Psychopharmacologia (Berl.) 43, 223-227 (1975) - 9 by Springer-Verlag 1975

Probenecid: Dosage, Levels in Plasma and Cerebrospinal Fluid (CSF) and Influence upon CSF Levels of Homovanillic Acid (HVA) and 5-Hydroxyindoleacetic Acid (5-HIAA) in the Rabbit J. P. MUIZELAAR and J. I. OBERINK Netherlands Central Institute for Brain Research, Amsterdam Received November 4, 1974; Final Version April 4, 1975

Abstract. Probenecid retards the efflux of acid monoamine metabolites from the brain tissue and CSF to the blood. The probenecid-induced accumulation of these metabolites is held to be indicative of the turnover rate of the corresponding amines. Although the penetration of probenecid into the CSF does not proceed at a constant rate, Korf et al. (1972) and Sj6strom (1972) have shown a correlation between CSF levels of probenecid and that of HVA and 5-HIAA. In this study an attempt was made to establish the relationship be-

tween doses of probenecid and levels of this compound in plasma and CSF; between levels in plasma and CSF; and between CSF levels of probenecid and of HVA and 5-HIAA. This study was performed in a homogeneous group of laboratory rabbits. All correlations proved to be significant. The implications of these results for studies using the probenecid technique are discussed.

Key words: Probenecid - Homovanillic Acid - 5-Hydroxyindoleacetic Acid - CSF - Rabbits.

Introduction Probenecid is a drug known to retard the efflux of homovanillic acid (HVA) and 5-hydroxyindoleacetic acid (5-HIAA) from the brain tissue to the blood (Neff etal., 1964, 1967; Werdinius, 1966) and from the cerebro-spinal fluid to the blood (Guldberg et al., 1966). Although there are arguments against the probenecid technique (Eccleston, 1973), it has been shown that probenecid does not influence serotonine (5-HT) turnover (Barkai et al., 1972; K o r f et al., 1972) and that the monoamine turnover rate as measured by the probenecid method corresponds well to results obtained using other techniques (Neff et al., 1969). Thus, the rate of HVA and 5-HIAA accumulation in CSF after probenecid treatment is held to be indicative of corresponding brain monoamine turnover (Olsson et al., 1968 ; Tamarkin et al., 1970). The probenecid method has been used to elucidate some aspects of monoamine turnover in a number of diseases in man, including Parkinson's disease (Olsson et al., 1968), Huntington's chorea (Lal et al., 1973), unipolar depression (Bowers, 1972), manicdepressive psychosis (Roos et al., 1969), acute psychosis (Bowers, 1973), Alzheimer's disease (Gottfries et al., 1974) and Jacob-Creutzfeldt's disease (Brun et al., 1971). Apart from these studies, carried out in patients where a diagnosis had already been provided, attempts have been made to classify certain disorders

using the probenecid technique (Van Praag etal., 1971). Results have been used to select patients with the best chance of responding well to certain drug therapies, both in psychiatry (Van Praag et al., 1974) and in neurology (Lakke et al., 1972). Much work has also been carried out in laboratory animals using the probenecid method to investigate physiological or pharmacological aspects of m o n o amine metabolism (e.g., Andersson et aI., 1973a, b). In most of these studies, however, the levels of probenecid in CSF are not mentioned, although it has been shown that even if probenecid penetration into the CSF does not proceed at a constant rate, there is a significant correlation between levels of probenecid and those of HVA and 5-HIAA in h u m a n CSF ( K o r f et al., 1971 a; Sj6strom, 1972). In this investigation we studied correlations between administered doses of probenecid and levels in plasma and CSF and the correlation between levels of probenecid and HVA and 5-HIAA in CSF. Methods Six male New Zealand white rabbits, weighing about 3 kg, housed in individual cages and receiving a diet of standard rabbit chow and water, were used. Sodium probenecid was dissolved in the minimum volume of 3 N NaOH, and the pH adjusted to 7.5 with diluted HC1. Distilled water was added to provide a solution containing 100 mg probenecid/ml. A probenecid dose of 25, 50, 100 or

224

Psychopharmacologia (Berl.), Vol. 43, Fasc. 3 (1975) Table 1

Eluate 0.01 N NH4OH 12 N HC1 0.2 ~ K~Fe(CN)6 Cysteine- KaFe(CN)6 ~ Stand. 5-HIAA 10 pg/ml 12 N HC1 6 ~ Cysteine O-Phthaldialdehyde 0,3 ~ 12 N HC1

Standard

Blank

(ml)

(ml)

1 0.1 0.01 0.5 0.05 1

1 0.1 -0.5 -0.05 1

Oxidation blank (ml)

Sample (ml)

Sample + standard (ml)

Tissue blank (ml)

1

t

-

1

1

.

0.5 0.05

0.1

0.5 0.05

-

0.1

-

-

0.01

--

0.05

0.5 0.05

0.5 0.05

0.05 0.05

1

1

i

1

0.05

.

.

.

-

-

Water bath 20 min 75~C. Cooling 3 rain 20 ~C. Read at 360/485 nm. Equal parts of 6 ~ cysteine and 0.2 ~ K3Fe(CN)6. 200 mg/kg body weight (b.w.) was injected intraperitoneally, and the same dose repeated 2 hrs later. The amount of probenecid administered therefore ranged from 50-400 mg/ kg b.w. Exactly 4 hrs after the first injection, the CSF sample was obtained. General anesthesia was induced, using 3 ~/o fluothane in a mixture of 7 0 ~ N20 and 30~o 02, inhaled via a mask over the animal's nose and mouth. A suboccipital puncture was performed following induction. In most cases 1 - 2 ml of CSF was obtained. Samples containing blood were not used. Blood samples were taken from the lateral ear vein in syringes containing dead-space heparine, and immediately centrifuged. The CSF and plasma samples were immediately stored at - 80~ and analysed within 2 weeks. There was at least t0 days between tests on the same animal. Chemical Analysis. Sephadex G10 columns, bore 1.6 cm and height 3 cm, were washed with 20 ml water as described by Korf et al. (1971 b,c). Cisternal CSF was adjusted to pH of 2.5 with concentrated formic acid and applied to the Sephadex GI0 columns. After the solvent had passed through it, the column was washed with 10 ml 0,01 N formic acid, 1 ml water and 2.5 ml phosphate buffer 0.5 M pH 8.5. In all procedures double quartz distilled water was used. HVA was eluted with 7 ml phosphate buffer 0.5 M pH 8.5 and 3 ml water. The HVA was analysed fluorimetricatly as described by Korf et al. (1971 a). The elution of 5-H1AA and probenecid was performed with 9 ml 0.01 M NH4OH solution. 5-HIAA was determined fluorimetrically according to a method described by And6n et al. (1967), as showa in Table 1. Probenecid in CSF was read spectrophotometrically at 242.5 nm in strong alkaline solution (Dayton et al., 1963). Two ml aliquots were taken from the eluate and mixed with 2 ml 6 N NaOH. Probenecid in plasma was also determined spectrophotometrically in strong alkaline solution. Plasma was adjusted to pH 2.5 with concentrated formic acid. This solution was applied to a Sephadex GI0 column, height 5 cm and bore I cm. After the sample had passed, the column was washed with 2 ml 0.01 N formic acid, 1 ml water, 5 mt 0.5 M phosphate buffer pH 8.5 and 1 ml water. Then probenecid was eluted with 8 ml 0.01 M ammonia and measured as described.

Statistical procedures (computing of correlation coefficients and regression) were performed on an IBM 1130 computer, using standard IBM programmes. Results

Results are shown in Table 2. Samples taken without probenecid pretreatment were paired for greater determination accuracy at these low concentrations o f H V A and 5 - H I A A . In some cases the dead-space heparine has not been able to prevent the blood f r o m clotting, and no accurate plasma probenecid concentration could be determined in these. The results are r a n k e d according to the increasing concentration o f probenecid in the CSF. It can be noted that in only 2 animals (numbers 1 a n d 2) could C S F be properly sampled in all cases o f probenecid treatment. 1. In untreated animals the m e a n C S F concentration o f H V A was 31 ng/ml and that o f 5 - H I A A , 39 ng/ml. 2. There is a significant correlation between the doses o f probenecid administered and the concentration o f probenecid in plasma ( r = 0.72, P < 0.01, N - - 11). 3. There is a significant correlation between the probenecid doses a n d the levels o f this d r u g in C S F (r = 0.66, P < 0.005, N -- t7). 4. There is a significant correlation between levels o f probenecid in plasma a n d in C S F ( r - - 0 . 9 5 , P < 0.0005, N = 11). A l t h o u g h the best fitting regression model is a 4th degree polynomial, it can be seen in Fig. 1 that the relationship m a y be described usefully using the linear transformation. 5. There are significant correlations between C S F levels o f p r o b e n e c i d a n d o f H V A (r = 0.81, P < 0.0005, N = 17) and o f 5 - H I A A (r = 0.77, P < 0.0005, N = 17). A fifth degree polynomial provides the best fit for the relationship between H V A and probenecid levels

J. P.

Muizelaar and

J. I.

225

Oberink: Probenecid, HVA and 5-HIAA in the Rabbit Table 2

Animal

Lot

1+2 3+4 5+6 1+3 2 1 1 5 2 5 1 2 4 2 6 1 3 5 4 6 3

0 0 0 0 3 1 2 2 1 3 3 6 3 2 1 4 1 4 1 2 3

Dose probenecid Concentration mg/kg body probenecid weight in plasma (gg/ml)

Concentration probenecid in CSF (gg/ml)

-50 50 100 50 100 100 200 400 100 200 200 400 200 200 400 400 400

-

38 23

42 40

-

31

37

--

30 2t0 20O 318 392 27O 340 440 330 647 323 420 369 510 483 530 556 711

39 63 58 17t 170 123 204 4O9 327 184 312 328 287 481 203 569 507 48t

---42.7 53.4 85.4 60.0 91.8 138.7 144.0 200.0 260.0 196.0 534.0

3.7 4.7 6.0 7.0 7.6 10.1 17.2 17.8 18.0 20.8 23.2 26.0 32.6 39.2 41.5 50.0 75.0

Concentration HVA in CSF

Concentration 5-HIAA in CSF

(ng/ml)

(ng/ml)

For explanation see text.

Pr~bn ;ecd iC / SF

/o

they have, however, not been used for the calculation of the correlation and regression.

60

Discussion

50

9

40'

"7

30'

/ r= P=

x

y =-1,8 +0.15/ 0.95

10-

/,",6/~ #

0.0005 N= n

( w

100

200

300

400

500

600

Probenecid/pt Fig. 1. Linear regression of probenecid levels in plasma on probenecid levels in CSF (Fig. 2a). A sixth degree polynomial provides the best fit for the relationship between 5 - H I A A and probenecid levels (Fig. 2b): The linear relationship is included in the graph for reference purposes. Note that in Fig.2a and 2b, the values of H V A and 5 - H I A A levels without probenecid treatment are designated;

The correlation between levels of probenecid and of H V A and 5 - H I A A in CSF has been stressed before (Korfetat., 1971a; Sj6strom, 1972). In these and other studies (Sharman, 1967) attempts have been made to correlate doses of probenecid, levels in plasma and levels in CSF of this drug. The fact that results using h u m a n subjects are inconsistent m a y be partly due to the use of patients and volunteers, forming heterogeneous samples with respect to sex, diagnosis, age and other factors. Perhaps work with laboratory animals, m a y provide results that are more readily reproduceable. Values of H V A in the CSF of untreated rabbits, that we obtained, do not greatly differ from those of Andersson et al. (1972, 1973a). Our values of 5 - H I A A are, however, considerably lower. This is probably due to breed differences. A variety of breed specific values have been reported: 39 ng/ml in the New Zealand white rabbit (present study), 68 ng/ml in the AlaskaF1 bastard rabbit (Muizelaar, 1974), 122 ng in the mongrel rabbit (Andersson et aL, 1973a) and 203 ng/ml in the white domestic rabbit (Andersson et al., 1973 a). Breed differences, however, do not explain all the

226

Psychopharmacologia (Berl.), Vol. 43, Fasc. 3 (1975)

HVA/CSF ng/ml 700"

600"

500'

400"

300" e9

200"

y=271+.6x r =0.81 P