- Email: [email protected]
High Pressure Liquid Chromatographic Determination of Penicillic Acid in Chicken Tissues
High Pressure Liquid Chromatographic Determination of Penicillic Acid in Chicken Tissues GARY D. HANNA, TIMOTHY D. PHILLIPS, 1 LEON F. KUBENA,2 SIGMUND J. CYSEWSKI,2 G. WAYNE IVIE, 2 NORMAN D. HEIDELBAUGH, DONALD A. WITZEL,2 and A. WALLACE HAYES Texas A&M University, Department of Veterinary Public Health, College Station, Texas 77843 and US Department of Agriculture, Science and Education Administration, Veterinary Toxicology and Entomology Research Laboratory, College Station, Texas 77841 (Received for publication March 3, 1981)
1981 Poultry Science 60:2246-2252 INTRODUCTION
Penicillic acid (PA, 3-methoxy-5-methyl-4oxo-2,5-hexadienoic acid) was first reported from corn contaminated with the mold Penicillium puberulum by Alsberg and Black (1913). Subsequently, PA has been shown to be a metabolite of other Penicillium and Aspergillus species (Wilson, 1976) with a "digitalis-like" action on cardiac muscle and a concurrent hypertensive effect when injected into experimental animals (Murnaghan, 1946). Also, PA reacts with sulfhydryl groups of glutathione and cysteine and the amine groups of lysine, histidine, and arginine (Ciegler et al., 1972). It acts as an enzyme poison by binding to the essential thiol groups of urease (Reiss, 1979) and by selectively inhibiting the (Na and K )dependent adenosine triphosphatase activity of membranes (Chan et al, 1979; Phillips et al, 1980). It has been reported that PA induces tumors at the site of subcutaneous injection in
'Correspondence to: Dr. Timothy D. Phillips. 2 Veterinary Toxicology and Entomology Research Laboratory.
rats (Dickens and Jones, 1961) and is cytotoxic in cultured cells from liver, kidney, and lung (Umeda, 1971). When compared to other mycotoxins such as aflatoxin B i , PA has minimal toxicity in the chicken (Huff et al., 1980); however, its potential occurrence in very high concentrations in corn (Ciegler and Kurtzman, 1970; Scott, 1978) and its occurrence in poultry feed (Bacon et al., 1973) emphasize the importance of this mycotoxin as a potential public health hazard. A rapid and sensitive method for detecting PA residues in the chicken is a critical need. This paper describes a method for extraction and analysis of PA (utilizing HPLC) and its application for detection of PA in tissues and biological fluids of the chicken. MATERIALS AND METHODS
Chemicals. All solvents were distilled in glass (Burdick and Jackson, Muskegan, MI). The PA was purchased from Makor Chemicals Ltd., Jerusalem, Israel. The purity of PA (99 to 100%) was confirmed by TLC, HPLC, GLC/ mass spectrometry, and nuclear magnetic resonance spectrometry. All other chemicals
2246
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
ABSTRACT Penicillic acid (PA) is a mycotoxin with reported cytotoxic, cardiotoxic, and carcinogenic activity and it can occur in high concentrations in corn. The occurrence of PA in contaminated poultry feed represents a potential public health hazard. A reverse phase high pressure liquid chromatographic (HPLC) method is proposed for determining PA residues in chicken tissues. Optimization of chromatography was achieved for PA using a mobile phase consisting of acetonitrile:H 2 0. PA was detected by ultraviolet absorption at 254 nm, identified by retention time, and quantitated by peak area integration. Blood, parenchymal tissues, muscle, and alimentary tract contents were homogenized, sonicated, and acid treated followed by extraction with ethyl acetate and analysis by HPLC. Acute oral dosing of chickens with PA over a range of 50 to 550 mg/kg body weight resulted in detectable levels of the mycotoxin (confirmed by gas liquid chromatography) in gizzard muscle and contents, liver, kidney, heart, and intestinal contents. This method should prove useful both for the rapid and sensitive detection of PA residues in poultry and in further studies on the distribution and metabolism of this mycotoxin. (Key words.- HPLC, penicillic acid, mold, mycotoxin, chicken tissue, blood)
PENICILLIC ACID IN CHICKEN TISSUES
100
with a sonic dismembrator (Fisher Scientific). An equal volume of 3 N HC1 was added to the homogenate with mixing and the resulting solution was incubated at 80 C for 15 min. The PA was extracted by the addition of equal volumes of ethyl acetate, with subsequent mixing and centrifugation for 5 min at 3,000 X g. The ethyl acetate layer was collected and the homogenate extracted a second time. The two extracts were pooled and evaporated to dryness under nitrogen (N 2 ) in a fume hood and the residue was dissolved in 1 ml of acetonitrilewater 60:40 (v/v) and washed twice with 2 ml of hexane. At this point, the extracts, except those from liver, were analyzed by HPLC. Liver samples, which required additional cleanup, were extracted with benzene to recover the PA. The benzene was evaporated under N 2 in a fume hood and the residue was redissolved in .5 ml of acetonitrile-water 60:40 (v/v) and analyzed by HPLC. Detection. The PA from tissue extracts was detected and analyzed with a Waters Model ALC-204 HPLC (Waters Associates, Milford,
AUTHENTIC PENICILLIC ACID
.01 AUFS
UJ
co
a 50
O oc
Injection
J v. RETENTION TIME (min)
FIG. 1. The HPLC resolution of authentic penicillic acid; 500 ng injected; fi Bondapak C,8 column; ultraviolet detection at 254 nrri; elution solvent system acetonitrile-water (60:40); flow-rate 1.0 ml/min.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
were from Sigma Chemical Co. (St. Louis, MO). Dosing and Tissue Collection. Seven-day-old, male laying strain chickens (Hyline W-36) were obtained from a local hatchery. The birds were housed in electrically heated batteries with feed (a commercial starter ration free of medications) and water available ad libitum. The PA was dissolved in sterilized double distilled, demineralized water and was administered in dosing experiments at levels of 50, 100, 200, 400, and 5 50 mg/kg body weight by intubation into the crop. Control chickens were dosed with equal volumes of solvent vehicle. All chickens were sacrificed by cervical dislocation 4 hr postdosing with PA, and kidneys, heart, gizzard muscle and contents, breast and thigh muscle, liver, and intestinal contents were collected, weighed, and frozen for subsequent PA analysis. Blood was collected in heparinized and nonheparinized tubes by cardiac puncture or rupture of the jugular vein. Tissue Extraction. Tissue samples were homogenized in distilled water with a blender, and the homogenate was sonicated for 1 min
2247
HANNA ET AL.
2248
HPLC was confirmed by gas liquid chromatography (GLC) employing methods recently described (Phillips et al., 1981). Briefly, PA peaks were collected and evaporated to near dryness under N 2 at 60 C. Residual water was removed by azeotrophing with excess acetonitrile under a steady stream of N 2 , and PA was converted to a pyrazoline derivative by addition of excess diazomethane to the dried sample with mixing for 15 to 20 min. (An ethereal ethanolic solution of diazomethane can be prepared from Diazald (iV-methyl-/V-nitroso-p-toluene sulfonamide); Aldrich Chemical, Milwaukee, Wl.) Reaction of PA with diazomethane results in a product having a shifted HPLC retention time and an increased ultraviolet absorbance, both of which represent properties useful in confirmation. Thus, derivatized PA residues were confirmed by rechromatographing on HPLC and by GLC utilizing a Varian Model 3700 equipped with a flame ionization detector and a
Weigh Tissue Samples
i Homogenize and Sonicate
i Hydrolyze with 3 N HCI 15 min at 80° C
Extract PA with Et Ac (twice) Pool Et Ac
i
Evaporate Et Ac under N,
i i Wash with Hexane (twice)
-•
Chromatograph on HPLC Reverse phase chromatography Mobile phase CH 3 CN-H 2 0 60:40(v/v)
Partition into Benzene (Liver extracts)
i Evaporate Benzene under N
i
Chromatograph (HPLC) FIG. 2. Diagrammatic illustration of the steps employed in sample preparation and extraction of penicillic acid from chicken tissues.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
MA) equipped with a Model 440 uv detector and 254 nm primary filter, U6K septumless injector, M-6000 and M-45 pumps, 660 solvent programmer, and a reverse phase octadecylsilane column (ju Bondapac). Chromatography was optimized for PA using a mobile phase of acetonitrile-water 60:40 (v/v) at a flow rate of 1 to 2 ml/min, ambient temperature, 10 /il injection volumes, and detector sensitivities from .01 to .005 absorbance units full scale (AUFS). Chromatograms were recorded on a Houston Instruments Series B-5000 Omni Scribe recorder (Austin, TX) and peak areas determined by digital integration using a Columbia Scientific Model CSI38 integrator (Austin, TX). Calibration curves were constructed using standard solutions of PA (1 mg/ml) made up fresh each day in mobile phase and injected over a range of 10 to 1000 ng/10 Ml Confirmation. Residual PA detected by
2249
PENICILLIC ACID IN CHICKEN TISSUES
20
30
40
50
60
HYDROLYSIS TWE (mil)
FIG. 3. Recovery of penicillic acid from spiked chicken serum and whole blood (heparinized) as a function of time of acid treatment (3N HC1 at 80 C). Values represent the mean ± SEM of three individual extractions. See text for details of analysis.
3% OV-101 chrom W-HP column (operating conditions: flow rate, 30 cc/min; column temperature, 175 C isothermal; injector temperature, 250 C; electrometer sensitivity, 10~ 12 amps/mV; chart range, 10 mV). RESULTS Penicillic acid was resolved as a sharp peak by reverse phase HPLC (Fig. 1). Retention times, peak heights, and peak areas were reproducible over a wide range of PA concentrations
100
.005 AUFS
GIZZARD
I HI
w z
o Q.
Aci
CO
£50Injection A
z
1
LU O
CC LU
\
il
0 0
-*n TISSUE SOURCE
FIG. 4. Tissue and alimentary tract levels of penicillic acid (detected by HPLC and confirmed by GLC) from chickens 4 hr postdosing with 550 mg/kg body weight. Values represent mean levels of PA (Mg/g) ± SEM of duplicate replicates each assayed in triplicate.
1
A £
\
\u / V y n v ? - ^ r t W
r
In ection B
o
IN
h-
J
1
/
\
A£ A A
k
W^l/l
1 2 3 4 RETENTION TIME (min)
FIG. 5. Representative HPLC chromatograms of gizzard extracts from a control chicken (A) and a chicken dosed with penicillic acid (B). For comparison purposes, the control HPLC tracing is superimposed over the chromatogram from the chicken treated with penicillic acid. See text for HPLC conditions.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
10
(Phillips et al, 1981). Mean retention times for PA at flow rates of 2 ml/min and 1 ml/min were 94 and 188 sec, respectively. At maximum sensitivity, 5 to 10 ng of PA could be easily detected and adequately separated from interfering peaks in biological samples. Ethyl acetate was used to extract PA from tissues and blood. A second extraction with benzene was required in the case of liver to eliminate interfering peaks. Figure 2 outlines the procedure of sample preparation, extraction, clean-up, separation, and quantitation employed in this study. Recovery of PA in spiked samples was enhanced by treatment with strong acid at 80 C. The optimum time of treatment was determined by spiking samples with PA at a level of 500 ng/ml and by incubating with 1.0 ml 3N hydrochloric acid for 0, 5, 10, 15, 20, 30, and 60 min at 80 C in a water bath with shaking before extraction and HPLC analysis. Maximum recovery (85%) was obtained after 5 min treatment in serum, whereas only 26% recovery was obtained with no acid treatment (Fig. 3). In whole blood, a maximum recovery of 27% was obtained only after 15 min of acid treatment followed by a significant decrease in recovery thereafter. Maximum recovery of PA from liver, kidney, gizzard, heart, and skeletal muscle samples (80 to 105%)
HANNA ET AL.
2250
line derivative of PA from HPLC peaks and analysis by GLC as described in the text; PA was not detected in blood or skeletal muscle at any treatment level. DISCUSSION Methodologies describing extraction and analysis procedures for the mycotoxin PA are numerous. Detection methods currently available include bioassay, colorimetry, thin layer chromatography, and GLC. Recently, methods using HPLC with detection of PA at 254 nm have been reported (Engstrom et al, 1977; Chan et al, 1980; Phillips et al, 1981). Moreover, HPLC methods are generally rapid, specific, sensitive, and less laborious than most previously reported procedures but have not been utilized in determining PA in tissues. Results from our study indicate that HPLC can be effectively utilized in the primary analysis of PA in chicken tissues, following appropriate extraction and cleanup procedures. The GLC of the pyrazoline derivative formed after reaction with diazomethane can also be used for PA confirmation. The extraction of PA from tissues and blood was enhanced by acid treatment of sample homogenates at 80 C. This effect has been previously noted for PA (Chan et al, 1980) and rubratoxin B, another mycotoxin which, like PA, is reactive with sulfhydryl groups (Unger and Hayes, 1978) and may be due to increased
.01 AUFS
HEART
GIZZARD C O N T E N T S
o Q.
K 50 O oc UJ 0.
0 RETENTION TIME (mki)
FIG. 6. Representative HPLC chromatograms of gizzard content extracts from a control chicken (A) and a chicken dosed with penicillic acid (B). See text for HPLC conditions.
1
2
3
4
RETENTION TIME (min)
FIG. 7. Representative HPLC chromatograms of heart extracts from a control chicken (A) and a chicken dosed with penicillic acid (B). See text for HPLC conditions.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
was obtained within 10 to 15 min of acid treatment and was unchanged after 60 min treatment (data not shown). Chickens were initially dosed with PA at a level of 550 mg/kg body weight to test the applicability of this method for detecting PA residues from various tissue sources including gizzard and intestinal contents. Figure 4 summarizes the results obtained from this experiment. Significant levels of PA were detected 4 hr postdosing in kidney, heart, gizzard, gizzard contents, and intestinal contents. The PA was not detected in skeletal muscle or blood. Typical HPLC chromatograms comparing gizzard muscle and contents and heart extracts from both control and PA treated chickens are illustrated in Figures 5, 6, and 7. The chromatograms of tissue extracts from untreated chickens are superimposed over chromatograms of extracts from chickens dosed with PA. Table 1 summarizes the results from a subsequent experiment where chickens were dosed with PA at levels of 50, 100, 200, and 400 mg/kg body weight. At all treatment levels tested PA was detected in kidneys at concentrations ranging from 1480 to 2910 ppb, in heart at the 400 mg/kg level at a concentration of 590 ppb, in gizzard at the 100, 200, and 400 mg/kg levels at concentrations ranging from 2.8 to 19 ppb, and in liver at the 200 and 400 mg/kg levels at concentrations of 5.1 and 6.8, respectively. The presence of PA was confirmed in positive samples by formation of the pyrazo-
PENICILLIC ACID IN CHICKEN TISSUES
2251
TABLE 1. Tissue concentrations of penicillic acid in chickens detected by HPLC34 hours after administration of levels from 50 to 400 mg/kg body weight Parts per billion PA at indicated treatment levels (mg/kg)D Tissue
c
Kidney Heart Gizzard Liver Blood Muscle
50 1480 ± 490 d ND e ND ND ND ND
100 1920
200
+340 ND 2.8 ± 1.8 ND ND ND
2160
+400 ND 3.6 ± .5 5.1 ± .6 ND ND
400 2910 ±240 590 ± 30 19 ± 2.9 6.8 ± .2 ND ND
See text for HPLC conditions.
Tissues were collected 4 hours after administration of penicillic acid. Detected levels of penicillic acid are expressed as ppb (ng/g) of tissue and represent the mean ± SEM from 3 chickens each assayed in duplicate. e
Penicillic acid not detected.
precipitation of interfering protein in samples, or to release of PA which may have been bound to protein components, or a combination of both. A more quantitative recovery of PA can be achieved with this added step. In whole blood, recovery was poor even after acid treatment, possibly indicating a strong association of PA with erythrocytes or the formation of adducts which escaped detection. This conclusion is consistent with the results of a distribution study by Park and coworkers (1977) wherein 14 C-PA exhibited a very high affinity for erythrocytes. Tissue concentrations of PA determined by HPLC from experiments where chickens were dosed with levels of PA ranging from 50 to 550 mg/kg body weight indicate that residue retention was highest in kidneys, heart, gizzard, and liver. The PA was not detected in heart below a level of 590 ppb. We have no obvious explanation for this observation. Also, PA was not detected in skeletal muscle or blood. Lack of detection in blood may be explained by the absence of quantitative recovery in this tissue. A previous report (Park et al, 1977) on the distribution of PA in rats indicated that maximum levels of 14 C-PA were reached in blood and tissues within 2 to 4 hr postdosing. The organs that contained the highest concentrations of radioactivity were liver, kidneys, heart, and bladder. Levels in skeletal muscle also were very low in comparison, which is consistent with the
results we obtained in the chicken. It is evident from the literature on mycotoxins that there is a lack of information regarding residue accumulation in foods of animal origin and their significance to human health through food chain interactions. Studies on the metabolic fate of penicillic acid in food animals are also lacking. Because the consumption of poultry and poultry products has greatly increased in the United States over the last 10 years, and because PA can occur in corn and in poultry feed in high levels, the potential for contamination of edible poultry tissues with subsequent public health hazards cannot be overlooked. The HPLC method reported in this study should prove useful in monitoring of PA in poultry and other food animals, either as proposed here, or in conjunction with standard methods of analysis, as well as aid in further studies on the distribution and metabolism of this mycotoxin. ACKNOWLEDGMENTS This research was supported by Texas Agricultural Experiment Station Projects H 6215, AH 6397, and AH 6529 and the Texas A&M University, Office of University Research, ORR 2-10 and 5-81. REFERENCES Alsberg, C. L., and O. F. Black, 1913. Contribution
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
Chickens (3 per group) were orally dosed with penicillic acid dissolved in water at levels of 50, 100, 200, and 400 mg/kg.
2252
HANNA ET AL. cillic acid. J. Pharmacol. Exp. Therap. 88:119— 132. Park, D. L., R. E. Dailey, L. Friedman, and J. L. Health, 1977. The absorption, distribution, and excretion of 14 C-penicillic acid by rats. Ann. Nutr. Aliment. 31:919-934. Phillips, T. D., P. K. Chan, and A. W. Hayes, 1980. Inhibitory characteristics of the mycotoxin penicillic acid on (Na + -K + )-activated adenosine triphosphatase. Biochem. Pharmacol. 29:19— 26. Phillips, T. D., G. W. Ivie, N. D. Heidelbaugh, L. F. Kubena, S. J. Cysewski, A. W. Hayes, and D. A. Witzel, 1981. Confirmation of penicillic acid by high pressure liquid and gas-liquid chromatography. J. Ass. Offic. Anal. Chem. 64(1):162165. Reiss, J., 1979. Inhibitory action of the mycotoxins, patulin and penicillic acid on urease. Food Cosmet. Toxicol. 17:145-146. Scott, P. M., 1978. Mycotoxins in feeds and ingredients and their origin. J. Food Prot. 41(5):385— 398. Umeda, M., 1971. Cytomorphological changes of cultured cells from rat liver, kidney and lung induced by several mycotoxins. Jpn. J. Exp. Med. 41:195-207. Unger, P. D. and A. W. Hayes, 1978. High-pressure liquid chromatography of the mycotoxins rubratoxins A and B, and its application to the analysis of urine and plasma for rubratoxin B. J. Chromatogr. 153:115-126. Wilson, D. M., 1976. Mycotoxins and other fungal related food problems. Pages 90—109 in Advances in chemistry. Series No. 149. J. Rodricks, ed. Amer. Chem. Soc, Washington, DC.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
to the study of maize deterioration. US Dept. Agr. Bull. Bur. Plant Ind. 270-J. Bacon, C. W., J. G. Sweeney, J. D. Robbins, and D. Burdick, 1973. Production of penicillic acid and ochratoxin A on poultry feed by Aspergillus ochraceus-. temperature and moisture requirements. Appl. Microbiol. 26:155-160. Chan, P. K., T. D. Phillips, and A. W. Hayes, 1979. Effect of penicillic acid on adenosine triphosphatase activity in the mouse. Toxicol. Appl. Pharmacol. 49:365-372. Chan, P. K., M. Y. Siraj, and A. W. Hayes, 1980. High performance liquid chromatographic analysis of the mycotoxin penicillic acid and its application to biological fluids. J. Chromatogr. 194:387— 398. Ciegler, A., and C. P. Kurtzman, 1970. Penicillic acid production by blue-eye fungi on various agricultural commodities. Appl. Microbiol. 20:761— 764. Ciegler, A., H. J. Mintzlaff, D. Weisleder, and L. Leistner, 1972. Potential production and detoxification of PA in mold fermented sausage (salami). Appl. Microbiol. 24:114-119. Dickens, F., and H.E.H. Jones, 1961. Carcinogenic activity of a series of relative lactones and related substances. Brit. J. Cancer. 1 5 ( l ) : 8 5 - 9 5 . Engstrom, G. W., J. L. Richard, and S. J. Cysewski, 1977. High pressure liquid chromatographic method for detection and resolution of rubratoxin, aflatoxin, and other mycotoxins. J. Agric. Food Chem. 25:833-836. Huff, W. E., P. B. Hamilton, and A. Ciegler, 1980. Evaluation of penicillic acid for toxicity in broiler chickens. Poultry Sci. 59:1203-1207. Murnaghan, M. F., 1946. The pharmacology of peni-
1981 Poultry Science 60:2246-2252 INTRODUCTION
Penicillic acid (PA, 3-methoxy-5-methyl-4oxo-2,5-hexadienoic acid) was first reported from corn contaminated with the mold Penicillium puberulum by Alsberg and Black (1913). Subsequently, PA has been shown to be a metabolite of other Penicillium and Aspergillus species (Wilson, 1976) with a "digitalis-like" action on cardiac muscle and a concurrent hypertensive effect when injected into experimental animals (Murnaghan, 1946). Also, PA reacts with sulfhydryl groups of glutathione and cysteine and the amine groups of lysine, histidine, and arginine (Ciegler et al., 1972). It acts as an enzyme poison by binding to the essential thiol groups of urease (Reiss, 1979) and by selectively inhibiting the (Na and K )dependent adenosine triphosphatase activity of membranes (Chan et al, 1979; Phillips et al, 1980). It has been reported that PA induces tumors at the site of subcutaneous injection in
'Correspondence to: Dr. Timothy D. Phillips. 2 Veterinary Toxicology and Entomology Research Laboratory.
rats (Dickens and Jones, 1961) and is cytotoxic in cultured cells from liver, kidney, and lung (Umeda, 1971). When compared to other mycotoxins such as aflatoxin B i , PA has minimal toxicity in the chicken (Huff et al., 1980); however, its potential occurrence in very high concentrations in corn (Ciegler and Kurtzman, 1970; Scott, 1978) and its occurrence in poultry feed (Bacon et al., 1973) emphasize the importance of this mycotoxin as a potential public health hazard. A rapid and sensitive method for detecting PA residues in the chicken is a critical need. This paper describes a method for extraction and analysis of PA (utilizing HPLC) and its application for detection of PA in tissues and biological fluids of the chicken. MATERIALS AND METHODS
Chemicals. All solvents were distilled in glass (Burdick and Jackson, Muskegan, MI). The PA was purchased from Makor Chemicals Ltd., Jerusalem, Israel. The purity of PA (99 to 100%) was confirmed by TLC, HPLC, GLC/ mass spectrometry, and nuclear magnetic resonance spectrometry. All other chemicals
2246
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
ABSTRACT Penicillic acid (PA) is a mycotoxin with reported cytotoxic, cardiotoxic, and carcinogenic activity and it can occur in high concentrations in corn. The occurrence of PA in contaminated poultry feed represents a potential public health hazard. A reverse phase high pressure liquid chromatographic (HPLC) method is proposed for determining PA residues in chicken tissues. Optimization of chromatography was achieved for PA using a mobile phase consisting of acetonitrile:H 2 0. PA was detected by ultraviolet absorption at 254 nm, identified by retention time, and quantitated by peak area integration. Blood, parenchymal tissues, muscle, and alimentary tract contents were homogenized, sonicated, and acid treated followed by extraction with ethyl acetate and analysis by HPLC. Acute oral dosing of chickens with PA over a range of 50 to 550 mg/kg body weight resulted in detectable levels of the mycotoxin (confirmed by gas liquid chromatography) in gizzard muscle and contents, liver, kidney, heart, and intestinal contents. This method should prove useful both for the rapid and sensitive detection of PA residues in poultry and in further studies on the distribution and metabolism of this mycotoxin. (Key words.- HPLC, penicillic acid, mold, mycotoxin, chicken tissue, blood)
PENICILLIC ACID IN CHICKEN TISSUES
100
with a sonic dismembrator (Fisher Scientific). An equal volume of 3 N HC1 was added to the homogenate with mixing and the resulting solution was incubated at 80 C for 15 min. The PA was extracted by the addition of equal volumes of ethyl acetate, with subsequent mixing and centrifugation for 5 min at 3,000 X g. The ethyl acetate layer was collected and the homogenate extracted a second time. The two extracts were pooled and evaporated to dryness under nitrogen (N 2 ) in a fume hood and the residue was dissolved in 1 ml of acetonitrilewater 60:40 (v/v) and washed twice with 2 ml of hexane. At this point, the extracts, except those from liver, were analyzed by HPLC. Liver samples, which required additional cleanup, were extracted with benzene to recover the PA. The benzene was evaporated under N 2 in a fume hood and the residue was redissolved in .5 ml of acetonitrile-water 60:40 (v/v) and analyzed by HPLC. Detection. The PA from tissue extracts was detected and analyzed with a Waters Model ALC-204 HPLC (Waters Associates, Milford,
AUTHENTIC PENICILLIC ACID
.01 AUFS
UJ
co
a 50
O oc
Injection
J v. RETENTION TIME (min)
FIG. 1. The HPLC resolution of authentic penicillic acid; 500 ng injected; fi Bondapak C,8 column; ultraviolet detection at 254 nrri; elution solvent system acetonitrile-water (60:40); flow-rate 1.0 ml/min.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
were from Sigma Chemical Co. (St. Louis, MO). Dosing and Tissue Collection. Seven-day-old, male laying strain chickens (Hyline W-36) were obtained from a local hatchery. The birds were housed in electrically heated batteries with feed (a commercial starter ration free of medications) and water available ad libitum. The PA was dissolved in sterilized double distilled, demineralized water and was administered in dosing experiments at levels of 50, 100, 200, 400, and 5 50 mg/kg body weight by intubation into the crop. Control chickens were dosed with equal volumes of solvent vehicle. All chickens were sacrificed by cervical dislocation 4 hr postdosing with PA, and kidneys, heart, gizzard muscle and contents, breast and thigh muscle, liver, and intestinal contents were collected, weighed, and frozen for subsequent PA analysis. Blood was collected in heparinized and nonheparinized tubes by cardiac puncture or rupture of the jugular vein. Tissue Extraction. Tissue samples were homogenized in distilled water with a blender, and the homogenate was sonicated for 1 min
2247
HANNA ET AL.
2248
HPLC was confirmed by gas liquid chromatography (GLC) employing methods recently described (Phillips et al., 1981). Briefly, PA peaks were collected and evaporated to near dryness under N 2 at 60 C. Residual water was removed by azeotrophing with excess acetonitrile under a steady stream of N 2 , and PA was converted to a pyrazoline derivative by addition of excess diazomethane to the dried sample with mixing for 15 to 20 min. (An ethereal ethanolic solution of diazomethane can be prepared from Diazald (iV-methyl-/V-nitroso-p-toluene sulfonamide); Aldrich Chemical, Milwaukee, Wl.) Reaction of PA with diazomethane results in a product having a shifted HPLC retention time and an increased ultraviolet absorbance, both of which represent properties useful in confirmation. Thus, derivatized PA residues were confirmed by rechromatographing on HPLC and by GLC utilizing a Varian Model 3700 equipped with a flame ionization detector and a
Weigh Tissue Samples
i Homogenize and Sonicate
i Hydrolyze with 3 N HCI 15 min at 80° C
Extract PA with Et Ac (twice) Pool Et Ac
i
Evaporate Et Ac under N,
i i Wash with Hexane (twice)
-•
Chromatograph on HPLC Reverse phase chromatography Mobile phase CH 3 CN-H 2 0 60:40(v/v)
Partition into Benzene (Liver extracts)
i Evaporate Benzene under N
i
Chromatograph (HPLC) FIG. 2. Diagrammatic illustration of the steps employed in sample preparation and extraction of penicillic acid from chicken tissues.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
MA) equipped with a Model 440 uv detector and 254 nm primary filter, U6K septumless injector, M-6000 and M-45 pumps, 660 solvent programmer, and a reverse phase octadecylsilane column (ju Bondapac). Chromatography was optimized for PA using a mobile phase of acetonitrile-water 60:40 (v/v) at a flow rate of 1 to 2 ml/min, ambient temperature, 10 /il injection volumes, and detector sensitivities from .01 to .005 absorbance units full scale (AUFS). Chromatograms were recorded on a Houston Instruments Series B-5000 Omni Scribe recorder (Austin, TX) and peak areas determined by digital integration using a Columbia Scientific Model CSI38 integrator (Austin, TX). Calibration curves were constructed using standard solutions of PA (1 mg/ml) made up fresh each day in mobile phase and injected over a range of 10 to 1000 ng/10 Ml Confirmation. Residual PA detected by
2249
PENICILLIC ACID IN CHICKEN TISSUES
20
30
40
50
60
HYDROLYSIS TWE (mil)
FIG. 3. Recovery of penicillic acid from spiked chicken serum and whole blood (heparinized) as a function of time of acid treatment (3N HC1 at 80 C). Values represent the mean ± SEM of three individual extractions. See text for details of analysis.
3% OV-101 chrom W-HP column (operating conditions: flow rate, 30 cc/min; column temperature, 175 C isothermal; injector temperature, 250 C; electrometer sensitivity, 10~ 12 amps/mV; chart range, 10 mV). RESULTS Penicillic acid was resolved as a sharp peak by reverse phase HPLC (Fig. 1). Retention times, peak heights, and peak areas were reproducible over a wide range of PA concentrations
100
.005 AUFS
GIZZARD
I HI
w z
o Q.
Aci
CO
£50Injection A
z
1
LU O
CC LU
\
il
0 0
-*n TISSUE SOURCE
FIG. 4. Tissue and alimentary tract levels of penicillic acid (detected by HPLC and confirmed by GLC) from chickens 4 hr postdosing with 550 mg/kg body weight. Values represent mean levels of PA (Mg/g) ± SEM of duplicate replicates each assayed in triplicate.
1
A £
\
\u / V y n v ? - ^ r t W
r
In ection B
o
IN
h-
J
1
/
\
A£ A A
k
W^l/l
1 2 3 4 RETENTION TIME (min)
FIG. 5. Representative HPLC chromatograms of gizzard extracts from a control chicken (A) and a chicken dosed with penicillic acid (B). For comparison purposes, the control HPLC tracing is superimposed over the chromatogram from the chicken treated with penicillic acid. See text for HPLC conditions.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
10
(Phillips et al, 1981). Mean retention times for PA at flow rates of 2 ml/min and 1 ml/min were 94 and 188 sec, respectively. At maximum sensitivity, 5 to 10 ng of PA could be easily detected and adequately separated from interfering peaks in biological samples. Ethyl acetate was used to extract PA from tissues and blood. A second extraction with benzene was required in the case of liver to eliminate interfering peaks. Figure 2 outlines the procedure of sample preparation, extraction, clean-up, separation, and quantitation employed in this study. Recovery of PA in spiked samples was enhanced by treatment with strong acid at 80 C. The optimum time of treatment was determined by spiking samples with PA at a level of 500 ng/ml and by incubating with 1.0 ml 3N hydrochloric acid for 0, 5, 10, 15, 20, 30, and 60 min at 80 C in a water bath with shaking before extraction and HPLC analysis. Maximum recovery (85%) was obtained after 5 min treatment in serum, whereas only 26% recovery was obtained with no acid treatment (Fig. 3). In whole blood, a maximum recovery of 27% was obtained only after 15 min of acid treatment followed by a significant decrease in recovery thereafter. Maximum recovery of PA from liver, kidney, gizzard, heart, and skeletal muscle samples (80 to 105%)
HANNA ET AL.
2250
line derivative of PA from HPLC peaks and analysis by GLC as described in the text; PA was not detected in blood or skeletal muscle at any treatment level. DISCUSSION Methodologies describing extraction and analysis procedures for the mycotoxin PA are numerous. Detection methods currently available include bioassay, colorimetry, thin layer chromatography, and GLC. Recently, methods using HPLC with detection of PA at 254 nm have been reported (Engstrom et al, 1977; Chan et al, 1980; Phillips et al, 1981). Moreover, HPLC methods are generally rapid, specific, sensitive, and less laborious than most previously reported procedures but have not been utilized in determining PA in tissues. Results from our study indicate that HPLC can be effectively utilized in the primary analysis of PA in chicken tissues, following appropriate extraction and cleanup procedures. The GLC of the pyrazoline derivative formed after reaction with diazomethane can also be used for PA confirmation. The extraction of PA from tissues and blood was enhanced by acid treatment of sample homogenates at 80 C. This effect has been previously noted for PA (Chan et al, 1980) and rubratoxin B, another mycotoxin which, like PA, is reactive with sulfhydryl groups (Unger and Hayes, 1978) and may be due to increased
.01 AUFS
HEART
GIZZARD C O N T E N T S
o Q.
K 50 O oc UJ 0.
0 RETENTION TIME (mki)
FIG. 6. Representative HPLC chromatograms of gizzard content extracts from a control chicken (A) and a chicken dosed with penicillic acid (B). See text for HPLC conditions.
1
2
3
4
RETENTION TIME (min)
FIG. 7. Representative HPLC chromatograms of heart extracts from a control chicken (A) and a chicken dosed with penicillic acid (B). See text for HPLC conditions.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
was obtained within 10 to 15 min of acid treatment and was unchanged after 60 min treatment (data not shown). Chickens were initially dosed with PA at a level of 550 mg/kg body weight to test the applicability of this method for detecting PA residues from various tissue sources including gizzard and intestinal contents. Figure 4 summarizes the results obtained from this experiment. Significant levels of PA were detected 4 hr postdosing in kidney, heart, gizzard, gizzard contents, and intestinal contents. The PA was not detected in skeletal muscle or blood. Typical HPLC chromatograms comparing gizzard muscle and contents and heart extracts from both control and PA treated chickens are illustrated in Figures 5, 6, and 7. The chromatograms of tissue extracts from untreated chickens are superimposed over chromatograms of extracts from chickens dosed with PA. Table 1 summarizes the results from a subsequent experiment where chickens were dosed with PA at levels of 50, 100, 200, and 400 mg/kg body weight. At all treatment levels tested PA was detected in kidneys at concentrations ranging from 1480 to 2910 ppb, in heart at the 400 mg/kg level at a concentration of 590 ppb, in gizzard at the 100, 200, and 400 mg/kg levels at concentrations ranging from 2.8 to 19 ppb, and in liver at the 200 and 400 mg/kg levels at concentrations of 5.1 and 6.8, respectively. The presence of PA was confirmed in positive samples by formation of the pyrazo-
PENICILLIC ACID IN CHICKEN TISSUES
2251
TABLE 1. Tissue concentrations of penicillic acid in chickens detected by HPLC34 hours after administration of levels from 50 to 400 mg/kg body weight Parts per billion PA at indicated treatment levels (mg/kg)D Tissue
c
Kidney Heart Gizzard Liver Blood Muscle
50 1480 ± 490 d ND e ND ND ND ND
100 1920
200
+340 ND 2.8 ± 1.8 ND ND ND
2160
+400 ND 3.6 ± .5 5.1 ± .6 ND ND
400 2910 ±240 590 ± 30 19 ± 2.9 6.8 ± .2 ND ND
See text for HPLC conditions.
Tissues were collected 4 hours after administration of penicillic acid. Detected levels of penicillic acid are expressed as ppb (ng/g) of tissue and represent the mean ± SEM from 3 chickens each assayed in duplicate. e
Penicillic acid not detected.
precipitation of interfering protein in samples, or to release of PA which may have been bound to protein components, or a combination of both. A more quantitative recovery of PA can be achieved with this added step. In whole blood, recovery was poor even after acid treatment, possibly indicating a strong association of PA with erythrocytes or the formation of adducts which escaped detection. This conclusion is consistent with the results of a distribution study by Park and coworkers (1977) wherein 14 C-PA exhibited a very high affinity for erythrocytes. Tissue concentrations of PA determined by HPLC from experiments where chickens were dosed with levels of PA ranging from 50 to 550 mg/kg body weight indicate that residue retention was highest in kidneys, heart, gizzard, and liver. The PA was not detected in heart below a level of 590 ppb. We have no obvious explanation for this observation. Also, PA was not detected in skeletal muscle or blood. Lack of detection in blood may be explained by the absence of quantitative recovery in this tissue. A previous report (Park et al, 1977) on the distribution of PA in rats indicated that maximum levels of 14 C-PA were reached in blood and tissues within 2 to 4 hr postdosing. The organs that contained the highest concentrations of radioactivity were liver, kidneys, heart, and bladder. Levels in skeletal muscle also were very low in comparison, which is consistent with the
results we obtained in the chicken. It is evident from the literature on mycotoxins that there is a lack of information regarding residue accumulation in foods of animal origin and their significance to human health through food chain interactions. Studies on the metabolic fate of penicillic acid in food animals are also lacking. Because the consumption of poultry and poultry products has greatly increased in the United States over the last 10 years, and because PA can occur in corn and in poultry feed in high levels, the potential for contamination of edible poultry tissues with subsequent public health hazards cannot be overlooked. The HPLC method reported in this study should prove useful in monitoring of PA in poultry and other food animals, either as proposed here, or in conjunction with standard methods of analysis, as well as aid in further studies on the distribution and metabolism of this mycotoxin. ACKNOWLEDGMENTS This research was supported by Texas Agricultural Experiment Station Projects H 6215, AH 6397, and AH 6529 and the Texas A&M University, Office of University Research, ORR 2-10 and 5-81. REFERENCES Alsberg, C. L., and O. F. Black, 1913. Contribution
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
Chickens (3 per group) were orally dosed with penicillic acid dissolved in water at levels of 50, 100, 200, and 400 mg/kg.
2252
HANNA ET AL. cillic acid. J. Pharmacol. Exp. Therap. 88:119— 132. Park, D. L., R. E. Dailey, L. Friedman, and J. L. Health, 1977. The absorption, distribution, and excretion of 14 C-penicillic acid by rats. Ann. Nutr. Aliment. 31:919-934. Phillips, T. D., P. K. Chan, and A. W. Hayes, 1980. Inhibitory characteristics of the mycotoxin penicillic acid on (Na + -K + )-activated adenosine triphosphatase. Biochem. Pharmacol. 29:19— 26. Phillips, T. D., G. W. Ivie, N. D. Heidelbaugh, L. F. Kubena, S. J. Cysewski, A. W. Hayes, and D. A. Witzel, 1981. Confirmation of penicillic acid by high pressure liquid and gas-liquid chromatography. J. Ass. Offic. Anal. Chem. 64(1):162165. Reiss, J., 1979. Inhibitory action of the mycotoxins, patulin and penicillic acid on urease. Food Cosmet. Toxicol. 17:145-146. Scott, P. M., 1978. Mycotoxins in feeds and ingredients and their origin. J. Food Prot. 41(5):385— 398. Umeda, M., 1971. Cytomorphological changes of cultured cells from rat liver, kidney and lung induced by several mycotoxins. Jpn. J. Exp. Med. 41:195-207. Unger, P. D. and A. W. Hayes, 1978. High-pressure liquid chromatography of the mycotoxins rubratoxins A and B, and its application to the analysis of urine and plasma for rubratoxin B. J. Chromatogr. 153:115-126. Wilson, D. M., 1976. Mycotoxins and other fungal related food problems. Pages 90—109 in Advances in chemistry. Series No. 149. J. Rodricks, ed. Amer. Chem. Soc, Washington, DC.
Downloaded from http://ps.oxfordjournals.org/ at New York University on June 2, 2015
to the study of maize deterioration. US Dept. Agr. Bull. Bur. Plant Ind. 270-J. Bacon, C. W., J. G. Sweeney, J. D. Robbins, and D. Burdick, 1973. Production of penicillic acid and ochratoxin A on poultry feed by Aspergillus ochraceus-. temperature and moisture requirements. Appl. Microbiol. 26:155-160. Chan, P. K., T. D. Phillips, and A. W. Hayes, 1979. Effect of penicillic acid on adenosine triphosphatase activity in the mouse. Toxicol. Appl. Pharmacol. 49:365-372. Chan, P. K., M. Y. Siraj, and A. W. Hayes, 1980. High performance liquid chromatographic analysis of the mycotoxin penicillic acid and its application to biological fluids. J. Chromatogr. 194:387— 398. Ciegler, A., and C. P. Kurtzman, 1970. Penicillic acid production by blue-eye fungi on various agricultural commodities. Appl. Microbiol. 20:761— 764. Ciegler, A., H. J. Mintzlaff, D. Weisleder, and L. Leistner, 1972. Potential production and detoxification of PA in mold fermented sausage (salami). Appl. Microbiol. 24:114-119. Dickens, F., and H.E.H. Jones, 1961. Carcinogenic activity of a series of relative lactones and related substances. Brit. J. Cancer. 1 5 ( l ) : 8 5 - 9 5 . Engstrom, G. W., J. L. Richard, and S. J. Cysewski, 1977. High pressure liquid chromatographic method for detection and resolution of rubratoxin, aflatoxin, and other mycotoxins. J. Agric. Food Chem. 25:833-836. Huff, W. E., P. B. Hamilton, and A. Ciegler, 1980. Evaluation of penicillic acid for toxicity in broiler chickens. Poultry Sci. 59:1203-1207. Murnaghan, M. F., 1946. The pharmacology of peni-