Separation of polypeptide antibiotics by reversed ... - Springer Link

0 downloads 0 Views 524KB Size Report
Column liquid chromatography. Adsorption effects. Polypeptide antibiotics. Quantitative determination. Summary. The influence of different reversed-phase ...
Separation of Polypeptide Antibiotics by Reversed-Phase High-Performance Liquid Chromatography

H.-P. F i e d l e r * / T . H S r n e r / A . WSrn Universitw TiJbingen, I nstitut fLir Biologie II, Lehrstuhl Mikrobiologie I, Auf der Morgenstelle 28, D-7400 TiJbingen, FRG

Key Words Column liquid chromatography Adsorption effects Polypeptide antibiotics Quantitative determination

Summary The influence of different reversed-phase packings and the addition of acidic modifiers to the mobile phase was observed on the separation of basic and neutral polypeptide antibiotics by gradient elution. A dependence of pore size, coverage, reaction type and endcapping of the packings was not observed. Nevertheless, not all reversedphase packings were suitable for the separation of polypeptides, especially of basic molecules. The addition of phosphoric or perchloric acid to the mobile phase prevented adsorption of the basic polypeptide antibiotics on the stationary phase.

I ntrod uction The 22-peptide epidermin belongs to the polycyclic heterodetic class of peptide antibiotics [1] and is related to n/sin [2] and subtilin [3]. The presence of two molecules of lysine cause the basic properties of epidermin. The substance is soluble in water, moderately soluble in methanol and has a molecular weight of 2136. Epidermin is produced by Staphylococcus epidermidis TO 3298 and can be isolated from the culture filtrate of the fermentation broth. The substance is antibiotically active against Gram-positive bacteria, especially against the anaerobic pathogen Prop~on~bacterium acnes occuring in ache disease. The quantification of substances during the fermentative production process demands a sensitive, selective, reproducible and rapid determination method. Reversed-phase high-performance liquid chromatography seemed to be a

suitable method, but the HPLC analysis of epidermin by gradient elution with water/acetonitrile failed due to total adsorption of the substance on the stationary phases. Addition of the acidic modifier phosphoric acid in a concentration of 0.1% achieved a convenient separation of epidermin, but failed in the batch-to-batch reproducibility of some packing materials. This effect was probably caused by adsorption of the basic molecule on residual silanol groups of the stationary phase or by complexation on the reversed-phase packings. The HPLC separation on amino bonded phases resulted in a good retention behaviour, but was unsatisfactory due to tailing of the epidermin peak. Polystyrene packings were not efficient, because of the rapid elution of the antibiotic. Basic compounds often show an unfavourable adsorption on the stationary phase [4--6]. In this paper we describe the adsorption of polypeptides depending on endcapping and surface treatment of different reversed-phase packings, as well as on the addition of acidic modifiers to the mobile phase. The influence of the sample was tested by comparing the separation of epidermin with other polypeptide antibiotics. Herb/colin A [7], which contains 1 molecule arginine, /st a basic compound like epidermin, whereas alamethicin [8] and suzukacillin [9] are neutral polypeptides. The structures of the antibiotics investigated are shown in Fig. 1.

Experimental Columns Hyperchrome HPLC columns ( 1 2 5 m m x 4 . 6 m m I.D.) filled with Nucleosil-100 C-18 (7 #m), Nucleosil-300 C-18 (7#m), Spherisorb ODS II (5/~m) and Shandon Hypersil ODS (5 #m) were obtained from Grom (Ammerbuch, F RG). All columns were fitted with a pre-column (20mmx 4.6mm I.D., Grom). A LiChroCART HPLC cartridge (125mmx 4 mm I.D.), filled with LiChrosorb RP Select B (7/~m) fitted with a pre-column (4mm x 4mm I.D.) was obtained from Merck (Darmstadt, FRG).

Chromatographia Vol. 24, 1987 0009 5893/87 0433406 ~ 03.00/0

433 9 1987 Friedr. Vieweg & Sohn Verlagsgesel[schaftrnbH

Eluents

Chemicals

The polypeptide antibiotics were separated by gradient elution, using water, 0.1% aqueous phosphoric acid, 0.05% aqueous perchloric acid, respectively, as solvent A and acetonitrile as solvent B, from 0% to 100% solvent B in 10 minutes and a flow-rate of 2ml/min.

Acetonitrile (HPLC grade), phosphoric acid and perchloric acid (analytical grade) were obtained from Merck. Water was purified by means of a Milli-Q system (Millipore, Eschborn, FRG). Herbicolin A, alamethicin and suzukacillin were a generous gift of Prof. G. Jung (Universtiy of T(Jbingen).

HPLC Equipment An HP 1090A liquid chromatograph (Hewlett-Packard, Waldbronn, FRG) was used with a DR5 solvent delivery system, variable autoinjector and autosampler. 10#1 of standards and centrifuged biological samples were injected. The substances were monitored by means of an HP 1040A diode array detection system. A detection wavelength of 225nm and a bandwidth of 4nm was used, with 550nm and a bandwidth of 100nm as referecne. Detector signals were processed and recorded on an HP 3092A integrator.

Results The Effects of Different Stationary Phases and Mobile Phases on the Separation of Epidermin Fig. 2 shows the separation of epidermin on a commercial column filled with the endcapped reversed-phase material Nucleosil-100 C-18 by gradient elution from water to acetonitrile compared with gradient elution from phosphoric

Fig, 1 Structures of the polypeptide antibiotics epidermin, herbicolin A, alamethicin and suzukacillin.

434

Chromatographia Vol. 24, t 987

Fig. 2 Separation of epidermin on Nucleosil-100 C-18, with different mobile phases. a) Gradient elution from water to 100% acetonitrile. b) Gradient elution from 0.1% phosphoric acid to acetonitrile. c) Gradient elution from 0.05% perchloric acid to acetonitrile.

gradient profile up to 100% acetonitrile. On the other hand, a successful separation was obtained by adding phosphoric acid or perchloric acid in a concentration of 0.1% or 0.05%, respectively, to solvent A. The same selectivity could be achieved using the wide-pore endcapped packing Nucleosil-300 C-18 or the non-endcapped reversed-phase material LiChrosorb RP Select B with a pore size of 60 A. These results indicate, that the separation of the basic polypeptide epidermin is neither affected by non-endcapping of the packings nor by the pore size of 60 A.

acid and perchloric acid, respectively, to acetonitrile. The data show that the basic compound is totally adsorbed on the packing when gradient chromatography was performed with water/acetonitrile. No peaks were eluted during the Chromatographia Vol 24, 1987

In contrast to these results, no separation of epidermin could be achieved with the endcapped packing Spherisorb ODSII and the non-endcapped material Hypersil ODS. Addition of acidic modifiers to solvent A could not prevent adsorption on the stationary phase. No peaks were eluted by linear gradient elution from 0% to 100% acetonitrile.

435

Separation of Other Polypeptide Antibiotics and Their Dependence of the Stationary and Mobile Phase The observed adsorption effects on the reversed-phase packings relating to the chemically nature of the sample were investigated. The retention behaviour of epidermin was compared with other polypeptide antibiotics, such as herbicolin A, alamethicin and suzukacillin and their dependence on the stationary phase and addition of acidic modifiers to the mobile phase. As mentioned above, herbicolin A is a basic molecule, whereas alamethicin and suzukacillin are neutral compounds. The chromatographic conditions used were the same as for epidermin. In Table I are described the properties of the investigated reversed-phase packings [10, 11]. All packings are C18 modified silica-based materials with exception of LiChrosorb Select B which has a C8 modification. The selectivity of the different reversed-phase packings for the separation of basic and neutral polypeptide antibiotics is summarized in Table II. Neutral compounds showed no adsorption on the stationary phases and a convenient seParation could be performed w i t h o u t addition of any modifier to the eluents to restric peak tailing using Spherisorb and Hypersil. In contrast to the neutral polypeptides, basic compounds showed a strong adsorption on the different reversed-phase packing. A separation could not be achieved when gradient elution was performed with water/acetonitrile, no peaks were

eluted during the gradient profile up to 100% organic solvent. A convenient separation of herbicolin A could be achieved with all the investigated reversed-phase packings when perchloric acid was added to solvent A as acidic modifier. Quantification of Epidermin in Biological Cultures The quantitative determination of the basic polypeptide epidermin during the fermentative production was performed with the reversed-phase packing Nucleosil- 100 C-18 and addition of an acidic modifier to the mobile phase to prevent adsorption, as shown in the results mentioned above. Gradient elution from 0.1% aqueous phosphoric acid to acetonitrile was characterized by a convenient separation of epidermin in biological samples, but was poor with respect to the batch-to-batch reproducibility of the Nucleosil packing. This disadvantage could be avoided using 0.05% aqueous perchloric acid instead of phosphoric acid. With those concentrations of aqueous perchloric acid, no corrosion problems were observed. Fig. 3 shows the HPLC analysis of an epidermin standard solution and of a centrifuged fermentation broth. The quantification of compounds is facilitated by a linear response in the range of analytical interest. With UV detection at 225 nm, linearity is achieved in a concentration range between 5mg/I and 1 g/t, as shown in Fig. 4.

Table h Properties of the investigated reversed-phasepackings.

Pore size (A.) Coverage Reaction type Endcapping

Nucleosil-100 C-18

Nucleosil-300 C-18

Spherisorb OOS II

Hypersil ODS

LiChrosorb RP Select B

100 14%C trifunctional +

300 6%C trifunctional

8O 7%C trifunctional

100 9%C monofunctional

4-

4-

60 12%C bifunctionai -)*

)*special treated to give homogeneity of the silica surface.

Table II. Selectivity of reversed-phasepackings and their dependence on the mobile phasefor separation of the potypeptide antibiotics epidermin, herbicolin A, alamethicin and suzukacillin. Gradient

A

Antibiotic Epidermin Herbicolin A Alamethicin Suzukacillin Epiderm in Herbicolin A Alamathicin Suzukacillin Epidermin Herbicolin A Alamethicin Suzukacillin

Nucleosil-100 C-18

Nucleosil-300 C-18

Spherisorb ODS II

Hypersil QDS

LiChrosorb RP Select B

m

+

4-

+)* +)*

44-

+

+)*

+)*

+

4-

+

+

+)* +)*

+)* +

-- adsorption; + convenient separation; + )* peak tailing Gradient A: water to 100 o~ acetonitrile; Gradient B: 0.1% aqueous phosphoric acid to 100% acetonitrile Gradient C: 0.05% aqueous perchloric acid to 100% acetonitrile.

436

Chromatographia VoI. 24, 1987

Fig. 3

HPLC analysis of epidermin on Nucleosil-100 C-18 by gradient elution with 0.05% perchloric acid and acetonitrile; gradient from 22.5% to 37% acetonitrile in 8 minutes, monitored at 225nm. a) Epidermin standard (0.Smg/ml). b) Centrifuged fermentation broth.

Table III. Precision of retention times and peak areas.

Relative standard deviation (%)* Retention time Peakarea Epidermin, standard (0.5mgJmi) Epidermin, culture filtrate

0.2 0.2

1.3 2.2

*based on twenty replicate analyses.

Discussion

Fig. 4 Linear dynamic range of epidermin determination.

The precision of the method was tested by assaying the antibiotic standard and biological samples twenty times, as shown in Table III. The reproducibility of both the retention time and the peak area proved the system to be suitable for routine analysis. Chromatographla Vol 24, 1987

Undesired adsorption, especially of basic compounds, and the ~ow hydrolytic stabitity of a~kyl bonded-phase ~igands have been intensivly studied by other workers [ 4 - 6 ] . Qften the surface of the silica support used for reversedphase packings is not homogeneous. The support contains unevenly distributed residual silanol groups as well as trace impurities, such as metal ions, which are able to interact w i t h the sample. Such deleterious interactions can be observed mainly with polar, basic solutes [4]. It has been shown, that aqueous-organic solvent systems at low pH are deleterious to bonded-phase stability. Hydrolysis of the alkyl ligands is observed using aqueous phosphoric acid, trifluoroacetic acid or other acids in the range of about

437

pH 2 [5]. But in contrast to the frequently-held opinion it was proved, that a fully bonded packing is not required to prevent deleterious interaction of basic compounds, proteins or other macromolecules with residual silanol groups on the packing surface. A good separation of certain proteins could be performed, even when 50% of the original covalently bonded stationary phase was lost from the packing. Undesired adsorption of basic compounds could be strongly affected by the presence of highly acidic, isolated silanol groups on the packings, which are not associated [6]. Those authors subdivided the silica supports in relatively adsorptive materials, e.g. Hypersil and Spherisorb, and less adsorptive materials, e.g. Nucleosil. The more adsorptive packings exhibit a higher concentration of highly acidic, isolated silanol groups than the less adsorptive packings, which are characterized by a higher content of associated or bonded silanol groups. In the light of these results, adsorption or peak tailing of our investigated polypeptide antibiotics on reversed-phase packings like Hypersil and Spherisorb can be explained by the existence of highly acidic, isolated silanol groups. The basic compound epidermin was totally adsorbed on these packings and adsorption could not be prevented by addition of acidic modifiers to the mobile phase, in contrast to the less adsorptive reversed-phase materials Nucleosil and LiChrosorb RP Select B.

presence of highly acidic, isolated silanol groups on the packings and not by the content of associated or bonded silanol groups. The separation of basic polypeptides can be performed more favourable using reversed-phase packings of the less adsorptive type, which are characterized by a smaller content of highly acidic, isolated silanol groups. The mobile phase has to be acidified to prevent adsorption of the basic polypeptides on the stationary phase. Addition of perchloric acid seemed to be more favourable than phosphoric acid, regarding the batch-to-batch reproducibility of the reversed-phase packings.

Acknowledgements This work was supported by the Deutsche Forschungsgemeinschaft (SFB 323) and the Bundesministerium fiJr Forschung und Technologie (F6rderkennzeichen 03 8759).

References [ 1]

[2] [3]

H. AIIgaier, G. Jung, R.G. Werner, U. Schneider, H. Z#hner,

Angew. Chem. Int. Ed. Engl. 24, 1051--1053 (1985). E. Gross, J. L. More/I, J. Am. Chem. Soc. 93, 4634--4635 (1971). E. Gross, H. H. Kiltz, E. Nebelin, Hoppe-Seyler's Z. Physiol. Chem. 354, 810-812 (1973).

[4]

J. K6hler, D.B. Chase, R.D. Far/ee, A.J. Vega, J.J. Kirkland,

[5]

J. L. Glajch, J. J. Kirk/and, J. KShler, J. Chromatogr. 384,

[6]

J. KShler,

J, Chromatogr. 352, 275-305 (1986).

Conclusions The separation behaviour of polypeptides on reversedphase packings is mainly influenced by the presence of basic amino acids. Neutral molecules showed no separation problems, on the other hand the basic compounds were adsorbed on to the stationary phase. Not all reversed-phase packings proved to be suitable for the separation of basic polypeptides. Pore size, coverage and type of material bonded to the packings had no effect on selectivity and resolution of the investigated compounds. Endcapping is not essential for the separation of basic molecules. Undesired adsorption is caused by the

438

81--90 (1987). J.J. Kirk/and,

J. Chromatogr. 385, 125--150

(1987). [7] [8]

M. Aydin, N. Lucht, W. A. K6nig, R. Lupp, G. Jung, G. Winkelmann, Liebigs Ann. Chem. 1986, 2285--2300. G. Jung, N. Dubischar, D. Leibfritz, M. Ottnad, H. Probst, C. Stumpf in: Y. Wolman (Ed.), "Peptides 1974", North-

Holland Publ., Amsterdam, 1975, pp. 347--366. [9] [10] [11 ]

E. Katz, It#. Aydin, N. Lucht, W. A. KSnig, T. Ooka, G. Jung, Liebigs Ann. Chem. 1985, 1041--1062. R.E. Majors, J. Chromatogr. Sci. 18,488-511 (1980).

E. Merck Product Bulletin, RP Select B. Received: June 29, 1987 Accepted; July 26, 1987 F

Chromatographia Vol. 24, 1987