вход по аккаунту


Flexible Drug Molecules and Dynamic Receptors.

код для вставкиСкачать
[2] R. G. Kirste, W A . Kruse, J . Schelten, Makromol. Chem. 162, 299
[3] a) A . I/: Tobolskg: Properties and Structure of Polymers. Wiley, New
York 1960; b) Revised edition by M . Hofmann, Berliner Union, Stuttgart 1967, pp. 118f.
[4] W W Grawsfey, Adv. Polym. Sci. 16, 1 (1974).
[5] a) F . Btreche: Physical Properties of Polymers. Interscience, New York
1962, p. 61; h) J. Chem. Phys. 20, 1959 (1952); 25, 599 (1956).
[6] M . Hofmann, Rheol. Acta 6. 92 (1967).
[7] P. J . Flory: Principles of Polymer Chemistry. Cornell Univ. Press,
New York 1953, p. 402.
[8] H . A . Sruurr: Die Physik der Hochpolymeren. Springer, Berlin 1953,
Vol. 2, p. 655.
[9] M . Huj>nann, Rheol. Acta 6, 377 (1967).
[lo] See ref. 171. p. 405; W Kuhn, F . Griin, Kolloid-Z. 101, 248 (1942):
cf. P. J . Flory, Angew. Chem. 87, 787 (1975).
[I 11 See ref. [3a], p. 94.
[12] M . Hofmann, Kolloid-Z. Z . Polym. 250, 197 (1972).
[I31 See ref. 171. p. 425; P . J . F l o r j , 7: G . Fox. J r . , J. Am. Chem. SOC. 73,
1904 (1951): J. Polym. Sci. 5 , 745 (1950).
1141 M . Huflmunn. H . Kriimer, R. Kuhn: Polymeranalytik. Thieme, Stuttgart
1977, Vol. 1 .
[l5] M . Hofmann, K . Rother, Makromol. Chem. 80, 95 (1964).
[I61 G. Dobrowofski, W Schnabef, Makromolekulares Colloquium, Freiburg,
March 3, 1977.
[17] M . Hofmann, Makromol. Chem. 153,99 (1972).
[18] M . Hoffmann. Rheol. Acta 6, 82 (1967).
[!9] M . Mooney. J. Appl. Phys. 1 1 , 582 (1940); R . S. R i d i n , D . $4Saunders,
Philos. Trans. R. SOC.London A243.251 (1951).
[20] M . Hqffmann, Makromol. Chem. 174. 167 (1973).
[21] a) A. S. Lodge, Rheol. Acta 7 , 379 (1968); b) R. Takserman-Krozer,
A . Ziabicky, J. Polym. Sci. A2, 7, 2005 (1969); A2, 8, 321 (1970); c)
F. S. Edwards, Proc. Phys. SOC.London 85, 613 (1965); 91, 513 (1967);
92, 9 (1967); Discuss. Faraday Soc. 49, 43 (1970); J. Phys. A 6 , 1169,
1186 (1973).
1221 See ref. 171, pp. 577, 512; P. J . Flory, J . Chem. Phys. 18, 108 (1950).
1231 M . L . Huggins, J. Phys. Chem. 46, 151 (1942).
[24] M . Dauuil, J . P Cotton, €3. Farnoux, G . Janninck, G . Sarmu, H . Benoit,
R. Dupfessiu, C . Picot, P. de Gennes, Macromolecules 8, 804 (1975).
[25] P. Debye, J. Phys. Colloid Chem. 51, 18 (1947).
1261 G. K Sehufz, 2. Phys. Chem. 193, 168 (1944).
1271 See ref. [5a], p. 19; P. Debye, F. Bueche, J. Chem. Phys. 11, 470
(1943); A. Ishihora. J. Phys. SOC.Jpn. 5, 201 (1950).
1281 See ref. [7], p. 599.
1291 G. Gee, L . R. G. Peloar, Trans. Faraday SOC.38, 147 (1942); G. Gee,
W I . C . Orr, ibid. 42, 507 (1946).
1301 C . E . H . Bawn, R . F . J . Freeman, A . R . Kamallidin, Trans. Faraday
SOC.46, 677 (1950).
1311 J . Furukawu, H . Inaguki, Kautsch. Gummi Kunstst. 29.744 (1976).
[32] U . Eisele, Lecture at the IISRP Conference Williamsburg (USA) 1976.
Flexible Drug Molecules and Dynamic Receptors[**]
By R. J. P. Williams[*]
When a small flexible drug molecule binds to its likewise mobile receptor (protein, membrane
etc.) the shape and function of both can change. The study of the nature and extent of these changes
by several independent methods gives an insight into the mode of action of drugs. The static
lock and key model will most probably have to be revised or be replaced by the nebulous
concept of dynamic states.
1. Introduction
The basic problem of drug action in biological systems
is easily formulated. In order to interact with a biological
system any drug must first bind and may then react, whence
a comprehensive minimal two-stage reaction path can be
where D is the drug which undergoes a rapid reversible binding
to a biological receptor L, i.e. a protein, DNA, RNA or
a membrane etc., giving DL. This reaction has an equilibrium
binding constant
[*] Prof. Dr. R. J . P. Williams
Inorganic Chemistry Laboratory
South Parks Road, Oxford OX1 3QR (England)
[**] This article is based on the Merck, Sharp and Dohme Scientific Lecture
1976 given in London. It was also given in outline at the Belgian Chemical
Society Meeting in Namur, 1976.
and there can be many successive steps of this kind before
the 'final' binding condition is reached. The second step in
the above simplified scheme [eq. (I)], which may or may
not be a required part of action, is an irreversible chemical
combination of D in the form D' with a part of the biological
system L', where L' may be a protein, RNA, DNA, a membrane etc. which has incorporated D (or D'). L and L' could
be the same receptor site of course. The reaction rate constant
in its simplest form is a first order rate constant, kDL, but
again many such steps could be involved.
My analysis of drug action will be based upon the structural
features of these reaction paths starting from the structures
of D and L themselves. By structural features I imply the
whole series of conformational states through which the two
species D and L must go in order to reach some final state
DL or D'L'. Recently the nature of such pathways has
been stressed by Feeney, Roberts and Burgen"', who were
motivated by their observations, using nuclear magnetic
resonance (NMR) spectroscopy, on the mobility of certain
drug and hormone molecules. My independent and parallel
interest in the problem of conformational mobility has arisen
through studies of both small and large molecules in solution
also using sophisticated NMR methods for conformational
analysis. In the past the solution structures of D and L (which
Angrw. Chem. I n t . Ed. Engl. 16, 766-777 i1977 J
are the ones to be discussed in a drug/receptor interaction)
have often been assumed to be known once X-ray crystal
structures had been completed. This approach ignores the
statistical occupancy and the time dependencies of many different conformations of a molecule in solution and may well
have led to over-precise thinking about the general need for
structural fitting, e.g. as in the notions of lock-and-key and
induced fit. Examples illustrate the point.
The study of different penicillins in the crystalline state
by X-ray crystallography showed that there were two possible
ring conformers and that different substituents on the fourmembered ring led to a switch from one conformation to
the other (Fig. 1). We have studied the conformation of some
of the same penicillins in aqueous solution by NMR probe
wriggling of the drug to a ‘best’ site. Drug design might
have to be based on selected dynamic properties of a molecular
frame as well as upon the possibility of providing a particular
structure to fit a supposedly rigid receptor. Finally it will
be shown in this article that the large molecules which form
receptors are not necessarily more rigid than the small molecule drugs, and the same statistical fluctuations and timedependent changes must be analyzed for them. This motion
also has an influence upon the constants in eq. (1).
The obvious need for a careful analysis of the mobilities
of drugs and receptors before indulging in a discussion of
their interaction leads me to turn away from drug action
for a while and to describe recently obtained knowledge of
structures in solution and of the dynamics of these molecular
2. General Comments about Rigidity of Small Molecules in Solution
Fig. I . General formula oi penicillins: spatial arrangement of the atoms
in crystalline penicillin G and ampicillin.
methodsr2]but our results show clearly that there is little
if any dependence of structure of the penicillins in solution
on substitution of its rings (see Fig. 1). We also noted that
the data suggested that although one conformation was
dominant there was a smaller percentage of the other conformation in fast equilibrium with it. In principle the rate constants for conformational flipping of the molecules could have
been determined too. We are forced to ask which particular
structure if any is important for drug action and, has the
nature and rate of conformational mobility an importance
of its own?
A rather general inspection of drug molecules shows in
fact that only a limited number of drugs have rigid frames
and obviously their conformations in the bound state are
easily described from crystal or solution studies (but see below).
In all other cases there are different conformations open to
the drug molecule and these may well be in rapid equilibrium,
as in the case of the penicillins. We need to consider the
relative energies of these different conformations (open to
the drug in the unbound state) since, presuming that only
a limited set of the conformational forms are bound by the
receptor, there is a population bias against the drug being
bound in these forms in direct proportion to the probability
of achieving the limited conformation set in the free state.
Drug design might then be based partly on fitting, but if
we are dealing with a conformationally mobile drug, partly
on the restriction of conformational freedom of the (free)
drug. However, this last approach could fail too for the very
flexibility of the drug might also be an unavoidable requirement in order to achieve final binding or reaction, and in
this case a knowledge of the conformational mobility of the
drug would be essential in an understanding of action. In
the above Scheme [eq. (1 )] certain steps, kf, would represent
necessary movements on the way to a final binding site-a
Angew. Chem. Int. E d . Engl. 16, 766-777 ( 1 9 7 7 )
Apart from the conformations of penicillins in solution
we have examined the conformations of a series of amino-acids,
some small peptides, some sugars, and a series of nucleotides,
where there is obviously a much greater variety of possible
mobilityr3~1‘. The general impression gained from these studies
is that even though there must be mobility in the small molecules (rotation about single bonds with limited steric hindrance) certain “conformational families” or a very small set
of families peculiar to each group of molecules in turn are
overwhelmingly favored, possibly through the solvation by
water. A family of conformations is to be thought of as a
relatively small group of structures related to one another
by minor changes in bond angles. We can illustrate the point
by reference to the simple nucleotides but rather than making
direct reference to the deduced nucleotide conformations, derived from NMR spectroscopic data, I shall use a comparison
of the experimental NMR information so that experiment
and interpretation are not confused.
Our experimental procedure starts from an examination
and an assignment of the NMR spectrum of a molecule in
solution in as much detail as possible. This means that each
‘H, I3C, 31P,19F nucleus in the molecule is studied as far
as is possible, but here I shall describe ‘H-NMR data in
large part. These spectra in themselves contain coupling constant data derived from spectral fine structure and the coupling
constants are related to the relative position in space of coupled
protons. For example, in such a system of atoms as C ( H ’ F
C(H’) the geometric relationship of H’ and H Zcan be uncovered. Restrictions upon free rotation about C-C bonds are
then obvious in a qualitative and sometimes in a semi-quantitative way. Again proton relaxation times of the spectral signals
can be used to calculate the non-bonded H-H distances
and further data on small local regions of the conformation
comes from the study of the nuclear Overhauser effect. All
these methods have been used for some time by a large number
of research workers. The considerable extension of the conformational analysis which we have made is to examine the
NMR spectra of the small molecule in the presence of bound
(in fast exchange) paramagnetic ions (see Fig. 2). The NMR
spectra are found to be perturbed in two ways: 1) shifts
of absorption lines ( 6 ) arise if the paramagnetic center has
Fig. 3. 5'-Adenosine monophosphate (5-AMP); bonds about which there
is apparent free rotation are indicated by arrows.
Fig. 2. The effect of a paramagnetic reagent placed at the origin of space
(a) upon the line position (shift b-c) and line of width (broadening b-dj
of an NMR spectrum (see ref. [3]). The broadening of lines is directly
related t o TZm,and the shift is given as 6 for the most general rhomhic
field case. In an axial field, R = 0.
Table 1. NMR shifts(6)of nucleotide sugar protons by lanthanoid(ir1j-containing probes given as the ratio R =6(i-H)/6(5'-H) [9] [see Fig. 2, Fig. 3 and
a fast electron relaxation time; 2) changes in relaxation time,
e.g. broadening of lines ( B ) arise when the paramagnetic
center has a long electron relaxation time. The equations
show that 6 is a vector quantity depending upon the distance,
r, of the nucleus under inspection from the paramagnetic
probe atom, and upon the angle 6 between the principal
axis of the paramagnetic probe and the direction from probe
to the nucleus (see Fig. 2), while B is a scalar quantity. In
the equations A and Care constants independent of the nucleus
under study. (The above shift equation [eq. (2)] assumes that
the bound paramagnetic ion generates an axially symmetric
magnetic field. Details of the methods and measurements
which allow the field to be so defined are given
Let us assume that the site of binding of the paramagnetic
ion to the molecule under study can be determined (which
is the case), then these shift and relaxation measurements
give directly a set of conformational parameters, 6 and B,
just as coupling constants are conformational parameters and
they can be used empirically as comparative molecular conformation data on very many different atoms which are separated
by considerable distances. Of course all the different NMR
methods add together to give the largest number of conformational parameters, and together should be used in the definition
of structure"].
[*] All NMR methods provide data which refer t o the average conformation.
It is a matter of convenience whether we search to describe the average
in terms of a single conformation with non-Idealized bond-lengths and bond
angles [2,3,5] or in terms of combinations of two, or three, or four idealized
conformations etc. It has often been the case that averaged conformations
have been analyzed usinysums oftwoor threeconformers where the contributing conformers introduced are those suggested by theoretical energy calculations in the gas phase. As shown by Feeney et d . [7] it is always possible
t o fit NMR data by such a procedure. Even though it is certain that better
fits of experimental parameters can be obtained by using such semi-theoretical
approaches, since larger numbers of variables are introduced, than using
single conformation fitt~ngthe fit must be suspect as the theory does not
apply to solvated species. Thus it is essential that many indept.ildent methods
are used to check structural representations of molecules in solution, e.9.
coupling constants, relaxation data, shift data, and nuclear Overhauser effects,
using first m e and only later combinations of two conformations.
R = 6(i-Hj/6(5'-H)
0.1 1
0.2 1
Table 1 gives shift data for the protons of some nucleotides
(Fig. 3). The observed shift 6(i-H) of the NMR signal of a
proton i-H due to presence of a bound lanthanide(rI1) ion
is expressed by the ratio R of this observed shift relative
to the shift of a chosen proton, here 6(5'-H) of the 5'-H
proton (see Fig. 2). The paramagnetic lanthanide ions have
been used to generate the shifts but a wide variety of reagents
can be employed. A striking result is that, almost within
experimental error limits, all the nucleotides give the same
set of conformation parameters. Broadening (relaxation) and
coupling constant data are also almost invariant which can
only mean that independent of the base (pyrimidine or purine)
and of the sugar (oxy- or deoxyribose), the general conformational outline of nucleotide (including mobility averaging)
is fixed, and it is surprisingly closely related to the single
conformation observed in crystals. (As an aside we observed
that although the crystal structures of deoxy- and oxy-uracil
monophosphate were different from that of deoxy- and oxycytidine monophosphate this difference was not seen in solution and it may well be that the difference seen is an "artefact"
of crystallization as in the case of the penicillin.) These results
were deduced independently from coupling constant data
alone by Sundara/ingam[6~who has gone so far as to state
that the nucleotides must be relatively rigid in solution.
As there is clearly a high degree of constraint in the mobility
of these molecules the relative stability of the observed (averaged) set of conformations seen in these nucleotides must
be due in part to the nature of water as a solvent. If we
change solvent we observe a different stable conformational
family. In dimethyl sulfoxide the (average) conformation of
the nucleotide monomers is different from that found in water,
and in some aqueous systems we find that even salt and
denaturants such as urea also affect the conformations of
small molecules, as indicated by probe measurementsr3I. We
Angrw. Chem. lnt. Ed. Enqi. 16, 766 777 (1977)
note that receptor sites may be very unlike water in their
“solvation” properties and may prefer energetically a quite
different conformational family from that found in aqueous
solution. All these data tell us nothing about the nature of
other non-preferred conformations or the rate of flipping to
lowly populated conformational states, although we may presume this flipping is fast since we see no separate signals
for different states (but of course we do not know that such
states of low probability exist). In fact many conformations
widely different from those determined by NMR methods
of analysis could be reached and for 1 % of the time almost
any conformation X could be present and it would go undetected by these solution methods. However, if X is the desired
conformation for action there is then a factor of 100 against
its formation for this particular molecule and it must be
possible to build a better molecule, drug, for binding alone
where this mobility factor is considerably reduced.
The implications of these findings for the study of drugs
is simple but has not been clearly stated before:
1. Only for those drugs which have rigid frames (Table
2), are we able to neglect mobility, but even here we must
consider the aqueous (or other) solvation environment. For
example our study of cyclic 3’,5’-adenosine monophosphate
(3’,5’-CAMP)gave the same structure for the two rings containing the phosphate and the sugar (X-ray structure analysis),
and we may assume that this frame is relatively rigid[’]. The
Table 2. Rigid and flexible drug molecules [a].
Relatively rigid
(Many alkaloids)
Straight-chain alcohols
Nitrogen mustards
(N-methyl bis(2-chloroethyl)amine) and derivatives
Linear polypeptides
[a] It is of great interest t o make molecular models of drugs and attempt
to visualize the number of different static and dynamic modes of action
which they could have.
receptor could bind exactly this frame, but just because the
hydration remains unknown it is conceivable that it is the
hydrated form of the molecule which generates action when
the stereochemistry of the receptor site can not be deduced
from the solution or crystal structure studies of the small
2. Other drug molecules may be much more mobile (Table
2); for example, we have been unable to determine a “structure”
for 2’,3’-CAMP by our methods, a fact which suggests that
the NMR datacan not be represented by a single conformation
family[91.For such a molecule (drug) which has considerable
mobility between two o r more rather different conformational
families knowledge of the conformations observed in crystals
or in solution is of much more limited value for we do not
know not only the hydration but which or how many conformationally different steps are required in each of the steps
Anguw. Chem. I n t . Ed. Engl. 16. 766 777 ( 1 9 7 7 )
of equation (l), and we d o not know which are the important
conformations in the final condition. The problem becomes
even more serious as we turn to molecules of increasing size,
and we note that it is not just a problem of drug binding
alone but is perhaps a general feature of biological activity.
3. From our studies, the orientation of the base in 3’3’CAMP is clearly not fixed relative to the two rings of the
cyclic system to the same degree as was found in the simple
nucleotides 5‘- or 3’-AMPr81. Thus it may be easy for this
part of 3‘,5’-CAMPto wriggle or corkscrew its way to a required
binding site, receptor. Thus, whether a molecule binds or
not can depend upon whether it can reach a certain site
or not, for the receptor site need not be on an open surface[”*].
Impressions as to the importance of molecular dynamics for
drug action are strongly supported by the inspection of hormones.
2.1. Hormones and Peptide Hormones
O n first inspection it is surprising that small molecules
such as adrenaline, 5-hydroxytryptamine, and acetylcholine
should be used as hormones and transmitters, for these molecules contain only a very small rigid unit-if
they contain
a rigid unit at all-and rather few sites (e.y. -OH or -NH2
groups) by which they can be recognized selectively. The
molecules also contain some bonds about which rotation
is relatively easy. Given the requirement for virtually specific
action of a hormone a much more rigid molecule would
seem to be appropriate. However, if we consider that a molecule has to “wriggle” to reach its site of action then rapid
on/off reactions using a series of different conformations could
be required[’? Thus as above a compromise may be reached
between a good match of hormone and binding surface in
some final state and a mobility required in order to reach
the site. This compromise will include a loss of some binding
capability which is governed by the adverse statistical weighting of conformers in free and bound forms. It is not easy
to work out an optimum system for such fast and selective
binding but, as it may be that the actual binding site is
closely related to the heavily weighted conformation in solution, it will still pay in designing drugs which compete with
hormones or transmitters to try first conformational matching
with these states.
It is remarkable too how many larger hormones are not
like 3,5’-cAMP or sterols, which must have very limited conformational mobilities. Instead many of them are apparently
disordered linear polymers ranging from tripeptides to higher
peptides such as glucagon, which have molecular weights
[**I A change of attitude may now be required on the part of chemists
who arc engaged in the design of drug molecules. The outline “shape” of
a molecule, at 0 K, when examined by instantaneous methods (time constant
< 10- ” s ) u~III conform to bond angle and bond length predictions based
upon our knowledge of simple molecules such as methane. Houever for
a mobile molecule.;mobile receptor interaction it could well be that a rather
different space-filling model would be more appropriate: for example. methane
would become a simple ball and an indole ring would be represented by
a flattened spinning disc. Dynamics in these cases involves rotational mocement about various axes, but no translation. However. translation could
also be important and perhaps we should seek for tunnels and channels
in receptors which may have gates at which “wriggling” motions may be
involved in sitc interactions. The shape of these channels must be such
that given wriggling the uhole molecule can pass through, but such a shape
is clearly not closely related to any single molecular structure.
7 69
of > 10000. These molecules must be flexible to a very high
degree, and in fact this mobility has been confirmed by NMR
measurements. The longer the chain of these hormones the
smaller the probability of a correct matching between the
hormone and receptor assuming that the receptor recognizes
all the hormone. It would seem that fast action should then
belong to small or rigid hormone molecules while slower
action, but perhaps very highly selective control, should belong
to longer flexible molecules, all static and dynamic ways of
achieving selectivity having then been found (Table 3).
Table 3. Naturally occurring rigid and flexible small molecules and hormones
Relatively rigid
Moderately flexible
Trypsin inhibitor
Coenzyme BIZ
Corticotropins (ACTH)
Releasing hormone
Many lipids
Vitamins E and K
Folk acid
[a] I t is assumed that steric hindrance as in thyroxine, ring formation as
in vasopressin, and cross-linking as in the trypsin inhibitor make a part
of a molecule effectively rigid.
Particularly interesting in this respect is the recent work
on insulin, which illustrates conformational mobility of at
least one tenth of the peptide“’! Here it is the cooperative
effect of metal ion and anion binding, at different sites, which
causes a rearrangement of the peptide. Using NMR methods
we have confirmed that a corresponding conformational switch
of insulin occurs in solution with addition of the same reagents[’’]. The changes are anion-, cation- (i. e. salt) and temperature-dependent. This result throws doubt again on the
detailed use of crystal structures in the discussion of the action
of such hormones-except
for the regions of immobilized
(rigid) surface (see the trypsin inhibitor in Section 2.2). The
observations can also be used to describe the synergism of
two small-molecule “drugs”, here X i (ZnZf) and Y - (C1or I-), on a receptor surface (insulin) when it is seen that
there can be large conformational effects in some regions
of the large molecule and that also general changes may
occur to a smaller degree over larger regions of the molecule
(see Section 2.2).
2.2. Small Molecules: Summary
Table 3 lists a variety of molecules found in biological
systems. Apart from peptides and polynucle~tides[~~
12] we
have looked at such molecules as adenosine triphosphate
(ATP)[13] and vitamin B12[’41by the same conformational
methods using paramagnetic probes. In the last two cases
the structures found in solution are again relatively closely
related to those in the solid state and we class these molecules
as largely rigid. Other molecules such as polyamines and
long chain fatty acids must have a much greater mobility
and Table 3 puts them into a different group. Not all control
peptidemolecules of higher molecular weight are of the mobile
kind such as is glucagon. The nature of the trypsin inhibitors
which are quite large peptides but not simple linear polymers
and which must act rapidly, suggest that they are rather
rigid, for their -S-S
cross-links and internal P-pleated sheet
restrict the conformation very greatly[15. l6]. The rigidity of
the trypsin inhibitor is shown too by the fact that even some
of the aromatic residues of amino-acid side chains do not
flip freely despite the fact that this is a small molecule[161.
We stress that the possible combinations of rigid and flexible
elements can lead to all kinds of intermediate behavior and
different parts of large molecules can behave differently.
3. The Receptor Site: Introduction
The approach in the discussion of the structure of the
receptor must be the same as that used in the study of small
molecules. We must go from an examination of constrained
molecules in crystals or free molecules in the solution phase
to a semi-speculative knowledge of a receptor geometry and
dynamics. Now we suspect that receptor sites are particular
regions of proteins, DNA, RNA, membranes or polysacchar-
Met 12
Val 109
Met 105
Leu 8
Fig. 4. Part of the ‘H-NMR spectrum of lysozyme with assignments (methyl groups in the given amino acids).
Angew. Chem. Int. Ed. Engl. 16, 766-777 ( 1 9 7 7 )
ides. What can we say about the structure of such large
molecules in biological systems? First we turn to the examination of the nature of these large molecules in solution.
3.1. Proteins as Flexible Molecules
In the last few years, 1974-77, a change away from rigid
pictures of the conformation of macromolecules has developed.
We pioneered these studies in so far as the use of high resolution
'H-NMR spectroscopy on proteins was concerned["! (Fig.
4), while parallel work by Sykes and by Wiitkvick[161was
underway on peptides, and many authors were carrying out
I3C-NMR studies. In other work a variety of techniques" 8.
have been employed which also demonstrate motion, such
as oxygen quenching of fluorescence and hydrogen exchange,
but this work is restricted to the general discussion of
the protein and it has not been possible to locate the mobility
within the protein sequence. Our work shows clearly the
nature of the motions and how they are distributed within
proteins. The types of motion observed are:
1. Rotation and libration (sometimes restricted) about certain
single bonds internal to the protein.
2. High mobility of protein side-chains on the surface.
3. General mobility of (hydrophobic) groups internal to proteins (breathing).
4. Occasional segmental movements of the main chain.
5. Some proteins are quite generally mobile i. e. close to random coil polymers. Table 4 gives some details of these
Table 4. Protein motion.
Type of Motion
Expansion (Vibration)
Temperature dependence of ring cnrrent
relaxation times
Insulin: transitions
Heme: resonances (hernoglobin)
Lysine: correlation times
Tryptophan: oscillation
Phenylalanine and tyrosine: flip
Temperature dependence of resonance
Relaxation times
Fast segmental motion
Surface motion
Slow motion of side chains
Random coil motions
3.2. Segmental Motion: Summary
In some proteins we must suppose that a part of the protein
may be mobile while another part is not and that these parts
can be of quite diverse sizes for different proteins[20-22! Vallee
et al. have stressed the mobility of the active site region
of carboxypeptidase in solution[231and it would appear that
the effect is seen even in the crystals. This mobility effects
a relatively small back-bone change, comparable with the
type of change which occurs on the binding of oxygen to
hemoglobin when the FG region undergoes perturbation.
In the meantime quite a number of proteins have been examined and they would appear to fall into very different classes
of behavior. Lysozyme, peroxidases, peptidases, carbonic anhydrase and cytochrome c show least conformational change
of the main frame; kinases, concanavalin A, colicin Ia, and
hemoglobin show rather larger changes; histones, chromaAngew. Chrm. Int. Ed. Engl. 16, 766-777 (1977)
granin A, and phosvitin are very adaptable, flexible, chains
(Table 5). Thus it seems that large molecules (receptors) show
the same range of flexibility as small molecules. There must
be some very rigid systems and some very mobile ones and
all intermediate behavior patterns will emerge.
Table 5. Rigid and flexible large molecules [a]
Relatively rigid
Moderately flexible
Cytochrome c
Neurotoxins [b]
Chromogranin A
Myelin protein A. 1 .
DNA-binding proteins
Kinases (?)
Antibodies (?)
[a] Generally, extracellular proteins are rigid and are often crosslinked by
S-S bridges. lntracellular proteins are more mobile especially if they are
required to equilibrate between bound and unbound states involving DNA,
RNA or membrane surfaces.
[b] Personal communication from Prof. C.Petsku.
The case of cytochrome c is of great interestrz1! Although
it is known that the two oxidation states of cytochrome c
have different physical properties e. g. solubility and chromatographic mobility we have been unable to detect differences
in the interior of the molecule in the two oxidation states
in solution. A similar result has been found in crystals. However
we do observe that the binding of various anionic and cationic
reagents on the surface of cytochrome c is different in the
two oxidation states which proves that their surface energies
are changed by the change in charge on the iron. Change
in surface energy will be reflected by change in charge, conformation, and hydration, but present methods do not detect
the surface states readily. Even so it is these surface states
which will be important in protein/protein and at least initially
in drug/protein interactions.
4. Small Molecule/Protein (Drug/Receptor) Combination
This article .has stressed dynamic features of both drugs
and receptors where drugs have been likened to many types
of small molecules and receptors to many types of large
polymers or assemblies. In discussing the combination of drug
with receptor it is the dynamic features which we wish to
keep to the forefront, although we realize that static matching
has a very important and proven role to play. The discussion
will proceed through examples, but before proceeding to these
examples the states of the receptor and small molecule must
be visualized as they approach one another. All the evidence
provided above indicates that the small molecule will be fluctuating very rapidly between at best a small number of similar
conformations and at worst a complete spectrum of very
different states. The receptor surface may be equally mobileespecially in the case of sugars and lysines-but in other regions,
pockets and grooves, may show limited mobility. An agonist
drug might bind and then pass through these mobile regions,
in some cases going to a site much deeper in the receptor
while other, antagonist, molecules may just block access of
the agonist to the deep-seated site. From here on we examine
the bound drug/receptor sites.
4.1. LysozymeSaccharide Reactions
The “receptor site” of lysozyme is the enzyme binding groove
for substrate. It is lined with readily recognizable (by NMR)
tryptophan residues (Fig. 5), and it is these groups which
we can follow most readily during the binding of the saccharide[24,25].
The NMR method is fast enough to follow steps
which occur with time constants of 5 1 0-4 s, and its specificity
allows a precise statement as to which groups move and,
if sufficient care is taken, the movement itself can be defined
with some precision (Fig. 6).
Fig. 5. The lysozyme pocket showing the binding of a trisaccharide (alter
Fig. 6. Aromatic ‘H-NMR spectral region of hen lysozyme (3mmol/l) at
pH 5.3 in the presence of tris(N-acetylglucosamine) (4 mmol/l) showing the
effect of temperature on the tryptophan > N H resonances at a) 3 7 T , b)
45T,c) 55°C.
In the initial state there is present say the disaccharide
bis(N-acetylglucosamine), which is a semi-mobile molecule
but which may well be largely constrained by interaction
with water, so that perhaps 95 % of the molecules belong
to a small family of conformers. The protein lysozyme has
a groove which is undergoing some dynamic motion including
fast motion of side chains such as valines and the slow movement of the tryptophans (see Fig. 10). In other words the
two molecules generate a set of geometries, some compatible
and some incompatible, and on binding they settle down
into the final energetically most favorable bound state, which-as we have shown-has rather less mobility. The overall
binding reaction path is then a succession of conformational
Ef S-r z E S ( l ) + xES(2)+
final state
final conformations)
where the sum sign, X, covers all the dynamic states of a
species. It is the dynamic nature of the initial and intermediate
states of the reaction sequence which allows all the steps
to the final condition to be rapid[241.The observed greater
rigidity in the last step (or stages) means that the reverse
process, the first dissociation step, may have a slow step
as its initial step. Ignoring the consequences of this description
as far as enzyme action is concerned we note here that an
inhibitor of the enzyme could be a molecule which reaches
states CES(1) or ZES(2) but cannot go through to the final
state for reaction, when the inhibitor need not be a very good
match for a substrate so long as it binds. As an almost trivial
example CN- is a very strong poison (drug) when it replaces
O2 in heme proteins. Sulfonamides do not match C 0 2 or
HCOy very well but they bind in the pocket of carbonic anhydrase in place of CO,. Thus the match of enzyme and substrate is not highly restricted and we visualize a series of poor,
moderate and good matches to different states of the enzyme
groove. A very interesting test of this situation has been
observed by accident.
We wished to study the surface of lysozyme using the
spin-labeled substrate 3-[4-(2-deoxy-2-acetylamino-l-glucosyloxy)-phenylcarbamoyl]-2,2,5,5-tetramethyl-3-pyrrolin-l-oxyl
( I )[261. However, we found that the binding was not through
the N-acetylglucosamine moiety, which is a portion of the
natural substrate, but was directly to the spin-label itself which
binds in site C. The fit can not be perfect but the binding
of the spin-label molecule (2) by itself, which was tested
as a result of the above finding, showed that it is bound
better than N-acetylglucosamine ! Thus by accident a potential
inhibitor (drug) has been uncovered which is chemically unlike
the substrate but binds better than the substrate analog. No
examination of the spin-label molecule (2) and the N-acetylglucosamine would have suggested the parallel. Detailed inspeG
tion ofpossible protein surfaces and the geometry of a molecule
such as N-acetylglucosamine or even more so a steroid reveals
how difficult it is for an enzyme to match precisely the spacefilling and bonding capabilities of a substrate or a drug. Restricted mobility of the surface then becomes a very great
advantage in assisting good (not excellent) matching. In fact
I believe the matching has been rather over-stressed, for all
that is required is good, relatively specific, binding and that
the kinetics of this binding should be under control. In both
respects mobility is highly valuable (Fig. 7).
Angew. Chem. I n t . Ed. Engl. 16. 766-777 (1977)
ring is a rigid frame and we have examined such frames
extensively in simple complexes, showing for example how
they form plane-to-plane stacked molecular complexes with
porphyrinscZ9].We then used these methods in the study of
peroxidase complexes‘301.
Peroxidases are very large glycoproteins which have some
obvious mobile features such as the saccharide side chains.
In the native state they are high-spin Fe”’ enzymes. Reaction
with cyanide converts them into the low-spin state, whereas
reaction with azide gives a mixture of the two forms (low-spins high-spin) in equilibrium. On the NM’R time-scale all these
Temperat ure
Fig. 7. Active center of lysoryme [25]. Overall survey of the conformational
states of lysozyme in solution. Each block represents a conformational state
and the linewidths referred to are those of the ‘H-NMR spectrum.
A somewhat more extensive rearrangement is seen on binding saccharides to concanavalin A[”]. Here binding results
in such changes to the protein structure that two complete
crystal structure analyses (protein with sugar and protein
without sugar) were required. No evidence is available about
the dynamics of the protein in this case. Note that here too
a carefully labeled reagent, 1-(o-iodopheny1)-P-D-glucopyranoside, designed to find the sugar binding site, actually became
bound to an abnormal site.
Another example of mobility at a receptor site has been
uncovered by Burgen et ~ l . [ They
~ ~ l observed
that the absorption spectrum of carbonic anhydrase (as the c o b a l t ( ~enzyme)
was changed by the addition of sulfonamides. The change
is consistent with the reaction
Native (largely high-spin)
+ Azide (high-spin)+ Azide (low-spin)
are rapidly in equilibrium so that both the on/off reactions
of azide and the spin-changes are very fast. In fact there
is probably present some 5 % of low-spin state in the native
enzyme, as is the case in myoglobin and hemoglobin. As
shown in Figure 8, high-spin and low-spin Fe”’ heme complexes have rather different structures. We must suppose that
iron(ri1) oscillates rapidly between the different states and
that the protein chain linked directly to the iron through
the proximal histidine also undergoes rapid changes‘”, 311.
(b) Low-spin Fern
Co(r1)(base-) + sulfonamide --t Co(ir)(sulfonamide-)
distorted geometry
tetrahedral geometry
for the 3- and 4-substituted sulfonamides; but the 2-substituted
sulfonamide complex has an absorption spectrum lying
between these extremes and thus indicating an “intermediate”
geometry at the cobalt. There is general agreement that the
metal site in carbonic anhydrases can flick between different coordination states. Although only a small change is apparent
this metalloenzyme is comparable with such enzymes as peroxidase and cytochrome P-450 in that the coordination sphere
of the metal is sensitive to drug binding (see also Table 6).
Tdble 6 Changes in metalloenzymes due t o drug binding
Cytochrome P-450
Low-spin H
g-value changes
Xe 1.3-diphosphoglycerate
Low-spin tt
Low-spin t - i
Distortion of
4.2. Peroxidase-Substrate Complexes
The complexes of peroxidase with substrates can be readily
studied since some of the substrates-indoles and phenolshave NMR spectra which are easy to interpret. The indole
Angew. Chem. I n t . Ed. Engl. 16, 766 777 ( 1 9 7 7 )
Fig. 8. Stereochemistry of the heme iron spin-states
When an aromatic substrate binds to the protein it does
not form a plane-to-plane stacked molecular complex like
in the model reactions with porphyrins. Taking peroxidase
complexes of 3-indolepropionic acid as an example we find
that the indole system lies as is shown in Figure 9I3O1. The
indole derivative is bound only by weak Van der Waals interactions in the pocket of the enzyme. Binding of other substrates
is in the same pocket but the relationship between these
substrates and their binding sites and between the indoles
and their binding sites differ considerably. We conclude that
the site of binding is more like a small mobile oily pool
than a lock and key fit. Thus it is anticipated that a large
number of organic chemicals can act as inhibitors of these
enzymes. Maybe enzymes for detoxification behave in this
rather general way, for similar features can be seen in the
binding sites of cytochrome P-450 and of alcohol dehydrogenases.
The effect of a drug can occur (i) at its immediate site
which may be a long way from the iron, i. e. 1,3-diphosphoglycerate, (ii) at the iron by way of the spin-state change-a
considerable relay, (iii) at the -SH group which is an even
greater distance away. Antagonists could block these processes
while agonists could potentiate them in a quite different way.
Moreover, there are several very different reagents which can
act on the same process, i.e. O2 uptake by hemoglobin, by
binding at totally different sites, e.g. 1,3-diphosphoglycerate,
xenon, HgI:, CO, and sulfydryl reagents. The very different
chemical behavior of these reagents as compared with 02,
the substrate which the drug challenges, shows that drug
design can be quite impossibly difficult. Note too that it
is the mobility of the receptor, here hemoglobin, which is
responsible for all these different “drug” effects.
4.3. Wriggling and Gating
Another illustration of this “wriggling” principle can be
given by examining again the mobility exhibited by the binding
groove of lysozyme (Fig. 10).This groove contains two aminodicarboxylate residues, Glu 35 and Asp 52, of which Glu
Fig. 9. The peroxidase complex of indolepropionic acid showing the disposition of heme and substrate, (a) and (b) are orthogonal views.
Now these oily pools in the proteins for taking up substrates
need not be immediately accessible to the small molecules
and it would be very interesting to observe exactly the time
course of approach to the final state of binding. An inhibitor
or a drug could act at any of the approach stages, or it
could act indirectly by preventing the conversion of one form
of a protein into another. Again it does not appear that
the mobility within the oily drop is much altered by the
insertion of the substrate even though the substrate binding
affects the iron(m) site somewhat.
A good example of the wriggling of a group to reach such
an “oily-drop’’ binding site is the uptake of the hydrophobic
anion HgI; into myoglobin. The binding site as observed
by X-ray crystallography is at the back of the heme. There
are gross restrictions to movement in this region of the protein
and even quite small ligands cannot bind to the iron itself
for this region of space is designed to accept oxygen only.
We have studied the effect of the uptake of Hgl; on the
properties of the Fe(r1r) form of myoglobin and have shown
that there is a small change in the spin-state equilibrium
(high-spin G low-spin) on binding. Thus the binding group
(drug), here HgI;, has to wriggle through the protein to its
binding center where it alters the conformation of the protein
slightly. Uptake of xenon (a very large atom) must take place
by similar wriggling. Again the effects of oxygen on the fluorescence of proteins demand a similar conformational mobility
allowing access by wriggling.
It is of interest to see how such binding could adjust the
chemistry of the receptor. In the P-chains of hemoglobin[321
there is a sulfydryl group F G (p-93) which is on the helix
which carries the Fe-linked histidine (proximal). Perutz has
shown that the degree to which this sulfydryl is exposed
depends upon the spin-state of the iron, for change of spin-state
causes rotation of the helix. Thus we have the following dependencies: a) drug bound-more low-spin Fe, exposed -SH;
b) drug absent -more high-spin Fe, hidden -SH.
Irp 6 2
A s p 101
I r p I08
Ala 107
V a l 109
Fig. 10. The active site region of lysozyme; section of the crystal structure.
A l l the amino acid residues can he detected in the NMR spectrum.
35 is held in the protonated form, against Trp 108. If a cation
is bound in this region of the groove it may interact with
sufficient strength that Glu 35 is ionized, and the resulting
carboxylate group moves away from Trp 108 (configuration
2 in Scheme 1). In the case of weakly interacting cations
this movement does not occur (configuration 1). We can therefore distinguish strongly interacting and weakly interacting
cations. Now let us suppose that exactly the same structural
relationship of two carboxylate groups occurs, not in lysozyme,
but in a region of space behind which there is a second binding
site for the cation. In the case of strong interaction (configuration 2) the actual binding at the two carboxylate centers
would prevent (gate) access to this site. In the case of weak
interaction (configuration 1) binding at the rear site would
occur as the gate would remain open. Clearly the cations
could act in very different ways. Gated reactions will always
be associated with wriggling paths. Of course the same pattern
can be imagined for the action of a cation such as acetylcholine.
Angew. Chem. Int. Ed. Engl. 16,766777 (1977)
T r p 108
T r p 108
. . .CH2-COZ-H
. . .CH,
Glu 3 5
Glu 35
A s p 52
A s p 52
Configuration 1
Configuration 2
Scheme 1
Another example concerns the passage of an ion, e.g. Li+,
through a membrane using an ionophore. There is a binding
on/off reaction of metal and ionophore followed by a release.
There is also an outward directed ion pump and we may
suppose a site of action for Li’ say in competition with
C a 2 + or Mg2+. It must be seen that the ionophore-governed
steps are very important in the overall effect of lithium.
As we have stated above the mobility of the surfaces of
proteins is much greater than that of the internal regions.
Our measurements show that surface lysine and sugar side
chains (in glycoproteins) have motions with correlation times
of I 0-9 s, i. e. as fast as collision diffusion rates. In an assembly
of proteins there must be a pathway between the protein
molecules which is diffusion restricted by the constraints
imposed by the packing of the mobile side chains. The general
point can be visualized from an examination of packing of
proteins in crystals. Examination of these crystals shows that
between protein molecules there are large channels of “disordered” solvent and protein side chains. Inhibitors and other
quite large groups such as heavy atom labels diffuse down
these channels quite readily to binding sites. But quite clearly
there are restrictions on this diffusion which can be of almost
any degree of selectivity. Thus only certain molecules will
be able to reach particular regions of proteins in an assembly.
It is possible to describe this as a “gating” of (drug) action,
and the actual motion as “wriggling” to a site. The importance
of this mobility of proteins extends to the binding reaction
when an antibody combines with an antigen. Huber et u ~ . [ ‘ ~ I
have suggested that the Fc fragment ofthe antibody is relatively
mobile so that it changes conformation on receiving the
4.4. More Complicated Receptors
A protein or a group of proteins may or may not be the
receptor site of a drug. Rather a drug may act on a similar
type of large molecule, e.g. RNA or DNA, or it could act
at a membrane or similar condensed phase. The effect of
small molecules on a condensed phase should now be examined
in the same way as we have examined their effect on large
molecules. The parallels between these problems has been
described in a recent review on phases in biological
We use lipid phases, membranes, as an example.
4.5. Membranes
Fast tumbling of groups such as we have described in
proteins is observed in artificial vesicle membranes of most
lipids. For example lecithin vesicles give a high resolution
Aiigrw. C h ~ w
Inf. Ed. Engl. 16, 7 6 6 7 7 7 ( 1 9 7 7 )
NMR spectrum which shows that motion must be fast for
the -N(CH,);
group on the head and the methyl group
on the tail of the fatty acid chains but is somewhat slower
for the region around the glycerol group. When the size of
the vesicle is changed or other molecules incorporated into
it, or the temperature changed, or ions bound to the surface,
these mobilities can undergo quite considerable changes.
Turning to biological membranes we and others have
observed that the ‘H-NMR spectrum of almost all cell membranes is very broad, even for membranes which contain
a high percentage of lecithin, and it is not possible to detect
in the NMR spectrum either the choline group on the head
or the methyl group on the tail of the fatty acid chains.
It is readily shown that inclusion of such molecules as cholesterol in the bilipid layer will cause a general reduction of
the motion of the lipid sections of the molecule but the headgroup remains free. We are therefore left with the problem
that there are gross restrictions on mobility in real biological
membranes which are not seen in vesicles. This restriction
could be related to size, curvature of membrane surface, or
to the binding of the head-groups.
The membrane lipids are also in rapid lateral motion, i. e.
in the plane of the membrane. Inversion of lipids is however
relatively slow; hence the membrane retains its asymmetry.
The time scale for lateral movement is not vastly different
from diffusion and it has frequently been supposed that drugs
can influence this motion. The inversion has a time constant
in the range of minutes. Thus although inversion is not important in any steady-state change and return, which takes place
within seconds, the steady-state itself can change in response
to ambient conditions. We proposed some years ago that it
would be interesting to study this asymmetry with the paramagnetic binding probes, but the first actual study was made
351. A major problem-as
in the
independently in
study of proteins-was (and still is) the question of the structure and the structural dynamics of the membrane and what
chemical factors control the dynamics. To this end we looked
at the binding of probe ions to a variety of phosphate esters
including phosphatidylcholines.
The first feature of binding is that the metal ions do not
bind to the phosphate of a membrane phosphatidylcholine
in the same way as they bind to phosphate esters such as
phosphoglycerolcholine in water. If we use our standard procedures for the study of biological molecules we obtain the
structure shown in Figure 11, in which the choline head-group
stretches away from the bilipid membrane to a large degree.
The result need not be analyzed in great detail, for there
is surely much motion, but it is important to observe that
the head-group of a lipid in a membrane does not behave
in the same way as the dissociated free head-group, say phosphoglycerolcholine, behaves in aqueous solution. Secondly,
the binding strength of cations at the two membrane surfaces
(inside and outside) is not equal. Thirdly, as the cations bind
selectively to different phospholipids they generate asymmetry
in the membrane if their concentrations are different on the
two sides of the
The effect of the metal binding
on the mobility of the membrane has not yet been assessed.
However, it will be recognized immediately that the action
of a chemical (drug) on such a system is not likely to be
dependent upon any one simple factor. Thus as with protein
receptors we anticipate that an extremely complicated story
drug action, for this description covers the whole of solid
state and solution chemistry with its wide variety of interactions which are expressed quantitatively only in terms of
free energy changes. Recently we have been using NMR to
examine whole organs and have found a remarkable variety
of mobilities in different phases[371.However, these features
of cells cannot be discussed at length here; but in conclusion
it should be pointed out that we must contend with many
new problems when dealing with added chemicals (drugs)
which can affect mobility, and static solid state models of
fitting the geometry and chemistry of one surface to another
may have to be supplanted by the more nebulous concepts
of dynamic (solution) states
The work described in this article was made possible by
grants from the Science and Medical Research Councils, England. The author wishes to acknowledge the great help he has
received from fellow members of The Oxford Eizqme Group
and from his students.
Fig. 1 1 . The structure of phosphatidylcholine. Note that this structure
has considerable mobility, and is for a cation-bound condition.
lies behind the drug action. Though it may be possible to
follow drug-protein binding atom by atom, this possibility
does not really exist in a membrane phase and we may have
to be content with a more phenomonological functional thermodynamic approach. This would mean that drug design
would remain very empirical. For example, fluidity in a cooperative liquid is open to a phase change of the solpgel kind
and it could be that a small quantity of an impurity could
shift the sol e gel equilibrium and thus change the dynamics
of the phase dramatically.
4.6. Mobility of Bound States
The conclusion seems inevitable that just as all degrees
ofmolecular motion are to be found in drugs (small molecules)
and in receptors (proteins, RNA, membranes) so in their combination we shall find all kinds of mobilities (Fig. 12), for
the binding site of a drug can vary from a condition in which
the best model is a molecule held on a rigid crystalline material
to one in which the drug has partitioned from water into
an organic medium. There cannot be a general theory of
Fig. 12. Schematic diagram for a mobile protein (linked by ari-ows)-mobile
drug interaction.
Received: November 8, 1976 [Z 187 IE]
German version: Angew. Chem. 89, 805 (1977)
!I Burgen, G. C. K . Roherrs, J. Feeriey, Nature 253. 753 (1975):
and references cited therein.
[2] C . M . Dohson. L. 0 . Ford, S . E. Summers. R. J . P. Williums. J. Chem.
Soc. Faraday Trans. 11 71, 1 I45 (1975).
131 B. A . Lrrine, R. J . P. Willirm~s,Proc. Roy. Soc. (London) 345, 5 (1975):
B. A . L e i w e , R . J . P. Williums in B. Pullman: Environmental Effects
on Molecular Structure and Properties. Reidel, Dordrecht 1976, p.
95 rr.
[4] See r. g. E. Bergmanri, B. Pullmuif: Molecular and Quantum Pharmacology. Reidel. Dordrecht 1974.
[5] C . D. Burry, J. A. Glo.sel. R. J . P. Williams, A. K Xur-ier. J . Mol.
Biol. 84, 471 (1974).
J. Am. Chem. Soc. 95. 2333 (1973).
[6] C . .4lronu, M. Sutidurulin~pn~,
171 B. Birrlsull, N. J . M . Birdsull, J. Feeney. J. Thornrun, J. Am. Chem.
Soc. Y7, 2845 (1975).
[ 8 ] C . D. Burr)., D. R. Murrin, R. J . P. Williums, 4 . K Xrriirr. J. Mol.
Biol. 84, 49 1 ( I 974).
[9] C. Geruldes, D. Phil. Thesis, Oxford 1976.
[lo] G. B e d r y , E. D o d m l . G. Dodson, D. Hodgkin. D. Merc~olo.Nature
-261, 166 (1976).
. J . P. William,$.K . Williumson. unpublished.
J. A . Gfascl, A. C. 7: Norrh, R. J. P. Williums. A. K
Xorirr. Blochim. Biophys. Acta 262, 101 (1972).
[I31 P. Trunswll, J. M . Thornron, A. T.: Kordu. R . J. P. U’illiani.~, Eur.
J . Biochem. 57, 135 (1975).
[I41 0. D. Hensens, H . A. 0. Hill, J. Thorriron, R. J. P. Willianis, Phil.
Trans. Roy. Soc. (London) 8 2 7 3 , 353 (1976).
11 51 R . H u h , D. Kirkla, M! Sriegrfnanri, J . Drirenhqfer, A . Jone.s: Bayer
Symposium V. Proteinase Inhibitors. Springer, Berlin 1973.
1161 K . Ct’iirhrich. G. Wugner, FEBS Lett. 50. 265 (1975).
[I71 1. D. Curuphell. C. M . Dohaon. R J . P. Wdliums, Proc. Roy. Soc. (London)
345. 23. 41 (1975).
[ I S ] K . Lindrrafrum-Lung, J. A . S~hrllmunriin P. D. Boyrr, H. Lorfly, K .
M ) r h u c k . The Enzymes. Vol. 1. Academic Press. Nor York 1959. p.
443 ff.
[Is] J . R . Lahowirz, G. Weher, Biochemistry 12. 4161 (1973)
[20] 4 . Care. C . M. Dohson, J . Porrllo, R. J . P. Williums, FEBS Lett.
65. 190 ( 1976).
[?1] G. R. M o o r e . R . J . P. Willrcrms, Coord. Chem. Re>. 18. 125 (1956).
[ZZ] K. J. P. Williams, Cold Spring Harbor Symp. Quant. Biol. 36. 53
[23] J . T Johunseri. B. L . W i l l e e ~Biochemistry 14, 649 11975)
[24] C. M. Dohsoii. R. J. P. CtElliun~s.FEBS Lett. 56. 362 (1975).
[25] 7: Inioro. L. N . Johnsor~. 4 C. T N o r t h , D. C. Pkiiiips. J. -1. Ruplr!,
in P D. B o w : The Enrymcs. Vol. 7. Academic Press, New York
1972, p. 665.
[26] J . P o ~ ~ l s e C.
n . M . Dohsori. R J . P. Williunis, unpublished.
1271 J. IV U e i ~ h r r G.
, N . Reeke. U. 4. Cuiininghurn, G. ,At. Erlchnun, Nature
259. 407 (1976).
[2X] R. M! King. A . S. K Bnrgt~n.Proc. Roy. SOC. (London) B193, 107
A. S.
Aiigew Chem. l n t . Ed. Engl. 16, 766-777 ( 1 9 7 7 )
1291 C. D . B u r r ] , H . A. 0.Hill,P. J . Sudler, R. J . P. Williams, Proc. Roy.
SOC.(London) A 334,493 (1973).
[30] P. S . Burns. R. J . P. Williums,P. E. Wright,J. Chem. SOC.Chem. Commun.
1975. 795.
[31] This problem has been studied in much detail recently by Prof. H .
Frauevifelder (Hamburg Int. Biochemistry Meeting 1976).
[32] M . F . Perirrz, L. F . Teneyck, Cold Spring Harbor Symp. Quant. Biol.
36. 295 (1971).
R. J . P. Williams, Biochim. Biophys. Acta 416, 237 (1975).
See H . Hauser, M . C . Phillipx. B. A. Leririr, R. J . P. Williums, Nature
261, 390 (1976).
1351 L. I . Bursirkor. Y. E. Sliayiro, A. V Viktoroc, A . F . Bystror. A.
Bergelson, Akad. Nauk USSR 208, 717 (1973).
1361 R. J . P. Williums, Physiol. Chem. Phys. 4, 427 (1972).
1371 A . Daniels. R. J . P. Williorm, P. E . Wrighr,Nature 261, 321 (1976).
C 0 M M U N I C AT1 0N S
Racemic proline ( 4 )
Simple Synthesis of Racemic Proline
By Ulrich Schmidt and Huns Poise/[*]
Dedicated to Professor Engelbert Broda on the occasion of
his 65th birthday
Racemic proline has so far only been accessible by time-consuming multi-step syntheses"! We have now found a simple
synthetic pathway starting from pyrrolidine:
N-Chlorination, preferably with tert-butyl hypochlorite,
leads to N-chloropyrrolidine ( 1 ) which is not isolated but
instead reacted directly with sodium methoxide to give l-pyrroline (2). This product trimerizes even in solution within
a few hours and cannot be isolated[']. However, we found it
to be sufficiently stable in dilute solution t o add reactive
compounds. Hydrogen cyanide adds smoothly to 1-pyrroline
to form the nitrile (3) whose hydrolysis affords racemic proline
( 4 ) in about 457, yield (based on pyrrolidine).
Hydrogen chloride must be eliminated from the N-chloropyrrolidine ( I ) prior to addition of hydrogen cyanide; direct
reaction of ( 1 ) with cyanide in alkaline solution yields N-pyrrolidinecarbonitrile. For large scale preparations it is recommended that chlorination of the pyrrolidine be carried out
in the two-phase system: ether (or toluene)/aqueous hypochlorite solution
[*] Prof. Dr. U. Schmidt and Dr. H. Poise1
Organisch-Chemisches Institut der Universitit
A-1090 Wien, Wiihringerstrazse 38 (Austria)
Pyrrolidine (7.1 g, 0.10 mol) and tert-butyl hypochlorite
(12.0 g, 0.11 mol) are simultaneously added dropwise to stirred,
ice-cooled ether (100 ml). After 5 min the colorless solution
is extracted once with dilute HCI and twice with a small
volume of water. The mixture is dried with sodium sulfate,
treated with sodium (2.76g, 0.12 mol) in methanol (70ml) and
refluxed for 25 min, whereupon NaCl is precipitated. After
cooling, a solution of anhydrous hydrogen cyanide (10 ml,
0.26mol) in ether (10ml) is added and reaction allowed to
continue overnight at room temperature. The solvent is
removed in a rotary evaporator and the residue dissolved
in methylene chloride, washed twice with water, and dried
with sodium sulfate. Vacuum distillation yields 5.6g (58 %)
of ( 3 ) , b.p. 77"C/12 torr.-Compound (3) (5.5g) is heated
with 19%aqueoushydrochloricacid(t00ml)for l 4 h a t 100°C
in a bomb tube. After evaporation of the acid the residue
is dissolved in water and decolorized with animal charcoal.
After concentration the product is dried over sulfuric acid.
Esterification of the salt mixture by Fischer's method with
HCl/ethanol affords 6.6g (81 %) of D,L-proline ethyl ester
which is allowed to stand overnight with 20ml of water.
Removal of solvent leaves 5.0g [76 'i:based on ( 3 ) ] of D.L-prOline ( 4 ) .
Received: September 22, 1975 [ Z 316 IE]
German version: Angew. Chem. 89. $ 2 1 ( 1 977)
Publication delayed at authors' request
CAS Registry numbers:
( I ) , 19733-68-7; (2). 5724-81-2; (3). 57015-08-4: ( 4 ) . 609-36-9
See, e.g., K . Hasse and A. Wirlund. Chem. Ber. 93. 1686 11960); K .
H . Buecliel and F . Korte, ihid. 95, 2453 (1962); R A . Srrojiii.. H . C .
Whitr, and E . Sirojnj, J. Org. Chem. 27, 1241 (1962).
[2] D. W Fulh~igeand C . A . liilii iler We~rf,J. Am. Chem. Soc. XO. 6249 (1958).
Photochemical [2 + 21 Cycloaddition between Parallel
CC and NN Double Bonds[**]
By Wilfr.ied Berning and Siegfried Hiinig[*]
Rigid polycyclic compounds containing two parallel C=C
bonds are well known and their chemistry, particularly photochemical [2 +2] cycloaddition to cage compounds, has been
[*] Prof Dr. S. Hunig. Dr. W. Berning
Institut fur Organische Chemie der Universitht
Am Hubland. D-8700 Wiiriburg (Germany)
[**I From the dissertation by W Beriii!ig. Universitiit Wiirrburg 1977. This
work was supported by the Fonds der Chemischeii Industrie and the BASF
AG. Ludwigshafen.
Без категории
Размер файла
1 265 Кб
drug, flexible, molecules, receptors, dynamics
Пожаловаться на содержимое документа