aDepartment of Foods and Drugs, Faculty of Pharmaceutical Sciences, Laboratory of Medicinal Plants and Herbal Medicines, Federal University of Alfenas, UNIFAL-MG, Gabriel Monteiro da Silva Street, 700, 37130-000, Alfenas, Minas Gerais, Brazil, bDepartament of Microbiology and Immunology, Institute of Biomedical Sciences, Laboratory of Research in Microbiology, Federal University of Alfenas, UNIFAL-MG, Gabriel Monteiro da Silva Street, 700, 37130-000, Alfenas, Minas Gerais, Brazil, cInstitute of Biosciences, Laboratory of Bioprospecting of Natural Products (LBPN), São Paulo State University, UNESP, Coastal Campus-IB-CLP, Infante Dom Henrique Square, s/n°, 11380-972, São Vicente, São Paulo, Brazil
Email: rgbastos.rb@gmail.com
Received: 25 Jan 2016 Revised and Accepted: 20 Apr 2016
ABSTRACT
Objective: The present study aimed to characterize and quantify the total phenolics, flavonoids and tannins in a hydroethanolic crude extract (70% (v/v) (EB)) of the leaves of E. florida DC, as well as to evaluate the antimicrobial activity of the extract against different species of micro-organisms.
Methods: EB was characterized using a mass spectrometer equipped with a direct insertion device for in-stream injection (FIA). Quantitative analyses of major compounds were carried out by spectrophotometry. In addition, we evaluated the sensitivity profiles of different strains of yeast and bacteria against different concentrations of EB.
Results: The classes found were in agreement with those described in the literature: flavonoids, tannins, phenolic acids and saponins. EB showed levels of phenolic compounds, flavonoids and tannins equal to 25.82 mg gallic acid equivalents per gram of extract (EAG/g), 8.42 mg quercetin equivalents per gram of extract (EQ/g) and 7.30 mg tannic acid equivalents per gram of extract (AT/g), respectively. In the analysis of antimicrobial activity, EB was more active against yeasts but was not effective against the bacteria used in the test.
Conclusion: We can conclude that E. florida DC has antimicrobial potential, due to the presence of bioactive secondary metabolites.
Keywords: Phenolics, Micro-organisms, Mass spectrometer, Yeasts, Bacteria, Potential
© 2016 The Authors. Published by Innovare Academic Sciences Pvt Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/)
INTRODUCTION
Medicinal plants are used in some of the oldest strategies employed by man in the treatment of diseases of all kinds, and their use in the prevention or cure of diseases is a habit that has always existed in the history of humanity [1]. In recent decades in Brazil, there has been a growing interest in the use of medicinal plants and their respective extracts for treatments, which can in certain circumstances, be of assistance in primary health care compatible with conventional medicine [2].The phytotherapy is intended to provide useful information for the development of pharmacological and phytochemicals studies on plant species, as well as providing therapeutic resources, conserving biodiversity, promoting social development, and fostering permanent education [2].
In contrast to synthetic drugs, therapeutic use of herbal products was materialized from the ancient texts and passed on through generations without scientific experimental evidence. For this reason, modern research on the pharmacology of herbal medicine is aimed to validate herbal medicine and to ensure rational use, safety and efficacy [3].Thus, due to the large diversity of Brazilian plants with therapeutic potential and considering the history and cultures of the population, some species of plants have aroused great interest among researchers [4].
The species Eugenia florida DC (Myrtaceae) is a small tree, approximately 3.5 m tall, that blooms in October and fruits at the end of November to mid-December. It is distributed throughout South America, from the Colombia to Bolivia, and from Guyana to the South of Brazil, growing in various ecological landscapes. It is popularly known as black cherry or guamirim cherry. Its fruits have a characteristic odor, and are edible and very tasty [5].The extensive therapeutic use of the leaves of the species by rural populations has been highlighted. It is used primarily as a hypotensive, hypoglycemic, and antipyretic; for reducing triglycerides and cholesterol levels; and in the treatment of infections, jaundice, heart disease and gastrointestinal disorders [6, 7].
Previous chemical studies of the genus indicate high levels of phenolic compounds such as flavonoids [8, 9], phenolic acids [10] and tannins [11, 12], as well as other classes of compounds, including saponins [13, 14] and terpenoids [15].
However, there are still no detailed chemical and biological studies on the species E. florida DC in the literature. Thus, chemical characterization and elucidation of the antimicrobial activity of the plant extract in question, to verify its potential as a medicinal agent, is justified due to the popular use of the plant by rural communities in the treatment of diseases.
MATERIALS AND METHODS
Plant material
Leaves of the species E. florida DC were collected in April 2014, the southern state of Minas Gerais, near the “Fazenda Cachoeirinha”, W045°56'07.8'', S21°33’51.7’’, Alfenas, Minas Gerais, Brazil. The details of the data collection and the coordinates of the site were provided by Dr. Flávio Nunes Ramos (Institute of Natural Sciences, UNIFAL-MG) and the species was identified by Dr. Geraldo Alves da Silva (Faculty of Pharmaceutical Sciences, UNIFAL-MG). A specimen was deposited and recorded in the Herbarium of the UNIFAL-MG under the registration number 2562. The leaves were dried in an oven with air renewal at 45oC for 72 h. After drying, the plant material was sprayed onto a knife mill.The powder obtained was subjected to exhaustive percolation using ethanol (70%) as a liquid extractor. The extraction solution was concentrated in a rotary evaporator under reduced pressure and subsequently lyophilized and stored, finally obtaining the crude hydroethanolic extract 70% (v/v) (EB).
Chemical characterization by mass spectrometry
For analysis by mass spectrometry, EB of E. florida DC (5.0 mg) was dissolved in methanol (3.0 ml), and filtered in a Sepak RP-18® cartridge and then through a Nylon membrane (Flow Supply®) with 22.25 mm diameter and 0.22 mm pore size. The extract (5 ppm) was analyzed online by the LCQ Fleet, Thermo Scientific® mass spectrometer, equipped with a direct sample insertion device for streaming injection analysis (FIA). The sample was ionized by electro spray (ESI) and fragmentations into multiple stages (MSn) were held in an ion-trap (IT) interface. The negative mode was chosen for the generation and analysis of all the spectra, and the experimental conditions were as follows: capillary voltage-35 V, spray voltage-5000 V, capillary temperature 350oC, drag gas (N2) and flow rate 60 (arbitrary units). The acquisition track was m/z 100-2000, with two or more scan events held simultaneously in the spectrum. The experiment was performed by Laboratory of Bioprospecting of Natural Products (LBPN), UNESP, Coastal Campus-IB-CLP.
Determination of total phenol content
The determination of total phenols was performed according to the colorimetric method based on the formation of complexes with molybdenum and tungsten [16]. Aliquots of the dry extract were diluted in absolute ethanol, to obtain a concentration of 250 mg/ml. 0.5 ml of the sample was added to 2.5 ml of Folin-Ciocalteau reagent 10% (v/v) and 2.0 ml of Na2CO3 4% (m/v). The samples were homogenized and kept at rest for 2 h at room temperature and under direct light. After this period, the absorbance at 750 nm was measured in a Shimadzu® TCC 240A UV-Vis spectrophotometer. The same procedure was performed with gallic acid in ethanolic solution (5, 10, 20, 30, 40, 50, 60, and 70 mg/ml), in order to obtain the analytical curve A = 0.0124C+0.0089 (R2 = 0.9877), where A is the absorbance at 750 nm and C is the concentration of gallic acid. The values obtained for the samples were interpolated on the calibration curve and the concentration values were expressed as milligrams of gallic acid equivalents per gram of extract (mg EAG/g). The experiment was performed in triplicate.
Determination of flavonoid content
Flavonoids were quantified based on the formation of flavonoid-aluminum complexes [17]. Extract aliquots were diluted in absolute ethanol to obtain a concentration of 500 mg/ml. 0.5 ml of sample was added to 0.1 ml of AlCl3. 6 H2O 10% (m/v), 0.1 ml of CH3COOK 1.0 mol/l and 2.8 ml of distilled water. The tubes were shaken to homogenization and kept at rest for 30 min. Absorbance at 425 nm were read in a Shimadzu® TCC 240A UV-Vis spectrophotometer. The same procedure was followed using quercetin in ethanolic solution (10, 20, 40, 50, 60, 70, 80, and 100 mg/ml), in order to obtain the analytical curve A = 0.0075C+0.0216 (R2 = 0.9982), where A is the absorbance at 425 nm and C is the concentration of quercetin. The values obtained for the sample were interpolated on the calibration curve and the concentration values were expressed as milligrams of quercetin equivalents per gram of extract (mg EQ/g). The experiment was performed in triplicate.
Determination of tannin content
Tannins in EB were determined using the method of Hagerman, based on precipitation with albumin and complexation with ferric chloride solution [18]. Extract aliquots were diluted with ethanol at a concentration of 1000 mg/ml. 2.0 ml of bovine albumin 1.0% (m/v) was added to 1.0 ml of extraction solution and, after precipitation, the samples were maintained at room temperature for 15 min and centrifuged at 3000 rpm. The supernatant was discarded and the precipitates were dissolved in 4.0 ml of sodium lauryl sulfate 0.5% (m/v). 1.0 ml of ferric chloride solution 1% (m/v) was added, and the tubes were homogenized and kept at rest for 15 min. Absorbance at 510 nm was measured in a Shimadzu® TCC 240A UV-Vis spectrophotometer. The same procedure was followed with tannic acid in ethanolic solution (10, 20, 40, 50, 60, 70, 80, and 100 mg/ml), in order to obtain the analytical curve A = 0.0059C+0.0014 (R2 = 0.9846), where A is the absorbance at 510 nm and C is the concentration of tannic acid. The values obtained for the sample were interpolated on the calibration curve and the concentration values were expressed as milligrams of tannic acid equivalents per gram of extract (mg EAT/g). The experiment was performed in triplicate.
Evaluation of antimicrobial activity against yeasts and bacteria
The extract was evaluated in vitro for antifungal and antibacterial activities using Mueller-Hinton Broth and the microdilution method following the methodology and interpretative criteria proposed by document M27-S4 for yeasts and document M7A6 for bacteria from the Clinical and Laboratory Standards Institute [19,20], with some modifications. The stock solutions of the extract were prepared in 1% DMSO and tested at the following concentrations (μg/ml): 1000; 500; 250; 125; 62.5; 31.25; 15.63; 7.81; 3.91; 1,95. The standard drugs chloramphenicol and fluconazole were applied as controls of bacteriostatic and fungistatic actions, respectively, at the following concentrations (μg/ml): 64; 32; 16; 8; 4; 2; 1; 0.5; 0.25; 0.125.
The tests were performed with opportunist and pathogenic yeast and bacterial standards from the American Type Culture Collection (ATCC): Candida albicans ATCC 10231, C. krusei ATCC 6258, C. glabrata ATCC 90030, C. parapsilosis ATCC 22019, Staphylococcus aureus ATCC 6538, and Escherichia coli ATCC 25922. The microorganism strains were obtained by Drª. Amanda Latércia Tranches Dias (Laboratory of Research in Microbiology, Institute of Biomedical Sciences, UNIFAL-MG). The microplates were incubated at 37oC for 24 h. Results were visualized and analyzed at 530 nm in an Anthos Zenyth 200rt Microplate Reader®. The concentrations of the compounds inhibitory to microbial growth were determined by identifying the concentrations at which a reduction of 50% (MIC50) and 90% (MIC90) in the absorbance values compared to the absorbance values of maximum growth of the isolates, in μg/ml. The tests were performed in triplicate.
Statistical analysis
Statistical analyses were performed using the Sisvar software Version 5.3/DEX-UFLA (Copyright, 1999-2010). Analysis of variance (ANOVA) was performed followed by Scott & Knott tests with a significance level of 5%. Correlation analyses were performed in the BioEstat software Version 5.0 and the results were presented as Spearman’s correlation coefficients (r).
RESULTS
The high percentage of yield observed for EB, equivalent to 23.89%, suggests that the most common compounds present in the extract are polar since the solvent used (ethanol 70%) offers a sufficient degree of polarity to extract such compounds [21].
Chemical characterization by mass spectrometry
The mass spectrum of EB, in full scan mode, is shown in fig. 1. The compounds obtained were registered as their respective ions [M-H]-. For these ions, partitions in multiple stages were obtained, which allowed the structures of some molecules to be suggested based on their fragmentation mechanisms, as described in table 1.
Quantitative analysis of main classes of compounds
The results of the quantification of the total phenols, flavonoids and tannins in EB of E. florida DC are shown in table 2. EB presented levels of each of these compounds, respectively, equivalent to 25.82 ± 0.85 mg EAG/g, 8.42 ± 1.13 mg EQ/g, and 7.30 ± 1.06 mg EAT/g.
Evaluation of antimicrobial activity
As shown in table 3, EB exhibited MIC50 values between 62.5 and 250 mg/ml. EB was more active against the yeast C. krusei, with an MIC50 calculated as 62.5 mg/ml. The extract did not achieve a growth reduction of 50 % at the concentrations tested for C. albicans and C. parapsilosis. EB showed a MIC90 for E. coli of 1000 mg/ml, demonstrating a lower activity against bacteria.
Fig. 1: Mass spectrum in full scan mode obtained by direct injection of EB of E. florida DC leaves
Table 1: Mass spectrometry by direct injection and electrospray ionization of EB of E. florida DC
Compound |
Fragmentation by FIA-ESI-IT-MSn |
Reference |
|
[M-H]- |
Major ions produced (MSn) |
||
Anhydrous hexose |
161 |
--- |
[22] |
Quinic acid |
191 |
173, 127, 111 |
[23] |
Resveratrol |
227 |
--- |
[24] |
5-hydroxyresveratrol |
243 |
227, 215, 137 |
[25] |
1,2-didehydrotanshinone IIA |
293 |
277, 249, 221, 207 |
[26] |
Quercetin |
301 |
--- |
[27] |
4-allyl-1-hydroxy-2-(2’-allyl-4’-hydroxy-5’-methoxyphenoxy)-benzene |
311 |
283 |
[28] |
4-O-p-coumaroylquinic acid |
337 |
191, 173, 127, 111 |
[29] |
5-caffeoylquinic acid |
353 |
179 |
[30] |
5-hydroxyeuphene |
427 |
409 |
[31] |
Quercetin-O-arabinose |
433 |
301 |
[32] |
Quercetrin |
447 |
301 |
[18] |
Rhododendrin penthoside |
459 |
327, 165 |
[33] |
Quercetin-O-glucose |
463 |
301 |
[18] |
Rhamnetin-O-glucose |
477 |
315 |
[34] |
Granatunflavanil-O-xylose |
481 |
349 |
[35] |
Dystenin |
514 |
482 |
[36] |
O-galloyl desbenzoylpaeoniflorin |
527 |
331, 169 |
[37] |
Procyanidin B2 |
577 |
533, 427, 289 |
[38] |
Hyperin |
595 |
433, 301 |
[39] |
Betulinic acid heteroside |
617 |
455, 411 |
[40] |
Cimicifoetiside A |
619 |
487 |
[41] |
3-O-D-glucopyranosylserjanic acid |
675 |
499, 455, 397 |
[42] |
3-O-caffeoyl-4,5-di-O-feruloylquinic acid |
705 |
543, 529, 367, 191, 173 |
[43] |
Kochianoside IV |
763 |
631, 455, 411 |
[44] |
27-O-D-glucopyranosyl physagulin D |
783 |
621, 459 |
[45] |
Procyanidin C2 |
865 |
823, 715, 577, 533, 425, 289 |
[46] |
Isoastragaloside I |
867 |
705, 489 |
[47] |
Clinopodiside E |
994 |
832, 670, 523 |
[48] |
Aesculuside B |
1006 |
844, 682, 506 |
[49] |
Phytolacagenic acid heteroside |
1016 |
854, 692, 530 |
[50] |
Guaianin D |
1026 |
880, 734, 572, 440 |
[51] |
Tetrameric cathequin |
1154 |
1110, 1002, 865, 823, 715, 577, 533, 425, 289 |
[46] |
Scillasaponin B |
1266 |
1104, 942, 780, 618, 486 |
[52] |
Table 2: Contents of total phenols, flavonoids and tannins in EB of E. florida DC.
Compounds |
Content |
Total phenols (mg EAG/g) |
25.82 ± 0.85c* |
Flavonoids (mg EQ/g) |
8.42 ± 1.13b* |
Tannins (mg EAT/g) |
7.30 ± 1.06a* |
Values expressed as average ± standard deviation (n = 3). The results of each test were analyzed separately. Averages with different letters in the same column are statistically different by Scott and Knott test, *p<0.05.
Table 3: Evaluation of the sensitivity profile of yeasts and bacteria on EB of E. florida DC
Microorganism |
MIC50 FCN (mg/ml) |
MIC50 CFN (mg/ml) |
MIC50 EB (mg/ml) |
MIC90 EB (mg/ml) |
Candida albicans |
0.5a* |
--- |
--- |
500b* |
Candida krusei |
32a* |
--- |
62.5b* |
500c* |
Candida glabrata |
16a* |
--- |
250b* |
500c* |
Candida parapsilosis |
1a* |
--- |
--- |
500b* |
Staphylococcus aureus |
--- |
0.975a* |
250b* |
500c* |
Escherichia coli |
--- |
0.975a* |
500b* |
1000c* |
Values expressed as average (n = 3). The results of each test were analyzed separately. Averages with different letters in the same line are statistically different by Scott and Knott test, *p<0.05. FCN, fluconazole; CFN, chloramphenicol.
DISCUSSION
When comparing the full scan spectrum of EB of E. florida DC (fig. 1) with the existing data in the literature, it might be said that there was agreement among the classes of secondary metabolites and putatively identified molecules, specially the flavonoids, tannins, phenolic acids, saponins and terpenoids [53].
Phenolic compounds are widely distributed in plants and have multiple biological activities, including prevention and treatment of many diseases, including microbial infections. They are a chemically heterogeneous group, in which most of the compounds are soluble in polar solvents, such as mixtures of water and ethanol [54]. This might account for the high concentrations of phenolic compounds in EB, since the efficiency of their extraction may be relatively higher.
The correlation of the results for total phenols in this study with the type of solvent used in the extraction, and with the chemical characterization by mass spectrometry, suggests the presence of polar phenolic compounds in EB, with an emphasis on those in the heterosidic form. Their relatively high concentration in this extract is likely due to the high solubility in ethanol 70% (v/v).
Among the naturally present compounds in the plants, the significant presence of flavonoids is due to their capacity for protection against ultraviolet radiation. Due to the presence of aromatic rings associated with hydroxyls in their chemical structures, flavonoids have the ability to capture and stabilize radiation avoiding damage to DNA. Flavonoids are relatively important phenolic compounds, especially with regard to their health benefits, which arise from their demonstrated activities [55, 56].
Tannins have been the subject of several studies, most of which discuss the different ecological interactions between vegetable tannins and herbivores. Some research on the biological activity of tannins has shown significant activity against certain microorganisms. In addition, they may act as anti-inflammatories, promote healing and inhibit HIV reverse transcription [57].
When EB obtained from the leaves of E. florida DC is compared with rich plant extracts such as green tea (Camellia sinensis), which features around 30 to 55 mg/g of phenolic compounds, it can be observed that EB had only slightly lower levels, equivalent to 25.8 mg/g [58].
Total phenols and flavonoids in a hydroethanolic extract of leaves of katukorandi (Eugenia singampattiana), a species found in India and used by indigenous tribes in the treatment of rheumatism and respiratory diseases, were determined to be, respectively, 21.58 mg EAG/g and 9.10 mg EQ/g [59]. These total phenol levels are very close to those found in the present study, however the flavonoid content is slightly higher than that found for EB.
Total phenols and flavonoids present in an aqueous extract of “pedra-ume-caá” (Eugenia punicifolia) leaves were determined as 21.6 mg EAG/g and 2.62 mg EQ/g, respectively [60]. The total phenol content is quite similar to that in the present study for EB, however the flavonoid content was very low.
However, one must take into consideration that the extract used by the researchers of this study was aqueous, suggesting that the extractor liquid used in the preparation of EB was more efficient in the removal of flavonoid glycosides.
The determination of total tannin content in samples of ethanolic extracts of Eugenia calycina revealed concentrations of these compounds equal to 2.43 mg EAT/g [18], lower than that found in EB in this study.
Differences in the levels of total phenols, flavonoids and tannins in this study, with those found in the literature, are due to differences in the various parameters employed in obtaining the extracts. For example, the processes of collecting plant material (date, time, site), the extractive method used, the solvent extractor, the processes of ecological plant interaction with the environment, the age and the stage of plant development can all have major effects. In addition, environmental conditions such as light, temperature, rainfall, radiation, seasonality, availability of water and atmospheric condition interfere directly in the production of secondary metabolites of biological and pharmacological interest. It should also be noted that some differences found in the levels of these metabolites in the literature might be related to different methodologies used for the quantitative determination of the compounds [61].
The extracts which have an MIC50 less than 100 mg/ml are considered to have excellent antimicrobial activity, between 100-500 mg/ml is considered moderate, those between 500-1000 mg/ml are slightly active and those with a MIC50 greater than 1000 mg/ml are inactive [62]. With respect to C. krusei, EB demonstrated excellent antimicrobial activity (MIC50 = 62.5 mg/ml). Against C. glabrata and S. aureus it was moderately active (MIC50 = 250 mg/ml), and it was slightly active against E. coli (MIC50 = 500 mg/ml). It was inactive against C. albicans and C. parapsilosis. The excellent antimicrobial activity of EB against C. krusei is important because this yeast is quite resistant to conventional antifungals, such as fluconazole (MIC50 = 32 mg/ml). This result demonstrates that the extract has significant potential as an antimicrobial against C. krusei.
The fungal inhibition capacity of phenolic compounds in an extract is due to the stabilization of cellular free radicals, preventing damage to infected cells. By contrast, tannins can inhibit fungal enzymes by complexation with enzyme substrates and metal ions, as well as modifying the microbial cell membrane, altering fungal metabolism [63].
Fungal infections in humans are caused mainly by allergic reactions to fungal proteins or toxins. These infections are responsible for the increased rate of morbidity and mortality in immunocompromised and severely ill patients. Candida sp. are yeasts that cause mycosis infections around the world. C. albicans and C. krusei are major pathogens responsible for the bulk of nosocomial infections. The yeasts C. glabrata and C. parapsilosis can grow as biofilms in materials used in hospitals such as catheters, needles, and other medical devices. They are considered as additional risk factors in the use of antibiotics [64].
EB of E. florida DC showed little activity on E. coli, which is a Gram-negative bacterium, compared with the Gram-positive bacterium S. aureus. This may be due to different compositions of the bacterial membranes. Gram-negative bacteria have an additional component in their structure: an outer lipid bilayer composed of lipopolysaccharides (LPS) firmly adhered to the peptidoglycan layer of the cell wall. This makes the bacterial cell membrane more lipophilic towards exogenous substances, whose power of penetration and/or action on the membrane determines their antibiotic ability. Therefore, it is suggested that the flavonoids and tannins have the ability to complex with and/or inactivate enzymes, extracellular proteins, soluble proteins and bacterial cell walls, which determines their likely mechanisms of antimicrobial activity [65].
The antimicrobial activity of a hydroethanolic extract 70% (v/v) of Eugenia uniflora leaves on strains of S. aureus and C. albicans revealed minimum inhibitory concentration values of 90% of the microbial growth (MIC90) of 80 and 500 mg/ml, respectively [66]. These values were demonstrated better antimicrobial activity than that evidenced by the MIC90 of EB for S. aureus and equal to C. albicans, in the present study.
Antimicrobial activity determination of ethanolic extracts of Eugenia calycina leaves against S. aureus, E. coli, C. albicans and C. parapsilosis demonstrated MIC50 values equal to 1000, 2000, 500 and 250 mg/ml, respectively [18]. These MIC50 values demonstrated weaker antimicrobial activity than for EB with respect to bacteria. However, they were more effective against the Candida yeasts, against which EB did not have significant activity.
It is important to consider the fact that the treatment of microbial infections is not always effective because conventional antimicrobials can cause recurrence of infections, induce resistance in microorganisms and generate toxicity. For these reasons, extracts and natural compounds have been investigated by researchers as new alternatives for the treatment of microbial diseases [66].
CONCLUSION
In this context, studies of chemistry and in vitro antifungal and antibacterial activities of EB of E. florida DC show relevant importance in phytotherapy. They can assist in improving the screening of chemical compounds in medicinal plants with antimicrobial activity and will aid the identification of potential plant sources for more effective and less toxic therapies.
ACKNOWLEDGMENT
The first author, Sr. Renan G. Bastos, would like to express heart full thanks to the Federal University of Alfenas (UNIFAL-MG) and to BIOTA-FAPESP, CAPES, FAPEMIG and CNPq for financial assistance.
CONFLICT OF INTERESTS
All authors have none to declare
REFERENCES