Molecular Capture Using Hydrophobic Space Formed By Tea Gallated Catechins in Water
Ishizu T
Published on: 2022-01-13
Abstract
The component of a precipitate resulted by creaming down, which was made from a catechin mixture and caffeine, was determined by an integrated value of a proton signals for H2 of tea catechins in the quantitative 1H NMR spectrum. The results showed that gallated catechins formed a precipitate by creaming down more predominantly than non-gallated catechins.
X-ray crystallographic analysis showed that the gallated catechin (-)-epigallocatechin-3-O-gallate (EGCg) formed 2:2 complex with caffeine, in which the caffeine moieties were located in the strong hydrophobic space surrounded by the bottom and top walls of the B’ rings of EGCg moieties and left and right walls of A and B rings of EGCg moieties. Therefore, the solubility of the 2:2 complex of EGCg and caffeine in water reduced rapidly, causing precipitation due to the creaming-down phenomenon.
The precipitates of the 2:2 complex occurred when aqueous solution of the heterocyclic compounds (Table 3) was poured into an aqueous solution of EGCg. The molecular capture abilities of the heterocyclic compounds were evaluated with the ratios of the amount of the heterocyclic compounds included in the precipitates to that of the total heterocyclic compounds used.
Furthermore, the molecular capture abilities of diketopiperazine cyclo(Pro-Xxx) (Xxx: amino acid residue) were evaluated, and the correlation between the chemical structures of cyclo(Pro-Xxx) and their molecular capture abilities was investigated.
When cyclo(Pro-Xxx) was taken into the hydrophobic space formed by EGCg to form the 2:2 complex of EGCg, conformation of cyclo(Pro-Xxx) other than cyclo(D-Pro-L-Ala) still maintained. And conformation of cyclo(D-Pro-L-Ala) changed, that is, its 3-position methyl group changed from the axial position to the equatorial position due to steric hindrance by EGCg moieties.
Keywords
Creaming down; (-)-epigallocatechin-3-O-gallate; Caffeine; Molecular capture ability; Q quantitative 1H NMR; X-ray crystallographic analysisIntroduction
Tea has been drunk in many countries throughout the world since ancient times for its taste, and to maintain and improve health [1]. Tea is commonly prepared by pouring hot or boiling water over leaves of the tea plant, Camellia sinensis, Theaceae, which includes caffeine, catechins, vitamins, theanine, etc.
The main eight tea catechins are classified into four categories by the existence of a galloyl group on the oxygen atom at the C3 position and the relative stereochemistry between the C2 and C3 positions: 2,3-cis gallate-type, 2,3-trans gallate-type, 2,3-cis-non gallate-type, and 2,3-trans non-gallate-type (Figure 1) [2]. Generally speaking, gallate-type catechins show higher physiological activities than non-gallate-type catechins [3-6].
When a hot tea beverage cools down, brown-white turbidity and precipitate occurs in the tea. This phenomenon is called “creaming down (reaction)”. Since the creaming down is a trigger that reduces the original taste and flavor of tea, it is one of the most serious problems in making a tea beverage. It has been suggested that caffeine formed precipitates of a cream down with tea polyphenols such as catechins, theaflavins and thearubigins [7].
Ina et al. reported that all the main signals were assigned to tea catechins such as 2,3-cis gallated catechins (-)-epigallocatechin-3-O-gallate (EGCg) and (-)-epicatechin-3-O- gallate (ECg) and caffeine in the 13C NMR spectrum of a hot water solution of a precipitate formed by the creaming down of a tea infusion [8]. Ina et al. also reported that creaming down eventually occurs when an aqueous caffeine solution is poured into an aqueous solution of EGCg, which is most abundant in tea catechins [8].
However, creaming down in tea beverages has not been chemically elucidated sufficiently. Thus, we investigated what kind of catechins predominantly precipitated using a catechin mixture and caffeine by the quantitative 1H NMR. The catechin mixture (purchased from Nagara Science Co., Ltd) is a fraction of extracted catechins included in green tea. We attempted crystallization of a precipitate formed by the creaming down of EGCg and caffeine, the stereochemical structure of the complex of EGCg and caffeine was investigated using X-ray crystallographic analysis. Furthermore, mechanism of creaming down was investigated using molecular interactions forming between tea catechins and caffeine.
Based on the mechanism of the creaming down found in this way, molecular capture of heterocyclic compouds by EGCg from an aqueous solution was investigated.
Precipitates occurred when aqueous solution of a variety of heterocyclic compounds poured into aqueous solution of EGCg. The molecular capture abilities of the heterocyclic compounds were evaluated with the ratio of the amount of the heterocyclic compounds included in the precipitates to that of the total heterocyclic compounds used.

Figure 1: The Eight Major Tea Catechins and Caffeine.
Results And Discussion
Analysis of the Precipitate of Creaming Down Made From Catechin Mixture and Caffeine
The content of tea catechins contained in the catechin mixture was investigated by the integrated value of the proton signal for H2 of each tea catechin in the quantitative 1H NMR spectrum [9]. No overlap of the methine H2 proton signal of each tea catechin in deuterated acetone (acetone-d6) was observed in the 1H NMR spectrum [10]. Then amount of each tea catechin included in the catechin mixture was measured with the integrated value of the H2 proton signal using the methyl proton signal (singlet) of Boc-glycine as an internal standard (Figure 2).
As shown in Figure 2, the H2 proton signal of 2,3-cis catechins of ECg, EGCg, EC, and EGC was observed as singlet-like, whereas that of 2,3-trans catechins of (+)-catechin (CA) and (+)-gallocatechin (GC) was observed as a doublet in the 1H NMR spectrum. Such a difference in coupling patterns resulted from differences in configuration. Judging from the coupling constant J2,3 of ECg, EGCg, EC, and EGC being ca. 0 Hz, it was thought that their dihedral angles ∠H2-C2-C3-H3 were approximately 90°, judging from the Kurplus equation [12]. On the other hand, since J2,3 of CA and GC were 7.8 Hz and 7.4 Hz, respectively, their dihedral angles ∠H2-C2-C3-H3 were expected to be considerably larger than 90°[12,13].

Figure 2: The methine proton H2 signal of each catechin of the catechin mixture in1H NMR Spectrum.
The content of the catechin mixture is shown in Table 1, indicating that it included a large amount of 2,3-cis-type catechins, such as EGCg, ECG, EC, and ECg [9]. Thus, a simple and easy method for quantitative analysis of a sample containing many kinds of catechins using the 1H NMR spectrum was developed. The reported contents of catechins in green tea are listed in Table 1 [11].
Table 1: The content of tea catechins in the catechin mixture.
|
EGCg |
EGC |
EC |
ECg |
GC |
CA |
|
|
Content (mg) |
5.13 |
2.7 |
0.99 |
0.97 |
0.43 |
0.17 |
|
Amount of substance (mmol) |
0.011 |
0.009 |
0.003 |
0.002 |
0.001 |
0.001 |
|
Relative content (%) |
49.37 |
25.99 |
9.53 |
9.34 |
4.14 |
1.64 |
|
Relative content (%) of green tea [17]) |
59.1 |
19.3 |
6.4 |
13.7 |
1.6 |
Equimolar amounts of the catechin mixture (15.00 mg, 2.76×10-2 mmol) and caffeine (5.36 mg, 2.76×10-2 mmol) were dissolved in D2O (100 ml) at 90? and left at room temperature for a day. The solution was divided into the supernatant liquid and a sticky precipitate, which is a precipitate formed by creaming down. The content of the supernatant liquid and the sticky precipitate was determined by a similar analytical method using an integrated value of the H2 proton signal of tea catechins and the H8 proton signal of caffeine in the quantitative 1H NMR spectra. The content of various tea catechins and caffeine was listed in Table 2.
As shown in Table 2, more than 80% of 2,3-cis gallated catechins ECg and EGCg included in the catechin mixture were present in the precipitate formed by creaming down, whereas only 45.9?64.7% of non-gallate-type catechins, EC, EGC, CA, and GC, were present. Quantitative analysis of the precipitate of the creaming from a catechin mixture and caffeine suggested that gallated catechins were predominantly responsible for creaming rather than non-gallated catechins [9].
Table 2: The content of tea catechins and caffeine in the supernatant liquid and the sticky precipitate made from catechin mixture and caffeine.
|
|
EGCg |
EGC |
EC |
ECg |
GC |
CA |
Caffeine |
|
Sample Content (mg) |
5.13 |
2.7 |
0.99 |
0.97 |
0.43 |
0.17 |
5.36 |
|
Amount of substance (mmol) |
0.0112 |
0.0088 |
0.0034 |
0.0022 |
0.0014 |
0.0006 |
0.0276 |
|
Sticky precipitate Content (mg) |
4.22 |
1.24 |
0.53 |
0.81 |
0.24 |
0.11 |
4.46 |
|
Amount of substance (mmol) |
0.0092 |
0.004 |
0.0018 |
0.0018 |
0.0008 |
0.0004 |
0.023 |
|
Relative content (%) |
82.3 |
45.9 |
53.5 |
83.5 |
55.8 |
64.7 |
83.2 |
|
Supernatant liquid Content (mg) |
0.88 |
1.46 |
0.45 |
0.13 |
0.19 |
0.06 |
0.9 |
|
Amount of substance (mmol) |
0.0019 |
0.0048 |
0.0016 |
0.0003 |
0.0006 |
0.0002 |
0.0046 |
|
Relative content (%) |
17.2 |
54.1 |
45.5 |
13.4 |
44.2 |
35.3 |
16.8 |
Analysis of the Complex of Tea Catechin Egcg, EC and Caffeine by X-Ray Crystallography
The molecular interactions between tea catechins and caffeine in the precipitate formed by creaming were investigated by X-ray crystallographic analysis. Crystals of the complexes of EGCg, EC and caffeine were prepared, and X-ray crystallographic analysis of the complexes was performed to determine the crystal stereochemical structures and to elucidate the detailed non-covalent interactions among EGCg, EC and caffeine moieties. Furthermore, we investigated differences in the stereochemical structures and non-covalent interactions, whether having a galloyl group or not.
Analysis of the Complex of Egcg and Caffeine by X-Ray Crystallography
Equimolecular amounts of EGCg and caffeine were dissolved in water by heating, and the aqueous solution was left at room temperature. It divided into a supernatant liquid and a sticky precipitate formed by creaming down, which were left at 10? for about 3 months. As a result, the sticky precipitate crystallized slowly to give a colorless block crystal, which was determined to be a 2:2 complex of caffeine and EGCg by X-ray crystallographic analysis [14,15].
An ORTEP drawing of the 2:2 complex of EGCg and caffeine is shown in Figure 3a. One unit cell contained four units of the 2:2 complex of EGCg and caffeine, and 60 water molecules as crystal solvent (Figure 3b).

Figure 3: 2:2 complex of EGCg and caffeine.
(a) ORTEP drawing with thermal ellipsoids at 30 % probability level.
(b) One unit cell Crystal solvent and hydrogen atoms are omitted for clarity.

Figure 4: Layered structure of the 2:2 EGCg complex of caffeine.
(a) Caffeine molecules are shown.
(b) Water molecules as a crystal solvent are shown but caffeine molecules are not.
The layer structure of the 2:2 complex of caffeine and EGCg consisted of two layers, which were in parallel in the same direction as the b-axis (Figure 4a). The caffeine molecule has an almost plain and rigid xanthine skeleton, was stacked between B’ rings of EGCg, and caffeine was located almost in the middle of two B’ rings of EGCg. In the 2:2 EGCg complex of caffeine, intermolecular interactions between caffeine and EGCg moieties were mainly p-p Interactions between a six-membered ring of caffeine and the B’ ring of EGCg. Furthermore, the caffeine moieties in the 2:2 complex were positioned in the space surrounded by the top and bottom walls of the B’ rings of EGCg moieties and right and left walls of the A and B rings of EGCg moieties as shown in Figure 4a. Resultly, the caffeine molecules were captured by the hydrophobic space formed with the three aromatic A, B, and B’ rings of EGCg in the 2:2 complex. Water molecules were not observed inside the space formed with the three aromatic A, B, and B’ rings of EGCg and existed outside the space, suggesting that the space had high hydrophobicity (Figure 4b). . It was therefore considered that the precipitate of the creaming down reaction occurred from the solution of EGCg and caffeine in water due to its high hydrophobicity. And I consider that this findings are the mechanism of the creaming down.
Analysis of the Complex of EC and Caffeine by X-Ray Crystallography
Equimolar amounts of EC and caffeine were dissolved in water by heating and the aqueous solution was left at room temperature, but no precipitate was afforded and it was still soluble. The mixture was lyophilized to afford a colorless powder, which was recrystallized from methanol to give a colorless block crystal, which was determined to be a 1:1 complex of EC and caffeine by X-ray crystallographic analysis [9,16]. One unit cell contained four units of the 1:1 complex of EC and caffeine and eight water molecules as crystal solvent (Figure 5).

Figure 5: 1:1 Complex of caffeine and EC.
(a) ORTEP drawing with thermal ellipsoids at 30 % probability level.
(b) One unit cell Crystal solvent and hydrogen atoms are omitted for clarity.

Figure 6: Layer structure of 1:1 complex of EC and caffeine.
In the layer structure shown in Figure 6, units of the 1:1 complex of EC and caffeine were stacked in parallel in the same direction as the b-axis. The A ring of EC and the six-membered ring of caffeine appear in turn along the b-axis, and the six-membered rings of caffeines were located in almost the middle of the A rings of EC moieties. The A rings of both upper and lower EC moieties faced the six-membered ring of caffeine. In the 1:1 complex of EC and caffeine, face-to-face p−p stacking interactions formed between the A ring of EC and the six-membered ring of caffeine.
Hydrophobic Space Formed Egcg and EC in Water
X-ray crystallographic analysis showed that in the 2:2 complex of EGCg and caffeine, EGCg moieties formed the hydrophobic space surrounded by the top and bottom walls of the B’ rings and right and left walls of the A and B rings (Figure 7), whereas, in the 1:1 complex of EC and caffeine, EC moieties formed the hydrophobic space only using the top and bottom walls of the A rings.
Therefore, it was thought that the hydrophobic effect in the 2:2 complex of EGCg and caffeine was much stronger than that of the 1:1 complex of EC and caffeine, with the result that the 2:2 complex of the gallated catechin EGCg and caffeine precipitated as a precipitate formed by creaming down more predominantly than the 1:1 complex of the non-gallated catechin EC and caffeine.

Figure 7: Hydrophobic space formed with the aromatic A, B, and B’ rings of EGCg.
Molecular Capture of Heterocyclic Compounds Using the Precipitate of Creaming Down
We investigated molecular capture using EGCg, which formed a strong hydrophobic space in water. Aqueous solution of various heterocyclic compounds (Table 3) was poured into an aqueous solution of an equimolar amount of EGCg, and it afforded sticky precipitates containing EGCg and the heterocyclic compounds at a molar ratio of 1:1, based on measurement of the integrated value of 1H NMR signals.
Analysis of the Complex of Egcg and 2-Chloropyrimidine by X-Ray Crystallography
To confirm that the precipitates made from an aqueous solution of equimolar of EGCg and the heterocyclic compounds were 2:2 complexes of the same mode as the 2:2 complex of EGCg and caffeine, X ray crystal analysis of the precipitate made from EGCg and 2-chloropyrimidine (Figure 8a) was performed [17].
Aqueous solution of 2-chloropyrimidine was poured into an aqueous solution of an equimolar amount of EGCg, and it afforded a precipitate as a colorless block crystal.
Based on an ORTEP drawing, this crystal contained two crystallographically different EGCgs and two 2-chloropyrimidines, and these formed a 2:2 complex which the aromatic ring of the 2-chloropyrimidine and the B’ rings of each EGCg faced (Figure 8). π-π Interactions formed between a six-membered ring of 2-chloropyrimidine and the B’ ring of EGCg, B rings of EGCg, and A rings of EGCg respectively.

Figure 8: 2-Chloropyrimidine in Hydrophobic space formed by the aromatic A, B, and B' rings of EGCg.
(a) 2-Chloropyrimidine.
(b) 2:2 complex of EGCg and 2-Chloropyrimidine Hydrogen atoms and crystal solvent are omitted for clarity.
It was found that 2-chloropyrimidine formed a 2:2 complex with EGCg in the same mode as the 2:2 complex of EGCg and caffeine. As a result, 2-chloropyrimidine molecules were captured by the hydrophobic space formed by the three aromatic rings of A, B, B’ rings of the EGCg as shown in Figure 8b, and it precipitated as the 2:2 complex from the aqueous solution of an equimolar amount of 2-chloropyrimidine and EGCg due to its high hydrophobicity.
Therefore, the precipitates were thought to be 2:2 complexes of EGCg and the heterocyclic compounds (Table 5) of the same mode as the 2:2 complex of EGCg and caffeine.
Molecular Capture of Heterocyclic Compounds Using Egcg
We investigated how much the heterocyclic compound could be captured in precipitate of the 2:2 complex from aqueous solution of EGCg and the heterocyclic compounds. The amount of the heterocyclic compounds were measured by an integrated value of corresponding proton signals in the quantitative 1H NMR spectrum in DMSO-d6. The molecular capture abilities of various heterocyclic compounds using EGCg were evaluated by the ratio of the amount of heterocyclic compounds included in the precipitates to that of the total heterocyclic compounds used (Table 3) [17].
In Table 3, a heterocyclic compound having A (%) of 0% means that an aqueous solution of the heterocyclic compound and EGCg did not give a precipitate.
However, no correlation was found between the chemical structures of the heterocyclic compounds and the molecular capture ability.
Table 3: The molecular capture abilities of heterocyclic compounds using EGCg.
|
Heterocyclic compounds |
A (%) |
B |
Heterocyclic compounds |
A (%) |
B |
|
pyridine |
48.73 |
1 |
2-Cyanopyridine |
20.91 |
0.49 |
|
2-Aminopyridine |
73.34 |
1.51 |
3-Cyanopyridine |
44.9 |
0.92 |
|
3-Aminopyridine |
66.96 |
1.37 |
4-Cyanopyridine |
51.14 |
1.05 |
|
4-Aminopyridine |
87.81 |
1.8 |
2-Nitropyridine |
49.7 |
1.02 |
|
2-Methylpyridine |
71.69 |
1.47 |
3-Nitropyridine |
45.56 |
0.93 |
|
3-Methylpyridine |
83 |
1.7 |
4-Nitropyridine |
76.67 |
1.57 |
|
4-Methylpyridine |
77.5 |
1.59 |
2-Pyridinecarboxamide |
0 |
0 |
|
2-Ethylpyridine |
46.13 |
0.95 |
3-Pyridinecarboxamide |
44.41 |
0.91 |
|
4-Ethylpyridine |
43.15 |
0.89 |
4-Pyridinecarboxamide |
82.65 |
1.7 |
|
3-Methoxypyridine |
78.67 |
1.61 |
2-Pyridinecarboxylic acid |
52.76 |
1.08 |
|
4-Methoxypyridine |
92.63 |
1.9 |
3-Pyridinecarboxylic acid |
73.24 |
1.5 |
|
2-Hydroxypyridine |
0 |
0 |
Isoniazid |
66.23 |
1.36 |
|
3-Hydroxypyridine |
34.73 |
0.71 |
Nicotinhydrazide |
57.47 |
1.18 |
|
4-Hydroxypyridine |
0 |
0 |
2-Methylpyrizine |
67.07 |
1.38 |
|
3-pyridineacetonitrile |
85.44 |
1.75 |
Chloropyrazine |
84.07 |
1.73 |
|
3-pyridinemethanol |
27.19 |
0.56 |
3-Amino-4-hydroxypyridine |
5.28 |
0.1 |
|
3-picolylamine |
82.59 |
1.69 |
|
|
|
* Mole number of each heterocyclic compound in a crude precipitate / total mole number
** A relative ratio when pyridine is set to 1.000.
The 2: 2 complexes dissolved in organic solvent, and readily decomposed into the constituent EGCg and the heterocyclic compound, which can be easily isolated by column chromatography. Therefore, the molecular capture ability is considered to be an important indicator of how much the heterocyclic compound can be isolated using EGCg from aqueous solution.
Molecular Capture of Diketopiperazines Containing Proline Residue Using the Precipitate of Creaming Down
We chose several types of diketopiperazine cyclo(Pro-Xxx) (Xxx: amino acid residue) as a substrate of EGCg to find some kind of correlation between the chemical structures and the molecular capture ability (Figure 9). This is because the chemical structure of cyclo(Pro-Xxx) consists of 5- and 6-membered rings, which is similar to that of caffeine, Furthermore, cyclo(Pro-Xxx) is advantage as because its side chain can be freely altered by substituting amino acid residues.

Figure 9: Cyclo(Pro-Gly) and Cyclo(Pro-Xxx) (Xxx=Amino Acid Residue).
Analysis Of The Complex Of Egcg And Cyclo(Pro-Gly) By X-Ray Crystallography
To confirm that the precipitate made from an aqueous solution of equimolar of EGCg and cyclo(Pro-Xxx) was 2:2 complexes of the same mode as the 2:2 complex of caffeine and EGCg, X ray crystal analysis of the precipitate made from EGCg and cyclo(L-Pro-Xxx), cyclo(D-Pro-Xxx) was performed [18,19].
An aqueous solution of diketopiperazine cyclo(L-Pro-Gly) was added to an aqueous solution of an equimolar amount of EGCg. The mixture afforded a colorless block crystal, which was determined to be a 2:2 complex of EGCg and cyclo(L-Pro-Gly) by X-ray crystallographic analysis (Figures 10a and 11a). Using the same method for crystallization using cyclo(D-Pro-Gly) and EGCg afforded a colorless block crystal, which was determined to be a 2:2 complex of EGCg and cyclo(D-Pro-Gly) (Figsures 10b and 11b).

Figure 10: Layered structures.
(a) 2:2 EGCg complex of cyclo(L-Pro-Gly).
(b) 2:2 EGCg complex of cyclo(D-Pro-Gly).

Figure 11: Intermolecular interactions.
(a) 2:2 EGCg complex of cyclo(L-Pro-Gly).
(b) 2:2 EGCg complex of cyclo(D-Pro-Gly) Black arrows and dotted lines indicate CH-π interactions and hydrogen bonds, respectively.
Upon the 2:2 complex formation with EGCg and cyclo(L-Pro-Gly), cyclo(D-Pro-Gly) were captured by the hydrophobic space formed by the three aromatic A, B, B’ rings of EGCg and were located almost in the middle of the two B’ rings of EGCg moieties in the same as caffeine in the 2:2 complex of EGCg (Figure 4a).
In the 2:2 complex of EGCg and caffeine, p−p interaction formed between caffeine and B’ rings of EGCg, while in the 2:2 complex of EGCg and cyclo(Pro-Gly), CH−p ineraction formed between methine C9-H of cyclo(Pro-Gly) and the B’ rings of EGCg. In the 1H NMR spectra in a solution containing equimolar amounts of EGCg and cyclo(L-Pro-Gly), cyclo(D-Pro-Gly), and EGCg in D2O, the upfield shift was observed in the proton signal for H9 of cyclo(L-Pro-Gly), cyclo(D-Pro-Gly), resulted from magnetic anisotropic shielding by the ring current from the B’ ring of the EGCg moieties (Table 4).
Table 4: Chemical shift of 1H NMR signals in a solution containing equimolar amounts (155.8 mM each) of EGCg and cyclo(L-Pro-Gly), cyclo(D-Pro-Gly) in D2O.

Molecular Capture Of Cyclo(Pro-Xxx) (Xxx=Amino Acid Residue) Using Egcg
Instead of the glycine residue of cyclo(Pro-Gly), diketopiperazine cyclo(Pro-Xxx) having amino acid residues with different side chains was variously synthesized [20].
An aqueous solution of diketopiperazine cyclo(Pro-Xxx) (Xxx: amino acid residue) was poured into an aqueous solution of an equimolar of EGCg. The resulting precipitate was dried to obtain a solid of the complex of cyclo(Pro-Xxx) and EGCg.
The obtained solid of the complex was measured by the quantitative 1H NMR. The proton signals for H3 and H9 did not overlap with other proton signals. However, H3 may have an abnormally long relaxation time and not be detected as a proton signal. Therefore, the integrated value of the proton signal of H9 was used to assess the molecular capture ability.
The molecular capture ability of cyclo(Pro-Xxx) using EGCg was evaluated by the ratio of the amount of the cyclo(Pro-Xxx) included in the solid of the complex to that of the total cyclo(Pro-Xxx) used (Table 5).
Table 5: Molecular Capture Ability (%) of Diketopiperazine Cyclo(Pro-Xxx) by EGCg.
|
Diketopiperazine |
|
||
|
Xxx |
Side Chain |
|
Molecular Capture Ability (%) |
|
-R |
Cyclo(Pro-Xxx) |
|
|
|
|
|
L-Pro-Gly |
0 |
|
Gly |
-H |
D-Pro-Gly |
0 |
|
|
|
L-Pro-L-Ala |
0 |
|
Ala |
0 |
D-Pro-L-Ala |
0 |
|
|
|
L-Pro-L-Val |
28.2 |
|
Val |
-CH(CH3)2 |
D-Pro-L-Val |
40.1 |
|
|
|
L-Pro-L-Pro |
50.6 |
|
Pro |
-CH2CH2CH2- |
D-Pro-L-Pro |
69.8 |
|
|
|
L-Pro-L-Leu |
59.7 |
|
Leu |
-CH2CH(CH3)2 |
D-Pro-L-Leu |
74.2 |
|
|
|
L-Pro-L-Ile |
60.9 |
|
Ile |
-CH(CH3)CH2CH3 |
D-Pro-L-Ile |
69.6 |
|
|
-CH2-CH2-C6H5 |
L-Pro-L-Phe |
83.2 |
|
Phe |
D-Pro-L-Phe |
68.7 |
|
|
|
|
L-Pro-L-Tyr |
63 |
|
Tyr |
-CH2-CH2-C6H4-OH |
D-Pro-L-Tyr |
52.6 |
|
|
|
L-Pro-L-Ser |
0 |
|
Ser |
-CH2-OH |
D-Pro-L-Ser |
0 |
|
Ser(Bzl) |
-CH2-O-CH2-C6H5 |
L-Pro-L-Ser(Bzl) |
69.5 |
|
D-Pro-L-Ser(Bzl) |
79.1 |
||
* Molecular capture ability (%) of 0% means that an aqueous solution of the cyclo(Pro-Xxx) and EGCg did not form a precipitate.
** Cyclo(L-Pro-L-Trp), cyclo(D-Pro-L-Trp), cyclo(L-Pro-L-Tyr(Bzl)), and cyclo(D-Pro-L-Tyr(Bzl)) were insoluble in water (80 ml) under this experimental condition.
Furthermore, the correlation between the chemical structures of cyclo(Pro-Xxx) and the molecular capture ability was investigated [20].
Based on comparison of the molecular capture ability of cyclo(Pro-Yyy) (Yyy: hydrophobic amino acid residue) by EGCg, a greater number of carbons in the side chain of the amino acid residue, such as Ala, Val, Pro, Leu, Ile, and Phe led to a higher molecular capture ability of cyclo (Pro-Yyy). Both proline and valine residues have 3 carbon atoms in their side chains. As the side chain of the proline residue is ring-shaped, the hydrophobicity of cyclo(Pro-Pro) was higher than that of cyclo(Pro-Val). Therefore, the molecular capture ability of cyclo(Pro-Pro) was considered higher than that of cyclo(Pro-Val).
In addition, cyclo (Pro-Zzz) (Zzz=Ser and Tyr), which has a hydrophilic hydroxyl group on the side chain, demonstrated a low molecular capture ability. The molecular capture ability of cyclo (Pro-Ser) was 0%, whereas, that of (Pro-Ser(Bzl)) was high. Cyclo(Pro-Tyr) had a lower molecular capture ability than cyclo (Pro-Phe).
Conformational Change of Cyclo (Pro-Ala) Upon Formation of a Complex with EGcg
1H NMR spectra of a solution of cyclo(Pro-Ala) and an equimolar solution of cyclo(Pro-Ala) and EGCg in D2O are shown in Figure 12 [20].
Table 6: Chemical Shift (ppm) of Cyclo(Pro-Ala) Alone and in the 2:2 Complex with EGCg in 1H NMR Spectra.
|
Position of proton |
Chemical shift of Cyclo(L-pro-L-Ala) |
Chemical shift of Cyclo(D-pro-L-Ala) |
||
|
Alone |
Compleax Formation |
Alone |
Compleax Formation |
|
|
CH3 |
1.397 |
1.582 |
1.444 |
1.58 |
|
7α,7β,8α |
1.871-2.115 |
1.819-2.093 |
1.872-2.121 |
1.696-2.099 |
|
8 |
2.328 |
2.345 |
2.355 |
2.374 |
|
6α,8β |
3.528 |
3.512 |
3.556 |
3.52 |
|
3 |
4.301 |
4.139 |
4.066 |
4.194 |
|
9 |
4.348 |
4.275 |
4.388 |
4.275 |

Figure 12: 1H NMR Spectra of a D2O Solution Containing.
(a) Cyclo(L-Pro-L-Ala) and EGCg (155.8 mM Each).
(b) Cyclo(L-Pro-L-Ala) (155.8 mM); (c) Cyclo(D-Pro-L-Ala) and EGCg (155.8 mM Each); (d) Cyclo(D-Pro-L-Ala) (155.8 mM) in D2O.
Based on comparison of the proton signals in 1H NMR spectra of cyclo(Pro-Ala) alone, those of an equimolar solution of cyclo(Pro-Ala) and EGCg were broaden, suggesting that cyclo(Pro-Ala) was taken into the three aromatic A, B, and B’ rings of EGCg, and formed the 2:2 complex with EGCg, resulting in strong restriction of their molecular motion.
Upon the formation of the 2:2 complex of cyclo(Pro-Ala) with EGCg, CH−p ineraction formed between methine C9-H of cyclo(Pro-Ala) and the B’ rings of EGCg. The proton signal for the H9 of cyclo(Pro-Ala) moiety in the axial position was observed upfield sift due to magnetic anisotropic shielding of the ring current from the B’ ring of the EGCg moiety (Table 6). On the other hand, the methyl proton signal of cyclo(Pro-Ala) moiety was observed downfield sift due to magnetic anisotropic deshielding of the ring current from the B’ ring of EGCg, suggesting that the methyl group existed outside the hydrophobic space.
The major difference in the 1H NMR spectra between cyclo(L-Pro-L-Ala) and cyclo(D-Pro-L-Ala) was the chemical shift value of the methine proton signal for H3. Although the methine proton signal for H3 of cyclo(L-Pro-L-Ala) was at 4.301 ppm, that of cyclo(L-Pro-L-Ala) was at 4.066 ppm , with a difference of 0.235 ppm (Table 6). NOE was performed to investigate the difference in conformation between cyclo(L-Pro-L-Ala) and cyclo(D-Pro-L-Ala) in D2O. As a result, characteristic NOE was observed between the 3-position methyl proton and methine proton H9 of cyclo(D-Pro-L-Ala) (Fig. 13). On the other hand, no NOE was observed between those of cyclo(L-Pro-L-Ala). As the methine proton H9 took an axial position, the 3-position methyl group of cyclo(D-Pro-L-Ala) was suggested to be in the axial position, whereas that of cyclo(L-Pro-L-Ala) was in the equatorial position (Figure 13).

Figure 13: Conformation of Cyclo(L-Pro-L-Ala) and Cyclo(D-Pro-L-Ala).
ROE measurement of cyclo(D-Pro-L-Ala) in the presence of EGCg in D2O was next performed. As a result, no ROE was observed between the 3-position methyl proton and the methine proton H9 of cyclo(D-Pro-L-Ala) moiety. When cyclo(D-Pro-L-Ala) was taken into the hydrophobic space formed by the three aromatic A, B, B’ rings of the EGCg, forming the 2:2 complex of cyclo(D-Pro-L-Ala) and EGCg, the 3-position methyl group in the axial position, which was highly sterically hindered by EGCg, was considered to conformationally change to the equatorial position (Figure 14).

Figure 14: Schematic Representation of 2:2 Complex of Cyclo(L-Pro-L-Ala) and EGCg, and 2:2 Complex of Cyclo(D-Pro-L-Ala) and EGCg.
Along with above change of the methyl group, the methine H3 changed from the equatorial position to the axial position and its proton signal was observed downfield shift from 4.07 ppm to 4.19 ppm (Table 6). As a result, the chemical shift value of the methine proton signal of H3 (4.194 ppm) of the cyclo(D-Pro-L-Ala) moiety in the 2:2 complex became almost the same as that for H3 (4.139 ppm) of cyclo(L-Pro-L-Ala) moiety in the 2:2 complex with EGCg. Therefore, a conformation change of the 3-position methyl group of cyclo(D-Pro-L-Ala) in axial position to equatorial position considered to have occurred upon the 2:2 complex formation with EGCg, whereas its diastereomer cyclo(L-Pro-L-Ala) was taken into the hydrophobic space formed by the three aromatic A, B, and B’ rings of EGCg with no conformationl change. Upon complex formation with EGCg, conformational change that occurred in cyclo(D-Pro-L-Ala) did not occur in other diketopiperazine cyclo(Pro-Xxx).
Conclusion
When an aqueous caffeine solution was poured into an aqueous solution of catechin mixture, precipitate by creaming down occurred. The gallated catechins EGCg and ECg were formed the precipitate of creaming down more predominantly rather than the non-gallated catechins, EC, EGC, CA, and GC.
X-ray crystallographic analysis showed that EGCg formed a 2:2 complex with caffeine, and EC formed 1:1 complex with caffeine. Upon the formation of the 2:2 complex, EGCg moieties formed the hydrophobic space surrounded by the top and bottom walls of the B’ rings and right and left walls of the A and B rings, whereas, upon the formation of the 1:1 complex, EC formed the hydrophobic space only using the top and bottom walls of the A rings. Therefore, it was thought that the hydrophobic effect in the 2:2 complex of EGCg and caffeine was stronger than that of the 1:1 complex of EC and caffeine, with the result that the 2:2 complex of the gallated catechin EGCg and caffeine precipitated as a precipitate formed by creaming down more predominantly than the 1:1 complex of the non-gallated catechin EC and caffeine.
Then we investigated molecular capture using EGCg, which formed a strong hydrophobic space in water. An aqueous solution of an equimolar amount of EGCg and a variety of heterocyclic compounds (Table 3) afforded sticky precipitates containing EGCg and the heterocyclic compounds of the 2:2 complex.
The molecular capture abilities of heterocyclic compounds using EGCg evaluated by the ratio of the amount of heterocyclic compounds included in the precipitates to that of total heterocyclic compounds used. However, no correlation was found between the chemical structures of the heterocyclic compounds and their molecular capture abilites.
Then the molecular capture abilities of diketopiperazine cyclo(Pro-Xxx) (Xxx: amino acid residue) were evaluated, and the correlation between the chemical structures of cyclo(Pro-Xxx) and the molecular capture abilites was investigated. A greater number of carbons in the side chain of the amino acid residue, such as Ala, Val, Pro, Leu, Ile, and Phe led to a higher molecular capture ability of cyclo (Pro-Yyy).
When cyclo(Pro-Xxx) was taken into the hydrophobic space formed by EGCg to form the 2:2 complex of EGCg, conformation of cyclo(Pro-Xxx) other than cyclo(D-Pro-L-Ala) still maintained. And conformation of cyclo(D-Pro-L-Ala) changed, that is, its 3-position methyl group changed from the axial position to the equatorial position due to steric hindrance by EGCg moieties.
Acknowledgements
All of the research in this review was conducted at the Laboratory of Organic and Bio-organic Chemistry in the Faculty of Pharmacy and Pharmaceutical Sciences, Fukuyama University. The author is grateful to the members and graduates of the Laboratory of Organic and Bio-organic Chemistry. In particular, the author would like to thank Dr. Hiroyuki Tsutsumi, Faculty of Pharmaceutical Sciences, Fukuoka University, for making a great contribution to the research results related to this manuscript.
Conflict of Interest
The authors declare no conflict of interest.
References
- Kuroda Y, Hara Y. Health Effects of Tea and Its Catechins. Kluwer Academic/Plenum Publishers, New York. 2004.
- Hayashi N, Ujihara T. A Water-Soluble acyclic phane receptor recognizing 2,3-trans-gallate-type catechins. J Org Chem. 2008; 73: 4848-4854.
- Okabe S, Suganuma M, Hayashi M, Sueoka E, Komori A, Fujiki H. Jpn J Cancer Res. 1997; 88: 639-643.
- Tezuka M, Suzuki H, Suzuki Y, Hara Y, Okada S, Jpn J Toxicol Environ Health. 1997; 43: 311-315.
- Mimura S, Watanabe J, Tomita T, Sano M, Tomita I, Biol Pharm Bull. 1994; 17: 1567-1572.
- Hara Y, Watanabe M. Antibacterial action of tea polyphenols against Clostridium botulinum. Nippon Shokuhin Kogyo Gakkaishi. 1989; 36: 951-955.
- Seshadri R, Dhanraj N. Proc. International Flavor Conference, Porto Karras Greece. 1987; 169-180.
- Maruyama N, Suzuki Y, Sakata K, Yagi A, Ina K, Proc. International Symposium Tea Science 1991; 145-149.
- Sato T, Kinoshita Y, Tsutsumi H, Yamamoto H, Ishizu T. Characterization of creaming precipitate of tea catechins and caffeine in aqueous solution. Chem Pharm Bull. 2012; 60: 1182-1187.
- Davis AL, Cai Y, Davies AP, Lewis JR. Magn Reson Chem. 1996; 34: 887-890.
- Muramatsu K. Chanokagaku Asakura Publishing. Tokyo; 1991: 124.
- Karplus J. Contact Electron?spin coupling of nuclear magnetic moments. J Chem Phys. 1959; 30: 11.
- Ishizu T, Kajitani S, Tsutsumi H, Sato T, Yamamoto H, Hirata C. Configurational studies of complexes of tea catechins with caffeine and various cyclodextrins. Planta Medica. 2011; 77: 1099-1109.
- Tsutsumi H, Sato T, Ishizu T. Stereochemical structure of the complex of (−)-epigallocatechin 3-gallate and caffeine. Chem Lett. 2012; 41: 1669-1671.
- Ishizu T, Tsutsumi H, Kinoshita Y, Mukaida H, Sato T, Kajitani S. Properties of precipitate of creaming down by (−)-Epigallocatechin-3-O-gallate and caffeine. Chem Pharm Bull. 2014; 62: 552-558.
- Ishizu T, Sato T, Tsutsumi H, Yamamoto H. Stereochemical structure determination of caffeine complexes with galloylated and non-galloylated catechins. Chem Lett. 2010; 39: 607-609.
- Tsutsumi H, Sato A, Fujino S, Fujioka Y, Ishizu T. Molecular capture using the precipitate of Creaming-Down by (−)-Epigallocatechin-3- O –gallate. Chem Pharm Bull. 2019; 67: 501-504.
- Ishizu T, Tsutsumi H, Yokoyama A. Chiral recognition of diketopiperazines cyclo(Pro-Gly) using (−)-epigallocatechin-3-O-gallate. Tetrahedron Lett. 2015; 56: 1111-1114.
- Ishizu T, Tsutsumi H, Yokoyama E, Tanabe H, Yokoyama A. Chiral recognition of diketopiperazine cyclo(Pro-Gly) and Propranolol Using (−)-Epigallocatechin-3-O-gallate. Chem Pharm Bull. 2016; 64: 142-149.
- Ishizu T, Tokunaga M, Fukuda M, Matsumoto M, Goromaru T, Takemoto S. Molecular capture and conformational change of diketopiperazines containing proline residues by Epigallocatechin-3-O-gallate in Water. Chem Pharm Bull. 2021; 69: 585-589.