View Report Details

The relevance of ATNC to colon carcinogenesis measurement of nitroso compounds by MS or LC/MS or LC/MS/MS

Project Code: T01031

31/03/2005

Central Science Laboratory
Dennis, M ; Clarke, D

1. There is considerable potential for nitrosation reactions to occur both in food and in the body, and a wide range of secondary amines could be precursors of nitrosamines. Measuring every nitrosamine in a sample would be time consuming and so an assay was developed which permitted an estimate of the “total” nitrosamines present. This was termed Apparent Total Nitroso Compounds (ATNC) because the exact nature of the compounds was not known. The search to identify faecal apparent total nitroso compounds (ATNC) is a challenging one. The increased sensitivity provided by the Quattro Ultima mass spectrometer over previous instruments offers the opportunity to characterise such molecules at the relatively low concentrations that are known to be present in faecal waters. However the difficulty is in determining which of the many compounds present are nitrosamines before embarking on a time consuming effort to fully characterise them. We have therefore adopted two somewhat different approaches to resolving this problem. The first was to guess at the most likely candidate molecules and synthesize a representative molecule for use as a standard. Such a standard would enable us to characterise the molecule by mass spectrometry, to ensure it was chromatographically resolved and ionised under the proposed LC-MS conditions, and if necessary to ensure it was not lost in any sample extraction or clean up procedures that may be required. It was decided to concentrate on the products generated by reaction between amines and fructose (Amadori products). Since these are secondary amines, their nitrosation would result in nitrosamines. These standards have been produced from tryptophan and phenylalanine and their mass spectra characterised.

2. Originally we had hoped that we could establish LC-MS conditions under which nitrosamines produced a neutral loss of NO. The loss of mass 30 is relatively unique and hence would be a good indication that any unknown compound under study was a nitrosamine. We found that MS/MS product ion spectra from both [M+H]⁺ and [M+Na]⁺ species in positive mode, and from [M–H]^– in negative mode were information-rich. However the anticipated loss of NO from the [M+H]⁺ ion (neutral loss of 30) was found not to occur, whereas a loss of 31 (H-N=O) gave rise to a major peak in the spectra. Many compounds give loss of 31 on MS/MS so that the strategy of employing ‘neutral loss’ scanning is not possible.

3. A second and rather different analytical approach was therefore undertaken. Most of the amines in food remain un-nitrosated even in cured products because of the relatively small amounts of nitrite used in their manufacture. We have assumed that any in vivo nitrosation of alimentary tract contents would be non-specific and depend on the amount of amine present (since concentration of nitrosating agent is likely to be the limiting factor). Hence by further nitrosating faecal waters (or ileostomy fluids) we could increase the amount of ATNC to enable them to be more readily detected. The amount and number of nitrosamines generated depends on the concentration of individual amines present. By using N14 and N15 labelled nitrite we demonstrated that mass spectral evidence of a unit mass difference provides a novel and practicable means for identifying unknown ATNC in faecal and ileostomy samples. We have demonstrated that this procedure is repeatable on a day-to-day basis for a single sample and have used it to identify 27 compounds (as parent masses) largely from ileostomy samples but also from faecal samples.

4. These putative nitroso compounds in different ileostomy and faecal samples tend to have different mass spectra; rather surprisingly we have not found the same candidate nitroso compound in both ileostomy and faecal samples. We have undertaken experiments to investigate reasons for this. It is apparent that modest changes in HPLC conditions lead to the expected masses of interest not being observed and this can include samples spiked with nitrosamine standard. Our explanation for this effect is that the large number of compounds in the faecal extract causes signal suppression in the mass spectrometer. We have made some progress to resolving this by taking extracts of faecal water into ethyl acetate, removing the ethyl acetate and resuspending in water. This clean-up reduced background in the mass spectrometer but was insufficient on its own to remove suppression entirely. We demonstrated that we extracted about half the ATNC into solvent and that the ATNC was stable to extraction, solvent removal and resuspension. Half of this ATNC extracted into ethyl acetate could then be transferred into dichloromethane.

5. We have developed conditions for a UV treatment of faecal/ileostomy samples, which enables us to differentiate N-Nitroso compounds from other compounds, which might have an N¹⁴O₂ / N¹⁵O₂ unit mass change on nitrosation but are not nitrosamines. (A typical example would be the C-Nitroso compound formed by nitrosation of the aromatic ring of tyrosine). This enabled us to pursue active mass measurements on putative nitrosamines with some confidence that they were compounds of interest.

6. In order to provide a pure chemical to the mass spectrometer, eliminating suppression effects, we undertook a series of experiments to fractionate and concentrate nitrosocompounds from a single large quantity of faecal aqueous extract. Attempts were made to enhance nitrosocompounds by nitrosation of the faecal water directly and then to isolate semi purified fractions using preparative scale-HPLC. We used a Greiss reagent procedure to identify fractions containing nitrosocompounds (this proving quicker than ATNC measurements). Our most successful experiment on a nitrosated faecal water identified a compound (m/z 348) that N¹⁴/N¹⁵ nitrosation demonstrated to be a nitrosocompound and that UV irradiation indicated was probably an N-Nitroso compound (because the peak disappeared ruling out the involvement of C-nitrosocompounds). While making preparations for accurate mass measurement, this compound was gradually lost from the preparation indicating limited stability even when frozen over a 3-week period. This came as a surprise since ATNC data had demonstrated the stability of this fraction as a whole. Further efforts to repeat this preparation were unsuccessful.

7. Efforts therefore turned to nitrosating ileostomy fluids since it was considered that these offered a better chance of providing material in sufficient quantity for an accurate mass measurement. In order to minimise potential stability issues, it was decided to perform the nitrosation immediately prior to analysis. This approach proved successful and a nitrosated ileostomy sample was generated in which compounds characterised by ions at m/z 242 and 264 and by m/z 636 and 658 could be identified. Reference material in the mass spectrometer run indicated that mass accuracy was only 15 ppm, which is insufficient to enable a reliable elemental composition to be calculated.

8. This project has demonstrated a successful strategy for identifying nitrosamine precursors in ileostomy and faecal waters. It is apparent that, unlike dialkyl nitrosamines, these have very limited stability and hence very considerable care will be required to identify the compounds concerned in order to establish their significance to colon cancer.

Main File

Some of the files on this site may be in a format that your computer can't read. However, you can download Readers and Viewers for the following document types below: