Kuca(09-044).

qxd:Article template 9/1/10 9:09 AM Page 1

Journal of Chromatographic Science, Vol. 48, October 2010

The pH and Mobile Phase Composition Effects

Ochratoxin A Fluorescence at Liquid Chromatography
Vlastimil Dohnal1,2,3, Lucie Pavlíková1, and Kamil Kuca2,3,*
1Department of Food Technology, Faculty of Agronomy, Mendel University of Agriculture and Forestry in Brno, Zemedelska 1, 613 00 Brno,
Czech Republic; 2Department of Chemistry, Faculty of Sciences, J.E. Purkinje University, Ceske mladeze 8, 400 96, Usti nad Labem, Czech
Republic; and 3Department of Toxicology, Faculty of Military Health Sciences, 500 01 Hradec Kralove, Czech Republic

Downloaded from https://academic.oup.com/chromsci/article/48/9/766/432171 by guest on 03 September 2022

Abstract probably one of the causes of Balkan endemic nephropathy dis-
ease (4). The suggestions about the possibility of provocation of
Changes in the fluorescence behavior of fungal toxic metabolite Parkinson’s disease symptoms also appeared (5). The toxicity of
ochratoxin were studied. The influence of ratio aqueous/organic OTA was illustrated in literature (3,6). Joint FAO/WHO Expert
solvent, the ionic composition of solution, and pH were Committee on Food Additives (JECFA) recommended provisory
investigated. The chromatographic separation under conditions tolerable daily intake to 100 ng/kg per week (7). Long-term expo-
yielding the highest fluorescence intensity were tested and sition of rats to OTA caused kidney and liver tumor induction.
compared with results obtained under commonly used Based on these facts, International Agency for Research on
chromatographic methods. The developed method was applied on Cancer (IARC) classified OTA as “potentially carcinogenic” (Class
analysis of spiked and naturally contaminated cereal samples. 2B). Moreover, in vivo and in vitro experiments demonstrated its
ability to cause DNA damage (8).

Introduction

Ochratoxin A (OTA, Figure 1) belongs to a family of toxic prod-

ucts of secondary metabolism produced by various fungi of
Aspergillus and Penicillium species. It is one of the most widely
occurring mycotoxins contaminating feed and food, such as
cereals, spices, wine grapes, etc. The entrance of OTA into the
food chain is possible in all its parts. Under proper conditions for
fungi growth, the feed or food of plant origin can be contami-
nated by OTA in each step of its production: on the field, during
harvest, transportation, and storage. The optimal temperature
for OTA production is 24°C (1), but there were reported cases
that OTA was produced at temperatures as low as 5°C (2). In
addition, meat and meat products made from animals fed with
OTA-containing feed are other sources of OTA and its metabo-
lites. Similarly to other mycotoxins, the appearance of OTA in
food can be earlier than the visual evidence of fungal infection.
OTA has a toxic effect on animal and human organisms. Toxicity
is related mainly to proteosynthesis inhibition. OTA (derivative
of isocumarin and L-β-phenylalanine) is competitive substrate to
amino acid phenylalanine in phenylalanine-t-RNA catalyzed
reactions due to its similarity with phenylalanine.
Acute toxicity of OTA is relatively low [LD50 (mouse) = 46–58
mg/kg of body weight]. Intoxication is accompanied by internal
bleeding and in certain cases liver and kidney necrosis. From
dietary point of view, the chronic toxicity is more significant
where nephrotoxic, hepatotoxic, teratogenic, and immunotoxic
effect is registered (3). Chronic exposition of human to OTA is
Figure 1. Chemical structure of ochratoxin A.
*Author to whom correspondence should be addressed.

766 Reproduction (photocopying) of editorial content of this journal is prohibited without publisher’s permission.
Kuca(09-044).qxd:Article template 9/1/10 9:09 AM Page 2

Journal of Chromatographic Science, Vol. 48, October 2010

With regards to cases of chronic exposition of OTA from con- 2.8 ng/mL. The mixture was injected into flow of water (flow rate
taminated food, there is a need to develop new faster and more 0.2 mL/min) and delivered to fluorescence detector. Then fluo-
sensitive methods for its determination. Chromatographic rescence spectra were collected in the range λex = 200–420 nm
methods, such as high-performance liquid chromatography (step 5 nm) and λem = 300–500 nm (step 5 nm). The bandwidth
(9–14) or thin-layer chromatography (15,16), are the most used of monochromator was 20 nm. Fluorescence spectra were evalu-
in OTA determination. Fluorescence detection is still preferred ated, and excitation and emission wavelength yielding maximum
beyond the more expensive mass spectrometry (17). fluorescence intensity were recorded.
Fluorescence intensity and sensitivity of analytical analysis
strongly depends on solution composition, such as concentra- Calibration solutions
tion and type of ions, pH, aqueous-to-organic phase ratio, Calibration solutions were prepared fresh from stock solution
cyclodextrin addition (10), terbium (III) ion presence (13), etc. In of OTA by dilution with an appropriate amount of mobile phase.
this work the influence of commonly used buffers, phosphate All solutions were stored at 4°C in dark place until the analysis.
and acetate, was investigated with relation to fluorescence inten- Spiked samples were prepared by addition of pure mycotoxin
sity of OTA. standard into wheat flour samples and stored for one day. After

Downloaded from https://academic.oup.com/chromsci/article/48/9/766/432171 by guest on 03 September 2022

this period, spiked samples were analyzed.

Samples of wheat flour

Experimental Twenty samples of milled wheat were collected during harvest
in 2008 in various parts of the Czech Republic.
Chemicals
The standard OTA solution in acetonitrile with concentration OTA extraction
10.04 ± 0.14 µg/mL and acetonitrile (LC-gradient grade) were The extraction of OTA and subsequent extract purification
obtained from Sigma Aldrich (Prague, Czech Republic). procedure described (21) was applied. Briefly, 20 g of sample was
Phosphoric acid, p.a. (85%), was supplied by Fluka; acetic acid, extracted by mixture of 144 mL of acetonitrile and 16 mL of 4%
p.a. (99–100%), and potassium hydroxide, p.a. was supplied by aqueous solution of potassium chloride acidified by 0.32 mL of
Riedel-de Haën (Seelze, Germany). Aqueous solution of 33% concentrated sulphuric acid. Next, the suspension was filtrated
ammonia, p.a. was obtained from Lachema (Zagreb, Croatia). through paper filter Whatman no. 4 and extracted for 1 min by
Demineralized water was prepared using Demiwa 10 ros (Watek, 100 mL of hexane. This step was repeated two times. Then, to the
s.r.o., Ledec nad Sázavou, Czech Republic). Standard solution of lower (aqueous) phase was added 20 mL of chloroform and
OTA was stored at –12°C. The working solutions were prepared shaken for 20 min. The chloroform phase was removed, and the
freshly before experiments. upper phase re-extracted two times.
Joined chloroform extracts were alkalinized and extracted
Apparatus three times with 50 mL of 5% sodium bicarbonate and shaken
pH meter pH526 (WTW, Weilheim, Germany) was used for for 10 min. Subsequent aqueous phases were combined and
control of mobile phase pH. Calibration buffers with pH 4, 7, and acidified to pH 1.5 with concentrated hydrochloric acid and
10 were supplied by Radiometer Analytical (Lyon, France). allowed to stand 20 min. This solution was three times extracted
Fluorescence spectra were recorded on HPLC chromatograph by chloroform (100, 50, and 50 mL). Chloroform extracts were
Agilent 1100 series (Agilent, Palo Alto, CA) consisting of vacuum pooled and evaporated near to dryness under vacuum on rota-
degasser (model G1322A), quaternary pump (G1311A), autosam- tory evaporator at 40°C. Two milliliters of methanol were added,
pler unit (G1313A), variable-wavelength detector (G1314A), flu- solution transferred to 1.8-mL vial, and evaporated to the dry-
orescence detector (G1321A), and mass spectrometer ness under nitrogen flow. Prior the analysis, the sample was dis-
(G1946VL). Nitrogen generator model NM18LA was delivered by solved in 500 µL of methanol.
Peak Scientific Ltd. (Renfrewshire, Scotland). Chromatographic
columns Zorbax Extend C18 (3.0 × 100 mm, 3.5 µm) (Agilent)
and ODS-Hypersil C18 (250 × 4.6 mm, 5 µm) (Supelco,
Bellefonte, PA) were used for chromatographic separations. Results

Preparation of solutions The combinations of two cations (potassium and ammonia),

Acetate and phosphate buffer solutions were prepared with two anions (phosphate and acetate), three pH values (3.0, 5.0,
concentration 83 mmol. pH value was adjusted by titration with and 9.0), and five acetonitrile-buffer ratios were prepared and
potassium hydroxide or ammonium solution to obtain pH 3.0, measured. Prior measurement the solutions were degassed by
5.0, and 9.0, respectively. All buffers were degassed by nitrogen bubbling with nitrogen to minimize molecular oxygen content
stream. that is a well-known fluorescence quencher.
The measured solutions were mixed directly in the chromato-
graph using a programmed autosampler unit and stop-flow Excitation/emission wavelength
method. Volume of 1 µL of standard OTA solution was mixed Two excitation wavelengths yielding maximum fluorescence
with an appropriate amount of acetonitrile and buffer to give 50 were found. The first one showed excitation maximum at 230
µL. Final OTA concentration in measured samples was 200.8 ± nm, and it was independent on pH, buffer composition, and ace-

767
Kuca(09-044).qxd:Article template 9/1/10 9:09 AM Page 3

Journal of Chromatographic Science, Vol. 48, October 2010

tonitrile-buffer ratio. The second excitation maximum was pH- at 230 nm with the exception of phosphate-acetonitrile ratio
dependent and observed at either 335 nm or 380 nm. Otherwise, 14:86. Acetates had significantly higher fluorescence emissions
the maximum emission wavelength was pH-dependent. While for acetate adjusted with potassium hydroxide than ammonium
solutions with pH 3.0 showed fluorescence maximum emission at 380 nm. Phosphates buffers adjusted with potassium
at 460 nm, the samples with higher pH (5.0 and 9.0) did it at hydroxide had 50% higher fluorescence intensity than those
445 nm. adjusted with ammonium in solutions with high aqueous por-
OTA is presented in two forms at pH 3.0, where the non-disso- tion content. This behavior is changing in solution with higher
ciated one is predominant (96%) and dissociated at carboxylic acetonitrile content.
group of phenylalanine moiety is in minority (4%) (Figure 2).
Influence of acetonitrile
Selection of pH and composition of buffer The fluorescence intensity of most of measurements per-
There were not significant differences between OTA fluores- formed was constant or decreased with increasing acetonitrile
cence intensity in buffers with pH 3.0. No influence of phos- content. Also, this rule is applicable for water-acetonitrile mix-
phate, acetate, potassium, ammonium, and oxonium ions were tures. There were some abnormalities registered at buffer-ace-

Downloaded from https://academic.oup.com/chromsci/article/48/9/766/432171 by guest on 03 September 2022

observed. tonitrile ratio 50:50. At pH 5.0 and 9.0, the fluorescence
On the contrary, a very important role in fluorescence inten- significantly increased for ammonium phosphate when λex was
sity plays an ionic composition of measured solutions adjusted to 230 nm. The same phenomenon was observed for this buffer also
pH 5.0 and higher. From a chemical point of view OTA is a deriva- at pH 9.0 and λex 380 nm. The highest fluorescence emission
tive of cumarine moiety amide linked to L-β-phenylalanine. It is from all tested solutions was obtained for ammonium phosphate
weak acid with dissociation constant pKa approximately equal to adjusted to pH 5.0 with 50% of acetonitrile and λex was 230 nm,
4 for carboxyl group of phenylalanine and 7 for phenolic groups. respectively, with 2% of acetonitrile and λex 380 nm.
The fluorescence emitted at pH 5.0 is a sum of fluorescence of
80% OTA1- and 20% of OTA2-. Generally, it was observed that in HPLC measurement
the case of phosphate buffers the fluorescence intensity is higher There were performed HPLC measurements under conditions
for ammonium ions in comparison with potassium ones. The that yield the highest fluorescence intensity. Chromatographic
fluorescence emissions of acetate buffers were more complex. column with stability extended to alkaline pH range was used,
When higher excitation energy is used (230 nm), the cation and flow rate was set to 0.4 mL/min. Generally, all measurements
influence is unremarkable instead acetate-acetonitrile ratio at λex 230 nm were very noisy, probably due to traces of impuri-
equal to 14:86, where fluorescence of potassium containing ties in solvents and OTA standard solution. For measurement,
buffer is about 80% higher in comparison with ammonium one. there were selected excitation wavelengths 335 (pH 5.0) respec-
Lower excitation energy (335 nm) did not show differences in tive 380 nm (pH 9.0) and emission 460 respective 445 nm. The
solutions with near to 100% of aqueous portion. With increasing results showed that the retention time of OTA in phosphate
content of organic modifier (acetonitrile), the fluorescence of ammonium buffer increases with increasing aqueous portion in
ammonium containing buffers decreased, which is not corre- mobile phase. For example, 83% of phosphate buffer content in
lated with potassium ones, where the influence was independent mobile phase resulted in retention time greater than 56 min.
with exception of near 100% acetonitrile solutions. With decreasing amount of buffer portion, the fluorescence
At alkaline pH 9.0 are carboxylic and phenolic groups of OTA intensity of OTA was lowered significantly. The retention time of
near to complete dissociated, and OTA undergoes a phototau- OTA was lower at pH 9.0 then at pH 5.0. With 83% of potassium
tomerization that results in an excitation and emission wave- acetate in mobile phase, its retention time was 7.2 min (Figure
length shift. Fluorescence intensity of both acetate and 3). Higher aqueous solution-acetonitrile ratio in extremely long
phosphate buffers had similar tendency and values at excitation retention times.

Figure 3. Chromatogram of OTA obtained under the conditions yielding the

highest fluorescence: 83 mM potassium acetate pH 9.0–acetonitrile (83:17,
Figure 2. Distribution diagram of ochratoxin A.
v/v) at a flow rate of 0.4 mL/min.

768
Kuca(09-044).qxd:Article template 9/1/10 9:09 AM Page 4

Journal of Chromatographic Science, Vol. 48, October 2010

Discussion The detection limit reached for samples with pure OTA was 0.3
ng/mL. This is comparable with results obtained by other
For buffer solutions, a concentration of 83 mM anion (acetate authors, which used pH 9.8 and 20 mM NH4Cl/NH3 buffer (20).
respective phosphate) was selected, which is the most often used Finally, the developed method was applied for analysis of
in OTA determination by high-performance liquid chromatog- spiked cereal samples, and obtained results were compared with
commonly used method (22). Reference method uses ODS-
Hypersil C18 reversed-phase column (Supelco; 250 × 4.6 mm, 5
raphy (18). OTA shows excitation maximum at 335 nm in acidic
solutions (pH 3.0 and 5.0) and 380 nm in alkaline buffers. All
these observations can be explained by phototautomerization µm). The mobile phase flow rate was 1 mL/min, and isocratic
(19, 20). OTA is a weak acid with dissociation constant pKa equal elution with a mixture of acetonitrile–water–acetic acid (99:99:2;
approximately to 4.4 for carboxyl group of phenylalanine and 7.1 v/v/v) was applied. The sample injection volume was 50 µL. OTA
for phenolic group (20). The measured fluorescence spectra at was detected by fluorescence detector with excitation wave-
pH 3.0 are mostly of non-dissociated OTA. The dissociated OTA length of 333 nm and emission wavelength of 477 nm.
undergoes a structural change in the excited state (phototau- While the standard method has the detection limit for OTA
tomerization) that increases conjugation to generate the red- determination in real samples 2.3 µg/mL, the newly developed

Downloaded from https://academic.oup.com/chromsci/article/48/9/766/432171 by guest on 03 September 2022

shifted emission spectrum. Phenolic system of cumaric moiety one was four times lower (0.6 µg/mL).
can form intramolecular hydrogen bonds and undergo an
excited state intramolecular proton transfer (ESIPT) (19).
Conclusion
Table I. Fluorescence Emission of Ochratoxin A*
The influence of buffer composition and pH used as mobile
Acetonitrile-83 mM buffer ratio chromatographic phase on OTA fluorescence was investigated.
Buffer composition 2:98 13:87 50:50 86:14 100:0 Two excitation maximums were observed. While the excitation at
pH anion cation LU LU LU LU LU wavelength 230 nm was pH-independent, the second one at 335
3.0 phosphate NH4+ 18.5 18 16.5 20.8 nm changed in alkaline pH to 380 nm due to the structural
3.0 phosphate K+ 19.8 16.5 14.6 24.6 changes in the excited state of OTA molecule. There was a trend
3.0 acetate H+ 10.9 16.9 15.1 19.6 of decreasing intensity of most tested solutions with increasing
5.0 phosphate NH4+ 23.1 19.7 50.5 34.6 acetonitrile content with certain exceptions. The buffers with the
5.0 phosphate K+ 16 11.8 16.9 20
5.0 acetate NH4+ 17.7 16.6 11.3 7.2
highest fluorescence emission were evaluated. The best chro-
35 matographic parameters showed 83 mM potassium acetate solu-
5.0 acetate K+ 18.1 18.1 19 7.6
9.0 phosphate NH4+ 25.4 23.6 34.8 13.9 tion at pH 9.0, where the detection limit 0.3 ng/mL was reached.
9.0 phosphate K+ 26.6 26.8 24 11.2 Developed method was compared with standard method, and
9.0 acetate NH4+ 23.9 23.5 21.6 8.1 the sensitivity was demonstrated. The method can be applied for
9.0 acetate K+ 30 27.4 19.9 11
5.5 water H+ 32.1 28.9 25.3 5.8
sensitive determination of OTA in samples of food or environ-
mental origin.
* Excitation 230 nm, emmision 460 (pH 3.0) respective 445 nm (pH 5.0; 9.0 and
100% acetonitrile).

Acknowledgment
Table II. Fluorescence Emission of Ochratoxin A*
This work was supported by Internal Grant Agency of Mendel
Acetonitrile-83 mM buffer ratio University of Agriculture and Forestry in Brno number
Buffer composition 2:98 13:87 50:50 86:14 100:0 MP32/AF-2008.
pH anion cation LU LU LU LU LU

3.0 phosphate NH4+ 10.5 10.3 9.3 11.6

3.0 phosphate K+ 11.1 9.2 8.3 12.5
3.0 acetate H+ 9 9 8 10.1 References
5.0 phosphate NH4+ 42.9 10.6 18.6 13.7
5.0 phosphate K+ 9.4 5.8 9.7 10.7 1. V. Betina. Mykotoxíny: chémia, biológia, ekológia, Alfa (1990),
5.0 acetate NH4+ 10.2 9.7 6.3 3.8 22.9 ISBN 80-05-00631-4, Bratislava, Czechoslovakia.
5.0 acetate K+ 10.5 11 11.3 3.7 2. International Program of Chemical Safety. Selected Mycotoxins:
9.0 phosphate NH4+ 22.9 21.9 32.5 5.5 Ochratoxins, tricothecenes, ergot. Environmental Health Criteria
9.0 phosphate K+ 33.2 31.9 26.7 4.8 105. World Health Organisation, Geneva, 1990. http://
9.0 acetate NH4+ 23.3 19.5 14.7 4.4 www.inchem.org/documents/ehc/ehc/ehc105.htm.
9.0 acetate K+ 35.6 27.1 20.2 10.5 3. A. Pfohl-Leszkowicz and R.A. Manderville. Ochratoxin A: An
5.5 water H+ 42.4 36.2 27.7 4.5 overview on toxicity and carcinogenicity in animals and humans.
Mol. Nutr. Food Res. 51: 61–99 (2007).
* Excitation 335 (pH 3.0 and 5.0) respective 380 nm (pH 9.0), emmision 460 nm 4. D. Puntaric, J. Bosnir, Z. Smit, and Z. Baklaic. Ochratoxin A in corn
(pH 3.0 and 5.0), or 445 nm (pH 9.0 and 100% acetonitrile).
and wheat: Geographical association with endemic nephropathy.
Croat. Med. J. 42: 175 (2001).

769
Kuca(09-044).qxd:Article template 9/1/10 9:09 AM Page 5

Journal of Chromatographic Science, Vol. 48, October 2010

5. V. Sava, O. Reunova, A. Velasquez, and J. Sanchez-Ramos. Can low performance liquid chromatography with enhanced fluorescence
exposure to ochratoxin A cause parkinsonism? J. Neurol. Sci. 249: detection and immunoaffinity column cleanup: methodology and
68–75 (2006). Swiss data. J. Chromatogr. B 666: 85–99 (1995).
6. F.N. Kotsinos and G.A. Burdock. Food Toxicology. In Klaassen, C.D. 14. S. Kumagasi and K. Airaba. Intestinal absorption and secretion of
(ed.) Casarett & Doull’s Toxiology, The Basic Science of Poisons, 7th ochratoxin A in the rat. Appl. Pharmacol. 64: 94–102 (1982).
Edition, McGraw Hill Medical, New York, pp. 1191–1236 (2008). 15. N.B. Büchmann and B. Hald. Analysis, occurrence and control of
7. JECFA. Evaluation of Certain Mycotoxins in Food. (56th Report of Ochratoxin A residues in Danish pig kidneys. Food Addit. Contam.
the Joint FAO/WHO Expert Committee on Food Additives). 2: 193–199 (1985).
Technical Reports Series No. 906, 2002. 16. S. Sforza, C. Dall’Asta, and M. Rosangela. Recent advances in
8. IARC: Ochratoxin A. In: IARC Monographs on the Evaluation of mycotoxin determination in food and feed by hyphenated chro-
Carcinogenic Risk of Chemicals to Humans, IARS Press Vol. 56, matographic techniques/mass spektrometry. Mass Spectrom. Rev.
Lyon, France, pp. 489–521 (1993). 25: 54–76 (2006).
9. P. Galtier, M. Alvinerie, and J.L. Charpenteau. The pharmacokinetic 17. H. Reinhard and B. Zimmerli. Reversed-phase liquid chromato-
profiles of ochratoxin A in pigs, rabbits and chickens. Food Cosmet. graphic behavior of the mycotoxins citrinin and ochratoxin A.
Toxicol. 19: 735–738 (1981). J. Chromatogr. A 862: 147–159 (2001).
10. V. Seidel, E. Poglits, K. Schiller, and W. Lindner. Simultaneous deter- 18. C. Frenette, R.J. Paugh, M. Tozlovanu, M. Juzio, A. Pfohl-
mination of ochratoxin A and zearalenone in maize by reversed- Leszkowicz, and R.A. Manderville. Structure-activity relationships

Downloaded from https://academic.oup.com/chromsci/article/48/9/766/432171 by guest on 03 September 2022

phase high-performance liquid chromatography with fluorescence for the fluorescence of ochratoxin A: insight for detection of ochra-
detection and beta-cyclodextrin as mobile phase additive. toxin A metabolites. Anal. Chim. Acta 617: 153–161 (2008).
J. Chromatogr. 635: 227–235 (1993). 19. C. Dall’Asta, G. Galaverna, A. Dossena, and R. Marchelli.
11. T. Toscani, A. Moseriti, A. Dossena, C. Dall’Asta, N. Simoncini, and Reversed-phase liquid chromatographic metod for the determina-
R. Virgili. Determination of ochratoxin A in dry-cured meat prod- tion of ochratoxin A in wine. J. Chromatogr. A 1024: 275–279
ucts by a HPLC–FLD quantitative Metod. J. Chromatogr. B 855: (2004).
242–248 (2007). 20. A. Molinié, V. Faucet, M. Castegnaro, and A. Pfohl-Leszkowicz.
12. H. Valenta. Chromatographic methods for the determination of Analysis of some breakfast cereals on the French market for their
ochratoxin A in animal and human tissues and fluids. J. Chromatogr. contents of ochratoxin A, citrinin and fumonisin B1: development
A 815: 75–92 (1998). of a method for simultaneous extraction of ochratoxin A and
12. B.I. Vazquez, C. Fente, C. Franco, A. Cepeda, P. Prognon, and citrinin. Food Chem. 92: 391–400 (2005).
G. Mahuzier. Simultaneous high-performance liquid chromato- 21. F. Karbancıoglu-Güler and D. Heperkan: Natural occurrence of
graphic determination of ochratoxin A and citrinin in cheese by ochratoxin A in dried figs. Anal. Chim. Acta 617: 32–36 (2008).
time-resolved luminiscence using terbium. J. Chromatogr. A 727:
185–193 (1996).
13. B. Zimmerli and R. Dick. Determination of ochratoxin A at the ppt Manuscript received February 17, 2009;
level in human blood, serum, milk and some foodstuffs by high- revision received March 31, 2009.

770