Int. J . Peptide Protein Res. 17, 1981,460-468

THERMAL PERTURBATION SPECTROPHOTOMETRY O F LULIBERIN A N D MODEL HIS-TRP PEPTIDES

JAKE BELLO, HELENE R. BELLO and SABURO AIMOTO’

Department of Biophysics, Roswell Park Memorial Institute, Buffalo, New York, U.S.A.

Received 9 July, accepted for publication 3 October 1980

Thermal perturbation ( T P ) spectra of luliberin were measured a t pH 5 -8, and compared with the model chromophores N-Ac-Tyr-NH2, N-Ac-Trp-NH?, t-Boc- His-Trp-NH2, H-His-Trp-OH, Ac-His-Trp-OH, tryptophan and cyclo- [His-Trp] (all L-isomers). Between pH 5 and neutrality, the major TP extremum of the Trp3 residue o f luliberin increases by about 50%. A similar effect is seen for luliberin acetylated on Tyr’. The effect with luliberin is attributed to the pro- tonation o f the His’ residue. One proposed explanation is that the protonated imidazole orients water around the nearby indole in a different way than does unprotonated imidazole. The Tyr’ residue of luliberin behaves like N-Ac-Tyr- NH2, and is considered to be well exposed to solvent. The TP spectra of N-Ac- Trp-NH, , t-Boc-His-Trp-NH2, Ac-His-Trp-OH, and cyclo-[His-Trp 1 are pH- independent f rom pH 5 to 8. The TP spectrum of H-His-Trp-NH2 has a bell- shaped pH dependence, rising f rom normal at pH 3.5 to above normal at pH 6 , and returning to normal a t pH 8. Luliberin and model peptides show that fluore- scence and TP spectra of His-Trp sequences can respond differently to pH.

Key words: luliberin; thermal perturbation spectra.

Luliberin (luteinizing hormone-releasing hor- mone, gonadotropin releasing hormone) is a decapeptide with the sequence <Glu-His- Trp-Ser-Tyr-Gly-Leu-Arg-Pro-C;iy-NH2 . N.m.r. studies have been interpreted as indicating a flexible structure. Wessels et a/. (1973) con- cluded that there are no specific sidechain interactions, no stacking of the aromatic rings and no electrostatic intramolecular interactions involving the charged imidazole ring with other residues at low pH. Deslauriers et al. (1975) concluded that the l3 C spectrum of luliberin in

’ Permanent address: Institute for Protein Research, Osaka, Japan.

aqueous solution shows no evidence for stacking of the rings of tryptophan and tyrosine, that the spectrum calculated from constituent amino acids resembles that of luliberin, and that the backbone of luliberin is mobile. The chemical shifts of the carbon atoms of Trp-3 were independent of pH over the range of titration of His-2. However, they noted: “Aromatic residues do not show a large degree of segmental motion; this may result from bulkiness of the side chain rather than from steric hindrance of neighboring residues.”

Mabrey & Klotz (1976) reported almost identical CD spectra of luliberin in water and in 6 M guanidinium chloride (GdmCI), and con-

460 0367 -83 77/8 1 /040460 -09 %02.00/0 @ 198 1 Munksgaard, Copenhagen

LULIBERIN THERMAL PERTURBATION SPECTRA

cluded that luliberin in water is a random coil. They measured (potentiometrically) the pK of the imidazole to be 6.6, a normal value. Cann et al. (1979) concluded from CD studies of luliberin and partial peptides that an ensemble of conformers exists in aqueous solution at pH 7. There are discrepancies between several published CD results. Mabrey & Klotz reported, for pH 7, a negative extremum at 204 nm with a molar ellipticity of - 18 x lo3 deg cm2 dmol-* . Marche ef al. (1973) reported an extremum at 195 nm with a molar ellipticity of -47 x lo3. Cann et al. (1979) reported an extremum at 196 nm with a molar ellipticity of - 8.5 x lo3 for luliberin in water at pH 7. Mabrey & Klotz reported that the CD spectra were the same at pH 2.7,7 and 11, except for a shoulder at 223 nm at pH 11. The CD spectra of Marche et al. are considerably different at pH 3 and 7.

Mabrey & Klotz reported “a slight decrease” in the fluorescence emission from the tryptophyl residue as the pH was reduced from 7 to 2; this was attributed to protonation of the histidine. Marche et al. (1976) and Shinitzky & Fridkin (1976) found a 30% decrease in emission intensity between pH 7 and pH 5, with a pK of 6, a value considerably lower than the potentiometric pK of Mabrey & Klotz for luliberin, or those of model peptides (Shinitzky & Fridkin, 1969; Shinitzky & Goldman, 1967; Schneider, 1967). Marche et al. proposed the existence of a charge-transfer complex between the side chains of protonated His-2 and Trp-3. Shinitzky & Fridkin (1976) had measured earlier the pH-dependence of the fluorescence of open chain and cyclic peptides containing hstidyl and tryptophyl residues. From their comparison between luliberin and the models, Shinitzky & Fridkin (1976) concluded that the interaction between His-2 and Trp-3 in luliberin is hindered, and that no charge-transfer complex is present.

Thus n.m.r., CD and fluorescence data have not been interpreted in a mutually consistent manner. We have applied the method of thermal perturbation (TP) difference spectra to luliberin and to some model peptides. In the TP method, a difference spectrum is generated from two portions of the same solution maintained at two temperatures. Ih the absence

of conformational change the difference spectra arise largely from differences in the interaction of the chromophore with its environment at the two temperatures. The TP method has been used for the estimation of exposed tyrosyl and tryptophyl chromophores of proteins (Bello, 1969, 1970; Misselwitz et al., 1975), for the study of the environment of the heme in leghemoglobin and myoglobin (Nicola & Leach, 1977), for the study of the solvation of bovine pancreatic ribonuclease in the crystallizing medium (Pittz & Bello, 1973) and for the study of peptide-detergent inter- actions (Bello & Bello, 1973).

MATERIALS AND METHODS

Luliberin Three batches of luliberin were used, one pur- chased from Bachem Chemical Co., and two obtained as gifts from the Contraceptive Development Branch, National Institutes of Health. No significant differences were observed in the results. Acetylated luliberin A well-stirred solution of luliberin, 2 mg (1.8 x mol) in l O m l of 0.02 M sodium acetate at pH 7.2, was treated with 0.01 ml

mol) of acetic anhydride added in five portions. The pH fell to 4. Sufficient 4 M KC1 was added to make the solution 0.1 M in KCl, and the solution was used for spectrophoto- metry. The concentration of acetic acid plus acetate ion was approximately 0.04 M .

Pep tides

t - B o c - ~ -His-t -Trp-NH2 . This was synthesized by standard procedures by condensing di(t- Boc)-His-OH and H-Trp-OEt with dicyclohexyl- carbodfmide. The di-(t-Boc)-His-Trp-OEt was converted to the amide with ammonia in methanol, with simultaneous removal of t-Boc from the imidazole. After evaporation, the oil was dissolved in ethyl acetate, filtered, con- centrated under vacuum, and solidified by trituration with petroleum ether. The product was crystallized from ether-petroleum ether and from ethyl acetateether. Yield: 79%, m.p. 149-150”. €274 was 5600, typical of a tryptophyl chromophore. The ninhydrin reaction was negative.

46 1

J. BELL0 ET AL.

cyclo-[His-Trp] . The t-Boc groups of di(t-Boc)- His-Trp-OEt were removed with formic acid (containing 4% anisole) at room temperature. After evaporation, the oil was dissolved in methanol-ammonia, and after 3 days the solvent was evaporated and the oil was chro- matographed on silica gel with methanol- tetrahydrofuran (1 : 1, v/v). After evaporation, the product was crystallized from methanol- ethyl acetate-ether, m.p. 157-165" (lit. 160- 165"; Shinitzky & Fridkin, 1976). The identity was established by spectrum, negative nin- hydrin and elemental analysis.

Measurements TP spectra were measured as described earlier (Bello, 1969), with temperatures of 26" vs. 2", using cells with Teflon stoppers. The compart- ment of the 2" sample was flushed with dry nitrogen. The pH was adjusted with 0.1 M NaOH, and was measured again after the TP spectrum was taken. A direct absorption spectrum was recorded for each TP spectrum. TP spectra were also obtained on the buffers, to ensure that these did not contribute to the spectra. Fluorescence measurements were made with an Aminco Bowman spectrophotofluori- meter fitted with a Princeton Applied Physics Model 1108 photon counter. Excitation was at 280 nm and emission was measured at 360 nm.

RESULTS

Background The extremum at 293 nm in the TP spectrum of luliberin arises almost entirely from Trp-3. The tyrosyl contribution to (from a TP extremum centered at 287nm) for the model mixture of equimolar N-Ac-Tyr-NH2 and N - Ac-Trp-NH2 is about 10% at pH 5-7 (Bello & Bello, 1973). However, a shift in the Tyr and/or Trp extremum could decrease or increase the Tyr contribution to the Trp extremum. Large contributions from tyrosyl develop when this residue begins to be titrated. Values of AA are normalized by dividing by A,, of the direct spectrum and by the temperature difference to give AAlAAT = Ae/eAT. A comparison of Ae/eAT for a peptide with that of a model compound may not be altogether valid, in that the extrema for the two may not have the same

shape, because of the influence of other groups. Also, when there is more than one chromo- phore of a given type in a peptide (not the case here), they may be differently affected by environmental conditions, resulting in wave- lengths for their extrema somewhat different from each other and from those of the models, resulting in a broader but shallower extremum. In such cases the integrated intensity may be more significant than A€. We have obtained integrated values by tracing the 293 nm extrema on cross-ruled paper and counting squares.

Results at 293 nm in ammonium acetate buffer Fig. 1 shows representative TP spectra of luliberin and of the model mixture in ammonium acetate buffer at several pH values between about 5 and 8. Fig. 2 shows Ae/eAT

r

W -0.02 l . I . l . l . 1

250 260 280 300 320 340 A, i t m

FIGURE 1 TP spectra of luliberin and the model mixture in 0.02 M ammonium acetate. A: luliberin, spectra 1 , 2, 3 at pH 5.0, 7.27, 8.0, respectively. B: equi- molar mixture of N-Ac-Trp-NH, and N-Ac-Tyr- NH,, spectra 4, 5 , 6 at pH 5.0, 7.8, 8.2, respectively.

462

LULIBERIN THERMAL PERTURBATlON SPECTRA

vs. pH at 293 and at 250nm for the model mixture. It should be noted that the sign of AE/eAT is the opposite of the sign of the observed AA since the temperature of the "sample" is lower than that of the "reference". Af293 /EAT for N-Ac-Trp-NH2 (not shown) is independent of pH. For a 1:l molar ratio of N - Ac - Trp - NHz and N - AC - Tyr- NH2 there is some change in A E ~ ~ ~ / E A T from pH 5 to about 7, followed by a steep decline at higher pH. Aczso /EAT changes sign as the pH is raised. AeZ50/~AT is near the begin- ning of a large extremum centered on 228 nm. The large changes in AE/EAT at 293 and 250 nm arise from the beginning of the titration of the tyrosyl phenolic group. The effect is seen for N-Ac-Tyr-NH, but not for N-Ac-Trp-NHz . This premature beginning of the titration of tyrosine arises from the change in the pK of NH;, from 9.25 at 25" to 9.9 at 5" (Sober,

5 6 7 0

PH

t 4 0

t2 0

0 0 X

-2

- 4

-6

-8

- 10 - I2

FIGURE 2 Dependence on pH of AEIEAT for luliberin and a model mixture in 0.02 M ammonium acetate. Luliberin: 0, A, with 0.1 M KCI; 0 , v, without KCl; 9, v , 0.02M n-butylammonium acetate with 0.1 M KCI; 0, model mixture ofN-Ac-Tyr-NH, and N-Ac-Trp-NH , , equimolar.

1968~) . The pK of N-Ac-Tyr-NH2 was little affected by temperature being 9.6 f 0.1 at 25" and 9.8 * 0.1 a t I" , in both buffers.

For luliberin in ammonium acetate there is an increase of about 50% in fkz93/eAT on going from pH 5 to 7 (Fig. 2). Af293 of the tryptophyl residue is strongly influenced by the titration of the adjacent histidyl residue, the only titratable residue in this pH range. At pH 5, A E ~ ~ ~ / E A T for luliberin is 64% of that of the model mixture. Above pH 7.25, A E ~ ~ ~ / E A T falls sharply as it does for the model mixture. For luliberin, as for the model mixture, Aezso/ EAT decreases sharply with pH above about pH 7.25.

The TP titration curve of luliberin has hardly begun at pH 6, at which the fluorescence titrations of Shinitzky & Fridkin (1976) and of Marche et al. (1976) have reached their mid- points. It is not possible to obtain a true pK for the imidazole from this experiment for several reasons. Firstly, the curve changes direction at pH 7.25, so that the titration of the imidazole cannot be completed. Secondly, TP generally cannot give a true pK since the pK usually changes with temperature, so that in the titration range we see superimposed the TP spectra of protonated and unprotonated species, the proportions of which are not the same at the two temperatures. The pK of imidazole (Sober, 1 9 6 8 ~ ) changes from 6.99 to 7.58 between 25" and 0", thus shifting the TP titration curve to a higher pH range. Thirdly, the pH of the buffer can change with tempera- ture.

Results at 293 nm in NaOAc In buffer of 0.1 M KCl and 0.02 M NaOAc, Ae293 /€AT for N-Ac-Trp-NH2 is independent of pH and is the same as in ammonium acetate. For N-Ac-Tyr-NH2 and for the model mixture Ae293/eAT is flat from pH 4.8 to pH 7 (Fig. 3). From pH 7 to 8 an increase in A E , ~ ~ / E A T occurs. The reversal of slope seen in ammonium acetate is not seen in NaOAc. Luliberin behaves similarly at 250nm, as shown in Fig. 4, but at 293 nm the TP titration is sigmoid. The values of Ae293/~AT and of the integrated intensity for luliberin rise from about 65% of that of the model mixture to about 90% as the pH is increased.

463

J . BELL0 ET AL.

'"-A

L I I 1

5 6 7 8 PH

FIGURE 3 Dependence on pH of AEIFAT for models and luliberin in 0.02 M sodium acetate: 0 , 0 , luliberin; A , A, N-Ac-Trp-NH, and N-Ac-Tyr-NH,; 0,

ace t yla ted luliberin.

Results in the 260-285 nm region For luliberin and the model mixture in both buffers, A&AT was calculated at 270 and 280nm. For luliberin the ratio of Ae/eAT to that for the model mixture was about 0.3-0.5 at pH 5-8, distinctly smaller than at 293nm.

0.02 L

-0.ozt I , , , I , , , , 1 260 280 300 320 340

A, nm FIGURE 4 TP spectra of luliberin and models in 6 M GdmC1;

NH, and N-Ac-Tyr-NH,, 2 X 10- ' M each. -, luliberin/2 X M; - - - - , N-A C-TI p-

464

No high accuracy can be claimed for the data at 260-285nm as the absolute values of AA observed are small, about 0.003-0.006. To attempt to increase the accuracy substantially by increasing AA would require unacceptably hgh absorbancies. We cannot yet apportion the extrema of luliberin in this spectral region between the two aromatic chromophores. It is clear, however, that at least one chromophore of luliberin differs from the model mixture by this criterion.

Results with 6 M guanidinium chloride Fig. 4 shows that the TP spectrum of luliberin at pH 5 in 6 M guanidinium chloride is very similar to that of the model mixture in this solvent. Thus, 6 M GdmCl effectively eliminates most of the constraints which exist in luliberin, resulting in the Trp-3 and Tyr-5 chrornophores becoming nearly as independent of intra- molecular influences as those of the model mixture.

Acetylated luliberin In order to avoid the complications arising from the Tyr residue, we acetylated the tyrosyl of luliberin with acetic anhydride. Acetylation of N-Ac-Tyr-NH2 reduces the direct and TP spectra by more than 90%. The absorption spectrum of acetylated luliberin showed the characteristic tryptophan peaks at 287 nm and at 280nm and the shoulder at 273nm. In native luliberin the overlapping tyrosyl spectrum hides the 273 and 287 nm peaks and broadens the main 280nm peak. That acetylation of Tyr-5 was complete was shown by the absence of the 293nm tyrosinate peak at pH 1 1 . Fig. 3 shows that the TP spectrum of acetylated luliberin is also pH-dependent. The AeleAT values for acetylated luliberin are not directly comparable with those of native luliberin, because for acetylated luliberin the value of E of the direct spectrum is almost entirely that of tryptophyl, while for native luliberin the whole E of the tyrosyl is included, although tyrosyl contributes to Ae to a minor extent. Using E of native luliberin to calculate Ae/EAT for acetylated luliberin, AE/EAT goes from 5.2 to 7.6 x as the pH isincreased. The latter is smaller than 8.5 x for N- Ac-Trp-NH2.

LULIBERIN THERMAL PERTURBATION SPECTRA

TP spectra o f model His-Trp peptides We recorded TP spectra of several His-Trp models and of tryptophan. Fig. 5 shows that tryptophan, t-Boc-His-Trp-NHz and cyclo- [His- Trp] show little or no pH-dependence between pH 4.8 and about pH 7.5. H-His-Trp-OH shows a bell-shaped pHdependence. The pH- dependence of the TP of this peptide is dif- ferent from that of luliberin in that at low pH AE/EAT is “normal” and rises to an above normal value at pH6 , before returning to normal at higher pH. TP for Ac-His-Trp-OH is essentially flat from pH4.8 to 7.8; the deviations of the points from the straight line are within the experimental error, namely 0.001A. The data for these models and for N- Ac-Trp-NHz show that there is no unique “normal” value, rather that Ae/eAT is influenced by the presence of other groups.

Fluorescence of models We measured the fluorescence of several models (Fig. 6), and found pH-dependence. Shinitzky & Goldman (1967) and Shinitzky & Fridkin (1976) found pH-dependence for cyclo-[His- Trp] and for N-Ac-His-Trp-OEt, but larger than we found for cyclo-[His-Trp] and t-Boc-His- Trp-NHz. The most interesting point about

I H-HIS TRP OH 16. 2 TRYPTOPHAN

3 CYCLO- LHlS-TRP] 4 I-EOC-HIS-TRP-NHz 3 nc-HIS-TRP-OH

~ 15 ’

0 . 14 r

‘0 t L 1 - I I I I

3 ‘4 5 6 7 8 9 10 11

PH

FIGURE 5 Dependence on pH of A E ~ ~ ~ / E A T for model peptides. Solvent: 0.1 M KCl, 0.02 M sodium acetate. The data for Ac-His-Trp-OH may be described by the straight line as well as by the curved line, since the deviations of the points from the former are close to experimental uncertainty.

Fig. 6 is the fluorescence of H-His-Trp-OH is pHdependent but in an Scurve, not the bell curve seen for TP.

DISCUSSION

At neutral pH the magnitude of the 293nm extremum of luliberin is near that of the mixture of N-Ac-Trp-NH2 and N-Ac-Tyr-NH2, indicating that Trp-3 is not under strong constraint, in agreement with the conclusions from n.m.r. (Deslauriers et ~ l . , 1975; Wessels el al., 1973). #en the normalization of Ae is done with E,, of Trp alone, without the Tyr contribution (1400 M - cm - I ) AE/EAT at 293nm is raised from 9 x close to those of N-Ac-Trp-NHz , t-Boc-His-Trp- NH2, and cyclo-[His-Trp] , further indication of a substantially normal Trp residue in luliberin. However, the TP data between 280 and 260 nm indicate some constraint, The results in 6 M GdmCl show that in this solvent almost all constraints are removed. The decrease in A€/ EAT for luliberin at low pH, at which the His-2 side chain is protonated, shows an influence of the imidazolium on the TP of Trp-3, similar to the effect seen by fluorescence (Marche et ~ l . , 1976; Shinitzky & Fridkin, 1976). The Tyr residue appears to be normal, since its response to buffer effects is very similar to that of N-Ac- Tyr-NH2. Also, the normal pK of the tyrosyl residue argues for a normal tyrosyl. The buffer- temperature effect is useful for studying Tyr in the presence of Trp.

to 11 x

-~

1 t:-HlS-TRP-OH 2 CYCLO- [HIS-TRP]

2 1 3 t -3oc-HIS-TRP-NHZ

a a

L” 5 1

4

x: 5 6 7 8 9 1 0

PH FIGURE 6 Dependence on pH of fluorescence emission of model peptides. Solvent as in Fig. 5 .

46 5

J. B E L L 0 ET AL.

With luliberin acetylated on Tyr-5, the strong pH-dependence of /EAT shows that the spectral effects observed with luliberin arise largely, if not wholly, from Trp not Tyr. A possible explanation for the difference between luliberin and acetylated luliberin is that acetylation of the hydroxyl of Tyr (and perhaps of Ser4) has introduced a constraint, perhaps arising from an increase in hydrophobic character. Despite some possible tyrosyl con- tribution to the pH-dependence of Ae of luliberin, it is clear that protonation of His influences Trp.

The data of Figs. 5 and 6 show that the pH- dependence of TP and fluorescence of luliberin do not necessarily have the same origin. The presence of neighboring His and Trp residues does not necessarily result in a pH-dependent TP but it does result in pH-dependent fluore- scence in all cases known to us. The two methods can give different views of chromo- phore environment.

The effect of imidazolium on Trp-3 might arise from several sources. Marche et al. (1 976) proposed a His-Trp charge-transfer, but, as Mabrey & Klotz (1976) noted, the absorption spectrum of luliberin is very close to the sum of those of N-Ac-Trp-NH2 and N-Ac-Tyr-NH2. A charge-transfer complex, if present, would shield part of the indole from solvent, and change the solvent-indole interaction, thereby altering the TP spectrum. A charge-transfer complex would be expected to raise the pK of the imidazolium by virtue of the greater electron density induced in the imidazolium. But all of the reported pK values (potentio- metric and fluorimetric) of the imidazolium group of luliberin are either normal, or are lower than normal for model peptides. Also, a strong direct interaction is not in accord with n.m.r. data. The pK values of the His-Trp peptides of Shinitzky & Fridkin (1 976), which were proposed to be charge-transfer complexes, were normal and were unrelated to the efficiency with whch the imidazolium group quenched the fluorescence of the indole. Shinitzky & Fridkm (1976) proposed that “Trp-3 is at a maximal distance from this unit [Arg-8, His-2 and Tyr-51 and may thus act as an independent unit”.

An effect of the imidazole in TP does not

necessarily require direct contact with the indole. The results might arise from an effect on intervening water. TP spectra arise in part from changes in the solvation of the chromo- phore with temperature (Bello & Bello, 1976; Demchenko & Zyma, 1977). Momany (1976) made a conformational energy analysis of luliberin. The three conformations of lowest energy have a bend at Gly-6, and in one of these the indole and imidazole rings are nearly parallel, with their C (7) atoms 4.5w apart. This distance is too small for water insertion, but somewhat different conformations might permit this. The presence of an imidazolium group near the indole could alter markedly the water orientation around both groups and change the response of the water t o tempera- ture. As the imidazolium is titrated the ordering of the water would be relaxed and become more like ordinary water, resulting in the TP spectrum of Trp-3 becoming more normal. Fluorescence, too, is sensitive to solvation, but the effects of water orientation might be quite different on TP and fluorescence. Our hypothesis appears t o be compatible with n.m.r. data, since changes in water structure would have a lesser effect on the n.m.r spectrum of the peptide than would direct rink interactions. Shinitzky & Fridkin (1976) state that the emission spectrum of luliberin excited at 295 nm is typical of tryptophan in a hydro- philic environment. This is compatible with our proposal for the influence of His-2 on Trp-3 via a water interaction. TP effects arising from structural changes in solvent have been observed for H2 0-alcohol mixtures, where A€/€ reaches a maximum at alcohol concentrations at which other properties of water show maxima or minima (Pittz & Bello, 1970). For N-Ac-Tyr-NH2 a 60% increase in A E ~ ~ , / E A T was observed at 0.05 mol fraction t-butanol.

The origin of the pH-dependence of luliberin may lie in conformational differences between pH 5 and 7.5. A conformational constraint at low pH might involve an interaction of the imidazolium with some group, other than the indole, this interaction indirectly affecting indole. This idea is not in accord with the interpretation of Wessels (1973) of their n.m.r. data obtained at low pH (see Introduction). Marche et al. (1973) showed significant changes

466

LULIBERIN THERMAL PERTURBATION SPECTRA

in the CD spectrum of luliberin between pH 3 and 7, which they interpreted as a shift from a disordered low pH conformation to a some- what ordered pH 7 conformation. The sugges- tion of Marche et al. that luliberin at pH7 is more ordered than at lower pH, is not in simple accord with our results which show that at pH 7-8 luliberin is more like the model mixture than at lower pH, nor with the interpretation of Deslauriers et al. (1975) of n.m.r. data at pH 6.4. Of course, the TP spectra report only on the state of the sidechain chromophores, while CD reports on the overall conformation, but is strongly influenced by the aromatic chromophores when the peptide is small. It is possible that at pH 7 the His-2-Trp-3 inter- action vanishes, and that a new backbone con- straint emerges, but of such a nature as not to prevent the Trp-3 and Tyr-5 side chains from behaving much like those of the model compounds.

The TP spectrum of H-His-Trp-OH is pH- dependent (Fig. 6). The low and high pH values of AEIEAT are close to that of N-Ac-Trp-NH2, while the peak value is 30% hgher, indicative of a Trp residue under abnormal influence. The pK of a typical carboxyl of an H-His-X-OH peptide is about 2.4 based on that of H-His- Gly-OH (Sober, 1968b). Therefore the increase in AE/EAT between pH 3.6 and 6 does not arise from the titration of the carboxyl, which is already largely dissociated at pH 3.6, but more likely from titration of the imidazolium group. A priori, we cannot be sure which cationic group is being titrated between pH 4 and 6, but normally the imidazolium has the lower pK, e.g. 5.8 for imidazole and 7.6 for ol-NH2 of H-His-Gly-OH (Sober, 19686). It may be that the abnormally high AeleAT for H-His-Trp-OH at pH 6 might arise from an interaction of the carboxyl with the (r-NHS, which is eliminated at higher pH by discharge of the (r-NH+,, and at lower pH by charging of the imidazole. The high Ae/eAT at pH 6-6.5 might arise from ordering of water by the charged groups. For H-His-Trp-OH the effect of deprotonation of imidazolium increases Ae of Trp, while for Ac- His-Trp-NH2 there is no effect. Clearly, the TP response is dependent on the local environ- ment, which may have a direct influence or

perhaps act by an effect on the water around the chromophore.

ACKNOWLEDGMENTS

This work was supported in part by Grant GM 24525 from the Institute of General Medical Sciences, National Institutes of Health.

REFERENCES

Bello, J. (1969) Biochemistry 8,4542-4549 Bello, J . (1970) Biochemistry 9,3562-3568 Bello, J. & Bello, H.R. (1973) European J. Biochem.

Bello, J. & Bello, H.R. (1976) Arch. Biochem. Bio-

Cann, J.R., Chanabasaviah, K. & Stewart, J.M. (1979)

Demchenko, O.P. & Zyma, V.L. (1977) Studio Bio-

Deslauriers, R., Levey, G.C., McGregor, W.H., Sarantakis, D. & Smith, I.C.P. (1975) Biochemistry

Mabrey, S. & Klotz, I.M. (1976) Biochemistry 15, 234-242

Marche, P., Morgat, J.L. & Fromageot, P. (1973) European J. Biochem. 40,5 13 -5 18

Marche, P., MontenayGarestier, T., Helene, C. & Fromageot, P. (1976) Biochemistry 15, 5730- 5737

Misselwitz, R., Zirwer, D., Frohne, M. & Hanson, H. (1975) FEBS Lett. 55,233-236

Momany, F.A. (1976) J. Am. Chem. SOC. 98,2990- 2996

Nicola, N.A. & Leach, S.J. (1977) Bi0chemistr.y 16,

Pittz, E.P. & Bello, J. (1970) Trans. Faraday SOC. 66,

Pittz, E.P. & Bello, J . (1973) Arch. Biochem. Biophys.

Schneider, F. (1967) Hoppe-Seylers Z. Physiol. Chem.

Shinitzky, M. & Fridkin, M. (1969) European J. Bio-

Shinitzky, M. & Fridkin, M. (1976) Biochim. Biophys.

Shinitzky, M. & Goldman, R. (1967) European J.

Sober, H.A. (19680) Handbook of Biochemistry, p.

Sober, H.A. (1968b) Handbook of Biochemistry, p.

34,535-538

phys. 174,532-540

Biochemistry 18,5 7 7 5 -5 7 8 1

phys. 64,143-150

14,4335-4343

50-58

537-545

156,437-447

334,26-43

chem. 9,176-184

Acta 434,137-143

Biochem 3,139-144

J 6 2 , Chemical Rubber Co., Cleveland Ohio

5-92, Chemical Rubber Co., Cleveland, Ohio

467

I . BELL0 ET AL.

Wessels, P.L., Feeney, J., Gregory, H. & Gormley, J . J . (1973) J. Chem. SOC. Perkin Trans. 11, 1691-1698

Address: JakeBello Biophysics Department Roswell Park Memorial Institute Buffalo, New York 14263 U.S.A.

468