вход по аккаунту


Analysis of Enzyme Structure and Activity by Protein Engineering.

код для вставкиСкачать
Volume 23
Number 7
July 1984
Pages 467-538
International Edition in English
Analysis of Enzyme Structure and Activity by Protein Engineering**
By Alan R. Fersht*, Jian-Ping Shi, Anthony J. Wilkinson, David M. Blow,
Paul Carter, Mary M. Y. Waye, and Greg P. Winter
Structure-activity relationships of enzymes can now be analyzed for the first time by the systematic alteration of protein structure. Recent developments in the chemical synthesis of
DNA fragments and recombinant DNA technology enable the facile modification of proteins by highly specific mutagenesis of their genes. Kinetic analysis of the mutant enzymes
combined with high-resolution structural data from protein X-ray crystallography allow direct measurements on the relationships between structure and function. In particular, the
strength and nature of enzyme-substrate interactions and their detailed roles in catalysis
and specificity can now be studied. We have developed such analysis of enzyme structurefunction by site-directed mutagenesis of the tyrosyl-tRNA synthetase from Bacillus steamthermophilus, concentrating so far on the subtle role of hydrogen bonding in both substrate
specificity and catalysis. We find that the energetics of tyrosine and ATP binding must be
analyzed in terms of an exchange reaction with solvent water. Based on this idea and structural data, we have engineered an enzyme of improved enzyme-substrate affinity, and there
thus appear to be real prospects of engineering proteins of new specificities, activities, and
structural properties. We are also using protein engineering to gather direct information on
the nature of enzyme catalysis. For example, we find the catalysis of formation of Tyr-AMP
from Tyr and ATP is due largely to electrostatic and hydrogen bonding interactions that are
stronger in the transition state than in the ground state-a “strain” mechanism rather than
acid-base or covalent catalysis.
1. Introduction
The study of enzymes has been revolutionized by two
major developments: knowledge of three-dimensional protein structure from X-ray protein crystallography and the
Prof. A. R. Fersht, J:P. Shi, A. J. Wilkinson
Department of Chemistry, Imperial College of Science and Technology
London SW7 2AY (UK)
Prof. D. M. Blow
Department of Physics, Imperial College of Science and Technology
London SW7 2AZ (UK)
P. Carter, Dr. M. M. Y. Waye, Dr. G . P. Winter
Laboratory of Molecular Biology, MRC Centre, Hills Road
Cambridge CB2 2QH (UK)
The review article is based on a lecture presented at The Third European
Symposium on Organic Chemistry (ESOC 111) in September, 1983. This
work was supported by the Medical Research Council of the UK.
Anyew. Chem. Int. Ed. Enyl. 23 (1984) 467-473
ability to produce large quantities of sometimes otherwise
inaccessible enzymes by genetic engineering. These methods are now being combined to produce enzymes having
novel properties by the new discipline of “protein engineering”.
Protein crystallography has provided, over the past
twenty years, the necessary structural information for modifying enzymes, but the
to change structure has been
the stumbling block to producing new enzymes. Although
there have been some isolated examples of synthesizing
enzymes de novo by classical peptide synthesis, this procedure is too laborious and too limited to be of general use
for producing novel enzymes. Further, synthetic proteins
often d o not fold into their native structures. The logical
Of protein synthesis is to
way to
nature and work at the level of DNA: DNA is the natural
0 Verlay Chemie GmbH, 0-6940 Weinheim, 1984
0570-0833/84/0707-0467 $ 02.50/0
genetic material and can be used indefinitely to generate
more supplies of protein; the simple rugged structure of
DNA causes no problems of folding; DNA may be mutated and recombined in vitro to generate novel genes in a
way analogous to the generation of new structures during
evolution. This process of producing new activities by mutation of existing structures is how the problem of producing novel enzymes is presently being approached. Although it is not yet possible to design new enzymes ab initio, since the rules governing the folding of protein chains
are not sufficiently well understood, it is possible to make
variations on existing structural themes by controlled mutation of the gene encoding the relevant protein. In this article, we outline the results obtained from experiments on
one particular enzyme and indicate the diversity and
wealth of information that can be obtained on enzymes
and enzymic catalysis.
2. Oligodeoxynucleotide-Directed Mutagenesis
The current method of generating specific mutations in
DNA uses a chemically synthesized fragment of DNA that
directs the mutation-oligodeoxynucleotide-directed mutagenesis. The mutator DNA fragment contains the codon
(or its complementary codon) for the amino acid that is to
replace the target amino residue in the native protein. The
procedure was introduced in 1973"l. The determination of
the DNA sequence of a bacteriophage (0x174) by Sanger's
laboratory in 1976-197812' enabled the method to be put
into practice. Oligodeoxynucleotide-directed mutagenesis
was used in 1978 to create point mutations in bacteriophage (DX174L3*41,
and has subsequently been fashioned
into a precision tool[51.At the end of 1982, three papers
were published reporting, for the first time, the site-directed mutagenesis of amino acid residues in the structural
genes of enzymes of defined me~hanism[~.'.~l.
One of these
examples involved an enzyme of known three-dimensional
structure, the tyrosyl-tRNA synthetase from BacilZus stearothermophilus161,and so it was possible to analyze the results in terms of structure and to initiate a systematic programme for dissecting the structure and activity of an enzyme.
The method of mutating amino acid residues in the tyrosyl-tRNA synthetase is i l l ~ s t r a t i v e(Fig.
~ ~ ~1).
~ ~The gene
coding for the enzyme was cloned into the bacteriophage
vector M13 (this is a single-stranded DNA bacteriophage
and is thus ideal for several of the following manipulations); a short oligodeoxynucleotide primer (12 to 20 residues) was synthesized that is complementary to the sequence of the DNA to be mutated, apart from a single mismatch; the mismatch was designed to convert the codon
for the target amino acid into that for the desired mutant
amino acid residue; the oligodeoxynucleotide was an-
Leu Tyr Cys Gly Phe
(+) - M 13 t e m p l a t e
1 I DNA polymerase I
2) DNA Ligase
closed circular
transformation o f E,coli JMlOl
Leu Tvr Cys Gly Phe
wild ( + ) - s t r a n d
Leu Tyr Gly Gly Phe
mutant (+)-strand
Fig. I. Schematic representation of oligodeoxynucleotide-directedmutagenesis. 'The mismatched primer is designed to
convert the codon for cysteine (TGC) into that for glycine (GGC). The gene for mutation must be obtained in the form of
single-stranded D N A before the primer can be annealed. This was done for the tyrosyl-tRNA synthetase gene by subcloning it from the plasmid pBR322 into bacteriophage M13, a parasitic bacteriophage that does not destroy its host. 'I Indithe portion of pBR322. The
cates the limits of the insert; the thickened line indicates the B . stearothermophilus DNA;
genome of this filamentous bacteriophage is a covalent circle of single-stranded DNA that becomes a covalent doublestranded circle (termed RF') during its replication in E. coli. The subcloning is performed o n the double-stranded form,
and the mutation on the single-stranded. A mutant gene is formed by copying the DNA of the bacteriophage and cloned
gene from the 3' end of the mismatched primer using D N A polymerase and ligating to the 5'end with D N A ligase. The resultant heteroduplex is introduced into E. coli (strain JMlOl in this case) to yield clones of cells; some contain the bacteriophage with the gene for the wild-type enzyme, others the gene for the mutant. The clones may be screened as described
in the text. The clones produce large quantities of tyrosyl-tRNA synthetase. For further details see [6,9].
Angew. Chem. Int. Ed. Engl. 23 (1984) 467-473
nealed to the single-stranded D N A and used as a primer
for replication catalyzed by a D N A polymerase; the new
strand was ligated using D N A ligase to produce a heteroduplex of wild-type and mutant DNA; host cells of Escherichia coZi were transformed by the heteroduplex to give
clones containing mutant or wild-type vector. Mutant
clones were identified using the 32P-labeled mutagenic
primer as a probe in a hybridization assay, since it binds
preferentially to the mutant DNA"']. The frequency of mutants varies from 1 to 50%. Mutant enzymes are produced
in high yield from the clones.
Oligodeoxynucleotide mutagenesis has been extended to
make deletions and insertions of sequences. For example,
a mutator D N A fragment containing an additional codon
somewhere in its middle anneals to the target with the
three additional bases looping out. After replication, the
synthetic strand contains the extra codon. Conversely, a
longer primer, e.g. a 24-mer, may be synthesized in which
the first 12 residues are complementary to one region of
the gene and the other 12 to a distal region. On annealing
the primer to the gene, the intervening section of the gene
is looped out. After replication, the synthetic strand lacks
the looped-out section. There are also more sophisticated
methods of causing deletions["].
3. The Tyrosyl-tRNA Synthetase from
B. stearothermophilus
The enzyme studied most extensively so far by oligodeoxynucleotide-directed mutagenesis is the tyrosyl-tRNA
synthetase (TyrTS)"' from B. stearothermophilus. To date,
we have prepared and analyzed some 20 mutants involving
single mutations, double mutations, and extensive deletions.
This enzyme catalyzes the aminoacylation of tRNATy' in
a two-step reaction [eqs. (a) and (b)]["I.
E + Tyr + ATP
+ E.Tyr-AMP + PP;
E.Tyr-AMP + tRNATyr-+ Tyr-tRNA'Yr + AMP + E
ing reaction are readily determined by straightforward assays; the pre-steady state kinetics of activation may be followed by stopped-flow fluorescence studies["] and those of
the transfer step [eq. (b)] by rapid quenching methods[''].
The enzyme also has interesting properties since it is a
dimer. Kinetic studies show that the two active sites interonly one tyrosine as well as one
act so that in
tRNATY"21.221 and one tyrosyl adenylate are bound tightly
per two active
It is a symmetrical dimer of M,= 2 x 47 500['3.'41.The nucleotide sequence of the gene has been determined from a
clone in the vector pBR322[l5I. X-ray crystallographic studies on the enzyme at 0.3 nm resolution have been published (Fig. 2)[16a1,and in subsequent work refinement has
been extended to a nominal 0.21 nm[16b1.Most importantly,
the crystal structure of the enzyme-bound tyrosyl-adenylate complex has also been
so that there is the
rare opportunity of direct knowledge of the interactions of
the enzyme with a substrate (Fig. 3).
Aside from this, the tyrosyl-tRNA synthetase has convenient kinetic properties that render it a suitable target for
mutagenesis studies: the enzyme is readily assayed by active-site titration, and hence accurate and reproducible
steady-state kinetic measurements may be made["]; the
partial reaction of activation [eq. (a)] and the overall charg[*I Abbreviations: TyrTS= tyrosyl-tRNS synthetase; mutants are indicated
thus: TyrTS(Cys35-Ser35)=TyrTScontaining the mutation Cys-35 to Ser35: k,.,, and K M are the catalytic and Michaelis constants, respectively, in
the Michaelis-Menten equation.
Angew. Chem. Int. Ed. Engl. 23 (1984) 467-473
Fig. 2. Crystal structure of one of the hymmetrical subunits of the tyrosyltRNA synthetase. Note that the tyrosyl adenylate (see Figs. 3 and 4) makes
hydrogen bonds with side chains from the helices spanning residues 47-57
and 164-182, and the strand of B-sheet 13.
Fig. 3. Tyrosyl-tRNA synthetase forms a stable, enzyme-bound tyrosyl adenylate complex. On addition of tyrosinc and ATP to the crystals of the enzyme, they first crack and then reanneal to produce the intermediate which is
stable in the absence of tRNA or high ince cent rations of pyrophosphate. The
sketch shows the backbone of the enzyme and the bound substrate which
were determined directly by X-ray diffraction studies.
The tyrosyl-tRNA synthetase from B. stearothermophiZus-a thermophilic organism-has one further very important property: the enzyme is thermostable and loses no
activity on prolonged incubation at 56"C, unlike the enzyme from Escherichia coZi which is rapidly denatured. As
the mutant tyrosyl-tRNA synthetases used in the studies
are produced from the clones in E. coli, provided the mutation does not reduce thermostability, any background native enzyme from E. coli can be readily removed by heating. Thus, weakly active mutants may be studied without
interference from active native enzyme.
4. Strategy for Studying Structure-Activity
The utilization of binding energy is at the heart of enzyme catalysis and specificity. There is a simple equation
relating the rate constant k,,,/K, of the Michaelis-Menten
equation to the binding energy of the enzyme and sub~trate["*~'~
[eq. (c)]
R T In (kcat/KM)= RT In ( k T / h ) - AG
'- AGs
where AGs is the binding energy between the enzyme and
substrate, AG is the energy of activation, and k and h are
the Boltzmann and Planck constants, respectively.
Using site-directed mutagenesis, we are able to alter the
binding energy term AG, by changing the side chains of the
amino acid residues that interact with the substrate. The
changes in binding energy may be calculated from k,,,/K,
using eq. (c) and the importance of the interaction evaluated. This provides an experimental means of measuring
AG, for the interactions. Theoretical c o n s i d e r a t i ~ n s [ ~ ~ ~ ~ ' ~
show that reaction rate is optimized when the binding energies of these interactions are realized in the enzyme-transition state complex rather than the enzyme-substrate complex-the concept of transition-state stabilization. Furthermore, hydrogen bonds should be particularly important in
mediating such differential binding effects because the
strength of hydrogen bonding varies strongly with interatomic distance and hence is sensitive to the movement of
Fig. 4. Active site of the tyrosyl-tRNA synthetase with the bound tyrosyl adeatoms during the reaction. Accordingly, the initial strategy
nylate. Top: an expanded and more detailed view of Figure 3. Bottom: schematic sketch of residues that probably form hydrogen bonds. These interacof our experiments was to alter residues that form hytions are tentative at the resolution of the structure, but most have now been
drogen bonds with the substrates.
confirmed by site-directed mutagenesis.
5. Nature of Hydrogen Bonding of Enzymes with
The mutation of Cys-35 to Ser-35 in tyrosine-tRNA synthetase was the very first experiment in which a residue at
the active site of an enzyme of known three-dimensional
structure was specifically mutated by protein engineering,
leading to an enzyme of altered activityf6].As shown in Figure 4, Cys-35 forms a hydrogen bond with the 3'-hydroxyl
of the ribose moiety of Tyr-AMP. The mutation experiment changes -SH to -OH in the amino acid side chain.
Although the absolute strength of the hydrogen bond
-OH.. .O- is greater than that of -SH.. .O- in simple
intermolecular associations, it is found on comparing the
wild-type TyrTS with the mutant TyrTS(Cys35-Ser35) (Table 1) that the former enzyme has a higher affinity (1.2 kcal
mol-') for the ATP substrate [calculated from eq. (c)].
There are two reasons for this:
1) It is not the absolute strength of a hydrogen bond in
an enzyme-substrate complex that affects the enzyme-substrate dissociation constant that is important; rather it is
Table 1. Activation of tyrosine hy TyrTS-wild type and mutants at 25°C (details see [9]).
the dgference in energies of the hydrogen bond donor/acceptor in the free enzyme and the acceptor/donor in the
free substrate when they are separately hydrogen-bonded
to solvent water relative to the situation when they are hydrogen-bonded to each other [eq. (d)]:
2) There are geometrical constraints on hydrogen-bond
formation in enzyme-substrate complexes because the donor and acceptor groups are held at fixed distances from
each other. For example, the S-0 distance in S H - e - 0
Angew. Chem. Int. Ed. Engl. 23 (1984) 467-473
bonds is 0.5 nm longer than the 0-0 in O H . . .O. The
substitution of the side-chain hydroxy group of a serine for
a cysteine residue is thus energetically unfavorable because the 0-0 distance is too long in a rigid active site"].
Comparison of TyrTS(Cys35-Ser35) and another mutant in
which the side chain has been completely removed,
TyrTS(Cys35-Gly35) (Table I), shows that the serine side
chain contributes no apparent binding energy with the
substrate and, if anything, the presence of the side chain
actually weakens binding.
It is concluded that the presence of a poor (too long) hydrogen bond in an enzyme-substrate complex can increase
the dissociation constant and the value of kca,/KM.Perhaps the removal of a weak hydrogen bond by site-directed
mutagenesis could be used to increase the affinity of an
enzyme for its substrate.
Wild t y p e
His48 T h r 5 l
SSt/KM = 8 , 4 0 0
s-' M-'
His48 Pro5 1
Gly48 T h r S l
2 0 8 , 0 0 0 s-' M-'
1 , 1 0 0 s-' M-'
Gly48 Pro5 1
Fig. 5. Use of double mutants for investigating structural changes. (Mutated
residues are shown in italics.) The high value of k,,,/KM for the mutant
TyrTS(Thr51-Pro51) requires the presence of both His48 and Pro51. If His48
is first converted into Gly48 (right), mutation of Thr51 to Pro51 to generate
the double mutant TyrTS(His48-Gly48, rhr51-Pro51) at the bottom leads to
little increase in the rate constant.
6. Improving the Affinity of the Enzyme for ATP
Preliminary crystallographic studies indicated that Thr51 makes a long hydrogen bond with the tyrosyl adenylate
(Fig. 4). Furthermore, although all the other side chains illustrated in Figure 4 are conserved in the E. coli enzyme, it
has a proline residue in this position that cannot possibly
make a hydrogen bond with the substrate and will also distort the secondary structure in this region. Accordingly, in
the light of the conclusion of the last section, we changed
Thr-51 first to Ala-51 to determine the effect of deleting
the hydrogen bond, and then to Pro-51 to examine the results of causing a larger structural change[261.
Table 2. Increase in affinity for ATP in tyrosine activation by TyrTs upon
mutation of Thr-51. Reaction at 25°C (details see [26]).
might thus result from a new contact formed between enzyme and substrate, from an improved interaction at an existing contact, or from a direct interaction of the substrate
and the pyrrolidine ring of the proline. An earlier mutagenesis experiment had shown that His-48 forms an Hbond to ATP which only gains 1.2 kcal mol-'. We were
therefore interested whether the effect of proline was to
distort the polypeptide backbone and thereby improve the
interaction of His-48 with the substrate. To prove this, we
constructed a double mutant TyrTS(His48,ThrSlGly48,Pro51). This mutant has approximately the same activity as the TyrTS(His48-Gly48) mutant, indicating that
the effect of Thr-51 is mediated via the imidazole group of
His-48 (Fig. 4). From the cycle shown in Figure 5, we can
calculate, using equation (c), that the imidazole group of
His-48 contributes 1.2 kcal mol ' to binding energy in the
wild-type enzyme and 3 kcal mol-' in the Pro-51 mutant.
This result shows how a second mutation introduced into
the active site of an enzyme can be used as a probe of the
structural change introduced by the first mutation.
K M (Tyr)
kc,t/KM (AT€)'
8 400
The results for activation of tyrosine by these mutants
are listed in Table 2. It is seen that the KM for ATP of
TyrTS(ThrS1-Ala51) is two times lower than that of the
wild type. It appears, in fact, that the interaction of threonine with the substrate is very weak and hence lowers the affinity of the enzyme for the substrate. Site-directed mutagenesis is thus a good fine-structure probe for hydrogen
Mutation of the wild type to TyrTS(Thr51-Pro51)
causes a massive increase in the value of k,,,/K,, involving both an increase in k,,, and a decrease in KM (for
ATP). Thus, engineering a single point-mutation has
caused a significant improvement in enzymic activity and,
hence, shows that there are real biotechnological possibilities of engineering the activities of enzymes in vitro.
What is the reason for this dramatic improvement in affinity of the enzyme for ATP? Thr-51 is located in an a-helix (amino acid residues 47-57) (Fig. 2), and the introduction of a proline would be expected to distort the helix
structure between residues 47-50. The improved affinity
Angew. Chem. I n t . Ed. Engl. 23 (1984) 467-473
7. Evidence for Transition-State Stabilization in
the Activation Reaction
How is the attack of the tyrosyl carboxy group on ATP
[eq. (a)] catalyzed? This reaction involves an associative inline displacement of pyrophosphate from ATP [eq. (e)]["].
I3O' O A d o
General base-catalysis of the nucleophilic attack is not
possible since the carboxylate is fully ionized. Nucleophilic catalysis by a group on the enzyme forming an inter47 1
mediate with A M P is unlikely since ATP is already an “activated” compound containing a good leaving group (pyrophosphate). The remaining possibilities are electrostatic
catalysis and transition state stabilization, whereby certain
groups on the enzyme make stronger interactions with the
substrate transition state. For transition state stabilization
it is predicted that when a group on the enzyme is removed, as in a site-directed mutagenesis experiment, the
affinity of the enzyme for the substrate will be decreased to
some extent, but the affinity for the transition state will be
decreased to a much greater extent than if the transition
state had been stabilized. This means that the value of KM
may be increased to a greater or lesser extent but the value
of k,,, will fall, as shown in Table 3. Particularly notewor-
so far in residues that hydrogen-bond with ATP. Conversely, k,,, is insensitive to changing a residue that forms a hydrogen bond with the tyrosine hydroxy group (Tyr-34 to
Phe-34 in Figure 4, (unpublished)). This suggests that during the activation reaction the bound tyrosine side-chain
may move relatively little, but that the bound ATP moves
towards the tyrosine carboxy as it makes a nucleophilic attack on the a-phosphoryl group of ATP. Since the strength
of the hydrogen bonding of the enzyme with ATP increases as the transition state is reached, the enzyme must
presumably be compressing the ATP and the tyrosine together.
8. Optimization of the Rate of Enzymic Reactions
during Evolution
Table 3. Influence of various mutations on the activation of tyrosine.
TyrTS [a]
TyrTS(cys35-Gly35) [a]
TyrTS(Cys35-Ser35) [a]
TyrTS(His48-Gly48) [b]
TyrTS(His45-Asn45) [b]
TyrTS(Glnl95-Glyl95) [b]
[a] Ref. [9]. [b] Unpublished results.
thy is the change from His-45 to Asn-45 leading to a 2000fold drop in k,,,, whilst the values of K , are hardly affected (J.-P. Shi,unpublished). Similarly, the rate of attack
of pyrophosphate on the tyrosyl adenylate complex is lowered 2000-fold in TyrTS(Hys45-Asn45) (unpublished data).
His-45 does not interact with the tyrosyl adenylate in the
crystal structure (Fig. 4). Model building studies by Dr. R .
J. Leatherbarrow (unpublished) show that the y-phosphoryl group of the pentacovalent intermediate in eq. (e) can
make a hydrogen bond with His-45[*].
Geometrical changes that occur during the activation
reaction may be inferred from some of the mutations in hydrogen-bonding groups we have engineered based on the
following reasoning. If an amino acid residue on the enzyme hydrogen bonds with the substrate equally well in
the enzyme-transition state complex as in the enzyme-substrate complex, removal of that residue will affect only K M
and not k,,, (see [**I pp. 244-247). Similarly, removal of a
residue that hydrogen bonds better in the enzyme-transition state complex will lower k,,,. We have found that k,,,
for the activation reaction is sensitive to all changes made
[*I As indicated in Table 3, changes in k,,, and K N may also arise from
mechanisms such as “induced fit” and “non-productive binding”. Because of this, the results of previous kinetic studies on structure-reactivity
relationships from studies where the structure of the substrate has been
altered have usually failed to provide unambiguous evidence for “strain”
mechanisms (transition-state stabilization) [21]. However, the changes induced by site-directed mutagenesis of the tyrosyl-tRNA synthetase are
more amenable for producing unambiguous analysis for the following
reasons: 1) It will be possible to solve the structures of the native enzyme
and enzyme-substrate complexes for the mutants by X-ray crystallography. 2) Many of the structural changes will be sufficiently small that the
binding energetics may be calculated. Even in the absence of precise
structural data, the combination of chemical intuition and inspection of
the enzyme structure by computer graphics strongly suggests that transition-state stabilization by electrostatics and hydrogen bonding is operating.
Various proposals have been made as to how enzymic
rate constants should respond to selective pressure in evolution to maximize the overall rate of reaction[20.21,281
. O ne
way of testing the validity of these models is by site-directed mutagenesis. Although the mutants we have generated are not necessarily the evolutionary precursors of the
present native enzyme, they are possible forms of the enzyme that have been rejected during evolution, since the
rate of spontaneous mutation is sufficiently high for them to
have arisen and hence been tested. One simple theory for
the effects of selective pressure on the evolution of maximum rate[20.2’1
predicts I ) that the catalytic term k,,,/K,
tends to a maximum (this is the consequence of the enzyme being complementary to the transition-state structure
of the substrate) and 2) that values of k,,, and K , tend to
increase in parallel so that the value of K , lies above the
normal concentration of the substrate encountered in vivo.
The net result is that the term [S], kcal/(KM+ [S],) is maximwhich gives the rate in vivo, where [S], is the
ized in eq. (9,
concentration of substrate in vivo. This is achieved by low-
ering the complementarity of the enzyme with the substrate in its unreacted structure while maintaining complementarity with the transition-state structure. Experimentally, the value of K M for ATP in the aminoacylation reaction
of native enzyme is 2.5 mM, a value close to the 2-3 mM
found for the concentration of ATP found in vivo for
many organisms. The full intrinsic binding energy of ATP
and a protein is clearly not utilized here-the KM for ATP
and myosin is
M, for example[291.In general, the values of k,,,/KM for the mutants are lower than that for wildtype. The notable exception is the mutant TyrTS(Thr51Pro51), where k , , / K , for the mutant is much higher (Table 4). However, the high value for kca,/KMis achieved by
compensation of a low value of k,,, by a very low value of
K , for ATP. At physiological concentrations of ATP (23 mM), the native enzyme is more active than the mutant
(as calculated by substituting the experimental values of
k,,,, K,, and [ATP] into eq. (9).
At lower concentrations of
ATP, however, the mutant TyrTS(Thr5 I-Pro51) is more active. This illustrates the importance of enzymes maximizing their rate by adjusting their KM values to the concenA n g e w . Chem. Int. Ed. Engl. 23 (1984) 467-473
Table 4. Kinetics of the aminoacylation of tRNA [a].
TyrTS [cl
TyrTS(Gln 195-Gly195)[e]
ATP conc.
K M kc,t/KM
[SKI] [mM] [ S - I M - ' ]
0. I5
0.019 95800
1.251 3200
[ATPI [ m ~ l
u [s-'1
[a] Rate of aminoacylation of tRNA at 25 "C [mol Tyr-tRNA. mol ' enzyme.
s-'I. k,,, and K N are rounded off, but k , , , / K M is calculated from the exact
values. [b] Approx. concentration in vivo. [c] Ref. [26]. [d] Ref. [9]. [el Unpublished results.
trations of substrates in vivo. One other mutant
TyrTS(Thr51-AlaSl) has similar activity to the native enzyme at 25°C in vitro.
in catalysis are being identified ; in more sophisticated experiments, the role of enzyme-substrate interactions in catalysis and specificity is being studied. Future experiments
will explore long-range effects of structural changes on activity and the transmission of the effects of substrate and
effector binding through protein subunits. Experiments are
being planned in various laboratories to change the thermostability of enzymes and to learn the structural basis of
enzyme stability. Undoubtedly, protein engineering will be
used to study enzyme folding. At the commercial level,
protein engineering will be used to alter the physical properties of enzymes, such as their thermostability, solubility,
p H optima etc., and hence optimize their use in biotechnological processes in vitro. Protein engineering will also be
used to alter kinetic properties, such as the values of k,,,
and K , for substrates, and to change even the specificities
of enzymes. The ultimate goal is to design tailor-made enzymes for every reaction.
Received: April 5, 1984 [A 497 IE]
German version: Angew. Chem. 96 (1984) 455
9. Structural Organization of the Tyrosyl-tRNA
The crystal structure of the enzyme consists of a well-ordered N-terminal domain comprising residues 1-319 that
binds tyrosyl adenylate, while the remaining 100 residues
of the C-terminus are disordered['61. The fragment of gene
coding for the N-terminal domain has been prepared by a
gene-deletion procedure, and the truncated protein isolated"']. The truncated enzyme is fully active in the activation reaction [eq. (a)], but does not catalyze the transfer of
tyrosine to tRNA or bind Tyr-tRNA. The structural division of the tyrosyl-tRNA synthetase is thus also a functional one in which an N-terminal domain is responsible
for amino acid activation and a C-terminal domain for
tRNA binding. A series of more general mutational experiments on the alanyl-tRNA synthetase, an enzyme of unknown three-dimensional structure, shows that this protein
also has distinct regions for activation and transfer reaction~[~'].
10. Summary and Outlook
Although we know from personal communications from
our colleagues that many other enzymes are now being
studied by oligodeoxynucleotide-directed mutagenesis,
there is little published work at present relating changes in
activity to changes in structure. The work on 0-lactam a ~ e ' ~cannot
, ~ ] be extended at present because the threedimensional structure of the enzyme is not known ; however, crystallographic work is in progress. Three mutants of
dihydrofolate reductase have been preparedL3'],and further
work is in progress by other groups. However, we anticipate that in the next year or two there will be a plethora of
exciting data published.
Protein engineering appears to be the answer to an enzymologist's prayer. The prospects are quite open-ended. At
the purely academic level, the relationship between structure and activity of enzymes is now being analyzed directly: in the most basic experiments, groups directly involved
Angew. Chem. I n [ . Ed. Engl. 23 (1984) 467-473
[I] H. Schott, H. Kossel, J . A m . Chem. Soc. 95 (1973) 3778.
[2] F. Sanger, A. R. Coulson, T. Friedmann, G. M. Air, B. G. Barrell, N. L.
Brown, J. C. Fiddes, C. A. Hutchinson, P. M. Slocombe, M. Smith, J .
Mol. Biol. 125 (1978) 225.
[3] C. A. Hutchinson 111, S. Phillips, M. H. Edgell, S. Gillam, P. Jahnke, M.
Smith, J . Biol. Chem. 253 (1978) 6551.
[4] A. Razin, T. Hirose, K. Itakura, A. D. Riggs, Proc. Null. Acad. Sci. USA
75 (1978) 4268.
[5] M. J. Zoller, M. Smith, Methods Enzymol. 100 (1983) 468.
[6] G. Winter, A. R. Fersht, A. J. Wilkinson, M. Zoller, M. Smith, Nature
(London) 299 (1982) 756.
[7] G. Dalbadie-McFarland, L. W. Cohen, A. D. Riggs, D. Morin, K. Itakura, J. H. Richards, Proc. Natl. Acad. Sci. USA 79 (1982) 6409.
[8] I. S. Sigal, B. G. Harwood, R. Arentzen, Proc. Natl. Acad. Sci. USA 79
(1982) 7175.
191 A. J. Wilkinson, A. R. Fersht, D. M. Blow, G. Winter, Biochemistry 22
(1983) 3581.
[lo] R. Wallace, M. Schold, M. J. Johnson, P. Dembek, K. Itakura, Nucleic
Acids Res. 9 (1981) 3647.
[I 11 M. M. Waye, G. Winter, A. J. Wilkinson, A. R. Fersht, EMBO J . 2 (1983)
(121 A. R. Fersht, R. Jakes, Biochemistr.v 14 (1975) 3350.
[I31 G. L. E. Koch, Biochemistry 13 (1974) 2307.
[I41 M. J. Irwin, J. Nyborg, B. R. Reid, D. M. Blow, J. Mol. Biol. 105 (1976)
[IS] G. Winter, G. L. E. Koch, B. S. Hartley, D. G. Barker, Eur. J. Biochem.
132 (1983) 383.
(161 a) T. N. Bhat, D. M. Blow, P. Brick, J. Nyborg, J. Mol. Biol. 122 (1982)
407; b) T. N. Bhat, P. Brick, D. N. Blow, unpublished.
[I71 J. Rubin, D. M. Blow, J. Mol. Biol. 145 (1981) 489.
[I81 A. R. Fersht, J. S. Ashford, C. J. Bruton, R. Jakes, G. L. E. Koch, B. S.
Hartley, Biochemistry 14 (1975) I.
[I91 A. R. Fersht, R. S. M. Mulvey, G. L. E. Koch, Biochemistry 14 (1975)
[20] A. R. Fersht, Biochemistry 14 (1975) 5 .
[21] R. Jakes, A. R. Fersht, Biochemisfrv 14 (1975) 3344.
[22] P. Dessen, G. Zaccai, S. Blanquet, J. Mol. Biol. 159 (1982) 651.
[23] R. S. Mulvey, A. R. Fersht, Biochemistry 16 (1977) 4005.
[24] A. R. Fersht, Proc. R. SOC.London Ser. B 187 (1974) 397.
[25] A. R. Fersht: Enzyme Structure and Mechanism, Freeman, New York
[26] A. J. Wilkinson, A. R. Fersht, D. M. Blow, P. Carter, G. Winter, Nature
(London) 307 (1984) 187.
[27] G. Lowe, G. Tansley, Tetrahedron 40 (1984) 113.
[28] W. J. Albery, J. R. Knowles, Angew. Chem. 89 (1977) 295; Angew. Chem.
Int. Ed. Engl. 16 (1977) 285.
[29] H. G. Mannherz, H. Schenk, R. S. Goody, Eur. J . Biochem. 48 (1974)
[30] M. Jasin, L. Regan, P. Schimmel, Nature (London) 306 (1983) 441.
[31] J. E. Villafranca, H. E. Howell, D. H. Voet, M. S. Strobel, R. C. Ogden,
J. N. Abelson, J. Kraut, Science 222 (1983) 782.
[32] R. Jaenicke, Angew. Chem. 96 (1984) 385; Angew. Chem. I n f . Ed. Engl.
23 (1984) 395.
Без категории
Размер файла
705 Кб
engineering, structure, enzymes, protein, activity, analysis
Пожаловаться на содержимое документа