Determination Of Nitroxynil Residues In Tissues Using High-performance Liquid Chromatography-thermospray Mass Spectrometry

ANALYST, SEPTEMBER 1989, VOL. 114

1013

Determination of Nitroxynil Residues in Tissues Using High-performance Liquid Chromatography - Thermospray Mass Spectrometry W. John Blanchflower and D. Glenn Kennedy Department of Agriculture, Veterinary Research Laboratories, Stormont, Belfast BT4 3SD, UK

A method is described for the determination of nitroxynil residues in muscle, liver and kidney. The samples were extracted into diethyl ether and cleaned-up using a simple liquid - liquid extraction step. Any nitroxynil present was separated from interfering compounds by high-performance liquid chromatography and detected using thermospray mass spectrometry. The assay is specific and sensitive, with a detection limit of 2 ng 9-1 in tissues. Keywords: Nitroxynil; residue; high-performance liquid chromatography; thermospray mass spectrometry

Nitroxynil (4-hydroxy-3-iodo-5-nitrobenzonitrile)is an anthelmintic mainly used to control liver fluke (Fasciola hepatica) in sheep and cattle.1 It is marketed as an aqueous solution of the N-ethylglucamine salt under the name "Trodax." As residues of the drug have been detected in tissues of animals up to 90 d after treatment,2 adequate withdrawal times must be observed before animals are slaughtered for human consumption. This has led t o the necessity for the development of methods to monitor residual levels of nitroxynil in meat products to ensure that they fulfil the relevant government tolerance limits. Most of the results reported for nitroxynil have been obtained using a polarographic method3 and, more recently, a gas chromatographic method has been described for determining residue levels in milk.4 However, for statutory residue testing purposes, mass spectrometric methods should be used if possible to ensure adequate specificity. This paper describes the development of a high-performance liquid chromatographic - thermospray mass spectrometric (HPLC - MS) method for the detection and quantification of nitroxynil residues in the tissues of slaughtered animals.

Experimental Apparatus

The HPLC system consisted of a Merck-Hitachi Model L-6000 pump (BDH, Romford, Essex, UK), a Rheodyne Model 7125 injector fitted with a 2 0 4 sample loop (BDH) and a LiChrosorb RP-18, 250 X 4 mm i.d. reversed-phase cartridge with holder (BDH). The HPLC - MS system was a Vestec Model 201A thermospray instrument (Vestec, Houston, TX, USA) complete with a Technivent workstation. The instrument was operated in the negative ion chemical ionisation mode (CI) using filamentinitiated ionisation with an electron beam current of 250 FA. The electron multiplier voltage was 2400 V. The temperatures of the vaporiser, block, tip heater and lens assembly were set at 180, 250, 280 and 150"C, respectively. The tuning parameters were checked weekly according to the manufacturer's instructions using a 50 mg 1-1 solution of polyethylene glycol 300 in 50% acetonitrile containing 0.1 M ammonium acetate solution. The instrument was used either in the full-scan mode to collect spectra or in the selected ion monitoring mode (SIM) for maximum sensitivity when analysing samples. For the former, a dwell time of 0.1 ms was used and for the latter the dwell time was 100 ms. In each instance the sweep window was set at 0.5 a.m.u.

Reagents

Acetonitrile was of HPLC grade; all other reagents were of AnalaR grade. Nitroxynil was obtained as the N-ethylglucamine salt from May & Baker, Dagenham, Essex, UK. A stock standard (1 -755 mg ml-1 of the salt, equivalent to 1mg ml-1 of nitroxynil) was prepared in methanol and was stable for 1 month if stored at 4°C. A dilute standard (100 ng ml-1 of nitroxynil) was prepared daily by dilution of the stock standard with the mobile phase. The mobile phase consisted of a mixture of 300 ml of acetonitrile and 700 ml of 0.1 M ammonium acetate solution. This solution was de-gassed and filtered under vacuum through a 0.45-pm filter using a Millipore - Waters solvent filtration system (Millipore Waters, Harrow, Middlesex, UK). The mobile phase was pumped at a flow-rate of 1 ml min-1.

Method

Cut the tissue samples into small cubes and store at -20°C until frozen or until required. Place the frozen cubes in a domestic food blender and pulverise until the tissue forms a fine powder. Weigh 2-g portions into 110 x 25 mm centrifuge tubes fitted with ground-glass necks. Add 15 ml of 0.4 M disodium hydrogen orthophosphate solution and homogenise for 1 min using a Silverson homogeniser. Add 2 ml of concentrated hydrochloric acid and cool the tubes by allowing them to stand in cold water for a few minutes. Add 8 ml of diethyl ether, stopper the tubes and shake them vigorously by hand for 30 s. Centrifuge at 10°C and 2000 g for 10 min. Remove the upper diethyl ether layer using a Pasteur pipette and transfer it into 100 x 16 mm centrifuge tubes fitted with ground-glass necks. Reduce the volume of the diethyl ether to a few millilitres under nitrogen at 40°C in a fume cupboard. Re-extract the homogenates once with further 5-ml aliquots of diethyl ether and combine these with the first extracts. Again reduce the volume of the diethyl ether to about 4 ml under nitrogen. Add 3 ml of 0.4 M disodium hydrogen orthophosphate solution, stopper the tubes and shake them vigorously by hand for 30 s. Centrifuge at 10°C and 2000 g for 5 min. Remove the upper diethyl ether layer and discard, taking care not to remove any of the aqueous layer. Add 0.4 ml of concentrated hydrochloric acid and cool the tubes by allowing them to stand in cold water for a few minutes. Add 2 ml of diethyl ether, stopper the tubes and shake them vigorously by hand for 20 s. Centrifuge at 10°C and 2000 g as before. Transfer the upper diethyl ether layer into 70 x 12 mm tubes, again fitted with ground-glass necks. Re-extract the acidified orthophosphate solutions once with further 1-ml aliquots of diethyl ether and combine these with the first extracts.

1014

ANALYST, SEPTEMBER 1989, VOL. 114

Evaporate to dryness at 40°C under nitrogen in a fume -upboard. Dissolve the residue in 100 pl of acetonitrile by eating at 60 "C for a few minutes and mix. Add 200 pl of 0.1 M mmonium acetate solution and mix. Inject 20-11 aliquots into ie HPLC - MS system using the Rheodyne injector. Collect ie peak data using the workstation and record the areas of the eaks. Compare these with peak areas for 20-pl aliquots of a 30 ng ml-1 nitroxynil standard similarly injected.

Results 'he HPLC-MS CI spectrum of nitroxynil obtained by ijecting 20 pl of a 10 pg ml-1 standard into the system and dlecting all the spectral data is shown in Fig. 1. Two main ms were observed, at rn/z 273 and m/z 290 (the molecular m). For optimum sensitivity when analysing tissue samples, ither ion could be monitored by STM. Alternatively, if more 48131 . 40000

-

30000

-

20000

-

10000

-

?H

0 '

I

I I

I

II,

I

1

I

definitive proof of the presence of nitroxynil residues in the tissue samples is required, both ions can be monitored and the ion ratio measurements compared. Typical computer outputs for the ion chromatograms at m/z 290 for a standard, a blank tissue extract and a tissue extract from a muscle sample spiked with 10 ng g-1 of nitroxynil are shown in Fig. 2. Nitroxynil elutes at 4.8 min. The linearity of the assay was checked by spiking duplicate aliquots of a blank tissue sample with 10, SO, 150,250,400 and SO0 ng g-1 of nitroxynil and carrying them through the procedure. The results are shown in Fig. 3 and it can be seen that the assay is linear up to at least 500 ng g- I of nitroxynil in tissues. For samples containing more than SO ng g-1 of nitroxynil, the final extracts were diluted with an appropriate volume of the mobile phase before injection to prevent saturation of the electron multiplier signal. The recovery of nitroxynil in the assay was determined using the results from the linearity test. The measured peak areas were compared with those of a 10 ng ml-1 standard and the amount of nitroxynil recovered was determined. The results are shown in Table 1; recoveries ranged from 71 to 95%. The precision of the assay was checked by spiking aliquots of a blank tissue sample with 10 and SO ng 8-1 of nitroxynil and then analysing each aliquot five times using the described procedure. The results for the mean (standard deviation) and coefficient of variation were 8.42 (1.03) ng g-1 and 12.2% for the 10 ng g-1 samples and 44.4 (2.53) ng g-1 and 5.7% for the SO ng g-1 samples.

g. 1. Full-scan negative-ion CI spectrum from mlz 250 to mlz 320 the nitroxynil peak at 4.8 min; 20 yl of a 10 yg ml-1 standard were

5000

'

jecred. The molecular structure of nitroxynil is also shown

67

.E 4000

K

3

2 .- 3000

e

16000

m' 2000

2

14000 12 000 10 000 0 -

8000

0

I

I

400

500

Linear regression of a series of extracts from muscle samples spiked with 1&500 ng g-I of nitroxynil and carried through the assay; injection volume, 20 yl. Extracts from samples spiked with >50 ng g-1 of nitroxynil were diluted with the mobile phase before injection to prevent saturation of the electron multiplier signal

52

a,

I

300

Fig. 3.

4000 a,

c

I

200

Amount of nitroxynilhg g-1

6000

.-w 2 3000

I

100

'

Table 1. Recovery of nitroxynil added to tissue samples I

I

I

I

1

Added/ng g-' 10 50 150 250 400 500

[r

13 000

- (''

11000

-

9000

-

Found& g8.4 44.5 106.2 236.7 284.9 413.6

Recovery, YO 84 89 71 95 71 83

Table 2. Nitroxynil levels in muscle and kidney from treated cattle

Nitroxynil/ng 8-1

3000

'

1

1

I

2

3

1

I

1

I

4

5

6

7

Run t i m e h i n

g. 2. Ion chromatograms at m/z 290 of a 2 0 4 injection of ( u ) a 100 ; ml-I nitroxynil standard solution (nitroxynil elutes at 4.8 min); (6) blank muscle extract; and ( c ) a tissue extract from a muscle sample iked with 10 ng g-I of nitroxynil

Sample 1 2 3 4 5 6

Muscle 300 130 540 540 360 510

Kidney 520 160 780 790 -

-

1015

ANALYST. SEPTEMBER 1989, VOL. 114

The results for a batch of samples submitted to this laboratory from an abattoir as part of a residue testing scheme are shown in Table 2. It was believed the animals concerned had been treated with the drug and submitted for slaughter within the recommended withdrawal period of 28 d . The values ranged from 130 to 540 ng g-1 for muscle and from 160 to 790 ng g-1 for kidney.

Discussion High-performance liquid chromatography - mass spectrometry offers analytical laboratories a further instrumental technique for the specific analysis of a wide range of chemical compounds. Using dedicated HPLC - MS systems such as the Vestec thermospray instrument described here, the analysis is easier than using gas chromatography coupled with mass selective detection. Derivatisation is generally not required and clean-up steps are often relatively simple. There are two main reasons for this: (1) the sensitivity of thermospray H P L C - M S using Cl is compound dependent, with some compounds such as nitroxynil giving good sensitivity whereas others are relatively insensitive; and (2) little fragmentation of the molecules takes place, with mainly the mass ion and a few large fragment ions being produced. In the proposed method, the only clean-up step required after the initial extraction is for the removal of fats. This was achieved by extracting the nitroxynil from diethyl ether into disodium hydrogen orthophosphate solution, acidifying the extract and re-extracting into diethyl ether. As can be seen from Fig. 2, no interfering compounds are present in the ion chromatograms of the ion at rnfz 290. The Vestec HPLC - MS instrument used offers three modes of ionisation: (1) pure thermospray; (2) filament-assisted CI;

and (3) Townsend discharge assisted CT. For nitroxynil we found that filamcnt-assisted CI gave the best sensitivity and this form of ionisation was therefore used for the assay. The instrument also allows the measurement of both positive and negative ions, and operation in the negative-ion mode was again found to offer the greater sensitivity for nitroxynil. The assay was found to be linear up to at least 500 ng g-1 of nitroxynil in tissues (Fig. 3) and recoveries averaged 82% (Table 1). It was not, therefore, necessary to construct a calibration graph for each batch of samples. A 100 ng ml-1 standard injected after each batch of five sample extracts was found to be sufficient. The proposed assay was used to confirm the presence of nitroxynil residues in muscle, liver and kidney samples submitted from abattoirs as part of our residue testing programme. A batch of 15 samples can easily be analysed by one operator in one working day and the results are more specific than those obtained with previously reported methods. The sensitivity is also higher, with a detection limit of 2 ng 6-1 in tissues, compared with 100 ng g-1 when using the polarographic method.3

References 1. Lucas. J. M . S . , Rr. Vet. J . . 1967, 123, 198. 2. Ekstrom, L. G., and Slania, P., Actu Ver. Scand., 1982,23,313. 3. Parnell, M. J., Pestic Sci., 1970. 1. 138. 4. Kazacos, M . , and Mok, V., Aust. J . Dairy Technol., 1986, 41, 82.

Paper 9i00835G Received February 23rd, 1989 Accepted April 17th, 1989