On Tannins, Tea, and Soymilk: Episode 1

I have this problem where I love strong black tea, but I occasionally turn green and vomit it back up. It's...not pleasant. Being stubborn and showing a complete disregard for my stomach lining led me to work through this minor bump in the road. Through trial and error, I found that I only had any nausea from tea if I am on an empty stomach. Not only that, but some foods hanging out in my gut seemed to do the trick, while some provided no protection at all. Milk, it seemed, worked really well. And eggs. Or pretty much any big, balanced meal. Toast and jam: not so much.

So I assumed it was protein, and dissolved protein at that, that buffered my belly from the inevitable puke-party.

Then the other day at work, a woman complained to me of two separate half-gallons of her soy milk spoiling before the expiration date.

"What were you using your soy milk for?" I had my suspicions.
"I put it in my tea."
"And it curdles?"
"It looks like cottage cheese and it sinks to the bottom."

Well, I thought, it actually looks more like tofu, since it basically is. And no, there is nothing wrong with your tasty soy beverage. It's a harmless, but annoying physical process, likely very similar to what was happening in my gut.

Here's my theory: Black tea (and green tea) contains considerable amounts of tannins, a fairly broad class of phenolic compounds so named because they were historically used for tanning leather; these tannins are binding to, aggregating, and generally dicking around with the proteins and mucus layer in my gut. Unless there is another more accessible (dissolved), abundant protein source for them to get to first!

So before I get into what tannins may be doing to proteins in my stomach and in soy milk, let's talk about what tannins are - I mean chemically - and why their structures determine this nauseating (and occasionally useful!) function.

Rule 1: Activity follows structure.

With Rule 1 in mind (which I just decided was a Rule...It seems true in my experience) let's look at a couple of different types of tannins.

Pentagalloylglucose

Pentagalloylglucose (PGG) is the per-esterified gallic acid ester of glucose. That just means at every available -OH group in glucose, a gallic acid group has been appended. It's also a prototypical hydrolyzable tannin, and it is capable of precipitating protein out of a solution (1, 2).

(4->8) Catechin Dimer

The catechin dimer shown above represents a member of another class of tannins, the flavanols. You're probably aware of this class from the ceaseless talk about antioxidants in tea, pomegranate, acai fruit, red wine...blah blah blah. Yes, they scavenge free-radicals, but they're also astringent tannins that are capable of binding to, aggregating, and precipitating protein (2)!

Similar to both classes of tannins? Lots of hydroxyl (-OH) groups, and a number of phenolic hydroxyl groups (an -OH sticking to a benzene).

OK. So. Tannins come in a couple of structural classes. They both have phenolic hydroxyl groups. They can mess with proteins. And they can be found in black tea. To understand how tannins might do this, let's think about the transient, non-covalent bonds a phenol could form with proteins.

In scenario 1, the aromatic ring itself uses its delocalized pi-bonds (benzene's most attractive feature, IMO) to "stack" on top of an aromatic ring on the outside of a protein, like tryptophan, tyrosine, or phenylalanine. In fact, some fancy scientists removed some of the aromatic residues out of a protein that is normally susceptible to tannins and found that the tannin (PGG) lost much of its ability to bind and aggregate that protein (1). Neat trick. What's next.

Scenario 2: Those tannins? They're covered in little polar -OH groups. The short version here is that "O" carries around a small negative electrical charge, and the "H" carries around a small positive electrical charge. There are buckets of residues on a protein which do the same, so you can imagine that enough of these weak, fleeting hydrogen bonds formed between tannins and a protein could become significant.

.....

I should mention how tempted I was to tell the lady with the chunky soy milk everything I've just explained here. Really, I think she would have felt much better knowing about pi-stacking as it relates to her morning cuppa (don't you?). I showed some restraint instead, which is something I scarcely have to worry about here.

In the next segment, I'm going to put the pieces together and get to the point...if there is a point...Oh yeah! How do the chemical properties of the tannins discussed above curdle a tasty vegan milk substitute? Why does tea make me want to vomit? Why are there so many coffee makers in my garage?! Who left the bathroom light on!!!?!


1.  Gyemant, G et al. (2008). Evidence for pentagalloyl glucose binding to human salivary α-amylase through aromatic amino acid residues. Biochimica et Biophysica Acta 2, 291-296.

2. Hagerman, AE et al. (1998). Mechanism of protein precipitation for two tannins, pentagalloylglucose and epicatechin16 (4->8) catechin (procyanidin). J Agr Food Chem 46, 2590-2595.

3. Chen, Z et al. (2006). Structural insights into histone demethylation by JMJD2 family members. Cell 125, 691-672.