Physicochemical Characterization of Creatine
`N-Methylguanidinium Salts
`
`Brandon T. Gufford
`Kamaraj Sriraghavan
`Nicholas J. Miller
`Donald W. Miller
`Xiaochen Gu
`Jonathan L. Vennerstrom
`Dennis H. Robinson
`
`ABSTRACT. Creatine is widely used as a dietary supplement for body builders to
`enhance athletic performance. As the monohydrate, its low solubility in water and high
`dose lead to water retention and gastrointestinal discomfort. Hence, alternative creatine
`derivatives with enhanced water solubility and potential therapeutic advantages have
`been synthesized. As a zwitterionic compound, creatine can form salts at the N-methyl
`guanidinium or carboxylic acid functional groups. In this study, we determined the
`aqueous solubilities and partition coefficients of six N-methyl guanidinium salts of cre-
`atine compared to those of creatine monohydrate; two of these were new salts, namely,
`creatine mesylate and creatine hydrogen maleate. The aqueous solubilities of the salts
`were significantly more than that of creatine monohydrate with the hydrochloride and
`mesylate being 38 and 30 times more soluble, respectively. The partition coefficients
`of the creatine salts were very low indicating their relatively high polarity. Perme-
`abilities of creatine pyruvate, citrate, and hydrochloride in Caco-2 monolayers were
`compared to that of creatine monohydrate. Aside from the creatine citrate salt form
`that had reduced permeability, there were no significant differences in permeability
`characteristics in Caco-2 monolayers. Typical of an amphoteric compound, creatine
`is least soluble in the pH region near the isoelectric point.
`
`Brandon T. Gufford, Kamaraj Sriraghavan, Jonathan L. Vennerstrom and Dennis H. Robinson
`are affiliated with the Department of Pharmaceutical Sciences, College of Pharmacy, University of
`Nebraska Medical Center, Nebraska Medical Center, Omaha, Nebraska, USA.
`Nicholas J. Miller, Donald W. Miller, and Xiaochen Gu are affiliated with the Department of
`Pharmacology and Therapeutics, University of Manitoba, Winnipeg, Manitoba, Canada.
`Address correspondence to: Dennis H. Robinson, Department of Pharmaceutical Sciences, College
`of Pharmacy, University of Nebraska Medical Center, 986025 Nebraska Medical Center, Omaha, NE,
`USA (E-mail: dhrobins@unmc.edu).
`We thank UNeMed, Omaha, NE for the financial support of this work.
`
`Journal of Dietary Supplements, Vol. 7(3), 2010
`Available online at www.informaworld.com/WJDS
`C(cid:1) 2010 by Informa Healthcare USA, Inc. All rights reserved.
`doi: 10.3109/19390211.2010.491507
`
`240
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`001
`
`Harvest Trading Group - Ex. 1109
`
`

`
`Gufford et al.
`
`241
`
`KEYWORDS. Creatine salts, creatine monohydrate, physicochemical properties,
`solubility, log P, thermal analysis, permeability
`
`INTRODUCTION
`
`Creatine, most commonly in the form of the monohydrate, is widely used as a dietary
`supplement for bodybuilding and for its potential to improve exercise and athletic
`performance (Persky & Brazeau, 2001). Creatine may also be useful in the treatment
`of certain diseases, especially those involving muscular atrophy or fatigue associated
`with impaired energy production (Persky & Brazeau, 2001; Wyss & Kaddurah-Daouk,
`2000). As creatine monohydrate is typically used at relatively high doses (5–25 g/day),
`its water solubility of 16.6 mg/ml (Dash & Sawhney, 2002) dictates that it is administered
`as an oral suspension, leading to water retention and possible gastrointestinal discomfort
`(Persky & Brazeau, 2001).
`It is possible to make salts of the zwitterionic creatine at either the carboxylic acid or
`N-methyl guanidinium functional groups (Figure 1). Due to relatively low aqueous sol-
`ubility of creatine monohydrate, there have been several creatine N-methylguanidinium
`salts introduced in the market as dietary supplements that are claimed to be more
`water-soluble than creatine monohydrate. These include the pyruvate, hydrogen citrate,
`and hydrochloride salt forms. Other creatine N-methylguanidinium salts have been de-
`scribed in patents and/or patent applications. These salt forms include the hydrobromide
`(Rudakova, Pospelova, & Yurkevich, 1967); nitrate (Dhar & Ghosh, 1961); mesylate
`(Blatt et al., 2006); dihydrogen phosphate and hydrogen oxalate (Dhar & Ghosh, 1961);
`malate (Boldt, 2004; Cornelius & Haynes, 2004; Qian, Ye, & Huang, 2005); maleate,
`fumarate, and tartarate (Boldt, 2004); lipoate (Buononato & Festuccia, 2003; Gardiner,
`2000); hydrogen maleate, hydrogen fumarate, hydrogen tartarate, and hydrogen malate
`(Negrisoli & Del Corona, 1996); ascorbate (Pischel, Weiss, Gloxhuber, & Mertschenk,
`1998); and bicarbonate (Kneller, 2008).
`Many of these creatine N-methylguanidinium salts are formed with acids, which are
`insufficiently acidic to directly form a salt with creatine (Figure 1). Thus, it is not clear if
`these “salts” are merely physical mixtures, weak complexes, or “addition salts” (Pischel
`et al., 1998) of creatine monohydrate and the acid in question. The only way to form such
`salts with weak acids is using an ion-exchange between a creatine N-methylguanidinium
`salt and a salt of the acid as described by Qian et al. (2005) in the synthesis of creatine
`maleate using creatine hydrochloride or creatine sulfate with sodium or calcium maleate;
`by Arnold (2001) in the synthesis of creatine pyruvate from creatine hydrochloride and
`sodium pyruvate; and by Gardiner, Heuer, and Molino (2006) in the synthesis of creatine
`citrate from tripotassium citrate and creatine hydrochloride.
`
`FIGURE 1. Creatine acid–base equilibria.
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`002
`
`Harvest Trading Group - Ex. 1109
`
`

`
`242
`
`JOURNAL OF DIETARY SUPPLEMENTS
`
`The purpose of this study was firstly, to synthesize additional salts, namely creatine
`mesylate and creatine hydrogen maleate, and secondly, to determine and compare the
`physicochemical properties of these and other commercially available salts, namely
`creatine hemisulfate, creatine hydrochloride, creatine pyruvate, and “creatine citrus”
`(citrate complex). The aim was to identify a water soluble salt that may have the follow-
`ing potential advantages over creatine monohydrate: (a) increase solubility promoting
`increased bioavailability (b) lower oral dose (c) decrease side effects (d) be able to be
`formulated into a more diverse range of formulations e.g., capsules or a topical product
`for (e) a wider range of therapeutic applications e.g., anti-inflammatory effects.
`
`MATERIALS AND METHODS
`
`Materials
`
`Creatine monohydrate (CreapureTM) and creatine pyruvate were obtained from
`Degussa, creatine citrate, i.e., “creatine citrus” was obtained from Peak Nutrition, and
`creatine hydrochloride (Miller, Vennerstrom, & Faulkner, 2009) was obtained from
`Vireo Systems. With the exception of the 1-octanesulfonic acid sodium salt that was
`obtained from Fluka, all other reagent solvents were obtained from Sigma-Aldrich and
`used as received.
`
`Nuclear Magnetic Resonance (NMR), Elemental Analysis, and Melting Point
`Determination
`
`Proton Nuclear Magnetic Resonance (1H NMR) spectra of each new salt in DMSO-d6
`were recorded on a 500-MHz spectrometer. All chemical shifts are reported in parts
`per million (ppm) and are relative to internal TMS (0 ppm). Elemental analyses were
`determined by M-H-W Laboratories. Melting points were obtained using differential
`scanning calorimetry and were uncorrected.
`
`Creatine Mesylate
`
`A suspension of creatine monohydrate (0.08 mol, 11.93 g) in deionized water
`◦
`(120 ml) warmed to 59
`C was added to a stirred solution of methane sulfonic acid
`◦
`C. The reaction mixture was stirred
`(0.08 mol, 7.69 g) in ethanol (EtOH) (200 ml) at 59
`◦
`C and then allowed to cool to room temperature. The solvents were
`for 10 min at 59
`removed in vacuo affording creatine mesylate (11.19 g, 58%) as a white crystalline solid
`◦
`C; 1H NMRδ 2.40
`that was filtered and washed with cold EtOH: melting point, 179.8
`(s, 3H), 2.95 (s, 3H), 4.16 (s, 2H), 7.44 (s, 4H). Anal. calculated for C5H13N3O5S: C,
`26.43; H, 5.77; N, 18.49; found: C, 26.60; H, 5.70; N, 18.39.
`
`Creatine Hydrogen Maleate
`
`To a stirred solution of maleic acid (0.04 mol, 10.40 g) in EtOH was added a
`◦
`suspension of creatine monohydrate (0.04 mol, 5.98 g) in water (80 ml) at 59
`C. The
`◦
`C before cooling to room temperature. The reaction
`solution was stirred for 10 min at 59
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`003
`
`Harvest Trading Group - Ex. 1109
`
`

`
`Gufford et al.
`
`243
`
`◦
`C for 24 hr after which a creatine monohydrate precipitate
`solution was then cooled to 5
`that got formed was collected using suction filtration. Cold acetonitrile (300 ml) was
`◦
`added to the filtrate and cooled to 0
`C to precipitate additional creatine monohydrate
`that was then removed using suction filtration. This process was repeated thrice until no
`additional precipitate was observed. After removing the solvents in vacuo, crude creatine
`hydrogen maleate (6.18 g, 59%) was filtered and washed with cold acetonitrile. Three
`crystallizations from 80% aqueous EtOH afforded analytically pure creatine hydrogen
`◦
`C; 1H NMRδ 2.95 (s, 3H), 4.14 (s, 2H), 6.03 (s, 2H), 7.33
`maleate: melting point 157.8
`(s, 4H). Anal. calculated for C8H13N3O6: C, 38.87; H, 5.21; N, 17.00; found: C, 38.71;
`H, 5.49; N, 16.85.
`
`HPLC Analysis of Creatine Salts
`
`The HPLC system consisted of a Shimadzu SCL-10A controller, an SIL-10AF au-
`tosampler, dual LC-10AT pumps, an SPD-10A UV-VIS detector set to monitor ab-
`◦
`C, with a
`sorbance at both 210 and 235 nm, and a CTO-10AS column oven set to 30
`T3 column (4.6 × 100 mm, 3 µm, C18). The isocratic mobile phase
`Waters Atlantis R(cid:1)
`consisted of 20% v/v acetonitrile, 5-mM formic acid, and 5-mM 1-octanesulfonic acid
`sodium salt (apparent pH 2.8) at a flow rate of 1.5 ml/min. Simultaneous quantitative
`determination of creatine and creatinine content in test solutions was based on the
`UV absorbance at 210 nm (creatine and creatinine) and 235 nm (creatinine only). The
`UV absorbance of creatinine at 235 nm allows for confirmation of creatinine concen-
`tration at both wavelengths. Calibration curves were generated using stock solutions
`(500 µg/ml) of creatine monohydrate and creatinine diluted in the mobile phase to
`concentrations of 3, 10, 30, and 50 µg/ml. For lower concentrations, calibration curves
`were prepared in mobile phase at concentrations of 0.1, 0.3, 1, and 3 µg/ml. Linear
`regressions gave excellent agreement between concentrations and detector response
`within the experimental concentration ranges (R2 = 0.999).
`
`Saturation Solubility Determination
`
`Using preliminary measurements as a guide, the saturated solubility of each creatine
`salt in deionized water was determined in triplicate by adding increasing amounts to 5
`◦
`C. After
`ml of solvent in screw-capped glass bottles placed in a shaking water bath at 25
`1.5 hr, the saturated solutions were vortexed and 2-ml aliquots removed and centrifuged
`in microcentrifuge tubes at 11,000 rpm for 5 min. Creatine concentrations were analyzed
`by HPLC by diluting 500 µL of supernatant with mobile phase (500 µL). The mean
`± standard deviation of the saturation solubility of each salt were calculated from the
`corresponding standard curves. Extended equilibration times for saturation were per-
`formed with no observable change in creatine concentration. Extending the equilibration
`time of creatine in these acidic saturated solutions did result in cyclization of creatine
`to creatinine as expected. Using the equilibration times outlined above, no significant
`creatinine concentration was observed in any of the saturation solubility experiments.
`
`Determination of the pH-Dependent Saturation Solubility of Creatine Monohydrate
`◦
`C in
`The pH-dependent solubility of creatine monohydrate was determined at 25
`triplicate. Excess creatine monohydrate was added to 20-ml glass vials containing
`10 ml of deionized water. The saturation solubilities of creatine monohydrate at pH
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`004
`
`Harvest Trading Group - Ex. 1109
`
`

`
`244
`
`JOURNAL OF DIETARY SUPPLEMENTS
`
`1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.4, and 8.5 were measured; pH was adjusted by adding
`0.1-M HCl or 0.1-M NaOH. The suspensions were equilibrated for 2 hr in a shaking
`◦
`water bath at 25
`C and the final pH recorded. Aliquots were then removed and after
`centrifuging at 3,900 rpm for 10 min, the supernatant was collected, filtered through
`a 0.22-µm hydrophilic AladnTM syringe filter, and diluted in mobile phase and the
`creatine monohydrate concentration determined by HPLC analysis.
`
`Thermal Analysis
`
`Melting points of each creatine salt were obtained by differential scanning calorimetry
`(DSC) using a Shimadzu model DSC-50 with a TA-50WS controller. Samples (∼5 mg)
`◦
`C at a rate of
`were placed in flat aluminum pans, crimped, and heated from 30 to 330
`◦
`10
`C/min in an atmosphere of nitrogen (20 ml/min). To ensure that the true saturation
`solubility, rather than the apparent solubility, i.e., of free creatine, was being measured
`at each pH (Pudipeddi, Serajuddin, Grant, & Stahl, 2002), 3 hr after equilibration,
`1-ml aliquot was removed, lyophilized for 72 hr, and the thermogram of the remaining
`compound obtained and compared to that of the original salt.
`
`Water–Octanol Partition Coefficient Determination
`
`To assay the concentrations of each salt in both the aqueous and octanol phase after
`equilibrium, the creatine salt in the octanol phase was extracted in water as described
`below. Standard solutions of creatine monohydrate in mobile phase were prepared in
`the range of 0.1 to 50 µg/ml by serial dilution of a 500-µg/ml stock solution. Two series
`of standards (0.1, 0.3, 1.0, 3.0 µg/ml and 3.0, 10, 30, 50 µg/ml) were prepared. For the
`partitioning studies, 25 ml of each creatine salt was prepared in octanol–saturated water
`at a nominal creatine concentration of 6.0 mg/ml. Due to very low partition coefficients,
`this relatively high concentration was used to ensure that each salt could be quantified
`in the octanol phase. Triplicate aliquots (5 ml) of each creatine monohydrate or creatine
`salt solution were placed in glass scintillation vials and then an equal volume (5 ml)
`of water-saturated octanol was added. Triplicate 5-ml samples of aqueous solutions
`of each creatine derivative served as controls to test for decomposition. The samples
`◦
`C water bath for 2 hr and then the
`were allowed to equilibrate in a slowly shaking 25
`aqueous phases were assayed after appropriate dilutions with mobile phase. To assay
`the concentration of the creatine salt in the octanol phase, 4 ml of the water-saturated
`octanol phase was removed and placed in scintillation vials and 4 ml of water was added
`to extract the water-soluble salt. The samples were equilibrated for an additional 2 hr
`◦
`C to extract the creatine from the octanol into the
`in a slowly shaking water bath at 25
`aqueous phase. The concentration of each derivative so extracted was then quantified
`after diluting 500-µL aliquots in 500-µL mobile phase. Due to the very small amount of
`creatine partitioning into the octanol phase, larger quantities of creatine (100, 200, 300,
`400 µL of a 10-µg/mL stock solution) were also employed to more accurately assay the
`aqueous phase. The mean of the partition coefficient for each creatine compound was
`calculated and recorded as a log P value.
`
`Caco-2 Monolayer Permeability
`
`The permeability of various creatine salt forms was evaluated using the established
`human intestinal epithelial cell line derived from a human colon carcinoma (Caco-
`2). Although derived from colon epithelial cells, the Caco-2 cell line has many of
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`005
`
`Harvest Trading Group - Ex. 1109
`
`

`
`Gufford et al.
`
`245
`
`the characteristics of absorptive intestinal enterocytes found in small intestine and
`consequently has been used extensively for examining intestinal absorption of drugs
`and nutrients (Batrakova, Li, Miller, & Kabanov, 1999; Chabane, Al Ahmad, Peluso,
`Muller, & Ubeaud, 2009; Tian, Yan, Yang, & Wang, 2009). The Caco-2 cells were seeded
`onto collagen-coated Transwell polycarbonate membrane inserts (12-mm diameter;
`1-µm pore size) (Fisher Scientific, Winnipeg, MB) at a density of 60,000 cells/cm2. The
`cells were grown in DMEM supplemented with 10% fetal bovine serum and maintained
`in 5% CO2 environment. Cells were cultured for a period of 18–21 days, after which
`confluent monolayers were used for permeability studies.
`For the permeability studies, the culture media was removed from the cells and
`replaced with Tyrode’s buffer consisting of 136-mM NaCl, 2.6-mM KCl, 1.8-mM
`CaCl2, 1-mM MgCl2, 0.36-mM NaH2PO4, 5.56-mM D-Glucose, and 5-mM HEPES
`and pH 7.4. Following a 30-min pre-incubation period, Tyrode’s buffer was removed
`from the apical compartment and replaced with 0.5 ml of Tyrode’s buffer containing
`creatine monohydrate, creatine hydrochloride, creatine pyruvate, and creatine citrate
`(10 mM). Samples (50 µL) were removed from the apical (donor) compartment at the
`start and conclusion of the permeability experiment. Samples (100 µL) were removed
`from the receiver compartment at 0, 15, 30, 60, and 90 min. Passage of creatine from the
`donor to the receiver compartments was analyzed using HPLC methods described above.
`◦
`C. Permeability coefficients
`All permeability studies were performed at pH 7.4 and 37
`were determined based on the following equation:
`Papp= dCr/dt(Vd/(A
`where Cr is creatine concentration in the receiver compartment, t is time, Vd is the
`volume in the donor compartment, A is area, and Cd is the concentration in the donor
`compartment at time zero.
`
`Cd)),
`
`∗
`
`RESULTS AND DISCUSSION
`
`Creatine Salt Synthesis
`
`Zwitterionic creatine (pKa 2.79 and 12.1) (Prankerd, 2007) has the potential to form
`N-methylguanidinium salts directly only with acids that are more acidic than its car-
`boxylic acid functional group. Examples of such salts are the hydrochloride, mesylate,
`and hydrogen maleate that were successfully synthesized from hydrochloric acid (pKa
`−6), methane sulfonic acid (pKa −1.2), and maleic acid (pKa1 1.92) (Table 1). How-
`ever, using acids that are more acidic than the acidic carboxyl group of creatine does
`not necessarily guarantee that a salt will be formed as illustrated by our numerous
`attempts to synthesize creatine dihydrogen phosphate by mixing creatine monohydrate
`with phosphoric acid (pKa1 2.15) at various temperatures and cosolvent mixtures with
`only creatine monohydrate being recovered after each attempt. By monitoring freezing
`point depression, Dhar and Ghosh (1961) reported that a stable complex, but not nec-
`essarily a salt, is formed from mixtures of creatine and phosphoric acid. It is important
`to note that creatine citrate salts cannot be synthesized by simply mixing solutions of
`creatine and the insufficiently acidic citric acid (pKa 3.1, 4.8, 6.4). Indeed, using proton
`NMR, no chemical shifts were observed in the spectrum of the “creatine citrus” used in
`this study; the spectrum was simply a superposition of the spectra of creatine monohy-
`drate and citric acid confirming that this product was a physical mixture of creatine and
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`006
`
`Harvest Trading Group - Ex. 1109
`
`

`
`−3.5
`297.1
`8.6
`
`–
`
`1.0
`
`17.1±0.4
`195.0
`87.9
`149.1
`
`−3.2
`141.2
`3.1
`
`3.1/4.8/6.4
`
`3.0
`
`44.8±2.4
`
`194.0
`58.1
`451.1
`
`(Degussa)
`
`Nutrition)
`
`Monohydrate
`
`“Citrus”(Peak
`
`−3.3
`123.5
`2.4
`2.4
`
`4.8
`
`91.6±7.7
`
`191.3
`59.8
`219.2
`
`Pyruvate
`
`−3.8
`157.8
`2.2
`1.9
`
`2.2
`
`37.4±7.3
`201.0
`50.4
`260.2
`
`Maleate
`Hydrogen
`
`CreatineSalt
`
`−3.3
`179.8
`1.1
`−1.2
`29.6
`
`588±8
`191.0
`54.6
`240.3
`
`−3.2
`166.1
`0.3
`−6
`37.9
`
`709±7
`196.0
`78.2
`167.6
`
`−3.5
`255.0
`2.0
`−9.0
`7.1
`
`121±1
`195.0
`72.8
`180.2
`
`25
`Octanol–WaterLogPat
`MeltingPoint(◦C)
`pHatSaturation
`pKaofAcid
`CreatineMonohydrate
`RatioAqSolubilityw.r.t.
`(mg/ml)
`AqueousSolubilityat25◦C
`λmax(nm)
`WeightCreatine(%)
`MW(g/mole)
`
`C
`
`◦
`
`Mesylate
`
`Hydrochloride
`
`Hemisulfate
`
`Property
`
`TABLE1.Physicochemicalpropertiesofcreatinesalts
`
`246
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`007
`
`Harvest Trading Group - Ex. 1109
`
`

`
`Gufford et al.
`
`247
`
`citric acid in a 2:1 molar ratio, and not a true salt. As a further example, we attempted
`to synthesize creatine lipoate (Buononato & Festuccia, 2003) by mixing a 1:1 molar
`ratio of creatine monohydrate and lipoic acid (pKa 5.4) in ethanol at room temperature
`for 12 hr, but recovered only the two starting materials. In summary, without using an
`ion-exchange process, it is not possible to form creatine salts with acids that are less
`acidic than creatine itself.
`
`Physiochemical Properties
`
`Table 1 lists the properties of the salts in order of the pKa values (Prankerd, 2007)
`of the conjugate acid. Specifically, these properties are molecular weight, percentage
`◦
`C, ratio
`weight of creatine in each salt, UV absorption maxima, aqueous solubility at 25
`of the aqueous solubility compared to creatine monohydrate, saturated solution pH, mp,
`and logarithm of the octanol:water partition coefficient.
`
`Aqueous Solubility
`
`Each creatine salt (or complex in the form of “creatine citrus”) was significantly more
`soluble than creatine monohydrate. A contributing factor could be that the pH values of
`all the saturated salt solutions in unbuffered water were strongly acidic in the range of
`0.3 to 3.1, whereas the pH of the saturated creatine monohydrate solution was 8.6. At
`neutral pH, or in biological fluid, creatine exists in the least soluble, electrically neutral,
`zwitterionic form because it has an isoelectric point of 7.4. The hydrochloride and
`mesylate had notably high water solubilities; the former was nearly 40-fold more soluble
`than creatine monohydrate. Although the lack of a common ion effect for mesylate salts
`is a potential advantage over hydrochloride salts, the former have the potential liability
`of contamination with residual genotoxic alkyl mesylate ester impurities (Snodin, 2006).
`The creatine salts formed from carboxylic acids were less soluble than the creatine salts
`formed from methanesulfonic acid and the two mineral acids, although the hemisulfate
`was only 7.1-fold more soluble than creatine monohydrate.
`
`Thermal Analysis
`
`Melting points of creatine monohydrate and salts are listed in Table 1. In addition,
`thermograms of creatine monohydrate and salts are shown in Figure 2. Thermograms
`of creatine monohydrate and salts are also shown after lyophization of the saturation
`solubility tests to confirm that what was measured in solution was the salt and not molec-
`ular creatine or creatinine. With the exception of creatine monohydrate, no significant
`change in the DSC thermograms was observed for the creatine salts before and after
`lyophilization of the saturated solution. This indicates that the solubility measurements
`are true representations of the salt saturation solubilities and not just the solubility of
`free creatine at that particular pH, described as the apparent solubility (Pudipeddi et al.,
`2002). The expected difference observed in creatine monohydrate is the result of the
`◦
`C
`water loss during lyophilization, as indicated by the absence of the peak near 100
`(Dash, Mo, & Pyne, 2002). HPLC analysis further confirms these conclusions as neither
`the 210-nm nor 235-nm trace shows significant creatinine content in any of the measured
`test solutions.
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`008
`
`Harvest Trading Group - Ex. 1109
`
`

`
`248
`
`JOURNAL OF DIETARY SUPPLEMENTS
`
`FIGURE 2. DSC thermograms of (a) creatine monohydrate; (b) creatine mesylate; (c) cre-
`atine hydrogen maleate; (d) creatine hydrochloride; (e) creatine hemisulfate; (f) creatine
`pyruvate; and (g) “creatine citrus.” Solid lines indicate thermal profiles of original samples
`and dashed lines represent thermal profiles of saturated solutions post-lyophilization.
`
`Octanol–water Partition Coefficients
`
`As expected, the octanol–water partition coefficients for the creatine salts were very
`low and essentially equivalent reflecting their high polarity. Due to dissociation of
`salts in aqueous medium, the measured partition coefficients reflect the free creatine
`concentration, and hence, yield similar values (log P – 3.8 to – 3.2) for each creatine
`salt or zwitterionic monohydrate. Although the aqueous solubilities of the creatine salts
`and monohydrate varied considerably, these very low log P values indicate that all had
`very low affinity for the lipophilic octanol phase. It should be noted that because of
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`009
`
`Harvest Trading Group - Ex. 1109
`
`

`
`Gufford et al.
`
`249
`
`the highly hydrophilic nature of creatine, it was necessary to use the extraction method
`to measure the creatine concentrations in octanol phases rather than estimating octanol
`concentrations by differences in the water and control samples.
`
`pH-Dependent Solubility of Creatine Monohydrate
`
`Figure 3 illustrates the solubility of creatine monohydrate from pH 1.0 to 8.5. The
`solubility of creatine monohydrate at pH 1.0 and 4.0 was 52.0 ± 3.3 mg/ml and 15.7 ±
`0.3 mg/ml, respectively. At pH 1.0, the carboxylic acid functional group of creatine is
`largely unionized and creatine exists predominately as the monocation, whereas at pH
`4.0, the carboxylic acid group is largely ionized and the zwitterionic form dominates.
`In the pH range of 4.0 to 9.0, where the zwitterionic form also predominates, the
`solubility remained essentially constant. The pH-dependent saturation solubility profile
`of creatine monohydrate follows expected results with the lowest measured solubility
`being at pH 7.4, the isoelectric point of the creatine zwitterion. The creatine salts would
`be expected to yield comparable profiles when the same acid or base is used for pH
`
`FIGURE 3. pH-dependent solubility of creatine monohydrate.
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`010
`
`Harvest Trading Group - Ex. 1109
`
`

`
`250
`
`JOURNAL OF DIETARY SUPPLEMENTS
`
`adjustment (Pudipeddi et al., 2002). The pH-dependent solubilities of the creatine salts
`were not determined, as all saturated aqueous solutions are strongly acidic and buffer
`concentrations required to maintain a constant pH would make the data meaningless.
`
`Caco-2 Monolayer Permeability
`
`The permeability of several creatine salts was determined in confluent Caco-2 mono-
`layers (Figure 4). For each salt form examined, the flux of creatine across Caco-2
`monolayers was linear over the entire permeability study. While there was a tendency
`for creatine citrate to have lower permeability in Caco-2 monolayers, there were no sig-
`nificant differences in the permeability coefficients for any of the examined creatine salt
`forms (Figure 4). The fact that permeability of all the creatine salt forms was relatively
`low is certainly consistent with the octanol–water partitioning studies. The results of the
`present study are also in agreement with the relatively low permeability of radiolabeled
`creatine monohydrate in Caco-2 monolayers reported previously by Dash, Miller, Han,
`Carnazzo, and Stout (2001). Furthermore, as the permeability studies were performed at
`
`FIGURE 4. Caco-2 monolayer permeability. Creatine hydrochloride (CHCl), creatine mono-
`hydrate (CM), “creatine citrus” (CCit), and creatine pyruvate (CPyr).
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`011
`
`Harvest Trading Group - Ex. 1109
`
`

`
`Gufford et al.
`
`251
`
`a concentration where all the salt forms examined were soluble, significant differences
`in permeability between the various salt forms were not anticipated.
`Permeability in Caco-2 monolayers has been used to predict oral absorption of a
`variety of dietary supplements. For example, Mandagere, Thompson, and Hwang (2002)
`have established and validated a model using Caco-2 monolayer permeability and in
`vitro metabolic stability to predict the oral absorption of drugs. Based on this model,
`the creatine permeability observed in Caco-2 monolayers in the present study (i.e., Papp
`of 10–15 × 10
`−6 cm/s) would predict a 20–50% oral absorption of creatine. Given the
`relatively large doses of creatine dietary supplements that are required for biological
`effects (i.e., 5–20 gm daily dose), identification of creatine salts with improved aqueous
`solubility would be anticipated to improve oral absorption.
`
`CONCLUSIONS
`
`Two new salts, specifically creatine mesylate and creatine hydrogen maleate, were
`synthesized. The physicochemical properties of six creatine salts were determined and
`compared with creatine monohydrate. All salts were significantly more water-soluble
`than the monohydrate and all had very low partition coefficients reflecting their highly
`hydrophilic nature. Saturation solubility was dependent on both pH of the solution at
`saturation and the specific salt derivative. Caco-2 permeability studies further confirm
`the hydrophilic nature of these compounds and with the exception of creatine citrate,
`no significant differences in permeability were observed between the salts and creatine
`monohydrate. Although all salts are similar in permeability when studied at soluble
`concentrations, the large doses of less soluble, commercially available creatine mono-
`hydrate compounds typically used for human consumption result in administration of
`suspensions expected to have reduced overall absorption. Potential advantages of crea-
`tine salts over creatine monohydrate include: enhanced aqueous solubility resulting in
`increased bioavailability, reduced adverse effects, additional formulation options, and
`wider range of therapeutic applications. Increasing water solubility would facilitate the
`administration of creatine salts as a solution potentially creating less GI disturbance and
`improved bioavailability. The enhanced water solubility makes the preparation of com-
`mercial oral and topical dosage forms more attractive than comparable dosage forms
`using creatine monohydrate. It is anticipated that tablet formulations of creatine salts
`would improve dissolution kinetics and oral absorption compared to creatine monohy-
`drate. In addition, the differences in crystalline structure between individual salts may
`allow for direct compression into a tablet. The potential for alternative formulations of
`creatine salts may also allow for a wider range of therapeutic applications e.g., topical
`anti-inflammatory. It is acknowledged that more complete in vivo pharmacokinetic anal-
`ysis of these compounds would be required to fully elucidate their potential advantages
`and disadvantages over currently available creatine supplement formulations.
`
`REFERENCES
`Arnold MJ. Preparation of stable ketals of pyruvic acid and its acid derivatives which are useful as
`dietary supplements. US Patent No. 6177576 B1, 2001.
`Batrakova EV, Li S, Miller DW, Kabanov AV. Pluronic P85 increases permeability of a broad spectrum
`of drugs in polarized BBMEC and Caco-2 cell monolayers. Pharm Res. 1999;16:1366–1372.
`Blatt T, Breitenbach U, Mummert C, Raschke T, Scherner C, Schulz J, Staeb F. Active ingredient
`combinations of glucosyl glycerides and creatine and/or creatinine. Patent No. WO 2006122668A1,
`Germany: Beiersdorf A.-G., 2006.
`
`Downloaded by [Reprints Desk Inc] at 10:50 11 November 2015
`
`012
`
`Harvest Trading Group - Ex. 1109
`
`

`
`252
`
`JOURNAL OF DIETARY SUPPLEMENTS
`
`Boldt M. Preparation of creatine salts of dicarboxylic acids. US Patent 2004133040A1, San Corpo-
`ration, 2004.
`Buononato A, Festuccia A. Process for the preparation of creatine lipoate salt. Patent No. WO
`2003-IT316 20030523, Italy: Licrea S.R.L., 2003.
`Chabane NM, Al Ahmad A, Peluso J, Muller CD, Ubeaud G. Quercetin and naringenin transport
`across human intestinal Caco-2 cells. J Pharm Pharmacol. 2009;61:1473–1483.
`Cornelius DW, Haynes GL. Dicreatine malate with increased bioavailability. US Patent No.
`2004220263A1, Creative Compounds, LLC, 2004.
`Dash AK, Miller DW, Han H-Y, Carnazzo J, Stout JR. Evaluation of creatine transport using Caco-2
`monolayers as an in vitro model for intestinal absorption. J Pharm Sci. 2001;90:1593–1598.
`Dash AK, Mo Y, Pyne A. Solid-state properties of creatine monohydrate. J Pharm Sci.
`2002;91:708–718.
`Dash AK, Sawhney A. A simple LC method with UV detection for the analysis of creatine and
`creatinine and its application to several creatine formulations. Pharmaceutical Biomed Anal.
`2002;29:939–945.
`Dhar NR, Ghosh GP. Complex compounds of acid, base, and salts with nitrogenous and other organic
`substances. Proc Natl Acad Sci India 1961;31A:74–77.
`Gardiner PT. Food supplements and methods comprising lipoic acid and creatine. US Patent No.
`6,136,339, 2000.
`Gardiner PT, Heuer MA, Molino M. Creatine hydroxycitric acid salts and methods for their production
`and use as food supplements e