NR AQEW
AU Jackman,R.; Everest,S.J.
TI Further development of the electrochemical analysis of urine from cows with BSE
QU Transmissible Spongiform Encephalopathies, A consultation on BSE with the Scientific Veterinary Committee of the Commission of the European Communities held in Brussels from 14 to 15 September 1993, European Commission Agriculture, S. 369-76
PT Article
AB The composition of urine from BSE-affected and clinically normal/non-affected cattle has been studied using cyclic voltammetry to measure the redox properties of excreted metabolites. A significant difference has been demonstrated in the relative concentrations of three urine constituents between suspect cases subsequently confirmed, by histological examination of the brain, as positive as those which are histologically negative. The analytical procedures involved are not, however, appropriate for large-scale screening of suspect cases of BSE and studies are underway to identify the three metabolites involved, to develop alternative tests and to assess the initial findings on a larger population of samples. Urine from confirmed BSE cattle has been subjected to a number of clean-up preparative procedures to isolate the metabolites for subsequent identification. Results from examination of the isolates by mass spectrometry, nuclear magnetic resonance spectroscopy and array voltammetry have indicated the class of chemical constituents implicated in two of the three fractions and this data is currently in use to assess comparisons with standards and type-compounds.
VT
INTRODUCTION
The diagnostic process for bovine spongiform encephalopathy currently consists of clinical assessment of a suspect case, slaughter and histopathological examination of the brain. This system, although definitive, results in unnecessary culling, disposal, compensation and case study of a significant number of animals subsequently confirmed as BSE-negative and represents a considerable extra cost to the control programme. The numbers of animals falsely identified as suspect BSE by clinical examination would be reduced if an effective further aid to diagnosis, such as a test, or series of tests, using aecessible body fluids, were to be available. Although any disease, with its accompanying symptoms, is an end result of alterations or malfunctions in metabolic or physiological processes, such changes need to be specific, as well as being reflected grossly, for the biochemical lesions detected to be useful in differential diagnosis. The, as yet, unidentified infectious agent produces no detectable immunological response and does not appear, by use of bioassay, to be present in body fluids. Immunoassay and bioassay, therefore, are of no value as diagnostic aids. However a number of other biochemical markers have been indicated as possible aids to diagnosis of the spongiform encephalopathies, including serotonin (Chatelain et al, 1984) choline acetyl transferase (McDermott et al, 1978) acetyl cholinesterase and glutamate decarboxylase (Iqbal e( al, 1985) neurone specific enolase (Jitawi el al, 1988; Yermuyten el al, 1990) creatine kinase isoenzymes (Agiroudis et al, 1982) and neurofilament (light chain) proteins (Nakazato et al, 1990).
Reported changes in the electrochemical properties of urine (Combrisson et al, 1991; Banissi-Sabourdy et al, 1992) taken from scrapie affected sheep appear, however, to offer the most appropriate analytical principle for a live animal test for BSE and the results obtained to date on the assessment of this approach are presented here.
MATERIALS AND METHODS
Samples were collected during normal urine flow at various, but unspecified, times of the day and transported at ambient temperature. Upon receipt samples were kept at 4° for up to one month and, if required, at -20° for longer term storage. Although the volumes of urine collected varied from 5-100 ml, only 200 µl only was required for analysis. Both clinically affected and clinically normal animals were sampled when appropriate at the Central Veterinary Laboratory (CVL). Field case assessment was carried out on samples obtained from animals suspected of BSE, by clinical examination and from age matched, unaffected animals in the same herd.
All animals showing clinical signs consistent with BSE were slaughtered, brain material was removed for histopathological examination, and the carcases incinerated.
Cyclic voltammetry is performed on diluted (1 in 5) urine acidified with 32 µl concentrated sulphuric acid/ml. The sample is pumped into the reaction chamber of an ESA Model 5010 triple electrode assembly. A potential difference is then applied across the sample in a linearly increasing gradient from 175 mV to 1325 mV at 10 mV/second. The potential is then immediately reduced, returning at the same rate to 175 mV. A second, and sometimes a third, similar sweep is made with the resulting redox changes being detected by the electrodes and the induced current responses transmitted to an on-line X-Y recorder.
Although the entire range of input potential was assessed initially, study has been concentrated on the oxidation of particular compounds in the urine, induced by increasing voltage between 500 and 800 mV. In particular, three peaks of interest have been identified. Peak 1 is seen on the first potential sweep at 750 mV, is present in all urines so far studied and appears to be an indicator of urine concentration. The second and third peaks are only recognised on the second and subsequent sweeps at 575 and 675 mV respectively. Peak 2 is again present in all urines so far tested but peak 3 offers the possibility of a discriminatory index between urines from BSE-affected and control animals by comparison of the response heights of peaks 1 and 2 in the formula: (peak 2 - peak 3) / (peak 1 - peak 2).
Although the ECD equipment produces extremely reproducible results, the passage of untreated urine through the test cell causes intermittent fouling of the electrodes on a scale inconsistent with its use for large scale screening of samples. Prolonged cleaning and subsequent quality assurance assessment result in unacceptable instrument down-time. As a result studies are underway to isolate, purify and identify the compounds of interest in order that a more robust analytical technique may be developed. To this end, quantities of urine from confirmed cases of BSE have been subjected to preparative purification procedures and analysis. No single-step procedure has yet been identified but the following techniques have been utilised in the preparation of isolates.
l. Ethanol precipitation of salts and insoluble materials.
2. DEAE ion exchange chromatography on sepharose in Tris buffer pH 7.0 and elution with a salt gradient.
3. Solid phase extraction on preparative C18 and NH2 column.
4. Silica gel ThC developed in chloroform/methanol mixtures.
5. HPLC.
RESULTS
Figure 1 demonstrates the voltammogram obtained by two successive sweeps of increasing, then decreasing voltage between 175 mV and 1325 mV on a urine from a BSE affected animal. Peaks 1,2 and 3 are indicated and by calculating the ratio of the distances between peaks 2 and 3 and that between 1 and 2 a figures of less than unity is obtained. An example of a similar scan from a control urine is given in Figure 2 and gives a peak ratio of greater than unity.
Fig. 1: Voltammogram or BSE positive urine
Fig. 2: Voltammogram of BSE negative urine
Using this Process 77 bovine urines from animals accommodated at CVL have been assessed. All samples, either from the live animal or taken at necropsy, were analysed blind within two weeks of collection. Subsequent histopathological diagnoses were then correlated with the urine test findings. Figure 3 shows the numerical distribution of the peak ratios of all the urines tested in this survey to date where histopathological examination of the corresponding brains has been completed.
Fig. 3: ECD RATIOS IN BOVINE URINE CASES SUBMITTED TO CVL
During 1992, 52 urine samples from other parts of Ihe UK were collected from suspected field cases of BSE and from their clinically normal herd mates.
Figure 4 shows the peak ratios distribution of the clinically affected animals. Only seven were subsequently diagnosed as non-BSE, six of which appear as negative by the urine test, one as highly positive.
Fig. 4: ECD RATIOS IN BOVINE URINE CASE AMPLED IN THE FIELD
Fig. 5: ECD RATIOS IN BOVINE URINE CLINICALLY NORMAL ANIMALS
A similar histogram Figure 5) shows the results of the urine samples taken from approximately age-matched animals within the same herds as the clinically affected cows of Figure 4. Here the distribution falls either side of the assigned discriminafion point of unity (1.0).
The critical phase of the isolation of the compounds corresponding to peaks 2 and 3 was accomplished via ion exchange chromatography of ethanol treated urine and gave fractions apparently pure by electrochemical analysis Figure 6). However, when these fractions were submitted for nuclear magnetic resonance spectroscopy (NMR) and mass spectroscopy (MS) other compounds were found still to be present. No useful data was forthcoming from MS, but even though impure, NMR indicated a possible compound-type. In addition these semi-purified fractions were submitted for array voltammetry following high performance liquid chromatography (HPLC) separation and compared to a library of known urinary metabolites without successful identification.
A second set of more highly purified fractions are currently in preparation.
Combined data from NMR studies, electrochemical properties and chromatographic comparisons indicate the presence in peak 3 of either a p-hydroxyphenyl nucleus with an acidic (carboxyl) substituent or a similarly derivatised indole. The "best-fit" agreement between the naturally occurring metabolite and standard compounds has been obtained with p-hydroxy phenyl pyruvic acid.
Voltammogram of urine extract following ion exchange chromatography
DISCUSSION
The electrochemical properties of the three urinary metabolites selected for study differ markedly. The compound producing the response at peak 1 is oxidised irreversibly during the first voltametric sweap at an input potential of 700 mV and does not therefore appear in successive cycles. However peaks 2 and 3 do not appear in the first oxidative sweep, presumably because they are already in the oxidised form in the original urine, or possibly that they are a result of chemically transformed oxidation from the first cycle. However, following the reduction phase they appear in the second and subsequent sweeps at 550 and 650 mV respectively indicating the presence of a fully reversible redox system. Although minor variations in the potential at which these peaks appear occur from urine to urine, in every sample so far studied the peaks are produced exactly 100 mV apart.
Such similar properties may point to similar metabolites being involved and this possibility is strengthened by their behaviour in various purification systems. Both have similar solubilities in ethanol and organic solvents, have extremely close salt-elution profiles from ion exchange resins and behave identically in the solid phase extraction protocols so far attempted. As the purification of fractions proceeds, however, peak 3 undergoes a shift in response to an applied potential of approximately 50 mV up-stream but returns to its original position when electrochemically assessed in the presence of untreated urine.
A further unusual aspect to the purification procedure as assessed by electrochemistry is the appearance of a peak, originally seen as only a minor shoulder on the up-grade of peak 2 in the original urine, apparently at the expense of the combined areas of 2 and 3, as the isolation procedure progresses. The significance of these observations remains to be determined.
This is also true of the, as yet, tentative identification of the metabolite of peak 3 as being a p-hydroxyphenyl or indole derivative. It is interesting to speculate that such compounds, being involved in tyrosine and tryptophan metabolism, may indicate a biochemical lesion in neurotransmitter production or degradation and correlate with the findings of Chatelain el al (1984) that serotonin levels in the blood of scrapie-affected sheep were reduced and that disturbances in tryptophan catabolism might lead to excitatory symptoms similar to those seen in BSE (A. Austin, Personal communication).
The data so far obtained from the urine test, as it is currently applied, have produced conflicting correlations with presence/absence of histological demonstration of BSE. During 1992 and early 1993, urine was obtained at necropsy from 77 animals kept at CVL, for different periods, under similar nutrition and husbandry. The correlation of urine test result and subsequent diagnosis Figure 3) is absolute.
At the same time samples were also taken from clinically suspect cases throughout the UK. Of the 52 urines analysed so far 45 were from animals subsequently confirmed to have BSE and seven were BSE-negative. Six of the seven BSE-negative samples were also deemed negative in the urine test; one was highly positive. However six of the 45 confirmed BSE cases were negative in the urine test Figure 4). The discrimination point of a peak rate of 1.0 has been assigned on the basis of the results of the cases studied at CVL. It is necessary to obtain a larger number of negative control urines - that is from both clinically normal animals with a guaranteed absence of any challenge by the infectious agent of BSE - before a statistically meaningful cut off point can be drawn.
This assumes greater importance in the light of the results depicted in Figure 5, the data being derived from applying the electrochemical urine test to 81 age-matched animals from the same herds and sampled at the same time as the BSE suspects. Here in excess of 56% of the urines were deemed positive and represents a considerable divergence in distribution and correlation from the previous data. Only two of these latter animals have since been available for brain histopathological examination.
The electrochemical urine test in its present form, therefore, appears to fail in its intention of discriminating between BSE affected and healthy control animals in the field although results from local (CVL) cases, carried out blind, offer a much more satisfactory outcome as does the correlation between the urine test and BSE positive and negative cases in clinically abnormal animals in the field.
Any one of the huge variety of husbandry regimes used in the cattle industry may duplicate the urine metabolite profile described here in the BSE affected animals at CVL, especially as not one but three electrochemically active species are being compared. The apparently false positive urine test results obtained from healthy control animals in the field may therefore be a result of naturally occurring variations in urine constitution. At present these healthy control animals remain alive, except for two animals subsequently slaughtered. Brain stem histopathology was carried out on both animals; one which had an original urine test ECD ratio of 0.64 was confirmed BSE positive, the other with a ratio of 2.05 was BSE negative. Identification of the metabolites involved and/or development, of a more rapid, robust and specific test based on the compound equivalent to peak 3 in the electrochemical test is essential for the resolution of what may be purely an analytical problem.
REFERENCES
Agiroudis S.A., Kent J.E. and Blackmore D.J. (1982). Observations on the isoenzymes of creatine kinase in equine serum and tissues. Equine Vet. J. 14(4), 317-321.
Banissi-Sabourdy C., Planques B., David J.-P., Jeannin C., Potel M., Bizien A., Di Menza C., BurgerePicoux J., Brugere M., Chatelain J. and Buvet R. (1992) Electroanalytical characterisation of Alzheimers disease and ovine spongiform encephalopathy by repeated cyclic voltammetry at a capillary graphite paste electrode. Bioelectrochem. & Bioenergetics 28, 127-147.
Chatelain J., Haimart M., Launay J.M., Baille V., DReux C. and Cathala F.(1984) Serotoninemie et histaminemie chez des ovins endimiquement atteints de tremblante (scrapie): premiers resultats. CR Soc. Biol. 178, 664-670.
Combrisson H, Robain 0., Brugere-Picoux J., Chatelain J. and Brugere H.(1991) Modifications des parametres urodynamiques chez la brebis atteinte de tremblante. Bull Acad. Vet. de France 64,
257-266.
Iqbal K, Somerville R.A., Thompson C.H. and Wisniewski H.M. (1985) Brain glutamate decarboxylase and cholinergic enzyme activities in scrapie. J. Neurol. Sci. 67. 345-350.
Jitawi S.A., Cochran A.J., Cancilla P.A. and Wen D.-R. (1988) The expression of S-100 protein and neuron-specific enolase in meningiomas. Disease Markers. 6, 109-118.
McDermott J.R., Frazer H. and Dickinson A.G. (1978) Reduced choline-acetyl transferase activity in scrapie mouse brain (letter). Lancet. 2, 318-319.
Nakazato Y., Hirato J., Ishida Y., Hoshi S., Hasegawa M. and Fukuda T. (1990) Swollen cortical neurons in Creutzfeldt-Jakob disease contain a phosphorylated neurofilament protein. J. Neuropath. Exp. Neurol. 49, 197-205.
Vermuyten K., Lowenthal A. and Karcher D. (1990) Detection of neurone specific enolase concentrations in cerebrospianl fluid from patients with neurological disorders by means of a sensitive enzyme immunoassay. Clin. Chem. Acta. 187, 69-78.
IN 200 µl Urin wurden mit je 6,4 µl Schwefelsäure angesäuert und auf 1:5 verdünnt. Anschließend wurden die Proben mit einem Dreielektrodensystem gemessen. Dabei wurde die Spannung mit einer Geschwindigkeit von 10 mV/Sekunde von 175 auf 1325 mV gesteigert und wieder abgesenkt. Dieser Zyklus wurde 2-3 mal wiederholt. Dabei fielen 3 Peaks auf, von denen der erste bei 750 mV in jeder Probe vorkam und als Maß für die Urinkonzentration genommen wurde. Der zweite und der dritte Peak bei 575 und 675 mV wurden erst ab dem zweiten Zyklus erkennbar. Allerdings kam auch der zweite Peak in jeder Probe vor. Nur der dritte Peak bei 675 mV zeigte eine gewisse, jedoch keine wirklich befriedigende Spezifität. Für Reihenuntersuchungen taugt diese Technik allerdings nicht, weil die Elektroden zwischen den Messungen aufwendig gereinigt und getrocknet werden müssen. Die Autoren experimentierten daher schon mit verschiedenen Reinigungstechniken und kündigten weitere Versuche zur Entwicklung einer robusteren und praktikableren Meßtechnik an. Per NMR, Massenspektrometrie, Chromatographie und der elektrochemischen Untersuchung wurde die angereicherte Substanz des Peak 3 mit Standardsubstanzen verglichen. Die größte Übereinstimmung wurde mit p-Hydroxyphenylpyruvat gefunden. Bei 77 im zentralen Veterinärlabor gehaltenen Rindern entsprach das elektrochemische Testergebnis immer richtig. Bei 45 auf normalen Farmen gesammelten Urinproben gab es ein falsch positives und 6 falsch negative Ergebnisse.
AD R. Jackman and S.J.Everest, Immunochemistry Group, Biochemistry Discipline, Central Veterinary Laboratory, New flaw, Addlestone, Surrey, KTl5 3NB UNITED KINGDOM
SP englisch
OR Prion-Krankheiten 5 und im violetten Proceedingsband