The Lock-and-Key Principle The State of the Art- 100 Years On EDITED BY JEAN-PAUL BEHR Universite Louis Pasteur de Strasbourg, Illkirch, France

Perspectives in Supramolecular Chemistry Volume I

John Wiley & Sons Chichester - New York - Brisbane - Toronto - Singapore

The Lock-and-Key Principle

Editorial Board

Founding Editor J.-M. Lehn, UniversitC Louis Pasteur, Institut le Bel, 4 Rue Blaisse Pascal, F-67070 Strasbourg, France

Editors J.-P. Behr, FacultC de Pharmacie, B.P. 24, F-67401 Illkirch, France

G. R. Desiraju, University of Hyderabad, School of Chemistry, Hyderabad 500134, India

A. D. Hamilton, University of Pittsburgh, Department of Chemistry, Pittsburgh, PA 15260, USA

T. Kunitake, Kyushu University, Faculty of Engineering, Hakozaki, Fukuoka 812, Japan

D. N. Reinhoudt, University of Twente, Faculty of Chemical Technology, PO Box 217, NL-7500 AE Enschede, The Netherlands

J.-P. Sauvage, Universitt Louis Pasteur, Institut le Bel, 4 Rue Blaisse Pascal, F-67070 Strasbourg, France

The Lock-and-Key Principle The State of the Art- 100 Years On EDITED BY JEAN-PAUL BEHR Universite Louis Pasteur de Strasbourg, Illkirch, France

Perspectives in Supramolecular Chemistry Volume I

John Wiley & Sons Chichester - New York - Brisbane - Toronto - Singapore

Copyright01994 by John Wiley & Sons Ltd, Baffins Lane, Chichester, West Sussex PO19 IUD, England

Telephone: National Chichester (0243) 779777 International +44 243 779777

All rights reserved.

No part of this book may be reproduced by any means, or transmitted, or translated into a machine language without the written permission of the publisher.

Other Wiley Editorial Offices

John Wiley & Sons, Inc., 605 Third Avenue, New York, NY 10158-0012, USA

Jacaranda Wiley Ltd. 33 Park Road, Milton, Queensland 4064, Australia

John Wiley & Sons (Canada) Ltd, 22 Worcester Road, Rexdale, Ontario M9W 1L1, Canada

John Wiley & Sons (SEA) Pte Ltd, 37 Jalan Pemimpin #05-04, Block B, Union Industrial Building, Singapore 2057

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN 0 471 93902 1

Typeset in 10/12pt Times by Dobbie Typesetting Limited, Tavistock, Devon Printed and bound in Great Britain by Biddles Ltd, Guildford, Surrey

Contents

Contributors vii

Preface ix

1 Emil Fischer’s Lock-and-Key Hypothesis after 100 Years-

Friedrich Cramer Towards a Supracellular Chemistry 1

2 Molecular Recognition in Biology: Models for Analysis of Protein-Ligand Interactions 25 Doron Lancet, Amnon Horovitz and Ephraim Katchalski-Katzir

3 New Biocatalysts via Chemical Modification Ian M. Bell and Donald Hilvert

73

4 Oligonucleotides: Superspecific Ligands for Targeting Nucleic Acids and Proteins and Development of Molecular Devices V. V. Vlassov

89

5 Macrocycles and Antibodies as Catalysts D. B. Smithrud and S. J. Benkovic

149

6 Lock-and-Key Processes at Crystalline Interfaces: Relevance to the Spontaneous Generation of Chirality Isabelle Weissbuch, Ronit Popovitz-Biro, Leslie Leiserowitz and Meir Lahav

173

7 A Model of the Origin of Life and Perspectives in Supramolecular Engineering H. Kuhn and J. Waser

8 Perspectives in Supramolecular Chemistry-From the Lock-and-Key Image to the Information Paradigm Jean-Marie Lehn

247

307

319 Index

V

Contributors

Ian M. Bell, Department of Chemistry and Molecular Biology, The Scripps Research Institute, 10666 North Torrey Pines Road, La Jolla, CA 92037,' USA

S. J. Benkovic, Chemistry Department, Pennsylvania State University, 152 Davey Laboratory, University Park, PA 16802-6300, USA

Friedrich Cramer, Max-Planck-Institut fur Experimentelle Medizin, H. Rein Strasse 3, D-37075 Gottingen, Germany

Donald Hilvert, Department of Chemistry and Molecular Biology, The Scripps Research Institute, 10666 North Torrey Pines Road, La Jolla, CA 92037, USA

Amnon Horovitz, The Weizmann Institute of Science, Rehovot 76100, Israel

Ephraim Katchalski-Katzir, The Weizmann Institute of Science, Rehovot 76100, Israel

H. Kuhn, (formerly Max-Planck-Institut fur biophysicalische Chemie, Gottingen) Ringoldwilstrasse 50, 3656 Tschingel ob Gunten, Switzerland

M. Lahav, The Weizmann Institute of Science, Rehovot 76100, Israel

Doron Lancet, The Weizmann Institute of Science, Rehovot 76100, Israel

Jean-Marie Lehn, UniversitC Louis Pasteur, Institut le Bel, 4 Rue Blaisse Pascal, F-67070 Strasbourg, France

L. Leiserowitz, The Weizman Institute of Science, Rehovot 76100, Israel

D. B. Smithrud, Chemistry Department, Pennsylvania State University, 152 Davey Laboratory, University Park, PA 16802-6300, USA

V. V. Vlassov, Institute of Bioorganic Chemistry, Lavrentiev Prospekt 8,630090 Novosibirsk, Russia

J. Waser, (formerly California Institute of Technology) 6120 Terryhill Drive, La Jolla, CA 92037, USA

vii

PREFACE

In 1894, a farsighted (bio)chemist came up with an idea that would provide the key to understanding the phenomena of molecular recognition that underlie life itself. After a century of refinement his concept still remains productive, as is evident from the many scientists who draw on it. This book is their tribute to the man who unlocked our thought.

ix

Chapter 1

Emil Fischer’s Lock-and-Key Hypothesis after 100 years- Towards a Supracellular Chemistry FRIEDRICH CRA MER Max-Planck-Institut fur Experimentelle Medizin, Gottingen, Germany

1. INTRODUCTION

Around 100 years ago Emil Fischer in his famous paper [ 1 ] proposed that enzyme and substrate can be compared to lock and key. Since that time this metaphor has been used to describe enzyme action. In this chapter we shall try to explore whether this metaphor still holds. On the one hand, enzymology has made enormous progress in these 100 years, but on the other hand the lock- and-key concept has greatly changed. In Figure 1 three types of ‘keys’ are shown.

The classical key, which Emil Fischer had in mind, is pushed into the lock and turned clockwise in order to open the lock. Thus, the process is chiral. The key shown in the figure was manufactured in the year 1853, one year after Emil Fischer was born; thus, probably this type of key was imprinted in the young boy’s conceptual memory. Today, the keys have become much more refined. Around the turn of the century the safety lock was invented, which does a kind of proof- reading of the key. At present an entirely new system is being installed; namely the magnetic card. This card can unlock money sources and hotel rooms. We shall discuss further these three key concepts and give a few examples of each of them.

2. CLASSICAL LOCKS

One almost classical molecule for lock-and-key studies is cyclodextrin. I entered the field of supramolecular chemistry 45 years ago with this molecule and it still

The Lock-and-Key Principle Edited by J.-P. Behr 0 1994 John Wiley & Sons Ltd

2 The Lock-and-Key Principle

Figure 1 Lock and keys from the last 100 years: (a) key from 1853; (b) safety lock and key; and (c) magnetic tape as a key

seems to be of interest. It has a well-defined hydrophobic cavity. It readily forms a complex with chlorobenzene but not with bromobenzene (Figure 2). Of special interest with respect to enzyme models are the kinetics of the formation of such inclusion compounds. The kinetics can be measured using azo dyes, which change their spectra on formation of the inclusion compound in solution. In Figure3, all three molecules independent of the substituent R have the same equilibrium constant. However, the rate constants are different by five orders of magnitude because the stereo- chemistry of the ‘threading-in process’ becomes more unfavorable the larger the substituent R.

Numerous enzyme models are based on this principle. For instance, pyrophosphates, if suitably substituted, are hydrolyzed by the cyclodextrin cavity, or rather by the OH group of the glucose in cyclodextrin, 400 times faster than without cyclodextrin (Figure 4).

On a similar basis, we have constructed a chymotrypsin model by furnishing /3-cyclodextrin with imidazole groups. Ester hydrolysis is 300 times faster, but does not yet come close to the actual catalytic rate of chymotrypsin (Figure 5) .

(b) Figure 2 Geometrical fit of chlorobenzene into a-cyclodextrin: (a) empty cavity and (b) cavity filled with chlorobenzene [2]

Fischer’s Hypothesis after 100 Years 3

+ a-Cyclodextrln

I -0,s

H 1.7x1@

Me 150

Et 2.0

Figure 3 Kinetic recognition of various azo dyes by the cyclodextrin cavity [ 31

CI

Figure 4 Catalysis of pyrophosphate splitting by cyclodextrin [ 4 J

3. CATENANES

It is tempting to try to synthesize catenanes according to the lock-and-key principle. An attempt made 35 years ago is illustrated in Figure 6. The substituted hydroquinone formed the inclusion complex readily. This was then oxidized in the hope that the ring would close to form a catenane. This, however, was done at a time when NMR was not available, a situation hard to think of nowadays. Therefore we could not finally prove that the catenane had been formed. Around 35 years ago I told as a kind of joke the following story. If one tried to give a rough estimation of the frequency of the molecular vibrations of the two rings against each other, one would arrive at the conclusion that this frequency was in the ultrasonic region between 10000 and 20000 Hz, a frequency which the human ear cannot perceive but the ear of a dog can.

4 The Lock-and-Key Principle

Ser 195

0

Cklodextrin (CD)

Water a-C D-OH &CDimidazole Chymotrypsin

Molecular weight 18 1000 1400 2.5 x 104

Km 2.6XlO-3 -1X1O-3 1 x l o d

kR >4 x 10' -1 xi07 1 xi07

kcat lk H20 1 .o 2.5 300 -1 x 10-4

pKof catalytic group 14.0 12.0 6.95 6.95

kcal 0.13 x lo-" 0.32 x lo-" 38.4 x 10-4 3.0

Figure 5 Chymotrypsin modeled by 8-cyclodextrin [ 5 1

One day, a friend with a dog came to visit me in the laboratory. When I took the beaker with the compound and shook it, the dog started barking; every time I shook the compound, the dog barked again. This for me was proof that the compound really had been formed. Proof of synthesis is now much easier, and recently Harada et al. [7] have published a paper about this type of catenane, the existence of which can be proven by NMR (Figure7).

Long molecular tubes can be formed in the following way [ 81. a-Cyclodextrin is threaded onto poly(ethy1ene glycol) of defined length. The complex is sealed at each end and the product is crosslinked with epichlorohydrin. Subsequently, the seals are removed and the empty tube can be isolated and filled with other long-chain molecules (Figure 8).

4. SAFETY LOCKS

A safety lock allows a much more precise recognition of its key by certain mechanical devices which are able to sense the small dents on the key. This type

Fischer's Hypothesis after 100 Years

-6A-

I 8 /

CH2.SH HSCH,

Figure 6 An early attempt to synthesize a catenane by the lock-and-key principle

5

0,0' 2 , 2 3 , s a, a' H3 (DM-P-CD) - - - - -

I' 8.10 8.00 7.90 7.80 7.70 7.60 <.60 4.50 4.40 4.30 4.20

I I I I

6 (PPm) - Figure 7 Catenane formed with cyclodextrin according to Harada [ 71 (DM-P-CD:

dimet hyl-0-cyclodextrin)

6 The Lock-and-Key Principle

O,N -Q NH

NO,

.... ....... ...... ... Q. .O .....'Po.... ............... N H ~ N O ~

NO2

CH&HCH&I \ / 0

..................

NO,

NaOH (25%)

Molecular tube

Figure 8 Preparation of a molecular tube according to Harada [ 81

of recognition is apparently used when very high precision in biosynthetic processes is required, especially in the synthesis of informational macromolecules. However, we are not going to discuss proofreading in DNA synthesis. We have done some work on the fidelity of protein biosynthesis, in particular on the aminoacylation of tRNA which is the key step in introducing the correct amino acid into its proper place in the protein [ 91. The aminoacyl tRNA synthetases have a twofold task in recognizing their substrates. At first they must recognize and activate the amino acid to form the adenylated amino acid, then in a second step this activated amino acid must be transferred to its proper tRNA (Figure 9). The three-dimensional structure of yeast phenylalanine tRNA is shown in Figure 10.

Amino acids are, in principle, very similar, and the differentiation between them, e.g. isoleucine and valine, seems to be hopeless using simple lock-and- key principles alone (Figure 11). One can estimate that the process should be erroneous by 20%; in other words, every fifth amino acid should be wrongly placed. In contrast, we can show that the isoleucyl synthetase selects its proper substrate against the similar valine with a fidelity of one error in 40 OOO. How is this possible? The recognition process goes through a selection cascade, as shown in Figure 12.

The primary selection, as measured by direct binding studies, is indeed very poor. There is only a threefold preference for isoleucine over valine. In a second selection step, this fidelity is improved by one error in 240. These two steps

7

Ir 0

Fig

ure

9 C

ours

e of

the

am

inoa

cyla

tion

of

tRN

A

8 The Lock-and-Key Principle

Figure 10 Three-dimensional structure of yeast phenylalanine tRNA (after S.-H. K :im)

are of the classical Michaelis-Menten type of recognition; that is, the recognition occurs in a reversible manner. One error in 240 is by far not good enough, because it would mean that each protein has more than one error. Therefore two proofreading steps follow in which the incorrect product is hydrolyzed preferentially. However, the correct product also has to be partially sacrificed. This occurs twice and finally one arrives at a fidelity of one error in 38 OOO. In this proofreading about 80% of the originally activated, correct isoleucyl AMP is sacrificed. Therefore this is an energy-consuming recognition. Five molecules of ATP are used for recognition and one molecule of ATP for

Fischer’s Hypothesis after 100 Years 9

Figure 11 Discrimination between isoleucine and valine [ 91

activation. This reaction is therefore an irreversible dissipative process in which ATP energy is pumping up the specificity (Figure 13).

With other synthetases there are similar mechanisms [ 101. In general, the fidelity of amino acid recognition is of the order of one error in 10OO0, thereby consuming energy. The aminoacyl synthetases are not classical enzymes with Michaelis-Menten microreversibility, rather they work in a unidirectional way.

What do we know about the structural details of the recognition process? Some synthetases have been crystallized and their structures been analyzed, e.g. seryl synthetase (Figure 14) [ 11 ] .

Aspartyl and glutamyl tRNA synthetases have also been co-crystallized with their tRNAs and analyzed, thus giving a picture of how these molecules are attached to each other. But this does not give any mechanistic details about the dynamics of the recognition process. Such information comes from kinetic studies. We have measured the relative AAG values for the activation of various amino acids, as compared with the substrate isoleucine, and plotted these values (Figures 15 and 16) against the available surface areas of the various amino acids, which one can take from the literature. The data very nicely lie on one line, at least those for amino acids which are shorter than isoleucine (glycine, leucine, alanine, valine). As the amino acids become longer a break of this line occurs and leucine joins the longer ones on a different line. This we interpret as a conformational change in the enzyme, like a plug controlling the substrate that must be removed. After that, the longer amino acids can also be accommodated, albeit in a less favorable situation. This applies for the first binding step (Figure 15). If one does the same for the second binding step (Figure 16) the story becomes more complicated, but still the various amino acids lie on defined lines. The very short ones, glycine and alanine, fit into the entrance part of the cavity, while the longer ones can only be accommodated if again two plugs are pushed aside. We arrive at such conclusions from our kinetic data. Only upon preparing this chapter was it realized that this is exactly the principle according to which the safety lock works, as shown in Figure 1.

10 The Lock-and-Key Principle

Amino acid pool

210 000 lle 21 0 000 Val

a, 140 000 - 70 000 1 :3

69 163 4 837 1 :240

8 1 : 12 000

1 Val

incorrect

21 0 000 a

1 :38000 a, L

L

38 000 lle

correct

Figure 12 Proofreading cascade in the selection of isoleucine by isoleucyl synthetase

5. MAGNETIC LOCKS

One can foresee that in a few years the mechanical locks will belong to the past. The magnetic card is coming. On this, the information is inscribed on magnetic tape in a linear fashion and read off in sequence. This is the principle of information laid down in DNA and RNA. This information can be transcribed in various ways such as through reverse transcriptase and DNA polymerase and the polymerase chain reaction (PCR). The PCR has become very important and was the subject of the 1994 Nobel Prize in Chemistry (Figure 17).

Fischer’s Hypothesis after 100 Years 11

I I

Destroy 4 No Release yes w

V V V Waste Product

Figure 13 Dissipative recognition process of isoleucine by isoleucyl synthetase [ 91

12 The Lock-and-Key Principle

Figure 14 Three-dimensional structures of (a) seryl synthetase in two different representations and (b) glutamyl synthetase as cocrystals with their substrate tRNAs [ 1 1 ]

Fischer's Hypothesis after 100 Years

(a)

14

12

10 A

E 2 Y

6

4 a

2

0

c

L

13

Figure 15 (a) Plot of AAG,* versus available surface area for the first activation step of several amino acids. Recognition mechanism for smaller (b) and larger amino acids (c) [11,121

14

10

n d a 4

The Lock-and-Key Principle

2

0

Figure 16 (a) Plot of AAGI, versus available surface area for the second activation step of several amino acids. Recognition mechanism for smaller (b) and larger (c) amino acids [11,121

Fischer’s Hypothesis after 100 Years 15

mRNA

mRNA-cDNA hybrid

Fragmented mRNA

Dou ble-stranded cDNA which can be cloned

Figure 17 Mechanism of polymerases in cloning procedures

A very important area of research involves the synthesis and application of antisense RNA and DNA. The idea is to synthesize pieces of nucleic acids which are complementary to certain genes or parts of them. By hybridization the piece of nucleic acid is attached to the gene and rigidly fixed, thereby blocking the reading of this gene. Blocking the reading of information is the key to this technique. We have done some experiments in this direction. An oligonucleotide was furnished with psoralene and then hybridized with another oligonucleotide (Figure 18). Psoralene has the capacity to intercalate the nucleic acid and on irradiation it reacts with the nuclear base (Figure 19). A reaction was carried out with an antisense oligonucleotide directed against the a-chain of hemoglobin, and indeed the in vitro biosynthesis of the a-chain in the rabbit reticulocyte system was inhibited with this antisense oligonucleotide (Figure 20).

16

-

S I 0- I"-0

Y

t

Fig

ure

18

Synt

hesi

s of

an

anti

sens

e ol

igon

ucle

otid

e fu

rnis

hed

wit

h ps

oral

ene

[ 131