Multisensory colorimetric analysis of drugs dydrogesterone, troxerutin and ademetionine using barcodes

Full Text

Abbreviations: RGB – red, green, blue; MBTH – 3-methylbenzothiazolinone hydrazone; PCA – Principal Component Analysis; PC1 – Principal Component 1.

INTRODUCTION

For a preliminary rapid detection of counterfeits (even before using more expensive analytical equipment), it is advisable to practise simple, accessible and express methods. One of these methods is a digital colorimetry, based on the registration of electromagnetic radiation in the visible range of wavelengths by digital devices to get color raster images [1–6]. Digital colorimetry has become widespread in pharmaceutical analyses. In this area, the method is used to: analyze medicinal plants [7, 8]; assess the quality of collections, which include powders of medicinal plants [9–11]; determine the whiteness of powdered and tableted drugs [12]; identify the biologically active substances and drugs both by their own color and by the color of the products of color reactions used in pharmacopoeial tests [12]; identify the DOA and banned products [13, 14].

Digital colorimetry combines the availability of chemical test methods with a visual detection and a good performance of instrumental methods, primarily optical molecular spectroscopy. The extremely low cost of the analysis by this method is due to the possibility to measure the analytical signal using consumer digital optical devices not certified as measuring instruments [1, 5, 15, 16]. Despite the obvious advantages, the colorimetric method is not devoid of a number of limitations, the main one of which is its low selectivity [4, 17].

To increase the selectivity of the method, the use of molecular sensors was proposed [18]. It is advisable to use a cell of several chromogenic agents as sensors, in which a series of analytical reactions can be carried out simultaneously. The multisensor colorimetry method [19–29] is based on obtaining colored products of an analyte interaction with molecular sensors, getting information about their color characteristics, and then converting them into a discrete substance “barcode” that can be used for chemical analyses [17, 30].

A unique colorimetric two-dimensional code makes it possible to estimate both the nature and the content of active substances in the drugs at the minimum level of information noise [17, 30]. To form the “barcodes”, it is advisable to choose such sensors and color channels, the values of the lightness of which correlate with the content of the analyte. Lightness shall mean the color coordinate on one of the color channels in the RGB system (varies in the range from 0 to 255).

The drugs of three different pharmacological groups were selected as the test objects. Dydrogesterone is synthetic progestogen that fully ensures the onset of the secretion phase in the endometrium in cases of endometriosis and dysmenorrhea. Troxerutin is a flavonoid, a phleboprotective drug that has venotonic, angioprotective, anti-inflammatory, anti-edema and antioxidant effects. Ademetionine is an antioxidant, hepatoprotective, detoxifying agent. The structural formulas of the active substances are shown in Fig. 1. The development of alternative methods for their identification, suitable for a preliminary screening analyses of drugs, is an important and urgent task of the pharmaceutical and analytical chemistry.

 

Figure 1 – Structural formulas of dydrogesterone (a), troxerutin (b), ademetionine (c)

 

THE AIM of this study is to develop a universal method by multisensor digital colorimetric analysis of drugs of various pharmacological groups using dydrogesterone, troxerutin and ademetionine as examples. The developed complex of molecular sensors in combination with new approaches to the processing of analytical signals will make it possible to identify the above-mentioned active substances in the drugs.

MATERIALS AND METHODS

Study objects

To implement the proposed approach, the substances of dydrogesterone (“Abbott Biologicals B.V.”, Netherlands), troxerutin (JSC “Interfarma”, Prague, Czech Republic) and ademetionine (LLC “Farmamed”, Moscow, Russia), troxerutin capsules 300 mg (LLC “Pranafarm”, Samara, Russia), lyophilisate for an intravenous solution and an intramuscular administration of “Heptral”^® (ademetionine) 400 mg (“Abbott Laboratories”, GMBH, Germany), tablets “Duphaston”^® (dydrogesterone) 10 mg (“Abbott Healthcare Products B.V.”, Netherlands) were used.

Materials

For a quantitative analysis, a series of calibration solutions was prepared using the substances troxerutin and ademetionine (4.0-20.0 mg/ml) in increments of 4 mg/ml, dydrogesterone (1.0-3.0 mg/ml) in increments of 0.5 mg/ml. The concentration range had been selected in such a way that the content of the active substance in the real drug would be in the middle of the calibration curve.

The calibration solutions were analyzed by multisensor colorimetry using the following set of 8 sensors (C₁–C₈): an intact solution – a 96% (v/v) aqueous ethanol solution – C₁; 1 mM alcoholic solution of anthraquinone green (CAS#4403-90-1) – C₂; a 0.2% aqueous solution of 3-methylbenzothiazolinone hydrazone (MBTH) (CAS#1128-67-2) – C₃; a 0.2% methyl orange aqueous solution (CAS#547-58-0) – C₄; a 1 mM alcoholic solution of sulforhodamine B (CAS#3520-42-1) – C₅; a 1 mM alcoholic solution of 1-hydroxypyrene (CAS#5315-79-7) – C₆; 1 mM alcoholic solution of allura red AC (CAS#25956-17-6) – C₇; a 1 mM aqueous solution of of iron (III) chloride – C₈.

Equipment

Transparent flat-bottomed polypropylene plates with 96 cells [31–33], cell volume 350 µl (Thermo Fischer Scientific, USA, cat. No. 430341) were used as a base for the chip.

Using Biohit mLine dispensers (Sartorius, USA), 100 μL of alcohol solutions of substances, sensor solutions (C₁–C₈), and purified water were placed into the cells of the plate. The number of sensors was determined so that it would be possible to analyze the maximum number of samples on one plate (8 sensors by the number of rows of the plate).

For obtaining raster images, an Epson Perfection 1670 office flatbed scanner (CCD-matrix) with a removable cover was used. The plate with the samples was scanned using the Epson Scan software in the Professional mode (600 dpi resolution, 24-bit color depth). “Color restoration”, “Unsharp mask filter” and “Descreening filter” options were disabled. To perform a digital colorimetric analysis using a 96-cell plate (Thermo Fischer Scientific, USA, cat. No. 1256604), a Teflon insert of 210×297×17 mm in size, was made with a center rectangular cut (128×86 mm) and was placed under the cover of an office A4 flatbed scanner. It made it possible to: expedite and formalize the procedure for placing the plate on the working glass table of the scanner; fix the coordinates and lighting conditions of the plate with an electroluminescent lamp built into the carriage; minimize the side stray illumination of the plate with substrates by external illumination sources; improve the accuracy of measuring results of plate raster images color channels lightness.

The difference in the color channels lightness between the analyte cell and the intact cell was used as an analytical signal. The obtained digital images of the cells were processed using the ImageJ software (Wayne Rasband, National Institutes of Health, USA; http://imagej.nih.gov/ij) using the RGB 24-bit color model (8 bits per channel), in each cell the central area was selected and 3 averaged values of lightness were obtained for it, one for each color channel of RGB. The choice of color channels was carried out empirically.

RESULTS AND DISCUSSION

Semi-quantitative colorimetric analysis of troxerutin, dydrogesterone and ademetionine

The obtained values of the lightness of RGB-channels were processed in the MS Excel spreadsheet editor, the optimal threshold values of the difference in the lightness of the channels for the analyzed solution and the intact cell were chosen. The values above them were conventionally designated as “1” and below them as “0” (Table 1) and colorimetric “barcodes” were created (Table 2). When choosing the optimal threshold value for the difference in lightness, the following requirements were met: (1) the code must be unique; (2) the difference in coding between adjacent concentrations should be minimal (1-2 values). To meet these requirements, it is advisable to set individual thresholds for each channel. This problem was solved using MS Excel (Add-in “Search for a solution”).

 

Table 1 – Colorimetric codes corresponding to various concentrations of dydrogesterone, troxerutin and ademetionine

Dydrogesterone
с, mg/ml	ΔR2	ΔG4			ΔG6		ΔB6			ΔR7
Threshold value of differences i n lightness	127	92			30		50			80
1.0	0	0			0		0			1
1.5	1	0			0		0			1
2.0	1	0			0		1			1
2.5	1	1			0		1			1
3.0	1	1			1		1			1
Troxerutin
с, mg/ml	ΔG4		ΔG5			ΔR7			ΔR8
Threshold value of differences in lightness	125		91			82			92
4 or less	0		0			0			0
8	1		0			0			0
12	1		1			0			0
16	1		1			1			0
20	1		1			1			1
Ademetionine
с, mg/ml	ΔR2	ΔG2		ΔR3		ΔB3		ΔG5		ΔG7
Threshold value of differences in lightness	127	92		30		50		80		101
4	0	0		0		0		1		0
8	0	1		1		0		1		0
12	0	1		1		1		1		0
16	0	1		1		1		1		1
20	1	1		1		1		1		1

 

Table 2 – Scale of “barcodes” corresponding to various concentrations of dydrogesterone, troxerutin and ademetionine

Dydrogesterone		Troxerutin		Ademetionine
c, mg/ml	“Barcode”	c, mg/ml	“Barcode”	c, mg/ml	“Barcode”
1.0		4		4
1.5		8		8
2.0		12		12
2.5		16		16
3.0		20		20

 

The presented one-dimensional “barcodes” can be сlustered into a two-dimensional code (Table 3), which makes it possible both to estimate the nature and the content of the active substance in the drugs at the minimum level of the information noise. The interpretation of a two-dimensional code for the identification of the substances is possible both in visual and “instrumental” mode, for example, using a software “barcode” scanner on a smartphone after its preliminary setup. The latter mode is especially useful when processing large data sets to increase the reliability of the analysis results.

 

Table 3 – Two-dimensional “barcodes” for simultaneous analysis dihydrosterone, troxerutin and ademetionine

Active substance, c, mg/ml

C1

C2

C3

C4

C5

C6

C7

C8

R

G

B

R

G

B

R

G

B

R

G

B

R

G

B

R

G

B

R

G

B

R

G

B

Dydroge- sterone (2.5 mg/ml)

1,0

1,5

2,0

2,5

3,0

Troxerutin (16 mg/ml)

4

8

12

16

20

Ademetionine (8 mg/ml)

4

8

12

16

20

Note: С₁–С₈ – sensors; the dark fill of the cell corresponds to the presence of a signal, the light one – to its absence

 

Thus, the technique of the semi-quantitative analysis of drugs can be reduced to comparing the code of the test solution with the corresponding code of a standard solution of the known concentration. Since the inaccuracy of the semi-quantitative analysis results is initially high, there is no need to use an inaccessible standard sample. It is just necessary to reproduce the described conditions of measuring the analytical signal and use a ready-made set of two-dimensional barcodes.

Colorimetric quantitative analysis of dydrogesterone, troxerutin and ademetionine drugs

For the quantitative analysis, it is advisable not to use all color channels and sensors, but only those the lightness values of which correlate with the analyte content. The coefficients of determination (r2) are calculated for all analytes, sensors and color channels, sensors and channels for which the value of r2 > 0.99 is a linearity criterion for the pharmaceutical analysis methods are identified. Thus, for the analysis of troxerutin, 4 color channels were selected (G4, G5, R7 and R8), for dydrogesterone – 5 channels (R2, G4, G6, B6 and R7), for ademetionine – 6 channels (R2, G2, R3, B3, G5 and G7).

To test the developed approach, a colorimetric analysis of the following drugs was carried out: tablets of dydrogesterone “Duphaston”^® 10 mg, capsules of troxerutin 300 mg and lyophilizate of ademetionine “Heptral”^® 400 mg. In order to select the optimal method for the identification of active substances, a comparison of the metrological characteristics of the methods using all the proposed color channels and sensors, was carried out. The content of the active substance in the drugs was determined by the calibration curve method. The results of the active substances identification in the indicated drugs using the developed approach, are presented in Table 4.

 

Table 4 – Results of active substances identification in medicinal products by multisensor digital colorimetry using various color channels and sensors

Sensor and color channel	Active ingredient content, mg/unit		Sr (for digital colorimetry)
	Spectrophotometry (n = 3, P = 0.95)	Digital colorimetry (n = 11, P = 0.95)
Dydrogesterone
R2	10.2 ± 0.1	11.1 ± 1.2	0.048
G4		8.4 ± 1.0	0.053
G6		7.0 ± 0.8	0.050
B6		6.2 ± 0.6	0.042
R7		14.4 ± 1.7	0.053
Troxerutin
G4	287 ± 2	294 ± 23	0.036
G5		284 ± 25	0.040
R7		290 ± 20	0.036
R8		291 ± 18	0.028
Ademetionine
R2	391 ± 4	393 ± 42	0.048
G2		395 ± 19	0.022
R3		389 ± 27	0.031
B3		388 ± 28	0.033
G5		400 ± 35	0.040
G7		376 ± 24	0.029

 

For all variants of colorimetric techniques, the equality of the means was proved using the modified Student’s t-test for independent samples (P=0.95). The table shows that methods of the troxerutin identification using the R-channel of sensor 7, of ademetionine – the G- channel of sensor 2, and of dydrogesterone – the R-channel of sensor 2, have the best metrological characteristics. The presented data show that the results of the analysis of the drugs by method of multisensor digital colorimetry, accord well with the data declared by the manufacturer (obtained by high performance liquid chromatography and spectrophotometric method).

Using the technique of the principal component analysis for the assay of dydrogesterone, troxerutin and ademetionine drugs.

An approach in which the set of lightness values of color channels is considered as a kind of “colorimetric spectrum”, seems promising. In this case the data can be processed using chemometric algorithms, of which the principal component analysis (PCA) is used most often. In this case, it is possible, on the one hand, to select all useful information from all sensors on all channels at once, on the other hand, the level of the information noise can be reduced and the accuracy of the analysis results can be increased.

To test chemometric approaches, a series of calibration solutions of the troxerutin and ademetionine substances (4.0–20.0 mg/ml) in increments of 4 mg/ml, and dydrogesterone (1.0–3.0 mg/ml) with a step of 0.5 mg/ml were used. The values of the first principal component (PC1) were calculated by the formulas.

For dydrogesterone:

PC1 = –0.01·∆G₁ – 0.31·∆R₂ – 0.02·∆G₂ – 0.23·∆B₂ –

– 0.01·∆R₃ – 0.01·∆G₃ – 0.35·∆B₃ – 0.01·∆R₄ – 0.21·∆G₄ – 0.01·∆B₄ +

+ 0.01·∆R₅ – 0.44·∆G₅ – 0.40·∆B₅ – 0.01·∆R₆ – 0.09·∆G₆ – 0.22·∆B₆ –

– 0.24·∆R₇ – 0.09·∆G₇ – 0.02·∆B₇ – 0.04·∆R₈ – 0.46·∆G₈ – 0.03·∆B₈

For troxerutin:

PC1 = 0.02·∆R₁ + 0.10·∆G₁ + 0.05·∆B₁ + 0.38·∆R₂ + 0.01·∆G₂ + 0.31·∆B₂ +

+ 0.17·∆R₃ + 0.48·∆G₃ + 0.21·∆B₃ + 0.13·∆G₄ + 0.14·∆B₄ +

+ 0.23·∆R₅ + 0.16·∆G₅ + 0.38·∆B₅ + 0.17·∆R₆ + 0.18·∆G₆ +

+ 0.23·∆R₇ + 0.01·∆G₇ + 0.02·∆B₇ + 0.27·∆R₈ – 0.02·∆G₈ + 0.02·∆B₈

For ademetionine:

PC1 = 0.02·∆R₁ + 0.01·∆G₁ + 0.27·∆R₂ + 0.54·∆G₂ + 0.03·∆B₂ +

+ 0.28·∆R₃ + 0.09·∆G₃ + 0.29·∆B₃ + 0.02·∆R₄ + 0.08·∆G₄ + 0.01·∆B₄ +

+ 0.01·∆R₅ + 0.16·∆G₅ + 0.01·∆G₆ + 0.13·∆B₆ +

+ 0.43·∆G₇ + 0.01·∆B₇ + 0.02·∆R₈ + 0.12·∆G₈ + 0.46·∆B₈

It can be notified that there is a linear correlation between the value of the first main component (PC1) and the content of dydrogesterone, troxerutin and ademetionine in calibration solutions (Fig. 2), which can be used to determine the content of these active substances in the drugs. The results of the analyses of the drugs for the identification of the active substances in the drugs using the developed approach, are presented in Table 5. The results obtained, accord well with the data declared by the manufacturer. Tables 4 and 5 show that the use of the principal component analysis improves the reproducibility of the analysis results in comparison with the use of the calibration dependence for the selected sensor and color channel.

 

Figure 2 – Dependence of the first main component vs concentration of dydrogesterone (a), troxerutin (b), ademetionine (c) in calibration solutions

 

Table 5 – Results of multisensor colorimetric identification of active substances in drugs by the principal component analysis

Active ingredient content, mg/unit		Sr (for digital colorimetry)
Spectrophotometry (n = 3, P = 0.95)	Digital colorimetry (n = 11, P = 0.95)
Dydrogesterone
10.2 ± 0.1	11.0 ± 0.8	0.031
Troxerutin
287 ± 2	290 ± 7	0.016
Ademetionine
391 ± 4	388 ± 9	0.020

 

CONCLUSION

An efficient approach (potentially having a wide application) has been proposed for a screening analysis of drugs of various pharmacological groups by multisensor digital colorimetry after a preliminary sample preparation. The simultaneous use of several chemical sensors in a chip provides sufficient selectivity. Discretization of the multisensor signal makes it possible to generate a unique barcode suitable for the identification of the active substances in drugs. The developed methods for the identification of active substances can serve as a good supplement to more expensive traditional methods.

FUNDING

The authors acknowledge the financial support of The Ministry of Science and Higher Education of the Russian Federation, the budget project of M.V. Lomonosov Moscow State University No. АААА-А21-121011590089-1 “Development of highly informative and high-tech methods of chemical analysis for the protection of ecosystems, the creation of new materials and advanced production technologies, the transition to environmentally friendly and resource-saving energy, the development of nature-like technologies, high-tech healthcare and rational use of natural resources”.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

AUTHORS’ CONTRIBUTION

All authors have equally contributed to the research work.

About the authors

Oksana V. Monogarova

Lomonosov Moscow State University

Author for correspondence.
Email: o_monogarova@mail.ru
ORCID iD: 0000-0002-5790-1462

Associate Professor, Candidate of Sciences (Chemistry), Department of Chemistry, Analytical Chemistry Division

Russian Federation, 1-3, Leninskie gory, Moscow, 119991

Aleksandr A. Chaplenko

I.M. Sechenov First Moscow State Medical University (Sechenov University)

Email: a.a.chaplenko@yandex.ru
ORCID iD: 0000-0003-1176-4658

Associate Professor, Candidate of Sciences (Pharmacy)

Russian Federation, 2-4 Bolshaya Pirogovskaya str., Moscow, 119435

Kirill V. Oskolok

Lomonosov Moscow State University

Email: k_oskolok@mail.ru
ORCID iD: 0000-0002-7785-4835

Associate Professor, Candidate of Sciences (Chemistry), Department of Chemistry, Analytical Chemistry Division

Russian Federation, 1-3, Leninskie gory, Moscow, 119991

References

Supplementary files