Regular Articles
Rapid Analysis of Cyclic Peptide Cyclosporine A by HPLC Using a Column Packed with Nonporous Particles
2018 Volume 66 Issue 8 Pages 805-809
Details
2018 Volume 66 Issue 8 Pages 805-809
We developed a rapid and efficient analytical technique for cyclosporine A using HPLC on a column packed with 2-µm nonporous octadecylsilyl silica particles. Under optimized conditions, cyclosporine A was separated with high resolution from other cyclic peptides within 3 min, because the mass transfer resistance in the stationary phase was reduced by the use of the small, nonporous particles. Although the plate number increased greatly with the increase in the column temperature, the retention times were not affected. This behavior is different from other cyclic peptides or linear peptides. Based on its physicochemical characteristics, cyclosporine A is a poor hydrogen bond donor, and has a small topological polar surface area, low rotatable bond count, and high log P value. These results show that cyclosporine A is structurally rigid and undergoes poor water solvation even at high temperature. In the context of the rapid development of cyclic peptides with similar physicochemical characteristics to cyclosporine A, our developed method is useful for the development of cyclic peptide therapeutics.
Most of the currently used drugs are classified into two main categories: small-molecule drugs with typical molecular weights of <500 Da, and much larger “biologics” with typical molecular weights of >5000 Da.1) The conventional small-molecule drugs have the advantages of membrane permeability and oral availability, but the disadvantage of reduced target selectivity, which can cause an off-target effect in humans. In contrast, biologics tend to be specific for their targets owing to the higher number of drug–target interactions.2,3) However, proteins suffer from low bioavailability, poor membrane permeability, and instability in vivo.3,4)
Recently, peptide drugs, especially cyclic peptides in the size range of 5–50 amino acids, have actively been developed.5–7) There is currently a gap between the properties of small molecules and biologics. Cyclic peptides offer the potential of filling this gap and represent a class of molecules that have the specificity and potency of larger protein biologics. However, they are smaller in size, more accessible and cheaper to manufacture using chemical methods, thus they potentially combine several of the advantages of proteins.8,9) A typical cyclic peptide with a long history of use as a pharmaceutical is cyclosporine A.10) This natural peptide has widely been used as an immunosuppressant and contributed to revolutionizing organ transplant therapy.
In the development of drugs, a thorough characterization and quantification of the candidate compounds is required. Chromatography, which offers high speed and efficiency, is one of the main tools for the characterization and analysis of drugs.11–13) The pharmacopoeial method,11–13) which is recommended for the determination of cyclosporine A, uses HPLC with acetonitrile, water, tert-butyl methyl ether, and phosphoric acid, as well as traditional porous silica gel columns, at a high temperature of 80°C. However, this technique requires approximately 25–30 min to elute the analytes.11,13)
We were recently successful in the HPLC separation of monoclonal immunoglobulin G2 (IgG2) isoforms on a column packed with nonporous particles.14) In the present study, we applied a column packed with nonporous particles to perform the fast and efficient analysis of the cyclic peptide cyclosporine A. Furthermore, we also studied the elution behavior of a cyclic peptide and a linear peptide with similar molecular weights.
Cyclosporine A and polymyxin B sulfate were purchased from Sigma-Aldrich Co. (St. Louis, MO, U.S.A.). Daptomycin was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Des-Arg9-[Leu8]-bradykinin was from Peptide Institute (Osaka, Japan). Acetonitrile for HPLC analysis and trifluoroacetic acid (TFA) were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). All samples to be injected into the HPLC apparatus were filtered by membrane filters (0.2 µm, Merck Millipore, Japan).
ApparatusThe HPLC analysis was performed on a Shimadzu Nexera-i LC-2040C 3DMT. The system was equipped with a photodiode array detector (Shimadzu, Kyoto, Japan). Data acquisition and processing were performed using the Shimadzu LabSolutions software (version 5.87). The samples were separated using a Presto FF-C18 column (50×4.6 mm; particle size, 2 µm; Imtakt Corp., Kyoto, Japan), with mixtures of water containing 0.1% TFA (mobile phase A) and acetonitrile containing 0.1% TFA (mobile phase B) at various ratios as indicated, at a flow rate of 600 µL min−1. The column temperature was maintained at 75°C, unless otherwise stated, the detector wavelength was set and operated at 210 nm, and the column pressure was approximately 7.6 MPa at a column temperature of 75°C.
We used a column packed with nonporous particles (2 µm, 50×4.6 mm). The mobile phases were mixtures of water containing 0.1% TFA (mobile phase A) and acetonitrile containing 0.1% TFA (mobile phase B). Cyclosporine A was used as the analyte. The concentration of acetonitrile was referred to that from the methods recommended in the Japanese pharmacopoeia.11)
We first examined the effect of column temperature on the plate number and retention times of cyclosporine A using an acetonitrile ratio of 43%. The column temperature was varied from 40 to 70°C in increments of 10°C, and from 70 to 80°C in increments of 5°C. As shown in Fig. 1a, the elution peak shape was sharpened with increasing the column temperature. This trend is consistent with that from the analysis of monoclonal IgG2 isoforms.14,15) Raising the temperature also resulted in a large increase in the plate number (Fig. 1b). However, despite the large increase in the plate number, the retention time was not greatly affected by the column temperature (Fig. 1c). This trend is not consistent with that from the analysis of monoclonal IgG2 isoforms.14)
(a) The effect of column temperature on the peak shapes of cyclosporine A. (b) Effect of column temperature on the plate number of cyclosporine A. (c) Effect of column temperature on the retention time of cyclosporine A. Analytical conditions: Sample, 80 µg/mL cyclosporine A; Column, Presto FF-C18 column (50×4.6 mm; particle size, 2 µm); Mobile phase, water/acetonitrile/trifluoroacetic acid=57/43/0.1; Flow rate, 0.6 mL/min; detection, UV 210 nm.
We then increased the ratio of acetonitrile from 43 to 53%. At a temperature of 40°C, this increase in the acetonitrile ratio accelerated the elution of cyclosporine A. At the higher acetonitrile ratio, increasing the temperature resulted in a minimal change in the retention times (Fig. 2), but a large increase of the plate number, similar to the result obtained using a mobile phase consisting of 43% acetonitrile. In both mobile phase compositions, the plate number was increased by approximately 7 times when raising the temperature from 50 to 75°C. These results showed that the eluotropic strength did not affect the temperature-dependent behavior of cyclosporine A elution.
Sample, 80 µg/mL cyclosporine A; Mobile phase, water/acetonitrile/trifluoroacetic acid=47/53/0.1. Other conditions are the same as those described in Fig. 1. Column temperature (a) 50°C and (b) 75°C.
The injection volume and response time did not affect the plate number of cyclosporine A at 75°C (Supplementary Table 1), which showed that the analytical conditions were well optimized. Therefore, we used 75°C for the further analysis of cyclosporine A.
We also studied the effect of the column temperature on the elution behavior of another cyclic peptide, daptomycin. The mobile phases were adjusted so that the peak was eluted at a similar time to that of cyclosporine A at 40°C (Fig. 3a). As shown in Fig. 3b, the plate number of daptomycin did not change in response to the increase of column temperature, in contrast to the behavior of cyclosporine A, although the peak shape was improved (Fig. 3a). One reason for this unchanged plate number may be that the large decrease in the retention time of daptomycin in response to the increase of the column temperature canceled out the contribution of the peak sharpening (Fig. 3c), because the plate number is proportional to retention time.11–13)
(a) Chromatograph of daptomycin at two different column temperatures. (b) Effect of column temperature on the plate number of daptomycin. (c) Effect of column temperature on the retention time of daptomycin. Analytical conditions: sample, daptomycin 25 µg/mL; mobile phase, water/acetonitrile/trifluoroacetic acid (73/27/0.1, v/v). Other conditions are the same as those described in Fig. 1.
By use of gradient elution, this system also enabled a simultaneous analysis of three different cyclic peptides in a mixture sample of 100 µg/mL cyclosporine A, 125 µg/mL polymyxin B, and 25 µg/mL daptomycin, within 3 min (Fig. 4). This showed that a high separation efficiency could be achieved by gradient elution.
Analytical conditions: mobile phase A, water/trifluoroacetic acid (100/0.1, v/v); mobile phase B, acetonitrile/trifluoroacetic acid (100/0.1, v/v); flow rate, 0.6 mL/min; gradient program, 15 to 100% mobile phase B for 2 min; column temperature, 75°C. Other conditions are the same as those described in Fig. 1.
It was also shown that the retention time, peak area, and plate number of cyclosporine A showed acceptable repeatability, as the relative standard deviations of all parameters were less than 1% (n=6), both in gradient and isocratic elution conditions (Table 1).
| (a) | |||
|---|---|---|---|
| Gradient elution | Retention time | Area | Plate number |
| RSD (%) | 0.19 | 0.033 | 0.82 |
| (b) | |||
| Isocratic elution | Retention time | Area | Plate number |
| RSD (%) | 0.20 | 0.66 | 0.45 |
Analytical conditions: (a) gradient elution: mobile phase A, water/trifluoroacetic acid (100/0.1, v/v), mobile phase B, acetonitrile/trifluoroacetic acid (100/0.1, v/v); gradient program, 15 to 100% mobile phase B for 2 min; column temperature: 75°C. (b) Isocratic elution: mobile phase, water/acetonitrile/trifluoroacetic acid=57/43/0.1; column temperature: 75°C. Other conditions are the same as those described in Fig. 1.
For further study of this unique elution behavior of cyclosporine A, we analyzed a linear peptide, Des-Arg9-[Leu8]-bradykinin, with a similar molecular weight to that of cyclosporine A. The mobile phases were adjusted so that the retention time of Des-Arg9-[Leu8]-bradykinin at 40°C was approximately 3 min, similar to that of cyclosporine A (Fig. 5a). As shown in Fig. 5, with the increase in the column temperature, the increase in the plate number was much smaller than that of cyclosporine A (2 times and 23 times from 40 to 80°C, respectively) despite the improvement of peak shape (Figs. 5a and b), while the retention time was substantially reduced (Fig. 5c). This result was similar to that of daptomycin, and the large reduction in retention time was presumed to prevent a large increase in the plate number, in a similar way to the trend observed for daptomycin.
(a) Chromatograph of Des-Arg9-[Leu8]-bradykinin at two different column temperatures. (b) Effect of column temperature on the plate number of Des-Arg9-[Leu8]-bradykinin. (c) Effect of column temperature on the retention time of Des-Arg9-[Leu8]-bradykinin. Analytical conditions: Sample, 10 µg/mL Des-Arg9-[Leu8]-bradykinin; Mobile phase, water/acetonitrile/trifluoroacetic acid=89/11/0.1. Other conditions are the same as those described in Fig. 1.
To elucidate the mechanism for the unique behavior of cyclosporine A, that is, an unchanged retention time and large increase in the plate number, the physicochemical characteristics of cyclosporine A were compared with those of daptomycin and Des-Arg9-[Leu8]-bradykinin (Table 2). The computed properties16) showed that the hydrogen bond donor count, rotatable bond count, and topological polar surface area of cyclosporine A were smaller, while the calculated log P value was larger, than those of daptomycin and Des-Arg9-[Leu8]-bradykinin (Table 2). These physicochemical characteristics, which contribute to the favorable pharmaceutical properties of cyclosporine A, result from the following key structural features—1: its cyclic backbone structure and incorporation of a D-amino acid, which protect against proteolytic degradation and increase structural rigidity, 2: seven N-methyl groups, which reduce the number of amide groups available to act as hydrogen bond donors, and 3: intramolecular hydrogen bonds, which reduce the availability of amide NH protons.8,9) These characteristics suppressed the solvation of cyclosporine A by water even at high column temperature, and therefore resulted in the unchanged retention times as a function of increasing column temperature. It is also known that the retention times of neutral compounds are less sensitive to the increase in the column temperature compared with charged compounds.17) In contrast to cyclosporine A, for the more hydrophilic peptides daptomycin and Des-Arg9-[Leu8]-bradykinin, the calculated log P values are both negative (and have similar values), as shown in Table 2. This may be the cause of the similar reduction in retention times of both peptides in response to increasing column temperature.
| Cyclosporine A | Daptocycin | Des-Arg9-[Leu8]-bradykinin | |
|---|---|---|---|
| Molecular weight | 1202.635 g/mol | 1620.693 g/mol | 870.022 g/mol |
| Hydrogen bond donor count | 5 | 22 | 9 |
| Rotatable bond count | 15 | 35 | 21 |
| Topological polar surface area | 279 A^2 | 702 A^2 | 325 A^2 |
| X Log P3-AA | 7.5 | −5.1 | −3.6 |
The data were collected from the database “Pubchem search.”
With regard to the large increase in plate number induced by an increase in the column temperature, the contribution of the stationary-phase particle characteristics to the column efficiency is important. The effect of the physical, kinetic, and thermodynamic parameters of separation on plate height (H) can be generally determined using the van Deemter equation18):
 | (1) |
Moreover, the C term is also affected by the sorption and desorption rate of analytes on the column packing materials.20) Therefore, at higher temperature, the increase in these rates results in an increase in the mass transfer rate, which in turn causes a sharpening of the peak shape and increase of the plate number.17) This temperature effect, in addition to the use of a nonporous column, must contribute to the highly efficient analysis of cyclosporine A. Furthermore, decreasing the particle size also contributes to the decreases in the A and C terms of Eq. 1, which in turn leads to a reduction in H and thus to a sharpening of the peak shape.21) Therefore, the use of a column packed with 2-µm nonporous particles is advantageous for the analysis of macromolecules.
In this study, we developed a rapid and efficient analytical technique for cyclosporine A by means of HPLC on a column packed with 2-µm nonporous octadecylsilyl silica particles. Under optimized conditions, cyclosporine A was separated with high resolution and within 3 min from other cyclic peptides. Recently, cyclic peptides have been developed to protect against proteolytic degradation and increase the cellular permeability. For this purpose, peptides with structural characteristics similar to cyclosporine A will be ideal.22–24) Therefore, the developed analytical method in our system is advantageous for the rapid and efficient analysis of these cyclic peptides, because our system enables satisfactory retention of the peak while realizing a high plate number at a high column temperature. Furthermore, our results suggest that studying the retention behavior of cyclic peptides on our HPLC system can be used for evaluation of the physicochemical characteristics (including primary, secondary, and tertiary structures) of cyclic peptides that contribute to the cellular permeability.
This work was supported in part by the Research on Development of New Drugs and Research on Regulatory Harmonization and Evaluation of Pharmaceuticals, Medical Devices, Regenerative and Cellular Therapy Products, Gene Therapy Products, and Cosmetics from the Japan Agency for Medical Research and Development, AMED (17ak0101074j0801), and the research grant from the Pharmaceutical and Medical Device Regulatory Science Society of Japan.
The authors declare no conflict of interest.
The online version of this article contains supplementary materials.
