MINUTES - 2ND OCTOBER 2003
Present:
| Chairman: Professor P Farmer |
Members: Dr B Burlinson Dr G Clare Dr J Clements Dr D Gatehouse Dr N Gooderham Ms M Langley Dr I Mitchell Professor D Phillips |
Secretariat: Dr R J Fielder (Scientific DH) Dr D Benford (Scientific FSA) Mr J Battershill (Scientific DH) Mr S Robjohns (Minutes) Mr K N Mistry (Administrative) |
Assessors: |
In attendance : Dr K Fuller (HSE) Ms C Mulholland (FSA item 4) |
CONTENTS
| Item | Paragraph |
|
1. |
Announcements/Apologies for absence | 1 |
2. |
Minutes of the meeting 29 May 2003 (MUT/MIN/03/2) | 5 |
3. |
Matters arising not covered by later agenda items: | 6 |
3.1 High dose bone marrow data that may not be indicative of in-vivo mutagenicity: 3rd draft statement (MUT/03/10) |
||
3.2 2-Phenylphenol: statement on mutagenicity. |
||
3.3 Statement on1,3-Dichloropropan-2-ol (1,3-DCP) (MUT/03/15) |
||
4. |
Mutagenicity of chromium picolinate (MUT/03/14) | 11 |
5. |
Mammalian cell mutation assays (MUT/03/12) | 25 |
6. |
Strategy for investigating germ cell mutagens (MUT/03/13) | 35 |
7. |
For information: | 41 |
| 7.1 Sudan I in chilli powder (MUT/03/16) | ||
8. |
Any other business | 46 |
9. |
Date of next meeting | 49 |
ITEM 1: ANNOUNCEMENTS/APOLOGIES FOR ABSENCE
1. The Chairman welcomed Ms C Mulholland (Food Standards Agency) and Dr K Fuller (HSE).
2. Apologies for absence were received from the COM member Dr E Parry and from the assessors Dr D Russell (National Assembly for Wales) and Mr A Browning (VMD).
3. The Chairman, Professor Farmer, was congratulated on receiving the Gerhard Zbinden memorial diploma at the recent EUROTOX meeting in Florence, Italy.
4. Members were reminded of the need to declare any interests before discussion of items.
ITEM 2: MINUTES OF THE MEETING ON 29 May 2003 (MUT/MIN/03/2)
5. The minutes were approved with minor amendments.
ITEM 3: MATTERS ARISING NOT COVERED BY LATER AGENDA ITEMS
3.1 High dose bone marrow data that may not be indicative of in vivo mutagenicity
6. The Committee was informed that its statement on interpreting high dose positive in-vivo mutagenicity data in the bone marrow assays had been provided for information to the COC. The COC had agreed that this was a very useful document and had no further comments. It would now be placed on the COM internet site.
3.2 2-Phenylphenol: statement on mutagenicity
7. The COM heard that its statement on the mutagenicity of 2-phenylphenol had been finalised by Chairman’s action and had been put on the COM internet site. This document would also be provided to the COC for information.
3.3 Statement on 1,3-Dichloropropan-2-ol (1,2-DCP) (MUT/03/15)
8. Dr Clements declared a non-personal interest. The FSA representative informed the COM that the draft statement on 1,3-DCP had been circulated to the Committee by post and that some editorial comments had been received. The revised draft included the suggested amendments. Details of the analysis of the dosing solutions had also been provided as background information.
9. Members commented that the concentration of 1,3-DCP in some of the dosing solutions was significantly below the nominal concentration and in one case it was nearly twice the nominal concentration. It was agreed that the quality assurance regarding the analysis of the dosing solution was of a questionable standard. However, this did not affect the validity of this study as clear evidence of toxicity had been demonstrated.
10. The Committee was also informed that the in vivo mutagenicity assays on 2,3-dichloropropan-1-ol requested by the COM and commissioned by the FSA had not yet been completed.
ITEM 4: MUTAGENICITY OF CHROMIUM PICOLINATE (MUT/03/14)
11. Members heard that the manufacturing company for chromium picolinate was Nutrition 21 based in the USA. Chromium picolinate is a widely available food supplement in the UK. No interests were declared.
12. The adverse effects of chromium were recently reviewed by the Expert Group on Vitamins and Minerals (EVM). The reports of genotoxicity associated with chromium picolinate were noted by EVM and chromium picolinate was excluded from their recommendations for a safe upper level for this dietary supplement. Following the publication of the EVM report, the Food Standards Agency advised that consumers should use other forms of trivalent chromium supplements until more detailed advice is available. It was known that hexavalent chromium compounds were established human carcinogens on the basis of both animal studies and epidemiological evidence of carcinogenicity (ie considered Group 1 carcinogens by the WHO International Agency for Research on Cancer (IARC http://monographs.iarc.fr/). Trivalent chromium compounds were considered by IARC to be not classifiable with regard to carcinogenicity in humans (ie Group 3).
13. The Food Standards Agency had asked the COM to review the available information on chromium picolinate in order to provide advice on mutagenicity and if appropriate to recommend what further mutagenicity studies, if any, would be required to draw definite conclusions on mutagenicity. Members were informed that US National Toxicology Program (NTP) carcinogenicity bioassays had been initiated with chromium picolinate but it would be over a year before a report was available.
14. The Chairman asked members to comment using the structure of the draft paper as a guide. He asked members to consider the data on chromium picolinate first and then to comment on studies which had used picolinic acid. He noted that a review of published literature on hexavalent and trivalent chromium compounds had been incorporated into the paper to provide background information for members.
Chromium picolinate and other chromium compounds
15. The Committee agreed that unlike other trivalent chromium compounds, chromium picolinate was soluble in water at neutral pH. Chromium picolinate contained trivalent chromium bonded to three molecules of picolinic acid. The formation of additional chromium co-ordination complexes with each molecule of picolinic acid through the lone pair of electrons present on the nitrogen aided the stability of the molecule. Picolinic acid was an isomer of niacin (vitamin B) and a minor metabolite of tryptophan metabolism. There were limited data on absorption, distribution, metabolism and excretion (ADME). The available oral absorption data in human volunteers and in experimental studies in rats suggested that chromium picolinate had a significantly greater absorption across the gastrointestinal tract than other forms of trivalent chromium and was comparable to hexavalent chromium. A recent in vivo ADME study in rats (Hepburn DD and Vincent JB Chemical Research in Toxicology, volume 15, 93-100, 2002) was available. The authors used repeated daily administration (intravenous) of laboratory synthesised radiolabelled chromium picolinate (51Cr or 3H- picolinate) for 14 days. Daily urinary and faecal excretion of 51Cr was approximately 10% at the beginning of the experiment and increased to approximately 20% by the end with the majority of radiolabel found in the urine. This material co-eluted on column chromatography with chromodulin. 3H-labelled material was more rapidly excreted via the urine. Faecal excretion was a minor route of elimination of 3H-labelled material. At the end of this treatment period, both 51Cr and 3H labels were widely distributed but predominantly in the liver. There was a different pattern of distribution to other tissues for these labels. Subcellular fractionation of hepatocytes isolated from animals treated with radiolabelled chromium picolinate for 14 days showed a different pattern of subcellular distribution of 51Cr and 3H radiolabels. These data suggested a complex pattern of ADME which included dissociation of some of the administered chromium picolinate.
16. Members were aware that hexavalent chromium was readily reduced to trivalent chromium both in vitro and in vivo and resulted in oxidative and cytotoxic damage to cells. Evidence for oxidative damage in-vivo in rats given daily intravenous doses of chromium picolinate for 60 days was reported by Hepburn DD et al (Polyhedron volume 22, 455-463, 2003). These authors reported increased urinary excretion of 8-hydroxy-2’deoxyguanosine after 31 days of treatment. In addition increased levels of lipid peroxidation were documented in liver and kidney cells of treated animals. There was evidence of oxidative damage in-vitro from other research groups which included damage to mitochondria from CHO cells and lipid peroxidation in cultured macrophage J774A.1 cells. However the only other in vivo study of oxidative damage had not reported any positive results with chromium picolinate or with other forms of trivalent chromium. Members concluded that overall the evidence was consistent with the conclusion that chromium picolinate induced oxidative damage in-vivo to a lesser extent than hexavalent chromium compounds (Preuss HG et al (Clinical Nephrology, volume 47, 325-330, 1997). Members noted that the evidence for oxidative damage in-vivo with chromium picolinate came from only one laboratory. The test material had been synthesised in the test laboratory.
17. The Committee was aware that trivalent chromium compounds were able to bind DNA and RNA in cell-free systems. Studies with chromium picolinate suggested little direct interaction with DNA (Hepburn DD et al 2003 Polyhedron volume 22, 455-463.) However Speetjens and colleagues (Chemical Research in Toxicology volume 12, 483-487, 1999) had shown a dose-dependent relaxation of supercoiled plasmid DNA to the circular nicked form by trivalent chromium picolinate in the presence of ascorbic acid and air. The authors noted that chromium picolinate was stable and thus could be incorporated into cells intact. They speculated that ascorbate reduced trivalent chromium to divalent chromium. The reduced form could then enter Fenton or Haber-Weiss reaction to produce hydroxyl radicals leading to oxidative damage. The Committee noted this observation but felt more direct evidence was required to confirm the suggestion. The Committee also noted that chromium picolinate induced DNA fragmentation in cultured J774A.1 murine macrophages (Bagchi et al, Mutation Research, volume 180, 5-22, 2002).
18. The Committee considered the available mutagenicity data on chromium picolinate and compared this with other chromium compounds. There were a large number of negative in vitro bacterial mutagenicity tests available where trivalent chromium compounds had been tested. Positive results had been documented for hexavalent chromium compounds in a number of in vitro tests in bacteria. An adequate study using chromium picolinate in Salmonella typhimurim strains which reported negative results had been undertaken as part of the US NTP. Other published bacterial tests with chromium picolinate had also yielded negative results.
19. Members noted that a positive result had been reported in an in vitro mutagenicity test in CHO AA8 cells at the hprt locus in the absence of exogenous metabolic activation, with up to 40-fold increase in mutations. Members noted that the published results were from a single, unusually long treatment time of 48 hours only. The test material had been synthesised in the testing laboratory (Stearns DM et al. Mutation Research, volume 513, 135-42, 2002).
20. The Committee was aware that positive results had been documented in in vitro tests for clastogenicity in mammalian cells using both hexavalent and trivalent chromium compounds. The significance of the results with trivalent compounds was uncertain as evidence of mutagenicity had been documented at high cytotoxic concentrations and under prolonged exposure conditions where it was considered that endocytotic uptake of test material had occurred. Members reviewed a study where chromium picolinate had been tested in both a solubilised and particulate form and noted dose-related positive results in CHO AA8 cells following incubation for 24 hours in the absence of exogenous metabolic activation. (Stearns DM et al. The FASEB Journal, volume 9, 1643-1649, 1995). Members noted that details of the types of aberrations had not been reported but agreed that the magnitude of the response (3-18 times control) clearly suggested a clastogenic effect. Members noted that the test material had been synthesised in the testing laboratory.
21. Chromium picolinate but not chromic chloride was active in a multi-generation Drosophila study where it was observed to delay pupation and decrease pupal viability. Further analysis indicated that chromium picolinate increased lethal mutations and dominant female sterility. (Hepburn DD et al. Proceedings of the National Academy of Science, volume 100, 3766-3771, 2003.) It is not possible to extrapolate such data to in-vivo exposure in mammals.
22. The Committee noted there were three in vivo mutagenicity studies in rats and one in mice. No evidence of chromosomal damage was reported in two studies in rats, but the data were available only in abstract form. The Committee was not able to draw any definite conclusions from these studies. No increase in micronuclei was reported in peripheral blood samples from F344 rats given gavage doses of up to 2500 mg/kg chromium picolinate for 3 days. This study had been undertaken as part of the preliminary studies of the US NTP prior to commissioning carcinogenicity bioassays. The ratio of polychromatic to normochromatic erythrocytes was not altered in this study. Members commented that it was not possible to conclude that the bone-marrow had been exposed in this study. Members noted that chromium picolinate may have been retained in the fatty tissue surrounding the bone-marrow. In a separate study B6C3F1 mice were given diets containing up to 50000 ppm chromium picolinate for 13 weeks and micronuclei in peripheral blood erythrocytes counted. There was no clear evidence for a mutagenic effect, although a statistical test for a dose-related trend had reported a positive finding in female mice (P=0.005). Pair-wise comparisons with control suggested no mutagenic effect.
Picolinic acid
23. An increase in hprt mutations and chromosomal aberrations was documented in CHO AA8 cells in the absence of exogenous metabolic activation in tests that were undertaken concurrently with tests with chromium picolinate. Members noted that the mutagenic effects of picolinic acid were reported only at concentrations that were higher than the lowest concentration of chromium picolinate giving rise to positive results in the tests. (Stearns et al 1995 and 2002). Picolinic acid was also studied in a multigeneration Drosophila study. Treatment with picolinic acid alone also increased the numbers of individuals arrested during pupation and reduced larval and adult viability. It is not possible to extrapolate such data to in-vivo exposure in mammals.
COM Discussion
24. The Committee agreed that the evaluation of the mutagenicity of chromium picolinate was complex and the available data were incomplete. Members considered that although some positive results had been documented with picolinic acid, the balance of data suggested that chromium present in chromium picolinate had given positive results in in vitro mutagenicity tests. The mechanism by which this occurred was unclear. However, it was difficult to evaluate the significance of positive results where the test material had been synthesised in the laboratory concerned and an adequate specification was not available. On balance members considered that chromium picolinate was an in-vitro mutagen but the results suggested that the relevant in vitro tests should be repeated using commercial grade material before any definite conclusions were drawn. The Committee noted the available in vivo tests in mammals with chromium picolinate had yielded negative results. However it was felt that any decision regarding the need for further in vivo tests should await results of in vitro tests using commercial grade chromium picolinate as supplied to the dietary supplement industry. In this respect members agreed that a further in vitro hprt assay and an in vitro chromosome aberration should be undertaken with both tests using CHO cells to ascertain whether the published results could be confirmed with a commercial grade sample. Both tests should be conducted to internationally accepted standards. The hprt assay should also include a 48-hour incubation experiment as well as the standard time points. The Committee was aware that US NTP carcinogenicity bioassays would provide important in vivo data in the future. The in-life phase of the NTP bioassays were due to end in July 2004. Members felt it would be several years before a finalised report of these carcinogenicity bioassays was available. The Committee were concerned that the mutagenicity of chromium picolinate should be resolved as soon as possible and recommended that the additional in vitro mutagenicity tests identified be carried out.
ITEM 5: MAMMALIAN CELL MUTATION ASSAYS (MUT/03/12)
25. The COM strategy for investigating the mutagenicity of chemicals has (both in the 1989 and the 2000 guidelines) indicated that the mouse lymphoma assay is the preferred mammalian cell assay. In the area of the approval of pesticide products however some companies frequently use the hprt assay in CHO cells. To support such use the Pesticide Safety Directorate had been provided with an expert opinion from Bayer that questioned this preference for the mouse lymphoma assay. This was provided at Annex A to MUT/03/12. PSD sought the advice of the COM on this paper, and whether there was a need for the COM to revise its advice on this issue.
26. Members questioned the basis for the Bayer paper. This was, as stated in the introduction, to question the claim that the mouse lymphoma tk+/- test might be of higher sensitivity for the detection of point mutations as compared to the hprt test and should therefore be better suited for genotoxicity testing. This was not the reason why the mouse lymphoma assay was the preferred choice for the 3rd in-vitro assay in stage one of the screening strategy.
27. The Committee agreed that that there was no evidence that the mouse lymphoma tk+/- assay had any higher inherent sensitivity for the detection of point mutations than the hprt assay, provided that adequate number of cells had been used in each assay. This presents, on statistical grounds, some practical difficulty with regard to the use of surface cultures of CHO cells, rather than with suspension cultures of mouse lymphoma cells. The number of cells needed is governed by the spontaneous mutation frequency of the cell system used. This is covered in some detail in the OECD test guideline in which it is stated that, as a general guide, the cell number should be at least 10 times the inverse of the spontaneous mutation frequency. The Committee endorsed this and indicated that, allowing for cytotoxicity, it would usually be necessary to use 107 cells or more in this assay. The Bayer paper made no mention of such statistical considerations and the practicalities of this with regard to use of the hprt assay in CHO cells.
28. Members also noted that another reason why the mouse lymphoma assay was preferred in the COM strategy was the fact that it detected a wider range of genetic endpoints, for example, chromosome deletions as well as gene mutations. This gives the assay added value as the second mammalian cell assay on the recommended strategy. Again this point was not considered in the Bayer paper.
29. It was accepted that the available published data on the higher ‘sensitivity’ of the mouse lymphoma assay than the hprt /CHO system, is based on the fact that the former assay detects a wider range of genetic end-points. There were no published studies that had compared the sensitivity of the two assays specifically with regard to the detection of gene mutagens.
30. Members noted that the arguments present in the Bayer paper were based on theoretical considerations, and no experimental evidence was provided to support the contention that the hprt assay was more sensitive specifically to the detection of gene mutations.
31. The Committee questioned the suggestion in the Bayer paper that the recommendation to use a top dose producing 80 – 90% cytotoxicity meant that this would lead to excessive false positives with regard to clastogenicity (where the top concentration recommended in the metaphase analysis guideline was 50%) and that the colonies indicative of clastogenicity should not be scored. The endpoints used to define cytotoxicity are different for metaphase analysis (usually mitotic index) to those used for the mouse lymphoma assays (usually relative total growth) and the high dose level was set to obtain adequate sensitivity in each assay. The mammalian cell mutation assays required viable cells, whereas the cytogenetic assay measured clastogenic events at metaphase; many of which were potentially lethal to the cells. It was noted that, where it had been tested, 50% inhibition of mitotic index could equate to over 90% reduction of cell viability.
32. Members accepted that the hprt assay might have a role in mechanistic studies where it was an advantage to specifically measure gene mutation.
33. The Committee briefly considered the question of genetic end-point specificity, and their earlier (1991) view that compounds producing gene mutations would also be expected to be clastogens and that differences between the results in the Ames test and the cytogenetic assay were most likely to be due to problems relating to metabolic activation or inadequacies in the test system, rather than genetic end-point specificity. They were not aware of any robust data that would contradict this issue, but a comprehensive ‘update’ of the literature had not been carried out and definitive conclusions could not be drawn.
34. In conclusion the Committee did not believe that any change in their recommended strategy was warranted and that the mouse lymphoma assay remained their preferred third in-vitro test. It was agreed that the Secretariat would write to PSD summarising their views on this issue.
ITEM 6: STRATEGY FOR INVESTIGATING GERM CELL MUTAGENS (MUT/03/13)
35. The COM strategy for investigating compounds for mutagenicity is based on the assumption that all germ cell mutagens will also be detected in somatic cell assays, and specifically the bone marrow assay for clastogenicity. For most chemicals recognised as in-vivo somatic cell mutagens, no further genotoxicity testing is required. However, in some specific cases germ cell studies may be undertaken to demonstrate whether a chemical is, or is not, a germ cell mutagen.
36. There is a considerable amount of data supporting the contention that all germ cell mutagens are detected in the bone marrow assay (micronucleus test or metaphase analysis for clastogenicity). The reverse is not true, and a compound may well be a somatic cell mutagen but not a germ cell mutagen.
37. A recent paper reported that N-hydroxymethylacrylamide produced positive results in the mouse dominant lethal assay, but gave negative results in the micronucleus test in both bone marrow and peripheral erythrocytes. It was claimed that this compound appeared to be unique in inducing genetic damage in germ cells but not in somatic cells of male mice. This may have implications for the COM recommended strategy for testing for chemical mutagenicity, and the advice of the COM was sought on this issue.
38. Paper MUT/03/13 reviewed the data on N-hydroxymethylacrylamide. In vitro studies indicated that this chemical had clastogenic potential. It was negative in a range of in vivo bone marrow assays (using acute, sub-acute and subchronic dosing regimes). Positive results, however, were obtained when an especially sensitive method involving flow cytometry was used. Positive results were obtained in the dominant lethal assay in mice after one week and 8 weeks dosing. At the latter time point negative results were obtained in a micronucleus test using peripheral blood. A single dose dominant lethal assay also gave negative results. The Committee was asked whether the data warranted any modification to the COM strategy of testing as outlined in the 2000 guidelines.
39. Members agreed that there were a number of possible reasons why this new study on N-hydroxymethylacrylamide induced genetic damage in germ cells but not in the bone marrow. In relation to the analysis for micronuclei in peripheral erythrocytes, it was noted that the spleen could have removed damaged cells. Members felt that accumulation in the germ cells (positive results had only been detected after repeated administration) could also be significant in producing positive effects on germ cell DNA. Additionally, a certain stage of a developing germ cell could be more susceptible to N-hydroxymethylacrylamide genetic damage, and germ cells might stay in a susceptible stage for a longer period of time than the faster dividing bone marrow cells. Members also felt that some of the bone marrow micronucleus tests involving repeated administration were equivocal, rather than negative.
40. The Committee felt that after consideration of all the available information on N-hydroxymethylacrylamide, following the COM strategy, a second in vivo assay in a different somatic tissue was needed. Overall members agreed that N-hydroxymethylacrylamide was not a germ cell specific mutagen. Members considered that this one study by itself did not warrant any change to the COM strategy as outlined in the 2000 guidelines, but that this issue should be kept under review in case other similar examples were found in the future.
ITEM 7: FOR INFORMATION
7.1 Sudan I in chilli powder (MUT/03/16)
41. Paper MUT/03/16 was tabled for information. The FSA representative informed the COC that the advice of the Chairmen of the COT, COC, and COM had been sought on the presence of SUDAN I in chilli powder.
42. In May 2003 the French Food authority had detected the azo dye Sudan I in a number of products containing chilli powder imported from India. Subsequently the FSA detected Sudan I in some food products containing chilli powder imported from India. Products affected included ‘tandoori sauce’, ‘mixed pickle’ and ‘flaming hot chutney’. Levels of Sudan I detected in the UK had been similar to those in France and in the range of 1000 ppm.
43. There is evidence that Sudan I is carcinogenic in rodents and mutagenic both in vitro and in vivo. Following FSA consultation with the Chairmen of the COT, COC, and COM, it was agreed that although there were some incomplete and possibly equivocal results, it was prudent to assume that Sudan I is a genotoxic carcinogen and that dietary exposures should be as low as reasonably practicable (ALARP).
44. Based on this advice and the fact that Sudan I is not a permitted food additive, the FSA had worked with Local Authorities and Industry to ensure that products containing Sudan I did not enter or were removed from the food chain. The Agency had also issued press releases informing consumers of the products contaminated with Sudan I, and advised people not to eat any of the affected products.
45. The COM was informed that Legislative action had been taken by the European Community (2003/460/EC), to try and prevent any further contamination of food with Sudan I. Analytical evidence would have to be provided on all hot chilli and hot chilli products entering the UK to show that they did not contain Sudan I. Investigations were ongoing to determine whether any other products were affected.
ITEM 8: ANY OTHER BUSINESS
8.1 Have environmental mutagens caused oncomutation in people? (MUT/03/17)
46. One member drew the attention of the secretariat to a recent paper by Thilly WG. (Have environmental mutagens caused oncomutations in people. Nature Genetics, volume 34, pp 255-259). This considered the evidence for environmental chemicals inducing point mutations in humans and questioned the value of mutagenicity assays in screening for potential carcinogens.
47. This paper was provided to members with the intention of having a discussion at the next meeting of the COM (February 2004) on whether it had any implications for the COM recommended strategy for screening for chemicals with mutagenic and carcinogenic potential.
COC draft revised guidelines on chemical carcinogen risk assessment
48. The Committee heard that a final draft of the COC guidelines for chemical carcinogen risk assessment had been produced. A copy was circulated to COM members essentially for information. The document would be made available for wider consultation. It was proposed that any comments would be reviewed by the COC at its April 2004 meeting.
ITEM 9: DATE OF NEXT MEETING
49. 5 February 2004.
ACTIONS
| Item | Action | Who |
| 4. Mutagenicity of chromium picolinate | Draft statement | FSA |
| 5. Mammalian cell muation assays | Draft statement | Secretariat |
| 6. Strategy for investigatingy Germ cell mutagens | Draft statement | Secretariat |