- Email: [email protected]
Spectrophotometric Determination of Cycloserine with 9-Methoxyacridine
Spectrophotometric Determination of Cycloserine with 9-Methoxyacridine JAMES T. STEWART.' AND GYURNG S YOO~ Heceived June 30, 1987, from the 'Department of Medicinal Chemistry and fharmacognosy. College of Pharrnac University of Georgia, Athens, GA 30602 Accepted for publication January 4, 1988 'Present address College of Pharmacy, Chonnam kational University, Yongbong dong, 300, Kwangju. 505 Korea Abstract _' Spectrophotometric assay for cycloserine based on the interaction of the drug with 9-methoxyacridineas a chromogenic agent
is described The highly colored substituted acridine product was identified as 9-(d-4 imino-3-isoxazolidinone)acridine Color development was affected by time and temperature of heating and by the quantity of 9-methoxyacridine reagent utilized The absorbance at 438 nm is linearly proportional to concentrations of cycloserine with a detection limit of 0 3 pg mL The optimum range for the assay of cycloserine was from 5 0 10 to 3 0 x 10 M (correlationcoefficient 09999, n = 6) When applied to cycloserine capsules labeled to contain 250 mg. t h e proposed method gave mean recoveries of 101 84 0 48% The procedure is sufficiently sensitive, precise, and accurate for the determination of cycloserine in its dosage form A
~
+
Cycloserine I ~d)-4-amino-3-isoxazolidinone] is a n antimicrobial substance producod by t h e growth of certain s t r a i n s of Struptomvc.e.s orchzdacrus or S. garyphalus, a n d is administered for t h e t r e a t m e n t of pulmonary tuberculosis and certain genitourinary infections, especially Rnterohacturia and Esrhrmchin coli.1 Cycloserine is characterized by an ultraviolet absorption band peak a t 226 n m in water. Because it h a s relatively low intensity ( e 3940),it is difficult to assay for t h e d r u g in dosage forms by UV spectrophotometry, even if it is separated from matrix materials. The USP method2 for t h e determination of cycloserine d r u g substance a n d in capsules is based on a microbial assay. In t h e British Pharmacopeia, t h e assay methods used a r e titrimetric:' and spectrophotometricJ procedures. The titrimetric method i performed in aqueous propanol and uses thymolphthale as indicator and sodium hydroxide as t i t r a n t . The spectrophotometric assay is based on t h e formation of a color complex with sodium nitroprusside in a n acetate buffer measured a t 6% n m . I n this method, t h e chromogenic agent is unstable and so must be used immediately and discarded after one set of determinations." The reacted colored complex cannot be diluted and t h u s , if samples w e too d a r k to read, t h e assay m u s t be repeated using small aliquots. Also, t h e cyc1oserine:nitritopentacyanoferroate complex is found to be sensitive to temperature changes. Other published quantitative determinations for cycloserine include GLC," spectrophotometric,' a n d spectrofluorometric methods.' These assay methods involve many procedural steps, including control of pH. a n d t h e need for a high reaction temperature. 9-Methoxyacridine ( % M A ) was developed as a chromogenic reagent in t h i s laboratory as a n alternative to B-chloroacridine!' becausc of stability problems associated with 9-chloroacridine chemistry. This paper describes a spectrophotometric determination of cycloserine and i t s dosage form based on derivative formation with 9-methoxyacridine. A discussion of t h e reaction mechanism is also included. 452 Journal of Pharmaceutical Soences Vol 77, N o . 5, May 1988
Experimental Section A p p a r a t u s - T h e spectrophotometer system consisted of a PerkinElmer model 7500 professional computer. a Beckman model DU-7 spectrophotometer. a Hewlett-Packard model 7470A plotter, and 10m m quartz cells. An American Optical shaking water bath (model 0 2 1 5 6 ~a n d Crest ultrasonics b a t h were used in t h e analytical procedure. All melting points were uncorrected and recorded on a Thomas Hoover capillary melting point a p p a r a t u s . Infrared spectra were run on a Perkin-Elmer model 684 spectrophotometer The ' H S M R spectra were recorded on a Jeol FX 90Q E T NMN spectrometer with tetramethylsilane used a s a n internal standard. Elemental analyses were performed by Atlantic Microlab, Inc., Atlanta, GA. R e a g e n t s a n d C h e m i c a l ~ 9 - . M e t h o x y a c r i d i n e(9-MA) was synthesized from 9-chloroacridine ( E a s t m a n Kodak Company) and sodium methoxide, using t h e method of Barber e t a1 ( m p 103 and recrystallized from petroleum e t h e r (30-60 'C I. Cycloserine reference standard was obtained from t h e L'nited States Pharmacopeia a n d cycloserine capsules were received from Eli Lilly. Indianapolis, IN. All other chemicals a n d solvents were the highest purity of commercially availablc materials. P r e p a r a t i o n of Stork Solutions-Stock solutions were prepared by dissolving weighed quantities of cycloserine powder in distilled water tl * 10 M I a n d 9-MA in acetonitrile ( 2 A 10 MI. The cyc!oserine solution was stored a t 5 'C and was stable for u p to 3 d. The 9 - M A stock solution was stable for up to one week a t room temperature ( 2 4 L 1 '('I. Assay ~'roceduri-Appropriate solutions of cycloserine ( 1.0 6 10 '-3.0 y l o - ' Mi were prepared by diluting the cycloserine stock solution with distilled water. Accurately pipetted quantities (5.0mI,) of each concentration were placed into 10-mL ground-glass stoppered volumetric flasks a n d 5.0 mI, of 9-hlA stock solution wag added The solutions were mixed well a n d heated i r i a shaking water bath a t 60 1 T for 60 rnin. Gpon cooling to room temperature and allowing to stand for a n additional 15 min a t room temperaturv, distilled water was added to volume. Each solution was then transferred into a cuvette a n d absorbance was measured a t 438 rim. A blank sample was run concurrently Assay of a Solid Douage Form-The powdered contents of 10 cycloserine capsules were mixed well and a quantity equivalent to -20 m g of cycloserine was accurately weighed and dissolved in distilled water with t h e aid of sonification for 15 min T h e solution was filtered into a 100-mL volumetric flask with the aid of distilled water washings, and water was added to volume. Five milliliters of t h a t solution was pipetted into a 100-mL volumetric flask and diluted to volume with distilled water. Five milliliters of t h e diluted cycloserine solution was then transferred to B 10-mL volumetric flask a n d assayed as described above under Assay Procedure. T h e concentration of cycloserine was calculated using slope and intercept constants derived from linear regression analysis of t h e calibration curve d a t a [concentration iabsorbance intercept )'slopeI. Identification of t h e Reaction I ' r o d u c t s T h i n - l a y e r chromatography (Silica gel GF,,,; Analtech; solvent system:chloroform:methanol:hexane, 80:15:5) of t h e analytical reaction mixture showed three bands, under long wavelength U V light, a t R,. 0.03, 0.5, and 0.9. T h e band a t R,0.03 appeared to be t h e reaction product of 9-MA +
c
0022-3549 88'0500-0452$07 00 0 1968. American Pharmaceutcal Association
and cycloserine. It was synthesized and the structure confirmed by UV, NMR, IR, elemental analysis, and TLC data (see below). Preparative TLC of the band at R f 0.5 was performed. After plate development, the band was scraped from the plate and the silica extracted with 1:l acetonitri1e:methanol. The pale yellow needles obtained were shown to be identical to 9-acridone by mp, NMR, and IR data. The band at R f 0.9 was identified as 9-MA based on comparison of R , values with an authentic sample. Synthesis of the reaction product of 9-MA and cycloserine was achieved by mixing equimolar quantities (1 x lo-' M)of cycloserine in distilled water and 9-MA in acetonitrile in a 100-mL roundbottomed flask and heating under reflw with stirring for 60 min. Cold distilled water was added and a yellow precipitate was collected. Recrystallization from acetonitri1e:water(1:U gave 94d-4-imino3-isoxazolidinone) acridine (4) as bright, yellow crystals, mp 221222 "C:UV Am=: 267 (a = 64,800), 414,438 nm; IR (KBr):3240 (NH), 1645 (C=O), and 1600-1500 cm-' (hydroxamate anion); 'H NMR (DMSO-d6):4.15-4.73 (m, 2H), 5.23 (t, lH), 6.83-7.20 (m, 2H), 7.227.70 (m,4H), and 8.03-8.44 ppm (m,2H). And-Calcd. for Cl6HI3N3o2:C, 68.74; H, 4.65; N, 15.04. Found: C, 68.72; H, 4.75, N, 15.04.
Results and Discussion The color intensity of the analytical reaction mixture was found to be pH dependent. Cycloserine in neutral aqueous solution (pH 5.5-6.5) occurs in a keto-enol equilibrium (1 and 2) which upon standing dimerizes to 2,5-bis(aminoxylmethyl)-3,6-diketopiperazine(3; Scheme I).11However, it is a reversible reaction given time and reaction conditions.12 At pH values <5, cycloserine is protonated and easily converts to the dimer, thus giving little or no measurable interaction with 9-MA. Thus, the analytical interaction between cycloserine and 9-MA is stability indicating with respect to any dimer that might be present in a degraded cycloserine sample. In addition, small quantities of 9-MA are easily converted to 9-acridone at pH <5 (Scheme 11). A t pH values >8, the reaction rate for cycloserine and 9-MA was very slow, and it was difficult to measure absorbance because of the low color intensity of the product obtained. The stock solution of cycloserine in distilled water was stable for 3 d a t ambient temperature and could be used for preparation of standard curves before decreasing absorbance readings were noted for the analytical solutions. Some quantity of water is essential for the interaction since i t was shown that cycloserine does not react with 9-MA in absolute methanol, ethanol, or acetonitrile, even when cycloserine is in solution. It is presumed that hydroxide ion plays an important part in the formation of the reaction intermediate. Proton NMR data was supportive of the 4b structure of the
t 0
3
Scheme I
0
& H
9-Acridone
t
&I
H
a
b
4
Scheme II
reaction product (Scheme 11). The three methoxy protons associated with 9-MA were absent. There was no evidence of substitution on the acridine side rings. The two protons at the 5' position of the oxazolidinonering appeared as multiple peaks at u 4.15-4.73, which was 0.65 ppm lower field than the same protons in cycloserine. The single proton at the 4' oxazolidinone position appeared as a clear triplet. These data supported a Schiff base form for the product 4b. The IR spectrum of 4 lacked an overtone band a t 2200 cm-', indicative of the quaternary ammonium ion13 form of cycloserine. In addition, there was the absence of the ether band a t 1100 cm-l, and the band pattern of the resonance hydroxamate anion of cycloserine a t 1600-1500 cm-' was not changed. These data appeared to substantiate that the bonding site of cycloserine to 9-MA is the 9-position of the acridine ring. The UV pattern of 4 was essentially identical to 9-acridone, except that all wavelength maxima were red shifted by 13,32, and 38 nm, respectively (see Figure 3). The similarity in the absorption spectrum also hinted strongly that the product was in a Schiff base form (4b). Fortunately, the absorbance bands of small quantities of 9-acridone formed during the analytical determination are a t h,, 380 and 400 nm and do not interfere with measurement of the absorbance intensity of the analytical solutions. The interaction between cycloserine and 9-MA was effected by the quantity of acridine, time of heating, and heating temperature. Absorbance intensities were measured at a fixed concentration of cycloserine with various concentrations of 9-MA, up to a 70-fold molar excess. As shown in Figure 1, maximum absorbance was obtained for the analytical solutions when a 50-fold or greater molar ratio of 9-MA was used. The large excess of 9-MA needed to maximize the absorbance intensity is probably based on some degradation of 9-MA to 9-acridone, as shown in Scheme 11. Analytical solutions were heated for various time periods at 60 "C. The results shown in Figure 2 indicate that maximum absorbance was obtained after heating the solutions for 50 min. A water bath temperature of 60 "C was selected for the analytical procedure, taking into consideration the maximum reaction time and boiling points of the respective solvents used in the assay procedure. 9-Methoxyacridine interacts with cycloserine in a water: acetonitrile system to yield a highly colored solution. The spectrum of the reaction mixture showed two absorption bands in the visible region at 414 and 438 nm which Journal of Pharmaceutical Sciences / 453 Vol. 77, No. 5, May 1988
10; W
u
0.8
1
..
WAVELENGTH (nm)
L! '0
2
- -2 -
i
10
20
-L
30
- 2 - - L_
40
50
9-METHOXYACRIDINE, x
_
-
I 2
60
70
M
Figure 1-Effect of 9-methoxyacridine concentration on color development of 70- M cycloserine at 60 1 "Cfor 50 min. Absorbance was measured at 438 nm. _f
1.1
I
w ogt
0
= I a
E 08t $ 1
5
071
L-
06j.II '0
10
20 TIME (minutes) 30 40
Table I-Analysls of Spiked Cycloserlne Samples by the 9-Methoxyacridine Method
Final Cycloserine concentration,
Concentration Found,
M x 10-5
M x 10-58
1.50 3.00 5.00 8.00 10.00 15.00
1.49 ? 0.002 3.02 t 0.008 5.04 ? 0.006 7.972 0.006 9.90 f 0.009 15.06 ? 0.010
a Mean t
50
$0
Flgure 2- Effect of heating time on color development of 1.2 x 10- M cycloserine with 10 M 9-methoxyacridine at 60 f 1 "C.Absorbance was measured at 438 nm.
corresponds to that of authentic 4 (see Figure 3). The absorbance of the analytical solution was measured at 438 nm in order to minimize the blank effect. A calibration curve for cycloserine was prepared in the 1.0 x 10-5-2.5 x M range. Beer's law was shown to be obeyed over the concentraM. The minimum to 3.0 x tion range from 5.0 x detectable quantity of cycloserine measured by this procedure was 0.3 pglmL (signa1:noise = 2). Linear regression analysis of absorbance versus drug concentrations gave a slope of 0.918, an intercept of 0.006, and a correlation coefficient, r, of 0.9999 (n = 6). The regression constants were applied to the analysis of spiked samples of cycloserine as "unknowns". The data revealed that the method provides good recovery and precision data (Table I). Under the analysis conditions utilized, a solution of the colored compound 4 was stable for at least 3 d at ambient temperature after color development. The analytical solutions were also shown to be stable in acidic medium in contrast with a solution of cycloserine. The mean percent recovery of cycloserine obtained by applying the assay procedure to a commercial capsule dosage form containing cycloserine (250 mg) was shown to be 101.84 2 0.48% (n = 6).The same lot of capsules assayed by the British Pharmacopeia aqueous titrimetry method3 gave a mean percent recovery of 100.86& 1.43% (n = 4).These data indicated that the spectrophotometric assay method described herein can be successfully applied to the determina454 / Journal of Pharmaceutical Sciences Vol. 77, No. 5, May 1988
Figure 3-Absorption spectra of 3 x 10- M 9-methoxyacridine ( ) , 4.8 x M 9-acridone (------), and 5 x M 9-(d-4-imino-3isoxazolidinone) acridine )-( .
Recovery, %
99.3 100.7 100.8 99.6 99.0 100.4
RSD,% 0.13 0.26 0.12 0.04 0.04 0.07
SD based on triplicate determinations of each sample.
tion of cycloserine in capsules with good recovery and precision. Experiments performed in this laboratory have indicated that none of the commonly employed capsule excipients, such as lactose, interfere with the assay since they do not interact with 9-MA to give colored derivatives. We have found that 9-MA is a very specific chromogenic reagent and only readily reacts with primary aliphatic amino functionalities in an organic structure. The advantage of the 9-MA method versus the USP XXI microbial assay lies mainly in the ability of the 9-MA procedure as a chemical method to show evidence of the presence of intact drug. The 9-MA method was not applied to the analysis of cycloserine levels in blood ( ~ 3 0pg/mL) during these studies, but there is enough specificity and sensitivity inherent in this spectrophotometric method to allow it to be useful as a clinical assay procedure. In summation, spectrophotometric measurements with 9-MA provided a relatively simple and rapid means of assay of cycloserine and its capsule dosage form.
References and Notes 1. Modern Drug Encyclopedia, Yorke Medical Books; Dun-Donnelley: New York, 1979;p 238. 2. US.Pharmacopeia, 21st rev.; U.S. Pharmacopeial Convention: Rockville, MD,1985;p 265. 3. British Phurmacopeia, 1980; Her Majesty's Stationary Office: London, England 1980;Vol. I, 134. 4. British Pharnacopeia, 1980; &er Majesty's Sbationary Office: London, England 1980;Vol. 11, p 528. 5. Jones, L.J. Anal. Chem. 1956,28,39-41. 6. Sondack, D.L.;Gainer, F. E.; Wesselman, H. J. J . Pharm. Sci. 1972,62,1344-1346. 7. El-Sayed, L.; Mohamed, Z. H.; Wahbi, A.M. Analyst 1986,I l l , 915-917. 8. Wahbi, A. M.; Mohamed, M. E.; Abounassif, M.J.; Gad-Kariem, E. Anal. Letter 1985,18(B3),261-267. 9. Stewart, J. T.;Parks, E. H. Znt. J . Phurm. 1983,17,161-166. 10. Barber, H.J.; Wilkinson, J. H.; Edwards, W. G. H. J . SOC. Chem. Zndust. 1947,66,411-415. 11. Hidy, P. H.; Hodge, E. B.; Young, R. L.; Harned, G. A. J . Am. Chem. Soc. 1955,77,2345. 12. Lassen, F. 0.; Stammer, C. H. J . Org. Chem. 1971,36, 26312634. 13. Florey, K. Analytical Profiles of Drug Substances; Academic: New York, 1972;Vol. 1, p 55.
is described The highly colored substituted acridine product was identified as 9-(d-4 imino-3-isoxazolidinone)acridine Color development was affected by time and temperature of heating and by the quantity of 9-methoxyacridine reagent utilized The absorbance at 438 nm is linearly proportional to concentrations of cycloserine with a detection limit of 0 3 pg mL The optimum range for the assay of cycloserine was from 5 0 10 to 3 0 x 10 M (correlationcoefficient 09999, n = 6) When applied to cycloserine capsules labeled to contain 250 mg. t h e proposed method gave mean recoveries of 101 84 0 48% The procedure is sufficiently sensitive, precise, and accurate for the determination of cycloserine in its dosage form A
~
+
Cycloserine I ~d)-4-amino-3-isoxazolidinone] is a n antimicrobial substance producod by t h e growth of certain s t r a i n s of Struptomvc.e.s orchzdacrus or S. garyphalus, a n d is administered for t h e t r e a t m e n t of pulmonary tuberculosis and certain genitourinary infections, especially Rnterohacturia and Esrhrmchin coli.1 Cycloserine is characterized by an ultraviolet absorption band peak a t 226 n m in water. Because it h a s relatively low intensity ( e 3940),it is difficult to assay for t h e d r u g in dosage forms by UV spectrophotometry, even if it is separated from matrix materials. The USP method2 for t h e determination of cycloserine d r u g substance a n d in capsules is based on a microbial assay. In t h e British Pharmacopeia, t h e assay methods used a r e titrimetric:' and spectrophotometricJ procedures. The titrimetric method i performed in aqueous propanol and uses thymolphthale as indicator and sodium hydroxide as t i t r a n t . The spectrophotometric assay is based on t h e formation of a color complex with sodium nitroprusside in a n acetate buffer measured a t 6% n m . I n this method, t h e chromogenic agent is unstable and so must be used immediately and discarded after one set of determinations." The reacted colored complex cannot be diluted and t h u s , if samples w e too d a r k to read, t h e assay m u s t be repeated using small aliquots. Also, t h e cyc1oserine:nitritopentacyanoferroate complex is found to be sensitive to temperature changes. Other published quantitative determinations for cycloserine include GLC," spectrophotometric,' a n d spectrofluorometric methods.' These assay methods involve many procedural steps, including control of pH. a n d t h e need for a high reaction temperature. 9-Methoxyacridine ( % M A ) was developed as a chromogenic reagent in t h i s laboratory as a n alternative to B-chloroacridine!' becausc of stability problems associated with 9-chloroacridine chemistry. This paper describes a spectrophotometric determination of cycloserine and i t s dosage form based on derivative formation with 9-methoxyacridine. A discussion of t h e reaction mechanism is also included. 452 Journal of Pharmaceutical Soences Vol 77, N o . 5, May 1988
Experimental Section A p p a r a t u s - T h e spectrophotometer system consisted of a PerkinElmer model 7500 professional computer. a Beckman model DU-7 spectrophotometer. a Hewlett-Packard model 7470A plotter, and 10m m quartz cells. An American Optical shaking water bath (model 0 2 1 5 6 ~a n d Crest ultrasonics b a t h were used in t h e analytical procedure. All melting points were uncorrected and recorded on a Thomas Hoover capillary melting point a p p a r a t u s . Infrared spectra were run on a Perkin-Elmer model 684 spectrophotometer The ' H S M R spectra were recorded on a Jeol FX 90Q E T NMN spectrometer with tetramethylsilane used a s a n internal standard. Elemental analyses were performed by Atlantic Microlab, Inc., Atlanta, GA. R e a g e n t s a n d C h e m i c a l ~ 9 - . M e t h o x y a c r i d i n e(9-MA) was synthesized from 9-chloroacridine ( E a s t m a n Kodak Company) and sodium methoxide, using t h e method of Barber e t a1 ( m p 103 and recrystallized from petroleum e t h e r (30-60 'C I. Cycloserine reference standard was obtained from t h e L'nited States Pharmacopeia a n d cycloserine capsules were received from Eli Lilly. Indianapolis, IN. All other chemicals a n d solvents were the highest purity of commercially availablc materials. P r e p a r a t i o n of Stork Solutions-Stock solutions were prepared by dissolving weighed quantities of cycloserine powder in distilled water tl * 10 M I a n d 9-MA in acetonitrile ( 2 A 10 MI. The cyc!oserine solution was stored a t 5 'C and was stable for u p to 3 d. The 9 - M A stock solution was stable for up to one week a t room temperature ( 2 4 L 1 '('I. Assay ~'roceduri-Appropriate solutions of cycloserine ( 1.0 6 10 '-3.0 y l o - ' Mi were prepared by diluting the cycloserine stock solution with distilled water. Accurately pipetted quantities (5.0mI,) of each concentration were placed into 10-mL ground-glass stoppered volumetric flasks a n d 5.0 mI, of 9-hlA stock solution wag added The solutions were mixed well a n d heated i r i a shaking water bath a t 60 1 T for 60 rnin. Gpon cooling to room temperature and allowing to stand for a n additional 15 min a t room temperaturv, distilled water was added to volume. Each solution was then transferred into a cuvette a n d absorbance was measured a t 438 rim. A blank sample was run concurrently Assay of a Solid Douage Form-The powdered contents of 10 cycloserine capsules were mixed well and a quantity equivalent to -20 m g of cycloserine was accurately weighed and dissolved in distilled water with t h e aid of sonification for 15 min T h e solution was filtered into a 100-mL volumetric flask with the aid of distilled water washings, and water was added to volume. Five milliliters of t h a t solution was pipetted into a 100-mL volumetric flask and diluted to volume with distilled water. Five milliliters of t h e diluted cycloserine solution was then transferred to B 10-mL volumetric flask a n d assayed as described above under Assay Procedure. T h e concentration of cycloserine was calculated using slope and intercept constants derived from linear regression analysis of t h e calibration curve d a t a [concentration iabsorbance intercept )'slopeI. Identification of t h e Reaction I ' r o d u c t s T h i n - l a y e r chromatography (Silica gel GF,,,; Analtech; solvent system:chloroform:methanol:hexane, 80:15:5) of t h e analytical reaction mixture showed three bands, under long wavelength U V light, a t R,. 0.03, 0.5, and 0.9. T h e band a t R,0.03 appeared to be t h e reaction product of 9-MA +
c
0022-3549 88'0500-0452$07 00 0 1968. American Pharmaceutcal Association
and cycloserine. It was synthesized and the structure confirmed by UV, NMR, IR, elemental analysis, and TLC data (see below). Preparative TLC of the band at R f 0.5 was performed. After plate development, the band was scraped from the plate and the silica extracted with 1:l acetonitri1e:methanol. The pale yellow needles obtained were shown to be identical to 9-acridone by mp, NMR, and IR data. The band at R f 0.9 was identified as 9-MA based on comparison of R , values with an authentic sample. Synthesis of the reaction product of 9-MA and cycloserine was achieved by mixing equimolar quantities (1 x lo-' M)of cycloserine in distilled water and 9-MA in acetonitrile in a 100-mL roundbottomed flask and heating under reflw with stirring for 60 min. Cold distilled water was added and a yellow precipitate was collected. Recrystallization from acetonitri1e:water(1:U gave 94d-4-imino3-isoxazolidinone) acridine (4) as bright, yellow crystals, mp 221222 "C:UV Am=: 267 (a = 64,800), 414,438 nm; IR (KBr):3240 (NH), 1645 (C=O), and 1600-1500 cm-' (hydroxamate anion); 'H NMR (DMSO-d6):4.15-4.73 (m, 2H), 5.23 (t, lH), 6.83-7.20 (m, 2H), 7.227.70 (m,4H), and 8.03-8.44 ppm (m,2H). And-Calcd. for Cl6HI3N3o2:C, 68.74; H, 4.65; N, 15.04. Found: C, 68.72; H, 4.75, N, 15.04.
Results and Discussion The color intensity of the analytical reaction mixture was found to be pH dependent. Cycloserine in neutral aqueous solution (pH 5.5-6.5) occurs in a keto-enol equilibrium (1 and 2) which upon standing dimerizes to 2,5-bis(aminoxylmethyl)-3,6-diketopiperazine(3; Scheme I).11However, it is a reversible reaction given time and reaction conditions.12 At pH values <5, cycloserine is protonated and easily converts to the dimer, thus giving little or no measurable interaction with 9-MA. Thus, the analytical interaction between cycloserine and 9-MA is stability indicating with respect to any dimer that might be present in a degraded cycloserine sample. In addition, small quantities of 9-MA are easily converted to 9-acridone at pH <5 (Scheme 11). A t pH values >8, the reaction rate for cycloserine and 9-MA was very slow, and it was difficult to measure absorbance because of the low color intensity of the product obtained. The stock solution of cycloserine in distilled water was stable for 3 d a t ambient temperature and could be used for preparation of standard curves before decreasing absorbance readings were noted for the analytical solutions. Some quantity of water is essential for the interaction since i t was shown that cycloserine does not react with 9-MA in absolute methanol, ethanol, or acetonitrile, even when cycloserine is in solution. It is presumed that hydroxide ion plays an important part in the formation of the reaction intermediate. Proton NMR data was supportive of the 4b structure of the
t 0
3
Scheme I
0
& H
9-Acridone
t
&I
H
a
b
4
Scheme II
reaction product (Scheme 11). The three methoxy protons associated with 9-MA were absent. There was no evidence of substitution on the acridine side rings. The two protons at the 5' position of the oxazolidinonering appeared as multiple peaks at u 4.15-4.73, which was 0.65 ppm lower field than the same protons in cycloserine. The single proton at the 4' oxazolidinone position appeared as a clear triplet. These data supported a Schiff base form for the product 4b. The IR spectrum of 4 lacked an overtone band a t 2200 cm-', indicative of the quaternary ammonium ion13 form of cycloserine. In addition, there was the absence of the ether band a t 1100 cm-l, and the band pattern of the resonance hydroxamate anion of cycloserine a t 1600-1500 cm-' was not changed. These data appeared to substantiate that the bonding site of cycloserine to 9-MA is the 9-position of the acridine ring. The UV pattern of 4 was essentially identical to 9-acridone, except that all wavelength maxima were red shifted by 13,32, and 38 nm, respectively (see Figure 3). The similarity in the absorption spectrum also hinted strongly that the product was in a Schiff base form (4b). Fortunately, the absorbance bands of small quantities of 9-acridone formed during the analytical determination are a t h,, 380 and 400 nm and do not interfere with measurement of the absorbance intensity of the analytical solutions. The interaction between cycloserine and 9-MA was effected by the quantity of acridine, time of heating, and heating temperature. Absorbance intensities were measured at a fixed concentration of cycloserine with various concentrations of 9-MA, up to a 70-fold molar excess. As shown in Figure 1, maximum absorbance was obtained for the analytical solutions when a 50-fold or greater molar ratio of 9-MA was used. The large excess of 9-MA needed to maximize the absorbance intensity is probably based on some degradation of 9-MA to 9-acridone, as shown in Scheme 11. Analytical solutions were heated for various time periods at 60 "C. The results shown in Figure 2 indicate that maximum absorbance was obtained after heating the solutions for 50 min. A water bath temperature of 60 "C was selected for the analytical procedure, taking into consideration the maximum reaction time and boiling points of the respective solvents used in the assay procedure. 9-Methoxyacridine interacts with cycloserine in a water: acetonitrile system to yield a highly colored solution. The spectrum of the reaction mixture showed two absorption bands in the visible region at 414 and 438 nm which Journal of Pharmaceutical Sciences / 453 Vol. 77, No. 5, May 1988
10; W
u
0.8
1
..
WAVELENGTH (nm)
L! '0
2
- -2 -
i
10
20
-L
30
- 2 - - L_
40
50
9-METHOXYACRIDINE, x
_
-
I 2
60
70
M
Figure 1-Effect of 9-methoxyacridine concentration on color development of 70- M cycloserine at 60 1 "Cfor 50 min. Absorbance was measured at 438 nm. _f
1.1
I
w ogt
0
= I a
E 08t $ 1
5
071
L-
06j.II '0
10
20 TIME (minutes) 30 40
Table I-Analysls of Spiked Cycloserlne Samples by the 9-Methoxyacridine Method
Final Cycloserine concentration,
Concentration Found,
M x 10-5
M x 10-58
1.50 3.00 5.00 8.00 10.00 15.00
1.49 ? 0.002 3.02 t 0.008 5.04 ? 0.006 7.972 0.006 9.90 f 0.009 15.06 ? 0.010
a Mean t
50
$0
Flgure 2- Effect of heating time on color development of 1.2 x 10- M cycloserine with 10 M 9-methoxyacridine at 60 f 1 "C.Absorbance was measured at 438 nm.
corresponds to that of authentic 4 (see Figure 3). The absorbance of the analytical solution was measured at 438 nm in order to minimize the blank effect. A calibration curve for cycloserine was prepared in the 1.0 x 10-5-2.5 x M range. Beer's law was shown to be obeyed over the concentraM. The minimum to 3.0 x tion range from 5.0 x detectable quantity of cycloserine measured by this procedure was 0.3 pglmL (signa1:noise = 2). Linear regression analysis of absorbance versus drug concentrations gave a slope of 0.918, an intercept of 0.006, and a correlation coefficient, r, of 0.9999 (n = 6). The regression constants were applied to the analysis of spiked samples of cycloserine as "unknowns". The data revealed that the method provides good recovery and precision data (Table I). Under the analysis conditions utilized, a solution of the colored compound 4 was stable for at least 3 d at ambient temperature after color development. The analytical solutions were also shown to be stable in acidic medium in contrast with a solution of cycloserine. The mean percent recovery of cycloserine obtained by applying the assay procedure to a commercial capsule dosage form containing cycloserine (250 mg) was shown to be 101.84 2 0.48% (n = 6).The same lot of capsules assayed by the British Pharmacopeia aqueous titrimetry method3 gave a mean percent recovery of 100.86& 1.43% (n = 4).These data indicated that the spectrophotometric assay method described herein can be successfully applied to the determina454 / Journal of Pharmaceutical Sciences Vol. 77, No. 5, May 1988
Figure 3-Absorption spectra of 3 x 10- M 9-methoxyacridine ( ) , 4.8 x M 9-acridone (------), and 5 x M 9-(d-4-imino-3isoxazolidinone) acridine )-( .
Recovery, %
99.3 100.7 100.8 99.6 99.0 100.4
RSD,% 0.13 0.26 0.12 0.04 0.04 0.07
SD based on triplicate determinations of each sample.
tion of cycloserine in capsules with good recovery and precision. Experiments performed in this laboratory have indicated that none of the commonly employed capsule excipients, such as lactose, interfere with the assay since they do not interact with 9-MA to give colored derivatives. We have found that 9-MA is a very specific chromogenic reagent and only readily reacts with primary aliphatic amino functionalities in an organic structure. The advantage of the 9-MA method versus the USP XXI microbial assay lies mainly in the ability of the 9-MA procedure as a chemical method to show evidence of the presence of intact drug. The 9-MA method was not applied to the analysis of cycloserine levels in blood ( ~ 3 0pg/mL) during these studies, but there is enough specificity and sensitivity inherent in this spectrophotometric method to allow it to be useful as a clinical assay procedure. In summation, spectrophotometric measurements with 9-MA provided a relatively simple and rapid means of assay of cycloserine and its capsule dosage form.
References and Notes 1. Modern Drug Encyclopedia, Yorke Medical Books; Dun-Donnelley: New York, 1979;p 238. 2. US.Pharmacopeia, 21st rev.; U.S. Pharmacopeial Convention: Rockville, MD,1985;p 265. 3. British Phurmacopeia, 1980; Her Majesty's Stationary Office: London, England 1980;Vol. I, 134. 4. British Pharnacopeia, 1980; &er Majesty's Sbationary Office: London, England 1980;Vol. 11, p 528. 5. Jones, L.J. Anal. Chem. 1956,28,39-41. 6. Sondack, D.L.;Gainer, F. E.; Wesselman, H. J. J . Pharm. Sci. 1972,62,1344-1346. 7. El-Sayed, L.; Mohamed, Z. H.; Wahbi, A.M. Analyst 1986,I l l , 915-917. 8. Wahbi, A. M.; Mohamed, M. E.; Abounassif, M.J.; Gad-Kariem, E. Anal. Letter 1985,18(B3),261-267. 9. Stewart, J. T.;Parks, E. H. Znt. J . Phurm. 1983,17,161-166. 10. Barber, H.J.; Wilkinson, J. H.; Edwards, W. G. H. J . SOC. Chem. Zndust. 1947,66,411-415. 11. Hidy, P. H.; Hodge, E. B.; Young, R. L.; Harned, G. A. J . Am. Chem. Soc. 1955,77,2345. 12. Lassen, F. 0.; Stammer, C. H. J . Org. Chem. 1971,36, 26312634. 13. Florey, K. Analytical Profiles of Drug Substances; Academic: New York, 1972;Vol. 1, p 55.