Toxicological screening models: Drug-induced oxidative hemolysis

153

Toxicology Letters, 15 (1983) 153-158 Elsevier Biomedical Press

TOXICOLOGICAL HEMOLYSIS (Methemoglobin; anemia)

SCREENING

osmotic

fragility;

MODELS:

DRUG-INDUCED

phenacetin;

U.A. BOELSTERLI, K.P. SHIE*, E. BRINDLE

reticulocytes;

OXIDATIVE

toxic-hemolytic

and G. ZBINDEN

Institute of Toxicology, Swiss Federal Institute of Technology and University of Ziirich (Switzerland) (Received August 16th, 1982) (Accepted August 27th, 1982)

SUMMARY In an attempt to obtain a simple screening system for the assessment of toxic-hemolytic effects of chemical substances, a battery of hematological tests was used. Phenacetin served as reference substance. The drug caused reversible formation of methemoglobin and Heinz bodies and an increase in peripheral reticulocytes after 2 and 4 weeks of treatment. Furthermore, an increase in the mean corpuscular volume of red blood cells (RBC) and the volume of RBC ghosts in hypotonic solutions, and a decrease of the mean corpuscular fragility was observed. The latter changes are considered to be a consequence of regenerative RBC compensation rather than due to structural membrane alteration caused by the drug. The results suggest that only a combination of several hematological tests can provide comprehensive information about the hemolytic potential of chemical substances, and that for screening purposes small numbers of animals are often sufficient.

INTRODUCTION

The main purpose of the toxicological screening is the detection of hazardous properties of chemical substances. Whenever possible, the models used should also give quantitative information permitting comparison of the adverse effects of a test compound with those of standard reference agents and the establishment of a ranking order of toxicity if several new chemicals are studied. Toxicological screening tests should be rapid, reliable and reproducible, and should require small numbers of animals. * Present address: Department of Pharmacology, Shanghai Second Medical College, Shanghai (People’s Republic of China). Abbreviations: Hkt, hematocrit; MCF, mean corpuscular fragility; MCV mean corpuscular volume; MHb, methemoglobin; PBS, phosphate-buffered saline; Ret, reticulocyte. 0378-4274/83/0000-0000/803.00

0 Elsevier Biomedical Press

154

In this paper we review the methods available for the detection of hemolytic properties of chemicals. A simple rat model system is presented which includes measurements of various hematologic parameters, i.e., RBC count, Hkt ratio, MCV of RBC, Ret count, Heinz body and MHb formation, osmotic fragility of erythrocytes and surface/volume ratio of erythrocyte ghosts. Phenacetin, an analgesic aniline derivative, was used as a reference substance. This drug is known to cause oxidative hemolysis in man when taken at high doses. In patients with intraerythrocytic defects (e.g., glucose-6-phosphate dehydrogenase deficiency) or other metabolic idiosyncracies, the toxic action is exacerbated [l , 21. MATERIALS AND METHODS

Animals and drugs Male Sprague-Dawley rats (SIV 50, Ivanovas, Kissleg, F.R.G.), initial weight 180-200 g, were kept in Macrolon@ cages with wood shavings as bedding under controlled conditions of humidity, temperature and light. They had free access to water and standard rat chow (Kliba 343, Kliba, Kaiseraugst, Switzerland). Phenacetin (Ph. Eur., Merck, Darmstadt, F.R.G.) was administered by stomach tube to 5 rats in a dose of 500 mg/kg, 5 times weekly, as a suspension in 0.5 % methyl cellulose in a volume of 2 ml/kg. Controls received the vehicle only (n = 5). Drug treatment was discontinued after 4 weeks. In a separate experiment, two groups of 5 rats were given phenacetin or 0.5 % methyl cellulose (same treatment as above) for osmotic fragility and surface area/volume ratio tests. Hematologic parameters Blood was obtained from the tail vein by cutting off the tail tip at 2 and 4 weeks of phenacetin treatment, and 2 weeks after discontinuation of the drug. RBC count, Hkt ratio, and MCV were measured with a Coulter Counter, Model S. The number of Ret and Heinz body formation were determined in freshly prepared blood smears after staining with brilliant Cresyl blue (1% solution in ethanol) and Nile blue sulphate (0.5% solution in ethanol), respectively [3]. MHb was assayed spectrophotometrically as the difference in absorbance of cyan-MHb with and without previous addition of ferricyanide [4]. Osmotic fragility and surface/volume

ratio of erythrocyte ghosts

Osmotic fragility tests as described by Beutler [5] for human RBC were adapted for rat RBC. 1.5 ml blood was drawn percutaneously into a heparinized 5-ml syringe from the V. subclavia in light ether anesthesia. The cells were spun down (5 min at

155

2000 x g) and washed twice with phosphate-buffered (0.01 M) isotonic (0.15 M) saline (PBS), pH 7.4, containing 0.25% BSA to remove plasma and buffy coat. The RBC were resuspended in PBS-BSA at a hematocrit of 40%. 50 ~1 of the RBC suspension were added to 5 ml of a series of graded NaCl solutions, made up from a stock solution of PBS (9% NaCl, 96 mM Na2HP04, adjusted to pH 7.4 with NaH2POd). After incubation with the RBC suspension for 60 min at room temperature, the tubes were centrifuged (5 min at 2000 x g) to sediment unlysed cells, and the hemoglobin in the supernatant was measured spectrophotometrically at 540 nm. The MCF was determined graphically and expressed as the NaCl concentration at which 50% of the RBC hemolyse. For evaluating in vitro interactions of phenacetin on the osmotic fragility of RBC, the drug was added to the graded PBS solutions in final concentrations of 10e6 to 10e4 M and incubated as described above. For estimating the surface/volume ratio of RBC, the maximal volume of globular red cell ghosts was determined. The suspension of washed RBC (40% Hkt) was diluted 1:50000 and incubated for 90 min in hypotonic (0.3%) PBS (0.01 M phosphate, pH 7.4). The volume of the RBC ghosts was measured with a Coulter Counter, Model S. The surface area (S) of the globular red cell ghosts was calculated from the volume (V’) by the formula: S = 3(4 7r/3)1’3 x V2’3 [6]. RESULTS

After 2 weeks of treatment MHb and Ret counts were both increased markedly (PcO.01) and a large number of Heinz bodies was readily recognized (Table I).

TABLE

I

METHEMOGLOBIN

BLOOD

CELL

(MHb),

(RBC)

RETICIJLOCYTE

COUNT,

AND

RATIO

HEMATOCRIT

(Ret), HEINZ (Hkt)

BODY RATIO

OF FRESHLY

(Heinz),

OBTAINED

RED ERY-

THROCYTES Duration

MHb

Ret

Heinz

RBC

of treatment

(070)

(per 103RBC)

(per 10’RBC)

(lO”/l)

156+ 19a

118* 16”

Hkt

2 weeks Phenacetin

7.2t

Controls

1.2 + 0.4

1.9”

28+

3

0

5.93 * 0.09”

0.40 f 0.004a

7.74f0.16

0.47~0.009

4 weeks Phenacetin

12.0+0.8a

Controls 2 weeks after

7.6+

1.7

41+

4=

7+

7.09+0.12a

0.44+ 0.009

29+

3

0

7.55kO.13

0.42 + 0.004

3”

drug

discontinuation Phenacetin

2.7i0.3

29+

la

0

7.19iO.16

0.40~0.009

Controls

2.9 + 0.2

24+

1

0

7.66+0.12

0.40+0.004

RBC were obtained from male Sprague-Dawley rats treated with phenacetin (500 mg/kg). The data are expressed as the mean + SEM; ‘PcO.05; Wilcoxon-Mann-Whitney test, 2-tailed; n = 5 rats.

156

MHb formation further increased after 4 weeks, but the incidence of Ret and Heinz bodies decreased to lower values although they were still significantly higher than in the controls. Drug treatment resulted in decreased RBC counts at 2 and 4 weeks and in reduced Hkt ratios at 2 weeks (Table I). Following discontinuation of drug administration, Heinz bodies, MHb, and Hkt returned to normal values, and only Ret and RBC counts were still slightly abnormal. The MCV of phenacetin-treated rat RBC in isotonic solutions was increased by approximately 12% after 2 and 4 weeks and still remained significantly higher than the control at 2 weeks after termination of drug treatment (Table II). Red cell ghost volume, as measured after incubation of RBC in hypotonic solution, was unchanged after 2 weeks but increased by approx. 14% after 4 weeks of phenacetin treatment. This corresponds to an approx. 9% increase in calculated ghost surface area (Table II). After 4 weeks, the osmotic fragility curve of RBC from phenacetin-treated rats was shifted to lower NaCl concentrations, i.e. the osmotic resistance increased (MCF decreased from 0.468 to 0.428 % NaCl, PcO.01) (Fig. 1). At 2 weeks no effect was shown. In vitro incubation of phenacetin with RBC did not alter the osmotic resistance (data not shown). TABLE II MEAN CORPUSCULAR VOLUME (MCV) OF RED BLOOD CELLS (RBC) IN ISOTONIC PBS, AND VOLUME AND SURFACE AREA OF RBC GHOSTS IN HYPOTONIC (0.3%) PBS Duration of treatment

MCV

Volume (V) of RBC ghosts (fim3)

Surface area (S) of RBC ghosts (&)

S/V ratio

(pm’)

2 weeks Phenacetin Controls

68 2 0.7” 60 i 0.2

124k 1.8 122kl.8

120* 1.2 llY_tl.i

0.968 0.975

4 weeks Phenacetin Control5

62*0X 55 s 0.3

134&2.2a 118,1.6

127* 1.4= 116tl.l

0.945* 0.986

2 weeks after drug discontinuation Phenacetin Controls

56kO.2” 53 f 0.4

nd

nd

nd

RBC were obtained from male Sprague-Dawley rats treated with phenacetin (500 mg/kg). Mean + SEM; aP<0.05; Wilcoxon-Mann-Whitney test, 2-tailed; n = 5 rats; nd, not determined.

DISCUSSION

The present study demonstrates

that toxic-hemol~ic

anemia induced by chronic

157

Hemolysis

0.35

(%)

0.40

cl.45

0.50 NaCl

0.55 (g/

1 OOml)

Fig. 1. Osmotic fragility curve of phenacetin-treated (500 mg/kg/day for 4 weeks) ( A---A) and control rats (0-e). MCF for phenacetin-treated rats, 0.428% NaCl; for control erythrocytes, 0.468 % NaCI. *P
treatment with phenacetin can readily be recognized in a small number of rats. The observation of MHb and Heinz body formation indicate that the mechanism of this type of drug-induced hemolysis is based on an oxidative damage to hemoglobin, leading to its denaturation and attachment to the red cell membrane [2, 71. These altered erythrocytes show a reduced lifespan and are readily taken up by the reticuloendothelial system. In the present rat model, all treated animals promptly responded to the drug. Calculated on a mg/kg base, the dose used in this study is 10 times higher than that which produces hemolysis in man 121.The toxic effect and the signs of compensatory regeneration were marked after 2 weeks of drug administration and less pronounced after 4 weeks. This could possibly be due to a reduction of the phenacetin plasma levels known to occur in chronically treated rats [8]. The increased MCV of RBC partly reflects the presence of large quantities of reticulocytes and young RBC featuring a different membrane lipid composition [9]. It cannot be excluded, however, that a possible structural change of the RBC surface membrane occurred during treatment with the drug leading to an increase in volume. This feature manifested itself by an increased critical volume of RBC ghosts which is the consequence of an enlarged surface area (Table II). The increased resistance to osmotic lysis of the RBC membranes in graded hypotonic PBS can be explained by this shift to a smaller S/V ratio. The concept of using osmotic-fragility tests as an indicator for damage to RBC leading ultimately to hemolysis needs reevaluation. Usually this test is performed in vitro, i.e. by incubating RBC in the presence of the test compound in hypotonic solutions. Structural changes of certain membrane components caused by insertion of the drug into the membrane bilayer and leading to expansion or condensation of the membrane can be recognized by this technique. Similarly, in the case of preexisting hereditary corpuscular abnormalities, the changes in RBC shape and/or size

158

and, thus, in the S/V ratio, can be readily detected. However, chronic damage due to oxidative membrane damage caused by metabolites of the test compound cannot be recognized with the in vitro test. Likewise, other pharmacological parameters influencing interactions of the drug with erythrocytes, such as binding to plasma proteins, development of tolerance phenomena, etc., cannot be shown with this test. Contrarily, a shift to higher MCV, caused by an increased population of reticulocytes or by physico-chemical structural changes of the RBC membrane might lead to a smaller MCF value in hypotonic solutions when compared with controls without taking into account that the lifespan of RBC might be reduced due to oxidative hemoglobin and membrane damage. Indeed, all the hematological data of the present study suggest toxic-hemolytic injury although the results from the osmotic-fragility tests would rather indicate a membrane ‘stabilization’ leading to increased osmotic resistance. A direct interaction of the phenacetin molecule with the red cell membrane can be excluded from in vitro fragility tests. It has been shown that a phenacetin metabolite, 2_hydroxyphenetidine, is responsible for oxidative hemolysis in vivo [2]. Therefore, in vivo osmotic-fragility tests reflect structural changes of the RBC membrane and provide useful information about alterations in the S/V ratio but do not necessarily run parallel with other signs of hemolysis. Hemolysis can be caused by many different mechanisms, and not in all cases is it accompanied by MHb and Heinz body formation [lo]. Therefore, for the toxicological screening it is necessary to use a set of tests which provide comprehensive information on the toxic and regenerative changes of the RBC system. The present study shows that this is feasible with relatively modest technical efforts. REFERENCES

1 E. Beutler, 2 E.C.

Drug-induced

Gordon-Smith,

3 L. Hallmann, 4 S. Frankel, Diagnosis,

Klinische S. Reitman

Chemie

Osmotic

Hematology, Streuli,

erythrocytes

J.R.

Rev., 21 (1969) 73-103.

haemolysis,

Clin. Haematol.,

Sonnenwirth MO,

in W.J.

(Eds.),

Stuttgart,

Gradwohl’s

9 (1980) 557-586.

1980, pp. 32-34.

Clinical

Laboratory

Methods

and

1970, pp. 400-401. E. Beutler,

A.J.

Erslev and R.W.

Rundles

(Eds.),

1977, pp. 1609-1610.

R.B. Gunn

the membrane

Thieme,

Williams,

Hill, New York,

Kanofsky,

following

Pharmacol.

St. Louis,

fragility,

McGraw

anemia, oxidative

und Mikroskopie,

and A.C.

Vol. 1, Mosby,

5 E. Beutler, 6 R.A.

hemolytic

Drug-induced

and S. Yachnin,

insertion

Diminished

of oxygenated

osmotic

fragility

sterol compounds,

of human

Blood,

58 (1981)

317-325. 7 H.S.

Jacob,

Hematol.,

Mechanisms

of Heinz body formation

and attachment

to red cell membrane,

8 G. Carro-Ciampi, Appl. Pharmacol.,

Tolerance to repeated 22 (1972) 641-648.

phenacetin

treatment

in albino rats and guinea pigs, Toxicol.

9 M.P. Westerman, L.E. Pierce and W.N. Jensen, Erythrocyte lipids: a comparison and normal old populations, J. Lab. Clin. Med., 62 (1963) 394-400. 10 E. Beutler

Seminars

7 (1970) 341-354.

and B.J. Mikus,

tion on blood

methemoglobin

The effect of sodium

nitrite and paraamino-propiophenone

levels and red blood

cell survival,

Blood,

of normal

young

administra-

18 (1961) 455-467.