Wednesday, February 4, 2015

Herbal supplements under suspicion

Some of you might have already read or heard that DNA Barcoding is in the news again (e.g. here and here). The New York State attorney general’s office accused four major retailers of selling fraudulent and potentially dangerous herbal supplements and demanded through an official cease and desist notification that they remove the products from their shelves. This new investigation was prompted by research at our institute. which back in 2013 found that about a third of herbal supplements tested did not contain the plants listed on their labels but cheap fillers instead.

It comes to no surprise that the industry reacted quickly to this new development and heavily criticized the attorney general. One of the many outlets that represent their views provided a list of 10 critical questions, and I thought it might be a good idea to respond to some of them as someone with considerable experience in DNA Barcoding. I usually don't put an emphasis on my credentials but since this is one angle of the criticism I thought I should state that upfront. And now without further ado the questions as found here (all questions are highlighted in italics to indicate that they are not mine - some were left out as they are outside my area of expertise):

1. DNA Barcoding is not an established method for extracts. Why did the AG and the researcher choose this method?
It is true that isolating DNA from herbal products such as the supplements in question is not an easy task. The fabrication process is often damaging DNA which usually means that it is broken down in smaller chunks (degradation). However, this does not necessarily mean that there isn't sufficient DNA information left to provide a species identification. Laboratory procedures need to be adjusted and tests need to be repetitive, a single attempt won't suffice. Still, there is a chance that no usable DNA is left but that hardly warrants this kind of criticism. The cases where the supposed ingredient was not found showed other species such as fillers like powdered rice, asparagus and houseplants. That tells me that there was indeed sufficient DNA left but unfortunately not from the right organism. Nevertheless, I think it is important to make one thing clear: The fact that a species was not detected can mean two things:
- it is indeed not present in the sample, or
- there wasn't enough specific DNA left to provide a sequence
The latter is what is in the center of the criticism. There is indeed a chance of a false negative result. However, if so it also means that the concentration is likely very low and - very important for any customer - it does not explain why there are contaminants present. Careful interpretation on both sides is needed here.

DNA Barcoding is currently the best method to do species identification especially for processed, fragmentary remains of organisms. The reference library is steadily growing and accessible to anyone. These might have been the reasons why it was chosen.

2. Why is the Attorney General withholding the data?
There isn't much I can say but the name of the researcher and his lab are known. I would love to see the data myself but I can't tell if it is common practice to withhold them for the time being. The blog author also complains that the data hasn't been published in a peer-review journal. Now that would be very unusual. If all legal case related data need to go through peer review that would defy the purpose of the system and keep every researcher busy reviewing paper that do not constitute basic research. There is simply no need for a scientific publication in such as case and it would make the scientific community part of the legal system. I wouldn't want to review a manuscript that harbors the risk of being drawn into a legal battle.

4. How qualified is Dr. Schulte to be testing botanical products?
Well, Dr. Schulte is very experienced in molecular biology and working with DNA and that is usually well enough to start with DNA Barcoding. After all we use it in schools and have kids participating in the work. The first cases of market product testing were actually done by grade 11 students, e.g. Sushigate, green tea. Given the challenges of DNA isolation as described above it would be advisable to have at least a university molecular lab working on such a study but there is no need to be an expert botanist if you focus solely on DNA Barcoding as it was the case. That changes once you include other means of identification such as morphology but that does not work with supplements. It all comes down to experience with the technology, knowing about its challenges, and how to interpret the results.

5. What were the quantities of the contaminations detected?
DNA Barcoding won't be able to provide this answer. It is used to find contaminants but not to check if label requirements are necessary. I see the point here and understand the interest in such data but it bothers me to know that presence of an unwanted species is not perceived as the problem but the quantity. That would mean that meeting regulatory requirements is more important than customer satisfaction.

7. This paper was widely criticized - were these criticisms known to the AG and the researcher?
This question refers to the paper by Newmaster et al. 2013 which I mentioned above. Problem is that the "wide" criticism came only from one source which is the American Botanical Council (ABC). I am not going to comment a lot on this organisation. Have a look at their website and especially at their sponsors list

8. Why did the researcher not use the accepted methods (with monographs)?
What are the accepted methods? I can only guess that this refers to chromatography and microscopy but the latter e.g. doesn't work at all with extracts. The blog author claims that there is little to no cellular material left in an extract for a DNA analysis. Well, how are we supposed to do morphological identification? Of all the things that could be left after processing supplements it is fragments of DNA and proteins. Other accepted methods have failed to detect contaminations. That's why a new method (now 12 years old) was applied.

10. Will the industry response be heard?
I would think so. A lot of the press articles include the criticism voiced by industry. There is nothing one can do about the negative coverage. I agree the echo is one-sided but that is the result of a general mistrust in industry in general caused by such incidents. The problem might still be very isolated but the customer doesn't know that. We don't know who sells quality and who not.  Naturally we react with suspicion. 

I'd like to make clear that I am far from demonizing the entire industry. I have enough experience with similar incidents to know that there are also some very reliable retailers and manufactures. However, there are no independent control mechanisms in place that would provide a clearer picture than what we see through these spot checks. The fact that unwanted material shows up in supposedly clean products is very concerning and needs to be addressed. I believe it would be advisable for industry to embrace DNA-based identification methods instead of condemning them each time when they provide inconvenient truths. If it doesn't work perfect then by all means lets sit down and work on improving it together. An industry that works proactively to single out the black sheep among them is what we need. If species identification of herbal extracts is an issue we need to work on that. Probably it is advisable to test before the product is processed beyond recognition. I am well aware that such changes and additional testing come with extra costs but there are perhaps ways to cover those. 

It all comes down to trust. Does a customer trust a company enough to buy their product? Nobody will gain this trust by referring to regulations that are met. We know that regulations in this market segment are insufficient which is why these studies find contamination in the first place. People will buy products if they can be reasonably certain that those aren't contaminated and contain a pure substance, not a mixture that sometimes might not even contain the ingredient they were looking for.

No comments:

Post a Comment