Validation of an ELISA for urinary dopamine: applications in monitoring treatment of dopamine-related disorders



Dopamine is a catecholamine that serves as a neurotransmitter in the central and peripheral nervous system. Non-invasive, reliable, and high-throughput techniques for its quantification are needed to assess dysfunctions of the dopaminergic system and monitor therapies. We developed and validated a competitive ELISA for direct determination of dopamine in urine samples. The method provides high specificity, good accuracy, and precision (average inter-assay variation < 12%). The analysis is not affected by general urinary components and structurally related drugs and metabolites. The correlation between ELISA and LC-MS/MS analyses was very good (r = 0.986, n = 28). The reference range was 64–261 μg/g Cr (n = 64). Week-to-week biological variations of second morning urinary dopamine under free-living conditions were 23.9% for within- and 35.5% for between-subject variation (n = 10). The assay is applied in monitoring Parkinson's disease patients under different treatments. Urinary dopamine levels significantly increase in a dose-dependent manner for Parkinson's disease patients under l-DOPA treatment. The present ELISA provides a cost-effective alternative to chromatographic methods to monitor patients receiving dopamine restoring treatment to ensure appropriate dosing and clinical efficacy. The method can be used in pathological research for the assessment of possible peripheral biological markers for disorders related to the dopaminergic system.

Abbreviations used

attention-deficit hyperactivity disorder


bovine serum albumin






3,4-dihydroxyphenylacetic acid


homovanillic acid




limit of blank


limit of detection


limit of quantification






Parkinson's disease


selective dopamine reuptake inhibitors

The catecholamines (dopamine, norepinephrine, and epinephrine; see Fig. 1), are a class of neurotransmitters and hormones that play a key role in the regulation of many physiological processes in both the central and peripheral nervous system. They are involved in the pathophysiology of many neurological, psychiatric, endocrine, and cardiovascular diseases (Eisenhofer et al. 2004; Goldstein 2010; Tayebati et al. 2011). A number of studies suggest that catecholamines are also key players in the modulation of immune responses (Bergquist et al. 1998; Franco et al. 2007; Haroon et al. 2011).

Figure 1.

Chemical structures of catecholamines and related tyramines.

The dopaminergic system is thought to affect a wide range of behaviors and functions, including cognition, motor activity, pleasure/reward, memory, eating and drinking behaviors, neuroendocrine regulation, and selective attention (Dalley and Everitt 2009). A deficiency in dopamine can cause impairment in these biological functions. In addition, abnormalities in the dopamine system have been associated with movement disorders (Parkinson's disease, Segawa disease), neuropsychiatric disorders (anxiety, depression, attention-deficit hyperactivity disorder (ADHD), drug, and alcohol addiction, suicide), and metabolic diseases (Willner 1983; Naranjo et al. 2001; Kienast and Heinz 2006; Goldstein 2010).

Treatment of dopamine deficiency-related diseases has been monitored by following levels of the major dopamine metabolite homovanillic acid (HVA) in cerebrospinal fluid and serum prolactin (Isaac et al. 2008). Yet, research suggests that the assessment of urinary dopamine might be used as a non-invasive alternative to HVA (Marc et al. 2011). The measurement is relatively easy to obtain, urinary concentrations are higher than those in plasma, and collection procedures are well established and non-invasive. In fact, some studies have evaluated the effects of l-DOPA treatment on urinary dopamine levels in Parkinson's patients (Routh et al. 1971; Dutton et al. 1993; Davidson et al. 2007). It has been demonstrated that after l-DOPA therapy, dopamine levels in the central and peripheral nervous system increase. Furthermore, preliminary evidence suggests that dopaminergic agents have the potential to improve quality of life for patients with major depressive disorder (IsHak et al. 2009; Howland 2012). The measurement of urinary dopamine might be used to monitor patients treated with selective dopamine reuptake inhibitors (SDRIs) in the clinical treatment of ADHD, narcolepsy, fatigue, obesity, mood disorders, anxiety disorders, and drug addiction. The availability of a simple and cost-effective analytical method for urinary dopamine screening would be helpful in assessing the suitability of patient's drug dosing and compliance, in avoiding overdose, and suggesting a change of therapy in case of ineffectiveness.

Several methods have been published on the analysis of dopamine (DA) and related catecholamines in biological fluids. Measurement of catecholamines and their metabolites in plasma and urine are commonly used to aid in the detection and monitoring of neuroblastoma and pheochromocytoma and the evaluation of hypotension or hypertension (Peaston and Weinkove 2004). Traditionally, high performance liquid chromatography (HPLC) combined with electrochemical or fluorescence detection is applied (Peaston and Weinkove 2004; Tsunoda 2006; Pussard et al. 2009). Recently, LC-tandem mass spectrometry is the method of choice because of increased selectivity and sensitivity (Kushnir et al. 2002; de Jong et al. 2010a, 2011; Moriarty et al. 2011). Automated on-line solid phase extraction and subsequent unique MS/MS fragmentation provided the possibility to replace laborious sample preparation, reduce extended LC separation time, and eliminate potential interferences from structurally related metabolites, drugs, and dietary constituents. Despite the technological advances that led to the expansion of the tandem mass methodology in clinical laboratories in the last decade, LC-MS/MS still requires expensive equipment and maintenance, highly qualified personnel, and cannot offer the high throughput provided by immunoassays.

The development and the application of immunoassays for catecholamine detection in biological fluids have been accompanied by difficulties related to their low physiological concentrations, the need of highly specific antibodies and the tendency of the catechol group to be easily oxidized (Knoll and Wisser 1984; Peaston and Weinkove 2004; Kim et al. 2008, 2010). Most of the undertaken approaches are based on the detection of their COMT–metabolites [3-O-methylated catecholamines (metanephrine, normetanephrine, 3-methoxytyramine) and 3-O-methylated acids (homovanilic and vanillylmandelic acid)] which are excreted in higher levels and are more stable. Although some ELISA and radioimmunoassay kits are commercially available, they still have limited clinical use because of complicated and time consuming sample preparation procedures. Free catecholamines are extracted using a cis-diolspecific affinity gel, followed by acetylation and enzymatic conversion (COMT methylation) into their n-acylated metadrenalines prior to immunochemical measurement (Westermann et al. 2002).

We present here the in depth validation of a high-throughput and cost-effective ELISA screening method for direct analysis of urinary dopamine (DA). We also demonstrate its application as a tool for monitoring dopamine concentrations in therapies with dopaminergic agents.

Materials and methods


Unless otherwise stated, all reagents were of analytical grade and purchased from Sigma Chemicals (St. Louis, MO, USA). Metanephrine was obtained from Santa Cruz Biotechnology (Dallas, TX, USA) and methylsynephrine from Maypro Industries (Purchase, NY, USA). l-DOPA (Mucuna cochinchinensis extract) was purchased from Jiaherb, Inc (Parsippany, NJ, USA). Rabbit polyclonal serum was obtained by immunization with l-dopamine (3,4-dihydroxyphenylethylamine) conjugated with glutaric aldehyde to a carrier protein bovine serum albumin (BSA) as reported (Huisman et al. 2010a). Specific polyclonal antibodies were isolated by serum depletion against carrier protein modified by glutaric aldehyde (Huisman et al. 2010a; see Electronic Supplementary Materials and Methods). In this way antibodies specific for the modified carrier protein are removed and the dopamine-specific antibodies are collected. The coating antigen (BSA-dopamine) was BSA coupled to dopamine via glutaric aldehyde linker as described in (Huisman et al. 2010a; see Electronic Supplementary Materials and Methods).

Samples and clinical study populations

Second void morning urine samples (spot samples) were collected in 5 mL tubes containing a filter disc impregnated with 250 μL of 3 mol/L HCl as a preservative. Samples were stored at −20°C until analysis. To evaluate the reference range second void urine samples were obtained from 64 apparently healthy volunteers as determined by the results of the Hopkins Symptoms Checklist-90 (HSCL-90) (Derogatis et al. 1974; Glass et al. 1978). Biological (week-to-week) variation of urinary dopamine was assessed for second morning urine samples collected from 10 individuals (apparently healthy based on the HSCL-90 assessment) once a week for 4 weeks. Diurnal variation was evaluated for 19 subjects who collected spot urine samples every 4 h for a period of 28 h. Urinary dopamine excretion was monitored after l-DOPA intake for 24 volunteers for 24 h. Parkinson's disease (PD) patients were diagnosed by their health care practitioner based on the International Classification of Diseases and Related Health Problems (ICD-9 code 332). In all studies, samples were analyzed in duplicate. Internal quality control materials at low and high levels were also analyzed in every run. All individuals gave their informed consent. The studies were performed following a protocol approved by the internal clinical ethics committee.

Derivatization of standards, controls and samples

Stock solutions for dopamine standards and controls were prepared in 150 mmol/L HCl. Urinary creatinine (Cr) was measured on Roche Integra 400/880 (Jaffe 2 Method) (Madison, WI, USA). Differences in urine excretion volumes were normalized using a volume of urine equivalent to 120 μg Cr. Standards (120 μL), controls (120 μL), and samples (40–300 μL) were diluted up to 300 μL with 150 mmol/L HCl. After neutralization with 60 μL neutralization buffer (1 mol/L NaHCO3 containing 0.3% BSA and 0.1%Tween-20), in situ conjugation was performed with 120 μL glutaric aldehyde (267 mmol/L). After 1 h incubation at 20°C in the dark, the reaction was stopped with 120 μL Tris-quenching buffer (666 mmol/L Tris-HCl pH 7.5; 344 mmol/L NaCl; 0.25% Tween-20; 0.5% BSA; 0.02% sodium azide). The derivatized standards and samples were incubated for 30 min at 20°C in the dark and then applied to the ELISA plates.

Competitive ELISA

Similar procedures have been described for other biogenic amines (Huisman et al. 2010a, b; Nichkova et al. 2012). Briefly, microtiter plates were coated overnight with 0.6 μg/mL BSA-dopamine in coating buffer (50 mmol/L sodium carbonate-bicarbonate buffer pH 9.6). Derivatized dopamine standards, controls, and samples (75 μL/well) were added to the coated plates followed by the dopamine specific antibody (2.5 μg/mL, 75 μL/well). Plates were incubated overnight at 32°C under shaking). After washing, a solution of secondary antibody (goat anti-rabbit IgG-alkaline phosphatase, 150 μL/well) was added and incubated for 60 min (20°C, shaking). Plates were washed after each incubation step six times with 250 μL/well of washing buffer (phosphate-buffered saline pH 7.5, 0.002% Tween-20) using a Tecan 96PW plate washer (Tecan Trading AG, Männedorf, Switzerland). Finally, the substrate was added (150 μL/well). Absorbance was read at 405 nm on a Spectramax microplate reader (Molecular Devices Corp., Sunnyvale, CA, USA). The competitive curves were analyzed with a four-parameter logistic equation using SoftmaxPro v5.4 (Molecular Devices Corp.). Concentration was calculated from the standard curve and expressed as μg dopamine/g Cr.

LC-MS/MS analysis

Urine samples were analyzed by ARUP Laboratories Inc. (Salt Lake City, UT, USA) on LC-MS/MS (Kushnir et al. 2002; see Electronic Supplementary Materials and Methods).

Statistical analysis

General statistical analyses were performed using GraphPad Prism 5 (GraphPad Software Inc, San Diego, CA, USA). Dopamine ELISA performance characteristics (linearity, sensitivity, reference range) were determined using EP evaluator Release 9 (Data Innovations LLC, South Burlington, VT, USA). Linear regression analysis was used to assess the correlation between ELISA and LC-MS/MS analyses. Linear regression and Pearson correlation coefficients were used for sample stability studies and comparison of dopamine levels in 24 h and spot urine. Biological variations [between- (CVG) and within-subject (CVI) variation] were calculated as the variances observed on the analyses of four weekly collections by 10 individuals. A one-way and repeated measures anova were used to compare dopamine levels in the biological (week-to-week) variation and diurnal variation studies. One-way anova and post hoc Tukey multiple comparison analyses were applied in the Parkinson's disease studies. p-values < 0.05 were considered statistically significant.



To evaluate the precision of the dopamine assay two urine samples (low and high dopamine levels) and two in-house controls were analyzed every day in duplicate on a single plate over a period of 20 days. Intra-assay variation (CV%) within plate was < 10% with an average intra-assay variation for samples and controls of 6%. The average inter-assay variation (n = 20) was less than 12% (see Table S1).

Linearity and detection limit

Assay linearity was assessed as suggested in CLSI Protocol EP6 (Clinical and Laboratory Standards Institute 2003). A series of samples of known concentrations (assigned values) were created by sequential mixing of low and high concentration pools of urine. Linearity at specific concentrations was considered acceptable if the % difference between the predicted 1st (linear) and 3rd order regressed values (deviation %) was less than ± 17% (see Fig. 2 and Table S2). The analytical measurement range of the assay was determined to be from 20 to 726 μg/g Cr. Urine samples with highly elevated dopamine levels (> 726 μg/g Cr) were measured accurately at dilutions up to 1 : 32 (for dilutional linearity see Table S3). The limit of blank (LoB) and limit of detection (LoD) were determined in accordance with CLSI EP17-A requirement (Clinical and Laboratory Standards Institute 2004). LoD (95% confidence) is based on a 2SD evaluation of the zero calibrator. LoB and LoD were determined to be 2.5 μg/g Cr and 7.3 μg/g Cr, respectively. The lower limit of quantification (LoQ - lowest concentration for which the 95% confidence interval of the CV% is less than 20%) was 18.8 μg/g Cr.

Figure 2.

Linearity (analytical measurement range). A series of urine samples with known concentrations of dopamine (assigned value) was created by sequential mixing of pools of urine with low and high concentration of serotonin. The linear (analytical measurement) range was from 20.0 to 726.2 μg/g Cr.

Analytical recovery

The analytical recovery (accuracy) was determined by adding dopamine to urine samples with low endogenous amounts of dopamine. The test sample concentrations were between 25 and 11808 μg/g Cr. Each sample was analyzed (spiked and non-spiked) in duplicate and the percent recovery was calculated. The amount of dopamine recovered ranged from 93.2 to 112.4%, and the average recovery was 101.7% (see Table S4).

Analytical specificity and interferences

The specificity of the dopamine antibody was evaluated by running ELISA standard curves of each possible cross-reactant in parallel to the dopamine standard curve. Cross-reactivity values were determined at 50% displacement [(IC50 dopamine/IC50 cross-reactant) × 100]. We tested 24 structurally related compounds, precursors, and/or metabolites related to dopamine, such as epinephrine, norepinephrine, metanephrine, normetanephrine, homovanillic acid, vanilmandelic acid, homoveratic acid, 3, 4-hydroxyphenylacetic acid, 3,4-dihydroxyphenylacetic acid (DOPAC), 3,4-dihydroxyphenylglycol (DHPG), 3-hydroxy,4-methoxyphenylglycol (MHPG), l-3,4-dihydroxyphenylalanine (l-DOPA), methyldopa, carbidopa, octopamine, synephrine, methylsynephrine, tyrosine, tryptamine, tyramine, 3-methoxytyramine, 3-methoxytyramine (3-hydroxy-4-methoxyphenylethylamine), phenylalanine, ß-phenylethylamine. All tested compounds had no cross-reactivity (< 0.01%), except for norepinephrine (4.7%) and l-DOPA (2.1%).

Interference studies were conducted according to CLSI Protocol EP7-A2 (Clinical and Laboratory Standards Institute 2005). Clinically, high concentrations of 47 potentially interfering substances (general urinary components, structurally similar urinary components, drugs and supplements) were spiked in urine samples with known levels of dopamine (67 and 300 μg/g Cr) and assayed along with non-spiked urine samples. The substances presented in Table 1 were found not to interfere at the concentrations indicated (< 15% Bias) (upper 95% confidence interval).

Table 1. Analytical specificity and interferences
  1. a

    Maximum urinary levels in subjects under maximum dose therapy.

Major urinary components
 Ammonium chloride15 mmol/LMagnesium sulfate15 mmol/L
 Barbiturate100 μmol/LOxalic acid0.7 mmol/L
 Bilirubin260 μmol/LProtein600 mg/L
 Calcium chloride10 mmol/LSodium chloride750 mmol/L
 Creatinine5 mmol/LSodium phosphate100 mmol/L
 Glucose200 mmol/LUrea600 mmol/L
 Hippuric Acid4 mmol/LUrobilinogen10 mmol/L
Structurally similar and other urinary components
 Alanine600 μmol/L3-Methoxytyramine800 nmol/L
 3,4-Dihydroxyphenylglycol180 nmol/LNorepinephrine120 nmol/L
 3,4-Dihydroxyphenylacetic Acid3 μmol/LNormetanephrine7 μmol/L
 Epinephrine90 nmol/Lβ-Phenylethylamine32 nmol/L
 Homovanillic acid120 μmol/LTryptamine375 nmol/L
 Metanephrine3 μmol/LTyramine2.5 μmol/L
  Vanilmandelic acid60 μmol/L
Drugs/Supplements a
 Acetominophen6 g/LMethyldopa45 mg/L
 Acetylsalicylic acid6 g/LMethylsynephrine50 mg/L
 Ascorbic acid800 mg/LOctopamine90 μg/L
 Caffeine15 mg/LPhenylalanine11 mg/L
 Carbidopa300 mg/LPhenylephrine100 mg/L
 5-Hydroxytryptophan20 mg/LSalicylic Acid2.5 g/L
 Histidine110 mg/LSynephrine40 mg/L
 Isoetharine3.6 mg/LTheophylline6 mg/L
 Isoprenaline90 μg/LTryptophan90 mg/L
 l-DOPA45 mg/Ll-Tyrosine4.8 g/L

Comparison of ELISA with LC-MS/MS

The dopamine ELISA analysis of 28 urine samples was compared to LC-MS/MS performed by ARUP Laboratories Inc. (Salt Lake City, UT, USA) (Kushnir et al. 2002). The regression analysis is shown in Fig. 3 (slope is 1.070; intercept is 18.3; r = 0.986, n = 28). Although dopamine concentrations obtained by the immunoassay showed a tendency to be slightly higher than those obtained by LC-MS/MS (slope is 1.070), both methods can be considered statistically identical.

Figure 3.

Comparison of urinary dopamine measurements by ELISA and LC-MS/MS (n = 28). Regression analysis: slope is 1.070 (95% confidence intervals 0.997 to 1.143); intercept is 18.3 (95% CIs −9.7 to 46.3); Pearson correlation (r) is 0.986. Dotted line represents x = y.

Stability of dopamine in urine

Freshly collected urine samples were aliquoted and stored at various temperatures (20°C, 4°C and at −20°C). The concentration of dopamine was measured in triplicate on the day of collection and at various times thereafter, ranging from 2 days to 1 month. Dopamine was stable in all urine samples for at least 30 days if stored at 20°C, 4°C or −20°C (see Figure S1 for stability at 20°C). The drift acceptance criterion was ± 15%.

Reference interval

Urinary DA levels were assessed in second morning urine samples of 64 apparently healthy individuals. Their age range was 18–63 years (43 females; 21 males). The parametric reference interval (95th percentile) was determined according to CLSI guideline C28-A3 (Clinical and Laboratory Standards Institute 2008). The reference range was from 64 to 261 μg dopamine/g Cr.

Comparison of dopamine levels in 24 h urine and spot urine

We conducted a preliminary pre-clinical study on 24 subjects to investigate the correlation between urinary DA levels in 24 h urine and spot urine collection. The participants (age 22–52 years; 4 males, 20 females) were not using medications/supplements that could affect DA levels. Strong correlation between DA concentrations in second morning spot urine (mean = 135.2, SD = 52.2) and 24 h urine (mean = 145.2, SD = 64.2) was observed (see Fig. 4; p < 0.001; r = 0.90; n = 24).

Figure 4.

Comparison of dopamine levels in 24 h urine and second morning spot urine (n = 24). Linear regression analysis (p < 0.001; r = 0.90; n = 24): slope is 1.1 (95% CIs 0.87 to 1.34); intercept is −4.29 (95% CIs −38.70 to 30.11)

Biological variation

Biological variation of urinary DA was assessed for 10 subjects (age 20–56 years; two males, eight females) considered healthy based on the HSCL-90. They were allowed to follow their ordinary diet without specific food restrictions and they were not using any medications/supplements during the study. Second morning urine specimens were collected weekly over a 4-week period at the same time of the day and on the same day of the week. The mean DA levels of these 10 individuals for the first, second, third, and fourth week were determined to be 132.5 ± 13.6, 121.4 ± 15.4, 127.3 ± 13.7 and 116.7 ± 13.2 μg dopamine/g Cr, respectively. A one-way anova (p = 0.866, F = 0.242) and a repeated measures anova (p = 0.784, F = 0.358) showed no statistical differences between mean dopamine weekly levels. The intra-individual variability (within-subject variation, CVI) for week-to-week collection was 23.9% and the inter-individual variability (between-subject variation, CVG) was 35.5%. The corresponding individuality index (I; CVI : CVG ratio) was 0.67.

Diurnal variation of urinary dopamine

Nineteen volunteers (age 23–58 years; 6 males, 13 females) with no chronic mental or physical illness collected urine samples every 4 h during 28 h time period. During the study period, the participants had no diet restrictions but they avoided excessive liquid intake and exercise. Figure 5 shows the diurnal excretion of dopamine (mean ± SEM) over a period of 28 h. One-way anova (p = 0.929, F = 0.350) and repeated measures anova (p = 0.155, F = 1.559) showed that the concentrations during these period were not statistically different.

Figure 5.

Diurnal (28 h) variation of urinary dopamine (mean ± SEM) for 19 individuals with no chronic mental or physical illness. Participants were allowed to follow their ordinary diet with the restriction to avoid excessive liquid intake and exercise. Urinary dopamine levels are not statistically different based on one-way anova (p = 0.3992) and repeated measures anova (p = 0.155, F = 1.559).

Monitoring urinary dopamine excretion after single l-DOPA administration

The dopamine ELISA was applied in the monitoring of DA excretion after intake of a single dose of l-DOPA in apparently healthy individuals. Urinary DA levels were assessed in 24 volunteers (age range: 18–54 years; 6 males; 18 females) with no physical or psychiatric disorders reported. A blinded crossover study was conducted where all participants took a placebo on the first day and a single low dose of l-DOPA (25 mg) on the next day. Both days the participants collected urine specimens at baseline, 2, 4, 6, and 24 h post-dose. Participants were restricted to use medications and supplements 24 h prior to the study and during the study. They were not allowed to ingest dopamine-rich food (e.g., bananas, pineapple, eggplant, walnuts). In Fig. 6 are presented the dopamine excretion profiles compared to the placebo levels at intake on the first day for each individual. It can be observed that the urinary dopamine excretion rates differ among subjects: 16 out of 24 subjects had maximum dopamine level at 2-h post-dose (Fig. 6a), 4/24 subjects – at 4-h post-dose (Fig. 6b), and one individual – at 6-h post-dose (Fig. 6c). For some individuals (3/24), the dopamine excretion was slower (Fig. 6d). In addition, the maximum dopamine levels observed ranged from 213 to 1407 μg/g Cr. The percent increase of dopamine levels versus placebo levels was in the range of 188–1825% with an average increase of 491%.

Figure 6.

Urinary dopamine excretion after single dose (25 mg) of l-DOPA administration compared to placebo levels. Twenty-four participants with no physical or psychiatric disorders collected urine samples before administration (0 h) and 2, 4, 6, and 24 h post-dose. Subjects with maximum dopamine levels at 2 h post-dose (a), at 4 h post-dose (b), at 6 h post-dose (c), and 2- to -4 h post-dose (d).

Urinary dopamine in Parkinson's disease patients under different treatment

Steady state dopamine levels were determined for three groups of PD patients (n = 162) subject to different treatments as described in Fig. 7. In group A were included patients receiving only SSRIs, SNRIs, tricyclics, benzodiazepenes, stimulant, or sleep medications; in group B–patients under anti-Parkinson's medications, such as dopamine agonists (Ropinrole), anticholinergics (Benzotropine), or mono-amine-oxidase (MAO) inhibitors (Selegeline), that do not contain DA precursors; and finally group C was formed by PD patients treated with the same medications as group B but in addition receiving DA support with DA precursors, such as l-Tyrosine, Carbidopa and/or l-DOPA with maximum doses of 2000 mg/day, 500 mg/day and 2000 mg/day, respectively. Second morning urine samples were collected at least 17 h post-dose of DA support intake. According to the one-way anova analysis the DA concentrations measured for the three groups were found to be significantly different (p = 0.0001, F = 41.61). Post hoc Tukey multiple comparisons analysis identified that patients under DA support therapy (group C) had significantly elevated urinary dopamine in comparison to the other two groups of PD patients (group A and B) that were not receiving DA support (all p-values < 0.05).

Figure 7.

Urinary dopamine levels (mean ± SEM) in Parkinson's disease patients under different treatment: group A – patients receiving only SSRIs, SNRIs, tricyclics, benzodiazepenes, stimulant, or sleep medications; group B – patients under anti-Parkinson's medications, such as dopamine agonists (Ropinrole), anticholinergics (Benzotropine), or mono-amine-oxidase (MAO) inhibitors (Selegeline), that do not contain DA precursors; group C – patients treated with the same medications as group B but in addition receiving DA support with DA precursors, such as l-Tyrosine, Carbidopa and/or l-DOPA with maximum doses of 2000 mg/day, 500 mg/day, and 2000 mg/day, respectively. According to the one-way anova analysis the DA concentrations measured for the three groups were found to be significantly different (p = 0.0001, F = 41.61).

Monitoring urinary dopamine in Parkinson's disease patients under l-DOPA therapy. Dose effect study

Urinary DA concentrations were measured in twenty PD patients prior to (baseline) and during l-DOPA therapy (retest). The study included non-treated group and three groups of treated patients: treated with low, mid and high daily dose of l-DOPA (see Fig. 8). The patients were taking anti-Parkinson's medications, such as dopamine agonists (Ropinrole), anticholinergics (Benzotropine), and/or MAO inhibitors (Selegeline) at the time of baseline testing and retest. Second morning urine (baseline and retest) were collected at least 17 h after l-DOPA dose. Figure 8 shows the percentage change in dopamine concentration for each group. Based on one-way anova, the percent changes in dopamine levels were found to be significantly different (p = 0.003, F = 6.942). Post hoc Tukey multiple comparisons analysis showed that Parkinson's Disease patients receiving either mid or high l-DOPA dose had a significantly higher increase in urinary dopamine than the patients without treatment or those treated with low l-DOPA dose (all p-values < 0.05).

Figure 8.

Monitoring urinary dopamine in Parkinson's disease patients prior to and during l-DOPA therapy. Circle symbols (•) represent the% change in dopamine concentrations from baseline to retest for each patient;% change = {[(Retest) − (Baseline)]/(Baseline)} × 100. The solid lines (—) represent the mean % change for each group.


A competitive ELISA for quantitative determination of dopamine in urine samples was established. The proposed method is very simple to perform and includes specimen neutralization followed by in situ glutaric aldehyde derivatization prior to the immunochemical quantification. The derivatization step with glutaric aldehyde mimics the preparation of the immunogen used to generate the anti-dopamine-specific antibody (Huisman et al. 2010a, b).

In contrast to the traditional approaches used in clinical urinary testing the creatinine normalization is an intrinsic part of the methodology described in this work. The volume of urine used in the analysis is equivalent to 120 μg Cr. In this way, samples are diluted according to their Cr level and the urinary matrix is normalized. This approach reduces matrix interferences and allows direct dopamine measurement in whole urine without the need of sample purification.

The analytical procedure was characterized by excellent precision, good sensitivity, and analytical recovery. The average inter-assay variability of 11.4% (see Table S1) was less than one half of the average within-subject biological variation (CVI = 23.9%) of urinary dopamine (see section ‘Biological variation’) which is an established criteria for desirable performance of laboratory measurements (Fraser et al. 1997). Although the sensitivity of the immunoassay (LoD of 7.3 μg/g Cr) is not as high as those reported for LC methods (Kushnir et al. 2002; Tsunoda 2006; de Jong et al. 2010a, 2011), it is certainly sufficient for the analysis of free dopamine in urine samples. The linearity studies showed that the analytical measurement range of the assay allows the direct quantification of endogenous urinary dopamine in the general population and also in patients under dopamine-modulating treatment where elevated dopamine levels might be expected. Very good correlation (r = 0.986, n = 28) was observed between urinary dopamine concentrations determined by the competitive ELISA and LC-MS/MS.

The present dopamine ELISA is based on the use of a highly specific antibody. The specificity studies suggest that the degree of recognition of the anti-dopamine antibody is directly related to the presence of the catechol (3,4-dihydroxy) group, to the ethyl-amino moiety available for amino coupling with the protein and to the glutaric aldehyde linker. If any of those three epitopes is absent in the analyte, no antibody binding is observed. For example, the lack of 3-hydroxy group in the molecules of tyramine and 3-methoxytyramine and the absence of 4-hydroxy group in the molecule of 3-hydroxy-4-methoxyphenylethylamine (4-methoxytyramine) result in no cross-reactivity with those compounds (see Fig. 1). Furthermore, molecules containing methylated imine group instead of amino group (such as epinephrine) were not detected in the assay because of weaker reactivity of the imine group in the glutaric aldehyde derivatization. Even more, any functional substitution in the ethylene arm (ex. norepinephrine) results in negligible interference.

The high selectivity of the anti-dopamine antibody used in the competitive assay allows for very specific dopamine determination in whole urine samples without sample pre-treatment. Thus, major urinary components do not affect the dopamine measurement (see Table 1). No significant interferences were observed by dopamine precursors, metabolites of the tyrosine pathway, related biogenic amines, dietary supplements, and drugs. The assay is not affected by urinary concentrations of l-DOPA up to 45 mg/L. Considering that ~ 1% of l-DOPA dosage is excreted as a parent compound in Parkinson's therapy (Routh et al. 1971), it can be expected that therapeutic doses up to 4.5 g l-DOPA per day will not interfere with the dopamine measurement. Therefore, the interference free quantification of urinary dopamine makes this ELISA a valuable tool in clinical settings for monitoring dopamine.

One of the aims of this study was to estimate if spot urine collection is an appropriate method to study biological variation of dopamine, to distinguish cause by disease from natural variations. Currently, the 24 h timed urinary collection remains the most common method for evaluation of dopamine excretion despite the practical difficulties in this sampling. However, in a study on Parkinson's patients random urine specimens were used and significant positive correlation with daily dose of l-DOPA was observed (Davidson et al. 2007). Thus, we have compared dopamine levels (expressed as μg of dopamine per g of creatinine) determined in 24 h urine collection and second morning spot urine. The results of this comparative study showed very good correlation (p < 0.01; r = 0.90; n = 24; see Fig. 3). Therefore, spot urine sampling could be considered an alternative collection for dopamine assessment. Furthermore, we observed that the week-to-week variation of dopamine in second morning spot urine collections for 10 individuals is very reproducible. We would like to note that the observed biological variations (within-subject: CVI of 23.9%; between-subject: CVG of 35.5%) and the corresponding individuality index (I of 0.67) for urinary dopamine are comparable to those seen for other types of urinary metabolites widely used in clinical laboratories, such as creatinine, urea, protein, sodium. (Ricós et al. 2012). Studies on day-to-day and within- and between-subject biological variations of catecholamine excretion under free-living conditions and on normal healthy subjects are very limited (Curtin et al. 1996; Souza et al. 1998). Souza et al. reported that the 24 h dopamine excretion during habitual daily activities for normal subjects had within-subject variation of 31% for women (n = 22) and 39% for men (n = 12) and between-subject variation of 42% for women and 39% for men (Souza et al. 1998). The biological variations for second morning urine collection observed by us are lower than the variations reported by Souza et al. for 24 h samples. A second morning urine sample seems to be reliable and more appropriate than 24 h collection for clinical use. The determination of biological variation of urinary DA established here will be of great value when considering patients with possible neuroendocrine disorders and other physiologic and pathologic conditions related to dopaminergic imbalances. Future studies are needed to assess the long-term biological variations in healthy population and in symptomatic patients to improve their diagnosis and treatment. No circadian rhythms were seen in the 28 h excretion pattern of free dopamine for 19 healthy volunteers. Our data are in agreement with earlier findings obtained by HPLC analysis that circadian variations in free dopamine excretion are relatively smaller compared to those found for norepinephrine and epinephrine (Fibiger et al. 1984; Westernik and ten Kate 1986; de Jong et al. 2010b).

The ability of the ELISA method to measure urinary dopamine was demonstrated by the excretion profiles obtained after very low dose (25 mg) of l-DOPA oral administration to 24 apparently healthy volunteers (Fig. 6). Maximum dopamine levels were found 2 h post-dose for most of the individuals similarly to a study reported by Brown and Collery (1981). Although significant inter-individual differences in the pharmacokinetics of l-DOPA has been observed, most of the ingested l-DOPA was completely eliminated as DA in ~ 17 h (data not shown). This suggests that in the case of steady state circulating levels of DA assessment urine sampling should be performed at least 17 h post-dose.

To evaluate whether the present ELISA is useful for detecting variations in DA levels under pathological conditions, we compared the steady state DA concentrations in second morning urine samples collected from 162 Parkinson's disease patients receiving different therapies (see Fig. 7). As demonstrated, long term DA support with DA precursors significantly increased DA concentration in the circulation and as a consequence higher DA levels in the urine were observed: the mean DA levels determined for the PD patients under DA support (group C) were ~ 8-fold higher than those observed for patients not taking anti-PD medications nor DA precursors (group A) and ~ 2-fold higher than those determined for the patients treated with anti-PD drugs other than DA precursors (group B). Interestingly, we have observed that intake of anti-PD medications such as DA agonists, anti-cholinergic drugs, SDRIs and/or MAO inhibitors (group B) resulted in significant but much smaller increase in circulating DA with respect to PD patients that were not taking any anti-PD drugs nor DA support (group A). The relationship between urinary DA levels and the clinical response of the patients remains to be established. Nevertheless, our observations suggest that the DA assay presented here would be a valuable tool in future clinical research on DA support therapies.

Finally, we have demonstrated the clinical utility of the dopamine immunoassay to monitor therapy of PD patients (n = 20) treated with different l-DOPA daily doses (Fig. 8). Pre-treatment (baseline) and under treatment (retest) DA levels were assessed and the relative increase in circulating DA was determined. Dose-dependent changes in urinary dopamine levels in response to l-DOPA administration were observed. We found that daily l-DOPA doses greater than 200 mg/day are necessary to significantly increase urinary dopamine concentrations in PD patients. Similarly, significant increase of urinary dopamine after l-DOPA treatment in a dose-dependent manner was reported in previous PD studies (Routh et al. 1971; Dutton et al. 1993; Davidson et al. 2007). Furthermore, it should be noted that the percent increase in DA levels is not only dose-dependent but also patient dependent. Our results indicate considerable differences in the individual responses to the same dose of l-DOPA. Also, the same % increase in DA levels can be achieved by low, mid or high dose depending on the patient. Both the pharmacokinetics disposition and the pharmacodynamic response to a given dose of dopamine modulators in individual patients may vary widely as a direct consequence of epigenetic differences. It is well known that the individual metabolism of patients may not only differ from one another but may change in the same patient during treatment. Thus, anti-parkisonian treatments lose effectiveness with progression of the disease. These strongly suggest the need of urinary dopamine assessment to determine optimal dosing and patient compliance of l-DOPA treatment. An understanding of the individual's metabolism of l-DOPA is crucial to establish the optimal therapeutic regimen for that agent.

In summary, this study outlines a validated ELISA method for the direct analysis of dopamine in urine samples. We conclude that determining the ratio of dopamine per g Cr in second morning urine samples presents an accurate, convenient, inexpensive, and reliable estimate of dopamine excretion. The method for urinary dopamine presented here is robust, sensitive and very specific and can be used in monitoring dopamine-modulating therapies. Comparing of urinary DA concentrations to a baseline level might be helpful in assessing the suitability of each patient's dosage, assessing patient's compliance, and avoiding overdose. The application of our method in overall clinical assessment still needs to be defined.

The immunoassay described here may provide a viable cost- effective alternative to chromatographic analysis as it offers higher throughput and there is no need of time-consuming and complicated sample pre-treatment. The dopamine ELISA can be run in parallel with other immunoassays for the detection of corresponding metabolites of interest (Huisman et al. 2010b; Nichkova et al. 2012). The method can be further validated as part of a biomarker panel for the determination of patterns of monoamines (norepinephrine, epinephrine, serotonin, etc.) that can be used for monitoring and treatment of neurological disorders.


We acknowledge the skillful technical assistance of Kathy Mork and Scott Howard. We thank Corena McManus and Jodi Branstad for their assistance in the clinical studies. The study was funded by Pharmasan Labs Inc. Dr. M. Nichkova and P. Wynveen have a conflict of interest working as scientists for Pharmasan Labs Inc. Dr. H. Huisman has a conflict of interest working as a consultant for Pharmasan Labs Inc.