Improved radioimmunoassay of melatonin in serum

CLIN. CHEM. 33/4, 604-605 (1987) Improved Radioimmunoassay of Melatonin in Serum W. Sleghart E. Ronca, G. Drexier, and S. Karall ton, MA) and incubated this with 150 tL of an antimelatonin antibody solution (WHB). After 1 h at 37 #{176}C, the samples were transferred to an ice bath and 400 p.L of icecold saturated ammonium sulfate solution was added. After 1 h at 4#{176}C, samples were diluted with 3.5 mL of ice-cold, 70% saturated ammonium sulfate solution, then ifitered through Whatman GF/B glass-fiber ifiters. We washed the filters once with 3.5 mL of the 70% saturated ammonium sulfate solution and then measured, by liquid scintillation counting, the radioactivity remaining on the ifiters. Additional Keyphrases: samplepreparation reversed-phase circadianrhythms The efficiency of the column extraction procedure was assessed by adding 34 pg of [3H]melatomn to 0.6 mL of serum and applying the mixture to Baker RP-18 columns. After the columns had been centrifuged for 3 mm at 80 to 100 x g, >99% of the applied [3H]melatomn was retained by the columns, only about 0.3% passing through them. In other experiments not shown here we found that, even after application of 5 mL of serum or 10 ng of melatonin, only 0.5% of the applied [3H]melatornn radioactivity passed through the columns. More than 5 mL of serum tended to clog the columns. We could prevent clogging by either using adequate ifiter discs on the top of the columns, precipitating the serum proteins before applying samples to the columns, or digesting the serum proteins with trypsin. Because the columns had to be washed to remove unbound proteins and lOWMr compounds in serum, we investigated various washing procedures that did not interfere with binding of [3H]melatonin to the Baker RP-18 columns-e.g., up to five washes with 1 mL of distilled water or one wash with 1 mL of pure n-hexane or 1 mL of aqueous methanol (100 mJ.JL). In other experiments, columns were washed with a 100 mLfL aqueous methanol that had been adjusted with buffer to pH 3.0 or 9.0. Under all these conditions >97% of the [3H]melatomn was still retained by the columns. Because the concentrations of melatonin estimated with use of 100 mL/L methanol as a washing fluid seemed to be slightly lower than those after the distilled water washes, we used a single 1-mL wash with 100 mIJL aqueous methanol for routine determination of melatomn in serum. Pure ethanol, methanol, or acetone completely eluted melatonin from the columns. However, because of the expense of pure ethanol and because a contaminant in commercial acetone interferes with radioimmunoassay of melatonin, we routinely used 300 ,uL of methanol as eluent, which rapidly and reproducibly eluted 97.4% (SD 0.8%) of the applied [3H]melatonin. The used columns could be reused at least 10 times without further purification, if pretreated with pure methanol and water as described above in Materials and Methods. To determine melatomn in the serum extract, we used a method originally described by Wetterberg et al. (4) and recommended by WFIB for use with their melatonin radioimmunoassay kit. We replaced the centrifugation with a ifitration step involving Whatman GFIB glass-fiber filters. This not only shortened the time necessary for the radioim- . chromatography, Because of the extremely low concentrations of melatonin in serum and the interference of serum components with the radioiminunoassay for melatonin, most studies performed so far have involved extraction of melatonin from serum before analysis (1). However, the usual liquid-liquid extraction is not only incomplete but also time consuming. Therefore, we investigated the usefulness of a rapid column-extraction method, such as has been developed for various other compounds (2, 3), for the extraction of melatonin. In addition, we modified a commercially available radioimmunoassay for melatonin to improve reproducibility and simplicity. Materials and Methods To extract the melatonin, we used Baker-lO SPET’ RP-18 columns (J. T. Baker Chemical Co., Phillipsburg, NJ 08865) pretreated by washing twice with 1 mL of pure methanol, and then twice with 1 mL of distilled water. These washing steps are absolutely essential for the correct function of the columns. After applying 0.1 to 5 mL of serum, we washed the columns with 1 mL of aqueous methanol (methanoll water, 10/90 by vol), then eluted melatonin from the columns by applying 300 L of pure methanol. The spontaneous flow of solvents or serum through the columns was rather low, but when we tried to speed the flow by applying suction, as recommended in the respective extraction protocols (2, 3), the proportion of melatonin extracted varied. Because it was essential to keep the columns wet throughout the washing and extraction procedure, we routinely centrifuged the columns at 80 to 100 x g for 3 miii to improve the flow rate of solvents or serum through them. After evaporating the methanol from the samples we determined the melatonin by using a modification of a commercially available radioimmunoassay (WHB, Bromma, Sweden). We redissolved the melatonin residue in 250 iL of sodium phosphate buffer (50 mmol/L, pH 7.5) containing, per liter, 154 mmol of NaC1, 15 mmol of sodium azide, 1 g of gelatin, 1 g of gamma-globulin, and 0.25 nmol of [3H]melatonin (76.3 kCilmol; New England Nuclear, BosPsychiatrische Universitataklinik, Allgemeines Krankenhaus der Stadt Wien, Wahringer GUrtel 18-20, 1090 Vienna, Austria. Received September 29, 1986; accepted January 27, 1987. 604 CLINICALCHEMISTRY, Vol. 33, No. 4, 1987 Results Downloaded from https://academic.oup.com/clinchem/article/33/4/604/5653239 by guest on 29 November 2021 Melatonin was extracted from serum by using Baker reversed-phase C-i 8 columns. More than 99% of the applied melatonin was retained by the columns, and more than 97% was eluted from the columns in 300 pt of methanol. We then determined melatonin in the serum extract by a modification of a standard radioimmunoassay, using filtration instead of centrifugation to collect the [3H]melatonin-antibody complex precipitated by saturated ammonium sulfate. These modifications allow more rapid, accurate, and reproducible determination of melatonin than do previously published procedures. Table 1. Accuracy and Precision of the Modified R1Afor Determining Concentrations of Melatonin In Seruma Concn In serum, nWL 22.4 34.9 62.4 102.4 Within run (n = 4) Mean, 22.0 33.5 64.5 108.0 First concentration Mean, CV, % ng Betweenrun(n = 11) 8.8 8.5 6.8 5.0 is for unsupplemented nL 22.4 34.0 66.1 106.0 CV, % 11.6 11.8 7.6 serum, others had added melatonin. Discussion This rapid column-extraction method for extracting melatonin from serum is very reproducible and yields a high proportion of applied melatonin in the column eluate. In easily evaporated than the relatively large volumes involved in liquid-liquid extraction of this compound. By using the described method, one can extract 40 samples within 90 miii and assay them within 3 to 4 h, with better reproducibility and accuracy than for other melatomn determinations. Thus, it is easily possible to assay melatonun in 40 samples every day. References 1. Arendt J. Assay of melatonin and its metabolites: results in normal and unusual environments. In: Wurtman R, Waldhauser F, eds. Melatomn in humans (Proc. First hit. Conf. on Melatonin in Humans). Vienna, Austria: Center for Brain Sciencesand Metabolism Charitable Trust, 1985:11-32. 2. Good TJ, Andrews JS. The use of bonded-phase extraction columns for rapid sample preparation of benzodiazepines and metabolites from serum for HPLC-analysis. J Chromatogr Sci 1981;19:562-6. 3. Application Note for Bond Elut#{174} extraction methods. AnalytichemInternational, Harbor City, CA 90710. 4. Wetterberg L, Eriksson 0, Friberg Y, Vangbo B. A simplified radioimmunoassay for melatonin and its applicationto biological fluids. Preliminary observations on the half life of plasma melatonm. Clin Chim Acta 1978;86:169-77. CLINICALCHEMISTRY, Vol. 33, No. 4, 1987 605 Downloaded from https://academic.oup.com/clinchem/article/33/4/604/5653239 by guest on 29 November 2021 munoassay but also decreased blank values, thus improving the accuracy and reproducibility of the assay. Using this extraction and radioimmunoassay procedure for melatornn, we recovered between 96 and 104% of applied melatonin in various experiments, independently of the volume of serum or buffer applied to the columns. The intra-assay CV was between 5 and 9%, the interassay CV between 5 and 12% (Table 1). Using the described method, we investigated circadian rhythms of melatomn concentrations in serum from 15 patients of the Psychiatric University Clinic in Vienna (experiments to be published elsewhere). The lowest concentrations (by day) were 5 to 21 pg/mL (mean ± SD: 13.5 ± 4 pg/mL), and the greatest values (during the night) were 31 to 180 pg/mL (86 ± 45 pglmL). addition, this extraction procedure is versatile. It can easily be modified by using various solvents or buffer solutions as washing media for removing from the columns compounds that might interfere with subsequent assay of melatonin. This column-extraction method thus might be beneficial for the quantification of melatonin, not only by radioimmunoassay but also by gas chromatography-mass spectrometry or by liquid-chromatographic methods. Finally, the relatively small volume of organic solvent needed to elute the melatonin from the columns can be more