Li WG et al / Acta Pharmacol Sin 2002 Aug; 23 (8): 727-732

Antioxidative activity of spin labeled derivatives of podophyllic acid hydrazide

LI Wen-Guang¹, ZHANG Xiao-Yu² , WU Yong-Jie, TIAN Xuan³

Department of Pharmacology, ²Department of Physiology, Lanzhou Medical College;

³National Laboratory of Organic Chemistry, Lanzhou University, Lanzhou 730000, China

¹ Correspondence to Prof LI Wen-Guang. Phn 86-931-862-3573. E-mail wyj@lz.gs.cninfo.net

Received 2001-05-10 Accepted 2002-05-21

KEY WORDS podophyllic acid hydrazide; spin labels; doxorubicin; free radicals; antioxidants; malondialdehyde

ABSTRACT

AIM: To study the relationship between structure and antioxidation activity of spin labeled derivatives of podophyllic acid hydrazide (GP) in tissues and red blood cells (RBC) from rats. METHODS: The homogenate of liver, heart, and kidneys of rats was used to measure malondialdehyde (MDA) spontaneous generated and induced by hydroxyl free radical generation system (Fe²⁺-ascorbic acid, FRGS) or doxorubicin (DOX) by TBA colorimetric method. H₂O₂-caused hemolysis was determined spectrophotometrically. Superoxide anion from zymosan-stimulated neutrophils of rats was evaluated by NBT-reduction assay. RESULTS: GP1 and GP1OH obviously inhibited MDA formation either spontaneously or induced by FRGS and DOX and antagonized hemolysis induced by H₂O₂, but GP and GP1H showed less potent activity. GP1 also inhibited the formation of superoxide anion from activated neutrophils of rats. CONCLUSION: Introduction of nitroxyl radical moieties into GP generated potent derivatives with antioxidative activity. The essential antioxidation active groups of spin labeled derivative of GP are NO or NOH group in nitroxyl radical moieties.

INTRODUCTION

The nitroxides, with low molecular weight and less toxic stable free radicals, had been widely used as spin labels^[1,2], which also had been demonstrated to possess potent antioxidative action either in biological or nonbiological test systems^[3,4], and antitumor action. Introduction of nitroxyl radicals into antitumor agent was demonstrated to enhance its antitumor activity and reduce its toxicity^[5,6], and could also be used for the study of pharmacokinetics of bioactive spin labels by electron paramagnetic resonance (EPR) technique. Based on this fact, a series of spin labeled derivatives of podophyllic acid hydrazine (GP) had been synthesized and showed that introduction of nitroxyl radicals into GP could enhance the antitumor activity and lower the toxicity. However, it is unknown how the antitumor activity was enhanced and the toxicity was lowered by introduction of nitroxyl radicals into antitumor agent. The present study were designed to evaluate the antioxidative effect of spin labeled derivatives of GP with following chemical structure and analyze the relationship between their structure and activity.

MATERIALS AND METHODS

Podophyllic acid hydrazide (GP), podophyllic acid [4-(2,2,6,6-tetramethyl-piperidinooxyl)] hydrazone (GP1), podophyllic acid [4 -(2,2,6,6 -tetramethyl-1-hydroxy piperidine)] hydrazone (GP1OH), and podophyllic acid [4-(2,2,6,6-tetramethyl-1-piperidine)] hydrazone (GP1H) were semi-synthesized^[1,2] by National Laboratory of Organic Chemistry, Lanzhou University, China. The purity of them was approximately 98 %. They were dissolved into 5 % Me₂SO. Doxorubicin (DOX) was the product of Shenzhen Main Luck Pharmaceutical Inc. Zymosan A (Sigma) was opsonized with rat serum^[7] and suspended in phosphate buffer 0.15 mmol/L (pH 7.4). All the other reagents were of analytical grade.

Spin labeled derivatives of podophyllic acid hydrazide (GP)

Wistar rats (¡â¡á, 8 weeks-old) weighing 182 ± s 16 g were provided by Animal Center of Gansu Academy of Medical Sciences (Grade II, Certificate ¡í 14-004).

Determination of malondialdehyde The homogenate of liver, heart, and kidneys of rats was pre-pared^[4] and MDA was assayed by thiobarbituric acid (TBA) method^[8].

Hemolysis test Rat RBC was washed 3 times with normal saline and made into 0.5 % suspension. H₂O₂ (100 mmol/L) induced hemolysis was tested after 1 h incubation of RBC suspension at 37 ¡æ with tested drugs as previously^[4]. The absorbance (A) at 415 nm of control tubes was defined as 100 %. The hemolysis extent was calculated by referring to control tube.

Superoxide anion formation analysis Rat neutrophils from abdominal cavity were prepared^[4] and the reduced NBT product formazan by O^-₂ from neutrophils was assayed by spectrophotometry at 515 nm.

Statistics Data were presented as mean±SD. Statistical analysis was performed using unpaired t-test. The IC₅₀ and its 95 % confidence limits were calculated by liner regression analysis^[9].

RESULTS

Effect on MDA formation MDA was spontaneously formed in liver homogenate after 2 h incubation. GP and GP1H 160 mmol/L inhibited spontaneous MDA formation by 26.5 % and 34.8 % respectively. In contrast, GP1 1.25, 2.5, 5, 10, and 20 mmol/L inhibited MDA formation by 7.6 %, 27.7 %, 59.8 %, 71.5 %, and 74.8 %; and GP1OH 2.5, 5, 10, and 20 mmol/L inhibited MDA formation by 21.9 %, 42.1 %, 75.2 %, and 84.2 % respectively (Tab 1).

Tab 1. The effect of spin labeled derivatives of GP on malondialdehyde (MDA) formation induced by Fe²⁺-ascorbic acid. n=4. Mean±SD. ^aP>0.05, ^bP<0.05, ^cP<0.01 vs control. ^dP>0.05, ^fP<0.01 vs basic tubes.

Drugs/ mmol¡¤L-1	MDA/nmol¡¤g-1 tissues
	Induced by Fe2+-ascorbic acid			Spontane- ously in liver
	Heart	Liver	Kidneys
GP
Basic	73¡À6c	51¡À3c	90¡À4c	117¡À10
Control	352¡À10	290¡À16	279.5¡À2.8
320	246¡À11c	196¡À9c	274¡À14a	68¡À5f
160	273¡À4c	242¡À9c	277¡À4a	86.2¡À2.9f
80	306¡À6c	294¡À8a	281¡À8a	102¡À5f
40	341¡À6b	298¡À8a	279¡À3a	118¡À6d
GP1
Basic	82¡À4c	58¡À7c	108¡À5c	128¡À8
Control	369¡À12	272¡À8	233¡À6
20	26¡À4c	43¡À5c	78¡À6c	32¡À9f
10	248¡À9c	78¡À8c	93¡À9c	37¡À9f
5	324.6¡À1.4c	133¡À12c	118.7¡À2.1c	52¡À5f
2.5	336¡À5c	223¡À15b	161¡À8c	93¡À4f
1.25	357¡À3b	240¡À18c	221¡À14a	118¡À13d
GP1H
Basic	92¡À6c	94¡À7c	96.1¡À1.5c	126¡À7
Control	295¡À13	252¡À8	294¡À8
160	278¡À10b	230¡À8c	219¡À8c	82¡À14f
80	282¡À8a	242¡À15a	234¡À9c	104¡À15f
40	289¡À9a	254¡À11a	279¡À8c	120¡À11d
GP1OH
Basic	46¡À6c	52¡À4c	110.5¡À2.1c	127¡À15
Control	339¡À9	225¡À7	275¡À11
20	90¡À8c	44¡À4c	109¡À7c	20¡À5f
10	212¡À7c	94¡À10c	246¡À8c	32¡À7f
5	331¡À7a	203¡À10c	273¡À8a	74¡À11f
2.5	335¡À4a	227¡À7a	274¡À8a	99¡À13f

After stimulated by FRGS (hydroxyl free radical generation system, Fe²⁺-ascorbic acid 50/50 mmol·L^-1) for 30 min, homogenate of heart, liver, and kidney produced enormous amount of MDA (Tab 1,2). GP 160 mmol/L inhibited FRGS-induced MDA formation from heart and liver by 28.3 % and 16.6 % respectively, but failed to affect MDA formation from kidneys. GP1H 160 mmol/L inhibited MDA formation from heart, liver, and kidneys by 5.9 %, 8.5 %, and 25.4 %, respectively. GP1 1.25-20 ¦Ìmol/L inhibited MDA formation from heart, liver, and kidneys concentration-dependently with MIC 1.25, 1.25, and 2.5 ¦Ìmol/L, respectively. All the MDA formation induced by FRGS were inhibited by GP1 20 ¦Ìmol/L in heart, liver, and kidneys. The MIC for GP1OH to inhibit MDA formation from heart, liver, and kidneys were 10, 5, and 10 ¦Ìmol/L, respectively. GP1OH 20 ¦Ìmol/L inhibited MDA formation from heart, liver, and kidneys by 85.2 %, 101.0 %, and 104.8 % respectively (Tab 1). MDA formation were elevated by DOX in rat heart and liver homogenate and were inhibited by all tested drugs with the potential rank order of GP1, GP1OH, GP1H, and GP (Tab 2).

Tab 2. The effect of spin labeled derivatives of GP on malondialdehyde (MDA) formation induced by DOX. n = 4. Mean±SD. ^aP>0.05, ^bP<0.05,^cP<0.01 vs control.

Drugs/ mmol¡¤L-1	Heart MDA/ nmol¡¤g-1	IR/%	Liver MDA/ nmol¡¤g-1	IR/%
GP
Basic	112¡À14c		135¡À9c
Control	160¡À11		198¡À4
160	135¡À14b	52.5	165¡À15c	51.9
80	150¡À6a	20.8	176¡À6c	34.0
40	153¡À5a	14.8	184.7¡À2.4c	20.8
20	163¡À5a	0	188¡À4c	15.6
GP1
Basic	174¡À4c		161¡À12c
Control	236¡À11		249¡À6
10	167¡À11c	110.7
5	204¡À11c	52.2	84¡À6c	189.2
2.5	209¡À8c	43.8	142¡À22c	122.8
1.25	226¡À5a	16.8	204¡À11c	51.6
0.625			216¡À13c	38.0
GP1OH
Basic	198¡À8c		117¡À9c
Control	250¡À16		283¡À11
20	196¡À13c	102.5	104¡À7c	107.9
10	196¡À5c	103.7	131¡À15c	91.8
5	206¡À5c	85.0	223¡À8c	36.4
2.5	213¡À13c	70.1	254¡À19b	17.4
GP1H
Basic	178¡À16c		104¡À26c
Control	219¡À8		160¡À17
250	146¡À6c	177.2	115.3¡À2.3c	80.2
125	168¡À6c	123.0	126¡À13c	60.1
62.5	178.1¡À2.1c	99.0	151.7¡À2.6a	14.2
31.25	190¡À10c	71.0	158¡À6a	2.0
15.625	204¡À7c	37.4

Effect on hemolysis induced by H₂O₂ GP1 and GP1OH 80 and 160 ¦Ìmol/L inhibited hemolysis of rat RBC, but GP and GP1H showed no effect (Tab 3).

Tab 3. Effect of spin labeled derivatives of GP on the hemolysis of rat RBC stimulated by H₂O₂. n=5. Mean±SD. ^aP>0.05, ^cP<0.01 vs control.

Drugs/ mmol¡¤L-1	Hemolysis		Extent/%
	GP	GP1H	GP1OH	GP1
Basic	5.9¡À1.2c	11¡À4c	8¡À4c	8.2¡À1.7c
Control	100	100	100	100
160	101¡À6a	102¡À4a	24.5¡À1.2c	10¡À7c
80	99¡À6a	99¡À5a	79.3¡À2.0c	82¡À8c
40	101¡À6a	102¡À11a	102¡À4a	91¡À4a

Effect on superoxide anion formation from activated neutrophils The reduced NBT product (formazan) from neutrophils of rats was markedly increased after stimulated by zymosan. The specificity of assay for O^-₂ was demonstrated by the fact that SOD 150, 300 and 600 kU/L inhibited formazan formation by 28.6 %, 40.8 %, and 84.5 %, respectively. GP1 160 and 320 ¦Ìmol/L also slightly inhibited superoxide anion formation from activated neutrophils by 18.3 % and 26.6 % with statistical significance (P<0.01), however, GP, GP1H, and GP1OH all did not show any action (Tab 4).

Tab 4. Effect of spin labeled derivatives of GP on the release of superoxide anion from rat neutrophils stimulated by zymosan. n=4. Mean±SD. ^aP>0.05, ^cP<0.01 vs control.

Drugs/ mmol¡¤L-1	A515
	GP	GP1H	GP1OH	GP1
Basic	0.154¡À0.023c	0.136¡À0.013c	0.154¡À0.023c	0.145¡À0.029c
Control	0.497¡À0.028	0.358¡À0.018	0.497¡À0.028	0.353¡À0.009
320	0.365¡À0.011c	0.355¡À0.020a	0.46¡À0.03a	0.357¡À0.010a
160	0.41¡À0.03c	0.354¡À0.010a	0.475¡À0.013a	0.359¡À0.011a

A comparison of IC₅₀ values In both experiments of MDA test and anti-hemolysis, the rank order of IC₅₀ values was GP1>GP1OH>GP1H>GP (Tab 5).

Tab 5. A comparison of IC₅₀ values (95 % confidence limits, ¦Ìmol/L) of spin labeled derivatives of GP.

Drugs	MDA formation
	Spontaneously  formed in liver	Induced by Fe2+-AA			Induced by DOX		Anti-hemolysis
		liver	heart	kidneys	liver	heart
GP	>160	>160	>160	NO	170.4 (65.2-445.5)	185.2 (107.2-319.8)	NO
GP1	5.4 (3.6-7.0)	3.8 (2.4-5.2)	7.5 (4.4-10.5)	2.4 (0.7-4.2)	0.9 (0.4-2.0)	3.1 (2.7-3.6)	88.9 (87.6-90.2)
GP1OH	5.9 (4.3-7.5)	7.3 (5.5-9.1)	10.5 (8.1-12.9)	10.9 (5.8-16.0)	5.3 (3.0-9.5)	2.0 (0.9-4.3)	103.0 (101.9-104.3)
GP1H	>160	>160	>160	>160	115.6 (67.3-198.7)	21.3 (14.3-31.7)	NO

AA: ascorbic acid. NO: no effect.

DISCUSSION

Fe²⁺-ascorbic acid system produces hydroxyl radicals according to Fenton reaction, and the latter caused lipoperoxidation and damage of tissues accompanying MDA formation^[10]. Therefore, MDA is a convenient index for indirectly detecting hydroxyl radicals. The spin labeled derivatives of GP were capable of inhibiting FRGS -induced MDA formation, and antagonizing H₂O₂- caused hemolysis, indicating that these drugs are scavengers against hydroxyl radicals.

DOX plays an important role in cancer chemo-therapy. But its clinical use has been limited by its irreversible cardiotoxicity. It has been believed that the cardiotoxicity of DOX is caused by free radicals, which has little relation to its anticancer effects^[11,12]. The semiquinone free radical formed from DOX in rat heart homogenate was demonstrated^[13]. And it may further transfer electron to oxygen or H₂O₂ to produce O₂^- or •OH and induce MDA formation. GP, GP1, GP1OH, and GP1H all concentration-dependently inhibited the MDA escalation caused by DOX from heart and liver homogenate of rat, among which GP1 was shown to be the most potent.

Our previous experiment demonstrated that nitroxides 4-oxy-2,2,6,6 -tetramethylpiperidinooxyl (4-O-TEMPO) and 4-oxy-2,2,6,6-tetramethyl-1-hydroxy piperidine (4-O-TEMPOH) inhibited FRGS-induced formation of MDA in the same test system as present one^[4]. However, they were relatively weaker than GP1OH and especially GP1 (Tab 4). Introduction of 4-O-TEMPO into GP produces GP1 and 4-O-TEMPOH into GP produces GP1OH ^[5,6]. In the experiment of FRGS-induced MDA formation in liver, heart, and kidneys of rats, the ratio of IC₅₀ values (¦Ìmol/L) for GP1/4-O-TEMPO were 3.8/22.2, 7.5/38.5, and 2.4/18.8 respectively, and for GP1OH/4-O-TEMPOH were 7.3/8.3,10.5/28.6, and 10.9/47.8, respectively, exhibiting that introduction of 4-O-TEMPO into GP1 or 4-O-TEMPOH into GP1OH greatly strengthened the antioxidative activity. However, GP1 was shown to be the most potent. Although the parent GP only had slightly inhibitory action against lipoperoxidation in FRGS system, its derivative GP1 was more potent than 4-O-TEMPO, also demonstrating that GP is capable of increasing the activity of nitroxides against oxidation. On the other hand, if the NO group of GP1 or NOH group of GP1OH was substituted by NH group, both GP1 and GP1OH were changed into GP1H almost without activity at concentration as high as 80 ¦Ìmol/L, suggesting that NO group or NOH group should be essential active groups of antioxidation in nitroxyl radical moieties.

Nitroxides were observed not to affect superoxide anion radical formation from rat neutrophils-zymosan system^[4]. The present study showed that GP1 inhibited superoxide anion radical formation in this test system. The results imply that GP1 is a scavenger for oxygen free radicals with more wide acting spectrum than nitroxides.

Podophyllotoxin and a number of its derivatives possess anticancer activity. It has been found that introduction of nitroxides into some antitumor drugs, such as podophyllotoxin, had significant antitumor activity with marked decrease in toxicity compared with the parent compounds^[5,6]. It is not very clear why introduction of nitroxides group enhances the antitumor activity but reduces the toxicity. It is possibly related to its antioxidative activity. Administration of antineoplastic agents such as GP derivative VP16 results in oxidative stress^[14], ie, the production of free radicals and other reactive oxygen species (ROS). ROS slow the rate of cell proliferation, and that occurring during chemotherapy may interfere with the cytotoxic effects of antineoplastic drugs, which depends on rapid proliferation of cancer cells for optimal activity^[15]. Therefore, simultaneous antioxidative effect may sensitize the reaction of tumor cells to cytotoxic drugs. Introduction of nitroxides group into GP made their derivatives possess dual actions, ie, as a potent antioxidant to possibly abolish the suppression of cell proliferation by ROS, and simultaneously as an inhibitor of topoisomerase II to enhance the anticancer effects of GP derivatives. Beyond their antioxidative activity, nitroxides also possess antitumor activities. This kind of active groups may synergize with original active groups in GP derivatives each other. On the other hand, ROS cause or contribute to certain side effects that are common to many anticancer drugs. The reduction of toxicity of GP derivatives with nitroxide group may also result from their antioxidative activity.

REFERENCES

• 1 Zhang JZ, Zhao BL, Zhang QG, editors. Basic theory and application of spin labeling ESR spectrum. 1st ed. Beijing: Science Press; 1987. p 1-5.
• 2 Hahn SM, Krishna CM, Samuni A, Degraff W, Cuscela DO, Johnstone P, et al. Potential use of nitroxides in radiation oncology. Cancer Res 1994; 54 (7 Suppl): 2006s-2010s.
• 3 Anastassopoulou JD, Rakintzis NT. Reaction of OH radicals with 2,2,6,6-tetramethyl-4-piperidine-oxyl (TEMPOL) in an aqueous solution. Z Phys Chem Neue Folge 1984; 141: s 53-8.
• 4 Wu YJ, Li WG, Zhang ZM, Tian X. Antioxidative activity o