| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1
Hygiene and Public Health Laboratory, Pharmacy Faculty, 75006 Paris, France
2
Biochemical Laboratory, Georges Pompidou Hospital, 75015 Paris, France
aaddress correspondence to this author at: Faculté de Pharmacie, 4 Avenue de l’Observatoire, 75006 Paris, France; fax 33-1-4325-3876, e-mail Isabelle.Momas{at}pharmacie.univ-paris5.fr
Nicotine and its metabolites (1), expired carbon monoxide, and thiocyanates (2) are the most widely used smoking biomarkers. Among these biomarkers, urinary cotinine has been one of the most representative and specific for tobacco smoke exposure (3)(4)(5) with regard to active or passive smoking. The methods most frequently used for cotinine quantification are gas chromatography (6) and HPLC (7), coupled or not with mass spectroscopy (8)(9). These methods, however, are difficult to use in large-scale epidemiological studies because they require specialized laboratories. In 1973, Langone et al. (10) proposed the assessment of cotinine by RIA, but RIAs also require specialized laboratories. This last technique was then extended to ELISA (11) and fluorescence polarization immunoassay (12).
Recently, an enzyme immunoassay (EIA) that is easier to perform (13) was developed to measure cotinine concentrations between 100 and 2000 µg/L, a range that exceeds concentrations observed in passive smoking. We thus propose an adaptation and automation of this EIA to assess urinary cotinine concentrations <100 µg/L to detect passive smoking. After analytical validation, this technique was applied to subjects exposed and not exposed to environmental tobacco smoke (ETS).
All chemicals, unless specifically noted, were from Prolabo. The cotinine assay (PROCLAIM; Servibio) is a liquid homogeneous EIA based on the competition of a cotinine-labeled enzyme, glucose-6-phosphate dehydrogenase, and the free cotinine in the urine for a fixed amount of cotinine-specific antibody binding sites. Glucose-6-phosphate dehydrogenase activity was determined spectrophotometrically at 340 nm by measuring its ability to convert NAD+ to NADH. Nicotine metabolites and various organic compounds were tested for cross-reactivity by the manufacturer (13); 3'-hydroxycotinine has a 50% cross-reactivity, but at a high concentration of 250 mg/L, a concentration not found in passive smoking.
Cotinine was measured on a Beckman CX7 automated analyzer, although
other automated analyzers (open systems) can presumably be used. The
analysis rate was 80 samples per hour. Briefly, 7 µL of
urine or calibrator was added to 210 µL of reagent A
(antibody/substrate reagent). Seventy microliters of reagent B (enzyme
conjugate reagent) was dispensed after a 500-s delay and incubated for
196 s. Spectrophotometric measurement was then performed at 340 nm
during a 376-s period (the use of a secondary wavelength at 650 nm is
recommended). The change in absorbance per minute was calculated,
and a four-parameter log-logit function was used to establish the
calibration curve. Each series of measurements needs to be calibrated.
The calibration curve (Fig. 1
) was constructed from a stock 20 mg/L methanolic calibrator
cotinine solution (Sigma), with final working solutions of 5, 15, 30,
60, 75, 150, 300, and 600 µg/L in pH 7 buffer. Data are expressed as
the mean ± SD.
|
To increase the sensitivity of the method, urines were concentrated fivefold using SepPack microcolumns (Waters) according to the following procedure: 1 mL of urine was applied to the preconditioned column and then eluted from the column with 0.2 mL of methanol. The eluate was evaporated under a nitrogen stream at 37 °C and redissolved in 0.2 mL of 67 mmol/L phosphate buffer, pH 7. Average recovery was >90%.
The influence of the matrix was studied by comparing the calibration
curve in pH 7 buffer and in a fivefold concentrated pooled urinary
sample from five nonsmokers not exposed to ETS. Each calibrator was
measured in triplicate. The Pearson correlations were r
= 0.99 in the range 0–600 µg/L and r = 0.97 in the
range 0–75 µg/L. The slopes of the calibration curves prepared using
pH 7 buffer or concentrated urine, although not strictly identical,
were not statistically different (Fig. 1
), which showed that
calibration in pH 7 buffer can, from an epidemiological point of view,
be used in routine practice.
The detection limit (mean + 3 SD from 20 determinations of the zero calibrator, pH 7 buffer) was 1.7 µg/L. The limit of quantification (three times the detection limit) was 5.2 µg/L, which actually represents 1 µg/L because the urine was concentrated fivefold. The within-run imprecision (CV) was, respectively, 20%, 19%, 18%, 13%, and 11% at cotinine concentrations of 5, 15, 30, 60, and 75 µg/L (n = 20).
The cotinine EIA was applied to urine samples from 53 Paris schoolchildren: 18 exposed to ETS (their exposure, as evaluated by questionnaire, being >10 cigarettes during 48 h before measurement), and 35 not exposed to ETS. This study was approved by the Human Investigation Committee, and informed consent was obtained from all children and parents. Morning urines (14) representing the exposure over the previous hours (15) were collected and stored at -20 °C until analysis (14). Creatinine was assessed in urine according to a kinetic method adapted from the Jaffe method (16). Urinary cotinine results, calculated taking into account the fivefold urine concentration, were expressed with reference to creatinine values. Results were compared using the Student t-test.
Urinary cotinine values in children exposed to ETS (14.1 ± 11.9 mg/mol of creatinine) were significantly higher (P <0.001) than those in nonexposed children (5.1 ± 5.4 mg/mol of creatinine). The proposed cotinine EIA thus enables nonsmokers exposed to ETS to be differentiated from those not exposed. The technique is simple, rapid, can be used in any laboratory having an automated analyzer (open system), and can be applied to large-scale studies.
This method can be recommended in epidemiological studies as a control for smoking, an important confounding factor often poorly described in self-administered questionnaires (17).
Acknowledgments
This study is part of the French multicentric study VESTA (Five Epidemiological Studies on Transport and Asthma) coordinated by Dr. D. Zmirou and supported by a grant from the French Ministry of Environment. We are grateful to Prof. P. Beaune for advice.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
DO, change
in absorbance at 340 nm.