It looks like a mouse should be blamed – at least partly – for events that led to the botulism and huge economic losses for Fonterra and some of its customers.
The role of the mouse was revealed when AgResearch publicly defended itself against suggestions its test findings contributed to the export market fiasco, now under official investigations, triggered by the Fonterra botulism scare.
The Hamilton-based science institute issued a statement to address the “public misunderstanding” surrounding its work in relation to suspect bug samples provided by Fonterra for testing.
AgResearch said contrary to some statements, it had not confirmed the presence of Clostridium botulinum in the Fonterra supplied samples.
AgResearch, said it had “reported that the research had potentially detected Clostridum botulinum and recommended that further testing be conducted”.
Government-commissioned tests of the whey protein concentrate samples, at the centre of last month’s baby formula botulism contamination alert by Fonterra, last week concluded it was a false alarm.
That news prompted science blogger Siouxsie Wiles to post an item in which she focused on the mouse.
The Fonterra ‘not botulism’ story continues to rumble on with Crown Research Institute AgResearch releasing a statement saying they didn’t confirm botulism in the Fonterra samples they ran through the mouse bioassay, merely flagged the need for further testing. Then came the revelation that it was the death of one mouse out of five injected that started the whole saga. It has also been stated in the press that AgResearch are not accredited to carry out the mouse bioassay for botulism testing.
Wiles took each point in turn and deconstructed them.
When is the botulism mouse bioassay not the botulism mouse bioassay?
The gold standard mouse bioassay involves injecting groups of mice with suspected Clostridial species of interest alone, and then in combination with anti-toxins, which neutralise each of the potential toxins carried by C. botulinum.
Depending on which animals die, and which ones live, you can then be fairly certain that you are dealing with a strain of C. botulinum and which toxin it is carrying. For example, its likely that groups of 3-6 mice would be injected with the mystery bug (lets call these mice group 1) and then mystery bug + antitoxin A (call these group 2), bug + antitoxin B (group 3), bug + antitoxin E (group 4), and bug + antitoxin F (group 5).
Another group of animals should be injected with an inert substance (group 6) to ensure there is no underlying health issues with the animals, which might cause them to die when injected. A further control group could be those injected with a known sample of C. botulinum (group 7) to ensure the assay works. What you can see from this is that with each group of animals that is added the time and money needed goes up.
So what would the results look like then if the mystery bacterium was C. botulinum carrying toxin A?
First off we would expect the animals in group 7 to get sick and 6 to stay well to show the assay is working as it should. Then we would expect animals in group 1 to get sick, as well as groups 3, 4 and 5.
However, animals in group 2 should remain well because they were given the antitoxin. If the mystery bug contained toxin B, then animals in group 1, 2, 4 and 5 would get sick, and the ones in group 3 would remain well.
It is worth noting at this point that from an ethics standpoint, the use of animals has to be minimised where possible (we call this the 3Rs) so in a case where one or more mystery bugs have been flagged as potentially toxigenic, it is not unreasonable to test them in mice alone without antitoxin as a first port of call.
We’ll call this the ‘lite’ version of the mouse bioassay. Crucial in this test though is the need for control animals (group 6 above) to rule out any underlying health issues in the animals used.
If there was any illness in any of the mice injected, this would then flag the need for more comprehensive testing, this time with the various anti-toxins.
AgResearch’s statement suggests they did the the ‘lite’ bioassay, flagging the sample as potentially toxigenic and suggested running further tests which we can only assume weren’t done. What we don’t know is whether any control animals were used in the ‘lite’ test and if not, who decided that was an appropriate way forward.
Wiles says she still thinks Fonterra was right in doing the recall, although (she contends) it should have either quarantined or downgraded the whey when their first tests came back as suspicious back in March.
Wiles then turns to the NZ Herald’s reporting that AgResearch was not accredited with the International Accreditation New Zealand to carry out the botulism mouse bioassay.
She tells her readers:
I think they have said that no lab in NZ is accredited for this test.
According to their website IANZ is “part of the Testing Laboratory Registration Council, an autonomous Crown entity established in 1972. IANZ is a full signatory member of the International Laboratory Accreditation Cooperation (ILAC) and the regional body, Asia Pacific Laboratory Accreditation Cooperation (APLAC).”
I think this is a red herring. AgResearch are not a bunch of cowboys doing tests in their garage. They are a CRI with the appropriate facilities and expertise.
What is needed is the lab equipment and expertise to grow clostridial species, as well as experienced people who can inject animals and take care of them during the experiment.
Accreditation is much important for independent companies who are offering testing as their core business. For a one-off situation like this, I don’t think the fact that AgResearch aren’t accredited is relevant.
Wiles then put matters in perspective.
So now we know that the contaminating organism was not the toxigenic Clostridium botulinum but C. sporogenes, how prevalent is this soil microbe in baby formula?
A study published in 2010 in the Journal of Pediatrics by Barash et al, found clostridial spores in 7 out of 9 cans they bought.
Most frequently isolated was C. sporogenes, 1.1 to >23 spores per 100 g of formula. So contamination of formula with this organism may well be a fairly common thing.
Barash JR, Hsia JK, Arnon SS (2010).Presence of soil-dwelling clostridia in commercial powdered infant formulas. J Pediatr. 156(3):402-8. doi: 10.1016/j.jpeds.2009.09.072.