Indian Journal of Experimental Biology Vol. 42, January 2004, pp. 81-85

Effect of potassium channel modulators on toxicity of Cleistanthus collinus

Vinu M Jose 1·, K N Anand 1, L Jeyaseelan2

, Kalpana Ernest1 & Alice Kuruvilla3

1Department of Pharmacology and 2Biostatistics, Christian Medical College and Hospital , Vellore 632 002, India 3Department of Pharmacology , P.S .G.I.M.S.R. , Coimbatore, 64 1 004, India

Received 16 June 2002; revised 31 July 2003

The study was conducted to determine the effects of boiled extract of C/eistcmthus co/linus on rats by observing ECG changes and electrolyte levels in serum and urine. lnfluence of minoxidil and glibenclamide on Cleisranrhus collinus induced toxicity was determined. ED50 for arrhythmia, changes in contractility and heart rate were recorded using the isolated frog heart. Cleisranthus at low doses caused transient tachycardia and increase in contractility and at high dose caused arrhythmi a and cardiac arrest in rat. LD50 was found to be 1690 mg/kg. Minoxidil potentiated cardiac toxicity, whereas glibenclamide did no\ produce any significant change. High concentration of potassium in Cleistanthus extract hindered comparison of its levels. There was excretion of sodium even in the presence of hyponatraemia. Cleistanthus at low dose caused transient tachycardi a and increase in contractility and at hi gh dose caused arrhythmia and cardiac arrest in isolated frog heart. ED50 for arrhythmia was found to be 1406 mg/kg. Acute toxicity was mainly due to depressive cardiac activity of Cleistanthus. It also caused renal failure . Potassium channel modulators did not have important role in acute cardiac toxic ity treatment. Probably in chronic toxicity, electrolyte level changes are involved and potassium channel modulators might have a ro le .

Key words : Acute toxicity, Cleisranrlws co/linus, Contracti lity, ECG, Electrolyte, Heart rate, Oduvanthalai

Cleistanthus collinus, belonging to Euphorbiaceae is a poisonous tree that causes mortality when consumed in fresh form or as extract in water1

-4. Treatment protocol for Cleistanthus poisoning includes stomach wash, activated charcoal, intravenous fluids and potassium replacement, cardiac pacing, atropine and the required supportive measures. The principle of treatment includes decreasing the absorption of Cleistanthus, maintaining normal potassium level, preventing bradycardia and providing supportive measures. No specific antidote is available till now.

In Cleistanthus poisoning, most of the deaths occur when hypokalemia is at its peak3 and is mostly due to anhythmia. Adenosine ttiphosphate sensitive potassium channel (KATr) modul ators can modify cardiac function by altering potassium level 5

·6 and thus it

may alter the toxic effect of Cleistanthus on heart. It has been observed that boiled extract of

Cleistanthus is more toxic to human beings as assessed by mortality rate3

. An animal study dos:~.e earlier used fresh ground extract It had minute particles that could obstruct the small blood vessels

*Address for correspondence: Senior Lecturer in Pharmacology. Government Medical College, Thrissur, 680 596, India E-mail: vishnujose@hotmail.com

when given intravenously and also its colour changed with time, as observed earlier in our laborator/. Effect of boiled extract has not yet been validated in animal models. Effects observed on patients following Cleistanthus poisoning is simu lated maximum by the use of boiled extract for the experiment rather than isolated active constituents. Hence, this study was aimed at determining the effects of boiled extract of Cleistanthus on the isolated perfused frog heart and to compare ED50 for arrhythmia of pH adjusted with pH unadjusted extract If comparable, the pattern of cardiac and electrolyte level changes after in vivo administration of extract to rat was also observed, since these were the usual effects observed in the poisoned patients. Influence of KATP modulators on Cleistanthus-induced cardiac toxicity in rat was determined and compared.

Materials and Methods Animal care and use - Frogs weighing 40 to 60 g

and Wistar albino rats of either sex weighing 250 to 350 g were used for the study. The rats were housed under standard laboratory conditions with 12 hr dark and light cycle. They were provided food and water ad libitum.

To detect the effect on isolated heart, six frogs were used in each group. For the determination of

82 INDIAN J EXP BIOL, JANUARY 2004

changes in electrocardiogram (ECG), serum and urine electrolyte values, six rats were used in each of the groups. The Christian Medical College animal ethical committee sanctioned the use of the above number of animals for the study.

Drugs and chemicals- Pentobarbitone sodium (Loba-Chemie lndoaustranal Co, Bombay), minoxidil su lphate (Dr Reddy's Laboratory, Hyderabad), glibenclamide (Nicholas Piramal India Ltd, Mumbai), heparin (Gland Pharma Limited, Hyderabad), double distilled water, 0.9% saline, and chemicals like sodium hydroxide, sodium chloride, potassium chloride, sodium bicarbonate, sodium dihydrogen phosphate, calcium chloride and D glucose from standard laboratory suppliers were used.

Preparation of extract- Fresh Cleistanthus leaves identified by a botanist was collected and used for the study. Double distilled water (300 ml) was taken in a porcelaine dish (Porcelaine dish of 15 em dia) and kept for boiling. Freshly collected Cleistanthus leaves (30 g) were added to the above boiling distilled water. Boiling was continued for 15 min. Then it was left to cool and the leaves removed. About 150 ml of yellowish brown extract with a characteristic odour and pH (Systronics, India) at 3.2 to 3.4 was obtained. lt was adjusted at 7.34-7.44 by titration and addition of 1 N sodium hydroxide. The volume of the extract was adjusted to 300 ml with distilled water, filtered and used for the study as the pH adjusted extract and without pH adjustment extract was also made up to 300 ml to obtain pH unadjusted extract. The above solution was calculated to have 100 mg of leaf extract in 1 ml of distilled water. Conventionally patients use similar boiled preparation, therefore the same method was followed for leaf extract preparation.

In vitro experiment on frog heart- Frog was pithed and pinned to frog board, thoracic cavity opened, pericardium removed, a thread passed below the inferior venacava and then nicked, Symes cannula introduced and tied. It was separated, and perfused with frog Ringer solution to obtain an isolated frog heart preparation ~ . Increasing closes of Cleistanthus were added to the isolated frog heart set up and the heart rate and force of contraction recorded till anhythmia started appearing. ED5o for anhythmia was determined9 and compared between the pH adjusted and unadjusted extracts.

Changes in ECG of rat heart after administration of extract- Rats were anaesthetised with pentobarbitone sodium (50 mg/kg; ip ). The external jugular vein and trachea were cannulated. For

artificial ventilation when required, the tracheal cannula was connected to artificial respirator (500 mllkg/min). Heparin solution (0.5 ml of 500 IU/ml ) was injected through venous cannula to prevent interference due to coagulation of blood in the tube.

The extract was given iv at the rate of 1m! in 3 min9

·10

. Limb lead II ECG recording was done every min for 3-5sec using the polygraph (Polyrite, Chandigarh) . The time taken for absence of P wave, for heart rate to decrease by 50%, and for cardiac asystole were determined. In some animals, occasional further transient contractions and electrical activities were observed after an initial cardiac arrest. Hence voltage recordings of less than 3 mY were considered as cardiac inactivity . The LD50 was also determined.

Pre-treatment groups were give n glibenclamide (5 mg/kg; ip) 11 or minoxidil (1.6 mg/kg; ip) 12 15 min prior to the experiment undertaken.

Measurement of electrolyte level- Sodium and potassium levels in the boiled extract, and serum and urine of rat were estimated. Blood was collected from the inferior venacava after the death of the rat. Urine samples were collected by supra pubic puncture.

Statistical method- Heart rate and contractility were compared using repeated measure ANOV A test followed by Dunnett multiple comparison test. Time changes in heart rate were compared using one-way ANOV A test followed by Tukey-Kramer multiple comparison test. Electrolyte levels were compared using Kruskal -Wallis test followed by Dunns test. Kaplan Meier survival curve, log rank and Cox regression analysis were also performed. The data was analysed using Graph Pad Instat Program and survival curve by SPSS. The dose causing arrhythmia in 50% of animals was determined by graphical method of Miller and Tainter9

.

Results Effect on isolated frog heart- ED50 fo r anhythmia

with Cleistanthus leaves pH-unadjusted extract on frog heart was found to be 753 ± 174j..l.l and for pH­adjusted extract was 703 ± 186j..1.1 which was not significantly different.

Iso lated frog heart rates , after exposure to various concentrations of pH adjusted boiled extract were compared (Fig . I). Extract ( 1.6 ml) exposed heart rate, showed significant decrease from pre-treatment heart rate (Table 1 ).

Concentration of the extract to which the frog's heart was exposed was compared with contractility (Fig. 1). Initial contractility was considered as 100%

JOSE er al.: A STUDY ON THE ACUTE TOXICITY OF CLEISTANTHUS COLLIN US 83

and subsequent contractility were determined compa~ing it with initial contractility. There was significant increase at 0.4 ml of extract and decrease noted at 1.6 ml of extract, when compared with pre­treatment contractility (Table I ).

In vivo effect of extract on rat heart-ECG changes were studied in 6 rats. Initial tachycardia was noted in 5 rats. Abnormal P waves and absence of P wave were observed in all the rats. Other changes included junctional escape in 4, atrioventricular block in 2, bundle branch block in 3, ev idence of multiform wide QRS escape beat in I , tall T wave in 1 and transient polymorphic ventricular tachycardi a in 1 rat. Finally all the rats developed bradysystolic card iac arrest. Similar ECG changes were observed in minoxidi l and glibenclamide pre-treated rats.

There was no significant difference between the three groups for time taken for 50% heart rate and for absence of P wave when compared between KATP

modulator pre-treated and untreated rats (Table 2). Kaplan Meier survival curve was done to compare

the survival of rats in different groups (Fig. 2). There was significant difference between the groups noted by Log rank statistics. Risk for asystole was 3.9 times for minoxidil group and 1.3 times fo r glibenclamicle

~ li!' ' Jl: I II ' I ' ' I 1 ' I'l l I ~ I I ' I

I •ill '•II i'f'lfl • '),• "1 111' !riff !1'1·,1• l•lh' l

o.a.J.

,, oL

' . .u-.t.

• ·t.J.

Fig. I- Effect of increasing doses of boiled extract of Cleisranrlws on isolated frog heart preparation

Table I - Effect of Cleisrantlzus on isolated frog heart

I Values are mean± SD of 6 animals]

Volume (ml ) Hean rate (min) Contractility(%)

0.0 57.67±10.99 100±0 0.1 59. 17±9.68 11 1.20±10.34 0.2 60.33±11.15 122.03±9.66 0.4 58.67±12.37 155.60±32.90* 0.8 51.50±12.39 119.2 1±63 .96 1.6 37 .33± 12.53** 30.42±25.97**

F ratio (df) 13.789 (5,30) 10. 198 (5,30) P value 0.0001 0.0001

*P<0.05 , **P<O.O I compared to pretreatment value (0.0 ml).

group when compared to Cleistanthus alone group by univariate Cox regression analysis. Survival curve with glibenclamide was found to cross Cleistanthus survival curve at certain points. LD50 with boiled Cleistanthus extract in rat was 1690 mg/kg.

In vivo study on serum and urine sodium and potassium levels - Sodium and potassium levels were compared with control values, Cleistanthus alone treated rats, those pretreated with glibenclamide and minoxidil. Cleistanthus extract contained potassium (11.57 ± 3. L1 milliequivalent per ml) and no detectable sod ium. In serum of Cleistanthus, glibenclamide and minoxiclil treated group, sodium values were s ignificantly low compared to control values. In serum of glibenclamide and minoxidil treated group, potassium was high compared to control values (Table 3). In urine of Cleistanthus, glibenclamide and minoxidil treated group, sodium levels were significantly hi gh compared to control values whereas

1.2

1.0

2 ~ 0.8

m >

-~ 0.8 ::J U'l (l)

> 0.4 715 ::J

E ::J 0 .2 ()

0.0

I I I

~---,

I : L----, ~--~

I L_l ~-----

1

L_ l I L_ ,

I I

-------Minoxidil . - - - Glibenclamide --Only Cleistanthus

extract

-0.2 +-------~------~--------~~ 10 20 30 40 50

Time for asystole (min)

Fi g. 2- Kaplan Meier surviva l curves showing time (min) to asystole fo r Cleistanthus alone, minoxid il and gl ibenclamide groups. Values are significant at P < 0.05 noted between the groups by Log rank stat istics.

Table 2-Time taken for changes in the heart rate in rat

I Values are mean ± SD of 3 repli cations]

Group (min) Time for rate Time for disappearance to fall below 50% of P wave

Cleistanthus 21.00±7.32 2 1.33 ±6.86

Gli benclamide 15.00± 10. 18 17 .17±9.45

Minoxidil 10.00±6.60 10.17±6.49

F ratio (df) 2.719(2, 15) 3.209(2, 15)

P value 0.0983 0.0691

84 INDIAN J EXP BIOL, JANUARY 2004

Table 3-Serum and urine electrolyte levels in rat

[Values are mean± SD of 4 replication]

Groups Serum sodium Serum potassium Urine sodium Urine potassium (meq/1) (meq/1) (mcq/1) (meq/1)

Control 146.50 ± 3.27 4.8 ±0.28 18.50±9.7 219.33±29.17 Cleistanthus 85.83 ± 8.80** 21.18±3.28 281 .67 ± 80.67** 491.50± 144.09* G1ibenclamide 90.50 ± I 3.56* 23.52±5.29** 258.83 ± 132.64** 484.67 ± 204.04* Minoxidil 89.83± 12.34* 24.65 ± 3.72** 259.83 ± 232.94* 447.00± 179.19 KW (df) 13.081 (3 ,20) 13.991(3,20) P value 0.0045 0.0029

** P < 0.0 I, * P < 0.05 compared to control group values.

in urine of Cleistanthus and glibenclamide treated group, potassium values were high compared to control values (Table 3).

Discussion The pH of the extract had to be adjusted to

7.4 ± 0.04 with sodium hydroxide since acidic substances can cause decrease in contractilitl. Since pH adjustment caused precipitation of certain proteins, the effects of pH adjusted and unadjusted extracts were compared. There was no significant difference noted for ED50 to produce arrhythmia and hence the effects of the two extracts were comparable.

In isolated perfused frog heart preparation, Cleistanthus extract produced transient insignificant tachycardia, significant increase in contractility followed by arrhythmia and significant decrease in heart rate, finally leading to cardiac arrest at higher dose. Similar findings have been noted earlier using ground ex trace. ED 50 for arrhythmia was 703 ± 186 !J.l (Dose of pH-adjusted extract was equivalent to 1406 mg/kg).

When Cleistanthus extract was given intravenously to anaesthetised rats, it produced changes in ECG. These effects can be summarised as due to initial stimulation followed by gradual inhibition of electrical activity of the heart, first affecting the pacemaker then gradually affecting the lower centres, finally causing total asystole. An earlier study done using oral ground extract in rat has shown a decrease in heart rate and arrhythmias, which are mainly junctional in origin, but does not show the initial tachycardia7

.

Increase in contractility and arrhythmia may be due d ' ff 1 .d . Cl . I 7 13- ls to 1 erent g yeast es m et.stant 1us extract · .

Direct toxicity of Cleistanthus might also be partly responsible for this, considering the immediate cardiac effect noted. Cardiac glycosides by inhibiting sodium-potassium ATPase leads to increased availability of calcium leading to increase in

13.336(3,20) 9.144(3,20) 0.004 0.0274

contractility and arrhythmia6. ECG changes noted in

patients are tachycardia, bradycardia, flat P wave, prolonged QT interval, ST segment depression and inversion ofT wave3

'16

. Our study showed evidence of bradyarrhythmias, proving clinical benefit of pacing.

There was no evidence of hypokalemia in the rat serum following extract administration. The earlier oral toxicity study using ground extract also does not show evidence of hypokalemia7

. A concrete conclusion on serum electrolyte level could not be made out due to high potassium and low sodium content in the extract. Hyperkalemia as the cause of cardiac arrest needs to be excluded since it was systolic cardiac arrest.

Urine examination of rat showed significant increase in excretion of sodium and potassium with extract treatment. Considering the fact that administering extract (3-5 ml) to blood (6-12 ml) of rat9 will cause a huge change in electrolyte make up and blood volume, the variation in electrolyte excretion has to be viewed with cauti on. Significant sodium excretion at normal, and low sodium level in serum is due to nephrotoxicity 17 of Cleistanthus as has been reported in 37% of patients3

. There are evidences of urinary potassium loss and absence of intracellular shift of potassium in patients 16

. In 72% patients, peak hypokalemia has been noted between 3'd and st" day and the peak mortality coincides with this period3

. From the observation in animals and patients, hypokalemia seems to be chronic rather than an acute cause. This is likely due to steroid-like structure of glycoside or secondary to activation of renin-angiotensin-aldosterone axis since it requires 3 to 4 days. In the present study pre-treatment drugs did not produce any characteristic alteration in ECG pattern compared to Cleistanthus group.

None of pretreatment drugs was protective. The significant toxicity of minoxidil observed in the survival curve might be due to exaggeration of cellular

JOSE et al.: A STUDY ON THE ACUTE TOXICITY OF CLEISTANTHUS COLLINUS 85

hypokalemia due to opening of potassium channel. Survival curve for glibenclamide seems to overlap with that of Cleistanthus though a significant difference was not noted. A bigger .sample size might be needed if at all a protective effect for glibenclamide is to be demonstrated. LD50 in rat was 1690 mg/kg.

It can be concluded that acute toxicity was mainly due to cardiac activity of Cleistanthus, that can also cause nephrotoxicity. Potassium channel modulators did not have important role in acute toxicity treatment.

Acknowledgement The authors express their gratitude to Mr V

Shankar, for helping in statistical analysis and Dr Ashish, for helping in ECG reading. We also express our sincere thanks to the Christian Medical College research committee for providing grant for the project. Thanks are also due to Mr. Raju John T., Department of Botany, Saint Aloysius College, Elthuruth, Thrissur, Kerala 680 011 for identification of the plant.

References 1 Nadkarni A K, Indian Materia Medica, 3'd edition, edited by

K M Nadkarni (Popular Book Depot, Bombay) 1954,349. 2 Modi Jaising P, Organic irritant poison (I), in Modi's

Medical Jurisprudence and toxicology, 22"d edition, edited by B V Subrahmanyan (Butterworths India, New Delhi) 1999,250.

3 Kurien Thomas, Dayal A K, Alan Gijsbers, Seshadri M S, Cherian AM, Oduvanthalai leaf poisoning, lAP!, 35 ( 1987) 769.

4 Aleem M A, Paramasivam M, Spectrum of acute poisoning in villagers, lAP!, 41(1993) 859 [Abstract] .

5 Gillian Edwards & Arthur H Weston. The pharmacology of ATP-sensitive potassium channels, Ann Rev of Pharmacal Toxicol, 33 (1995) 597.

6 William F Ganong, Review of medical physiology, 19'h edition (Prentice-Hall International Inc, USA.) 1999.

7 Sathiakumar N B, Studies on Cleistanthus collinus­A pharmacological and toxicological characterisationo Ph.D. thesis, submitted to University of Madras, Madras, 1988.

8 Muralikrishna R, Vade mecum of experimental pharmacology (ESPE Media and Printing, India.) 1999, 37.

9 Ghosh M N, Fundamentals of experimental pharmacology, 2nd edition (Scientific Book Agency, Calcutta) 1984.

10 The Staff of the Department of Pharmacology, University of Edinburgh and L 1 McLeod, Phannacological experimenrs on intact preparations, (E and S Livingstone, Edinburgh and London .) 1970.

II Lepran I, Baczko l, Varro A & Papp J G, ATP sensitive potassium channel modulators: Both pinacidil and glibenclamide produce antiarrhythmic activity during acute myocardia l infection in conscious rats, JPET, 277 (1996) 1215.

12 Manoj G Tyagi & Yinu M Jose, Anticonvulsant effect of cytoskeletal depolymerizers in combination with potassium channel opener and adenylate cyclase activators: A causative link with nerve growth factor? Indian J Exp Bioi, 39 (2001) I.

13 Ragupathi G., Prabhasankar P, Sekharan P C, Annapoorani K S & Damodaran C, Enzyme-linked immunosorbent assay (ELISA) for the determination of the toxic glycoside cleistanthin B, Foren Sci bu. 56 (1992) 127.

14 Viswanathan N & Joshi B S, Toxic constituents of some Indian Plants, Curr SciL 52 ( 1983) I.

15 Annapoorani K S, Periakali P, Ilangovan S, D amodaran C & Sekharan PC, Spectrofluorometric determination of the toxic constituents of Cleistanthus collinus, J Anal Toxicol , 8 (1984) 182 .

16 Kurien Thomas, Dayal A K, Narasimhan, Ganesh A, Sheshadri M S, Cherian AM, Kanakasabapathi & Bhanu M, Metabolic and cardiac effects of Clistanthus collinus poisoning, lAP!, 39 (1991) 312.

17 Hugh R Brady & Barry M Brenner, Acute renal failure in Harrison's principles of internal medicine, 14'h edition, edited by Anthony S Fauci, Eugene Eraunwald , Kurt J lsselbacher, Jean D Wilson, Joseph B Martin, Dennis L Kasper, Stephen L Hauser and Dan L Lango (McGraw-Hill Companies Inc, USA) 1998, 1505.