Field of the Invention

The invention relates to the method of determination of hyaluronic acid in complex samples by cleavage with SpnHyl enzyme followed by the reaction with 3-methyl-2- benzothiazolinonhydrazone (MBTH) to form a coloured product analysed by

spectrophotometry. In case of high content of glycosaminoglycans, these can be separated by precipitation before the cleavage, e.g. with use of cetyltrimethylammoniumbromide (CTAB) in acetate buffer and salt medium. In case of high content of reducing sugars, e.g. glucose, their influence can be further eliminated, after removing the glycosaminoglycans and before the cleavage, by precipitating hyaluronic acid from the formed supernatant and then redissolving this sediment in acetate buffer and performing the cleavage step.

Background of the invention Hyaluronic acid (hyaluronan, hyaluronate, HA) is a linear glycosaminoglycan consisting of alternating molecules of glucuronic acid and N-acetylglucosamine. It is one of the basic building materials of connective tissue thanks to its exceptional rheological and mechanical properties. Thus, hyaluronic acid is widely used in pharmaceutical and cosmetics industry (Shimmura et al. 1995; Kogan et al. 2007); its production has been constantly increasing. At the same time, the concentration of hyaluronic acid in blood serves as a biological marker of physiological state of joints and the whole moving apparatus. Depending on the needs, many methods of determination of hyaluronic acid concentration have been developed, differing in both their sensitivity and basic principles.

According to their sensitivity the methods of determination of HA can be divided into nanogram, microgram, and milligram scales. According to the principle, the methods of determination of HA can be divided into the following categories:

1) Gravimetric

2) Turbidimetric

3) Spectrophotometric

4) Chromatographic 5) Immunoprecipitation (ELISA)

Gravimetric methods belong to the oldest and the least accurate methods, where HA is precipitated with alcohol, the precipitate is then dried and weighed. The processes of precipitation and drying are slow, which takes up to two days. The sensitivity of the method is low, usually hundreds of milligrams per litre, moreover, other polysaccharides can precipitate concurrently during the precipitation. Gravimetry is conventionally used in biotechnological production of HA, where the composition of the culture medium is known and constant, other polysaccharides are absent, and the analysis time does not play an important role. Turbidimetric methods exploit the properties of the complexes of glycosaminoglycans and aliphatic ammonium salts comprising at least one long aliphatic chain (cetylpyridinium chloride - CPC, or cetyltrimetylammonium bromide - CTAB). Under certain conditions these complexes form very stable colloid solutions. The content of HA in a sample can be determined according to the level of turbidity of the colloid solution of "HA - aliphatic ammonium salt" (Oueslati et al., 2014). This method has a great advantage of a very short time of analysis. Its main disadvantages are low sensitivity (tens milligrams per litre) and very low specificity due to the simultaneous precipitation of other negatively charged glycosaminoglycans and, at high pH, even non-ionic polysaccharides such as starch (Scott, 1960). Turbidimetry progressively replaces gravimetry in biotechnological production of HA, where the demands for the specificity are lower.

Nowadays, the spectroscopy methods are used the most often, thanks to their simplicity, rapidity, and sensitivity (several micrograms per liter). The first colorimetric methods for determination of HA were based on the methods for detection of specific monosaccharides, such as N-acetylglucosamin in Elson-Morgan method with the Reissig's modification (Elson&Morgan, 1933; Reissig et al., 1955), and D-glucuronic acid in Dische method

(Dische, 1947). HA, being the polysaccharide, must be depolymerized into monosaccharides prior the colour reaction. Depolymerization is performed by 80% sulfuric acid at 60 °C (Dische, 1947), or 2N-hydrochloric acid at 100 °C (Reissig et al., 1955). In some

modifications of the Elson-Morgan method, the acid hydrolysis was replaced by enzymatic hydrolysis by hyaluronidase from Streptomyces hyalurolyticus (Lee, 1966). However, this lead to prolongation of the analysis, as the enzymatic degradation itself proceeded 20 hours. In both cases, the depolymerisation of HA resulted in a mixture of products: oligosaccharides of various lengths in case of the enzymatic hydrolysis, and side products of the degradation in case of the acid hydrolysis. Subsequently, various non-specific colour products are formed in the colour reaction, which leads to the variable extinction coefficient of Elson-Morgan reaction of 18000-21000 mol ^"1 (Reissig et al, 1955). In case of Diche method, the accuracy is even lower, as the carbazole reaction was originally developed for determination of D- galacturonic acid resulting in the coloured product with the extinction coefficient of 11200 mol ^"1 , whereas the extinction coefficient of the coloured product of D-glucuronic acid is 1750 mol ^"1 only. Furthermore, glucose, fructose, and sucrose are able to react with the carbazole, and the reaction itself is sensitive to the ionic strength of the solution. Elson-Morgan reaction is sensitive to the presence of protein in the samples which causes the increased turbidity of the samples (Greilig, 1965). Both methods are time-consuming and unsafe due to the use of hot concentrated mineral acids. The necessity of the precise timing in each step significantly limits the flexibility of both methods.

Thiobarbiturate method (Jourdian et al., 1979) is worth mentioning, too. In this method, HA is cleaved with a lyase from S. hyalurolyticus into 4, 5 -unsaturated oligosaccharides. These are subsequently oxidized with periodic acid to formylpyruvate that reacts with

thiobarbiturate. The sensitivity of this method is comparable with the Elson-Morgan method, but it is more user friendly. The drawbacks of the method are "slow" hyaluronan lyase and the use of highly toxic arsenic compounds. Chromatographic methods use a direct detection of HA at 201 nm, or at 232 nm in case of unsaturated HA forms (Gassier et al., 1993; Alkrad et al., 2002; Ruckmani et al., 2013). In both cases HA can be determined in pure solutions only, whereas the sensitivity is slightly lower than in colorimetric methods. On the contrary, the chromatography has the advantage of determination of molecular weight of HA. That is the main reason why the

chromatographic methods are commonly used for determination of the molecular weight. There is also the disadvantage of high running costs and costs of equipment.

The progress in the knowledge about the role of HA in the metabolism leads to the demand of determination of HA in nanogram scale, as the common physiological HA concentrations in blood are within the range of tens of nanograms per litre. These demands were fulfilled with immunoprecipitation techniques (enzyme-linked immunosorbent assay -ELISA) (Haserodt et al., 2011). Currently, ELISA is the most sensitive and the most accurate method. Unfortunately, the extremely high price of analysis and the necessity of highly professional operators limit the use of this perspective method.

The preliminary analysis of the existing methods of determination of HA shows that each of them has serious limitations in sensitivity, range of detection, time, price, and safety of analysis, and so far, there is no versatile method that would maximally integrate these parameters in a satisfactory way.

Summary of the Invention

Spectrophotometry is considered to be a very accurate and sensitive tool. At the same time, spectrophotometry is relatively cheap and simple and thus is widely used in laboratories. In order to keep the affordable price while maintaining the high sensitivity, the spectrophotometric determination of HA must be based on the reaction forming an intensively coloured product that absorbs at longer wave lengths. There are many agents reacting with saccharides to form coloured products. Some of them have been used for determination of other polysaccharides, such as metyl-2-benzothiazolinonhydrazon (MBTH) which is characterized by a low interaction with salts, proteins, and nucleic acids (Moretti&Thorson, 2002).

MBTH has been used primarily for the determination of aldehydes (Hauser&Cummins, 1964). The analysed aldehyde condenses with MBTH, and then follows the conversion into formazan upon reaction with the diazonium salt which is generated in the reaction medium by oxidation of the excess reagent. (Scheme 1). That forms highly a coloured product with the highest absorbance at620 nm (blue colour).

Scheme 1. The scheme of reaction between MBTH and acetaldehyde

The reducing sugars in the open chain form have an active aldehyde group, and thus they can react with MBTH (Honda et al., 1981). Only a very small fraction of the open chain form exists in solution; thus alkali media and high temperatures are necessary for a quantitative condensation of MBTH with saccharides. The resulting product of the condensation is yellow and Amax = 390 nm. According to G. E. Anthon and D. M. Barrett, MBTH-saccharide compound undergoes a further oxidative adduction of the second molecule of MBTH, which leads to an intensively blue coloured product 620 nm). The sensitivity of this reaction for glucose is 2.5 μΜ of reducing ends and the range is from 2.5 to 400 μΜ.

HA of various molecular weight (24, 73, 211, 400, 549, 875, 1690, and 2049 kDa) were subjected to the reaction with MBTH according to the original Anthon-Barrett method protocol (Anthon & Barrett, 2002). This proved the possibility of the use of MBTH for the detection of HA. However, HA, as a polymer, comprises very few reducing ends making it impossible to directly determine HA due to its weak signal, especially in the case of high molecular weight HA (Fig. 1).

Furthermore, HA in alkali media, as well as the other glycosaminglycanes, undergoes the so-called peeling reaction and β-elimination reactions (Kiss, 1974; Whistler and BeMiller, 1958; Yang and Montgomery, 2001). This leads to an increase of absorbance in time during heating (Fig. 2). This effect is more pronounced at higher temperatures. This demands very precise timing of the reaction, which is highly inconvenient and decreases the practical applicability of the assay.

Controlled degradation of HA leading to the formation of only one low molecular product solves simultaneously three problems. First, an increase of the concentration of the reducing ends amplifies the analytical signal strength and thus the sensitivity of the analysis. Second, the subsequent formation of only one coloured product enhances the repeatability and robustness of the measurement. Third, the risk of the peeling reaction is eliminated, which also contributes to the robustness of the method.

An acid hydrolysis is not suitable for a controlled degradation of HA. Besides hydrolysis, HA also undergoes racemization and random degradations in acid media (Reed and Reed, 1989; Lapcik et al, 1998), so the products of the acid hydrolysis are not identical with the structural components of the native HA. This results in different extinction coefficients of the coloured products and inaccuracy of the analysis. This probably explains why the extinction coefficient of the Elson-Morgan reaction is not constant.

First experiments with an enzymatic degradation of HA were not promising. The commonly used hyaluronidase from S. hyalolyticus had a low activity and formed a mixture of tetra- and hexasaccharides, so there were essentially the same problem as in case of the acid hydrolysis (Lee, 1996). Later experiments with other enzymes cleaving HA were also limited by an insufficient activity and variability of the products of digestion. Processive lyase SpnHyl from Streptococcus pneumoniae belongs to the group of phage-derived hyaluronanlyases (Baker et al., 2002). In contrast to other lyases, it comprises two domains and cleaves HA not only randomly but also in a processive way. It has an outstandingly high activity and forms 4, 5 -unsaturated disaccharide (ΔΗΑ2) as the only product of the reaction (Fig. 3).

Wide testing of SpnHyl activity revealed the highest specific activity of the enzyme of

1,000 U/mg being demonstrated at 37°C in 50mM acetate buffer at pH 6. The enzyme is dependent on the ionic strength of the solution; in distilled water its activity lowers 15times. However, most of the activity restores in lOmM acetate buffer. The addition of S0 ₄ ^2" a Ca ²⁺ ions enhances the SpnHyl activity by approximately 20 %. A high concentration of the salt does not influence the activity significantly; the enzyme remained very active even in 1 M NaCl. 4 U of SpnHyl (1 L of the solution with the activity of 2,000 U/mg and concentration of 2 mg/mL) were sufficient for the full digestion of 200 L of 0.2 % (w/v) HA. Under these conditions HA completely degraded in less than 5 minutes. The product of the first condensation of ΔΗΑ2 and MBTH is a yellow intermediate having the maximum absorbance at 390 nm. According to the literature (Honda et al., 1981), this yellow product can be used for the detection without the necessity of a further reaction. To verify this option, different amounts of HA cleaved in 200 were mixed with 200 μΐ _^ of 1M NaOH and 600 μΐ of 0.1 % (w/v) MBTH. Then the reaction mixture was heated at different temperatures from 60 to 95 °C. The samples were then cooled to room temperature and measured at 390 nm. It was found that the dependence of the absorbance on the concentration is not linear, the intensity of colouring is not stable during the time, and the sensitivity of the reaction is very low (extinction coefficient of 1100 mol ^"1 ). This experiment proved the necessity of the second reaction.

After adding an acid ferric solution into the reaction mixture after 5-10 minutes of cooling, development of blue colour was observed, with maximum absorbance at 620 nm and a small peak at 654 nm. Further experiments on repeatability showed, that the conditions of the first reaction are the most important for the method's robustness. Several combinations of conditions were tested with temperatures of 60, 80 and 95°C; times of 20, 30, 40, 50 and 60 minutes. The analysis of recorded spectra revealed, that the highest repeatability was achieved at 80°C and 30 minutes, when the absorbance was measured at 654 nm. Further investigation of "heating time"-"absorbance" dependence showed, that the absorbance reaches a plateau between 30 and 50 minutes during heating at 73-77°C (Fig. 5). Under these conditions, the measurement error was less than 1%. Based on these results, the following parameters were chosen for the analytical method: the first reaction is performed at 75°C, for not less than 30 and not more than 50 minutes. Such conditions offer allowance of ± 2°C and the sufficient time for the manipulation with the samples, which significantly improves the analysis repeatability and robustness.

During the experiments, the role of DTT (dithiotreitol) in the first reaction was revealed.

The presence of 10 mM DTT in MBTH solution greatly improves the stability of background samples, presumably by maintaining the reducing environment during the first colour reaction. The addition of 1% (w/v) sulfamic acid in the acidic iron solution prevents iron oxide precipitation in the second color reaction. Several optimizations in concentrations and volumes of sample and reagents were made for better manipulation and compatibility with laboratory equipment.

When the optical density of samples exceeded the capabilities of the spectrophotometer, the samples were diluted 2 to 10 times prior to measurement. The absorbance of the diluted samples however, deviated from Beer-Lambert-Bouguer law, i.e. the concentration- absorbance dependence was nonlinear. These deviations were caused by a pH shift due to the low buffering capacity of the acidic Fe ³⁺ solution (solution E, see Methods). The buffering capacity was improved by an addition of 0.2 M citric acid that allowed samples to be diluted to twenty times. Finally, the extinction coefficient was calculated for blue-colored ΔΗΑ2- MBTH product, which equaled 34,735 mol ^"1 . Direct comparison with the Elson-Morgan method (ε=18,000-21,000 mol ^"1 ) showed that new method is 1.5-2 times more sensitive.

The effect of the molecular weight of HA on the determination accuracy is quite small. The background absorbances of low-MW HAs were higher because of a higher count of reducing ends. However, the differences were of a very small magnitude. It was found that the variations were minor in all samples except of HA 11 kDa. Thus, it can be concluded that MW of HA does not influence the results of measurements unless MWs are lower than 50 kDa

The gathered data resulted in the first version of the SpHyl-MBTH assay:

- 0.2 mL of the sample containing HA is mixed with the same volume of lOOmM acetate buffer pH 6 with lOOmM of NaCl. The resulting 400 μΐ, is divided into two equal parts of 200 μΐ,. One is marked as the "sample" (S), the other as the

"background" (B).

- 4 U of SpnHyl (1 jiL of 2 mg/ml SpnHyl, 2000 U/mg) is added to the "sample" and the digestion occurs for 10 minutes at 37 °C.

- Then 200 μΐ, 1M NaOH and 400 μΐ. of the solution comprising 0.2% (w/v) MBTH and 1 OmM DTT is added into both the tubes S and B. The tubes are heated at 75°C for at least 30 minutes but not more than 50 minutes.

- After the heating, 500 μΐ, of the solution having the following composition is added to the warm reaction mixture:

50mM FeCl ₃ , 1M HC1, 1% (w/v) sulfamic acid, 0.2M citric acid.

- After 10-15 minutes, after the samples are cooled to the room temperature, the absorbance is measured at 654 nm. After subtracting the absorbance of the "background" from the absorbance of the "sample", the concentration of HA can be calculated based of the extinction coefficient.

All percentages mentioned herein are weight-volume percents, i.e. % w/v, if not indicated otherwise. For example the value 0.1 % MBTH means the concentration of

1 g/1 MBTH.

The hyaluronan-based medical remedies and nutritional supplements for joint care and support usually contain high amounts of other glycosaminoglycans along with HA, the most often it is chondroitin sulfate (CS) (Table 1). Glycosaminoglycans have similar physical and chemical properties and play a similar role within the organism, especially in connective tissues. Therefore, the next step of investigation was to determine conditions of the SpnHyl- MBTH method which would be suitable for HA detection in the presence of high amounts of CS.of SpnHyl-MBTH

Table 1. Composition of commercial preparations for supporting the joints

It was found out that SpnHyl has a weak nonspecific activity towards CS, 2000x lower than towards HA. The enzyme activity towards CS, as well as the lower molecular weight of CS comparing to HA (10-20 kDa and 1-2 MDa) and the generally higher concentration of CS in tissues and medical preparations (Table 1) lead to a significant effect on the HA detection. In solutions with the ratio of HA:CS from 1 :2 to 1 : 10, the measured concentration of HA was overestimated by 5-30%, wherein the rate of the overestimation was directly proportional to the concentration of CS. The negative influence of the CS presence was eliminated by selective fractional glycosaminoglycan precipitation with CTAB.Their solubility increases in solutions with high ionic strength. For a complete dissolution of complexes of CTAB and anionic polysaccharides, the higher is the negative charge of the polysaccharide molecule, the higher should be the ionic strength of the solution. HA, as carboxylate, has a much lower negative charge than polysulfated CS, allowing their separation by controlled precipitation by CTAB (Scott, 1960).

The inhibitory effect of CTAB on SpHyl was not observed and the enzyme was able to digest HA even in the presence of 2% CTAB. Due to the requirements of the subsequent enzymatic reaction, all precipitations were performed in 50 mM acetic buffer pH 6 instead of the recommended alkaline environment. Under these conditions the HA-CTAB complex precipitates at concentrations of NaCl lower than 370 mM, CS-CTAB remained insoluble even at 1.5 M NaCl. In the presence of NaCl the HA-CTAB and CS-CTAB complexes existed in the form of a very stable colloidal solution, which were not removable by centrifugation. According to the literature (Scott, 1960), monovalent ions stabilize such complexes, while counter-ions favor the precipitation. Among all tested electrolytes the best results were achieved with Na ₂ S0 ₄ used for ionic strength regulation. The HA-CTAB complex was soluble in 50 mM acetic buffer pH 6 at concentrations of Na ₂ S0 higher than 76 mM, whereas CS- CTAB dissolves at concentrations higher than 700 mM of Na ₂ S0 ₄ ..

The following conditions were optimal for selective HA and CS precipitation: 80 mM Na ₂ S0 ₄ , 50 mM acetic buffer pH 6. To prevent problems related to the local oversaturation, the precipitation occurred by mixing two solutions with the equal ionic strengths. Under such conditions only CS precipitated from the mixture of HA and CS. The sediment was easily removed by centrifugation at 12,100 g for 5 minutes. The supernatant underwent further analysis steps, as described above. The ratios of tested HA to CS concentrations varied from 1 :1 to 1 :20 with the concentrations of HA from 100 to 500 μg ml. The recovery of HA was 98- 102%For this reason, the analysed sample is premixed with the same volume of 160mM Na ₂ S0 ₄ in lOOmM acetate buffer of pH 6 (referred to as the "reagent A") - in this case 100 and 100 μΐ. The thus treated sample is then mixed with l%w/v CTAB in 80mM Na ₂ S0 ₄ , 50mM acetate buffer of pH 6 (referred to as the "reagent B"). The hindering low-molecular-weight impurities such as reducing sugars could be removed by fractioning the glycosaminoglycans with the CTAB precipitation. First, the CS is removed by precipitation as described above. Then, after the centrifugation, 400 μΐ of supernatant containing HA and reducing sugars was transferred into a pure microtube and treated with 25 μΐ of 1 M NaOH. This increases the negative charge of HA, which forms an insoluble complex with the excesses of CTAB. The sediment of HA-CTAB was subsequently washed several times with 500 μΐ of 10 mM NaOH, dried and then redissolved in 400 μΐ of reagent A. Its higher ionic strength and buffer capacity allow the CTAB-HA complex to be effectively dissolved and neutralize excesses of alkali. After the complete dissolution of the HA-CTAB complex the sample was divided into two parts - "sample" and "background" - and processed further according to the protocol. These precautions allowed HA to be quantified in the presence of reducing sugars up to the concentrations equivalent to 5% w/v glucose. Successful implementation of the steps intended for the removal of hindering impurities resulted in the final version of the MBTH-SpHyl method protocol. If the analysed sample contains low amounts of other glycosaminoglycans or reducing sugars, the simplified protocols I and II should be used (see the Methods).

Thus, the object of the invention is a method of determination of hyaluronic acid in a sample comprising the following steps:

a) the solution of a sample is prepared by mixing with acetate buffer containing monovalent salt so that pH of the final solution is within the range of 5 to 7, the concentration of acetate buffer is within the range of lOmM to 1M, and the ionic strength of the mixture corresponds to the solutions of monovalent salts with the concentration between 10 mM and 1M,

b) SpnHyl hyaluronan lyase from Streptococcus pneumoniae is added to the solution from the previous step, in the amount of at least 20 U per 1 mL of hyaluronic acid and the mixture is incubated at the temperature of 25 to 38 °C for 5 to 60 minutes, c) then a alkali and 0.1-l%w/v aqueous solution of 3-methyl-2- benzothiazolinonhydrazone (MBTH) are added and the reaction mixture is heated at the temperature in the range of 73 to 77 °C for 30 to 50 minutes or at the temperature in the range of 94 to 96 °C for 4 to 6 minutes,

d) then an acidic solution of a ferric salt comprising citric acid and sulfamic acid is added to the mixture, the mixture is then cooled for 5 to 15 minutes to the room temperature,

e) then absorbance is measured at 654 nm, and

f) the content of HA in the sample is calculated from the measured data.

In a preferred embodiment of the method of the invention, the solution of MBTH in step c) also includes dithiotreitol (DTT) in the concentrations 2-20mM. The monovalent salt in the step a) is preferably in the concentration 50 mM and is selected from the group comprising NaCl, KC1, CH ₃ COONa, NaN0 ₃ , Na ₂ S0 ₄ , and K ₂ S0 ₄ . In step c), 1M aqueous solution of NaOH, KOH, or LiOH is preferably added as an alkali, wherein its final concentration in the reaction mixture is 0.25M. In a preferred embodiment of the method, 50mM aqueous solution of FeCl ₃ , 0.2M aqueous solution of citric acid, and l%w/v aqueous solution of sulfamic acid are added in step d).

In order to eliminate the influence of anionic glycosaminoglycans with a negative charge higher than HA, the step a) comprises a set of the following sub-steps:

al) the sample is mixed with acetate buffer and aqueous solution of Na ₂ S0 ₄ or K ₂ S0 ₄ so that the final concentration of acetate buffer after mixing is in range of lOmM to lOOmM and the concentration of the sulfate is between 75 and 300mM,

a2) 0.5-2%w/v aqueous solution of cetyltrimethylammoniumbromide or cetylpyridiniumchloride, comprising the same concentrations of acetate and sulfate as the sample after mixing in step al), is added to the sample solution from step al) in order to precipitate anionic glycosaminoglycans with a negative charge higher than HA, which lead to the sediment formation,

a3) the sediment formed in step a2) is removed.

In a preferred embodiment of the invention, Na ₂ S0 ₄ is added in steps al) and a2) so that its final concentration in the sample solution after mixing is 80mM, and acetate buffer is added so that its concentration in the sample solution after mixing is 50mM and pH is 6, and then the sediment is removed by centrifugation in step a3).

In order to eliminate the influence of reducing sugars, further sub-steps a4) to a6) are performed within the step a) and after step a3), as follows:

a4) an alkali, in an amount sufficient for reaching pH >12, is added to the supernatant obtained in step a3) to precipitate hyaluronic acid in form of a sediment,

a5) the sediment formed in step a4) is isolated and washed with 5-20mM of an alkali solution,

a6) the sediment from step a5) is dissolved in acetate buffer with the concentration from 10 to 100 mM, pH between 5.5 and 6.5, and containing sodium sulfate or potassium sulfate in the concentration 100-300 mM.

In a preferred embodiment of the method, the alkali in step a4) is selected from the group comprising NaOH, KOH, and LiOH, the concentration of the alkali is 1M, and the added volume is 20 to 40 μΐ, per 200 to 500 ]xL of the supernatant, preferably 25 μΐ _^ of 1M NaOH per 400 μΐ, of supernatant. The sediment is separated preferably by centrifugation in step a5) and is washed with 5-20mM alkali selected from the group comprising NaOH, KOH, and LiOH, more preferably it is washed at least two times with lOmM NaOH. The acetate buffer in step a6) preferably has pH 6, its concentrationis lOOmM, and the sulfate is preferably Na ₂ S0 ₄ with the concentration 160 mM.

The concentration C(HA) of hyaluronic acid in the sample in mg/mL is calculated in step f) preferably according to the formula: r _j r _& _ {A _s -Ab)*V {colored sample ^~ )*M( HA2)

^ ' e*V HA sample) ' where A _s is the absorbance of the sample, Ab is the absorbance of the background taking into account the dilution after the colour development; V(colored sample) is the final volume of the reaction mixture; Μ(ΔΗΑ2) is the molecular weight of the unsaturated disaccharide of hyaluronic acid, 379 g/mol; ε is the extinction coefficient, 34735 mol ^"1 ; V(HA sample) is the volume of the original sample in the final volume of the reaction mixture.

Methods

Determination of the activity of SpnHyl

The activity of SpHyl was assayed on pure 0.1% HA, pure 0.1-2% chondroitin sulfate (CS) or on their mixes with constant 0.1% HA and 0.1-2% CS in a variety of buffers, which, however, always contained 50 mM sodium acetate buffer pH 6. The activity of SpHyl was defined as an increase in samples absorbance at 232 nm referred to 1 mg of enzyme. To 1 ml of 0.1% HA preheated to 37°C, the 1 μΐ of enzyme solution was added. Then the changes in A232 were monitored and recorded for 10 to 20 minutes. The slope of the linear part of the kinetic curve corresponded to enzyme activity according to the following formula: ΔΑ ₂₃ 2 ·1000 _{T T r} , -.

Activity =— [U/mg ] where ΔΑ232 - absorbance difference between the sample and background at 232 nm; At - time, min; ε - extinction coefficient, 5500 mol.l^.cm ^"1 ; C - protein concentration, mg/mL. The unit of the activity is defined as the amount of enzyme necessary for the formation of one μηιοΐ of unsaturated bonds in 0.1%w/v HA at pH 6 at 37°C in one minute.

The effect of the ions (S0 ₄ ^2" , Ca ²⁺ , Mg ²⁺ ) and CTAB (up to 2%w/v) on the SnpHyl activity was evaluated by graphical comparison of the kinetic curves without numerical calculations. Anthon-Barret method

To 0.1 mL of the sample containing HA, 0.1 mL of 0.5 M sodium hydroxide and 0.1 mL of a freshly prepared 1 :1 mixture of 0.3% (w/v) 3-methyl-2-benothiazolinonehydrazone (MBTH) and 1 mg/mL dithiothreitol (DTT) was added. The mixture was heated for 30 min at 80 °C and 0.2 mL of 0.5% (w/v) FeCl ₃ , 0.5% (w/v) sulfamic acid, then 0.25 M hydrochloric acid was added to the warm mixture. The solution was cooled, the sample was diluted ten times, and the absorbance was measured at 620 nm (Anthon & Barrett, 2002). SpHyl-MBHT test

Complete protocol

The complete protocol was prepared for the HA samples contaminated with high concentrations of sulfated glycosaminoglycans (up to 5%w/v of CS) and reducing sugars (up to 5 %w/v of glucose). It describes the most preferred conditions of the method according to the invention. The process volumes were optimized for the standard spectrophotometric cell with the light path of 10 mm and volume approx. 1 mL.

1. Mix 100 L of the analysed sample with 100 μL of the solution A (lOOmM acetate buffer pH 6, 160mM Na ₂ S0 ₄ ), vortex for 5 seconds.

2. Add 300 μL of the solution B (1% CTAB, 50mM acetate buffer pH 6, 80mM

Na ₂ S0 ₄ ), vortex for 5 seconds.

3. Remove the sediment by centrifugation at >12000 g for 5 minutes.

4. Transfer 400 μL of the supernatant into a new microtube and add 25 μΐ, of 1M NaOH.

5. Spin at > 12000 g for 5 minutes, remove the supernatant.

6. Wash the sediment twice with 500 μΐ of 1 OmM NaOH.

7. Dissolve the sediment in 400 μL of the solution A, mark the microtube as the "sample", transfer 200 μL of the sample to the tube marked as the "background".

8. Add 1 μL of SpnHyl enzyme (2000 U/mg, 2 mg/ml) to the "sample" tube, incubate for 5-10 minutes at 37°C.

9. Add 200 μΐ of the solution C (1M NaOH) and 400 μL of the solution D (0.2%w/v MBTH and lOmM DTT) to both tubes - "sample" and "background". Mix thoroughly, then incubate at 75°C for at least 30 minutes but not more than 50 minutes. Alternatively, incubate at 95°C for 5 minutes but not more than 6 minutes. W

10. Add 500 μΐ of the solution E (50mM of FeCl ₃ , 1M HC1, l%w/v sulfamic acid, 0.2M citric acid) to the "sample" and the "background". Let it cool to the ambient temperature for 5 - 10 minutes .

11. Measure absorbance at 654 nm, dilute if necessary.

12. Calculate HA in mg/niL or g/L:

(A _s -Ah)*V(colored sample)*M{A A2)

C HA) =

ε*ν(ΗΑ sample)

where A _s and Ab are absorbances of the "sample" and "background", taking into account the dilution after the colour development; V(colored sample) is the final volume of the reaction mixture, it is 1.3 mL for this protocol; M(AHA2) is the molecular weight of unsaturated disaccharide, 379 g/mol; ε is the extinction coefficient, 34735 mol ^"1 ; V(HA sample) is the volume of the sample in the final volume of the reaction mixture, 0.04 mL for this protocol. The simplified formula after all the transformations is:

C(HA) = 0,3546135 * (A _s - A _b )

Simplified Protocol I

The original protocol can be simplified, if the sample comprises small amounts of reducing sugars. The concentration of the reducing sugars can be considered low, if the absorbance of "background" at 654 nm is less than 0.7. The absorbance of the "background" is determined as follows: mix 100 μΤ of the sample with 100 of the solution A; add 200 μΐ^ of the solution C and 400 μL of the solution D; incubate for 5 minutes at 95°C; let it cool and measure at 654 nm.

In case of low concentration of reducing sugars, steps 4-7 can be skipped and 200 μΐ _^ of the supernatant from step 3 are pipetted to both the "sample" and the "background" tubes, then the procedure continues from the step 8 to 12 of the complete protocol. The modifications do not affect the formula; HA concentration is calculated in the same way as in the complete protocol. Simplified Protocol II

If the sample contains neither reducing sugars, nor any sulfated glycosaminoglycans, the protocol can be simplified further. The concentration of sulfated glycosaminoglycans can be considered low, if the addition of 300 μΤ of the solution B into the mixture of 100 μL of the sample and 100 μΐ, of the solution A does not lead to the formation of a precipitate or turbidity. In that case, steps 2-3 are also omitted, besides the already omitted steps 4-7. The first step changes: 200 iL of a sample is mixed with 200 of the solution A, and after vortexing the solution is divided into equal parts: the "sample" and the "background". Then the procedure continues from step 8. An increase of the volume in the first step leads to a change of the formula for HA calculation:

C(HA) = 0,1418454 * (A _s - A _b )

Brief description of the drawings

Fig. 1 represents the result of the reaction of 0.5% HA (weight average molecular weights from 24 kDa to 2049 kDa) with MBTH according to the Anthon-Barrett method. The first condensation reaction was performed at 80°C for 30 min.

Fig. 2 represents the absorbance of the HA sample with the concentration of 0.1% after incubation with MBTH at various temperatures. The optical density increases in time due to the degradation of HA in a peeling reaction.

Fig. 3 represents the enzymatic activity of S. pneumoniae hyaluronan lyase SpHyl.

A. Chromatogram of the products of cleavage of 50 mg of HA with 15 U of SpnHyl after two hours at 37 °C. The higher peak corresponds to 4,5-unsaturated disaccharide; the lower peak corresponds to impurities.

B. Polyacrylamide gel of HA cleaved with the same amounts of bovine hyaluronidase (left) and SpnHyl (right). PV - original sample. SpnHyl can cleave the same amount of HA in 1 minute as bovine hylauronidase does in 24 hours.

Fig. 4 represents the repeatability of the presented method at various temperatures and times of the reaction. HA samples were mixed with the solution A, digested with SpnHyl, then solutions C and D were added into the reaction mixture according to the protocol (see the Methods). Then the samples were subjected to heating at 73, 75, and 77°C for 30, 40, and 50 minutes. Immediately after heating, the solution E was added into the heated reaction mixture. After cooling to the room temperature and colour development, the absorbance spectra were measured in the range of 500-700 nm.

A. Three HA concentrations were tested: 25, 50, and 100 g/mL. The samples were heated at 75°C for 30 minutes (solid line), 40 minutes (dashed line), and 50 minutes (dotted line). The highest repeatability is at 654 nm. B. The combinations of three temperatures (73, 75, and 77°C) and three times (30, 40, and 50 min) give nine spectra. The signal intensity is the same at 654 nm, but shows significant deviations at 620 nm.

Fig. 5 represents the influence of the molecular weight of HA on the accuracy of the presented method.

Examples

Example 1

The analysis of pure HA samples was performed according to the simplified protocol II. Solutions of a known HA concentration from 1 to 2000 μg mL were analysed.

The process volumes were optimized for the standard photometric cell with the light path of 10 mm and volume approx. 1 mL.

First, 100 μΐ, of the analysed sample are mixed with 100 μΐ _^ of the solution A (lOOmM acetate buffer pH 6, 160mM Na ₂ S0 ₄ ) and the mixture is vortexed for 5 seconds.

Then 200 μΐ _^ of the mixture are transferred to a tube marked as the "sample" and 200 μΐ, of the mixture are pipetted to a second tube marked as the "background". Then 1 μί of SpnHyl enzyme (2000 U/mg, 2 mg/mL) is added to the tube "sample" and the mixture is incubated for 5-10 min at 37°C. Then 200 μΐ, of the solution C (1M NaOH) and 400 μΐ, of the solution D (0.2%w/v MBTH and lOmM DTT) are added to both tubes - the„sample" and the „background". The mixture is thoroughly mixed and incubated at 75°C for at least 30 minutes but not more than 50 minutes. Alternatively, it can be incubated at 95°C for 5 minutes but not more than 6 minutes.

Finally, 500 xL of the solution E (50mM FeCl ₃ , 1M HC1, l%w/v sulfamic acid, 0.2M citric acid) are added to the„sample" and„background", and the mixture is cooled to the laboratory temperature for 5- 10 minutes. Then absorbance at 654 nm is measured. If necessary (in case of the absorbance exceeds the detection range of the spectrophotometer), both the sample and the background should be diluted and measured once more.

The results are presented in the Table 2.

Table 2. Determination of pure HA.

20 19.87 -0.6

30 30.14 +0.5

40 39.39 -1.6

100 100.56 +0.6

200 196.38 -1.8

300 304.95 +1.6

1000 1000.01 0

2000 2028 +1.4

The results show that HA in pure solutions can be reliably determined within the range of 2-1000 μg/mL. Therefore the presented method it is suitable for the purity verification of the commercially available HA.

Example 2

The solutions of pure HA with the concentrations from 8 to 72 μ /ιηΐ, and various ratios of HA:CS from 1 :2 to 1 :10 were processed according to the Simplified Protocol I.

The process volumes were optimized for the standard photometric cell with 10 mm light path and volume of approx. 1 mL.

First, 100 μΐ _^ of the analysed sample are mixed with 100 ΐ. of the solution A (lOOmM acetate buffer of pH6, 160mM Na ₂ S0 ₄ ) and the mixture is vortexed for 5 seconds. Then 300 μΐ, of the solution B (l%w/v CTAB, 50mM acetate buffer of pH6, 80mM Na ₂ S0 ) are added and the mixture is vortexed again for 5 seconds. Then the sediment is removed by centrifugation at >12000 g for 5 minutes.

Then 200 of supernatant are transferred into a microtube referred to as the„sample", and 200 μΐ, of supernatant are pipetted into the second tube referred to as the„background". Then 1 μΤ of SpnHyl enzyme (2000 U/mg, 2 mg/ml) is added into the "sample" tube and the mixture is incubated for 5-10 minutes at 37°C. Then 200 μΤ of the solution C (1M NaOH) and 400 μΐ, of the solution D (0,2% w/v MBTH a lOmM DTT) are added into both tubes - the „sample" and the„background". The mixture is mixed thoroughly and incubated at 75 °C for 30 min but no more than 50 min. Alternatively it can be incubated at 95°C for 5 min but no more than 6 min.

Finally, 500 μΐ, of the solution E (50mM FeCl ₃ , 1M HC1, l%w/v sulfamic acid, 0.2M citric acid) is added into the "sample" and the "background" and the mixture is cooled to laboratory temperature for 5-10 min. Then absorbance is measured at 654 nm. If necessary (in case of the absorbance exceeds the detection range of the spectrophotometer), both the sample and the background should be diluted and measured once more. During the analysis, CS was completely removed from the samples by fractional precipitation with use of CTAB in form of the solution B. The measured data show a high efficiency of the chosen method (see the Table 3). Table 3. Determination of HA in the mixture with CS

This example proves the applicability of the presented method for the analysis of complex samples comprising sulfated glycosaminoglycans along with HA Example 3

Solutions of pure HA with various additions of glucose (up to 50 g/L) were used as a model of influence of the reducing sugars on determination of HA in real samples. The complete protocol had to be used for removing the reducing sugars, where HA is precipitated in form of insoluble HA-CTAB complex and glucose is washed out with the supernatant.

The process volumes were optimized for the standard photometric cell with 10 mm light path and volume of approx. 1 mL.

First, 100 μΤ of the analysed sample are mixed with 100 μΐ, of the solution A (lOOmM acetate buffer of pH 6, 160mM Na ₂ S0 ₄ ) and the mixture is vortexed for 5 seconds. Then 300 μΤ of the solution B (l%w/v CTAB, 50mM acetate buffer pH 6, 80mM Na ₂ S0 ₄ ) are added and the mixture is vortexed again for 5 secdnds. Then the sediment is removed by centrifugation at >12000 g for 5 minutes.

Then 400 μΐ, of supernatant are transferred into a new microtube and 25 μΤ of 1M NaOH are added to form an insoluble HA-CTAB complex. The complex is separated from the supernatant by centrifugation at > 12000 g for 5 minutes and then it is washed with 2x 500 μΐ, of lOmM NaOH. In the next step, the sediment (complex) is dissolved in 400 μΐ, of the solution A, the microtube is marked as the„sample", and 200 μΐ _^ of the solution are pipetted to the second microtube marked as the background". Then 1 μΐ, of SpnHyl enzyme (2000 U/mg, 2 mg/mL) is added to the "sample" and the mixture is incubated for 5-10 minutes at 37°C. Then 200 μΐ, of the solution C (1M NaOH) and 400 μΐ of the solution D (0.2%w/v of MBTH and lOmM of DTT) are added to both tubes - the "sample" and the "background". The mixture is mixed thoroughly and incubated at 75°C for 30 minutes but not more than 50 minutes. Alternatively, it can be incubated at 95°C for 5 minutes but not more than 6 minutes.

Finally, 500 μ]1 of the solution E (50mM FeCl ₃ , 1M HC1, l%w/v sulfamic acid, 0.2M citric acid) are added to the "sample" and the "background" and the mixture is cooled to the laboratory temperature for 5-10 minutes. Then absorbance is measured at 654 nm. If necessary (in case of the absorbance exceeds the detection range of the spectrophotometer), both the sample and the background should be diluted and measured once more.

The results of the analysis are presented in the Table 4.

Table 4. Determination of HA in the mixture with glucose.

The reducing sugars are often used as a component of cultivation media and veterinary and pharmaceutical preparations. The use of SpnHyl-MBTH method enables determination of HA in such samples.

Example 4

Sample of HA without the reducing sugars and sulfated glycosaminoglycanes

The process volumes were optimized for the standard photometric cell with 10 mm light path and volume of approx. 1 mL.

First, 200 μί, of the analysed solution and 200 μΐ, of the solution A (20mM acetate buffer pH 6.5, 50mM NaCl) are mixed and the mixture is vortexed for 5 seconds. Then 200 μΐ, of the mixture are transferred into the microtube marked as the "sample", and 200 μΐ _^ of the mixture are pipetted into the second tube marked as the "background". Then 1 μΐ, of SpnHyl enzyme (1000 U/mg, 1 mg/mL) is added into the„sample" tube, and the mixture is incubated for 20 minutes at 37°C. Then 100 iL of the solution C (1M NaOH) and 500 \x of 0.1%w/v MBTH are added into both tubes - the "sample" and the "background". The mixture is mixed thoroughly and incubated at 75°C for 30 minutes.

Finally, 500 μΤ of the solution E (50mM FeCl ₃ , 0,5M HC1, 0.5%w/v sulfamic acid, 0.1M citric acid) are added into the "sample" and the "background" and the mixture is cooled to the laboratory temperature for 5-10 minutes. Then absorbance at 654 nm is measured. If necessary (in case of the absorbance exceeds the detection range of the spectrophotometer), both the sample and the background should be diluted and measured once more.

Example 5

Sample of HA comprising sulfated glycosaminoglycans but not comprising reducing sugars

The process volumes were optimized for the standard photometric cell with 10 mm light path and volume of approx. 1 mL.First, 100 μΐ, of the analysed sample and 100 μΐ, of the solution A (lOOmM acetate buffer of pH 6,5, 200mM Na ₂ S0 ₄ ) are mixed and the mixture is vortexed for 5 seconds. Then 300 μΐ. of the solution B (l%w/v CTAB, 50mM acetate buffer pH 6, lOOmM Na ₂ S0 ₄ ) are added and the mixture is vortexed again for 5 seconds. Then the sediment is removed by centrifugation at >12000 g for 5 minutes.

Then 200 μΐ, of the supernatant are transferred into the microtube marked as the "sample", and 200 μί of the supernatant are pipetted into the second tube marked as the "background". Then 1 μΐ, of SpnHyl enzyme (1000 U/mg, 2 mg/mL) is added into the„sample" tube, and the mixture is incubated for 10 minutes at 37°C. Then 100 μΐ, of the solution C (1M NaOH) and 400 μΐ, of the solution D (0.3%w/v MBTH and 5mM DTT) are added into both tubes - the "sample" and the "background". The mixture is mixed thoroughly and incubated at 75°C for 30 minutes.

Finally, 500 xL of the solution E (50mM FeCl ₃ , 0,5M HC1, l%w/v sulfamic acid, 0.2M citric acid) are added into the "sample" and the "background" and the mixture is cooled to the laboratory temperature for 5-10 minutes. Then absorbance at 654 nm is measured. If necessary (in case of the absorbance exceeds the detection range of the spectrophotometer), both the sample and the background should be diluted and measured once more. Example 6

Sample of HA with a high content of reducing sugars and sulfated glycosaminoglycans The process volumes were optimized for the standard photometric cell with 10 mm light path and volume of approx. 1 mL.First, 100 μΐ of the analysed sample and 100 μΐ, of the solution A (80mM acetate buffer of pH 5.5, 250mM Na ₂ S0 ₄ ) are mixed and the mixture is vortexed for 5 seconds. Then 300 of the solution B (1.3%w/v CTAB, 40mM acetate buffer pH 5.5, 125mM K ₂ S0 ₄ ) are added and the mixture is vortexed again for 5 seconds. Then the sediment is removed by centrifugation at >12000 g for 5 minutes.

Then 400 μΐ, of supernatant are transferred into a new microtube and 30 μΐ, of 1M KOH are added to form an insoluble HA-CTAB complex. The complex is separated from the supernatant by centrifugation at >12000 g for 5 minutes and then it is washed with 3x 200 μΐ of 15mM KOH.

In the next step, the sediment (complex) is dissolved in 400 μΐ, of the solution A, the microtube is marked as the "sample" and 200 μΤ of the solution is pipetted into the second tube marked as the "background". Then 1 ih of SpnHyl enzyme (1000 U/mg, 2 mg/mL) is added into the„sample" tube and the mixture is incubated for 15 minutes at 37°C. Then 200 μ∑ of the solution C (1M NaOH) and 400 μΐ, of the solution D (0.3%w/v MBTH and 5mM DTT) are added into both tubes - the "sample" and the "background". The mixture is mixed thoroughly and incubated at 95°C for 5 minutes.

Finally, 500 μΐ, of the solution E (50mM FeCl ₃ , 0,5M HC1, l%w/v sulfamic acid, 0.2M citric acid) is added into the "sample" and the "background" and the mixture is cooled to laboratory temperature for 5-10 minutes. Then absorbance at 654 nm is measured. If necessary (in case of the absorbance exceeds the detection range of the spectrophotometer), both the sample and the background should be diluted and measured once more.

Example 7

The final experiments before the analysis of real samples were performed in a model system comprising HA, CS, and glucose. Various concentrations of HA from 10 to 1000 μg mL were tested, the samples comprised also 50 mg/mL of glucose and CD of the concentration of 20times higher than the concentration of HA. The SpnHyl-MBTH method effectively separates HA, reducing sugars, and sulfated glycosaminoglycans, and enables determination of HA even at the presence of a high excess of interfering substances (Table 5). W 201

Table 5. Determination of HA in the mixture with glucose and CS.

Example 8

Geloren (Contipro, a.s.) is a gelatine gel tablet for dogs and horses. The "dog" version comprises 0.6% HA, 5% CS, 1.24% glucosamine, 7% hydrolysed collagen, 28% gelatine (http-V/wwvv^aktivnizvirexz/kloubni-vyziva-pro-psy). The "horse" version comprises 0.4% HA, 1.9% CS, 15.5%> gelatine, and also a high amount of glucose (http://www.aktivnizvire.cz vyziva-kloubu-kone). All percents indicated in the Geloren composition are weight percents, i.e. w/w. A slice of gel of the size of approx. 0.1 g was weighed and then melted in 9 volumes of 20mM Tris-HCl buffer pH 8 at 80°C. Then the sample was digested with pancreatin (10 mg of pancreatin per 1 mL of the melted gel) at 37°C for 1 hour. The digestion with pancreatin is necessary as the gelatine creates complexes with CTAB and thus inhibits the analysis (Ji et al., 2004), Then the sample is processed according to the Complete Protocol SpnHyl-MBTH including the removal of reducing sugars and CS. The measured concentration of HA was 0.408%> and 0.59% of HA for the horse and the dog version of Geloren, respectively.

Example 9

The MBTH-SpHyl method was used for the monitoring of HA content during the whole manufacturing process beginning from Streptococcus zooepidermicus (SEZ) cultivation, through to the downstream processes and ending with the final product. More than 500 samples were analyzed in total.

The absence of reducing sugars and CS in the analysed samples and low concentrations of HA in in some phases of isolation allowed the use of "Simplified protocol II". That allowed the detection range to be expanded to 1-2,000 μg/ml. This obstacle was eliminated by incubation of the samples at 95° for 5 minutes prior to digestion. The total assay time of one hour was unsatisfactory for mass analysis. Boosting the temperature from 75°C to 95°C allowed significant reduction of the reaction time from 30-50 minutes to 5-6 minutes. This along with the reduction of the digestion time to 5 minutes dropped full assay duration down to 20-30 minutes without any significant negative impacts on the sensitivity. This is currently the most efficient time in practice of colorimetric methods of HA quantification. Only the turbidimetric determination of HA is faster, however at the costs of the accuracy and robustness.

Literature:

Alkrad JA, Merstani Y, Neubert RH, (2002) New approaches for quantifying hyaluronic acid in pharmaceutical semisolid formulations using HPLC and CZE. J Pharm Biomed Anal 7;30(4):913-9.

Anthon GE, Barrett DM, (2002) Determination of reducing sugars with 3-methyl-2- benzothiazolinonehydrazone. Anal Biochem 305: 287-289.

Baker JR, Dong S, Pritchard DG, (2002) The hyaluronan lyase of Streptococcus pyogenes bacteriophage H4489A. Biochemical Journal 365: 317 - 322.

Dische Z, (1947) A new specific color reaction of hexuronic acids. J Biol Chem. 167(1): 189- 98.

Elson LA, Morgan W, (1933) A colorimetric method for the determination of glucosamine and chondrosamine. Biochem J 27: 1824-1828.

Gassier N, Reissner C, Janzen N, Kahnert H, Kleesiek K, (1993) A high performance liquid chromatography method for the determination of glycosaminoglycans in human blood. Eur J Clin Chem Clin Biochem31(8):503-51 1.

Haserodt S, Aytekin M, Dweik RA, (201 1) A comparison of the sensitivity, specificity, and molecular weight accuracy of three different commercially available Hyaluronan ELISA-like assays. Glycobiology 21(2):175-83. doi: 10.1093/glycob/cwql45.

Honda S, Nishimura Y, Chiba H, Kakehi K, (1981) Determination of carbohydrates by condensation with 3-methyl-2-benzothiazolinonehydrazone. Anal Chim Acta 131(C): 293- 296.

Ji Y, Zhang XH, Guo R, (2004) Interaction between Gelatin and Cationic Surfactant CTAB. Acta Chimica Sinica, 62(4): 345-350.

Jourdian GW, Wolfman M, Sarber R, Distler J, (1979) A specific, sensitive method for the determination of hyaluronate, Anal Biochem 96(2): 474-480, http://dx.doi.org/10.1016/0003- 2697(79)90609-2. Kogan G, Soltes L, Stern R, Gemeiner P, (2007) Hyaluronic acid: a natural biopolymer with a broad range of biomedical and industrial applications. Biotechnol Lett 29: 17-25. doi: 10.1007/sl 0529-006-9219-z.

Lapcik Jr L, Lapcik L, De Smeedt S, Demeester J, Chabrecek P, (1998) Hyaluronan: preparation, structure, properties, and applications. Chem Rev 98:2663-84.

Lee VL, (1966) Automated Analysis of Amino Sugars using the Elson-Morgan Reaction. Yale Medicine Thesis Digital Library. Paper 6.

Oueslati N, Leblanc P, Harscoat-Schiavo C, Rondags E, Meunier S, Kapel R, Marc I, (2014)

CTAB turbidimetric method for assaying hyaluronic acid in complex environments and under cross-linked form. Carbohydr Polym 4(112):102-1088. doi:10.1016/j.carbpol.2014.05.039.

Reed CE and Reed WF, (1989) Light scattering power of randomly cut random coils with application to the determination of depolymerization rates. J Chem Phys 91 :7193-9.

Reissig JL, Strominger JL, Leloir LF, (1955) A modified colorimetric method for the estimation of N-acetylaminosugars. J Biol Chem 217: 959-966.

Ruckmani K, Shaikh SZ, Khalil P, Muneera MS, Thusleem OA, (2013) Determination of sodium hyaluronate in pharmaceutical formulations by HPLC-UV. J Pharm Anal 3(5):324-

329, http://dx.doi.Org/10.1016/j.jpha.2013.02.001.

Shimmura S, Ono M, Shinozaki K, Toda I, Takamura E, Mashima Y, Tsubota K, (1995) Sodium hyaluronate eyedrops in the treatment of dry eyes. The British journal of ophthalmology 79 (11): 1007-1011. doi: 10.1136/bjo.79.11.1007.

Whistler RL and BeMiller JN, (1958) Alkaline degradation of polysaccharides. Adv Carbohydr Chem 13:289-329.

Yang BY and Montgomery R, (2001) β-Elimination of glucosyluronic residues during methylation of an acidic polysaccharide from Erwinia chrysantemi CU 643. Carbohydr Res332:317-23.