I originally wrote this article back in 2011 and I received some very accurate and fair criticism for having not grasped all of the concepts involved in the use of transglutaminase a/k/a “meatglue” by restaurants. So, after revising the article to try to correct those problems I still did not believe that I had done the topic justice, so I took the article down for about one year. Because of a renewed interest in transglutaminase, I thought that I ought to try to rework the article to be as fair and accurate as possible. Again, I want to thank David Arnold from French Culinary Institute blog that was the source of my original musings . See the comment chain below. I’ve now read the GRAS (Generally Recognized as Safe) filings by Ajinomoto available on the USDA‘s website. Although they omit any specific reference to the parts per million level of enzyme in a given amount of product, other than to state that there are a certain number of “units,” I have learned that enzymes are measured by activity or effectiveness and not so much by quantity. Hence the use of the undefined term “unit” in their filings and on their website. Because the company does not define what a unit is, I imagine that bloggers will be going down this path for years to come until Ajinomoto clarifies in writing somewhere how it defines a unit and, perhaps more importantly, points out that this enzyme is safe to eat.
Most important to me, GRAS 00095 at page 114 also notes that enzymes have not been implicated in any human allergic reactions, though the dust can can cause skin and inhilation irritation. http://www.accessdata.fda.gov/scripts/fcn/gras_notices/grn0095.pdf
Ajinomoto should clarify in writing somewhere how it defines a unit and, perhaps more importantly, points out that this enzyme is safe to eat.
The USDA states that food processors selling meat treated with meat glue must not use more than 65 parts per million (65 ppm) of meat glue to the total weight of the treated product. What about restaurants? They are not bound by those rules. Do they use more than 65 parts per million of meat glue or transglutaminase in their products?
As I mentioned before, the restaurant industry is not subject to USDA regulations governing Transglutaminase or meat glue if the product is sold as a meal and not as a meat product for cooking by a consumer. However, the USDA only allows Transglutaminase to be used as a binder at levels of up to 65 ppm in beef. Here is one place where that limitation is set out in the regulation:
(d) Fabricated steak. Fabricated beef steaks, veal steaks, beef and veal steaks, or veal and beef steaks, and similar products, such as those labeled Beef Steak, Chopped, Shaped, Frozen, Minute Steak, Formed, Wafer Sliced, Frozen, Veal Steaks, Beef Added, Chopped/Molded/Cubed/Frozen, Hydrolyzed Plant Protein, and Flavoring shall be prepared by comminuting and forming the product from fresh and/or frozen meat, with or without added fat, of the species indicated on the label. Such products shall not contain more than 30 percent fat and shall not contain added water or extenders. Transglutaminase enzyme at levels of up to 65 ppm may be used as a binder. Beef cheek meat (trimmed beef cheeks) may be used in the preparation of fabricated beef steaks only in accordance with the conditions prescribed in paragraph (a) of this section.
The same 65 parts per million shows up for sausage and that rule is used by many other products as well.
Another stated assumption that I made in an earlier version of this article was that the meat glue products are 100% product and that they are not full of other substances, such as binders or preservatives. If so, then more powder would need to be applied to meet the requirement. It turns out that they are diluted with other ingredients, namely Maltodextrin and Casein, which substantially reduces the amount of meat glue in the final product. According to Dave Arnold of the French Culinary Institute in a comment to this article: “Ajinomoto is well aware of the approval limits for their enzyme. Their usage guidelines for Activa RM are .75-1.0% (for most applications), which is well within 65 ppm limits for pure enzyme.”
As noted above, the actual effective level of the enzyme is measured by the manufacturer on a per-bag basis – 65 units per gram for one formulation as an example – so that if a restaurant follows the directions it should end up using no more transglutaminase than is recommended. In fact, if a chef uses too much transglutaminase, then the product is actually less effective and it falls apart.
The recent spike in concern over meatglue and frankenmeat may stem in part from two separate developments to hit the news. First, there is the pink slime issue. I assume, but do not know, that transglutaminase is used as a binder to put the “slime” back together so that it approximates a regular meat product. But the issue is not about the use of transglutaminase to bind the meat but the source of the meat. Some consumers have a well-founded belief that hamburger is made of “beef” and that “beef” is something different than what is typically used to make hot dogs and sausage. It does not good for Governors like Terry Branstad to wolf down pink slime meat at public relations events because it is not the fact that all sorts of meat products go into the mix but the fact that it has been legally marketed as regular hamburger that is the problem. It really comes down to a matter of properly labeling the product. if people want sausage then they will buy it knowing full well (or not caring) that it contains cartelage, sinus tissue, mouth parts, and all other scrapings and cuttings.
The second thing to pop up in the news is the very intersting development of meat protein in a test tube. I have my own doubts that meat will be made this way on an industrial scale, but one can never second-guess technology and economics, so lets assume it is possible. If test-tube meat is indeed produced on an industrial scale, then new issues arise as to what sort of labeling will need to be used. Will the product be secretly added to our sacred hamburgers? Or will it have to be labeled and sold as something entirely different. After all, there is no reason that the meat even has to be from a specific species of animal. It might be possible, through recombinent DNA science to mix and match the best of several animals to make a final product. It will be neither fish nor fowl, and pose problems for Catholics on Friday.
As it stands now, meat glue is, thankfully, only being used to glue real pieces of honest-to-God meat from real animals. As I’ve tried to make clear, any product sold in store shelves has to be labeled as containing transglutaminase. Not so with restaurant dishes. Here is where things can get a bit murky. It is one thing for a culinary chef to make a special product using science. It is another for a restaurant to glue smaller, cheaper cuts together and sell them as a more expensive cut of beef. That is one area of concern that I think consumers may have. One chef is trying to make a reputation as a great chef and the other is just trying to make a buck. The latter is what I think that we need to carefully watch as this and related issues come to the market.
At this time I consider this posting to be just a statement of opinion, subject to comment and criticism. I do not have any ax to grind but I am interested in making sure that consumers have access to accurate information. If I am wrong about anything in this blog I will, after investigation, post the correction. If I disagree, I will note why I disagree.
Distilled Opinion 
Your argument is incorrect in several respects.
Most USDA TG approvals extend up to 65 ppm, but chicken breast is allowed to be glued at 100 ppm (because chicken breast doesn’t glue so well). Most importantly, chefs always follow these codes (even if they don’t know it). The USDA codes refer to the straight enzyme Transglutaminase, not the product ACTIVA RM, which contains Maltodextrin, Casein, and Transglutaminase (in small quantities).
Ajinomoto is well aware of the approval limits for their enzyme. Their usage guidelines for Activa RM are .75-1.0% (for most applications), which is well within 65 ppm limits for pure enzyme.
In fact, chefs mostly use much, much, less than this –usually on the order or 0.1 percent or less (because we are binding large pieces of muscle together).
Mr. Arnold: I changed the article to reflect the fact that Ajinomoto’s Activa RM is not 100% transglutaminase. However, I need to find out exactly what percentage of the product is in fact transglutaminase. If it is less than 1% by weight then it would go a long way toward convincing me that the product is being safely used by restaurants. I do have a Freedom of Information Request pending with USDA for some information on the research on the product. Assuming that I don’t get stonewalled (a poor assumption), I may find that information in the research. — D. O.
Ajinomoto’s documents are available online. I’m also sure they would be happy to discuss this matter with you (if just to dispel your concern). They don’t print an absolute level of enzyme in their documents. Instead they give a usage level –usually up to 1 percent except for chicken breast, and then repeat the 65 ppm maximum enzyme level. (I once had the actual documents used in the GRAS submisissions. They might have had proportion ranges. Maybe I got them from USDA’s website… don’t remember. You also might trypatent documents). The reason the datasheets don’t give absolute ratios of enzymes to maltodextrin/helper protein is that Activa is standardized to a certain level of enzymatic activity. So batch A might have twenty percent less enzyme than batch B because it was 20 percent more active. All the batches have to be blended so the company recommended usage levels for Activa RM(1 percent) fall under the USAD’s limits (usually 65 ppm –which Ajinmoto helped set by the way).
Again, thank you for the clarification. I’ve read the GRAS (Generally Recognized as Safe) filings by Ajinomoto available on the internet. As you note, they omit any specific reference to the level of enzyme in a given amount of product, other than to state that there are a certain number of “units.” Units are undefined. But what you say makes sense as the enzyme’s effectiveness is measured not so much by quantity but by activity. Hence the use of the undefined term “unit.” I have independently confirmed this fact, but the definition of unit is still undefined and I imagine that bloggers will be going down this path for years to come until Ajinomoto clarifies in writing somewhere how it defines a unit and, perhaps more importantly, points out that this enzyme is safe to eat.
Most important to me, GRAS 00095 at page 114 also notes that enzymes have not been implicated in any human allergic reactions, though the dust can can cause skin and inhilation irritation. http://www.accessdata.fda.gov/scripts/fcn/gras_notices/grn0095.pdf
I will remove this particular article, or put up a brief explanation, sometime this week after you have had a chance to read my comment. I will also edit the main article that is generating all of the traffic to remove the reference to a threat of a potential health risk. Ajinomoto should clarify in writing somewhere how it defines a unit and, perhaps more importantly, points out that this enzyme is safe to eat.