Individual assignments of amide proton resonances in the proton NMR spectrum of the basic pancreatic trypsin inhibitor

177

Biochimica et Biophysica Acta, 577 (1979) 177--194 © Elsevier/North-Holland Biomedical Press

BBA 38125

INDIVIDUAL ASSIGNMENTS OF AMIDE PROTON RESONANCES IN THE P R O T O N NMR SPECTRUM OF THE BASIC PANCREATIC TRYPSIN INHIBITOR

ANDREAS DUBS, GERHARD WAGNER and KURT WUTHRICH

Institut fiir Molekularbiologie und Biophysik, Eidgen6ssische Technische Hochschule, 8093 Ziirich-HSnggerberg (Switzerland) (Received July 24th, 1978)

Key words: NMR, Nuclear Overhauser effect; Trypsin inhibitor; Amide proton Conformation

Summary Studies of proton-proton nuclear Overhauser effects were used to obtain individual assignments of 17 amide proton resonances in the 360 MHz proton nuclear magnetic resonance spectrum of the basic pancreatic trypsin inhibitor. First, optimizing the conditions for obtaining selective nuclear Overhauser effects in the presence of spin diffusion in macromolecules is discussed. Truncated driven nuclear Overhauser experiments were used to assing the amide p'roton resonances of the /3-sheet in the inhibitor. It is suggested that these techniques could serve quite generally to obtain individual resonance assignments in fl-sheet secondary structures of proteins. Combination of nuclear Overhauser studies with spin decoupling further resulted in individual assignments of the 7-methyl resonances of the two isoleucines and numerous Ca and C ~ protons.

Introduction

The basic pancreatic trypsin inhibitor is a small globular protein with a highly refined single crystal X-ray structure [1], which has recently been much used as a model c o m p o u n d for theoretical and experimental studies of fundamental aspects of protein conformation [2--8]. Particular emphasis was on the investigation of internal mobility in proteins [2--4,9--12]. For experimental studies in this field, high resolution NMR spectroscopy is a powerful m e t h o d [13]. In the inhibitor, important structural information has come from IH NMR studies of the labile protons (Refs. 14--21, and Wiithrich and coworkers,

178 unpublished results), whereby the analysis of the experimental data was greatly aided by the identification of the resonances. The present paper describes the techniques used to obtain individual assignments of numerous amide proton lines. Early experiments showed that the 1H NMR spectrum of the inhibitor contains numerous resolved resonance lines which correspond to labile protons [22,23]. Previously, individual assignments for five of these resonances were obtained with the use of lanthanide shift reagents in studies of chemical modifications of the inhibitor [24,25]. Here, individual assignments for these five and twelve additional lines were independently obtained from studies of protonproton nuclear Overhauser effects [26,27]. Much care was exercised in the selection of the experimental conditions, so that specific Overhauser effects were obtained in spite of spin diffusion [28]. Technical details of these measurements are described in Materials and Methods. Interpretations of individual experiments and the strategy used to obtain individual assignments for numerous polypeptide backbone proton resonances in the inhibitor are presented in the Results section. Materials and Methods The basic pancreatic trypsin inhibitor (Trasylol ®, Bayer Leverkusen, F.R.G.) was obtained from the Farbenfabriken Bayer AG. For the NMR studies, the lyophilized protein was dissolved either in 2H20 or in H20. In some experiments, praseodymium(III) was added as a NMR shift reagent as described previously [29]. While 0.005 M inhibitor solutions were used for all the other experiments, nuclear Overhauser difference spectra were obtained with 0.02 M solutions in order to obtain a satisfactory signal-to-noise ratio within reasonable periods of time. ~H NMR spectra were recorded on a Bruker HXS-360 spectrometer. Sample tubes with 10 mm outer diameter were used for measurements of nuclear Overhauser effects, standard 5-mm tubes for all the other experiments. Nuclear Overhauser difference spectra [30] were obtained by subtracting spectra with Overhauser effects from reference spectra. Spin decoupled Overhauser difference spectra [30] were obtained with the application of a selective decoupling field during data acquisition. The free induction decays with and without Overhauser effects were accumulated alternately in order to minimize instrumental drifts. Accumulation times of approximately 90 min were used to obtain difference spectra with a satisfactory signal-to-noise ratio. Selective proton-proton Overhauser effects in the presence of spin diffusion. The nuclear Overhauser effect is the fractional change in intensity of one NMR line when another resonance is irradiated. It has long been a valuable tool for structural studies of small molecules, where the conventional 'steady-state' Overhauser effects are simply related to the distance between irradiated and observed nuclei [26,27]. When working with macromolecules at high frequencies, however, spin diffusion is of considerable importance [28,31,32], which makes steady-state Overhauser effects less specific. For the basic pancreatic trypsin inhibitor, this is born o u t by the data in Fig. 1. It is seen that the steady-state Overhauser difference spectrum obtained with selective presatura-

179

°

B

Fig. 1. 3 6 0 M H z 1 H N M R s p e c t r a o f t h e i n h i b i t o r r e c o r d e d in 0 . 0 2 M s o l u t i o n in 2 H 2 0 in 1 0 - m m s a m p l e t u b e s , p 2 H 4 . 5 , T = 1 5 ° C . ( A ) N o r m a l F o u r i e r t r a n s f o r m s p e c t r u m . (b) S t e a d y - s t a t e n u c l e a r O v e r h a u s e r d i f f e r e n c e s p e c t r u m o b t a i n e d w i t h p r e s a t u r a t i o n o f t h e a m i d e p r o t o n line a t 1 0 . 6 p p m , as i n d i c a t e d b y t h e a r r o w . • a n d ~, i d e n t i f y t w o a m i d e p r o t o n lines a n d a • a n d • , t h e t h r e e c~-proton lines w h i c h will b e further discussed below.

tion of a well resolved one-proton resonance contains numerous lines in different spectral regions, so t h a t a n unambiguous identification of protons located near the irradiated amide proton is n o t possible. Theory shows t h a t in contrast to the steady-state Overhauser effects, the initial build-up rates of nuclear Overhauser effects are simply related to the inverse sixth power of the distance between the observed and the presaturated p r o t o n even in the presence of spin diffusion [26--28]. It is hence of great practical interest to develop techniques where upon selective irradiation of individual resonance lines, spectral features can be observed which are in a .simple way related to the initial build-up rates of the Overhauser effects. In the following two fundamentally different experiments are distinguished. In studies of 'transient nuclear Overhauser effects', no radio frequency field is applied during the build-up of the Overhauser effects. 'Radio frequency-driven nuclear Overhauser effects' or simply 'driven Overhauser effects' result when a radio frequency field is applied during the build-up process. Transient nuclear Overhauser effects. Transient Overhauser difference spectra were obtained with the pulse sequence (1): (--180°(¢OA)

--

Tl --

Observation Pulse -- v: -- 180°(¢Ooff.res.) -- r,

-- Observation Pulse -- r: --)n

(1)

The experiment is initiated by a selective 180 ° pulse of short duration, typically 10 ms, at the resonance frequency of a spin A. This is followed, after a delay time rl during which the nuclear Overhauser effects are built up in the absence of a radio frequency field, by an observation pulse. After the data acquisition, the spin system is allowed to recover during a waiting period V2, where v2 was typically of the order of 2 s in the experiments with the inhibitor.

180

The second half of the pulse sequence is identical, except that the 180 ° pulse at ¢Oo~-res. is applied in an e m p t y region of the spectrum, so that no Overhauser effects are produced. The t w o free induction decays obtained in one cycle are stored in two different sections of the c o m p u t e r memory, and after accumulation of n cycles they are subtracted to get Overhauser difference spectra. When the experiment (1) is repeated with different delay times T1, the time course of the magnetizations of the irradiated nucleus A and nuclei located near A in the protein structure can be recorded {Fig. 2) [32]. Fig. 2 illustrates that in the sequence of transient Overhauser difference spectra recorded at different times v, after the 180 ° pulse, the intensity of the pulsed line decreases, while the intensities of other lines build up by spin diffusion. If the observation had been continued for sufficiently long times T,, the pulsed line would have further decreased to zero, while the other lines would first have passed through an intensity maximum before decaying to zero by spin relaxation [32]. Comparison of Figs. 1 and 2 clearly illustrates the improved selectivity of the transient experiment. Between 5 and 10 ppm only the three lines marked with ~, <> and • appear in Fig. 2. Since the initial buildup rates of the resonance intensities are simply related to the inverse sixth p o w e r of the distance between the pulsed proton and the observed proton [26,27], transient Overhauser experiments reliably manifest nearest-neighbor relations between individual protons in proteins [32]. While transient Overhauser difference spectra are a particularly straight-forward technique for measurements of initial build-up rates, the continued practical applications with proteins showed that a reasonable compromise between high selectivity of the presaturation pulse and workable signal-tonoise ratio was difficult to obtain when working in crowded spectral regions. Thus, except for the experiment of Fig. 2, where the presaturation pulse was applied to the well separated lowest field amide proton resonance, transient Overhauser experiments were only of limited use for the present investigation, and more information was obtained from truncated driven Overhauser experiments.

I !

,'

IOPPM

o

8

6

4

2

0

F i g . 2, T r a n s i e n t n u c l e a r O v e r h a u s e r d i f f e r e n c e s p e c t r a a t 3 6 0 M H z o f t h e i n h i b i t o r s o l u t i o n in F i g . 1, obtained after a 10 ms inversion pulse was applied to the amide proton line at 10.6 ppm (arrow). The d e l a y t i m e r I ( E q n . 1) is i n d i c a t e d o n t h e r i g h t . T h e s a m e r e s o n a n c e s are l a b e l l e d b y O, 0, • a n d • as in F i g . 1,

181

Driven nuclear Overhauser effects. The experiment (2) was used to obtain driven Overhauser difference spectra. (--tl (~A) -- Observation Pulse -- T2 -- t~(COoff.res.) -- Observation Pulse -- T2 --),

(2) The nuclear Overhauser effects are built up during the period of time t~, while a selective low power radio frequency field is applied to an individual resonance A. This is followed immediately by the observation pulse. After data acquisition the system is allowed to recover during a waiting time T2, where T2 was typically of the order of 2 s in the experiments with the inhibitor. If t~ is sufficiently long, the driven nuclear Overhauser experiment provides a steady-state Overhauser effect [27]. It was found that with sufficiently low power for obtaining selective irradiation of individual lines, steady-state Overhauser effects for the inhibitor resulted with t~ >~2 s (Fig. 1). On the other hand, highly selective Overhauser difference spectra are obtained when the driven Overhauser effects are 'truncated' after a short time t~. From a series of truncated driven Overhauser difference spectra recorded with different preirradiation times tl, the build-up rates of the Overhauser effects for individual nuclei are obtained (Fig. 3). Fig. 3 shows that the truncated driven Overhauser difference spectrum recorded with a pulse length of 150 ms was essentially identical to the transient Overhauser difference spectrum obtained after a delay time of 70 ms (Fig. 2). A more reliable basis for the analysis of driven nuclear Overhauser effects was obtained by comparison of the experimental data with model calculations

"IOPPM 8

6

4

2

o

F i g . 3. D r i v e n O v e r h a u s e r d i f f e r e n c e s p e c t r a o f t h e i n h i b i t o r s o l u t i o n in Fig. 1 o b t a i n e d w i t h i r r a d i a t i o n o f t h e a m i d e p r o t o n r e s o n a n c e a t 1 0 . 6 p p m ( a r r o w ) . T h e i r r a d i a t i o n t i m e t I ( E q n . 2) is i n d i c a t e d o n t h e r i g h t . T h e s a m e r e s o n a n c e s are l a b e l l e d b y $ , •, A • a n d • as in F i g . 1.

182

for a three-spin system [33]. In Fig. 4 the experimental and calculated Overhauser effects are shown for the s-protons of Phe 22, Tyr 23 and Cys 30 in an experiment where the irradiation frequency COA(Eqn. 2) was on the amide proton of Tyr 23. The interatomic distances used in this calculation were taken from the refined crystal structure [1]. The algebraic form of the function plotted in Fig. 4A is given in the Appendix. Fig. 4A shows that the computed relative build-up rates are 1.0, 0.27 and 0.04 for the o-protons of Phe 22, Tyr 23 and Cys 30, respectively, which are at distances of 2.3, 2.6 and 4.6/~ from the irradiated proton. The agreement with the corresponding experimental data in Fig. 4B, which gave relative build-up rates for the three protons of 1.0, 0.38 and 0.07, is quite satisfactory.

NOE

A

C°tH F22

.8

.t, .2

i

NOE

B

.3

C°tH F22

.2

0

~ C ~ H

.1

.2

.3

.4

y2e'-'-

sec

F i g . 4 . P l o t o f t h e r e s o n a n c e i n t e n s i t i e s o f t h e s - p r o t o n s o f P h e 22, T y r 2 3 a n d C y s 30 v s . t I in t h e d r i v e n O v e r h a u s e r e x p e r i m e n t o f Fig. 3, w h e r e t h e i r r a d i a t i o n f r e q u e n c y ~ A w a s o n t h e a m i d e p r o t o n o f T y r 23. ( A ) C a l c u l a t e d w i t h t h e t h r e e s p i n a p p r o x i m a t i o n , w h e r e t h e n e a r e s t - b y v i c i n a l ~ - p r o t o n w a s t a k e n as t h e third proton ( ). T h e i n t e r a t o m i c d i s t a n c e s u s e d in t h e c a l c u l a t i o n w e r e t a k e n f r o m t h e X - r a y s t r u c ture [1]. - . . . . . , the result of a calculation for the s - p r o t o n of Phe 22, w h e r e besides the ~-methylene p r o t o n t h e i n f l u e n c e o f five a d d i t i o n a l p r o t o n s w a s c o n s i d e r e d (see t e x t ) . (B) E x p e r i m e n t a l l y o b s e r v e d d r i v e n O v e r h a u s e r e f f e c t s f r o m F i g . 3. N o t e t h a t t h e scale e x t e n d e s o n l y o v e r 0 . 5 s, as c o m p a r e d to 5 s in ( A ) . T h e n o n o l i n e a r i t y o f t h e i n i t i a l t i m e c o u r s e o f t h e r e s o n a n c e i n t e n s i t i e s is a c o n s e q u e n c e o f t h e n o n - i n s t a n t a n e o u s s a t u r a t i o n o f t h e i r r a d i a t e d p r o t o n , w h i c h w a s also c o n s i d e r e d in t h e t h e o r e t i c a l calculation.

183 The description of the situation in a protein by a three-spin system is of course a rather crude approximation, in which the influence of the bulk of the protons is accounted for by the location in space of a single proton relative to the irradiated proton and the observed proton. Yet, since the initial build-up rate of the Overhauser effect on the observed proton depends only on the inverse sixth power of the distance from the irradiated proton and is essentially unaffected by the presence of additional nuclei [26,28,32,33], this simple model provides an adequate description of the initial stages of a driven Overhauser experiment. On the other hand it cannot account for the observed steady-state Overhauser effects, which are largely affected by the presence of additional protons. This is readily seen from Fig. 4. The solid curves in Fig. 4A were computed for a three-spin system consisting of the irradiated amide proton of Tyr 23, the observed a-proton and the nearest-by vicinal fi-methylene proton. They largely overestimate the steady-state Overhauser effects. If the influence of the bulk protons in the protein is accounted for more realistically by inclusion of the effects of additional neighboring protons on the relaxation time of the observed nucleus, which enters as a parameter into the three-spin model calculation (see Appendix), a more realistic value for the steady-state Overhauser effects is obtained (Fig. 4a, dashed line). Results

The resonance assignments described in this paper are compiled in Table I. These results depended critically on certain properties of the amide protons in the protein, certain features of the spatial polypeptide structure and the combined use of proton-proton nuclear Overhauser and spin decoupling experiments. These general aspects are in the following briefly considered, followed by the description of individual resonance assignments in the fi-sheet and s-helix secondary structures contained in the inhibitor [1]. Amide protons in the basic pancreatic trypsin inhibitor In freshly prepared solutions of the inhibitor in 2H20 , numerous 'H NMR lines of relatively slowly exchanging labile protons are observed between 6.5 and 11.0 ppm [22,23]. From comparison with model peptides [34] these resonances must correspond to amide protons (Refs. 22, 23 and 35, and Wiithrich and coworkers, unpublished results). In all, the inhibitor contains 53 backbone amide protons and eight amide protons of asparagine and glutamine side chains. Depending on the location in the protein structure, largely different exchange rates with the solvent prevail for the individual protons; quite generally, protons on the protein surface exchange too rapidly to be seen in 2H20 solution of the protein. As is discussed in detail elsewhere (Ref. 35, and Richarz, R., Sehr, P., Wagner, G. and Wiithrich, K., unpublished results), the amide proton exchange rates are further strongly dependent on pH and temperature. At the pH minimum and ambient temperature, 33 amide proton resonances were identified in a freshly prepared 2H20 solution of the inhibitor. These have been numbered in the order of decreasing chemical shifts (Refs. 19 and 35). Under the experimental conditions of Fig. 5, twelve of these protons

184 TABLE I P R O T O N S P I N S Y S T E M S IN T H E B A S I C P A N C R E A T I C PROTONS WERE INDIVIDUALLY ASSIGNED

TRYPSIN

INHIBITOR

IN W H I C H A M I D E

T e m p e r a t u r e : 3 6 ° C , p 2 H 4 . 5 . Spin s y s t e m s w h e r e the a m i d e p r o t o n c o u l d he o b s e r v e d in 2 H 2 0 are n u m bered a c c o r d i n g t o Fig. 5, t h o s e w h e r e t h e y w e r e , b e c a u s e o f fast e x c h a n g e , o b s e r v e d o n l y in H 2 0 are i n d i c a t e d b y a star. C h e m i c a l shifts (5) are in p p m relative to internal s o d i u m - 2 , 2 , 3 , 3 - t e t r a d e u t e r o - 3 - t r i m e t h y l s i l y l - p r o p i o n a t e . Spin-spin c o u p l i n g c o n s t a n t s (J) are in Hz.

5 (NH)

6 (C~XH)

1

10.55

4.31

2

9.94

5.12

3

9.79

5.25

4 5

9.37 9.39

4.89 4.89

6

9.18

5.70

7 9 10 *

8.77 8.61 8.58 8.40

4.87

11 15 16 18 19 22 24

8,39 8,25 8.12 8.07 7.98 7.78 7.78

4,70 4,31 4.25 5.28

Spin

6 (CJfH)

3JHNc~

3Jail

Assignment *

system ( 2.72 ~3.45 2.79 ~2.87 L2.81 2.69 2.52 2.69

7,0 11.0 9.0

T y r 23 { 13.0 2.5 { 3.0 3.0

7.0 10,5 9.5

4.41

2.67 3.69 1.62

6.5 6.5

1.86 4,00

f 3,5 ~14,0 9.0 8.5 9.0

2.69

Phe 2 2 (Phe 2 2 ) Phe 3 3 ( P h e 3 3 ) Tyr 35

-~4.0 5.60

Phe 4 5

T y r 21 ( T y r 2 1 ) Gin Gly Met Cys

31 ( A r g 2 0 ) 36 52 30

Arg 20 (Gin 31) Cys 55 Ile 1 8 Thr 3 2 A s n 43 N S H 1 Ash 24 Ash 43 N b H 2

* T h e e x p e r i m e n t s leading to t h e s e a s s i g n m e n t s are d e s c r i b e d in the t e x t . The previous a s s i g n m e n t s b y Marinetti e t al. [ 2 4 , 2 5 ] are i n c l u d e d in p a r e n t h e s e s for c o m p a r i s o n .

were already completely exchanged, i.e. the proton 8, 12, 14, 17, 18, 20, 21, 23 and 29--33. Fig. 5 shows that certain amide proton lines overlap mutually, so that selective irradiation in nuclear Overhauser or spin decoupling experiments is difficult. It was therefore important for certain experiments that, because of the largely different exchange rates of different protons, the total number of amide proton lines is selectively reduced with time after the sample preparation, resulting in improved resolution of the remaining lines [22,23]. On the other hand additional resolved lines, besides the 33 resonances observed in 2H20, can be seen in H20 solutions of the inhibitor (Ref. 35, and Wiithrich, K. and Wagner, G., unpublished results).

Useful features of the protein conformation In the interpretation of the nuclear Overhauser experiments it was assumed that the refined single crystal atomic coordinates [1] are conserved in the globular solution conformation of the inhibitor. Evidence in support of this assumption was presented elsewhere [29,35--37]. It is then of interest that entirely different proton-proton nearest-neighbor relations occur in the/3-sheet

185

r3. PPM 4

I 0

1

PPM

/ j/

/

/

/

/, /

I

!

QH I PPM 6

6

32

i

!

i

NH

1

23

PPMIO

I

2 PPM

6

i 7

101

6 PPM

Fig. 5. 3 6 0 M H z I H N M R s p e c t r u m o f t h e i n h i b i t o r in 2 H 2 0 s o l u t i o n a t 4 5 ° C , p 2 H 4 . 5 . T h e s p e c t r u m was recorded 90 min after sample preparation. The spectral resolution was improved by multiplication of t h e f r e e i n d u c t i o n d e c a y first w i t h a n i n c r e a s i n g e x p o n e n t i a l w i t h t i m e c o n s t a n t 4 s-1 a n d t h e n w i t h a s h i f t e d sine b e l l [ 1 9 ] . F r o m t o p t o b o t t o m t h e s p e c t r a l r e g i o n s o f t h e a l i p h a t i c side c h a i n r e s o n a n c e s , t h e ~-protqn resonances, and the aromatic and amide proton resonances are presented. The resonances of the labile p r o t o n s a r e n u m b e r e d in t h e o r d e r o f t h e c h e m i c a l s h i f t s , as p r o p o s e d p r e v i o u s l y ( R e f s . 1 9 a n d 3 5 ) , C e r t a i n n u m b e r s are m i s s i n g b e c a u s e t h e s e p r o t o n s w e r e a l r e a d y f u l l y e x c h a n g e d w h e n t h i s s p e c t r u m w a s r e c o r d e d . S o m e o f t h e c o r r e s p o n d e n c e b e t w e e n r e s o n a n c e s o f t h e s a m e s p i n s y s t e m s , w h i c h w e r e evid e n c e d b y s p i n d e c o u p l i n g a n d are listed in T a b l e 1, a r e i n d i c a t e d b y t h e lines c o n n e c t i n g t h e d i f f e r e n t spectral regions.

and in the a-helix of the inhibitor [1]. This is illustrated in Fig. 6 for all the amide protons involved in hydrogen bonds of the {]-sheet or the a-helix. The sixth p o w e r of the distances between the amide protons and the neighboring protons was plotted, since the initial build-up rates for nuclear Overhauser effects are simply related to this quantity [26,28,32]. In the {]-sheet there are, with the exception of residues 16 and 27, always a-protons among the nearest-by protons (Fig. 6). More precisely, the nearest neighbor of the amide proton of residue n is almost exclusively the a-proton of residue (n -- 1). The second-nearest a-proton is that of the residue n, which can

186

p-pleated-stoat 22 I 16

2J6 I o

3.O I

o

**•

3.4 I

0 20

•

21

o ~o

22

•

23 24

•

[]

O

oo

0

o

o

•0

• o o

o•

0

O

•

0

O

•

o

• o

0

0

0

0 0

o

0

•

•

0 0

0

0

& A0

27

~

28

2

33

~

35

o

O

•

0

•

[]

o .e

se~ 45

@ 0

0

I 0 •

o

o

•

o

o

•

o

0

0

.,

o

•

o •

o

o

•

e °

•

~o

~

0

o

°o •

moo

o

o

I o~-I"~ix 2B I

3.O

I

J

51

oo

=o

52

~?~

..

53 i 54

~

o

[]] l ~

•

ooS ~j

•

•

* aA

I

o

o

i

•

o

.

D

: o •

o

o •

*

o

o

•

• •

n

0

.

O

Fig. 6. R e p r e s e n t a t i o n of t h e sixth p o w e r o f t h e d i s t a n c e s b e t w e e n t h e h y d r o g e n - b o n d e d p e p t i d e p r o t o n s of t h e /3-sheet a n d t h e s - h e l i x in t h e i n h i b i t o r ( i n d i c a t e d in t h e first c o l u m n ) and t h o s e s - p r o t o n s (A), ~ - p r o t o n s (D), a r o m a t i c p r o t o n s (0) a n d o t h e r h y d r o g e n - b o n d e d a m i d e p r o t o n s ( e ) w h i c h are l o c a t e d within 4.0 A f r o m the respective peptide proton, The distances which correspond to the plotted sixth p o w e r v a l u e s are also i n d i c a t e d .

easily be identified by spin decoupling. The next nearest s-proton is in no case closer than 3.0 .~ and can therefore by nuclear Overhauser studies be distinguished from the nearest-by a-proton, which is typically at a distance of 2.2 (Fig. 6). Hence, once a backbone resonance in the fi-sheet is assigned one can obtain additional assignments in a sequential way: from the assigned amide proton n, the neighboring s - p r o t o n of residue (n -- 1) is identified b y nuclear Overhauser experiments; the amide p r o t o n (n -- 1) is then identified b y spin decoupling and is next used to find the s - p r o t o n ( n - 2) by nuclear Overhauser studies, etc. In the a-helix the nearest neighbors of the peptide protons are mostly fi-protons (Fig. 6). Thus a sequential assignment as in the ~-sheet was n o t possible.

187

Selective nuclear Overhauser effects In spite of the strong spin diffusion in the inhibitor (Fig. 1), b o t h transient (Fig. 2) and truncated driven Overhauser experiments (Fig. 3) can give reliable information on intramolecular proton-proton distances [26,28,32,33]. However, while the t w o experiments were readily feasible when the lowest field amide proton 1 at 10.6 ppm was irradiated (Figs. 2 and 3), it proved much more difficult, if n o t impossible, to obtain selective transient Overhauser effects when irradiating in more crowded spectra regions (Fig. 5). Therefore, truncated-driven Overhauser experiments were used in the following. From the above considerations on the protein structure it is evident that individual assignments of backbone resonances will largely depend on the determination of interatomic distances. While strictly speaking only the initial buildup rates of the Overhauser effects are simply related to the proton-proton distances [26--28,32], it is of considerable practical importance that the resonance intensities in truncated driven nuclear Overhauser difference spectra are simply related to the build-up rates when the preirradiation time tl is shorter than approximately 400 ms (Fig. 4). Therefore, rather than recording the time course of the driven Overhauser effect in each individual case, an optimal value for tl was determined from Fig. 4 and the relative intensities of individual lines (in percent of the corresponding resonance intensity in the unperturbed Fourier transform NMR spectrum) in a single truncated Overhauser experiment were used to compare relative distances from the irradiated nucleus. In all these experiments, t~ was 400 ms.

Proton-proton spin decoupling Spin decoupling data were obtained either with the use of conventional difference spectra [38] or with spin decoupled nuclear Overhauser difference spectra [30]. Spin decoupling experiments served two purposes. Firstly, once the amide proton or a-proton line was identified b y nuclear Overhauser experiments in the sequential procedure used for the ~-sheet, they provided resonance assignments within the spin systems of the individual amino acid residues. The correlations between amide, a- and ~-protons, which were thus established, are listed in Table I and in part also shown in Fig. 5. Secondly, numerous a- and ~-proton lines had previously independently been assigned [29,35,36,38] with different techniques. Spin decouplings with these resonances then provided double checks for some of the resonance assignments described in this paper.

Distinction between amide protons of the a-helix and the {J-sheet Three criteria were used for this global distinction. The first one is a direct consequence of the structural features presented in Fig. 6: when u p o n irradiation of an amide proton the strongest Overhauser effects were exclusively on fl-protcns, it was concluded that the amide proton in question is n o t in the ~-sheet and hence could quite likely be located in the a-helix. Secondly, the vicinal amide p r o t o n ~ proton spin-spin coupling constants 3JNHa are expected to be of the order of 7--11 Hz in the ~-sheet and, with the exception of Cys 55, smaller than 5 Hz in the a-helix [13]. Thirdly, potential lanthanide binding sites in the native inhibitor are all far away from the ~-sheet, whereas some are at a short distance from the a-helix [29]. Hence addition of praseodymium(III)

188 as a shift reagent [29] should n o t affect any of the amide proton lines in the B-sheet, b u t might well cause shifts for amide protons in the a-helix. On the basis of these three criteria the most slowly exchanging amide protons could all be located in the B-sheet. Indication for a location in the a-helix was obtained for three resonances, i.e. line 10 and one line of each of the two groups 14 and 15, and 26--28. Individual resonance assignments in the Blsheet The antiparallel B-sheet in the inhibitor is formed by the residues 1 6 - 3 6 . It is strongly twisted and extends through the full length of the molecule [ 1 ]. The above-mentioned sequential procedure for resonance assignments was started from the amide proton 6 at 9.25 ppm (Fig. 5), which was assigned to Tyr 21. The corresponding a-proton is at 5.70 ppm (Fig. 5). This assignment of the Tyr 21 amide proton is identical to that obtained previously from a different approach [24,25]. It was here evidenced by nuclear Overhauser experiments with the previously assigned [10] aromatic protons of Tyr 21, and confirmed by the entity of the experiments described in the following.

A16e

A[•

•A16

G36• •G36

G36e

R17&

Y~

nso/~ 1

V34D

T32G ~R20

Q31• Y21 C30i "

' Y23

"

F 4 5 ! ~

•

F22 N4~ ~'23

•N~ Fig. 7. P l a n a r p r o j e c t i o n o f t h e c e n t r a l p a r t o f t h e ~-sheet in t h e i n h i b i t o r . T h e a m i n o a c i d s are d e n o t e d u s i n g t h e I U P A C o n e - l e t t e r c o d e . T h e d i s t a n c e s b e t w e e n a m i d e p r o t o n s ( e ) , (~-protons (A) a n d ~ - p r o t o n s ([]) a r e p r o j e c t i o n s o f t h e a c t u a l d i s t a n c e s in t h e t h r e e - d i m e n s i o n a l s t r u c t u r e . T h r e e p a t h w a y s f o r s e q u e n tial r e s o n a n c e a s s i g n m e n t s , all s t a r t i n g f r o m Y 2 1 , a r e i n d i c a t e d b y t h e a r r o w s (see t e x t ) . T h e o p e n a r r o w s indicate the locations where the resonance assignments were double checked by spin decoupling with prev i o u s i y a s s i g n e d side c h a i n r e s o n a n c e s [ 2 9 , 3 6 , 3 8 ] .

189 Starting from the amide proton and the a-proton of Tyr 21, further assignments were obtained by sequential use of Overhauser and spin decoupling measurements along the three pathways outlined in Fig. 7. The first pathway goes from Tyr 21 to the amide p r o t o n of Phe 22, from where it branches o u t into t w o directions, i.e. via the a-proton of Phe 22 to Tyr 23 and via Gln 31, Cys 30 and Asn 24 to T y r 23, From the a-proton of Tyr 21 the amide resonance 3 at 9.79 ppm was assigned to Phe 22 by an Overhauser experiment. From there the a-proton resonance of Phe 22 was identified by spin decoupiing at 5.25 ppm. Having identified the a-proton 22, the amide resonance 1 at 10.6 p p m was assigned to Tyr 23 (Figs. 2--4), and then the a-proton signal of Tyr 23 was identified at 4.30 ppm by spin decoupling. The second branch started with the Overhauser effect from the amide proton resonance of Phe 22 to the amide resonance 7 at 8.77 ppm, which was thus assigned to Gln 31 in the other peptide strand. From spin decoupling the a-proton of Gln 31 was then localized at 4.87 ppm. Overhauser effects obtained with irradiation of the amide proton resonance of Gln 31 assigned the a-proton signal at 5.60 ppm to Cys 30. The amide proton of Cys 30 exchanges rapidly, b u t could be observed in H20 solution at 8.40 ppm. The t w o fl-proton resonances of Cys 30 are largely inequivalent. The fl-proton in gauche configuration to the a-proton was localized at 2.67 p p m and that in trans configuration at 3.69 ppm. From the a-proton resonance of Cys 30 the amide resonance 22 at 7.78 ppm was assigned to Asn 24 by an Overhauser experiment and subsequently the a-proton resonance of Asn 24 was found at 4.41 ppm by spin decoupling. Overhauser effects observed with irradiation of the amide proton resonance of Asn 24 resulted in the assignment of the a-proton resonance of Tyr 23, confirming the assignm e n t via the first branch. The t w o singlet-like resonances 19 and 24 at 7.98 and 7.78 p p m were identified as the N ~ amide protons of Asn 43 by Overhauser effects obtained with irradiation of the amide proton of Tyr 23 (see Figs. 2 and 3). In the crystal structure the proton of signal 19 forms a hydrogen bond to the carbonyl oxygen of Glu 7, and that of resonance 24 is in a hydrogen bond to the carbonyl oxygen of Tyr 23 [1]. The second assignment pathway goes from the amide proton of Tyr 21 to Phe 45 and Asn 44. The amide proton of Phe 45 at 9.94 ppm was identified by Overhauser effects. Spin decouplings showed that the a- and 13-proton signals of Phe 45 are at 5.12 ppm and 2.79 ppm, respectively. Overhauser effects obtained with irradiation of the amide proton of Phe 45 identified the a-proton resonance of Asn 44 at 4.88 ppm. The third assignment pathway goes from the amide proton of Tyr 21 to Arg 20 and is branching at the amide proton of Arg 20 into three directions. The first branch leads to Phe 33 and Thr 32, the second branch to Tyr 35, Val 34, Gly 36, Ile 18 and Arg 17, and the third branch to Ile 19. From the amide proton of T y r 21 the a-proton resonance of Arg 20 was localized at 4.70 ppm by an Overhauser experiment. By spin decoupling the amide proton resonance of Arg 20 was found to be line 11 at 8.39 pprn. Next, the amide resonance 4 at 9.37 ppm was identified as that of Phe 33 from Overhauser measurements. The a-proton signal of Phe 33 was detected at 4.89 p p m with spin decoupling. From the amide proton of Phe 33 the sharp a-proton resonance at 5.28 ppm, which overlaps with the a-proton resonance of Phe 22, was assigned to Thr 32

190 by an Overhauser experiment. The small value of approx. 2.0 Hz for the spinspin coupling constant 3 j ~ of Thr 32 was previously described [36]. The a-proton of Thr 32 is coupled to the amide proton resonance 18 at 8.07 ppm, which exchanges t o o fast to be seen in Fig. 5. The second branch of the third pathway starts with the assignment of the amide proton line 5 at 9.39 ppm to Tyr 35 by the Overhauser technique. By spin decoupling in nuclear Overhauser difference spectra [30], the a-proton and the two /~-protons of Tyr 35 were then localized at 4.89, 2.69 and 2.52 ppm. Overhauser effects between the a-proton resonance of Tyr 35 and the amide proton resonance 9 at 8.61 led to the assignment of the latter to Gly 36. In Fig. 5 the triplet-type fine structure of the amide proton resonance 9 can be recognized. From nuclear Overhauser effects obtained with irradiation of the amide proton of Tyr 35, the a-proton of Val 34 was identified at 3.92 ppm. The latter assignment was confirmed by spin decoupling with the previously assigned fi-proton of this residue at 1.95 ppm [36,38]. From Overhauser effects between the amide proton of Tyr 35 and the amide resonance 16 at 8.12 ppm, the latter was assigned to Ile 18. The a-proton resonance of Ile 18 was then detected at 4.25 ppm by spin decoupling. The third branch of the third pathway consists of a single step, i.e. the Overhauser effect built up from the amide proton of Arg 20, resulted in the identification of the a-proton resonance of Ile 19 at 4.30 ppm. In previous work, the resonances of the Ca, C~1, C~2 and C8 protons of the two isoleucines 18 and 19 in the inhibitor were identified but not individually assigned [36,38]. Spin decoupling with the two lines which were previously assigned to the Ca protons of the two isoleucines now provided on the one hand a double check of the assignments of the backbone protons (Fig. 7), and on the other hand resulted in the individual assignments of the/3-methine and the 7-methyl resonances of Ile 18 and Ile 19 (Table I). For Ile 18, the ~-proton line is at 1.87 ppm and the ~/-methyl at 0.97 ppm, for Ile 19 the corresponding chemical shifts are 1.96 ppm and 0.73 ppm. Individual resonance assignments in the a-helix Evidence for individual assignments was found for two among the three resonance lines for which a location in the a-helix appeared likely. When the amide proton resonance 15 at 8.25 ppm was irradiated, the previously identified [36] spin system of the side chain of Thr 54 appeared in the nuclear Overhauser difference spectrum. From inspection of the atomic coordinates and comparison of the expected coupling constants with the experiment, the amide proton resonance 15 was thus assigned to Cys 55. Irradiation of the amide proton resonance 10 at 8.58 ppm produced Overhauser effects on two spin systems which would be compatible with the location of this amide proton between the side chains of a cysteinyl and a methionyl residue. Line 10 was therefore tentatively assigned to the amide proton of Met 52.

Discussion The main purpose of this paper was to describe the experiments used to obtain individual assignments for 18 amide proton resonances in the inhibitor.

191 These data complement earlier work which resulted in the individual assignments of all the aromatic resonances (Refs. 10,19--21,29,35, and Wagner, G., Tschesche, H. and Wiithrich, K., unpublished results) and most of the methyl resonances of the aliphatic side chains [29,36,38]. Individual assignments of proton resonances of the polypeptide backbone are for a fundamental reason even more difficult to obtain than those for amino acid side chains. While each type of amino acid side chain occurs usually in a quite small number in a protein, the backbone protons present a large number of resonances with very similar spectral properties. Therefore, comparison of different homologous or chemically modified proteins may often result in unambiguous assignments of side chain resonances [13,36], but will not usually be a suitable method for assignments of backbone proton lines. It would appear that the sequential use of Overhauser and spin decoupling experiments should be a generally applicable technique for identifying nearestneighbor relations in antiparallel ~-sheets. The perspective drawing of the

V34

R20

C30 Fig. 8. P e r s p e c t i v e c o m p u t e r d r a w i n g o f t h e c e n t r a l p a r t o f t h e ~-sheet in t h e i n h i b i t o r . This figure illust r a t e s t h e c l o s e d i s t a n c e s b e t w e e n t h e a m i d e p r o t o n s o f r e s i d u e s n a n d t h e a - p r o t o n s o f r e s i d u e s (n - - 1), w h i c h w a s e s s e n t i a l for t h e s e q u e n t i a l r e s o n a n c e a s s i g n m e n t s b y a l t e r n a t i v e use of spin d e c o u p l i n g a n d n u c l e a r O v e r h a u s e r e f f e c t m e a s u r e m e n t s . I n t r a m o | e c u l a r h y d r o g e n b o n d s a~e i n d i c a t e d b y IlllI[ [L

192 twisted ~-sheet in the inhibitor (Fig. 8) clearly illustrates two structural features which were essential for the use of the spectroscopic techniques described here for resonance assignments. First, within each peptide strand of the fl-sheet, the crucial property was the close proximity of the amide proton with the s-proton of the preceding residue (indicated by arrows in Fig. 8). Second, the arrangement of the hydrogen bonds brings always one amide proton of the first strand into close proximity with an amide proton of the antiparallel second strand, which allowed to extend the sequential assignment procedure also across the fl-sheet. In certain cases this provided new additional assignments, in other cases important double checks on assignments made within each individual peptide strand. It may well be that close examination of other regular polypeptide structures could open similar avenues. Since spin decoupling can in m a n y cases be used to identify the spin systems of entire amino acid residues [13], the techniques described here could in principle probably also be applied to establish the amino acid sequences in l-sheet regions of proteins. In view of the other methods available for peptide sequencing, such applications would possibly be of rather limited practical interest. Using lanthanide shift reagents with a chemically modified inhibitor, assignments were previously obtained for five amide proton resonances in the inhibitor [24,25]. From Table I it is seen that three of these earlier assignments could be confirmed by the present experiments. On the other hand, the resonances 7 and 11 (Fig. 5) are now found to correspond to Gln 31 and Arg 20, respectively, rather than to Arg 20 and Gln 31, as suggested previously [24,25]. From an examination of the assumptions used for the analysis of the lanthanide shifts [24,25], it would appear that this technique could not really differentiate between these two protons. Hence, one should probably argue that there is satisfactory agreement between the two sets of data, since the same two lines were assigned to the pair of amide protons of Arg 20 and Gln 31. Structural interpretations of the NMR properties of the amide protons in the inhibitor, which depend to a large measure also on the assignments in Table I, are described elsewhere (Refs. 14--18, and Wiithrich and co-workers, unpublished results). Therefore, only two outstanding observations shall be listed here. One is that all the most slowly exchanging labile protons of the inhibitor are in the ~-sheet. In particular, all the amide protons in the peptide fragment 20--24 exchange very slowly (Richarz, R., Sehr, P., Wagner, G. and Wiithrich, K., unpublished results). This fragment occupies a quite central position in the molecule. On the one side it has the second strand of the /3-sheet as a near neighbor, on the opposite side the peptide fragments 6--9 and 43--46. Secondly, it is quite striking that the six resonances at lowest field (Fig. 5) correspond to those six aromatic residues which are preserved in homologous inhibitors [19,20] and which appear from this and other observations to be essential for the architecture of the molecule [14--20].

Appendix This appendix describes the algebraic functions which were used to compute tl~e time course of the driven nuclear Overhauser effects in Fig. 4A. A threespin system was considered, including the irradiated spin, the observed spin and

193

a third spin which represents the bulk of the protons in the protein. The time dependence of the magnetisation of the non-irradiated spins i in a driven Overhauser experiment is determined by Eqn. 3 [28]. oijM j (3) dt j¢i Mi is the difference between the actual magnetisation M z i of spin i and its equilibrium magnetisation M °, and correspondingly Mj = M z j - - M ° . Pi a n d oij describe the spin lattice relaxation and the spin diffusion.

dM---~i=--'piMi - - D

~2")'4 i ~ 1 I 3T¢ 6re 7 Pi = -~-~- .__. rT6j re + 1 + (osrc) ~ + 1 + 4(¢ore)2J

(4)

fi27' 1 [ 6re oij - 10 r6j 1 + 4(6o%) 2

(5)

1 Te

W is the Larmor frequency, r~j the distance between spins i and j, 2rrh the Planck constant, 7 the gyromagnetic ratio and re the effective rotational correlation time. In the three-spin system, we denote as spin 1 the irradiated nucleus, as spin 2 the observed nucleus and as spin 3 the nucleus which represents the influence of additional spins upon the driven Overhauser effect. The two coupled differential Eqn. 3 for the spins 2 and 3 were solved with the assumption that the time course of the magnetization of the irradiated spin 1 can be described by 0,

for t < 0

M, = M°( 1

_ e-et)

for t/> 0 ,

,

(6)

where M ° is the equilibrium magnetization of spin 1 and c is a constant which determines the rate of saturation of spin 1. The expression (6) for M, was selected empirically on the basis of the time course of the saturation observed with the experimental conditions used [33]. For the magnetization Ms of the observed spin one thus obtains M2 -

M°

0"21

(b--a) +

(e - a t - - e - b t ) + (o21P3 - - 0"23031) ( b e - a t - - a e - b t + (a - - b ) ) ( a - - b)ab ((b--c)

0"21P3 - - 0230.31

(a

--

b)(b

--

c)(c

--

e-at+(c--a)e-bt+(a--b)e

-ct )

(7)

a)

0"2, (c(a - - b) e - c t + a ( b - - c ) e - a t + b ( c - - a ) e - b t (a - - b ) ( b - - c ) ( c - - a)

with a = {[(P2 + P3) - - ~/(P2 - - P 3 ) ~ + 40"~3]

(8)

and

b = ½[(p: + P3) + x/(P2 --P3) 2 + 40"]3]

(9)

The t, dependence of the driven Overhauser effects in Fig. 4A was computed with Eqn. 7.

194

References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

Deisenhofer, J. and Steigemann, W. (1975) Acta Cryst. B 3 1 , 2 3 8 - - 2 5 0 Gelin, B.R. and Karplus, M. (1975) Proc. Natl. Acad. Sci. U.S. 72, 2002--2006 Hetzel, R., WUthrich, K., Deisenhofer, J. and Huber, R. (1976) Biophys. Struct. Mech. 2, 159--180 McCammon, J.A., Gelin, B.R. and Karplus, M. (1977) Nature 267, 585--590 Levitt, M. and Warshal, A, ( 1 9 7 5 ) Nature 2 5 3 , 6 9 4 - - - 6 9 8 Levitt, M. ( 1 9 7 6 ) J. Mol. Biol. 104, 59--107 Chothia, C. ( 1 9 7 6 ) J. Mol. Biol. 105, 1--14 Creighton, T.E., Dykes, D.F. and Sheppard, R.C. (1978) J. Mol. Biol. 1 1 9 , 5 0 7 - - 5 1 8 WQthrich, K. and Wagner, G. (1975) FEBS Lett. 50, 265--268 Snyder, G.H., R o w a n Ill, R., Karplus, S. and Sykes, B.D. (1975) Biochemistry 14, 3765--3777 Wagner, G., De Marco, A. and Wiithrieh, K. (1975) J. Magn. Resonance 20, 565--569 Wagner, G., De Marco, A. and Wiithrich, K. (1976) Biophys. Struct. Mech. 2, 139--158 W(ithrich, K. (1976) NMR in Biological Research: Peptides and Proteins, North-Holland, Amsterdam Wagner, G. and Wiithrich, K. (1978) Nature 275, 247--248 W~ithrich, K. and Wagner, G. (1978) Proceedings Int. Symp. on Biomolecular Structure, Conformation, F u n c t i o n and Evolution, Madras, India, in press Wilthrich, K. and Wagner, G. (1978) Trends Biochem. Sci. 3 , 2 2 7 - - 2 3 0 Wilthrich, K., Wagner, G. and Bundi, A. (1978) In Nuclear Magnetic Resonance Spectroscopy in Molecular Biology, Proceedings of the 11th Jerusalem S y m p o s i u m on Q u a n t u m Chemistry and BiDchemistry (Pullman, B., ed.), pp. 201--210, Reidel, Dordrecht, Holland WUthrich, K., Wagner, G. and Richarz, R. (1978) Proceedings of the 12th FEBS Meeting in Dresden, in press Wagner, G., Wtithrich, K. and Tschesche, H. (1978) Eur. J. Biochern. 86, 67--76 Wagner, G., Wiithrich, K. and Tschesche, H. (1978) Eur. J. Biochem. 89, 367--377 Brown, L.R., De Marco, A., Richarz, R., Wagner, G. and Wiithrich, K. (1978) Eur. J. Biochem. 88, 87--95 Masson, A. and Wiithrieh, K. (1973) FEBS Lett. 31, 114--118 Karplus, S., Snyder, G.H. and Sykes, B.D. (1973) Biochemistry 12, 1323--1329 Marinetti, T.D., Snyder, G.H. and Sykes, B.D. (1976) Biochemistry 15, 4 6 0 0 - - 4 6 0 8 Marinetti, T.D., Snyder, G.H. and Sykes, B.D. (1977) Biochemistry 16, 647---653 Solomon, I. (1955) Phys. Rev. 9 9 , 5 5 9 - - 5 6 5 Noggle, J.H. and Schirmer, R.E. (1971) The Nuclear Overhauser Effect, Chemical Applications, Academic Press, New York Kalk, A. and Berendsen, H.J.C. (1976) J. Magn. Resonance 24, 343--366 Perkins, S.J. and Wilthrich, K. (1978) Biochim. Biophys. Acta 5 3 6 , 4 0 6 - - 4 2 0 Richarz, R. and Wfithrich, K. (1978) J. Magn. Resonance 30, 147--150 Sykes, B.D., Hull, W.E. and Snyder, G.H. (1978) Biophys. J. 21, 137--146 Gordon, S.L. and Wiithrich, K. (1978) J. Am. Chem. Soc. 100, 7094--7096 Wagner, G. and Wfithrich, K. (1978) J. Magn. Resonance, in press Englander, S.W., Downer, N.W. and Teitelbaum, H. (1972) Annu. Rev. Biochem. 41, 903--924 Wagner, G. (1977) Ph.D. Thesis 5992, E.T.H. Zfirich W/ithrieh, K., Wagner, G., Richarz, R. and Perkins, S.J. (1978) Biochemistry 17, 2253--2263 Perkins, S.J. and Wilthrich, K. (1978) Biochim. Biophys. Aeta 5 7 6 , 4 0 9 - - 4 2 3 De Marco, A., Tschesehe, H., Wagner, G. and Wfithrieh, K. (1977) Biophys. Struct. Mech. 3 , 3 0 3 - - 3 1 5