Degradation of acylaminopenicillins with regard to their pH dependency

Zbl. Bakt. Hyg. A 267, 531-536 (1988)

Degradation of Acylaminopenicillins with Regard to their pH Dependency j . KNOLLER, W. SCHONFELD, K. D. BREMM, and W. KONIG Lehrstuhl fiir Medizinische Mikrobiologie und Immunologie, AG Infektabwehrmechanismen, Ruhr Universitat, D-4630 Bochum With 1 Figure' Received March 20, 1987 . Accepted July 23, 1987

Summary Determination of antibiotic concentration is performed in many biological fluids and tissues whichall havedifferent pH values. Therefore, we investigated the in vitro stabilityof three acylaminopenicillins (azlocillin, mezlocillin and piperacillin) in borate buffer by the HLPC technique with regard to pH dependency. HPLC allows the detection of all three substances togetherwith their metabolites, penicilloate and penilloate, within 15 min. Decomposition was monitored at 37°C during a 24 h incubation period (pH values ranged between pH 3.0 and 10.0). The highest degradation rates were observed with a buffer solutionof pH = 10.0: 50% of the azlocillin and 83% of the mezlocillin weredecomposed after 8 h while under the same conditions, piperacillin was completely decomposed already after 1 h. The higheststability was detected in borate bufferat pH values of 4.0, 5.0, and 6.0. At pH = 3.0, degradation was determined as follows : 31% of the piperacillin, 39% of the mezlocillin, and 45% of the azlocillin were decomposed after 24 h. Penilloic acid was identified as the main metabolite in contrastto buffersolutions with higherpH values which only revealed negligible amounts of this compound. Zusammenfassung Die Erfassung von Antibiotikakonzentrationen wird in vielen biologischen Fliissigkeiten und Geweben durchgefiihrt, die aileverschiedene pH Werte besitzen. Da im Gegensatz zum Serum und zur Pufferlosung (pH = 7,4) im Urin (pH = 4,5-5,5) keine Zersetzung der Acylaminopenicilline festgestellt werden konnte, untersuchten wir die in vitro Stabilitatder drei Acylaminopenicilline Azlocillin, Mezlozillin und Piperacillin in Boratpuffer in Abhangigkeit vom pH-Wert. Mit Hilfe der HPLC-Technik war es moglich, die Metabolite Penicilloat und Penilloat zusammen mit den Penicillinen selbst innerhalb von 15 min in einem Analysengang zu bestimmen. Der Abbau wurde bei 3rC und verschiedenen pH Werten (pH = 3,0 - pH = 10,0) iiber einen Zeitraum von 24 h verfolgt, Die hochsteAbbaurate wurde beipH = 10,0 gefunden: 50% des eingesetzten Azlocillins und 83% desMezlocillins waren nach 24 h abgebaut, wahrend sich das Piperacillin unter diesen Bedingungen schon nach einerStundefast vollstandig zersetzt hatte. DiegroBte Stabilitatkonnteim pH-Bereich von 4,0 bis 6,0 festgestellt werden. Bei pH = 3,0 hatte sich folgender Prozentsatz der eingesetzten Penicilline zersetzt: 45% des Azlocillin, 39% des Mezlocillin und 31% des

532

J. Kneller, W. Schonfeld, K.D. Bremm, and W. Konig

eingesetztenPiperacillin.1mGegensatz zu Pufferlosungenmit hoheren pH Werten wurde bei pH = 3,0 das Penilloat als Hauptmetabolit identifiziert, das bei hoheren pH Werten nur in vernachlassigbar geringen Mengen nachgewiesen werden konnte. Introduction In the last 10 years, the HPLC technique has become an excellent tool in microbiology as well as a valuable supplement to conventional microbiological assays (8, 10) because of its speed, specificity and sensitivity. One of the major benefits of the HPLC method is the possibility of detecting metabolites of the original compound and to separate these from biological matrices and the antibiotic drug itself. This question is of particular importance in cases where the determination of ~-lactam antibiotics is performed in biological fluids and tissues which all have different pH values: e.g. serum (pH = 7.4, urine (with reported values between 5 and 8.5) (4) as well as pancreatic fluid (pH = 8.3-9.0). Therefore, we investigated the in vitro stability of three acylaminopenicillins (azlocillin, mezlocillin and piperacillin) in borate buffer by HPLC with regard to their pH dependency. HPLC allows the detection of all three substances together with their metabolites, penicilloate and penilloate within 15 min. In earlier studies, no degradation of the acylaminopenicillins in urine (pH = 4.5-5.5) was observed in contrast to serum (pH = 7.4) and buffer (pH = 7.40) (6, 7, 9). Furthermore, the stability of the penicilloic acid of azlocillin (which was analyzed as a representative of the other penicilloic acids) was assayed in borate buffer and serum to investigate the further degradation pathway of the penicillins at low (pH = 3) and high (pH = 10) pH values.

Material and Methods

Study design: Serum specimens were obtained from 3 healthy donors. Azlocillin, mezlocillin and piperacillin as well as azlocillin-penicilloate, respectively, (250 ug/ml) were added to three buffer and three serum samples and divided in to 200 ul portions. For serum and buffer kinetics, samples containing the antibiotic drugs were incubated under sterile conditions at 37°C in a water bath (pH values ranged between pH 3.0 and 10.0). Aliquots of the samples were analysed before incubation, directly after addition ofthe penicillins and after 2, 4, 8 and 24 h. Sample preparation: Sample preparation for HPLC required only dilution of the serum (1:5). This buffer is fully compatible with the biological samples, so that no crystallisation was observed. The diluted samples were centrifuged at 9300 x g for 4 min; subsequently,20 J1.l of the resulting supernatant were then subjected to HPLC. Analysis by HPLC: Azlocillin, mezlocillin as well as the corresponding penicilloates and penilloates were kindly provided by Bayer (Leverkusen F.R.G.), piperacillin was obtained from Cyanamide Lederle (Miinchen, F.R.G.). Methanol, dipotassium hydrogen phosphate and phosphoric acid were purchased from Riedel de Haen (Seelze, F.R.G.), acetonitrile was purchased from Baker Chemicals (GroS-Gerau, F.R.G.). HPLC was performed with a Constametric III-G pump, a variable wavelength UV-detector Spectromonitor D (LDC-Milton Roy, Hasselroth, F.R.G.), an SP 8780 XR autosampler (Spectra, Physics,Darmstadt, F.R.G.), as well as a Nelson Analytical System (AS Analysensysteme, Wuppertal, F.R.G.). The column (4 x 200 mm) was filled with reversed phase material Nucleosil5C18 (Macherey und Nagel, Duren, F.R.G.). The solvent systems consisted of a mixture of acetonitrile and phosphate buffer (43 mMol K2HP04 ad 11 H 20):

Degradation of Acylaminopenicillins

Azlocillin Mezlocillin Piperacillin AzlocillinPenicilloate

acetonitrile

phosphate buffer

pH

20% 25% 25%

80% 75% 75%

5.00 4.50 4.50

17%

83%

5.50

533

The mobile phases were degassed and delivered with a flow of 1 mUmin at room temperature. The absorbance of the column effluent was monitored at 220 nm (0.2 absorbance units full scale). Standards (5 flg/ml,25 ug/ml, 50 ug/ml, 100 ug/ml and 250 ug/ml)of azlocillin, mezlocillin, piperacillin and of the metabolites were prepared by serial dilution of a stock solution (freshly prepared every day) containing 1 mg/ml of the drug with borate buffer (pH 7.40). All samples were stored at -70°C until analysis and kept at 4°C prior to injection, Peaks obtained by chromatography were identified as to their retention times (monitoring at 220 nm). The concentrations were determined by external standardisation. Calibration curves were obtained by plotting the peak area against the concentrations of the standard solutions. Statistical analysis was performed using Students t-test.

Results The linear relationship between penicillin concentration and peak area for all three penicillins ranged between 0.5 and 250 ug/ml with a correlation coefficient of more than 0.99. For mezlocillin e.g. the following coefficients of variation were determined: 3.8% (at a mezlocillin concentration of 1 ug/ml), 1.4% (50 ug/ml) and 1.1 % (100 !1g! rnl). The recovery was calculated by comparing samples from penicillin-spiked serum with those of buffer solutions: at a concentration of 1 !1g!ml the recovery amounted to 97% and one of 100 ug/ml almost to 100%. The detection limits for the three penicillins are 0.05 !1g!ml, 0.05 ug/ml (penicilloates) and 0.2 ug/ml (penilloates), respectively. An interference of the quantitation with coeluting peaks derived from other compounds such as additionally administered drugs or with peaks of the biological matrix was not observed. In Tables 1-3, the degradation rates of the antibiotic drugs have been shown with regard to their pH-dependency: At pH values between 4.0 and 6.0, the maximum

Table 1. Degradation of Piperacillin at 37°C in borate buffer of various pH values: Percentage of the initial concentration (100%); mean values; standard deviation (n = 3) was ± 5% Piperacillin pH 0 2h 4h 8h 24h

3.0

4.0

5.0

98.0 94.9 83.0 69.4

101.2 100.8 97.8 99.4

103.5 99.4 98.4 101.3

6.0 100.00 102.2 103.2 99.4 98.6

8.0

9.0

10.0

97.9 90.9 78.6 44.2

92.8 39.8 8.4 0

0 0 0 0

534

]. Kneller, W. Schonfeld, K.D. Bremm, and W. Konig

Table 2. Degradation of azlocillin at 37°C in borate buffer of various pH values: Percentage of the initial concentration (100%); mean values; standard deviation (n = 3) was ± 5% Azlocillin pH 0 2h 4h 8h 24h

3.0

4.0

5.0

102.0 96.4 92.6 54.4

98.8 96.3 92.7 78.6

102.3 99.5 101.3 102.4

6.0 100.00 104.3 101.7 99.3 100.6

8.0

9.0

10.0

99.4 97.3 85.9 80.5

98.3 89.1 83.8 45.3

99.4 92.2 49.9 16.7

Table 3. Degradation of mezlocillin at 37°C in borate buffer of various pH values: Percentage of the initial concentration (100%); mean values;standard deviation (n = 3) was ± 5% Mezlocillin pH 0 2h 4h 8h 24h

3.0

4.0

5.0

98.3 91.9 79.3 61.3

99.5 101.7 102.4 101.3

98.4 99.2 103.1 99.7

6.0 100.00 97.8 98.7 99.6 102.4

8.0

9.0

10.0

98.1 99.7 97.4 83.2

99.3 83.3 72.3 2.0

98.7 63.7 27.1 0.8

stability for all three pencillins could be observed in which piperacillin was the most stable acylaminopenicillin at low pH values. At pH = 3.0,45% of azlocillin, 39% of mezlocillin and 31% of piperacillin were found to have been degraded after 24 h at 3rC. The penilloic acid was identified as the main metabolite at this pH value in contrast to previous degradation studies in buffer solutions (pH = 7.40) which showed onlynegligible amounts of this compound. At pH values of 9.0 and 10.0, the antibiotic drugs were decomposed to a large extent: 16% of the azlocillin and 18% of the mezlocillin had been degraded after 8 h, up to 55% and 98% after 24 h at pH = 9.0, respectively. Under these conditions, piperacillin was completely decomposed; at pH = 10.0 piperacillin was no longer detected already after 1 h. In serum and in buffer solution of high pH values (pH = 10.0), no degradation of the penicilloic acid (e.g, azlocillin) could be observed; 80% of the penicilloate were degraded to penilloate pH = 3.0 after 8 h 37°C (half-life at this pH value was 228 min)

Discussion The present study confirms that the HPLC method is an excellent tool for degradation studies due to a rapid and sensitive determination of the metabolites combined with the antibiotic drugs. An additional advantage is the simple sample preparation in contrast to a number of HPLC methods described earlier (1, 2, 3, 5); the latter require

Degradation of Acylaminopenicillins

PC

535

AZ

PN PC

® «

o

!

,

,

5

!

!

I

10 min

I

o

!

(

J

5

,

.,

10min

Fig. 1. Representative chromatograms (azlocillin as a representative of the other acylaminopeniciIlins): A. after 24 h at pH = 9.0. B. after 24 h at pH = 3.0. PC = Penicilloate; PN = Penilloate; AZ = Azlocillin.

either time-consuming extraction steps with poor recovery or solid phase extraction. The results obtained in this study have confirmed previous observations (6), i.e. that the acylaminopenicillins in urine (pH = 4.70-5.80) are more stable than in serum (pH = 8.80) and buffer solution (pH = 7.40). As shown maximal stability for the acylaminopenicillins was obtained between pH = 4.0 and 6.0. Lower pH values led to an acid-catalyzed hydrolysis of the ~-Iactam ring; this results in a subsequent dearboxylation of the penicilloic acid to penilloate, which is the main metabolite at this pH value. These data were supported by degradation studies with penicilloate, which showed the decomposition of the penicilloic acid at pH = 3.0 via decarboxylation to the penilloic acid. At low pH values the free carboxylic acid group may form a six-center transition state involving the amino group of the thiazole ring which is thermodynamically favourable for decarboxylation. Therefore we could not detect any degradation of the penicilloate in alkaline solution because the carboxylic anion is not able to form a comparably suitable transition state for decarboxylation. At higher pH values, a rapid basic catalyzed hydrolysis of the ~-lactam ring took place, especially at pH = 9.0 and 10.0. With regard to azlocillin and mezlocillin, our results support the data of Gundert-Remy and Weber (4). It appears that the in vitro degradation of the acylaminopenicllinins in serum is not only due to aqueous hydrolysis (the pH value of the serum was 8.85) but also to the nucleophilic attack of polar groups like amino and hydroxy groups of proteins which then leads to the generation of protein-bound penicilloate: In previous studies (6) on the degradation of mezlocillin, it was observed that an additional peak appeared in the polar region of the chromatograms; our data suggested that this peak represented a degradation product of the acylaminopenicillins in serum. A confirmation was now obtained by the following

536

J. Kneller, W. Schonfeld, K. D. Bremm, and W. Konig

facts: 1. The obvious stability of the penicilloic acid in serum, i.e, that the penicilloate in alkaline solution is not able to bind to serum proteins because of the unreactive carboxylic anion. 2. In buffer solution with high pH values , the penicilloate was the onl y degradation product. In contrast to the high stability at low pH values, piperacillin was the most unstable acylaminopenicillin in alkaline solution: At pH = 9.0, only negligible amounts of piperacillin were detected after 8 h, while at pH = 10.0, the total amount of the compound was degraded already after 60 min. Our results clearly indicate, that acylaminopenicillin-containing serum samples have to be analyzed with meticulous care: these samples should be rapidly worked on and instantly processed or otherwise kept cool (-70°C) until the processing of the samples can be carried out.

References

1. Fasching, C. E. and L. R. Peterson: High pressure liquid chromatographic assay of azlocillin and mezlocillin with an anion-exchange extraction technique. J. Liq. Chromatogr. 6 (1983) 2510-2520 2. Fiore, D., F. A. Auger, G. L. Drusano, and L. l- Lesco: Improved micromethod for mezlocillin quantitation in serum and urine by high pressure liquid chromatography. Antimcrob. Agents Chemother. 26 (1984) 775-777 3. Gundert-Remy, U. and j. X. DeVries: Determination of the ureido-peniciIlins azlocillin, mezlocillin and Bay K 4999 in plasma by high performance chromatography. Brit. J. Pharmacol. 8 (1979) 589-592 4. Gundert-Remy, U. and E. Weber: Degradation von Azlocillin and Mezlocillin. Arzneim.-Forsch.lDrug Res. 31 (1981) 2041-2044 5. Hildebrandt, R. and U. Gundert-Remy: Improved procedure for the determination of the ureidopenicillins azlocillin and mezlocillin in plasma by high performance liquid chromatography. J. Chromatogr. 228 (1982) 409-412 6. Knoller, ]., W. Schonfeld, K. D. Bremm, and W. Konig: Determination of aspoxicillin by high-performance liquid chromatography. Stability at different temperatures. Zbl. Bakt. Hyg. A 265 (1987) 176-181 8. Nilson-Ehle, I.: High performance liquid chromatography for analysis of antibiotics in biological fluids. J. Liq. Chromatogr. 6 (1983) 251-293 9. Schonfeld, W., j. Knoller, K. D. Bremm, and W. Konig: Degradation of azlocillin hin human serum. Zbl. Bakt, Hyg. A 260 (1985) 254-259 10. Weber, A., K. E. Opheim, K. Wong, and A. L. Smith: High-pressure liquid chromatographic quantitation of azlocillin. Antimicrob. Agents Chemother. 24 (1983) 750-753 Dr. rer. nat. j. Kneller, Prof. Dr. med. W. Konig, Medizinische Mikrobiologie und Immunologie, AG Infektabwehrmechanismen, Ruhr-Universitat Bochum, Universitatsstr. 150, 0-4630 Bochum