l-histidine-containing peptides as models for the ... · histidine and...

THE JOURNAL OF BIOLOQICAL CHEMISTRY Vol. 241, No. 5, Issue of March 10, 1966

Printed in U.S.A.

L-Histidine-containing Peptides as Models for the Interaction of Copper (II) and Nickel (II) I ons with Sperm Whale Apomyoglobin*

(Received for publication, September 7, 1965)

GRAEME F. BRYCE,~ ROGER W. ROESKE,~ AND FRANK R. N. GURD

From the Department of Biochemistry, Indiana University School of Medicine, Indianapolis, Indiana 46207

SUMMARY

1. The association and ionization constants for nickel (II) and a selection of L-histidine-containing peptides have been computed together with the constants for copper (II) and acetylglycylglycyl-L-histidylglycine.

2. A close parallel between the titration behavior of copper (II) apomyoglobin complexes and model peptides has been obtained on the assumption that 1 of the 4 bound metal ions is situated at the NHz-terminal portion of the polypeptide chain.

3. Visible absorption spectra of nickel (II) and copper (II) apomyoglobin complexes containing 4 metal ions per mole have been measured and found to be consistent with the binding of 1 metal ion at the NHr-terminal locus and 3-metal ions in the interior of the peptide chain at histidine loci.

dichroism spectra of complexes of copper (II) and nickel (II) ions with sperm whale apomyoglobin and selected n-histidine- containing peptides. In particular, it deals with acetylglycylglycyl-n-histidylglycine, which may be regarded as perhaps a better model than those considered hitherto for the histidine environment in a polypeptide chain. As an adjunct to the present study, the titration behavior of the complexes of nickel (II) and n-histidine-containing peptides has been investigated together with the copper (II) complex of acetylglycylglycyl-n- histidylglycine which extends work of this type already reported (4, 5).

EXPERIMENTAL PROCEDURE

4. Optical rotatory dispersion and circular dichroism spectra in the visible region have been measured for both the protein and the peptide complexes. Strict similarities were not, in general, found and the various explanations for this failure are discussed.

Physical Chemical fMeasurements-Titrations were carried out as previously described (4, 6). The spectropolarimetry with the Bendix Polarmatic instrument and spectrophotometry with the Cary model 14 spectrophotometer were normally applied to the same solution under conditions described previously (5, 6). Results are reported in terms of molar absorbance e and molecular rotation

Nickel (II) ions have been shown to promote the ionization of protons from peptide bonds in complexes of simple dipeptides and polyglycines (2) and from complexes of histidine-containing dipeptides (3). Previous communications (4, 5) demonstrated the adequacy of peptides such as acetylglycylglycyl-n-histidine as models for the interaction of copper (II) ions with sperm whale metmyoglobin and apomyoglobin as judged by titration studies. The present communication deals with comparisons of the visible absorption spectra, optical rotatory dispersion,, and circular

* This is the 20th paper in a series dealing with coordination complexes and catalytic properties of proteins and related sub- stances. See Reference 1. The work was supported by United States Public Health Service Research Grants HE-05556, HE- 06308, and GM-10591.

t Present address, Department of Molecular Biophysics, Yale University, New Haven, Connecticut.

$ Research Career Development Awardee of the United States Public Health Service.

in which (01) = (observed rotation in degrees/&), d is the path length in decimeters, c is the concentration in grams per ml and M is the molecular weight. A light path of 10 mm or 1 mm was generally used with concentrations of the complexes in the 0.01 M range. Rotations were measured on samples with an absorbance in the visible region of generally less than 0.2. Meas- urements of circular dichroism spectra were made on the Rous- se1Jouan instrument kindly made available by Dr. T. S. Piper of the University of Illinois. Results are quoted in terms of EL - en, the difference in molar absorbance between the left and right circularly polarized beams.

The apomyoglobin complexes were studied at about 4 x 1O-4 M metal ion concentration in a 100~mm cell for the spectrophotometry, a 50-mm cell for the circular dichroism, and a 20-mm cell for the spectropolarimetry.

Preparation of Sperm Whale Apomyoglobin-The protein was prepared from metmyoglobin by the modified method of Bres- low (7). A further modification was introduced by diluting the metmyoglobin solution to a concentration of about 2 mg per

1072

by guest on August 25, 2020

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Issue of March 10, 1966 Bryce, Roe&e, and Gurd 1073

ml during the extraction with Z-butanone. The resulting pro- filtered, dried in a vacuum, dissolved in water (25 ml), and tein solution was completely colorless after two extractions. relyophilized to a white hygroscopic powder. Rr 0.06 in Solvent

Preparation of Apomyoglobin Complexes-Copper (II) and A and 0.22 in Solvent B. nickel (II) complexes (4: 1) of apomyoglobin were prepared by taking the protein solution to above pH 11 and adding the metal CI&GNGOB.H~O

ion as the chloride. The pH was then readjusted to 11 over a Calculated: C 43.52, H 5.74, N 21.75 period of a few hours. No added electrolyte was present. The Found : C 43.83, H 5.85, N 21.13

final protein concentration was 0.107 mM. Synthesis of Acetylglycylglycyl-L-histidylglycine-In the fol-

RESULTS AND DISCUSSION

lowing procedures, the RF values refer to thin layer chroma- Titration Studies-Table I lists the formation and ionization

tograms with the use of Silica Gel G (Merck, Darmstadt). constants for 1: 1 complexes of nickel (II) and selected L-histidine-

Solvent System A is I-butanol-acetic acid-water (4:l :l, by containing peptides together with the results for copper (II)

volume) ; Solvent System B is I-propanol-water (2: 1, by volume). The chromatograms were developed by the hypochlo- TABLE I

rite method (8). Formation and ionization constants for 1:1 complexes of nickel N-tert-Butoxycarbonylglycyl-~-iminobenzylhistidylglycine Benxyl (II) and r,-histidine-containing peptides at ~25~ and 0.16

Ester (Compokd I)-A suspension of L-iminobenzylhistidylglycine benzyl ester dihydrobromide (4) (2.34 g) in chloroform (50 ml) was treated with triethylamine (1.2 ml) and N-tert- butoxycarbonylglycine p-nitrophenyl ester (4) (1.24 g) and allowed to stand at room temperature overnight. The solution was diluted with chloroform (300 ml), washed with cold 0.1 N

sodium hydroxide solution (4 X 5 ml), and dried over sodium sulfate. Evaporation of the solvent in a vacuum left a syrup which crystallized from ethanol-water. After two recrystallizations, the product (1.39 g, 60%) melted at 132.5-134.5”. RF 0.75 in Solvent B and 0.50 in Solvent A.

C~G&NSOB Calculated: C 63.37, H 6.42, N 12.74 Found : C 63.67, H 6.36, N 12.26

Acetylglycylglycyl - L - iminobenzylhistidylglycine Benzyl E.&t (Compound II)-Compound I (2.18 g) was added to a 1 N solution of HCl in acetic acid (20 ml) and allowed to stand for 30 min. The dihydrochloride was precipitated by the addition of ether (100 ml), filtered, dissolved in a mixture of water (40 ml) and chloroform (150 ml), and converted to the free base by the addition of 5% aqueous NaHC03 (75 ml). The residue obtained by evaporating the chloroform layer was dissolved in dimethylformamide (10 ml), treated with acetylglycine p- nitrophenyl ester (0.95 g), and allowed to stand for 18 hours. Ethyl acetate (100 ml) was added, and the precipitated product was filtered and washed with pH 9 carbonate buffer (50 ml), then with water (20 ml), and allowed to dry in the air. After three recrystallizations from 95% ethanol, the product (0.68 g, 30%) melted at 193-195”. RF 0.60 in Solvent B.

Peptide

Glycyl-L- histi- dinet.. .

Glycyl- dYCYl- L-histidine..

Glycyl- dYCYl- dYCYl- L-histidine.....

Acetylgly- CYklY- cyl-L- histidine

Acetylgly- CYklY- cyl-L- histidylglycine.

Acetylgly- CYMY- cyl-L- histidylglycine + cop-

per (II).

ionic strength*

c. E

M a 9 3 I-1

2.66 6.77 8.24 NW ND 6.10 6.70 9.25

2.84 6.87 8.22 ND ND 6.20 6.30 6.35

3.02 6.85 8.11 ND ND 8.50 8.75 9.10

3.08

3.29

3.29

7.18 2.94

2.93

3.95

2.30 8.65 8.95 9.20

6.90 8.50 8.60

6.90 3.80

8.40

5.95 6.45 9.00

Calculated: C 61.30, H 5.88, N 15.32 Found : C 61.04, H 5.97, N 15.02

Acetylglycylglycyl-L-histidylglycine (Compound III)-A solution of Compound II (600 mg) in liquid ammonia (150 ml) was treated with small pieces of sodium until a deep blue color persisted for 5 min. Dowex 5OW-X2 on the ammonium cycle (6 g) was added, and the mixture was stirred for 30 min and allowed to evaporate. The residue was extracted with deionized water (2 x 50 ml), and the pH (7.5) was adjusted to 4.8 by the addition of a few drops of 1 N HBr. The solution was lyophilized and the product was dissolved in water (9 ml) and precipitated as a gel by the addition of acetone (50 ml). The product was

* Equilibrium constants are defined as follows (4) :

cu2 + L- * cuL+ [cuL+l

cuL+ + L- * CULZ

K1 = w&H;-1

K2 = [cuL+] [L-l

cuL+ * CuL + H+ K = [CuLl [H+l

c [CuLfl

CUL = CuL- + H+ K , c = [CuL-1 tH+l [CULI

CuL- * CuL*- + H+ K ,,

c = lCuLa-1 lH+l

[CuL-I t Taken from Martin and Edsall (3). $ Not determined.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

1074 Copper and Nickel Complexes of Sperm Whale Myoglobin Vol. 241, No. 5

and acetylglycylglycyl-L-histidylglycine. No attempt was made to compute formation constants for peptides containing a free a-amino group. For the acetylated peptides, formation constants were computed by methods outlined previously (1). The pK, values for nickel (II) complexes were computed by the method used for the analysis of nickel (II) and triglycylglycine (2) and should be regarded with a degree of uncertainty.

The log K1 values for nickel (II) and acetylglycylglycyl+ histidine and acetylglycylglycyl-L-histidylglycine are comparable to those found for the formation constants for nickel (II)- imidazole complexes (9). The log K1 value for copper (II) and acetylglycylglycyl-L-histidylglycine of 3.95 agrees favorably with the value of 3.89, predicted from the correlation with the pK of the imidazole group found from previous studies (4).

The pK, values of nickel (II) and glycylglycyl-n-histidine exhibit the same extensive overlapping that was found for the copper (II) complexes resulting from the cooperative effect of initial association at both the or-amino group and imidazole nitrogen atom. However, the pK, values for the interaction of nickel (II) with the acetylated peptides and glycylglycylglycyl-n-histidine also show this overlapping, which can be attributed in this case to a change in stereochemistry from octahedral to square planar, the latter form being characterized by an intense yellow color (2, 3).

For the copper (II) complex of acetylglycylglycyl-L-histidylglycine, the predicted pK, value is 6.35 (4), which is significantly larger than the observed value of 5.95. The pK,’ value of 6.45 is also about 0.7 log unit larger than the expected value, but the pKZ value of 9.00 is comparable to others found in similar peptides (4). It is not entirely clear from structural considerations why these pK, values should differ by such an amount. Sub- stitution of a glycyl residue on the L-histidyl carboxyl group would not be expected to affect the chelation of the copper (II) to the peptide bonds on the NHz-terminal side of the n-histidine residue.

In general, the pK, values for the nickel (II) complexes were about 2 log units higher than for the copper (II) complexes. A comparison of the titration curves of nickel (II) and copper (II)

=‘I

0 ; I ; 4

MOLES NaOH REACTED PER METAL ION

FIN. 1. Titration of equimolar mixtures of copper (II) and nickel (II) and acetylglycylglycyl-n-histidylglycine with sodium hydroxide. The ionic strength was 0.16 and the temperature 25 f 0.1”.

TABLE II Visible absorption characteristics of 1:l copper (II) and nickel

(II) complexes of L-histidine-containing peptides and 4:l complexes of apomyoglobin at pH 11 .O

Peptide Metal

Glycyl-L-histidine. Glycylglycyl-n-histidine Glycylglycylglycyl-n-histi-

Ni (II) Ni (II)

dine....................... Acetylglycylglycyl-r-histidine Acetylglycylglycyl-L-histidyl-

glycine Acetylglycylglycyl-n-histidyl-

glycine Tetraglycylglycine Tetraglycylglycine Apomyoglobin Apomyoglobin Tri-n-alanine + copper (II). Tri-n-alanine + nickel (II). .

Ni (II) 430 Ni (II) 425

Ni (II) 425

cu (II) 555 Ni (II) 412 cu (II) 510 Ni (II) 425 cu (II) 530 cu (II) 555 Ni (II) 426

- -

+w

450 425

e max

u-1

480 62

480 64 485 83

485 85

485 100

-

with acetylglycylglycyl-L-histidylglycine is shown in Fig. 1. The overlapping pK, values, noted in Table I, are reflected in the extreme flatness of the titration curve of the nickel (II) complexes between 1 and 4 equivalents of NaOH added.

Visi6Ze Spectral PropertiesTable II lists the visible absorption characteristics of nickel (II) complexes of n-histidine-containing peptides together with the copper (II) complex of acetylglycylglycyl-L-histidylglycine, all at pH 11. In effect these represent the spectra of the MV- species for all except glycyl+ histidine, where the NiL- species is represented, according to the nomenclature used previously (5). Also included are the results for the copper (II) and nickel (II) complexes of apomyoglobin and tetraglycylglycine at the same pH. AL,, refers to the position of a second absorption band on the long wave length side of the major peak in the nickel (II) complexes. For the complex of glycyl-n-histidine, this band is absent. For glycylglycyl-n-histidine and glycylglycylglycyl-L-histidin?, it is a poorly defined shoulder; hence, the corresponding emax values must be regarded as only approximate. But for the acetylated peptides the spectra are readily resolved into contributions from two absorption bands of comparable strength. As discussed further below, this latter observation is true to a slightly lesser extent for the apomyoglobin complex.

Resolution of the spectra of intermediate nickel (II) complexes was not attempted. Only the square planar Nip- species possesses any significant color, whereas the octahedral species are pale blue and would be difficult to resolve accurately because of the uncertainty in the ionization constants.

Copper (II) and Nickel (II)-Apomyoglobin Complexes-The visible absorption spectrum of a 4:l copper (II)-apomyoglobin complex at pH 11 is shown in Fig. 2, Curve 1; E is expressed per metal ion. The single peak has a X,,, of 530 rnp which is exactly similar to the X,,, found for the complex at pH 10 by Breslow (10). However, the higher E,,, value of 112 compared to a value of 98 found by Breslow (10) can be accounted for partially by the narrower peak. Fig. 2 also includes the spectrum of the CuL+ species of acetylglycylglycyl-L-histidylglycine with a X,, of 555 rnp and emax of 90 (Curtie 2). I f one considers


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


120

80

W

40

0 I I I I I I

400 500 600 700

1 (ml

FIG. 2. Visible absorption spectra of: Curve 1, a 4:l complex of copper (II) and apomyoglobin at pH 11.0; Curve 2, the CuL2- species of the copper (II) acetylglycylglycyl-L-histidylglycine complex; Curve 9, the computed spectrum of a mixture of 1 mole of the CuLz- species of the copper (II)-tetraglycylglycine complex and 3 moles of the CuLZ- species of the copper (II) acetylglycylglycyl-n-histidylglycine complex. E values are expressed per copper (II).

the X,,, to be the more important parameter, this observed difference is quite significant. Consequently, the environment of the copper (II) ion in the protein complex cannot be simulated exactly by a model peptide of the kind studied. However, the work of Breslow (10) offered no evidence that the NHz-terminal portion of the molecule was not a chelation locus, and indeed there is no a priori reason for its exclusion. This region of the molecule has been implicated in the binding of the single copper (II) ion to bovine serum albumin (10) and a logical approach to the present case would be to suggest that one of the 4 copper (II) ions is chelated to the NHS-terminal residues and the remain- ing three to histidine loci in the interior of the peptide chain. A model for this system would thus be a combination of 1 mole of the CuL2- species of tetraglycylglycine, X,, 510 rnp, ernSx 147l and 3 moles of the CuL2- species of acetylglycylglycyl-n-histidylglycine, imax 555 w, emax 90. The resultant spectrum, shown in Fig. 2, Curve S, is seen to have a X,, of 530 rnp and an emax of 98, resembling much more closely the spectrum of the protein complex, particularly with respect to the more relevant parameter, X,,,.

The available evidence is that strong binding of copper (II) ions to metmyoglobin levels off at 7 per molecule which should apply approximately to apomyoglobin (10, 11). Hence, ceteris paribus, with 4 metal ions bound, the probability of there being one at the NH2-terminal site is only one-third of that for binding at histidine loci. However, at this pH there is no element of competition by protons for the basic form of the free a-amino group or the imidazole nitrogen so that the distribution of copper (II) between these two sites should be in favor of the a-amino

1 G. F. Bryce, unpublished observation.

group by virtue of a higher association constant. Log KI for copper (II) triglycylglycine is 4.93 (12), while the value for acetylated n-histidine peptides is about 4.0. Thus, it is not unreasonable to consider the copper (II) complex of apomyoglobin as made up of a population of molecules with the pre- dominant species carrying 1 metal ion at the NH2 terminus and 3 metal ions somewhere in the interior of the peptide chain at histidine loci.

Further evidence of this type is presented in Fig. 3, which shows the analogous spectra for nickel (II) complexes. The spectra in this case have the added advantage for interpretation of a second absorption band, albeit overlapping with the major peak. This second, long wave length absorption is not present in the NiL2- species of tetraglycylglycine. Hence, the presence of this type of complex in the protein would be expected to reduce the absorbance at 485 rnE.1 compared to that at 425 rnp. This is found to be the case. Examination of Fig. 3 reveals that the computed spectrum closely parallels the protein complex spectrum in profile. All three spectra have maxima at 425 rnE.1 and, as closely as can be determined, also at 485 rnp. However, the ratios of absorbance at these two wave lengths are not the same. For the protein complex, Curve 1, and the computed spectrum, Curve 8, the ratios of absorbance at 425 rnp to absorbance at 485 rnp are 1.61 and 1.64, respectively. The ratio of 1.13 for the acetylglycylglycyl-n-histidylglycine complex, Curve %‘, rules out this peptide as being typical of all the binding sites.

160'

120

w a0

40

II 400 500

X (mfi)

600

FIG. 3. Visible absorption spectra of: Curve 1, a 4:l complex of nickel (II) and apomyoglobin at pH 11.0; Curve 2, the NiL2- species of the nickel (II)-acetylglycylglycyl-L-histidylglycine complex; Curve 9, the computed spectrum analogous to Curve .!? in Fig. 2.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

1076 Copper and Nickel Complexes of Sperm Whale Myoglobin Vol. 241, No. 5

TABLE III Values of (-a?&~) as function of pH for copper (II) complexes

of various peptides, of certain mixtures of peptides, and of metmyoglobin and apomyoglobin

The anionic form of the peptide is taken as the starting point.

Compound

Triglycylgly- tine.

Tetraglycylgly- tine. . . . . .

Acetylglycylgly- cylglycyl-n-histidine. . . . . . .

Acetylglycylgly- cyl-L-histidine

Acetylglycylgly- cyl-n-histidylgly-

tine . . . . . Tetraglycylgly- : tine + three

complexes of acetylglycylgly- cylglycyl-n-histidine. . . .

Tetraglycylgly- tine + three complexes of acetylglycylglycyl-L-histidylglycine . . . . .

Triglycylgly- tine + three complexes of acetylglycylgly- cylglycyl-n-histidine. . . . .

Triglycylgly- tine + three complexes of acetylglyoylgly- cyl-n-histidylglycine . . . . . .

Triglycylgly- tine + three complexes of acetylglycylglycyl-n-histidine

Tetraglycylgly- tine + three complexes of acetylglycylgly- cylglycyl-n-histidine. . . . . .

Metmyoglobin . . Apomyoglobin . .

- 6.50 7.00 7.50 8.50 9.20 11.00

1.12 1.52 1.79 2.12 2.41 2.79 3.00

1.16 1.79 2.14 2.77 3.00 3.00 3.00

0.56 1.09 1.53 2.32 2.69 2.94 3.00

0.50 1.08 1.56 2.07 2.46 2.85 2.98

1.16 1.79 1.95 2.28 2.65 2.98

0.71 1.25 1.68 2.43 2.77 2.96

1.16 1.77 2.00 2.40 2.74 2.99

0.70 1.20 1.60 2.27 2.62 2.90

1.15 1.72 1.91 2.24 2.59 2.91

0.66

0.67 1.25 1.38

1.19

1.24 1.50 1.44

1.62 2.08 2.45 2.84

1.71 2.25 2.60 2.89 1.56 2.12 2.48 2.50 1.56 2.17 2.60 2.86

3.00

3.00

3.00

3.00

3.00

3.00

3.00 3.15 3.47

value Of (-&&,I) at pH

These two sets of evidence provide strong indications for the nature of the chelation loci for copper (II) and nickel (II) in apomyoglobin.

Comparison of Titration Behavior of Complexes of Copper (II) with Apomyoglobin and Model Peptide--It was noted in an earlier

communication (4) that the titration behavior of the copper (II) complex of apomyoglobin could be paralleled by complexes of acetylated histidine peptides as judged by variation with pH of ( -AT~,/F~), the average number of protons liberated per metal ion bound. The agreement was good above pH 7.5, but there was a deficit of protons displaced from the model peptide complexes at lower pH values. Now, if account is taken of binding to the NH&erminal portion of the molecule then the most accurate model should contain the contribution of one peptide not containing histidine. Data are available for triglycylglycine (12), tetraglycylglycine,l and a selection of acetylated L-histidine peptides described here or in a previous study (4). The values of (- AF,/F& computed for “mixtures” representing 1 copper (II) at the NH2 terminus and 3 at histidine loci are presented in Table III in comparison with the observed values for metmyoglobin and apomyoglobin. Fig. 4 is a graphi- cal representation of the data for the three best fitting models and for apomyoglobin. It is evident that a close parallel throughout the entire pH range is not obtained from any of the models. However, for reasons discussed elsewhere (4), exact similarity is not to be expected because of difficulties in defining the environment in the protein. The discrepancy is partially alleviated by inclusion of either a triglycylglycine (Curve 2)

11.0 --

10.0 --

“,I s*“--

s.o--

7.0--

I I I I I I I I

1.0 2.0 3.0

- AT~sH VM

FIG. 4. Values of (-AIH/GM) as a function of pH. Curve 1, copper (II)-apomyoglobin with 4 copper (II) ions bound; Curve d, computed curve for a mixture of 1 mole of 1: 1 copper (II)-triglycylglycine and 3 moles of 1:1 copper (II)-acetylglycylglyoyl-n- histidine; Curve 5, computed curve for a mixture of 1 mole of 1:l copper (II)-tetraglycylglycine and 3 moles of 1:l copper (II)- acetylglycylglycyl-n-histidine; Curve 4, computed curve for a mixture of 1 mole of 1: 1 copper (II)-tetraglycylglycine and 3 moles of 1:l copper (II)-acetylglycylglycyl-n-histidylglycine.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


or tetraglycylglycine (CUUJ~ S) complex but is nevertheless significant. The unusual titration properties of the complex of acetylglycylglycyl-L-histidylglycine (Curve 4) almost match the protein at pH 6.50, but an overcompensation at pH 7 to 8 results in excessively large values for ( -A?J~,/?,). The best over-all fit is obtained from three complexes of the acetylglycylglycyl-L-histidine type and one of either triglycylglycine or tetraglycylglycine (Curves d and 3).

One very important factor so far has been omitted from con- sideration, the influence of the side chains on the ionization equilibria. The bulky aliphatic side chains of valine and leucine cause an increase in the pK, value in copper (II) complexes of glycyl-L-valine (13) and glycyl-L-leucine (14) to 4.75 and 4.80, respectively, compared to 4.25 for glycylglycine. This could be the result of the expenditure of energy in the reorganization of the hydrotactoid environment of these side chains in free solution. No data are available for similar effects on pK, values in larger peptides. It is to be anticipated that the same process would operate for aliphatic residues on the NHz-terminal side of an L-histidine residue to raise the pK, values. Little can be said concerning the effect of other side chains but, as noted previously (4), structural factors could be present in the protein which would promote or favor apposition of the first peptide bond to the copper (II), thereby lowering the pK, value. A much greater “proximity factor” would be required to cause the same reduction in pK,‘. A more definitive description must await a study of n-histidine peptides representing sequences of myoglobin.

Optical Rotatory Dispersion and Circular Diehroism Xpectra- Complexes such as the CuL2- species of acetylglycylglycyl-n- histidine exhibit well defined Cotton effects around the visible absorption band (5), and it would not be unreasonable to expect the chelation loci in copper (II) apomyoglobin complexes to behave in a similar manner. However, from the foregoing discussion, it is apparent that account must be taken of the rotatory properties of a complex representing the NH2 terminus. A suitable model would be tri-L-alanyl-n-alanine. We have employed here di-n-alanyl-n-alanine, the properties of which should not be too dissimilar and from which it should be fairly easy to extrapolate to the more representative peptide. Anoma- lous dispersion in the NHz-terminal complex of apomyoglobin should be greater in magnitude and displaced towards shorter wave lengths than those of di-n-alanyl-n-alanine complexes.2

Fig. 5 illustrates the ORD3 spectra of the CuL- species of di-n-alanyl-L-alanine and the CuL’- species of acetylglycylglycyl-L-histidine (A) and the CD spectra of the same (B). The CULT- species of the latter peptide complex is the form existing at pH 11, and the CuL- species of di-L-alanyl-n-alanine represents maximum involvement of peptide nitrogen atoms without the complication of ionizations from coordinated water molecules. The negative ORD Cotton effect of this complex centered at 570 rnp is reproduced as a negative CD Cotton effect at 565 mM. The absorption spectrum (Table II) consisted of a single peak at 555 rnp. By extrapolation to tri-n-alanyl-n- alanine, we may predict that the Cotton effect for this complex would also be negative, centered at about 520 to 530 rnp and perhaps of twice the magnitude. It was suggested earlier (5)

2 Based on a trend observed with other simple peptides (G. F. Bryce and F. R. N. Gurd, in preparation).

a The abbreviations used are : ORD. ontical rotatorv disnersion: I _ ” a

and CD, circular dichroism.

that the ORD spectrum of the CUP- species of acetylglycylglycyl-n-histidine was the resultant of two Cotton effects of opposite sign located so that the positive extrema overlapped. The CD spectrum in Fig. 5B confirms this interpretation where it is seen that there is a negat,ive Cotton effect centered at 590 rnp and a positive one at 500 rnp. The crossover points in the ORD spectrum are 600 m,u and 490 mp, respectively. The slightly skew shape of this CD curve indicates partial overlap so that the exact wave lengths of the transitions cannot be determined. It is apparent that they would not differ by more than 5 rnw from the wave lengths quoted above. It should be men- tioned here that the CD spectrum of copper (II)-acetylglycylglycyl-L-histidylglycine was identical with the above in the X,,, values and only slightly different in the magnitudes of EL - CR.

Fig. 6A and B illustrates the ORD and CD spectra of the analogous nickel (II) complexes. The CD Cotton effects of the nickel (II)-acetylglycylglycyl-L-histidine complex appear at 420 rnp and 485 rnp, which are very close to the estimated positions of the resolved partial absorption bands (Table II) and not very different from the crossover points in the ORD

I I I I I I

+4000--

-lOOO--

A I I I

I I I I 360 400 500 600

X (mp)

u.7 I

w”

I I I I +1.20-- +1.20--

eO.80 -- CO.80 --

+0.40-- +0.40--

0 0

- 0.40-- - 0.40--

- 0.00 -- - 0.00 --

- 1.20-- - 1.20-- B B

‘ ‘ I I I I I I I I I I I I

300 300 400 400 500 500 600 600

x (mfl)

FIG. 5. A, optical rotatory dispersion spectra: --, CuL2- species of copper (II)-acetylglycylglycyl-L-histidine; - - -, CuL- species of copper (II)-di-n-alanyl-n-alanine. B, circular clichro- ism spectra; nomenclature as in A.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/


or tetraglycylglycine (Curve S) complex but is nevertheless significant. The unusual titration properties of the complex of acetylglycylglycyl-L-histidylglycine (Curve 4) almost match the protein at pH 6.50, but an overcompensation at pH 7 to 8 results in excessively large values for (-A~n/ii,). The best over-all fit is obtained from three complexes of the acetylglycylglycyl-L-histidine type and one of either triglycylglycine or tetraglycylglycine (Curves d and S).

One very important factor so far has been omitted from con- sideration, the influence of the side chains on the ionization equilibria. The bulky aliphatic side chains of valine and leucine cause an increase in the pK, value in copper (II) complexes of glycyl-n-valine (13) and glycyl-L-leucine (14) to 4.75 and 4.80, respectively, compared to 4.25 for glycylglycine. This could be the result of the expenditure of energy in the reorganization of the hydrotactoid environment of these side chains in free solution. No data are available for similar effects on pK, values in larger peptides. It is to be anticipated that the same process would operate for aliphatic residues on the NHz-terminal side of an L-histidine residue to raise the pK, values. Little can be said concerning the effect of other side chains but, as noted previously (4), structural factors could be present in the protein which would promote or favor apposition of the first peptide bond to the copper (II), thereby lowering the pK, value. A much greater “proximity factor” would be required to cause the same reduction in pK,‘. A more definitive description must await a study of L-histidine peptides representing sequences of myoglobin.

Optical Rotatory Dispersion and Circular Dichroism Spectra- Complexes such as the CuL2- species of acetylglycylglycyl-L- histidine exhibit well defined Cotton effects around the visible absorption band (5), and it would not be unreasonable to expect the chelation loci in copper (II) apomyoglobin complexes to behave in a similar manner. However, from the foregoing discussion, it is apparent that account must be taken of the rotatory properties of a complex representing the NH2 terminus. A suitable model would be tri-L-alanyl-n-alanine. We have employed here di-L-alanyl-L-alanine, the properties of which should not be too dissimilar and from which it should be fairly easy to extrapolate to the more representative peptide. Anoma- lous dispersion in the NH2-terminal complex of apomyoglobin should be greater in magnitude and displaced towards shorter wave lengths than those of di-n-alanyl-L-alanine complexes.2

Fig. 5 illustrates the ORD3 spectra of the CuL- species of di-L-alanyl-n-alanine and the CuL” species of acetylglycylglycyl-L-histidine (A) and the CD spectra of the same (B). The CuU- species of the latter peptide complex is the form existing at pH 11, and the CuL- species of di-n-alanyl-L-alanine represents maximum involvement of peptide nitrogen atoms without the complication of ionizations from coordinated water molecules. The negative ORD Cotton effect of this complex centered at 570 rnp is reproduced as a negative CD Cotton effect at 565 rnp. The absorption spectrum (Table II) consisted of a single peak at 555 rnp. By extrapolation to tri-n-alanyl-n- alanine, we may predict that the Cotton effect for this complex would also be negative, centered at about 520 to 530 rnp and perhaps of twice the magnitude. It was suggested earlier (5)

2 Based on a trend observed with other simple peptides (G. F. Bryce and F. R. N. Gurd, in preparation).

3 The abbreviations used are: ORD, optical rotatory dispersion; and CD, circular dichroism.

that the ORD spectrum of the CUP- species of acetylglycylglycyl-L-histidine was the resultant of two Cotton effects of opposite sign located so that the positive extrema overlapped. The CD spectrum in Fig. 5B confirms this interpretation where it is seen that there is a negative Cotton effect centered at 590 rnp and a positive one at 500 mp. The crossover points in the ORD spectrum are 600 ml and 490 rnw, respectively. The slightly skew shape of this CD curve indicates partial overlap so that the exact wave lengths of the transitions cannot be determined. It is apparent that they would not differ by more than 5 rnp from the wave lengths quoted above. It should be men- tioned here that the CD spectrum of copper (II)-acetylglycylglycyl-L-histidylglycine was identical with the above in the X,,, values and only slightly different in the magnitudes of CL - CR.

Fig. 6A and B illustrates the ORD and CD spectra of the analogous nickel (II) complexes. The CD Cotton effects of the nickel (II)-acetylglycylglycyl-L-histidine complex appear at 420 mp and 485 rnp, which are very close to the estimated positions of the resolved partial absorption bands (Table II) and not very different from the crossover points in the ORD

1 I I I I I

+4000--

-lOOO-- A

I 1 I I I I I

360 400 500 600 X(w)

w”; I

w”

, , I I I I +1.20-- +1.20--

+0.80 -- +0.80 --

+0.40-- +0.40--

0 0

- 0.40-- - 0.40--

- 0.80 -- - 0.80 --

- 1.20 - 1.20 -- -- B B

L L I I 1 1 1 1 I I I I I I I I

300 300 400 400 500 500 600 600

x (Iw)

FIG. 5. A, optical rotatory dispersion spectra: -, CuLz- species of copper (II)-acetylglycylglycyl-L-histidine; - - -, CuL- species of copper (II)-di-L-alanyl-L-alanine. B, circular clichro- ism spectra; nomenclature as in A.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Issue of March 10, 1966 Bryce, Roeske, and Gurd 1079

plex curves shown in Fig. 6B provided allowance is made for a much larger contribution from the di-L-alanyl-L-alanine complex, based on the premise that the tri-L-alanyl+alanine would have higher eL - eR values.

Conjiguration about Copper (II) and Nickel (II) Ions in Apo- myoglobin ComplexeesIt is pertinent to the discussion at this stage to examine certain properties of the model peptides to establish just what can be expected from the ORD and CD of the protein complexes. In a previous communication (5) it, was suggested that the origin of the optical activity in the visible region of copper (II) complexes of acetylated L-histidine peptides was due to asymmetry induced around the metal ion by a partially tetrahedral array of the ligand nitrogen atoms. The 4 nearest ligand atoms in several related crystalline copper (II) complexes have been found to be not exactly coplanar (15). Now, since this configuration appeared to be controlled by the conformation of the chelate ring containing the L-histidine residue, it was necessary to postulate an interaction between the carbonyl oxygen of the L-histidine carboxyl group (or of the peptide bond linking the carboxyl-terminal glycine as in acetylglycylglycyl-L-histidylglycine) and the metal ion. From a study of molecular models, the interaction can occur with the carboxyl group in an axial orientation. An equatorial orientation leads to the alternative conformation of the six membered ring and an effectively inverted configuration around the metal ion.4 The extent of distortion of the square plane into a flattened tetrahedral arrangement need not be exactly comparable in the two alternative configurations. In the interior of a peptide chain, the multitude of inter-side chain interactions and other factors stabilizing the protein conformation probably impose severe constraint on the mobility of the polypeptide components and may limit the extent of formation of each of the puckered forms of the ring.

If we are to entertain the possibility of an NH&err&al binding site, an explanation for the optical activity in copper (II) and nickel (II) di-L-alanyl-L-alanine complexes must be sought. In this case, no simple interaction is present in the complex which would produce configurational dissymmetry. Preliminary studies on tripeptides made up of glycine with one L-alanyl residue in each of the three positions indicates that the magnitudes of the Cotton effects are dependent on which chelate ring contains the asymmetrical center. These Cotton effects are all of the same sign so that we are forced to conclude that

4 This is not the only possible bonding arrangement giving rise to configurational dissymmetry. The four corners of the distorted square plane could be occupied by the 3 peptide nitrogen atoms and the carbonyl or carboxyl oxygen atom. (Preliminary electron spin resonance studies appear to exclude 5-fold coordination with nitrogen atoms.) Binding of the imidazole side chain, through the l-nitrogen atom, in an apical position would produce a distorted square pyramidal configuration and hence optical activity. The sign of the optical rotation is still dependent on the absolute configuration of the L-histidine residue. If this were of the D configuration then the imidazole would bind to the copper on the obverse side of the distorted molecular plane to give the optical antipode. A type of 5-coordinated copper (II) has been found in the crystalline structure of the diglycylglycine complex (15). One other possibility must be entertained, namely the formation of a dimeric complex in which the imidazole group of one complex is bonded to the copper (II) of another complex. Such is the case in the copper (II)-carnosine complex in which the imidazole group occupies, through the 3-nitrogen, the fourth position in the distorted square plane around the copper (II) of the second complex (H. C. Freeman and J. T. Szymanski, unpublished results).

there are interactions between the side chain and components of the peptide, perhaps the carbonyl oxygen atom in the peptide bonds. This finding, coupled with the observation (6) that the ORD curves for the copper (II) complexes of glycyl+alanine and glycyl-n-alanine were exactly equal and opposite in sign, establishes that the sign of the Cotton effects is controlled purely by the configuration of the constituent amino acids and that there is little possibility of a structural factor which could cause inversion around the metal ion as there is in the case of histidine- containing peptides.

As has been pointed out, the option of assuming some contribution from an inverted configuration in the histidine-containing loci of the apomyoglobin complexes, coupled with a contribution from the NH2-terminal complex locus, allows a fairly good summation of elements in Figs. 5B and 6B to give 7B. Perhaps a more straightforward way of explaining the relatively small contributions of histidine loci to the optical rotatory properties (but not to the absorption spectra, Figs. 2 and 3) would be to assume less distortion of the puckered forms in the protein complexes. Constraints imposed by the polypeptide backbones and the multitude of side chain interactions may not permit the formation of a complex having the same degree of distortion as a small model peptide. Unfortunately, the overlay of the NH*-terminal complex spectrum in the case of both copper (II) and nickel (II) is such that the separation of transitions asso- ciated with the histidine loci cannot be estimated. Such a

separation, for example that of about 90 rnp for the CuLt- species of acetylglycylglycyl-L-histidine (Fig. 5B), could indicate the relative degree of distortion if it were observable in the protein cases.

From the absorption spectra, the nickel (II) complexes can be taken confidently as square planar, so that the position of the donor nitrogen atom of the imidazole ring is in the approximate plane of the peptide donor nitrogen atoms in both the small peptide complexes and in the protein complex. This appears to limit the possibilities for variation in optical activity to varia- tions in degree and direction of distortion induced by neighboring structures. In the copper (II) complexes, it is possible in addition to have some contribution from forms in which the imidazole nitrogen bonds to the copper in an apical position.4 The relative importance of such forms could well differ between the small peptides and the protein. There is some crystallographic evidence that in copper (II) complexes with the square planar array somewhat distorted, the apical bond in the square pyramidal arrangement need not be normal to the approximate plane of the square base (16).

There are undoubtedly many other factors which preclude strict parallels between the optical rotatory behavior of the protein complexes and the peptide models considered here. Paramount among these is the influence of side chains. It is premature at this stage to assess the influence of side chains on the optical rotatory properties of model peptides, and a more definitive answer must await the study of small segments of the myoglobin polypeptide chain.

Acknowledgments-The technical assistance of Miss Anne Kask is gratefully acknowledged. We are deeply indebted to the late Professor T. S. Piper for advice concerning the per- formance and interpretation of the measurements of circular dichroism. Dr. H. C. Freeman has contributed valuable sugges- tions about the potential structures of the complexes.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Copper and Nicks1 Complexes of Xperm Whale Myoglobin Vol. 241, No. 5

REFERENCES

1. HARDMAN, K. D., EYLAR, E. H., RAY, D. K., BANASZAK, L. J., AND GURD, F. R. N., J. Biol. Chem., 241, 432 (1966).

2. MARTIN, R. B.. CHAMBERLIN, M., AND EDSALL, J. T., J. Am. Chem.‘Soc., s2, 495 (1960). ’

3. MARTIN. R. B.. AND EDSALL. J. T.. J. Am. Chem. Sot.. 82. 1107 (i960). ’

,

4. BRYCE, G. F., ROESKE, R. W., AND GURD, F. R. N., J. Biol. Chem., 240, 3837 (1965).

5. BRYCE. G. F., AND GURD, F. R. N.. J. Biol. Chem., 241, 122 (1966). ’

6. BRYCE, G. F., PINKERTON, J. M. H., STEINRAUF, L. K., AND GURD. F. R. N.. J. Biol. Chem.. 240, 3829 (1965).

7. BRESLO;, E., J. &oZ. Chem., 239,‘486 .(1964).‘ ’ 8. MAZUR, R. H., ELLIS, B. W., AND CAMMARATA, P. S., J. BioZ.

Chem., 237, 1619 (1962).

9. LI, N. C., CHU, T. L., FUJII, C. T., AND WHITE, J. M., J. Am. Chem. Sot., 77, 859 (1955).

10. BRESLOW, E., J. Biol. Chem., 239, 3252 (1964). 11. BRESLOW, E., AND GURD, F. R. N., J. BioZ. Chem., 238, 1332

(1963). 12. KOLTUN, W. L., ROTH, R. H. AND GURD, F. R. N., J. Biol.

Chem., 238, 124 (1963). 13. KOLTUN, W. L., FRIED, M., AND GURD, F. R. N., J. Am. Chem.

Sot., 82, 233 (1960). 14. DOBBIE, H., AND KERMACK, W. O., Biochem. J., 69, 246 (1955). 15. COOPER, T., FREEMAN, H. C., ROBINSON, G., AND SCHOONE,

J. C., Nature, 194, 1237 (1962). 16. FREEMAN, H. C., in P. AISEN, W. E. BLUMBERG, AND J.

PEISACH (Editors), Proceedinge of the International Sym- posium on the biochemistry of copper, Academic Press, Inc., New York, 1966, in press.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Graeme F. Bryce, Roger W. Roeske and Frank R. N. GurdNickel (II) Ions with Sperm Whale Apomyoglobin

l-Histidine-containing Peptides as Models for the Interaction of Copper (II) and

1966, 241:1072-1080.J. Biol. Chem.

http://www.jbc.org/content/241/5/1072Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/241/5/1072.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/content/241/5/1072

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;241/5/1072&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/241/5/1072

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=241/5/1072&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/241/5/1072

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/241/5/1072.full.html#ref-list-1

http://www.jbc.org/

l-histidine-containing peptides as models for the ... · histidine and...

Documents