Heterogeneous cation induced clusters formed at surfaces of micro-droplets

Abstract

The surface of a micro-droplet can be treated as a simplified model to study reactions on a cell membrane. A novel group of heterogeneous cations induced clusters (HeteroCICs) that formed at the surface of micro-droplets was discovered by mass spectrometry (MS). This novel category of clusters are produced through two types of cations co-induced self assemblies of thymine molecules. Applying this phenomenon, an on-line derivatization and quantitation method of “ambient MS” style has been established, which can be used to real time monitor fish freshness changes during shelf-life.

---

Molecular recognition at the cell membrane lays the foundation for intercellular communications.^1–3 To shed light on the pathways of serious chemical reactions at the cell surface would improve our understandings of basic living processes. Micro-droplets generated from electrospray or neutral spray, whose volume resemble those of single cells, can be regarded as model systems for studying the chemistry both “on” and “in” cells.⁴ Previously, the different pH values “on” and “in” micro-droplets have been stressed.⁵ The unique acceleration effect within micro-droplets⁶ and at the micro-droplets’ surface⁷ has also been discussed. Specific reactions that only occurred at the surface of the micro-droplets can be of significance. Although proton and electron transfers^8–10 on a micro-droplet surface have been reported, the potential of non-covalent molecular recognitions at a micro-droplet surface has not been considered. Non-covalent interactions play key roles in intercellular communications, so specific non-covalent reactions that only occur at the surfaces of micro-droplets are supposed to be highly interesting and important.

The formation process of magic number clusters is a typical example of supramolecular recognition. Previously, the homogeneous cations induced thymine (and its analogs) clusters have been studied in our group.^11–14 Ammonium ions, alkali metal ions, and ions of organic primary amines can independently induce thymine (or its analogs) clusters. It has been shown that such homogeneous cations induced clusters (homoCICs), in which only one type of central cation is involved, are formed in liquid phase. In this study, we will show heterogeneous cations induced clusters (HeteroCICs), in which two types of central cations are involved, are formed only at the micro-droplets’ surface. Moreover, a quantitation method applying this newly discovered phenomenon has also been established. This novel method is another example of “ambient mass spectrometry (MS)”^15–17 detection, which features in very few sample-preparation and quick procedures.

Experiments were conducted on a home-built multichannel electrospray ionization (ESI) array¹⁸ (MRESI, Scheme S1†), upgraded with a neutral desorption (ND) system (Fig. 1a). This design was similar to neutral desorption extractive electrospray ionization¹⁹ (ND-EESI), except that multiple ESI emitters were used here. In the ND system, volatile compounds from liquid or solid samples can be desorbed and entrained in the neutral inert gas (nitrogen gas in this case) flow. The aerosol/molecules within the neutral gas flow can later be merged with charged micro-droplets generated from the ESI plumes, and after the ionization process, the MS signals corresponding to the desorbed compounds can be obtained. The locations of the ND tube end (I.D. 2 mm), ESI emitters and their positions relative to the MS inlet are shown in Fig. 1b.

Fig. 1 (a) The neutral desorption system coupled to the multichannel ESI array. Both liquid and solid samples can be desorbed by nitrogen gas flow. The neutral plume can be post-ionized by ESI spray plumes. (b) ND-MRESI configuration. The relative positions of the neutral desorption tube outlet, the three ESI sprayers and the MS inlet are demonstrated.

Trimethylamine (TMA) is a typical example of a tertiary amine, and according to our former experimental results, it cannot induce homoCICs in traditional ESI condition. Herein, if TMA (10 mL, 0.6 mM) was desorbed by nitrogen gas flow and thymine (0.2 mM) solution was used in ESI post-ionization sprays, as shown in Fig. 2a, no magic number clusters could be obtained. The ion m/z 610 was a background ion most probably originated from the plastic ND tube. However, when the TMA solution was spiked with ammonia (0.1% v/v), and a similar experimental procedure was applied, a TMA and ammonium co-induced thymine cluster, m/z 1110, could be obtained in the mass spectrum (Fig. 2b). This is the first heteroCIC discovered in our lab. To our surprise, if TMA, ammonia and thymine were pre-mixed and then the mixed solution was put to an ESI-MS analysis, only homoCICs of [Na + T₄]⁺, [NH₄ + T₅]⁺ and [2NH₄ + T₁₅]²⁺ could be obtained in the mass spectrum, as shown in Fig. 2c. These results indicated the heteroCIC m/z 1110 was unique compared to all the thymine clusters discovered previously. The structure, the way it formed and the applications of this special cluster are our major concerns.

Fig. 2 (a) ND-MRESI-MS spectrum obtained when the TMA solution was neutrally desorbed and the thymine solution was used as the ESI post-ionizing spray. (b) ND-MRESI-MS spectrum obtained when the TMA and ammonia mixed solution was neutrally desorbed and the thymine solution was used as the ESI post-ionizing spray. (c) ESI-MS obtained when mixed solution of TMA, ammonia and thymine was analyzed. (d) ND-MRESI-MS/MS spectrum of m/z 1110 in the experiment of (b).

Previously, we have produced evidence that magic number thymine clusters are commonly constructed from building blocks of 5 or 6 membered thymine circles.^11–14 Central cations were usually located at the cavities of such circles. Fig. 2d illustrates the fragments of the cluster m/z 1110 obtained under collision induced dissociation (CID) conditions. Based on previous data and the MS/MS results, the plausible structure of the ion m/z 1110 is proposed in Fig. 3. The model structure was optimized via molecular mechanical calculations as described previously.²⁰ The cluster is induced by one protonated TMA cation and one ammonium cation. The two cations are nearly aligned if we observe them in the downward direction (the ammonium part defined as top, Fig. 3b). 17 thymine molecules constitute, from top to bottom, “5 + 6 + 6” three consecutive circles, which surround the two central cations, as shown in Fig. 3a. This structure model can well explain the obtained fragmentation pathways (Fig. S1†).

Fig. 3 Proposed structure of the HeteroCIC m/z 1110. The structure has been optimized via molecular mechanical calculations. (a) Side view and (b) Top view of the HeteroCIC.

The next question is where does this heteroCIC form. We propose three possibilities: (1) it is formed in the liquid phase, (2) it is formed in the gas phase, and (3) it is formed at the liquid/air interface, also described as the surfaces of the micro-droplets. EESI methods generate ions on a milliseconds time scale. Recently, dual nano-ESI via theta capillary tips has been established as an in situ liquid-mixing method to generate micro-droplets with a life-time of also milliseconds.^21–23 Therefore, we tried to implement dual nano-ESI via theta capillary tips to synthesize the cluster m/z 1110. As shown in Fig. S2,† under all possible conditions, no such cluster could be obtained. These experiments indicate that the cluster m/z 1110 is not formed in the liquid phase. Furthermore, the reagents in the neutral spray and the ESI sprays were systematically changed and the influence on the signal intensity of m/z 1110 was investigated. The combinations of reagents and the experimental results are presented in Fig. 4a. To ensure the concentrations of the reagents introduced into the ionization zone in all the conditions are the same, instead of the ND system, a neutral spray (NS) assisted by nebulizing gas was implemented (details in the ESI†). When TMA and ammonia were co-sprayed in the NS plume and the thymine solution was applied as the ESI post-ionization plume (condition 1), the highest signal intensity of m/z 1110 was obtained. However, when TMA was sprayed in NS plume and ammonia/thymine mixed solution was applied as the ESI post-ionization plume, or when ammonia was sprayed in NS plume and TMA/thymine mixed solution was applied as the ESI post-ionization plume (conditions 2 and 3, respectively), the signal intensities of the cluster m/z 1110 were an order of magnitude lower. When TMA/ammonia/thymine were all sprayed in the NS plume and pure solvent (1 : 1 M/W) was applied as the ESI post-ionization plume (condition 4), the signal of m/z 1110 was absent. If the cluster m/z 1110 was formed in the gas phase, under conditions 1–4, the obtained signal intensities of m/z 1110 would have been similar. However, the experimental data denies such possibility, because under conditions 1–4, the signal intensities of m/z 1110 differed greatly. Considering the abovementioned evidence, we propose that the heteroCIC m/z 1110 is formed at the liquid/air interface (Fig. 4b).

Fig. 4 (a) Signal intensities of the HeteroCIC in different conditions. (b) Schematic illustration of the HeteroCIC formation at the micro-droplets surface.

The formation of the cluster m/z 1110 at the surfaces of the micro-droplets can explain the following two facts well. First, in conditions 2 and 3, because some of the cations “buried” in the charged micro-droplets, the surface concentrations of the cations were lower than those in condition 1, which resulted in lower signal intensities of m/z 1110. Second, similar to condition 2, various ammonium salts were used instead of ammonia. With the carbon chain of the anion in the ammonium salts increasing, the signal intensities of both m/z 963, a cluster formed in liquid, and m/z 1110 decreased, but the latter decreased much more intensely (Fig. S3†). As the carbon chain of the anion in the ammonium salts approach 10, these salts can be considered as surfactants, and the micro-droplets’ surfaces would be largely occupied by such anions. Because the heteroCIC is formed at the micro-droplets’ surfaces, the occupancy of surfactant at the surface would reasonably inhibit its formation. While for homoCIC (m/z 963, for instance), because it is formed “in” micro-droplets, the negative effect of the surfactant would be much less (note that generally, surfactants would cause ion suppression in MS analyses).

TMA is an important biogenic amine and can be regarded as the indicator of fish freshness. According to the Chinese National Standard, when the concentration of TMA exceeds 10 ppm, the fish is not eligible for selling.²⁴ Fish meat of 3 g was cut from a fish body, spiked with ammonia and then was directly put into the ND bottle. The amount of ammonia that the fish substrate was spiked with and the concentration of thymine in the ESI spray solution was optimized (Fig. S4†). When the fish was fresh, as shown in Fig. 5a, the signal of m/z 1110 was weak. When the fish was spoiled, as shown in Fig. 5b, the signal of m/z 1110 was strong. As the EIC in Fig. 5b shows, the signal intensity of m/z 1110 would decrease after a period of gas desorption, but if the position of the fish meat relative to the gas outlet in the ND bottle was changed to another spot, the signal intensity of m/z 1110 would return to the similar level. For each sample, quantitative experiments were performed by repeating this procedure 5 times and the data were collected at the beginning 60 s at each spot. A calibration curve was established, as shown in Fig. 5c. The limit of detection (LOD) is 0.2 ppm, which is as good as or slightly better than those obtained from published methods.^25–27

Fig. 5 (a) ND-MRESI-MS when fresh fish pieces (Cod) were analyzed. The total ion chromatogram (TIC) and extracted ion chromatogram (EIC) are shown in the upper panels. (b) ND-MRESI-MS when spoiled fish pieces (Cod, 24 h at room temperature) were analyzed. The TIC and EIC are shown in the upper panels. (c) Calibration curves established using Crucian fish substrate. LOD: 0.2 ppm. (d) Real time monitoring of the fish freshness changes during their shelf-life.

In real cases, the real time monitoring of fish freshness within the first few hours is challenging. The ideal method should be quick, sensitive and selective. By implementing our instrument, one fish sample can be analyzed within minutes and no labor-intensive sample preparations are needed. Because only TMA with ammonium cation can induce the cluster m/z 1110, high selectivity is guaranteed. Fig. 5d shows the real time monitoring of the freshness changes of three types of fish, under either room temperature or mild cold storage conditions. The detected concentrations were all below 10 ppm, demonstrating our method can be implemented to trace the spoiling process of fishes during their shelf-life. Dimethyl amine and imidazole can also induce heteroCICs with the assistance of ammonium, as shown in Fig. S5 and S6.† Similar quantitation methods²⁸ can be established accordingly.

The cell surface is a special platform for molecules/ions to react. Electrospray and neutral spray can generate micro-droplets that can be treated as simplified models to study reactions “in” and “on” cells. In this study, the first example of non-covalent clusters formed only at the surfaces of micro-droplets is demonstrated. Moreover, an “ambient MS” quantitation method of TMA was established and the real time monitoring of fish freshness has been conducted. The present study has offered a novel practical strategy to derivatize and quantify small molecules and will also motivate forthcoming studies to investigate supramolecular chemistry on the liquid/air interface.

Acknowledgements

We thank Dr Xin Zhang and Prof. Yuanhua Shao for access to theta pipettes. We also thank Dr Yafeng Li, Dr Anyin Li and Prof. Meiping Zhao for inspiring discussions. The financial support from the National Natural Science Foundation of China (no. 20727002 and 21075005) and the NSFC Funding (21445005) are acknowledged.

Notes and references

1. D. A. Doyle, A. Lee, J. Lewis, E. Kim, M. Sheng and R. MacKinnon, Cell, 1996, 85, 1067–1076.
2. J. Voskuhl and B. J. Ravoo, Chem. Soc. Rev., 2009, 38, 495–505.
3. F. X. Contreras, A. M. Ernst, P. Haberkant, P. Bjorkholm, E. Lindahl, B. Gonen, C. Tischer, A. Elofsson, G. von Heijne, C. Thiele, R. Pepperkok, F. Wieland and B. Brugger, Nature, 2012, 481, 525–529.
4. T. Muller, A. Badu-Tawiah and R. G. Cooks, Angew. Chem., Int. Ed., 2012, 51, 11832–11835.
5. H. Mishra, S. Enami, R. J. Nielsen, L. A. Stewart, M. R. Hoffmann, W. A. Goddard and A. J. Colussi, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 18679–18683.
6. M. Girod, E. Moyano, D. I. Campbell and R. G. Cooks, Chem. Sci., 2011, 2, 501–510.
7. Y. Li, X. Yan and R. G. Cooks, Angew. Chem., Int. Ed., 2016, 55, 3433–3437.
8. S. Enami, Y. Sakamoto and A. J. Colussi, Proc. Natl. Acad. Sci. U. S. A., 2014, 111, 623–628.
9. S. Enami, M. R. Hoffmann and A. J. Colussi, J. Phys. Chem. Lett., 2010, 1, 1599–1604.
10. H. Mishra, S. Enami, R. J. Nielsen, M. R. Hoffmann, W. A. Goddard and A. J. Colussi, Proc. Natl. Acad. Sci. U. S. A., 2012, 109, 10228–10232.
11. B. Qiu, J. Liu, Z. Qin, G. B. Wang and H. Luo, Chem. Commun., 2009, 20, 2863–2865.
12. B. Qiu, Z. Qin, J. Liu and H. Luo, J. Mass Spectrom., 2011, 46, 587–594.
13. Z. Qin, J. Liu, B. Qiu and H. Luo, J. Mass Spectrom., 2012, 47, 552–554.
14. Z. Qin, B. Qiu, J. M. Sun, W. B. Zhao and H. Luo, J. Mass Spectrom., 2014, 49, 266–273.
15. R. G. Cooks, Z. Ouyang, Z. Takats and J. M. Wiseman, Science, 2006, 311, 1566–1570.
16. G. A. Harris, A. S. Galhena and F. M. Fernandez, Anal. Chem., 2011, 83, 4508–4538.
17. K. Chingin, J. C. Liang and H. W. Chen, RSC Adv., 2014, 4, 5768–5781.
18. R. Qiu, C. S. Zhang, Z. Qin and H. Luo, RSC Adv., 2016, 6, 36615–36622.
19. H. Chen, S. Yang, A. Wortmann and R. Zenobi, Angew. Chem., Int. Ed., 2007, 46, 7591–7594.
20. R. Qiu, J. M. Sun, X. Zhang, W. B. Zhao, Z. Qin and H. Luo, Analyst, 2016, 141, 1641–1644.
21. D. N. Mortensen and E. R. Williams, Anal. Chem., 2014, 86, 9315–9321.
22. D. N. Mortensen and E. R. Williams, Anal. Chem., 2015, 87, 1281–1287.
23. C. M. Fisher, A. Kharlamova and S. A. McLuckey, Anal. Chem., 2014, 86, 4581–4588.
24. Number, GB/T 2733-2015.
25. S. Bourigua, S. El Ichi, H. Korri-Youssoufi, A. Maaref, S. Dzyadevych and N. J. Renault, Biosens. Bioelectron., 2011, 28, 105–111.
26. C. Carrillo-Carrion, B. M. Simonet and M. Valcarcel, Analyst, 2012, 137, 1152–1159.
27. G. M. Bota and P. B. Harrington, Talanta, 2006, 68, 629–635.
28. J. M. Sun, Z. Qin, J. Liu, C. S. Zhang and H. Luo, Analyst, 2014, 139, 3154–3159.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra06903g

---