Binding of N-methyl nicotinamide chloride by tryptophan residues of -lactalbumin.

Abstract Studies of the binding of N-methyl nicotinamide chloride by α-lactalbumin show that complex formation occurs with 1 of 4 tryptophan residues at pH 6 whereas two form complexes at pH 2 as a result of acid denaturation. Alkaline denaturation likewise leads to an increase in exposure of tryptophans, and three of the four groups form complexes at pH 11. From a consideration of previous studies and a recently proposed molecular model for α-lactalbumin, the residues involved in binding N-methyl nicotinamide chloride have been tentatively identified as tryptophan-118, tryptophan-104, and tryptophan-60.

From the Pioneering Research Laboratory, United States Army Satick Laboratories, Natick, Lllassachasefts 01760 SUMMARY Studies of the binding of N-methyl nicotinamide chloride by cu-lactalbumin show that complex formation occurs with 1 of 4 tryptophan residues at pH 6 whereas two form complexes at pH 2 as a result of acid denaturation.
Alkaline denaturation likewise leads to an increase in exposure of tryptophans, and three of the four groups form complexes at pH 11. From a consideration of previous studies and a recently proposed molecular model for cr-lactalbumin, the residues involved in binding N-methyl nicotinamide chloride have been tentatively identified as tryptophan-118, tryptophan-104, and tryptophan-60.
Previous studies with a-lactalbumin have shown changes in the environment of tryptophanyl side chains as a result of acid and alkaline conformational changes which occur with this protein (l-7).
Studies utilizing acid and alkaline difference spectra (1, 5) have shown changes in the neighborhood of one or more of the tryptophanyl side chains relative to the native protein.
Similar results have been obtained by fluorescence spectroscopy (3) and circular dichroism measurements (6). Solvent perturbation difference spectral measurements at acid pII indicate, however, that the changes observed during the acid transition do not involve a net change in exposure to solvent of tryptophanyl side chains relative to the native protein.
Interference due to tyrosyl ionization precludes the use of the latter t.echnique at alkaline pI-I. Since the solvent perturbation method does not unambiguously differentiate between partial exposure of many groups or complete exposure of a few groups (8), we have investigated the use of N-methyl nicotinamide chloride for titrating exposed tryptophanyl side chains (9-11).
This reagent forms a 1: 1 chargetransfer complex with tryptophanyl residues, and because of the planar interaction of the two ring systems needed for complex format,ion, requires that the tryptophanyl group be exposed at the protein surface and essentially unhindered. N-Methyl nicotinnmide chloride has been reported to form a charge-transfer complex with other aromatic side chains; however, t,his does not interfere with its use as a probe for tryptophanyl residues (12). This paper concerns itself with the use of N-methyl nicotinamide chloride as a reagent for titrating the exposed tryptophanyl side chains of cr-lactalbumin, and illustrates the potential of this  (14).
Binding Studies-For the binding experiments protein concentrations were 5 mg per ml in aqueous solution.
Protein solutions were adjusted to the appropriate pH by the addition of 0.1 M HCl or 0.1 M NaOH and measured on a Radiometer Instrument, model PH 4. Solid N-methyl nicotinamide chloride was added to the protein solutions; the amount added was determined by weighing the cuvettes before and after each addit,ion, and concentrations were corrected for volume changes occurring (10). Spectra were measured at room temperature 011 a Gary 15 recording spect,rophotometer. Eight to ten spect.ra were obtained for each protein solution at a variety of molar excesses.
The assumed molecular weight for cr-lnctnlbumiu was 14,500, and for lysozyme, 14,400.

N-Acetyl Tryptoplla?a Ethyl Ester arrd Lysoz~~tlf--Preli~llillar~
to binding studies with cr-lactalbumin, we investigated the pH dependency of complex formation between N-acetyl tryptophan ethyl ester and N-methyl nicotinamide chloride at pH 2, 6, and 11. We also repeated t,he binding studies of Deradeau et al.  The data for N-acetyl tryptophan ethyl ester when plotted according to Equation 1 gave a straight line at all pf1 values. The ralues of the constants K and E (Table I) show little or no pH dependence of the molar extinction coefficient; however, the association constant is significantly higher at high pl1. The values of the constants at pI1 6 are also in good agreement with those found for N-acetyl tryptophan and A-acetyl tryptophanamide in water (9).
The lgsozyme data were linear when plotted according to Equation 1 (Fig. 1). The value of K was 3.7 11-l and E was 1230 81-1 cn-1. These findings are in reasonable agreement with those of Deranlenu et al. (10) ( Table I).
Brad&am and Deranleau (11) have demonstrated that only 1 mole of nicotinamide is bound to lysozyme at tryptophan-62. Therefore, the molar extinction coefficient of the tryptophan- lysozyme as compared to model tryptophan conq~o~ds. There are no data available to judge whether this is a general rule with proteins or whether it is a reflection of the environment of the complex in lysozyme.
With cr-lactalbumin, however, as shown below, the binding data are most reasonably interpreted if the number of binding sites are estimated with the molar extinction coefficient determined for lysozyme rather than those from model compounds.
Native c+lactalbumin-Titration of a-lactalbumin at pII 6 with N-methyl nicotinamide chloride results in formation of a yellow-colored complex due to charge-transfer complex formation with tryptophan (9-11) (Fig. 2). Complex formation was instantaneous and was monitored by measuring the increase in absorbance at 350 nm upon addition of N-methyl nicotinamide chloride, as described by Deranleau et al. (10, 11).
The titration data plotted according to Equation 1 (Fig. 1) give a linear plot indicating a single class of binding sites in the native protein.
The regression line was determined by the method of least squares, and the values of the constants estimated  Table I).
The results agree with those for hen lysozyme (Table I) and suggest' t'hat only one of the four tryptophanyl side chains in native oc-lactalbumin is available for binding with N-methyl nicotinamide.' This contrasts with data obtained by solvent perturbation difference spectroscopy (2) which indicates that two tryptophanyl groups are exposed at the protein surface. Since the latter method cannot differentiate between partial exposure of several groups or complete exposure of a few groups, it appears from a comparison of the results of the two methods that only one tryptophanyl side chain is completely exposed in the native protein, but that some of the three remaining tryptophanyl groups are partially exposed at the protein surface.
Our results with native oc-lactalbumin are not in agreement with previously reported work.
Bradshaw and Deranleau (11) found that native a-lactalbumin formed only a very weak complex n-it,h N-methyl nicotinamide while Guire (17) reported that native protein did not form a complex with N-methyl nicotiuamide, and that urea-denatured protein did form a complex.
1 We have estimated the nmnber of binding sites in a-lactalbumiu utilizing the molar extinction coefficient for complex formation between AT-met,hyl nicotinamide and lysozyme (10).
The number of completely exposed tryptophanyl groups estimated by this means (Table I) seems to offer a reasonable explanation of both t,he physical and chemical data as to the state of these side chains in a-lsctalbumin.
By contrast, the use of the extinction data from the 1node1 compomlds appears t.o overestimate t.he number of exposed tryptophan groups at acid pH when compared with solvent perturbation data. It also indicat,es that all four tryptophanyl side chains are completely exposed at pH 11, which does not appear likely in light of other observations under these conditions (5, 6).
Although we cannot offer a firm explanation for this anomalous behavior, some recent observations made in our laboratory indicate that it may arise from the well known tendency of a-lactalbumin to associate near pH 6 (15). This latter behavior appears to be related to t.he method of purificatiou of the protein.2 Denatured Lu-La&albumin-Previous studies (1, 4-6) have shown that a-la&albumin undergoes conformational chnuges below pH 4 and above pH 10. These two processes have many similar characteristics, and in each case a change in the environment of one or more tryptophanyl side chains occurs. With the acid transition, solvent perturbation difference spectroscopy indicates no net change in exposure of tryptophanyl side chains relative to the native protein at pH 6. In the alkaline transition little is known regarding the state of the tryptophanyl side chains, since the solvent perturbation cannot be used because of interference from ionized tyrosyl residues. For these reasons we have examined the binding of N-methyl nicotinamide t,o the acid-and alkaline-denatured proteins. The titration results with the actid-denatured protein (Fig. 3) are linear when plotted according t,o Equation 1 :LII~ suggest a single class of binding sites. The constants obtained from t,lle data are presented in Table I. The value of t#lie :\pp:\rent molar extinction coefhc~inlt 2277 M-I (~11~~ suggests the prraellce of two binding sites on the acid denatured protein.
These sites have an average association constant of 4.9 ~-1.
The titration data for the alkaline-denatured yroteiu (Fig. 3) are also linear, and the apparent molar extinction coefficieut of the complex 3411 M-' cm-l (Table I) indicates that probably three binding sites are present with an average association COnStaId Of 6.6 M-l. These results show that as a consequence of the conformational changes which occur a.t acid a.nd alkaline $1 values, additional tryptophanyl residues become completely exposed at the surface of the protein.
The significantly higher association constants of the alkaline-denatured protein compared to the native and acid-denatured proteins also indicate that the conformation in the vicinity of the tryptophauyl side chains of the alkaline-denatured proteins facilitate binding of Xmethyl nicotinamide. These observations are in general agreement with previous studies of the state of the tryptophnnyl residues in oc-lactalbumiu. DISCUSSION The present study iudicntes that native cr-lnct,nlbumin has only one tryptoplinnyl group sufficientjly esposed at the surface so that the planar geometry of the indole and pyridinium rings required for charge-transfer complex formation (9) ('an occur. l3y contrast solvent perturbation measurements show that. an average of two tryptophanyl groups are exposed in the native protein (2). Chemical reactivity studies with A--bromosuc-2 In cursory experiments Iye have observed that a snmple of a-lactnlbumin purified by DEAE-chromatography in Tris-chloride buffer (18), \T-hich had been treated in essentially the sixme manner as the sample used by Guire (171, in contrast t,o the sample purified by ammonium sulfate recrystallizabion, showed little or no evidence of complex formation at pH 6 or 2. It, did appear to form a strong complex at pH 11 It can therefore be concluded that in addition to the completely exposed tryptophanyl side chain, another group is sufficiently exposed to react with N-bromosuccinimide, and be "seen" as a completely exposed group by solvent perturbation, yet hindered enough by its environment to prevent complex formation with N-methyl nicotinamide. As shown below, these conclusions are in accord with the model of cr-lactalbumin proposed by Browne et al. (13) on the basis of its homology with hen lysozyme.
The present study shows that the conformational changes occurring at acid pH with a-lact,albumin result in complete exposure of an additional tryptophanyl side chain for complex formation with N-methyl nicotinamide. Solvent perturbation also shows an average of two exposed groups, and suggests that, as at pH 6, tryptophanyl residues 104 and 118 are most probably the groups perturbed.
These observations agree with circular dichroism results (6) which indicate that one or more tryptophanyl side chains are in a less czonstrained rnvirornnent relative to the native protein as a result of the acid transition and are therefore capable of forming a coml)lex with i2T-methyl nicotinamide. Chemical modification studies :lt acid pH with 2hydroxy-5-nitrobenzyl bromide (a0) show that tryptophan-118 and l.ryptophan-104 are reactive; ho\vever, tryptophari-26, which is thought to be completely buried, is unexpectedly reactive (see below).
As pointed out by Barman (20), the relatively high concentration of acetone (1070) used in the reaction mixture may induce additional (Lonformationsl changes which could lead to further exposure of groups, n11d thus account for the unexpected reactivity of tryptophan-26.
At alkaline pH three tryptophangl side chains form a complex with N-methyl nicotinamide, showing their exposure at the protein surface. This is in keeping with the results etf circular dichroism measurements which show that the tryptophanyl side chains of t,he alkaline-denat,ured protein are less constrained at high l&I (6) than at pH 6 and that the environment in the vicinity of a disulfide bridge of cr-lactalbumin is changed at high pH (see below).
Although no definitive information is available at present on the tertiary structure of oc-lactalbumin it will be useful to consider the environment of the tryptophanyl side chains as seen in the model of ar-lactalbumin ljroposed by Browne et al. (13) and based on the similarity in amino acid sequence of bovine ar-lactalbumin and hen lysozyme, and to compare it, with the observations of this study and 1)revious work.
The dcncription of the model is taken from the work of Kronman et al. (7) and is presented below. ar-Lactalbumin contains four tryptophnnyl side chains, and the model indicates that only one of them, tryptophan-26, can be considered as being completely buried in the interior of t'he protein.
Tryptophan-118 lies exposed on the surface with little apparent hindrance from other side chains. Tryptophan-104 and tryptophan-60 appear to be less exposed. These residues lie in a cleft-like region of the molecule, homologous with the active site of lysosyme.
Tryptophan-104 is somewhat hindered by tyrosine-103 and by other side chains in the cleft. Tryptophan-60 appears to be less accessible than tryptophan-104, and is surrounded by several hydrophobic residues in addition to tyrosine-103.
The proximity of disulfide bridge 73-92 might also shield tryptophan-60 from contact with solvent. On the basis of the chemical reactivity studies described previously and the environment of the tryptophanyl residues seen in the oc-lactalbumin model, it is most likely that tryptophan-118 is the exposed group forming a complex at pH 6 in the native protein.
Similar reasoning suggests that tryptophan-104 is most likely the other group perturbed by solvent in the native protein, and that this group is subsequently available for complex formation at acid and alkaline pH as a result of the couformational changes which occur under these conditions. The previously described observations from circular dichroism studies at high pH, as well as the environment surrounding tryptophan-60, as seen in t.he molecular modt*l, make it likely that this residue is involved in binding at, alkaline pII along with tryptophan-1 I8 and tryptophm~-104.
The observations of this study when interpreted in conjunction with previous work on oc-lactalbumin provide support, for the proposed molecular model of this protein.
IIowe\rr, a direct verification of the sites involved in binding will require tile preparation and titration of suitably modified derivat,ives.