5-Hydroxytryptophol in human cerebrospinal fluid: Quantitative determination by gas chromatography-mass spectrometry using a obliterated internal standard

55

Clinica Chimica Acta, 84 (1978) 55-62 @ Elsevier/North-Holland Biomedical Press

CCA 9081

5-HYDROXYTRYPTOPHOL IN HUMAN CEREBROSPINAL FLUID: QUANTITATIVE DETERMINATION BY GAS CHROMATOGRAPHYMASS SPECTROMETRY USING A DEUTERATED INTERNAL STANDARD

SABURO TAKAHASHI *, DAMODAR and HARVEY C. STANCER

D. GODSE, AHMAD NAQVI, JERRY J. WARSH

University of Toronto, Clarke Institute of Psychiatry, Section on Neurochemistry, 250 College Street, Toronto, Ontario M5T lR8 (Canada) (Received July 27th, 1977)

Summary Procedures for the quantification of 5-hydroxyindole-3-ethanol, or Fi-hydroxytryptophol (5-HTOL), in human cerebrospinal fluid are described. 5-HTOL was determined as its di-pentafluoropropionyl derivative. Deuterium labelled 5-hydroxyindole-3-ethanol-a,a,b,b-d, (5-HTOL-&) was used as internal standard. Mass fragmentography was performed by double ion monitoring each for 5-HTOL and 5-HTOL-d4 and their ratios were determined for specificity. Assay sensitivities of 0.15 ng/ml were achieved using 2.0 ml of cerebrospinal fluid. Free 5-HTOL concentrations in human cerebrospinal fluid were determined to be 0.73 ? 0.44 ng/ml (mean t S.D.) (range 0.33-2.11 ng/ml) from 15 patients with various neurological disorders, and 0.85 + 0.30 ng/ml (range 0.48-1.32 ng/ml) from 9 subjects who complained of low back pain but did not show signs of neurological illnesses.

Introduction A possible derangement of brain 5-hydroxytryptamine (5-HT) metabolism in depressive illnesses has been proposed over a period of twenty years, since the discovery that reserpine, a drug which lowers its concentration in the animal brain, causes depression in‘ some patients [ 11. This hypothesis has been supported by a study on the brains of depressive suicides, indicative of a decrease in brainstem concentrations of 5-HT or its major metabolite, Fj-hydroxyindole3-acetic acid (5-HIAA) [2], followed by a series of reports on decreases in cerebrospinal fluid (CSF) concentrations of 5-HIAA in the depressed patients [ 3-51. However, the data which have emerged in CSF studies are still equivocal * Present address and address for correspondence: Prefectural University of Medicine. Kawaramachi

Department of Psychiatry and Neurology. and Hire-koji, Kyoto, 602, Japan.

Kyoto

56

[6]. It is possible that biogenic amine metabolites found in CSF might be irrelevant to the events present in the CNS. Another source of uncertainty lies in the lack of specificity of the analytical methods that have been used. Several papers describe a method for the quantitative analysis of indole amine metabolites such as 5-HIAA and indole-3-acetic acid (IAA) in CSF by a gas chromatographic-mass spectrometric (GC-MS) technique [7-lo]. On the other hand, only one paper has described a GC-MS method for measuring 5-hydroxyt~ptophol (5HTOL), a major neutral metabolite of 5HT, in human CSF [ 111. This paper reports for the first time the occurrence and actual concentrations of 5-HTOL in human CSF, employing &fluoro-a-methyltryptamine hydrochloride as internal standard. However, the need for deuterated 5-HTOL as internal standard has been strongly suggested in order to establish the specificity of the analytical procedures. Very recently, tetradeutero-&hydroxytryptophol (5-~TOL~~, R-C2~*-CzH~-OH) was successfully synthesized in our laboratory and this prompted us to develop assay procedures for 5-HTOL concentrations in various biological samples. In this communication a GC-MS method is described for the assay of 5-HTOL in human CSF. This method has achieved satisfactory specificity as well as sensitivity, requiring 2 ml of CSF. It is expected that this method will be a new tool for the study of 5-HT metaboiism in affective disorders. Materials and methods Standards and reagents A stock solution of 5-HTOL (Regis chemical Co,, Morton Grove, Ill.) was prepared in a concentration of 100 pg/ml in ethyl acetate. It was found to be stable at 4°C for at least six months. Working standard solutions, 100 ng/ml in redistilled water, were prepared every four weeks by diluting the stock solution. Ethyl acetate (Pesticide Grade, Fisher Scientific Co., Fair Lown, N.J.) and pentafluoropropionic anhydride (PFPA) (Pierce Chemical Co., Rockford, Ill.) were used without prior purification. All other reagents were of analytical grade. In ternal standards Deuterium labelled 5-hydroxyindole-3-ethanol-a,a,b,b-d4 was synthesized from 5-hydroxyindole-3-acetic acid-a,a-d, ( 5-HIAA-d2, Merck, Sharp and Dohme, Canada, Ltd., Isotope Division, Pointe Claire, Quebec) in our laboratory. In brief, methyl ester of 5-HIAA-d2 was prepared and then was reduced to 5-HTOL-d& with deuterated borane methyl sulfide. The isotopic purity of the product was assessed by converting the product to its pentafluoropropionyl and trifluoacetyl derivatives and analysing on GC-MS. The 5-HTOL-d, employed for this analysis was 99.7% enriched. The concentrations of 5-HTOL-d4 were evaluated by comparison with known concentrations of the respective protium compound. It was finally dissolved in redistilled water to make a concentration of 200 ngfml, and stored in a refrigerator. CSF samples Remainders of lumbar CSF samples after laboratory tests were supplied by courtesy of the Biochemistry Laboratory, Toronto General Hospital. They were obtained from inpatients admitted to the neurology ward. CSF samples

57

were also obtained from the Department of Radiology, Toronto General Hospital. These patients were examined by lumbar tap, because of their complaints of severe low back pain, and their medical examination indicated that no significant neurological disorder was present. CSF samples were collected between 1O:OO a.m. and 2:00 p.m. by routine lumbar puncture procedures, and kept in the refrigerator for 2-3 h. To 2 ml of each CSF sample were added 2 mg of cysteine - HCl, and the mixture was stored at -20°C until assayed. Analytical 2 ml of CSF were introduced into a stoppered centrifuge tube containing 2 mg of cysteine - HCl, and 0.2 ml of pooled serum (mixture prepared from human blood samples for laboratory tests, kept at -20°C in aliquots, which contained no detectable amount of 5HTOL by these assay procedures) and 10 ng of 5-HTOLd4 were added to each sample. The mixture was saturated with NaCl, then the pH was adjusted to 8.0 with 0.05 M NaOH and HCl. The samples were extracted successively with 3-ml and 2-ml portions of ethyl acetate, with 7 mm shaking each time and centrifuging for 5 min at 2500 X g. The organic layers were transferred to a tube containing 0.5 g Na2S04, and the combined mixture was evaporated in a l-ml reactival under nitrogen until the entire extract was reduced to dryness. Each residue was treated with 0.2 ml of PFPA/ethyl acetate mixture (1 : 1, v/v) by heating at 45°C for 45 min. The excess of reagents was evaporated under nitrogen and samples were reconstituted in 15 ~1 of ethyl acetate and analysed by GC-MS. A set of standards, 0, 1, 2, 5 and 10 ng of 5-HTOL, was prepared for each assay in 2 ml of redistilled water containing 0.1% (w/v) cysteine - HCl. The standards were extracted and derivatized as described above. GC-MS Analyses were performed with a Finnigan Model 3200 GC-MS system, a combined gas chromatograph-quadrupole mass spectrometer equipped with programmable multiple-ion monitors. Separation of di-pentafluoropropionyl5-hydroxytryptophol (di-PFP-5-HTOL) was achieved on an isothermal glass column, 1.5 m long, 2 mm i.d., packed with 3% OV-17 on Gas Chrom Q, lOO120 mesh (Applied Science Laboratories Inc., State College, Penn.), at 180°C with a carrier gas (helium) flow rate of 20 ml/min. The injector and separator temperatures were 250°C and 23O”C, respectively. The mass spectrometer trap current was 0.75 mA and electron energy was 70 eV. Results Mass spectrometry of 5-HTOL The mass spectrum of the di-PFP derivative of 5-HTOL is shown in Fig. 1 along with the proposed fragmentation pattern. The molecular ion (M’) at m/e 469 and other major positive ions at m/e 305 (M’ - C,F&OOH) and m/e 292 (M’ - C,F,COOH-CH). Mass spectrometry of 5-HTOL was carried out by selective monitoring of the ion currents representative of m/e 469 and 305 for 5-HTOL and m/e 473 and 308 for 5-HTOL-&. In case of di-PFP derivative of 5-HTOLd,, fragmentation occurs on the aliphatic chain to split -C,F,COO’H

Fig. 1. Mass spectrum and proposed fragmentation patterns for the di-pentafluoroproPionY1 derivative of 5-HTOL prepared from a reference standard. The instrumental scan range was set from m/c 60 to m/e 475. Relative intensity is expressed in percent to the highest yield of 5-HTOL derivative at m/e 292. m/e, mass-to-charge ratio.

away, so that the latter fragment (AI’- CzF,C002H) loses one atom of deuterium, and hence has the difference of 3 a.m.u. The retention times of the ion current peaks were identical for authentic 5-HTOL standard, for the compound extracted from a CSF sample, and for the CSF sample extract to which 2 ng of 5-HTOL had been added (Fig. 2). The ratios of the ion currents occurring at

B

/I

123

0123

d f’

1 3:

0 123

0 12

3

0123

0

1

i23

MINUTES

Fig. 2. Mass fragmentograms obtained from the di-PFP derivatives of: (A) 2 ng of reference 5-HTOL; (B) material extracted from human cerebrospinal fluid; (C) material extracted from human cerebrospinal fluid to which 2 ng reference 5-HTOL has been added. The ion currents at m/e 469 and m/e 305 were monitored for protium 5-HTOL and at m/e 473 and m/e 306 for 5-HTOL-dq. Retention time was 152 set at the column temperature, 1 80°C.

59 TABLE

I

SPECIFICITY

OF CSF 5-HTOL

Data are expressed Fragment

BY GC-MS

DETERMINATION

as the mean + S.D. (N = 12).

ratio

Standard

2 ng

CSF samples

m/e 469 m/e 305

0.144

+ 0.012

0.147

+ 0.013

m/e 473 * m/e 308

0.185

2 0.007

0.185

f 0.001

* 10 ng of 5-HTOL-dq

was used as internal standard.

m/e 469 and m/e 305 were identical for the authentic 5-HTOL standards and the derivatized CSF extracts, which secured the specificity of the assay (Table I). Linear calibration curves were obtained for the peak height ratio of standard (H) to internal standard (D) vs. concentration for both the ion current at m/e 469/m/e 473 and that at m/e 305/m/e 308 (Fig. 3). In these graphs, correction was made for the slight errors in accuracy which are attributable to the presence of protium 5-HTOL in the synthesized 5-HTOL-d4. Routinely this was substituted by intercept of the regression line, which was ‘obtained without making any correction and was found to be 1%. Based on the intensities of ion currents, the limit of accurate quantification was calculated as 1.0 ng in a sample when monitoring the M’ ion at m/e 469 (where the signal-to-noise ratio of the ion currents was more than 5 : 1). However, monitoring the M’ - CzFSCOOH ion at m/e 305 gave a current of sufficient magnitude for quantification of 5-HTOL to the level of 0.2 ng. In the actual determination, concentrations of as little as 0.3 ng of 5-HTOL in a sample, or 0.15 ng/ml using 2 ml of CSF, were measurable with good accuracy and reproducibility (Fig. 3). The overall yield of added standard, 2 ng, was determined using internal standard which was added before and after the extraction procedures. It was found to be

5-HYDROXYTRYPTOPHOL

lng)

Fig. 3. Standard curve for the quantitative determination of L-HTOL. Solid circles fragment peak height ratios (H/D) at m/e 469/m/e 473 and open circles (04 m/e 308 versus 5-HTOL concentration (0.3-10 ng).

(0~ ) those

) represent at m/e 3051

60 TABLE II CSF

LEVELS

OF 5-HTOL

IN PATIENTS

WITH NEUROLOGICAL

DISORDERS

AND

WITH LOW

BACK PAIN Group and subject

B-HTOL

Group and subject

(m/ml) Neurological disorders

5-HTOL (w/ml)

Low back pain

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

0.33 0.14 0.49 0.69 0.53 1.26 0.39 0.79 2.11 0.66 0.84 0.49 0.54 0.46 0.60

Mean ? S.D.

0.73 c 0.44

1.32 0.48 0.73 1.30 0.66 0.95 0.56 0.90 0.72

Mean r S.D.

0.85 i 0.30

71.5% + 3.3 (mean + S.D., N = 12). Assay recovery of added standard taken through the procedures with pooled CSF was 102.1 + 5.3% for 2 ng (mean C S.D., N = 13). Quantification of CSF levels of 5-HTOL The concentration of 5-HTOL in human CSF are given in Table II. Free 5-HTOL levels in lumbar CSF were in the range of 0.33-2.11 ng/ml. There was no significant decrease in the 5-HTOL concentrations in CSF from patients with neurological disorders, as compared to those from patients with low back pain who were supposedly free of neurological abnormalities. After hydrolysis with sulfatase at 37°C for 20 h, no detectable increase in the concentrations of 5-HTOL was observed. Since the presence of various amine metabolites in glusulase was identified by means of mass fragmentography [ 121, we have examined the content of 5-HTOL in sulfatase (Type H-l, 18600 units/g, Sigma Co., St. Louis, MO.). The amount of enzyme used (1 mg) for hydrolysis of 2 ml CSF contained recognizable quantities of 5-HTOL (0.13 ng/mg sulfatase). 5-HTOL in CSF was found to be stable during storage at -20°C for at least 4 weeks. A series of pooled CSF samples carried through the whole procedures gave the mean f S.D. as 0.504 ng/ml + 0.055 (N = 7), and the coefficient of variation was calculated as 10.9%. Discussion Occurrence of free 5-HTOL in human CSF was demonstrated for the first time by Curtius et al. in 1975 [ll]. In their analytical procedures, Fi-fluorocY-methyltryptamine was used as internal standard. They suggested that deuter-

61 TRI-PFP-5HTOL

DI-PFP-5HTOC

75%

45%

6

6DQC

6% 45%

75% 'Y

_ 15 30

w

90

120 MINUTES

Fig. 4. Formation of di-PFP and tri-PFP derivatives of S-HTOL after incubation at different temperatures: ) and 75’C (A). Ordinate represents relative amount of PFP 45% ,(0-----o ). 60°C (.derivatives formed in reference to PFP derivations of 5-HTOL-dq, which were added after incubation. Abscissa represents incubation time in minutes.

ated 5-HTOL is preferable for the assay, particularly for the quantitative determination of conjugated 5-HTOL. At that time, however, this compound was not available. Deuterium-labelled 5-HTOL, which has been synthesized in our laboratory, was used in the present study. Enrichment of the deuterated compound was satisfactorily high for employment of this compound as internal standard for the assay, securing its specificity with double-ion monitoring (Fig. 2, Table I). Thus, occurrence of free 5-HTOL has been confirmed by us. For the quantitative determination of 5-HTOL on GC-MS, either the di-PFP or tri-pentafluoropropionyl (tri-PFP) derivative of 5-HTOL can be used. Both were found to be useful and allowed multiple ion detection. Ion currents at m/e 469, the molecular ion of di-PFP-5-HTOL, were 4-5 times stronger than those at m/e 438 and 451, major positive ions of tri-PFP-5-HTOL. Therefore, the di-PFP derivative was preferable for the assay of a small amount of 5-HTOL. When incubated with PFPA, both di-PFP-5-HTOL and tri-PFP-5-HTOL were formed at different rates according to the different incubation temperatures and times. Formation of di-PFP-5-HTOL was found to be greatest when incubated at 45°C for 30-60 min (Fig. 4). On the other hand, the higher the temperature and the longer the time, the more tri-PFP-5-HTOL was formed. Longer heating is required for complete derivatization of the indole compounds

1131. Mass fragmentograms of di-PFP-5-HTOL were not interfered with by those of tri-PFP-5-HTOL, which were also largely formed in the derivatized specimens, as an ample interval of retention times achieved clear separation of these two (relative retention time 3.2 : 1) when injected on the column. One of the properties of di-PFP-5-HTOL convenient for routine procedures was that it remained stable for 2 weeks when kept in the refrigerator. Trifluoroacetyl (TFA) derivatives appeared to be less suitable for analysis by GC-MS. Their response was considerably lower than that of the PFP derivatives. Protection of 5-HTOL against oxidative decomposition during the extraction

62

procedures is another important factor for increasing the sensitivity of the assay procedure. We have examined several reagents and materials, and found addition of human serum gave the best yield of 5HTOL from the extraction procedures. Thioglycolic acid was also a good preservative of 5-HTOL. Other antioxidative agents such as ascorbic acid or cysteine were not satisfactory for this purpose when used alone. Our data for the concentrations of 5-HTOL in human CSF are, as a whole, comparable to values reported previously [ll]. These measurements of 5HTOL using GC-MS method gave a range of 0.1-5.8 ng/ml using 1 ml of CSF samples in 7 younger subjects with neurological disorders. They obtained the data from two boys suffering from leukaemia who showed very high values of 5-HTOL (10.2 and 33.4 ng/ml). Our data showed that adult neurological patients had lower 5-HTOL concentrations in a range of 0.33-2.11 ng/ml (Table II). If 5-HTOL concentrations in a group of patients with low back pain are regarded as representing normal levels, 5-HTOL in CSF may not be affected by some neurological disorders, though further study is necessary on this point. Our data indicate that the concentrations of 5-HTOL, the reduced catabolite of 5-HT, are considerably lower in human CSF than those of 5-HIAA, which have been reported to be around 20-60 ng/ml [3-6]. This suggests that human brain 5-HT may preferentially be metabolized to 5-HIAA rather than to 5-HTOL. The ratio of 5-HTOL : 5-HIAA measured in CSF or in the brain may be an index of derangement of brain 5-HT metabolism. It would be intriguing to see whether or not the 5-HTOL : 5-HIAA ratio is altered in affective disorders. Acknowledgements We are grateful for the skilful technical assistance of Mrs. M. Lee in performing the gas chromatography-mass spectrometry. Dr. G. Wortzman, Department of Radiology, and Mr. G. Bardeleben, Biochemistry Laboratory, Toronto General Hospital, kindly supplied the CSF samples. References 1 2 3 4 5 6 7 8 9 10 11 12 13

Pletsher, A., Shore, P.A. and Brodie, B.B. (1956) J. Pharmacol. Exp. Ther. 116, 84-90 Shaw, D.M., Camps, F.E. and Eccleston. E.G. (1967) Br. J. Psychiatr. 113,1407-1411 Ashcroft, G.W. and Sherman, D.F. (1960) Nature 186,1050-1051 Van Praag, H.M. and Korf, J. (1971) Psychopharmacologia 19,148-152 Coppen, A., Prange, Jr., A.J., Whybrow, P.C. and Noguera. R. (1972) Arch. Gen. Psychiatr. 26, 47447a Schildkraut, J.J. (1973) Ann. Rev. Phannac. 13, 427-454 Bertilsson. L., Atkinson, Jr.. A.J., Althaus, J.R., Htifast, A., Lindgren, J.-E. and Holmstedt, B. (1972) Anal. Chem. 44.1435-1438 Bertilsson, L. and Palmer. L. (1972) Science 177, 74-76 Fri. C.-G., Wiesel, F.-A. and Sedvall, G. (1974) Life Sci. 14.2469-2480 Swahn, C.-G., Sandgiirde. B., Wiesel, F.-A. and SedvaB, G. (1976) Psychopharmacology 48.147-152 Curtius. H.C., Wolfensberger. M., Redweik, U., Leimbacher. W., Maibach. R.A. and Isler, W. (1975) J. Chromatogr. 112, 523-531 Gordon. E.K.. Oliver, J. and Kopin. I.J. (1975) Life Sci. 16, 1527-1532 Godse, D.D., Warsh. J.J. and Stancer. H.C. (1977) Anal. Chem. 49.915418