Evaluation of Pyrimido[ 5,4-d ]pyrimidine Derivatives as Peroxyoxalate Chemiluminescence Reagents Using a Flow Injection System Kenichiro Nakashima,? Kouichi Maki and Shuzo Akiyama Faculty of Pharmaceutical Sciences, Nagasaki University, 1-14, Bunkyo-Machi, Nagasaki 852, Japan
Kazuhiro Imai Branch Hospital Pharmacy, Tokyo University, 3-28-6, Mejirodai Bunkyo-Ku, Tokyo 112, Japan
Eighteen kinds of pyrimido[5,4-d ]pyrimidines together with several commercially available fluorescent compounds such as perylene, Rhodamine B, etc., were evaluated as the reagents for a peroxyoxalate chemiluminescence (CL) detection system by using a flow injection method. The peroxyoxalate CL reaction employed consists of bis(2,4,6trichlorophenyl)oxalate, hydrogen peroxide, triethylamine, and a fluorophore. Under the conditions used, 2,6-bis[di(2-hydroxyethyl)amino]-4,8-dipiperidinopyrimido[5,4-d]pyrimidine (Dipyridamole) and 2,4,6,8-tetrathiomorpholinopyrimido[5,4-d ]pyrimidine (li) gave very intense chemiluminescence intensities which were larger than those of any other commercially available fluorescent compounds tested (e.g., 10 times larger than that of perylene).
INTRODUCTION The peroxyoxalate chemiluminescence (CL) detection system is known to be sensitive and useful for the determination of trace amounts of hydrogen peroxide (Scott et al., 1980; Zoonen et al., 1985; Beltz et al., 1987; Abdel-Latif and Guilbault, 1988) and fluorescent compounds (Honda et al., 1985; Imai et aL, 1987). Recently, the detection system has been applied to high-performance liquid chromatography (HPLC) and flow injection analysis (Mellbin and Smith, 1984; Koziol et al., 1984; Imai, 1986; Abbott et al., 1986; Mann and Grayeski, 1987; Nakashima et aL, 1989a). For the peroxyoxalate CL detection system, the selection of the reagents, such as oxalates and fluorophore, is very important to yield high CL intensity. Therefore, many kinds of aryl oxalates and fluorophore have been synthesized and their CL properties examined (Rauhut et al., 1967; Tseng et al., 1979; Dowd and Paul, 1984; Imai et al., 1986; Givens et al., 1989). Pyrimidopyrimidines having coronary vasodilatory and diuretic effects were first synthesized by Fischer et al. (1951). Among them, 2,6-bis[di-(2-hydroxyethyl)amino] - 4 , s - dipiperidinopyrimido [ 5 , 4 - d ] pyrimidine (Dipyridamole), a platelet inhibitor, produces an intense blue-green fluorescence (Schmid et al., 1979). Dipyridamole has shown the strongest peroxyoxalate CL intensity among the known fluorescent compounds examined (Imai et al., 1987). In a previous paper (Nakashinia et al., 1989b), the syntheses of several kinds of pyrimido[5,4dlpyrimihines and their ultraviolet absorption and fluorescence spectral properties were reported, and it was found that some of them gave strong fluorescence. For the sensitive determination of hydrogen peroxide using a peroxyoxalate CL system, a fluorescent reagent t Author to whom correspondence should be addressed.
having a high fluorescence intensity is preferable. Therefore, in this work, eighteen pyrimido[5,4-d ]pyrimidines together with several commercially available fluorescent compounds were examined to evaluate their properties as the reagents for the peroxyoxalate CL detection system in the flow injection (FI) method developed previously.
EXPERIMENTAL Reagents. Pyrimido[5,4-d ]pyrimidines were synthesized by the method reported previously (Nakashima et al., 1989b).
Dipyridamole was purchased from Sigma Chemicals (St. Louis, MO, USA). Bis(2,4,6-trichlorophenyl)oxalate(TCPO) and 30% hydrogen peroxide were obtained from Wako Pure Chemicals (Tokyo, Japan). Acetonitrile was HPLC grade (Wako) and filtered through a membrane filter (0.45 pm) before use. All other chemicals were of reagent grade. FI-CL apparatus. The FI-CL apparatus was equipped with three Shimadzu LCdA high performance liquid chromatographs (Shimadzu, Kyoto, Japan) connected to a system controller (Shimadzu SCL-6A), and using a flow rate of 1.0 mL/min. The sample injector (Rheodyne Model 7125, Cotati, CA, USA) with a 50pL loop and an AL-2200 luminomonitor (Atto, Tokyo, Japan) with a 60 p L flow cell were used. Recorder was Nippon Densi Kagaku Model U-228 (Tokyo, Japan). Stainless steel tubings with 0.5 mm diameter were used in all the flow lines. The lengths of the delay coils connected to the mixing tees were set at 1.0 and 0.1 m. Procedure for evaluation of fluorophore. A fluorescent compound in chloroform or methanol was injected into a flow line of 3 mM triethylamine in acetonitrile, which was combined with a flow line of premixed solution of 0.15 mM TCPO and 4.5 mM hydrogen peroxide in acetonitrile. The intensity of CL generated was measured by a luminomonitor. Standard curves were prepared using known concentrations of each fluorescent compound. From the slopes of the standard curves, CL intensity of each fluorescent compound was estimated.
CCC-O269-3879/90/0105-0107$1.50 @ Heyden & Son Limited, 1990
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 3. 1990
105
K. NAKASHIMA ET AL.
RESULTS AND DISCUSSION The chemical structures of the pyrimido[5,4dlpyrimidines are shown as compounds 1 and 2. As commercially available fluorescent compounds, 9aminoacridine hydrochloride (AA), sodium fluorescein (FL), Rhodamine B (RB), 9,10-bis(phenylethynyl)anthracene (PEA), 9,lO-diphenylanthracene (DPA), and perylene (PE) were used. For the selection of oxalate
2
1
DP
R anilino p-anisidino m-anisidino morpholino pyrrolidino piperidino hydroxyethylamino dihydroxyethylamino thiomorpholino diphenylamino thiazolidino benzylamino
R 2(a) 2(b) 2(c) 2(d) 2(e)
anilino p-anisidino m-anisidino morpholino pyrrolidino
and/or fluorophore in a peroxyoxalate CL detection system, a simple and rapid method could be useful. Recently, a convenient evaluation method of oxalates using an FI system has been developed (Nakashima et al., 1989~). The method also seemed to be applicable to the evaluation of fluorescent compounds. Therefore, it was applied to the evaluation of pyrimido[5,4d ]pyrimidines as peroxyoxalate CL reagents. In this
work, TCPO was used as an oxalate, since it is one of the most popular oxalates for the peroxyoxalate CL system. As described previously (Nakashima et al., 1989~1,the recommended concentrations of TCPO, H z 0 2 , and triethylamine were 0.15, 4.5, and 3.0 mM in acetonitrile, respectively, and the flow rate of each solution was set at 1.0 mL/min. The selection of solvent for the analytes was examined by using chloroform, dioxane, acetronitrile, N , N dimethylformamide, methanol, and ethyl acetate. Among these, acetonitrile, chloroform, ethyl acetate and methanol were found to be favorable, since these solvents showed only a weak background CL. In this work, chloroform or methanol was selected as solvent in view of the solubility of fluorescent compounds tested. The CL intensity of each fluorophore was measured according to the evaluation procedure described in the text. The conce;itrations of standard solutions of fluorophores were adjusted so that the amount injected ranged from 0.1 to 25pmol in chloroform or methanol. The recorder responses obtained by l(i) are shown in Fig. 1. Standard curves of these compounds were linear. For example, the linear regressions were Y = 27.75X + 0.10 ( I = 0.998) for Dipyridamole, Y = 30.69X - 0.77 ( r = 0.996) for l(i) and Y = 3.13X - 0.18 ( I = 0.995) for PE, respectively ( Y is the peak height (cm) and X is the concentration (pmol/injection)). The correlation coefficients of the others were > 0.980. The relative CL intensities of the fluorophores examined are given in Table 1. Under the conditions used, Dipyridamole and l(i) showed almost the same strong CL intensity, i.e., an order of magnitude larger than that of PE. l(d), l(e), l(f), l(k), l(h) and 2(e) also showed greater C1 intensity than that of PE. Among the commercially available fluorescent compounds, PE, PEA, and RB gave strong CL. As a result, it was found that pyrimido[5,4d ]pyrimidine having a heterocyclic ring as a substituent gave a larger CL. On the other hand, those having aromatic rings such as l(a) and l(c), which showed strong fluorescence (Nakashima et a l , 1989b), unexpectedly exhibited a lower CL intensity than l(i). The reasons for the weak CL emission are not clear. However, these results would suggest that a strong fluorescer may not always yield strong CL emission under these restricted conditions.
Table 1. Comparison of relative chemiluminescer ce intensities of fluorescent compounds Compound
DP
W) 1(4 l(4 1(e) 1( f ) l(g) l(h) l(i)
W) l(k) 1(1) Figure 1. Recorder response of chemiluminescence with l(i):TCPO, 0.05 mM; H20,, 1.6 mM; TEA, 1.O mM.
106 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 3, 1990
a
RCP
Compound
RCI'
Compound
RCI'
100.0
2(a) 2(b) 2(c) 2(d) 2(e)
0.2 cO.01
Kazuhiro Imai Branch Hospital Pharmacy, Tokyo University, 3-28-6, Mejirodai Bunkyo-Ku, Tokyo 112, Japan
Eighteen kinds of pyrimido[5,4-d ]pyrimidines together with several commercially available fluorescent compounds such as perylene, Rhodamine B, etc., were evaluated as the reagents for a peroxyoxalate chemiluminescence (CL) detection system by using a flow injection method. The peroxyoxalate CL reaction employed consists of bis(2,4,6trichlorophenyl)oxalate, hydrogen peroxide, triethylamine, and a fluorophore. Under the conditions used, 2,6-bis[di(2-hydroxyethyl)amino]-4,8-dipiperidinopyrimido[5,4-d]pyrimidine (Dipyridamole) and 2,4,6,8-tetrathiomorpholinopyrimido[5,4-d ]pyrimidine (li) gave very intense chemiluminescence intensities which were larger than those of any other commercially available fluorescent compounds tested (e.g., 10 times larger than that of perylene).
INTRODUCTION The peroxyoxalate chemiluminescence (CL) detection system is known to be sensitive and useful for the determination of trace amounts of hydrogen peroxide (Scott et al., 1980; Zoonen et al., 1985; Beltz et al., 1987; Abdel-Latif and Guilbault, 1988) and fluorescent compounds (Honda et al., 1985; Imai et aL, 1987). Recently, the detection system has been applied to high-performance liquid chromatography (HPLC) and flow injection analysis (Mellbin and Smith, 1984; Koziol et al., 1984; Imai, 1986; Abbott et al., 1986; Mann and Grayeski, 1987; Nakashima et aL, 1989a). For the peroxyoxalate CL detection system, the selection of the reagents, such as oxalates and fluorophore, is very important to yield high CL intensity. Therefore, many kinds of aryl oxalates and fluorophore have been synthesized and their CL properties examined (Rauhut et al., 1967; Tseng et al., 1979; Dowd and Paul, 1984; Imai et al., 1986; Givens et al., 1989). Pyrimidopyrimidines having coronary vasodilatory and diuretic effects were first synthesized by Fischer et al. (1951). Among them, 2,6-bis[di-(2-hydroxyethyl)amino] - 4 , s - dipiperidinopyrimido [ 5 , 4 - d ] pyrimidine (Dipyridamole), a platelet inhibitor, produces an intense blue-green fluorescence (Schmid et al., 1979). Dipyridamole has shown the strongest peroxyoxalate CL intensity among the known fluorescent compounds examined (Imai et al., 1987). In a previous paper (Nakashinia et al., 1989b), the syntheses of several kinds of pyrimido[5,4dlpyrimihines and their ultraviolet absorption and fluorescence spectral properties were reported, and it was found that some of them gave strong fluorescence. For the sensitive determination of hydrogen peroxide using a peroxyoxalate CL system, a fluorescent reagent t Author to whom correspondence should be addressed.
having a high fluorescence intensity is preferable. Therefore, in this work, eighteen pyrimido[5,4-d ]pyrimidines together with several commercially available fluorescent compounds were examined to evaluate their properties as the reagents for the peroxyoxalate CL detection system in the flow injection (FI) method developed previously.
EXPERIMENTAL Reagents. Pyrimido[5,4-d ]pyrimidines were synthesized by the method reported previously (Nakashima et al., 1989b).
Dipyridamole was purchased from Sigma Chemicals (St. Louis, MO, USA). Bis(2,4,6-trichlorophenyl)oxalate(TCPO) and 30% hydrogen peroxide were obtained from Wako Pure Chemicals (Tokyo, Japan). Acetonitrile was HPLC grade (Wako) and filtered through a membrane filter (0.45 pm) before use. All other chemicals were of reagent grade. FI-CL apparatus. The FI-CL apparatus was equipped with three Shimadzu LCdA high performance liquid chromatographs (Shimadzu, Kyoto, Japan) connected to a system controller (Shimadzu SCL-6A), and using a flow rate of 1.0 mL/min. The sample injector (Rheodyne Model 7125, Cotati, CA, USA) with a 50pL loop and an AL-2200 luminomonitor (Atto, Tokyo, Japan) with a 60 p L flow cell were used. Recorder was Nippon Densi Kagaku Model U-228 (Tokyo, Japan). Stainless steel tubings with 0.5 mm diameter were used in all the flow lines. The lengths of the delay coils connected to the mixing tees were set at 1.0 and 0.1 m. Procedure for evaluation of fluorophore. A fluorescent compound in chloroform or methanol was injected into a flow line of 3 mM triethylamine in acetonitrile, which was combined with a flow line of premixed solution of 0.15 mM TCPO and 4.5 mM hydrogen peroxide in acetonitrile. The intensity of CL generated was measured by a luminomonitor. Standard curves were prepared using known concentrations of each fluorescent compound. From the slopes of the standard curves, CL intensity of each fluorescent compound was estimated.
CCC-O269-3879/90/0105-0107$1.50 @ Heyden & Son Limited, 1990
BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 3. 1990
105
K. NAKASHIMA ET AL.
RESULTS AND DISCUSSION The chemical structures of the pyrimido[5,4dlpyrimidines are shown as compounds 1 and 2. As commercially available fluorescent compounds, 9aminoacridine hydrochloride (AA), sodium fluorescein (FL), Rhodamine B (RB), 9,10-bis(phenylethynyl)anthracene (PEA), 9,lO-diphenylanthracene (DPA), and perylene (PE) were used. For the selection of oxalate
2
1
DP
R anilino p-anisidino m-anisidino morpholino pyrrolidino piperidino hydroxyethylamino dihydroxyethylamino thiomorpholino diphenylamino thiazolidino benzylamino
R 2(a) 2(b) 2(c) 2(d) 2(e)
anilino p-anisidino m-anisidino morpholino pyrrolidino
and/or fluorophore in a peroxyoxalate CL detection system, a simple and rapid method could be useful. Recently, a convenient evaluation method of oxalates using an FI system has been developed (Nakashima et al., 1989~). The method also seemed to be applicable to the evaluation of fluorescent compounds. Therefore, it was applied to the evaluation of pyrimido[5,4d ]pyrimidines as peroxyoxalate CL reagents. In this
work, TCPO was used as an oxalate, since it is one of the most popular oxalates for the peroxyoxalate CL system. As described previously (Nakashima et al., 1989~1,the recommended concentrations of TCPO, H z 0 2 , and triethylamine were 0.15, 4.5, and 3.0 mM in acetonitrile, respectively, and the flow rate of each solution was set at 1.0 mL/min. The selection of solvent for the analytes was examined by using chloroform, dioxane, acetronitrile, N , N dimethylformamide, methanol, and ethyl acetate. Among these, acetonitrile, chloroform, ethyl acetate and methanol were found to be favorable, since these solvents showed only a weak background CL. In this work, chloroform or methanol was selected as solvent in view of the solubility of fluorescent compounds tested. The CL intensity of each fluorophore was measured according to the evaluation procedure described in the text. The conce;itrations of standard solutions of fluorophores were adjusted so that the amount injected ranged from 0.1 to 25pmol in chloroform or methanol. The recorder responses obtained by l(i) are shown in Fig. 1. Standard curves of these compounds were linear. For example, the linear regressions were Y = 27.75X + 0.10 ( I = 0.998) for Dipyridamole, Y = 30.69X - 0.77 ( r = 0.996) for l(i) and Y = 3.13X - 0.18 ( I = 0.995) for PE, respectively ( Y is the peak height (cm) and X is the concentration (pmol/injection)). The correlation coefficients of the others were > 0.980. The relative CL intensities of the fluorophores examined are given in Table 1. Under the conditions used, Dipyridamole and l(i) showed almost the same strong CL intensity, i.e., an order of magnitude larger than that of PE. l(d), l(e), l(f), l(k), l(h) and 2(e) also showed greater C1 intensity than that of PE. Among the commercially available fluorescent compounds, PE, PEA, and RB gave strong CL. As a result, it was found that pyrimido[5,4d ]pyrimidine having a heterocyclic ring as a substituent gave a larger CL. On the other hand, those having aromatic rings such as l(a) and l(c), which showed strong fluorescence (Nakashima et a l , 1989b), unexpectedly exhibited a lower CL intensity than l(i). The reasons for the weak CL emission are not clear. However, these results would suggest that a strong fluorescer may not always yield strong CL emission under these restricted conditions.
Table 1. Comparison of relative chemiluminescer ce intensities of fluorescent compounds Compound
DP
W) 1(4 l(4 1(e) 1( f ) l(g) l(h) l(i)
W) l(k) 1(1) Figure 1. Recorder response of chemiluminescence with l(i):TCPO, 0.05 mM; H20,, 1.6 mM; TEA, 1.O mM.
106 BIOMEDICAL CHROMATOGRAPHY, VOL. 4, NO. 3, 1990
a
RCP
Compound
RCI'
Compound
RCI'
100.0
2(a) 2(b) 2(c) 2(d) 2(e)
0.2 cO.01
