategy was to attach peptides to the surface of a target protein either covalently to a surface cysteine in the case of staphylococcal nuclease. SHP-1. MBP(N+C)SH2. deformylase or noncovalently in the case of streptavidin. After attaching the peptide by brief incubation, the excess peptide was removed by ion exchange or gel filtration chromatography. Tsp was added to the protein in a slight molar excess and incubated to allow proteolysis to occur, and finally the reactions were quenched by the running of a SDS-PAGE gel to monitor proteolysis.

Results and discussion

We failed to observe proteolysis of any targets tested for a variety of reasons. .All of the experiments lacked a thorough systematic approach as they were carried out when little was known about Tsp and many other projects were on going. However a summar>- of the findings follows. Streptavidin and staphylococcal nuclease are probably too stable for Tsp to effectively proteolyze these targets as proteolysis was not observed for either of the targets. It has previously been demonstrated that Tsp has a limit in terms of its ability to proteolyze stably folded proteins (Keiler et al.. 1995) so both of these targets probably represent a very high stringency test of this system. In the case of SHP-1. deformylase. and MBP(N+C)SH2, the thiol-activated peptides caused the proteins to denature. So. the system could not be thoroughly tested due to the lack of a proper protein target that remained stably folded after derivatization. The derivatized denatured deformylase did demonstrate some susceptibility to cleavage but the results were inconsistent. The two main insights that were gained is that it is unlikely that surface

110 attachment of Tsp specific peptides will result in wholesale proteolysis of stably folded proteins. In order to properly demonstrate the possibility of Tsp targeted proteolysis, a target that is not remarkably stable and that will maintain its tertiary structure after derivitazation should be chosen. The second insight is that the negative results reported here should not be regarded as meaning that the general idea will not work. As pointed out previously, the work lacked a systematic thorough experimentation.

Many possibilities exist aside from simply finding an appropriate target protein.

Since this work was carried out we have learned much about Tsp. Future work should factor in the information we have gained into the Tsp mechanism, keeping in mind that short peptides lacking a cleavage site should be used to direct Tsp to a protein target.

New derivitazation strategies can be examined for adding peptides to their targets. If possible, enzymes whose activity can be monitored should be used, as the gel assay for proteolysis may not be sufficiently sensitive to detect finite peptide directed Tsp proteolysis events.

I ll Part II. Peptide-Inhibitor library selection against protein tyrosine phosphatases

Introduction

The search for protein tyrosine phosphatase (FTP) inhibitors has many challenges that primarily include cellular permeability and specificity (Burke and Zhang, 1998). Our lab has developed PTP inhibitors with very high specificity with regards to dual­ specificity phosphatases and other distantly related phosphatases (Figure .A..1) (Arabaci et al.. 1999). They are >200 fold more potent towards PTPs compared to dual-specificity phosphatases and a K| towards alkaline and acid phosphatases can not be determined due to lack of inhibition even at - 10 mM (Arabaci et al.. 1999). Despite the potency of these inhibitors and relative specificity, they remain rather nonspecific across different PTPs.

Aside from potential therapeutic uses, generating inhibitors with high specificity for use in cell biological studies is of great utility. To improve the potency but more importantly the specificity of these inhibitors, we attempted to graft specificity elements to them in the form of peptides (Figure A.l). These a-bromoacetophenone-based inhibitors are designed to be a pY substrate mimic and therefore, the addition of peptide sequences with specificity for one type of PTP to the inhibitor, should improve their specificity . Since all

PTPs utilize pY. the only potential determinants of specificity could be the surrounding peptide sequence. In an attempt to search for high affinity/specificity inhibitors, we sought to employ our library strategy, described in chapter 4. to this problem by

1 1 2 synthesizing a random tetrapeptide library, attaching the inhibitor to the amino-terminus and selecting against specific PTPs. These types of compounds would be useful leads in the search for fully non-peptide inhibitors.

Strategy and methods

All library synthesis, screening and sequencing methods were described in chapter 4 with the modification that Lys was excluded from the library. Purification and biotinylation of proteins was carried out by the addition of a 1-2 equivalents of N- hydroxysuccinimidobiotin as detailed in chapter 4. Inhibitor 1 and a p-N- hydroxysuccininmide version were synthesized as described previously (.Arabaci et uL.

1999; Guo. 1999). The inhibitor was added to the free N-terminus of the synthesized library by adding 1-2 equivalents o f the p-N-hydro.xysuccinamide version of I in Figure

A.I. The reaction took place in DMF, at room temperature for 30 minutes. .After the reaction, the library beads were washed 3 times with DMF and 20 times deionized water.

The library beads were subsequently blocked as indicated in chapter 4 with buffer containing 0.1% gelatin. Screening was carried out with biotinylated SHP-1 (ASH2) or

PTP IB by adding 5 nM-1 pM enzyme and incubating 10-60 minutes. The resin was washed, inhibitors quenched with excess 2-mercaptoethanol. washed again, and incubated with streptavidin-aikaline phosphatase. The staining with BCIP was carried out as described in chapter 4. All steps were carried out in the dark with only the briefest exposure to light as the inhibitor is light sensitive (Arabaci et al.. 1999).

113 Results and discussion

Initially, strong selection was observed with a discrete number of distinctly blue beads occurring as a result of the screening procedure. A putative consensus sequence was obtained of either Inhibitor-QRRR or Inhibitor-VRRR. However, upon resynthesis and testing of these peptide-inhibitor compounds, poor inhibition properties were exhibited. It was concluded that these represented an artifact in the screening process. library excluding both Arg and Lys was synthesized and screened under the same conditions as above but the strong selection in terms of blue beads was no longer observed. Using this library, excluding Arg and Lys, very poor selectivity was observed as evidenced by the beads tending to be uniformly colored rather than having only a discrete number of distinctly colored beads. To obtain only a small population of colored beads, the enzyme concentration had to be dramatically reduced (5 mVI) and at this level, there was very little difference compared to the control screening reaction (excluding biotinylated PTP).

The conclusion of this work is that there is very little selectivity observed for residues C-terminal to the pY. This is also validated by reported specificity studies in which most of the specificity for PTPs exists at positions N-terminal to the pY (Burke and Zhang, 1998). Assuming our inhibitor is indeed a pY mimic, our lack of selectivity can be rationalized based on the reports in the literature. Further work in this area should include designing an inhibitor analogue that can be added to a peptide library in such a way that it is C-terminal to the library peptides.

114 O . / ^ B r

OH 1

T O R1 0 0 TVV'.'W- 0 R2 0 R4

0 R1 R3

r T T t( I T NHj 0 R2 0 R4

Figure A.I. I. Parent inhibitor characterized previously (Arabaci et al.. 1999). 2. Inhibitor attached to N-terminal peptide. 3. Native pY peptide.

115 REFERENCES

Adachi. T.. Flaswinkel. H.. Yakura, H.. Reth, M. and Tsubata, T. (1998). The B cell surface protein CD72 recruits the tyrosine phosphatase SHP-1 upon tyrosine phosphorylation. J. Immunol.. 160.4662-4665.

.\kaishi. Y.. Hattori. Y., Yoshimoto. K... Kitabatake. A.. Yasuda. K.. and Kanno. M. (2000). Involvement of tyrosine phosphorylation in the positive inotropic effect produced by H( 1 )-receptors with histamine in guinea-pig left atrium. Br. J. Pharmacol.. 130. 907-915.

.Arabaci. G.. Guo. X.-C.. Beebe. K.D.. Coggeshall. K.M. and D.. P. (1999). a- haloacetophenone derivatives as photoreversible covalent inhibitors of protein tyrosine phosphatases. J. Am. Chem. Soc.. 121. 5085-5086.

Bachmair. .A.. Finley. D. and Varshavsky, A. (1986). In vivo half-life of a protein is a ftmction of its amino-terminal residue. Science. 234. 179-186.

Banville. D.. Ahmad. S.. Stocco. R. and Shen. S.H. (1994). A novel protein-tyxosine phosphatase with homology to both the cytoskeletal proteins of the band 4.1 family and junction-associated guanylate kinases [published erratum appears in J Biol Chem 1995 Apr 28;270(17):10359].J. Biol. Chem.. 269. 22320-22327.

Barford. D. and Neel. B.G. (1998). Revealing mechanisms for SH2 domain mediated regulation of the protein tyrosine phosphatase SHP-2. Structure. 6. 249-254.

Baumgartner, S., Martin, D., Chiquet-Ehrismann, R., Sutton, J., Desai, A.. Huang, I.. Kato, K.. and Hromas, R. (1995). The HEM proteins; a novel family of tissue- specific transmembrane proteins expressed from invertebrates through mammals with an essential function in oogenesis. J. Mol. Biol., 251,41-49.

Beeson, D.. Brydson, M. and Newsom-Davis, J. (1989). Nucleotide sequence of human muscle acetylcholine receptor beta-subunit. Nucleic Acids Res.. 17. 4391.

Beebe, K.D. and Pei, D. (1998). A continuous fluorimetric assay for tail-specific protease. Anal. Biochem., 263, 51-56.

116 Beebe. K.D.. Shin, J.. Peng, J., Chaudhury, C., Khera, J. and Pei, D. (2000). Substrate recognition through a PDZ domain in tail-specific protease. Biochemistry 39. 3149-3155.

Blasioli. J.. Paust, S. and Thomas, M.L. (1999). Definition of the sites of interaction between the protein tyrosine phosphatase SHP-1 and CD22. J. Biol. Chem.. 274. 2303-2307.

Bléry. M., fCubagawa. H., Chen, C.C., Vély, F., Cooper, M.D. and Vivier. E. (1998). The paired Ig-like receptor PIR-B is an inhibitory receptor that recruits the protein- tyrosine phosphatase SHP-1. Proc. Natl. Acad. Sci. USA. 95. 2446-2451.

Bone. H.. Dechert. U.. Jirik. P.. Schrader. J.W. and Welham. M.J. (1997). SHPl and SHP2 protein-tyrosine phosphatases associate with betac after interleukin-3- induced receptor tyrosine phosphorylation. Identification of potential binding sites and substrates, y. Biol. Chem.. 272. 14470-14476.

Bork. P.. Schultz. J. and Ponting, C.P. (1997). Cytoplasmic signalling domains: the next generation. Trends Biochem. Sci.. 22. 296-298.

Bradford. M.M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. .Anal. Biochem.. 72. 248-254.

Burke. T.R.J. and Zhang, Z.Y. (1998). Protein-tyrosine phosphatases: structure. mechanism, and inhibitor discovery. Biopolymers. 47. 225-241.

Burshtyn. D.N.. Lam. A.S.. Weston. M.. Gupta. N.. Warmerdam. P.A. and Long. P.O. (1999). Conserved residues amino-terminal of cytoplasmic tyrosines contribute to the SHP-1-mediated inhibitory function of killer cell Ig-like receptors. J. Immunol.. 162. 897-902.

Burshtyn, D.N.. Scharenberg, A.M., Wagtmann, N.. Rajagopalan. S.. Berrada. K.. Yi. T.. Kinet. J.P. and Long, E.O. (1996). Recruitment of tyrosine phosphatase HCP by the killer cell inhibitor receptor. Immunity, 4, 77-85.

Burshtyn. D.N., Yang, W., Yi, T. and Long, E.O. (1997). A novel phosphotyrosine motif with a critical amino acid at position -2 for the SH2 domain-mediated activation of the tyrosine phosphatase SHF- \.J. Biol. Chem., 272. 13066-13072.

Chang, C.C.. Huh, H.Y., Cadigan. K.M. and Chang, T.Y. (1993). Molecular cloning and functional expression of human acyl-coenzyme A:cholesterol acyltransferase cDNA in mutant Chinese hamster ovary cells. J. Biol. Chem.. 268.20747- 20755.

117 Chen, C.J., Clark, D., Ueda, K., Pastan, I., Gottesman, M.M. and Roninson, I.E. (1990). Genomic organization of the human multidrug resistance (MDRl ) gene and origin of P-glycoproteins. J. Biol. Chem.,165, 506-514.

Chen, T.Y., Peng, Y.W., Dhallan, R.S., Ahamed, B., Reed, R.R. and Yau, K.W. (1993). A new subunit of the cyclic nucleotide-gated cation channel in retinal rods. Nature. 362, 764-767. Classon, B.J., Williams, A.F., Willis, A.C., Seed, B. and Stamenkovic. 1. (1990). The primary stmcture of the human leukocyte antigen CD37. a species homologue of the rat MRC OX-44 antigen../ Exp. Med., 172, 1007,

Cuppen, E., Nagata, S.. Wieringa, B. and Hendriks, W. (1997). No evidence for involvement of mouse protein-tyrosine phosphatase-BAS-like Fas-associated phosphatase-1 in Fas-mediated apoptosis. J. Biol. Chem., 272, 30215-30220.

Carpenter. L.R.. Farruggella, T.J., Symes, A., Karow, M.L.. Yancopoulos. G.D. and Stahl, N. (1998). Enhancing leptin response by preventing SH2-containing phosphatase 2 interaction with Ob receptor. Proc. Natl. Acad. Sci. US.4, 95. 6061- 6066.

Celia, M., Dohring, C., Samaridis, J., Dessing, M.. Brockhaus. M.. Lanzavecchia. .A. and Colonna. M. (1997). A novel inhibitory receptor (1LT3) expressed on monocytes, macrophages, and dendritic cells involved in antigen processing. J. Exp. .Vied.. 185, 1743-1751.

Cho. K.O., Hunt, C.A. and Kennedy, M.B. (1992). The rat brain postsynaptic density fraction contains a homolog of the Drosophila discs-large tumor suppressor protein. Neuron. 9, 929-942.

Colonna, M.. Navarro, F., Bellon, T., Llano. M., Garcia, P., Samaridis, J., Angman. L.. Celia. M. and Lopez-Botet, M. (1997). A common inhibitory receptor for major histocompatibility complex class I molecules on human lymphoid and myelomonocytic cells [see comments]. J. Exp. Med, 186. 1809-1818.

Craven, S.E. and Bredt, D.S. (1998). PDZ proteins organize synaptic signaling pathways. Cell. 93. 495-498.

D'Ambrosio, D.. Fong, D.C. and Cambier, J.C. (1996). The SHIP phosphatase becomes associated with Fc gammaRUBl and is tyrosine phosphorylated during 'negative' signaling. Immunol. Lett.. 54, 77-82.

Deka, N., Wong, E., Matera, A.G., Kraft, R., Leinwand, L.A. and Schmid. C.W. (1988). Repetitive nucleotide sequence insertions into a novel calmodulin-related gene and its processed pseudogene. Gene, 71,123-134.

118 Drayna, D.. Jamagin, A.S., McLean. J., Henzel, W., Kohr, W., Fielding, C. and Lawn. R. (1987). Cloning and sequencing of human cholesteryl ester transfer protein cDNA. Nature, 327, 632-634.

Dorman. G. and Prestwich, G.D. (1994). Benzophenone photophores in biochemistry. Biochemistry, 33, 5661-5673.

Doyle. D.A.. Lee. A., Lewis. J., Kim. E.. Sheng, M. and MacKinnon. R. (1996). Crystal structures of a complexed and peptide-free membrane protein-binding domain; molecular basis of peptide recognition by PDZ. Cell, 85. 1067-1076.

Eck. M.J.. Pluskey. S.. Trub. T.. Harrison. S.C. and Shoelson. S.E. (1996). Spatial constraints on the recognition of phosphoproteins by the tandem SH2 domains of the phosphatase SH-PTP2. Nature, 379. 277-280.

Eck. M.J.. Shoelson. S.E. and Harrison. S.C. (1993). Recognition of a high-affinity phosphotyrosyl peptide by the Src homology-2 domain of p561ck. Nature, 362. 87-91.

Fairman. W.A.. Vandenberg, R.J.. Arriza. J.L.. Kavanaugh. M.P. and Amara. S.G. (1995). An excitatory amino-acid transporter with properties of a ligand-gated chloride channel. Nature, 375. 599-603.

Fleckner. J.. Rasmussen. H.H. and Justesen. J. (1991). Human interferon gamma potently induces the synthesis of a 55-kDa protein (gamma 2) highly homologous to rabbit peptide chain release factor and bovine tryptophanyl-tRNA synthetase. Proc. Natl. Acad. Sci. USA, 88. 11520-11524. '

Felder. S.. Zhou. M.. Hu. P.. Urena. J.. Ullrich, A.. Chaudhuri. M.. White. M.. Shoelson. S.E. and Schlessinger. J. (1993). SH2 domains exhibit high-affinity binding to tyrosine-phosphorylated peptides yet also exhibit rapid dissociation and exchange. XIol. Cell. Biol., 13. 1449-1455.

Fournier. N.. Chalus, L.. Durand, I., Garcia, E., Pin, J.J.. Churakova. T.. Patel. S.. Zlot. C.. Gorman, D., Zurawski, S., Abrams, J., Bates, E.E. and Garrone. P. (2000). FDF03, a novel inhibitory receptor of the immunoglobulin superfamily, is expressed by human dendritic and myeloid cells. J. Immunol., 165. 1197-1209.

Furka, A.. Sebestyen, F., Asgedom. M. and Dibo, G. (1991). General method for rapid synthesis of multicomponent peptide mixtures. Int. J. Pept. Protein Res., 37.487- 493.

Gamer. C.C.. Nash, J. and Huganir, R.L. (2000). PDZ domains in synapse assembly and signalling. Trends Cell Biol, 10,274-280.

119 Gee. S.H.. Sekely, S.A., Lombardo, C., Kurakin, A., Froehner, S.C. and Kay. B.K. (1998). Cyclic peptides as non-carboxyl-terminal ligands of syntrophin PDZ domains. J. Biol. Chem., 273, 21980-21987.

Gottesman. S. (1996). Proteases and their targets in Escherichia coli. .-innii. Rev. Genet.. 30. 465-506.

Gottesman. S.. Roche. E.. Zhou, Y. and Sauer, R.T. (1998). The ClpXP and ClpAP pro teases degrade proteins with carboxy-terminal peptide tails added by the SsrA- tagging system. Gene.9 Dev.. 12. 1338-1347.

Gram. H.. Schmitz. R., Zuber. J.F. and Baumann. G. (1997). Identification of phosphopeptide ligands for the Src-homology 2 (SH2) domain of Grb2 by phage display. Eiir. J. Biochem., 246. 633-637.

Grucza. R.A.. Bradshaw. J.M.. Futterer. K. and Waksman. G. (1999). SH2 domains; from structure to energetics, a dual approach to the study of structure-fimction relationships. Med. Res. Rev., 19. 273-293.

Guo. X.-C. (1999). unpublished results.

Ha. J.. Daniel. S.. Kong, I.S.. Park. C.K.. Tae. H.J. and Kim. K.H. (1994). Cloning of human acetyl-CoA carboxylase cDNA. Eur. J. Biochem.. 219. 297-306.

Hagen. F.S.. Grant. F.J.. Kuijper. J.L., Slaughter. C.A.. Moomaw. C.R.. Orth. K.. O'Hara. P.J. and Munfbrd. R.S. (1991). Expression and characterization of recombinant human acyloxyacyl hydrolase, a leukocyte enzyme that deacylates bacterial lipopolysaccharides. Biochemistry, 30. 8415-8423.

Hagenbuch. B. and Meier, P.J. (1994). Molecular cloning, chromosomal localization, and fimctional characterization of a human liver Na-H/bile acid cotransporter. J. Clin. Invest.. 93, 1326-1331.

Hamajima. N.. Matsuda. K., Sakata, S., Tamaki. N., Sasaki. M. and Nonaka. M. ( 1996). A novel gene family defined by human dihydropyrimidinase and three related proteins with differential tissue distribution. Gene, 180, 157-163.

Horton. Y.M.. Sullivan. M. and Houslay, M.D. (1995). Molecular cloning of a novel splice variant of human type IVA (PDE-IVA) cyclic AMP phosphodiesterase and localization of the gene to the pl3.2-ql2 region of human chromosome 19 [corrected] [published erratum appears in Biochem J 1995 Dec 15:312(Pt 3):991 ]. Biochem. J., 308 ( Pt 2), 683-691.

120 Hao, B., Gong, W., Rajagopalan, P.T., Zhou, Y., Pei, D. and Chan, M.K. (1999). Structurai basis for the design of antibiotics targeting peptide deformylase. Biochemistry, 38, 4712-4719.

Hata. Y.. Butz, S. and Sudhof, T.C. (1996). CASK: a novel dlg/PSD95 homolog with an N-terminal calmodulin-dependent protein kinase domain identified by interaction with neurexins. J. NeuroscL, 16, 2488-2494.

Hichuli, E., Dobeli, H. and Schacher, A. (1987). New metal chelate adsorbent selective for proteins and peptides containing neighbouring histidine residues. Jnurnol o f Chromatography, 411, 177-184.

Hof, P., Pluskey, S., Dhe-Paganon, S., Eck, M.J. and Shoelson. S.E. (1998). Crystal structure of the tyrosine phosphatase SHP-2. Cell. 92, 441-450.

Hu, Y.J., Wei, Y.. Zhou, Y., Rajagopalan, P.T.R. and Pei. D. (1999). Determination of substrate specificity for peptide deformylase through the screening of a combinatorial peptide library. Biochemistry. 38, 643-650.

Huber. M., Izzi, L„ Grondin. P.. Houde. C., Kunath. T., Veillette, A. and Beauchemin. N. (1999). The carboxyl-terminal region of biliary glycoprotein controls its tyrosine phosphorylation and association with protein-tyrosine phosphatases SHP-1 and SHP-2 in epithelial cells. J. Biol. Chem.. 274, 335-344.

Huyer, G.. Li. Z.M.. Adam. M.. Huckle. W.R. and Ramachandran. C. (1995). Direct determination of the sequence recognition requirements of the SH2 domains of SH-PTP2. Biochemistry'. 34, 1040-1049.

Huyer, G. and Ramachandran, C. (1998). The specificity of the N-terminal SH2 domain of SHP-2 is modified by a single point mutation. Biochemistry. 37,2741-2747.

Itoh, M., Nagafuchi, A., Yonemura. S., Kitani-Yasuda. T. and Tsukita, S. (1993). The 220-kD protein colocalizing with cadherins in non-epithelial cells is identical to ZO-1. a tight junction-associated protein in epithelial cells: cDNA cloning and immunoelectron microscopy./. Cell Biol.. 121, 491-502.

Jin, C.J., Miners, J.O., Lillywhite, K.J. and Mackenzie, P.l. (1993). cDNA cloning and expression of two new members of the human liver UDP-glucuronosyltransferase 2B subfamily. Biochem. Biophys. Res. Commun., 194, 496-503.

Keiler, K.C. and Sauer, R.T. (1995). Identification of active site residues of the Tsp protease. J. Biol. Chem., 270, 28864-28868,

Keiler, K.C. and Sauer, R.T. (1996). Sequence determinants o f C-terminal substrate recognition by the Tsp protease. J. Biol. Chem., 271,2589-2593.

121 Keiler. K.C.. Silber, K.R.. Downard, K.M., Papayannopoulous, I.A.. Biemann. K. and Sauer, R.T. (1995). C-terminal specific protein degradation: activity and substrate specificity of the Tsp protease. Protein ScL. 4. 1507-1515.

Keiler. K.C.. Waller, P.R. and Sauer, R.T. (1996). Role of a peptide tagging system in degradation of proteins synthesized from damaged messenger RNA [see comments]. Science, 271. 990-993.

Keilhack. H.. Tenev. T.. Nyakatura. E.. Godovac-Zimmermann. J.. Nielsen. L... Seedorf. K. and Bohmer. F.D. (1998). Phosphotyrosine 1173 mediates binding of the protein-tyrosine phosphatase SHP-1 to the epidermal growth factor receptor and attenuation of receptor signaling. J. Biol. Chem., 273. 24839-24846.

Kelly. M.A.. Liang, H.. Sytwu, II, Vlattas, I., Lyons, N.L.. Bowen. B.R. and Wennogle. L.P. (1996). Characterization of SH2-ligand interactions via library affinity selection with mass spectrometric detection. Biochemistry. 35. 1 1747-11755.

Kim. E.. Niethammer. M.. Rothschild. A.. Jan. Y.N. and Sheng, M. (1995). Clustering of Shaker-type K+ channels by interaction with a family of membrane-associated guanylate kinases. Nature, 378. 85-88.

Klingmuller. U.. Lorenz. U.. Cantley. L.C.. Neel. B.G. and Lodish. H.F. (1995). Specific recruitment of SH-PTPl to the erythropoietin receptor causes inactivation of JAK2 and termination of proliferative signals. Cell, 80. 729-738.

Koike. J.. Takagi. A.. Miwa. T., Hirai, M.. Terada. M. and Katoh. M. (1999). Molecular cloning of Frizzled-10. a novel member of the Frizzled gene family. Biochem. Biophys. Res. Commun., 262, 39-43.

Komau. H.C.. Schenker. L.T., Kennedy, M.B. and Seeburg, P.H. (1995). Domain interaction between NMDA receptor subunits and the postsynaptic density protein PSD-95.5’c/e«ce,269, 1737-1740.

Komau, H.C., Seeburg, P.H. and Kennedy, M.B. (1997). Interaction of ion channels and receptors with PDZ domain proteins. Curr. Opin. NeurobioL, 7. 368-373.

Kuriyan, J. and Cowbum. D. (1997). Modular peptide recognition domains in eukaryotic signaling. Annu. Rev. Biophys. Biomol. Struct., 26. 259-288.

Kuroiwa. A.. Yamashita, Y., Inui, M., Yuasa, T., Ono, M.. Nagabukuro. A.. Matsuda. Y. and Takai, T. (1998). Association of tyrosine phosphatases SHP-1 and SFIP-2. inositol 5-phosphatase SHIP with gp49Bl, and chromosomal assignment of the gene. y. Biol. Chem., 273, 1070-1074.

1 2 2 Ladbury, J.E.. Lemmon, M.A., Zhou, M., Green, J., Botfield, M.C. and Schlessinger. J. (1995). Measurement of the binding of tyrosyl phosphopeptides to SH2 domains: a reappraisal. Proc Natl Acad Sci USA, 92, 3199-3203.

Lam. K.S.. Salmon, S.E., Hersh, E.M., Hruby, V.J., Kazmierski, W.M. and Knapp. R.J. (1991). A new type of synthetic peptide library for identifying ligand-binding activity [published errata appear in Nature 1992 Jul 30;358(6385):434 and 1992 Dec 24-31;360(6406):768]. Nature. 354. 82-84.

Lanzetta. P.A., Alvarez. L.J., Reinach, P.S. and Candia, O.A. (1979). An improved assay for nanomole amounts of inorganic phosphate. Anal. Biochem.. 100. 95-97.

Law. C.L.. Sidorenko. S.P.. Chandran. K.A.. Zhao. Z., Shen. S.H.. Fischer. E.H. and Clark. E.A. (1996). CD22 associates with protein tyrosine phosphatase 1C. Syk. and phospholipase C-gamma(l) upon B cell activation. J. Exp. Med. 183, 547- 560.

Lee. C.H.. Kominos. D.. Jacques, S.. Margolis. B.. Schlessinger. J.. Shoelson. S.E. and Kuriyan, J. (1994). Crystal structures of peptide complexes of the amino-terminal SH2 domain of the Syp tyrosine phosphatase. Structure. 2. 423-438.

Laxminarayan. K.M.. Chan. B.K.. Tetaz, T.. Bird. P.L and Mitchell. C.A. (1994). Characterization of a cDNA encoding the 43-kDa membrane-associated inositol- polyphosphate 5-phosphatase. J. Biol. Chem.. 269. 17305-17310.

Lissy. N.A.. Bellacosa. A.. Sonoda, G.. Miller. P.O.. Jhanwar. S.C. and Testa. J.R. (1996). Isolation, characterization, and mapping to human chromosome 1 lq24-25 of a cDNA encoding a highly conserved putative transmembrane protein. TMC. Biochim. Biophys. Acta. 1306. 137-141.

Liu. D.. Martino. G.. Thangaraju. M.. Sharma, M.. Halwani, P.. Shen. S.H.. Patel. Y.C. and Srikant. C.B. (2000). Caspase-8-mediated intracellular acidification precedes mitochondrial dysfunction in somatostatin-induced apoptosis. J. Biol. Chem.. 275. 9244-9250.

Malmqvist, M. (1993). Biospecific interaction analysis using biosensor technology. Nature. 361, 186-187.

Maranto. A.R. (1994). Primary structure, ligand binding, and localization of the hiunan type 3 inositol 1.4,5-trisphosphate receptor expressed in intestinal epithelium. J. Biol. Chem.. 269, 1222-1230.

Mauzy, C.A., Hwang, O., Egloff, A.M., Wu, L.H. and Chung, F.Z. (1992). Cloning, expression, and characterization of a gene encoding the human angiotensin II type 1A receptor. Biochem. Biophys. Res. Commun., 186, 277-284.

123 Miranda-Vizuete, A., Gustafsson, J.A. and Spyrou, G. (1998). Molecular cloning and expression of a cDNA encoding a human thioredoxin-like protein. Biochem. Biophys. Res. Commun., 243, 284-288.

Morrione, A., Plant, P., Valentinis, B., Staub, O., Kumar, S., Rotin. D. and Baserga. R. (1999). mGrblO interacts with Nedd4. J. Biol. Chem.. 274. 24094-24099.

Matayoshi. E.D.. Wang, G.T.. Krafft. G.A. and Erickson, J. (1990). Novel fluorogenic substrates for assavina retroviral oroteases bv resonance enerav transfer. .Science. 247. 954-958.

Matthews. R.J., Bowne, D.B., Flores, E. and Thomas, M.L. (1992). Characterization of hematopoietic intracellular protein tyrosine phosphatases: description of a phosphatase containing an SH2 domain and another enriched in proline-, glutamic acid-, serine-, and threonine- rich sequences. iWb/. Cell. Biol.. 12. 2396-2405.

Mayer. B.J.. Jackson, P.K.. Van Etten. R.A. and Baltimore. D. (1992). Point mutations in the abl SH2 domain coordinately impair phosphotyrosine binding in vitro and transforming activity in vivo. Mol. Cell. Biol.. 12. 609-618.

Morals Cabral, J.H.. Petosa, C., Sutcliffe. M.J.. Raza. S.. B>Ton. 0.. Poy. P.. Marfatia. S.M.. Chishti. A.H. and Liddington, R.C. (1996). Crystal structure of a PDZ domain. Nature. 382, 649-652.

Mousseau. D.D.. Banville, D., L'Abbe. D., Bouchard, P. and Shen. S.H. (2000). PILRalpha. a novel immimoreceptor tyrosine-based inhibitory motif-bearing protein, recruits SHP-1 upon tyrosine phosphorylation and is paired with the truncated counterpart PILRbeta. J. Biol. Chem.. 275.4467-4474.

Mulhem. T.D.. Shaw. G.L.. Morton, C.J.. Day, A.J. and Campbell. I.D. (1997). The SH2 domain from the tyrosine kinase Fyn in complex with a phosphotvTosyl peptide reveals insights into domain stability and binding specificitv. Structure. 5. 1313- 1323.

Muller, K.. Gombert, P.O., Manning, U., Grossmuller, P., Graff, P.. Zaegel, H.. Zuber. J.P., Preuler, P.. Tschopp, C. and Baumann, G. (1996). Rapid identification of phosphopeptide ligands for SH2 domains. Screening of peptide libraries by fluorescence-activated bead sorting. J Biol. Chem.. 271. 16500-16505.

Neel. B.G. and Tonks, N.K. (1997). Protein tyrosine phosphatases in signal transduction. Curr. Opin. Cell Biol., 9, 193-204.

124 Nomura, N.. Sasamoto, S., Ishii, S., Date, T., Matsui, M. and Ishizaki, R. (1989). Isolation of human cDNA clones o f ski and the ski-related gene. sno. Nucleic Acids Res., 17, 5489-5500.

Oelmuller, R., Herrmann, R.G. and Pakrasi, H.B. (1996). Molecular studies of CtpA. the carboxyl-terminal processing protease for the 01 protein of the photosystem II reaction center in higher plants. J. Biol. Chem., 271, 21848-21852.

Olcese. L.. Cambiaggi, A., Semenzato, G., Bottino, C., Moretta. A. and Vivier. E. ( 1997). Human killer cell activatory receptors for MHC class I molecules are included in a multimeric complex expressed by natural killer cells. J. Immunol.. 158. 5083- 5086.

Ottinger. E.A.. Botfield. M.C. and Shoelson, S.E. (1998). Tandem SH2 domains confer high specificity in tyrosine kinase signaling. 7. Biol. Chem.. 273. 729-735.

Ottinger. E.A.. Shekels. L.L.. Bemlohr. D.A. and Barany. G. (1993). Synthesis of phosphotyrosine-containing peptides and their use as substrates for protein tyrosine phosphatases. Biochemistry. 32.4354-4361.

Paetzel. M. and Dal bey, R.E. (1997). Catalytic hydroxyl/amine dyads within serine proteases. Trends Biochem. Sci.. 22. 28-31.

Pallen. M.J. and Wren, B.W. (1997). The HtrA family of serine proteases. .Viol. Microbiol.. 26. 2D9-22\.

Parsed. D.A.. Silber. FC.R. and Sauer, R.T. (1990). Carboxy-terminal determinants of intracellular protein degradation. Genes Dev.. 4, 277-286.

Pawson. T. (1995). Protein modules and signalling networks. Nature. 373. 573-580.

Pawson. T. and Gish. G.D. (1992). SH2 and SH3 domains: from structure to tunction. Cell. 71. 359-362.

Pawson, T. and Scott, J.D. (1997). Signaling through scaffold, anchoring, and adaptor proteins. Science, 278,2075-2080.

Pei. D.. Lorenz, U., Klingmuller, U., Neel, B.G. and Walsh, C.T. (1994). Intramolecular regulation of protein tyrosine phosphatase SH-PTPl: a new function for Src homology 2 domains. Biochemistry, 33, 15483-15493.

Pei, D.. Neel, B.G. and Walsh, C.T. (1993). Overexpression, purification, and characterization of SHPTPl, a Src homology 2-containing protein-tyrosine- phosphatase. Proc. Natl. Acad. Sci. USA, 90,1092-1096.

125 Pei. D.. Wang, J. and Walsh, C.T. (1996). Differential functions of the two Src homology 2 domains in protein tyrosine phosphatase SH-PTP1. Proc Nall Acad Sci i ’ SA. 93. 1141-1145.

Piccione, E., Case. R.D., Domchek, S.M., Hu, P., Chaudhuri, M., Backer. J.M.. Schlessinger, J. and Shoelson, S.E. (1993). Phosphatidylinositol 3-kinase p85 SH2 domain specificity defined by direct phosphopeptide/SH2 domain binding. Biochemistry. 32. 3197-3202.

Plutzky. J.. Neel, B.G. and Rosenberg, R.D. (1992). Isolation of a src homology 2- containing tyrosine phosphatase. Proc Natl .AcadSci US.4. 89. 1123-1127.

Ponting. C.P. ( 1997). Evidence for PDZ domains in bacteria, yeast, and plants. Protein Sci.. 6. 464-468.

Ponting. C.P. and Phillips. C. (1995). DHR domains in syntrophins. neuronal NO synthases and other intracellular proteins [letter] [see comments]. Trends Biochem. Sci.. 20. 102-103.

Poy. F.. Yaffe. M.B., Sayos. J., Saxena. K., Morra, M., Sumegi. J.. Cantley. L.C.. Terhorst. C. and Eck. M.J. (1999). Crystal structures of the XLP protein S.AP reveal a class of SH2 domains with extended, phosphotyrosine-independent sequence recognition. Mol Cell. 4. 555-561.

Pumphrey. N.J.. Taylor. V.. Freeman. S.. Douglas. M.R.. Bradfield. P.F.. Young. S.P.. Lord. J.M.. Wakelam. M.J.. Bird. I.N.. Salmon. M. and Buckley. C.D. (1999). Differential association of cytoplasmic signalling molecules SHP-1. SHP-2. SHIP and phospholipase C-gamma 1 with PECAM-1/CD31. FEBS Lett.. 450. 77-83.

Rajagopalan. P.T., Datta, A. and Pei, D. (1997). Purification, characterization, and inhibition of peptide deformylase from Escherichia coli. Biochemistrv. 36. 13910- 13918.

Ramamoorthy. S., Leibach, F.H.. Mahesh. V.B.. Han, H., Yang-Feng, T.. Blakely. R.D. and Ganapathy. V. (1994). Functional characterization and chromosomal localization of a cloned taurine transporter from human placenta. Biochem. J.. 300 ( Pt 3), 893-900.

Rasola, A.. Galietta, L.J.. Barone, V., Romeo, G. and Bagnasco. S. (1995). Molecular cloning and functional characterization of a GABA/betaine transporter from human kidney. FEBS Lett., 373. 229-233.

Ren. H. and Stiles, G.L. (1994). Characterization of the human A1 adenosine receptor gene. Evidence for alternative splicing. J. Biol. Chem., 269, 3104-3110.

126 Rowley, R.B., Burkhardt, A.L., Chao, H.G., Matsueda, G.R. and Bolen, J.B. (1995). Syk protein-tyrosine kinase is regulated by tyrosine-phosphorylated Ig alpha/lg beta immunoreceptor tyrosine activation motif binding and autophosphorylation. J. Biol. Chem., 270, 11590-11594.

Ranganathan. R. and Ross, E.M. (1997). PDZ domain proteins: scaffolds for signaling complexes. Curr. Biol., 1, R770-773.

Sadowski, I.. Stone, J.C. and Pawson, T. (1986). A noncatalytic domain conserved among cytoplasmic protein-tyrosine kinases modifies the kinase ftmction and transforming activity of Fujinami sarcoma virus P130gag-fps. Mol. Cell. Biol.. 6. 4396-4408.

Samson, M., Libert, P., Doranz, B.J., Rucker, J., Liesnard, C.. Farber, C.M.. Saragosti. S.. Lapoumeroulie, C., Cognaux. J., Forceille, C., Muyldermans, G., Verhofstede. C., Burtonboy, G.. Georges. M.. Imai, T., Rana. S.. Yi. Y.. Smyth. R.J.. Collman, R.G., Doms, R.W., Vassart, G. and Parmentier. M. (1996). Resistance to HIV-1 infection in Caucasian individuals bearing mutant alleles of the CCR-5 chemokine receptor gene [see comments]. Sature. 382. 722-725.

Schaap. D., de Widt, J., van der Wal, J., Vandekerckhove, J.. van Damme. J.. Gussow. D.. Ploegh. H.L.. van Blitterswijk, W.J. and van der Bend, R.L. (1990). Purification, cDNA-cloning and expression of human diacylglycerol kinase. FEBSLett..215.\5\-\5%.

Shafqat, S., Tamarappoo, B.K., Kilberg, M.S., Puranam, R.S., McNamara, J.O.. Guadano-Ferraz, A. and Fremeau, R.T.J. (1993). Cloning and expression of a novel Na(+)-dependent neutral amino acid transporter structurally related to mammalian Na+/glutamate cotransporters. J. Biol. Chem.. 16^. 15351-15355.

Shashidharan. P. and Plaitakis, A. (1993). Cloning and characterization of a glutamate transporter cDNA from human cerebellum. Biochim. Biophvs. .Acta, 1216. 161- 164.

Shaw. S.K., Cepek, K..L., Murphy, E.A., Russell, G.J., Brenner. M.B. and Parker. C.M. (1994). Molecular cloning of the human mucosal lymphocyte integrin alpha E subunit. Unusual structure and restricted RNA distribution. J. Biol. Chem.. 269. 6016-6025.

Svoboda, M., Tastenoy, M., Van Rampelbergh, J., Goossens, J.F.. De Neef. P.. Waelbroeck, M. and Robberecht, P. (1994). Molecular cloning and functional characterization of a human VIP receptor from SUP-Tl lymphoblasts. Biochem. Biophys. Res. Commun., 205, 1617-1624.

127 Sato, T.. Irie. S.. BCitada, S. and Reed, J.C. (1995). FAP-1: a protein tyrosine phosphatase that associates with Fas. Science, 268, 411-415.

Sawyer, T.K. (1998). Src homology-2 domains: structure, mechanisms, and drug discovery. Biopolymers, 47, 243-261.

Schepens. J., Cuppen, E., Wieringa, B. and Hendriks, W. (1997). The neuronal nitric oxide synthase PDZ motif binds to -G(D,E)XV* carboxyterminal sequences. FEBS Lett.,m . 52-56.

Schultz, J., Hoffmuller, U., BCrause, G., Ashurst, J., Marcias, M.J., Schmieder, P., Schneider-Mergener, J. and Oschkinat, H. (1998). Specific interactions between the syntrophin PDZ domain and voltage-gated sodium channels. Nat. Struct. Biol.. 5, 19-24. Selvin, P.R. (1995). Fluorescence resonance energy transfer. Methods Enzvmoi. 246. 300-334.

Shen, S.H., Bastien, L., Posner, B.I. and Chretien, P. (1991). A protein-tyrosine phosphatase with sequence similarity to the SH2 domain of the protein-tyrosine kinases [published erratum appears in Nature 1991 Oct 3 1;353(6347):868|. Nature. 352, 736-739.

Shoelson. S.E., Sivaraja, M., Williams, K.P., Hu, P., Schlessinger, J. and Weiss. M..A. (1993). Specific phosphopeptide binding regulates a conformational change in the PI 3-kinase SH2 domain associated with enzyme activation. EMBOJ.. 12, 795- 802.

Slegal, G., Davis, B., Kristensen, S.M., Sankar, A., Linacre, J., Stein, R.C., Panayotou, G., Waterfield, M.D. and Driscoll, P.C. (1998). Solution structure of the C- terminal SH2 domain of the p85 alpha regulatory subunit of phosphoinositide 3- kinase. J. Mol. Biol., 276,461-478.

Silber, K.R., Keiler, K.C. and Sauer, R.T. (1992). Tsp: a tail-specific protease that selectively degrades proteins with nonpolar C termini. Proc. Natl. .Acad. Sci. US.A. 89, 295-299.

Silber, K.R. and Sauer, R.T. (1994). Deletion of the prc (tsp) gene provides evidence for additional tail-specific proteolytic activity in Escherichia coli K-12. Mol. Gen. Genet., 242, 237-240.

Siminovitch, K.A. and Neel, B.G. (1998). Regulation of B cell signal transduction by SH2-containing protein-tyrosine phosphatases. Semin. Immunol., 10, 329-347.

128 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. USA, 96, 6678-6682.

Songyang, Z.. Fanning, A.S., Pu, C.. Xu, J., Marfatia, S.M., Chishti. A.H.. Crompton. A.. Chan. A.C., Anderson, J.M. and Cantley, L.C. (1997). Recognition of unique carboxyl-terminal motifs by distinct PDZ domains. Science, 275. 73-77.

Songyang. Z.. Shoelson. S.E., Chaudhuri, M., Gish. G.. Pawson, T.. Haser. W.G.. King. P.. Roberts, T., Ratnofsk}', S., Lechleider, R.J. and et al. (1993). SH2 domains recognize specific phosphopeptide sequences. Cell, 72. 767-778.

Songyang, Z.. Shoelson. S.E., McGlade. J.. Olivier. P., Pawson. T.. Busteio. X.R.. Barbacid. M.. Sabe. H., Hanafusa, H., Yi. T. and et ai. (1994). Specific motifs recognized by the SH2 domains of Csk, 3BP2. Ips/fes. GRB-2. HCP. SHC. Svk. and Vav. Mol. Cell. Biol.. 14. 2777-2785.

Strieker, N.L.. Christopherson. K.S.. Yi, B.A., Schatz, P.J., Raab. R.W.. Dawes. G.. Bassett, D.E.J., Bredt, D.S. and Li, M. (1997). PDZ domain of neuronal nitric oxide synthase recognizes novel C-terminal peptide sequences. Nat. Bioiechnol.. 15.336-342.

Sugimoto. S.. Wandless. T.J.. Shoelson. S.E., Neel. B.G. and Walsh. C.T. (1994). .Activation o f the SH2-containing protein tyrosine phosphatase. SH-PTP2. by phosphotyrosine-containing peptides derived from insulin receptor substrate-1. J. Biol. C hem ..169. 13614-13622.

Takiguchi. M.. Haraguchi. Y. and Mori, M. (1988). Human liver-type arginase gene: structure of the gene and analysis of the promoter region. Nucleic Acids Res.. 16. 8789-8802.

Tamir. I.. Dal Porto. J.M. and Cambier. J.C. (2000). Cytoplasmic protein tyrosine phosphatases SHP-1 and SHP-2: regulators of B cell signal transduction. Curr. Opin. Immunol., 12, 307-315.

Taylor. V.C.. Buckley, C.D.. Douglas, M., Cody, A.J.. Simmons, D.L. and Freeman. S.D. (1999). The myeloid-specific sialic acid-binding receptor. CD33. associates with the protein-tyrosine phosphatases. SHP-1 and SHP-2. J Biol. Chem.. 274. 11505- 11512.

Tonks, N.K. (1996). Protein tyrosine phosphatases and the control of cellular signaling responses. Adv. Pharmacol., 36, 91-119.

Van Melderen, L. and Gottesman, S. (1999). Substrate sequestration by a proteoiytically inactive Lon mutant. Proc. Natl. Acad. Sci. USA, 96, 6064-6071.

129 Varshavsky, A. (1996). TheN-end rule: functions, mysteries, uses. Proc. Natl. Acad. Sci. USA, 93, 12142-12149.

Veillette, A.. Thibaudeau, E. and Latour, S. (1998). High expression of inhibitory receptor SHPS-1 and its association with protein-tyrosine phosphatase SHP-1 in macrophages. J. Biol. CAem.. 273, 22719-22728.

Waksman. G.. Kominos. D.. Robertson. S.C.. Pant. N.. Baltimore. D.. Birge. R.B.. Cowbum. D.. Hanafusa. H.. Mayer. B.J.. Overduin. M. and et al. ( 1992). Crystal structure of the phosphotyrosine recognition domain SH2 of v- src complexed with tyrosine-phosphorylated peptides [see comments]. Nature. 358. 646-653.

Wallace. A.M.. Dass, B., Ravnik, S.E.. Tonk. V.. Jenkins, N.A.. Gilbert. D.J.. Copeland. N.G. and MacDonald. C.C. (1999). Two distinct forms of the 64.000 Mr protein of the cleavage stimulation factor are expressed in mouse male germ cells. Proc. Natl. Acad. Sci. USA. 96. 6763-6768.

White. A.. Lamb. P.W. and Barrett, J.C. (1998). Frequent doumregulation of the KAI1(CD82) metastasis suppressor protein in human cancer cell lines. Oncogene. 16.3143-3149.

Williams. M.E.. Marubio. L.M.. Deal. C.R.. Hans. M.. Brust. P.P.. Philipson. L.H.. Miller. R.J.. Johnson. E.C.. Harpold. M.M. and Ellis. S.B. (1994). Structure and functional characterization of neuronal alpha IE calcium charmel subtypes. J. Biol. Chem.. 269. 22347-22357.

Woods. D.F. and Bryant, P.J. (1991). The discs-large tumor suppressor gene of Drosophila encodes a guanylate kinase homolog localized at septate junctions. Cell. 66. 451-464.

Woods. D.F. and Bryant. P.J. (1993). ZO-1, DlgA and PSD-95/SAP90: homologous proteins in tight, septate and synaptic cell Junctions. V/ec/i. Dew. 44. 85-89.

Ye. Q. and Worman. H.J. (1994). Primary structure analysis and lamin B and DNA binding of human LBR, an integral protein of the nuclear envelope inner membrane. J. Biol. Chem.. 269.11306-11311.

Yi. T.L., Cleveland. J.L. and Ihle, J.N. (1992). Protein tyrosine phosphatase containing SH2 domains: characterization, preferential expression in hematopoietic cells, and localization to human chromosome 12pl2-pl3. Mol. Cell. Biol.. 12. 836-846.

Youngquist, R.S., Puentes, G.R., Lacey, M.P. and Keough, T. (1995). Generation and screening of combinatorial peptide libraries designed for rapid sequencing by mass-spectrometry. J. Am. Chem. Soc.. 117, 3900-3906.

130 Yu, Z., Ahmad, S., Schwartz, J.L., Banville, D. and Shen, S.H. (1997). Protein-tyrosine phosphatase SHP2 is positively linked to proteinase-activated receptor 2-mediated mitogenic pathway. J. Biol. Chem., 272. 7519-7524.

131