Biochimica et Biophysica Acta, 492 (1977) 245-253

© Elsevier/North-Holland Biomedical Press BBA 37667 I N F L U E N C E OF H Y D R O G E N B O N D I N G ON T H E R O T A M E R DISTRIBUTION OF T H E H1ST1D1NE SIDE CHAIN IN PEPTIDES: 1H N M R AND CD STUDIES

TOAN TRAN, KARL LINTNER, FLAVIO TOMA and SERGE FERMANDJIAN* Service de Biochirnie, D~partement de Biologic, Centre d'dtudes nucl~aires de Saclay. B.P. No. 2, 91190 Gif-sur- Yvette (France)

(Received November 9th, 1976)

SUMMARY Both 1H N M R and circular dichroism pH titration studies on histidine, HisGly, Gly-His and Gly-His-Gly indicate that the side-chain spatial orientation depends strongly on the vicinal charges. The arrangement of the imidazole side-chain (rotamer population) is shown by the histidine fl and fl' and the glycine methylene proton chemical shifts as well as the vicinal XH-1H coupling constants 3Jca_n.~.H. ~'-u. For His-Gly and Gly-His-Gly a good correlation can be found between the ionization of the glycine COOH group and the increase of rotamer III (g-g) which is also visualized by circular dichroism through an enhancement of the ellipticity at 212 rim. In these two peptides a hydrogen bond between the imidazolium and the carboxylate group is supposed to stabilize rotamer III at pH 4-5.

INTRODUCTION Our interest in histidine was aroused by the fact that it enters the composition of many natural peptides studied in this laboratory, such as thyroliberin, angiotensin 11 and corticotropin, in which it plays an important role [I-3]. Although histidine has been studied in many ways [4, 5], recent developments in high field N M R spectroscopy now allow a more detailed analysis of the aliphatic region of the spectrum. Free histidine, and small peptides with a histidine residue in the sequence: L-histidylglycine (His-Gly), glycyl-L-histidine (Gly-His) and glycyl-L-histidylglycine (Gly-His-Gly) were studied. Full analysis of the proton N M R spectra gave, among other data, the vicinal coupling constants between the a and fl, fl' protons of the histidine residue. Internal chemical shifts for the fl and fl' and glycine methylene protons (nonequivalences) were also obtained. As the histidine side chain is a chromophore of its own and gives rise to circular dichroism (CD) signals, it was interesting to study the four compounds mentioned by this technique, too. In these simple model peptides the side chain * To whom correspondence should be addressed.

246 chromophore contribution is relatively strong and can be followed on its own as a function of conformational parameters. EXPERIMENTAL The compounds used (Sigma Chem. Comp.) were of the highest purity and in the form of the free base. Solutions were prepared in deuterium oxide (Commissariat /~ l'Energie Atomique, France; isotopic enrichment 99.8~) at concentrations ranging from 0.10 to 0.15 M, except for Gly-His-Gly which was at a concentration close to its isoionic point, 0.03 M. The pH was adjusted with 2HCI or NaOZH solutions (Commissariat ~ l'Energie Atomique, France; isotopic enrichment superior to 97 ~o). None of the measurements was corrected for p2H. Spectra were obtained at 90 M H z on a Bruker H X 90 spectrometer equipped with a deuterium lock or at 250 M H z on a Cameca TSN 250 spectrometer using a homolock on the tertiary butanol signal. Frequency readings were accurate to ± 0.05 Hz. "[he choice of spectrometer depended on the observability of the nonequivalence of the /4-CH2 protons of histidine. In almost all the experiments the proton C-5-H of the imidazole ring was decoupled during observation of the fl-CH2 protons. All spectra were analysed with the LAOCN3 program [6]. Circular dichroism measurements were made on a Jobin Yvon Dichrograph Ill model using fused quartz cells of 0.1 mm path length. Concentrations used were in the order of I mg/ml, p H adjustments were effected with concentrated HCI and N a O H solutions so that the peptide concentration did not change measurably. Results are expressed in molar ellipticity [0]. RESULTS AND DISCUSSION NMR

Fig. 1 shows the spectrum of His-Gly at 250 MHz, obtained in 2H20 at p H 4.2, as an example. It displays the following characteristics: Protons

Structure

Remarks

C-2-H C-5-H ~t-C-H

doublet unresolved quintuplet quadruplet /

coupled with C-5-H

fl-CH2 CH2 (glycine)

multiplet / quadruplet

form an ABX system AB system, well separated

The assignments of the fl and/3' protons of histidine in the three rotamers are represented in Table I. They are based on generally accepted arguments, e.g. rotamer l which presents the least steric hindrance is almost always preferred over the two others [5-7, 8]. This finds support in the 3ff13co.13c4 values in histidine and histidine derivatives [9] which allow the determination of the rotamer I fraction, and in studies of phenylalanine with a selectively deuterated fl-carbon [10]. All the proton chemical shifts as well as signal aspects and coupling constants vary greatly depending on the p H ; thus

247

•

6;0

'

i

6~0'

+

660

r

6;0

Hz 6+0

j•

5Hz

++0

~

'

+70'

i

~

++0 Hz +~0

~ . ~

t,, C-2-H .

.

C--5-H 7.

.

.

6.

C-a-H .

5

'

~,

'

3

CH.

~ C-#-H a '

ppm/ t-SuON

Fig. 1. 250 MHz ~H NMR spectra of L-histidylglycine in 2 H 2 0 at pH 4.2. All parts of the survey spectrum (bottom) are presented expanded. At this pH the AB quadruplet of glycine methylene protons exhibits the greatest nonequivalence (0.226 ppm). the determination o f chemical shifts and coupling constants for each p H value allows the plotting o f the curves shown in Figs. 2, 3 and 4. N o n e o f these curves shows m o n o t o n o u s variation. All represent, to a varying extent, the ionization states o f the different groups with privileged experimental points, e.g. the p H values for which the monocations are in a pure state; isoionic

248 TABLE I R O T A M E R P O P U L A T I O N S O F T H E H I S T I D I N E S I D E C H A I N O B T A I N E D A T pH V A L U E S CORRESPONDING TO PURE IONIC SPECIES Compound

Ionic species

"pH"

co

co

co

Im

Ha

I

II

I11

Nr ~1~ "Ha Ha '

Histidine

NH; ~ NH~ NH3+ NHz

lmH + lmH + Im lm

COOH COOCO0COO-

0.5 3.7 7.8 12.0

0.38 0.40 0.53 0.48

0.36 0.32 0.16 0.22

0.26 0.28 0.31 0.301

His-Gly

NH~ NH~ NH~ NHz

lmH + ImH ÷ lm Im

COOH CO0CO0COO-

1.3 4.2 6.9 12.0

0.49 0.37 0.34 0.33

0.29 0.15 0.35 0.39

0.22 0.48 0.3 I 0.28

Gly-His

NH;NH3+ NH3 ~ NHz

lmH + hnH ÷ lm lm

COOH COOCOOCO0-

1.3 5.0 7.5 10.0

0.52 0.52 0.56 0.56

0.22 0.22 0.18 0.16

Gly-His-Gly

NH + NH~ NH; NH2

lmH + ImH ÷ Im lm

COOH CO0CO0COO-

1.0 4.9 8.0 11.0

0.50 0.32 0.52 0.55

0.33 0.33 0.26 0.23

, I , t

0.26 0.26£ 0.26 0.28 0.17 i 0.35 0.22' 0.22 ~

!

....

yiJ

i

........+:+++:J

\ Hi+

+

H,S Gly " my - Hi+

D -flU

X+

•

+

\L

-

t

÷',~'

}~,_

+

I 1

: z

,

J

,

1

s

6

~,

i

L,

i

l,

,i

"pW

Fig. 2. Chemical shift nonequivalence curves o f histidine side chain fl a n d ~' protons. As an e x a m p l e of an a r o m a t i c a m i n o acid without ionizable f u n c t i o n s the curve o f phenylalanine is also presented. N o t e that in the case of His-Gly the ~ da, n o n e q u i v a l e n c e is zero at p H 6.4 where the glycine methylene proton 6 nonequivalence is also zero a n d Jaa -- Jaa" = 6.3 H z (see Figs. 3 a n d 4). F o r such a coupling c o n s t a n t the fractions of r o t a m e r s I, II, III are 0.33.

249 I

0,2~

i ,-~+%

/

0.20

/

\ \

/

/

I

+

His - Gly

o

Gly - His

•

~

o

GLy - His - Gly

- His - G~y

/

I I t

s÷ /

J

O.

'~ a05 ._o_

.o=. - . o -

~,O-- 0-O --IL..e

\\

0.00

- o,o~

P

i

i

i Ii

\--

i

i

-o,0

J

1

i

z

,

i

\I ..... "

i i

I

i

s

3

11

"pH"

I

12

Fig. 3. Chemical shift nonequivalence curves of glycine methylene protons. points, where neutral zwitterions constitute the majority if not all the molecules. Comparison of the nonequivalence curves of the glycine methylene protons (Fig. 3) with those of the fl and fl' protons of the histidine residue (Fig. 2) reveals a more or less pronounced analogy between their respective behaviour depending on the position of the glycine residue in the peptide (N- or C-terminal). Primarily for His-Gly, the two curves translate the course of the imidazole and carboxyl ionizations I +

0

7.7 o His L3

- Gly

. . . . . . o't

JaP'i

..~,/\,,

03

+

o

0.1 ~.÷

s7

! \N

'\ \

J~p

/,3

~.,

1

2

3

/

/

/

~pH"

Fig. 4. Jaa" and Jr~a in His-Gly as a function of pH.

+

.+

..o

-+'+

ooooi

250 dramatically. Simultaneously the values of Ja~ and Ja~" vary significantly (Fig. 4). When the histidine residue is at the C-terminal, the nonequivalence varies little, even if the curves do show the titration of the ionizable groups of the molecule. Finally, a histidine residue in the middle of the peptide is submitted to the effects of glycine at the N- and C-ends; the resulting curves have several of the characteristics and exhibit some of the trends of the curves of both His-Gly and Gly-His. From the coupling constants Ja~ and Ja~', we determined the rotamer populations for histidine at several of these selected points (Table 1). Among the existing methods, we chose the one of Pachler [7], taking Jg - 3.25 Hz and Jt - 12.40 Hz, which are the values resulting from the calibration curve by Kopple et al. [11]. The peptides whose rotamer distributions show the highest pH sensitivity are His-Gly and Gly-His-Gly (in both, a glycine residue is found at the C-terminal) concerning the ionization of the carboxyl and the imidazole groups. On the other hand, only small changes are observed during the titration of free histidine and the peptide Gly-His. Several kinds of interactions such as hydrogen bonding, van der Waal's forces and salt bridges are generally considered to be present in amino acids and peptides. In addition, although neglected in this study, the role of water molecules that may solvate and bridge the different groups, is probably important. The influence of hydrogen bonds seems, nevertheless, to emerge clearly, as follows from the close observation of the variations of rotamer populations versus pH. For His-Gly the rotamer population changes are particularly evident during the deprotonation of the COOH- and the imidazolium groups. If we assume the two groups, C O 0 - and imidazolium to interact at pH 4.2 (Table 1), then the preponderance of the retainer Ill over the other two could be explained. In the proposed interaction, the CH2COOpart of glycine is "frozen"; this finds support in the nonequivalence of the glycine methylene protons (0.266 ppm at pH 4.2), which is very similar to that found in cyclo-Gly-His [12]. On the other hand, 13C N M R measurements of His-Gly have shown that with increasing pH it is almost exclusively the imidazole N~ (bl) site which is deprotonated, rather than N~ [13]. Since for sterical reasons it can be assumed that it is the N.~-H which is hydrogen bonded to the carboxylate, the important loss of rotamer IlI with increasing pH above 4.5 would be a logical consequence. Whereas in His-Gly we found a strong parallelism of the h-nonequivalence curves of the histidine /3, /3' and glycine methylene protons, these curves vary in directions opposed to each other in Gly-His-Gly. Notably at pH 5 the nonequivalence of the/~,/3' protons is at a minimum; the nonequivalence of the C-terminal glycine protons, however, mimicking well the one found in His-Gly, is at a maximum. Still, as in the case of the dipeptide, one notes an increase in rotamer llI between p H I and 5, although at pH 4 and 5 the rotamer distribution which in the dipeptide was very uneven, is almost statistical in the tripeptide: 0.37, 0.15, 0.48 and 0.32, 0.33, 0.35 respectively (Table l). Consequently whereas the nonequivalence of the methylene protons of the C-terminal glycine can be explained here, too, by the interaction of its carboxylate group with the imidazolium group in the rotamer Ill state, the decrease of nonequivalence of the/3 and/3' protons is explained by the statistical distribution of the three rotamers. For His-Gly at the isoionic point (pH 6.9) all rotamers are about equally populated which contrasts with the findings for histidine and the other two peptides,

251 this shows that steric factors are compensated by interacting forces in the rotamers, ~.s found in Gly-His-Gly at pH 5. During the deprotonation of the NH + group the distribution of the rotamer population changes little in all compounds examined, which indicates that the NH + group can be substituted by NH2 without affecting the interactions with the side chain. In summary we find that (a) nonequivalence of C-terminal glycine methylene protons is greatest when there is an extra stabilization of the side chain in the rotamer llI conformation as shown by His-Gly (pH 4.2) and to a lesser degree by Gly-His-Gly (pH 5). It becomes minimal when the fractions of rotamer populations tend to be equal (His-Gly at pH 6.5-6.9). (b) Nonequivalence of N-terminal glycine methylene protons is always weak and remains almost constant for Gly-His, where practically no change of rotamer population is observed during titration. The situation is about the same for GlyHis-Gly. (c) Nonequivalence of the histidine fl and/3' protons depends on whether or not a rotamer is preferred over the other two. This is the case for Gly-His for which the unvarying nonequivalence corresponds strongly to the almost total pH independence of the preferred rotamer I. For His-Gly the nonequivalence reaches its maximum when rotamer III becomes preponderant (pH 4.2); when the fractions of rotamer populations tend to be equal (His-Gly at pH 6.5-7 and Gly-His-Gly at pH 5) the nonequivalence is minimal. We can therefore conclude that nonequivalence of the glycine methylene as well as the histidine/3 and/3' protons can be attributed largely to the restriction of rotation around the bonds of the molecule [14]. Anisotropy of the peptide bond, orientation of the carboxylic or carboxylate group and anisotropy of the imidazole or the imidazolium ring are all factors influencing the observed nonequivalence. In a molecular model of His-Gly showing the histidine side chain in the rotamer III conformation several of the observed effects can be visualized: the magnetically anisotropic imidazolium ring, hydrogen bonded to the carboxylate group, contributes to the nonequivalence of the vicinal fl and /3' and, to a lesser degree, the glycine methylene protons. Each of them is differently influenced by the carbonyl group of the peptide linkage and the carboxylate group. The deprotonation of the imidazolium group induces the rupture of the carboxylate-imidazolium hydrogen bond, thus leading to concerted changes in the molecule.

Circular dichroism To the N M R studies of these peptides we have associated circular dichroism measurements of their aqueous solutions as a function of pH. These studies allow, by the parallelism observed to the N M R results, the corroboration of the conclusions that have been drawn. Circular dichroism spectra of histidine and its derivatives consist of the peptide chromophore contributions and the imidazole side-chain contribution. The latter has only recently been studied in relation to its conformation, and efforts were concentrated on cyclo-dipeptides [15, 16]. Theoretical interpretation of the side chain contribution has been restricted to the 207 nm (lowest energy ~t--~*) transition. In the study of His, Gly-His, His-Gly and Gly-His-Gly by CD we, too, mainly followed this wavelength region.

252 [0]2,~ 10-~ I

I

I

I

I

I

L

i

/ ....

"\'x ,......

-.

Y

.

o

o.~.,.o-- • --

/i

"\.. .... i12 \

I

\', \+÷

!

/i ~.I)~D'--I)--()

+

-~-~-~,\~

/"

--

!

\ _,_

i

\

i

\

\ \

i /

\

-10

-

-

i

/

\,._ ....

i

I

I

I

t

I

I

t

t

1

2

3

t.

6

6

7

8

i pH

9

Fig. 5. Titration curves: Ellipticities 0212 . . . . plotted as a function of pH. Histidine (F7 -iT3), Gly-His (O---O), His-Gly ( L [ ), Gly-His-Gly (tD--ID).

At pH 4.8 (NH~, ImH +, C O O - ) the side chain contribution of the imidazole ring appears positive in the free amino acid and in Gly-His, whereas it is strongly negative in His-Gly and almost annihilated in the tripeptide. Fig. 5 shows the titration curves of these four compounds, obtained at 212 nm. We notice that all compounds translate the titration of the carboxyl and the imidazole group through spectral changes, whereas only the NH~- group attached to histidine (free or in His-Gly) influences the spectrum by its deprotonation. More important, however, are the relative differences observed from one titration curve to another and the spectral changes they reflect. Indeed, the positive band of histidine and of Gly-His undergoes only small intensity changes during variations of pH, much as the already very weak band of Gly-His-Gly at 212 nm decreases in intensity to almost zero at either pH 1 or pH 8. The spectacular spectral modifications during the titration of His-Gly present a strong contrast to the former 3 compounds. Although the imidazole chromophore is now removed from the C-terminal carboxyl group by an optically inactive glycine residue, the deprotonation of this group between pH 1 and pH 4 changes the positive band at 212 nm to a strong, negative one at pH 4-5. In the same way, this band changes back to positive during the imidazole titration, before undergoing the influence of the transition NH~ . . . . . . NH2. It is known that rotational power increases with increased asymmetry of a chromophore, induced, for instance,

253 by a decrease o f r o t a t i o n a l f r e e d o m a r o u n d dihedral angles. F r o m the intensities o f the C D b a n d s at 212 n m where we expect the m a i n c o n t r i b u t i o n o f the imidazole c h r o m o p h o r e , we find the strongest signal for H i s - G l y at p H 4.5. Thus the spectral changes are in full agreement with the N M R results showing that a change in p r o t o n a t i o n state o f either o f the groups l m H + -+ Im, C O O - - 7 C O O H destroys the p r o p o s e d interaction and leads to radical changes in the r o t a m e r p o p u lations. CONCLUSION Thus, even if r o t a m e r p o p u l a t i o n s o f the histidine side chain are not the only factors influencing the intensities o f the C D b a n d at 212 nm, the variations o f N M R p a r a m e t e r s such as coupling constants, nonequivalence and a m p l i t u d e s o f chemical shift variations can be c o r r e l a t e d well with the changes in the C D spectra as reflected in Fig. 5. It a p p e a r s clearly t h a t the restriction o f r o t a t i o n a r o u n d the b o n d s o f this side chain, which is magnetically a n i s o t r o p i c as well as an ultraviolet c h r o m o p h o r e , constitutes the m a i n factor influencing nonequivalence and optical activity. In these experiments, u n f o r t u n a t e l y , we did n o t e n c o u n t e r peptides whose side chain is stabilized in one single c o n f o r m a t i o n ; this w o u l d have allowed us to better correlate o u r results and p e r h a p s even to quantify them. Studies, however, on several peptides c o n t a i n i n g a r o m a t i c side chains with restricted r o t a t i o n a r o u n d the bonds, are in progress. ACKNOWLEDGMENT W e are grateful to Dr. J. Thi6ry for valuable discussions. REFERENCES 1 Haar, W., Fermandjian, S., Vicar, J., Blaha, K. and Fromageot, P. (1975) Prcc. l~'atl. Acad. Sci. U.S. 72, 4948-4952 2 Fermandjian, S., Lintner, K., Haar, W., Fromageot, P., Khosla, M. C., SnTeby, R. R. and Bumpus, F. M. (1976) in: Proc. 14 Eur. Peptide Symp. (Loffet, A., ed.), Editions de l'Universit6 de Bruxelles, Belgium 3 Greff, D., Toma, F., Fermandjian, S., L6w, M. and Kisfaludy, L. (1976) Biochim. Biophys. Acta 439, 219-231 4 McDonald, C. C. and Phillips, W. D. (1963) J. Am. Chem. Soc. 85, 3736-3742 5 Weinkam, R. J. and Jorgensen, E. C. (1973) J. Am. Chem. Soc. 95, 6084-6090 6 Bothner-By, A. A. and Castellano, S. M. (1968) in: Computer Programs for Chemistry (DeTar, D. F., ed.), Vol. 1, pp. 10-53, W. A. Benjamin, New York 7 Pachler, K. G. R. (1964) Spectrochim. Acta 20, 581-587 8 Feeney, J. (1975) Proc. R. Soc. London A 345, 61-72 9 Fermandjian, S. (1975) in: Abstracts of Soviet-French Symposium on the Physical Chemistry of Proteins and Peptides, Pushchino-on-Oka, Sept. 9-12 10 Kainosho, M. and Ajisaka, K. (1975) J. Am. Chem. Soc. 97, 5630-5631 11 Kopple, K. D., Wiley, G. R. and Tauke, R. (1973) Biopolymers 12, 627-636 12 Kopple, K. D. and Ohnishi, M. (1969) J. Am. Chem. Soc. 91,962-970 13 Reynolds, W. F., Peat, I. R., Freedman, M, H. and Lyerla, J. R. Jr. (1973) J. Am. Chem. Soc. 95, 328-331 14 Nakamura, A. and Jardetzky, O. (1967) Proc. Natl. Acad. Sci. U.S. 58, 2212-2219 15 Grebow, P. E. and Hooker, Jr., T. M. (1975) Biopolymers 14, 871-881 16 Grebow, P. E. and Hooker, Jr., T. M. (1975) Biopolymers 14, 1863-1883