Combating vancomycin-resistant bacteria with catalytic antibodies

MISSING IMAGE

Material Information

Title:
Combating vancomycin-resistant bacteria with catalytic antibodies the synthesis of a transition-state analog
Physical Description:
xii, 99 leaves : ill. ; 29 cm.
Language:
English
Creator:
Vedha-Peters, Kavitha
Publication Date:

Subjects

Subjects / Keywords:
Antibiotics -- Synthesis   ( lcsh )
Chemistry thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Chemistry -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2002.
Bibliography:
Includes bibliographical references (leaves 91-98).
Statement of Responsibility:
by Kavitha Vedha-Peters.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 028838975
oclc - 50448861
System ID:
AA00020484:00001

Full Text










COMBATING VANCOMYCIN-RESISTANT BACTERIA WITH CATALYTIC
ANTIBODIES: THE SYNTHESIS OF A TRANSITION-STATE ANALOG
















By

KAVITHA VEDHA-PETERS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2002



























To my beloved husband and friend, Matt














ACKNOWLEDGMENTS

First and foremost I would like to thank my advisor, Dr. Jon Stewart, for all of his

guidance, encouragement, and patience during the course of my studies. He has taught

me a tremendous amount, and I have enjoyed working in his lab. I would also like to

express my appreciation to all of the past and present Stewart group members for their

helpful discussions and friendship. Sonia Rodriguez, Jennifer Tonzello, Carlos Martinez

and Aris Polyzos provided a very supportive and open environment for me when I first

joined the group and was very nervous and intimidated; for this and their friendship I will

be forever grateful. My deepest gratitude goes to Carlos for all of his help with my

project and for his patience with my daily barrage of questions during my first three years

in the lab. Iwona Kaluzna has kept me entertained the past two years with the interesting

things that she does, but she has also become a good friend who always listens to me,

which I truly appreciate. Brian Kyte, whose daily "gestures" will be missed dearly,

deserves acknowledgement for the idea for the labeling experiment for the Mitsunobu

reaction product. Adam Walton has been a constant source of facts about California and

I have enjoyed hearing about his quests in the lab to make Dr. Stewart do a jig on the

roof. Brent Feske has actually made me put the words "dude" and "whoadie" in my

working vocabulary and I wish him luck with the continuation of the vancomycin-

resistance project.








I am very grateful to the Horenstein, Hudlicky, Scott, and Wright groups, who let

me borrow many chemicals and equipment over the years that helped me with my

project.

I don't think I would have made it to this point without the continual love and

support of my parents. They always encouraged my curiosity as a child and have taught

me the importance of an education.

Finally, my husband Matt has been an integral part of my graduate career on

many different levels. As a colleague he has provided a constant source of discussion for

my research ideas and problems, and would often give me very helpful suggestions for

my chemistry. I have also found Matt's enthusiasm and love of science inspiring and

appreciate the many new things I have learned from our conversations on his new book

of the week (even though he thought I was never listening). But above all, I thank Matt

for being the most wonderful husband and friend in the world. His endless love and

support helped make this Ph.D. possible.














TABLE OF CONTENTS

pae

ACKNOW LEDGM ENTS.................................................................................................. iii

SYM BOLS AND ABREVIATIONS................................................................................ vii

ABSTRACT.......................................................................................................................xi



CHAPTERS

1 BACKGROUND AND INTRODUCTION .................................................................... 1

A Brief Introduction to Vancomycin .............................................................................. 1
Bacterial Cell W all Biosynthesis..................................................................................... 2
Vancomycin-Resistant Bacteria...................................................................................... 5
Strategies to Fight Vancomycin-Resistant Bacteria........................................................ 7
Introduction to Research Goal....................................................................................... 13
Catalytic Antibodies...................................................................................................... 14
Introduction to Synthetic Goal...................................................................................... 18
M ethods for the Synthesis of Phosphonate Esters........................................................ 19
Synthesis of Unsymmetrical Phosphonate Diesters.................................................. 19
Synthesis of Phosphonate M onoesters...................................................................... 25
Solid-Phase Peptide Synthesis ...................................................................................... 26
Segment Condensation.................................................................................................. 30


2 MODEL STUDY OF THE MITSUNOBU REACTION AND SIDE-CHAIN
DEPROTECTION STEPS FOR THE TRANSITION-STATE ANALOG
SYNTHESIS ..................................................................................................................... 32

M itsunobu Reaction with a M onomethyl Phosphonate Ester....................................... 33
Oxygen-18 Labeling Experiment.............................................................................. 36
Chiral GC Experiment............................................................................................... 41
Side-Chain Deprotection Steps ..................................................................................... 43
Phosphonate Diester Demethylation with Bromotrimethylsilane............................. 43
Saponification of Carboxylate Ester......................................................................... 43








3 SYNTHESIS OF THE TRANSITION-STATE ANALOG.......................................... 46

Solid-Phase Synthetic Approach to the Transition-State Analog ................................. 47
Protection of L-Lactic Acid....................................................................................... 48
Coupling to the W ang Resin..................................................................................... 49
Convergent Approach to the Transition-State Analog.................................................. 52
Solid-Phase Peptide Synthesis of the Tetrapeptide Fragment................................... 53
Synthesis of the Phosphonate Ester Fragment.......................................................... 58
M odel Studies of the Segment Condensation Step................................................... 59
Coupling of the Tetrapeptide and Phosphonate Fragments ...................................... 61
Deprotection of the Phosphonate Methyl Ester and Carboxylate Esters .................. 62
Conclusion..................................................................................................................... 67


4 CONCLUSIONS AND FUTURE W ORK ................................................................... 68


5 EXPERIM ENTAL ........................................................................................................ 71

General M ethods ........................................................................................................... 71
Experimental Procedures............................................................................................... 72


REFERENCES.................................................................................................................. 91

BIOGRAPHICAL SKETCH............................................................................................. 99













SYMBOLS AND ABREVIATIONS


Ac acetyl

ATP adenosine triphosphate

Bn benzyl

Boc tert-butoxycarbonyl

BOP (1 H-benzotriazole-1 -yloxy)tris(dimethylamino)phopshonium
hexafluorophosphate

BSA bovine serum albumin

Bu butyl

Cbz benzoxycarbonyl

DBU 1,8-diazabicyclo[5.4.0]undec-7-ene

DCC dicyclohexylcarbodiimide

DIAD diisopropyl azodicarboxylate

DIC diisopropylcarbodiimide

DIEA diisopropylethylamine

DMAP 4-dimethylaminopyridine

DMF N, N-dimethyl formamide

EEDQ 2-ethoxy- 1 -ethoxycarbonyl- 1,2-dihydroquinoline

ESI electrospray ionization

Et ethyl

FAB fast atom bombardment








Fmoc 9-fluorenylmethoxycarbonyl

GC gas chromatography

HBTU O-( 1H-benzotriazol-1 -yl)-N,N,N',N'-tetramethyluronium
hexafluorophosphate

HMPA hexamethylphosphoramide

HOBt 1 -hydroxybenzotriazole

HPLC high performance liquid chromatography

IBCF isobutyl chloroformate

IR infrared

KLH keyhole limpet hemocyanin

Lac lactate

Me methyl

MS mass spectrometry

MSRA methicillin-resistant Staphylococcus aureus

NAD+ nicotinamide adenine dinucleotide, oxidized form

NADH nicotinamide adenine dinucleotide, reduced form

NADP+ nicotinamide adenine dinucleotide, phosphate, oxidized form

NAG N-acetylglucosamine

NAM N-acetylmuramic acid

NMM N-methyl morpholine

NMP N-methyl pyrrolidinone

NPEOC 4-nitrophenethyloxycarbonyl

PAM phenylacetamidomethyl

PCR polymerase chain reaction








Ph

Pr

PyBOP


SPPS

TBAF

TBDMS

TFA

THF

TMS

TSA

VRE



alanine

arginine

asparagine

aspartic acid

cysteine

glutamine

glutamic acid

glycine

histidine

isoleucine

leucine


phenyl

propyl

(IH- 1-benzotriazol-I -yloxy)tripyrrolidinophosphonium
hexafluorophosphate

solid-phase peptide synthesis

tetrabutylammonium fluoride

tert-butyldimethylsilyl

trifluoroacetic acid

tetrahydrofuran

trimethylsilyl

transition-state analog

vancomycin-resistant enterococci

Amino Acid Codes

Ala A

Arg R

Asn N

Asp D

Cys C

Gin Q

Glu E

Gly G

His H

lie I

Leu L









methionine Met M

phenylalanine Phe F

proline Pro P

serine Ser S

threonine Thr T

tryptophan Trp W

tyrosine Tyr Y

valine Val V














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

COMBATING VANCOMYCIN-RESISTANT BACTERIA WITH CATALYTIC
ANTIBODIES: THE SYNTHESIS OF A TRANSITION-STATE ANALOG

By

Kavitha Vedha-Peters

May 2002

Chairman: Dr. Jon D. Stewart
Major Department: Chemistry

Gram-positive bacterial infections are commonly treated with the antibiotic

vancomycin. Resistance to vancomycin has arisen in some bacteria as a result of the

replacement of an amide bond with an ester linkage in the pentapeptide of the bacterial

peptidoglycan cell wall precursor. This modification prevents the antibiotic from

hindering cell wall synthesis, but leaves the bacterium vulnerable to an ester hydrolysis

catalyst. This project was designed to exploit this weakness by raising catalytic

antibodies that selectively catalyze ester hydrolysis of this pentapeptide. Hydrolysis of

this moiety will prevent essential crosslinking with other pentapeptide strands in the cell

wall, ultimately causing the bacteria to lyse. A peptidylphosphonate ester transition-state

analog was synthesized in order to raise catalytic antibodies against it. In the successful

convergent approach to the transition-state analog, a tetrapeptide fragment was

synthesized using manual solid-phase synthesis and a phosphonate ester fragment was

synthesized in solution using a Mitsunobu reaction. The two fragments were then








coupled using solution-phase techniques to provide the transition-state analog.

Antibodies raised against this transition-state analog will be tested for catalytic and

antimicrobial activity.













CHAPTER 1
BACKGROUND AND INTRODUCTION


A Brief Introduction to Vancomycin

Scientists at Eli Lilly first discovered the glycopeptide antibiotic vancomycin

(Figure 1 -1) in 1956. It is produced by the microorganism Amycolatopsis orientalis

which was found in a soil sample collected from the jungles of Borneo. Vancomycin was


Figure 1-1. Structure of Vancomycin.


first used clinically in 1959 to treat Gram-positive bacterial infections and was especially

important for fighting against penicillin-resistant Staphylococcus aureus (S. aureus). In

later years vancomycin became an antibiotic of last resort when methicillin-resistant S.








aureus (MRSA) had also developed resistance to a variety of other antibiotics, including

cephalosporins, tetracyclines, aminoglycosides, erythromycin and sulfonamides.23


Bacterial Cell Wall Biosynthesis

Vancomycin interferes with the biosynthesis of the Gram-positive bacterial

peptidoglycan cell wall (Figure 1-2). The rigid, heteropolymeric peptidoglycan cell wall

of these bacteria contains a repeating disaccharide unit of N-acetylglucosamine-N-


peptidoglycan
cell wall
cytoplasm\








embedded
proteins plasma
membrane

Figure 1-2. Morphology of Gram-positive bacteria. Figure adapted from
www.hhmi.org.

acetylmuramic acid (NAG-NAM) (Figure 1-3). Attached to the lactyl ether carboxylate

of each N-acetylmuramic acid is the pentapeptide, L-Ala-D-y-Glu-L-Lys-D-Ala-D-Ala.

The heteropolymeric structure of the cell wall is generated by the polymerization of the

NAG-NAM disaccharide monomer unit and by the crosslinking of the L-lysine e-amino

group of one pentapeptide, via a peptide linker, to the penultimate D-alanine of another

pentapeptide, resulting in the loss of the terminal D-alanine residue. This crosslinking of











N-acetyl
glucosamine
'-__-_ _


Pentapeptide


N-acetyl
muramic acid

HO


NH __,)
H3C : N
0 Cell Wall
Crosslink
HN
).-CH3
-02C


Figure 1-3. Structure of the peptidoglycan cell wall of Gram-positive bacteria.


the pentapeptide chains gives the bacterial cell wall rigidity and protects the bacteria from

4
osmotic lysis.

Bacterial cell wall biosynthesis begins in the cytoplasm where the NAM-NAG

disaccharide unit with an attached pentapeptide unit and lipid carrier is assembled. This

lipid-coupled disaccharide-pentapeptide is then transported through the membrane to the

cell surface via a translocase enzyme. A transglycosylase enzyme adds the NAG-NAM-

pentapeptide monomer unit to the growing cell wall and finally a transpeptidase enzyme


-020.-









membrane


cytoplasm


NHAc
0=023-ONAG CH3 0 H 0
OH N ."%yr y' co2-
HO0 1- H 0 CH3
HO

"' 0 NAM H2N
3 NHAc t Kr0
e-- 0 O./ '

transpeptidase NH)

= lipid carrier 0 NH
R = pentaglycine
S CH3 H 0 H O CH3

0 O CO2-" O CH3
0C2- H CH3 H

; C H 0 C02 H 0 CH3
OH3 O CH3

(N 0

I H
NTCO2-
R transglycosylase ( I \

D-Ala-D-AIa-L-Lys-Di-fGIu-L-AIa-A 0 o0_,


H H H C -2- H 0 O3 H

OH3 H 0 ~ 0 OH3




'NH)4
(-0
H2N
diasaccharide precursor unit
Figure 1-4. Biosynthesis of the peptidoglycan cell wall. Figure adapted from reference 3.








catalyzes the crosslinking reaction, forming the peptidoglycan cell wall (Figure 1-4).'3

Vancomycin binds tightly to the terminal D-Ala-D-Ala residues of the peptidoglycan

pentapeptide, thus sterically preventing the transglycosylase and transpeptidase enzymes

from binding to their substrates, which leads to the breakdown of the peptidoglycan.

NMR6' and X-ray crystallography studies have shown that vancomycin binds to the

L-Lys-D-Ala-D-Ala fragment of the pentapeptide by forming five distinct hydrogen

bonds and demonstrates additional effectiveness as a result of its ability to dimerize in

solution. 11-14


Vancomycin-Resistant Bacteria

After almost 30 years of clinical use, strains of vancomycin-resistant enterococci

(VRE) were first reported in 1988.5 While enterococci are not normally considered

pathogenic bacteria this development sparked concern because these bacteria could be

lethal to immunodeficient patients. In addition, the possibility that VRE might transfer

the genetic information responsible for resistance to a more serious pathogen, such as

MRSA, was alarming.- Vancomycin-resistant enterococci are divided into five

categories, designated VanA, VanB, VanC, VanD and VanE.16-18 In the VanA, VanB

and VanD phenotypes, bacterial resistance to vancomycin arises from the replacement of

the terminal D-alanine residue in the peptidoglycan pentapeptide with D-lactate (D-Lac).

This change of an amide bond to an ester bond results in the loss of a crucial hydrogen

bond with vancomycin, reducing the binding affinity by 1000-fold (Figure 1-5). 19 In

VanC and VanE resistance, which is chromosomal rather than plasmid-born, D-Ala-D-

























Figure 1-5. Vancomycin binding with N-Acyl-D-Ala-D-AIa (left) and N-Acyl-D-Ala-D-
Lac (right). The hydrogen bond between the amide proton of D-alanine and a carbonyl of
vancomycin is lost when D-alanine is replaced by D-lactate. Figure adapted from
reference 27.

Ala is replaced with D-Ala-D-Ser in the peptidoglycan pentapeptide, resulting in only a

6-fold decrease in the binding affinity of vancomycin.2021

The genetic mechanism of vancomycin resistance has been well-studied by the

22-26
Courvalin and Walsh groups. The five plasmid-bom genes (vanS, vanR, vanH, vanA,

and vanX) responsible for resistance are expressed only in the presence of vancomycin.

The VanS and VanR proteins form a two-component regulatory system that activates the

transcription of vanH, vanA, and vanX. The VanH protein, an NADP dependent a-

ketoreductase, catalyzes the reduction of pyruvate to D-lactate and VanA acts as a

depsipeptide ligase producing the D-Ala-D-Lac unit that is incorporated into the resistant

bacterial cell wall from D-Ala, ATP and D-Lac. VanX, a zinc-dependent dipeptidase,

selectively hydrolyzes the D-Ala-D-Ala dipeptide used in the normal peptidoglycan

biosynthetic pathway to prevent its incorporation (Figure 1-6).27









0 0 CH3a 0
OY O-. H0- CH3 CH3 D-Ala + ATP 0 CH3


CH3 0 0
_N_ 2 +N VanX 2 +'3N 0-
0 CH3 CH3

Figure 1-6. Role of VanH, VanA and VanX proteins in VRE peptidoglycan biosynthesis.



Strategies to Fight Vancomycin-Resistant Bacteria

Since the emergence of vancomycin-resistant bacteria several approaches have

been taken to overcome this resistance. One strategy has involved the chemical

modification of naturally occurring glycopeptide antibiotics to produce semi-synthetic

derivatives. An exceptionally successful candidate, designated LY333328, was

developed by scientists at Eli Lilly by integrating a chlorobiphenyl moiety into the

28
structure of the glycopeptide chloroeremomycin, via a reductive amination. This

compound shows activity against vancomycin-sensitive and resistant bacteria and is

29
currently in clinical trials. The Kahne group has also made modifications to the sugar

residues of vancomycin and found the most potent derivatives against VRE also
30-32
contained this chlorobiphenyl group. Another approach involved the alteration of the

amino acid backbone of teicoplanin aglycon which produced a compound, MDL63166,

that displays enhanced effectiveness against VRE (Figure 1-7).33 Further examples of

these semi-synthetic derivatives can be found in recent reviews by Malabarba et al.34 and

Nicolaou et al.


















H N. I = m 2. NaBH3CN H H

K H J H
H 'OH
chloroeremomycin LY333328



H H
IC
II.-


H H H HMDL6366
N N *,NH2 *N N NH
H H H 1 H H H

H ~ H2N
H HHH
teicoplanin MDL63 166
aglycon


Figure 1-7. Semi-synthetic glycopeptide antibiotics.


The investigation of covalently linked dimers and timers of vancomycin to fight

resistant bacteria was inspired by studies showing that dimerization of glycopeptide

3,35-37
antibiotics results in enhanced efficacy. Sundaram and Griffin connected the

carboxyl-terminus of two vancomycin molecules with a diamine linker to produce a

head-to-head dimer that displayed a 60-fold improvement over vancomycin against

38
VRE. The Whitesides lab constructed vancomycin dimers and trimers and studied the

affinity of these derivatives to the dimeric and trimeric ligands, L-Lys-D-Ala-D-Ala and

L-Lys-D-Ala-D-Lac (Figure 1-8). These polyvalent derivatives showed enhanced

binding to the ligands compared to the monomeric species, with the dimer demonstrating


increased ability to combat vancomycin-resistant bacteria.3942














2 H H vH ye 0 Vancomyci H
2.. o ,iS -- -- I,,o0NH>
H H& H Me '-2 "HN-Vancomydn
-oH' 9i H H,: Me'
H2N MPH

H H Vancomycin

Figure 1-8. Synthesis of Vancomycin Dimer.


An alternative strategy to fight VRE is to attack the source of the resistance by

targeting the Van proteins. Information obtained from the extensive mechanistic and

structural studies of some of these enzymes can be used toward the rational design of

inhibitors and several small molecule and dipeptide-like inhibitors have been developed
20,27.43 44,45 46
for these proteins.2743 The Walsh group445 and Yang et al.46 constructed phosphorus-

containing dipeptide analogs and studied their inhibition properties on the VanX protein.

The phosphinate (1-la), phosphonate (1-1b), and phosphonamidate (1-1c) dipeptides

were designed to mimic the tetrahedral transition-state of hydrolysis of the natural D-Ala-

D-Ala ligand and showed Ki values of 0.3 pM, 300 giM, and 36 PM, respectively. Based

on previous evidence that VanX also displays activity against the dipeptide D-Ala-D-Phe,

researchers in France developed the dipeptide-like inhibitor 1-2, which showed a Ki value
47
of 30 pM (Figure 1-9).47 Compound 1-2 is proposed to function as a mechanism-based

inhibitor by forming an irreversible covalent bond with the VanX enzyme, as shown in
47
Figure 1-10.47 Several dithiol compounds, such as dithiothreitol, have also been found to

inhibit VanX by possibly acting as zinc binding ligands, although the exact mode of











0\ 0_- CH3
H2NY P- x C02-
CH3
1-la-c
a: X = CH2
b: X=O
c: X=S


Inhibitor Ki (jipM)
1-la 0.3
1-lb 300
1-ic 36
1-2 30


CHF2

0 S"
IH2N N OH

CH3 0
1-2


Figure 1-9. Inhibitors of the VanX protein.


D-Ala

VanX


NH3 + HCOCO2H


Enz-Nu,

F F- F
SD^J~


Enz- Nu


HS&


Figure 1-10. Postulated mechanism of VanX inhibition by 1-2.


action these inhibitors have not been elucidated.48 The Bartlett group investigated the

inhibitory effect of various dipeptide phosphinates and phosphonates on the VanA

protein and found phosphinates 1-3a and l-3b to be the best inhibitors with K, values of

49
750 nM and 1600 nM, respectively. Compound 1-4, which was previously discovered

as an inhibitor of a two-component signal transduction pathway, was identified as an

inhibitor of the VanR protein (Figure 1-11). Through the use of a phage-display












VanA inhibitors:

0 o.
H2N P" CO2"

CH3

l-3a-b
a: R = CH2Ph
b: R = i-Bu


VanR inhibitor:


SN'
F 4


F

1-4


Figure 1-11. Inhibitors of the VanA and VanR proteins.






0

0 N-o )-N|o
N )- CH/
F'
0
Linezolid


NMe2


CH3 N 0


0 N 0
I -i T N
N0 0 0
0 N J // 0

N Nr
Et2N

Quinupristin Dalfopristin


Figure 1-12. Structures of the antibiotics linezolid and quinupristin/dalfopristin.













S0 NH2
0


0 G 0














HNJ

0

HH
I- 0 N N
0 Co2-


V 0-Lac


IH20
HONH2

0

00
H H
N N
I-HHN ^ INs


0 C02- 0

Figure 1-13. Proposed mechanism of SProC5 catalyzed hydrolysis of the
D-Ala-D-Lac ester bond.








combinatorial peptide library, the dodecamer peptide SLCHDSVIGWEC was also found

to inhibit VanR.51 While several of these compounds have successfully inhibited the Van

proteins in vitro, none are able to pass into the bacterial cytoplasm and function in vivo.20

Other approaches to circumvent resistance have also been investigated. The

development of new antibiotics led to the discovery of two drugs, the oxazolidinone

linezolid and quinupristin/dalfopristin (Figure 1-12), which fight VRE by inhibiting

protein biosynthesis.523 Chiosis and Boneca have recently shown the ability of small

molecules with enhanced nucleophilicity, such as SProC5 (Figure 1-13), to selectively

mediate the cleavage of the D-Ala-D-Lac ester bond of the resistant peptidoglycan. The

proposed mechanism of lactate cleavage mediated by SProC5 is shown in Figure 1-13,

where the formation of an internal hydrogen bond between the hydroxyl group and the

amide carboxyl group enhances the nucleophilicity of the alcohol to form a

transesterification intermediate that is hydrolyzed upon reaction with water. These small

54
molecules in combination with vancomycin show activity against VRE.54


Introduction to Research Goal

The long-term goal of this research project is to develop an alternative strategy to

combat vancomycin-resistant bacteria. The replacement of an amide bond (D-Ala) with

an ester bond (D-Lac) in the peptidoglycan pentapeptide leaves the bacteria vulnerable to

ester hydrolysis. Hydrolysis of the lactate ester would produce a free carboxylate at the

terminus of the pentapeptide, which would prevent subsequent crosslinking and

ultimately lead the resistant bacteria to lyse. The aim of this project is to exploit this

weakness by raising catalytic antibodies to selectively catalyze ester hydrolysis of the

pentapeptide lactate ester (Figure 1-14).












HN H Catalyti HNj HN
_CH3 )-'CH3 Antibody 0=CH3 H=
HH3 OOO= :
020.. -.020.. HONCOH -020"- -020.
0 2 H3
4HN 0 D-Lactate o





Unreactive Free Carboxylate -
N NHNH H H34, H ^ NH N)H3 H ^ NH 0)
03- )0 H3C-< 0 - )0 3- )0
>* 0 0 0 \NH 4 Q7 \NH 14 HA ^ H/
*-CH3 = >-CH3 =OP0<
-02C ^NH2 _02d 0 NH2 OCNg / VCNH

Unreactive Free Carboxylate -
No Crosslinking -
CELL LYSIS
Figure 1-14. Hydrolysis of the ester bond in the peptidoglycan of vancomycin-resistant
bacteria with a catalytic antibody prevents crosslinking and leads the cell to lyse.


Catalytic Antibodies
Antibodies produced by the immune system bind foreign molecules tightly and

selectively. Jencks first proposed in 1969 that by taking advantage of their tight binding

and selectivity properties, antibodies with catalytic ability could be developed.55 Catalysis

is believed to occur by lowering the activation energy (Ea) of the reaction by either

raising the ground state energy of the reactants or by binding and stabilizing the
56
transition-state (Figure 1-15). Based on this theory of catalysis, an antibody that is

generated against a transition-state analog (TSA) will bind preferentially the reaction

transition-state and will thus show catalytic activity. The first examples of catalytic

57 58
antibodies were reported in 1986 from the labs of Schultz57 and Lerner58 for the catalysis

of carbonate and ester hydrolysis, respectively. In addition to these reactions, antibodies

have also been produced that catalyze a variety of other reactions including amide










/ Uncatalyzed rxn

Catalyzed rxn
Ea Activation energy
Ea / \ \ = Transition-state

E/ \







Rxn Coordinate

Figure 1-15. Energy diagram of a catalyzed and uncatalyzed reaction.


hydrolysis, phosphate ester hydrolysis, sigmatropic rearrangements, cycloadditions, aldol

reactions, and transesterifications.592

The most commonly used method for the generation of catalytic monoclonal

antibodies utilizes hybridoma technology.63 This is accomplished by first immunizing a

mouse with the transition-state analog to stimulate antibody production. Because small

molecules are not immunogenic, the TSA must be coupled to a carrier protein, usually

bovine serum albumin (BSA) or keyhole limpet hemocyanin (KLH), to obtain an

antibody response. The antibody-forming cells are then isolated from the spleen and

fused with tumor cells to provide the hybridoma cells. Individual hybridoma cells are

cloned and screened for antibody production and binding affinity toward the desired
64
ligand (Figure 1-16). The antibodies that successfully bind the ligand are then screened

for catalytic activity. An alternative method to generate antibodies uses phage-display









Immunize mouse
with TSA-carrier protein

4,1

0 +
Antibody-forming cells Tumor cells
from spleen |Fusion




Hybridomas
S1) Clone
2) Screen
Monoclonal Antibodies

Figure 1-16. Hybridoma method for the production of monoclonal antibodies.


65,66
libraries. A gene library of the antibody variable fragment is constructed from either

the DNA of immunized or non-immunized mouse spleen cells by using reverse

transcriptase and polymerase chain reaction (PCR) to provide the complementary DNA.

This DNA is used to express the antibody fragment as a protein on the surface of the

phage, which is then screened for binding affinity to the ligand of interest followed by
. 67
catalytic activity.67








0
R0O' \-
OH
I I
OH
Transition-State Analog

Figure 1-17. A phosphonate ester mimics the transition-state of ester hydrolysis.









Table 1-1. Ester hydrolysis reactions catalyzed by antibodies.

TSA Reaction k w_=s
l- 0
/ NHCO(CH2)3COOH -4\ H \/ -& NHCO(CH2)3COOH
H3C H3C
| 83,000
\ OH -02C
+ /. -NHCO(CH2)3COOH
H3C


o 0
0 \ FN'CO-NH-CH2-COOH N' ~CO-NH-CH2-COOH
IOHI
02N 02NO"| 2.6 x 10

0
O2N 'OH- + NO CO-NH-CH2-COOH
I +H'




HO N _. P N"S N v "_N 0 v PNh
"^ ^ *a,/ K H O-H 1 H

OPh 4300

1 I 1J o ,o H I I
~ P N -~ h N N'- N~ OHP


0 0 + o


H
Several antibodies have been developed to act as ester hydrolysis catalysts. Most

of these catalytic antibodies were raised against phosphonate esters, which mimic the

electronics and tetrahedral geometry of the transition-state of ester hydrolysis (Figure 1-

17). Peptidylphosphonate esters have shown to be successful transition-state analogs by

their potent inhibition of zinc peptidases, in which the peptide cleavage also occurs via a

68,69
tetrahedral transition-state. A few examples of ester hydrolysis reactions catalyzed by

antibodies are shown in Table 1-1.70-72









Introduction to Synthetic Goal

In order to develop antibodies capable of catalyzing ester hydrolysis of the lactate

bond in the VRE peptidoglycan pentapeptide (Figure 1-14), a transition-state analog must

be synthesized. The proposed transition-state analog, 1-5, incorporates a phosphonate

ester into the structure of the peptidoglycan pentapeptide. Although the use of a

phosphonate ester dipeptide analog as the TSA may be sufficient to obtain catalytic'

antibodies, the larger TSA should provide antibodies with increased substrate selectivity

due to an increased number of binding interactions. The additional L-cysteine residue at

the N-terminus of TSA 1-5 is included to allow coupling to a carrier protein by utilizing

the reactive sulfhydryl moiety. The acetyl-protecting group at the L-lysine E-amine

mimics the peptide linker, which is attached to the peptidoglycan L-lysine before

crosslinking occurs (Figure 1-4).

Two approaches toward the synthesis of TSA 1-5 are described in this

dissertation. Initially, a stepwise approach using manual solid-phase peptide synthesis

,SH
O 0 C02- 0 CH3
HH
H:3C N f; N1'k N 'N 1 N(0C0
H H H H0 O~
0 CH3 0 0CH3
1-5

S t-Bu Ph HN,1 O
I 0 CH3 OH3
HHiII H 0 0 CO2Et
H N N CO02H + H2N < T
H I _= II O / CH3
0 CH3 0 CH3
1-6 1-7

HN O 0

CH3


Scheme 1-1









(SPPS) was attempted. The details of this work will be discussed in Chapter 3. Several

problems were encountered with this route, primarily in the protection of the first residue,

L-lactic acid, and in the coupling of this residue to the solid-support. These problems led

to the adoption of the successful convergent approach in which the phosphonate fragment

1-7 was synthesized in solution, the tetrapeptide fragment 1-6 was synthesized using

manual SPPS and the two fragments were coupled using solution phase techniques

(Scheme 1-1). The convergent approach will also be discussed further in Chapter 3.


Methods for the Synthesis of Phosphonate Esters

The interest in the utility of phosphonate esters as transition-state analogs has led

to the development of several synthetic methods for the construction of these types of

compounds. A few of these methods are presented in the following sections.

Synthesis of Unsymmetrical Phosphonate Diesters

Several methods have been devised for the synthesis of mixed phosphonate

diesters. In a common synthetic approach, phosphonate diesters are prepared from the



o
II
1R/P-OR2
H S
CC14


o 0 0
I1 PCI5 II R30H
1R -OR2 1R-- NPOR2 ROH \ 1R-OR2
OR2 Cl OR3


SOC2 or (COCI)2
0
II
1Ri'P\-OR2
OH


Scheme 1-2









reaction of alcohols with phosphonochloridates.73 Phosphonochloridates can be

synthesized by oxidative chlorination of phosphinate monoesters with carbon

74 75-
tetrachloride, from symmetrical phosphonate diesters with phosphorus pentachloride,

77and from the reaction of phosphonate monoesters with either thionyl chloride or oxalyl

78.79
chloride (Scheme 1-2). The use of silver ion catalysts in the reaction of

phosphonochloridates with alcohols and amines has been shown to improve reaction


CH3 CH3 CH3
CbzHN P H2 ObzHN .P H Cl CbzHN P C
C OH CH30H 0 OCH3 O3N \OCH3


HO C02CH3 CH3 CH3
CH3 0 o C02OCH3 1) ULSPr, HMPA H3N,. pO, ,00,CO-
54% bzHN 2) H2, Pd/C H* *"-
0 0CH374 0 0- CH3
CH3 1-lb

Scheme 1-3


0 CH3

CH3CHO+ P(OPh)3 + V O NH2 AcOH CbzHN KPh CH3ONa
S600 OPh CH3OH; 72%


CH3 CH3 CH3 HO C02CH3
C .0 'OCH3 NaOH CH OH SOd2Cb 1 CH3
CbzHN ,P\ ObzHNP--- -^bzN P^ ---2--
OC OH3OH; 86% bzHN Et3N 53%
(/ \(XH3 COS/ o/OCH3 CO^CH3^


CH3 CH3
C PHN RO 00CO2CH3 1) LiSPr, HMPA ;60% HN F Y 002-
O C 2) H2, Pd/C H 0'/ O
OH3 CH3
CH3 1-lb


Scheme 1-4







80 49,68 46
yields. The Bartlett group (Scheme 1-3) and Crowder group (Scheme 1-4) have

synthesized the free carboxylate derivative of peptidylphosphonate 1-7 (1-lb) from

phosphonochloridate intermediates.

The (IH-benzotriazol- 1-yloxy)tris(dimethylamino)phosphonium

hexafluorophosphate (BOP) and (1H- 1-benzotriazol- 1-yloxy)tripyrrolidinophosphonium

hexafluorophosphate (PyBOP) reagents commonly used in peptide synthesis have also

been employed to synthesize unsymmetrical peptidylphosphonate diesters by coupling

alcohols to peptidylphosphonate monoesters in the presence of diisopropylethylamine
81.82
(DIEA) (Scheme 1-5).,2 This method was shown to give less than 0.5% racemization

R1 R1
CbzHN L pOH + R3OH BOP or PyBOP CbHN ROR3
CbzHN DIE+A'O ---^- CbzMN'>
0 P OR2 DIEA 0/ OR2


N (N)3
o- P+ 0- P+
c N PF6- /P6-

N C
BOP PyBOP

Scheme 1-5

and better yields than reactions carried out with other coupling reagents used for peptide

synthesis, such as dicyclohexylcarbodiimide (DCC), DCC/4-dimethylaminopyridine

(DMAP), DCC/1-hydroxybenzotriazole (HOBt), and O-(IH-benzotriazol- 1-yl)-

N,N,N',N'-tetramethyluronium hexafluorophosphate (HBTU). The mechanism of the

reaction with PyBOP was studied using methyl phenylphosphonate (1-8) and methanol













1-8
0-


0 0 N
II N- 0



Ph P- MeOH ,= Phl P O + C
PyBOP --- p^-(
OMe OMe N
1-9
o ,,N
Ph" /C _N MeOH P
OMe OMe
1-9
Scheme 1-6

and was found to proceed via the benzotriazole ester intermediate 1-9 rather than a

82
pyrophosphonate intermediate (Scheme 1-6).

83
The Mitsunobu reaction83 has been utilized in the synthesis of mixed phosphonate

diesters (Scheme 1-7).84 This reaction is carried out under mild conditions and provides

0 o
p Ph3P, DIAD II
1R \ OR2H P02
OCH3 R2OH 'R
OCH3

Scheme 1-7
the product in good yields, presumably with an inversion of configuration at the alcohol

attached carbon atom. The proposed mechanism of the Mitsunobu reaction with

phosphonate monoesters is similar to the well-studied mechanism of the reaction with

carboxylic acids, which show that two pathways are possible (Scheme 1-8). In path

A, the betaine 1-10 formed from triphenylphosphine (Ph3P) and diisopropyl

azodicarboxylate (DIAD) is protonated by the acidic component of the phosphonate

monoester (pKa < 11) to provide betaine 1-11, which then reacts with the alcohol to

afford 1-13. The SN2 reaction of the deprotonated phosphonate monoester with 1-13

provides the products. Path B is operable when there is no acidic component present to









Ph3P + ,-PrO2C-N=N-CO2i-Pr



i-PrO2C- N-N- C02i-Pr
0
11 / PPh3
/ OH 1-10 o20H
H OCHA BR2

1 1 -0
R \ II
1-11 J-PrO2C-N-N-COzi-Pr Ph3P, I1-1


OCH3 R21 I'- OH
o ++-O OCH3

00
RHR /O 1-13
OCH3




O
II
'R"\-OR2 + Ph3P=0
OCH3


Scheme 1-8

protonate betaine 1-10. This betaine undergoes reaction with two equivalents of the

alcohol to give 1-12, which then proceeds to products via 1-13.

The Mitsunobu reaction is usually carried out using an excess of alcohol, Ph3P,

and DIAD relative to the phosphonate monoester. Campbell and Bermak developed

efficient alcohol-limiting conditions for the Mitsunobu reaction by adding triethylamine

89
or DIEA and replacing Ph3P with the more electrophilic tris(4-chlorophenyl)phosphine.

These modified conditions were used to synthesize peptdiylphosphonates with solid-

phase synthesis, where the alcohol-attached solid support is the limiting reagent. This

solid-phase peptidylphosphonate synthesis used standard Fmoc (9-

fluorenylmethoxycarbonyl) chemistry (Scheme 1-9); however the phosphonate











Fmoc-NH-(peptide)--@ 1) 30% piperidineINMP
2) HBTU, HOBt, DIEAJFmocO 0CO2H
R1


0

RFmOy N--(peptide)--
R1 H


02N 0 R2 0
1) 30% piperidineiNMP 0 R2 0

2) DIAD, DIEA, H I0 E H
H101 0 I H 1
CH3
2Ny 0 R2
10"- ,.',.O N J,-.p O H
H 0" 'OCH3

R2 0
1) 5% DBU/NMP H2N-(peptide)--N 'OP N-(peptide --
2) N-Fmoc-amino acid. HBTU, HOBt, DIEA H 0/ 0 H "W
3) 30% piperidine/NMP CHR
CH3

R2 0
1) 1:2:2 PhSH-Et3N-dioxane H N (d )-'O. A
----- -- --- *H2N-(peptide-N .R P N- 2) scavengers, TFA H 0/ \OH H
R1


Scheme 1-9

monoester was protected with the 4-nitrophenethyloxycarbonyl group because under

90
Mitsunobu conditions the Fmoc group was cleaved via P-elimination by betaine 1-10.9

Mixed phosphonate diesters have also been synthesized from phosphonic

dichlorides by the separate addition of different alcohols and using tetrazole as a

9'
catalyst.91 Hirschmann and coworkers have shown the use of

phosphonyltriethylammonium salts as reactive compounds to produce phosphonate

diesters.92
diesters.









Synthesis of Phosphonate Monoesters

Phosphonate monoesters can be synthesized by base hydrolysis of symmetrical

phosphonate diesters. This method provides the monoester rather than the phosphonic

acid because of electrostatic repulsion between the resulting monoester and the incoming

hydroxide ion (Scheme 1-10). Hydrolysis of unsymmetrical phosphonate diesters often

o 0
II It
RP-OR2 -OH -OR2 X-H
OR" o-
OR2 0-

Scheme 1-10

results in a mixture of the two possible monoesters and therefore has limited utility.

However, an example of hydroxide-mediated cleavage of a mixed phosphonate diester

will be discussed in Chapter 3 in which reasonable selectivity was obtained, presumably

due to steric factors.

Bromotrimethylsilane (TMS-Br) has been found to show selective cleavage of

93
methyl esters of unsymmetrical phosphonate diesters by nucleophilic displacement.93

The relative reactivities of TMS-Br with dimethyl, diethyl and diisopropyl

acylphosphonates are approximately 1: 0.25: 0.04. The more reactive iodotrimethylsilane

does not show this selectivity and chlorotrimethylsilane shows little or no reactivity. This

selective demethylation of phosphonate diesters can also be accomplished with thiolates,

such as thiophenoxide, thioethoxide, and propanethiolate, and t-butylamine (Scheme 1-

).94-97
11).

t-butylamine
0 / \ o
II I
R,, OR2 TMS-Br R \
OCH3 OH
S\ _/

Scheme 1-11








Solid-Phase Peptide Synthesis

Solid-phase peptide synthesis, synthesis by the sequential addition of amino acids

to an insoluble polymeric support, was introduced by Merrifield in 1963 in order to

bypass some of the problems of solution phase synthesis ofpeptides and was one of the

first examples of polymer-supported chemistry.9 Since its inception SPPS has shown to

be advantageous due to the elimination of isolation and purification steps for intermediate

products and the ability to accelerate and automate the synthesis process. The general

strategy of SPPS is shown in Figure 1-18, where an amino acid with the amino group and

the side chain protected orthogonally is attached to the polymer support (resin) from the

C-terminus. The amino group is selectively deprotected and then coupled to the next

amino acid in the sequence. The deprotection and coupling steps are repeated until the

desired peptide has been synthesized. The peptide is then cleaved from the solid-support

with simultaneous removal of the side-chain protecting groups. One problem often

encountered with SPPS is the formation of internal deletion peptides, which are shorter

by one or more amino acids than the desired sequence and result from an incomplete

coupling or deprotection step during the synthesis. This problem is avoided by obtaining

high yielding reactions by adding an excess of reagents. An additional step in the

synthetic scheme is also included in which an unreacted amino group is capped with an

acetyl group to block the formation of a deletion sequence during the coupling of the
99
subsequent residue.

The most common N-a-protecting groups used in SPPS are the tert-

100
butoxycarbonyl (Boc) group, which is cleaved with trifluoroacetic acid (TFA) and the

9-fluorenylmethoxycarbonyll (Fmoc) group, which is deprotected with piperidine










side-chain protecting group

__________R 0
A1 Q1
i -- I II
N-a-protecting group -N-C-C-OH +
--H H


IlinkerI ply e|


Attach to linker
I side-chain protecting group

'R 0
N-a-protecting group N-C-C-0I r-
H H


Deprotect amino group

side-chain protecting group

I I 1
R' 0

H


R2 0
N-a-protecting group -N-C-C-0- activating group
H H


Couple


side-chain protecting group

R20 R' 0
N-a-protecting group -N-C-C-N-C--C--Op
I ---H H H H -- '


Repeat deprotection
Sside-chain protecting group and coupling steps

__________R f 2 0 1R' 0 __ __
N-a-protecting group -N-C-C N-C-CO- ner I polymer
'-----------'\H H / H H l_ -_' ,
\ n
SCleave

R2 0\ R' 0
+2N-C- ColN-C-C-OHHr +
\ /n



Figure 1-18. General scheme for solid-phase peptide synthesis. Figure
adapted from reference 108.









Boc Fmoc
r----- A I
0 R 0 R
OkN JC02H O'kN JC02H
H 06^1 H



Scheme 1-12
(Scheme 1-12). The peptide-resin bond must be stable under these N-a-deprotection

conditions, thus in the Boc strategy the peptide is cleaved from the resin with hydrogen

fluoride (HF) and with the milder TFA when using the Fmoc strategy. Of the several

polystyrene-based resins developed for SPPS, the Wang (4-alkoxybenzyl alcohol) resin102
103
and the PAM resin are most frequently used for peptide synthesis with Fmoc and Boc

chemistry, respectively (Scheme 1-13).




HO -^
Wang Resin PAM Resin


Scheme 1-13

The peptide is assembled on the solid-support with coupling reactions in which

the carboxyl group of the N-a-protected amino acid is activated. A variety of methods

have been developed to perform this coupling reaction. Carbodiimides are widely used

reagents in solution and solid-phase peptide synthesis and react with N-a-protected

amino acids to form the O-acylisourea 1-14, which then reacts with the amino component

to form the peptide bond. When two equivalents of the N-a-protected amino acid is used

relative to the carbodiimide, the second equivalent of the amino acid can react with 1-14








to form a symmetrical anhydride (1-15) that also subsequently reacts with the amine

group. Dicyclohexylcarbodiimide (DCC) and diisopropylcarbodiimide (DIC) are

primarily the carbodiimides used in peptide synthesis, however DIC is chosen over DCC

for SPPS because of the solubility of the resulting urea in the reaction solvent. The

formation of active esters (1-16) from O-acylurea 1-14 with 1-hydroxybenzotriazole

(HOBt) is an alternative and efficient method for peptide bond formation (Scheme 1-14).


ol
0
2RA OH +RNCNR


^0
2R.kO
1 I
0 R-N=C-N-R'
0 H
R H, 1-14
2R OH \ HOBt


0 0 H2NR30
2RW, 0 R2 2R OBt
1-15 1-16

H2NR3 H2NR3

2R NHR3

Scheme 1-14

These HOBt active esters can also be formed in situ with the reaction of the N-a-

104 1016
protected amino acid with the coupling reagents BOP PyBOP105, and HBTU106 in the

presence of a tertiary base. The addition of an extra equivalent of HOBt can accelerate

h 99,107,108
the reaction further. The coupling reaction can also be performed by the










(9)3
0- P+ OH
NN PF6- (V N=C=N
/N LZJL^ vN7
PyBOP HOBt DCC



0 CI (
NI
..-N I

KN)^ 0U, P. NCOEt
CH3 O- CO2Et
IBCF NMM HBTU EEDQ


Figure 1-19. Structures of reagents used in coupling reactions.

formation of reactive mixed anhydrides, which is often accomplished by using isobutyl

chloroformate (IBCF)19 or isopropyl chloroformate with a tertiary base.


Segment Condensation
Large peptides can be synthesized using solution phase synthesis by dividing the

sequence into segments and coupling these fragments together using the methods

described above. Fischer and coworkers have investigated different reaction conditions

for the segment condensation of a tripeptide with a dimeric dipeptide.I I The coupling

reaction was studied with several reagents (Figure 1-19), including PyBOP/HOBt/tertiary

base, PyBOP/HOBt, DCC/HOBt, IBCF/N-methylmorpholine (NMM),

HBTU/HOBt/DIEA, and 2-ethoxy-1 -ethoxycarbonyl- 1,2-dihydroquinoline (EEDQ). The

reactions with PyBOP/HOBt/tertiary base and HBTU/HOBt/DIEA gave moderate yields,

but resulted in a high amount of racemization, which was detected and quantitated by RP-

HPLC to give 16% and 24%, respectively. Racemization was lowest for the reaction










O H
H2N "o"A N N" ,C'O2CH3
0z 0 =0 "
/c -v O< =
CH3



0H
N,(K CO2H IBCF
NOY
/H zI




H 0 Q0 iN
N o o" ,
H I H Ko H I "
^ ~CH3 l)'



Scheme 1-15

with IBCF/NMM at 0.5%, however the reaction yield was only 16%. The results of the

reaction with EEDQ, a reagent that forms mixed anhydrides, also exhibited a low amount

of racemization but provided a higher yield. Morgan et al. also found a solution-phase

coupling reaction with DCC/HOBt resulted in racemization, while none was obtained
withIBCFNMM112
with IBCF/NMM. The use of IBCF in the segment condensation of a tripeptide with a

peptidylphosphonate has been demonstrated (Scheme 1-15). 13

The use of the synthetic techniques discussed in this chapter toward the synthesis

of transition-state analog 1-5 will be discussed in the following chapters.















CHAPTER 2
MODEL STUDY OF THE MITSUNOBU REACTION AND SIDE-CHAIN
DEPROTECTION STEPS FOR THE TRANSITION-STATE ANALOG SYNTHESIS

In principle, the synthesis of transition-state analog 1-5 could be accomplished

with a stepwise or convergent approach (Scheme 2-1). We had originally planned to

employ manual solid-phase peptide synthetic techniques to build the TSA on a solid-

support resin from the C-terminus; however, we later adopted a convergent approach.

This latter route required the synthesis of peptide fragment 1-6 and phosphonate fragment


-SH
0 H 0
S I -" H II : H 61, 0.-
0 CH, 0 CH

1-5
Stepwiset /0N, Convergent
Approach// n, \ A approach


t-Bu Ph
SBn,
0 f 0 CH3 0 0S 0
II H | 0 .0 H
H3C N N N O' N Nv p.O COEt
0 CH. 0 0 0N 0 C o 5 0 0 H
3 H 0 C 0H3 0 0CH






OH, 2-6
2-1 2-5
HNS 0 0
11 Ph CH3
0 n 0 I 9H
j0 9H3 0 0 ( 5' 9 0 HSM
H3C N 1,PkO, HCN1k NC0,H H2
Hlw~ ,OH,^^<^ HW -T0 00"y
0 CMH 0 0 0 6H, 0 OH,
2-2 S 1-6 S 1-7
I N I 0 HN 0
jJ 0CH CH, 00H3 143
OzNCH, 0 N .- .O H + H 0, O E t
N 0* 0083
O CH, CH,

2-3 2-46


Scheme 2-1








1-7, then coupling these fragments along with subsequent side-chain deprotection steps.

Both routes utilize the Mitsunobu reaction between L-lactate and a phosphonate

monoester, which was predicted to result in inversion at the lactate residue to provide the

desired D-isomer, allowing the less expensive L-lactate isomer to be used as the starting

material for this reaction. The final steps in the synthesis of 1-5, using either the stepwise

or convergent method, involve the cleavage of the side-chain protecting groups of 2-1,2-

2, and 2-5. It is essential that the phosphonate ester bond remain intact under these

deprotection conditions. Thus, a model study was carried out with methyl methyl

phosphonate (2-8) in order to optimize the Mitsunobu reaction conditions for the

synthesis of 1-7 and to study the stability of the phosphonate under conditions for the

selective cleavage of the phosphonate methyl ester and carboxylate ester saponfication.


Mitsunobu Reaction with a Monomethyl Phosphonate Ester

Methyl methyl phosphonate 2-8 was prepared by esterification of

methylphosphonic dichloride with methanol to provide 2-7, which was then selectively

hydrolyzed to the monomethyl ester according to the procedure developed by Christol et



O\ /Cl CH3OH 0 /OCH3 4% NaOH, dioxane
H3C Cl 68% H3c *OCH3 95%
2-7


CH3 C02-Pr CH3
1N=N OH3
0^ .OCH3 HO' ECO..t -PrO2C' 0 0 \. 3
H3C" 'OH/ / \H3C 0 'O CO2Et
2-8 C I_( -P ;79% 2-9
C /


Scheme 2-2








114
al. (Scheme 2-2). The Mitsunobu reaction of the monomethyl phosphonate ester 2-8

with (L)-ethyl lactate was initially carried out according to Campbell with

triphenylphosphine and diisopropyl azodicarboxylate (DIAD) to afford the mixed

84
phosphonate diester 2-9. However, the yield of the Mitsunobu reaction was improved

when tris(4-chlorophenyl)phosphine was used in place of triphenylphosphine due to an

improved separation between the product (2-9) and the resultant phosphine oxide during

purification by column chromotagraphy. The product was obtained as a 1:1 mixture of

diasteromers as a result of the newly formed chiral center at the phosphorus moiety.

The stereochemical configuration of the lactate residue was examined because in

addition to the normal Mitsunobu mechanism to provide 2-9, an alternative pathway for



Ar3P + -PrO2C-N=N-CO2i-Pr



FPrO2C-N-N-CO2i-Pr
I H
+ PAr3
2-10


OH3 O\ OCH3

Ar3P-O "CO2Et H3C O-PAr3
2-11 2-12


Op /OCH3 inversion Retention ICH3
H3C' 0- HO CO2Et

CH3 CH3
0~ 0 CN
/ -H3 cx O
P-'
H3CP 0 CO2Et H3C" 0 PO"CO2Et
2-9 2-13


Scheme 2-3










CO2iPr
N=N
, -PrO2C
Ph3P


No Reaction


Scheme 2-4
the reaction is possible where the phosphonate is activated by the phosphine (2-12)

followed by nucleophilic attack by the lactate alcohol, which would result in retention of

configuration (2-13) (Scheme 2-3). Campbell demonstrated this alternative pathway was

not likely by observing that no reaction occurred when methyl benzyl phosphonate (2-14)

was reacted with tert-butyl alcohol, DIAD and triphenylphosphine (Scheme 2-4).

Mitsunobu reactions do not usually take place with tertiary alcohols, thus any formation

of the t-butyl phosphonate ester would have been the result of nucleophilic attack by the

84
alcohol moiety. While this evidence suggests that the coupling of 2-8 and (L)-ethyl



CH3 CO2i-Pr
1 N=N
O9\ /OCH3 H1tCO2Et PrO2C6N=N
H3C" "-OH /
2-8 (C-0 P


OH3
0 cH3
H3c' p.. 1 CO2Et
2-16


or


+ \ /P=
2-17


RETENTION


Scheme 2-5


ON /CH3 HO0
Bn P" OH P
2-14


CH3
0 /H3 8
SN + \C /-- P- INVERSION
H3C P0 CO2Et 3
2-9 2-15








lactate proceeded by the typical Mitsunobu mechanism, it was not conclusive. To rule

out a pathway involving phosphonate activation, a labeling experiment was performed.

The Mitsunobu reaction between oxygen-18 (180) labeled (L)-ethyl lactate and

monomethyl phosphonate ester 2-8 would provide the 18O label in the phosphine oxide

product (2-15) if the reaction occurred as expected. By contrast, the 180O labeled

Mitsunobu product (2-16) would be produced with retention of configuration at the

lactate residue if the alcohol moiety acts as the nucleophile (Scheme 2-5).

Oxygen-18 Labeling Experiment

Oxygen-1 8 labeled (L)-ethyl lactate was synthesized (Scheme 2-6) from 180

labeled pyruvate (2-18), which in turn was prepared by stirring sodium pyruvate with

H2 O80 for two days at room temperature. 5 This reaction has been reported to label only

the keto-position (CH3C8OCOONa). 116 Reduction of the ISO sodium pyruvate with L-

lactic dehydrogenase and NADH117 resulted in the sodium (L)-lactate (2-19) with
OH3 O H3 P
OH3 Na H21- O nH3 L-Lactic Dehydrogenase
O=aONa
0 rt, 2 days 18,-'6 NADH
0 0
2-18

OH3 CH3CH2I, 15-Crown-5 H3
H0 CH3CN H18 CO2Et
0
2-19 2-20

Scheme 2-6
approximately 33% 180O labeling, as determined by HPLC/electrospray ionization mass

spectrometry. Subsequent esterification of the labeled lactate with iodoethane and 15-
118 1
crown-5 afforded the 0O (L)-ethyl lactate (2-20) and the amount of labeling was

measured with GC-MS. The M+l and M+2 peak intensities of the m/z 75 fragment ion










possible m/z 75 fragment ions from ethyl lactate

H H3C
oi I.< CH2
.HO -O HO 0 2
OH
2-21 2-22

Scheme 2-7


unlabeled
(L)-ethyl lactate
75




















77 %(M+2) = 0.76%


A
75


"S0 labeled
(L)-ethyl lactate
75
















77 %(M+2) = 28.1%





B-
75


Figure 2-1. Mass spectra ofm/z 75 fragment peak from unlabeled (L)-ethyl lactate (A)
and 8O labeled (L)-ethyl lactate (B).
119
peak (2-21 or 2-22) of unlabeled (L)-ethyl lactate (Scheme 2-7) were compared with

the same peaks from the labeled material (Figure 2-1). The %(M+l) was 3.25% + 0.26








and the %(M+2) was 0.76% + 0.11 for the (L)-ethyl lactate standard and 3.7% 0.4 and

28.1% + 2.6, respectively, for 2-20. This corresponds to approximately 28% 180 labeling

in 2-20. After 2-20 was diluted with unlabeled (L)-ethyl lactate to provide a sample with

roughly 15% 180, the Mitsunobu reaction was carried out as before and the isolated

phosphonate and phosphine oxide were examined by mass spectrometry for 180O content.

The M+l and M+2 intensities of the m/z 93 and 137 fragment ion peaks, which

are assumed to have the structures of 2-23 and 2-24 (Scheme 2-8), obtained from the GC-


Proposed fragment ions from 2-9
miz 93 m/z 137
CH3
0 %v n u
0 0
II o,\, OH
H3C-" POCH3 H3C" ,O,
+ +
2-23 2-24

Scheme 2-8
MS of unlabeled Mitsunobu product were compared with the product obtained from the

180 labeling experiment (Figure 2-2). For the unlabeled Mitsunobu product the %(M+l)

and %(M+2) for the m/-z 93 peak were 15.01% + 0.89 and 3.64% + 0.84 and for the m/z

137 peak were 13.67% + 1.15 and 3.92% + 0.21, respectively. The Mitsunobu product

isolated from the reaction with 18O (L)-ethyl lactate resulted in a %(M+1) of 15.44% +

1.66 and %(M+2) of 4.07% + 0.38 for the m/z 93 fragment peak and 13.46% 0.81 and

4.85% + 0.65 for the m/z 137 fragment, respectively. These M+l and M+2 peak

intensities for the labeled and unlabeled Mitsunobu product were within experimental

error of each other, thus no 180 label was incorporated into the product obtained from the

labeling reaction (2-9). The tris(4-chlorophenyl)phosphine oxide (2-15) isolated from the

labeling experiment was found by FAB-MS to contain 8% 1O label by determining the









unlabeled
Mitsunobu product 2-9
93



















95 %(M+2) = 3.64%
Aa
93
137

















139 %(M+2) = 3.92%

140
140


4itsunobu product 2-9
from labeling reaction
93


95 %(M+2) = 4.07%


137
















/139 %(M+2) = 4.85%

140
140


Figure 2-2. Mass spectra of m/z 93 and 137 fragment peaks from unlabeled Mitsunobu
product 2-9 (A & C) and Mitsunobu product 2-9 from labeling reaction (B & D).











unlabeled
tris(4-chlorophenyl)
phosphine oxide

381 UsI


ISO0 labeled
tris(4-chlorophenyl)
phosphine oxide


ISO label = 8%


382 382

Figure 2-3. Mass spectra ofm/z 381 peak from unlabeled (A) and 10 labeled (B)
tris(4-chlorophenyl)phosphine oxide (2-15).

amount the M+2 intensity of the m/z 381 peak increased compared to an unlabeled tris(4-

chlorophenyl)phosphine oxide sample (Figure 2-3). Thus, the results of the labeling

experiment suggest that the reaction proceeded with inversion of configuration at the

lactate residue resulting in the desired D-isomer because the phosphine oxide 2-15

contained the 180 label, although the amount was lower than expected, and no label was

seen in the Mitsunobu product 2-9 (Scheme 2-5). In order to further verify the lactate

stereochemistry an additional experiment was performed using chiral gas

chromatography.








Chiral GC Experiment

Before the Mitsunobu product 2-9 was analyzed by chiral-phase GC, a racemic

standard racemicc 2-9) was synthesized from the Mitsunobu reaction between 2-8 and

racemic ethyl lactate. Only three of the four diastereomers were resolved in the gas

chromatogram of racemic 2-9 (Figure 2-4A) and the retention times of the peaks were

56.76, 57.70, and 58.23 minutes, with two of the diastereomers overlapping at 57.70

minutes. The GC of 2-9 (Figure 2-4B) contained two peaks for the two diastereomers at

57.60 and 58.19 minutes. This chromatogram demonstrated that only one mechanism

was occurring in the Mitsunobu reaction, which resulted in either all inversion product

(2-9) or all retention product (2-13). In order to determine the stereochemistry of the

product obtained, a coupling reaction was performed with 2-8, (L)-ethyl lactate and

PyBOP. This type of coupling reaction with PyBOP has been shown conclusively by
82
Campagne et al. to occur solely with retention of configuration at the alcohol center,

therefore the reaction provided the L-isomer at the lactate center of 2-13. The chiral GC

of the PyBOP coupled product 2-13 should show different peaks than those obtained

from 2-9. The gas chromatogram of 2-13 (Figure 2-4C) contained two peaks for the two

diastereomers at 56.62 and 57.73 minutes. In order to further demonstrate that 2-9 and 2-

13 contained a different stereochemistry at the lactate center, a co-injection was carried

out and the gas chromatogram (Figure 2-4D) was identical to the GC of racemic 2-9

(Figure 2-4A), with peaks at 56.76, 57.70 ad 58.20 minutes. Thus, it was concluded that

the Mitsunobu reaction provided the desired D-isomer at the lactate center of 2-9.













CH3

H3C" P O' CO2Et
racemic 2-9


CH3
O\ /O0H3 HO "CO2Et i-PrO2C
H3C" P-OOH
2-8 ( \/(
C


C02i-Pr
N=N


CH3
O, /OCH3 HO 'CO2Et, PyBOP
H3C 'OH DIEA, DMF
2-8


CH3
X CH3
01, 10 T3
H3C' P'0 CO2Et
2-9
Inversion


CH3
0 O H3
H3C" P"O' 0 CO2Et
2-13
Retention


56.76


CH3
Ox /O -CH
"<0m "


CH3
+ 0 0 r.3


58.20


H3CO 0 CO2Et H3C- O-'COOEt
2-9 2-13 D


Figure 2-4. Chiral gas chromatograms ofracemic 2-9 (A), 2-9 (B), 2-13 (C), and
a co-injection of 2-9 and 2-13 (D).


57.70


56.76







57.60


58.23


58.19


56.62


1LB


57.73


IUc
57.70









Side-Chain Deprotection Steps

Following the optimization of the Mitsunobu reaction and determination of its

stereochemrnical outcome, the conditions for the cleavage of the side-chain protecting

groups of TSA 1-5, especially the selective demethylation of the phosphonate methyl

ester and saponification of the carboxylate ester, were examined.

Phosphonate Diester Demethylation with Bromotrimethylsilane

Bromotrimethylsilane (TMS-Br) demethylation of the unsymmetrical model

phosphonate diester 2-9 was investigated initially (Scheme 2-9). The reaction was


CH3 TMS-Br, THF Na Na
Na 3C 60 CH Na+
0 OH3 60% 0-. CH3 0 0- CH3
HC- O26EtO .. O" P a 0.2 M UOH '
H3CP" 0'CO2Et H3CP" 0 CO2Et Dowex" H3CP 0 CO"Na+
2-9 TMS-Br / 2-25 cation-exchange 2-26
\_TMS-Br, DMF/ "column *v
isobutylene; 62% 94%


Scheme 2-9
carried out in THF using an excess of TMS-Br followed by workup to form the sodium

salt, which afforded the phosphonate monoester 2-25. 4,2 Because this demethylation

reaction with a peptidylphosphonate, such as 2-2 or 2-5, would be performed in DMF, the

reaction was attempted with this solvent using the same conditions as described for THF.

The demethylation in DMF resulted in the monoester 2-25 in addition to the product

obtained from cleavage of both ester linkages, which is possibly due to the formation of

HBr in the reaction. This problem was avoided by using an equimolar amount of TMS-

Br and bubbling the reaction mixture with isobutylene 3 to consume any HBr produced.

Saponification of Carboxylate Ester

The lactate ethyl ester of 2-25 was hydrolyzed with a 0.2 M aqueous solution of

LiOH121 and the resulting product was passed through a cation-exchange column to form









52.09





CH3 52.65
O ) cH3 51.12
H3C O' CO2CHa3
racemic 2-27
A


51.92 52.53



Na+ CH3
0 +- CH3 O /) 'CH3
OP /O -3 CH31,15-crown-5 -
H3C 0P CO-Na+-- H3C 0 CO2CH3
2-26 2-27

nopeakat51.1 min >B



51.90
52.52

CH3 N=NcO2Pr CH3
N=N \ o cH
0 /OCH3 HO CO2CH3 ,-PrO2C 0 =
H3C0 POH H3c' O"0 CO2CH3
2-8 2-27 standard


c

Figure 2-5. Chiral gas chromatograms of racemic 2-27 (A), 2-27 (B), and
a 2-27 standard (C).
the disodium salt 2-26 (Scheme 2-9). In order to determine whether any racemization

occurred at the lactate residue of 2-26 during this saponification step, the stereochemical

configuration of this center was examined. Derivitization of 2-26 was required in order








to analyze the configuration by chiral-phase GC. The phosphonate and carboxylate

methyl esters were prepared by esterification of 2-26 with methyl iodide and 15-crown-5

to provide 2-27 (Figure 2-5B). A racemic chiral GC standard racemicc 2-27) was

synthesized from the Mitsunobu reaction between 2-8 and (D, L)-methyl lactate. Three

of the four diastereomers of racemic 2-27 were resolved in the GC (Figure 2-5A) with

retention times of 51.12, 52.09 and 52.65 minutes. Two of the diastereomers were

overlapping at 52.09 minutes. The gas chromatogram of 2-27 (Figure 2-5B) showed only

two peaks for the two diastereomers at 51.92 and 52.53 minutes. Because no peak at 51

minutes was seen in the GC of 2-27 it was concluded that racemization did not occur at

the lactate center during the saponification of 2-25. To further verify the stereochemistry

of the lactate center of 2-27 a standard (2-27 standard) was synthesized from a

Mitsunobu reaction between 2-8 and (L)-methyl lactate and the GC (Figure 2-5C)

corresponded to the GC of 2-27 (Figure 2-5B), with peaks at 51.90 and 52.52 minutes.

This model study allowed the optimization of the Mitsunobu reaction and

demonstrated the desired lactate configuration was achieved. The stability of the

phosphonate under demethylation and saponification conditions was also illustrated in

addition to the lack of racemization obtained during the saponification step. These

optimized steps were then used toward the synthesis of transition-state analog 1-5.













CHAPTER 3
SYNTHESIS OF THE TRANSITION-STATE ANALOG

Catalytic antibodies raised against transition-state analog 1-5 might be used to

combat vancomycin-resistant bacteria by catalyzing ester hydrolysis of the lactate bond

of the bacterial peptidoglycan. The necessary synthesis of 1-5, in order to produce these

antibodies, could have been approached from several directions, with a total solution-

phase route or a total solid-phase route as two synthetic extremes. A total solid-phase

synthesis was initially adopted due to the advantages of SPPS, such as the elimination of

purification steps and efficient and high yielding reactions. However, protection of the

first residue in the synthesis, L-lactic acid, proved to be difficult and the incompatibility

of the protecting group ultimately used with the conditions for the coupling to the solid-

support provided an additional problem. These problems led to the employment of an

alternative route, in which the synthesis of the TSA was approached convergently. A

peptide fragment and a phosphonate fragment were synthesized with SPPS and solution-

phase synthesis, respectively, and the two pieces were then coupled together in solution.

This route allows the benefits of SPPS to still be utilized for the peptide fragment

synthesis and avoids the problems associated with the protection of the L-Lac residue,

which is necessary for SPPS, by synthesizing the phosphonate fragment in solution. The

details of the initially attempted solid-phase approach and the ultimately applied

convergent approach toward the synthesis of TSA 1-5 will be discussed in this chapter.









Solid-Phase Synthetic Approach to the Transition-State Analog
90
The methodology developed by Campbell and Bermak for the solid-phase

synthesis of peptidylphosphonates facilitated the construction of these compounds for

their use as transition-state analogs. This method inspired the solid-phase synthetic

approach to TSA 1-5, which would provide the traditional benefits of SPPS. The

retrosynthetic analysis of this approach is shown in Scheme 3-1, where the TSA was to


SH
0 O CO 2- H 0 CH3
N J, N,, 11" .10C02- >
H3C N'- N N P ,
H H I H "> I.
0 CH3 0 0 -, CH3

1-5
HN O

CH3
SBn
0 H 0 CO2CH3 0H CH3 0
aCNN N 0)".N
H H II H
0 CH3 0 CH3
S CH3
2-2
HN O

CH3
02N 0 OH3 0
o-- P .o.o ,


2- OH3





HO~

CH3
2-4


Scheme 3-1








be synthesized on the Wang resin using Fmoc chemistry. The Mitsunobu reaction

between the L-lactic acid (L-Lac) on the resin (2-4) and the 4-nitrophenethyloxycarbonyl

(NPEOC)-protected aminophosphonate would result in the required D-lactate isomer. In

order to couple L-Lac to the Wang resin the alcohol moiety must first be protected with

the Fmoc group.

Protection of L-Lactic Acid

Initial attempts to protect (L)-lactic acid using Fmoc-Cl and pyridine122 led to the

recovery of only starting material. When the reaction was carried out according to

Campbell and Bermak under the same conditions using (L)-ethyl lactate as the starting

material, the Fmoc protected (L)-ethyl lactate was synthesized. Attempts at ethyl ester

cleavage of this product under acidic conditions (1 M aqueous HC1 and dioxane) resulted

90
in both ethyl ester hydrolysis and cleavage of the Fmoc group. Thus, a different

protecting group had to be selected for the L-Lac residue.

The protecting group used for the alcohol moiety of L-Lac was selected carefully,

because the linkage between the resin and lactate must be stable to the conditions of the

subsequent cleavage of this protecting group. Carboxylates attached to Wang resin are

cleaved by anhydrous TFA; therefore any hydroxyl protecting groups that are deprotected

with acid could not be used. In addition, protecting groups that are removed with basic

hydrolysis or catalytic hydrogenation were also not considered because L-Lac would be


0 O O
HO "AO TBDMS-CI, imidazole TBDMSO ,'- 0.2 MUOH TBDMSO -J,-
DMF, 96% O THF, 80H
CH3 H3 CH3
3-1 3-2


Scheme 3-2









attached to the Wang resin via a benzyl ester bond (2-4) and these conditions would

destroy this linkage. After taking these restrictions into consideration, the tert-

butyldimethylsilyl (TBDMS) group, which is deprotected with tetrabutylammonium

fluoride (TBAF), was chosen. TBDMS-L-lactic acid (3-2) was synthesized according to
121
Mayer et al. by protecting (L)-ethyl lactate using TBDMS-C1 and imidazole, followed

by hydrolysis of the ethyl ester (Scheme 3-2).

Coupling to the Wang Resin

The attachment of the first monomer to the Wang resin can be achieved by several

108
methods.8 The mixed anhydride approach developed by Sieber is often used because it

eliminates racemization and dipeptide formation, two problems that can be encountered

in the loading of the first residue. 23 The initial attempt to couple the TBDMS-protected

L-lactic acid to the Wang resin used the Sieber method, in which the coupling was carried

out with 2,6-dichlorobenzoyl chloride and pyridine in DMF (Scheme 3-3). The reaction


CI 0


0 -0 -'' pyridine .I "
-00
OH X 'A
CH3
HO DIC/DMAP OH3 3-3a-b
a: X =-NHFmoc
b: X = -OTBDMS

Scheme 3-3
was monitored for the appearance of the ester carbonyl band of the product at -1735 cm'1

in the IR spectrum of the resin. Unfortunately, no carbonyl band was seen in the IR,

indicating that the reaction was unsuccessful. A model study for this method was

performed under the same conditions with Fmoc-protected L-alanine (Fmoc-Ala). Once








again, according to the IR spectrum, no coupling was achieved. An alternative and also

frequently used procedure for the coupling was then performed using 3 equivalents of

Fmoc-Ala and DIC and 0.3 equivalents of DMAP, relative to the resin, in
124
dichloromethane. The amount of coupling was quantitated by measuring

spectrophotometrically the amount of piperidine/dibenzofulvene adduct (E290nm = 4950 L

mol"1 cm') released after Fmoc deprotection (Scheme 3-4). This method provided a 10%



0 -~0
\~ _^qh' 4- H1 ^ s ^fz
0/11\ -1j o -)------ 1 lo I /+ 2N +0C02
H
-0 CH3 CH3
H H







N
0 0







piperidine/dibenzofulvene
adduct

Scheme 3-4
coupling yield, although the IR spectrum of the resin from this reaction did not show a

carbonyl band. However, when 3 equivalents of Fmoc-Ala, 6 equivalents of DIC and 0.1

equivalents of DMAP in DMF were used relative to the resin the coupling reaction was

successful (3-3a), as shown by the carbonyl band at 1726 cm' in the IR spectrum (Figure

3-1). The reaction was repeated with this resin to maximize the amount of coupling,

which provided a 72% anchoring yield of Fmoc-Ala to the resin.














Sj I i

/- x /i
'I 'I




s- \ I




V f
s-A ,; >g!______ g s g s s







carbonyl -1 *~I croy
Sand I




'" B -0 I" IJ Oi
., . ,, ,V ,,, ....
L/I S. O*, n,"fl-'l n n' -






Figure 3-1. Infrared spectra of Wang resin (A) and Fmoc-Ala attached
Wang resin (B).


The TBDMS-protected L-lactic acid was coupled to the Wang resin (3-3b) using

DIC and a catalytic amount of DMAP, as determined by the appearance of the carbonyl

band at 1746 cm- in the IR spectrum of the resin. The coupling reaction was repeated to

maximize loading on the resin, but only a 20% level of attachment of the TBDMS-L-Lac

was achieved, which was determined by mass recovery after cleavage from the resin with

1:1 TFA:CH2CI2. A double coupling reaction was repeated with the L-Lac residue and

the resin using a higher concentration of reagents in an attempt to increase the anchoring








yield; however after the residue was cleaved from the resin, only 14% coupling resulted.

The material isolated from this cleavage reaction contained not only TBDMS-L-Lac, but

also some unprotected L-lactic acid, suggesting that the TBDMS group was removed

during cleavage from the resin with TFA, as expected. However, TBDMS cleavage

during the coupling reaction was also a possibility. In order to determine whether the

TBDMS was cleaved during the coupling step, the Wang resin and the protected L-Lac in

DMF-d7 were shaken for one hour. The 'H NMR spectrum of the solution indicated that

more free L-lactic acid was present than TBDMS-L-Lac. The next step in the synthetic

plan would be to acetylate any unreacted resin hydroxyl groups to prevent the formation

of deletion sequences. The premature loss of the TBDMS group, however, would result

in the alcohol moiety of the L-Lac also being capped and thus unable to undergo the

Mitsunobu reaction in the next step. This problem demonstrated that the TBDMS group

was an unpromising choice for the protection of L-lactic acid for SPPS.

Due to the low amount of coupling obtained and the limitations in the selection of

a new protecting group for L-Lac, as described above, the stepwise approach to the

synthesis of TSA 1-5 was abandoned. The desire for a synthetic strategy that avoided the

protection of L-Lac led to the development and use of a convergent approach to the

transition-state analog.


Convergent Approach to the Transition-State Analog

The most logical disconnection of TSA 1-5 for a convergent synthesis was

between the lysine residue and the phosphoalanine residue (Scheme 3-5), because

solution-phase synthesis of the phosphonate fragment made the problematic protection of

the L-Lac residue unnecessary and the remaining peptide fragment contained standard









.SH
0 0 C02- 0 CH3
)J H HL ~j : 0 o.zz
H3C N f;HN N N

H II H 7 H O-/ 0O
0 CH3 0 CH3
1-5 1

HN 0
t-Bu Ph Y
0 00 CH3 CH3
1H INI OH + H2 P,"" CO2Et
H3C kN '>NT AHNNC02H + H2N ,
H II H II CH3
0 CH3 0 OH3
1-6 1-7

HN. 0

CH3

Scheme 3-5
amino acids to be used in the SPPS. This route also minimized the number of solution

phase couplings required, with one to provide 1-7 and an additional step to couple the

fragments together. In this convergent synthetic strategy toward 1-5, a solution phase

Mitsunobu reaction between (L)-ethyl lactate and phosphonate monoester 2-6 was used to

provide the phosphonate ester fragment 1-7. The remainder of the peptide (1-6) was

synthesized with SPPS using the Wang resin and Fmoc chemistry and then these two

fragments were coupled using solution-phase techniques, followed by cleavage of the

amino acid side-chain protecting groups (Scheme 3-5).

Solid-Phase Peptide Synthesis of the Tetrapeptide Fragment

Before synthesizing the tetrapeptide 1-6, the side-chain groups of the amino acids

were protected in order to prevent these side-chains from interfering with the coupling to

the phosphonate fragment. The protecting groups were carefully selected to ensure that

these groups remained attached during removal of the peptide from the resin support.

This protection strategy differs from traditional SPPS, where the side-chain groups are








99
usually deprotected simultaneously as the peptide-resin bond is cleaved. Certain

protecting groups were also chosen to minimize the number of final deprotection steps

required. N-a-Fmoc-N-e-acetyl-L-lysine was synthesized due to its expense (Scheme 3-

6), while the D-glutamic acid residue with the desired side-chain protecting groups was

synthesized (Scheme 3-7) due to its lack of commercial availability.

Fmoc-protected L-Lys residue was synthesized from N-s-acetyl-L-Lys by first

forming 0, N-bis-trimethylsilyl-lysine in situ with TMS-C1 followed by reaction with

Fmoc-Cl to provide 3-4. This method of Fmoc protection eliminates the formation of

dipeptides which is sometimes observed when the reaction is carried out under Schotten-

Baumann conditions (Scheme 3-6).125


H2N ,C02H FmocHN .CO2H

1) TMS-C1, DIEA
2) Fmoc-Cl, 92%/

HN 0 HN 0

CH3 CH3
3-4

Scheme 3-6
The glutamic acid residue was prepared by protecting the amino group of D-Glu

99
with the Boc group and then selectively benzylating the a-carboxylic acid with benzyl

bromide and triethylamine, according to the procedure of Pawelczak and coworkers,126 to

provide 3-6. The Boc group was removed with 25% (v/v) TFA:CH2Cl2 which resulted

in TFA salt 3-7 followed by reprotection of the amine with the Fmoc group to afford the

N-a-Fmoc-D-glutamic acid-a-benzyl ester 3-8 (Scheme 3-7).125 The D-Glu residue was









H2N 02H BoCHN C02H BocHN CO2Bn
(Boc)20, K2003 BnBr, Et3N TFA:CH2C2, 25% vlv
86% 50%0/ 98%
CO2H CO2H CO2H
3-5 3-6

/H2N C02Bn FmocHN CO2Bn
i 1)TMS-CI. DIEA
STFA 2) Fmoc-CI, 87% CH
\ co H/ CO H

3-7 3-8


Scheme 3-7

not protected with Fmoc initially because the triethylamine in the benzylation step would

have been sufficiently basic to cleave it.

With the protected L-Lys and D-Glu residues in hand, the SPPS of tetrapeptide 1-

6 was carried out manually on Wang resin (Scheme 3-8). The first amino acid, 3-4, was

anchored to the Wang resin using DIC and a catalytic amount of DMAP to give 3-10.

Over multiple batches, the coupling yield ranged from approximately 50% to 85%, as

determined by quantitation of the piperidine/dibenzofulvene adduct produced upon Fmoc

deprotection. The remaining unreacted hydroxyl groups of the resin were capped with

127
acetic anhydride and a catalytic amount of DMAP. 27 The Fmoc group was deprotected

from the lysine residue using 20% (v/v) piperidine in DMF, which was followed by the

124
coupling of 3-8 to the Lys(Ac)-resin using DIC and HOBt to give 3-11. Any

remaining free amino groups were capped with acetic anhydride and pyridine to prevent

128
the formation of possible deletion sequences. The Fmoc group was deprotected from

the glutamate residue with piperidine and N-a-Fmoc-L-alanine was coupled to the D-y-

Glu(OBn)-Lys(Ac)-resin with DIC and HOBt to provide 3-12. The capping and

















FmrnocHN CO2H
0 ~N
i-Pr"



1 HHN O
HO

34 CH3

0
FmocHN^A ^iO,. 1) Ac2O, DMAP(cat.)
2) 20%(v/v) piperidine in DMF
S3) 3-5, DIG, HOBt



HN"

3-10 CH3


1) Ac20 pyridine
2) 20%(v/v) piperidine in DMF
CH3 D
3)FmocHN 00CO2H, DIC, HOBt


=C=N' i-Pr (DIC)
DMAP(DIG)
DMAP(cat.)


HN 0


3-9 G"3

COzBn 0
O H
FmocHN 0
0

3-11 A

HN O 0

CH3


0 CO28Bn 0
FmocHN NN, H
_- H nI :. -
H3C 0O

3-12

HN 0O
(-.&


1)Ac2C
2)20%


3)Fmoc


s--'"5
t-Bu 3


H 0 CO2Bn 0

NFmocHNf N N- N 1) 20%(v/v) piperdine in DMF
II H I 2) Ac2O pyridine,
0 CH3 0 3) 70%(v/v) TFA in CHFCI2

3-13

HN 0O

CH3


Scheme 3-8

deprotection steps were repeated followed by the coupling with N-ax-Fmoc-S-t-butylthio-


L-cysteine to give 3-13. The Fmoc group of the cysteine residue was then deprotected


followed by reprotection of the amino group with acetic anhydride to provide the


), pyridine
(v/v) pipenridine in DMF
S. t-Bu

:HN CO2H, DIC, HOBt










1-6











14.25











0. 00 10.00 Tim (in.) 20.00 28.00


62o i
B 91.0370 E 06
""I | 1.33
B 100 1.33


80-i


|[M + Na-] =
60- 734.2844


40-


20- 10U292
131.0109|
I 217.0759 JL
26L..442 385. 826 4 .1648 589.2312
AJ OL. i -. L J I F.. P 0. ', P,.,,,n, i 0,1 I I ,. I 1 0 1& .
100 230 300 400 500 600 700

Figure 3-2. Analysis oftetrapeptide 1-6. A) The peak at 14.25 min in HPLC
chromatogram is 1-6. The peaks seen before 10.00 min are due to the injection
and solvent. B) The FAB-MS showed the [M + Na] = 734.2844 for 1-6;
calculated [M + Na] = 734.2869.


tetrapeptide with an acetylated N-terminus. Every coupling step was performed twice to

ensure a high level of attachment, and the Kaiser ninhydrin test, which is an indicator of

free amino groups, was carried out on the resin beads to confirm the completeness of the

99
coupling reaction. Finally, the peptide was cleaved from the resin with 70% (v/v)

129
TFA:CH2Cl2 to give tetrapeptide 1-6 in 36% overall yield (relative to 100% attachment








to the resin and 100 % coupling of the amino acids). The purity of 1-6 was determined

by HPLC analysis, and its identity was confirmed with FAB-MS (Figure 3-2).

Synthesis of the Phosphonate Ester Fragment

The synthesis of the phosphonate ester 1-7 of the transition-state analog 1-5 is

shown in Scheme 3-9. (S)-(1-aminoethyl)-phosphonic acid was protected with the

benzoxycarbonyl (Cbz) group to afford 3-14. Initial attempts at isolating 3-14 from the

reaction mixture by extraction resulted in poor yields due to its partial water solubility.

CH3 CH3 CH3
POH Cbz.C, 4MNaOH CbzHN.p OH CH2N2 CHN P .0CH3 a. KOH
HNCzN PCbzHN !P, a.KO
O0P'OH 84% O \OH 94% OCH3 79%
3-14 3-15


QH3 HO..vCO2Et CO2i-Pr CH3
N=N
-'.10H CH3 2.p2 P 1-7
CbzHN pO 3 \Pr2C CbzHN 0 CO2Et H2, Pd/C 1-7
0 OCH3 / \ O' 95%
2-6 /C, J)P;90% CH3
3H %CH3

3-16

Scheme 3-9
Separation of the product from the starting material was achieved by passing the reaction

mixture through a cation-exchange resin (H+ form) to result in an 84% yield. 3 The

dimethyl phosphonate ester 3-15 was produced by methylation of 3-14 with

diazomethane. The monomethyl ester 2-6 was synthesized according to Campbell and

90
Bermak9 by selective hydrolysis of 3-15 with aqueous KOH. The Mitsunobu reaction of

2-6 with (L)-ethyl lactate yielded the unsymmetrical phosphonate diester 3-16 and the

results of the model study with methyl methyl phosphonate suggest that the desired D-

S4
isomer of the lactate residue was produced. Compound 3-16 was obtained as a 2:1

mixture ofdiasteromers. Note that this is not a problem since final demethylation of the









phosphonate methyl ester will provide the product as a single enantiomer, due to the loss

ofchirality at the phosphorus center. Finally, the Cbz group of 3-16 was cleaved by

catalytic hydrogenation to afford the phosphonate fragment 1-7. While cyclization of 1-7

was believed to be a potential problem in this step, no cyclized product was seen.

Model Studies of the Segment Condensation Step

Before the segment condensation of the tetrapeptide and phosphonate fragments

was carried out, two model studies were performed by coupling dimethyl-S-(l-

aminoethyl)phosphonate (3-17) with N-a-Boc-D-glutamic acid-ao-benzyl ester (3-6) and

glycine methyl ester with 1-6 in order to verify that product could be synthesized when

1,3
using isobutyl chloroformate (IBCF) as the coupling agent. Catalytic hydrogenation of

the Cbz protected 3-15 produced the dimethyl-S-(l-aminoethyl)phosphonate 3-17

(Scheme 3-10), which was then added to a solution of the glutamic acid residue 3-6

preactivated with IBCF and N-methylmorpholine (NMM) to afford the coupled product


CH3 CH3
,C-b .OCH3 H2, Pd/C .. .OCH3
CbzHN .-^ H2N oPo3
O OCH3 42% 0 OCH3
3-15 3-17

Scheme 3-10





CH3
0 N
BOCHN CO2Bn 1) DMF, O "Cl I 0 -OH3
0 BOCHN A OCH3
N "P
2) 3-17 ; 42% CO2Bn H OCH3
CO2H 3-18
3-6 3-18


Scheme 3-11









3-18 in 42% yield (Scheme 3-11). The same conditions were used to couple glycine

methyl ester with tetrapeptide 1-6 to provide 3-19 (Scheme 3-12), as shown from


t-Bu Ph



H3C H -H 2) cr +H3NCO2CH3
0 CH3 0 %
1-6

HN 0O

$ At-Bu Ph CH3

0 f N T0 N 0

H3C N NT NAC02CHn
:- I II ., 2CH3


3-19


Scheme 3-12


20.54


w~~~~~ I IN

Figure 3-3. HPLC chromatogram of 3-19. The peaks seen before 8.00 min are
due to the injection and solvent. The peak at 20.54 min is 3-19.








HPLC/ESI-MS. The HPLC chromatogram demonstrated that nearly all of 1-6 had been

reacted to form 3-19 (Figure 3-3). All compounds synthesized from reactions with

tetrapeptide 1-6 were analyzed with HPLC for purity and MS for identity due to the small

scale of the reactions, which prevented the acquisition of sufficiently resolved NMR

spectra.

Coupling of the Tetrapeptide and Phosphonate Fragments

With evidence from these model studies that the coupling reaction with IBCF

resulted in product with reasonable yields, the reaction was carried out with the

tetrapeptide fragment 1-6 and the phosphonate fragment 1-7 (Scheme 3-13). HPLC

St-Bu Ph
,
0 0 0 CH3
1) IBCF, NMM, DMF jH H 0 = 2Et
1-6 N T:- Hy -
2) H3 H30 NNN
H H 11 H 0 0
..; 0 CO2Et 0 CH3 0 CH3
H2N ",P\' O H3
OH3 2-5
OCH3
CH3
1-7 HN
CH3
Scheme 3-13
analysis of this reaction showed the diastereomeric product 2-5 was obtained, however

the tetrapeptide starting material (12.69 min) was still present along with other impurities

(Figure 3-4). Isolation of the HPLC peaks (17.91 min and 18.29 min) believed to

correspond to 2-5 followed by MS analysis confirmed their identities. When the crude

reaction mixture was subjected to the reaction conditions once more with freshly

prepared 1-7, the remaining tetrapeptide starting material was consumed. This re-

reaction is believed to provide 2-5 in a higher yield, although it is possible the starting

material was consumed due to degredation. Finally, compound 2-5 was purified from the










12.69


17.91

S18.29


Figure 3-4. HPLC chromatogram of crude 2-5. The peaks seen before 8.00 min are
due to the injection and solvent. The peak at 12.69 min is tetrapeptide 1-6. The
peaks at 17.91 min and 18.29 min are the diasteromeric 2-5.

crude reaction mixture by RP-HPLC (Figure 3-5); however only a 6% yield of product

was obtained. Reaction yields reported in the literature for segment couplings with IBCF

range from 16% to 92%.

Deprotection of the Phosphonate Methyl Ester and Carboxylate Esters

Based on our previous success with the model system, selective demethylation of

the phosphonate methyl ester of 2-5 was initially attempted with bromotrimethylsilane

(TMS-Br) and isobutylene in DMF; however, no reaction occurred even when a large

excess of TMS-Br had been added. Thus, an alternative route was investigated where

both the phosphonate methyl ester, glutamate benzyl ester, and lactate ethyl ester of 2-5

were hydrolyzed simultaneously with LiOH. Although saponification of unsymmetrical

phosphonate diesters usually results in the non-selective cleavage of both ester bonds, we





























Figure 3-5. HPLC chromatogram of purified 2-5. The peaks seen before 10.00
min are due to the injection and solvent. The peaks at 28.26 min and 28.64 min
are due to the diasteromeric 2-5.

proposed that selectivity toward methyl ester cleavage in 2-5 might be achieved due to

steric factors. During hydrolysis, the attacking hydroxide ion and the phosphonate ester

bond being cleaved must be in the axial position of the trigonal bipyramidal intermediate

(Scheme 3-14). When the lactate residue is in an axial position an unfavorable steric

interaction with the peptide chain in an equatorial position may occur (3-20). It is

possible that this steric interaction is decreased when the methyl ester is axial (3-21) and

is therefore more favorable for hydrolysis.

This simultaneous phosphonate and carboxylate ester cleavage with LiOH was

first attempted with 3-16 as a model study because it was believed the Cbz group would

provide a sufficient amount of steric interaction with the lactate ester to provide

selectivity toward methyl ester hydrolysis. However, the reaction provided a mixture of

the two possible phosphonate monoesters (3-22 and 3-23) and the hydrolysis product of















\ ,/ lactyl
0 H3CH30 group
cleaved
" kN P-OCHa 3-27
E H -0 I
OH


HN.0

CH3

potentially interacting
groups separated by 120


methyl
group


3-25


3-21


Scheme 3-14


CH3

CbzHN ypOyC02Et
0 0
/ CH3
CH3
3-16

0.2M UOHj


CH3

CbzHN P-p YC02 +
2 OU CH3
3-22


CH3
CL A .,OLi
CbzHN .PO
O OCH3
3-23


Scheme 3-15


CH3

+ CbzHN '-/"O"0c2U
0 ? OH3
/ CH3
CH3
3-24


0

H3C N


3-20








only carboxylate ethyl ester cleavage (3-24) (Scheme 3-15). Although this model study

did not display a great deal of selectivity, the reaction was attempted with purified 2-5 to

determine if more selectivity would be achieved (Scheme 3-16). The reaction was


St-Bu
S0 OH
0 0 2OH 0 CQH3
OW U fl H .-COH
2-5 0... N N,_A M ";-N P., o..


3-25


Scheme 3-16


20.88


0.0 wee. 10!00 35}.0

Figure 3-6. HPLC chromatogram of crude mixture of 3-25. The peaks seen before
6.00 min are due to the injection and solvent. The peak at 20.88 min is 3-25.


monitored with HPLC, which showed that all of the starting material was consumed

(Figure 3-6). The major products from the reaction were determined by MS to be the









St-Bu

S O-. .0OH
0 0 0 CH3
N N p0" C02H
H3C N' ~ N N /P 02
CH3 O 0 1 CH3
3-26 CH3

HN 0

CH3
t-Bu
S O .OH
0 0 0 CH3
H3C N' N HN PH
0 NH 3 0 A
H S CH3 HO"

3-27 A l

HN.0

CH3

0^ ,OH
0 H CH3
,,,:N/J Y N _OA 0, CO2H
H3C N ",K HO/'\OH,
H II = H II i >
0 CH3 0 0 OH CH3
3-28

HN O 0

CH3
Scheme 3-17
desired product with the carboxylate esters and phosphonate methyl ester cleaved 3-25

and the product from exclusive hydrolysis of the lactate ethyl ester and glutamate benzyl

ester (3-26) (Scheme 3-17). Only a small amount of the hydrolysis of the phosphonate

lactate ester (3-27) was seen. As predicted partial selectivity toward the methyl ester was

observed. Attempts to further hydrolyze 3-26 to product failed and instead resulted in

elimination of the disulfide group by abstraction of the a-proton of the cysteine residue to

produce 3-28. This elimination product was also seen if the reaction time was extended








beyond the time necessary for the formation of 3-25. Finally, in order to reduce the

number of purification steps, this LiOH hydrolysis reaction was carried out with crude 2-

5 and the product was purified by HPLC to provide the cysteine-protected TSA 3-25 in a

2% yield (relative to 100% yield from the coupling reaction). The cysteine disulfide

protecting group will be removed in situ when attaching the TSA to the carrier protein.


Conclusion

Two approaches to the synthesis of TSA 1-5 have been discussed in this chapter.

The initial solid-phase route was abandoned after encountering difficulty with protection

of the L-Lac residue and the coupling of the eventually protected L-Lac to the solid-phase

resin. The successful convergent approach was then adopted to avoid these problems.

The tetrapeptide fragment 1-6 was synthesized efficiently using standard SPPS and the

optimized Mitsunobu reaction provided the phosphonate fragment 1-7 in high yields with

the desired D-isomer at the lactate residue. The segment coupling of the tetrapeptide and

phosphonate fragments was successful, although the reaction was low yielding, and

finally the simultaneous and selective deprotection of the phosphonate methyl ester and

carboxylate esters provided the TSA 3-25. Once the TSA is attached to a carrier protein,

antibodies will be raised against it and tested for their ability to catalyze the cleavage of

the resistant bacterial peptidoglycan cell wall.













CHAPTER 4
CONCLUSIONS AND FUTURE WORK

The emergence of high-level vancomycin-resistance in gram-positive bacteria

occurred as a result of the replacement of an amide bond with an ester bond in the

pentapeptide of the bacterial peptidoglycan cell wall. The long-term goal of the project

detailed in this dissertation is to develop antibodies to catalyze the hydrolysis of this ester

bond and thus combat vancomycin-resistant bacteria. In order to raise catalytic

antibodies, a transition-state analog (1-5) was designed that incorporated a phosphonate

ester, an ester hydrolysis transition-state mimic, into the resistant bacterial peptidoglycan

pentapeptide sequence. The model study and synthesis of this transition-state analog was

detailed in the previous chapters.

The synthesis of TSA 1-5, which proved to be challenging, was originally

attempted with a stepwise solid-phase synthetic approach and finally accomplished with

the successful convergent route. The solid-phase approach was discontinued when the

TBDMS-protecting group used for the alcohol moiety of the first residue in the sequence,

L-lactic acid, was found to be unstable to the coupling conditions. The search for

alternative compatible protecting groups was unpromising; consequently the convergent

approach was developed. This route consisted of the fairly straightforward SPPS of a

tetrapeptide fragment (1-6) and the solution-phase synthesis of a phosphonate fragment

(1-7). The model study described in chapter 2 demonstrated that by utilizing the

Mitsunobu reaction in the synthesis of 1-7 the desired D-isomer of the important lactate

residue was obtained. The coupling of the two fragments was performed, although with








low yield, followed by the one step hydrolysis of the carboxylate esters and selective

hydrolysis of the phosphonate methyl ester to provide the cysteine-protected TSA (3-25).

Experiments performed in the model study also confirmed that racemization of the lactate

center did not occur during this hydrolysis with LiOH.

The most problematic step in the synthesis of 1-5 with the convergent approach

was the extremely low yielding segment condensation of the tetrapeptide and

phosphonate fragments with isobutyl chloroformate (IBCF). This low yield affected the

ability to perform exploratory work on the subsequent deprotection steps due to the lack

and value of the material. Therefore, attempts to optimize the yield of this reaction could

begin with an investigation of alternative reagents for this coupling which also do not

result in racemization, such as EEDQ.

The TSA will be prepared for immunization by reducing the disulfide-protecting

group of the cysteine residue in 3-25 with tris (2-carboxyethyl) phosphine (TCEP). Once

this step has been performed to provide 1-5 it will be coupled to the carrier protein,

keyhole limpet hemocyanin (KLH), using a maleimide-containing bifunctional linker.

This KLH-antigen conjugate will be used to immunize mice and the hybridoma method

will be used to develop antibodies. All antibody development work will be carried out at

the Interdiciplinary Center for Biotechnology Research (ICBR) laboratories at the

University of Florida. Once antibodies that to bind 1-5 are isolated they will be tested for

the ability to catalyze ester hydrolysis of the resistant bacterial peptidoglycan

pentapeptide, which will be synthesized using manual SPPS. The rate of this hydrolysis

reaction will be assayed by using D-lactic dehydrogenase and the cofactor NAD+ to

catalyze any D-lactate hydrolyzed by the antibody to pyruvate and NADH. The NADH





70

growth can be monitored spectrophotometrically and thus will provide kinetic data for the

antibody. Although the goal of this project is to obtain antibodies capable of catalysis,

antibodies that only bind tightly to the resistant bacterial peptidoglycan pentapeptide

substrate would sterically prevent crosslinking from occurring and would have a mode of

action similar to vancomycin. The application of these catalytic and substrate binding

antibodies to studies of vancomycin-resistant bacteria will then be explored.













CHAPTER 5
EXPERIMENTAL


General Methods

Reagents were obtained from commercial suppliers and used as received. L-

Lactic dehydrogenase (rabbit muscle; EC 1.1.1.27) was obtained as a solution in 50%

glycerol containing 10 mM potassium phosphate buffer (pH 7.5) from Sigma. All

reactions were run using flame dried glassware, dry solvents and under an inert

atmosphere of argon. Thin layer chromatography was performed on pre-coated silica gel

60 plates with a fluorescent indicator. Compounds were visualized with UV light at 254

nm and with 5% phosphomolybdic acid in absolute ethanol. Flash column

chromatography was carried out using 60 angstrom silica gel (200-425 mesh) from

Fisher. Tetrahydrofuran was distilled from sodium benzophenone ketyl. DMF was

distilled over anhydrous MgSO4 and stored over 4 angstrom molecular sieves.

NMR spectra were obtained on Varian Mercury or VXR-300 instruments (300

MHz). Unless otherwise specified, all 'H NMR spectra were recorded in CDC13 or

CD3OD and chemical shifts are reported in ppm downfield relative to tetramethylsilane

(0 ppm) as an internal standard. 31P NMR spectra are reported in ppm relative to

phosphoric acid (85% solution in a sealed capillary, 0 ppm). IR spectra were recorded

from thin films or KBr pellets on a Bruker Vector-22 instrument. High-performance

liquid chromatography (HPLC) was performed on a Beckman System Gold or a Gilson

system. All solvents used for HPLC were passed through a 0.45 micron filter. Capillary








gas chromatography was carried out on a Hewlett-Packard 5890A instrument equipped

with a flame ionization detector. Chiral separations were performed on a Chrompack

0.25 mm x 25 m CP chirasil-Dex CB GC column using a gradient from 60 C (2 min) to

120 C (5 min) at 1 C/min followed by a 10 C/min gradient to 180 C (5 min). The

injector and detector temperatures were maintained at 225 C and 220 C, respectively.

GC-MS spectra were recorded on a Hewlett-Packard 5890 series II equipped with a 5971

series mass selective detector and a 0.32 mm x 30 m DB-17 GC column. A gradient

from 60 C (2 min) to 250 C (10 min) at 10 C/min was used. The injector and detector

temperatures were maintained at 250 C and 280 C, respectively. All other mass

spectroscopy and elemental analysis was performed by the University of Florida

Department of Chemistry Spectroscopic Services Labs.


Experimental Procedures

Dimethyl methylphosphonate (2-7). To a solution of CH3OH (9 mL) cooled to

0 C, methylphosphonic dichloride (1.00 g, 7.5 mmol) was added. After the reaction had

been stirred at rt for 16 h, the solution was concentrated and neutralized to pH 7 with

saturated aqueous NaHCO3. The aqueous layer was extracted three times with CHC13 (15

mL). The combined organic extracts were dried over MgSO4, filtered, and concentrated

under vacuum to give the product as a yellow oil (0.629 g, 68%). The spectral data were

in agreement with published data.132 'H NMR (CDCI3) 6 1.48 (d, 3H, J = 17 Hz), 3.75

(d, 6H, J = 11 Hz); 31P NMR (CDCI3) 5 33.86.

Methyl methylphosphonate (2-8). Using a procedure described by Christol et
114
al,. 2-7 (0.745 g, 6 mmol) was dissolved in a 1:1 mixture of 4% NaOH:dioxane (16

mL). The reaction mixture was held at reflux for 18 h at 90 C and then cooled in an ice








bath. The mixture was acidified to pH 2 with concentrated HC1 and the volume was

reduced by rotary evaporation. The product was extracted from the precipitated NaCI

with acetone (30 mL). The organic solution was filtered, dried over Na2SO4 and

concentrated in vacuo to give the product as a brown oil (0.628 g, 95%). The spectral

data were in agreement with published data."4 'H NMR (CDC13) 8 1.51 (d, 3H, J = 18

Hz), 3.73 (d, 3H, J = 11 Hz); 3P NMR (CDC13) 8 34.88; IR (neat): 2284,1678,1187,

994 cmn'.

Ethyl -(R)-2-[1Methoxy-methyl-phosphorylloxylpropionate (2-9). The

84
synthesis of 2-9 was adapted from a procedure described by Campbell.84 To a solution of

2-8 (0.330 g, 3 mmol), (S)-ethyl lactate (0.531 g, 4.5 mmol), and tris(4-

chlorophenyl)phosphine (1.65 g, 4.5 mmol) dissolved in THF (30 mL) was added

diisopropyl azodicarboxylate (0.909 g, 4.5 mmol). The reaction was stirred at rt for 16 h

and then concentrated. The crude product was purified by flash column chromatography

(55 mm diameter) eluting with EtOAc to produce 2-9 as a yellow oil (0.499 g, 79%). The

product was analyzed with chiral GC. The retention times of the diastereomers were

57.60 and 58.19 min. The spectral data were in agreement with published data.33 'H

NMR (CDC13) 5 1.30 (t, 3H), 1.54 (d, 1.5H, J = 18 Hz), 1.54 (d, 1.5H, J = 2 Hz), 1.56 (d,

1.5H, J = 2 Hz), 1.60(d, 1.5H, J = 18 Hz), 3.71 (d, 1.5H, J = 11 Hz), 3.78 (d, 1.5H, J = 11

Hz), 4.23 (q, 2H), 4.98 (m, 1H); 13C NMR (CDC13) 8 10.32,10.44, 12.25, 12.39, 13.99,

19.05, 19.12, 19.35, 19.41, 51.25, 51.34, 52.14, 52.23, 61.41, 61.48, 69.93, 70.02, 70.37,

70.44,170.88, 170.93, 170.99, 171.05; 3"P NMR (CDC13) 5 32.61, 34.10; IR (neat):

1751, 1233, 1040, 990 cm1.








Oxygen-18 labeled sodium L-lactate (H"OCH(CH3)CO2Na) (2-19). The

synthesis of 2-19 was adapted from a procedure described by Fromm.17 The 8O labeled

sodium pyruvate 2-18 was first synthesized using previously described procedures.""'116

Sodium pyruvate (0.090 g, 0.82 mmol) was dissolved in H2180 (0.2 mL) and stirred at rt

for 2.5 days in a sealed vial. This solution was added to potassium phosphate buffer (17

mL, 24mM, pH 7) followed by NADH (0.575g, 0.81 mmol) and L-lactic dehydrogenase

(196 units). The pH of the reaction was maintained at 7 by periodic addition of 0.96 M

HC1. When constant absorption at 340 nm was obtained (30 min) the reaction was

stopped by boiling the solution for 1 min followed by acidification to pH 2 with 6 M HC1.

The acidified reaction mixture was continuously extracted with Et20 (20 mL) for 2 days.

The Et20 layer was concentrated by rotary evaporation and then I M NaOH (0.81 mL)

was added to the resulting residue followed by lyophilization to give the product as a

yellow solid (0.089 g, 98%). 'H NMR (CD30D) 6 1.34 (d, 3H), 4.05 (q, 1H); 13C NMR

(CD3OD) 5 22.05, 69.69, 182.87; HPLC/ESI-MS m/z: 43.06 (M HCOOH, 67%), 45.08

(M HCOOH, 33%).

Oxygen-18 labeled ethyl L-lactate (H'8OCH(CH3)CO2CH2CH3) (2-20). The

synthesis of 2-20 was adapted from a previously described procedure. To a mixture of

15-crown-5 (0.174 g, 0.79 mmol) and CH3CN (6 mL) was added 2-19 (0.089 g, 0.79

mmol) and the solution was stirred at rt for 30 min. lodoethane (1.47 g, 9.48 mmol) was

added and the reaction mixture was held at reflux for 1.5 days and then concentrated.

The resulting residue was purified by flash column chromatography (20 mm diameter)

eluting with EtOAc to yield the product as a yellow oil (0.039 g, 41%). 'H NMR









(CDC13) 1.29 (t, 3H), 1.40 (d, 3H), 4.10 (q, IH), 4.22 (q, 2H); GC-MS m/z: 75 (M -

C3H7 or C2H30, 72%), 77 (M C3H7 or C2H30, 28%); tR = 3.53 min.

Ethyl -2-([Methoxy-methyl-phosphorylJoxylpropionate racemicc 2-9). The

synthesis was followed as described for 2-9 except starting with racemic ethyl lactate

(0.055 g, 0.5 mmol). Overall yields were similar to those obtained for compound 2-9.

The product was analyzed with chiral GC. The retention times of the diastereomers were

56.76 min, 57.70 min and 58.23 min (only three of the four diastereomers were resolved).

The spectral data matched the values reported for 2-9.

Ethyl -(S)-2-JIMethoxy-methyl-phosphoryljoxylpropionate (2-13). The
82
synthesis of 2-13 was adapted from a procedure described by Campagne et al. To a

solution of 2-8 (0.110 g, 1 mmol), PyBOP (0.781 g, 1.5 mmol) and

diisopropylethylamine (0.517 g, 4 mmol) dissolved in DMF (10 mL) was added (S)-ethyl

lactate (0.177 g, 1.5 mmol). The reaction was stirred at rt for 16 h and then concentrated.

The resulting residue was dissolved in EtOAc (20 mL) and washed twice with 5%

KHSO4 (15 mL), three times with 5% NaHCO3 (15 mL), and twice with brine (15 mL).

The EtOAc layer was dried over Na2SO4, filtered and concentrated by rotary evaporation.

The resulting residue was purified by flash column chromatography (20 mm diameter)

eluting with EtOAc to yield the product as a brown oil (28.4 mg, 14%). The product was

analyzed with chiral GC. The retention times of the diastereomers were 56.62 and 57.73

min. The spectral data matched the values reported for 2-9.

Ethyl-(R)-2- [[Sodium oxy-methyl-phosphoryl]oxylpropionate (2-25).

Using THF as solvent. This synthesis of 2-25 was adapted from previously
described procedures.84120 To a mixture of 2-9 (0.049 g, 0.23 mmol) and THF (3 mL)
described procedures. 'To a mixture of 2-9 (0.049 g, 0.23 mmol) and THF (3 mL)








was added TMS-Br (0.115 g, 0.75 mmol). The reaction was stirred at rt for 4 h and the

volatiles were evaporated. The resulting residue was treated with a solution of NaOH (0.

009 g, 0.23 mmol) in CH3OH (3 mL) and stirred for 30 min. The solution was

concentrated, Et20 (8 mL) was added to the residue and the resulting brown solid was

filtered to give 0.030 g (60%) of the product. 'H NMR (D20) 8 1.32 (t, 3H), 1.34 (d, 3H,

J = 17 Hz), 1.50 (d, 3H), 4.28 (q, 2H), 4.78 (m, 1H); 3C NMR (DO20) 6 12.18, 13.99,

14.18, 20.17, 63.41, 69.99; 3P NMNIR (D20) 5 27.69.

Using DMF as solvent. This synthesis of 2-25 was adapted from previously
113.120
described procedures. Isobutylene was bubbled into DMF (1.3 mL) for 2 min

followed by the addition of TMS-Br (0.044 g, 0.286 mmol). Five minutes later, a

solution of 2-9 (0.060 g, 0.286 mmol) in DMF (0.2 mL) was added to the TMS-Br

solution. The reaction was stirred at rt for 3 h and then concentrated in vacuo. The

resulting residue was treated with a solution of NaOH (0.011 g, 0.286) in EtOH (1.5 mL)

and stirred for 30 min. The solution was concentrated, Et2O (10 mL) was added to the

residue and the resulting yellow solid was filtered and then passed through a column (10

mm diameter) of Dowex 50X8 200 (Na) resin. The compound was eluted with

distilled water (15 mL), which was then lyophilized to give the product as a white solid

(0.038 g, 62%). The spectral data matched the values reported above.

Lithium-(R)-2-1 [Lithium oxy-methyl-phosphoryl]oxy]propionate (2-26). The

synthesis of 2-26 was adapted from a procedure described by Mayer et al. 121 To 0.038 g

(0.176 mmol) of 2-25 was added a 0.20 M aqueous solution of LiOH (0.9 mL) cooled to

0 C and the reaction was stirred at rt for 2 h. The reaction mixture was lyophilized and

the resulting solid was dissolved in water and passed through a column (10 mm diameter)








of Dowex 50X8 200 (Na) resin. The compound was eluted with 20 mL distilled water,

which was then lyophilized to give the product as a white solid (0.036 g, 94%): 'H NMR

(D20) 8 1.25 (d, 3H, J = 17 Hz), 1.39 (d, 3H), 4.46 (m, 1H); "3C NMR (D20) 8 12.51,

14.23, 21.12, 72.94,131.03; 31P NMR (D20) 826.5; IR (KBr): 3448,1685,1208,1064

cm'1; HRMS calcd for C41-H7Na2O5P (M + Na) 234.9724, found 234.9696.

Methyl-2-[[Methoxy-methyl-phosphorylloxylpropionate racemicc 2-27) and

Methyl -(R)-2-[[Methoxy-methyl-phosphoryl]oxylpropionate (2-27) chiral GC

standards. The chiral GC standards were synthesized as described for 2-9 except

racemic or S-methyl lactate (0.055 g, 0.5 mmol) were used as the starting material.

Overall yields were similar to those obtained for compound 2-9. The standards were

analyzed with chiral GC. The retention times of the diastereomers were 51.12 min, 52.09

min and 52.65 min for racemic 2-27 and 51.90 min and 52.52 min for the 2-27 standard.

Methyl -(R)-2-[[Methoxy-methyl-phosphoryl]oxy]propionate (2-27). The

1 18
synthesis of 2-27 was adapted from a previously described procedure.118 To a mixture of

15-crown-5 (0.025 g, 0.114 mmol) and CH3CN (0.4 mL) was added 2-26 (0.012 g, 0.057

mmol) and the solution was stirred at rt for 30 min. lodomethane (0.456 g, 3.21 mmol)

was added and the reaction mixture was stirred in a sealed vial at 80 C for 20 h and then

concentrated. The resulting residue was purified by flash column chromatography (10

mm diameter) eluting with EtOAc to yield the product as a colorless oil (1.2 mg, 11%),

which was analyzed with chiral GC. The retention times of the diastereomers of 2-27

were 51.92 min and 52.53 min.

N-ax-Fmoc-L-Alanine-Wang Resin (3-3a). The synthesis of 3-3a was adapted

from a previously described procedure. 27 To a manual solid-phase peptide synthesis








vessel 0.100 g of the Wang Resin (substitution: 0.71 mmol/g, 0.071 mmol) was added

followed by 4 mL of DMF. The suspension was agitated with argon for 1 h at rt to swell

the resin. The solution was drawn off and Fmoc-Ala (0.066 g, 0.213 mmol) was added to

the resin followed by DIC (0.054 g, 0.426 mmol) and 3 mL DMF. The suspension was

agitated with argon for 5 min, then 0.9 mg DMAP (0.0071 mmol) dissolved in 0.2 mL

DMF was added to the resin and the suspension was agitated with argon at rt for 2 h. The

solution was drawn off and the resin was washed 5 times with DMF (35 mL) then the

procedure was repeated. A small amount of resin was washed with CH2C12 and

isopropanol and dried: IR (KBr): 1726 cm-'.

Estimation of level of first residue attachment by quantitation of the

piperidine/dibenzofulvene adduct. To a sample of the Fmoc-amino acid-resin that was

washed with isopropanol (5 mL), dried and weighed (approximately 1 mg) was added

freshly prepared 20% piperidine in DMF (3 mL). The mixture was agitated with a

pasteur pipet for 3-5 min and then the resin was allowed to settle to the bottom of the

vial. The UV absorbance of the solution was recorded at 290 nm along with a reference

containing only 20% piperidine in DMF. The level of attachment was calculated

according to the following formula obtained from the 2000 NovaBiochem catalog:

mmol/g loaded = (Abssampie- Absref)/(1.65 x mg of dry Fmoc-amino acid-resin).

TBDMS-L-Lactate-Wang Resin (3-3b). The synthesis of 3-3b was followed as

described for 3-3a except starting with 0.563 g of the Wang Resin (substitution: 0.71

mmol/g, 0.4 mmol) and TBDMS-L-lactic acid (0.245 g, 1.2 mmol). A small amount of

resin was washed with CH2C12 (5 mL) and isopropanol (5 mL) and dried: IR (KBr):

1746 cm'.








Estimation of level of first residue attachment of 3-3b. The TBDMS-L-Lac-

resin (0.563 g) was washed 5 times with CH2Cl2 (25 mL), 5 times with isopropanol (25

mL) and then dried under vacuum. To the dry resin 1:1 TFA:CH2CI2 (10 mL) was added

and the suspension was stirred at rt for 1 h. The resin was filtered and washed with 1:1

TFA:CH2Cl2 (10 mL). The filtrates were combined and concentrated to give a 0.0146 g

(18% coupling yield) of a white solid.

Determination of TBDMS cleavage during the coupling reaction. A

suspension of the Wang resin (0.015 g) in DMF-d7 (1 mL) was agitated at rt for 30 min.

Then TBDMS-L-lactic acid (0.02 g) was added, the suspension was agitated for 1 h and

the solution was analyzed with 'H NMR: 'H NMR (DMF-d7) 8 0.08 (d, 7.86H), 0.87 (s,

11.8H), 1.31 (d, 3H), 1.34 (d, 0.93H), 4.18 (q, 1H), 4.34 (q, 0.31H).

N-ca-Fmoc-N-e-acetyl-L-lysine (3-4). The synthesis of 3-4 was adapted from a

125
procedure described by Bolin and coworkers. To a solution of N-E-acetyl-L-lysine

(0.500 g, 2.66 mmol) in CH2C12 (15 mL) was added diisopropylethylamine (0.685 g, 5.31

mmol) and chlorotrimethylsilane (0.574 g, 5.31 mmol). The reaction mixture was held at

reflux for 3 h and then cooled in an ice bath. Fmoc-Cl (0.618 g, 2.39 mmol) was added

and the reaction mixture was stirred on ice for 30 min and then at rt for 2 h. The reaction

mixture was concentrated and then distributed between Et20 (30 mL) and 2.5% NaHCO3

(25 mL). The phases were separated and the aqueous layer was washed twice with Et2O

(15 mL). The Et20 layers were extracted three times with water (20 mL). The combined

aqueous layers were acidified to pH 2 with 1 M HC1 and extracted three times with

EtOAc (30 mL). The combined EtOAc layers were washed with brine, dried over

Na2SO4, filtered and concentrated in vacuo to give 0.898 g (92%) of the product as a








white solid. 'H NMR (CD3OD) 6 1.43 (m, 2H), 1.51 (m, 2H), 1.70 (m, 1H), 1.85 (m,

1H), 1.91 (s, 3H), 3.16 (t, 2H), 4.14 (m, 1H), 4.21 (t, 1H), 4.35 (d, 2H), 7.30 (t, 2H), 7.38

(t, 2H), 7.67 (t, 2H), 7.78 (d, 2H).

N-a-t-Boc-D-Glutamic Acid (3-5). The synthesis of 3-5 was adapted from a

99
previously described procedure. To a solution of D-glutamic acid (3.00 g, 20.4 mmol),

water (60 mL) and dioxane (60 mL) was added K2C03 (5.6 g, 40.8 mmol) and (Boc)20

(4.9 g, 22.4 mmol). The reaction mixture was stirred at rt for 16 h and then washed three

times with Et20 (70 mL). The aqueous layer was cooled on ice, acidified to pH 2 with

solid citric acid, and extracted three times with EtOAc (100 mL). The EtOAc layers were

washed with brine until the brine wash was no longer acidic, dried over Na2SO4, filtered

and concentrated to give 4.31 g (86%) of the product as a gummy solid: 'H NMR

(CD3OD) 8 1.45 (s, 9H), 1.89 (m, 1H), 2.14 (m, 1H), 2.40 (t, 2H), 4.14 (m, 1H).

N-a-t-Boc-D-glutamic acid-a-benzyl ester (3-6). Using a procedure described

126
by Pawelczak et al., 3-5 (4.11 g, 16.6 mmol) was dissolved in DMF (7 mL) and Et3N

(1.68 g, 16.6 mmol) was added followed by the slow addition of benzyl bromide (3.12 g,

18.3 mmol). The reaction was stirred at rt for 15 h. The reaction mixture was diluted

with water (50 mL) and the aqueous layer was extracted with EtOAc (3 x 20 mL). The

combined EtOAc layers were extracted with 1 M Na2CO3 (3 x 20 mL) and then the

combined aqueous extract was acidified to pH 2 with 1.3 M HC1. The aqueous layer was

extracted with EtOAc (3 x 20 mL) and these combined EtOAc layers were washed with

0.02 M K2C03 (6 x 25 mL), 0.02 M K2C03 saturated with NaCi (25 mL) and brine (2 x

15 mL). The EtOAc layer was dried over Na2SO4, filtered and concentrated to give 2.81

g (50%) of the product as a white solid. The spectral data were in agreement with








published data.134 'H NMR (CDC13) 8 1.41 (s, 9H), 1.93 (m, 1 H), 2.14 (m, I H), 2.38 (m,

2H), 4.38 (m, 1H), 5.15 (s, 2H), 7.34 (s, 5H); IR (KBr): 3356,1745,1681 cm'.

D-Glutamic acid-a-benzyl ester trifluoroacetate salt (3-7). The synthesis of 3-

99
7 was adapted from a previously described procedure.99 To 3.29 g (9.8 mmol) of 3-6 was

added a 25% (v/v) solution of TFA:CH2Cl2 (30 mL). The reaction mixture was stirred at

rt for 2 h and concentrated to give the product as a yellow oil (3.37 g, 98%): 'H NMR

(CD3OD) 8 2.17 (m, 2H), 2.48 (m, 2H), 4.15 (t, 1H), 5.28 (s, 2H), 7.39 (m, 5H); IR

(neat): 3037,1746,1674cm'.

N-a-Fmoc-D-glutamic acid-a-benzyl ester (3-8). The synthesis of 3-8 was

adapted from a procedure described by Bolin and coworkers.125 To a solution of 3-7

(0.203 g, 0.58 mmol) in CH2CI2 (10 mL) was added diisopropylethylamine (0.223 g, 1.73

mmol) and chlorotrimethylsilane (0.125 g, 1.16 mmol). The reaction mixture was held at

reflux for 3 h and then cooled in an ice bath. Fmoc-Cl (0.099 g, 0.38 mmol) was added

and the reaction mixture was stirred on ice for 40 min and then at rt for 2 h. The reaction

mixture was concentrated and then distributed between Et20 (20 mL) and 2.5% NaHCO3

(15 mL). The phases were separated and the aqueous layer was washed twice with Et2O

(15 mL). The Et20 layers were extracted three times with water (20 mL). The combined

aqueous layers were acidified to pH 2 with 1 M HC1 and extracted three times with

EtOAc (20 mL). The combined EtOAc layers were washed with brine, dried over

Na2SO4, filtered and concentrated in vacuo to give 0.151 g (87%) of the product as a

white solid. The spectral data were in agreement with published data.135 'H NMR

(CD3OD) 8 1.94 (m, 1H), 2.16 (m, 1H), 2.38 (t, 2H), 4.15 (t, 1H), 4.30 (m, 1H), 4.32 (d,








2H), 5.14 (s, 2H), 7.29 (m, 5H), 7.34 (m, 4H), 7.63 (d, 2H), 7.76 (d, 2H); IR (KBr):

3331,1740,1690cm-1.

N-a-Fmoc-N-e-acetyl-L-Lysine-Wang Resin (3-9). The synthesis of 3-9 was

followed as described for 3-3a except starting with 0.155 g of the Wang Resin

(substitution: 0.71 mmol/g, 0.110 mmol) and Fmoc-Lys(Ac)-OH (0.135 g, 0.330 mmol).

Capping of remaining hydroxyl groups of the resin. This method was adapted

127
from a previously described procedure. To 3-9 (0.155 g, 0.110 mmol) was added 3 mL

DMF followed by acetic anhydride (0.067 g, 0.660 mnmol) and DMAP (1.34 mg, 0.011

mmol). The suspension was agitated with argon at rt for 1 h. The solution was drawn off

and the resin was washed 5 times with DMF (15 mL).

General Procedure for Fmoc deprotection. This general method was adapted

from a previously described procedure.136 To the resin was added enough 20% (v/v)

piperidine/DMF to cover it and the suspension was agitated with argon at rt for 30 min.

The solution was drawn off and the resin was washed 5 times with DMF (15 mL).

General Procedure for Coupling of Amino Acid to the Peptide Resin. This

general method was adapted from a previously described procedure.124 To a mixture of

DIC (4 eq) in DMF (5 mL) was added HOBt (4 eq). The solution was stirred on ice for

10 min, then the Fmoc-amino acid (4 eq) was added and the solution was stirred on ice

for an additional 10 min. This mixture was then added to the resin and the suspension

was agitated with argon at rt for 2 h. The solution was drawn off and the resin was

washed 5 times with DMF (15 mL). This procedure was then repeated.

General Procedure for the Kaiser test. This general method was followed as

99
described by Stewart and Young. After each coupling a small sample of the peptide-








resin (-2 mg) was placed in a test tube and washed three times with 1:1 EtOH:AcOH (2

mL) and three times with EtOH (2 mL). The washes were decanted each time. To the

washed peptide-resin was added 3 drops of a cyanide solution (2 mL 0.01 M KCN diluted

to 100 mL with pyridine), a ninhydrin solution (0.5 g ninhydrin in 10 mL n-BuOH), and a

phenol solution (80 g phenol in 20 mL n-BuOH). The test tube was then placed in a 100

C oil bath for 5 min. The coupling was complete and no free amines were present if the

solution was yellow and the resin beads were white. If the solution and beads were blue

the coupling was not complete.

General Procedure for Capping of remaining free amino groups. This

128
general method was adapted from a previously described procedure. To the resin was

added enough DMF to cover it, followed by acetic anhydride (6 eq) and pyridine (12 eq).

The suspension was agitated with argon at rt for 30 min. The solution was drawn off and

the resin was washed 5 times with DMF.

Cleavage of tetrapeptide from resin. This method was adapted from a
129
previously described procedure. 29 The resin (0.155 g) was washed several times with

DMF (15 mL), acetic acid (15 mL), and CH2Cl2 (20 mL). The resin was then shrunk by

washing with CH3OH (20 mL) several times and dried under vacuum. To the dry resin

70% (v/v) TFA/CH2Cl2 (5 mL) was added and the suspension was stirred at rt for 4 h.

The resin was filtered and washed with 70% (v/v) TFA/CH2C12 (5 mL). The filtrates

were combined, concentrated and 1-6 was isolated from the resulting residue by

precipitation with cold Et20. The solid was filtered and isolated to give 0.028 g of the

product in an overall yield of 36%. The peptide was analyzed by analytical RP-HPLC

(Vydac Cl 8 column, 4.6 x 250 mm). A linear gradient using 0.1% TFA in water (solvent








A) and 0.1% TFA in acetonitrile (solvent B) was programmed to increase from 35 to 55%

B over 20 min with a flow rate of 1 mL/min. The separation was monitored at 220 nm.

The peptide had a retention time of 14.25 min. HRMS calcd for C32H49N509S2 (M + Na)

734.2869, found 734.2844.

(S)-(1-(N-Carbobenzoxyamino)ethyliphosphonic Acid (3-14). The synthesis of

3-14 was adapted from previously described procedures. '30,131 To a solution of (S)-(1-

aminoethyl)phosphonic acid (0.500 g, 4 mmol) in 4 M NaOH (3.3 mL) cooled to 0 C

was added benzyl chloroformate (0.887 g, 5.2 mmol) and the reaction was stirred at rt for

14 h. The reaction mixture was washed three times with Et2O (10 mL), then the

combined Et20 layers were extracted once with saturated aqueous NaHCO3 (5 mL). The

aqueous layers were combined, acidified to pH 2 with concentrated HC1, and passed

through a column (30 mm diameter) of Dowex 50X8 200 (H) resin using distilled

water as the eluent. Acidic and UV active fractions were lyophilized to give 0.869 g

(84%) of the product as a white solid. The spectral data were in agreement with

published data.'30 'H NMR (CD3OD) 8 1.34 (dd, 3H, J = 7 Hz, 16 Hz), 3.98 (m, 1H),

5.08 (s, 2H), 7.34 (m, 5H); 3C NMR (CD3OD) 8 15.99,44.46,46.54,67.71, 128.86,

128.99, 129.44, 138.18; 31P NMR (CD3OD) 25.01; IR(KBr): 3294,1688, 1541, 1258

cmT'.

Dimethyl-(S)-1l-(N-Carbobenzoxyamino)ethyl]phosphonate (3-15). To a

solution of 3-14 (0.869 g, 3.35 mmol) in CH3OH (3 mL) was added excess ethereal

diazomethane prepared from Diazald. 137 The solution was concentrated in vacuo to

afford 0.909 g (94%) of the product as a white solid. The spectral data were in agreement

with published data.13 'H NMR (CDC13) 8 1.39 (dd, 3H, J = 7 Hz, 10 Hz), 3.73 (d, 3H, J








= 11 Hz), 3.77 (d, 3H, J = 10 Hz), 4.20 (m, 1H), 5.12 (s, 2H), 7.34 (m, 5H); 3C NMR

(CDC13) 8 15.91, 41.70, 43.81, 53.33, 53.45, 53.61, 53.92, 67.15, 128.08, 128.22, 128.50,

136.15; 31P NMR (CDC13) 8 28.67; IR (KBr): 3250, 1719,1543,1252,1043 cm-'.

Monomethyl-(S)-1l1-(N-Carbobenzoxyamino)ethyllphosphonate (2-6). Using

a procedure described by Campbell and Bermak9 3-15 (0.857 g, 2.99 mmol) was

dissolved in THF (7 mL) and aqueous KOH (14.95 mmol) was added. The reaction was

stirred at rt for 4.5 h. The THF was evaporated and then the reaction mixture was washed

three times with Et20 (15 mL). The aqueous layer was acidified to pH 2 with

concentrated HC1 and extracted three times with EtOAc (25 mL). The EtOAc layers

were combined and washed once with brine (15 mL), dried over Na2SO4, filtered and

concentrated to give the product as a white solid (0.64 g, 79%). The spectral data were in

agreement with published data.139 'H NMR (CDC13) 8 1.35 (dd, 3H, J = 7 Hz, 10 Hz),

3.72 (d, 3H, J = 11 Hz), 4.19 (m, 1H), 5.11 (s, 2H), 7.33 (m, 5H); 31P NMR (CDCI3) 8

28.87; IR (KBr): 3282, 1688, 1542, 1159, 1046, 991 cm'.

Ethyl-(R)-2-1 Methoxy-((S)- 1-IN-Carbobenzoxyaminolethyll phosphoryll

oxyl propionate (3-16). The synthesis of 3-16 was adapted from a procedure described
84
by Campbell. To a solution of 2-6 (0.603 g, 2.21 mmol), (S)-ethyl lactate (0.391 g, 3.32

mmol) and tris(4-chlorophenyl)phosphine (1.21 g, 3.32 mmol) dissolved in THF (22 mL)

was added diisopropyl azodicarboxylate (0.669 g, 3.32 mmol). The reaction was stirred

at rt for 14 h and then concentrated. The crude product was purified by flash column

chromatography (55 mm diameter) eluting with 7:3 EtOAc:hexane to yield the product as

a yellow oil (0.742 g, 90%): 'H NMR (CDCI3) 8 1.27 (t, 3H), 1.41 (dd, 1H, J = 8 Hz, 10

Hz), 1.42 (dd, 2H, J = 8 Hz, 9 Hz), 1.45 (d, 1 H, J = 7 Hz), 1.56 (d, 2H, J = 7 Hz), 3.74 (d,








2H, J = 11 Hz), 3.83 (d, 1H, J = 11 Hz), 4.22 (q, 2H), 4.25 (m, IH), 4.91 (m, 0.34H), 5.04

(m, 0.66H), 5.12 (s, 2H), 7.35 (m, 5H); 13C NMR (CDCI3) 8 14.38, 14.41,16.39,16.73,

19.35,19.42,19.51, 19.57,43.30,43.34,45.39,45.41, 52.73, 52.83,53.54, 53.64,61.95,

62.24,67.39,67.51, 71.11, 71.19,71.46,71.57,128.36, 128.39,128.43,128.46,128.53,

128.80, 129.28, 129.42, 129.46, 129.59, 133.64, 133.72, 133.78, 133.86; 31P NMR

(CDC13) 8 27.06, 27.94; IR (neat): 3251, 1719, 1233, 1100, 1046,991 cm'-; Anal. calcd.

for C1i6H24N07P: C, 51.47; H, 6.48; N, 3.75; found: C, 51.45; H, 6.60; N, 4.10.

Ethyl-(R)-2-([Methoxy-[(S)-l-aminoethyll phosphoryl] oxy] propionate (1-7).

To a solution of 3-16 (0.034 g, 0.091 mmol) in EtOAc (2 mL) was added 5% Pd/C (0.023

g). The reaction was stirred at rt under a balloon of H2 gas for 4 h. The reaction mixture

was then filtered through a bed of celite and the EtOAc filtrate was dried over Na2SO4

and concentrated by rotary evaporation to afford the product as a yellow oil (0.021 g,

95%). 'H NMR (CD3OD) 6 1.28 (t, 3H), 1.32 (dd, 1.5H, J = 8 Hz, 11 Hz), 1.34 (dd,

1.5H, J = 8 Hz, 11 Hz), 1.55 (d, 1.SH), 1.57 (d, 1.5H), 3.20 (m, 1H), 3.78 (d, 1.5H, J = 11

Hz), 3.85 (d, 1.5H. J = 11 Hz), 4.24 (q, 2H), 4.96 (m. IH); '3C NMR (CD3OD) 5 14.55,

14.57, 16.83, 17.18, 19.62, 19.71, 43.92, 44.13, 45.94, 46.09, 53.14, 53.24, 53.95, 54.05,

62.91, 63.14, 72.51, 72.59, 72.79, 72.89, 172.68, 172.71, 172.84; 3"P NMR (CD3OD) 8

31.86, 32.37; IR (neat): 3384, 1748, 1221, 1100, 1035, 987 cm-1.

Dimethyl-(S)-(1-aminoethyl)phosphonate (3-17). To a solution of 3-15 (0.615

g, 2.14 mmol) in EtOAc (30 mL) was added 5% Pd/C (0.547 g). The reaction was stirred

at rt under a balloon of H2 gas for 3 h. The reaction mixture was then filtered through a

bed ofcelite and the EtOAc filtrate was dried over Na2SO4 and concentrated by rotary

evaporation to afford the product as a yellow oil (0.138 g, 42%). The spectral data were








in agreement with published data.'140 'H NMR (CDaOD) 8 1.31 (dd, 3H, J = 8 Hz, 11

Hz), 3.18 (m, 1H), 3.79 (d, 3H, J = 11 Hz), 3.80 (d, 3H, J = 11 Hz); 13C NMR (CD3OD) 8

17.01,42.53,44.52, 52.54, 52.61, 52.64, 52.71; "P NMR (CD3OD) 8 32.96.

Dimethyl-(S)-I l-[N-(N-a-Boc-D-y-glutamyl-a-benzyl ester)]aminoethyl]

phosphonate (3-18). The synthesis of 3-18 was adapted from a procedure described by

Bartlett et al. To a solution of 3-6 (0.044 g, 0.131 mmol) in DMF (1 mL) was added 4-

methylmorpholine (0.013 g, 0.131 mmol) and the mixture was cooled to -10 C. Then

isobutyl chlorofonrmate (0.018 g, 0.131 mmol) was added and stirred for 2 min. This

solution was then transferred to a vial containing 3-17 (0.02 g, 0.131 mmol), which was

cooled to -10 C. The reaction mixture was stirred at -10 C for lh and at rt for 15 h and

then concentrated. The resulting residue was taken up in EtOAc (3 mL) and the white

solid that formed was filtered off. The EtOAc layer was then washed three times with

saturated aqueous NaHCO3 (1 mL), dried over MgSO4 and concentrated by rotary

evaporation to provide the product as a colorless oil (0.026 g, 42%). 'H NMR (CD3OD)

6 1.33 (dd, 3H. J = 8 Hz, 10 Hz), 1.42 (s, 9H), 1.89 (m, 1H), 2.14 (m, 1H), 2.32 (t, 2H),

3.75 (d, 3H, J = 11 Hz), 3.76 (d, 3H, J = 11 Hz), 4.15 (m, 1H), 4.45 (m, 1H), 5.17 (s, 2H),

7.35 (s, 5H); 3C NMR (CD30D) 8 15.19, 19.30, 22.25, 27.97, 28.29, 28.70, 30.97, 31.19,

32.85, 40.57, 42.68, 52.18, 53.84, 53.93, 53.96, 54.06, 54.43, 54.62, 60.61, 67.86, 80.73,

84.73, 129.22, 129.28, 129.55, 137.29, 158.12, 173.74, 173.89, 173.96; 31P NMR

(CD3OD) 8 29.19; IR (neat): 3278, 1740,1713,1165, 1034 cm'; HPLC (Vydac C18

column, 4.6 x 250mm; solvent A: H20 with 0.1% TFA, solvent B: CH3CN with 0.1%

TFA; gradient: 35 to 55% B over 20 min; flow rate: ImL/min; detection: 261 nm): tR=

10.95 min; HRMS calcd for C21H33N208P (M + Na) 495.1872, found 495.1614.








N-a-acetyl-cysteinyl(S-t-butylthio)-alanyl-D-y-glutamyl(ca-benzyl ester)-

lysyl(N-e-acetyl)-glycine methyl ester (3-19). The synthesis of 3-19 was adapted from a

procedure described by Bartlett et al.113 To a solution of 1-6 (10 mg, 0.014 mmol) in

DMF (0.4 mL) was added 4-methylmorpholine (2.8 mg, 0.028 mmol) and the mixture

was cooled to -10 C. Then isobutyl chloroformate (1.9 mg, 0.014 mmol) was added and

stirred for 2 min. This solution was then transferred to a vial containing glycine methyl

ester hydrochloride (1.8 mg, 0.014 mmol), which was cooled to -10 C. The reaction

mixture was stirred at -10 C for I h and at rt for 15 h and then concentrated to afford the

crude product as brown oil. HPLC (Vydac C18 column, 4.6 x 250mm; solvent A: H20

with 0.1% TFA, solvent B: CH3CN with 0.1% TFA; gradient: 35 to 55% B over 20 min;

flow rate: lmL/min; detection: 220 nm): tR= 20.54 min; HPLC/ESI-MS m/z: 783 (M+

H), 805 (M + Na).

Ethyl-(R)-2-[[Methoxy-[(S)- 1-IN- IN-a-acetyl-cysteinyl(S-t-butylthio)-alanyl-

D-y-glutamyl(a-benzyl ester)-lysyl(N-e-acetyl)]]aminoethyl] phosphoryl] oxy]

propionate (2-5). The synthesis of 2-5 was adapted from a procedure described by
al113
Bartlett et al. To a solution of 1-6 (53 mg, 0.074 mmol) in DMF (0.5 mL) was added

4-methylmorpholine (7.5 mg, 0.074 mmol) and the mixture was cooled to -10 C. Then

isobutyl chloroformate (10.1 mg, 0.074 mmol) was added and stirred for 2 min. This

solution was then transferred to a vial containing freshly prepared 1-7 (17.8 mg, 0.074

mmol), which was cooled to -10 C. The reaction mixture was stirred at -10 C for lh

and at rt for 15 h and then concentrated and subjected to the reaction conditions again

until 1-6 was consumed to afford the crude product as a brown oil.