Research ArticleGlycosylation of Aromatic Amines I: Characterization of Reaction Productsand Kinetic SchemeMadhushree Y. Gokhale,1,3,4William R. Kearney,2and Lee E. Kirsch1Received 21 August 2008; accepted 19 February 2009; published online 21 March 2009Abstract. The reactions of aliphatic and aromatic amines with reducing sugars are important in both drugstability and synthesis. The formation of glycosylamines in solution, the rst step in the Maillard reaction,does not typically cause browning but results in decreased potency and is hence signicant from theaspect of drug instability. The purpose of this research was to present (1) unreported ionic equilibria ofmodel reactant (kynurenine), (2) the analytical methods used to characterize and measure reactionproducts, (3) the kinetic scheme used to measure reaction rates and (4) relevant properties of variousreducing sugars that impact the reaction rate in solution. The methods used to identify the reversibleformation of two products from the reaction of kynurenine and monosaccharides included LC massspectrometry, UV spectroscopy, and 1 D and 2 D 1H 1H COSY NMR spectroscopy. Kinetics was studiedusing a stability indicating HPLC method. The results indicated the formation of and glycosylaminesby a pseudo rst order reversible reaction scheme in the pH range of 1 6. The forward reaction was afunction of initial glucose concentration but not the reverse reaction. It was concluded that the reactionkinetics and equilibrium concentrations of the glycosylamines were pH dependent and also a function ofthe acyclic content of the reacting glucose isomer.KEY WORDS: glucose; glycosylamines; glycosylation; kinetics; Maillard reaction; reducing sugararomatic amine.INTRODUCTIONThe reactions of aliphatic and aromatic amines withaldehydes to form imines are important in both drug stabilityand synthesis. The reactions of glycosidic aldehydes withamines involve the added complexity of monosaccharideequilibria in solution whereby the reversible formation ofan imine (acyclic sugar moiety) is in equilibrium with theglycosylamine isomers (the reduced cyclic sugars) (1,2).Amadori products may subsequently form and give rise toadditional degradation products. The complex so calledMaillard reaction pathways result in potency loss andproduct discoloration. All reducing sugars have beenimplicated in these browning reactions including glucose,lactose, galactose, mannose, and dextrates (3). Glucose iscommonly used as a solute in solutions for parenteraladministration; lactose is used as an excipient in bothparenteral and enteric formulations.Examples of drug instability due to the Maillardreaction include both primary and secondary amines as inthe case of amphetamines, insulin, uoxetine, sulfa drugs,procainamide, metoclopropamide, and daptomycin to namea few (4 9). For example, Duval et al. studied the browningof dextroamphetamine sulfate solutions containing lactoseand found that the solution darkens on storage at 50C withthe formation of brownish precipitate (4). Using infrareddata and thin layer chromatography, they concluded that theproduct was a Schiff base formed from the reaction of theprimary amine in dextroamphetamine and the carbonylgroup of lactose. Secondary amines in generic formulationsof uoxetine HCl containing lactose as a common ingredientwere shown to be less stable than the formulations containing starch due to the Maillard reaction between the drug andlactose (5). Dextrates obtained by the controlled hydrolysisof starch used as excipients are also known to exhibit thebrowning reaction due to the presence of varying amountsof dextrose (4,10,11).The majority of previous studies primarily focused onreactions of glucose with model aliphatic amines such asamino acids (12 16) and the browning process due to theformation of Amadori products and advanced Maillardreaction products (15,17 21). However, the formation ofglycosylamines in solution, the rst step in the Maillardreaction, does not typically cause browning but results indecreased potency and is hence signicant from the aspect ofdrug instability. The glycosylamines of cyclic sugars can existin equilibria in solution predominantly as the and forms317 1530-9932/09/0200-0317/0 # 2009 American Association of Pharmaceutical ScientistsAAPS PharmSciTech, Vol. 10, No. 2, June 2009 (#2009)DOI: 10.1208/s12249 009 9209 21Division of Pharmaceutics, College of Pharmacy, The University ofIowa, Iowa City, Iowa 52242, USA.2The Nuclear Magnetic Resonance Facility, College of Medicine, TheUniversity of Iowa, Iowa City, Iowa 52242, USA.3Bristol Myers Squibb Company, 1 Squibb Drive, New Brunswick,New Jersey 08903, USA.4To whom correspondence should be addressed.(e mail: madhush reeg@gmail.com)similar to the parent sugar though the equilibrium compositions may differ. The , glycosylamines mutarotate andhydrolyze through the acyclic imine (22,23).In the area of drug stability, only a few cases have beenreported for reactions of weakly basic aromatic aminecontaining drugs in the presence of reducing sugars likeglucose forming glycosylamines. For example, the reversibleformation of glycosylamines for procainamide in the presenceof 5% dextrose solution was reported in water (24,25). Inanother study, a 20 30% procainamide potency loss (aromatic amine pKa=2.75) in 24 h was reported following admixingwith glucose solutions (26). The formation of glucosylamineshas also been reported for the oral suspension of sulfamethoxazole (aromatic amine pKa=1.69) in the presence ofglucose and a limit for the amount of glycosylaminesallowed in the suspension has been reported for suchformulations (27). In another study involving dissolutiontesting of Bactrim DS tablets (sulphamethoxazole, 800 mgand trimethoprim, 160 mg) in the presence of 5% glucose,glucosylamines readily formed with the sulphamethoxazoleand suggested the potential of reduced bioavailability ofsulphamethoxazole in the presence of dietary glucose (7).Lucida et al. studied the reaction of glucose and sulphamethoxazole and showed a pH and temperature dependentrate of reversible formation of glycosylamines in the acidicto neutral pH range of 1 6 (6). Thus, the kinetics offormation of glycosylamines with respect to solution conditions such as pH, buffers and temperature, effect ofaldehydic content of sugar, and an understanding of themechanism of their formation is a key factor in predictingreactivity and degradation kinetics of these sugar aminereactions. A systematic study involving formation andcharacterization of the glycosylamines and a detailed kineticand mechanistic understanding of their formation for weaklybasic aromatic amines and sugar carbonyls has not beenreported.Daptomycin, a cyclic lipopeptide antibiotic, has beenreported to react with 5% dextrose and other reducing sugarsto form reversible products that decrease the potency of thedrug in acidic solutions (28). The peptide portion ofdaptomycin consists of 13 amino acid residues connected atthe N terminal tryptophan to a decanoyl aliphatic group. Ithas six ionizable groups: four carboxylic acid side chains andtwo primary amines from the side chains of ornithine and aweakly basic amine kynurenine. The pKa of kynurenine indaptomycin was determined to be 0.8 using UV spectroscopy(29). Reactions of daptomycin with glyceraldehyde resultingin the formation of Schiffs base are also reported (9).Reaction products formed with dextrose were proposed tobe the glycosylamines formed at the kynurenine aromaticamine by reaction with the aldehydic glucose.This paper represents the rst in a series of papersdescribing the kinetics and mechanisms of the reactions ofreducing sugars and weakly basic amines in aqueous solution,solid state, and pharmaceutical formulations. The mainobjectives of this rst paper is to present (1) unreported ionicequilibria of model reactant (kynurenine), (2) the analyticalmethods used to characterize and measure reaction products,(3) the kinetic scheme used to measure reaction rates, and (4)relevant properties of various reducing sugars that impact thereaction rate in solution.Kynurenine is a weakly basic aromatic amino acid withthree ionizable groups viz. an aromatic amine, an alphacarboxylic acid, and an alpha amine (Fig. 1) with a molecularweight of 208.1. Glucose is one of the most common reducingsugars (MW=180) that exists in solution as cyclic sixmembered rings as or glucose which equilibrates throughan acyclic aldehyde by mutarotation. The and glucosediffer in the position of the anomeric proton (proton at the C1position in the ring) being in the equatorial or the axialpositions, respectively. The acyclic form of the sugar is thereactive aldehyde in carbonyl amine reactions. All reducingsugars can undergo reactions with amines owing to thealdehydic form. Thus, isomers of glucose like galactose,mannose, gulose, and allose with varying proportion ofthe acyclic form can all undergo reactions with amines (30).Although the six membered ring forms are the more stablethermodynamically, glucose also exists as ve memberedring forms in very low concentrations (30). The reactionbetween kynurenine and glucose (Fig. 2) gives rise tothe imine that exists in equilibrium with the cyclic , glycosylamines.Reactions were initiated with kynurenine and thesemonosaccharides to isolate the reaction products and identifythem. A reaction scheme was proposed based on observedkinetics for these sugar amine reactions. The aromatic aminepKa for kynurenine was determined using UV spectroscopy;the amine acid pKa was determined using potentiometrictitration and the carboxylic acid pKa was determined using2D HMQC nuclear magnetic resonance (NMR) spectroscopy.Reaction products were generated by heat stressing mixturesof kynurenine in the presence of excess sugars under acidicconditions at various pH values. Separation of the reactionproducts was achieved by reversed phase high performanceliquid chromatography (RP HPLC) and product characterization was done by liquid chromatography mass spectrometry (LC MS). For reactions of kynurenine with glucose,additional characterization studies were conducted by collecting fractions of products which were lyophilized and evaluated by UV spectroscopy, 1 D 1H NMR, 2 D 1H 1H COSY, anddecoupling NMR experiments.After product identication, a reaction scheme wasproposed and conrmed by determining the effect of varyingconcentrations of glucose on the apparent pseudo rst orderloss of kynurenine. The effect of pH on the extent and halflife of the reaction was determined in acidic aqueoussolutions (pH values between 1 and 6). The effect of varyingacyclic content of various sugars on reactivity was alsostudied.COCH2NH3+CHNH3+COOHFig. 1. Structure of kynurenine318 Gokhale, Kearney, and KirschMETHODS AND MATERIALSKynurenine was obtained from ICN Biomedicals. Anhydrous glucose was obtained from Fisher (Springeld, NJ,USA) as was hydrochloric acid (1 N), sodium hydroxide(0.1 N), and sodium chloride. D allose was obtained fromTCI America; D gulose was obtained from O Micron Biomedicals, Inc.; and D Mannose and D galactose wereobtained from Sigma Aldrich. All chemicals were reagentgrade. Solvents used for chromatography were HPLCgrade.pKa DeterminationsThe pKa value of the aromatic amine was determinedusing a UV spectrophotometric method (Hewlett Packard HP8453) at 401C (31). A 5.07 mM of L kynurenine stocksolution was used to prepare a series of 0.507 mM solutions ofkynurenine in the pH range of 0.46 3.58 with buffers(hydrochloric acid, acetate buffer, and phosphate buffer)and sodium chloride (to adjust to a constant ionic strengthof =0.5). The absorbance at 281 nm was the wavelength ofmaximum difference between the ionized and unionizedspecies and was used to estimate the apparent pKa of thearomatic amine by non linear regression (JMP) using anequation relating the absorbance to the fraction and absorptivity of each species (31).The pKa value for the alpha amine group on the aminoacid side chain was determined using potentiometric titration(DL25, Mettler Inc.). All measurements were carried out at401C for a 0.01 M L kynurenine solution. The pKa wasdetermined using the Gran plot method.The alpha carboxylic acid pKa was determined usingNMR spectroscopy (heteronuclear multiple quantum correlation spectroscopy) by following changes in chemical shifts ofthe alpha carbon proton and beta carbon protons on the sidechain of kynurenine as a function of pH. The HMQCexperiment monitors protons attached to only the carbonskeleton of the molecule, specically by one bond in theversion used.Kynurenine stock solution (0.020 M) was prepared indeuterated water. The pH was measured at 40C. The pHelectrode was equilibrated and calibrated using pH 7 andpH 4 standard buffer solutions also equilibrated at 401C.Kynurenine solution was pipetted into a water jacketedvessel, equilibrated to 401C and titrated using standardized1.0 N HCl added in aliquots of 0.010 mL using an automatictitrator (DL25, Mettler Inc.). One millimolar sodium 2, 2dimethyl 2 silapentane 5 sulfonate (DSS) was also added tothe solution for referencing the NMR spectra. During thetitration, 0.65 mL of kynurenine solution was removed at theOHHHOHHOOOHHHOHOHHHOHHOHOHHOHCOCH2NH2CHNH3+COO-COCH2NCHNH3+COO-imineD-glucose kynurenine+OHHOHHOHHOHH HOHOHHOHHOHHOHH NHOHCOH2C CHNH3+COO-OHHOHHOHOHH HOHCOH2C CHNH3+COO-NH-glycosylamine -glycosylamine Fig. 2. Reaction of kynurenine and glucose forming imine and and glycosylamines319 Glycosylation of Aromatic Amines I: Characterization of Reaction Products and Kinetic Schemedesired pH and subjected to a 1D proton and a 2D HMQCNMR experiment using a 500 MHz NMR, Varian Inc.,spectrometer at 401C. The parameters for the 1 Dexperiment were: averaged number of transients (nt)=8,spectral width (sw)=6,000 Hz, number of points (np)=6,000points processed with 0.2 Hz line broadening (lb) and zerolled to 128 k points, and a recycle delay of 5 s (d1) in D2O.For the 2 D HMQC experiment the parameters were asfollows: sw1=30,000, ni =128 s phase 1, 30,000/125.659=12 ppm in proton. The spectra for the HMQC experimentswere directly referenced to DSS at high pH and indirectlyreferenced at low pH.After obtaining the NMR spectra, the sample wasreintroduced into the titration vessel and the titration wascontinued until the next desired pH value was obtained. Theproton on the chiral carbon and the protons on the carbon in the side chain were monitored for changes inchemical shifts as a function of the deprotonation of the alphacarboxylic acid group. Plots of chemical shifts versus pD foreach of the two groups of protons were sigmoidal in nature.pKa was estimated using non linear regression (JMP V5.0.1,SAS Institute) using the following equation: (32,33)

obs acid10pH10pH10pKa base10pKa10pH10pKa 1where obs is the observed chemical shift, acid is the chemicalshift of the most acidic species (pH* 1.02) and base is thechemical shift of the most basic species (pH* 3.75) (pH*denotes the uncorrected pH in deuterium oxide).HPLC AnalysisHPLC analyses were performed using a Shimadzu RPHPLC system consisting of an SCL 10AVP system controller,LC 10ATVP pumps, SIL 10ADVP auto injector, SPD 10AVPUVVIS detector, CTO 10ASVP column oven, and a FRC 10Afraction collector. Chromatograms were integrated and datastored using Class VP Chromatography Data System software(Version 4.2). The column used was a Phenomenex HydroRP 18,4.6250 mm, 4 column.For reactions of kynurenine and glucose, an isocraticmethod with a mobile phase composition of 1% methanoland 99% water, a ow rate of 1 mL/min, injection volume of10 lt, run time of 35 min, sample temperature of 4C, columntemperature of 25C, and detection wavelength of 257 nmwas used. The HPLC method was validated using standardcompendial methods. Calibration curves for kynurenine inthe range of 0.2 mM to 1 mM were used for calculatingconcentrations. The precision of the analytical method wasdetermined by multiple analyses of 0.50 mM solutions ofkynurenine. The coefcient of variation was determined to be3.6%. The accuracy of the method was determined byestimating the percent recovery for 44 kynurenine testsamples of known concentrations in the range 0.2 to1.0 mM. The mean, range, and CV for the test set were98.9%, 91 110%, and 4.8%, respectively. This method wasused for identication of products for kynurenine glucosemixtures using LC MS and all further kinetic analyses.For separation and LC MS analysis of reaction productsof kynurenine with galactose, mannose, gulose, and allose thecolumn temperature and run time for the HPLC method wasfurther modied to range from 25 40C and 35 to 50 min,respectivelyReaction Product Formation and IsolationFor reactions of kynurenine with glucose, a reactionmixture of 1.2 mM kynurenine and 0.5 M glucose in hydrochloric acid (pH 3.45) was prepared and stored at 401C in aTeon coated rubber stoppered glass vial. Aliquots of reaction mixture were removed and diluted in acetate buffer(0.5 M, pH 5.8) to quench the reaction. Reactions of 1.2 mMkynurenine with 0.5 M glucose solution were conducted in5.0010 4N HCl at 40C. A semi preparative Shimadzu RPHPLC system consisting of a SCL 10A system controller, twoLC 10ATVP pumps, and SPD 10A UV VIS detector wasused. Separation of reactants and products were obtainedusing a Phenomenex Synergi semi prep C18 column, 25021.2 mm using a 1% methanol water mobile phase, at a owrate of 7 mL/min, detection wavelength of 257 nm and a runtime of 80 min. Two products (I and II) were formed.Products were collected over dry ice, transferred to a benchlyophilizer (Virtis lyocentre) and lyophilized under vacuumuntil the product appeared to be completely dry upon whichthey were removed from the lyophilizer, sealed in plastictubes and stored at 20C until use. The integrity of thelyophilized products was reconrmed by HPLC and theproducts were used for identication by UV and NMRspectroscopy.Reaction Product Identification and CharacterizationReaction products were identied using UV, NMR, andmass spectrometry. Reaction mixtures of 1.2 mM kynurenineand 0.5 M monosaccharides (glucose, mannose, galactose,gulose, and allose) were subjected to a quadrupole ion traptype mass spectrometer (Thermo Finnigan, San Jose, CA,USA), with a LCQ Deca Ion trap MS, a Surveyor LC, andPDA detector. A positive electrospray type ESI ionizationmethod was used. The sheath gas and auxiliary gas ow rateswere 40 and 10, respectively. Nitrogen was used as the sheathand the auxiliary gas. The capillary temperature was 350Cand the capillary voltage was 3 V.The isolated, lyophilized products from reactions ofkynurenine with glucose were dissolved in water and scannedby UV spectroscopy from 190 450 nm using a UV spectrophotometer (Hewlett Packard HP 8453 Diode Array) todetect any observable spectral changes between the reactant(kynurenine) and the products.NMR spectroscopy was conducted on kynurenine andthe lyophilized products of kynurenine glucose reactions todetermine their structure. Kynurenine and each isolatedlyophilized product (I and II) were dissolved in 0.8 mL ofdeuterated solvent dimethylsulfoxide (DMSO). A 0.65 mLaliquot of this solution was introduced into the NMR tubeand subjected to 1 D 1H NMR experiments using a 500 MHzNMR (Varian, Inc.) spectrometer. Kynurenine and product Iwas also subjected to 2 D 1H 1H COSY NMR experiments,but not product II due to its low concentration. Decoupling320 Gokhale, Kearney, and Kirschexperiments were also performed on products I and II settingthe decoupler frequency at the frequency of the aromaticamine proton of kynurenine. This simplied the splittingpattern of the protons on the anomeric carbon for products Iand II. The coupling constants for the simplied decoupleddoublets were calculated for the two products and comparedto literature values for the / form of glucose to conrm theidentity of the / isomers. The NMR parameters for the 1 D1H NMR experiments for determination of kynureninestructure were as follows; np=60,000, number of transients=4, spectral width=6,000, d1=5 s, temperature=25.4, and lb=0.2. Similarly parameters for 2 D 1H 1H COSY experimentfor kynurenine were as follows: nt=4, sw=60,000, np=2,048,lb=not used, fn=2,048, temperature=25.4, and d1=4. For the2 D acquisition sw1=6,000 and ni=512 in seconds. Similarly,parameters for 1 D 1H NMR determination for product I were,np=60,000, nt=16, sw=6,000, d1=10 s, temperature=25.4, lb=0.2, and fn=131,072. The parameters for 2 D 1H 1H COSYexperiment were nt=4, sw=6,000, np=4,096 (direct dimension)lb=not used, fn=2,048, temperature=25.4, and d1=7. For the2 D acquisition sw1=6,000 and ni=1,024 in seconds (indirectdimension). For the decoupling experiment for product I thepeak at 9 ppm was decoupled and the following parameterswere used: nt=8, sw=6,000, lb=0.2, fn=131,072, temperature=25C, and d1=10 s. The parameters for 1 D 1H spectraldetermination for product II were nt=256, sw=6,000, d1=10 s,temperature=25.4, lb=0.25, and fn=131,072. For the decouplingexperiment for product II the peak at 9.01 ppm was decoupledand the following parameters were used: nt=256, sw=6,000,lb=0.25, fn=131,072, temperature=25C, and d1=10 s.Kinetics of Kynurenine and Reducing MonosaccharidesReaction mixtures of 1.2 mM kynurenine and 0.50 Mglucose in hydrochloric acid or phosphate buffer at thedesired pH and ionic strength of 0.10 or 0.50 were stored at401C in Teon coated rubber stoppered glass vials. Thereactions were carried out in the pH range of 1 6. Thesolution conditions are given in Table I. Aliquots wereremoved from reaction mixtures at appropriate timesquenched with acetic acid/sodium acetate quench buffer(0.5 M at pH 5.8) and stored at 4C in auto injector andanalyzed by HPLC. Standard calibration curves were used forcalculating the concentration of kynurenine and the reactionproducts in the reaction mixtures. The concentration timeproles for the loss of kynurenine and the appearance ofglycosylamines were constructed and reaction extent andhalf lives were measured as function of pH.Reactions were also conducted with 1 mM kynurenineand various initial concentrations of glucose ranging from0.19 0.70 M in sodium acetate/acetic acid buffer (total bufferconcentration 0.279 M) at pH 3.4 and a temperature of 40C.The reaction conditions are described in Table II.To study the kinetics with other glucose isomers, reactionmixtures of 1.2 mM kynurenine and 0.5 M of monosaccharides(galactose, mannose, allose, and gulose) in hydrochloric acid(pH 2.7) at an ionic strength of 0.1 were stored at 401C inTeon coated rubber stoppered glass vials. Aliquots of thereaction mixture were removed, diluted with acetate quenchbuffer, and HPLC analyses were performed. The extent ofreaction and reaction half life as a function of the acycliccontent were determined.RESULTSpKa Determinations for KynurenineThe apparent Ka value of 0.01445 (pKa=1.84) at 40C forthe aromatic amine was calculated from the UV absorptiondata at the analytical wavelength of 281 nm and the Ka valueof 8.99510 10(pKa=9.05) at 40C of the alpha amino groupwas determined using potentiometric titration.The pKa of the alpha carboxylic acid group wasdetermined using 1 D 1H NMR spectroscopy. The protonon the chiral carbon and the protons on the carbon in theside chain were monitored in the 2 D HMQC experiment forchanges in chemical shifts as a function of the deprotonationof the terminal carboxylic group.The two monitored protons showed a characteristic pHdependent change in chemical shift. The chemical shifts versuspD(pH+0.4) for each of the two groups of protons resulted in asigmoidal curve. The proton on the chiral carbon gave a pKavalue of 2.57 (Fig. 3a), while the protons on the carbon in theside chain gave a pKa value of 2.50 (Fig. 3b). The averagecarboxylic acid pKawas estimated to be 2.53 in deuteriumoxideor 2.11 in water (as described in the DISCUSSION section).Table I. Experimental Conditions to Determine the Scheme for the Reactions of 1.2 mM Kynurenine and 0.5 M Glucose at 40C and IonicStrength of 0.1 0.5 in Hydrochloric Acid or Phosphate BufferBuffer Total concentration M103NaCl (M) Glucose (M) Kynurenine (mM) pHHCl 30.0 0.0902 0.501 1.20 1.66HCl 10.0 0.0891 0.505 1.25 2.11HCl 0.100 0.100 0.514 1.25 4.39NaH2PO4/Na2HPO4 0.121 0.032 0.508 1.19 5.94Table II. Experimental Conditions of Reactions of 1 mM Kynureninewith Varying Concentration of Glucose in the Range of 0.19 0.70 Min Acetic Acid/Sodium Acetate Buffer at pH 3.4 at 40CNo. Glucose (M) Kynurenine (mM) pH1 0.193 1.05 3.492 0.323 1.06 3.503 0.504 1.05 3.404 0.701 1.05 3.49321 Glycosylation of Aromatic Amines I: Characterization of Reaction Products and Kinetic SchemeIsolation and Characterization of Reaction ProductsThe HPLC chromatograms for reactions of kynurenineand glucose showed the presence of three peaks; kynurenine and two reaction products; product I and product IIin the order of their elution at retention times of 17.3 min,23 min, and 27.4 min, respectively (Fig. 4). The twoproducts were formed reversibly and were isolated forfurther characterization.Kynurenine and reaction mixture samples were analyzedusing a LC mass spectrometer. The kynurenine sample showeda peak with m/z ratio of 209.1, which corresponded to themolecular ion of kynurenine and one proton (MH)+. The rstpeak in the reaction mixture chromatogram corresponded tothe kynurenine molecular ion with m/z of 209.1. The tworeaction product peaks that eluted from the reaction mixture ofkynurenine and glucose corresponded to m/z ratio of 371(MH)+.Reactions of kynurenine with galactose, mannose, allose,and gulose also gave two major products in addition to thekynurenine peak. In reactions of galactose, the two peakswere not completely separated and co eluted as a single peak.The LC MS analysis indicated that the products formed withthe other monosaccharides (galactose, mannose, gulose, andallose) also corresponded to a m/z ratio of 371 (MH)+indicating that the same products were formed for reactionsof weakly basic amine with all reducing monosaccharides.The UV scans of the isolated products looked similar tothe UV scan of kynurenine exhibiting the same threewavelength maxima of 228 230 nm, 257 260 nm, and 360362 nm. These results are consistent with the formation of theglycosylamines.The 1 D proton spectrum of kynurenine, showed twodistinct regions of peaks: one at 6 8 ppm with ve peaks thatcorresponded to the aromatic protons and the other region at3 4 ppm with three peaks that corresponded to the side chainprotons. The assignments of the protons for kynurenine wereconrmed by performing a 2 D 1H 1H COSY experiment(Table III). The COSY spectrum also showed two distinctregions: the aromatic protons in the region of 6 8 ppm andthe side chain protons in the region of 3 4 ppm. The 1 Dspectra and the 2 D spectrum for product I both showed tworegions associated with the aromatic protons at 6 9 ppm andthe side chain protons at 3 4 ppm similar to the kynureninespectra. In addition to these protons, the 1 D spectra forproducts I and II showed the sugar protons in the region of 34 ppm overlapping with the side chain protons of the aminoacid. Also, the 1 D spectra showed protons at 4.5 and 5.1 ppmrepresenting the anomeric protons for glucose and protons at9 and 9.1 ppm representing the aromatic amine protons forproducts I and II, respectively. The COSY spectrum forproduct I showed a cross peak between the anomeric proton(4.5 ppm) and the aromatic amine (9 ppm) thereby conrmingthe identity of the reaction product to be a glycosylamine andnot an imine (Fig. 5). The identity of the and glycosylamineswas conrmed based on the coupling constants of 7.98 and 3.674.254.34.354.44.454.54.554.64.651 1.5 2 2.5 3 3.5 4 4.5chemical shift (ppm)pD 3.83.853.93.9544.051 1.5 2 2.5 3 3.chemical shift (ppm)pDb a. .5 4 4.5Fig. 3. Chemical shift in the proton frequency for the a chiral carbon proton and b carbon protons as a function of pD in the pH* range of1.02 3.76. The solid circles are the experimental data and the solid line was estimated based on non linear regression using JMPFig. 4. Representative chromatogram of reaction mixture of kynurenine and glucose peak at 17.3 min is kynurenine, at 23 min is productI and at 27.4 min is product II322 Gokhale, Kearney, and Kirschdetermined for the products I and II, respectively, by decouplingexperiments conducted for products I and II. The assignmentsfor the glycosylamines are listed in Table IV.Determination of Reaction SchemeReactions of kynurenine in the presence of excessglucose were carried out in the pH range of 1 to 6.5. Thekinetic studies were performed under pseudo rst orderconditions by maintaining constant pH, temperature, andionic strength and glucose concentrations in high molarexcess. The chromatograms for the reaction mixtures showedthe same three peaks at all pH values; kynurenine and thetwo products indicating that the reaction proceeded to the and glycosylamines at all pH values. Typical calibrationcurves of peak area versus concentration for kynurenine andthe glycosylamines (concentration range of 0.2 1 mM) werelinear (R2>0.99). The concentration time proles (Fig. 6)indicated that kynurenine reversibly formed the twoglycosylamines and that equilibrium between all threespecies was reached over the duration of the reaction. The and glycosylamines equilibrated rapidly with each otherand existed as a constant ratio of 6:1( glycosylamine glycosylamine) at all pH values. The simplest schemeconsistent with the observed kinetics was:kynurenine !kfobs

krobs glycosylamines 2where kfobs is the rate constant for the forward reaction andkrobs is the rate constant for the backward reaction. TheTable III. Chemical Shift Assignments in ppm for KynurenineProtons for the 2 D 1H 1H COSY Spectrum in Deuterated DMSORegion Position Chemical shift (ppm)Aromatic region H1 6.75H2 7.25H3 6.5H4 7.7NH 7.2Side chain region CH 3.6CH2 3.5, 3.25 aromatic amine protonanomeric protonsugar regionaromatic regionOHHOHHOHHOHHHOHOHHOHHOHHOHHNHOHCOH2C CHNH3+COO- 1 234aromatic amine protonanomeric protonsugar regionaromatic regionOHHOHHOHHOHHHOHOHHOHHOHHOHHNHOHCOH2C CHNH3+COO- 1 234H4H2H1H3H44H2H1H3Fig. 5. COSY spectrum of product I showing the region of 3 4 ppm corresponding to overlapped peaks and cross peaks of the side chain andsugar protons and 6 9 ppm corresponding to peaks and cross peaks for the aromatic protons. (Bottom right) Expanded region showing theregion of 6 7 ppm corresponding to overlapped peaks and cross peaks of the aromatic protons. The peaks at 6.7, 6.9, 7.4, and 7.8 ppmcorrespond to the aromatic ring protons H3, H1, H2, and H4, respectively. (Top left) Expanded region showing the region of 3 3.7 ppm. Twodistinct spin systems corresponding to the aromatic side chain and sugar protons323 Glycosylation of Aromatic Amines I: Characterization of Reaction Products and Kinetic Schemeobserved rate constant (kobs) for the overall loss of kynureninewas determined based on the pseudo rst order reversiblekinetics of the reaction. Table Vlists the reaction half lives andequilibrium concentrations of kynurenine for the reactions ofkynurenine and glucose in the acidic pH range of 1 to 6.Reactions were conducted with 1.2 mM kynurenine withvarious initial concentrations of glucose from 0.19 0.70 M.The data was t using reversible reaction kinetics and theforward and reverse rate constants were determined using theEq. 2 (Table VI). Typical CV values for rate constantestimates were 10%.Reactions of kynurenine with other reducing sugars viz.galactose, mannose, gulose, and allose also resulted in theformation of the glycosylamines. The reaction kinetics forthese reducing monosaccharides was also described by thereversible reaction scheme shown in Eq. 2. The total acycliccontent for the various isomers of glucose were obtained fromliterature, based on measurements in aqueous solutions indeuterium oxide at 30C obtained by 13C NMR for thesesugars (34). The acyclic content was reported to be lowest forglucose and allose at 0.009% while gulose had the highest at0.082%. Equilibrium concentrations of kynurenine forreactions with allose, gulose, galactose, and mannoseindicated that the equilibrium concentration was constantthrough all the sugars averaging at 64.3% (3.4%) of theinitial amine concentration. Reaction half lives indicated thelongest half life for glucose and allose at 0.6 h and 0.5 h,respectively, and the shortest for gulose at 0.06 h.DISCUSSIONDetermination of pKa ValuesKynurenine is an aromatic amino acid with threeionizable groups; the aromatic amine, the alpha amine, andthe alpha carboxylic acid. The aromatic amine pKa wasestimated to be 1.84 at 40C using UV spectroscopy.Kynurenine is a weakly basic aromatic amine similar inbasicity with primary aromatic amine containing drugs suchas sulfamethoxazole (pKa=1.69) (6), procainamide (pKa=2.75) (26), benzocaine (pKa=2.49), or procaine (pKa=2.28)(35) to name a few (31). The alpha amine pKa value wasdetermined to be 9.05 at 40C by potentiometric titration.Protonation of the reactive amine dramatically reduces itsnucleophilicity thereby substantially minimizing its reactivity.Hence, the alpha amine is largely unreactive in the acidic pHrange of 1 6 and the reactions with glucose were limited tothose involving the weakly basic aromatic amine. The alphacarboxylic acid pKa for amino acids typically lies in the rangeof 2 3. This is very close to the aromatic amine pKa andhence to avoid interference by the aromatic amine pKa, anHMQC type NMR experiment was used to determine thepKa for the carboxylic acid. In the HMQC experiment,protons attached to the carbon skeleton of the molecule weremonitored. Since the carboxylic acid proton is exchangeable,the protons on the carbon atoms ( and carbons of the sidechain) neighboring to the carboxylic acid group weremonitored as a function of pH. The protonation or deprotonation of the carboxylic acid changed the shielding of theneighboring protons and caused a change in chemical shift ofthose protons as a function of pH. This chemical shift wasplotted as a function of pD to yield a sigmoidal curve that wasused to estimate the pKa value. The standard correction forestimation in deuterium was applied to the observed meterreading (36,37). pKa value of 2.57 for the carbon protonand 2.50 for the carbon protons gave an average pKa valueof 2.53 for the carboxylic acid pKa at 40C in D2O. However,the values for ionization constants of carboxylic acids indeuterium oxide and water differ by a factor of 0.5 0.6 dueprimary isotope effects (38,39). The difference in the ionization constant (pKa) of carboxylic acids of amino acids inTable IV. Chemical Shift Assignments for the Glycosylamines Basedon 1 D 1H NMR, 2 D NMR in Deuterated DMSORegion Position Chemical shiftAromatic region form NH 9.1 form NH 9.0H1 6.9H2 7.4H3 6.7H4 7.8Side chain region CH 3.6CH2 3.5, 3.25Sugar region form H1 5.1 form H1 4.5H2 3.1H3 3.16H4 3.3H5 3.3H6 3.65H6 3.4400.511.50 8 10Conc (mM)time (hrs)2 4 6Fig. 6. Typical concentration area time proles of reaction of 1.2 mMkynurenine (solid circle) and 0.5 M glucose in HCl/NaCl at 0.1 ionicstrength and 40C. Appearance of glycosylamine (solid square), glycosylamine (solid triangle), and mass balance (multiplicationsymbol) are depicted324 Gokhale, Kearney, and Kirschwater and deuterium oxide can be described by a linearrelation of the form (y=mx+h) as:pKD pKH $pK a b pKH 3where a=0.332 and b=0.040 (38). Substituting the pKa valuein D2O (pKD) of 2.53 for kynurenine in the Eq. 3, the nalpKa value for kynurenine in water (pKH) was 2.11. This isconsistent with the typical pKa values in the range of 2 3reported for the alpha carboxylic acids of amino acids (31).Identification of Reaction ProductsIn the pH range of 1 6, the aromatic amine of kynurenineis largely unprotonated and therefore present in a reactiveform, whereas the aliphatic alpha amine is largely protonatedand unreactive. The reaction of kynurenine and glucose gaverise to two products I and II. LC MS, UV spectroscopy, andNMR were used to determine the structure of the twoproducts. Both the products, I and II, showed a m/z of 371 oridentical molecular mass of 370 indicating that the productswere either the glycosylamines or the imine. The m/z of 393 inthe mass spectra for the two products corresponds to molecularion plus sodium. Though the expected products were theglycosylamines, as has been reported for reactions of amineswith sugars (6,8,26) this was further conrmed using UV andNMR spectroscopy. The two products showed similar UVspectral properties to that of kynurenine. The presence of anadditional double bond in an imine would result in a shift of theUV wavelength maxima to a longer wavelength. Since no suchshifts were observed, the two products were most likely the twoglycosylamines. This supposition was further conrmed usingNMR spectroscopy. 1 D 1H NMR, 2 D 1H 1H COSYexperiments and decoupling experiments were conducted toconrm the identity of the products.The chemical shift assignments for the protons on theamino acid kynurenine are listed in Table III. The proton at3.6 ppm was typical of the chiral carbon proton and showeda doublet of doublets due to the two nonequivalent protonson the side chain that appeared at a chemical shift of 3.5 and3.2 ppm. The proton at 7.2 ppm showed a typical broadsinglet characteristic of an amine proton. A triplet at 6.5 ppm,a doublet at 6.75 ppm, another triplet at 7.3 ppm, and adoublet at 7.7 ppm each corresponded to one proton onthe aromatic ring of kynurenine. In the COSY spectrum, thepeaks across the diagonal represent the 1 D NMR peaks. Thecross peaks perpendicular to the peaks along the diagonalrepresented connectivity between corresponding peaks alongthe diagonal. Protons up to two to three bond distances awaytypically show cross peaks. Occasionally longer range couplings also show in a COSY. In the aromatic region of 68 ppm, cross peaks were observed for four peaks along thediagonal except one peak which was the broad amine peak.The assignments of the aromatic region protons in kynurenine were based on the splitting pattern of peaks in the 1 Dspectrum for kynurenine and product I, and the cross peaks inthe COSY spectrum. Substituent group rules and resonanceeffects were also used to verify the assignments in thearomatic region.The identity of the glycosylamine was determined usingthe following:1. The presence of the anomeric proton and aromaticamine proton indicates the presence of an intact sugarring.2. A cross peak in the COSY spectrum for product Ishowed that the anomeric proton and the amineproton are connected in the structure through two tothree bonds.3. Decoupling experiments identied the and formsof the glycosylamines.4. Chemical shifts were assigned to complete the structuralidentication.1 D NMR spectrum of product I and II in deuteratedDMSO showed a doublet at 9.0 and 9.1 ppm that wereassigned to the aromatic amine protons for products I and II,respectively. The triplet at 4.5 ppm for product I and at5.1 ppm for product II was typical of the anomeric proton onthe glucose and was assigned to be the C1 proton (anomericproton) on the glucose ring. A 2 D 1H 1H COSYexperimentwas conducted for product I (Fig. 5). It showed a distinctcross peak between the proton at 9.0 ppm and the proton atTable V. Overall Observed Rate Constant (kobs), Half Lives (h), and Percent of Initial Concentration of Kynurenine at Equilibrium for theReactions of 1.2 mM Kynurenine with 0.5 M Glucose in Hydrochloric Acid and Phosphate Buffer at an Ionic Strength of 0.1 0.5 and 40C inthe pH Range of 1 to 6.5Buffer pH kobs (h 1) Half life (h) Percent of initial concentration at equilibriumHCl 1.66 3.87 0.179 78.58HCl 2.11 2.43 0.285 72.20HCl 4.39 0.118 5.87 65.24NaH2PO4/Na2HPO4 5.94 0.026 26.86 63.27The estimated rate constant values for the forward and reverse reactions are contained in paper Glycosylation of aromatic amines II: Kineticsand Mechanisms of the Hydrolytic Reaction between Kynurenine and Glucose (in press)sTable VI. Estimated Values for the Forward (kfobs) and ReverseRate (krobs) Constants for Reactions of 1 mM Kynurenine with 0.190.70 M Glucose in Acetic Acid/Sodium Acetate Buffer at pH 3.4 and40CConcentration of glucose (M) kfobs (h 1) krobs (h 1)0.193 0.25 1.840.323 0.50 2.560.504 0.75 2.290.701 1.19 2.77325 Glycosylation of Aromatic Amines I: Characterization of Reaction Products and Kinetic Scheme4.5 ppm, which indicated that they were two to three bondsaway, and hence the two peaks were assigned to be the amineand the anomeric proton, respectively. In order to furtherconrm these assignments and identify the and forms, thedecoupling experiments were used that provided two piecesof information:1. Decoupling of either of those protons affected thesplitting pattern of the other proton, indicating thetwo protons were connected.2. It could be used to distinguish between the isomericforms i.e. / forms.The decoupling experiment was performed by setting thedecoupler frequency at the frequency of the aromatic aminepeak for each of the products I and II. Thus, the effect of thearomatic amine proton on the anomeric proton of glucosewas removed and simplifying the splitting pattern. Theanomeric proton splitting pattern changed from a triplet to adoublet. This conrmed that the protons at 9.0 and 9.1 ppmand the anomeric protons at 4.5 and 5.1 ppm were connectedin the structure for products I and II, respectively. Thecoupling constant (J) for the decoupled doublets wasdetermined to be 7.98 for product I and 3.67 for product II.These coupling constants were typical to that of the glucose(J~8) and the glucose (J~4), respectively (6,30,40). Theresult supported the assumption that product I was the glycosylamine and the product II was the glycosylamine.The more stable form, the glycosylamine, appeared in ahigher ratio in the reaction mixture, in agreement with theobservations of Sunderland and coworkers for the reaction ofglucose and sulfamethoxazole (6).Figure 5 shows the expanded COSY spectrum in theregion of the aromatic protons for product I. The chemicalshift assignments were done using the kynurenine assignments as a reference and also considering the splittingpatterns of the peaks in the 1 D NMR spectrum. The regionof 3 3.7 ppm in the COSY spectrum showed two spin systemsoverlapping each other, one from the protons in the sidechain of the amino acid and the other from the sugar protons(Fig. 5). One of the spin system overlayed with thekynurenine side chain region and was assigned accordingly.The other spin system was due to the sugar protons. Specicproton assignments for the sugar part of the glycosylamine(Product I) was done by initially assigning the anomericproton (C1) at 4.5 ppm followed by assigning the proton at3.1 ppm as C2 based on the cross peak observed in the COSYspectrum. Similarly the subsequent protons were assignedusing cross peaks, peak integrations, and comparing typicalassignments reported in literature for a D glucose (30).Proton signals for the amino acid side chain and sugar ringwhich lie around 3.2 3.35 ppm are also overlapped by thewater peak due to interference of water signals in this regionof the spectrum. The chemical shift assignments for theglycosylamines are reported in Table IV. The chemical shiftsfor the glycosylamines for the anomeric protons correlatedwell with those determined by Lucida and co workers (6).The product II was identied based on the mass, the couplingconstant (J), and similar chemical linkages identied for the1 D spectra of products I and II. However, individualchemical shift assignments for product II were not donebecause the low concentrations and instability of the sampleresulted in a noisy spectrum. The characterization studiesdemonstrated that for the reactions of kynurenine andglucose, product I formed was the glycosylamine and theproduct II was the glycosylamine.Reaction Scheme and KineticsThe reaction of kynurenine and glucose reversiblyformed glycosylamines in the pH range of 1 to 6. Based onthe concentration time proles, only three species existed insolution in detectable quantities, kynurenine and the twoglycosylamines. The glycosylamines equilibrated in solutionwith each other through the acyclic imine. This is analogousto the , glucose that exists in equilibrium in solution with0123450 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8k (h-1)sugar Conc (M)Fig. 7. The forward (solid circle) and reverse (solid square) rateconstants as a function of initial glucose concentration in sodiumacetate/acetic acid buffer at pH 3.49 and 40C00.10.20.30.40.50.60.70 0.02 0.04 0.06 0.08 0.1half life (hr)total acyclic content (%)glucoseallosemannosegalactoseguloseFig. 8. Half life (h) versus total acyclic content (%) for reactions of1.2 mM kynurenine with 0.5 M monosaccharides for reactions atpH 2.7, 0.1 ionic strength, and 40C326 Gokhale, Kearney, and Kirschthe aldehydic form (23). However, no detectable concentrations of the imine were obtained under the experimentalconditions. No loss in mass balance was observed. A simplereversible scheme (Eq. 2) was used to describe the data. Thepseudo rst order rate constant (kobs) for the loss ofkynurenine was determined based on reversible kinetics.The rate of formation of glycosylamines was pH dependent.For example, the reaction half life at pH 1.66 was 0.179 hwhile the half life at pH 5.94 was 26.86 h (Table V). pHdependent reaction rates for the reactions of daptomycin withsome monosaccharides were also reported earlier by Inmanand Kirsch (28). The reaction extent was also a function ofpH as shown in Table V.Reaction scheme conrmation was obtained by conducting a series of reactions with various concentrations of excessglucose. The forward rate constant (kfobs) was determined tobe a function of sugar concentration; however, the reversereaction rate constant (kfobs) was independent of the glucoseconcentration (Table VI, Fig. 7).Reactions of kynurenine with the other monosaccharides indicated that the extent of reaction was not a functionof acyclic content; however, the reaction half life was afunction of the total acyclic content (Fig. 8). This indicatedthat the reaction kinetics for these sugar weakly basic aminereactions were a function of the inherent sugar propertiesand a higher equilibrium acyclic concentration resulted inincreased reactivity.CONCLUSIONSKynurenine contains a weakly basic aromatic amine thatreacts with the aldehydic moiety of reducing sugars undermildly acidic conditions to form glycosylamines. The reactionoccurs reversibly via a steady state imine intermediate, andthe reaction rate is dependent on the concentration of theacyclic form of the sugar. The pH dependent mechanisms ofthis reaction are the subject of the second paper in this series.REFERENCES1. G. P. Ellis, and J. Honeyman. Advances in carbohydratechemistry. Glycosylamines. 10:95 125 (1955).2. W. Pigman and D. Horton. The carbohydrates: Chemistry andBiochemistry 2nd edition, IB ed., Academic Press, New York,1980, pp. 881 921.3. V. Kumar, and G. S. Banker. Maillard reactions in chemistry, foodand health; Maillard reaction and drug stability, The RoyalSociety of Chemistry, Cambridge, 1994, pp. 20 27.4. R. N. Duvall, K. T. Koshy, and J. W. Pyles. Comparison ofreactivity of amphetamine, metamphetamine, and dimethylamphetamine with lactose and related compounds. J. Pharm. Sci.54:607 611 (1965b).5. D. D. Wirth, S. W. Baertschi, R. A. Johnson, S. R. Maple, M. S.Miller, D. K. Hallenbeck, and S. M. Greg. Maillard reaction oflactose and uoxetine hydrochloride, a secondary amine. J.Pharm. Sci. 871:31 39 (1998).6. H. Lucida, J. E. Parkin, and V. B. Sunderland. Kinetic study ofthe reaction of sulfamethoxazole and glucose under acidicconditions I. Effect of pH and temperature. Int. J. Pharm.202:47 61 (2000).7. J. E. Parkin, K. F. Ilett, and D. A. Joyce. Reaction ofsulphamethoxazole in Bactrim DS with carbohydrates to formglucosylamines in vitro and in vivo. Eur. J. Pharm. Sci. 5:347 353(1997).8. A. Sianipar, J. E. Parkin, and V. B. Sunderland. The reaction ofprocainamide with glucose following admixture to glucoseinfusion. Int. J. Pharm. 176:55 61 (1998).9. W. Muangsiri, W. R. Kearney, L. M. Teesch, and L. E. Kirsch.Studies on the reactions between daptomycin and glyceraldehyde. Int. J. Pharm. 289:133 150 (2005).10. S. M. Blaug, and W. T. Huang. Interaction of dextroamphetamine sulfate with dextrates in solution. J. Pharm. Sci. 62:653655 (1973).11. S. M. Blaug, and W. T. Huang. Browning of dextrates in solidsolid mixtures containing dextroamphetamine sulfate. J. Pharm.Sci. 63:1415 1418 (1974).12. C. Petriella, J. Chirife, S. L. Resnik, and R. D. Lozano.Correlation between induction time and rate of browning inheated model solutions of glucose and lysine. Int. J. Food Sci.Tech. 23:415 418 (1988).13. H. C. Warmbier, R. A. Schnickels, and T. P. Lazuba. Nonenzymicbrowning kinetics in an intermediate moisture model system:effect of glucose to lysine ratio. J. Food Sci. 48:981 983 (1976).14. C. M. Lee, B. Sherr, and Y. N. Koh. Evaluation of kineticparameters for a glucose lysine Maillard reaction. J. Agric. FoodChem. 32:379 382 (1984).15. T. P. Lazuba, and S. A. Massaro. Browning and amino acid lossin model total parental nutrition solutions. J. Food Sci. 55:821826 (1990).16. L. N. Bell, K. L. White, and Y. H. Chem. Glycine loss andMaillard browning as related to the glass transition in a modelfood system. J. Food Sci. 63:625 628 (1998a).17. T. P. Lazuba. Interpreting the complexity of the kinetics of theMaillard reaction. In T. P. Lazuba, G. A. Reineccious, V. M.Monnier, J. OBrien, and J. Baynes (eds.), Maillard reactions inchemistry food and health, 5th ed., The Royal Society ofChemistry, Cambridge, 1994, pp. 176 181.18. C. G. A. Davies, B. L. Wedzich, and C. Gillard. Kinetics modelof the glucose glycine reaction. Food Chem. 60:323 329 (1997).19. S. E. Fayle, and J. A. Gerrard. The Maillard reaction, The RoyalSociety of Chemistry, Cambridge, UK, 2002.20. W. M. Baisier, and T. P. Lazuba. Maillard browning kinetics ina liquid model system. J. Agric. Food Chem. 40:703 713(1992).21. J. A. Stamp, and T. P. Lazuba. Kinetics of the Maillard reactionbetween aspartame and glucose in solution at high temperature.J. Food Sci. 48:543 545 (1983).22. H. S. Isbell, and H. L. Frush. Mutarotation, hydrolysis andrearrangement reactions of glycosylamines. J. Org. Chem.23:1309 1319 (1958).23. B. Capon and B. E. Connett. The mechanism of hydrolysis of Naryl D glucosylamines. J. Chem. Soc. 4497 4502 (1965).24. V. Das Gupta. Complexation of procainamide with dextrose. J.Pharm. Sci. 719:994 996 (1982).25. V. Das Gupta. Complexation of procainamide with hydroxidecontaining compounds. J. Pharm. Sci. 722:205 207 (1983).26. A. Sianipar, J. E. Parkin, and V. B. Sunderland. Chemicalincompatibility between procainamide hydrochloride and glucose following intravenous admixture. J. Pharm. Pharmacol.46:951 955 (1994).27. The United States Pharmacopeia. 23:1463 (1995).28. E. L. Inman and L. E. Kirsch. An improved diluent formulationfor daptomycin., in PCT Int. 1990: European.29. W. Muangsiri, and L. E. Kirsch. The kinetics of the alkalinedegradation of daptomycin. J. Pharm. Sci. 908:1066 1075 (2001).30. H. S. Khadem. Carbohydrate chemistry, monosaccharides andtheir oligomers, Academic Press Inc, San Diego, California, 1988,pp. 11 56.31. A. Albert, and E. P. Searjeant. The determination of ionizationconstants. A laboratory manual, 3rd ed., Chapman and Hall, NewYork, 1984, pp. 70 73.32. W. Schaller, and A. D. Robertson. pH, ionic strength, andtemperature dependences of ionization equilibria for the carboxyl groups in Turkey ovomucoid third domain. Biochemistry.34:4714 4723 (1995).33. W. R. Forsyth, and A. D. Robertson. Insensitivity of perturbedcarboxylic pKa values in the ovomucoid third domain chargereplacement at a neighboring residue. Biochemistry. 39:80678072 (2000).327 Glycosylation of Aromatic Amines I: Characterization of Reaction Products and Kinetic Scheme34. Y. Zhu, J. Zajicek, and A. S. Serianni. Acyclic forms of [1 13C]aldohexoses in aqueous solution: Quantication 13C NMR anddeuterium isotope effects on tautomeric equilibria. J. Org. Chem.66:6244 6251 (2001).35. G. Volgyi, R. Ruiz, K. Box, J. Comer, E. Box, and K. TakacsNovak. Potentiometric and spectrophotometric pKa determinationof water insoluble compounds: Validation study in a new cosolventsystem. Anal. Chim. Acta. 583:418 428 (2007).36. P. K. Glasoe, and F. A. Long. Use of glass electrodes tomeasure acidities in deuterium oxide. J. Phys. Chem. 641:188190 (1960).37. D. D. Perrin, and B. Demsey. Buffers for pH and metal ioncontrol, Chapman and Hall, London, 1974, pp. 81 82.38. I. N. Gordon and B. M. Lowe. The dissociation constants ofamino acids in deuterium oxide. J. Chem. Soc. Chem. Commun.13(803 804) (1970).39. G. C. K. Roberts. NMR of macromolecules. A practicalapproach. In D. Rickwood, and B. D. Hames (eds.), ThePractical Approach Series, Oxford University Press, New York,1993, pp. 22 23.40. P. A. Harmon, W. Yin, W. E. Bowen, R. J. Tyrell, and R. A.Reed. Liquid Chromatography mass spectrometry and protonnuclear magnetic resonance characterization of trace levelcondensation products formed between lactose and the aminecontaining diuretic hydrochlorothiazide. J. Pharm. Sci. 897:920929 (2000).328 Gokhale, Kearney, and KirschBrief/Technical NotePhysicochemical Properties of Solid Dispersions of Gliclazidein Polyvinylpyrrolidone K90S. Biswal,1,2J. Sahoo,1and P. N. Murthy1Received 27 June 2008; accepted 13 February 2009; published online 25 March 2009KEY WORDS: dissolution; gliclazide; polyvinylpyrrolidone K90; solid dispersion.INTRODUCTIONGliclazide is a second generation hypoglycemic sulfonylurea which is useful in the treatment of type 2 diabetesmellitus (1). Following oral administration, however, gliclazide exhibits slow rate of absorption and interindividualvariation in bioavailability. Stated problems of gliclazide mightbe due to its poor water solubility and slow dissolution rate (24). But gliclazide exhibits good tolerability, low incidence ofhypoglycemic effect, low rate of secondary failure, and low rateof progression of diabetic retinopathy (2,5). Hence, gliclazideappears to be a drug of choice in long termsulfonylurea therapyfor treatment of type 2 diabetes mellitus. Few attempts havebeen made for improvement of solubility and bioavailability ofgliclazide including complexation with cyclodextrin (6,7) orcyclodextrin hydroxypropylmethylcellulose (8) and using PEG400 (9) as per present literature. The authors investigated thephysicochemical characteristics and dissolution behaviors ofgliclazide in physical mixtures as well as solid dispersions withpolyethylene glycol 6000 in a previous study (10).The main objective of this work was to investigate thephysicochemical characteristics of gliclazide in physical mixtures (PMs) and solid dispersions (SDs) prepared by usingpolyvinylpyrrolidone K90 (PVP K90). The possible interactions between gliclazide and PVP K90 in both solid and liquidstates were investigated. Interaction in solid state was investigated by Fourier transform infrared (FT IR) spectroscopy, Xray diffraction (XRD) analysis, and differential scanningcalorimetry (DSC). Interaction in solution was studied byphase solubility analysis and dissolution experiments.MATERIALS AND METHODSMaterialsA gift sample of gliclazide was received from AristoPharmaceuticals Ltd. (Mumbai, India). PVP K90 was received from BASF (Germany). Double distilled water wasused throughout the study and all the other chemicals usedwere of analytical grade.Preparation of SDsThe SDs of gliclazide with PVP K90 containing threedifferent weight ratios (1:1, 1:2, 1:5; gliclazide/PVP K90) anddenoted as SD 1/1, SD 1/2, and SD 1/5, respectively, wereprepared by solvent evaporation method. In the solventevaporation method, to a solution of gliclazide in chloroform,an appropriate amount of PVP K90 in solution was added.The solvent was evaporated under reduced pressure at 40Cby using a rotary evaporator and the resulting residue wasdried under vacuum for 3 h. The mixture was storedovernight in a desiccator. The hardened mixture was powdered in a mortar, sieved through a 100 mesh screen, andstored in a screw cap vial at room temperature until furtheruse (11,12).The PMs having the same weight ratio as SDs wereprepared by thoroughly mixing the required amount ofgliclazide and PVP K90 for 10 min in a mortar. The resultingmixtures were sieved through a 100 mesh sieve and denotedas PM 1/1, PM 1/2, and PM 1/5, respectively. The mixtureswere stored in a screw cap vial at room temperature untiluse.Phase Solubility of GliclazideSolubility determinations were performed in triplicateaccording to the method of Higuchi and Connors (13). Inbrief, an excess amount of gliclazide was taken into a screwcapped glass vial to which 20 ml of aqueous solutioncontaining various concentrations of PVP K90 was added.Then, the samples were shaken at 370.5C for 72 h in awater bath (Remi Pvt Ltd, Mumbai). After 72 h, sampleswere ltered through a 0.45 m membrane lter. The ltratewas suitably diluted and analyzed spectrophotometrically at thewavelength of 227 nm using a UV VIS spectrophotometer(Shimadzu 1700, Pharm Spec, Japan).329 1530-9932/09/0200-0329/0 # 2009 American Association of Pharmaceutical ScientistsAAPS PharmSciTech, Vol. 10, No. 2, June 2009 (#2009)DOI: 10.1208/s12249 009 9212 71Royal College of Pharmacy and Health Sciences, AndhapasaraRoad, Brahmapur, 760002, Orissa, India.2To whom correspondence should be addressed. (e mail: sudarsanmpharm@yahoo.co.in)Dissolution StudiesDissolution studies of gliclazide in powder form, SDs,and PMs in triplicate were performed by using the USPharmacopoeia (USP) model digital tablet dissolution testapparatus 2 (Lab India Ltd, Mumbai) at the paddle rotationspeed of 50 rpm in 900 ml 0.1 N HCl containing 0.25% (w/v)of sodium lauryl sulfate as a dissolution medium at 370.5C(14). The SDs or PMs equivalent to 30 mg of gliclazide wereweighed using a digital balance and added into the dissolutionmedium. At the specied times (every 10 min for 1 h), 10 mlsamples were withdrawn by using syringe lter (0.45 m;Sepyrane, Mumbai) and then assayed for gliclazide contentby measuring the absorbance at 227 nm using the UV Visiblespectrophotometer (Shimadzu UV 1700, Pharm Spec). Freshmedium (10 ml), which was prewarmed at 37C, was added tothe dissolution medium after each sampling to maintain aconstant volume throughout the test.Differential Scanning CalorimetryThe DSC measurements were performed on a DSC 6100(Seiko Instruments, Japan) differential scanning calorimeterwith a thermal analyzer. All accurately weighed samples(about 1.541 mg of gliclazide or its equivalent) were placed insealed aluminum pans, before heating under nitrogen ow(20 ml/min) at a scanning rate of 10C min 1from 25C to250C. An empty aluminum pan was used as a reference.X-ray DiffractionThe X ray powder diffraction patterns were obtained atroom temperature using a PW1710 X ray diffractometer(Philips, Holland) with Cu as anode material and graphitemonochromator, operated at a voltage of 35 kV, current20 mA. The samples were analyzed in the 2 angle range of5 70, and the process parameters were set as: scan step sizeof 0.02 (2) and scan step time of 0.5 s.Fourier-Transform Infrared SpectroscopyFT IR spectra were obtained by using an FT IR spectrometer 430 (Jasco, Japan). The samples (gliclazide or SDsor PMs) were previously ground and mixed thoroughly withpotassium bromide, an infrared transparent matrix, at 1:5(sample/KBr) ratio, respectively. The KBr disks were prepared by compressing the powders at a pressure of 5 tons for5 min in a hydraulic press. Forty scans were obtained at aresolution of 4 cm 1, from 4,600 to 300 cm 1.Dissolution Data AnalysisPhase Solubility StudiesThe value of apparent stability constant, Ks, betweendrug carrier combinations was computed from the phasesolubility proles, as described belowKs SlopeIntercept 1 Slope : 1The Gibbs free energy of transfer (Gtr) of gliclazidefrom pure water to the aqueous solutions of carrier wascalculated as$Gtr 2:303 RTlogSoSs2where SoSsis the ratio of molar solubility of gliclazide inaqueous solution of PVP K90 to that of the same mediumwithout PVP K90.In Vitro Dissolution DataIn the present investigation, model independent andmodel dependent approaches are used for comparison ofdissolution proles. Model independent approaches arebased on the ratio of area under the dissolution curve(dissolution efciency) or on mean dissolution time (15,16).Percent dissolution efciency (%DE) and mean dissolutiontime were also computed to compare the relative performance of various concentrations of carrier in SDs and PMs.The magnitude of %DE at 10 min (%DE10 min) and 30 min(%DE30 min) for each formulation was computed as thepercent ratio of area under the dissolution curve up to time t(10 and 30 min), to that of the area of the rectangle describedby 100% dissolution at the same time. The magnitude ofmean dissolution time for each formulation was calculatedusing PCP Disso v3 software (Pune, India).%DE Rt0YdtY100t 3In the model dependent approaches, release datawere tted to ve kinetic models including the zero order(Eq. 4), rst order (Eq. 5), Higuchi matrix (Eq. 6), PeppasKorsmeyer (Eq. 7), and Hixson Crowell (Eq. 8) releaseequations. PCP Disso v3 software (Pune, India) was used tond best t model.R k1t 4log UR k2t2:303 5R k3 tp 6log R log k4nlog t 7UR 1=3 k5t 8where R and UR are the released and unreleased percentages, respectively, at time t; k1, k2, k3, k4, and k5 are the rateconstants of zero order, rst order, Higuchi matrix, PeppasKorsmeyer, and Hixon Crowell models, respectively.330 Biswal, Sahoo and MurthyRESULTSPhase Solubility StudiesThe phase solubility diagram investigated in 0.1 N HCl(pH 1.2) was linear in a wide range of PVP K90 concentrations and correspond to AL type proles (13). The stabilityconstant was found to be 1.20 ml 1mg. At 18% (w/v)concentration of PVP K90, the solubility of gliclazideincreased by 4.6 fold (Table I). Table I presents the valuesof Gibbs free energy associated with the aqueous solubility ofgliclazide in the presence of PVP K90.Dissolution StudiesQ10, Q20, and Q30 values (percent drug dissolved within30 min) are reported in Table II (coefcient of variation isless than 10% in each case). The onset of dissolution of puregliclazide is very low, about 18.46% of drug being dissolvedwithin 10 min (Q10). SDs of gliclazide with PVP K90considerably enhanced dissolution rates within 10 min compared to pure gliclazide such as SD at 1:1 (gliclazide/PVPK90) ratio up to 76.29%, SD at 1:2 ratio up to 77.12%, andSD at 1:5 ratio up to 77.73%, respectively. In case of PMs ofgliclazide, Q10 values increased up to 25.36% at 1:1, up to33.1% at 1:2, and up to 43.88% at 1:5 (gliclazide/PVP K90)ratio, respectively. From Table II, Q20 as well as Q30 valuesindicate, in general, PVP K90 based formulations (SDs andPMs) at high carrier levels exhibited higher dissolution ratesthan those at low polymer levels (PM 1:1 and SD 1:1 with PM1:5 and SD 1:5, respectively).The %DE values were computed at two times, showingthe early and late phase of dissolution study for comparativeanalysis of all the formulations. The %DE values in the initialtime period of dissolution study, i.e., %DE10 min, providecomparative information for very fast releasing formulations,whereas %DE30 min provides relative information about bothfast and slow releasing formulations. The value of %DE10 minfor pure gliclazide (9.16%) was enhanced in PMs (21.94%) aswell as in SDs (49.87%). The value of %DE30 min for the puredrug increased to 45.11% in PMs and up to 73.20% in SDs(Table II).The obtained values of mean dissolution time (MDT) forpure gliclazide, PMs, and SDs are presented in Table II. TheMDT of gliclazide was 12.5 min and it decreased to 7.09 minin SD with PVP K90 at 1:5 (gliclazide/PVP K90) ratio,whereas in case of PMs, MDT decreased to 9.43 min.Table III shows the regression parameters obtained aftertting various release kinetic models to the in vitro dissolutiondata. In vitro release data of the drug are best tted toKorsemeyer Peppas model with n value of 0.7249, henceexhibits non Fickian diffusion. For all PMs (PM 1:2 and PM1:1), the Higuchi matrix is best tted except PM 1:5 whichexhibits Korsemeyer Peppas model. All the SDs best tted toKorsemeyer Peppas model with n values less than 0.4500tended to exhibit Fickian diffusion characteristics except SD1:5 which exhibits rst order release kinetics (17,18).Differential Scanning CalorimetryThe DSC curve of pure gliclazide exhibited a singleendothermic response corresponding to the melting of thedrug. Onset of melting was observed at 170.8C, thecorresponding heat of fusion (HF) was 171.2 J/g (Fig. 1 A)(10). During scanning of PVP K90, a broad endothermranging from 80C to 120C was observed, due to thepresence of water. DSC thermograms of PMs and SDs ofgliclazide and PVP K90 always exhibited complete absence ofmelting peak of the drug at 170.8C and the broad endothermdue to the presence of water ranging from 60C to 120C(Fig. 1 D). Repeated scanning of PVP K90 based formulaTable I. Effect of PVP K90 Concentration and Gibbs Free Energy onSolubility of Gliclazide in 0.1 N HClConcentrationof PVP K90(% w/v)Concentrationof gliclazide(mg ml 1)at 37C Gtr (J/mol)1 0 0.812 2 0.81 03 4 1.21 1,143.554 6 1.53 1,749.305 8 1.87 2,266.56 10 2.28 2,672.737 12 2.61 3,020.838 14 2.96 3,345.229 16 3.33 3,651.2110 18 3.7 3,920.44PVP K90 polyvinylpyrrolidone K90Table II. In Vitro Dissolution Prole of drug, Physical Mixtures, and Solid Dispersions of Gliclazide in pH 1.2 (0.1 N HCl)Sr. no. FormulationDissolution parametersQ10 min Q20 min Q30 min %DE10 min %DE30 min MDT (min)1 Drug 18.46 32.67 40.82 9.16 23.67 12.52 PM 1:1 25.36 38.93 56.68 12.68 30.88 13.663 PM 1:2 33.1 48.8 59.5 16.55 37.21 11.244 PM 1:5 43.88 58.54 65.77 21.94 45.11 9.435 SD 1:1 76.29 80.36 88.24 38.15 66.92 7.256 SD 1:2 77.12 87.12 92.45 45.56 70.16 7.247 SD 1:5 77.73 93.95 95.84 49.87 73.20 7.09PM physical mixture, SD solid dispersion of gliclazide prepared by the solvent evaporation method, %DE percent dissolution efciency, MDTmead dissolution time331 Properties of Gliclazide in Polyvinylpyrrolidone K90tions led to disappearance of the endotherm, which is dueto evaporation of water during the rst run. The presenceof broad endothermic peak of PVP K90 and formulationson DSC thermogram was reported by several researchers(19 21).X-ray DiffractionsFigure 2 shows the X ray diffraction pattern of puregliclazide, PVP K90, its physical mixture, and solid dispersion.In the X ray diffraction pattern of gliclazide, sharp peaks arepresent at 2 of 10.59, 14.98, 17.2, 17.85, 18.15, 22.07, 25.42,26.25, 26.75, and 29.5 (Fig. 2 A), and it suggests that gliclazideis a crystalline material. Pure PVP K90 shows absence ofpeaks in diffraction spectrum (Fig. 2 B). Peaks characteristicof the drug were observed in the X ray diffractogram ofphysical mixture and solid dispersion.FT-IR SpectroscopyThe IR spectra of SDs and PMs were compared with thestandard spectrum of gliclazide (Fig. 3 A). The IR spectrumof gliclazide is characterized by the absorption of carbonyl(C=O) sulfonyl urea group at 1,706 cm 1(6). In the spectra ofSDs and PMs, this band was shifted towards higher wavenumber at 1,711 and 1,709 cm 1, respectively (6). Also, theNH group which is located at 3,265 cm 1from the IRspectrum of gliclazide was shifted to 3,630 cm 1in SDs. Thesulfonyl group bands are located at 1,349 and 1,162 cm 1inpure gliclazide. In SDs, the asymmetrical vibration peak ofS=0 band was shifted from 1,349 to 1,342 cm 1withdecreased frequencies. In SDs, the symmetrical stretchingvibration band of S=0 was shifted from 1,162 to 1,113 cm 1with decreased frequencies. The spectrum of PVP K90exhibited important bands at 2,953 cm 1(C H stretch) and1,652 cm 1(C=O). A very broad band was also visible at3,446 cm 1which was attributed to the presence of water,conrming the appearance of broad endotherm in DSC rundue to the presence of water (20).DISCUSSIONThe results of phase solubility are in accordance with thewell established formation of soluble complexes betweenTable III. Statistical Parameters of Various Formulations After Fitting Drug Release Data to Various Release Kinetic ModelsFormulationsZero order model First order model H M model P K model H C modelR k1 R k2 R k3 R k4, n R K5Drug 0.9813 1.4736 0.9939 0.0184 0.9943 7.1358 0.9952 3.4734, 0.7249 0.9905 0.0057PM 1:1 0.8256 1.2789 0.8975 0.0190 0.9764 8.6575 0.9560 8.5742, 0.5041 0.8774 0.0055PM 1:2 0.7319 1.3776 0.8582 0.0213 0.9695 9.4130 0.9608 15.1667, 0.3712 0.8214 0.0061PM 1:5 0.5379 1.5087 0.7312 0.0247 0.9265 10.4269 0.9421 26.6687, 0.2439 0.6742 0.0069SD 1:1 0.1647 2.0615 0.7982 0.0533 0.8589 14.3741 0.9812 57.9627, 0.1171 0.6450 0.0123SD 1:2 0.2008 2.1641 0.8956 0.0685 0.8646 15.0883 0.9809 58.4002, 0.1287 0.7345 0.0143SD 1:5 0.1099 2.2337 0.9362 0.0903 0.8560 15.5967 0.9166 60.4486, 0.1284 0.7612 0.0165H M Higuchi matrix, P K Peppas Korsmeyer, H C Hixon Crowell, R correlation coefcient, k1 k5 constants of release kinetics, PM physicalmixture, SD solid dispersion of gliclazide prepared by the solvent evaporation methodFig. 1. DSC thermograms of pure gliclazide a, pure PVP K90 b,gliclazide PVP K90 PMs at 1:2 ratio c, and gliclazide PVP K90 SDsat 1:2 ratio dFig. 2. X ray diffractograms of pure gliclazide a, pure PVP K90 b,gliclazide PVP K90 PMs at 1:2 ratio c, and gliclazide PVP K90 SDsat 1:2 ratio d332 Biswal, Sahoo and Murthywater soluble polymeric carriers and poorly water solubledrugs (17). Gtr values were all negative for PVP K90 atvarious concentrations indicating the spontaneous nature ofthe drug solubilization. The values decreased by increasingPVP K90 concentration, demonstrating that the reactionbecame more favorable as the concentration of PVP K90increased.The results of the dissolution study indicate an improvement of dissolution rate of gliclazide in solid dispersion. Therate of dissolution increases as the concentration of PVP K90increases in SDs. The improvement of dissolution rate ispossibly caused by several factors. Such factors are: (a) thestrong hydrophilic character of PVP K90, which improves thewater penetration and the wettability of the hydrophobicgliclazide; (b) the optimal dispersion of gliclazide to PVP K90;(c) the absence of crystals (amorphous dispersions) corresponds to lower energy required for dissolution; and (d) theintermolecular hydrogen bonds and the molecular dispersionof gliclazide on PVP leads to partial miscibility, improving thehydrophilic characteristics of the drug substance via interactions within the polymer (22). The improvement ofdissolution rate of gliclazide in PMs is due to increasedwettability of the drug powder (23).Thermograms of SD (Fig. 1 D) and PM showed theabsence of a gliclazide melting endothermic peak. Theabsence of drug melting endothermic peak in PM and SDindicate that gliclazide is present in amorphous form withinPMs and SDs. The inhibition of crystallization of drug indispersions results in the amorphous form of gliclazide.Crystallization inhibition is attributed to two effects: interactions such as hydrogen bonding between the drug and thepolymer and the entrapment of the drug molecules in thepolymer matrix during solvent evaporation or a combinationof both (24). Numerous studies have shown that polymerslike povidone used in solid dispersions can inhibit thecrystallization of drugs resulting in an amorphous form ofthe drug in the solid dispersions (25,26). The formation of anamorphous form of gliclazide in SDs is due to the combination of the two stated effects (22,24). In order to verify theDSC results and to exclude the possibility of existence ofcrystalline material in solid dispersion, these systems wereagain evaluated by using X ray diffraction analysis.The prominent peaks from pure gliclazide such as at 2of 10.59, 14.98, 17.2, etc. were observed clearly at the sameposition in the PMs and in SDs, and their intensities weredecreased by 55 60% and by more than 80%, respectively.The X ray diffraction ndings suggested that some portion ofgliclazide still existed in the same crystal structures of thepure drug, but the relative reduction of diffraction intensity ofgliclazide in SDs at these angles suggests that the crystal sizewas reduced to microcrystal form (27). The nding of theXRD such as existence of some portion of gliclazide in thesame crystal structures of the pure drug is not in agreementwith the DSC nding (complete conversion of drug toamorphous form) (28). The XRD ndings again suggest thatthe melting peak of gliclazide in SDs and PMs containingsome portion of crystalline gliclazide was also absent. This isdue to the interaction between gliclazide and polymer in solidstate (28). This further conrms that DSC is not useful forexamining the solid state of drug within the PMs and SDs(28). The existence of some portion of gliclazide in the samecrystal structures of pure drug was conrmed by the XRDstudy but not by the DSC study. The absence of anendothermic peak of gliclazide in SDs and PMs suggests anamorphous form of the drug, but again the presence of asharp peak of the drug with lesser intensity indicates someportion of crystalline drug. DSC and XRD results are inagreement when the crystalline form of the drug is convertedinto amorphous form completely. Absence of endothermicand diffraction peak of the drug in dispersion indicate anamorphous form, where DSC and XRD results are inagreement with each other. Absence of an endothermic peakof the drug in dispersion and presence of a diffraction peak inXRD with less intensity indicate that the drug is partiallyconverted into amorphous form, and the DSC and XRDresults, which are not in agreement with each other, could bedue to the presence of some portion of the crystalline drugand the presence of polymer. Microcrystals are formed as aconsequence of evaporation of solvent during the preparationof solid dispersions by the solvent evaporation technique.Evaporation of solvent increases the viscosity very rapidlyleading to a decrease in drug mobility preventing recrystallization. When the solvent is evaporated completely,drug molecules are frozen in the polymer. A crystal lattice isnot formed, but the drug molecules are of randomlydispersed order over only a few molecular dimensions(20). The existence of interaction between the drug andpolymer is again suggested by FT IR study.The shift in the peaks associated with C=O, S=O, andNH groups of gliclazide indicates some sort of solid stateinteractions between the drug and the polymer in SD andPM. The interactions are due to intermolecular hydrogenbonding between the drug and polymer. An intermolecularhydrogen bond was expected to occur between the hydrogenatom of the NH group of gliclazide and one of the lone pairelectrons of C=O group of polymer and/or C=O group ofgliclazide and one of the hydrogen atoms of PVP K90 (20).The shift in the peaks associated with the S=O group ofgliclazide could be due to involvement of complexation withPVP K90 (29,30). The FT IR data suggest the formation ofFig. 3. FT IR spectrograms of pure gliclazide a, pure PVP K90 b,gliclazide PVP K90 PMs at 1:2 ratio c, and gliclazide PVP K90 SDsat 1:2 ratio d333 Properties of Gliclazide in Polyvinylpyrrolidone K90intermolecular hydrogen bonding between gliclazide andPVP K90. The absence of the endothermic peak of the drugin SDs or PMs is due to intermolecular hydrogen bonding andpartial conversion drug into amorphous form.ConclusionsThe solubility and dissolution rate of gliclazide can beenhanced by formulating SDs of gliclazide with PVP K90.The solubilization effect of PVP K90, reduction of particleaggregation of the drug, formation of microcrystalline oramorphous drug, increased wettability and dispersibility, anddevelopment of intermolecular hydrogen bonding are responsible for the enhanced solubility and dissolution rate ofgliclazide from its SDs and to some extent in PMs. Theresults of infrared spectroscopy, X ray diffractometry, andDSC indicate that some sort of interactions such as intermolecular hydrogen bonding or complexation between thefunctional groups of gliclazide and PVP K90 have occurredin the molecular level and show microcrystallinity of gliclazide in the solid dispersion, which increased the solubility andthe dissolution of gliclazide from the solid dispersion.ACKNOWLEDGMENTSThe authors are grateful to Aristo Pharmaceuticals Pvt.Ltd, Mumbai, India. Mr. S. Biswal acknowledges the Principalof the Royal College of Pharmacy and Health Sciences,Brahmapur, for providing the necessary facilities.REFERENCES1. Reynolds JEF, editor. Martindale. The Extra PharmacopoieaXXX. 30th edn. London: The Pharmaceutical Press; 1993. p.279 80.2. Palmer KJ, Brogden RN. Gliclazide, an update of its pharmacological properties and therapeutic efcacy in NIDDM. Drugs1993;46:92 125.3. Lobenberg R, Amidon GL. Modern bioavailability, bioequivalence and biopharmaceutics classication system: new scienticapproaches to international regulatory standards. Eur J PharmBiopharm. 2000;50:3 12.4. Desai KGH, Kulkarni AR, Aminabhavi TM. Solubility ofrofecoxib in presence of methanol, ethanol and sodium laurylsulfate at (298.15, 303.15 and 308.15) K. J Chem Eng Data.2003;48:942 5.5. Harrower AD. Comparison of efcacy, secondary failure rateand complications of sulfonylurea. J Diabetes Complicat.1994;8:201 3.6. zkan Y, Atay T, Dkman N, Isimer A, Aboul Enein YH.Improvement of water solubility and in vitro dissolution rate ofgliclazide by complexation with cyclodextrin. Pharm ActaHelv. 2000;74:365 70.7. Winters S, York P, Timmins P. Solid state examination of agliclazide: beta cyclodextrin complex. Eur J Pharm Sci.1997;5:209 14.8. Aggarwal S, Singh PN, Mishra B. Studies on solubility andhypoglycemic activity of gliclazide beta cyclodextrin hydroxypropylmethylcellulose complexes. Die Pharmazie. 2002;57:191 3.9. Hong SS, Lee SH, Lee YJ, Chung SJ, Lee MH, Shim CK.Accelerated oral absorption of gliclazide in human subjects froma soft gelatin capsule containing a PEG 400 suspension ofgliclazide. J Control Release. 1998;51:185 92.10. Biswal S, Sahoo J, Murthy PN, Giradkar PR, Avari JG.Enhancement of dissolution rate of gliclazide using soliddispersions with polyethylene glycol 6000. AAPS PharmaSciTech.2008;9(2):563 70.11. Trapani G, Franco M, Latrofa A, Pantaleo MR, ProvenzanoMR, Sanna E, et al. Physicochemical characterization and in vivoproperties of Zolpidem in solid dispersions with polyethyleneglycol 4000 and 6000. Int J Pharm. 1999;184:121 30.12. Trapani G, Franco M, Latrofa A, Tullio C, Provenzano MR,Serra M, et al. Dissolution properties and anticonvulsant activityof phenytoin polyethylene glycol 6000 and polyvinylpyrrolidoneK 30 solid dispersions. Int J Pharm. 2001;225:63 73.13. Higuchi T, Connors K. Phase solubility techniques. Adv AnalChem Instrum. 1965;4:17 123.14. Damian F, Blaton N, Naesens L, Balzarini J, Kinget R,Augustinjns P, et al. Physicochemical characterization of soliddispersions of the antiviral agent UC 781 with polyethyleneglycol 6000 and Gelucire 44/14. Eur J Pharm Sci. 2000;10:311 22.15. Khan CA, Rhodes CT. The concept of dissolution efciency. JPharm Pharmacol. 1975;27:48 9.16. Arias MJ, Gines JM, Moyano JR, Rabasco AM. Dissolutionproperties and in vivo behaviour of triamterene in soliddispersions with polyethylene glycols. Pharm Acta Helv.1996;71:229 35.17. Costa P, Lobo JMS. Modeling and