Preparation of an amino acid-based polymer monolith for trimodal liquid chromatography

Abstract

A versatile method for preparing a reverse phase/weak anion exchange/hydrophilic interaction trimodal polymer monolith was proposed by the one-step copolymerization of amino acid-based monomer and N,N′-methylenebisacrylamide. The retention mechanisms of the resultant monolith were evaluated using different series of small molecules. Furthermore, the separation of protein mixtures could be easily achieved under isocratic elution. The proposed polymer monolith, which has the merits of easy preparation, low mass transfer resistance and good biocompatibility, is a promising separation medium for small and large molecules.

---

Introduction

Multimodal liquid chromatography has been increasingly attractive because it can resolve a wide range of molecules with distinct properties using a single column.¹ To date, several such “universal stationary phases” have been developed by modifying the column matrix with ligands containing multiple interaction sites, for example, poly(ionic liquid).² Despite some progress, multiple synthetic steps are usually required. In addition, restricted mass transfer is often observed because very distinct adsorption–desorption kinetics may occur at each active site, hindering applications in separations of biomacromolecules.³ Monolithic materials, which were developed two decades ago, have held important positions in analytical chemistry.^4,5 The prominent feature distinguishing a monolithic column from a traditional packed column is the fast mass transfer, which originates from a large number of through pores in its structure.^6,7 Polymer monoliths, compared with their silica-based counterparts, are more suitable for biological applications because of mild preparation conditions and high stability over a wide pH range.⁸ Nevertheless, one-step preparation of polymer monoliths for multiple chromatographic modes still remains a challenge because of the limited number of multifunctional monomers.⁹

Amino acids and their derivatives, as a class of readily available and bioinert molecules, have been regarded as the potential sources for biomaterials.¹⁰ Till now, several kinds of amino acids have been introduced in polymer monolith for the different chromatographic purpose (e.g. chiral selector,¹¹ ion exchange sites,¹² and so on). However, it is more attractive to know whether the various chromatographic modes could be integrated into a single amino acid-monolithic column. Herein, we report a versatile method to produce trimodal polymer monolith by a one-step copolymerization of amino acid moiety-containing monomer with N,N′-methylenebisacrylamide (MBA). The monolith prepared from N-methacryloyl-L-phenylalanine methyl ester (MA-L-Phe-OMe) was taken as an example. Due to the presence of various functional groups (e.g. benzene ring and amine groups), the poly(MA-L-Phe-OMe-co-MBA) monolith could be operated in reverse phase/weak anion exchange/hydrophilic interaction chromatography (RP/WAX/HILIC). Moreover, the proposed polymer monolith was applied in separation of protein mixtures.

Experimental

Materials and apparatus

L-Phenylalanine methyl ester hydrochloride, N,N′-methylenebisacrylamide (MBA), and methacryloyl chloride were obtained from Aladdin Co. (Shanghai, China). 2,2′-Azobisisobutyronitrile (AIBN) was purchased from Shanghai Chemical Plant (Shanghai, China) and purified by recrystallization from ethanol before use. Concanavalin A, pepsin, ovalbumin, and human serum albumin (HSA) were purchased from Xinjingke Biotechnology Co., Ltd. (Beijing, China). All organic solvents used in chromatographic system were HPLC grade and provided by Beijing Wonder Technology Co., Ltd. (Beijing, China). Deionized water used in all experiments was obtained from Milli-Q system (Billerica, MA, USA). Other reagents not mentioned were all provided by Beijing Chemical Plant (Beijing, China) and directly used without purification.

Pore size distributions of the polymer monoliths were measured by exploiting an Autopore III 9220 mercury intrusion porosimeter (Micromeritics, USA). Fourier transform infrared spectra (FT-IR) were recorded on a Tensor-27 FT-IR spectrophotometer (Bruker, Germany). Scanning electron microscopy (SEM) photograph was obtained with a Model S-4300 scanning electron microscope (Hitachi, Japan). Chromatographic performance of the polymer monoliths was evaluated using a LC-20A HPLC system (Shimazu, Japan). The data processing was carried out with a HW-2000 chromatography workstation (Nanjing Qianpu Software, China).

Preparation of N-methacryloyl-L-phenylalanine methyl ester (MA-L-Phe-OMe)

N-Methacryloyl-L-phenylalanine methyl ester (MA-L-Phe-OMe) was synthesized through the acylation reaction between L-phenylalanine methyl ester hydrochloride and methacryloyl chloride. The synthesis procedure was according to the literature reported by Li et al. with some modifications.¹³ Briefly, L-phenylalanine methyl ester hydrochloride (6.15 g, 28.49 mmol) and triethylamine (5.30 mL) were added in dichloromethane (CH₂Cl₂, 150.0 mL). The above mixture was cooled in ice bath, and then added with methacryloyl chloride (2.76 mL, 28.49 mmol) under stirring. After the reaction solution was further stirred for 12 h at room temperature, the crude product was successively washed with HCl (1.0 M, 200.0 mL × 2), a saturated NaHCO₃ solution (200.0 mL × 2), and a saturated NaCl solution (200.0 mL × 2). Then, the organic layer was dried by MgSO₄. Finally, the pure product was obtained by column chromatography (20–25% ethyl acetate in petroleum ether).

Preparation of the polymer monolith

Poly(MA-L-Phe-OMe-co-MBA) monolith was synthesized within stainless column (50 × 4.6 mm i.d.). Typical procedure was as follows: the mixture of AIBN (10.0 mg, 0.06 mmol), MA-L-Phe-OMe (200.0 mg, 0.81 mmol), MBA (300.0 mg, 1.94 mmol), dodecanol (2.0 mL), and dimethyl sulfoxide (DMSO, 1.8 mL) was sonicated for 10 min. The resulting homogeneous solution was introduced into a column by a syringe. Then, the column was sealed at both ends and placed in oil bath (80 °C). After 12 h of polymerization reaction, the monolithic column was connected with HPLC system, washed with methanol at the flow rate of 0.1 mL min⁻¹ to remove the unreacted residues.

Results and discussion

Preparation and characterization of the polymer monolith

As displayed in Scheme 1, poly(MA-L-Phe-OMe-co-MBA) monolith was prepared via one-step free radical copolymerization with DMSO and dodecanol as the porogens.

Scheme 1 Synthesis of poly(MA-L-Phe-OMe-co-MBA) monolith.

The effect of preparative compositions on the physical properties of the resultant monoliths (monolith 1, monolith 2, monolith 3, monolith 4 and monolith 5) was investigated. As shown in Table S1,† for the monolith prepared with low content of poorer solvent (dodecanol, monolith 1), “stone-like” hard structure and high back pressure were observed, because the absence of poorer solvent hindered the formation of large pores in polymerization process.¹⁴ Then the effect of monomer/crosslinker ratios was studied at constant porogen percentage. We found that increase of the monomer/crosslinker ratios in polymerization mixture led to higher mechanical strengths and back pressures. When the weight percentage of monomer was increased to 60% (monolith 5), the back pressure of the resultant monolith was higher than the tolerant limitation of HPLC systems. The increasing back pressure alonged with monomer content was due to that larger amount of MA-L-Phe-OMe caused smaller pore size of the monolith, which could be further confirmed by the pore size distribution studies. Fig. S1† displays that average pore sizes of monolith 2 (40% monomer, wt%), monolith 4 (20% monomer, wt%) and monolith 3 (0% monomer, wt%) are 11.8 nm, 464.1 nm and 1676.3 nm, respectively. Moreover, the back pressures at different flow rates were studied (Fig. S2†), the linear correlations between back pressure and flow rate indicated that no breakdown occurred for the three monoliths. Comparison of separation efficiencies among these different monoliths prepared with various monomer/crosslinker ratios was performed. Tables S2 and S3† exhibit the chromatographic data for separating polycyclic aromatic hydrocarbons and nucleobases/nucleosides on monolith 2, monolith 3 and monolith 4, respectively. It could be found that the three nucleobases/nucleosides could be baseline separated on all of the three polymer monoliths under the same elution conditions. However, for polycyclic aromatic hydrocarbons, lower content of the monomer MA-L-Phe-OMe (monolith 3 and monolith 4) led to broader peaks and lower separation resolution, which is due to the insufficient amount of hydrophobic interaction sites. Therefore, the monolith 2 was selected for following studies as a compromise between back pressure and separation efficiency. FT-IR spectra of typical poly(MA-L-Phe-OMe-co-MBA) monolith (Fig. S3†) confirmed the presence of multiple functional groups.¹⁵ The peaks at 1542 cm⁻¹ and 1268 cm⁻¹ were for ΙΙ band, and ΙΙΙ band of secondary amine groups (–NH–). The peaks at 1739 cm⁻¹ and 1686 cm⁻¹ were characteristic peaks of carbonyl groups. The peak at 790 cm⁻¹ was attributed to the double bond in aromatic ring. Fig. 1 shows SEM photograph of the prepared monolith, similar with typical rigid polymer monoliths, numerous of microglobules were aggregated into globules clusters, among which large pores were formed, leading to the fast mass transfer of the monolith. It should be noted that the different pore sizes of monolith 2 between the SEM photograph and mercury intrusion study were observed, probably due to that the SEM exhibit superiority in observation of macropores (larger than 50 nm), whereas the mercury porosimetry provides the average pore size.

Fig. 1 SEM photograph of typical poly(MA-L-Phe-OMe-co-MBA) monolith (monolith 2).

Retention mechanisms of the polymer monolith

In the poly(MA-L-Phe-OMe-co-MBA) monolith, MA-L-Phe-OMe has hydrophobic domain (e.g. aromatic ring) and could attract hydrophobic molecules through hydrophobic interaction, whereas MBA was more hydrophilic and could interact with hydrophilic analytes. Therefore, hydrophobic–hydrophilic interaction retention mechanism of the monolith was studied. In Fig. 2(a), we found that the retention times of toluene and propyl-4-hydroxybenzoate significantly decreased with increasing acetonitrile (ACN) percentage, suggesting a reverse-phase retention behavior. It was also observed that propyl-4-hydroxybenzoate was eluted later than toluene, though toluene was more hydrophobic. The atypical retention behavior was due to hydrogen bonds formed between the monolith and phenolic esters. For acrylamide (AAm), the retention times obviously elongated with increasing ACN percentage from 80% to 95%, indicating a hydrophilic interaction-driven retention. Interestingly, separations of polycyclic aromatic hydrocarbons and nucleobases/nucleosides, which are typical hydrophobic and hydrophilic analytes, respectively, could be achieved on a single column by just altering the ACN content in the mobile phase (Fig. S4†). Anion exchange retention property of the monolith was further evaluated with three benzoic acids. As displayed in Fig. 2(b), we observed that the retention times of the analytes increased with decrease of pH from 6.0 to 4.0. In the pH range, the amine groups became protonated and positively charged, resulting in the stronger electrostatic interactions with the analytes and longer retention times. Additionally, in order to test the ion exchange contribution in the retention of benzoic acids, effect of salt concentration (c) in mobile phase on the retention factor (k′) was studied. Fig. S5† exhibited the linear relationships between log k′ value and log c value, which could be described by the following equation: log k′ = −s log c + b (s and b are constants). In pure ion exchange mode, s represents the number of charges involved in the ion exchange process. In this work, s was less than 1, explained by the fact that the hydrophobic interaction was still involved in the retention of the charged molecules. Our results indicated that the proposed monolith indeed could be operated in RP/WAX/HILIC trimodal chromatography. Series of non-polar, polar, and charged molecules could be resolved in different modes. The repeatability of the monolith was then tested. As shown in Table S4,† the run-to-run (N = 5) and column-to-column (N = 3) relative standard deviations (RSDs) for retention times of analytes were less than 2.59% and 4.01%, respectively, indicating an acceptable repeatability of the prepared monolith.

Fig. 2 (a) Effect of ACN percentage on the retention times of toluene, propyl-4-hydroxybenzoate, and AAm. Mobile phase: ACN–water. (b) Effect of pH on the retention times of benzoic acids. Mobile phase: 20.0 mM sodium phosphate buffer (pH = 3.0–9.0). Chromatographic conditions for (a) and (b): flow rate, 1.0 mL min⁻¹; UV detection, 254 nm; column, monolith 2. Error bars represent the standard deviations for three replicate determinations.

Versatility of the proposed method was demonstrated by replacing MA-L-Phe-OMe to another amino acid-based monomer, N-methacryloyl-_L-proline methyl ester. Fig. S6† exhibits the effect of mobile phase composition on the retention behavior of the modal analytes. Similar with the poly(MA-L-Phe-OMe-co-MBA) monolith, the proline-based monolith could also be used in the RP/WAX/HILIC trimodal chromatography. Moreover, separations of hydrophobic and hydrophilic analytes could be achieved (Fig. S7†).

Separation of protein mixture

As it is well known, separation and isolation of biomacromolecules, such as proteins, has been paid great concern in many bio-related research areas.^16,17 However, development of efficient separation medium for proteins is still a challenge because of the non-specific adsorption toward complex target analytes and the limited mass transfer. The application of the proposed monolith in biomacromolecules separation was further demonstrated. The other two monoliths, poly(MBA) monolith and poly(MA-L-Phe-OMe-co-ethylene dimethacrylate (EDMA)) monolith, were synthesized and used as the control monoliths. Among them, poly(MBA) monolith had the same preparative composition as that of poly(MA-L-Phe-OMe-co-MBA) monolith except the absence of MA-L-Phe-OMe; poly(MA-L-Phe-OMe-co-EDMA) monolith was synthesized with a more hydrophobic crosslinker EDMA instead of MBA. Four acidic proteins with different molecular weights and isoelectric points, including concanavalin A (Con A), pepsin (Pep), ovalbumin (Ova) and human serum albumin (HSA), were selected as model analytes. As displayed in Fig. 3(a), separation of proteins could be achieved on poly(MA-L-Phe-OMe-co-MBA) monolith. The detailed chromatographic information for protein separation is listed in Table S5.† However, for poly(MBA) monolith (Fig. 3(b)), the proteins could not be completely separated. The poor separation ability was due to the insufficient hydrophobic interaction between the proteins and the poly(MBA) monolith. For poly(MA-L-Phe-OMe-co-EDMA) monolith (Fig. 3(c)), only one small peak was observed within 60 min of retention time, because non-specific adsorption of the monolithic column was so strong that the proteins were difficult to be eluted. Based on the results, poly(MA-L-Phe-OMe-co-MBA) monolith greatly exhibited its advantages in protein separation, originated from the good permeability, reduced non-specific adsorption, and multiple interaction. Moreover, it should be mentioned that in the previous reports, protein separations in HPLC were mostly performed under gradient elution with increasing percentage of organic solvent or salts.^18,19 The harsh elution conditions were protein-unfriendly.²⁰ In the present work, protein separation could be easily achieved using organic solvent/salt-free mobile phase under isocratic elution. Compared with the reported protein separations under gradient elutions, comparable separation resolutions with acceptable broader peaks were observed, indicating that the proposed poly(MA-L-Phe-OMe-co-MBA) monolith has great potential in the application of all-aqueous chromatography for biomacromolecules.

Fig. 3 Chromatogram for separation of standard proteins on (a) poly(MA-L-Phe-OMe-co-MBA) monolith, (b) poly(MBA) monolith, and (c) poly(MA-L-Phe-OMe-co-EDMA) monolith. Chromatographic conditions: mobile phase, 20.0 mM sodium phosphate buffer (pH = 7); flow rate, 1.0 mL min⁻¹; UV detection, 280 nm. Peaks: 1. Con A; 2. Pep; 3. Ova; 4. HSA.

Conclusions

A kind of amino-acid monolithic column was developed for RP/WAX/HILIC trimodal chromatography. Different modes could be achieved for separating various series of analytes by altering the mobile phase composition. Compared with other reported multi-modal separation medium, the present one has following advantages: firstly, simple and convenient one-step preparation process; secondly, large analyte scope including not only small molecules but also proteins; thirdly and the most importantly, the MA-L-Phe-OMe could be extended to other amino acid-based monomers to produce trimodal polymer monolith. Therefore, the work provides a versatile method to synthesize the RP/WAX/HILIC trimodal monolithic column and the proposed monolith is a promising separation media in separation science.

Acknowledgements

We gratefully acknowledge the financial support from National Natural Science Foundation of China (No. 21175138, No. 21375132, No. 21135006, No. 21475137, No. 21205125 and No. 21321003).

Notes and references

1. F. R. Mansour and N. D. Danielson, Anal. Methods, 2013, 5, 4955–4972.
2. H. Qiu, A. K. Mallik, M. Takafuji, S. Jiang and H. Ihara, Analyst, 2012, 137, 2553–2555.
3. M. Lämmerhofer, R. Nogueira and W. Lindner, Anal. Bioanal. Chem., 2011, 400, 2517–2530.
4. R. Meinusch, K. Hormann, R. Hakim, U. Tallarek and B. Smarsly, RSC Adv., 2015, 5, 20183–20194.
5. R. F. Yu, W. L. Hu, G. Lin, Q. Xiao, J. N. Zheng and Z. A. Lin, RSC Adv., 2015, 5, 9828–9836.
6. N. Li, W. Zhang, Y. Shen, L. Qi, Y. P. Li, J. Qiao, F. Y. Wang and Y. Chen, J. Sep. Sci., 2014, 37, 3411–3417.
7. M. H. Wu, H. Q. Zhang, Z. X. Wang, S. W. Shen, X. C. Le and X. F. Li, Chem. Commun., 2013, 49, 1407–1409.
8. J. Urban, F. Svec and M. J. Fréchet, Anal. Chem., 2010, 82, 1621–1623.
9. H. Huang, Z. Lin, Y. Lin, X. Sun, Y. Xie, L. Zhang and G. Chen, J. Chromatogr. A, 2012, 1251, 82–90.
10. Z. Y. Qiao, C. Y. Hou, D. Zhang, Y. Liu, Y. X. Lin, H. W. An, X. J. Li and H. Wang, J. Mater. Chem. B, 2015, 3, 2943–2953.
11. J. He, X. Wang, M. Morill and S. A. Shamsi, Anal. Chem., 2012, 84, 5236–5242.
12. C. Aydoğan, A. Tuncel and A. Denizli, J. Sep. Sci., 2012, 25, 1010–1016.
13. D. Yu, C. Luo, W. Fu and Z. Li, Polym. Chem., 2014, 5, 4561–4568.
14. F. Svec, J. Chromatogr. A, 2010, 1217, 902–924.
15. J. C. Yang, M. J. Jablonsky and J. W. Mays, Polymer, 2002, 43, 5125–5132.
16. Y. Li, B. Gu, H. D. Tolley and M. L. Lee, J. Chromatogr. A, 2009, 1216, 5525–5532.
17. Y. Liu, L. Ren and Z. Liu, Chem. Commun., 2011, 47, 5067–5069.
18. J. Zhao, F. Gao, Y. Fu, W. Jin, P. Yang and D. Zhao, Chem. Commun., 2002, 752–753.
19. B. Mayr, R. Tessadri, E. Post and M. R. Buchmeiser, Anal. Chem., 2001, 73, 4071–4078.
20. Z. Liu, R. Su, X. Liang, Y. Liang, Y. Deng, Y. Li and R. Dai, RSC Adv., 2014, 4, 15830–15834.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra12203a

---