This protocol describes a novel colorimetric method for antimalarial primaquine (PMQ) detection in synthetic urines and human serums.
Primaquine (PMQ), an important anti-malarial drug, has been recommended by the World Health Organization (WHO) for the treatment of life-threatening infections caused by P. vivax and ovale. However, PMQ has unwanted adverse effects that lead to acute hemolysis in patients with glucose-6-phosphate dehydrogenase (G6PD) deficiency. There is a need to develop simple and reliable methods for PMQ determination with the purpose of dosage monitoring. In early 2019, we have reported an UV-Vis and naked-eye based approach for PMQ colorimetric quantification. The detection was based on a Griess-like reaction between PMQ and anilines, which can generate colored azo products. The detection limit for direct measurement of PMQ in synthetic urine is in the nanomolar range. Moreover, this method has shown great potential for PMQ quantification from human serum samples at clinically relevant concentrations. In this protocol, we will describe the technical details regarding the syntheses and characterization of colored azo products, the reagent preparation, and the procedures for PMQ determination.
PMQ is one of the most important antimalarial drugs, it works not only as a tissue schizontocide to prevent relapse but also as a gametocytocide to interrupt disease transmission1,2,3,4. Intravascular hemolysis is one of the concerning side effects of PMQ, which becomes extremely serious in those deficient in G6PD. It is known that the G6PD genetic disorder is distributed worldwide with a gene frequency between 3-30% in malaria endemic areas. The severity of PMQ weakness depends on the degree of G6PD deficiency as well as the dose and the duration of PMQ exposure5,6. To lower the risk, the WHO has recommended a single low dose (0.25 mg base/kg) of PMQ for malaria treatment. However, this is still challenged by the variations in patient drug sensitivity5,7. Dose monitoring is necessary to assess the pharmacokinetics after PMQ administration, which can effect dosage adjustment for a successful treatment with limited toxicity.
High-performance liquid chromatography (HPLC) is the most widely used technique for PMQ clinical determination. Endoh et al. reported a HPLC system with a UV detector for serum PMQ quantification using a C-18 polymer gel column8. In their system, serum proteins were first precipitated with acetonitrile, and then the PMQ in the supernatant was separated for HPLC. The calibration curve was linear over the concentration range from 0.01-1.0 μg/mL8. Another method based on a reverse-phase HPLC with UV detection at 254 nm has been reported for the quantification of PMQ and its major metabolites9. The calibration curve for PMQ was linear in the range between 0.025-100 μg/mL. An additional liquid-liquid extraction with mixed hexane and ethyl acetate as organic phase was used for PMQ separation with percentage recovery reached to 89%9. More recently, Miranda et al. developed an UPLC method with UV detection at 260 nm for PMQ analysis in tablet formulations with a detection limit at 3 μg/mL10.
Though HPLC methods exhibit promising sensitivity in drug determination and the sensitivity can be further improved if the HPLC is equipped with a mass spectrometer, there are still some disadvantages. Direct drug measurements in biological fluids are usually inaccessible by HPLC, since many biomolecules can greatly influence the analysis. Additional extractions are required to remove endogenous molecules before HPLC analysis11,12. Moreover, PMQ detection by a HPLC-UV detector is typically performed at its maximum absorption wavelength (260 nm).; however, there are many endogenous molecules in biological fluids with a strong absorbance at 260 nm (e.g., amino acids, vitamins, nucleic acids and urochrome pigments), thus interfering with PMQ UV detection. There is a need to develop simple and cost-effective methods for PMQ determination with reasonable sensitivity and selectivity.
The Griess reaction was first presented in 1879 as a colorimetric test for nitrite detection13,14,15,16. Recently, this reaction has been extensively explored to detect not only nitrite but also other biologically relevant molecules17,18,19,20. We have previously reported the first systematic study of an unexpected Griess reaction with PMQ (Figure 1). In this system, PMQ is able to form colored azos when coupled with substituted anilines in the presence of nitrite ions under acidic conditions. We have further found that the color of azos varied from yellow to blue when increasing the electron donating effect of the substituent on anilines21. A UV-vis absorption based colorimetric method for PMQ quantification has been developed through the optimized reaction between 4-methoxyaniline and PMQ. This method has shown great potential for sensitive and selective detection of PMQ in bio-relevant fluids. Here, we aim to describe the detailed procedures for PMQ determination based on this colorimetric strategy.
1. Synthesis of Colored Azos
2. UV-Vis Measurements and Theoretical Calculation
3. PMQ Determination
To optimize the reaction conditions (Figure 2), various anilines were used to couple with PMQ through the Griess reaction. We have achieved a series of azos with different colors. It has been found that anilines with an electron donating substituent can cause a red-shift in the UV-vis absorption spectrum. Theoretical calculations were carried out through time dependent density functional theory (TD-DFT). As presented in Figure 2A, the calculation result was in good agreement with optical measurements with average error of 3.1%. 4-methoxyaniline was then used to conduct the PMQ detection reaction due to its good performance in reaction rate, product solubility, and stability21. Moreover, the azo product from 4-methoxyaniline is red in color, which is easy to distinguish with naked eyes. Therefore, this reaction offers potential for naked-eye PMQ detection (Figure 3).
Figure 4A shows the pH effect on the UV-vis absorption spectrum of the azo product 3d. I504 does not change when increasing pH from 1.0 to 6.0. I504 under pH 7.0 exhibits a slight decrease, while a basic pH (8.0 and 9.0) greatly affects the absorption. Figure 4B shows the pH effects of PMQ solutions on the Griess reaction. PMQ (50 μM) in PBS buffer with various pHs (4.0, 5.0 ,6.0, 7.0, 8.0, 9.0) were individually mixed with the testing reagent as described in section 3.1. I504 was then measured after 15 min at room temperature. As indicated, basic pHs (8.0, 9.0) of PMQ solutions potentially influence the reaction. Figure 5 shows the general procedure to perform the Griess reaction for PMQ detection. As described in the protocol section, four steps are required to obtain the absorption data I504 for analysis. Figure 6A and 6B show the calibration curves for direct detection of PMQ from urine and serum samples, respectively, without sample pretreatments. An excellent linear relationship (R2 = 0.998) was found when PMQ in synthetic urine ranges from 0 to 200 μM. In term of the serum sample, a linear relationship was found at the concentration ranging from 10 to 200 μM.
Figure 7A shows the procedure to extract PMQ from serum. The residues were redissolved in distilled water after extraction and concentration, and then filtrated. To simulate a real PMQ-containing serum, PMQ was added into human serum with final concentrations at 0, 0.2, 0.5, 1.0, 2.0 μM. Using steps 3.4 and 3.5, the concentrations of PMQ in serums were found to be 0.02, 0.14, 0.44, 0.90 and 1.78 μM, respectively (Figure 7C). Based on the result, the percentage of PMQ recovery was found to be around 90% when PMQ was over 0.5 μM in serum, which was comparable to previous reports9.
Figure 1: Schematic of the Griess reaction on PMQ. (A) A classical Griess reaction for nitrite analysis. (B) The Griess reaction in the proposed PMQ detection method. This figure has been modified with permission from previous work21. Please click here to view a larger version of this figure.
Figure 2: Photophysical properties of synthetic azos. (A) UV-vis measurement and theoretical calculation of the maximum absorption of azos generated from different anilines. The numbers outside the brackets represent for the maximum absorbance measurement in distilled H2O near neutral pH conditions; the numbers in the brackets refer to the measurement in 5% H3PO4 solution (pH ≈ 1.1). λabs/nm exper. represents the experiment data and λabs/calc. represents the theoretical calculation data. Eexc is the excitation energy (eV), and f is the oscillator strength. (B) Photo images of PMQ and the azo products with different substituents, 50 μM in 5% phosphoric acid solution. (C) UV-vis spectra of the synthetic products. The values were normalized to a range between 0 and 1. This figure has been modified with permission from previous work21. Please click here to view a larger version of this figure.
Figure 3: Colorimetric determination of PMQ. (A) Monitoring the absorbance changes at maximum I504 in a time dependent way. The reaction was performed using 4-methoxyaniline, and PMQ was used at 100 μM; (B) Color changes of the reaction with different concentrations of PMQ: 400 µL of 4-methoxyaniline solution (200 mM in 0.2 M HCl) and 200 µL of sodium nitrite in water (5 mM), with 200 µL of PMQ solution of different concentrations (0, 1, 2, 5, 10, 20, 50, 100 μM). Please click here to view a larger version of this figure.
Figure 4: pH effect on PMQ detection. (A) pH effects on the UV-vis absorbance of azo product 3d (50 μM); (B) PMQ (50 μM) in PBS buffer with different pHs (4.0, 5.0, 6.0, 7.0, 8.0, 9.0) were used to perform the reaction as described in step 3.1. Fifteen min later, the absorbance at 504 nm was measured. Please click here to view a larger version of this figure.
Figure 5: PMQ determination through a Griess reaction on a 96-well plate based system. R1 refers to 200 mM 4-methoxyaniline solution in 0.2 M HCl; R2 refers to 5 mM sodium nitrite in distilled water. Please click here to view a larger version of this figure.
Figure 6: Calibration curves for PMQ determination from (A) synthetic urine and (B) human serum samples. The concentration of PMQ ranges from 0-200 μM. Please click here to view a larger version of this figure.
Figure 7: PMQ determination from serum samples. (A) Schematic illustration of PMQ extraction from serum samples for the quantitative analysis. (B) The linear relationship found between I504 and PMQ concentration within the range from 0 to 100 μM. (C) PMQ in serum was quantify by the Griess reaction-based method in comparison with the exact amount added into the serum. This figure has been modified with permission from previous work21. Please click here to view a larger version of this figure.
Table 1. Theoretical calculation of Log D and the percentage of water distribution of PMQ and CPMQ.
We described a colorimetric method for convenient PMQ quantification. It is potentially the most simple and cost-effective current method. More importantly, this method offers enables naked-eye based PMQ measurement without using any equipment.
The optimized Griess reaction for PMQ detection can generate a red color azo with a maximum absorption at 504 nm. The potential influence from UV-vis absorption of endogenous biomolecules is limited, thus making the method promising for direct measurement of PMQ in biological fluids. As indicated by the result, an excellent linear relationship (R2 = 0.998) was found for urine PMQ detection over the concentration range of 0-200 μM (Figure 6A). The limit of detection (LOD) for PMQ was found to be 0.63 μM. This method has also shown great potentials for direct measurement of PMQ in human serum. An excellent linear relationship was found in the concentration ranging from 10 to 200 μM for serum PMQ detection (Figure 6B). We can further improve the sensitivity by pre-treating the serum sample through extraction and concentration. As Figure 7 shows with a simple extraction process, this method can quantify serum PMQ at clinically relevant ranges. Based on the reaction mechanism, the main carboxyl metabolite of PMQ (CPMQ) can potentially form an azo product with similar UV-Vis properties. However, the liquid-liquid extraction under basic pH conditions can potentially minimize the interference from CPMQ. Table 1 shows the calculated log D and water distribution of both PMQ and CPMQ. As shown, at pH >10, less than 6.33% of PMQ will be found in the water phase, whereas over 98.54% of CPMQ will be in the water phase. Therefore, theoretically, more than 93.7% of PMQ and less than 1.56% of CPMQ could be extracted out for test. It can be concluded that the interference from the main metabolite CPMQ is limited.
The procedure for PMQ detection is very easy to handle. Taking the 96-well plate-based system as an example, the entire procedure consists of four steps: 1) adding 100 µL of 4-methoxyaniline solution (200 mM in 0.2 M HCl) R1 into a 96-well plate; 2) adding 50 µL of PMQ concentration-unknown sample to mix with R1; 3) adding 50 µL of R2 (5 mM sodium nitrite solution) to perform the reaction at room temperature; and 4) recording the UV-vis absorption at 504 nm using a spectrometer. The concentration of PMQ from an unknown sample can be calculated based on the absorption intensity I504 and the linear equation from calibration curve. The entire procedure is performed at room temperature without the need of incubation. A dark environment is not necessary for the entire procedure, as the colored product is not sensitive to room light.
It should be noted that the time for the reaction solution to reach its saturated I504 is temperature dependent. As shown in Figure 3, at least 12 min was required at room temperature (25 °C). The reaction time would be longer if performing the reaction at temperatures below 25 °C. The basic pH condition of PMQ solutions can potentially affect the absorbance I504. To address this issue, adjust the pH of PMQ solution to be less than 7.0. Otherwise, a new calibration curve is needed for the solution with pH over 7.0. In addition, intrinsic nitrites in the tested samples can influence the detection. However, this may only occur when the concentration of intrinsic nitrites is extremely high since a high concentration of nitrite (5 mM) was used in a standard test.
The authors have nothing to disclose.
The authors acknowledge the Start-Up Grant from Guangzhou University of Chinese Medicine and the youth scientific research training project of GZUCM (2019QNPY06). We also acknowledge the Lingnan Medical Research Center of Guangzhou University of Chinese Medicine for the support on facilities.
4-Methoxyaniline | Aladdin | K1709027 | |
2,4-Dimethoxyaniline | Heowns | 10154207 | |
3,4-Dimethoxyaniline | Bidepharm | BD21914 | |
4-Methylaniline | Adamas-beta | P1414526 | |
4-Nitroaniline | Macklin | C10191447 | |
96-wells,Flat Botton | Labserv | 310109008 | |
Gaussian@16 software | Gaussian, Inc | Version:x86-64 SSE4_2-enabled/Linux | |
Hydrochloric acid | GCRF | 20180902 | |
Marvin sketch (software) | CHEMAXON | free edition: 15.6.29 | |
Phosphoric acid | Macklin | C10112815 | |
Primaquine bisiphosphate | 3A Chemicals | CEBK200054 | |
Sodium nitrite | Alfa Aesar | 5006K18R | |
Sulfonamides | TCI(shanghai) | GCPLO-BP | |
Varioskan LUX Plate reader | Thermo Fisher | Supplied with SkanIt Software 4.1 |