Mass spectrometric detection of enantioselectivity in three-component complexation, copper(II)-chiral tetradentate ligand-free amino acid in solution

Abstract

By using electrospray ionization mass spectrometry coupled with the use of deuterium-labelled quasienantiomers, enantioselective coordination of a free amino acid as the second ligand of a copper(II) complex with a novel chiral tetradentate ligand was evaluated quantitatively in a short measurement time.

---

Metal complexes with chiral amino acid derivatives as the stereogenic centres have been developed to exhibit various promising functions. Asymmetric catalytic ability,^1–4 stimulus-responsive helicity control,^5–10 stereoselective encapsulations in organic zeolite¹¹ and metal organic frameworks (MOFs),¹² circular polarized luminescence induction^13,14 and collagen synthesis in a cell were typical examples.¹⁵ Chiral molecular recognition of the free amino acid is one of the most important functions,¹⁶ because it is applicable to understanding biological metabolism/molecular recognition systems, and developing sophisticated industrial systems. In this paper, we report for the first time detecting enantioselective coordination of a free amino acid as the second ligand of a metal complex with a simple chiral tetradentate ligand⁵ by using electrospray ionization mass spectrometry (ESI-MS) coupled with the use of deuterium-labelled quasienantiomers.^17–20 In general, a chiral discrimination ability of a chiral host toward chiral guest in solution is evaluated by a spectroscopic titration method using NMR, UV-vis and CD measurements. However, these methods basically require high concentration and are time consuming. Compared to these titration methods, the ESI-MS method can determine chiral discrimination ability of the complex toward free amino acids more rapidly at smaller scale.

As the suitable chiral tetradentate ligand of a chiral metal complex, simple S,S-chiral ligands (S,S-L1 and S,S-L2) and the corresponding R,R-enantiomers (R,R-L1 and R,R-L2) were designed (Chart 1) and synthesized using the conventional method.⁵ Achiral ligands (L3 and L4) were also synthesized as the reference compounds.

Chart 1 Chiral and achiral tridentate ligands.

This type of ligand coordinates with a metal cation in the tetradentate fashion, and the resulting metal complex can provide two binding sites suitable for chiral recognition of a bidentate amino acid as the second ligand. In fact, the X-ray crystal analysis of the copper(II) complex with S,S-L2 showed that tetradentate S,S-L2 coordinated with the copper cation in the helical cis-α Λ form and two further methanol molecules occupied two remaining binding sites of the copper cation (Fig. 1). These methanol molecules could be replaced by a bidentate amino acid to form a three-component complex. In order to evaluate enantioselective complexation with chiral amino acid, a three-component complex ion composed of a metal cation, ligand (L) and amino acid is required to be detected with high sensitivity in ESI-MS. Firstly, metal cations providing 1 : 1 complex ions with the achiral ligand (L3) in a mass spectrum were systematically examined using the following metal salts: CrCl₃, MnCl₂, FeCl₃, CoCl₂, NiCl₂, CuCl₂, NiCl₂ and LaCl₃. In the case of MnCl₂, CoCl₂, NiCl₂, CuCl₂ and ZnCl₂, 1 : 1 complex ions of [M(L3) + Cl]⁺ were observed (Fig. S7, ESI†). In particular, CuCl₂ gave a highly intense ESI mass spectrum without contaminating peaks and unassignable signals. Therefore, the copper cation was selected as the suitable metal ion. Second, different anions were screened using the copper salts CuCl₂, Cu(CH₃COO)₂, Cu(NO₃)₂, Cu(acac)₂ and Cu(OTf)₂ in the presence of L3 and S-Ala. As a result, it was found that CuCl₂ provided desirable ions, [Cu^II(L3)(S-Ala–H)]⁺ as the base peak with high intensity and with almost no contaminating ions including cluster peaks corresponding to [Cu^II(L3)(S-Ala–H)₂] complex in the mass spectrum (Fig. S8, ESI†).

Fig. 1 (a) ORTEP drawing (50% probability) and (b) chemical structure of cation part of [Cu(S,S-L2)(CH₃OH)₂](ClO₄)₂(CH₃OH)₂. In (a), most hydrogen atoms except for amide hydrogen atoms are omitted for clarity. Selected bond lengths (Å) and angles (deg); Cu1–O1 2.115(3); Cu1–N2 2.105(4); Cu1–O2 2.064(3); O1–Cu1–O1′ 178.23(18); O1–Cu1–N2 78.80(14); O1–Cu1–O2 90.17(13); O2–Cu1–N2 92.23(14); N2–Cu1–N2′ 85.9(2); O2–Cu1–O2′ 91.53(19).

We chose the MS/enantiomer-labelled (MS/EL) method¹⁸ to discriminate chiral compounds rapidly with high sensitivity. In this method, an equimolar mixture of deuterium-labelled S- and unlabelled R-enantiomers of chiral amino acid (AA) ([²H]_n-S-AA and R-AA, n = number of deuterium atoms) were used to evaluate the chiral discrimination ability of a chiral Cu^II-(L) complex toward chiral amino acid on the basis of the relative peak intensity of diastereomeric three-component complex ions, [Cu^II(L) + ([²H]_n-S-AA–H)]⁺ and [Cu^II(L) + (R-AA–H)]⁺, observed in mass spectra.

Sample solutions for the MS/EL method were prepared by mixing a methanol solution of CuCl₂ and ligand (S,S-L1, S,S-L2, L3 or L4) and an aqueous solution of quasiracemic mixture of amino acid (an equimolar mixture of deuterium-labelled S-AA and unlabelled R-AA) and K₂CO₃. Fig. 2 shows the typical mass spectra in which phenylalanine (Phe) was used as the quasiracemic mixture of amino acid. In the case of achiral ligands L3 and L4, the relative peak intensity values of two diastereomeric complex ion peaks (I[Cu^II(L)(R-Phe–H)]⁺/I[Cu^II(L)([²H]₅-S-Phe–H)]⁺ = I_R/I_S value) were unity due to no selectivity (Fig. 2(c) and (d)). In the case of chiral ligands S,S-L1 and S,S-L2, the I_R/I_S values were larger than 1.00 (Fig. 2(a) and (b)), indicating that R-Phe preferentially coordinates with the Cu^II-S,S-L1 and Cu^II-S,S-L2 complexes compared to S-Phe. Inversely, S-Phe preferentially coordinates with Cu^II-R,R-L2, which is the enantiomer of Cu^II-S,S-L2 (Fig. S9, ESI†). The multiplication of each I_R/I_S value (1.55 × 0.63 = 0.98) from [Cu^II(S,S-L2)(Phe–H)]⁺ and [Cu^II(R,R-L2)(Phe–H)]⁺ under the same sample concentration conditions and spectral measurement conditions was unity. It shows that differentiation of the peak intensity related to the diastereomeric complex ions is caused not by random complexation but by highly controlled stereo-specific complexation. Thus, the chiral discrimination ability of the complex ion toward amino acid corresponds to the I_R/I_S values.

Fig. 2 ESI mass spectra of mixing system of CuCl₂/L/R-Phe/[²H]₅-S-Phe in water/methanol (1/100, v/v). [CuCl₂]₀ = 1.19 × 10⁻⁴ M, [L]₀ = 9.90 × 10⁻⁵ M, [R-Phe]₀ = 4.95 × 10⁻⁵ M, [[²H]₅-S-Phe]₀ = 4.95 × 10⁻⁵ M and [K₂CO₃]₀ = 9.90 × 10⁻⁵ M. [CuCl₂]₀/[L]₀/[R-Phe]₀/[[²H]₅-S-Phe]₀ = 1.2/1.0/0.5/0.5. (a) L = S,S-L1, (b) L = S,S-L2, (c) L = L3, (d) L = L4.

As the I_R/I_S value reflects the concentration ratio of the diastereomeric three-component complex ions [Cu^II(S,S-L1)(R-AA–H)]⁺ and [Cu^II(S,S-L1)([²H]_n-S-AA–H)]⁺, the value changes depending on the concentration ratio of the quasiracemic mixture of amino acid to the Cu^II-S,S-L1 complex (Fig. S11, ESI†). Amount of amino acid to be added was optimized to evaluate chiral discrimination ability (I_R/I_S value). As a result, CuCl₂/L1/R-AA/[²H]_n-S-AA = 1.2/1.0/2.0/2.0 (mole ratio) system ([L1]₀ = 7.14 × 10⁻⁵ M) was found to be the most suitable condition. Under this condition, the highest I_R/I_S value for each amino acid was obtained which was stable during the MS measurement. When the mole ratio of the quasiracemic amino acid was more than 1.0 equivalents, the I_R/I_S value became slightly larger and the signal-to-noise ratio became smaller due to formation of several interfering chemical species such as [Cu^II(AA–H)₂] and S,S-L1 (Fig. S10, ESI†).

The formation of [Cu^II(AA–H)₂] was also confirmed by other spectroscopic analyses. As illustrated in Fig. S3 and S4 (ESI†), UV-vis spectral changes were induced by reacting chiral ligand (S,S-L) with CuCl₂ in methanol, indicating the generation of the [Cu^II(S,S-L)]²⁺ complex. Adding a solution of equimolar S-Phe into the resulting [Cu^II(S,S-L1)]²⁺ solution in methanol clarified that the [Cu^II(S,S-L1)(S-Phe–H)]⁺ complex was produced (Fig. S5, ESI†). Further adding S-Phe changed UV-vis and CD spectra. Both titration studies, UV-vis and MS measurements (Fig. S10g, ESI†), suggest that the chiral tetradentate ligand S,S-L1 was dissociated from the [Cu^II(S,S-L1)(S-Phe–H)]⁺ complex and [Cu^II(S-Phe–H)₂] was generated simultaneously. Its UV-vis and CD spectra also gave close agreement with that of in situ generated [Cu^II(S-Phe–H)₂] (Fig. S6, ESI†).²¹

Signals of the complex ions ([Cu^II(S,S-L)(R-AA–H)]⁺ and [Cu^II(S,S-L)([²H]_n-S-AA–H)]⁺) were clearly detected in all cases of the nine given amino acids, Ala, Leu, Val, Met, Orn, Lys, Phe, Hyp and Trp (Fig. 3 and Fig. S12, ESI†).²² The I_R/I_S values are summarized in Table 1. The I_R/I_S values were larger than 1.00 for most of the given amino acids, indicating R-selectivity. However, Cu^II-S,S-L1-Leu, Cu^II-S,S-L2-Ala, Cu^II-S,S-L2-Leu and Cu^II-S,S-L2-Orn gave 0.95–1.05, meaning no selectivity. Especially, the I_R/I_S values for Val, Phe, Hyp and Trp were larger than those for the other amino acids. These amino acids have bulky substituents at the α-position, suggesting that the main factor of chiral discrimination toward amino acids is steric repulsion between the chiral ligand and the substituents of amino acids. I_R/I_S values changed depending on enantiomer excess (ee) of added Phe with linear correlation (Fig. 4). Thus, the MS/EL method is suitable for evaluation of the concentration ratio, [Cu^II(S,S-L2)(R-Phe–H)]⁺/[Cu^II(S,S-L2)([²H]₅-S-Phe–H)]⁺, in solution, precisely.

Fig. 3 ESI mass spectra of CuCl₂/S,S-L1/R-Phe/[²H]₅-S-Phe in water/methanol (2/5, v/v). [CuCl₂]₀ = 8.57 × 10⁻⁵ M, [S,S-L1]₀ = 7.14 × 10⁻⁵ M, [R-Phe]₀ = 1.43 × 10⁻⁴ M, [²H]₅-S-Phe]₀ = 1.43 × 10⁻⁴ M and [K₂CO₃]₀ = 2.86 × 10⁻⁴ M. [CuCl₂]₀/[S,S-L1]₀/[R-Phe]₀/[[²H]₅-S-Phe]₀ = 1.2/1.0/2.0/2.0.

Table 1 I _R/I_S values of a Cu^II-chiral ligand (S,S-L1, S,S-L2)-amino acid (AA) complexation system in the MS/EL method

Complex	AA	I R /I S	Complex	AA	I R /I S
a All errors measuring the same sample were within 5%. Intensity of overlapped isotopic peak of the lower m/z signal on the higher one was corrected based on the natural abundance of isotope.
[Cu(S,S-L1)]^2+	Ala	1.09	[Cu(S,S-L2)]^2+	Ala	1.01
	Leu	1.01		Leu	0.97
	Val	1.69		Val	1.56
	Met	1.20		Met	1.07
	Orn	1.21		Orn	1.02
	Lys	1.22		Lys	1.16
	Phe	1.68		Phe	1.62
	Hyp	3.11		Hyp	1.41
	Trp	1.86		Trp	1.31

---

Fig. 4 Correlation between intensity excess (Ie) {(I_R/I_S − 1)/(I_R/I_S + 1) × 100} and enantiomer excess (ee) of Phe in Cu^II-S,S-L1-Phe (1.2 : 1.0 : 4.0) system. The ee value of Phe was confirmed by HPLC with chiral column.

As mentioned above, I_R/I_S values obtained under the optimized Cu–ligand–amino acid concentration conditions reflects the concentration ratio of the diastereomeric three-component complex ions [Cu^II(S,S-L1)(R-AA–H)]⁺ and [Cu^II(S,S-L1)([²H]_n-S-AA–H)]⁺, meaning I_R/I_S ≈ K_R/K_S.^18,23 This concludes that free energy difference between diastereomers ΔΔG(R/S) is also affected by I_R/I_S value. Then, we studied structures of the diastereomeric complex ions, [Cu^II(S,S-L2)(R-Phe–H)]⁺ and [Cu^II(S,S-L2)(S-Phe–H)]⁺, in methanol by DFT calculations in which four conformers for the [Cu^II(S,S-L2)(R-Phe–H)]⁺ complex and three conformers for [Cu^II(S,S-L2)(S-Phe–H)]⁺ were obtained (see each Boltzmann population in Table S1, ESI†). Total population percentage of [Cu^II(S,S-L2)(R-Phe–H)]⁺ and [Cu^II(S,S-L2)(S-Phe–H)]⁺ in solution was calculated as 62.7% and 37.3%, respectively (Table S1, ESI†); the complex ion [Cu^II(S,S-L2)(R-Phe–H)]⁺ was slightly more stable than [Cu^II(S,S-L2)(S-Phe–H)]⁺. Fig. 5 shows the most stable structure for each complex ion. This calculated population ratio of the diastereomeric complex ions (62.7/37.3 = 1.68) is in good agreement with the I_R/I_S value (1.62) observed in the MS/EL method.

Fig. 5 Molecular structures of the most stable conformer of (a) [Cu^II(S,S-L2)(R-Phe–H)]⁺ and (b) [Cu^II(S,S-L2)(S-Phe–H)]⁺ optimized by DFT calculation at B3LYP/LANL2DZ.

As described above, we have succeeded in quantitative evaluation of the enantioselective coordination behaviours of free amino acids with chiral organometallics by mass spectrometry. Several researches have been conducted on the optical resolution of amino acids with metal complex since Gillard et al. reported in 1968.^24–27 Although mass spectroscopic techniques have been widely used in determining the enantiomeric purity of chiral compounds,^28–30 ESI-MS method is the most convenient and sensitive for detecting metal complex ion, which exist in the equilibrium between diastereomeric metal complexes in solution. Therefore, establishment of a new system for evaluating the enantiomeric purity of amino acids more easily and quickly based on the research results described here will enable microanalysis of free amino acid in the human body, which will be widely used in the medical field.

This work was partially supported by JSPS KAKENHI Grant Numbers JP18K0595 to M. S. and JP19K05505 to H. M.

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. A. Sato, M. Yoshida and S. Hara, Chem. Commun., 2008, 6242–6244.
2. Y. Kawato, N. Takahashi, N. Kumagai and M. Shibasaki, Org. Lett., 2010, 12, 1484–1487.
3. M. Yoshida, M. Narita and S. Hara, J. Org. Chem., 2011, 76, 8513–8517.
4. M. M. Sheeba, M. M. Tamizh, L. J. Farrugia and R. Karvembu, J. Organomet. Chem., 2017, 831, 45–49.
5. H. Miyake, K. Yoshida, H. Sugimoto and H. Tsukube, J. Am. Chem. Soc., 2004, 126, 6524–6525.
6. H. Miyake, H. Sugimoto, H. Tamaki and H. Tsukube, Chem. Commun., 2005, 4291–4293.
7. H. Miyake, H. Kamon, I. Miyahara, H. Sugimoto and H. Tsukube, J. Am. Chem. Soc., 2008, 130, 792–793.
8. S. Zahn and J. W. Canary, Science, 2000, 288, 1404–1407.
9. H. Miyake, H. Hikita, M. Itazaki, H. Nakazawa, H. Sugimoto and H. Tsukube, Chem. – Eur. J., 2008, 14, 5393–5396.
10. J. Gregoliński, M. Hikita, T. Sakamoto, H. Sugimoto, H. Tsukube and H. Miyake, Inorg. Chem., 2016, 55, 633–643.
11. R. Nomura, H. Mitsukuri, A. Ohtaka and O. Shimomura, Trans. Mater. Res. Soc. Jpn., 2008, 33, 399–402.
12. S. Wang, M. Wahiduzzaman, L. Davis, A. Tissot, W. Shepard, J. Marrot, C. Martineau-Corcos, D. Hamdane, G. Maurin, S. Devautour-Vinot and C. Serre, Nat. Commun., 2018, 9, 1–8.
13. P. Atkinson, Y. Bretonniere, D. Parker and G. Muller, Helv. Chim. Acta, 2005, 88, 391–405.
14. B. T. Nguyen, A. J. Ingram and G. Muller, Chirality, 2016, 28, 325–331.
15. H. Tsukube, Y. Noda, Y. Kataoka, H. Miyake, S. Shinoda, A. Kojima-Yuasa, Y. Nishida and I. Matsui-Yuasa, Dalton Trans., 2008, 4038–4043.
16. J. Wang, H.-B. Liu, Z. Tong and C.-S. Ha, Coord. Chem. Rev., 2015, 303, 139–184.
17. A. Horeau and A. Nouaille, Tetrahedron Lett., 1990, 31, 2707–2710.
18. M. Sawada, Y. Takai, H. Yamada, S. Hirayama, T. Kaneda, T. Tanaka, K. Kamada, T. Mizooku, S. Takeuchi, K. Ueno, K. Hirose, Y. Tobe and K. Naemura, J. Am. Chem. Soc., 1995, 117, 7726–7736.
19. M. Sawada, Mass Spectrom. Rev., 1997, 16, 73–90.
20. M. Shizuma, Chiral Recognition in Mass Spectrometry, Focusing on FAB Mass Spectrometry, in Chiral Recognition in the Gas Phase, ed. A. Zehnacker, CRC Press, New York, 2010, ch. 5, pp. 61–86.
21. It was unable to estimate the stability constants of the complexation of [Cu(S,S-L)(CH₃OH)₂]²⁺ with S-Phe in solution, since the curve fitting of the titration plot could not be performed precisely due to the formation of the [Cu^II(S-Phe–H)₂] complex. Therefore, enantioselectivity could not be evaluated by UV-vis and CD measurements.
22. Structures of deuterium-labelled S-amino acids are shown in Table S2 (ESI†).
23. K _R and K_S are defined as described in page S9 of ESI†.
24. R. D. Gillard, P. R. Mitchell and H. L. Roberts, Nature, 1968, 217, 949–950.
25. S. Tashiro, Y. Ogura, S. Tsuboyama, K. Tsuboyama and M. Shionoya, Inorg. Chem., 2011, 50, 4–6.
26. J. Chin, S. S. Lee, K. J. Lee, S. Park and D. H. Kim, Nature, 1999, 401, 254–257.
27. A. Fujihara, N. Maeda and S. Hayakawa, Chirality, 2015, 27, 349–352.
28. W. A. Tao, D. Zhang, D. Wang, F. Thomas, P. D. Cooks and R. G. Cooks, Anal. Chem., 1999, 71, 4427–4429.
29. D. Schröder and H. Schwarz, Int. J. Mass Spectrom., 2004, 231, 139–146.
30. Y. Takai, K. Iguchi, H. Yamada, M. Shizuma, R. Arakawa and M. Sawada, J. Mass Spectrom., 2006, 41, 266–268.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental procedures including general methods and materials, synthesis of ligands, UV and CD spectra of complexes in solution, X-ray crystal structure analysis, DFT calculation, and mass spectra. CCDC 1912468. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c9cc07231d

---