The cellobiose lipid of Cryptococcus humicola, 16-(tetra-O-acetyl-β-cellobiosyloxy)-2-hydroxyhexadecanoic acid, is a natural fungicide. Sensitivity of the cells of Saccharomyces cerevisiae to the fungicide depends on a carbon source. Cellobiose lipid concentrations inducing the leakage of potassium ions and ATP were similar for the cells grown in the medium with glucose and ethanol. However, the cells grown on glucose and ethanol died at 0.05 mg ml-1 and 0.2 mg ml-1 cellobiose lipid, respectively. Inorganic polyphosphate (PolyP) synthesis was 65% of the control with 0.05 mg ml-1 cellobiose lipid during cultivation on ethanol. PolyP synthesis was not observed during the cultivation on glucose at the same cellobiose lipid concentration. The content of longer-chain polyP was higher during cultivation on ethanol. We speculate the long-chained polyP participate in the viability restoring of ethanol-grown cells after treatment with the cellobiose lipid.
Some yeasts and mycelial fungi produce glycolipids of various types such as mannosylerythritols, sophorolipids, and cellobiose lipids. They possess multiple biological activities: they act as biosurfactants, facilitate dissolution and consumption of organic hydrophobic compounds, and display fungicidal activity (Kitamoto et al.2002; Cameotra and Makkar 2004; Golubev 2006; Rodrigues et al.2007). Cellobiose lipids display antifungal activity against many species of ascomycetous and basidiomycetous yeast and mycelial fungi including know pathogens, Filobasidiella (Cryptococcus) neoformans and Candida albicans (Puchkov et al.2002; Kulakovskaya et al.20052009; Mimee et al.2005; Bölker et al. 2008; Hammami et al.20102011). The broad spectrum of activity, pH and temperature stability allows considering cellobiose lipids as promising compounds for the development of novel fungicides for medical and agricultural applications. So, many studies are now performed in their biochemistry, genetics and possible ecology role (Teichmann et al.2007, Bölker et al.2008; Hammami et al.20102011).
The mechanism of action of cellobiose lipids on yeast cells is based on enhancement of nonspecific permeability of the cytoplasmic membrane, which results in the rapid leakage of ATP and potassium ions from the yeast cells treated with these compounds (Kulakovskaya et al.20052008). The glycolipids are surface-active compounds reducing the surface tension of water solutions. Cellobiose lipids of Cr. humicola have a high surface activity comparable with that of SDS (Puchkov et al.2002). The intercalation of glycolipid of Cr. humicola into liposomes containing diphytanoylphosphatidylcholine, ergosterol, and phosphatydilserine was demonstrated (Puchkov et al.2002). These data suggest that the mycocidal effect of cellobiose lipids is associated with its detergent-like properties. Based on these observations and on the electrical measurements on planar phospholipid bilayers, which showed glycolipid-induced membrane permeabilization, it was suggested that the cytoplasmic membrane is the primary target of cellobiose lipid activity (Puchkov et al.2002).
The fungal species are known to have different sensitivity to cellobiose lipids (Kulakovskaya et al.20052009; Mimee et al.2005). For example, the effective concentrations against basidiomycetes (Filobasidiella neoformans) and ascomycetes (Candida spp.) are 0.03 mM and 0.1-0.4 mM, respectively (Kulakovskaya et al.2009). The causes of such difference have not yet been investigated. It is unknown whether cultivation conditions, including those affecting the state of the cytoplasmic membrane, influence the sensitivity of target cells.
The cultivation in ethanol-containing media substantially changes the properties of the cytoplasmic membrane of Saccharomyces cerevisiae compared to cultivation in glucose-containing media (Susan et al.1978; Beaven et al. 1982; Mishra and Prasad 1989; Walker-Caprioglio et al.1990; Herve A et al.1994; Kubota et al.2004). During the cultivation on ethanol, the proportion of ergosterol and mono-unsaturated fatty acid residues in cellular phospholipids increases and the fluidity of membrane decreases (Susan et al.1978; Beaven et al.1982; Mishra and Prasad 1989; Walker-Caprioglio et al.1990; Herve A et al.1994; Kubota et al.2004). Inorganic polyphosphate (PolyP) is an energy reserve and a stress-protective compound for microbial cells (Kulaev et al.2004; Rao et al.2009; Achbergerová and Nahálka 2011). The content and chain length of these bioactive polymers in Saccharomyces cerevisiae depend on carbon source (Vagabov et al.2008). So, cultivation on glucose or ethanol allows obtaining the cells of S. cerevisiae which differ in membrane fluidity and PolyP content.
The objective of this work was to compare the sensitivity of S. cerevisiae cells grown on glucose and ethanol to the fungicide 16-(tetra-O-acetyl-β-cellobiosyloxy)-2-hydroxyhexadecanoic acid secreted by Cryptococcus humicola (Kulakovskaya et al.2009; Morita et al.2011). An attempt was made to assess the relationship between PolyP accumulation and sensitivity to cellobiose lipid.
Results and discussion
The cellobiose lipid preparation used in the work was obtained from the culture liquid of Cr. humicola strain 9-6 (All-Russian Collection of Microorganisms). Mass spectrometry shows that the major component of the preparation has a molecular mass 781 kDa. This compound is a 16-(tetra-O-acetyl-β-cellobiosyloxy)-2-hydroxyhexadecanoic acid (cellobiose lipid) according to earlier data (Puchkov et al.2002; Kulakovskaya et al.2009) (Figure 1).
Figure 1
The structure of cellobiose lipid secreted byCryptococcus humicolastrain 9-6. R - acetate.
The cells of S. cerevisiae proved to have different survival capacities at the same cellobiose lipid concentrations depending on the carbon source used. The cells grown on glucose died at a concentration of 0.05 mg ml-1, while the cells grown on ethanol died at 0.2 mg -1 (Table 1).
Table 1
The viability of the cells ofSaccharomyces cerevisiaegrown in the media with glucose and ethanol treated with cellobiose lipid
Cellobiose lipid, mg ml-1
Cell viability, % of control
Glucose grown cells
Ethanol grown cells
0
100
100
0.025
60 ± 12
-
0.034
11 ± 1.1
-
0.050
4 ± 1.7
100
0.100
0.2
16 ± 5.1
0.200
0.1
2 ± 0.5
0.400
0
1.5
0.800
0
0
- not assayed.
Cellobiose lipid shows fungicidal activity in acidic medium, where it is a weak acid due to dissociation of the carboxyl group (Puchkov et al.2002; Kulakovskaya et al.2009). The average values of electrokinetic potential (EKP) were calculated to be 18.8±1.2 and 23.6±3.0 mV for the cells of S. cerevisiae grown on glucose and ethanol, respectively. It is probable that the high negative surface charge decreases the binding of negatively charged molecules of the fungicide.
One of the known criteria of yeast cytoplasmic membrane integrity damage is the leakage of potassium ions into the medium (Kulakovskaya et al.2008; Shirai et al.2009). The effective cellobiose lipid concentrations inducing K+ leakage were not different for the cells of S. cerevisiae grown on both carbon sources (Figure 2). Consequently, the high stability of S. cerevisiae cells grown on ethanol cannot be explained solely by intensification of the barrier functions of the membrane.
Figure 2
Potassium leakage from the cells ofSaccharomyces cerevisiaetreated with cellobiose lipid ofCr. humicola: 1 and 3 - the cells were grown in the medium with glucose, 2 and 4 - the cells were grown in the medium with ethanol. The concentration of cellobiose lipid was 0.24 mg ml-1 (1 and 2) and 0.12 mg ml-1 (3 and 4).
The energy of the phosphoester bond in PolyP is similar to that of ATP. PolyP is known to be a factor of microbial cell resistance to stress conditions (Kulaev et al.2004; Rao et al.2009; Achbergerová and Nahálka 2011). Hence, the effect of cellobiose lipid on the PolyP, Pi and ATP content in the cells has been studied. The experiments were performed under the conditions of PolyP synthesis. The cells with the PolyP content of 50–65 µmole P/g dry biomass (not shown) were cultivated in the complete medium for 30 min. Then PolyP fractions with different chain lengths were extracted. PolyP accumulation was observed in both media: with glucose and with ethanol (Table 2, Figure 3). PolyP synthesis was almost completely suppressed by cellobiose lipid in the medium with glucose (Figure 3, Table 2) but only by 35% lower in the medium with ethanol.
Table 2
The effect of cellobiose lipid on the content of Pi, PolyP and ATP (µmol g-1dry biomass) in the cells ofS. cerevisiaeunder PolyP synthesis: the cells after Pistarvation were cultivated in the complete medium with 10 mM Pifor 30 min
Culture condition
Glucose ethanol
Glucose ethanol
Glucose ethanol
Pi
PolyP
ATP
Control
45
41
870
825
5.1
13
Cellobiose lipid, 0.05 mg ml-1
3.4
20
80
560
1.2
6.0
Figure 3
The accumulation of different fraction of inorganic polyphosphate (PolyP) in the cells ofSaccharomyces cerevisiaefor 30 min of cultivation in phosphate-rich medium with glucose (A) and ethanol (B). (□) - control, (■) - in the presence of cellobiose lipid, 0.05 mg ml-1.0.
The chain length of PolyP of different fractions determined by electrophoresis in PAAG did not depend on the carbon source in the presence and absence of cellobiose lipid (Table 3). The content of longer-chain fractions was higher during cultivation on ethanol (Figure 3 and Table 3). During the cultivating on glucose, the Pi and ATP content decreased in the presence of cellobiose lipid much more than during the cultivation on ethanol (Table 2).
Table 3
Polyphosphates inSaccharomyces cerevisiaeafter 30 min cultivation in the media with glucose or ethanol: the average chain length and proportion of PolyP of different fractions
Fraction
Average chain length (n)
% of total PolyP content
Glucose
Ethanol
PolyP1
15
37
20
PolyP2
25
16
9
PolyP3
65
35
41
PolyP4
75
3
6
PolyP5
>200
9
24
The cells after Pi starvation were cultivated in the complete medium with 10 mM Pi.
We have also determined the effect of cellobiose lipid on ATP leakage from cells in phosphate-citrate buffer (pH 4.0). The effective glycolipid concentrations were different for the cells grown on glucose and ethanol (Table 4). The addition of glucose decreased the effective concentration of the fungicide only for glucose-grown cells. This effect was not observed in the cells grown on ethanol.
Table 4
The concentrations of cellobiose lipid (mg ml-1) causes the maximal leakage of ATP in phosphate-citrate, pH 4.0 fromS. cerevisiaecells
Incubation medium
Cells grown in medium with glucose
Cells grown in medium with ethanol
The concentration of cellobiose lipid
0.04 M citrate-phosphate, pH 4.0
0.6
0.6
0.04 M citrate-phosphate, pH 4.0, 2% glucose
0.15
0.6
The sensitivity of the cells of S. cerevisiae to 16-(tetra-O-acetyl-β-cellobiosyloxy)-2-hydroxyhexadecanoic acid depends on the carbon source used for cell cultivation. The cells grown on ethanol are more resistant to this membrane damaging fungicide. It is probably due to the increase of the negative charge of cell surface (EKP) during cultivation on ethanol and to the change in membrane lipid composition. However, the effects of cellobiose lipid on potassium leakage were similar for the cells grown in the media with glucose or ethanol. It suggests the existence of additional factors increasing the resistance of yeast grown on ethanol to cellobiose lipid. We speculate the long-chained polyP participate in the restoring of viability of ethanol-grown cells after treatment with the cellobiose lipid.
Conclusion
The sensitivity of yeast cells to antifungal cellobiose lipids depends on culture conditions especially on carbon source. The peculiarities of growth conditions of target microorganisms should be taken into account when assessing effective concentrations of these new fungicidal compounds.
Materials and methods
Strains and growth conditions
The yeast Saccharomyces cerevisiae strain VKM Y-1173 was grown in a shaker in the Reader medium with 0.2% yeast extract, 2% glucose (120 r.p.m.), or 1% ethanol (200 r.p.m.). The medium contained (g l-1): (NH4)2SO4, 3; MgSO4, 0.7; Ca(NO3)2, 0.4; NaCl, 0.5; KH2PO4, 1; K2HPO4, 0.1; (NH4)2SO4. FeSO4. 6H2O, 0.00025; and trace elements (Vagabov et al.2000).
Polyphosphate (PolyP) biosynthesis was studied using the cells with low PolyP levels grown in a phosphate-free medium as described in (Vagabov et al.2000). Then the cells were cultivated in complete medium for 0.5 h. Biomass samples were harvested at 3000 g for 10 min, washed twice with distilled water at 4°C, and used for PolyP extraction. Dry cell mass was determined after drying cell aliquots at 85°C under vacuum.
Purification of cellobiose lipids
The cellobiose lipid of the yeast Cryptococcus humicola 9-6 (All-Russian Collection of Microorganisms, VKM) was obtained as described (Kulakovskaya et al.2009). After the cultivation, the culture supernatant was separated by centrifugation at 5000 g for 40 min, filtered through a Whatman glass fiber filter GF/A from Sigma-Aldrich Rus (Moscow, Russia), and lyophilized. The residue was extracted with methanol for four to five days at 5°C and filtered. The filtrate was evaporated at 50°C, and the resulting product was suspended in deionized water. The suspension was kept for 24 h at 5°C, and the resulting precipitate was separated by filtration through a glass filter, washed twice with cooled deionized water, and dissolved in methanol. The concentration of glycolipids was determined by weighing after methanol evaporation. In the course of purification, the antifungal activity was assayed by placing the aliquots of preparations on glucose-peptone agar (GPA) containing 0.5% glucose, 0.2% yeast extract, 0.25% peptone, 2% agar, 0.04 M citrate-phosphate buffer, pH 4.0 and inoculated with S. serevisiae.
ESI-MS analysis
The ESI-MS spectra were recorded with a Finnigan MAT LCQ (San Jose, CA, USA) mass spectrometer as described earlier for positive ions (Kulakovskaya et al.20052009. For direct (syringe) inlet, the methanol solution of a sample was injected at 10 µL/min. MS spectra were measured in positive mode.
Inorganic polyphosphate (PolyP) assay
Five separate polyP fractions differ in the chain length were obtained from Saccharomyces cerevisiae cells as described in (Vagabov et al.2000). Acid-soluble polyphosphates (PolyP1) were extracted with 0.5N HClO4. Salt-soluble polyphosphates (PolyP2) were extracted with saturated NaClO4 solution. Two fractions of alkali-soluble polyphosphates (PolyP3 and PolyP4) were extracted with the weak NaOH solution (pH was adjusted to 9–10) and 0.05 M NaOH (pH 12), respectively. All extractions were performed twice at 0°C under stirring for 10 min. The PolyP contents in the fractions PolyP1, PolyP2, PolyP3 and PolyP4 were quantified as a difference in Pi amounts before and after hydrolysis of the samples in 1N HCl for 10 min at 100°C. The level of PolyP5 fraction was determined by treating residual material with 0.5N HClO4 at 90°C twice for 20 min and assaying the released Pi. Pi was determined according to (Vagabov et al.2000). The data in the tables and figures are the average values of three experiments.
The chain length of PolyP from different fractions was determined by electrophoresis in polyacrylamide gel according to (Kumble and Kornberg 1995). For PolyP5 electrophoresis, residual biomass was extracted with distilled water for 12 h (Vagabov et al.2008). PolyP standards with the average chain lengths of 15, 25, 45, 75 phosphate residues were from Sigma (St Louis, USA), and with the average chain lengths of 208 phosphate residues were from Monsanto (St Louis, USA).
ATP assay
The ATP content in the cells was assayed after treating biomass samples with dimethylsulfoxide (0.2 ml / 25–50 mg of wet biomass). The effects of cellobiose lipids on ATP leakage from the cells was assayed as described (Kulakovskaya et al.2003). ATP was assayed by the luciferin-luciferase method using a Sigma assay kit and a LKB 1250 Luminometer (Sweden).
Potassium ion leakage
The leakage of K+ from the yeast cells was registered with a K+-selective electrode (Orion, USA). The measurements were made in a thermostatically controlled 2.5 ml cell at 25°C under stirring. The measuring medium containing 0.01 M citrate-phosphate buffer, pH 4.0, was injected with 50 µl of cell suspension to a final cell concentration of 6–6.5 · 108 ml-1. The maximum quantity of K+ found in the medium was taken as 100%.
Measurement of EKP (electrokinetic potential)
The cells were suspended in 0.01 M citrate buffer, pH 4.0, to a concentration of 107 – 5 × 107 cells ml-1. The EKP of yeast cells was measured with a Zetasizez nano ZS (Malvern, Great Britain) by the method of laser Doppler spectroscopy at 25°C. The average EKP value was calculated from three repeated measurements in each population of yeast cells.
The assay of cell viability
Yeast cell viability assay was performed as follows. The starting cell suspension was diluted in distilled water (1:100). Then the cells were treated with different cellobiose lipid concentrations (0.025 to 0.8 mg ml-1). The incubation mixture contained 0.5 ml of 0.04 M citrate-phosphate buffer, pH 4.0, and 0.1 ml of cell suspension. The mixture without cellobiose lipid was used as a control (100% viability). After the treatment, the cells were incubated at 30°C for 1 h. Then the cell suspensions were diluted in the citrate-phosphate buffer to different ratios and deposited on Petri dishes. The Petri dishes were incubated at 28°C for 3 days and the number of colonies was calculated.
All experiments were performed in triplicate. The biochemicals except those which are listed separately were obtained from Sigma-Aldrich-Rus (Moscow, Russia).
The work was supported by the Russian Foundation for Basic Research (grants 11-04-01009 and 12-04-32138). We are thankful to Dr. W.I. Golubev for providing of the strain 9-6 of Cr. humicola and to E. Makeeva for the help in preparation of the manuscript.
Authors’ Affiliations
(1)
Skryabin Institute of Biochemistry and Physiology of Microorganisms, Russian Academy of Sciences
(2)
Institute of Cell Biophysics, Russian Academy of Sciences
References
Achbergerová L, Nahálka J: Polyphosphate - an ancient energy source and active metabolic regulator.Microb Cell Fact 2011, 10: 63-70. 10.1186/1475-2859-10-63View ArticleGoogle Scholar
Beaven JM, Charpentier C, Rose AH: Production and tolerance of ethanol in relation to phospholipid fatty-acyl composition in Saccharomyces cerevisiae.J Gen Microbiol 1982, 128: 1447-155.Google Scholar
Bölker M, Basse CW, Schirawski J: Ustilago maydis secondary metabolism-from genomics to biochemistry.Fungal Genet Biol 2008, 45: 588-593.View ArticleGoogle Scholar
Cameotra SS, Makkar RS: Recent applications of biosurfactants as biological and immunological molecules.Cur Opin Microbiol 2004, 7: 262-266. 10.1016/j.mib.2004.04.006View ArticleGoogle Scholar
Golubev WI: Antagonistic interactions among yeast. In Biodiversity and Ecophysiology of the Yeasts. Edited by: Rosa CA, Peter G. Berlin: Berlin Verlag; 2006:197-219.View ArticleGoogle Scholar
Hammami W, Chain F, Michaud D, Bélanger RR: Proteomic analysis of the metabolic adaptation of the biocontrol agent Pseudozyma flocculosa leading to glycolipid production.Proteome Sci 2010, 8: 7. 10.1186/1477-5956-8-7View ArticleGoogle Scholar
Hammami W, Castro CQ, Rémus-Borel W, Labbé C, Bélanger RR: Ecological Basis of the Interaction between Pseudozyma flocculosa and Powdery Mildew Fungi.Appl Environ Microbiol 2011, 77: 926-933. 10.1128/AEM.01255-10View ArticleGoogle Scholar
Herve A, Rousseaux I, Charpentier C: Relationship between ethanol tolerance, lipid composition and plasma membrane fluidity in Saccharomyces cerevisiae and Kloeckera apiculata.FEMS Microbiol Lett 1994, 124: 17-22. 10.1111/j.1574-6968.1994.tb07255.xView ArticleGoogle Scholar
Kitamoto D, Isoda Y, Nakahara T: Functions and potential applications of glycolipid biosurfactants - from energy-saving material to gene delivery carriers.J Biosci Bioeng 2002, 94: 187-191.View ArticleGoogle Scholar
Kubota S, Takeo I, Kume K, Kanai M, Shitamukai A, Mizunuma M, Miyakawa T, Shimoi H, Iefuji H, Hirata D: Effect of ethanol on cell growth of budding yeast: genes that are important for cell growth in the presence of ethanol.Biosci Biotechnol Biochem 2004, 68: 968-972. 10.1271/bbb.68.968View ArticleGoogle Scholar
Kulaev IS, Vagabov VM, Kulakovskaya TV: The Biochemistry of Inorganic Polyphosphates. Chichester: John Wiley & Sons Ltd.; 2004.View ArticleGoogle Scholar
Kulakovskaya TV, Kulakovskaya EV, Golubev WI: ATP leakage from yeast cells treated by extracellular glycolipids of Pseudozyma fusiformata.FEMS Yeast Res 2003, 3: 401-404. 10.1016/S1567-1356(02)00202-7View ArticleGoogle Scholar
Kulakovskaya EV, Ivanov AY, Kulakovskaya TV, Vagabov VM, Kulaev IS: Effects of cellobiose lipid B on Saccharomyces cerevisiae cells: K + leakage and inhibition of polyphosphate accumulation.Microbiology 2008, 77: 288-292. 10.1134/S0026261708030065View ArticleGoogle Scholar
Kulakovskaya T, Shashkov A, Kulakovskaya E, Golubev W, Zinin A, Tsvetkov Y, Grachev A, Nifantiev N: Extracellular cellobiose lipid from yeast and their analogues: structures and fungicidal activities.J Oleo Sci 2009, 58: 133-140. 10.5650/jos.58.133View ArticleGoogle Scholar
Kumble KD, Kornberg A: Inorganic polyphosphate in mammalian cells and tissues.J Biol Chem 1995, 270: 5818-5822. 10.1074/jbc.270.11.5818View ArticleGoogle Scholar
Mimee B, Labbe C, Pelletier R, Belanger RR: Antifungal activity of flocculosin, a novel glycolipid isolated from Pseudozyma flocculosa.Antimicrob Agents Chemother 2005, 49: 1597-1599. 10.1128/AAC.49.4.1597-1599.2005View ArticleGoogle Scholar
Mishra P, Prasad R: Relationship between ethanol tolerance and fatty-acyl composition of Saccharomyces cerevisiae.Appl Microbiol Biotechnol 1989, 30: 294-308.View ArticleGoogle Scholar
Morita T, Ishibashi Y, Fukuoka T, Imura T, Sakai H, Abe M, Kitamoto D: Production of glycolipid biosurfactants, cellobiose lipids, by Cryptococcus humicola JCM 1461 and their interfacial properties.Biosci Biotechnol Biochem 2011, 75: 1597-1599. 10.1271/bbb.110036View ArticleGoogle Scholar
Puchkov EO, Zahringer U, Lindner B, Kulakovskaya TV, Seydel U, Wiese A: Mycocidal, membrane-active complex of Cryptococcus humicola is a new type of cellobiose lipid with detergent features.Biochim Biophys Acta (Biomembranes) 2002, 1558: 161-170. 10.1016/S0005-2736(01)00428-XView ArticleGoogle Scholar
Shirai A, Sumimoto T, Kurimoto M, Maseda H, Kourai H: The mode of the antifungal activity of gemini-pyridinium salt against yeast.Biocontrol Sci 2009, 14: 13-20. 10.4265/bio.14.13View ArticleGoogle Scholar
Susan TD, Hossack JA, Rose AH: Plasma-membrane lipid composition and ethanol tolerance in Saccharomyces cerevisiae.Arch Microbiol 1978, 117: 239-245. 10.1007/BF00738541View ArticleGoogle Scholar
Teichmann B, Linne U, Hewald S, Marahiel MA, Bölker M: A biosynthetic gene cluster for a secreted cellobiose lipid with antifungal activity from Ustilago maydis.Mol Microbiol 2007, 66: 525-533. 10.1111/j.1365-2958.2007.05941.xView ArticleGoogle Scholar
Vagabov VM, Trilisenko LV, Kulaev IS: Dependence of inorganic polyphosphate chain length on the orthophosphate content in the culture medium of the yeast Saccharomyces cerevisiae.Biochemistry-Moscow+ 2000, 65: 349-355.Google Scholar
Vagabov VM, Trilisenko LV, Kulakovskaya TV, Kulaev IS: Effect of carbon source on polyphosphate accumulation in Saccharomyces cerevisiae.FEMS Yeast Res 2008, 8: 877-882. 10.1111/j.1567-1364.2008.00420.xView ArticleGoogle Scholar
Walker-Caprioglio HM, Casey WM, Parks LW: Saccharomyces cerevisiae Membrane sterol modifications in response to growth in the presence of ethanol.Appl Environ Microbiol 1990, 56: 2853-2857.Google Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
