Lactic acid can be easily and precisely determined by reversed-phase high performance liquid chromatography with pre-column derivatization

Simple, sensitive and rapid methods of lactic acid (LA) quantifi cation in the presence of other short-chain fatty acids in specimens of biological origin were developed to facilitate investigation of metabolism in body tissues and fl uids of farm animals, fermentation processes, and ensiled products by reversed-phase high-performance liquid chromatography (RP-HPLC). LA and other acids in assayed samples were converted to their sodium salts (RCOONa). Derivatization of RCOONa was carried out with of a solution of 2,4’-dibromoacetophenone (DBAP) (2.2–2.4 g DBAP/100 ml of acetone or 20 g DBAP/100 ml of chloroform) and triethylamine. The reaction mixture was vigorously mixed and then reacted for 1 h at 45oC. The separation of derivatized acids was performed on a Nova Pak column (4 μm, 300 x 3.9 mm I.D., Waters) using a binary gradient elution program and photodiode detection at 259 nm. Lactic, acetic and propionic acid peaks were eluted at 11.05±0.10, 18.0±0.1 and 19.6±0.1 min, respectively, while the iso-capronic acid peak (an internal standard) at 23.1±0.1 min. The total run time of the HPLC analysis was 59 min. LA detection at 259 nm assures the excellent selectivity and high sensitivity of the proposed method; the limit of detection (LOD) and the limit of quantifi cation (LOQ) were equal to 0.04 and 0.13 ng/ml, respectively. The use of chloroform as a solvent for DBAP in the derivatization of LA in samples followed by RP-HPLC offers the best sensitivity of LA determination in biological samples. The utility of the method was demonstrated by LA analysis in different materials such as silage, fermented dairy and vegetable products, intestinal digesta, blood and etc. The presented methods based on an inexpensive HPLC column, simple and rapid processing of samples provide accurate and sensitive analytical tools for routine determination of LA, particularly in specimens of biological origin.


INTRODUCTION
Lactic acid (LA) has many important biochemical and physiological properties.It is chiral and has two optical isomers: D-(-)-lactic acid (R-lactic acid) and its mirror image, L-(+)-lactic acid (S-lactic acid), which is the biologically important isomer (Ewaschuk et al., 2004).In animal bodies, L-LA is constantly synthesized from pyruvate via lactate dehydrogenase in a fermentation process during metabolism and, especially, during intensive exercise like sprinting.DL-LA is also an intermediate in animals, particularly ruminants (Ewaschuk et al., 2004;Davies et al., 2007).Formation of DL-LA from feed components is highly correlated with the sum of the soluble sugar and soluble protein content in forage (Weinberg et al., 2004a).Moreover, silage, a major component in the rations of dairy cattle, is forage preserved through lactic acid fermentation (Ricke, 2003).Therefore, inoculants comprising mainly lactic acid bacteria (LAB) are applied as silage supplements to stimulate ensiling fermentation.Some LAB strains produce via lactic acid an aromatic organic acid, phenyllactic acid, possessing fungicidal activity (Greifova et al., 2007).
Lactic acid, a product of fermentation of lactose, can also be found in many milk products (koumiss, kefi r, yogurt or several cottage cheeses) as well as in ensiled vegetables, grasses, etc. LA can also be used as a food preservative or as a pH adjusting additive to various processed foods (Ricke, 2003).
Considering the above, monitoring concentrations of LA and other organic acids in specimens of biological origin is essential in physiological and nutritional studies.Therefore, titration after steam distillation, enzymatic, spectrometric, colorimetric and electrophoresis methods, chemiluminescence biosensor assay, gas (GC) and liquid (HPLC) chromatography have been used for the determination of LA and other shortchain fatty acids (C 2 -C 6 ) in specimens of biological origin (Gomez-Alvarez et al., 1999;Czauderna andKowalczyk, 2001, 2002;Czauderna et al., 2001Czauderna et al., , 2002;;Mataix et al., 2001;Ewaschuk et al., 2004;Wu et al., 2005;Resende et al., 2006;Inoue et al., 2007).The main advantages of high-resolution HPLC over the above-mentioned methods are simplicity, the possibility of using selective HPLC columns and very sensitive photodiode detection having excellent response linearity over a wide concentration range of acids in samples (Sivakesava et al., 2001;Lin and McKeon, 2005).Unfortunately, LA as well as other organic acids and their methyl/ethyl esters have high molar absorptions only at short UV wavelengths (<205 nm).Therefore, for quantifi cation of LA in the presence of other volatile (VFA) and short-chain fatty acids (SCFA) by HPLC it is better to convert them with reagents having high molar absorptivity at longer UV wavelengths (>250 nm) (Czauderna et al., 2002).
For these reasons it seemed desirable to develop simple and more selective derivatization methods for quantifi cation of LA, particularly in the presence of other VFA and SCFA, using selective reversed-phase liquid chromatography (RP-HPLC) with sensitive photodiode array detection.To obtain satisfactory stability of the formed LA derivative we used 2,4'-dibromoacetophenone as the derivative reagent, and a long HPLC C 18 column for the highly effi cient resolution of LA from other acids and endogenous species present in biological samples.

Chemicals and materials
All chemicals were of analytical grade and organic solvents were of HPLC grade.Lactic acid (LA), acetic acid (AA), propionic acid (PA) and chloroform were purchased from POCh (Gliwice, Poland), while iso-capronic acid (an internal standard; IS) was from Sigma (St. Louis, MO, USA).Methanol and acetonitrile were purchased from Lab-Scan (Eire), whereas 2,4'-dibromoacetophenone (DBAP) and triethylamine were from Merck (Darmstadt, Germany).Water used for the preparation of mobile phases (Table 1) and chemical reagents was prepared using an Elix TM water purifi cation system (Millipore, Canada).

Chromatographic equipments
An alliance separation module (model 2690, Waters) was equipped with a Waters 996 photodiode array detector (DAD) operated in the UV range from 195 to 400 nm.Development of the gradient systems, collection of chromatograms, peak integrations and peak purity analyses (in the UV range from 230 to 290 nm) were performed using a Pentium III computer (Compaq) and Millennium 2001 software (version 2.15).Acid derivatives were quantifi ed at 259 nm.UV spectra of analytes in an eluate with a spectral resolution of 1.2 nm were recorded every second and were electronically stored on a computer hard disk.Fractionations and quantifi cations were performed on a Nova Pak C 18 column (4 µm, 300 × 3.9 mm I.D., Waters) in conjunction with a Waters guard column of 10 × 6 mm I.D. containing reversed-phase C 18 (30-40 µm) pellicular packing material.1 041 875 22.3 × 10 6 1 1.07 mg of LA was used for the derivatization procedure according to the new derivatization method; intra-assay and inter-assay coeffi cient of variations based on 3 repeated analyses were equal to 1.2 and 2.5%, respectively 2 110 g of carefully grinded grass silage was extracted with 75 ml of ethyl alcohol; the resulting mixture was centrifuged at 2000 g for 10 min.The supernatant was alkalized with 6 M NaOH to ~pH 9. Then the resulting mixture was centrifuged at 2000 g for 10 min.Finally the supernatant ethyl alcohol was removed under a stream of argon at ~30 o C. Obtained residue was used for the derivatization procedure according to the new derivatization method 3 sheep were fed standard forage without extra carbohydrate; 11 ml of a rumen fl uid was alkalized with 6 M NaOH to ~pH 9. Then the resulting mixture was centrifuged at 2000 g for 10 min.
Finally the supernatant water was removed under a stream of argon at ~30 o C. Obtained residue in a vial was used for the derivatization procedure according to the new derivatization method 4 11 ml of a rumen fl uid from sheep were fed standard forage enriched in a carbohydrate (10 g/kg forage); fl uid was processed according to the protocol described in the superscript 3 5 2 ml of plasma was used for the derivatization procedure according to the new derivatization method 6 5 ml of digesta from caecum or ileum of a broiler were mixed with 4 ml of water and then acidifi ed with 0.5 ml of formic acid.The resulting mixture was centrifuged for 10 min at ~ 5000 g and next the obtained supernatant was centrifuged for 15 min at ~ 6000 g.Finally, 0.6 ml of obtained supernatant was used for the derivatization procedure according to the new derivatization method 7 5 g of kefi r were used for the derivatization procedure according to the new derivatization method 8 6 ml of fermented cucumber, cabbage or red beet juice were used for the derivatization procedure according to the new derivatization method

Preparation of samples and derivatization of acids
Specimens of biological origin (Table 2) with 10 µl of IS solution (~1 mg of IS in 50 µl of acetone) were alkalized with 6 M NaOH to ~ pH 9 and then the resulting mixtures were centrifuged at 2000 g for ~10 min.Supernatants were separated from solid residues and the supernatant water was removed under a stream of argon at ~30 o C. Obtained residues in reacti-vials were used for derivatization with 2,4'-dibromoacetophenone (DBAP) (2.2-2.4 g DBAP/100 ml of acetone) and triethylamine.The derivatizing procedure for assayed LA, AA, PA and IS was the same as for biological samples.

Analytical solvents and gradient elution systems
Two HPLC grade solvents were used in chromatographic analysis of derivatized acids.Solvent B was water with 5% methanol (v/v), while solvent A was methanol.For analysis of all derivatized acids in standards and biological samples, a binary gradient elution program was used (Table 1).Analyses of derivatized acid in standards and biological samples were performed at a column temperature of 39°C, while analysed samples in the autosampler were maintained at 30 o C. Injection volumes were 1-50 µl.The maximum pressure of the HPLC system was 39 MPa.The limits of detection (LOD) were calculated as a signal-to noise ratio of 3, while the limit of quantifi cation (LOQ) was defi ned as 10 times the noise under a peak (Gratzfeld-Husgen and Schuster, 1994;Meyer, 1999).
LA, AA and PA in biological samples were identifi ed by the retention time of processed standards injected separately or/and by adding standard solutions to biological samples.It is recommended that all derivatized samples be protected from light and stored at about -20°C when not in use.

RESULTS AND DISCUSSION
The main aim of the present studies was to develop pre-column derivatization methods for quantifi cation of LA in the presence of other short-chain organic acids in biological samples using high-resolution RP-HPLC with sensitive photodiode detection.Initially, all acids in assayed samples were converted to their sodium salts (RCOONa).Therefore, no loss of LA and, especially, volatile fatty acids was found in derivatized samples (Lin and McKeon, 2005).To obtain a satisfactory yield of product formation, the derivatization of sodium salts of LA and other assayed acids was carried out with an excess of 2,4'-dibromoaceto-phenone (DBAP) and triethylamine.Based on our recent studies, derivatization should be performed at a lower temperature than the original reaction (i.e.80 or 100 o C) and for a shorter time -less than 2 h (Czauderna et al., 2001(Czauderna et al., , 2002)).It thus seemed reasonable to convert LA and other assayed acids at ~ 50 o C and for only ~ 60 min.Our preliminary results are in agreement with our previous studies (Czauderna et al., 2001(Czauderna et al., , 2002) ) in which the yield of the acid derivatization reaction positively correlated with medium alkalinity.Therefore, to achieve a satisfactory yield of the derivatization procedure, the conversion of assayed acids was performed using 2,4'-dibromoacetophenone and an excess of triethylamine.It is also recommended to add iso-capronic acid as an internal standard (IS) to assayed samples.

New derivatization method
Based on systematic investigations concerning optimization of the derivatization reaction and our recent studies (Czauderna et al., 2001(Czauderna et al., , 2002) ) following new method for the derivatization of LA and other assayed acids as presented below is recommended.
To a residue in a reacti-vial (see section "Preparation of samples and derivatization of acids), 800 µl of a solution of DBAP (2.2-2.4 g DBAP/100 ml of acetone) and 200 µl of triethylamine are added together with vigorous mixing.The resulting solution in a sealed tube is again mixed and reacted for 60 min at 45 o C. All derivatized samples are protected from the light.The derivatization procedure for standards is the same as for biological materials.The resulting standard solutions (2-50 µl) or processed biological sample solutions (2-6 µl) are injected onto a C 18 column (Nova Pak).All processed biological samples are protected from light and stored at about -20 o C until analysed.For the HPLC assay of LA and other acids, frozen samples should be re-warmed to ~30 o C to avoid the absorption and/or precipitation of analysed LA, AA, PA and the added internal standard (IS).To determine the yield of the acid separation followed by derivatization of acids, it is recommended to add IS to assayed standards and specimens of biological origin.

Ultra-effi cient derivatization method
To facilitate more sensitive and accurate analyses of very low concentrations of LA in specimens of biological origin, the objective of the current study was also to further improve the yield of the derivatization reaction.Modifi cations to the new derivatization method were required, specifi cally to select the solvent for ultra-effi cient derivatization of LA in biological samples.Based on systematic investigations concerning optimization of the solubility of DBAP and derivatization reaction products, we recommend chloroform as the solvent for the ultra-effi cient derivatization of LA and other assayed acids.Considering the above, the residue was used for the ultra-effi cient derivatization method as below is recommended.
To a residue in a reacti-vial (see section "Preparation of samples and derivatization of acids), 800 µl of a solution of DBAP (20 g DBAP/100 ml of chloroform) and 200 µl of triethylamine are added together with vigorous mixing.The resulting solution in a sealed tube is mixed again and reacted for 60 min at 45 o C. All derivatized samples are protected from the light.The derivatization procedure for standards is the same as for biological materials.The resulting biological sample solutions are injected (1-2.5 µl) onto a C 18 column (Nova Pak).All processed biological samples are protected from light and stored at about -20 o C until analysed.For the HPLC assay of LA and other acids, frozen samples should be re-warmed to ~20 o C to avoid the absorption and/or precipitation of analysed LA, AA, PA and IS, especially in assayed acid standards.To determine the yield of the acid separation followed by derivation of acids, it is recommended to add IS to assayed standards and specimens of biological origin.

Evaluation of the new derivatization method
A binary gradient system (Table 1) composed of H 2 O in methanol was found to provide a wide range of solvent strengths and excellent baseline stability (Figure 1).LA, AA, PA and IS can be satisfactorily separated using a long column packed with a strongly hydrophobic silica-based bonded phase.Satisfactory peak shapes, close to symmetrical, were observed with sample elution times even up to 24 min.The derivatized LA as well as other acids were substantially retained on the reversed-phase column and were distinct from background interference or from endogenous substances present in assayed biological samples (Figure 1).In our procedure the LA peak was eluted at 11.05±0.10min, the AA peak at 18.0±0.1 min, the PA peak at 19.6±0.1 min, while the IS peak at 23.1 ± 0.1 min.As expected, the peaks of LA, AA, PA and IS (monitoring the yield of the procedure) were absent from the blank when the proposed gradient elution program and photodiode detections at 205 and 259 nm were used.Moreover, the responses of the photodiode array detector to the concentrations of LA, AA and PA in standards were linear functions (Figure 1).Indeed, as illustrated in Figure 1A, the relationship between the LA concentrations (y µg/ml ) and LA peak areas (S n ) was linear over a wide range of concentrations of LA in standards (i.e. from 5 to 970 µg LA/ml), so, the correlation coeffi cient (r LA ) was equal to 0.990.Indeed, the use of monitoring at the maximum of LA derivative absorption assures excellent selectivity and high sensitivity of the proposed method.Consequently, the LOD and LOQ values obtained by the use of detection at 259 nm were equal to 0.04 and 0.13 ng/ml, respectively.Chromatographic analyses (Figure 1, Table 2) evidenced that background fl uctuations and the presence of endogenous species cannot interfere in the accurate and precise integration of the LA peak in all assayed specimens of biological origin.Analyses of LA purity and peaks based on Figure 1.A -a chromatogram for derivatized lactic acid (LA), acetic acid (AA), propionic acid (PA) and iso-capronic acid (IS) using the new derivatization method and the elution program (Table 1; the detection at 259 nm; AU -an absorption unit).B -UV spectra with their maxima (λ max ) of the derivatized standards of LA, AA, PA and IS.Chromatograms for derivatized acids in: C -grass silage; D -kefi r; fermented: E -cucumber, F -cabbage G -red beet; H -ovine ruminal fl uid; I -ovine blood plasma; J -broiler caecum digesta; K -broiler ileum digesta.Injection volumes: 2-6 μl relationships between the monitoring wavelength (λ) and the LA peak area (S n ) in assayed biological samples (S n sample ) and the LA standard (S n standard ) (Czauderna andKowalczyk, 2001, 2002) indicate that in all assayed biological samples, the LA and IS peaks are pure (~100%) in the UV range from 230 to 290 nm.Detection at 259 nm provides the greatest response of the detector (i.e. S n /background) as compared with detection at a shorter wavelength (λ≤244 nm).Moreover, LA and, particularly AA, PA and IS monitoring at a short UV wavelength (λ≤225 nm) resulted in a signifi cant decrease in the signal-to-noise ratio and/or an increase in interference from endogenous species present, particularly in rumen fl uid, caecum and ileum digesta, silage, kefi r and fl uids from ensiled cucumbers or cabbage.

Comparison of the new derivatization method with the ultra-effi cient derivatization method
The key step in the ultra-effi cient derivatization method is complete dissolution of a high amount of derivatizing reagent (20 g DBAP in 100 ml of chloroform) and forming derivatives of LA, AA, PA and IS as well as all endogenous substances present in analysed biological samples.Fortunately, chloroform is an excellent solvent for the derivatizing reagent, acid derivatives, and endogenous substances of assayed biological samples containing trace or high concentrations of LA.As can be seen from detailed chromatographic studies of the ultra-effi cient derivatization method, particularly excellent fractionation of LA and IS from endogenous species of samples was achieved using the proposed binary gradient elution program (Table 1).Based on the UV spectra of assayed acid derivatives formed in the chloroform medium, we can conclude that the ultra-effi cient derivatization method was also found to produce acid derivatives possessing high molar absorption with a maximum at 259 nm.Indeed, excellent, nearly symmetrical LA, AA, PA and IS peak shapes, and satisfactory separation of these derivatives from other endogenous substances were observed with sample elution times even up to 24 min (Figure 2).Moreover, LA, and IS derivative peak areas in chromatograms of samples processed using the ultra-effi cient derivatization method are considerably larger than the peak areas of these acids in samples processed using the current new derivatization method (Figure 2, Table 3).Considering the above, it can be concluded that a greater yield of acid derivatives can be achieved using the ultraeffi cient derivatization method compared with the new derivatization method.
The reproducibility of acid derivative analyses was examined in detail with respect to the storage of processed samples at -20 o C for 24 h.The obtained results of chromatographic analyses documented that areas of LA and IS monitored at 259 nm were practically constant when samples were processed by the ultra-efficient derivatization method and warmed to only ~20 o C before HPLC analysis.Purity analyses of LA and LA peaks based on relationships between the monitoring wavelength and the LA peak area (S n ) in assayed biological samples (S n sample ) and the LA standard (S n standard ) also indicate that in the assayed biological samples the LA and IS peaks are pure (~100%) in the UV range from 242 to 278 nm.Thus, the above results documented that the insolubility and/or adsorption of acid derivatives in chloroform medium is lower than in acetone medium (i.e. the solvent used for processing samples according to the new method of derivatization).Therefore, the ultra-effi cient derivatization method provides a very accurate and sensitive (LOD: 0.014 ng/ml; LOQ: 0.046 ng/ml) analytical tool for routine determination Figure 2. Dependence of the yield derivatization of the same amounts of LA and IS standards upon the derivatization procedure: A -the ultra-effi cient derivatization method; B -the new derivatization method.Parts of a typical chromatogram for derivatized acids in 1 g of kefi r processed using the ultra-effi cient derivatization method (the chromatogram C) and the new derivatization method (the chromatogram D).All derivatized acids were separated using the binary gradient elution program (Table 1) and the UV detection at 259 nm.Injection volumes 2.5 µl 18.5 × 10 6 1 0.25 mg of LA was used for the derivatization procedure according to the new derivatization method and the ultra-effi cient derivatization method 2 the inter-day coeffi cient variations based on 3 repeated analyses -3.1% (1 g kefi r samples were stored for 24 h at -20 o C) 3 the inter-day coeffi cient variations based on 4 repeated analyses -2.7% (1 g kefi r samples were stored for 24 h at -20 o C) 4 the limit of detection (LOD) -0.014 ng/ml; the limit of quantifi cation (LOQ) -0.046 ng/ml of very low concentrations of LA in specimens of biological origin.This method is also recommended for small samples (<1g) analysed in order to quantify LA.

CONCLUSIONS
The proposed new and ultra-effi cient derivatization methods are universal and simple analytical tools for accurate and precise determination of lactic acid (LA) in particular, as well as acetic acid (AA), propionic acid (PA), in various types of biological samples.LA, AA, PA and iso-capronic acid (IS) derivatives were satisfactorily separated using widely available and inexpensive C 18 columns and a typical UV detector.The total run time of HPLC analyses including the conditioning and equilibration was 59 min.The use of derivatives possessing a chromophoric group responsible for the very high molar absorptivity make these compounds almost ideally suited for RP-HPLC with detection at a relatively transparent UV region (i.e. at 259 nm).Moreover, due to aromatic derivatization, these acids are substantially retained on the C 18 column compared with underivatized acids analysed by liquid-or gas-chromatography.The use of our liquid chromatography is more rapid, simple and selective than colorimetric or enzymatic methods, chemiluminescence biosensor assay, or titration after steam distillation.
The use of the ultra-effi cient derivatization method followed by high-resolution RP-HPLC in conjunction with selective UV detection at 259 nm offers the best accuracy and sensitivity for the determination of LA in biological samples.Considering large LA peak areas in all assayed biological samples and small injection volumes (1-6 µl), we claimed that small amounts of biological materials (~100 mg) are suffi cient for the LA analysis using the new derivatization method or particularly the ultra-effi cient derivatization method followed by RP-HPLC with DAD.

Table 1 .
The binary gradient elution program used for analysis of lactic acid, acetic acid, propionic acid and iso-capronic acid in standards and specimens of biological origin (the column temperature: 39°C)

Table 3 .
Comparison of the effectiveness of lactic acid (LA) determination in the LA standard 1 and in kefi r using the new derivatization method (S n new ) 2 and the ultra-effi cient derivatization method