Chem. Senses 26: 167-177,
2001
© Oxford University Press 2001
Structure–Sweetness Relationship in Thaumatin: Importance of Lysine Residues
Research Institute for Food Science, Kyoto University, Uji, Kyoto 611-0011, Japan
Correspondence to be sent to: Naofumi Kitabatake, Research Institute for Food Science, Kyoto University, Uji, Kyoto 611-0011, Japan. e-mail kitabata{at}food2.food.kyoto-u.ac.jp
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
To clarify the structural basis for the sweetness of thaumatin I, lysine-modified derivatives and carboxyl-group-modified derivatives were prepared by chemical modification followed by chromatographic purification. The sweetness of derivatives was evaluated by sensory analysis. Phosphopyridoxylation of lysine residues Lys78, Lys97, Lys106, Lys137 and Lys187 markedly reduced sweetness. The intensity of sweetness was returned to that of native thaumatin by dephosphorylation of these phosphopyridoxylated lysine residues except Lys106. Pyridoxamine modification of the carboxyl group of Asp21, Glu42, Asp60, Asp129 or Ala207 (C-terminal) did not markedly change sweetness. Analysis by far-UV circular dichroism spectroscopy indicated that the secondary structure of all derivatives remained unchanged, suggesting that the loss of sweetness was not a result of major disruption in protein structure. The five lysine residues, modification of which affected sweetness, are separate and spread over a broad surface region on one side of the thaumatin I molecule. These lysine residues exist in thaumatin, but not in non-sweet thaumatin-like proteins, suggesting that these lysine residues are required for sweetness. These lysine residues may play an important role in sweetness through a multipoint interaction with a putative thaumatin receptor.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Thaumatin is a sweet-tasting protein isolated from the arils of Thaumatococcus daniellii Benth, a plant native to tropical West Africa (Van der Wel and Loeve, 1972
). Naturally occurring thaumatin consists of six closely related proteins (I, II, III, a, b and c), all with a molecular mass of 22 kDa (207 amino acids) (Van der Wel and Loeve, 1972; Ledeboer et al., 1984). Neither protein contains bound carbohydrate or unusual amino acids (Van der Wel and Loeve, 1972). The proteins have an isoelectric point of 12 (Van der Wel and Loeve, 1972). The sweet taste of thaumatins can be detected at threshold amounts 1600 times less than that of sucrose on a weight basis (equivalent to 105-fold less on a molar basis); the threshold values are near 10–4% or 48 nM, making these proteins the most intensely sweet substances known (Van der Wel and Loeve, 1972; Higginbotham, 1979). The three-dimensional structure of thaumatin I has been determined at high resolution (De Vos et al., 1985; Ogata et al., 1992), revealing that the protein consists of three domains: (i) an 11 strand, flattened ß-sandwich (1–53, 85–127 and 178–207, domain I); (ii) a small disulfide-rich region (54–84, domain III); and (iii) a large disulfide-rich region (128–177, domain II). A family of pathogenesis-related (PR) proteins has been referred to as thaumatin-like proteins, due to the remarkable homology between PR-5 proteins and thaumatin. Despite their structural relatedness to thaumatin, none of the PR-5 proteins has been reported to have a sweet taste (Dudler et al., 1994; Peng et al., 1997; Tattersall et al., 1997).
Despite extensive studies in this field, the structural basis of thaumatin’s sweetness is still uncertain. There is no obvious structural similarity between the three-dimensional structures of thaumatin I and another sweet protein, monellin (Ogata et al., 1987; Somoza et al., 1993). Chemical modification studies have suggested that thaumatin’s basicity is not the dominating factor in determining the intensity of its sweetness (Van der Wel and Bel, 1976; Shamil and Beynon, 1990). These previous chemical modification studies did not identify the labeled amino acid residue, and so did not identify a structural determinant for sweetness. Slootstra et al. (Slootstra et al., 1995) identified epitopes for monoclonal antibodies that reacted with both thaumatin and monellin, and suggested that two regions of thaumatin, amino acids 19–29 and 77–84, are the important determinants of sweet taste. A site-directed mutagenesis study has reported that mutation at Lys67, Lys137 or Tyr169 reduced sweetness by five-fold (Kim and Weickmann, 1994); the authors concluded that these residues contribute to determining sweet taste. Thus, the determinants of thaumatin’s sweet taste remain controversial and obscure.
Two intensely sweet proteins, thaumatin and monellin, are both basic proteins (Van der Wel and Loeve, 1972; Morris et al., 1973), suggesting that basicity could be important for sweetness. Sequence analysis has revealed substantial homology between thaumatin and thaumatin-like proteins (Koiwa et al., 1999), but there are differences in some amino acid residues; in particular, there is a characteristic pattern of lysine residues in thaumatin, which suggests that some lysine residues may play an important role in sweetness. Van der Wel has suggested in his review (Van der Wel, 1994) that an increase in the basicity of thaumatin, by amidation of its carboxyl groups, can increase its sweetness. This could suggest that positive charge on a thaumatin molecule could play some role in the increase in sweetness. Therefore, in this study the roles of positive charge and lysine residues in sweetness were investigated. Our approach was to prepare a variety of lysine-modified or carboxyl-group-modified thaumatin derivatives by chemical modification followed by chromatographic purification, and to examine the effects of the modifications on sweetness.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Materials
Thaumatin I was purified from crude thaumatin powder (Sunsweet T; San-Ei Gen FFI, Osaka, Japan) by ion-exchange column chromatography followed by gel filtration column chromatography as described previously (Kaneko and Kitabatake, 1999). L-1-Tosylamido-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin (type XIII), chymotrypsin (type II) and potato acid phosphatase were purchased from Sigma (St Louis, MO). Urea, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), dithiothreitol and iodoacetamide were of specially prepared reagent grade; all other chemicals were of guaranteed grade from Nacalai Tesque Inc. (Kyoto, Japan).
Modification of lysine residues by pyridoxal 5'-phosphate and dephosphorylation
Thaumatin I was incubated with pyridoxal 5'-phosphate (PLP) at 25°C in 50 mM sodium phosphate (NaPi) buffer, pH 7.0 (Lundblad and Noyes, 1984). The final concentrations of thaumatin I and PLP were 200 µM and 3 mM, respectively. After 15 min of reaction, a freshly prepared solution of sodium borohydride (30 mg/ml) was added to the reaction mixture to achieve a final concentration of 0.6 mg/ml. This concentration of sodium borohydride was sufficient to reduce the Schiff base completely without substantial loss of sweetness due to this reagent (addition of sodium borohydride without addition of PLP did not affect the sweetness of thaumatin I).
Phosphopyridoxylated thaumatin I was separated by ion-exchange column chromatography and extensively dialyzed against 5 mM NaPi buffer, pH 5.0, at 4°C. The dialyzed sample (200 mg of protein in 50–60 ml) was applied on to a 1.1 x 5.0 cm column of SP Sephadex C-25 (Pharmacia, Uppsala, Sweden) previously equilibrated with 5 mM NaPi buffer, pH 5.0. Elution was carried out using a stepwise gradient elution with buffers of increasing pH and ionic strength as follows: the flow-through was the first fraction; fractions 2–4 were eluted, respectively, with 100 ml of 5 mM NaPi buffer, pH 7.0; 50 ml of the same buffer containing 50 mM NaCl; and 100 ml of 20 mM NaPi buffer, pH 8.0, containing 100 mM NaCl. These fractions were dialyzed overnight against water at 4°C, lyophilized and stored at 4°C until use.
Mono-phosphopyridoxylated thaumatin I, in which one lysine residue was modified, was further purified using a hydroxyapatite column on fast protein liquid chromatography (FPLC) (Pharmacia). Phosphopyridoxylated thaumatin I was dissolved in 5 mM NaPi buffer, pH 7.0, and applied to a hydroxyapatite column (TSKgel HA-1000, 7.5 x 75 mm; TOSOH, Tokyo, Japan) connected to an FPLC. The flow rate was 0.5 ml/min. Elution was monitored by a UV detector at 278 nm. Elution was performed with the following linear gradient of NaCl in 5 mM NaPi buffer, pH 7.0: (i) 0 M from 0 to 15 min; (ii) 0–275 mM over 40 min; (iii) 275–500 mM over 5 min; and (iv) 500 mM for 10 min. Five peaks (a–e) were observed. Peak fractions were concentrated and desalted using a Centricon YM-10 centrifugal filter device (Millipore, Bedford, MA, USA) at 4°C.
These fractions were further purified by ion-exchange column chromatography on an FPLC equipped with a Cosmogel SP (7.5 x 75 mm; Nacalai Tesque) or a Cosmogel CM (7.5 x 75 mm; Nacalai Tesque) ion-exchange column. To purify fractions a–d, a Cosmogel SP ion-exchange column and two solvents [5 mM NaPi buffer, pH 7.0 (solvent A) and 0.2 M NaPi buffer, pH 8.0 (solvent B)] were used. The flow rate was 0.5 ml/min. Elution was monitored by a UV detector at 278 nm. The elution was performed with a linear gradient from 100% solvent A to 37.5% solvent B over 60 min, and then to 100% solvent B over 10 min. A Cosmogel CM-ion exchange column was used to purify fraction e. The elution was performed as described above, except for the gradient program, which was a linear gradient from 100% solvent A to 15% solvent B over 60 min, and then to 100% solvent B over 20 min. Peak fractions were pooled and desalted using a Centricon YM-10 centrifugal filter device.
5'-Phosphates on the PLP group of phosphopyridoxylated derivatives were eliminated using acid phosphatase (Sigma) (Bingham and Farrell, 1976). In brief, 1 mg of phosphopyridoxylated derivative was incubated with acid phosphatase (5 units) in 500 µl of 0.1 M sodium acetate buffer, pH 5.5, at 37°C for 24 h. The dephosphorylated derivative was separated by ion-exchange column chromatography using SP Sephadex C-25 resin. The sample was eluted with 0.2 M NaPi buffer, pH 8.0. Samples were pooled and desalted using a Centricon YM-10 centrifugal filter device.
Modification of carboxyl groups
The carboxyl groups of thaumatin I were modified with pyridoxamine (PM) using EDC as coupling agent (Carraway and Koshland, 1972). Thaumatin I (225 µM) was incubated for 2 h at 25°C with PM (0.2 M) and EDC (7.0 mM) in 5 mM NaPi buffer, pH 5.0. The reaction mixture was extensively dialyzed at 4°C against 5 mM NaPi buffer, pH 7.0. The dialyzed sample (200 mg of protein in 30–40 ml) was applied on to a SP Sephadex C-25 column (1.1 x 5.0 cm) equilibrated with 5 mM NaPi buffer, pH 7.0. Elution was carried out using stepwise gradient elution with buffers of increasing the ionic strength as follows: the flowthrough was the first fraction, and the second, third and fourth fractions were eluted respectively using 100 ml of 5 mM NaPi buffer, pH 7.0, containing 70 mM NaCl; 100 ml of the same buffer but containing 90 mM NaCl; and 50 ml of 5 mM NaPi buffer, pH 7.0, containing 400 mM NaCl, consecutively. The third fraction contained mono-PM-thaumatin I, whose one carboxyl group was modified with PM; this was dialyzed overnight against water at 4°C. The dialyzed sample was then lyophilized and stored at 4°C until use.
Mono-PM-thaumatin I was further purified by ion-exchange column chromatography on high performance liquid chromatography (HPLC) (LC-6A; Shimadzu, Kyoto, Japan) and then dissolved in 5 mM NaPi buffer, pH 7.0, and applied to a Cosmogel SP (20 x 100 mm; Nacalai tesque) ion-exchange column connected to a HPLC. Elution was monitored by a UV detector at 278 nm. The flow rate was 3.0 ml/min. To perform ion-exchange column chromatography, the same two solvents (solvents A and B) were used and the elution was performed with a linear gradient: 100% solvent A for 5 min, 0–30% solvent B in Solvent A over 5 min, 30–50% over 50 min, then 50–100% over 10 min. Peak fractions were pooled, concentrated and desalted using a Centricon YM-10 centrifugal filter device at 4°C. They were then further purified by rechromatography using ion-exchange column chromatography, performed on HPLC apparatus equipped with a Cosmogel SP (7.5 x 75 mm; Nacalai Tesque) ion-exchange column. The flow rate was 0.5 ml/min. The elution was monitored by a UV detector at 278 nm. The same two solvents as used in the first chromatography were used and the elution was performed with a linear gradient from 100% solvent A to 25% solvent B over 15 min and to 35% solvent B over 60 min, and then to 100% solvent B in 70 min. Peak fractions were pooled and desalted using a Centricon YM-10 centrifugal filter device and were immediately used for further analysis.
Spectrophotometric quantification
The concentration of native thaumatin I and its derivatives were determined spectrophotometrically with a UV-160A spectrophotometer (Shimadzu) in 5 mM NaPi buffer, pH 7.0, using a molar extinction coefficient of 
278 of 17 000/M/cm (Van der Wel and Loeve, 1972). To obtain 278, a mass of 22 209 daltons (Iyenger et al., 1979) was used. Chemical modification did not affect the 278 value, as shown by the bicinchoninic acid (BCA) protein assay (Pierce, Rockford, IL, USA) procedure. Protein-bound PLP was determined at 325 nm using an 325 of 4800/M/cm (Lundblad and Noyes, 1984; Green et al., 1992).
Proteolytic digestion and peptide purification
Tryptic digestion was performed as follows. The lyophilized sample (final concentration 2 mg/ml) was dissolved in 8 M urea and 50 mM Tris–HCl buffer, pH 8.0. The sample was incubated with dithiothreitol (2 mM) for 15 min at 50°C and then carboxamidomethylated by the addition of freshly prepared 0.1 M aqueous iodoacetamide solution to a final concentration of 8 mM. The sample was incubated at 37°C for 20 min and was diluted five-fold with 50 mM Tris–HCl buffer, pH 8.0. TPCK-treated trypsin was added to a ratio of 1:7 (w/w) protease:substrate and the solution was incubated for 22 h at 37°C. Chymotryptic digestion was performed in the same manner as tryptic digestion, except that the ratio of protease:substrate was 1:2 (w/w).
The products of proteolytic digestion were applied on to a reversed-phase HPLC column (Cosmosil 5C18-AR-II: ODS, 4.6 x 150 mm; Nacalai tesque) connected to an HPLC apparatus (L-6200; Hitachi, Tokyo, Japan). The flow rate was 0.42 ml/min and elution was performed in 0.1% trifluoroacetic acid (TFA) with a linear gradient of 10–38% acetonitrile over 35 min, then to 90% acetonitrile in an additional 10 min. The eluent was monitored using an RF-530 fluorescence detector (Shimadzu): excitation, 327 nm; emission, 395 nm for detection of phosphopyridoxylated lysine (Lo et al., 1992), or excitation, 315 nm; emission, 395 nm for pyridoxamine-labeled residues. The peaks were further purified by rechromatography using a reversed-phase HPLC column (Cosmosil 5C18-AR: ODS, 4.6 x 150 mm) and eluted in 0.1% TFA with a linear gradient of 10–50% methanol over 35 min, and then to 90% methanol in an additional 10 min. Flow rate and detection were the same as above.
Sensory analysis
The sweetness threshold of samples was evaluated by means of a triangle test for taste absolute threshold (Horio and Kawamura, 1998). Five healthy subjects, three males and two females, aged 25–48 years, participated in this trial. The samples were tasted in order of their concentration, starting with those of the lowest concentration. After each taste, subjects thoroughly rinsed their mouths with distilled water. The test solutions used were 20, 50, 100, 200 and 500 nM for native thaumatin I; 100, 200, 500, 1000, 2000, 5000 and 10 000 nM for the di- or tri-phosphopyridoxylated derivatives; 20, 50, 100, 200, 500 and 1000 nM for the mono-phosphopyridoxylated derivatives; 20, 50, 100, 200 and 500 nM of the pyridoxylated derivatives; and 20, 50, 100 and 200 nM of the PM-labeled derivatives. Protein was dissolved in 5 ml of 5 mM NaPi buffer, pH 7.0, and kept at room temperature until testing. Relative sweetness (%) was calculated as 100 x (threshold of native thaumatin I)/ (threshold of sample).
Other methods
Non-denaturing PAGE (native PAGE) was performed using a system for basic proteins (Reisfeld et al., 1962). Five micrograms of protein sample was applied on to each lane of a 7.5% homogeneous native polyacrylamide gel. After electrophoresis, gels were stained with Coomassie Brilliant Blue R-250. Circular dichroism (CD) spectra were obtained with a J-720 spectropolarimeter (Jasco, Tokyo, Japan) at 25.0°C in 5 mM NaPi buffer, pH 7.0. Far-UV CD spectra were recorded at a protein concentration of 4.5 µM with a 1 mm cell at wavelengths from 250 to 200 nm. The data were collected three times and are given as the average mean residue ellipticity. Matrix-assisted laser-desorption/ ionization-time-of-flight mass spectrometric (MALDI-TOF-MS) analyses were carried out using a Voyager RP BioSpectrometry Workstation (PerSeptive Biosystems, Inc., Framingham, MS, USA). Automated sequencing of peptides was performed in an Applied Biosystems 492 Protein Sequencer (PE Biosystems, Foster City, CA). Amino acid compositions were analyzed with an amino acid analyzer (L-8500A; Hitachi) after acid hydrolysis in the gas phase with 6 M hydrochloric acid containing 0.1% (v/v) phenol for 22 h at 110°C under vacuum.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Importance of lysine residues
Preparation of mono-, di- and tri-phosphopyridoxylated derivatives
To prepare lysine-modified derivatives to examine the role of lysine residues in sweetness, thaumatin I was reacted with PLP. The reaction of thaumatin I with PLP generated many phosphopyridoxylated derivatives (Figure 1A, lane 2). Each of these derivatives was isolated by SP Sephadex C-25 ion-exchange column chromatography and the resulting samples were analyzed on a native polyacrylamide gel (Figure 1A, lanes 3–5). Judging from the electrophoretic mobility, the first, second and third fractions consisted mainly of tri-, di- and mono-phosphopyridoxylated derivatives, respectively. The number of modified residues was further confirmed by amino acid analysis (data not shown).
|
Sweetness of mono-, di- and tri-phosphopyridoxylated derivatives
The sweetness of native thaumatin I and its derivatives was evaluated. The resulting sweetness threshold values and the relative sweetnesses are listed in Table 1. Native thaumatin I showed intense sweetness at 500 nM (data not shown) and a sweetness threshold of 50 nM. This latter value corresponds to that previously reported by Van der Wel and Loeve (Van der Wel and Loeve, 1972
).
|
To examine the effect of phosphopyridoxylation of lysine residues on sweetness, the sweetness of the mono-, di- and tri-phosphopyridoxylated derivatives was investigated. As the number of the modified lysine residues increased, the sweetness decreased. Mono-, di- and tri-modification of lysine residue reduced sweetness by five-, 30- and >200-fold, respectively. These results suggest that lysine residues are involved in determining sweetness.
The position of functionally important lysine residues
Isolation of specifically phosphopyridoxylated derivatives
It is probable that the sites of modification of derivatives differ from each other, even in a mono-phosphopyridoxylated derivative obtained above. To examine whether a variety of derivatives exist, we subjected mono-phosphopyridoxylated derivative to hydroxyapatite FPLC. We observed five peaks, named a–e (Figure 2), indicating the existence of a variety of derivatives whose modified sites are different from each other. Each peak was further purified by rechromatography using ion-exchange FPLC. The purified derivatives were named phosphopyridoxylated derivatives a–e. Each derivative showed a single peak on analytical ion-exchange FPLC analysis with different retention times (data not shown) and gave a single sharp band on native PAGE (Figure 1B, lanes 3–7), suggesting the homogeneity of these derivatives. Incorporation of one PLP per mole of thaumatin I was further confirmed spectrophotometrically and by amino acid analysis (data not shown).
|
Next, we identified the residue that was modified in these mono-phosphopyridoxylated derivatives. Each derivative was reduced with dithiothreitol, carboxamidemethylated, digested with trypsin (derivatives a and c-e) or chymotrypsin (derivative b) and resolved by reversed-phase HPLC. The elution profiles of protease digests of derivatives a–e are shown in Figure 3A–E, respectively. Each profile showed one or two major peaks (Roman numerals) and some minor peaks. The eluted fractions corresponding to the major peaks were collected and subjected to additional reversed-phase HPLC. The fluorescent peptide was isolated and sequenced (Table 2, peaks A-I to E-I; X represents a modified residue). Table 2 also indicates the corresponding sequences in thaumatin I. The results show that a specific single lysine residue was modified in each derivative, revealing the homogeneity of all derivatives. Aspartic acid was observed at position 113 in this study, although other authors have reported asparagine at this position (Iyengar et al., 1979
). The incorrectness of the assignment by Iyengar et al. has been pointed out before (Lee et al., 1988). The phosphopyridoxylated lysine residues in derivatives a–e were identified as Lys106, Lys78, Lys187, Lys137 and Lys97, respectively. Thus, these phosphopyridoxylated derivatives were named PLP-Lys106-thaumatin I, PLP-Lys78-thaumatin I, PLP-Lys187-thaumatin I, PLP-Lys137-thaumatin I and PLP-Lys97 thaumatin I, respectively.
|
|
Sweetness of specifically phosphopyridoxylated derivatives
To examine the effect of phosphopyridoxylation of each lysine residue on sweetness, the sweetness of specifically phosphopyridoxylated derivatives was investigated (Table 1). The sweetness intensities of the phosphopyridoxylated derivatives were reduced by 60–81%, almost similar to that of mono-phosphopyridoxylated thaumatin and significantly different from that of native thaumatin I. The extent of the decrease in sweetness of these derivatives was comparable to that of mutant thaumatin produced by Kim and Weickmann (Kim and Weickmann, 1994
). Since the CD spectrum of these specifically phosphopyridoxylated derivatives was unchanged (see below), the decrease in sweetness caused by the incorporation of PLP could reflect the importance of these lysine residues for sweetness. There are two possible interpretations of this effect: (i) that lysine residues at these sites are necessary to exhibit sweetness fully; or (ii) that positive charges at these sites are necessary for sweetness, that is, replacement of positive charge (on a lysine residue) by a negative charge (on a phosphopyridoxylated lysine residue) should reduce the sweetness. Role of lysine residues in sweetness
Preparation of specifically pyridoxylated derivatives
To clarify the possibilities above, the phosphate group of phosphopyridoxylated lysine residues was eliminated by phophatase to alter charge of this position from negative back to positive (see Discussion) and the effect on sweetness was examined. Elimination of the phosphate group was confirmed by a decrease in absorbance at 325 nm and by an increase in electrophoretic mobility on native PAGE (Figure 1C), relative to those of the phosphopyridoxylated derivatives. The dephosphorylated derivatives were named pyridoxylated (PL)-Lys106-thaumatin I, PL-Lys78-thaumatin I, PL-Lys187-thaumatin I, PL Lys137-thaumatin I, and PL-Lys97-thaumatin I, respectively, because a PL-lysine residue was formed from a PLP-lysine residue.
Sweetness of specifically pyridoxylated derivatives
The sweetness of the pyridoxylated derivatives is shown in Table 1. All of the pyridoxylated derivatives, except the PL-Lys106-thaumatin I derivatives, exhibited sweetness equivalent to that of native thaumatin I, i.e., the sweetness intensity of these derivatives was restored to that of native thaumatin by dephosphorylation of PLP-thaumatin.
The sweetness of PL-Lys78-, PL-Lys97-, PL-Lys137- and PL-Lys187-thaumatin I derivatives was equal to that of native thaumatin I. This indicates that the increase in sweetness was due to the positive charge on these lysine residues. Therefore, positive charges at Lys78, 97, 137 and 187 are important for sweetness and the distance between the 
-carbon and the -amino group of these lysine residues may not be important for sweetness.
In contrast, the sweetness of PLP-Lys106-thaumatin I was not affected by dephosphorylation. The effect of dephosphorylation of PLP-Lys106 on sweetness was different from that of the other four residues. This may suggest that, at position 106, the side-chain structure is important for sweetness.
Effect on sweetness of elimination of a negative charge from the carboxyl group
Preparation of specific carboxyl-group-modified derivatives
In the experiments above, it was shown that the positive charges distributed on the thaumatin molecule are important for sweetness. The effect of elimination of negative charge from the thaumatin molecule was examined next. The carboxyl group of thaumatin was modified with pyridoxamine (PM). By this modification, a pyridoxal moiety was attached to a carboxyl group and then a negative charge was shielded. The reaction generated a variety of modified derivatives. To investigate if the effect on sweetness depended on the position of the modified residue, a mono-modified derivative was purified by SP Sephadex C-25 column chromatography followed by ion-exchange HPLC (Figure 4). The peak indicated by an arrow represents unmodified thaumatin I. Three other peaks, denoted a–c, were further purified by rechromatography using ion-exchange HPLC. The samples obtained were called PM-thaumatin a–c, respectively. Although PM-thaumatin b was heterogeneous (see below), we were unable to separate this fraction further. Incorporation of one mole of PM into the thaumatin I molecule was confirmed by MALDI-TOF-MS analysis (data not shown). Amino acid analysis and MALDI-TOF-MS analysis confirmed that there was no occurrence of side reactions, including modification of tyrosine residues (Carraway and Koshland, 1968) and chemical cross-linking (Carraway and Koshland, 1972).
|
To identify the modified residues, each derivative was reduced with dithiothreitol, carboxamidemethylated, digested with trypsin (derivatives a and c) or chymotrypsin (derivative b) and subjected to reversed-phase HPLC. The elution profiles are shown in Figure 3F–H. The profiles of tryptic digests of derivatives a and c showed one major peak (Roman numerals) and three minor peaks (arrows). The profile of chymotryptic digestion of derivative b showed five large peaks (Roman numerals) and three small peaks (arrows). The peaks marked with arrows also appeared in a control elution profiles (proteolytic digests of unmodified thaumatin I) and so were not attributed to any modified species. The eluted fractions corresponding to the peaks marked with Roman numerals were collected, and the fluorescent peptide was isolated by additional reverse-phase HPLC purification (data not shown). The amino acid sequences of the peptides in the isolated peaks are shown in Table 2 (peaks F-I to H-I; X represents a modified residue). The modification at Ala207 (C-terminal) in peptide H-I was further confirmed by a comparison of the protonated ion peaks observed at m/z 946.4 with the theoretical values of 946.1 for (M + H+), which was calculated from a peptide fragment from Val201 to Ala207 plus one molecule of PM. The results showed that Asp60 and Ala207 were specifically modified in derivatives a and c, respectively, indicating the homogeneity of these derivatives. In contrast, three modified residues, Asp21, Glu42, and Asp129, were found in derivative b. Together with the fact that one PM molecule was incorporated per one thaumatin I molecule, three kinds of mono-PM-thaumatin I were involved in the derivative b. It was possible to evaluate the relative proportions of each derivative by the fluorescence intensity of chymotryptic peptides: we found 44% of Asp21-modified derivative, 31% of Glu42-modified derivative, and 25% of Asp129-modified derivative. Thus, these derivatives a–c were denoted as PM-Asp60-thaumatin I, PM-Asp21/Glu42/Asp129-thaumatin I, and PM-Ala207-thaumatin I, respectively. Native PAGE analysis of these derivatives showed the increase in net charge of these molecules (Figure 1D).
Sweetness of specific carboxyl-group-modified derivatives
The effect on sweetness of PM modification of the carboxyl group was examined (Table 1). Although modification of the carboxyl group of Asp21, Glu42, Asp60, Asp129 or Ala207 (C-terminal) did not significantly affect the sweetness of thaumatin I, PM-Asp21/Glu42/Asp129-thaumatin I was a little sweeter. It is certain that elimination of negative charge or a decrease in the number of negative charges on the surface of thaumatin molecule does increase sweetness in specific conditions, but the details of how altering the surface charge of a thaumatin molecule affects the sweetness, including the specificity of the topological site, need to be clarified.
Structural properties of thaumatin I derivatives
To determine whether the chemical modifications affected the secondary structure of thaumatin I, far-UV CD spectra were measured. None of the modifications were associated with any detectable changes in the dichroic content (Figure 5), indicating that no major conformational change was induced by the modifications. These observations show that the loss of sweetness must have been due to local changes in structure, rather than being the result of major disruptions in protein structure.
|
Location of modified residues on the three-dimensional structure of thaumatin
The location of modified residues on the three-dimensional structure of thaumatin is shown in Figure 6. Lys106 is represented in red and Lys78, Lys97, Lys137 and Lys187 in blue. Five other positions, Asp21, Glu42, Asp60, Asp129 and Ala207 (C-terminal), are colored green. One side of the thaumatin molecule (shown at the front in Figure 6A and on the left in Figure 6B) contains all of the important lysine residues shown in this study. Asp60, Asp129 and Ala207 (C-terminal) are located on the opposite side. Lys78, Lys97, Lys137 and Lys187 are located in the flexible loop region, but Lys106 is not. A previous study suggested that the determinant of thaumatin’s sweet taste was located in the flexible loop region (Kim and Weickmann, 1994), but our study does not support this suggestion.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In this study, we investigated the role of positive charge and lysine residues in determining sweetness. Although thaumatin I contains 11 lysine residues (Iyengar et al., 1979
), only five specifically phosphopyridoxylated derivatives, modified on Lys78, Lys97, Lys106, Lys137 or Lys187, were isolated in this study. This indicates that these lysine residues are more reactive with PLP than are the other six lysine residues, which may indicate (Lundblad and Noyes, 1984) that these modified lysine residues are located on the surface of the molecule. Indeed, Kim and Weickmann (Kim and Weickmann, 1994) have reported that these five lysine residues are exposed more than the other six lysine residues. This could explain why these lysine residues play an important role in the perceived sweetness of thaumatin.
The electric charge of lysine residues was changed in this study. First, lysine residues were phosphopyridoxylated, which introduces a negative charge into the phosphate group on lysine residues. Second, the phosphate group was eliminated using phosphatase. The pyridoxal moiety contains one positively charged group (the NH group in the pyridine ring) and one negatively charged group (the OH group in phenol) (Fischer et al., 1958). In addition, one positively charged group (NH of imino group) is thought to be involved in a pyridoxylated lysine residue, since an alkylamino group would be positively charged (Van der Wel, 1994). Thus, elimination of the phosphate group alters the charge at this position from negative back to positive. We examined how these changes in the electric charge of lysine residues affected sweetness.
Phosphopyridoxylation of Lys78, Lys97, Lys106, Lys137 or Lys187 reduced sweetness to 
20% and dephosphorylation of these phosphopyridoxylated lysine residues restored the intensity of sweetness to that of native thaumatin, except in the case of Lys106. This shows that the positive charges on these sites are important in determining sweetness. Although others have reported that the net charge of the molecule seems not to be the dominant factor determining the intensity of sweetness (Shamil and Beynon, 1990; Van der Wel, 1994), we have shown that the positive charges on Lys78, Lys97, Lys137 and Lys187 are required for sweetness.
The sweetness of PLP-Lys106-thaumatin I was not returned to that of native thaumatin by dephosphorylation of PLP-Lys106 , in contrast to the other four residues mentioned above. This may be because Lys106 itself plays a critical role in sweetness, or because an amino acid residue that is critical for sweetness is located near Lys106, or there may be other reasons. Although more extensive studies are needed to elucidate the exact role of Lys106 in determining sweetness, it seems that the role of Lys78, Lys97, Lys137 and Lys187 is, at least in part, to confer the positive charges required for sweetness.
We also investigated the role of carboxyl groups in determining sweetness. Van der Wel has suggested that an increase in the basicity of thaumatin, by amidation of its carboxyl groups, can increase its sweetness to some extent (Van der Wel, 1994). However, details have been obscure until now. Our study has shown that PM-Asp21/Glu42/Asp129-thaumatin I is sweeter than native thaumatin I, indicating that altering the charge of these sites from negative to neutral increases the sweetness. Elimination of negative charge from some region of the surface of thaumatin molecule could, therefore, increase the sweetness of thaumatin or reduce the threshold of sweetness, and nonspecific positive charge(s) on the surface of thaumatin molecule could affect sweetness. It would be interesting to determine if altering the charge of a specific carboxyl group from negative to positive increased the sweetness. Although the present study showed that Asp60 and Ala207 (C-terminal) were not critical for sweetness, more extensive studies are needed to examine the roles of carboxyl group in determiming sweetness.
Of the five lysine residues we have shown to be important for sweetness, Lys106 is located in a cleft region in the three-dimensional structure of thaumatin, and Lys78, Lys97, Lys137 and Lys187 surround the cleft region. An X-ray crystallographic study of thaumatin and two thaumatin-like proteins, PR-5d and zeamatin, showed that the cleft region and the surrounding lysine residues in thaumatin are more basic than the corresponding region in PR-5d and zeamatin (Figure 7), although there were no large differences in the surface electrostatic potential of the opposite side among these three proteins. The corresponding positions are occupied by acidic or neutral residues in six other thaumatin-like proteins, namely PR-5c, NP24, PR-5a, Arabidopsis thaliana thaumatin-like protein, PR-5K receptor kinase and Oryza sativa thaumatin-like protein (Koiwa et al., 1999). These could be important features, because no other members of the family have been reported to be sweet (Dudler et al., 1994; Koiwa et al., 1997; Peng et al., 1997; Tattersall et al., 1997). Thus, the basicity of the cleft region and the four lysine residues surrounding the cleft region seem likely to be structurally required for sweetness.
|
The present results suggest a possible mechanism for the determination of sweetness in thaumatin. In this study, monophosphopyridoxylation of Lys78, Lys97, Lys106, Lys137 or Lys187 reduced sweetness significantly, but not completely. The results showed that these lysine residues contributed to sweetness almost equally, and increase in phosphopyridoxylation of lysine residue reduced the sweetness markedly. These findings suggest that sweetness could be determined by multipoint interaction of thaumatin with a putative receptor molecule. It is likely that the low threshold value of thaumatin, 50 nM, is accomplished by a high-affinity interaction between thaumatin and a putative receptor site. A number of high-affinity interactions occur through multipoint interaction, in which many amino acid side chains on a broad surface region are in contact with their counterparts, each one contributing partly to the total binding energy; examples include interactions between interleukin-3 and its receptor (Klein et al., 1997
), antigen and antibody (Amit et al., 1986; Colman et al., 1987), neurotoxin and its receptor (Kharrat et al., 1989; Zilberberg et al., 1997), cAMP dependent protein kinase and its peptide inhibitor (Knighton et al., 1991), and collagen and integrin (Kamata et al. 1999). Thus, it is conceivable that the mechanism by which the sweetness of thaumatin is perceived also occurs through multipoint interaction of the molecule with a putative receptor molecule.
For other sweet proteins, monellin and brazzein, it has also been suggested that many residues spread over a broad surface region of the molecule are important in determining sweetness (Kohmura et al., 1992; Somoza et al., 1995; Assadi-Porter et al., 2000). It has been suggested that two remote regions of the molecule determine sweetness. These findings are not inconsistent with a multipoint interaction model.
The five lysine residues, modification of which reduced sweetness, are all located on one side of the thaumatin molecule, i.e., the cleft-containing side, while the other residues studied, modification of which did not affect sweetness, are located mainly on the opposite side. These results suggest that the cleft-containing side of the thaumatin molecule is important for sweetness; this may, therefore, be the side of the molecule that interacts with a putative thaumatin receptor. This is not inconsistent with previous studies. Lys67 and Lys137, mutation of which reduces the sweetness of thaumatin (Kim and Weickmann, 1994), are located on the cleft-containing side. It is likely that many amino acid residues on the cleft-containing side of the molecule contribute, by some cumulative effect, to determining sweetness through multipoint interaction with the putative receptor molecule.
In this study, PLP, PL and PM were used as modifying agents. Since these compounds are natural coenzyme forms of pyridoxine (vitamin B6), thaumatin derivatives modified with these compounds were considered safe for use in taste tests. The additional benefit of using these compounds is their fluorescence, which makes it easier to detect modified residues. Further studies of other chemical modifications or site-directed mutagenesis of residues on the cleft-containing side of the molecule may be useful for characterizing further the factors that determine the sweetness of thaumatin.
|
Acknowledgments |
|---|
This work was supported by a Grant-in-Aid for Scientific Research (C) and a Grant-in-Aid for the ‘Research for the Future’ Program from the Japan Society for the Promotion of Science, and by the San-Ei Gen Foundation for Food Chemical Research.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Amit, A.G., Mariuzza, R.A., Phillips, S.E. and Poljak, R.J. (1986) Three-dimensional structure of an antigen–antibody complex at 2.8 Å resolution. Science, 233, 747–753.
Assadi-Porter, F.M., Aceti, D.J. and Markley, J.L. (2000) Sweetness determinant sites of brazzein, a small, heat-stable, sweet-tasting protein. Arch. Biochem. Biophys., 376, 259–265.[Web of Science][Medline]
Bingham, E.W. and Farrell, H.J. (1976) Removal of phosphate groups from casein with potato acid phosphatase. Biochim. Biophys. Acta, 429, 448–460.[Medline]
Carraway, K.L. and Koshland, D.E.J. (1968) Reaction of tyrosine residues in proteins with carbodiimide reagents. Biochim. Biophys. Acta, 160, 272–274.[Medline]
Carraway, K.L. and Koshland, D.E.J. (1972) Carbodiimide modification of proteins. Methods Enzymol., 25, 616–623.
Colman, P.M., Laver, W.G., Varghese, J.N., Baker, A.T., Tulloch, P.A., Air, G.M. and Webster, R.G. (1987) Three-dimensional structure of a complex of antibody with influenza virus neuraminidase. Nature, 326, 358–363.[Medline]
De Vos, A.M., Hatada, M., Van der Wel, H., Krabbendam, H., Peerdeman, A.F. and Kim, S.H. (1985) Three-dimensional structure of thaumatin I, an intensely sweet protein. Proc. Natl Acad. Sci. USA, 82, 1406–1409.
Dudler, R., Mauch, F. and Reimmann, C. (1994) Thaumatin-like proteins. In Witty, M. and Higginbotham, J.D. (ed), Thaumatin. CRC Press, Boca Raton, FL, pp. 193–199.
Fischer, E.H., Kent, A.B., Snyder, E.R. and Krebs, E.G. (1958) The reaction of sodium borohydride with muscle phosphorylase. J. Am. Chem. Soc., 80, 2906–2907.
Green, P.C., Latshaw, S.P., Ladror, U.S. and Kemp, R.G. (1992) Identification of critical lysyl residues in the pyrophosphate-dependent phosphofructo-1-kinase of Propionibacterium freudenreichii. Biochemistry, 31, 4815–4821.[Medline]
Higginbotham, J. (1979) Protein sweeteners. In Hough, C.A.M., Parker, .J. and Viltos, A.J. (eds), Developments in Sweeteners 1. Applied Science Publishers, London, pp. 87–123.
Horio, T. and Kawamura, Y. (1998) Influence of physical exercise on human preferences for various taste solutions. Chem. Senses, 23, 417–421.
Iyengar, R.B., Smits, P., Van der Ouderaa, F., Van der Wel, H., Van Brouwershaven, J., Ravestein, P., Richters, G. and Van Wassenaar, P.D. (1979) The complete amino-acid sequence of the sweet protein thaumatin I. Eur. J. Biochem., 96, 193–204.[Web of Science][Medline]
Kamata, T., Liddington, R.C. and Takada, Y. (1999) Interaction between collagen and the 2 I-domain of integrin 2ß1. Critical role of conserved residues in the metal ion-dependent adhesion site (MIDAS) region. J. Biol. Chem., 274, 32108–32111.
Kaneko, R. and Kitabatake, N. (1999) Heat-induced formation of intermolecular disulfide linkages between thaumatin molecules that do not contain cysteine residues. J. Agric. Food Chem., 47, 4950–4955.[Web of Science][Medline]
Kharrat, R., Darbon, H., Rochat, H. and Granier, C. (1989) Structure/ activity relationships of scorpion -toxins. Multiple residues contribute to the interaction with receptors. Eur. J. Biochem., 181, 381–390.[Web of Science][Medline]
Kim, S.H. and Weickmann, J.L. (1994) Crystal structure of thaumatin I and its correlation to biochmical and mutational studies. In Witty, M. and Higginbotham, J.D. (eds), Thaumatin. CRC Press, Boca Raton, FL, pp. 135–149.
Klein, B.K., Feng, Y., McWherter, C.A., Hood, W.F., Paik, K. and McKearn, J.P. (1997) The receptor binding site of human interleukin-3 defined by mutagenesis and molecular modeling. J. Biol. Chem., 272, 22630–22641.
Knighton, D.R., Zheng, J.H., Ten, E.L., Xuong, N.H., Taylor, S.S. and Sowadski, J.M. (1991) Structure of a peptide inhibitor bound to the catalytic subunit of cyclic adenosine monophosphate-dependent protein kinase. Science, 253, 414–420.
Ko, T.P., Day, J., Greenwood, A. and McPherson, A. (1994) Structures of three crystal forms of the sweet protein thaumatin. Acta Crystallogr., D50, 813–825.
Kohmura, M., Nio, N. and Ariyoshi, Y. (1992) Highly probable active site of the sweet protein monellin. Biosci. Biotechnol. Biochem., 56, 1937–1942.[Medline]
Koiwa, H., Kato, H., Nakatsu, T., Oda, J., Yamada, Y. and Sato, F. (1997) Purification and characterization of tobbacco pathogenesis-related protein PR-5d, an antifungal thaumatin-like protein. Plant Cell Physiol., 38, 783–791.
Koiwa, H., Kato, H., Nakatsu, T., Oda, J., Yamada, Y. and Sato, F. (1999) Crystal structure of tobacco PR-5d protein at 1.8 Å resolution reveals a conserved acidic cleft structure in antifungal thaumatin-like proteins. J. Mol. Biol., 286, 1137–1145.[Web of Science][Medline]
Kraulis, P. (1991) MOLSCRIPT: A program to produce both detailed and schematic plots of protein structures. J. Appl. Crystallogr., 24, 946–950.
Ledeboer, A.M., Verrips, C.T. and Dekker, B.M. (1984) Cloning of the natural gene for the sweet-tasting plant protein thaumatin. Gene, 30, 23–32.[Web of Science][Medline]
Lee, J.H., Weickmann, J.L., Koduri, R.K., Ghosh, D.P., Saito, K., Blair, L.C., Date, T., Lai, J.S., Hollenberg, S.M. and Kendall, R.L. (1988) Expression of synthetic thaumatin genes in yeast. Biochemistry, 27, 5101–5107.[Medline]
Lo, B.M., Petruzzelli, R., Reale, L., Ricci, G., Barra, D. and Federici, G. (1992) Chemical modification of human placental glutathione transferase by pyridoxal 5'-phosphate. Biochim. Biophys. Acta, 1121, 167–172.[Medline]
Lundblad, R.L. and Noyes, C.M. (1984) The modification of lysine. In Lundblad, R.L. and Noyes, C.M. (eds), Chemical Reagents for Protein Modification. CRC Press, Boca Raton, FL, pp. 127–170.
Merritt, E.A. and Murphy, E.P. (1994) Raster3D version 2.0. A program for photorealistic molecular graphics. Acta Crystallogr., D50, 869–873.
Morris, J.A., Martenson, R., Deibler, G. and Cagan, R.H. (1973) Characterization of monellin, a protein that tastes sweet. J. Biol. Chem., 248, 534–539.
Nicholls, A. and Honig, B. (1991) A rapid finite difference algorithm, utilizing successive over relaxation to solve the Poisson–Boltzman equation. J. Comput. Chem., 12, 435–445.[Web of Science]
Ogata, C., Hatada, M., Tomlinson, G., Shin, W.C. and Kim, S.H. (1987) Crystal structure of the intensely sweet protein monellin. Nature, 328, 739–742.[Medline]
Ogata, C.M., Gordon, P.F., De Vos, A.M. and Kim, S.H. (1992) Crystal structure of a sweet tasting protein thaumatin I, at 1.65 Å resolution. J. Mol. Biol., 228, 893–908.[Web of Science][Medline]
Peng, Z., Pocock, K.F., Waters, E.J., Francis, I.L. and Williams, P.J. (1997) Taste properties of grape (Vitis vinifera) pathogenesis-related proteins isolated from wine. J. Agric. Food Chem., 45, 4639–4643.
Reisfeld, R.A., Lewis, U.J. and Williams, D.E. (1962) Disk electrophoresis of basic proteins and peptides on polyacrylamide gels. Nature, 195, 281–283.[Medline]
Shamil, S. and Beynon, R.J. (1990) A structure–activity study of thaumatin using pyridoxal 5'-phosphate (PLP) as a probe. Chem. Senses, 15, 457–469.
Slootstra, J.W., De Geus, P., Haas, H., Verrips, C.T. and Meloen, R.H. (1995) Possible active site of the sweet-tasting protein thaumatin. Chem. Senses, 20, 535–543.
Somoza, J.R., Jiang, F., Tong, L., Kang, C.H., Cho, J.M. and Kim, S.H. (1993) Two crystal structures of a potently sweet protein, natural monellin, at 2.75 Å resolution and single-chain monellin at 1.7 Å resolution. J. Mol. Biol., 234, 390–404.[Web of Science][Medline]
Somoza, J.R., Cho, J.M. and Kim, S.H. (1995) The taste-active regions of monellin, a potently sweet protein. Chem. Senses, 20, 61–68.
Tattersall, D.B., Heeswijck R.V. and Hoj, P.B. (1997) Identification and characterization of a fruit-specific, thaumatin-like protein that accumulates at very high levels in conjunction with the onset of sugar accumulation and berry softening in grapes. Plant Physiol., 114, 759–769.[Abstract]
Van der Wel, H. (1994) Structure–activity relationships in the thaumatin molecule. In Witty, M. and Higginbotham, J.D. (eds), Thaumatin. CRC Press, Boca Raton, FL, pp. 115–122.
Van der Wel, H. and Bel, W.J. (1976) Effect of acetylation and methylation on the sweetness intensity of thaumatin I. Chem. Senses, 2, 211–218.
Van der Wel, H. and Loeve, K. (1972) Isolation and characterization of thaumatin I and II, the sweet-tasting proteins from Thaumatococcus daniellii Benth. Eur. J. Biochem., 31, 221–225.[Web of Science][Medline]
Zilberberg, N., Froy, O., Loret, E., Cestele, S., Arad, D., Gordon, D. and Gurevitz, M. (1997) Identification of structural elements of a scorpion -neurotoxin important for receptor site recognition. J. Biol. Chem., 272, 14810–14816.
Accepted September 22, 2000
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. Masuda, N. Ide, and N. Kitabatake Structure-Sweetness Relationship in Egg White Lysozyme: Role of Lysine and Arginine Residues on the Elicitation of Lysozyme Sweetness Chem Senses, October 1, 2005; 30(8): 667 - 681. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Sano, T. Egashira, Y. Kinekawa, and N. Kitabatake Astringency of Bovine Milk Whey Protein J Dairy Sci, July 1, 2005; 88(7): 2312 - 2317. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Masuda, N. Ide, and N. Kitabatake Effects of Chemical Modification of Lysine Residues on the Sweetness of Lysozyme Chem Senses, March 1, 2005; 30(3): 253 - 264. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. M. Assadi-Porter, F. Abildgaard, H. Blad, and J. L. Markley Correlation of the Sweetness of Variants of the Protein Brazzein with Patterns of Hydrogen Bonds Detected by NMR Spectroscopy J. Biol. Chem., August 15, 2003; 278(33): 31331 - 31339. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Jin, V. Danilova, F. M. Assadi-Porter, J. L. Markley, and G. Hellekant Monkey Electrophysiological and Human Psychophysical Responses to Mutants of the Sweet Protein Brazzein: Delineating Brazzein Sweetness Chem Senses, July 1, 2003; 28(6): 491 - 498. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||